(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "[Publications] ser. 1 [no.] 1-99; ser 2 [no.] 1-56"

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 
to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 
to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 
are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other marginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 
publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 

We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain from automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attribution The Google "watermark" you see on each file is essential for informing people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liability can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 



at |http : //books . google . com/ 








w 



. .■■•V. *'•-.. ■ 



. ^.^ 



t:..:*t-^ 



e»«*«i 



$ ■ ^;^^^^^^^P 




-\' ■ , . 








^m 


4 






^^B 

•<.' 
•/ 




k»l ■ .-i ^ 




• .••'> 




^//iBL 




• • 


av/'^b^H 


• ■ ' •, .• • ■'^ . * 


*'^^•••• . 


r 






ft*: 




^ 


b/ : " 


■K f •: "^ 


#'1 


3^ 




v9^ 


'^^^ V 'i ' /.' ' 


• • '• ■:'._ 


w^^ 


1 < 


«.# ^& 


orCm- 


1 






•• 







^ \ t. c, >. 



OK 



EAELT ENGLISH PRONUNCIATION, 



WITH ESPECIAL EBPEEENCE TO 



SHAKSPERE AND CHAUCEE, 



OONTAININQ AN UnTESTIQATION OF THE COBRBSPONDENOB OF 
WHITING WITH SPEECH IN ENGLAND FBOM THE ANGLOSAXON 
PEBIOD TO THE PRESENT DAY, PRECEDED BY A SYSTEMATIC 
NOTATION OF ALL SPOKEN SOUNDS BY MEANS OF THE ORDINARY 
PRINTING TYPES. 

mCLUDINO 

A BX-ISRAITGEICENT OF PBOF. F. J. CHILD'S UXM0IR8 ON THE LANaXTAGE OF 
CKkVaSR AND OOWEB, AlTD BEPBDrTS OF THE BABE TBACT8 BY 8ALESBUBT 

oir xNoiisH, 1547) Ain> ttelch, 1667, and by babiiley ok fbengh, 1521. 



ALEXANDER J. ELLIS, F.R.S., 

wmudjw ov nn oambbisob rmhoaomoAL booutt, Mnonm ov m lohdov kacbxmatxoax. 

•OOXBTT, XUCan OV TBM OOVITCXL OV VHX FHILOLOOXOAL ■OCXITT. VOBKBBLT 
•CBOULB OV TBXVXT7 COLX.BOB, OAMBBIDGB, BO. 18S7. 



PART L 

ON THE PRONUNCIATION OP THE XIV TH, XVI TH, XVII TH, AND 
XVIII TH CENTURIES. 



LONDON: 

PUBLI8HBB FOB THB PHILOLOGICAL 80CIBTT BT 

A8HER & CO., LONDON & BERLIN, 

AMD FOB THB BABLT BKOLISH TEXT BOCIBTT, JlSJ} THB CHAVCBB BOCEBTT, 

TRiJBNER k CO., 60, PATERNOSTER ROW. 
1869. 



KU 



Speake the Speech I peat toxt, as I peonoitnc'd it to you. 

Shaksperey Tragediet^ p. 266, fo. 1623. 



LeGENDI 8EHPEB OGCASIO EST, ATDIENDI TUGS SEICPEB. PbAETEEEA, 

MTLTO ICAOIS (tT TTLGO DICITUK) TIVA VOX AFEICIT. NaM, LICET 

ACBIOEA SINT, QYAE LEOAS, ALTIYS TAKEIT IN AlokO SBDSNT, QTAS 

PBONTirnATIOy VVLTV8, HABTTYS, GE8TYS STIAK DICENTIB AFFIOIT : 

NISI TEBO EALSTH PTTAMTB ILLTI) AeSCHINIS, QTI, CTK LEOISSET 

KhODICS OBAnOITEM DeMOSTHENIS, ADMTRATTTIBYS CYNCnS, ADIECISSE 

PEBTYB, TI AE, £1 ATTOT TOY OHPIOT AKHKOEITE ; Et ERAT 

AeSCHINES, si DeMOSTHENI CBEDDITS, AAMIIP(>»aNOTATOS : FATE- 

BATYB TAMENy LONOB HELIYS EADEIC ILLA PEONTNTIASSE IPSYIC QTI 

PEPEBERAT. 

0, PUnii Oaeeilii Seemdi JBpist, ii. 3. 



159613 



YeETX OBTHOOBAPHIA QTOQYE OOKSYETTDINI SEBTITy IDEOQYE 8AEPE 
XVTATA EST. KaH ILLA VETVSTISSDCA TBAN8E0 TElfPOBA, QTXBTS ET 
PATCIOBES UTEBAE, NEC fiTMTT.TM HIS NOSTBIS EABTX POBMAE PYEBTNT, 
ET YIS QTOQTE DIYEBSA .... FOBTASSE SICVT SCRIBEBANT, ETIAM 
ITA' LOQYEBANTYB .... EOO (nISI QTOD CONSVETTDO OBTINTEBIT) 
SIC SCBIBENDTX QTIDQyE ITBICO, QTOHODO SONAT. HjC ENIH EST TSYB 
LTTKBABTM, TT CTSTOBIANT TOCES, ET TELTT DEPOSITTIC BEDDANT 
LEOENTEBTS; ITAttYE ID EXPBDCEBE DEBENT, QTOB BICTYBI STMYB. 

M. Fab, QuinetUianif Inst, Orator, L 7. 



NOTICE. 



The first portion of the Chaucer Society's publications 
being ready for delivery to its members, it has been thought 
adyisable to issue at the same time the first four chapters 
of the present work^ which contain an investigation of 
Chaucer's pronunciation and Prof. F, J. Child's Memoir 
upon his language. The MS. of the remainder of the work^ 
which will be of about the same extent as the present part, 
18 so far advanced, that it will possibly be ready for issue 
before the close of the present year ; but as the revision at 
press and the construction of the indices will be very 
laborious, it may have to be delayed beyond that time. A 
brief summary of the contents of both parts, and an out- 
line index, is here annexed. Complete Indices will be added 
to make reference to the great variety of matters treated 
upon, ready and convenient, as the work is intended to 
give in a small space the greatest possible amount of in- 
formation upon a subject hitherto almost untreated. 

This treatise also replaces the paper on the Pronunciation 
of the Sixteenth Century, etc., which was read by the 
Author before the Philological Society, on 18 January and 
1 February, 1867. 

A. J. E 

Kbtbinotok, 

1 Feb., 1869. 



CORBIGENDA IN PART I. 

\* Beader8 observing any misprints in Part L ar$ rsspeetfuUy 
requested to eommunieate with tJ*e author, 25, Argyll Road, 
Kensington, W, 

p. 5. under Orb, read A. i 

p. 7; 1. 5, for A8BA read vaBA. 

p. 53, 1. 6, for atikirk read atiku^h. 

p. 57, line 9 from bottom, for oo'w read oo^to, 

p. 60, 1. \1, for' read ^ 

p. 70, 1. 18/or ut it read ut in. 

p. 80, 1. 20, for inclined suspect read inclined to suspect. 

p. 85, L 12, /or tbat be read tban be. 

p. 89, n. 1, L 2, for be a read be is a. 

p. 106, L 18,ybr refuse so say read refuse to say. 

p. 113, 1. 21, /or does seem read does not seem. 



ADDENDA, 
p. 12. After the paragraph commeneing ** add : 

I evanescent, made from [, before a single letter or combination, 
denotes tbat it is scarcely audible, altbougb tbe speaker is 
conscious of placing bis organs in tbe proper position for 
speaking it. 

1^1 eyanescents, made from [], enclose more tban one evanescent 
element, or entire evanescent words, as (|^'n it1 k^^m |^t' 
paabs,) = and it came to pass. 

p. 12. After the paragraph commencing add: 

(') prominent, tbe acute accent may be placed over any element of 
a dipbtbong or tripbtbong, wbcn it is considered desirable, 
to ^ew tbat it bas tbe cbief stress of tbe inter-gliding 
vowels, but not necessarily tbe cbief stress in tbe wbole 
word, as, for example, to distinguisb tbe pairs of dipbtbongs 
(fu iu, ui uf, ^a ed). 

p. 273. Add to note 2. Compare also : wbitlow, wbitsour, wbitster, 
wbitsul; Wbitacre, Wbitbarrow, Wbitburn, Wbitcburcb, Wbit- 
field, Wbitgift, Wbitbom, Wbitland, Wbitley, Wbitmore, Wbit- 
ney, Wbitstable, etc. etc. 



CONTENTS. 



PAET I. (Now Publishbd). 

INTBODUCnON. Palabottpb, or thb Ststexatic Notatiok of all 
Spoken Souifos bt mhans of thb Obdinabt Primtiko Ttfbb, pp. 1-16. 

CHAPTER I. On Pbonunolition and its Chanobs, pp. 17-30. 

CHAPTER II. Authobitibs fob thb Pbontjnciation of English dubino 
THB Sixtbenth, Sbtbntbbnth, and Eiohtbbnth Centubibb, pp. 81-48. 

CHAPTER m. On the Pbonunciatiok of English in the Sixteenth 
Centubt, and its Gradual Changs dvrino the S even t ee nt h and 
Eiohtebnth Centuries, pp. 49-240. 
} 1. Introduction, pp. 49-50. 
{ 2. Combined Speech Sonnds, pp. 51-59. 
i 3. The Vowels, pp..69-184. 
{ 4. The Consonant, pp. 184-223. 
{ 5. Realisation of the Pronunciation of English in the xvi th, xyu th, 

and xTmth oentories, pp. 223-225. 
{ 6. The Direction of Change, pp. 225-240. 

CELiPTER rV. On thb Pronunciation of English durino the Four- 
teenth Century, as Deduced fbom an Examination of the Rhymes 
IN Chauceb and .Qower, pp. 241-416. 
i 1. Principles of the Inyestigation, pp. 241-257. 
i 2. The Vowels, pp. 258-307. 
] 3. The Consonants, pp. 308-317. 

I 4. On the Pronunciation of E Final in the xir th Century, pp. 318-342. 
{ 5. Professor F. J. Child's Obsenrations on the Langoage of Chaucer 

and Oower, pp. 342-397. 
{ 6. Chaucer's Pronunciation and Orthography, pp. 397-404. 
{ 7. Pronunciation during the Fifteenth Century, pp. 405 6. 
] 8. Pronunciation during the Earlier Part of Uie xnr th Century, with 
mnstrations from Dan Michel of Korthgate, and Riohaid RoUe 
de Hampole, pp. 406-416. 



VI CONTENTS. 

PART II. (UWPUBLISBDED.) 
CHAPTER V. On thb P&oNUNaATioN of English dubino the Thib- 

TBENTH AND PuBYIOUB CsNTU&IBS, AND OF THB TEUTONIC AND SCANDI- 
NAVIAN S0UBCB8 OF THB English Language. 

{ I. Rhymed Poems of the xmth Century and earlier. 

No. 1. The Cnckoo Song (with the Mnsic), ciick A.D. 1240. 
No. 2. The Prisoner's Prayer (with the Music), droit A.D. 

1270. 
No. 3. Miscellanies of the xinth Centory, from the Reliqais 
AntiquBd and Political Songs, with an Examination 
of the Norman French EI, AI, 
No. 4. The Story of Genesis and Exodus, circle A.D. 1200. 
No. 6. Hayelok the Dane, circ^ A.D. 1200. 
No. 6. King Horn, circle A.D. 1290. 
No. 7. Moral Ode, Pater Noster and Orison, xn th Centory. 
{ 2. Unrhymed Poems of the xm th Century and Earlier. 

No. 1. Orrmin*s Orrmulum, end of xuth Century. 
No. 3. Layamon's Brut, beginning of xni th Century. 
§ 3. Prose Writings of the xni th Century and Earlier. 

No. 1. Only English Proclamation of Henry III, 18 Oct. 1258. 
No. 2. The Ancren Riwle, xm th Century. 
No. 3. Old English Homilies, xn th Century. 
i 4. Teutonic and Scandinarian Sources of the English Language. 
No. 1. Anglosaxon. 
No. 2. Icelandic and Old Norse. 
No. 8. Gothic. 

CHAPTER YI. On the Cobbmpondbnce of Obthogbafht with Pbonun- 

CIATION FBOM THB AnOLOSAXON PbBIOD TO THB PbBSBNT DaT. 

} 1. The Value of the Letters. 

} 2. The Expression of the Sounds. 

{ 3. Historiod Phonetic Spelling. 

{ 4. Etymological Spelling. 

{ 5. Standard or Typographical Spelling. 

§ 6. Standard Pronunciation. 

CHAPTER VII. Illustbations of the Pbonunciation of English 
DURING the Fourteenth Century. 

i 1. Chaucer (Prologue to the Canterbury Tales). 

{ 2. Gower (Punishment of Nebuchadnezzar, and Message from Venus 

to Chaucer.) 
§ 3. Wycliflfe (Parable of the Prodigal Son.) 

CHAPTER VIII. Illustrations of the Pronunciation of English 
during the Sixteenth Century. 
{ 1. William Salesbury's Account of Welsh Pronunciatioii, 1567. 
\ 2. William Salesbury's Account of English Pronunciatioii, 1547. 
§ 3. John Hart's Phonetic Writing, 1569, and the Pronunciation of 
French in the xvith Century, inolnding Alexander Bardey's 
French Pronunciation, 1521. 



CONTENTS. Vll 

} 4. William Bnllokar't Phonetic Writing, 1580, and the PronTinoia- 

tion of Latin in tiie xyi th Century, 
} 5. Alexander Gill's Phonetic Writing, 1621, with an Szamination of 

Spenser's Bhymes. 
{ 6. Charles Bntler's Phonetic Writing and list of Words Like and 

Unlike, 1633-4. 
^ 7. Pronoimdng Yocabnlary of the xvith Ceniory, collected from 

Palsgraye 1530, Salesbnry, 1547, Smith 1568, Hart 1569, Bnl- 

lokar, 1580, GiU, 1621, and Butier 1633. 
{ 8. Conjectured Ftonnnciation of Shakspere, with an Examination of 

his Bhjmes and Puns. 

GHAPTEB IX Illustrations op thb Pbonunciation op Enolibh dxtbino 
TBI Sbyentbbnth Csntubt. 
I 1. John Wilkins* Phonetic Writing, 1668. 
} 2. Noteworthy Pronundations of the xth th Century, — 

No. 1. Pronouncing Vocabulary, collected from WaUis 1653, 
WiUdns 1668, Price 1668, Cooper, 1685, Miege 
1688, and Jones 1701. 
No. 2. Price's Difference between Words of Like Sound, 1668. 
No. 3. Cooper's Words of Like or nearly like Sound but Dif- 
ferent Spelling, 1685. 
No. 4. Cooper's Words of Like or nearly like SpeUing but 
Different Sound, 1685. 
{ 3. Conjectured Pronunciation of Dryden, with an examination of his 
Bhymes. 

CHAPTER X. Illitst&atioks op thb PRONUKciATioir op English dubdco 

thb Eiohtxbnth Centubt. 
{ i^^aiBV Buchanan's Phonetic Writing, 1766. 
{ 2. Benjamin Franklin's Phonetic Writing, 1768. 
{ 8. Noteworthy Pronunciations of the xriu th Century, collected from 

the Expert Orthographist 1704, Dyce 1710, Buchanan 1766, 

Franklin, 1768, Sheridan 1780. 

CHAPTER XI. Illust&ations op English and Lowland Scotch Pbo- 

NUNCIATION DITBING THB Nl^pSTBENTH CbNTUBT. 

} 1. Varieties of English Pronunciation in the xixth Century. 
{ 2. Prof. S. S. Haldeman's Phonetic Writing, 1860. 
§ 3. Mr. A. Helrille Bell's Phonetic Writing, 1867. 
f 4. Dialectic Varieties of Pronunciation, English and Lowland Scotch, 
in the xti th and xix th Centuries. 

CHAPTER Xll. Results op thb Pbbcbding Inybstigation. 
Index of Authors cited. 
Phonetic Index. 

Index of all the Words of which the Pronunciation ii described, indicated, 
or oGnjectuTed. 



OuTLiNB Ikdex to the Lettebs Explained in Pakt I. 

Xofe.—Uhe fignrea 14., 16., 17., 18., with periods after them, refer to the centuries, 
the other figures to the pages. 



A 

14. 259, 

16. 59, 

17. 65, 

18. 74. 

AI, AY 
14. 263, 

16. 118, 

17. 124, 

18. 129. 

AXJ, AW 
14. 263, 

16. 136, 141, 

17. 147, 

18. 149. 

B 
14. 308, 
16. 17. 18. 203. 

C 

14. 308, 
16. 17. 18. 203, 
214. 
CH 
14. 308, 
16. 17. 18. 203. 

D 
14. 308, 
' 16. 17. 18. 203 

E 
14. 260, 318, 

16. 77, 

17. 81, 

18. 88. 
EA 

14. 1260, 
16. 77, 

81, 

88. 

EE 
14. 260, 

77, 

81, 

88. 



17. 
18. 



16. 
17. 
18. 



EI, EY 
14. 263, 

16. 118, 

17. 124, 

18. 129. 

EO 
14. 260. 

EXJ, EW 
14. 301, 

16. 136, 137, 

17. 139, 

18. 141. 

F 
14. 308, 
16. 17. 18. 219. 

G 
14. 308, 
16. 17. 18. 203. 

GH 
14. 310, 
16. 17. 18. 209. 

GN 
14. 308. 

H 
14. 314, 
16. 17. 18. 220. 

I, Y 
14. 270, 

16. 104, 

17. 116, 

18. 117. 

IE 
14. 260, 

16. 104, 

17. 116, 

18. 117. 

J 
14. 314, 
16. 17. 18. 203. 

K 
14. 315, 
16. 17. 18. 203. 

L 
14. 315, 
16. 17. 18. 193, 



M 

14. 315, 

16. 17. 18. 188. 

14. 315, 

16. 17. 18. 188. 

NG 
14. 315, 
16. 17. 18. 188. 


14. 266, 

16. 93, 

17. 99, 

18. 103. 

OA 
14. 266, 

16. 93, 

17. 99, 

18. 103. 

OE 
14. 260. 

01, OY 
14. 268, 

16. 130, 

17. 133, 

18. 135. 

00 
14. 266, 

16. 93, 

17. 99, 

18. 103. 
OXJ, OW 

14. '303, 

16. 136, 149, 

17. 156, 

18. 160. 

P 
14. 316, 
16. 17. 18. 203. 

PH 
14. 316. 
Q 
14. 316, 
16. 17. 18. 203. 



196. 



R 

14. 316, 
16. 17. 18. 

S 
14. 317, 
16. 17. 18. 214. 

SH 
14. 317, 
16. 17. 18. 214. 





T 


14. 


317, 


16. 


17. 18. 203 




TH 


14. 


317, 


16. 


17. 18. 219 




XT 


14. 


298, 


16. 


160, 163. 


17. 


171, 


18. 


184. 



m, iTY 

14. 269, 

16. 17. 18. 135. 

V 

14. 317, 

16. 17. 18. 219. 

W. 
14. 317, 
16. 17. 18. 184. 

WH 
14. 317, 
16. 17. 18. 184. 

X 
14. 317, 
16. 17. 18. 214. 
Y vowel, see I 
Y consonant. 
14. 310, 317, 
16. 17. 18. 184. 

Z 
14. 310, 317, 
16. 17. 18. 214. 



INTRODUCTION. 



Paiabotyfe, or the Systematic Notation of All Spoken 
Sounds by means of the Ordinary Printing Types. 

In order to write intelli^bly on speech sounds, some 
eystematic means of representmg them must be adopted. In 
order to understand the mode in which speech sounds change, 
delicate physiological actions of the vocal organs must be 
indicated. In order to be generally intelligible, the letters of 
ihe^Koman Alphabet in their original Latm senses, as nearly 
as may be, should form the nucleus of the system of symbo- 
lisation. In order to be convenient to tne Printer and 
Writer, the old types, irdXcuoX rxnroi, (paleii* tii*pi), should 
be used, and no accented letters, few turned, and still fewer 
mutilated letters should be employed. The system of writing 
here proposed to fulfil these conditions will, in consequence 
of the last, be termed Palaeotype (psel'iotoip). It is essen- 
tially a makeshift scheme, adapted solely to scientific, not 
popular use, not pretending to supersede any existing system 
of writing, but sufficing to explain all such systems, and to 
indicate the pronunciation of any language with great 
minuteness and much typographical convenience.^ 

The reader will have no occasion to study the whole of the 
following list before beginning to read the book. The nature 
of the symbols allows by far uie greater number of them to 
be arranged alphabetically, so that the reader can imme- 
diately discover the meaning of any symbol or usual combi- 
nation, and any unusual symbol is generally explained when 
it first occurs in the following pages. It is only necessary 
to bear in mind that the Roman vowels (a, e, i, o, u,)^re 
pronounced as in Italian, and (y, oe) as the Oerman ii, o, that 

^ A fbU acoonnt of the principles improyementB. Ab now presented, 
of tbe notation is given in the JVofM- Palaeotype is believed to contain cha- 



of th$ Fhilohgieal Society for tacters for all the sounds considered 

1867, Supplement, Part I. The sub- by Rapp, Lepsius, Briioke, Max Miiller, 

teaoent appearance of Mr. Melville daldeman, Merkel, and Melville Bell, 

Bell's VitibU ^^k, and the elabo- and hence to be the most complete 

ration of the following pa^, have series of phonetic symbols which has 

occasioned a few modifications and been published. 

1 



INTRODUCTION. 



•^ ';the italics and small capitals indicate certain modifications 

/>.*'•* of these sounds^ that (h, j, it) are always diacritical, havings 

;'«./• *' no meaning of their own but serving to modify the meanings 

of the preceding letter, and that (h, j, w, q, 9, oi, on) repre- 

,•. *v** sent the sounds in (Aay, yea, iray, sin^, bwt, btte, hotc). 

, ••'•/. * Long vowels are indicated by reduplication, as (aa, ee, ii) ; 

•' X • repeated vowels are separated by a comma as (a,a, e,e, i,i). 

f I * The other common symbols are well known. 

The explanation is given by keywords, the letters ex* 
^ / pressing tne sounds in question being italicised, and by the 

symbols (* M t-|- + ^0~) which shew how some of the 
letters are formed from others, (*) by attempting te pronounce 
simultaneously the two letters between which it is plaoed, 
by teking the contact ( f) nearer the mouth, or (1 ) nearer the 
throat, (t) by protruding, or (4.) by inverting the tonffue» 
it) by clicking, («?) by 'rounding' or labial modification, 
(0) by 'widening' or distending the pharynx and oral pas* 
sages, (~) by removing the effect of the diacritic before which 
it 19 pla(M3d, and which is inherent in the preceding letter, as 
(-4r) with opened lips, (-q) with narrowed pharynx, eto. For 
all English sounds, numerous other examples will be found 
in Chapter YI, § 2. On p. 15, there is furnished a com- 
plete comparison of Palaeotype with Visible Speech, whence 
the exact value of the former can be determined by a refer- 
ence to Mr. Melville Bell's work. Diagrams of the positions 
of the tongue and lips during the pronimciation of the vowels, 
are given on p. 14. 

In the course of the following pages many explanations 
and discussions of phonetic subjecte become necessary. See 
the nature of glides, diphthongs, and combined speech 
sounds explained in Chapter III, § 2, the principal vowels 
and diphthongs in the same chapter, § 3, especially under 
the heading U, the nature of palatisation (j) and labialisation 
(w) in the same chapter, § 4, under P, B ; T, D ; C, E, Q; 
CH, J, and GH, and the nature of aspiration under H. The 
Tables in Chapter YI, §§ 1 and 2, and the footnotes to 
Chapter YIII, § 1, may also be consulted. 

Examples of the use of Palaeotype in continuous writing 
will be found in Chapter Y, §§ 1, 2, 3, 4; Chapter VII; 
Chapter YIII, §§ 3, 4, 5, 6, 8 ; Chapter IX, §§ 1, 3 ; Chap- 
ter X, §§ 1, 2, Chapter XI, g§ 1, 2, 3. In this Chapter XI 
will be found examples of modern English and Spoteh, form- 
ing a convenient exercise for those who wish to study the 
nature of this system of writing, and allowing of a airect 
comparison with Visible Speech. 



KEY TO PALABOTYPB — LETTEBS. 3 

The mode of writing tHe '^ turned" or inverted letters is 
explained in each particular case. Italic letters have one 
horizontal line below them, as i ; small capitals have either 
two horizontal lines, or one short oblique line, as i, below 
th^n, tailed letters as g, j, p, q, j, when they have to be 
printed as small capitals, may have a horizontal stroke abow 
them, like I. The letter h may be also written with its stem 
crossed like t, and/ with two cross bars. 

For the purposes of alphabet arrangement, se, ob are con- 
sidered to be the same as ae, oe, and the turned letters 
e9s«XA0Q)2^i modifications of 
e^EeLnooerrr respectively. 
Isolated letters, words, and phrases in palaeotype occurring 
in the midst of ordinary spelling are enclosed in a paren- 
thesis ( ) to prevent confusion. 



KEY TO PALAEOTYPE. 

Ahhrmaiume. — ^A. arabic, C. chinefle, E. english, F. french, 

G. germaa, I. italian, P. provinoal, S. sansorit. 
occ. occasional, = interchangeable with. 

L Leitebs. 

= (ao)» I- nifltto, F. chatte, (matix), shat) 

=s (o50), G. nunm, F. matelas, (m<in, matla) 

= (<KM?), E. w<mt, wluit, tftfgust', (wAnt, whAt, Agast*), 

see (o) 
Gaelic nwith, good, (ma,) ; nasal twang 
long of (a), E. father, I. numo, (faadh'j, maa-no) 
long of {a)y G. nuz^nen, (inaa*n^ 
long of f a), E. awTLf (AAn), see (oo) 
long of (a j 
long of (ah) 
long of (ah) 
long of (aA), see (a) 

= (eo), E. man, cat, sad, (maen, kaet, saed) 
long of f 8b), p. E. Bath, (BaeaBth) 
long of (aBn) 

= {sdw) = (pho) labially modified (ae) or widened (^h) 
= (90)> occ. E. flsk, staff, grant (ahsk, stahf, grahnt) 
= (ahir), Irish str, Austrian man (sahr, mahn) 

E. ayey G. ham, (ai, nain), see (ai) 

F. an, temps, cent, (aA, taA, saA), see (a) 

G. hauB, (nans), see (an) 
theoieticaL G. euch (jBLjkh) 



A 


a 


A 


a 


:A 


A 


A. 


a* 


Aa 


aa 


Aa 


aa 


:Aa 


AA 


Aa. 


aa, 


A»h 


aah 


Aah 


aah 


Au. 


aaA 


JB 


2d 


JEa 


aBae 


JEahsBsili 


Mi 


seh 


Ah 


ah 


Ah 


ah 


Ai 


ai 


Aa 


aA 


Au 


au 


Ay 


ay 



B 


b 


S 


i 


:B 


B 


•B 


'b 


Bh 


bh 


Bj 


l)j 


Brh 


brh 


Bw 


h» 


C 








c 


D 


d 


2) 


d 


:D 


D 


.D 


.d 


'D 


'd 


Dh 


dh 


Dh 


<fli 


iDVb 


I Dbh 


Dj 


dj 


Ptr 


dw 


Dzh 


dzh 


E 


e 


S 


« 


•.E 


E 


a 


a 


a: 


9 


:a 


a 


t) 


B 


Ee 


ee 


£^ 


M 


:Eb 


EE 


39 


99 


a-" 


M 


:aa 


az 


t)« 


OT 


aah 


99h 


ar»t 


wh 


Eei 


eei 


^'j 


^^•j 


EeA 


eeA 


a^A 


e9A 


ah 


9h 


^h 


«h 


£i 


ei 



INTRODUCnON. 



E. hee, (bii) 

sonant of (p), whicli see, ? = (bir) 

== (hi), lower lip against teeth, Briicke's l^ 

= (b*p), flat Saxon b, Rapp's ir 

G. tt' in the middle and sou^ (t) without the teeth 

= (b*0 

= (bh^), ]ip trill, G. hrr for stopping horses, Briicke's k 
hmB, (btra) 



= (bhi),Up1 
=(b»w), F. 



=z (b\)? nearly (th), Spanish s, and ^ before e, •', Badajot, 

(Baadaaxooc*) 
= (zf) ? nearly (dh), Spanish <^ (?), ciadad (ciu^aa^*) 

E. doy (duu) 

= (d*g)» usually accepted A. tp, Lepsius's A. L 

= m, 8. 'f 

= (d|-), tip of tongue on gums 

= (d*t), flat Saxon (d), Rapp's t 

E. ^Aee, Danish yed, {dim, yedh), Welsh id 

(dh*gh), Newman's A. ^, Lepsius's A. I9 

Lepsius's Dravidian sound, nearly (nsh) 

= (d*j), Hungarian yy, E. yercfure, (vJ'djj) 

= (d*w), F. doit (dira) 

E.jui^ing, (dzh9dzh*»4) 

= {eo)f E. nwt, G. f(rtt, F. j^tte, (met, fet, zhet), see (e) 
= (©-0)> E. serial, F. ^t^ (Ifer'isl^ ete), I. ^ chiuso 
= (8B-o)» I- « aperto, occ. E. nwt, G. frtt, (mBt, fet) 
= (ah-o) turned e, written 9, E. but (hot), see (a) 
= [ew) = (oD-o), turned e, F. qu* j* m^ r^penttf {h 

zhp mp r^paAtp) 
= (a-o) turned a, occ. E. birt; (bat) 
= (a?o) = (w-*^)? turned a, written e , E. mentum, 

real, (men'shnn, rii'Bl) 
long of (e), E. nuire, Mary, (meei, MeeiTi) 
long of M, E. atling (ee'liq), see (eei, w'j) 
long of (b), like a bleat 
long of fa), replaces (j, oi, a>i) in South E. 
long of (9) 
long of (a^ 
long of (b) 
long of fah) 
long of (tfh) 

occ. E. tlwy, (dheei), for (dhw) 
occ. E. fate, {Mjt), for (fd^t) 
long of (eA), see TaJ 
long of (8a), see (a) 

= (ah-M>), West E. sir, first (sahr, fahrst) 
= {bw), occ. F. sH 
Scotch time (teim), Portuguese e% 



KEY TO PALABOTYPE^LETTERS. 5 

gi di agaal E. ey$, tune, (ai, taim) 

£▲ eiA F. vw (veA), see (a) 

gA 9A F. tdi empTMfit, (9Aii-aApr9A), see (a) 

£a en I. -Ekropa, (Euroo-pe), Cockney and Yankee toton (teun) 

an 9u mnial E. lumse, shotit (naus, shaut) 

F f E./)e, (few), gentle hiss 

-F / = (ff ), upper lip against lower teeth 

.F .f violently hissed (f ) 

Fh fh = rf*kh) 

Fi^ tw = (f*wh), the back of the tongue in the (u) position, 
F./ots, {tw9) 

G g E. yo, (goo) 

g =^ (gj) == (g*j), occ. E. yward, (^d), F. yweux, (yoe) 
:Q e sonant of (x) 

*G 'g = (g*k), flat (g), Eapp's ic 

Gh g^ G. tfi^e, (taaghv), Dutch y, 8. f 

61i yh = (gjh) = (gh*j), G. wieye, (bhiiyh-*) 

:Gh eh buzz of (xh) . 

.Gh .gji violently buzzed (gh) 

Gj gj = (a\ which see 

Gjh gjh = (yh), which see 

Grh grh = (gh^), A. ^, heard in gargling 

Gw gio = (g*w), F. goitre, (gwatr*) 

:GiCF GfT = (g*w) 

Gvh gie^h = (gji*w), G. au^, (au'gu^h^) 

:Giph Qwh = (Gh*w) 

H H E. Ae (mi), 8. H ^ ^, (bn, dn, gn), jerked utterance 

H* H* jerked whisper 

h with no capital, diacritic, with no meaning by itself, 
but modifying the meaning of the preceding letter in 
any manner that is convenient, see (ah, th, ^ 'h), &c. 

H h A. ^ {haa) 

'h a scarcely audible (a) as Cockney park, (paa'hk) 

hh with no capital, diacritic, variety of (h), see (Ihh) 

Hip hio a voiced whistle 

Hifh Hi^h an ordinary whistle, distinct from (wh, ktrh) 

1 i = (^), E. ^ent, F. fim, fiche, (ivent*, fini, fish) 
/ I = (io), E. rtver, ftnny, frsh, (rtvi, fih'i, fish) 

:I I =(it^), occ. G. ii, Swedish y 

li ii long of fi), E. ^e, (iiv) 

E ii long of fi\ E. happy... (Hcep'tV), in singing 

:Ii n long of (i) 

lu iu E. futility, (fiutil'iti*) 

7u tu American variety of (iu) 

luu iuu E. fictile, (fiuu'til) 



IKTBODUCnON. 

E. yet, G.ya, (jet, Jaa) 

witb no capital, diacritic, palatal modification of pre- 
ceding letter. 

£unt sound of (j, i) into wliicli £. {$e) occasionally 
tapers, see (m'j) 

occ. E. hue (jhiuu), occ. G. ja (jIum), occ. F. cetZ (^/h) 

E. key, ^an, eoal, (kii, ksen, kool) 

= (kj) = (k*j), occ. E. cart (ifart), F. queue (ite) 

= (k1),A. J(Maf) 

Q. ds/ehf Scotch lo^A, (doUi, lokh) 

= (kjh) = (kli*yh), G. sieM, (szuAh) 

related to (k) as (Kh) to (k) 

S. ^, upper G. ^omm, (knom) 

Tiolently hissed (kh) 

= (^), which see 

= (kh)f which see 

= (kh^), Swiss ehf A. ^ (krha«) 

= (k*w), E. ^tieen, F. ^tioi, (ktriin, kti^), Latin $w 
ktrh = (kh^h\ G. au^A, (auktah), WeL^ ehw, Scotch quh 
Kwh = (Kh*wh) 

E. fow, (loo) 
Polish harred i 
= W), 8. « 

whispered (1), hreatii escaping on both sides the tongue, 
E. feh = (fellht) at fall, occ. F, table, (tablh) 

whisper of (A 

according to Lepsius, Dravidian I in (TamiLh) 

whisper of (i) 

=x (ISh), breath escaping on the right side of the tongue 
(Hily, Welsh U 

= (l*/),Lyfiaji) 

whisper of (Ij) 

a (l^)i F. lei {lwa)f Anglosaxon ip^ 

= ll^w) 

«= (Ih^wh) 

= (/h^wh) 

.M m E. me, (mii) 

m no capital, diacritic, « (a), which see 

Mh mh voiceless (m), E. tem^it (temmht) at full 

Mm7 mw = (ni*ir), F. wun, (nwa) 

N n E. nap (naep) 

iV » =r (n*q), see (i) 
:N K = (n+), S ^ 



J 


J 




J 




'J 


Jll 


;h 


K 


k 


^ 


A 


:E 


K 


Kh 


kh 


Kh 


Ah 


:ini 


xh 


Eh 


kH 


.Kh 


.kh 


%h 


% 


Krh 


krh 


Ktp 


ktc 


Kwh 


kirli 


:K<p1i 


I KWI 


L 


1 


Z 


; 


:L 


L 


i 


X 
.1 


Lh 


lh 


Zh 


A 


:Lh 


Lh 


fihh 


lh 

Ihh 


l-j 


Ij 


Lih 


Iji 


Im 


Iw 


Lw 


Up 


Ltoh 


l»h 


Zwh 


^h 



.N 


.n 


Kh 


nil 


:Nh 


Kh 


Njh 


njh 















KEY TO PAXASOTTPE — ^LETTEBS. 7 

= no capital, written ff not joined to the following 
letter, diacritic^ Prench nasality, the four French 
nasals, ym, on, on, tin, are written for conyenience 
(ycA, aA, OA, oa), though perhaps more properly 
(a»a, oa, oa, 9a), according to Mr. Melyille BeU 
(y»A, ahA, ohA, 9a) 

= (n^, see (.d) 

yoiceless (n), £. tent «= (tennht) at fdll 

according to Lepsius, Drayidian nasal before (nh) 

= (n*y), F. and I. yn, Spanish n, Portuguese nh 

whispered (nj) 

= (n*w), F. noix, (jawa) 

= (air) s=B (oo)> !• aperto, F. homme (om) 
=3 (xiff) s= (o-o) £. omit, American stone, whole, 
(omtV, ston, hoI) 
Q c= {aw) s= Ao), turned c, written o, being used for 
small capital o which is not sufficiently distinct from 
the small o, £. on, odd, (on, od) 
(E OB = (ew) = (po), F. jtftme, G. b^icke, (zhoen, bcek'^), 
Feline writes (zh«n, zhoeoBn), for F. j^tme, j^ne 
= (u-io), Galic lao^ (^h) 
= (a-o)= (A-to), Kumanian or Wallachian 'a, 'e, 4, 

'o, 'u 
ss (oho), written a>, £. ftrst, (foxist), see (j) 
= occ. F. odl, (oei, OBiih, ceilj) or (ri), occ. Dutch uy 
long of (ob), F. j^ne, (;di(BOBn} 
long of (m) 
long of (<e) 
long of (so) 
occ. Butch «y 

«= (ahto) = (ohj), (o) modified by raising the tongue 
= (ofo), (o) modified by raising the tongue 
= (goto) = (aho), (o) modified by raising the tongue 
North G. neu, (noi), see (ay, oy) 
P. E. boy, (boi) 
usual E. o^ter, (oistu) 
F. bon (boA), see (a) 

long of (o), I. uomo, (uoo-mo), P. E. home, (noom) 
long of (o), E. home, (noom), see (oo'fo) 
long of (o), drawled E. odd, God, (ood, Good), different 

from E. atoed, gaiod (AAd, gAAd) 
long of (oh) 
long of (oh^ 
long of (oh) 
long of (oa), see (a) 
occ. E. knoio, (noou) 
more usual E. knoeo, (noou) 
occ. E. no, (noo'io), for (noo) 
Dutdi oti, P. E. oiit, (out), see (ou) 



as 


<B 


:(E 


(E 


3i. 


00 


(Ei 


od 


(Eoa 


(BOB 


(BCB 


#0 


'.(Em 


<B(B 


^ 


OBJ 


Oh 


oh 


Oh 


oh 


:0h 


oh 


Oi 


oi 


(H 


oi 


:0i 


oi 


Oa 


OA 


Oo 


00 


Oo 


00 


:Oo 


00 


Ooh 


ooh 


Ooh 


ooh 


:Oah 


ooh 


OOA 


OOA 


Oou 


oou 


Oau. 


oou 


Ww 


oo'io 


Ou 


ou 



INTRODUCTION. 

P. E. house, (hous) 

occ. upper Ot. euch, {ojkh) 

E. peeL (pii) 

= (p*k)? = (ptr)?, Lepsius's Peruvian or (Kheteh*- 

wa)^ 
= (pI ), lower lip against teeth 
whisper of (bh), an old sound of ^ ? 
S. "^ Bavarian ^erd, (pneerd), Schmeller Gh*. p. 137. 

= (p*,) 

= (P^)» whisper of (brh), which see 
= (p*w), F. pois, (j^a) 

E. sinyer, linger, sinker, (stq'j, l*'q*gJ) sfok'j), S W 
= (qj) = (q*j), distinct from (nj), S. ^ 

= (q1) 

= voiceless (q), E. sink = (stqqhk) at foil 
= (jj) which see 

E. ray (ree), breath passes over the tip of the tongue 

which trembles slightly, Spanish r suave. 
uvula trill, F. r provencal or grassey6, Paris, (Pari) 

= (4), 8. T 

turned r, written as r with ^ above, E. vocal r when 
not preceding a vowel, ear, air, are, oar, poor, (iii, 
eei, aai, ooi, putir), heariug, airing, mooring, (hm*- 
rtq, eei-rtq, mwta'rt'q,) p^rv^, murmw = (peiveit*, 
mwmoi) or (p.<v.^", nuiiu), or (pxvjt*, mj'mj), see (^) 
21 J^ turned r, written as r with ^ above, E. palatal voccd 
r when not preceding a vowel, ear, air = (ii^, eeu) 
more accurately than (iii, eei), and (serf, surf) may 
be distinguished as (se^, soxf ) or (s^, sjf ), this dis- 
tinction is frequently neglected in speech. 
1 1 turned l, written as r with ^ below, glottal low Ger- 
man tnll, nearly (g) 

= (r^) strongly tnUed Italian, Spanish, Scotch r 

whisper of (r) 

whi^r of (r) 

whisper of (b) 

Lepsius's Dravidian sound, nearly (Bsh) 

whisper of (i) 

= (r*J) 

Polish pnsez, (prshez), (r) very brief, (sh) distinct 
«= (r*w), F. roi, (rera), Anglosaxon, and early E. irr- 
= (j*w), occ. E. (oii^) in place of (oiu) = our 
Polish rsaz, (rzhaz), (r) brief 

E. 80j (soo) 

= (8*kh), Lepsius's and usually received A. ^ 

=. E. «Ae, F. cAant, O. «<;Aein, (shii, shaA, shain) 



Ou 


on 


Oy 


oy 


P 


P 


P 


P 


:P 


p 


Ph 


ph 


Ph 


pH 


Pj 


PJ 


Prh 


prh 


P«; 


pu; 


Q 


q 


Q 


q 


•.a 


Q 


Qh 


qh 


Qj 


qj 


R 


r 


R 


r 


:R 


B 


a 


1 



.R 


.r 


Rh 


rh 


Bk 


rh 


:Rh 


Bh 


:RUi 


Bhh 


t> 


ih 


Esh 


rsh 


^w 


TW 


TL^ 


iw 


Rzh 


rzh 


S 


8 


8 


S 


Sh 


sh 



KEY TO PALAEOTYPB — LETTERS. 9 

Sh ih =(8l4), s. "^ 

Shj shj =s (Bh*jh), occ. G. «telleii, sprechen, (shjtel'wi, 

shjprei&h'en) 

8j sj = (s*jh), Polish i 

&p Btp = (8*wh), F. 8oi == (swa) or (sua), not (sira) 

Swh Bwh == (sh*wh), F. ehmjL = (shwa) or (shua), not (mho) 

82 8z = G. initial «, 80^ (bzoo) 

T t E. fea, (tii) 

T t = (t*k), Newman's and nsoally received A. 1^ 

:T T = (t|), 8. Z 

.T .t = (t\)y tip of tongae on gams 

Th th = E. thin, (thin), modem Greek d 

Th ih = (th*kh), Newman's A. ^ 

:Thh Thh Lepsios's Dravidian sound, nearly (x^h) 

Tj tj = (t*j) whisper of (dj), occ. E. vir^e, (vj-tjiu) 

I^ tsh E. ^Aest, msdeh, ca^Aing, (tshest, mstsh, kstsh'tq) 

Tw tw = (t^w), F. toi, (Utm) 

TJ u = (flw), F. p<wl^, E. Lotdsa, (pnl, Lu,ii*za), see (t«) 

U u = I'Bw) = (uo), E. pwU, cook, (pwl, kuk), generally 

confiised wil^ (u) 

:TJ u = (ytr), Swedish u short 

Uh uh = (yw) = (uo), I. chiuso, (0) yerging into («) 

Ui ui F. out = (ui), F. oui == (u,i) 

TJn uu long of (u), E. pool, (punl) 

Uu uu long of (ti) 

:TJu w long of (v) 

Uuh uuh long of («h) 

V V E. real, (viil), F. v, North G. to, see (bh) 

V V =s (v|-), buzz of (/), which see 
.V .V buzz of (.f ), which see 

Vh vh = (v*gh), buzz of (fh), which see 
Yw yW = (v*w), F. roix, (two) 

W w E. iritch, (witsh) 

W w diacritic, labial modification of preceding letter 

j[ m turned m, written la, defective lip tnll, occ. E. veiry 

time, (vera'f turau) 
Wh wh whisper of (w), E. loAich, (whttsh) 

X X Spanish x, j\ QuLrote, Me^co, or Qui;bte, Me;ico, 

(Kiixoo't^, Mee'xiikoo) 
X X buzz of (x) 

Y y = (iw) = (10), F. htrtte, G. Ittcke, (yt, lyk-e) 

T y = (to), Welsh u, and final y, pump, ewyllys, (pymp, 

ewsdhh'ys), E. hous^, goodness, (H9uz*yz, gudnys) 
:Y T Polish, Bohemian, Hungarian y, Eussian (/err) 



z 


z 


z 


t 1 


Zh 


zh 


Zh 


sh 


Zhj 


zhj 


Zj 


2] 


Zs 


zs 


Zw 


zw 


Zwh 


zwh 



10 unPRODUcmoN. 

Ti yi F. l«t, ennw, (lyi, aAnyi) 

Yy yy long of (y), F. fltlte, G. gemath, (flyyt, g«nyyt-) 

Ty yy long of (y) 

: Yt tt long of (t) 

buzz of (b), E. seal, muer, (ziil, moi'zi) 

buzz of («), Newman's and uzoally received A. 1^, Lop- 

sios's A. (jo 
buzz of «A, E. vwion, F. yens, (vizh-wi, zhaA) 

== (zhj), buzz of («h) 

= (zh*y), buzz of (shj) 

= (z*j), buzz of (sj) 

final E. «, s, when fully pronounced, day«, flie«, bu», 
(d^zs, flaizs, bazs) 

= (z*w), see {%u)) 

=e (zh*w), see (sirh) 



2. SiOKS. 

(*) turned comma, when final, simple whisper, as E. bi^, (bit*) ; 
before a vowel, diacritic^ attempt to whisper the vowel, 
as ('a), whispeied (a) ; before a sonant, diacritic, semi- 
vocalise, see (*b, *d, *g) 

(') apostrophe, simple voice, F. abl^, (abl'), E. little, rhythm, 
open =(lit-'l, rtth-'m, oo-p'n), often written (l«t*l, n'th-m, 
oop-n), S. ^ ^ = Ce, '1) 

(") double apostrophe, long of ('), S. ^ ^ « ("b, "1) 

(-) hyphen, read words or letters that are written apart as if 
tiiey were written close, opposed to (,), letter elided, as 
F. nous avons un ami, dit-il k I'homme, (nuz- avoAz- oau- 
ami, dit- il a 1- om) 

(-) minus, before a diacritic, remove its effect from the pre- 
ceding letter in which it is inherent, thus (a?==u-fr means 
that the sound of a? is heard, when (u) is first pronounced 
and then the lips opened 

(l) turned 1, A. \ (jaa'lef), Hebrew K» Gbeek soft breathing (?) 

(,) comma, diaeresis, begin the following letter as if it had no 
connection with the preceding, E. minutisD = (mtniuu*- 
shiji), E. unening, unowned =(9n,er'fq, 9n,oond*) 

(„) double comma, commence the following letter so gently that 
its commencement is difficult to determine, spiritus lenis (?) 

(.) period, pronounce the following letter emphatically 

(.,) period and comma, commence the following letter with great 
abruptness, strongly marked hiatus 

(;) semicolon, open tiie glottis suddenly, A. *■ (nam-za). 



KBY TO PALABOTTPB — SIGNS. 11 

(t) tamed semioolon, dose the glottis suddenly as in stammer- 
ing, or suddenly cease any sound, as when startled, 
leaving a sound half uttered; (hS) is a suddenly checked 
emission of breath, strongly resembling a dick ({), as in 
Zulu (ik.Htwa), FmhU Speech, p. 126. 
8 turned 3, A. ^, bleat haa ^^ (bsagfleg) 

(^) turned comma and apostrophe, speak the following word in 

a subdued tone or voix voiUe. 
(f) turned apostrophe, nasalize the preceding letter, but not as 

in F. nasalisation (a) 
(l) turned !, attempt to pronounce the preceding lettev with 

inspired breatii, (f{, ph;), calling a bird 
(X) attempt to pronounce tiie preceding letter with the air in 
the mouth without inspiring or expiring, click, E. tut =^ 
(tt), E, d*ck (tjSJ) 
S turned 5, Caffir d^tal dick, Appleyaid's ^ = (tj), or (tfj), 

^ as in (iqgbha'tt), Fmble Speech, p. 126. 
S turned 2, Caffir cerebral (Lepsius) or palatal (Appleyard) 
click, Appleyard's q = (ij-t), as in (Egu^al^V), Vieihle 
Speech, p. 126. 
I turned 7, Caffir (uni-) lateral dick, Appleyards x, = (tjSJt) 
wil^ prolonged suction, as in (gaq^an-ji), Fmhle Speech^ 
p. 126. 
f turned 4, Hottentot palatal dick, Boyce's qe, = (tjj) pro- 
bably, Lq>sius'8 Stimdard Alphabet, 2nd ed., p. 79. 
8 turned 8, Waco click»c= (k{), Haldeman, Analytic Ortho- 
graphy, p. 120. 
turned 0, distend the pharynx and cheeks, ^ widen' the 

sound, 
j made from f , take the preceding letter nearer the throat and 

farther from the lips, inner position. 
I- made from f , take the preening letter frirther from the 

throat and nearer to the lips, outer position. 
\ turned f , inyert the tongue so that the under part strikes 
the piedate, when pronouncing the preceding letter, see 
(d, l, n, k, *h, t) 
protrude the tongue when pronouncing the preceding letter, 
bi-lateral, allow the breath to escape on both sides of the 
tongue or mouth, but not over the tip of the tongue or 
through the middle of the mouth, 
made fi^m §, uni-lateral, allow the breath to escape on one 

side of the tongue or mouth only, 
turned ?, trill any free part during the utterance of the pre- 
ceding consonant. 

link, form a new position by attempting to pronounce the 
two letters between which it is placed, at the same instant, 
but giving prominence to the first letter named, see (Ij) 

- (1*0 



12 iirrRODUcnoN. 

*♦ governor, placed between two letters at the beginning of a 
phrase, shews that the first is to be pronounced like the 
second thronghont, indicating a defect of utterance, as 
(1**1,), (1) pronounced with a nasal twang; when no 
letter prcK^edes, it indicates that the effect of tiie following 
letter is heard in all letters, (**.p) close lips, (**tt) pro- 
truded tongue, (*♦,) general nasal quality, (**.') strained 
voice, etc., Vmbls Speech^ p. 81. 

(') turned period, before a word, speak the word emphatically as 

(*Hii did ft, Hii *did tt) ; after a letter, (*) shews that it 

• occurs in an accented syllable, as (bii'tq, m^k'tq, ripooz*) 

(:) colon, before a capital letter, (jr which case it is written 
below it, as o,) shews that it is the capital of a small 
capital letter, see (:E) capital of (e) ; after a letter, shews 
that it occurs in a secondarily accented syllable, as 
(inkom:priHen:sfbtl*ft$, Hoi-wtftftmaBu:) 

^^^^^^^ written under a word indicates spaced letters, ttsed 
to give prominence to a word in palaeotype, answering 
to ittUicB in ordinary printing. 



Following a Word. 

(..) low level tone, C. high (pmq) 

{'•S high level tone, C. low (pniq) 

(.*) rising tone, G. high (shaq) 

(..•) tone rising from low pitch, C. low (shaq) 

(.-.) rise and fall, circumflex, C. (fdTkjen shaq) 

(*.) falling tone, G. high (knoece, kmu, km) 

(•..) falling tone to low pitch, C. low (kHOBoe) 

(*.') fall and rise, inverted circumflex 

{i') stop voice in high pitch, C. high (shuf, zhif , n^ip') 

(i.) stop voice in low pitch, G. low (diui, zhii, njipf) 

Preceding a Word, 
(':') speak in a high key 
(.:.) speak in low key 

Palaeottfe Ain) Visible Speech Goxpased. 

The diagrams on p. 14, transferred by Mr. Melville Bell's per- 
mission from p. 8 of his English Vieihle Speech^ will be the best 
guide to the pronunciation of the vowels. Each of the first nine 
diagrams represents the position of the tongue for the four vowels 
written below it. For the first and third vowels in each diagram, the 
passages behind the nairowcst part of l^e channel formed by the 
tongue are in the usual condition, but for the second and fourth 
vowel in each diagram, they are distended, making the vowels 
' wide.' For the first and second vowel in each diagram, the lips are 
open. For the third and fourth vowel in each dia^am, the lips are 
more or less rounded, — ^namely, for Nos. 1, 2, 3, as in No. 10, for 



PALAEOTTPE AND YISIBLB SPEECH CX>MPAEBD. 13 

Kos. 4, 5, 6y as in No. 11, and for Nos. 7, 8, 9 as in No. 12. As the 
principal interest in the following investigation attaches to changes 
in the vowel system, a caiefnl study of these diagrams will be of 
material assistance. If any reader pronounce the key words with a 
vowel requiring a different position from that here pointed out, his 
pronunciation differs from the author's, and the value of the symbol 
is to be determined from the diagram in preference to the key word. 

In order to fix the value of the palaeotypic letters, they are on 
p. 15 compared with those of Mr. Melville Bell's VtsihU Speech^ by 
means of his " Cosmopolitan Telegraphic Table," which has been 
here reprinted by his permission. The figures indicate the columns 
and the letters the lines. The following is Mr. Bell's classification, 
which will be frequently alluded to. 

Columns 1, 2, 3, 4 contain consonants, lines a, b, e, t^ e, / axe 
wieelessy lines y, A, i, k, I, m, are voiced; lines a, y are primary, 
lines b, h are mixed, lines e, i are divided, lines d, h, are mixed 
divided, lines e, 1 9ie shut, lines/, m are nasal. 

Column 5 consists of glides, which are represented in palaeotype 
on a different principle, see below. Chapter III, § 2. The letter 
(h), 5f, is considered as the true English aspirate in palaeotype, but 
Mr. M. Bell considered (h*), or 9a, to be the more correct form. 

Columns 6, 7, 8 are vowels, column 6 back vowels, column 7 
mixed vowels, column 8 front vowels, and in each column lines a, b, 
e, are primary, lines d, e, f are wide, lines g, h, i are round, lines 
k,l, m are wide round, lines a, d, g, k are high, lines 5, $, h, I are 
mid, and lines c, /, i, mere low vowels. 

Columns 9, 10 contain the aspirates and modifiers. 

GLOSSOTTPE. 

An investigation of historical English spelling in Chapter YI, § 3, 
suggested the possibility of enlarging the alphabet required for 
writing the theoretically received pronunciation of literary English, 
so as to meet the requirements of writers of our provincial dialects, 
who endeavour to preserve the analogies of ordinary spelling. It 
was found necessary to deviate from these slightly for the repre- 
sentation of our complicated diphthongal system, and some foreign 
sounds, which occur provincially, but are imrecognized in our or- 
thography.' The use of the short mark C*') to indicate the provin- 
cial shortening of vowels generally long in the literary dialect, and 
of the long mark (") for the lengthening of vowels generally short, is 
hardly a deviation from ordinary usage. The principles of tliis 
scheme are explained in Chapter YI, ^ 3, where the exact value of 
the letters is explained, and its use is exemplified in Chapter XI. 
But for convenience, a very brief key is- given on p. 16. The name 
Glossottte refers to the chief use for which it was intended — ^the 
writing of provincial Glossaries. It is hoped, however, that such 
a scheme, although designedly incomplete, may be found usefol to 
all who may occasionally wish to indicate pronunciation with some 
degree of exactness, but do not care to enter upon general phonetic 
investigations. 



14 



nCTBODUCnOH* 



LINGUAL POSITIONS OF THE' VOWELS. 



HaL 

ce, B, u, u. 

ft 




No.S. 


Ho. 4. 

a, a, o, o. 


9y ah| ^hy oh. 


No. 6. 

'. e, *, ce. 


NcT. 
CE, a. A, 0. 


No. 8. 
9h, 30, ^zh, oh. 


Ha 9. 



LABIAL POSITIONS OF THE VOWELS. 



No. 10. 




u, p ; u, «A ; i,y. 



No. 11. 




#, o; ^h,oh; p, oe. 



Naia. 




A,o; ah, oh; ^h,aeh. 



PALABOTTPE ADD VCOBLK SPEECH OOHFARED. 



15 



Mb. MELVILLE BELL'S VISIBLE SPEECH LETTEBS. 





I 


2 


3 


4 


5 


"is- 


7 


T" 


9 


"0 




a 


c 


O 


O 


ro" 


I 


1 


I 


I 


"o" 




a 


b 


G 


a 


U 


D 


I 


] 


I 


t 





\ 


r 


c 


C 


o 


CO 


3 


I 


JL 


J- 


_L 


X 


i 


c 


PS 


e^ 


3 


J 


1 


r 


~ 


i 


t 


e2 


e 


a 


Q 


a 


D 


2 


3 


X 


C 


f 


. 


6 


j_ 


a 


Q 


o 


O 


> 


_L 


I 


I 


o> 


< 


a 


e 


o 


<i) 


3 


i 


i 


i 


{ 


i 


^ 


.^ 


h 


e 


^ 


Cl3 


3 


f 


i 


\ 


{ 


e 


<* 


h 

i 


i 


8 


CD 


© 


3 


^ 


J_ 


I 


t 


{ 


c 


k 




so 

CD 


ei3 






i 


I 




» 


D 


k 


} 


I 


£ 


n 


1 


I 


m 


Q 


CD 


G) 


O 


I 


i 


i J 


L-^- 


1 


^ 


m 




1 


2 


3 


4 


6 


6 


7 


8 


9 


|0 





PALAEOTTPIO EQUIVALENTS OP VISIBLE SPEECH 
LETTERa 





1 


2 


» 


4 


6 


6 


7 


8 


9 







a 


kh 


jh 


rh 


ph 




(^ 


Y 


i 


H' 




a 


b 


k«,h 


8 


Bh 


wb 


r 


a 


9 


^ 


< 


doub 


b 


c 
d 

e 


l\x 


Ijh 


Ih 


f 


1,J 


(E 


9h 


E 


> 


> 


c 
d 


Ivoh 


th 


th 


fh 


J 


B 


^ 


i 


< 


•> 


k 


tj 


t 


P 


la 


a 


ah 


e 


A 


( 


e 


l_ 


±. 


Djh 


nh 


mh 


H 


^ 


eo 


ae 


iitt?h 


i 


Z 


9 
h 


g^ 


J 


r 


TT 


w 


u 


u 1 


I 


i 


ui 


ff 


gioh 


z 


zh 


w 


rw 





^h 


^ 


4» 


t 


h 


t 


/ 


u 


1 


V 


y 


A 


ah 


^h 


1 


i 


t 


k 


Iw 


dh 


dh 


vh 


aw? 


u 


uh 


y 


r 


t 


I 


s 


d) 


d 


b 


u,w 





oh 


oe 


• 


§8 


I 


m 


s 


",j 


n 


m 


*ti 





oh 


£h 


» 


• 


m 




1 


2 


3 


4 


5 


^ 


7 


8 


9 1 





16 



IMTRODUCnOK. 



Key to Glossottpe. 

See p. 13. Isolated letters and words in glossotjpe should be inolosed in ( ). 

i£) is never mute; all vowels and combinations having C*') or (~) over them, except 
tt)^ are the short or lonf sounds of the vowels and comoinations without these marks, 
which should not be used for any other letters, thus : (A) is the long sound of (a) ; (^) 
the short sound of (ee) ; (tt) is to be used whenever it is thought that the pn^^er form 
(ou) might create concision. 

(7. Cocbiey, D, Butch, B, English, F, French, (?. German, /. Italian, P. Provincial, 
8, Scotch, Sw. Swedish, W, Welsh. 



VOWBLB. 



a P. 

aa ask 
^ask 
ae ware 
U8.e 
ah father 
thF.O.S.a 
ai watt 
&i 8. ai 
air F. an 
ao 8. 

&o S, man 
aa aU 
&a want 
e net 
e 8* 
ee meet 
«e 8J.F. 
ew F. in 
eu,F.eu 
4uF.eu 



When more than two vowels 
come together and the first 
two form one of the pre- 
ceding combinations, read 
them as such, as (reeent'er 
( ssree-ent'er) =rd-«i<#r. 



i knit 
IS. 
ih,rhP.G.il 
not 
oP. 

oa, 6a 7. 6 
oe, 6e 0. 6 
oh rose 
6hS. 
ON F. on 
00 pool 
6o 8. hook 
oUf a cotdd 
buF. 
u nut 
uP. 

ue,iiei9tr.« 
uh worth 
uhP. 
ui, iii F. u 
UN F. un 

C) 



DiPHTHONOS. 



aiy may 
aj 8.C. 
aay high 
aey 8, 
thj O. ai 
ahj aye 
auy P. 
ey 8. tide 



oy hoi/ 
oyP. 
ohyP 
oojZEF. 
uy high 
uiy F. ui 
euy F. eui 



aiw C. 
aw C. 
aaw how 
aew C. 
Hhw O, au 
ahwP. 
auw P. 
ew7. eu 
eew J. iu 
iw mew 
owP. 
owP. 
ohw know 
•••••• 

uw how 

euw 2>. 



In all these diphthongs 
the first element has the 
sound assigned in the 
preceding column, which 
IS run on quickly, with a 
glide, to a following (ee) 
or (oo) written (y) or (w). 

Diphthongs are also formed 
f, by affixing (*) as 
(roh*^ almost (rohud) = 
roadj and by affixing (ui), 
which should then l>e 
written ({ii), as 2). (HeuOis) 
s huit, tiieoretical (?. 
(froQind) ^ freund. 



Consonants. 



b hee 
eh chest 
d doe 
^the 
i fee 

g 9<^ 
ghD.Q. 

H he 
(written h) 

j y«y 

k 000 

kh Q.C.ch 

I lo 

1 UtUe 

Ih W.ll 

m me 

*m rhythm 

n no 

'n op^n 

N P. n 
(written »^) 

ng ^Ain^ 

Foreign and Oriental sounds 
represented by Italics and 
small capitals, by special 
convention. 

Accent the first syllable, un- 
less or (') is written after 
some other syllable, as: 
august, august*, august'. 



n-g tngratn 

nk think 

n-k in-eome 

P PM 

T ray 

'r air 

r 1.8. r 

rh P.P. r 

B see 

sh she 

t tin 

th thin 

V vale 

w wail, or 

-w (after 
vowels) 

wh why 

y yet, or 

-y (after 
vowels) 

z zeal 

zh. vision 



17 



CHAPTER I. 
On Pronunciation and its Changes. 

Thought may be conveyed from mind to mind by various 
systems of symbols, each of which may be termed language. 
A real, living, growing language, however, has always been 
a collection of spoken sounds, and it is only in so far as they 
indicate these sounds that other symbols can be dignified 
with the name of language. But a spoken sound once 
written ceases to grow. Even when an orthography is 
chosen which varies with the sounds from day to day, each 
written word is, as it were, but an instantaneous photograph 
of a living thing, fixing a momentary phase, while the organ- 
ism proceeds to grow and change till all resemblance to the 
old form may in course of time be obliterated. The systems 
of writing which have been generally adopted, far from 
acknowledging this fact, force us, as it were, to recognize 
mature or ancient men from the portraits of youths or 
children, and ignore the ever-active irrepressible vitality of 
language. We G^>eak of the " dead" languages of Rome and 
Athens, imconscious that our own English of a few years 
back has become as dead to us, who can neither think in the 
idiom nor speak with the sounds of our forefathers. 

Spoken language is bom of any two or more associated 
human beings. It grows, matures, assimilates, changes, incor- 
porates, excludes, developes, languishes, decays, dies utterly, 
with the societies to which it owes its being. It is difficiUt 
to seize its chameleon form at any moment. Each speaker 
as thought inspires him, each listener as the thought reaches 
him with the sound, creates some new turn of expression, 
some fr^sh alliance of thought with sound, some useful modi- 
fication of former custom, some instantaneous innovation 
which either perishes at the instant of birth, or becomes part 
of the common stock, a progenitor of future language. The 
different sensations of each speaker, the different apprecia- 
tions of each hearer, their intellectual growth, their environ- 
ment, their aptitude for conveying or receiving impressions, 
their very passions, originate, change, and create language. 



18 PRONUNCIATION AND ITS CHANGES. Chap. I. 

Without entering on the complex investigation of the 
idiomatic alterations of language, a slight consideration will 
shew that the audible forms in which these idioms are clothed 
will also undergo great and important changes. The habit 
of producing certain series of ^ken sounds is acquired 
generally by a laborious and painml process, beginning with 
the first dawn of intelligence, continued through long stages 
of imperfect powers of appreciation and imitation, and be- 
coming at last so fixed that the speaker in most cases either 
does not hear or does not duly weigh any but great devia- 
tions from his own customary mode of speech, and is rendered 
incapable of any but a rude travesty of strange sounds into 
the nearest of his own familiar utterances. 

We may apparently distinguish three laws according to 
which the sounds of a language change. 

First, the chronological law. Changes in spoken sounds 
take place in time, not by insensible degrees, but per 
saltum, from generation to generation. 
Second, the indkndual law. A series of spoken sounds 
acquired during childhood and youth remains fixed in 
the individual during the rest of his life. 
Third, the geographical law, A series of spoken sounds 
adopted as the expression of thought by persons living 
in one locality, when wholly or partly adopted by an- 
other community, are also changed, not by insensible 
degrees, but per saltum, in passing from individual to. 
individual. 
At any one instant of time there are generally three gene- 
rations nving. Each middle generation has commenced at 
a difierent time, and has modified the speech of its preceding 
generation in a somewhat different manner, after which it 
retains the modified form, while the subsequent generation 
proceeds to change that form once more. Consequentlv 
there will not be any approach to uniformity of speech 
sounds in any one place at any one time, but there will be a 
kind of mean, the general utterance of the more thoughtful 
or more respected persons -of mature age, round which the 
other sounds seem to hover, and which, like the averages of 
the mathematician, not agreeing precisely with any, may for 
the purposes of science be assumed to represent all, and be 
call^ the language of the district at the epoch assigned. 
Concrete reality is always too complex for science to grasp, 
and hence she has to content herseli* with certain abstractions, 
and to leave practice to apply the necessary corrections in 
individual cases. Thus, if we descended into every minute 



Chap. I. PBONUNCIATION AND ITS CHANGES. 19 

sbade of spoken sound, the variety would be so interminable, 
each individual presenting some fr&ii peculiarities, that all 
definite character would be lost. In actual life this necessary 
abstraction is replaced by the second law which gives fixed- 
ness of utterance to the individual, regardless of surrounding 
change. Indeed, few persons of mature years, even in the 
most civilized communities, think of the sounds they utter. 
They speak to communicate thought, not to examme the 
instrument which they employ for. that purpose, and they 
would be constantly checked, and irritated by thinking of 
how they speak, rather than of what they speak. 

It is this individual fixity of habit, and powerlessness of 
adaptation that operates in producing the per saltum geogra- 
phical changes, in which must be included, not only the 
changes made in foreign words, but also those resulting 
fix)m any society within a society, — schools, colleges, cliques, 
coteries, professions, trades, emigrations, — in short any means 
of isolating some companies of speakers from others. Slang 
is only a form of dialect. 

One marked result of the third law is that a \miform 
system of spoken sounds cannot extend over a very large 
(fistrict. AU the speakers must have frequent opportunities 
of hearing the sounds from youth up, or they will be unable 
to appreciate and imitate them. Education, which sends 
teachers as missionaries into remote districts to convey the 
required sounds more or less correctly, but, more safely and 
certainly, rapid communication of individuals, such as rail- 
roads now effect, does much to produce uniformity of speech. 
How far, however, even in small, educated and locomotive 
England we are yet removed from uniformity of speech, may 
be learned by a very slight attention to the sounds heard in 
different districts, each of which has its own characteristic 
burr or brogue, less marked perhaps than it was in Higden's 
and Cax ton's time, but still unmistakable.^ 

The results of emigration and immigration are curious and 
important. By emigration is here specially meant the sepa- 
ration of a considerable body of the inhabitants of a country 

* Treuisa in hia translation of Hi?- to relate how when " certayn mer- 

dcn*8 Polychronicon, 1386, says **afle chatuites .... taryed atte forlond . . . 

^ lanniges of ]-e nor^hnmhrrs k and axed for mete, and specyaUy . . . 

ffiecialich at jfixVe is so scharp slittinge axyd after eggys ... the gooae wyf 

airotyng^&vnschape; }>at we sou}ffren answerde that she coude speke no 

men may ^at langage Tnne)>e vndfr- frenshe .... and thenne at last a 

stonde.*' And Caxton (Prologue to nother sayd that he wolde haue eyren, 

£nrydos) complains that *' comyn £n- then the good wyf sayd that she ynder- 

glysshe that is spoken in one shyre stod hym.*' See Chapter XI for ex- 

vnyeth from a nother,'* and goes on isting Tarieties of pronunciation. 



20 PRONUNCIATION AND ITS CHANGES. Chap. I. 

from the main mass, without incorporating itself with another 
nation. Thus the English in America have not mixed with 
the natives^ and the Norse xn Iceland had no natives to mix 
with. In this case there is a kmd of arrest of development, 
the language of the emigrants remains for a long time in the 
stage .at which it was when emigration took place, and alters 
more dowly than the mother ton^e, and in a different 
direction. Practically the speech of the American English 
is archaic with respect to tnat of the British English, and 
while the Icelandic scarcely differs from the old Norse, the 
latter has, since the colonization of Iceland, split up on the 
mainland into two distinct litera^ tons^ues, the Danish and 
Swedish. Nay, even the Irish English exhibits in many 
points the peculiarities of the pronunciation of the xvii th 
century. 

By immigration, on the other hand, is meant the introduc- 
tion of a comparatively small body into a large mass of 
people, with whom they mix and associate. This may be 
commercially (as when German emigrants settle in the 
United States), or by conquest (as when the Norsemen settled 
first in the north of France, and secondly in England, or 
when the Goths ruled in Italy). In these cases the immigrant 
language is more or less lost cmd absorbed, especially if it is 
not so developed as the language among which it enters, and 
into which it introduces comparatively little change. The 
French element of our language, for example, is only indi- 
rectly traceable to the Norman Conquest^ for we find it very 
slightly marked, even in the xiii th century. The Boman 
occupation of England and the English domination in India 
have produced very little effect upon eith^ the immigrant 
or receiving language, principally from the want of associa- 
tion. The languages have remained practically unmixed. 
The Roman language in France and Spain de facto ousted 
the Celtic of the inhabitants, and, after natural chan^, 
altered by the absorption of the Frankish and Moorish im- 
migrations. 

The alterations thus introduced into a language produce 
but little effect on the idioms (that is, the expression of the 
relations of conceptions), but principally affect the words 
employed. Thus English has remain^ a Low German 
dialect through all the introductions of French, Latin, and 
Greek elements^ and French, Spanish, and Italian remain 
Latin notwithstanding the Frankish, Moorish, and Gothic 
additions which they have received. But in all these 
languages great changes have £Etllen upon the forms of the 



Chat. I. PRONUNCIATION AND ITS CHANGES. 21 

words used. We are apt to regard (btsh'ep, bish'of, bis-po, 
Yes'kovo, dVEEk, obhiis'po, epiis'kop, epis'kopus, epis'kopos) 
as entirely different words, and to call (br^ek briik, keez kiiz, 
oblaidzh* obliidzh') etc., different pronunciatkxis of the same 
words. But the latter are reaUy only less marked examples 
of the same phenomenon as is exhibited in the former. If 
the latter pairs of words are to be regarrded as the same, the 
former nine must also be classed as one. In the latter we 
have chiefly chronological^ in the former we have chiefly 
geographical changes. In both cases we have examples of 
the variation of one sound as it passes through various 
mouths — volitat vtw^ per ora virum. 

Even without reference to written forms, the conception 
of altered forms of one original sound (that is, of various 
pronunciations of the same word), naturally arises in men's 
minds, but when languages come to be written as well as 
^ken, this is more strongly forced upon them — at least in 
those cases which the writing notices. Writing, that won- 
derful method of arresting sound which has made human 
memory independent of life, and has thus perpetuated know- 
ledge, was necessarily at first confined to the learned alone, 
the priest and the philosopher. These fixed, as nearly as 
they could appreciate, or their method of symbolisation, 
which was necessarily insufficient, would allow, the sounds 
of their own language as they heard them in their own day. 
Their successors venerating the invention, or despairing of 
introducing improvements, trod servilely in their steps and 
mostly used the old symbols while tne soimds changed 
around them. Within the limits of the powers of the old 
symbols some changes were made from time to time, but 
very slowly. Then in quite recent days, the innovation of 
diacritical signs arose as in French and German, whereby a 
modem mo<Sfication of an ancient usage was more or less 
indicated. Occasionally, whole groups of letters formerly 
correctly used to indicate certain sounds came to be con- 
sidered as groups indicating new sounds, — ^not in all cases, 
but in many perhaps, where the sounds had changed by re- 
gular derivation. Before the invention of printing, writers, 
become more numerous, had become also less controlled by 
the example of their ancestors, and endeavoured as well as 
they could, with numerous conventions, inconsistencies, im- 
peifections, and shortcomings, rendered inevitable by the 
inadequacy of their instrument, to express on paper the 
sounds they heard. When we are fortunate enough to find 
the real hcmdy work of a thoughtful writer, as Orrmin, we see 



22 PRONUNCIATION AND ITS CHANORS. Chap. I. 

how much might have been done to clear our mode of writing 
from inconsistencies. But with the invention of printing, 
came a belief in the necessity of a fixed orthography to 
facilitate the work of the compositor and reader. The re- 
gulation of spelling was taken from the intellectual and given 
to a mechanical class. Uniformity at all hazards was the 
aim. And uniformity has been sained to a s^reat extent in 
late years, but at a sacrifice which uniformity is far from 
being worth — loss of a knowledge of how our ancestors spoke, 
concealment of how we ^eak at present, innumerable diffi- 
culties to both reader and writer, and hence great impedi- 
ments to the acquisition of knowledge. The numerous 
societies for printing old English books which are now at 
work, and especially the Early English Text Society , have, 
by conscientiously printing manuscripts literatim^ done much 
to restore our knowledge of ancient sounds as well as ancient 
sense. But the veil of our modem spelling lies over our 
eyes, and it is not easy to gain the key to the mystery which 
these texts are calculated to display. 

" Nobody," says Archdeacon C. J. Hare,^ " who has a due 
reverence for his ancestors or even for his own spiritual 
being, which has been mainly trained and fashioned by his 
native language, — nobody who rightly appreciates what a 
momentous thmg it is to keep the unity of a people entire 
and imbroken, to preserve and foster all its national recol- 
lections, what a glorious and inestimable blessing it is to 
* speak the tongue that Shakspere spake,' will ever wish to 
trim that tongue according to any arbitrary theory." But 
the English of to-day do not know * the tongue that Shak- 
spere spake.' They may be familiar with the words of his 
plays according to their own fashion of speech, but they 
know no more how Shakspere would have uttered them than 
they know how to write a play in his idiom. The language 
of Shakspere has departed from us, and has to be acquired 
as a new tongue, without the aid of a living teacher. What 
this means can only be justljr appreciated by observing how 
foreigners, after most laborious study of our own modem 
language from books and grammars, proceed to write and 
speak it. You will read and hear whole sentences in which 
every phrase shall be in accordance with grammar, and yet 
perhaps not a single sentence so composed as an Englishman 
would have penned it, or so uttered as an Englishman would 
have spoken it. A langua^ can only be learned by ear. 
But how did our glonous old writers speak P What 

^ On Englibh Orthography^ Philological Museum, Vol. 1, p. 645. 



Chip. I. PBONUNCIATION AND ITS CHANGES. 23 

sounds did Goldsmith, Pope, t)ryden, Milton, Shakspere, 
Spenser, Chaucer, Langland, call the English language? 
Or if we cannot discover their own individual peculiarities, 
what was the style of pronunciation prevalent at and about 
their time among the readers of their works ? The inquiry- 
is beeet with difficulties. It would be almost impossible to 
determine the pronunciation of our contemporary laureate, 
but surely with our heap of pronouncing dictionaries, it 
would seem easy to determine tlmt of his miders. Yet this 
is far from being the case. It is difficult even for a person 
to determine with accuracy what is his own pronunciation. 
He can at best only give an approximation to that of others. 

In the present day we may, however, recognize a received 
pronimciation all over the country, not widely diflTering in 
any particular locality, and admitting a certain degree of 
variety. It may be especially considered as the educated 
pronunciation of the metropolis, of the court, the pulpit, and 
the bar.^ But in as much as all these localities and pro- 
fessions are recruited from the provinces, there will be a 
varied thread of provincial utterance running through the 
whole. In former times this was necessarily more marked, 
and the simultaneous varieties of pronunciation prevalent 
and acknowledged much greater. In the xiiith, xivth, 
and XV th centuries it is almost a straining of the meaning 
of words to talk of a general EnjgLish pronunciation.* There 
was then only a court dialect of the south, and the various 
" uplandi" northern, eastern, and western modes of speech. 
And hence we can only seek to discover the court dialect, 
and then, having partly ascertained the value of the letters, 
endeavour to ascertain the pronunciations meant to be in- 
dicated by such writers as Dan Michel and Orrmin.® 

But how are we to arrive at a knowledge of the court 
dialect P Molidre ridicules the notion of having a master to 
teach pronnnciation, and certainly the analysis of speech 
sounds, was at no time, and is not even at the present day, 
notwithstanding the appearance of so many treatises in quite 
recent times, down to that of Mr. Melville Bell, 1867, a 
favorite subject of investigation. It is voted tiresome or 
unnecessary, and the greater number of even those who 

' The pronunciation of the stage is mm modnm loqnendi solum sum secu- 

inclined to be archaic, except in the tus, quern solum ah infancia didici, 

modemest imitations of evenr day life. et solotenus plenius perfectiusque cog- 

* Thus in 1440 the author of the novi." 

Promptorium l^nmlorum sajSj ** Conn' ' The subject of a standard pro- 

tatas Northfolcie** or^ according to nunciation is specially considered below, 

another reading, " Orientalium A^lo- Chap. YI, { 6. 



24 PRONUNCIATION AND ITS CHANGES. Chap I, 

touch upon it incidentally, in grammars and orthoepical 
treatises, are profoundly ignorant of the nature and mechan- 
ism of speech, and the inter-relations of the sounds which 
constitute language.^ The consequence is that writers being 
imaware of the mechanism by which the results are produced, 
were constrained to use a variety of metaphorical expressions 
which it is extremely difficult to comprehend, and which 
naturally have different meanings in the works of different 
authors. Thus sounds are termed thick, thin, fat, full, 
empty, round, flat, hard, soft, rough, smooth, sharp, clear, 
obscure, coarse, delicate, broad, fine, attenuated, mincing, 
finical, aflected, open, close, and so on, till the reader is in 
despair. For example, in English, German, Italian, Spanish, 
'hcurd c* is (k), but 'soft c' is (s) in English, (ts) in Ger- 
man, (tsh) in Italian, (c), that is, nearly (th), in Spanish. The 
Germans call (g) the 'soft' of (k), and (^h) the 'soft' of 
(g). But the English call (ff) 'hard g,' and (dzh) 'soft g,' 
and ' soft g' is (x), or nearly (kh), in Spanish. Most writers 
term (s, th) hard sounds, and (z, dh) soft, but Dyche' finds 
(s, th) soft, and (z, dh) hard. One writer calls o obscure 
when it sounds as (a) or (uu), no matter which, but p final 
obscure when (i), and sharp and clear when (ei). 

Some writers, again, content themselves with using key 
words. This is indeed the easiest method for the writer, and 
conveys very fair notions to contemporary readers. It has 
been adopted in the description of Paiaeotype to avoid prolix 
explanations. But the publication of Mr. Melville ^Bell's 
Visible Speech has enabled me by referring to his symbols to 
fix the sounds with accuracy, for Visible Speech contains an 
exact account of the disposition of the organs for producing 
the sounds, and hence by carefully studying that work at 
any time— centuries hence — the exact sound could probably 
be recovered. Not so with key words, for they involve the 

* The beantifiil phonetic short-hand Bounds, bnt with very small saccess, 
invented by Mr. I. Pitman, under the even amon^ those who were most 
name of Phonography^ and developed earnest in die use of phonetic types 
by the assistance of many co-workers, as an educational appliance. The sub- 
gave rise to a desire to print phoneti- ject was not sufficiently attractive. At 
caUy, in consequence of which a pho- present Mr. Melville BeU*s recent 
netic English alphabet was invented treatise on VitibU Speeehy renders a 
hj Mr. I. Pitman and myself, which, study of the whole subject compara- 
with various subseauent modifications, tively easy. And he has supplemented 
has been extensively used in England it by a svstem of shorthand writing 
and America. From the first I en- which will be applicable with almost 
deavoured Tin my treatises on the equal fiicility to all lang^uages in the 
Alphabet of Nature^ 1845, and Em^H' world, rendering his svstem extremely 
tiaU of Fhonetiesy 1848,) to make this easy to write even at full, 
alnhaliet a means of extending a know- * Omd4 to th4 EngliMh Jhrtfue, 1710. 
leoge of the inter-relations of speech 



Chaf. I. PRONUXCIATION AND ITS CHANGES. 25 

very riddle which we have to solve. Only those who, like 
the present writer, have spent hours in endeavouring to dis- 
cover what was meant by a simple reference to a key word 
given three hundred years ago, can fully appreciate the ad- 
vantage of an exact description like that furnished by Visible 
Speech.^ There is some relief when many key-words are 

fiven, or when contemporary languages are cited. But 
ere the imperfect appreciation of uie citer is painfully con- 
spicuous, and allowances have always to be made on that 
account. Many writers, too, content themselves with re- 
ferences to the Latin, Greek, or Hebrew sounds, apparently 
forgetting that the older pronunciation of these languages is a 
matter of dispute, and that the modem pronunciation varies 
from country to country and century to century. Let any 
one begin by studying Sir T. Smith, Hart, Bullokar, Gill, 
and Butler, in order to determine the pronimciation of 
Shakspere from these sources alone, — or even with the as- 
sistance of Palsgrave, — -and he will soon either find himself 
in the same slough of despond in which I struggled, or will 
get out of his difficulties only by a freer use of hypothesis 
and theory than I considered justifiable, when I endeavoured 
to discover, not to invent, — to establish by evidence, not to 
propound theoretically, — the English pronunciation of the 
XVI th century. 

The first ray of light came to me from a comer which had 
hitherto been very dark. While searching for information, 
some book or other led me to consult William Salesbury's 
Welsh and English Dictionary, 1547. The introduction 
contains a very short and incomplete introduction to English 
pronunciation, written in quaint old Welsh. My imperfect 
knowledge of the language was sufficient for me to perceive 
the value of this essay, which mainly consisted in the 
transcription of about 150 typical English words into Welsh 
letters. Now the Welsh alphabet of the present day is re- 
markably phonetic, having only one ambiguous letter, ^, 
which is sometimes (a) , or (a), and at others {y) . Did Salesburv 
pronounce these letters as they are now pronounced in Nortn 

^ At the latter end of his treatise of the speech organs, — or if possible 

Mr. Melville Bell has given in to the also from the living voice of some one 

practice of key words, and assigned thoroughly acquainted with the system 

them to his symbols. Let the r^er — and then determine Mr. Bell's own 

be carefnl not to take the value of the pronunciation of the key word from 

Smbol from his own pronunciation of the known value of the symbol. This 

e key words, or from any other per- pronunciation in many instances differs 

son's. Let him first determine the nx)m that which I am accustomed to 

value of the symbol from the exact give it especially in foreign woxds. 

description and diagram of the position £otii of us may be wrong. 



26 PRONUNCIATION AND ITS CHANGES. Chap. I. 

Wales P Most fortunately he has answered the question 
himself in a tract upon Welsh pronunciation written in 
English, and referring to many other languages to assist the 
English reader. The result was that with tne exception of 
y, the soimds had remained the same for the last 300 years. 
Here then we have a solid foundation for future work, — ^the 
pronunciation of a certain number of words in the xvi th 
century determined with considerable certainty ; and Ax>m this 
we are able to proceed to a study of the other works named, 
with more hope of a satisfactory result. These tracts of 
Salesbury are so rare, and one of them so little intelligible 
to the mass of readers, that at the suggestion of the Philo- 
logical Society, they will be transferred to the pages of this 
essay, — the English treatise almost entire, the Welsh treatise 
complete with a translation.^ 

The pronunciation of English during the xvi th century 
was thus rendered tolerably clear, and the mode in which it 
broke into that of the x\aith century became traceable. 
But the XVII th century was, like the xv th, one of civil war, 
that is of extraordinary commingling of the population, and 
consequently one of marked linguistic change. Between the 
XIV th and xvith centuries our language was almost bom 
anew.* In the xvn th century the idiomatic changes are by 
no means so evident, but the pronunciation altered distinctly 
in some remarkable points. These facts and the breaking 
up of the xvii th into the xviii th century pronimciation, 
which when established scarcely differed from the present, 
are well brought to light by Wallis, Wilkins, Owen, Price, 
Cooper, Miege, and Jones, followed by Buchanan, Franklin, 
and Sheridan. It became therefore possible to assign with 
considerable accuracy, the pronunciation of Spenser, Shak- 
spere, Milton, Dryden, and Pope, or rather of their con- 
temporaries. 

This was much, but it was not enough. No treatise on 
Early English pronimciation could be satisfactory which did 
not include Chaucer. But here all authorities failed. Pals- 
grave is the earliest author from whom we learn distinctly 
how any English sound was pronounced, and then onlv 
through the analogy of the French and Italian. Two princi- 
ples, however, suggested themselves for trial. In tracing 
the alteration of vowel sounds from the xvi th through the 
XVII th to the XVIII th century a certain definite fine of 
change came to light, which was more or less confirmed by 
a comparison of the changes, as far as they can be traced, in 

1 See Chapter VIII, J§ 1 and 2. « See Chapter IV § 1. 



Chap. I. PRONUNCIATION AND ITS CHANGES. 27 

other languages. Hence the presumption was that from the 
XIV th to the XVI th centuries, if the sounds had altered at all, 
they would have altered in the same direction. But a second 
principle was necessary to make the first available. This 
was found in the fact that since writing was confined to a 
comparatively small number of persons, the majority of those 
who heard and enjoyed poetry would be ignorant of the 
spelling of the words. Hence the rhymes to be appreciated 
at all must have been rhymes to the ear, and not the modem 
monstrosity of rhymes to the eye. If we could have a manu- 
script in Chaucer's own handwriting, we should therefore 
expect to find all the rhymes perfect. Hence we might 
conclude that when two words rhymed together in one of 
Chaucer's couplets, they also rhymed together in his pro- 
nunciation, and if they would not have mymed together in 
the XVI th century, one of them must have altered in the 
definite line of change already discovered. In conformity 
with these principles the whole of the rhymes in Chaucer's 
Canterbury Tales as exhibited in the best available manu- 
script, together with those in all his other poems as edited 
by Mr. Morris, and those in Gower's Contessio Amantis, 
have been carefully examined, and a system of pronuncia- 
tion deduced for the xiv th century.^ 

Much uncertainty must necessarily prevail concerning the 
pronunciation of English from 1400, the death of Chaucer, to 
1530, the date of Palsgrave's French Grammar, as the 
changes were numerous and rapid, both in language and 
pronimciation. Similarly if we had lost the xvii th century 
books on English pronunciation, it would have been impos- 
sible to restore it, from a knowledge only of the pronuncia- 
tions in the xvi th and xvni th centuries. But standing on 
the secure groimd of the xivth century we can, without 
much doubt penetrate into stiU more remote regions, espe- 
cially with the help of Orrmin's orthography, which lands 
us into Anglosaxon. 

Before proceeding to the detailed investigation, it may be 
convenient to present the main results in a tabular form. 
This has been attempted in the merest outline, on the two 
following pages. An explanation of the construction of the 
table is added on p. 30. 

^ For a detafled account of this inTestigation, Bee Chapter lY. 



28 



PRONUNCIATION AND ITS CHANGES. 



Chap. I. 



Modem Spelling 


Chancer 

XIV 


Shaks] 
Spenser 

XTI 


Kmton ^0P« 

Dryden Goldsmith 

xvn xvm 


a short 


a 


a 


n 


8B 


along 


aa 


aa 


8Bfl9 


ee 


aiy ay 


ai 


ai, aai 


aesei, ee 


eei, ee 


auy aw 


an 


an, aaa 


AA 


AA 


e short 


e 


e 


e 


e 


tf long 


ee 


ee, ii 


ee,ii 


ii 


ea 


ee, e 


ee, e 


ee, e 


ii,e 


ee 


ee 


ii 


ii 


ii 


ei.ey 


ai 


ei, eei, ai 


eei, ee 


eei, ee, ii 


m, ew 


en,yy 


yy, en 


in, en 


in 


gh 


kh 


kh,H* 


H* - 


- - 


i, y short 


» 


1 


t 


i 


Sylong 


ii 


ei,ai 


ai 


ai 


short 


0, n 


0, u 


A, 0,9 


0,0 


olong 


00 


00 


00 


00 


oa 


00 


00 


OOy kh. 


OOf AA 


oi.oy 


ui 


oi, ni 


Ai, oi ; ni, oi 


oi 


00 


00 


nu, u 


nn, 


nn, a 


OH, OW 


uu, oou 


on, oon 


an, o<m 


an, 00 


u short 


u; », e 


n; se 


n,o; 1, e 


n,a; ^i 


Mlong 


yy 


yy 


yy,in 


in 



Chap. I. 



PRONUNCIATION AND ITS CHANGES. 



29 



Modem SpOUng 


Chaucer 
xnr 


Sbaks] 
Spenser 

XTI 


>cre p 

tfUton ^ 

Dryden 

xvn 


ope 

Goldsmith 

xvm 


kmmd 


Hand 


Hand 


Hffind 


Hssnd 


UUe 


taal-e 


taal 


toffil 


teel 


ram, u^if] 


rain, wai 


rain, waai 


neffiin, wsbsbi 


reen, weei 


taWf mw$ 


Ban, an 


sau, aau 


SAA, AA 


8AA, AA 


W 


^ 


eg 


eg 


eg 


thstCf W4 


dheez, wee 


dheex, wii 


dbeez, wii 


dhiix, wii 


meamy head 


meen, Heed 


meen,Hed 


meen, ned 


miin, ned 


seen 


seen 


siin 


siin 


siin 


fviMy Tdoeive 


obai*, dhai, 
Tain,re6aiY* 


obei- dbeei, 
Tain,re8eeiT* 


obeei*, dheei, 
Teen,re8eeT' 


obee', dheei* 
Teen, risiiv 


feWftiew 


feu, 8lyy 


feu,8tyy 


feu, stiu 


fiu,Btiu 


nighi 


naht 


m*ht, niH*t 


niH% nait 


nait 


hit 


btt 


bft 


bit 


btt 


hiu 


bitHe 


belt 


bait 


bait 


keUy, wonder 


Hol'i,wun*der 


Hol'i^ wun'der 


Rol'tfWan'der 


Hol'f, wander 


hope 


HOop 


HOop 


Hoop 


Hoop 


eoapf hroad 


floop, brood 


soop, brood 


soop, brAAd 


soop, brAAd 


Jomt, boil 


dzhuinty bull 


dzhoint, buil 


dzhoint, buil ; 
dzhaint, bail 


dzhaint, bail 


fool, blood 


fool, blood 


fiiul,blud 


fuul,blad 


fuul,blad 


HOWf KttOW 


nuu, knoou 


nou, knoou 


nan, noon 


nau, noou 


jma, but, buey, 
burp 


pul, but, biz'f 
bert 


pul,but,btzf, 
bert 


pul,bat,btz-t, 
ber-t- 


puU bat. hiti, 
ber'i 


wntoe 


myyz'e 


myyz 


myyz, miuz 


miuz 



30 PRONUNCIATION AND ITS CHANGES. Chap. L 

Taking the principal modem combinations of vowels, and the 
one consonant combination, gh, for which the pronunciation 
of successive centuries have mainly differed, I have arranged 
them in the first column of the preceding table. It must be 
borne in mind that these spellings are modem, and in many 
cases replace at present other spellings which were current 
in the xrvth to the xvith centuries. In the four next 
colimms I give in palaeotype, as explained in the introduc- 
tion, the pronunciations prevalent during the xiv th, xvi th, 
xvn th, and xviii th centuries. For this rough and general 
view of the subject there is no perceptible difference between 
the xviii th and xix th centuries. It must not be supposed 
that the pronunciation here indicated prevailed throughout 
the centuries to which they are attributed. The xivth 
century pronunciation refers only to the latter half of that 
century. The xvith century is represented rather in its 
former half and middle than m the latter part when it was 
verging to the xvii th century pronunciation. The xvii th 
century pronunciation represents the fully established pro- 
nunciation of the time in the middle and latter part of the 
century. And the xviii th century pronunciations is that of 
the latter part. Hence we may roughly term the pronuncia- 
tions exhibited those of Chaucer, Spenser, Dryden, and Gold- 
smith. Shakspere and Milton are transitional between 
Spenser and Dryden, while Pope lies between Dryden and 
Goldsmith. These names are therefore placed at the top of 
the columns, and between the columns, as an assistance to 
the reader. As single letters are more difficult to appreciate 
than entire words, examples of each mode of speech are 
given. The same combination of letters was not always 
pronounced in the same way in all positions, even in the 
XIV th century ; hence it is sometimes necessary to give two 
sounds and two examples, and in this case the more usual 
(not the older) sound is put first. In the latter part of the 
XVI th, in the xvii th and later centuries, anomalies of pro- 
nunciation became more common, and nothing but detailed 
lists of words, such as will be furnished hereafter, will serve 
to explain them. The reader must therefore remember that 
this table gives merely a general view to serve as a guide in 
studying the subsequent details. 



31 



CHAPTER 11. 

AXTTHORITIES FOR THE PRONUNCIATION OF ENGLISH DURING 

THE Sixteenth, Seventeenth, and Eighteenth Cen- 
turies. 

§ 1. Sixteenth Century. 

1530, 22 Henry VIII. Palsgrave, Jolm. 

Lesclarcissement de la Langue Francoyse ; compose par 
maistre Jehan Palseraue Angloys natyf de Londres, 
et gradue de Paris, London, 4to. 

19 folios unmarked, 473 folios nmnbered, the English in black 
letter, the French in Bpman characters. The book is written in 
English although the title is French. It was reprinted by the 
French Government, and edited by F. G^nin, in 1862. 

Palsgrave graduated at Cambridge as well as in Paris, and was 
appointed French tutor to the princess Mary, sister of Henry VIII, 
when a marriage was negociated between her and Louis XII of 
France in 1514. He was made a royal chaplain, and on going to 
live at Oxford in 1531, there took the degrees of M.A. and B.D. 
He is supposed to have died in 1554. He must consequently have 
spoken the educated southern and court dialect of the latter part 
of the XV th, and the early part of the xvith century. 

This work contains a very elaborate account of French pronuncia- 
tion, frequently elucidated by reference to contemporary English 
and Ittdian. The pronunciation of several English words is thus 
incidentally established with more or less certainty. 

To the French reprint is added a reprint of 

An Introductorie for to leme to rede, to pronounce and 
to speke French trewly, compyled for the right high, 
excellent and most vertuous lady The Lady Mary of 
Englande, doughter to our most gracious soverayn 
Lorde Kyng Henry the Eight. 
By Giles du Guez or du Wes, vrith no author's name, except as 
shewn by an initial acrostic, and no date, but apparently about 
1532. The rules for pronunciation are few and insufficient, ex- 
tending over three quarto pages. 

1545, 37 Henry VIII. Meigret, Loys. 

Traits touchant le commvn vsage de Tescrityre francoise, 
faict par Loys Meigret, Lyonnois: auquel est debattu 



32 AUTHORITIES — SIXTEENTH CENTURY. Chap.II. }!• 

des faultes, & abus en la ysage, & ancienne puissance 
des letres. Auecqpriuilege de la court. Pans, 12mo, 
in Italics, pp. 128 unnumbered. 
This little book incidentally enters into a discussion of the pro- 
nunciation of the French language, and thus renders Palsgrave's 
English analogues more certain. Where Meigret differs ^m Pals- 
grave, it is difficult to decide whether Palsgrave is in fault through 
want of appreciation and English habits, or Meigret &om being 
a Lyonnese instead of a Parisian. See another work by Meigret 
described under its date 1550. This little work is also remarkable 
as having in some way suggested Hart's English work on Ortho- 
graphy, 1569, subsequently described. Hart says, translating his 
phonetic spelling into modem EngHsh orthography: "You may 
see by this little treatise I have been a traveller beyond the seas, 
among vulgar tongues, of which that small knowledge I have, hath 
been the cause of this mine entreprize. And therewithal the sight 
of a treatise set forth in print at Paris, Anno 1545, by a worthy 
man, well learned both in Greek and Latin, named Leuu Meigret 
of LyoUf touching the abuse of the writing of the French tongue, 
whose reasons and arguments I do here before partly use, as he did 
Quintilian's, whom it appeared he had well studied. And I have 
seen divers French books put forth in print in that his manner of 
Orthography, of some well liked of, and received, and of others left 
and repugned. But what good & notable thing can take a speedy 
root, amongst a multitude, except the princes & governors, (by 
the grace which God may give them) do favour & somewhat 
coimtenance it." 

1547, 38 Henry VI & 1 Edward VI. Saleshury, W. 

A Dictionary in Englyshe & Welshe London, 

4to, black letter. 

The complete title is given below, Chapter VIII, § 2, which 
contains a transcript of the preliminary Welsh essay on Ijiglish pro- 
nunciation, with a translation. 

From Anthony a "Wood's Athenae Oxonienses by Philip Bliss, 
London, 1813, voL i, p. 358, we leam that Salesbury was bom of 
an ancient family in Denbighshire, studied at Oxford, and was 
entered at Thavies Inn, Holbom, London. In his latter days he 
lived with Humph. Toy, a bookseller in St. Paul's Churchyard. He 
translated the New Testament into Welsh, and obtained a patent 
for printing it, from Queen EUzabeth, 1567. He wrote also other 
works, see under 1567. 

As a Welshman, Salesbury was of course liable to mispronounce 
English, but he was so early removed to England, and had so long 
an opportunity of studying the Southern English pronunciation to 
which his treatises shew that he was fiilly aHve, that any assertion 
of his must carry great weight with it, however much opposed it 
might be to theory. His pronunciation is evidently more modem 
tluui Palsgrave's. 



Ckaf. n. § 1. AUTHOBTTIESHHSIXrESNTH CENTUBT. 33 

1550, 4 Edward VI ; 4 Henri II of France. Meigret, Leys. 
Le trEtt^ de la OranmiEre Fran90Eze f£t par Louis 
MeigTEt, LionoEs. Paris, 4to of a folio shape. 

This very curious Prench Grammar, (which is not noticed by M. 
Oemn in his introduction to Palsgraye, although it was so nearly 
contemporary,) is entirely printed phonetically, apparently to carry 
out the suggestions of Meigret's little hook already described, bettCT 
than he had done in a former work, which he alludes to thus : 
^Tecritture qe j'ey obsEru^ (combien q'slle ne sost pas du tout 
selon qe reqerost la rigeur de la pronon9{a9(on) eu la translaqfon du 
Msnteur de Lu^lan," (fo. lOi.) His alphabet consists of the letters 
"a, £ ouuErt, e clos, i Latin, o ouuErt, ou clos, u, y Grsc de 
ni£me puissance qe I'i, b be, p pe, f ef, ph phi, u conso., c ca 
Latin, k ca GrEc ou kappa, q qu, g ga ou gamma, ch cha aspir6, 
d de, t te, th the aspir^, (^ 9, s, es, z zsd, qh ^he, 1 e1, l el moUe, 
m Em, n En, K EU moUe, r Br, i ji consonante, x, cs, ks, gs, ix," 
(fo. \bh) where I haye used b for an e with a tail like 9, l for an 
1 with a short mark over it like 1, and k f or an n with the second 
stroke produced and terminating in a backward hook, which re- 
sembles the letter c, and with a short mark over it like ii. The 
powers of these letters, taken in order, appear to have been, (a, b, 
^ i, o, u, y, i; b, p, f, f, v, k, k, k, g, k, d, t, t, s, z, sh, 1, Ij, 
m, n, nj, r, zh, ks, gz). 

La Grammaire Fran9aise et les Grammairiens au XYI* si^le, par 
Ch.' Z. Livety Paris, 1859, gives an abstract of all Meigret's works 
and of his controversies with G. des Autels, and J. PeUetier, &om 
which it appears tbat Meigret lived in Paris, and had been an 
assiduous fi^uenter of the court of Fran9oiB I, (p. 139). The dis- 
pute principally affects Meigret's e, e, (pp. 127, 132, 140), o, ou, 
(p. 139), ai, (p. 130), ao, (p. 122), eu, (p. 130), and shews the transi- 
tional state of French pronunciation at the time. M. Livet's book 
also contains notices of Jacques Dubois (Jacobi Sylvii Isagoge, 1531), 
J. Pdletier (Dialogv^ d^ I'orthograf^ et prononciacion fran9oes«, 1555, 
a year after Meigret had been forced by his publisher te use the 
ordinary orthography), Pierre Eamus ou de la Eam^e (Grammaire, 
1 ed. 1562, 2 ed. 1572, last 1587,) Jean Gamier (Institutio gallicsB 
linguse, 1558), Jesen PiUot (Gallicee linguse institutio, 1581), Abel 
Hi^hieu (Devis de la langue fran9oyse, 1559), Eobert Estienne 
(Dictionnaire fran9.-lat., 1539, Traicte de la Gram. fran9. without 
date), Henri Estienne (H. Stephani Hypomneses, 1582, Traict6 de la 
conformity, Beux Dialogues, without date, 1578?, Pr^cellence, 1579), 
Claude de Saint-Lien (Claudii II Sancte Vinculo de pronunciatione 
ling. gall. 1580), Theodore de B^ze (De Francicae linguse recte pro- 
nunciatione tractatus, Theod. Beza auct. 1584). If te these we add 
Palsgrave & du Guez, neither of whom are abstracted by M. Livet, 
we can trace the change of French pronunciation from the earlier to 
the later part of the xvith century, till it subsided inte a form 
practically the same as the present, by a course remarkably similar 
to that pursued by the contemporary English pronunciation. 

3 



34 AUTHORrnES — SIXTEBNTH century. Chap. II. } 1. 

1555, 3 Mary. Cheke, Sir Jolm. 

Joannis Cheki Angli de pronunoiatione Ghraecae potiBsi- 

miiiii linguae disputationes cum Stephano Yuintoniensi 

Episcopo. Basle, 24mo. 

In this work seyeral iUustrationB of Greek eonnds are drawn from 

English words which are printed phonetically in Greek letters, to 

give a conception of the autiior's theoretical pronunciation of Greek. 

Adolph Mekerch of Bruges, in H. Stephanus's collection Be vera 

pranuneiatume Graeeae et Latinae Linguae^ 1587, adopts in many 

places the very expressions of Cheke, but changes his iUustrative 

words from. English to Flemish, which he again prints phonetically 

in Greek letters. In this way a comparison of English and Flemish 

in the xnth century is instituted. Cheke bom at Cambridge in 

1514, moved in the best literary society, was secretiEiry of state 

1552, and died 1557. 

1667, 10 Elizabeth. Salesbury, W. 

A playne and familiar Introduction, teaching how to 
pronounce the letters in the Brytishe tongue, now com- 
monly called Welsh .... London, 4to, English in 
black letter, Welsh in Boman. 
All the portions of this rare book which are useful for the present 

investigation are reprinted, with illustrative notes, below. Chap. 

Vni, § 1. See 1547, suprk p. 32. 

1668, 11 Elizabeth. Smith, Sir Thomas. 

De recta et emendata lingveB anglicaa scriptione, dia- 
logue, Thoma Smitho Equestris ordinis Anglo authore. 
Lutetise. Ex officina Boberti Stephani Typographi Rem . 
Paris, folio, 44 folios. Date of colophon, 13 Nov 1568. 
A beautifully printed book in large Eoman letters with tables of 
illustrative words printed according to a phonetic alphabet, without 
the ordinary spelling, Smith's object being to improve the ortho- 
graphy not explain ^e pronunciation. The value of his 34 letters 
in tiie order of his alphabetic table (fo. 41) is apparentiy as follows, 
(a, aa, b, tsh, d, dh, e, ee, u, f, v, g, dzh, n, ♦, ei, k, 1, m, n, o, 
00, p, k, r, s, £, sh, t, th, u, uu, yy, ks.) 

Smith uses e for (tsh), which has occasioned many misprints, % 
for (dh), a letter like the Anglosaxon e with a diaeresiB for (ii), an 
inverted ^ or p^ for v, the Anglosaxon ^ for (dzh), a reflected z for 
(sh), for (th), V for (yy). The long vowels he has represented 
by a diaeresis, and as he considers (ei) to be the long of (t ), he prints it I. 
Since then (ee) is S, and (ii) is a character almost identical m appear- 
ance, misprints occasionally occur. In all cases of phonetic writing 
when diacritic accents are employed, misfortunes of this kind are fre- 
quent. Hence the importance of indicating length by reduplication, 
as in palaeotype, or by some constant additional sign, as in Fit. Speech. 
Sir Thomas Smith was bom at Saffiron Walden, Essex 1515, 
was fellow of Queen's Collegei Cambridge 1531, public orator 



Chap. II. } 1. ATTTHOMTIES — SIXTEENTH CENTUET. 35 

1536, proTOst ai Eton, master of requests ta Edward TE, secretary 
of state 1548, priry councillor and assistant secretary of state 1571, 
sacceeded Burleigh, and died 1577. Hence bis pronunciation must 
be accepted as tbe most literary and courtly of a time somewbat 
subsequent to Palsgrave's. He was not mucb acquainted with 
Froicb,' or probably with any other living language, and conse- 
quently without the assistance of Salesbury great doubts would be 
felt as to many of hi» pronunciations. 

1569, 12 Elizabeth. Hart, John. 

Ad Orthographie, conteyning^ihe due order and reason, 

howe to write or paini;e thimage of mannes voice, most 

like ta the life or ncUnire. Composed by J. H. Chester, 

Heralt The contents whereof are next folowing. Sat 

oitosi (aic) sat bene. Anno. 1569. London, 12mo. 

The first part in black letter, the latter part in italics- with new 

letters for (sh, dzh, tsh, dh, th, %) and a dot under a t^ort vowel 

sign to lengthen it. Eeprinted in lithography by 1. Pitman, 1850, 

the first port in the phonogn^y^or phonetic uiorthand of that date, 

the latter part in a longhand writing imitating the italic original. 

The name John Hart is token from the British Museum catalogue. 
Dr. GKll calls him *' e fecialibus vnus, qm eorum more ex gnvdu 
officii nomen abi Chester assumpsit." Be is cited as "Master 
Chester" by Bullokar. It seems probable that he was a Welshman, 
as he writes (uuld) for (would), that is, he did not pronounce (wuu) 
as distinct from (uu). 

This is a most disappointing book. The writer knew several 
languages, as French, German, Italian, Spani^ and there is little 
or no doubt as to the general value of his symbols, but in the words 
of I>r. Gill, ''sermonem nostrum characteribus suis non ssqut sed 
iueere meditabatur." He has in fact chosen a pronunciation then 
coming in, heard by few, and distasteful to the old school. See 
below. Chapter III, § 3, EI, AI, and Chapter VIII, § 3. One of 
the causes of the writing and publication of this work, was Hart's 
acquaintance with Meigret's book of 1545, see above p. 31. 

It appears that this book of Hart's was twenty^ years older than 
its real date, which would bring it up to 1549, for he says (fo. 5^) : 
''The lining doe knowe themselues no frirthir bounde to this our 
instant maner, than our predecessors were ta the S^on letters and 
writing, which hath bene altered as the speach hath chaunged, much 

^ This he informs m of in the be* been iminteHigible most probably to 

ginning of his ti^atise De reeta et emen- Aristophanes, as it certaiiuj^ would be 

data, lingutB Oracm pronundatUme to any modem Qreeh. While he was 

EpitUla, 1668, in whicii also several in Paris he met with a modem Oreek, 

passages occur which are nsefnl in the who was ftzrious at the notion of in- 

determination of English pronnnciation. trodncing ** tarn vastos sonos et absonas 

The two treatises are bonnd in one diphthongas in G^rsscam lingnam," but 

Tohime in the British Museum Library, the two mspotants could not argue the 

He introduced Erasmus's system of point, ^'(^uoniam ego Qallicd param ad- 

Gre^ pronunciation, which is similar to modum, lUe non ita multd plus, Latind 

fiiat now used at Eton, and would haTO nihil oallebat^" fo. 6^. 



36 ATJTHOEinES — SIXTEEMTH CENTUBY. Chap. IL } 1. 

differing from that which was vsed with in these fine hnndreth, I 
maye say within these two hnndreth yeares: which I considered 
of ahout .z X. yeares passed, and thought it worth my labonr, if I 
collide finde the meane of remedie, of our present abuse. And so 
framed a treatise therevpon, and would then it had bene published, 
but I am the gladder it hath bene stayed yntill this time, wherein 
«o well a learned gentilman, in the Gr6eke & Latino tongues, & 
trauailed in certain vulgares sir Thomas Smith knight, hath written 
his minde, touching this matter, in hys booke of latejset forth in Latin, 
entituled, De recta Sf emendata lingua Anglica seriptume. Where- 
of and of this my treatise the summe, effect, and ende is one. 
Which is, to Yse as many letters in our writing, as we doe voyces 
or breathes in our speaking, and no more ; and neuer to abuse one 
for another, and to write as we speake : which we must needes doe 
if we will euer haue our writing perfite." 

1570, 13 Elizabeth. Levins, Peter. 

Manipulos Yocabulorum : a Rhyming Dictionary of the 
English Language by Peter Levins. 4to. 
This book has been reprinted by the Early English Text Society, 
under the able editorship of Mr. Henry B. Wheatley. The words 
are arranged according to their orthographies, so that very little assist- 
ance is given towards determining ^e pronunciation. The place of 
the accent, however, is generally marked, but as evident errors are 
committed, no reliance can be placed on it. It is chiefly valuable 
for shewing the received orthography of that period, and as such 
will be frequently cited. 

1573, 16 Elizabeth. Baret, Jolin. 

An Alvearie or Triple Dictionarie, in Englishe, Latin 
and French : very profitable for all such as be desirous 

of any of these three languages London, fo. 

The introductory remarks upon each letter afford some slight 
assistance. John Baret, was fellow of Trinity College, Cambridge, 
and graduated in 1554. His pronunciation belongs therefore to tiie 
middle of the xnth century, and to the educated class, but his 
county is not known. 

1580, 23 Elizabeth. Bullokar, William. 

Bulhkara Booke at large for the Amendment of Ortho- 
graphic for English speech: wherein, a most perfect 
supplie is made, for the wantes and double sounde of 
letters in the olde Orthographie, with Examples for the 
same, with the easie conference and vse of both Ortho- 
graphies, to saue expences in Bookes for a time, vntill 
this amendment fi^w to a generall vse, for the easie, 
speedie, and perfect reading and writing of English, 
(the speech not changed, as some vntruly and maliciously^ 



Chap. IL § !• AUTHORinES — SIXTEENTH CENTURY. 37 

or at the least ignorantlie Howe abroade,) by the which 
amendment the same Anthoor hath also framed a niled 
Grammar^ to be imprinted heereafter, foT the same 
speech, to no small commoditie of the English Nation, 
not oidy to come to easie, speedie, and perfect vse of 
our owne language, but also to their easie, speedie, and 
readie entrance into the secretes of other Languages, and 
and easie and speedie pathway to all Straungers, to vse 
our Language, neeretofore very hard vnto them, to no 
small profite and credite to this our Nation, and stay 
thereynto in the weightiest causes. There is also im- 
printed with this Orthographic a short Pamphlet for all 
Let^nera, and a Primer agreeing to the same, and as 
learners shall go forward therein, other necessarie Bookes 
shall spedily be prouided with the same Orthoffraphie. 
Herevnto are also ioyned written Copies with the same 
Orthographic. Giue God the praise, that teacheth 
alwaies. When truth trieth, errour flieth. Scene and 
allowed according to order. Imprinted at London by 
Henrie Denham 1580. London 4to.r 

In black letter, the new characters being also in black letter, 
with divers points, hooks, etc., placed above and below. His 
object was to keep as closely as possible to the existing orthography, 
and inark the pronunciation, and also certain grammatical forms. 
The union of these two objects serves greatly to complicate his 
orthography, which perhaps no one but the mventor could have 
used. He reckons 37 letters, most of which have duplicate forms 
•* for help in eqiu'oc'y." These 37 letters in order apparently re- 
present the soui^ (a, b, s, k, tsh, d, e, ii, f, dzh, g, h, i, 1, '1, m, 
'm, n, 'n, o, uu, p, k«?, r, s, sh, t, dh, th, yy, u, v, w, wh, ks, j, z) 
Bullokar admits seven diphthongs (ai, au, eei, eu, oi, oou, uui) 
with tit <' seldom in use," and rather uncertain in his text. The 
reduplicated forms and the fineness of the diacritical strokes, render 
his book troublesome to the reader, but the above interpretation, 
founded on Salesbury's information, fiirmshes a tolerably consistent 
account of EngHsh pronunciation. There are some long vowels not 
included in the scheme, namely (aa, ee, oo) which are generally 
represented by accents, as d, 6, f , f, 6, although 8b is conmionly em- 
ployed for (ee). In the case of long f and ou, he seems to have re- 
tained the ancient sounds (it, uu,) in place of the (ei, ou) given by 
Salesbury and Smith, see Chapter III, § 3, I, but he unfortunately 
generally neglects to write the accent on t. 

The pronunciation of BuUokar was certainly antiquated in some 
particulars, agreeing better with Palsgrave's than with that of any 
intermediate author, and proceding in a direction contrary to Hart's. 
Hence Gill looked upon him with favour, and says, " Bulokerus vt 
paucula mutavit, sic multa fildeliter emendavit.'' Altogether the 



38 AtTTHOBITIES — SL&XKKHTU CEHTUBT. Chat. ILi I. 

book is rery Taliiable for detennmiiig the pnmmici«tion of the early 
part of the xn till eentoiy. See Chap. YUI, § 4L 

1611, 9 James L Cotgrave, Bandle. 

A Dictkmarie of the Frencli and Englisb TongueSi 

London imprinted by Adam lalip. Fo. 
There is a short account of French pronnndation which inciden- 
tally gives some assistance towards the determination of English 
sounds. Although this book appeared in the xmtili century, its 
pronunciation belongs to the xnth. 

1611, 9 James I. Fhrio, John. 

Queen Anna's New World of Words, or Dictionarie 

of the Italian and Englisb tongaes, collected, and newly 

much augmented b^ J. F., B^aAer of the Jtalian vnto 

the Soneraigne Maiestie of ANNA, crowned Queene of 

England, Scotland and Ireland, &c., and one of the 

Gentlemen of hir Boyall Priuie Chamber. Whereonto 

are added certame necessarie rules and short obserua- 

tions for tbe Italian tongue. Fo. 

The first edition appeared in 1598, and of course had no reference 

to James's queen, Anne of Denmark. It also did not contain any 

account of the pronunciation. This second edition, in treating of 

the Italian pronimciation of e, o, discriminates their open and 

close sounds, which are mailed throughout the book, and exempli^ 

fies them, together with some of the consonants by a reference to 

English, which, allowing for Italian errors, is useful 

1619 first ed., 1621, second ed., 17-19 James I, GV//, Alexander. 

LogonomiaAnglica. Qu&gentissermofaciliiisaddiscitur 

Conscripta ab Alexandro Gil, Paulinse Scholaa magistro 

primano. Secundd edita, paul6 oorrectior, sed ad TSum 

communem acconmiodatior. Small 4to. 

This second edition differs from the first mainly in the characters 

employed ; there are, however, a few verbal differences in the text. 

The pronunciation exhibited, with perhaps two exceptions, that of 

long i and of au, was that of the mid<Ue of the xnth century, 

although the book appears in the xvn th, for Dr, Gill evidently re- 

siBted all modem mmcing and effeminacy of speech, as the new 

fEtfhions appeared to him. He was bom in Lincolnshire, 1564, the 

same year as Shakspere, became a student of Corpus Ohristi 

College, Cambridge, and graduated in 1583, and was made head 

master of St. Paul's school in 1608. He died 1635. Milton is 

said to have been one of his pupils. Dr. GKll had several fancies 

besides old pronunciations, thinking it best to speak '*ut docti inter- 

dAm" — anglici, pedantically — ^rather than like the "indoctus," 

although if the latter followed his ears in phonetic spelling the doctor 

says : ** susque deque habeo." 



Chap. II. { 2. ATJTHORITIBS — SBYEKTBBNTH CENTURY. 39 

Br. OiU's alphabet of 40 letters will be rendered in order by the 
following palaeotypic symbolB, — (a aaAAbtshddheeefTg dzh 
H kh f ii ai k kw Imnqoooprsshtthyyunuw wh ks j z). 

Dr. Gill's book enters at great length on the subject of pronun- 
ciation, without, howeyer sufficiently describing the sounds, and is 
peculiarly valuable in giving numerous passages &om Spenser and 
the Psalms written phonetically. See below Chapter YIII, § 5. 

1633, 9 Charles I. Butler, Charles. 

The English Grammar, or the Institution of Letters 
Syllables, and Words in the English tongue. Where- 
unto is annexed an Index of Words Like and TTnlike. 
Oxford. 4to. 
Printed phonetically with new characters for (ii, uu, dh, tsh, kh, 
^ ph, sh, wh) and a mark of prolongation. There is great dif- 
ficulty in determining the value of his vowel system. He was of 
Magdalen, Oxford, an M.A. and a country clergyman. His pro- 
nunciation belongs to the end of the xvi th century, as he clearly 
fights against many of the new pronunciations which were starting 
up, and the true xvnth centujy pronunciation seems not to have 
developed itself till the civil war had fairly begun. Butler pub- 
Hshed a work on the management and habits of bees, The Feminine 
Monarchy or Hietory of the Beee, Oxford, 1634, both in the ordinary 
and in his phonetic character. These are the first English books 
entirely printed phonetically, as only half of Hart's was so presented. 
But Meigret's works were long anterior in French. See below 
Chapter VIQ, § 6. 



§ 2. Seventeenth Century. 

1640, 16 Charles I, Jonson, Ben. 

The English Ghrainmar. Made by Ben. Johnson. For 

the benefit of all Strangers, out of his observation of 

the English Language now spoken, and in use. Fo. 

This was published two years after Jonson's death, and the text 

is known to have been altered fix)m his MS. in some parts. Jonson's 

pronunciation ought to have belonged to the xvi th century, as he 

was bom 1574, only ten years after Shakspere, but he seems to 

have inclined towards the xvnth century use. 

1646, 22 Charles I. Oataker, Thomas. 

De Diphthongis Bivocalibus, deqe Literarum qarundam 
sono germano, natur& genmn& figur& nov&, idoneft, 
8criptur& yeteri ver&qe. London, 24mo. 
This is useful for a few diphthongs, but is not of much value 

generally. 



40 



AUTHOEITIES — SEVE^TKENTH CENTURY. Chap. II. j 2. 



1661, 3 Commonwealth. JFillis, Thomas, of Thistlewood, 

Middlesex. 
Vestibulum Linguae Latinae. A Dictionarie for children 
consisting of two parts : 1. English words of one syllable 
alphabetically with the Latino W ords annexed. 2. W ords 
of more syllables derived from the Latine words adjoined. 
This first part consistfl of a vocabulary of more than 4000 mono- 
syllables, professedly arranged in order of rhyme, but with very 
few exceptions arranged only according to the spelling. In some 
of these exceptions we find real rhymes with differing spelling, but 
on the other hand we have words classed together which do not 
rhyme, so that there is by no means so much to be learned from it, 
as was to be hoped. The following are the only rhymes which 
are noticeable throughout the whole vocabulary. The initial syllable 
in italics as -affe is that under which these words and others having, 
the same termination are arranged. It is to be understood that 
only such words in each list are given in this extract as were in 
some respect curious or irregular, and that all other monosyllables 
having the prefixed termination are to be supplied by the reader. 



-afiy langhf cbafe, safe, Baphe 

-oip, = -fly, treie, weigh, whay 

•am, feign 

-flir, heir, major 

•aity eight, heieht, sleight, straight 

•arre, = -cr, wr, tar, warre 

-fli/tf, dwarfe, scarfe, wharfe 

'•arm, swarm, warm 

-am, warn 

'Urp, warp 

•art, heart, thwart 

-a«A, quash, wash 

'Ostey the waste medituUium 

-atte, Wat, what 

-atehj watch 

draught, naught 

&ult, yault 

-M, keie, the, yea 

'ead, head, knead, lead plumbum 

•earn, dream, phleagm, realm 

-Mr, blear, pear 

-tfa«, ceas, greas, leas, peace 

-w^, hee^ brief, chief, gne{, theef 

-eid, yeeld field, shield 

'^ndf friend 

-^^, here, there, where 

-«r, dew, due, few, glue, Jew, lieu, 

rue, sew «tMr#, sue, ^ew, shrew, 

▼iew, yew 
-t = -lij = -y, eie, buy, by, high, my, 

ni^h, Tie, side, why, wry 
'HSf guile, style 
'tit, guilt 

"immg = -I'm, hymne 
-wfitf, climbe 
'ine, siffne 
'irre, firre, myrrhe, sir 



-M>, giy, liy, seiy 

-0 =5 .^twe = -oe, bowe, blowe, crowe, 
glowe, flTOwe, knowe, lowe, mowe, 
rowe, uowe, sowe, snowe, towe, 
throwe 

-oad, broad, goad, load 

-oA, chouffh, cough, doueh, though, 
trough, rough, tiirougn 

-owle = -oo/ = -oU, bowle eraUr, 
jowle, powle tondire, prowl, rowle 
rotula, sole, soul, scrowle sehedula, 
toll, towle 8<mu8, trowle advoher$ 

--on = -onn^ John 

..one =: -ooif, bone, groan, Joan 

-0, s -oe, to, toe, doe offere, woe prth 
care 

'Oom, loom. Borne, toomb 

-ooj, goose urutr, loos, noos noduB 

"Oov, mooY, move, prooT, prove 

-ord, cord, foord, horde, sword 

'Oree, hors eguui 

•ote, prose, rose, those, whose 

'Oath, oath, both, frothe, growth, loath, 
mothe, slothe 

-othi, bothe, cloathe 

-or, doy, gloT, lov, shoy 

-otc, bougn, bow, brow, cow, how, mow 
fmnile, mow ttruere, now, plough, 
prow, BOW, thou, trow, vow 

-ous, a hous 

-ouae, to house 

-urn s 'umme, some, summe, thumb 

-firs, burs emporium, curs, nurs, purs, 
to purs reponere 

"Urtt, burst, cmrst^ worst 

'Uee, bruise. 



Chap. II. { 2. AtrTHORITIES — SEYEITTEEKTH CENTURY. 41 

1653-1699, 1 Protectorate— 11 William and Mary. Wallis, 

John. 
Joannis Wallisii Grammatica Lingvuae Anglicanae 
Cvi praefigitur De Loqvela ; sive de sonorum omnivm 
loquekriym formatioiie: Tractatvs Grammatico-Physi- 
CYS. Editio Sexta. Accessit Epistola ad Thomam 
Beverley; de MYtis SvrdisqYe informandis. Londini, 
excYdebat Gvil. Bowyer, prostant apYd A. Millar, 
1765. First edition 1653, second 1664, sixth 1699, the 
Oxford reprint of this edition 1765. The latest edition 
shews no variation in pronunciation from the second. 

Wallis was bom at Ashford in Kent 1616, and died in 1703. 
In 1649 he was appointed Savilian Professor of Geometry at Oxford. 
Buiing the civil war he made himself nsefnl to the parliamentary 
party by decyphering letters in secret characters. His chief feme 
rests on his mathematical powers. 

The introductory treatise on sound is of great importance, and 
^tablishes with much certainty the meaning of every symbol used. 
He did not attempt an alphabet, and consequently did not write 
out complete passages according to the pronimciation, which is 
greatly to be regretted. This work is the chief authority for the 
middle of the xvnth century. 

1668, 9 Charles II. Wilkins, John. 

An Essay towards a Real Character, And a Philosophical 
Language. Folio. 

Wilkins was bom in Northamptonshire 1614, and was therefore 
older than Wallis, although his work was not published till much 
later. His father was a goldsmith at Oxford. He graduated at 
Oxford 1631, and was made warden of Wadham College, Oxford, in 
1648, just before Wallis came to Oxford. The two must have been 
weU acquainted, and were among the original promoters of the 
Eoyal Society. In 1668 he was made Bishop of Ripon. He died 
1672. 

In this curious work, there is a very good English treatise on 
phonetics. He used a complete phonetic alphabet, and wrote the 
Lords prayer and Greed in his character, reproduced in palaeotype, 
below Chapter IX, § 1. 

The alphabeticad scheme on p. 358 of his work when translated 
into palaeotype will read thus — 

(k g qh q kh gh h 9 

t dnh n th dhlhlrhrshzhszjhieaA 
pbmhmfv whuoy) 

The short sound of (o) is not recognized in English. Long vowels 
are imperfectly represented by accents. Confusing, as so many have 
done, (j w) with (i u) he writes (i-i i-u u-u u-i) for (ji ju wu wi). 



42 AT7TH0BITIES — SEVENTEENTH CENTUET. GiUP. II. i 2. 

1668, 9 Charles U. Price, Owen. 

English Orthographie or The Art of rigM spelKng, read' 
ififfy pronouncing, and uniting all sorts of English Words. 
Wherein Such, as one can possibly mistake, are digested 
in an Alphabetical Order, under tneir several, short, yet 
plain Rules. Also some Rules for the points, and pro- 
nunciation, and the using of the great letters. Together 
WTFH The difference between words of like sound. All 
which are so suited to every Capacitie, that he, who 
studies this Art, according to the Directions in the 
Epistle, may be speedilv, and exactly grounded in the 
whole Language. Oxford 4to. The author's name is 
given on the authority of the British Museum copy in 
which it is pencilled. 
As interpreted by Wallis and WiUdns, this book is of great use 
in discriminating the exact sounds of ihe different vowel digraphs 
in the xmth century, famishing ahnost a pronouncing vocabulary 
of the period. The author was probably a Welshman. 

1669, 10 Charles EC. Bolder, William, D.D., F.R.S. 
Elements of Speech, an Essay of Inquiry into the 
natural production of Letters with an appendix concern- 
ing persons Deaf and Dumb. 8vo. 

Eeprinted by Isaac Pitman, 1 865. Not a very important treatise for 
our purpose, but useful in helping to fix some of the vowel sounds. 

1677, 18 Charles EC. PooU, Josua. 

The English Parnassus : Or a Help to English Poesie. 
Containing a Collection of all the Bhythming Mono- 
syllables, &c Svo. 
Not much confidence can be placed on the classifications of words, 
though they are not so purely orthographical as Willis's. Thus 
hass, hays, hlate, ease, are made to rhyme ; calf, half, Ralph are 
entered both under afe and alfe; Alice, else, ails, balls, which cer- 
tainly never rhymed, are placed together ; similarly ant, aunt, pant, 
vaunt, want ; words with ee and simple e are separated from words 
with ea, so that the different uses of ea are not shown ; and so on. 
The Hst seems to be rather one of allowable, than perfect rhymes, 
and consequently is of little service. 

1685, 1 James II. Cooper, C, A.M. 

Grammatica Linguae Anglicanae. Peregrinis eam ad- 
discendi cupidis pemecessaria, nee non Anglis praecipud 
scholis, plunmjlm profuturu. Cum Prcefatione & Indice, in 

Iuibus, quid in hoc libello perficitur, videatur. London, 
6mo. 
The first 94 pages, out of the 200 which this book contains, are 
devoted to a consideration of the sounds of speech, and peculiarities 



Ohap. n. § 2. AUTHORITIES — SEVENTEBNTH CENTURY. 43 

of orthography and pTonimciation, with long lists of words contain- 
ing the several vowel sounds, which render it of great use for the 
determination of the pronunciation of the xvn th century. I am in- 
debted to Mr. J. Payne, of the Philological Society, for my acquaint- 
ance with this valuable work. 

1688, 3 James II. Miege, Guy, gent. 

The Great French Dictionary. In Two parts. The 
first French and English; the second English and 
French ; according to the Ancient and Modem Ortho- 
graphy. Fo. London. 
There is much valuable information prefixed to each English 

letter and digraph, concerning the customary pronunciation, written 

in Flench. 

1700, 12 William and Mary. Lane, A. 

A Key to the Art of Letters ; or, English a Learned 
Language, Full of Art, Elegancy and Variety. Beinff 
an Essay to enable both Foreiners, and the English 
Youth of either Sex, to speak and write the English 
Tongue well and leamedhr, according to the exactest 
Bules of Ghranmxer .... London, 24mo, pp. xxiv, 112. 
A meagre treatise on Grammar by way of question and answer, 
in which 16 pages are devoted to spelling. The vowels are six, 
y being admitt^ and w excluded, although it is said that ''we 
usually sound to like the vowel tf, and for the most part we 
write it instead of u, in the middle and end of words, as in 
Vawsl, Law, Bow^ etc.," and ** when y begins a syllable, we sound 
it as in the word yea, and then it is a real Consonant; every- 
where else it is a vowel, and is soonded like t ; and is always 
writtai at tlie end of words instead of «', as in my, thy, &c." The 
liquids are three, m being excluded '' because a Mute before it can- 
not, without force, be sounded with it in the same Syllable with 
the Vowel after it.'' This should imply that n can be so sounded, 
and hence that h, g were pronounced in knot, gnat. The change of 
H- before a vowel into (sh) is not recognized ; "we sound ti before 
a Vowely like *», as in the word Eelation.'* The following assertion 
and its justification are curious : " E Servile is of great use in the 
Ei^glieh Tongue ; for by its help we can borrow the most significant 
and useful Words from other Languages, to iorich our own ; and so 
far disguise and transform them into good English^ that others can- 
not lay claim to them as theirs ; as for Example^ these Latin words, 
Candela, Vinea, Linea, Brutum, Centrum, are made good English, 
by the help of e Servile, thus ; a Candle, a Vine, a Line, a Brute, 
a Centre. Q. JFTutt need is there to disguise words borrowed from 
other Languages ? A. It is necessary to disguise Words borrowed 
from other Languages, because no free People should have a Foreign 
Face on their current Words, more than on their current coin, both 
being Badges of Conquest or Slavery.^^ The following is a curious 



44 AUTHORITIES — SEVENTEENTH CENTUEY. Chap. II. } 2. 

conceit : '^ E Subjunctive is written at the end of a word after a 

siagle Consonant, to make the single Vowel before it long 

E Suhjunctive is really sounded with the single Vowel before the 
Consonant, and so makes the Subjunctive or latter Vowel of a Diph- 
thong : otherwise it could not make the Syllable long, as in the 
words, Fire, more, pale, read, Fier, moer, pad" This leads us to 
suppose that he said (faiar, mooar, peeal) ; the two former are com- 
mon, the last is adduced by Cooper (p. 42). 

Tliis author is cited by the Expert OrthograpHst (p. 46). In 
the title he is called, ** M.A. late Master of the Free-School of Leo- 
minster in Herefordshire, now Teacher of a private School at Mile* 
end-green near Stepney." There is a certificate at the back of the 
title from the Masters of Merchant-Taylors, Charterhouse, Christ's- 
Hospital, and "Westminster, in favour of the use of this book to " all 
who desire to learn, pronounce, and write the English Tongue 
exactly." It is, of course, dedicated to the young Duke of Glouces- 
ter, and is of extremely little use as regards pronunciation, but 
belongs, like the following, to the xvntii century, whereas the 
Expert Orthographist who cites it, belongs entirely to the xvmth 
century. 

1701, 13 William and Mary. Jones, John, M.D. 

Practical Phonography: or, the New Art of Rightly 
Speling (sic) and Writing Words by the Sound thereof. 
And of Rightly Sounding and Reading Words by the 
Sight thereof. Applied to The English Tongue. De- 
sign'd more especially for the Yse and Ease of the Duke 
of Glocester, (sic). iBut that we are lamentably disap- 
pointed in our Joy and Hopes in him. By tf . Jones, 
M.D. You may read the Preface, where you have an 
account of what the Book performs ; which ('tis hoped) 
will not only answer Men s Wishes, but exceed their 
Imaginations ; that there could be such mighty Helps 
contrived for Reading, Spelling, and Writing Englisn, 
rightly and neatly ; with so much Ease. London. 4to. 
The above title is transcribed from a copy I have in my posses- 
sion. The Duke of " Glocester*' referred to, died 29th July, 1700, 
aged 11. In the copy in the British Museum, dated 1704, of which 
the whole text is identical with mine, the title runs thus — 

" The New Art of Spelling. Designed chiefly for Persons of 
Maturity, teaching them how to speU and write Words by the 
sound uiereof^ & to sound & read words by the sight thereof, 
rightly neatly and fSashionably. I. It will instruct any person 
that can read & write to spell & write most languages that he 
can speak & uses to read in a few hours by a gener^ rule con- 
tained in two or three lines, & the use of a spelling alphabet, 
which may be written on the 12th part of a sheet of paper to 
carry about them. II. Short & easy directions whereby any 



Chap. II. } 2. AUTHORITIES — SEVENTEENTH CENTURY. 



45 



one may be tanglit to spell tolerably well in a few days, & in 
balf a year's time may be perfected m the art of true spelling. 
m. A child or any person who can read or write may by the 
help of this book learn to spell & write perfectly in a small 
time. IV. Roles for foreigners by which they may sweeten 
their language, & directions how to invent a uniyersal one. 
Applied to the English Tongue by J. Jones^ M.D." 

Notwithstanding the prolixity of the title it gives but a very 
inadequate conception of the book, which is a sort of pronouncing 
dictionary arranged under the simple sounds and their various re- 
presentations, in the form of a dialogue. Thus he asks ** when U 
iAs sound of a written aa, ah, ac, ad, ada, ac, ae, ag, agh, ah, aha, 
ai, aia, aie, aig, aigh, al, alf, ana, ao, ap, ath, au, ave, aw, ay, ayo, 
e, ea, ei, ena, exa, ey, ha, i, ia, ina, . ioa, o, oa, ua, wa, wha ?" 
And to each of these questions he gives an answer, often containing 
a long list of words, from which may be inferred, not always the 
pronunciation generally received as best, but certainly the different 
pronunciations which were more or less prevalent. This is in fact 
the peculiar value of the book to those who seek to know how 
people actually pronounced at the time when Dryden died (1700) 
and Pope (b. 1688) was in his teens. 

His single rule for spelling is as follows : — AU JTordi which can 
he sounded uveral way»^ must he written according to the hardest^ 
harshest, longest^ and most unusual Sound, And the Spelling Alpha- 
bet, spoken of on his second title, runs thus : — 



Tbtt tmtUr wad 
pUaumUr 



Th.t harder 

and kartktr 

Boundfl written 



A SpeUing ALPHABET. 



A 
b 

d 

e 
ee 

% 
m 

ng 

00 

8h 
t 

v 

tf 

s 





r *,o . 






p . 






t,th. 




s 


h<hu 




o 


0, i, 




a 


Cf eh 






n 
n 




n 


a,fi. , 




B 


chy 9 




s 


th . 






f.ph 






a, 0,4,0 , 




L * 


. . 



as in CUrk, Wagon 

as in Cupid, Deputy , 

u in Hatton, MurtKer 

as in Oirl, Fagot, injure .. 
as in A^, Shire, Women .. 
as in Clyeter, Nortoieh,,,,, 

as in Banbury , 

as in Ink, sink 

as in to. Bull 

as in Bench, Issui 

as in Thomas 

as in Foes, Nephew 

as in Evan, even. Sir, Son 
as in Bate, cause 





a 
b 




d 


a 


e 


^ 


ee 


^ 


g,g« 


V 


m 


00 


^g 


8h 


i3 


t 


> 


v 




a 




z 



Then upon the principle of the grammarian 

Yisom est GrammaticflD metricis lenire laborem Pneceptis, 
he proceeds "for Memory's sake'' to reduce the above to verse. 
Afterwards come long explanations of the use of this alphabet in 
teadiing spelling, the last of which is, as he says, " more a Shift 
than a Rute," and is simply this : 

" When you are (notwithstanding all that is directed^ in Douht 
of spelling a Word rightly, the last Shift wiU be to change the 



46 AI7TH0RITIES — EIOHTBENTH CBlTrUBY. Chap. II. § S. 

JTord or Expreisiony so as to prcfleire the Sense or Meaning; as 
suppose that you cannot, or are in Doubt of spelling the Word 
jiffeetion, write Kindness^ Zove^ Favour^ &c. instead thereof;" .... 
This was the '^ shift" employed in q>ea]dng by the deafinute Br. 
S[itto, when he wished to use words that he knew well by sight 
but had never heard during his youth before the accident wluch 
made him stone deaf. — See Kitto's Lost Senses. 

This book closes the xyn th century and trenches on the xrin th, 
because the Author was compelled by his plan to introduce all the 
most altered forms of speech as well as the least unaltered. 

§ 3. Eighteenth Century. 

1704, 3 Anne. Anonymous. 

The Expert Orthographist : Teaching To Write True 
English Exactly, By Rule, and not by Bote. According 
to the Doctrine of Sounds. And By such Plain Ortho- 
graphical Tables, As Condescend to the Meanest Capa- 
city. The Like not Extant before. For the Use of 
such Writinfi^ and Charity Schools which have not the 
Benefit of the Latin Tongue. By a Schoolmaster, of 
above Thirty Years Standing, in London. Persons of 
Quality may be attended at their Habitations ; Boarding 
Schools may be taught at convenient times. London : 
Printed for, and Sold by the Author, at his House at 
the Blue' Spikes in SpreadSagk'Couri in Orays^Inn' 
Lane. Where it is also Carefully Taught. 
This little book, 8vo, 112 pages, for a knowledge of which I 
have been indebted to Mr. Payne of the Philological Society, is fall 
of tables, but does not enter with sufficient minuteness into the 
<< Doctrine of Bounds " (which is paraded in capital letters in the 
title page) to render deUcate points at all appreciable. The great 
peculiarity of the work is, that though it beiars date 1704 the same 
year as that on Jones's second title page, it belongs exclusively to 
the xvmth century, and differs as much from Jones, as Hart from 
Smith in the xvith century. Thus Jones only allows eighteen 
words containing m to be pronounced with (ii), this author (whom 
I shall call the Orthographist) gives a list of 255 such words, and 
allows only four words in m, to have the sound of (ee), viz. hear s. 
and v., swear y tear v., wear. Again, Jones distinctly asserts that ei is 
"never" pronounced (u), the Orthographist gives ten words in 
which ei is so spoken, lliese shew totally different systems of pro- 
nunciation. Dr. Jones was a physician, and hence we may better 
trust his pronunciation than that of a visiting schoolmaster living in 
a court turmng out of Grays-Inn-Lane, who, attending *' persons of 
quality" would naturally adopt the thinnest pronunciation for fear 
of being thought vulgar. The curious thing, however, is, that 
though Dr. Jones endeavoured to collect, and did actually collect 



Chap. II. { 3. AUTHORITIES — ^EIOHTEEKTH CENTTTRY. 47 

a great yariety of even ridiciiloiis pronunciatioiis, for the purpose of 
assuBtmg pronouncers of all kinds to spell, he seems to be entirely 
unconscioiis of these sweeping innoyations, which are yaluable as 
the fot-eshadowB of coming events. 

1710, 9 Anne. Ano nym ous. 

A Short & easy Way for the Palatines to learn Eng- 
lish. Oder erne knrze Anleitung zur englischen 
Sprache znm Nutz der annen Pfalzer, nebst angehang- 
ten Englischen und Teutschen ABC. London, 8yo, 
pp. 64 and 18. 
A UUle tract in which the pronimciation of several words is ap- 
proximatively given in German letters. The Upper Palatinate was 
wasted by Louvois, general of Lonis XIY. in 1688, and 5000 of the 
distressed people for whom this tract was intended emigrated to 
England in 1709. 

1710, 9 Anne. Dyche^ Thomas. 

Chiide to the il^glish Tongue, London 12mo. 
The pronunciation of nearly 200 words is imperfectly indicated 
by re-spelling them. E. Coote's English Schoolmaster 1673, which 
is bound up in the same volume in the British Museum, and is often 
referred to, contains no information on pronunciation. The four- 
teenth edition of Dyche's Quide^ 1729, also in the British Museum, 
contains a few alterations, and has been chiefly followed. 

1713, 12 Anne. Anonymous. 

A Grammar of the English Tongue. With the Arts of 
Logick, Bhetorick, Poetry, &c. Sixth edition. Svo. 
There is no date throughout the book, but as it is dedicated to 
the Queen, and as the example given for finding '' the Moon's Age 
at any time," refers to 1 Jan. 1713, it was probably published 
about that time. The first part, consisting of 52 pages is devoted 
to Spelling and Pronunciation. The latter agrees almost exactiy 
with that of the Expert Orthographist (1704), but in the notes and 
especially from p. 43 to 52, there is a translation of many of Wallis's 
observations on phonetics and on English pronunciation, generally 
without acknowledgement, and evidentiy in happy ignorance of the 
£Bct that they belonged to a different stage of pronouncing English, 
and in several cases directiy contradicted the rules which tiie author 
Tifwiflftlf had previously given. It is a mere compilation, but cor- 
roborates other accounts of the xvmth century pronunciation. 

1766, 7 George III. Buchanan, James. 

Essay towards establishing a standard for an elegant 
and uniform pronunciation of the English Language, 
throughout the British Dominions, A Work entirely 
new ; and whereby every one oan be his own private 



48 AXrrHOBITIES— EIGHTEENTH CENTUEY, Chap. II. i 8; 

teacher. Designed for the Use of Schools, aad of 

Foreigners as well as Natives, especially such whose 

Professions engage them to speak in Public. Extera 

quid quaerat sua qui Yemacula nescit ? As practised 

by the Most Learned & Polite Speakers. London, 8yo. 

This almost amounts to a pronouncing dictionary, and like it, 

aspires rather to lead than follow general usage. The pronunciation 

it exhibits does not materially differ from that now heard, except 

in admittiQg many usages as <' learned and poUte," which would 

probably be considered much the contrary by modem Orthoepists. 

The xnnth century pronunciation is fiilly established in this work. 

But allowances must be made for certain Scotticisms, which will be 

more particularly pointed out in Chapter X, § 3. 

1768, 9 George III. Franklin, Benjamin. 

A Scheme for a Kew Alphabet & reformed mode of 

Spelling, with Remarks & Examples concerning the 

same, and an Enquiry into its Uses, in a correspondence 

between Miss Stephenson & Dr. Franklin written in the 

Characters of the Alphabet. 

From the Complete Works in Philosophy, PoUtics, & Morals of 

the late Benjamin Franklin ; now first collected and arranged, with 

memoirs of his early life, written by himself, 3 vols, London Svo. 

Johnson, 1806. Vol. ii. p. 357. 

The preceding works from the time of Wilkins, exactly 100 years 
preriously, have furnished us with no connected specimen of English 
speech. They have generally contented themselves with giving 
lists of words illustratiug particular usages. By this means the 
whole pronunciation of a word had to be collected from different 
lists, and some parts of it remained doubtfuL This is not the case 
in Buchanan's book, because he gives the pronunciation of every 
part of the word. But even then the isolated words do not seem to 
convey the same idea as connected sentences. The paper of Dr. 
Franklin therefore, is very acceptable, and wiU be printed at length 
in Chapter X, § 2. Being the pronunciation of a man of 62, who 
had passed his me among colonial English, it has necessarily rather 
an old appearance, and, notwithstanding the actual date, must be 
considered as belonging to the earUer part of the xvinth century. 

1780, 21 Georffe III. Sheridan, Thomas. 

A General Dictionary of the English Language, One 
main Object of which, is, to establish a plain and per- 
manent Standard of Pronunciation. To which is pre- 
fixed a Ehetorical Grammar. London, 4to. 
This is the first of the modem army of pronouncing dictionaries, 

and indicates a pronunciation which only d^ers in isolated instances 

from that now in use. It is therefore unnecessary to pursue the 

list further. 



49 



CHAPTER m. 

Oh the Pronitnciation of English ik the Sixteknth 
Cbntukt, and its Gradual Change duking the Seven- 
teenth AND Eighteenth Centuries. 

§ 1. Introduction. 

The authorities enumerated in the preceding chapter, 
enable ns to form a tolerably correct conception of the pro- 
nunciation of English during the xvi th century, and to note 
the principal changes which it underwent in the xvu th and 
XTiu th centuries. It is the object of this chapter to shew 
as precisely as possible — althougn of course far from as pre- 
dsely as desirable — ^what the pronunciation indicated for 
each period really was. The results which have been c^iyen 
by anticipation at the end of Chapter I, are arranged cupha- 
betically. But it will be far more convenient to adopt a 
different order in the present chapter, and revert to the 
alphabetical in a subsequent recapitulation. See Chapter YL 

The principal authorities described in the last chapter 
will be better appreciated by arranging them chronologically 
in connection with the names of the contemporary sovereigns 
and the chief contemporary writers. Any statement can 
thus be immediately refeired to its proper political and 
literary epoch. 

It must be remembered that the authorities for a period 
are necessarily somewhat more recent in date than the period 
itself, for the account which an elderly man gives of pro- 
nunciation refers in general to that which he acquired as a 
youth. It is in most instances safe to assume that a man's 
system of pronunciation is fixed at twenty to twenty-five 
years of age. The first ten years of his life are spent in 
acquiring sounds from his nurse, his mother, and his family. 
In the next ten, he is jostled with his schoolmates or work- 
mates, and he will probjEtbly adapt his mode of speech to his 
environment. After the mental faculties have matured, the 
acquired habits have become settled, and the environment 
fixed at twenty to twenty-five, little change may be expected, 
except under rare and peculiar circumstances. It is probable, 
therefore, that each of^ the authorities on the next page, re- 
fers to a pronunciation prevalent twenty or thirty years 
before the actual date. 



50 PRONTTNCIATION IN XVI-XVUITH CENTURIES. Chap. III. § 1. 



CHRONOLOGICAL TABLE OF AUTHORITIES. 



Date 






o» 


AVTHORITT. 


Wbitbbs. 


SOTBBBIONS. 


Work 










For the XVIth Century, 


1530 


Palsgrave, London 


LordSurrey, 1516-46 
Tyndale's Bible, 1535 


1509 Hen. VIII 


1545 


Meigret, Lyons 


Sydney 1544-86 




1547 


Salesbury, Wales 


Spenser 1553-98 


1547 Edw. VI. 


1550 


Meigret, Lyons 






1555 


Cheke, Cambridge 




1558 Elizabeth 


1567 


Salesbury, Wales 


Shakspere 1564-1616 




1568 


Smith, Essex 






1569 


Hart 






1570 


Levins 






1573 


Baret 


Ben Jonson 1574-1638 




1580 


Bullokar 


Massinger 1584-1640 








Milton 1608-1674 


1603 James I 


1611 


Cotgrave 


Authorized Version 




1611 


Florio, Italy 


[1611 




1621 


Gill, Lincolnshire 


Butter 1612-80 


1625 Charles I 


1633 


Butler 









For the Xrilth Century. 


1640 


Jonson, Westminster 


Diyden 1631-1700 


1646 


GatAker 




1651 


Willis, Middlesex 




1653 


Wallis, Kent 




1668 


WilkiuR, Oxford 




1668 


Price 




1669 


Holder 




1685 


Cooper 




1688 


Miege, France 


Pope 1688-1744 


1701 


Jonetf, Wales | 1 


For the XVIIIth Century. 


1704 


Expert Orthographist 




1710 


Dyche 


S. Johnson 1709-84 
Goldsmith 1728-74 


1766 


Buchanan, Scotland 




1768 


Franklin, U.S. 




1780 


Sheridan, Ireland 





1649 Common- 
[wealth 
1660 Charles n 



1685 James 11 
1688 Wm. Ill 
1702 Anne 



1714 George I 
1727 George II 
1760 George ni 



Chap. III. § 2. COMBINED SPEECH SOUNDS. 51 

§ 2. Combined Speech Sounds^ 

It is a favourite, and occasionallj convenient theory, to 
suppose that there are three principal vowels (a,, i, u), as 
that there are three principal colours, or rather pigments, 
blue, red, and yellow, whence the rest are formed by mixture. 
Neither theory must be taken literally,, or be supposed to 
represent a fact in nature. Both partake of the same degree 
of partial truth and complete error, as the still older theory 
of the four elements. But as earth,, water, air, fire, still re- 
present solids, liquids, gases and chemical action, so the (a, 
i, u) represent the most open position of the mouth with 
respect both to tongue and Kps, and the two most closed 
positions with respect to tongue and lips respectively through 
which a vowel sound can be produced. A vowel sound is 
properly a musical tone with a definite quality or timbre,^ 
and, to be distinctly heard and recognized, the position of the 
vocal organs must be kept fixed for an appreciable duration 
of time, the longest time being really a small fraction of a 
second.* But vocal sounds may be also heard through 
changing positions. These are the "glides,"' which are 
naturally generated in passing &om any position of the 
organs of speech to any other, while the vocal ligaments of 
the elottis continue to act. The best mechanical illustration 
of this effect is obtained by sliding the fii^g^r down a violin 
string, while the bow is kept in action. This glide is the 
essence of aU combination of vocal elements ; the cement, as 
it were, which binds them into masses. In diphthongs, as 
Cai, au), the action is most clear, and Mr. Melville BeU has 
introduced a series of glide signs for exclusive use in diph- 
thongs. But the same action is audible in (pa, ka), the 
glide commencing with the loosening of the contact, and 
continuing until the full sound of (a) is produced. It is this 

flide which alone gives soxmd and meaning to the (p, k). 
n palaeotype the isolated letters all mark fixed positions, 
whether initial or final, and their combination indicates the 
glide occurring between them, in addition to their own value, 
unless a comma (,) be interposed, which cuts out the glide, 
and thus distinguishes the dissyllable (u,i) French ou'i, from 
the monosyllable (ui) French out, which again must be dis- 

* This is Sir Charles Wheatstone's deliberately^ three times, and rapidly, 
theory, snbseqnently verified by Prof, four times in a second. 

H. Helmholtz, Die Lehre Yon den > This phonetic term was introduced 

Tonempfindnneen,2nded. 1865, p.l63. and explained by myself, Univtrsal 

* The word eat^ although contain- Writing and Printing, 1856, p. 6, col. 
ing A long Towel, can be pronounoed 2,andJ^/tiAPA0fM<t«f, 1854,p.8,{61. 



52 COMBINED SPEECH SOUNDS. Chap. m. { % 

tinguislied carefully from the monosyllable (wii), English toe, 
where the first element is a buzz and not a voweL This 
conyention in notation will be strictly carried out and should 
be careAilly observed by the reader. As a necessary conse- 
quence (aa, nn, ss) represent prolonged (a, n, s), but (a,a, 
n,n, s,s) repeated (a, n^ s). The prolongation of consonantal 
sounds may appear strange, but if unowned is compared with 
unknoum, or missile with misaent, it will be readily perceived 
that the (n, s) in the second of each pair is really prolonged, 
thus (en^oond* ennoon*, mts'tl mtssent*), and that the ormo- 
graphy (9n,noan*, m«,sent') would not quite meet the latter 
case, as there is no cessation of sounds, no ending of the one 
(n, s) and beginning of the following. Again, in comparing 
open opening ; stable stabling^ schism schismatic ((XTpim oop'ntq ; 
st^^b'U st^^bitq, stz'mm sizmaet'tk), the greater length of 
soimd of (n,l,m) in the first three words over that which it 
has in the second three, will be apparent. Generally, how- 
ever, it is sufficient to mark ((jopn, st^b'l, siz'm), oecause 
the effort to pronounce (n, 1, m) independently of any follow- 
ing vowel will necessarily lengthen the sound, nnt that 
some attention to this difference is occasionally necessary, is 
shown by such French words as stable^ schisme, which French 
orthoepists also mark (stabl, shizm), although their sound is 
not at all (stabll, shizmm), but either (stabF, shizm*) with the 
faintest vowel murmur following, thus making (1, m) initial 
and consequently shortening the sound, or (stablh, shizmh) 
with an entire remission of the vocal murmur. In palaeotype 
the distinction will often be made thus : Engliim (st^^bl, 
s/Vm), French (stabP, shizm'), so that ('1, 'm, *n) = (ll,mm,nn). 

The glide which connects two vocal elements has a ten- 
dency to draw those elements into nearer relation than they 
would have had if pronoimced apart ; that is, as in the course 
of speech it is necessary to pass rapidly from one position of 
the vocal or^;ans to the other without intermitting the voice, 
the two positions naturally draw nearer to each other. It 
has long been observed that certain vowels affect certain 
consonants. Thus, in Polish, it is laid down as a rule in 
language, that " hard consonants when brought by inflection 
or derivation before high vowels are changed into softer or 
weak consonants."^ 

The other Sclavonic lanc:uage8 have similar rules. In the 
Gaelic language there is also a division of vowels into broad 
a, o, u, and small e, i — " leathan agus caoV^ — ^with the celebrated 
rule which so singularly influences their orthography, ^' broad 

1 •/. BUmackL TheoretiBoh-praktisohe Orammatik der polmsohen Sptadie, 



CHiLP. m. { 2. OaBCBINED SPBBCH SOUNDS. 53 

to broad and small to smaU, — ^leathan ri leathan, an 'us oaol 
ri caoL"^ Of course, this rule only indicates a change of the 
intermediate consonant in actual speech. In German we 
find ach, loch with one sound of eh (kh), ich, dcht, euch^ locher, 
tmeher with another {kh)y and auch^ tuch with a third (kfrh), 
thus (akh, lokh ; iA;h, eiht, oiA;h, loeA^h'er, tyyA^h'er ; auktrk, 
tuukt&h); so that the Germans find a natural character in this 
change. But no such change occurs in Dutch, or in Swiss 
patois, which do not possess {kh). Again, a modem Greek 
informs me that (kh) is always replac^ by (Arh) in his Ian- 
mage, whatever be the adjacent yoweL lliis seems also to 
have been the case in old Sanscrit, where (Ah) has given way 
to (sh), just as most Englishmen hear a Saxon say (ir'iiBh- 
mtidrntisht) for (ir ikk mi^ niA;ht) trr* ich mich nicht, (dtosh) 
for (durArh). llie old Germans nad also a feeling of attrac- 
tion in the vowel sounds in succeeding syllables, as »ahn 
zdhne,Ju9z/iis»e, bock bocke, mann manner , (tsaan t8EE*n^, fuus 
tyj'Be, bok boek*^, man mEn*er) which the modems have lost, 
and which is simply unintelligible in the modem English 
tooth teeth, foot feet, man men, (tuuth tilth, fut fiit, msBn men). 
The initial consonant is in European languages mostly 
altered to suit the following voweL We are familiar witn 
the change of sound of c in the first and second syllable of 
caneei = (ksen'sel), and are accustomed to regard it as a me- 
chanical rule of pronunciation, whereas it is the modem pro- 
duct of an action of a vowel on the preceding consonant. 
Sometimes the action takes place by an appEurent desire to 
avoid this attraction. Most persons are familiar with (Aaaid, 
^aaid) for card, guard, but few are aware that it was through 
a precisely simflar change that Latin cantus, campus fell 
through (A»nt, Aamp) into French chant, champ, botib being 
now (shaA). In Arabic, however, the vowel yields to the 
consonant, and it is chiefly by the ** widening'^ of the follow- 
ing vowet properly due to extending the pharynx for the 

1837, p. 8. The dmnon of Towels and consonants referred to is, in palaeotype 

deM» Towels (a, ai, «, o, nh, t, n) 

high Towels (e, oa, -e, «, o, i, ..) 

hard tonsonants (bdgHkhk/mnprs tbhz) 

softer {.. dz dz ih sh ts rs sh ts .. zh) 

weak (hj dri zh sj sj tsh ^ mj nj pj .. sj tsj bjh zj) 

Such a eombination as ^) is impossible to a Pole, wno is compelled to say 
either {It) or (Iji). 

> This IS thus explained in J. Forbes*s same class, i.e, both broad or both 
Double Grammar of English and Gae- small ; as catl^ag, a ftrl, feorag, a 
lie, 1843, p. 28 : " In words of more tquiml. It would be false orthogra- 
thaa one syllable, the hat Towel of phy to write wordi thus: eailag, 
eadk prseeemig syllable, and ihe/hrtt feor-eaf, cui-lag, lur-eag, cir-adh, 
of iteh iveoeedisg one nnwl be of the bAorea^" 



64 COMBINED SPEECH SOUNDS. Chap. III. § 2. 

pronunciatioii of the consonant, that an Englishman distin- 
guishes Arabic t ^ ^ b, whatever sounds Arabic scholars 
may finally agree that the latter symbols represent, from 
(t d s z).^ The rounding of the lips has often a similar effect 
in English, as in war^ wan, what, wash, squall, = (waaj, WAn 
won, whAt whot, wAsh wosh, skw^AAl). 

A final consonant may yield to the vowel, or force the 
vowel to consort with it. Both cases are common, the 
French fait as derived from loLtm factum shews both effects.* 
In English, and also in French, (1, r, r, i) have had very 
disturbmg effects on the preceding vowel. But the greatest 
changes ensue when two vowels come together, first as pure 
diphtnongs, and afterwards degenerating into a single derived 
vowel sound. It is precisely because (1, i) are so vowel-like 
in soimd that they react so strongly on the preceding vowel. 

Glides and mutual actions do not occur only between two 
vowels or vowel and consonant, but are also frequent between 
two consonants, and are especially marked where one ia a 
mute (p t k), or sonant (b d g), and the other continuous. In 
German the soimd (ts) initial is a true diphthong, like (tsh) 
initial in English. Many writers have considered (tsh, dzh) 
initial to be simple sounds in English, while (tsh, dzh) final 
as in watch, grudge, are generally recognized to be com- 
poimds. This is explained by a consideration of the nature 
of a syllable. 

When a number of pure vowels come together with glides 
between them, it may so happen that there is a gradual 
change from a close to an open, an open to a close, or a 
close to an open and thence to another close position, as 
in (ia, ai, iai), or (ua, au, uau), or (iau, uai), etc. In all 
these cases the ear recognizes one undivided group {avXXa/3)i) 
or syllable. But if the transition be from open to close and 
thence to open, as (aua, aia), the ear immediately recognizes 
two groups or syllables, and the division between them is 
felt to be the moment of the smallest opening of the vocal 
organs, thus in (aua) the syllable does not divide before or 
after (u), but during the pronunciation of the pure (u) as 
held fixed without any precedent or subsequent glide from 
or to the (a). There is in this case a decided mterv^ between 
the two glides. In attempting to make the separation of the 
groups more evident, a speaker woidd either smiply prolong 
(u), thus (auua), or prolong it with a cessation of force in 

^ See (^ ^h <; <fli « s) in the palaeo- forms seem to have been (fakt, &JH, 

typic alphabet. feUt, feit, feet). The form (feUt) pro. 

* Omitting the last syllable, the bably originated the ^Id spelling /ai^. 



Chap. HI. $ 2. COMBINED SPEECH SOUNDS. 55 

the middle, which miffht be expressed by (au-na), or would 
absolutely pause and thus repeat the (u), as au,ua). In this 
way orthographers, by separating the glides, arrive at the 
conception of doubling the letter which indicates the smallest 
opening* This, however, becomes more strongly marked 
when the division of the two glides is a mere buzz, as (ava), 
or sonant as (aba), or mute as (apa), for in these cases pro- 
longation being either difficult or impossible, the ortho- 
grapher, trying to ascertain the letters, says (av,va, ab,ba, 
ap,pa), and by thus separating the glides, actually alters the 
whole character of the word. In the English and other 
Teutonic languages real cases of prolonged medial consonants, 
or really separated glides, are rare, not occurring except in 
compound words or connected words, compare soappot, boot- 
tree, bookcase, penknife, till late, till eight. Miss Smith, yes sir, 
etc^ Hence these nations readily adopted a system of 
doubled consonants for those cases where the first glide was ^ 
onnustakeable ; that is, where the first vowel being short 
and accented, it was difficult to leave out the glide and pro- 
nounce it independently of the vowel ; for example (a,ba) is 
more difficult than (ab,a).' The doubling of consonants came 
finally to be considered the mark of a short accented vowel, 
and is so consistently applied by Kapp,' who, adopting the 
usual German grammatical term, calls this effect a ** sharpen- 
ing" {schdrjung) of the vowel. But Orrmin had used the 
same means of indicating short vowels even in unaccented 
syllables, in the first attempt at a regular English ortho- 
graphy, and lays the greatest stress upon this mode of mark- 
ing short vowels.* 

To continue the theory of the eyllable. The separation 
can be made, as we have seen, by a buzz, whisper, sonant, or 
mute, as well as by a vowel, and several of these being inter- 
posed, the syllable divides on the least vocal or narrowest 
aperture. Thus in watching (wAtshtq), the syllable divides 

^ ManT speakers say (peii'tY) for Towel as could be used at the begin- 

rpeniidif), waiters are apt to fall into ning of a word, — except in the case of 

ueri) for (je6*&^, uid few care to mamfiBst compounds — to belong to the 

iMngimhMus Smith from Miss M^th syllable containing that Yowel, thus 

(mtnnilh*, mts^mrth*). In snch a diteiplins bsging^ he would diyide dft,st> 

eommon name no mistake is likely^ bat pltn be^gto). Snch divisions are mere 

woold Miss Sterry be distingmshed matters of practice, and are beside the 

from Miss Teny, or Miss Stmt from scientific inTeetig^ation of the natural 

Miss Tsntf T&d names from the London diyision of words into groups of sounds. 
Directory ? • Jf . Sapp. Versuch einer Physiolo- 

s Mr. Melyille BeU finds the division gie der Sprache, 1836-1841. 
(a,ba) quite as easy as (ab,a), and hence ^ See the passage from the Omnu- 

always considers so much of the con- lum quoted in Chapter Y, } 2. 
Bonantal group which precedes any 



56 OOMBIKBD SPEECH SOUNDS. Chap. III. { 2. 

between the glide from (a) to (t), and the glide, in this case 
non-YOcal, from (t) to (sh). The orthographer dividing the 
syllables then says (wAt,tshiq), and hears nrst a (t) and th^i 
ms presumed simple sound (tsh) ; whence the orthography 
tehy which never occurs initially. Between ch in chin^ and 
tch in watching, there is this dinerenoe, that in (tshtn) there 
is onlv the gUde from (t) to (sh)» but in (wAtshiq] tiiere is 
also the gUde from (a) to (t). llie palaeo<ypic orthography 
(wAtshtq) implies all this, for to remove the last named glide 
in the last word we must write (wA^tshrq). 

In (wAtsh) we have the same effect of the (t) with its 
double glide, but as the second glide is entirelv unvocal, the 
ear does not recognize a distinct group, and hence receives 
(wAtsh) as a single group or monosyllable. Indeed so little 
is a final whisper accounted, that it is generally introduced 
in English after final mutes, to give them the double glide 
and make them more audible ; thus Wat ! would be uttered 
(WAt* !) not (WAtf) as we should be almost forced to write 
u we wished to imply the absence of the ('). In the word 
act (eekf) we have first a mute (k) with only a precedent 

flide, so that the (t) would be inaudible without the ('). 
iut to say (aek't^) would be unpleasant and affectedly pe- 
dantic. This mode of overcoming a difficulty, which is so 
common and natural in Teutonic nations, is unknown in the 
Romanic or Semitic. The French say (akf), or in poetry 
(akt^), and are inclined even to (ak't'). The Italians assi- 
milate the (k) to the (t), and dividing the glides say (at,to). 
The consequence is that consonants have more weight in 
BrOmanic than in Teutonic tongues, and not only cannot so 
many be pronounced in succession, but when two consonants 
that cannot be pronounced as an initial combination follow a 
vowel, they necessarily lengthen the syllable^ — ^not the vowel, 
as grammarians erroneously assert. 

The hisses are never felt to produce new groups, and hence 
are added on with the ^eatest liberality before as well as 
after close positions. Thus wrists, scrips, (rtsts, skrtps), 
and in Polisn szczkac (shtshkatsj), to hiccough, in which we 
have a frequent combination (sntish) containmg one stop (t), 
preceding the stop (k) with the same ease to a Pole, as the 
simple (sh) before (t) and (p) in atehen, sprechen, (sht«tf,«n, 
shpre^h'm) presents to a German or Englishman, who are 
unaware of the difficulties which such combinations offer 
to Frenchmen and Italians, and to Arabs, whose easy sounds 
are in turn a very shibboleth to Europeans. 
The division of syllables to the eye is therefore a great 



Crap. m. {2. COMBINED SPEECH 80VKDS. 57 

difficnltVy unless some mark be placed oyer or nnder the 
letter or diyisioiii or miless this mark, placed for conveni- 
ence of printing before or after the letter of division^ is to be 
miderstood as merely pointing that letter out. Thus writing 
the hyphen as usual for this purpose, (wA-tshtq) or (wat-shtc]^ 
might be used, but the latter is objectionable as it divides 
a Tery close glide. In palaeotype it is not necessary to 
divide syllables, and when they are divided in speech, the 
consonants are really doubled, as already mentioned, thus 
(wAt,tshtq). When the accent mark is written in palaeotype 
it is generally placed where it is convenient to the printer or 
writer, but as it forms a break to the eye it should not be 
interposed between close glides, so that either (wA'tshtiq) or 
(wAtsh*iq) is preferable to (wAt'shtq). 

TJnaccentea short vowels do not generally glide on to the 
following consonant ; but this follows them legato (smoothly) 
and not staccato (abruptly), to use musical terms. Thus in 
event, society, (i,vent', so,s9i',e,tO we have in English no glides 
— altiiough it is seldom necessary to indicate their absence 
as above. On the other hand, the absence of marked accent 
in French makes the glide distinct, as in 4vinement, socidt^ 
(rv^en'maA, sosi,d;^). (irammarians, as usual, do not recog^ 
nize these distinctions. 

A short accented vowel is in English always followed by a 
consonant on to which it glides, almost before it begins to be 
heard; whereas a long accented vowel can be distinctly 
heard before the glide to the consonant. Consequently the 
elide with us a£^ts the short more than the long vowel. 
One result of this is that English long and short accented 
vowels do not form precise pairs. Thus peat pit, gate get, 
father gather, sought sot, pool pull = (piit pit, ged; get, faadh'J 
geedh'J, sAAt sot, puid pul). The distinction is here made 
clear to the eve. The vowel (p6) does not occur as a short 
vowel in closed syllables in recognised English, but hole whole 
are not unfrequently distinguished as (hooI, hoI). The long 
vowels (ee, oo) are also very frequently pronoimced {eei, oon) 
or (ee'j, oo'w) with a faintly indicated (i, u), following them 
with the utmost rapidity just as the soimd is expiring. It 
is only before the letter r (i) that this effect is generally 
avoided, and then the vowel sounds are changed, thus more, 
Mary^ door, glory are properly (meei, MeeJ'ri, dooJ, glooi'ri), 
although (moo' J, Mee*'ri, doo'j, gloo*'rt) and even (Mee'ri, 
glooTiJare sometimes heard. This diversity of long and short 
vowels, similar to that which probably prevailed in Greece 
when iJie diBtinctions tje,o>o were introduced, while no written 



68 COMBINED SPEECH SOUNDS. Chjlf. III. { 2. 

di£Eereiice was made between a c v long and short, serves 
to mark the difference between syllables with long and short 
vowels very clearly. If a foreigner neglects the distinction we, 
in the ignorance of our ears, often accuse him of lengthening 
the vowel, thus we write his pitp (pit'i) as peetee, confounding 
it with (pii'tii), and we make a Scotchman speak of his meenis- 
terr and his bodk (mii'ntste.r, buuk) when he only says 
(min-jste.r, buk) in place of our (mm'tsti, huk). Most of the 
old English writers thought that the vowel sounds in bite bit 
formed a pair, and we shall find Sir T. Smith completely 
puzzled with the English ee (ii) of which he knew no short 
sound. In languages like the Italian, where the short and 
long vowels exist in perfect pairs (ii i, ee e, ee e, aa a, oo o, 
uuh wh, uu u) the distinction of long and short vowel is not 
much perceived, except before separated glides or doubled 
consonants, as they are termed, and consequently no necessity 
fof indicating them orthographically has been felt. In 
Italian also, final short accented vowels occur unprotected by 
a following consonant, as cittd amd cid (tshit,ta' amo* tsho*) 
which however take a doubled consonant when followed by 
an enclitic syllable as amowi (amov,vi). 

These different usages are important to be allowed for, 
when we derive the pronunciation of any language through 
the observations of one who is not a native. He necessarily 
hears the sounds incorrectly and imitates them at first, if not 
always, with more or less reference to those with which he 
is familiar. Those Englishmen who hear a Scot or German 
say (man, man), hear the words as either (maen) or (mon), 
soxmds which being unfamiliar to the Scot and German are 
liable to sound in their ears as (mEn, mon).^ It is this dif- 
ficulty in appreciating foreign sounds which renders the use 
of any universal system of writing so difficult. Tet indistinct 
and imperfect as a foreigner's accounts must necesssurily be, 
it is almost entirely bv their means that we are able to 
arrive at a conception of the old soimds of our language. It 

1 An anmsing instance of the diffi- intended to mean fbakhfihiiBh*}, itself 

culty of hearing foreign sounds is an error for (bakrhsniish*). This letter 

Quoted in Max Mailer's Lectures on Horh) ^ is almost invariably con- 

tne Science of Language, 2nd series, rounded with (k) by Englishmen. 

1864, p. 169, from Marsh's Lectures, Similarly, if an Engushman asks a 

and taken by him from *^ Constantino- Saxon to repeat after him I had a hat 

pie and its Enyirons, by an American on my heady instead of (di Heed o Htet 

hng rendent,'* New York, 1836, ii. 161. on mai ned) he will probably obtain (ai 

The writer is certain that he spells at Hst a HEt on mai HEt), where the thi^e 

least one word correctljr, for it had English unusual sounds (nsd Hcet ned) 

been so impressed on his mind ; this are reduced to the one common Ger- 

word Ib haetthtateh! letters which man (HEt) =M/^'. 
ought to mean (b»ktsht»sh), but were 



Chap. III. § 3, A — XVI TH CENTUBT. 59 

is the foreigner who generally wants to haye the sounds 
explained, and we find the writers of pronouncing diction- 
aries of English to be mainly Welsh, Scotch, Irish, American, 
French, and German. Those early English writers who 
gave an account of our pronunciation had not studied the 
nature of spoken sounds sufficiently to refer them to any fixed 

S^tional scale, such as we now possess in Visible Speech. 
ence they illustrated them as they best could by reference 
to other tongues ; frequently indeed by Latin and Hebrew, 
which being yery differently pronoimced in different countries 
gaye but an indifferent clue. It ia only by making allow- 
ances for old habits, that we can hope to arrive at an ap- 
{roximate conception of the soimds they had in their mind.^ 
t ia not therefore to be expected that we can assign the 
older pronunciation of oiu* language with anything luke the 
minute accuracy with wliich the modem pronimciation of 
English can be indicated by means of Palaeotype and Visible 
Speech. We can, however, approximate to the sounds so 
nearly that one who thus pronounced them would appear to 
utter familiar words in perhaps rather a singular manner, but 
not so strangely by far as a foreigner's attempts at modem 
Engb'sh, or as the modem English would have sounded in 
the ears of our ancestors.* 



§ 3. The Vowels. 
A — XVI TH Century. 

1530. pALseEAYE says: ''The soundyng of a, whiche is most 
generally vsed through out the frenche tonge, is suche as we vse 
with Ys, where the best englysshe is spoken, whiche is lyke as the 
Italians sounde a, or they witii vs, that pronounce the latine tonge 
aryght" 

The Italians at present always say (a), and never (a). The 
French at present generally say (a) but sometimes (a). The 
reference to Latin, as pronounced "aryght" ought -to imply 
the existence of another English pronunciation in common 
use, which was not (a). This wrong pronunciation we have 
no means of eliciting. Then again tne English pronuncia- 
tion referred to is a theoretical standard, ''where the best 

^ The key-words in VUibU Speteh, who saw them would have read (wan 

p. 94, are pronoimced differentiy by Mr. totsh ay ntfrtshj), sonnds which wonld 

MeliiUe Bell and myself, (p. 26, n. 1.) hare probably been nnintelligible to 

' While writing this I saw tiie tiieir author (Shakspere, T. & u. iii, 8, 
words "One tou^ of nature," pla- 176), who would have certainly under- 
carded on the streets of London, as stood (oon tutsh ot naa'tyyr), strange 
tiie name of a drama. Most of those as this may now seem to our ears. 



60 A — XVI TH CENTURY. Chap." III. { S. 

engl^sshe is spoken," implyinfi^ that there was another pro- 
nunciation which Pakgrave did not approve of. The only clear 
result we obtain is negative, — the long sound was certainly 
not that now in use in England, " where the best englysshe 
is spoken," that is not (ee, ee, ed). But could we trust JPals- 
ffrave to have heard the difference between (sb, a, a), or if he 
had heard it, to have thought it worth noting P In the next 
century at least Wallis heard the French a as (a), and we 
know that even at the beginning of the present century the 
French SmigrSa heard the En^ish a in all as their a, and 
gave that as the French sound in their Grammars. Walker 
gives (iiklAA*) as the pronunciation of dclat, though Smart 
writes (^klaa')i the Frenchmen Feline ^ and Tarver giving 
(^kla). 

The sound (a) is more marked^ and was probably more 
ancient than the finer sound (a), for which the ton^e has to 
be raised from a " low back" to a " mid back" position.^ It 
is very possible that the French may have used (a) and have 
subsequently refined it into (a). It is very probable that 
the Anglosaxons used (a), as the present Germanic nations, 
and the Scotch, have still a neat tendencv so to do. Perhaps 
one of the sounds {a, ah, a) was the faulty pronunciation of 
the Latin a, to which Palsgrave objected. Either (a) or (a) 
is still used in Scotch Latin. It is not likely that at so early 
a period the very thin (8b), — a sound which Englishmen from 
historical tradition connect with (a), but which foreigners 
consulting their ears, refer to (e, e) — was recognized as the 
use of those who spoke English best. It seems safest to con- 
clude that Palsgrave, living in the latter part of the xvth 
and early part of the xvi th century, recognized (aa) lonp^ 
and (a) short as the best pronunciation of English a, and 
that he would at any rate have accepted that pronunciation. 
This view is confirmed by Gilles du Guez's account of French 
pronunciation, probably printed in 1532, and reprinted at 
the end of the French reprint of Palsgrave. He says : " Ye 

^ Walker^s PronouMifip Dietwnar^^ dples of the FreiKilL prammcifttioB, 

and Smart's Walker BemodelUdj are dereloped in a short treatise by J. C. 

well known. Adrim F4Un$j Diction- Tarver^ French Master, Eton, iJondon, 

naire de la pronunciation de la (Longman) 1847, C. O, Joberfs CoUo- 

langae Fran9ai8e, indiqu^e aa moTen qnial Yimck, London, (Whittaker) 

de caract^es phon^tiqnes pr^c^4 avai 1864, and ThMaft Le Phonographe 

m^moire snr la r^forme de Y alphabet, on la Fronondation Fran^aise rendue 

Paris, 1851. This and Tart^t Ezpla- fadle k tons les ^tranffers, Paris, (ehea 

natory prononncing dicticmarj of the les autenrs, roe de rOuest^ ll,y 1867, 

French langoaffe in French and Eng- are the best goides to modem French 

lish, wherein we exact sound and ar- pronunciation that I hare seen, 

tionlation of eyery syllable are dis- 'These technical terms are explained 

tincUj marked, aooordxng to the prin^ in tiie introdnotion p. 13. 



Chap. HI. § «. A — XVITH CBNTUBT. 61 

dial pronoiinoe your a as wyde open mouthed as ye can/' 
whicn ought to make French a ^ {a)\ ** your e, as ve do in 
latyn, almost as brode as ye pronounce your a in englysshe." 
Tins makes French e = (b), and proves that English a was 
not (ad), because Gilles du Guez, as a Frenchman, would not 
have distinguished (e^ sb). Neither du Ghiez nor PaLsgrave 
sCTarate the close from the open French e (e, b) which 
Moigret has found necessary to distinguish by two signs. 
CHlles du Ghiez was French master to Henry VlII. and his 
daughter, afterwards Mary I. 

1667. Salesbubt says of the Welsh sound of a that " it 
hath the true pronunciation of a in Latin/' meaning of course 
Aw pronunciation of that letter, and that it is never sounded 
" so fully in the mouth as the GFermaynes sound it in this 
word wagen.'' He also distinguishes it clearly from (a) with 
a following (u) or (i). This distinction, hereatfter considered, 
leads me to suppose that his Welsh a was neither (a) nor 
(se), and consequently that it was then true (a). The con- 
clusion is not very safe, because certainly, in the next century, 
Wallis makes the Welch a very " thin," that is closer thaii 
(a), and probably (aa), a sound said to be often heard in 
Wales to this day.^ 

1547. Salesbury heard no difference between the English 
and Welsh a, whether long or short. He says : — 

^^ A in English is of the same sound as <i in Welsh, as is evident 
in these words of English alb, aaly cervisia, pale, pool, salb, «a/." 

It is not u9ual in Welsh orthography to distinguish the 
long and short vowels, although Grammarians say that 
the former have an acute accent mark. In his account 
of English pronunciation, Salesbury does not always dis- 
criminate the long vowel, though, as here, he occasionally 
doubles the vowel sign to represent length, and doubles 
the consonant sign to imply the brevity of the preceding 
ToweL We must not suppose, however, that where he has 
neglected to double either, the sotmd was necessarily 
ei£er long or short. No doubt sale was (saal), if ak, pale 
were (aal, paal). Again he writes narrw and sparu? for 
narroipe, aparrowe, although no doubt the consonant was not 

1 During a short residence in An- monly heard in Monmouthshire, just 

glesea ab<mt ten years ago, I did not bordering on tiiose Western T?ngit<>h 

recognize (ee) as in general nse ia counties where (se) prcTails. A gentle- 

WeUh, although I was familiar with man from Cardigan when asied to 

the sound, bo& long and short, from name the first letter in the Welsh al- 

haring resided two years in Bath, phabet, naturally called it (sbsb), though 

where (too!) is the regular sound of a three other Welsh ^tlemen present 

long, as (BsesBih, ktetejd). I have at the same time said (aa). 
linoe been informed that it is com- 



62 



xviTH cEamrr. 



Chat. m. i 3. 



realty doobled in either and the rowel was short in both. 
Nnmeroas examples of such carelessness occur in the short 
list of words with which Salesbory has faTonred ns.^ 

SALESBrByS KXATfPT.ES OF A. 



Old Spsllixo. 


Xbajono. 


WsLffs Liriutt. 


TALAxorm. 


ale 


cenrisia 


aal 


aal 


pale 




P..1 


paal 


sale 




8b1 


saal 


babe 




baab 


baab 


face 


fecies 


£Eas 


faas 


gracyouse 


comis 


gnaws 


graa-8i,u8 


able 




abl 


aaVl 


sable 




sabl 


saab'l 


bake 


coquerepanem 


baak 


baak 


galaunt 




galawnt 


gal'aunt 


plage 


pestis 


plaag 


plaag 


have 


accipere 


haf 


Haay 


lady 


domina 


ladi 


laa-di 


papyr 




papyr 


paaTwr 


mase 


stupor 


infi|m 


maaz 


ahappe 


forma 


ssiapp 


shaap 


ape 


wiTfiia 


ap 


a^ 


naiTowe 


angustus 


narrw 


nam 


gpaiTowe 


passer 


sparw 


spam 


laddre 


scala 


lad-dr 


lad-er 


bladd' 


vesica 


blad-der 


blader 


nagge 


mannuB 


nag 


nag 


pappe 


mamma vel 
inf antium cibus 


papp 


pap 


quarter 


quartapars 


kwarter 


kwar*ter 


band 




bande 


Hand 


bandes 


duae y.plures 
manus 




wandg 


Thomas 




tomas 


tom'as 


flaxe 


linum 


fflacA 


flaks 


axe 


securis 


ags 


aks 


man 






man 


that 




ddat 


dbat 


kappe 


pila 


cap 


kap 


Agnus 




angnuB 


aq*nu8 



* A complete alphabetical list of all 
thete woidi will be found in Chapter 



Yin, { 2, at the dose of the tnuis- 
lationof hisinct 



Chap. III. § 3. A — XVI TH CBNTUBY. 63 

The preceding are all Salesbury's words containing a^ in 
his English selling, Welsh transcription^ and my palaeo- 
typic translation of the last. The meaning is given in 
Latin where he has given it in Welsh, but not otherwise. 
The lonff a, so far as I can conjecture from other sources, 
is placed first. Words with the combinations aly an, ash, 
etc., which will be considered hereafter, are omitted. 
This long list of words in which the long and the short 
sound of a is represented by the same letter, occasionally 
doubled for the long sound, is conclusive in shewing that 
long a and short a were to Salesbury's ears, sounds differing 
only in duration. And as there could be no reasonable doubt 
that short a was then, as it still is generally in the provinces, 
and is admitted to be by some of our orthoepists in a great 
number of words,^ the true Italian (a), so we are led to con- 
oonclude that the long a was also the true Italian (aa), to 
Salesbury. 

1568. Sib T. Smith says : ''A igitur TiatimiTn Angli habent tam 
breue quam lowguin," and after giving some examples, adds : " et 
alia sexcenta, vbi nullius literarom sonns auditor in lingua nostrati 
nisi a vocalis RomansB longSB breuisque." 

This ought to be decisive, but unfortunately we shall find 
that Smith considered the Latin i long to be the English i 
long, that is (ei) according to Salesbury, and hence he might 
have considered the Latin a long to be (ee) as in England to 
this day. Hence it is only by comparison with Sedesbury 
and others that we can interpret his examples thus :-^ 

"A breuis (man) homo, (far) long^, (nat) petaso aut galems, (mar) 
cormmpere, (pas) superare, (bar) vectis, (bak) dorsum. 

"A longa (maan) juba equi, (faarwel) vale bene, (naat) odisse, 
(maar) equa, (paas) passus, (baar) nudus, (baak) in fUmo coquere.*' 

The words (man, baak) being given in Salesbury interpret 
all the rest. Smith does not give the ordinary spelling, but 
always adds the Latin signification. 

1569. Habi, in describing the "due and auncient soundes" of the 
five vowels, says of A, " the first, with wyde opening the mouth, 
as when a man yauneth," and he identifies it with the German, 
Italian, French, Spanish, and Welsh a. 

This identification has the misfortune of being too wide 
and again leaving us in doubt as to (a, a, se). But (aa, a) 
seems the most probable. Still Gill's censure of Hart, whicn 
we shall find justified for ai, would make us doubtful of a, 
were not Hart confirmed by Palsgrave and Salesbury. 

1 Those of which ttaf, bath, bask, demand, are types. Other orthoepists, how- 
era*, prefer (ah) in these words. 



64 A — XVI TH CENTtJET. Chap. III. { 8. 

1580. Bttllokab sap, ^'that there be eight vowels of differing 
Bouiids in Inglish speech : may app^ere by these wordes following, 
wherein are eight notes in voice differing one from another as diners 
notes in musicke." 

The words are given in his phonetio orthography and are 
arranged in this order, '' to lack, to leak, a leek, to lick, a 
lock, to look, luck, Luke," which, for reasons which will 
appear hereafter, I believe are meant for (tu lak, tu leek, 
a liik, tu Ilk, a lok, tu luuk, Itik, Lyyk). The long a, the 
short e, and the long i, all of which BuUokar uses, are not 
noted in this list. Bullokar's sign for (ii) is a modification 
of (e^, and hence there is no security that he should have 
considered (aa) to be the Ions of (a), although he so notes 
it. Perhaps his observation that ^ b dfk are the only "per- 
fectly perfect" letters, that is, used according to their alpha- 
betic names on all occasions, is meant to imply that long a 
is the sound of short a produced. 

1621. Gill says, "In $ et o, duplicatis, sonus ii proprio aliquanta- 
Imn distat ; vt in gbik laqueus,' et eBESNB viridis, sonus vnns est, 
sed in voce priori correptus, in altera longus. Sic in bi7CI3b hie dama, 
et BOOEE liber : neque in his vUa soni differentia est, prseter iUam 
quflB in quantitate percipitur." 

As then he has a proper feeling for vowel pairs, we may 
feel sure that, when he says — 

"A, est tenuis, aut lata: tenuis, aut brevis est vt in (taloou) 
lALLOWB sebum, aut deducta, ut in (taal) ialb fabula aut com- 
putus : lata, vt in (tAil) talle procerus — " 

the two first sounds really only differ in length, but the last 
differs in quality. We cannot, however, feel sure that the 
two first sounds were (a, aa) as written above. In fact, the 
sounds (se, eesD) must have begun to be prevalent at the time 
GKll wrote, and it is only because he decidedly opposed in- 
novations that I consider he really pronounced (a, aa) as was 
probably customary in the days of his youth.* 

1633. BxTTLEB (translating his phonetic spelling) says : " A is in 
English, as in all other languages, the first vowel, and the first 
letter of the Alphabet ; the which, like % and u, hath two sounds, 
one when it is short, an other when long, as in man and mcme, hat 
and haU:' 

* In LeyiM, 1670, we hare "Grume, Pet Why there's a wench : Come on, 
pedica^* on which Mr. Wheatley cites and idsse mee Kate, 

Cotgraye, "Lags, a snare, gum or Lwt, WeU go thy wales oldeLadfor 
grinn." thon shalt hit, 

* Shakn>ere*8 rhyme at the dose of indicates the pronimciation (kaat, 
Tmmnp the Shrew, according to the Haa-t). 

folio 1628,— 



Chap. m. § 3. A — XVII TH CENTURY. 66 

I cannot find any confirmation of this even in later writers, 
until t^e time of Cooper, 1685, who admits a double use of a 
long, pairing can east, ken eane, as will be presently con- 
siderod. Wnat Butler's pair was, whether (aeee, a) or (aa, ae) 
I cannot guess. But as nis book was published about the 
time when a began to change from (a) to (sb), he probably 
did not adopt either of the true pairs (aa, a) or (sdsb, se). 

The effect of the L, N, Nge, Sh upon a preceding A, 
changing it to (au, ai) or (aa, ee) will be most conveniently 
considered under Au, Ai and the above consonants. Omit- 
ting these from consideration, the best conclusion I have 
been able to draw from a consideration of the preceding 
authorities after repeated examination of all their passages 
bearing even remotely on the subject, is that — 

A long and A short during the xvi th century had in 
eeneral the sounds of (aa, a) ; but (aa, a) may have been 
frequent at the beginning and (aah, ah) towards the 
dose of that period. 

A — xviiTH Century. 

1640. Bew Joksow says: "A, with us, in most words is pro- 
nounced lesse, then the French d, as in, art, act apple, ancient. But, 
when it comes before /. in the end of a syllable, it obtaineth the full 
French sound, and is utter'd with the mouth, and throat wide 
open'd, the tongue bent backe from the teeth, as in ah tmal. gal. 
fal. tal. eai:' 

The description of French d would answer for either (a) 
or (a). Although the sound had perhaps not broadened more 
than to (a) during Jonson's lifetime, it would not be safe to 
assume any other sound than (a) for Ben Jonson^s concep- 
tion of the French sound, which must have been opener than 
the English. The precise value of the latter, however, is not 
fixed; but as Jonson was bom in 1574, his pronunciation was 
probably that of the close of the xvi th century, and he there- 
fore perhaps retained (aa, a). 

1663-1699. Wallis is the great authority for the ftilly 
developed pronunciation of the xvii th centiuy. He recog- 
nizes nine vowels, being, according to my mterpretation, 
three guttural (a, ob, e), three palatal (sb, e, i), and three 
labial (o, u, y), so that the sounds of {a, a) are both lost. The 
sound (a) occurs only in the combinations al, au, aw, under 
which it will be considered. Of the palatal vowels he says : 

" Vocales Palatinae in Palato formantur, aere scilicet inter palati 
^ linguae medium moderate compresso : Dum nempe concavom 



66 A — XVII TH CENTURY. Chap. III. § 8. 

palati, elevato lingaae medio, minus redditur, qn^m in guttnralibus 
proferendis. Suntque in triplici gradu, prout concayum magis 
minusve contrahitur. Quae quidem diyersitas duobus modis fieri 
potest ; vel fauces contrahendo, manente lingua in eodem situ ; vel 
faucibus in eodem situ manentibus, linguae medium altius et ad 
interiores palati partes elevando : utrovis enim modo fiat, vel etiam 
si utroque, perinde est. 

" Majori apertura formatur Anglorum a, boe est d exile. Quale 
auditur in vocibus, hat, vespertilio ; hate, discordia ; pal, paUa Epis- 
copalis; pale, pallidus; ^Sa^Ti (Samuelis contractio); same, idem 
lamb, agnus ; lame, claudus; dam, mater (brutorum); dame, domina; 
bar, vectis; hare, nudus; ban, exsecror; bane, pemicies; etc. 
Differt bic sonus a Germanorum d pingui seu aperto ; eo quod 
Angli linguae medium elevent, adeoque* aerem in Palato compri- 
mant; Germani vero linguae medium deprimant, adeoque aerem 
comprimant in gutture. Galli fere sonum ilium proferunt ubi e 
praeccdit literam m yel n, in eadem syllaba ut entendement, etc. 
Cambro-Britanni, boc sono solent suum a pronunciare." Here the 
paragraph ends in the editions of 1653, 1664, 1674, which are all 
I have been able to find that were published during Wallis's life 
time ; but the Oxford reprint of 1765 adds the words : " Italique 
suum." Again he says in another plfiwje ** A plerumque pronuncia- 
tur sono magis exili quam apud alias plerasque gentes : eodem fere 
modo quo Gallorum e sequente n in voce entendement, sed paulo 
acutius et clarius ; seu ut a Italorum. Non autem ut Germanorum 
d pingue ; quem sonum nos plerumque exprimere solemus per au 
vel aw, si pitMiucatur; aut per 6 breve si corripiatur." 

Now if we omit the reference to the Italian, and confine 
ourselves to the description, it certainly ought to give (ee) 
rather than (a). The tongue is, of course, more raised for (a) 
than for (ja) or (a). The two latter are low vowels, the for- 
mer is a mid vowel, but all are hack vowels, that is, the 
nearest approach of the tongue and palate is made with the 
back not the middle of the tongue, as Wallis strictly points 
out. The three vowels made with the middle of the tongue, 
disregarding the effect of widening, are (ae, e, i), or, taking 
the widening into effect, the three normal (e, e, i) and the 
three wide (ae, e, t). Of these (sd) has the greater opening, 
"majori apertura formatur." With this view agrees the 
pairs of words he gives, which must have been either (aa, a) 
or (sBBB, ae). That a change was taking place we have seen 
by the citation from Butler, (p. 64) and it will appear by 
Miege, (p. 71) that the sounds (bbsb, bb) were fully established 
in 1688, before the death of Wallis, and this view agrees with 
all the following accounts. At the present day the sounds (a, 
aa) are almost unknown in the pronimciation of many per- 

^ The Oxford reprint erroneously inserto tit. 



Chap. III. § 3. a — XVHTH CEKTUBY. 67 

sons,^ and except in a few classee of words they are unknown 
among those who pride themselves on exact speaking. Hence 
we need not feel suiprised that the fashion of (a, aa) had en- 
tirely gone out in Wallis's time, and had b^n supplanted 
by (se, aesD.) Nor is there any other period to wnich the 
change, which certainly occurred, can be distinctly traced. 

It is a remarkable &ct that in Somersetshire where the 
sound of (aece) is very common, replacing all sounds of (aa) 
in use in the east of England, as (Bseasth, baeses'ket, SBSBsk, 
kaes&id, HseaBid) = Bath^ basket, ask, card^ hard^ the sound of 
(aa) or (ai) degenerates into (aa) or (aaa), as (laa, draa, 
kaaid) = law, draw, cord.^ But in WalHs's time the true 
sound of (aa) and not (aa) is guaranteed by his vowel pairs, 
" fall foUy, call collar, cause cost, aw'd odd,, saw'd sod/' 

The reiference to the French entendement is of very little 
assistance. We know how the present English stumble over 
the French nasals. We may hear now (ontandmon, oqtoqd- 
maq, SBqtseqdmaeq), and it is very difficult to determine what 
is the oral basis of the orinasal vowel, so strangely is it modi- 
fied by the nasal vibration. Most French writers refer the 
sound to (a), thus (aA), but English people refer it to (o), 
thus (oa), very few keeping it distinct from on (oa, oa P) As 
frequent allusions will be made to the four French nasals in 
f4n, an, on, un,, which are palaeotypically represented by (ca, 
aA, OA, oa), it may here be stated that I)r. Kapp writes (ea, 
aA, OA, <BA 9a), M. Feline seems to mean (ea, aA, oa, 9a), 
Mr. Melville Bell uses (8ba, ohA, ohA, ha), M. Favarger, a 
Swiss gentleman, who has carefully studied the relation of 
French and English soimds, gives as the normal sotmds (ea, 
EA, OA, dhA). The differences are here more apparent tnan 
real, and probably all sets may be heard coexisting in France 
at the present day. 

The reference to Welsh indicates certainly a very thin 
palatal (a) which must have closely approached to the (sb), 
if not exactly reached it, (p. 61 n.). The final reference to 
the Italian may have arisen from Wallis's mispronoimcing the 
Italian long a, making it as thin as the English long a. 

» Walker, 1732-1807, says that " the mar, &c., and in the word fatherr— 

second sound of a ... answers nearly to PrincipleB, 77. 

the Italian a in Toseano^ Romana &o., * The fact was first forced on my at- 

or to the final a in the naturalized tention by being asked in Bath for a 

Greek words ^m^ and mamma ; and in piece of card as I imagined, when a 

hma ; the word adopted in almost all piece of cord was really wanted. Other 

Jiagnages to express the cry of sheep, old pronunciations in use at Bath, are 

We Midom find the Umg tound of this (fair)/flfr, (keez) ksysj (beek-nj beacon. 

Utter in our languagCy except in mono- but (bflesek'n) bacon; while (aa) almost 

syllables ending with r, as far, tar, reappean in (naon) know. 



f 



68 A — XVn TH CENTURY. Chap. III. j 8. 

In Irelandi where we shall see that the English pronuncia- 
tion consorts in many other respects also with tnat of the 
XYiith century, the name soimd of the first letter of the 
alphabet is (sbsb), as was spontaneously pointed out to me 
by an Irish clergjrman, the five vowels a e i o u being called 
(aeee, ee, oi, oo, ran), instead of {ee, ii, oi, oo, juu). A Danish 
lady informed me that the sound of (8D8d) in lieu of (aa) was 
fashionable in Copenhagen. That the transition is easy and 
is not much perceived by the generality of speakers is evident 
from the present scarcely noticed co-existence of both soimds.^ 
But the transition frt)m the xvi th century (aa) to the xviii th 
and xixth century (ee, ee) is scarcely intelligible without 
the intermediate (seee). 

1668. WiLKJSSy after describing the vowel (aa) as formed with 
the tongue in ^'a more concave posture and removed further from the 
palate," says that ** the Yowel a is framed by an emission of the 
Breath, betwixt the tongue and the concave of the palate ; the upper 
superficies of the tongue being rendered less concave, and at a less 
distance from the palate," and he does not allow of any convexity 
of the tongue till he reaches (ee). 

Now it is only for some very unusual mixed vowels that 
there is any approach to a concavity of the tongue, with 
respect to the palate, so this may be regarded as a theoretical 
error. His description must be considered to leave the 
question of (se, a) in doubt. Although it will be seen that 
Wilkins and Wallis occasionally disagree, I am inclined to 
interpret Wilkins in this case by Wallis, and to consider that 
Willans's examples batt bate, val-ley vale, fatt fate, mat mate, 
pal pale, Rad-nor T-rade, implied the pairs (IwDt l^eset, vflel'i, 
vaDSBl, faet faeeet, meet ma^t, psel paeael, Baodnur traesDd). 

1669. HoLDEK writing at the same time says ""We may imagine 
the vowel a to be made hy the freest and openest passage of the 
throat through the mouth and so to have a kind of natural articula- 
tion without art, only hy opening the mouth ; a to be a little strait- 
ened by the boss of the tongue near the throat, and therefore if you 
try to pass from a to a you will find you thrust the end of your 
tongue something forward to raise the boss of the tongue towards the 
palate to straiten the passage." " In a the mouth is more open, 
in a, $. f . the straitenings of the concavity of the mouth between 
the tongue and palate are gradual, both forward & nearer the roof." 

B^ actual trial, I find that this would serve just as well 
to distinguish (a, se), (aa, aa), or (aa, aeee). It is therefore 
not decisive. The illustrative words for a are fall folly, for 
a dJ^fate fat. 

^ The wordB ti<u$y ttaffy demand, are even (ah, oh) are in occasional nse bf 
prononnced with (aa, a, ah, aah, as, others. 
mm)j by different careM speakers, and 



Chap. m. {3. A — XTHTH CBirnJKT. 69 

1685. Cooper seems to mark the begmning of a cliange 
which was not complete till the next century, and does not 
appear to be noticeli by Miege or even Jones, for he gives 
two sounds to a Ion?, generally (smb) as I conjecture, and 
occasionally (ee). In this respect Cooper bears a resemblance 
to Hart, wno anticipated the general pronunciation of at as 
(ee) by a century. Cooper sajB : 

*'A formator k medio lingose ad concavum palati paululibn elevate. 
In his eon possum, pass hy pnetereo, a comptur ; in east jsoeo,^ past 
pro passed praBteritus, prodndtur. Frequentissimus auditur hie 
sonus apud Anglos j qui semper hoc modo pronimciant a latinum ; ut 
in amabam. Sic etiam apud Camhrehritannos ; quandoqiM apud 
OaUos; ut in ammaiy demands j mh autem aut nimquam apud 
Germanos, Hunc sonum correptum & productum semper scribimuB 
per a; at huic chaiacteri pneterea adhibeutur aonus imus & alter : 
prior, qui pro vocali ejus longa habetur ut in eane, definitur sect, 
sequenti ; posterior ut m was sect, septimft sub o gutturalem." 

He here implies that cane although considered the long 
of can is not so. He also for the first time makes teas =: 
(wAz), whereas Wilkins wrote «az = (usez) meaning (wsez). 
These are both anticipations. He implies that though short 
(sb) was common, long (sdgb) was uncommon, and identifies 
the sound with that of me Welsh a, which he must have taken 
as (8B8b). He allows that it "sometimes" is in use in French, 
in which language it is to be supposed he called a generally 
(aa). The two examples animal, demande are insufficient to 

S>e assistance. He says that it never occurs among the 
ermans. The present German sound in great part of Ger- 
many is (oa, a), and in Austria it becomes (oah, a) or perhaps 
(a A, a). But throughout North Germany the sounds (aa, a) 
are constantly heard from the more educated and refined 
speakers, and though Schmeller distinguishes the Italian from 
the common German a, neither Rapp nor Lepsius notice the 
difference.* Yet in the xvn th century the general impression 
seems to have been that the French and Germans said (i^A). 
Was this really the case P I think not.^ I woidd rather trace 

^ Hisprint foTJaeio ? by the researches of Serffarth, lisoov, 
* SekmelieTy Die Mrmdarten Bay ena, etc., that lon^ a in Greek had the 
Miixichen 1821, Nos. 62. 66. £appf BOimd of Itahan a in amare, that is, 
Phjsiologie der Sprache, passim. X^ (aa). And then he imMediatelj said, 
9tu$, Standard Alphabet, London and "the lon^ a should always be pro- 
Berlin, 1863, espedaUy p. 50, where noonced like the English ocr or aM, as 
tiie Rnglish sounds are taken into con- in eawlj maul, ^'U ^^^ ^. ^^' 
eideration. (Proceedings of the Boyal Institution, 
> Mr. Blackie, the Professor of Greek vol. t. p. 149.) Here then we have a 
in tike University of Edinburgh, when recent example of a lecturer upon pro- 
leeiaiing on the pronunciation of Greek nunciation, confusing the two sounds(aa, 
before we Boyal Institution, 3rd May, aa). We must not expect our ancestors 
1S67| said that it had been established to nave been much more particular. 



70 A — XVII TH CENTUKY. ' Chap. III. } 8. 

it to the loss of the pure (aa, a) in refined English, and its 
separation into (aa) on the one hand, and (seed, se) on the 
other. To those accustomed to say (aecB, aa) the intermediates 
(aa, aa) would both be referred to (aa) rather than (aeae). 

The opinion that a long had become (aeee) seems to derive 
additional force from the fact, first mentioned by Cooper, 
that a long had in many words become (ee). He says — 

** E formator k lingua magis eleyata et expanse quam in a 
propriiis ad extremitatem, unde concayum palati minus rcdditur & 
sonus magis 6tcutus ; ut in ken video. Sic apud Qermanos menschen 
liomines. Apud GaUos raib at in exc^s^ protiste, aessianj & Benjamin 
obsolete. Hunc sonum correptum Angli semper exprimunt per e 
brevem ; & e brevem nunquam aliter pronunciant nisi ante r, ubi 
propter tremulam ipsius motionem, & yocalis subtilitatem subita cor- 
reptione comitatam, vix aliter efferri potest quam ur ; ideo per in 
pertain pertinco, & pur in purpose propositimi ejusdem sunt yaloris. 
Vera hujusce soni productio scribitur per a, 9.U\ue a longom false 
denominatur ; ut it eane canna, wane deflecto ; & ante ge ut age 
setas ; in cseteris autem yocabulis, (nifallor) omnibus ubi e quiescens 
ad finem syllabae post a, adjicitur ; u gutturalis . . . inseritur post a ; 
ut in name nomen, quasi scriberetur na-um dissyUabum." He pro- 
ceeds to say that this sound is usually written ai or atfy sometimes 
eg and rarely ea. 

Here we have two curious facts, first the clear recognition 
of an (ee) sound of long a, and secondly the insertion of (e) 
after (be) in all but a certain class of words. Thus cane, 
name =z{kEEii, nEEom). The peculiarity here is, that so far 
from inserting (o) in modem times, the tendency is to palat- 
ize the sound still more by inserting (i) tnus (ne^m). 
Cooper returns to this point again, saying — 

" Post a in omnibus, nisi in eane canna, u?ane deflecto, stranger 
adyena, strange alienus, manger prsesepe, mangy scabiosus, & ante 
ge; ut age setas; inseritur u gutturalis, quse nihil aliud est qu^ 
continuatio nudi murmuris postquam a formatur, nam propter exili- 
tatem, ni accuratitis attenditur ; ad proximam consonantem, sine 
interveniente u non-facile transibit lingua. DifTerentia auribus, quae 
sonos distinguere possunt, manifest^ apparebit in exemplis sequent! 
ordine dispositis. 

a breyis. a longa. a exilis. 

Bbt yectis Bdrge nayicula Bare nudus 

hhh eflutio hldist flatus llaxmi diyulgo 

eop pileum carking anxietas cape capa 

car carrus carp carpo care cura 

cat catus cast j actus case theca 

dash allido dart jaculum date dactylus 

Jlash Mguro flasket corbis genus Jlake flocculus 

gash caesura gasp oscito gate janua 



Chap. III. § 8. a — XVII TH CENTURY. 71 

a breyis a longa a exilis 

ffrand grandis (frsLnt concedo grange villa 

IsMd terra hnch solvo lane viculus 

m^h farrago mo^k larra maeon lapidarius 

poi aptus pdih semita pate caput 

te^r pix fluida tfxrt scriblita tares lolia 

Si quid amplius ad banc veritatem confirmandam yelles, accipe 
exempla sequentia; in quibus ai leniter pronunciata sonum babet 
a pujSB ; ut in cane^ a verb post se admittit u gutturalcm ut, 
Bain balneum Hail grando Maid virgo 

lane venenum hale trabo made factus 

main magnus lay'n jacui pain dolor 

mane juba lane viculus pane quadra 

plain manifestus epaid castratus tail cauda 

plane Isevigo epade ligo tale fabala." 

Here I interpret a brevis = (se), a loDga = (aeae), a exilis 
= (ee), thus (baer, baeaerdzh, bEEr), and in the last list I read 
(bEEn bEEan, msEn mEEdn, plEsn plEEan) or (b£En b£E'n),etc. 

1688. 2dj£6£ says : Dans la langue Anglaise cette voyelle A 
s'appelle et se prononce ai, Lors qu'elle est jointe avec d'autres 
Lettres, elle retient ce m^me Son dais la plupart des Mots ; mais il 
se prononce tantot long, tantot bref. • Ua se prononce en ai long 
generalcment lorsqu'il est suivi immediatement d'une consonne, et 
d'une e final. Exemple^/or^, tare, care, grace, fabUy qui se pronon- 

cent ainsi, faire, taire, caire, graice, faille D'ailleurs, a se 

pronounce en ai bref ou en e ouvert, lorsqu'il se trouve entre deux 
Consonnes, au nulieu des MonosyUabes ; comme hat, cap, mad, Mais 
il approcbe du Son de notre a, si la fin des I^oms en al, ar, & ard 
qui ont plus d'une syllabe. Exemple general, special, animal. 
Grammar, altar, singular, particular ; mustard, custard, lastard, 
vizard, & autres semblables. Excepts regard, qui se prononce re- 

gaird ; award & reward oii il sonne comme en Fran96ds Dans 

le mot de Jane Ya se prononce on e masculin, Dgine^ 

To understand this we miLst remember that English hat, 
cap, mad were never, and are not now, called (HEt, ksp, msd) 
but that Frenchmen, and even Germans, do not distinguish 
them from these soimds. Indeed the true soimds (HSDt, keep, 
maed) only differ from the former by the widening of the 
pharyngal aperture. My own pronunciation of (ae) nas been 
constantly misunderstood, and considered as (e) or (e). As 
to the long sound (sdsd) it is now so little known in the East 
of England and on the continent, that it would be invariably 
taken for (ee) or (ee). When then Mi^ge distinguishes 
Jane = Dg^ne (Dzh^^n) from grace = graice (grees, grEEs), 
we may feel pretty sure that, since in modem English (grEEs) 
is as difficult to English organs as (graasDs) would be to 



72 A — XVnXH CENTUKT. Chap. III. § 8. 

Frencli organs, the words containing a to which he assigns ai 
long and &ort, were really pronounced with (esed, ae). 

As to those words in which he considered the a to be pro- 
nounced as in French, we know they had the sound (aa) and 
not (aa) and we also know that at present most Frenchmen 
pronounce our (aa) as (aa) or {aa\ neglecting the labial 
effect. The exception regard, was probably (re^raeeBrd'), with 
the palatial (g) which is still so prevalent in this word, and 
whicn may have caused the pure sound of (edSd) to be pre- 
served. Whether the sound of (aa) occurrcd in mmtard, 
cmtard, etc., we cannot tell. At any rate, this notice is 
not sufficient to establish the fact. 

1701. Jones's book is so curiously arranged that it is diffi- 
cult to determine the sound of a long from it except by in- 
ference. It is certain that at this time ai was sounded (ee) 
or (ee), probably the former. When Jones therefore gives 
a list 01 words in which ai has the sound of a, but may be 
soimded as ai, he certainly distinguishes the two sounds. 
That is although in some words ai was bv some people 
sounded as a, this was not universal or considered best, even 
in those words. They are Abigail, aid, bargain, captain, 
certain^ chair, complaisant, fair, glair, hair, laid, maid, pain, 
pair, plaister, stairs, etc., (32 examples are given) of which 
plaister is now generally pronounced (plaas'ti). Then he 
adds this note : 

^' The capacity of being sounded ai distuigiiishes them from such 
as are written with an <i ; because these cannot be sounded at, as are, 
chare, fare, glare, hare, lade, made, pane, pare, stares, etc." 

Again, the question, " when is the sound of ai written a P* 
is not a^ed, and the answer to the question, '^ when is the 
sound of e written a ?" is only answered by the cases of im- 
accented -ar as altar, beggar^ emissarg, bastard, etc. As then 
Jones could not have said (ee) or (aa), I conclude that he said 
(sdgb), and this agrees with the fact that Jones only recog- 
nizes two sounds of a as in an, as, at, and as in all, ball, so 
that his sound of a long, when evidently not (aa), should be 
the long sound of his a in a^ which was certainly (8b). 
From all these considerations I conclude that 

a short was (sb) very early in the xvii th century, and 
that it has retained that sound to this day, except in the 
provinces, and also that a long was generally (sesD) from 
at least the middle of the xvn th century to its close, 
although about the close it began to degenerate into (ee) 
in many words. It is possible, however, that the soimd 
of (aa) may have remamed imrecognized before r when 



Chap. IIL § 8. A — XVnXH CBNTUBY. 73 

not followed by a vowel, and even in several of those 

wordsi as bath, ask^ grant, etc., because it may still be so 

heard in the xixth century. 

Bhvmes at the latter end of the xvi th and during the 

XVII th centuries are not of much use in determining sound, 

unless thev are frequent usual normal rhymes. Thus from 

Shakspere s rhymes in — 

Vmus and Adonis v. 47, broken open, 134 voice juice, 419 young 
strong, 592 neck back, 773 nurse worse; and m Luerece v. 13 
beauties duties, 62 fight white, 72 field killed, 78 tongue wrong, 
113 hither weather, 303 ward regard heard, 408 blue knew, 554 
dally folly, Swmet 20 created defeated; Lover's Cotnplaint 302 
matter water ; Passionate Pilgrim 308 talk halt, 

nothing could be inferred. But when on looking through 
the whole of his poems (exclusive of his plays) 1 find only 
the following examples of long a rhyming to ai, Venus v. 271 
mane again, 529 gait lato, Lucrece v. 6, waist chaste, Sonnet 
128 state gait, of which gait and waist are only modem forms 
for gate tcaste,^ so that there i^ only one real example left 
(mane again), we may safely conclude that Shakspere pro- 
nounced the sounds dififerently, that is, as I believe (aa, ai). 
When in the xvuth century, a long and ai altered, as I 
think, to (aesB, sei) and in the latter part of the century ai 
became (eei) or (ee), we may well expect to find these rhymes 
more abundant. In Milton's rhymed poems I find only — 

Lyeidas care hair, raise blaze praize, L^ Allegro maid shade, fail 
ale, cares airs, 77 penseroso cares airs, state gait, fail pale, Arcades 
blaze, praise, Sonnets 8 spare air bare, 15 praise amaze raise displays, 
19 state wait, 20 air spare. Nativity y near the end, pale jail, Fair 
Infant air care, Solemn Music made sway'd, Anno JEtatis xix (1627) 
aid made, Psalm 2 made sway'd, 4 spare prayer, 80 declare prayer, 
laid made, 83 said invade, strays blaze, 88 prayer are. 

These cannot be considered numerous in such a large col- 
lection of verses. But Milton's contemporary Waller has, 
in some 130 pages of his works which 1 have examined, 21 

* In Merry Wives, act i., sc 8, 1. 41 dyl," and Palsgrave "fr«»< a middle;** 

(Globe edn.) according to the old ouarto the word is not in Levins in this sense. 

of 1619, supposed to be the first sketch, In the same 4to. of 1630, act 1, sc. 4, 

we have the following orthography of 1. 31 (Olobe edn.) and act 3, sc. 3, 1. 68, 

tcMt : *' jPo/. WeU mj honest lads, He we have first '* I should remember him, 

teU yon what I am about. Fit. Two do's hee not hold vp his head (as it 

yards and more. Fal. No gibes now wereF) and strut in his gaUf** and 

PisioU ; indeed I am two yturds in the seconmy ** the firme fixture of thy foote, 

wU, but now I am about no wuU : would give an exceUent motion to thy 

hdeAj, I am about thrift you rogues pate in a semidrcled farthingale." I do 

j€n» In the quarto of 1630 the two not find the word in this sense in 

words are tcattf uxuU. The Promp- Promptorium, Palsgrave, or Levins, 
toiium has *^wasU of a mannyi myd- 



74 A — XVIII TH CENTURY. Chap. III. { 8. 

cases of a similar kind. Dryden has 27 instances in his 
Fable of Palamon and Arcite alone, which belonged to the 
close of the xviith century. 

Now (8B8b) and (ee) are not very unlike, and before (i) it 
is difficult to distinguish them, as carey air (kseaw, eej), 
especially if the (ee) be deepened into (be) as is sometimes 
done.^ Hence we must not be surprised that poets to whom, 
as Byron confesses 

" sometimefl 
Monarchfl are less imperious than rhymes," 

should take the liberty of considering these sounds as 
identical. If they had been (aeaB, aBaei) they would have 
passed for rhymes, just as few of those who now insert 
an (i) after (ee) as in (werit, str^rit) wait^ straight, are even 
aware of the fact, much less would feel that the rhyme 
were injured, if others said (st^^t, g^^^t) or even (steet, greet) 
for state, great. The German habit of rhyming (ob, e) and 
(y, i) although justified by the pronunciation of the unlettered, 
is yet admitted by the best poets. In this case the vowels 
difier by the important distinction of labialisation, whereas 
(ee, aeae) as they may have been soimded, differ only by the 
effect of widening, which is constantly disregarded. 

A — xviii TH Century. 

1704. The Expert Orthographist talks of the "short 
and long soimd common to all the vowels in rat & rate" 
This ought to mean that these words were (raet, rsBset), but 
with a person so destitute of real phonetic feeling, (raet, reet) 
might have been thought to have a " common sound." His 
expression also might not have meant that the long soimd 
and the short sound were the same. The following passage 
is noteworthy. 

" Take special notice that the Dipthong at and the Vowel a are 
very apt to be mistaken," i,e,, confused one for the other, " the 
[iOndoners, affecting (as they think) a finer pronunciation, would 
Utiite lose the sound of the proper diphthong at, as too broad and 
clownish for their fine smooth Tongues ; but the honest Countryman, 
not to say our Universities will (by no means) part with authentick 
Custom, time out of mind, according to its natural sound ; however, 
t<J reconcile this difference, you must be sure to keep close to the 

^ The story that Kin? James I., ye sail hae," and united the bishoprics, 

wbihin? to bestow the Dishopric of although it labours under the historical 

I'Uher Bath or Wells on a west country difficulty of uniting the sees 500 years 

divine, asked him which he would haye, after their union, scnres to shew the 

[iiicl on being told Bath (Bs&aeth^, re- near coincidence of the sounds. 
]i\iiA **Baith (beeth) say ye, then baith 



Chaf. m. § 3. A — XVIII TH CENTURY. 75 

orthography, which that you may the better do ; always remember 
that the single a mnst end no English word ; but if thii/ wiU speak 
Jht0j yet be sure that you write true, by adding y, not da but day. 
Observe that tho' many times this Diphthong at is parted in proper 
names, as «7tf-ir, La-ishy Sepharva-im &c. yet f is usually swallowed 
up, in the sound of the forgoing a, especially when the word ends 
in aA as Benai-ah^ Serai-ah &c. the t is not sounded." 

This feeble attempt to keep long a and ai apart seems to 
be dictated by theoretical grounds. He had previously said 
there were 16 sounds : " five short and five long sounds be- 
longing to the vowels, besides five such proper diphthongs as 
make five other distinct sounds, differing from the foregoing 
ten sounds." And he assigns as his first reason for admitting 
none other but at, aw, oi, oo, and ot^ to be proper diphthongs, 
that "none but these five have such a plain distinct sound, 
different from the five vowels." Hence it was important 
for him to distinguish long a and at, though in pronunciation, 
the utmost difference which I can suppose him, with his 
palatal tendencies, to have made, is to have called long a (ee) 
and ai (eei). The first conclusion is strengthened by his 
identifying his long a with the vowel in there, were, where, 
which was certainly (ee). 

1710. Dyche distinctly says ai, ^ry = a in care, and as 
Cooper in 1685 had given the pairs sell sail, sent saint, tell 
tail, tent taint, there ought to be no doubt that at this time 
the change of the sound of long a from (aa) to (ee) was fully 
established, notwithstanding that Jones only nine years be- 
fore would not allow that long a was pronounced as ai. At 
the same date as Dyche, the anonymous instructor of the 
Palatines writes the words / make, I have, care in German 
letters ei mdhk, ei hdhf, kdhr which should mean (ai mEEk, 
91 HEEf, kEEr), but would have been written even if the real 
sound had been (aese). Here hate is made to have long a, 
as it used to have ; it is now (hsdv) and the pronunciation, 
(Heev), indicated by the German letters is very doubtful. 

1766. Buchanan always uses ai to represent the long 
sound of a. 

1768. Franklin simply gives men, lend, name, lane as 
examples of the same soimd, and this is nearly the modem 
practise. 

This change of (a) into (e) has also occurred in French. 
Chevallet^ says : " Le changement de a en ^ est frequent dans 
le langage du peuple de Paris : . . . . des le conmiencement 

' Origine et formatioii de la langue Fran9ai8e. Paris, 1853-7, vol. L, part 
3, p. 69. 



76 A — XVni TH CENTUEY. Chap. III. } 3. 

da XV* sidcle Q^offiroi Tory observe ohez lee dames de Paris 
la tendance que je viens de signaler. . . . ^Les dames de 
Paris au lieu de a prononcent e bien souvent, <]^aant elles 
disent: 'Mon mery est & la porte de Peris ou il se faict 
peier* . . . telle maniere de parler vient d'accoustement de 
jeunesse ;' Geoffix)i Tory, Ohampfleury, fo. xxxiii, P." The 
same writer quotes (vol. i, part 2, p. 65) from various imitators 
of popular pronunciation, eri^re, tramontane, terrir, douainier, 
errnes, ou^te, plaine, cl^rinette, ^paimeul, for arrifire, tramon- 
tane, tarir, douanier, arrhes, ouate, plane, clarinette, ^pagneul. 

1780. Sheridan seems altogether to ignore the sound of 
(aa) in English, allowing only (seee) to the English a in /ar, 
bary psalm, balm. Being an Irishman who had devoted his 
attention for years to English pronunciation, while his fre- 
quent residences in Ireland kept his ear alive to the Irish 
pronunciations of English then current in educated society, 
his remarks upon Irish pronimciation are of considerable 
importance. They serve to shew generally that the Irish 
peculiarities arose partly from the persistence of xvii th cen- 
tury pronunciations, and partly from an endeavour to correct 
that pronunciation by the then current English usage, which, 
learned rather by rule than custom, was carried to an excess. 
There will be frequent occasion to notice this as we proceed* 
With respect to a, long a is frequently (absb) in Irisn where 
it is (ee) in English, and sometunes (seas) in Irish against 
(8b) in English. He instances patron, matron, rather, 
which in England were (pee'tron, mee'tron, reedh'j) and in 
Ireland (paet'ran, msBt'ren, raeaadhu). These were evidently 
the older, xvii th century sounds, which have arain become 
current in England, where even the older (raa'dhj) is com- 
mon. The pronunciation (raDdhu), may be heard from 
Americans, among whom there is also a great tendency to- 
wards the pronunciation of the earlier setders, 1628. Thus 
the true sound (naert) may be heard in America, which is 
very rare in England. 

As a general rule the words in -aim, which Sheridan pro- 
noimced (-aDeem), were according to him, called (-AAm) in 
Ireland, as (bAAm, sAAm, kwAAm, kAAm, kAAf ) for balm, 
psalm, qtmlm, calm, calf, and this was a distinct xvn th cen- 
tury sound. In the following words, which he cites, there 
is sometimes an ** overcorrection'^ of the kind above alluded 
to : gape, gather, catch, quash, clamour, wrath, tcroth, farewell, 
squadron, were then pronoimced in England (gseeep, geBdh'or, 
kfetsh, kw8Bsh, klaem-or, rAAth, rAth, faDr'wel, skwAd'ran) 
and in Ireland (geep, gedh'ar, kEtsh, kwAsh, klseae'mar. 



Chap. HI. } 8. B, BB, EA — XVI TH CENTURY. 77 

Tseeeih, reeth^ feerwel, skwaeee'drdn). The received usage of 
the XIX th century varies between the two^ and may be ^en 
as {geep, gsedh*!, ksetsh^ kwAsh^ klsem'i, raath^ rAAth^ feei'- 
wel", skfTAd-ron.) 

The recognized pronunciation in the xviu th century seems 
then to have been, short a = (se) in all cases^ lon^ a 

Senerally = (ee), the exact quaKty (ee, ee, eb) bemg 
oubtfiily and in those cases in wnich (aa) is now fre- 
quently heard, as in dart, father, etc., long a was = (seas), 
as it always was in the xvii th century. 

E, EE, EA — XVI TH Century. 

1530. Paubgravb says: "J? in the frenche tong hath thre 
dyverae sowndes, for somtyme they sownde hym lyke as we do in 
our tonge in these words, a beere, a heest a peers, a heme and suche 
lyke .... The sowndyng of e, whiche is most generally kepte with 
them, is suche as we gyre to « in our tong in these wordes aboue 
rehersed, that is to say, lyke as the ItaHanes sounde e, or they with 
V8 that pronounce the latine tonge aright : so that e in frenche hath 
neuer suche a sownde as we vse to gyue hym in these wordes, a bee 
suche as maketh bonny, a beere to lay a deed corps on, a peere a 
make or felowe, and as we sounds dyuers of our pronownes endynge 
in «, as we, me, the, he, she, and suche lyke, for suche a kyude of 
soundynge both in frenche and latine, is allmoste the ryght pronun- 
ciation of t, as shaU here after appere." 

Here are laid down two sounds of English e long, as (ee) 
in bear, beast, pear, bean, and as (ii) in bee, bier, peer; we, nie, 
thee, he, she, but the spelling of the two sets of words is not 
distinguished. We shall see that in the xrv th century all 
these words were pronounced with (ee) and that they were 
spelled indifferently with e or ee, sometimes with ie, and rarelv, 
if ever, with ea. In Palsgrave's text ea is very rare, but m 
his vocabularies he uses it freely. The following words taken 
from his vocabulary of substantives will illustrate his con- 
fused use of e, ee, ea. To shew a fiirther advanced state of 
spelling I add Levins's orthography 1570 of the same words 
preceded by two dots, after Palsgrave's explanations. 

^*Bee a flye .. bee, heche tree .. bech, heed of stone or wode .. bead, 
beane come .. beane, befe meate .. beefe, beaky n fev au guet .. beacon, 
beame of an house .. beame, beare a he beest .. beare, beere for deed 
men .. beare, beest .. beast, beatyng .. beate, dede acte .. deede, deed 
body .. dead, deane of a church, defnesse lacke of heryng .. deafe, 
demyng judgying .. deeme, derenesse chiert^ .. deare, derlyng a man 
m^non .. darling, eare of a man or beeste .. eare, ease rest ..ease, 
easter a hye feest .. easter feast, feanyng faincte .. fain, feate of arms 



78 E, EB, BA — XVI TH CENTUEY. Chap. III. { 3. 

.. feate, fedyng place .. feede, felyng .. feele, fearyng ., fear, fesant 
coke faisant .. fesant, feeit .. feast, f ether plume ., fetber, g&re cloth- 
ing .. geare, geet a blake stone, heed pate or nob .. bead, hepe of 
money .. beape, heale of body..beale, heele of tbe fote .. heele, 
helthe .. bealtby, heape a great quantite .. beape, heer of tbe beed 
ehettevl .. beyry, heree^ a deed body .. berse, heerrgng a fyssbe .. ber- 
ring, hearyng tbe place wbereby we bere ovye .. beare, hert of any 
beest eveur .. beartie, her the of a cbymney .. bertb, heate .. beate, 
hevyn ciel .. beaven, i^/ou^y.. jelouse, kepyng obseruation .. keepe, 
leehe a sorgion .. lecbe, leed a metall .. leade, leee pasture, leafe of a 
tree .. leafe, lefenesse cberete .. liefer, leage two mile .. league, leaning 
to,, leane, leke an berbe .. leeke, lenenesse maigret6 .. leane, lepe or 
start savlt .. leape, leaue lycence .. leave, leven for brcdde .. leven, 
leaner to lyfte witb .. lever, meale of meate .. meale, meane of a 
songe moyen ,, meane, measwre of two gallons .. measure, mede drinke, 
mede rowarde .. meedo, medowe fclde .. medowe, tnekenesse bumilit^ .. 
meeke, nede besoing .. neede, nedyll to sowe witb .. needil, neare of 
a beest roignon^ nesyng witb tbe nose eitemuement .. sneeze, neatee 
ledder cordovayny peace,. ^pcace, pece or parte of a tbyng..pece, 
peache a frute .. peacbe, pecocke a byrde, peake of a ladycs moumyng 
beede .. peake, peele of belles, pele for an ovyn .. peale, peerle a stone 
.. pearle, pese frute poys ,, pease, pescoddey queue lady .. qucene, queane 
garse..queane, realme roiaulme, rede to playe or pype witb.. rede, 
reed berryng.. redde, reed breest a byrde .. brest, reedmsse rovgevr, 
redy money .. reddy, rele for yame*.. reele, rehener ,, reberso, release 
forgyvenesse, reame of paper .. reame, rere banket ralias, rerewarde of 
men arriere garde,, rerewarde, resonableneas ., reasonable, reason .. 
reason, season tyme .. season, see water mer,, sea, secole charbon de 
terre, sede of herbes .. seede, sege before a castell .. sogc, sekenesse 
maladie .. sicknessc, seeke, sekyng or serchyng .. seekc, seals a fyssbe .. 
seale, seame of sowyng .. seame, seme for to frye witb seyn de povrceau 
[saindoux], semelynesse .. semely, see breatne a fyssbe, sertche enquyre 
.. searebe, seats a place .. seate, teching leming .. teacbe, tediousnesse „ 
tedious, teele a byrde plignon., teale, tele a byrde plinget ..tealey 
ieme of a plougb or oxen .. teame, teere of wepyng.. teare, tete, 
pappe or dugge, a womans brest.. teate, tethe dens .. teetbe, veele 
flessbe .. vealc, wede clothyng .. weedc, weke for candels .. weak, 
weykenesse flebesse .. wayk, weke a senygbt .. weeke, weUhe ,, weltb, 
tp^yng pleur.. weepe, were to take fyssbe, werynesse or grefe .. 
wearie, wesant tbe pype .. weysand, wesyll a beest .. wesyll, wevyng 
frame .. weave, whele of a carte .. wbeele, whete come .. wbeate, yere 
xii monetbes .. yeare, yest or barme for ale, zele love or fi^nshyp .. 
zeele, Zealande a countrcy. 

This long list will shew that in Palsgrave's time no definite 
rule had been laid down for the spelling of these words, and hence 
the reader could not discriminate the sounds. It was not till 
after the middle of the xvi th century that anything like a rule 
appeared, and then ee was used for (ii), and ea for (ee). But 
Levins shews that the rule was by no means consistently 



Chap. HI. § 3. E, EB, EA — XVI TH CENTURY. 79 

applied so early as 1570. And even at a later period ea was 
orten used for (e) the short vowel, and simple e often repre- 
sented (ee) and sometimes perhaps, but not often, (ii). We 
often find hee, niee written like thee to give the full sound of 
(ii) and prevent the pronunciation (ee), which was given to 
tlie. The introduction of the difference ee^ ea was therefore 
a phonetic device, intended to assist the reader. Great diffi- 
culty again arose as many words in ea came to be pronounced 
(ii) without any chsmge being made in the spelling, and we 
find orthoepists obliged to give long lists of words with ea as 
(ee), as (e) and as (ii). If it had only been recognized that 
ea was a modem innovation, introduced with a phonetic 
purpose,* writers and printers might not have hesitated to 
replace ea by ^, ee in the two last cases. It is now perhaps 
too late to vfvitQfeesty beesty reepy beenty etc., but there is no 
reason but habit against this spelling, and abundance of 
historical authority in its favour. 

Palsgrave in saying that e was sounded as in Italian, takes 
no notice either in French or Italian of the double sound 
(Cy e) into which (e) splits, although Meigret, 1550, finds it 
necessary to use two distinct vowel signs for the two sounds. 
In modern English we distinguish a»7, atr, = (eel, eei), but 
in some parts in the north of England I find this distinction 
unknown, and (ee) alone pronounced. Hence I suspect that 
the older English sounds were (ee, e). The short sound (e) 
has remained, apparently unchanged, from the earliest Eng- 
lish times to the present day. 

1547. Salesbury gives the two sounds (ee, ii) and also 
notices the mute or tmpronounced e. He scarcely ever uses 
ee or ea. As examples of (ee, e) he gives in his Welsh pro- 
nunciation A were, wreke, breke, wreste = a weir, wreak, 
break, wrest, and calls attention to the difference of meaning 
in bere, pere, hele, mele according as they are pronounced 
with (ii) = bier, peer, heel, meel (to meddle ?), or with (ee) 
=i:bear, pear, heal, meal. Omitting mute e and ea, the fol- 
lowing are all the words containing e, of which he gives 
the sounds ; the old spelling is in small capitals, and the 
Welsh transcription in italics : — 

Brede bred (bred) panis, laddre lad-dr (lad'er), ettermore efer- 
mwor (evermoor) in aetemum, thojtdre thwndr (thund'er), wondre 
umdr (und'er = wund'er), chese tsis (tshiiz) caseus, yrendes frinds 

^ This was so little suspected that ciato/' and when he says it was then 

we find Wallis imagining that ea was **nunc dienun" pronounced (ee) he adds 

properly pronounced as (eea) or (ee*) "sono ipsius a penitus suppresso/* as if 

*'per ^ masculinum, adjuncto etiam si it eyer had been sounded since the xnith 

libet exilis d sono raptissimo pronun- century, except in provincial dialects. 



80 By BE, EA — ^XVITH CENTURY. Chip. HI. } 8. 

(Mindz) amid, tbsbs iriyi (trii'iz) arbores, suffbe iwfffre (saf'er) 
siiiere, gelding gelding ^geld-iq), Gylbebt Gilbert (Gil'bert), gtngeb 
Uini9%r (dzhin-dzher) zinziber, beggynoe begging (beg'iq), bgge eg 
(eg) ovnin, Jbsu tsieeuw (Dzheczyy), quene kwin (kwiin) regina, 
BENT rent (rent), tbesube treeuwr (trez*yyr) theBaurus, velxtet velfet 
(vel'vet) holoaericum, tebtue vertuw (ver-tyy), the dde (the), to- 
gether with the Latin ego egu (eg'u), Dei deei (dee'i). 

Of these the words chese^ frendes^ quene have the sound of 
(ii). It should be observed that Bullokar also gives (friindz), 
and so does Wallis, and so late as 1701 Jones admits this 
sound, thus making the new spelling ie indicate (ii) in 
" Algier, bier, canonier, friend, Aisilier, grenadier, Tangier," 
and harmonizing ,/W^rf,,/fenrf, both formerly (frec^d, feend), 
but then (friind, fiind), and now (frend, fiind). 

As respects ea Salesbury agrees with others in giving sea 
see (see) mare, yea ie {see), season seespn (seez'in) tempestas 
vel occasio, but he is peculiar in ease iea (jeez) otium, leaue 
lief (Ijeev) licentia, since Hart gives easy (ee*zi), and GKll 
writes leaf>e (leev). I can find no authority for the insertion 
of « = (j), and am inclined suspect a misprint, because the 
four words ease, leaue, sea, yea are given together and 
transcribed ies, lief, see, ie, so that the last ie may have 
occasioned the two former, and he introduces them by 
saying: ''In certain words they place a sometimes, as we 
should consider it, rather carelessly according to our custom, 
out of its own power and rather metamorphosed into the 
vowel e," this should merely imply that ea was written for 
ee, meaning prolonged e (ee), and not that in two of the 
words e was also altered into the Welsh t, meaning English y. 
If then we read ees, leef for ies, lief, in Salesbury's Welch 
transcription, we shall reconcile it with his observation and 
with the usages of other orthoepists. 

1568. Smith, agreeing generally with Salesbury, calling 
the English e *^e Latina,*' pronoimces yet, yes (sit, jie), but 
pves aUo the pronunciation (jet, jes), though by introducing 
it with an " alii vocant," he clearly prefers the former. 

1669. Hart says, describing this vowel: "The seconde 
with somewhat more closing the mouth," than for a, "thrust- 
ing softlye the inner part of the tongue to the inner and 
vpper great t^eth, (or gummes for want of teeth) and is 
marked e." He writes (dheez) for these, and (mii'terz, 
Hier) for metres, here. In 1680, Bullokar writes both (neer) 
and (niir) for here,^ and has also (siil'dum) for seldom. 

> Henry lY., part 1, act L, sc. 2, 1. apparant that thon art heire apparant,'* 
66, Quarto 1613} : "were it not heere ought to have been pronounced (wer ft 



Chap. HI. { 8. E, EB, EA — XVII TH CENTURY. 81 

1621. Gill says, "E, breois est hac formft (e), vt in (net) rete : 
et longa sic, (ee), vt in (neet) iteatb. i. nitidus adiectiumn : Sub- 
stantiunm keaie significat omne genus bouum.'' 

The pronunciation in the xvi th century is therefore toler- 
ably certain. All words now spelled with ee had (ii), 
a few final e as A^, m^, %he^ we, had also (ii), almost every 
word now written with ea, or words written with ea in the 
latter part of the century had (ee) though some had (e). 
All simple e long were (ee). Exceptions were here 
(mir) occasionally, hear^ year (niir, jiir) in BuUokar, 
appear is markea (apiir*) in Butler 1633, who also dis- 
tinguishes (teer) lacerare, (tiir) lacryma, and wishes 
dear, weary, hear to be called (deer, wee*r», Heer) instead 
of (diir, wii'ri, niir) which he therefore implies to have 
been the more usual pronunciation. 

E, EE, EA. — xvTiTH Century. 

It would be waste of time to establish that through the 
xvnth century and down to our own times short e has 
remained (e) and ee has been (ii). The difficulty only turns 
upon the pronunciation of long e and of ea. 

1653. Walus says : " « profertur sono acute claroque ut Gkl- 
lorom i masculinum," except before r as will be hereafter con- 
sidered; ^^ ea effertur nunc dierum ut 6 longum: sono ipsius a 
penitus suppresso, et sono Hterae 6 producto. Nempo iUud solum 
pnestat a ut syllaba reputetur longa. Ita met obyiam factus, meat 
victus, iet sisto, sedere facio. Mat sella, etc., non sono differunt nisi 
quod vocalis illic correpta, hie producta intelligatur." 

He however gives the exceptions near, dear, hear = (niir, 
diir, Hiir). WiBdns has (ii'vU) for evil,^ but he writes Jesm 
as (Dzhee'ses), where the first (s) is probably a mere over- 
sight for (z). 

1668. Pbice says : " E soundes like, ee, (ii, i) in he, even, evening, 
England, JEnglM, he, liere, me, she, we, ye," probably the complete 
list at that tune. He also says : **ea soundes e, d-r-a-w-n out long 
as lead, weak." And then subjoins the following list : — 

Appeal, appease. Bean, bear, beast, beat, beneath, breach, break, 

not Heer apar-ent, dhat dhou art Hair in alluding to raisins, pronounced in 

aparent), but for the sake of the joke the usoal but unrecognized manner 

we may suppose Falstaff to have pro- (reez'nz),apronunciationnTen by Price 

nonnced in Hart*B waj, and called heir 1668 as the correct sound, and, as we 

(Heer), a pronunciation certainly well see by Hart, well known at the time, 
known in Shakspere's time, altnou^h * The ags. forms pfel, eofel, point 

censured by Gill so late as 1621. Again, to the sounds (yr'vel, ee'-vel), at a very 



in the same play, act ii., sc. 4, 1. 264 : early period, and consequently to a con- 

'* If reasons were as plenty as blacke- current (ii'yl, ee'yl) in old English, 

berries," was (if reez*nz wer az pleu'tt The contracted form ill shews that the 

az blak'bera), and the joke consisted (ii) sound had the preference. 





82 E, EE, BA — XVII TH CBNTUEY. Chap. III. i 8. 

to break. Cease, cheat, clean, cleave, compleat, conceal, congeal. 
Deal, decrease, defeat, displease, dream. Eager, ean, ear, earn, 
easie, Easter, endeavour, estreat, eat, eaves. Feature, forswear. 
Glean. Heal, heap. Jealousie. Meal, mean. Beach, reveal. To 
sheath, speak, spear, spread, squeak, seam, seamstress, streak, 
surcease, swear. Teach, teazils, treatise. "Weave, weaver. Zeal." 
Of these the following are still either (ee, ee) or (e), hear, break, 
earn, endeavour, forswear, jealousie, spread, swear, while the rest have 
become (ii). "-Eb sounds short (e) in head, dead, ready. Bed- 
stead, beard. Earl. Feather. Heaven. Measure. Pearl, pleasure. 
Search, stead, sweat. Thread, threaten, treasurie, treasure. "Wealth, 
weary, weather," of which only heard, weary have now changed. 

John Hemble used to be laughed at for speaking of his 
bird, meaning heard \ we have here old authority for the 
sound.^ Price makes ea soimd as a and there is consider- 
able probability that he meant (sb) and neither (a) nor (aa), 
in heard, heart, hearken, searge. Jones said both hard and herd 
for heard (p. 86) ; serge, is borne out by the modem (klaaik, 
saaJ*dzh«nt) for clerk, sergeant The only words in which 
Price admits ea to sound as ee (ii) are dear, appear ; blear-eyed, 
chear, clear, hear, near, read, year, which short list also em- 
braces all Wallis's exceptions. 

1685. CoopEE has not named any instances in which e 
long is (ii), but he enters fiilly into ea. 

First ea = (e) in already ^ behead, bread, breadth, breakfast, 
breath, cleanse, deadly, dearth, death, dread, earth, endeavour, 
feather, head-y, health, heaven, heavy, leather, leaven, leaver\leYeT] 
leaveret [leveret], pageant, reachles [reckless], ready, realm, 
spread, stealth, threaten, treachery, tread, wealth. Here en- 
deavour has (e) instead of (ee) as in Price; breakfast is 
shortened as at present, and lever has now become (ii). 

Second ea = (ee), of which more presently. This is a 
long list beginning with appeal, appease, beacon, etc. Most 
of the words now have (ii), except break, forswear, great, 
sweat, wear. The words ean = yean, enitor, eavs = eaves, 
subgrunda, learn lampas, lease formula locationis, deserve note. 

Third ea = (ee), of which more presently. With the 
single exception of scream clamo, all the words have the com- 
bination ear, as bear, beard, earl, early, earn, earnest, learn, 
rehearse, scarce cribrum, search, shear, potsheard, swear, tear, 
wear. 

^ Sheridan, 1780, gmne a list of Irishmen, who, wishing to imitate the 

Irishisms, notes (hiird) as the Irish and English (ii) pronunciation of tfo, carried 

(hsrd) as the English pronunciation of it too far, as Sheridan points out in 

beard. Most prohahly (biird) was at some other cases, (p. 92). 
that time one of the mistakes made by 



Chap. III. § 8. E, BE, EA — XVH TH CENTURY. 83 

Fourth ^ = a, which we have identified with (sd\ (p. 71), 
in hearties^ hearten^ hearth. 

Fifth ea = (ii) in arrear^ besmear, blear^et/'d, char, ear^mg, 
fear^ gear, hear, near, sear, shears, spear, tear lacrynu, icear^f, 
whereas Price speaks toeary with (e). Here arrear, ear^mg, 
fear, gear, sear, shears, spear^ tear b., weary, are in addition to 
Price's list, which also contains words not here found. It is 
clear that the (ii) sound was beginnings to assert its ehdms 
to the domain which it has since almost entirdy conquered, 
and from which the orthography ea was intended to driye it, 
so powerless is the artificial barrier of spelling, to arrest the 
natural flow of roeech. 

Cooper's vowel system is peculiar, and is clearly founded 

upon a careful analysis of his own pronimciation. His list 

of exact pairs of long and short vowed sounds is as follows : 

12 3 4 6 6 7a 

can kffli will folly full up meet foot 

cast cane weal fall foale — need fool. 

Now there can be little doubt that the series of short 
vowels in the upper line was meant for (aB> e, i, a, u, % 
i, u), although (e, a), may have been used for (e, a). Hence 
tiie long vowels should be (aesB, ee, mV aa, uu, — , ii, uu). The 
second may of course have been (ee), and the third may 
have been (ee) rather than (ii). The two sounds are closely 
enough allied for even a careful analyzer to confuse. In 
order to bring a Frenchman to the sound of ($) it is necessary 
to exaggerate the sound into (e). Persons endeavouring to 
prolong (i) are very apt to fall into (ee). Other orthoepists 
seem to have confused Cooper's sec(md long vowel with (aBse) 
when it was iq>elt a as in cane, and with (ee) in other cases. 
It is to be remarked also that Cooper finds his second Jong 
vowel expressed by ea almost only before r. This rather 
points to (8B0Q, EE, ee) as his first three vowels, which others 
reduced to two (8B8B, ee). There is no evidence, beyond 
Cooper, for (u) occurring long, or (e) short, in English. The 
inference is that Cooper had either a peculiar pronunciation, 
or that vowel sounds appeared to him exact pairs, which do 
not so appear to us. He seems not to have been satisfied 
with the pair (ii, i), which is even now commonly adopted, 
and hence he tried to find (ii, i) in the English (need, meet), 
although he owns that in this case '' minima datur differentia 
inter oorreptionem et productionem," and indeed the differ- 
ence is rather due to the consonants than to the vowels, the 
sonant (d) having a sound of its own in addition to the glide 
from (ii). Again he strove to find a proper long vowel to 



84 £y BEy EA — XVII TH CKMTURY. Chap. III. § 3. 

{(), and, observing a difference then between weal and wear, 
corresponding to the modem difference between ail and air 
(eel, eej), he assumed that the finer sound was the real long 
^f (i), and thus paired (ee, t). Acting upon this conclusion 
I shall transcribe Cooper s vowels accordingly. He seems, 
precisely in the same way, to have heard the difference (uu, u) 
fuid refusing to consider them as pairs, endeavoured to hear 
<u) in foot as distinct from fool and full, and then, not find- 
ing the real long soimd of his (u), took (oo) in foal as its 
nearest representative. This would reduce his vowel scale to 
the following, which I shall adopt in future citations. 
12 3 4 5 6 7 8 

keen km wH hli ful ap mit fat 
kaeaest k£En w^l fiAl tool — niid fiiul 
The distinction between the words in ea which Cooper pro- 
nounces {ee), and those in ea which he pronounces (eb), 
may have been a step in the direction of change from (ee) to 
(ii) which may have oeen commencing at his tune in the long 
list of words to which he assigns (ee), although it was not 
accomplished tiU much later. 

Holder, 1669, does not make these distinctions, contenting 
himself with fate fat, seal sell, eel ill (fsecBt faet, seel sel, iil tl), 
but admits that some vowel may lie between (ad) and (ee). 
In comparing Cooper with his contemporaries we must then 
consider his [ee, ee) as represented by their single (ee). 

1688. Mieoe after laying down the rule that e long is (ee), 
the French ^ aigu, and e short is (e), the French I ouvert, 
excepts the following which have the sound of (ii, i), be, he, 
she, me, we, "qui s'^crivaient autrefois avec deux e," yes, besom, 
evil; eve, even, evening, here ; the termination -eotis', employ- 
ment, enquiry, " qui s'^crivent indifferemment avec un ^ ou 
avec un i,'* ten, linnen, penny, hence, then, thence, when, whence, 
which he transcribes in French letters " tinn lininn, peny, 
hinnce, denn, dence, hoinn, hoinnce,^' so that he gives e and 
not t in three of the words (by mistake P). This last list is 
peculiar to this author. 

Miege gives long i mascuKn, (ee), as the general pronuncia- 
tion of ea, but says that the a counts for nothing in the fol- 
lowing words, for which ea therefore = (e), beard, bread, 
breakfast, breath s., dealt, dearth, death. Earl, early, to earn, 
earnest, earth, feather, head, health, heard, hearken, hearth, 
heaven, heavy, leap, learn, leather, leaven, leaver, meadow, 
pageant, peasant, pillow-bear, potsheard, read " le Preterit et 
Participe,"r^fl</y, realm, to rehearce, scarce, search, stead, stealth, 
threaten, treachery, tread, wealth, weather; of which beard. 



Chap. III. § 3. E, EE, EA — XVn TH CENTUEY, 85 

Ie<q>f lever, pillouhbeer, have now (ii). It is observable that 
lie gives hearken to (e), and also tnat the vowel in breakfast 
was shortened at so earl^ a period. 

Miege makes ea = (ii) in these words only, besmear, blear- 
eyed, clear, dear, gear, hear, near, shears, spear, in which we 
miss some of Price's words, though the list is increased by 
besmear, gear, shears, spear. 

" Bear nn ours et pear une poire, se prononce bair, pairj^ 
There is a modem American pronunciation, probably (beesei), 
but generally heard bv Englishmen as (baaj), whien may 
date from this time, for as Miege evidently means bear to 
have a broader soimd that he heard in other words, the real 
sound may have been (bseeer). See Cooper's third list as 
noted above, (p. 82). 

1701. Jones says that the sound of e (ee) is written ea 
" in all words or syllables, that are, or may be sounded lone," 
except a certain number of words where it is written e only, 
and it is perhaps worth giving these lists as shewing many 
words in e, e-e, now mostty pronounced with (ii), which had 
all (ee) so lately as the end of the xvii th century, because 
the fact is little known, and its announcement is generally 
received with incredulity. Those marked (*) have still (ee) 
or (e). 

1) eke, *e're (ever), *e're (before), mere, rere, the, *there, these, 
♦were, ♦where; glebe, Medes, mete, nepe, scene, scheme, sphere, 
Swede, Thebe, Theme. 

2) adhere, antheme, austere, blaspheme, ♦cherub, cohere, com* 
plete, concede, ♦credit, discrete, *felo, female, ♦ferule, frequent, 
Hebrew, imp^e, negro, *nephew, obscene, *pedant, pedee, poeme, 
serous, sincere, supreme, systeme, ♦tenet, terrene, ♦treble, ♦venew; 
— *crevice, crewel, menow, *nether, ♦plevin, ♦whether. 

3) " all Scripture names and proper names from other languages, 
as Belus, Jehu, Jesus, &c." 

4) " all that begin with the sound of ee, de, e, per, pre, re, w." 

With these we must contrast the words in which e had the 
sounds (ii, i) ; 

1) the termination -ecus. 

2) initial he- as become, bedew, before, &c. 

3) the six words, be, he, me, she, we, ye, 

4) the ten words, ehesel [chisel], crete, Midland, English, here, 
mere, metre, Peter, saltpetre, Ticede. 

5) the six words, Ihan, Eve, EceUng, even, evening, evil. To 
which in another place he adds devil} 

In the following list e is said to be sounded as a, which 

^ Jones says that devil is " sounded de^U^ are curious in connection with the 
dm sometimes." This, and the Scotch derivation of iU from ml. 



86 B, IB, BA, — XVnTH CENTUEY. Chap. III. § 8. 

was most probably sliort (sb) : Berks, clerk, eleven, Herbert, 
merchant, mercy, Owen, phrentick, verdict, yeUow, etc. ; of which 
phrentick has asserted itself in the orthography .^n^tc ; mercy, 
vellaw, and sometimes verdict are known as yulgarisms; eleven, 
Herbert are now nnknown, merchant is known as an archaism, 
and Berks, clerk are yery common. This list seems to shew 
that Miege's service, bear, pear in which he makes e = at 
French, had the same sound, especially as (saai'vis) is a 
well-known yulgarinn at the pres^it day. 

The only woras in which /ones allows ea to be like a (ae) 
are heard, heart ''to distinguish them from hard (not sojfi), 
Hart (or Stag)," but he also gives heard the sound of (Herd). 

Jones makes ea short = (e), in beard, bread, breadth, breast, 
breath, cleanse, dead, dealt, dear, dearth, death, dread, earl, 
earn, earth, head, heard, hearth, lead, leap, meant, meash, pearce, 
pearl, reach, read, reath, realm, scarce, search, searge, sheard, 
shread, slead, spread, stead, stealth, sweat, thread, threat, tread, 
icealth, yearn; — bedstead, bestead, heaven, heavy, leacher, leather, 
leaven, measure, peasant, pheasant, pleasant, steady, treasure, 
weapon, weasand, weather; most of which have preserved their 
sounds, though some have changed their spellmg. 

The only words in which Jones aUows ea to have the 
sound (ii) are chear, clear, dear, ear, gear, hear, mear, near, 
year; — appear, beadle, beaw (biu) now (Jboo), instead, stead, 
steam, team, yea, yeast. 

Collecting together all the words spelled with ea and pro- 
nounced witn (ii) as given in the preceding lists, we find them 
limited to the folio wmg — ^all others in ea having (ee) or (e). 

^pear dear mear^ steam 

arrear ear near team 

beadle earwig read a tear 

besmear fear sear* weary 

blear-eyed gear shears yea' 

chear* hear spear year 

clear instead* stead* yeast* 

Those marked (*) are now spelled cheer, mere, sere ; those 
marked (^) had often the sound (e) at that time, and perhaps 
more regularly ; {^) the word yea is not marked (jii) except 
by Jones. 

This list must be borne in mind in judging of rhymes in 
the xvn th century. Li Oroker's Johnson, ed. 1848, p. 57, 
it is said respecting Rowe's couplet 

As if misfortune made the throne her seat, 
And none could be unhappy but the great, 
which Dr. Johnson in his Flan of a Dictionary in 1747 had 



Chap. IIL } 3. E, EB, BA — XVII TH CBMTURT. 87 

adduced to shew that gretd had sometimeB the sound (griit), 
that Lord Chesterfield remarked it was " Undoubtedly a bad 
rhyme, tho* found in a great poet/* — an observation which 
shewed first that Lord Chesterfield did not know the pro- 
nunciation of English when Rowe was young, and secondly 
that he was so little aware of the habits of great poets (at 
least if he reckoned Shakspere and Dryden among them) that 
he looked to their greatness as a guarantee for the perfection 
of their rhymes. STow Rowe liv^ from 1673 to 1718. We 
may therefore expect to gather his pronunciation from Cooper, 
Miege, and Jones. The first gives (s^^t, gteei\ the rules of 
the others would imply (sect, greet). The rhyme was there- 
fore perfect. While Pope's couplet, adduced by Johnson to 
shew the other sound of great. 

For Swift and him despis'd the farce of $taU 
The sober follies of the wise and great j 

would have been to Rowe a somewhat imperfect rhyme (©ae, 
ee), and one which I have but rarely found when examining 
the rhymes of this period. 

As the point has been so much disputed, the orthoepical 
accounts have been given at great length, and it will be in- 
teresting to add the result of an examination of Dryden's 
rhvmee in his Absalom and Achitophel, Annus Mirabilis, 
Palamon and Arcite, Wife of Bath, Good Parson, Theodore 
and Honoria, Religio Laici, Flower and Leaf, Cjrmon and 
Iphigenia, with respect to the pronunciation of the long e 
and ea. Rejecting those in which both spelling and sound 
were, as far as is known, identical in the rhyming termina- 
tions, the following are the results. 

1) Begular rhymes, (ee, ee) ; ease with these seize, sea with 
eaivey prey weigh key lay way sway, wear despair, reveal frail, 
leave with deceive receive, mean obscene, congeal hail, remained 
glean'd, there hair, please these, theme dream, bear heir ; 

2) Nearly regular rhymes, a long with its corresponding short 
vowel (ee, e) ; feast with breast guest addressed rest, set with great 
retreat, increase less, heat with sweat threat, beat threat, conceal 
with tel dispel, appeal rebel r., zeal dwell, please with grievances 
images, yet great, extreme stem, supreme them ; 

3) Begular rhymes (ii, ii), cheer with clear year, years ears, 
appear with year ear tear 8, steer gear cheer clear, near with clear 
ear, dear here, clear ear, career spear, fear with leer cheer near steer 
tear s. ear ; 

4) Possibly regular rhymes owing to variety of pronunciation, 
(ii, ii) ; rear with fear appear, to bear with hear year tear «. hear 
appear spear, hut also bear with heir hair fair were, and were with 
career spear appear ; where with clear near, there with spear appear 



88 E, EE, BA — XVni TH CENTURY. Chap. III. § 8. 

disappear clear fear ; for we still hear were^ where, there pronounced 
(will will dhiii) as yulgarisms ; 

5) Bare irregular rhymes (ee, ii) now hecome regular as (ii, ii) ; 
heap sweep, retreat feet, deal wheel, disease degrees (?), severe hier, 
plead freed, repeat sweet, unclean seen ; 

6) Faiilty rhymes, (e, ii) petitioners years, pensioners fears, steed 
wiih fled head, feet sweat, field heheld, kneel'd compell'd, unseen 
men, reed head, — (e, i^ contest resist, sense prince, hut civil devil, 
does not belong to this place, for the rhyme was perfect (♦, ♦) ; — 
(ee, 8B8e) wear care, tears v, spares. 

These rhymes, notwithstanding an occasional laxity which 
Dryden seems to have preferred as a relief,* serve to shew the 
general correctness of the rules laid down by the orthoepists 
on this point. 

E, EE, EA — xviiiTH Century. 

1704. The Expert Orthographist dashes at once into 
the full sounds of the xviiith century. "Tho' ee be 
reckoned among the Dipthongs," says he, " yet what differ- 
ence is there in the sound of meet to come together, and mete 
to measure, in proceed and intercede?^' Hence making the 
exceptions that there, were, where, " though they have e at 
the end, yet it serveth only to lengthen the foregoing e into 
a long," that is (ee), he gives the following 17 monosyllables 
and 26 polysyllables as having the sound (ii), which may 
be contrasted with Jones's lists, (p. 85 : Bede, Crete, ere 
even now (eei), glebe, glede a kite, here, Mede, mere, mete, 
Pede, rere now rear, scene, scheme, sphere, these, Vere ; adhere, 
apozeme, austere, blaspheme, cohere, complete, concede, concrete, 
convene, extreme which Jones spelled extream, grete "or Lord," 
impede, intercede, interfere, intervene, Nicene, obscene, portgreve, 
precede, recede, replete, revere, severe, sincere, mpercede, supreme, 

Jones gives only 18 words out of the 28, (p. 86), in which 
he and preceding orthoepists aUow ea to have the sound of (ii), 

^ Besides the faulty rhymes named resemhlance between the Towels; thus 

in the text the following have been Dryden conld not have rhymed son 

noted : (awe, aa) prepare war, — (e, with seen pain cane, or beat with coat, 

arae) possess, place, — (o, n) Uood with etc. Some eyen of the above may be re- 

eood wood, — («, a) took, flock, — ferred to peculiar or archaic pronuncia- 

(fi, oo) shook with broke spoke, poor tions, so that Dryden's rhymes are not, 

with more swore ; — (a, a) strung wrong, properly speaking, the monsters of mo- 

retum scorn, turn bom,— (a, oo) lost aem times, known as rhymes to the eye, 

with boast coast ; god abode ; — (9, 9n) as move love grove, has was gas, seat 

won mith town crown, son with crown, g][^^ W^^ flour, changed hanged, 

— (uu, an) swoon %cith drown'd sound. That keep the word of promise to our tye 

We also twice find (oon, Am^ none And break it to our ear, 

Absalom. Notwithstanding tne di- See a further examination of Dryden's 

versity there is always some point of rhymes in Chap. IX, } 3. 



Chap. HI. § 8. E, EB, EA — XVIII TH CBNTUEY. 89 

69 others Iiavinff short (e) and all the rest having long (ee) 
for ea. The orthographist only admits 4 words in which ea 
is sounded like a long, that is (ee) ; viz. bear s. and v., BweaVy 
tear v., icear ; 3 woroB in which ea "is sounded like a short," 
that is (ao), viz. hearken^ heart and its derivatives, hearth; 
but gives 95 examples of ea soimded as (e) short including 
heard ; and then no less than 255 in which '^ ea is sounded 
e^ or 6 long '' that is (ii). This last list of ea = (ii), includes 
Uie words breakf deaf, deafen^ great, indeavour, — ^but endeavour 
is in the list of ea = (e), — leassee, pear, shear, pea, yearn, in 
all of which, except shear which is often (shiij), and peam 
which is (jJn), the old lons^ (ee) is still preserved; and 
though (briik, griit) may stifi be heard from a very few, I 
have not been so fortunate as to hear (diif, indii'vJ, liisii*, 
pill, jii, jiijn). We can imagine a GiU of the period ex- 
claiming again : "Non nostras hie voces babes, sed Mopsarum 
fictitias!" It is impossible to believe that this represented 
the generally-received pronunciation of the time. 

1710. Dyche, so far as I can imderstand his notation, 
agrees with Jones, but between him and Buchanan 1766, 
were fifty years, which seem to have had a great eflfect on our 
pronunciation, in settling long a to {ee) and long e and ea to 
(ii). They were years in which there was a remarkable ten- 
dency to thinness and mea^eness of sound owing to a pre- 
dilection for the higher ungual or palatal vowels. The 
change from (ee) to (ii) was attempted to be carried much 
further than actually succeeded. Thus chair,^ steak, break, 
great were (tshiii, stiik, briik, griit), obliae was (obludzh*)^ 
and (k, g) before (aa), where the sound of (aa) really re- 
mained, were palatalised into {k, ^) as in (X^aaad, ^aajd). All 
these sounds might have been heard from elderly speakers 
some thirty years ago, and those which have remained to 
the present day, are accoimted old pronunciations. In the 
XVII th century however, they were modemizms which did 
not set through, and our present pronunciations (tsheer, 
steek, breek, greet, oblaidzh*) were older, although not all 
of them the oldest forms. In the provinces (tshiii) is stiU 
frequent, and (obliidzh*) is nearly universal in Scotland. 

1710. The anonymous instructor of the Palatines, writes 
me, he, we, she, be in German letters mi, hi, wi, schi, bi as par- 
ticular exceptions, and gives as examples of ea sounding 

^ **Whyi8 a stout man alwaTS happy P (tsheer, tshiir^ the latter being one of 

Because he a cheerfnl (chair full)." This the words which had then chimged its 

is a conundrum of that period, and could sound, notwithstanding the spelling 

not have belonged to any otiier, for in cAtferr, since altered to eheer, 

the xynth oentury, ehmr, chear were * So pronounced by Byche. 



90 B, BB, SA — XVni TH CaSlf TUKY. Chap. III. } 3. 

sometimes almcNit (bmoeilen fast) as G^man i (ii)^ the watdB 
heap, heat, cheap, dean, dear* 

176&-8. Buchanan and Franklin may be said to haye 
completely adopted the present nsage respectiiig e long 
and ea. The following are all the words in Franklin's 
examples, with his transcriptions, translated into palaeotype, 
and allowing all his inaccuracies : 

Long e, serene siriin, editione iidishans, religien rUidshan, idea oidia; 
— ea long, pleased pliiz'd, eiream striim, clear kliir, meaning miiniq, 
easieei iiziiest, Uaet Hist, increasing inkriisiq, speaker spikar, readers 
riders, to read riid, dear diir ; — greater greetar greter ; — ea shoitr 
heaven nev'n, already alreadi Alreadi, / have read red, unlearned 
anlam'd. 

An Irish gentleman, bom in 1755, told me he remembered 
the change. It is to be observed that the change is not yet 
made among the less educated class in Irehmd, and was 
probably imiyersal in Ireland when this gentleman was a 
youth. He came to England as a young man, and observed 
the custom Rowing. He distinctly remembered a youth who 
asked for (piiz) peas, being told to say (peez) " like a man." 
The thinner voice of woman has perhaps occasioned all thin- 
ness of utterance to be called effeminate. Thus Meigret says : 

" Je vou' bas* a pEiiser qElle gra^* aora Te dos ku se' vocables 
mES, tEs, SES, si nou' Fy pron(m9on8, come nou' fezons En pare mere : 
E come font je ne sey qels effeminez miNons [n = (nj)] auEq vn 
prEsqe clos resErremEnt de bouqhe : creKans a mon aufs qe la voes 
virille de Thome ne soEt point taut harmonieuze, ny aggreabl' ao' 
dames q'une la9he, fosbF b femenine. Or quant a moE ie ne 
poursuy pas icy ^ete dokett' [l «- Ij] e effeminle fa^on de parler : 
car je la Less' aoz amoureuz poursnyuant tant seulement gete 
jenerall' e comune &i9on, qi asnt son home, b qi Bt re^u' Eutre Ie* 
mieus appriz." 

Just in the same way Smith exclaims against the " mulier- 
culaB delicatiores et nonnuUi qui volunt isto mode videri loqui 
urbaniis" who use (ei) for (ai). And Dr. Gill works him- 
self up into absolute rudeness, in the following noteworthy 
passa^. After observing that the eastern English are fond 
of thinning their words, saying (fir, kiver, deans) for (feier, 
kuver, dans), fire, cover, dance, he goes on to say : 

" Urxvimfra ^ autem illam magnopere affectant wvyoaroXjot, * 

* Printed Iffx^^rtip by an error, but means " with a sweeping train,'* as a 
corrected in the errata. All palatalis- parody of the Homenc i\K€aiirtw\os, 
atlon or diminution of the lingual " if it be not rather lewd, leehfrous.** 
aperture in vowels produces this effect The allusion is evidently to «iry4, and 
of meanness, thinness of sound. the word might be translated ** wrig- 

* This is an unusual word found in gling," as a mark of affectation. 
Hes. Op. 371, which Liddell says 



Chap. HI. } 3. B, BE, BA — XVm TH CENTURY. 91 

nostrse Kop088 ' qnsB qxddem ita omnia attenuant, \t a et Of non 
aliter perhorrescere yideantor quam Appius Claudius z. sic enim 
nostiSB non emunt (Iaau) laum\ et (kaambiik) camhrte, sindonis 
species; sed (leen) et (keembrik); nee edunt (kaapn) capon caponem, 
Bed (keepn) et fer^ (kiipn) ; nee unquam f butsherz meet) butchbbs 
iCEAT camem k lanijs, s^ (bttsherz miit). Et quum sunt omnes 
(dzbmtlfmfh) non (dzhentlwimen') genilnoommy i.e. matronsB no- 
biles, nee maids anciUas yocant (maidz) sed (meedz). Quod autem 
dixi de a, recanto ; nam si quando 6 gravistrepum audiretur, locum 
concedunt ipsi a, sic enim ^diquoties ad me pippiunt^ (ai pre la gii 
Jar skalerz liiv ta plee) pro (oi prai jou * g»V juut skolare leey tu 
plai), / pray you give your acholara leave to play. Quaeso concede 
tuis discipulis veniam ludendi." 

We cannot but regret that Dr. Gill had not greatly ex- 
tended his list. (Leen) does not seem to haye suryiyed, but 
(keCTQ-brtk) is now the recognized pronunciation, though I 
haye heard (kaam'bnk). So with (ke^'p'n). This anticipa- 
tion of the change from (aa) to (ee), which was not fully ac- 
complished till nearly a century after GHll's time, is remark- 
able. It must, however, be considered as a xvii th and not a 
XVI th century sound. (Bttsher, meeds, plee) will be con- 
sidered herearfcer. Here we are principally interested in the 
anticipations (miit, liiv) for, (meet, leev), meat, leave, which 
are not named as exceptions by any professedly xvii th cen- 
tury writers, and (meet, leev) being then the rule, would 
have sounded most probably as affected to Price, Cooper, and 
Jones as they did to GKll. 

Generally with regard to the change of (ee) into (ii) it is 
observable that in Modem Greek (as has been probably the 
custom for nearly 2000 years), rf is pronounced (ii), while 
there seems reason to suppose that it was originally (ee) or 
perhaps {ee), although, at least in one word, it was confounded 
with (ii) at an early period.* Also in the passage fit)m Latin 
to the modem Bomance language, (ee) fell not unfrequently 

* It would be difficult to find any present day, ignorant as we are of the 
authority for this piece of Latin. The effect that onr pronunciation would 
English is mopseyi, sluts, which may haye produced on our ancestors. 

be related to mop^ mope, ^ Probably an inaccuracy for (ju). 

* llie pronunciation is an exact < The old quotation d 8' ^X/9ioySov<p 
pal»tot;^ic reproduction of 6ill*s, and wp6fiarov firj firj \4yi»v $aSt(€t, does 
the orunary spelling in italics is my not absolutely establish Ue) or eyen 
addition throuniout. (be) as the sound. The latter is (ax 

* Both words require to be written more bleating, and Schmeller calls it 
with ('I), or else to haye (,) inserted that yowel which an^ lamb can teach 
after (1), as (dzhtntl,imen, azhentl,w»- us, " tiber den uns jedes Lammchen 
men,) to ayoid a pronunciation in three belehren kann." The well-known pas- 
tyUaoles. sage in Plato, Crat. c. 15, otop, ol /tiv 

^ This pipping, chirpiuff effect is &f^cu^aroi lfi4pap r^y hf^pop iKdlKovy^ 
precisely that now produced upon our only shews that some old people pro- 
ears by the flunkey (Dzhiunz; of the nouucedthatparticularwordintnatway. 



92 



By EE^ EA — XVni TH CENTURY. Chap. HI. i 3. 



into (ii),^ and as the Latin me, te, ae became the Italian mi, 
it, at, 80 the English pronouns he, she, me, toe, thee, as some of 
the conmionest words, were the first which fell into (nii, shii, 
mii, dhii), having remained as (hoOi shoe, mee, dhee) to the 
close of the xiv tn century. 

1710. Sheridan's usa^e agrees with the modem, but his 
observations on educated Irish usage are .important. He 
says that ee^ ie were pronounced as (ii) both in England and in 
Ireland, but that ea, ei, e when sounded with (ii) in England 
''almost universally" received the sound of (ee) in Ireland, 
as (tee, see, pleez) tea, sea, please. But he adds that " gentle- 
men of Iremnd, after sometime of residence in London, are 
apt to Ml into the general rule, and pronounce these words" 
great, a pear, a hear, to hear, forhear, swear, to tear, tcear, 
which were exceptionally pronounced with (ee) in England, 
" as if spelled greet, heer, sweer,*' that is, as (griit, piir, biir, 
swiir, tiir, wiir). Omitting these mistakes, which had nothing 
to do with ihe true Irish habits of the time, we see that the 
latter really belonged to the xvii th century. Again Sheridan 
says : '* the final mute e makes the preceding e in the same 
syllable, when accented, have the sound of (ii) as in the words 
supreme, sincere, replete. This rule is almost imiversally 
broken through by the Irish, who pronounce such words as 
if written suprame, sinsare, replate" that is with (ee) as in the 
XVII th century. In Sheridan's list of miscellaneous words 
with Irish pronunciations, we find several examples of forcing 
a rule too far, as above stated (see also p. 76). The complete 
list is as follows, to which I have axmexed my own pro- 
nunciation in the present century : — 



Writtm. 


Irish. 


En^liah 1780. 


2^/mA, 1868. 


cheerful 


tflhiirful 


tflhET-ful 


tshiiifwl 


fearful 


fiir-fiil 


&rful 


fiiifwl 


beard 


biird 


bKrd 


biud 


leisure 


lEzh-ar 


lii*zhar 


lezh-i 


search 


seertsh 


sErtsh 


sitsh 


tenure 


tEUjar 


tiiiuar 


ten'iui 


tenable 


tEn-aebl 


tii-naBbl 


t^n-Bb'l 



^ Dies, Gram, der rom. Sprachen, 
2nd ed., vol. i., jp. 139, giyes as ex- 
amples, Italian CornigUa {Cornelia,) 
Messina (Messene), sarracino (sara- 
cenus) — ^to which the initial di-y n- 
andsereral others may be added. — Span, 
consigo fsecom), yenino (venenom) ; 
port siso (sensns sesus). — FVoy. berbitz 
(Tenrecem), ponzf (pnllicenns), razim 
(racemus), sarracl. — French, brebis, 



cire (cera), marqnis (marchensis), merci 
(meixedem), pns (prensns), poussin. 
raisin, tapis (tapetum), venin ; old 
French, pais (pagense, now pays), seine 
(sagena), seri (serenns). He also re- 
marks on the same tendency in the old 
high German fira (feriae), ptna (Ital. 
pena), spisa (spesa), which hare onder- 
ffone another change in modem times, 
Deooming Finer, Pmpm, Speisi. 



Chap. HI. } 3. 


0, 00, OA 


— XVITH CENTUEY. 


Writim, 


Iriah. 


JBitffUih, 1780. 


JBftffiiah, 1868. 


wherefore 


whiirfoor 


whBr-foor 


wheei-fooi 


therefore 


dhiir'foor 


dhKT'foOT 


dheei'fooi 


breadth 


brEth 


brEdth 


bredth 


endeavour 


endii'var 


endBvar 


endevT 


miAchieYOUS 


mistshii'Tds 


mis-tshivas 


mis-tshivBa 


reach 


retsh 


riitsh 


riitsh 


zealous 


zii'bs 


Z£l*88 


zel-BS 


zealot 


zii'lAt 


Z£l'9t 


zel'Bt 



93 



0, 00, OA — XVI TH CENTURY. 

1530. Palsgbave says: "0 in the frenche tong hath two diners 
manors of soundynges, the soundyng of o, whiche is most generall 
with them, is lyke as we sounde o in these words in our tonge a 
boore, a soore, a coore, and suche lyke, that is to say, like as the 
Italians sounde o, or they with vs that sounde the latin tong aright." 

1567. Salesbuby says : " in Welsh is sounded according to the 
right sounding of it in Latin : eyther else as the sound of o is in 
these Englyshe wordes : a Doe, a Eoe, a Toe : and o never soundeth 
in Welsh as it doth in these wordes of Englysh : to, do, two.^^ And 
again, 1547, speaking of English, he says : **0 takes the sound of 
[Welsh] (o) in some words, and in others the sound of to (mi) ; 
thus TO, to, (too\ digitus pedis ; so, so, (soo), sic ; two, tw, (tuu) 
duo ; TO, tw (tu) ad ; schole, ecwl, (skuul) schola .... But two oo 
together are sounded like to in Welsh, as good gwd (guud) bonus ; 
pooKE pwr (puur) pauper." 

1568. — Snt T. Smith simply says : " Latina," giving as ex- 
amples the following words, which he only writes phonetically, but 
are here given in ordinary spelling — 

Short — smock, horse, hop, sop, not, rob, hot, pop. 
Long — smoke, hoarse, hope, soap, note, robe, boat, pope. 
Smith makes oo in hoot, look, mood, fool, pool, too the same as the 
Latin u long, meaning (uu). See under U. 

1569. Hart says : " The fourth [vowel], by taking awaye of all 
the tongue, cleane from the t6eth or gummes, as is sayde for the a, 
and turning the lippes rounde as a ring, and thrusting forth of a 
sounding breath, which roundnesse to signifie the shape of the 
letter, was made (of the first inuentor) in like sort, thus o." And 
his English examples are no, not, so. 

1580. BuLLOEAB says: **0 hath thr^e soundes, and all of them 
vowels ; the one sound agreeing to his olde and continued name, 
another sound, betw6ene the accustomed name of, o, and the old 
name of, v, and the same sound long, for which they write oo,* (as I 
do also, but giuing it a proper name, according to the sound thereof), 
the thirde sounde is as, v, flat and short, that is to say, as this 
Billable ou, short sounded : for which some of the better learned did 
many times use, oo, &, v, according to their sounds, but most times 

^ The two o's are united in one type as the o and are in the type ce. 



94 Oy 00, QA — XVI TH CENTURY. Chap. III. § 8. 

with superfluous letters." He illustrates the l^iree sounds by the 
words. 

1^ Sonne Alius, vpony hoiome (flirst vowel), eome, eloae, 

2) Sonne sol, out, hosome (second vowel), come. 

3) loked, toke, hoke, eone. 

1611. Flohio says, speaking of the Italian (wh, o) : "So likewise 
to the close 0, I have throughout my book given this oualle forme 
0, and to the open this round form 0. The fiirst close or oualle is 
euer pronounced as the English single Y. in these wordes, Bun, Bug, 
Flud, Gud, Rud, Stud, Tun, &c., whereas the other round or open 
is euer pronounced as our 0. in these words Bone, Dog, Flow, Gfod, 
Eod, Stone, Tone &c. as for example in these Italian wordes, lo 
hondro il mio Bio c6n dgni diuotidne, where euer, 0. is close and 
ouaUe. And in these, lui mi vuole torre la mia tdrre ; or else, lui 
mi hk rdsa la mia rosa ; where Torre with an open or round 0. is 
a verbe and signifleth to take, and t<5rre with a close or oualle 0. 
is a noune substantiue, and signifleth a tower ; and B<5sa with an 
oualle and close 0. is a participle of the verb Bddere, and signifleth 
Gnawne or Nibled, and Eosa with a round or open 0. is a noune 
substantiue, and signifleth the floure that we call a Rose." 

1621. Gill gives as key words for his long and short o, "coale, to 
coll," and calls them a>, o. 

In endeavouring to discoyer what are the sounds intended, 
it is necessary first to examine what sounds of o exist. They 
are all round vowels, that is, the action of the lips with a 
tolerably round opening is necessary. The tongue must also 
not be much raised, or the sound flsils into (u, u) or at least 
(tih) the Italian o chimo. At the same time the tongue must 
not be too much depressed, or the sounds become (a, o), the last 
of which is the modem English o in odd, which Mr. M. Bell 
considers to be a wide form of (a), and which is generally, 
though inaccurately, confounded with (a), just as (t) is 
usuaUy confounded with (i). Hence we obtain two forms, by 
raising the hack of the tongue to a mid position, and round- 
ing the lips in a medium manner, namely (o, o), the latter 
bemg the wide of the former. In present English {o) only 
occurs as a Ion? vowel, and in the south it usually has a 
faint sound of (u) after it, thus (Hooum, Hoo'trm) homey but 
this is unhistorical, except where a tr is written; thus we 
may distinguish no, know as (noo, noou). The other sound 
(oo) is often heard long in provincial English as (noom) home. 
XJnaccustomed ears then confound it with (aa) or (oo). The 
long sound (oo) is also sometimes heard from those London 
spe^ers who wish to prolong the sound of o in dog^ cross, 
off, office, without degenerating into (dAAg, krAAs, AAf, XAf'ts), 
or being even so brcxad as (doog, kroos, oof, oof-ts). It is also 
the sound now most esteemed in oar, gUyry, story, memorial, 



Chap. IIL § 8- O, 00, OA — XVITH CaBNTUBT. 96 

onoe called, and still so called by elderly people, (oo'j, gloo'j-ri, 
stoo'irt, memoo'i'TTBl), but now professedly called (ooi, glooiTi, 
stooi'rt, memooJ'ml), the action of the glide from (oo) to 
(j) haviQg resulted in widening the Towel.* Mr. M. BeU 
recognizes two other sounds (oh, oh) related to (o, o) by being 
mixed instead of back vowels. The former he hears in the 
French hamme, where I hear (o), and the latter in the 
American aione^ where I hear (o). The soimds are unusual 
to English ears, and it will be imnecessary to distinguish 
(p, oh) or (o, oh) for any purpose in this treatise. Generally 
(ston) is heard as (stan), which is the modem English form 
in such phrases as to weigh twelve stone (tu weei twelv stan). 
The sound (hoI) for (hooI) whok^ is by no means unconmion, 
although most persons hear it as (hoI), and it is imitated by 
writing " the hull of a thing." 

Now long being {oo) and short o in closed syllables beinfi; 
(o), as note, not (noot, not), English writers have got so much 
into the habit of considering mese two sounds as a pair, that 
when they speak of long and short o we naturally expect 
these soimds and not (oo, o). This creates the difficulty. 
The ear and judgment are confused. Sir T. Smith may 
haye pronounced his key words (smok smook, Hors Hoors, 
Hop, Hoop), and yet haye considered them as pairs, for he 
actually has so considered the more distant sounds (beit, btt). 
As the Welsh at the present day, so far as I haye obseryed, 
say (oo, o) and do not use either (o) or (o), they probably so 
pronounced in Salesbury's time. But Salesbury would in 
that case have heard {po^ o) as (oo, o), so that his identifi- 
cation of the English with the Welsh o, although probably 
correct, would not suffice to decide so delicate a point. Quite 
recently I haye heard Welsh gentlemen who seemed to me 
to say (poob) and not (poob) declare that the yowel sounded 
to them the same as that in my pronunciation of robe (roob). 
Hart's description, giving the lingual positions for a (a) and 
the rounding of the lips should produce (o) exactly. And I 
am inclined to think that the normal English sound up to 
the end of the xvith century was (oo, o), both long and 
short. This would make sense of Hart's examples no, not, so 
as (noo, not, soo), and would make l^mith's and Gill's long 
and short o, perfect pairs, thus : Gill oo//, coaly (kol, kool) ; 
Smith smock, smoke, (smok, smook). 

1 Of course this somid degenerates anxious to correct this, say (gloo,Tt\ 

into {9s!) or (aa), so that (gUA'rt) or without any (j), the effect ol whicn 

even (dUArt) may often be heara in was decidedly unpleasant. 
London. I haye heard olergymen, who, 



96 O, 00, OA — XVI TH CENTUEY. Chap. HI. § 3. 

My own impression, after considerable thought on the sub- 
ject, though it would be difficult to enumerate all the reasons 
which have led me to this conclusion, is, that (oo, o) must be 
considered as the normal sound, intermediate to (a) and (u) ; 
and that (o, u) are felt as approximations towards (u), and 
(o, a) as approximations towards (a). To me the Italian 
sounds chimo and o aperto, close and open o, are respectively 
(wh, o), the former coming from Latin u, the latter from Latin 
0. The regular short German and French o I also consider to 
be (o). To shew however the ease with which sounds so 
near may be con£used, I may mention that Mr. MelviUe Bell 
in taking down sounds from my dictation, heard my (o, on) 
as (oh, un).^ 

I shall assume as at least most likely that (oo, o) was the 
original sound of long and short o previous to the xvith 
century, but that (oo) inclining often towards (u) had 
become (uu) in many words in the xvith centurv, other 
words retaining the pure (oo).* It was, I beheve, to 
separate these two effects that a diversity of spelling 
was introduced. The o which became (uu) was written 
00, and the o which remained unchanged became oa. The 
change was precisely similar to the introduction of the two 
fipelmigs ee, ea at the same period, and the device was 
the same, viz., the more guttural sounds of each, that is, the 
sounds more nearly approaching to a, were represented by 
adding on a as ^, oa, and the other sounds further from a, 
were represented by simple duplication as ee, oo. When o 
had changed to (u) the spelling u gradually prevailed, but 
sometimes simple o and sometimes oo was employed. The 
older spelling ou also occasionally remained. We have seen 
that the orthography ee, ea was not fixed in Palsgrave's time. 
Similarly we find him writing in the passage first quoted 
under this letter, (p. 93), boore, soore, coore for boar, sore, core. 
Reverting to Palsgrave'^s vocabulary of nouns, we find the fol- 
lowing spellings, to which I add Levins's, as under EA (p. 77) : 

* * JBoke ,. hooke, hoke oth^..oihe, hokeram^ hochette for a well., 
bucket, hokyll .. buckle, hocler for defence .. bockler, hone a request .. 

^ See Viiible Speech, Plate viii. con- and that nyen by Mr. M. Bell, must 
taining the speech of Portia on Mercy, generaUj be attributed to farther in- 
written in VUible Speech letters from yesti^ition on my part, 
my dictation, where (noht, drohpeth) * In the examination of Chaucer's 
are written for what 1 intended to pro- pronunciation I shall endeayour to 
nounce as (not, drop*eth.) This speech shew that in his time the sound of o 
will be found as an example in Chap, had not split into two, although I think 
VIII, § 8, Ex. 1. The differences be- that o was written not unfirequently for 
tween tiie pronunciation there exhibited an original (u). 



Chip. HI. i 3. O, 00, OA — XVITH CBNTUKY. 97 

boone, l^ttra^e berbe, hoore beest .. bore, boorde for bnylding .. boord, 
horde doth nappe .. borde, hoorder that gothe to borde .. border, 
hoeter uantevr, hotfe to rowe in httteav .. bote, boty that man of wane 
take .. booty, hoilar.. butler, hottrae .. bnttresse, heUrye .. bntterie, 
hoete of lether .. boote, heeihe .. boothe, huUyen in a woman's girdle, 
haitie of clothes, ehke a garment .. cloke, coke that seUeth meate .. 
oooke, cole^ of fyre .. cole, coupe [coop], core of frate .. core, corBe a 
deed body .. corse, courser of horses .. course, cosyn kynsman .. 
cousin, costee charge .. coste, coet of a countre .. coaste, cote a byrde .. 
coote, cote for a ladde .. cote, ewer .. cover, couple .. couple, course .. 
course,^ doo a beest .. doe, dokelyng .. duckling,' dole .. doole, dome 
jugement .. 4oome, dong hyll .. dungil, dore a gate., door, doublet^ 
dove .. doove, doute .. doubte, fole .. foole, foole a colte .. fole, foome 
.. feme, foo .. foe, for owe .. furrowe, fete .. foote, foMe for shepe .. 
fould, fouU .. foule, good., good, golde a metall .. golden, goulfe of 
oome, so moche as may lye bytwene two postes, otherwyse a baye .. 
golfe, gode for a carter .. gode, goore of a smock .. gore, goee a foule 
.. goose, goseherrg .. gooseberrie, goost .. ghoste, gate a beest .. gote, 
gotUemglk, grome .. groome, grote money .. grote, hode .. hoode, hoke .. 
hooke, ^^..hole, Ao/y.. holy, Atmy ..honye, honny combe, honny^ 
euekell .. honysuckle, hare .. whore, hope .. hope, Ao^ house .. bote, 
horse a beest .. horse, hoorsnesse of the throte .. horse, host of men .. 
hoste, hose for ones legges .. hose, houpe [hoop], ttmiltf^.. junkets, 
iotiM..juce, ^<i!^.. lode [load], ^<^ of bredde .. lofe, ^^..looke, 
lope [loop], lome [loam], losyng perdition .. lose, love .. loved, mole 
moule a beest .. moule, moleyne an herbe, molet a fysshe .. mullet, 
moone a planet .. moone, moneth .. month, mode in a verbe .. moode, 
wHfre a fen .. moore, mote a dytche .. mote, mote in the sonne .. mote, 
wtoton [mutton], mou^y^tM^^ .. multitude, moulde a form., mould, 
mcumyng .. moume, noone mydday.. noone, ncntte a relvgious wo- 
man .. nunne, norisshyng .. nourish, nose [in the body of his work 
constantly written noose] .. nose, ore of a bote .. ore,' ote come .. otes, 
othe sweryng .. othe, oulde mayde .. ould, plome a frute .. ploume, 
podyng .. pudding, poddeU a slough .. puddel, poke or bagge .. poke, 
pocke or blayne .. x>ocke, pole a staffe .. pole, pompe [pump], ponde .. 
ponde, pore .. pore, poors [poor], profe .. proofe, prose, rho bucke a 
beest .. roe buck, r(S>e .. robe, roche a fysshe .. rochet, rode a crosse .. 
roode, ro/»..roofe, roil* .. rooke, roptf .. rope, ro«tf..rose, ro^ of a 
tree .. roote, sloo worme ..*sloe, smoke .. smooke, sokelyng .. souke, 
sole a fysshe .. sole, sole of a fote .. sole, sole of a shoo .. sole, somme 
[sum], Sonne .. sonne,^ sope to wasshe with .. sope, soper .. supper, sore 
a woimd .. sore, sote of a chymney .. sooty, sothenesse [soothness], 
sodayne [sudden] .. sodayne, so%de [soul] .. soule, souldier .. soldionrie, 
eouter sauetier, soveraynte of a Kynge .. soveraygne, spoke of a 
wheel .. spoke, stoble .. stubbil, stone .. stone, store .. store, tode [toad] 
.. tode, too of ones fote .. toe, toost of breed .. toste, tothe dent .. 

^ The adjectiye eoarMS is also spelled > LeriiiB uses oor^ for a metaUie 

ceurte boUibr PalBgrare and Leniis. ore. 

* The TerD to duek is spelled douh « Both Palsgrare and Lerins use 

both by Palsgrare and Levins. mime for botii ton and nm. 

7 



98 



O, 00, OA XVI TH CENTUEY. 



Chap. Ul. § 3. 



toothe, votU under tlie ground .. yalte, wode fwoad] .. woodwasse, 
wodwosse, wood or tre that is fallen .. wood, toodde to bnme .. wood, 
icoodnssse rage .. woode, wolfe .. wolfish, woman .. woman, wombe, 
wonders .. wonder, wo sorowe .. woe." 

It is evident that long o and oo were not yet separated by 
Palsgrave to whom also the device of oa or oe final, (see doo^ 
foo, wo) had not yet occurred, and although oo was freely 
used by Levins, oa was almost unknown to mm. 

A comparison of Bullokar's notation of the three classes of 
words he cites, leads me to the conclusion that their sounds 
were, in palaeotype — 

1) son, upon, boz'um, koom, kloos. 

2) sun, ut, boz'um, kum. 

3) luuked, tuuk, buuk, suun. 

The pronunciation (son) is however peculiar. Smith gives 
(sun). Where direct authority cannot be obtained it is ex- 
tremely difficult to distinguish which of these sounds should be 
given to in any words of the xvi th century. Generally we 
may conclude that the o, oa, — ^not the ow, — which is now {oo) 
or (oou) was then (oo), bein^ the old sound but very slightly 
altered ; what is now (uu) it is not so safe to conclude was then 
(uu) unless in the course of the century we find the spelling 
00 adopted. What is now (o) was pretty certainly (o) at that 
time, being almost the old sound preserved. But it is not 
quite so certain that what is now (a) was formerly (u), for 
some of these may have been (o), or both sounds may have 
prevailed, thus Bullokar and Smith difiPer respecting son, and 
none, one were (noon, oon). It is also very probable that 
many o represented (u) even as early as Chaucer s time. The 
following cases of o, oo, oa = (u) or (uu) are taken from the 
authorities for this century. 



above 


cook 


hood 


ooze 


some 


two 


wood 


afford 


cool 


hoof 


other 


soon 


whom 


woof 


among 


coot 


hoop 


pool 


soothe 


whoop 


wool 


blood 


cover 


loof 


poor 


stood 


whore 


Worcester 


board 


do 


look 


prove 


stool 


wolf 


word 


bombast 


done 


loose 


rook 


sword 


womb 


work 


book 


food 


loving 


room 


thorough woman 


worm 


boot 


foot 


mood 


root 


to 


won 


worship 


brood 


forth 


mother 


shoe 


ton 


wonder worst 


broom 


good 


mouth 


shovel 


too 


wont 


worth 


come 


goose 


move 


smother 


took 


woo 


wost 



conjurer government 

To these Shakspere authorises the addition of Rome} 

I Julius Csesar act i. sc. 2, v. 166 : — 

Now is it Rome indeed, and Roome enough 
When there is in it but one onely man. 



Chap. III. § 3. o, 00, OA — XVnTH CENTURY. 99 

The following are all the words containing o which Salet- 
burj adduces, leaving ou, ow, ai, ol to he considered hereafter. 

QoT) Ood (Gk>d); coNDicroN eondisyum (kondis'nm) ; ettebmobe 
efermww (eY'ermoor);* Tnoin)BB thumdr (thun-der), wondbb wnder 
(wan'der);' hope hoop (noop); obakoes oreintsys (or'aindzhw), pole 
jjiwl (fdul) ; HOLT holi (hoo'Ii, Hol'i),' hoitest anest (on-est); hokouse 
onor (on'or) ; EXHiBmoN ecsibmum (eksibis-i,im) ; peohtbition pro- 
ihisiwn (proo,ibi8*i,im) ; John tsionf $um (Dzhon) ; boke hwk (bunk) ; 
TO, to (too) meaning a toe; so so (soo) ; two tw (tuu), to to (tn) the 
preposition ; schole aewl (skniil) ; good, gwd (gaud) ; poobe pwr 
(puur) ; bos roe (rooz) a rosej season seesyn (seez'tn);^ top top (top) ; 
Thoxas tomas (Tom'as) ; thbone tncn (truun) ; oxe ocs (oks). 

Florio (p. 94,) evidently heard bone, dog as (boon, dog), 
and, if Q^an) had been said, he would have most probably- 
heard that sound as (In^t^hn), just as at present Englishmen 
confuse the Italian (mUtiL, o), o chiuso long and o aperto short, 
with their own {oo, o). Hence his remarks give a presump- 
tion in favour of (oo, o). 

O, 00, OA — XVII TH Century. 

1653. Wallis says of the guttural vowels "dd" aperta: Si 
apertudL majori seu pleno rictu spiritus exeat, formatur Germanorum 
d vel d* apertum. Neque Germani soliiiii sed et Galli, aliique non 
pauci, eodem sono suum a plenunque proferunt. Angli sonum 
ilium correptum per 6 breue ; productum ver6 plerumque per au 
vel aw^ rarius per d exprimunt. ITam in fdll, foUy ; hdU^ haul, 
hoUy; cdU, collar; lawesy loaae; cause, cost; aw*d, odd; sawd, sod; 
aliisque similibus ; idem prorsus Yocalium sonus auditur in primis 
syllabis, nisi qubd illic producatur his corripiatur. Atque hinc est 
quod Hebrsei suum camets longum, et camsts breve seu camets chatuph, 
(hoc est, nostrum d apertum et 6 breve,) eodem charactere scribunt. 

Kam eorum 7^ et 7^ non alitor differunt quam nostrum cdU et eoU. 
" 6 rotundum, Majori labiorum apertura formatur 6 rotundum ; 
quo sono plerique proferunt Greecorum a>. Hoc sono Gralli plerum- 
que proferunt suum au, Angli ita fere semper proferunt pn)i-» 
ductum vel ctiam oa (ipso a nimirum nunc dierum quasi evauescente; 
de quo idem hie judiciiun ferendum est ac supr^ de ea^) : Ut, one, 

> The inserted w is perplexing, it words were meant. This shews that the 

ihonld give the sound (mnor), and quality of the long and riiort was the 

Price nses wo to indicate (uu). But same to him. 

Smith pronounces (moor). * The origin of tiiis y b not appa* 

* The initial (w) has been supplied, rent. The real sound of the word 
because its omission has been regarded seems to have been (seezni). 

as a Welsh habit, and Salesbury's mode ^ The Oxford reprint has 6 in each 

of writing did not gire him the means case, which is erroneous. 

of representing (wii). • We haye seen that the a was never 

* Salesburj does not distinguish pronounced in either case ; that it was 
koiljf, hoty either in sound or speUin^, a mere orthographical device. 

but his interpretation shews tluit both 



100 O, 00, OA — XVnTH CENTURY. Chap. HI. § 3. 

muis ; none^ nnlliis ; whoh, totus ; hole foramen ; eoaly carbo ; hoot, 
cymbBL ; oai, avena ; those, illi ; ehose, eligi ; etc. At nbi o breve est, 
at plarimum per 6 apertnm (de quo supra) rarius per 6 rotundum 
prouunciatur. 

'* Oo sonatur ut Germanorum H pingue, seu Galloram ou, TJt in 
Tocibus good bonus, stood stabam, root radix, foot pes, hose lazus, 
loose laxo, amitto. 

'' ^onnunquam o & ou negligentius pronunciantes eodem sono" 
& {L obscure » (a), ^'efferunt, ut in cdmSf venio ; sdme, aliquis ; ddne, 
actum; edmpany, consortium; country, ros; couple, par; cdvet^ 
ccHicupisco ; Idvs, amo ; aliisque aliquot ; quae alio tamen sono rectius 
proferri debent." 

These extracts seem to make long o a true labial {oo),^ 
short a true gutturo-labial (a) — ^for which however the softer 
(o) may have been really sounded, and occasionally (e), a 
new sound, which will be considered under TJ, — and long or 
short 00 the true (uu, u), which however may have been 
{vM, u). Hence long and short o had ceased to be a pair 
(oo, o), and had become the diflferent vowebi (oo, o) or {oo, a). 
This fiilly agrees with Wilkins, 1668, who gives the follow- 
ing pairs, leaving {po) without a mate, 

^ ( short ho\rtom fol-fy fot mot Pol rod 

^011^ bought fiEdl fought Paule Bawd 

short 

long bote foale vote mote pole rode 

(short foil fut pul 

{long boote foole foote moote poole roode 

but he also gives amongst as containing (oa). 

1668. Price distinguishes three sounds of o, long as in no, 
"/o,** more, most = {oo) according to Wallis ; short as in lot^ 
not, for = (o) ; " obscure like short u (e) as in son, tongue^ 
London, above, ^approve, *behoveth, brother, come, companie, 
eonie, conduit, dosen, dost, doth, love, mother, *move^ plover, 
pomel, *prove, *remove, shovel, some, venom^ *«7Aom," all of which 
with the exception of those marked * retain the sound of (o).' 

Price also says : "o after tc, soundes like short u, (a) as 
worldy *8word, *woman, toon, except, o, soundes, ee, in women, 
and long in wo, wore, woke," (swaid, wom'un) are uncom- 
mon. Then follows a long list of final om, on sounded as 
as (am, en), including some words in which the sound is 
now ('n). 

1 The French distmffoish two soTuidi of Cambridge, tiiat he used to saj: 

of 0, the close au and the open o, which " If a man say I lie, I say (proT) it ; 

to my ears sonnd as (o, o). if he (prev) it. then I lie ; if he 

* As regards prove, it is an ancient don't (prey) it, then he lies, and there's 

uniTondty story of the late Pro! Yince, an end on't" 



Chap. m. } 8. O, 00, OA — XVn TH CENTURY. ' ..-. 101 

" 0, soiindes like (woo)^ oo in ^JRome^ do, shoe, cucko^'/^g0^ 
*hord, mushrom, undo, who, *whore.*^ (Ruum) we have tb^ 
was heard in Shakspere's tuae, and may still occasionally be 
heard ; (gnu) is mentioned by Wallis in terms of disapproval ;.' 
(Huurd) may be classed with (afuurd) afford ; and mushroom ' 
has changed its spelling. 

Price makes oa the lone^ o, (oo), and oo generally ''like 
fffoo** (uu), but "like w" (o) in good, wool, ho^, wood, stood. 

1685. Cooper pairs the vowels fall folly, and foal fulL 
By the latter p^ he could not have meant (ftiul ful), or 
(ftiuhl fuhl). His (ful, fwl, ftthl, fol) whichever way he 
pronounced it, contained the nearest vowel sound to (fool) 
that he was acquainted with (p. 84). He says : — 

*' fonnatur 2i labiis paululil^in contractis, dam spiritus orhumlatus 
emittor : ut in hope spes ; productum semper, (nisi in paucis qusB 
per 00 (uu) sonantur ; et ante / per ou (fm, auS labiales : ut in told 
audax) hoc modo pronunciant Angli, quem auqnando scribunt per 
oa ; ut coach currus ; correptus rarb auditor, nisi in paucis, quee k 
consonante labiali incipiunt ; ut post w in foo(f lupus, wonder nurum; 
& in syllaba war ; plura non memini : in quibus^Eim u hoc modo 
pronunciatur, ubi prsecedens vocalis est labialis ; ut pull, veUo, full 
plenus ; non quia debet, sed quoniam aliter fiEiciliiis efferri nequit : 
Et 00 in ffood bonus, hood cucullus, wood lignum ; / stood steti ; 
Galli per o ut ^lohe globus, proteste protestor; in copy exemplar 
corripitur. Oermani per o, ut ostem pentecoste ; quem in prinapio 
dictionum fer^ producant : in toor^ verbum ; Gott Deus corripitur." 

Whence it appears that Cooper did not distinguish (u) 
from (o) or even (o). In fact he hardly knew the true short 
(u) for after describing oo he says '' inter sonum correptum 
& productum minima datur differentia/' and he pairs foot 
short, fool long, where the difference of length is solely due 
to the following consonant. As I have found it necessary to 
suppose that Cooper paired {ee, %), see p. 83, so here I pre- 
sume he paired {oo, ti), sounds which have nearly the same 
degree of diversity. This occasions a slight difficulty in his 
diphthong om, which will have to be afterwards considered. 

Cooper gives the following list of words in o, oa which 
have the sound of (uu), those marked * being unusual: 
^aboard, *afford, ^behoves, *boar,^ *bom carried, *force, *forc€S, 
move, *sword, *swom, tomb, two, who, whom, whore, whosoever, 
womb, *worn. The words *board, *forth, prove, stoup he says 
are also written boord, foorth, proov, stoop. In the following 
words he hears his short o ==(t«) ; bloodily, good-ly-ness, flood, 

> Prioe^s own notation, not palaeo- * This is hooTy the animal, not hoar 

type. Ab a Welshman he evidently ^boor as given afterwards by Jonm. 
called 1^00 (an), the same as pa. 



lO!^*-.^ * ^9 ^ ^^ — XVnTH CENTURY. Chap. IIL § 3. 

Juiod^ brotherhood, sisterhood, neighbourhood, falsehood, soot, 

'stodd, wood, wool. The exceptions damosel, women (daem'zel, 

^ wtm'en) are notecL After givinff examples of oa as (oo), 

'•*wliich are often written with o-^, ne says, as cloak, cloke, he 

admits the sound of (aa), as now usual, in abroad, broad, groat. 

1686. MiEGB agrees in the main with the former, but he 
hears long o as French o (oo), and the short o when it was 
(a) as the French short o also, that is either (o) or (o) while 
he says : ^'il y a bien des mots ou T o a un son mM6 de celui 
de V a, et oil sans scrupule on le pent sonner conmie im a," 
that is, he confused (a, o) or (a, a). Interpreting his signs 
by former explanations we find the following novelties. O is 
short = (e) in compoimds of most, as hithermost. Borne = 
{hoorn), bom = (bAAm) ; form a bench = (foorm), form a 
shape = (f^Arm) ; holy = (hoo'Ii), holy day = (haH dee). 
Tolk, maggot, anchor, women = (jelk, msBg'et, sen'ker, wtm'en). 
Rome = (Ruimi). On = ('n) in capon, mutton, lesson, reckon, 
reason, season, apron, citron, saffron, iron, fashion, cushion, 
puncheon. 

1701. Jones confirms the others. The following is his list 
of long sounded as (uu) afford, bomb, comb. Ford, ford, gam- 
boya, gold, Monday, More, Rome, tomb, womb, in which most 
are unusual, and gold, Monday are noteworthy. The oa as 
(uu) are ** aboard, boar a clown,*' now written boor, " board." 
The words doe, does, doest, doeth, shoe, woe, he likewise hears 
pronounced with (uu), although he also gives (daz) for does. 
He admits the soimd of (a) for o in "the beginning*' of 
colonel, colour, etc., comfort, company, etc., coney, conjure, etc., 
money, monkey, etc., mongcom, monger, etc. ; blomary, bombast, 
borrage, bosom, botargo, brocado, chocolate, cognisance, colander, 
coral, coroner, cozen, Devon, dozen, forsooth, gormandize, gromel, 
London, onion, poltroon, pomado, poniard, porcelane, potato, 
recognisance, sojourn, Somerset, stomach, tobaco ; in final -come, 
'dom, -some, -son ; in the hat syllables of chibol, gambol, 
symbol. Even the unusual cases will be recognized as still 
occasionally heard, but they evidently bear the same relation 
to the present pronunciation with (o), as (griit, briik, tshiii) 
do to {greet, br^k, tsheei). Both resulted from overdriving 
a new attenuative habit. 

In the xvn th century then the change fix)m (oo, o) into 
(oo, a) or (oo, o) was complete ; a few more of the (oo) had 
advanced into (uu), more indeed than those which maintained 
their position, and those formerly heard as (u) or (u) had 
become (e), a change to be considered under U. 



Chap. m. { 3. O, 00, OA — XVIII TH CBNTURY. 103 



0, 00, OA — xvniTH Cbntuey. 

Daring the xviii th century the change in the use of these 
letters as just described, was so slight that it will be quite 
unnecessary to enter into many particulars. It will be suf- 
ficient to note some examples, chiefly of exceptions to the 
general rule that o long and oa = {oo\ o short = (o) or (a), 
and 00 long and short = (uu, u), or of exceptions to the pre- 
ceding exceptions to this rule. 

1704. The Expert Orthoorafhist gives oo m flood, blood 
the sound of (a), and in door, fl^oor, moor, poor the sound of 
{oo). He also makes o = (uu) in " tool/, toolves, Borne, comb, 
tomb, divorce, force, forge, form to sit on, bom endured, sup- 
ported, forth abroad, port and its compounds com, de, in, mp, 
trans-port, fport, shorn and torn, engross. Ghost, most, post, 
rost, and o between w and r for the most part is sound^ oo 
(uu) as word, work, world, worm, worry, worship, worse-st, 
worsted, worst, and ivorth ; and in approve, behove, move, prove, 
remove, reprove ; but like short u (o) in chve, glove, love, cover, 
covet, groveling" He admits oa to be a mX>de of lengthening 
0, but says **oa in abroad, broad, and groat, have a pecidiar 
broad sound" without saying that it is the same as au (aa), 
and *^oa sounds ai in goal pronounced jail, (dzheel)." 

1766. Buchanan writes London Lon*en, toon won, lot lot ; 
dost dost, work work, worship wor'ship, woman wem'tn, women 
wtm*m, wonder won'dtr, mouth mouth, money mon'i, son son ; 
twopence top'ins, poltroon poltruun, forth foorth ; globe glooh, 
robe Tooh, whole whool; who huu, do duu, tomb tuum, gold 
guuld, liome Ruum ; move muuv, one wsen, once wsbus, only 
on'li, come kom; soap Boap, broad hrood, oats oota; loath 
lAAth, groat, grsDeet. 

1768. Franklin has of at, bosom boz'om, compared 
kompeerd', other odh'or, government govomment, London 
Lon'don ; only ootl'U, spoke sp^ok, wrote root, some som, one 
won, once wans, to too, in which will be found some uses 
different from Buchanan's. 

1780. Shbridan notes the Irishisms : (duur) door, (fluur) 
floor, (kuurs) both coarse and course, (stray) strove, (drav) 
drove, (rod) rode, (strood) strode, (shoon) shone, (fat) foot, 
which he says were pronounced in England {Aoot, fLoofv, 
\lootb, str(wv, droov, roodi, strAd, shAn, fwt). Most of these 
Irishisms are clearly, all of them are probably, as usual, 
remnants of the xviith century. 



104 



Y, I, IE — XVI TH GENTU&T. 



Chap. III. i S. 



Y, I, IE — xviTH Century. 

When y, i were consonants, they were employed like the 
modem y, J = {j, dzh), and were never interchanged in the 
old writers, although the sound of (j) was not usually con- 
sidered a consonant, as will be noted under y, ta. When y, t 
were vowels they were used indiscriminately, except perhaps 
that / was always^ used as the personal pronoun, and was 
not employed at the end of any otner word. For the present 
section they must be considered as identicaL 

Table Shewbto the Iktbodfction of IE fob E, EE. 



Price. 


Minshew. 


Levins. 


Palsgrave. 


Promptornim, 


1668. 


1617. 


1670. 


1630. 


1440. 


believe 


belecTe 


beleeve 


beleve 


beleneness 


beriege 


besiege 








bier 


beere (biere) 
briefb (breefe) 


beare 


beere 


been 


brief 


breefe 


brefe 




cayfdier 










casbier 


casbeere 








ebief 


cheife (cbiefe) 


cbeefe, cbief 


cbefe, chief 


cbenetna 


field] 


field (feeld) 


feeld, field 


felde 


feelde 


'fienol 


feend 




fende 




■fierce] 


fierce 


fierse 


iyers 


VBTBSe 


riend 


frend (friend) 


frende 


fiende 


freende 


frontier 


frontier 








tS 


CTeeue (grieve) 


greeve 
kercber 


kerchefe 


^Sde 


riief] 
Uege 




liefer 


lefe 


lefe 


liege 




lege 


lycbe 


niece 


neece 


neece 


neyoe 




piece 


peeoe (piece) 


peeoe 


pece 


pece 


fierce 


pearce (pierce) 


perse 


perce 


peercyii 


Bbid!^ 


prieste 


preestly 


preest 


preest 


sbeeld 






scheeld 


siege 


"lege 


sege 


iege 


oege 


nerse 




cearse 






rieye 


due (dene) 


seefe 




n 


tbief 


theef 


iheefe 


tbefe 


Tiew 


view 


vewe 






yield 


yeeld 


yeeld 


yelde 


yddoM 



IE was often used at the end of words where we now use 
y. IE in the middle of words was employed in the xiv th 
century indiscriminately with e or ee, but not very frequently. 
In the XV th and xvr th centuries it had Mien out of use, 
though we find it fully established with the modem sound 
of (ii) in the xvii ih century, in which is included also the 
word /riend as already noted (p. 80). The preceding table 
containing all Price's list and a few other words in bi^ckets, 

1 In MSS. If was not nnfreqnently used even for ^e personal pronoun in the 
XT th century and earlier. 



Chap. III. i 8. Y, 1, IB — XVI TH CENTURY. 105 

will shew the corresponding spellings in the Fromptorium 
1440, Pabgraye 1530, and Levins 1570, and Minshew 1617; 
the spellings in parenthesis in Minshew's column, are spellings 
which he recognizes and gives in cross references, but the 
other spellings are those under which he explains the words. 
It will be seen that MinsheVs book shews the exact period of 
the transition, when generally both spellings were sufficiently 
known to require noUce, but one was decidedly preferred bv 
the author, and that one was only occasionally ie. The French 
mice, piice, Jier, stiff e and occasionally chief maj have in- 
fluenced some vfordA, but others, as believe, bier, friend, field, 
lief, thief, yield, seem to have no reason, either in sound or 
etymology, for this curious change of custom in spelling. 
For our present purpose, then, we may dismiss ie, consider- 
ing it, in the midile of words, as a fiEmciful variation of ee 
and having precisely the same value ^ii) towards the close 
of the XVI th centuiy, and, at the end of words as an archa- 
ism for y, having the same sound {i)} 

There seems to have been only one sound of short i and, 
with rare exceptions, such as machine, only one sound of long 
i, during the xvi th and subsequent centuries. At the pre- 
sent day, English short i or (i) is the wide sound of the 
Italian or European short i or (i). The fine sharp clear (i) is 
very difficult for an Englishman to pronounce, and although 
the Scotch can and do pronounce it,^ they not unfrequenfly 
r^lace it with (e) or (e), not (e). In this respect they re- 
semble the Italians who have so frequently rejnaced Latin i 
by their e chiuso or (e). The Dutch may be said not to 
mow (i), as they regularly replace it by {e). The English 
sound (i) lies between (i) and (e). The position of the tongue 
is the same as for (i), but the whole of the pharynx and 
back parts of the mouth are enlarged, making the sound 
deeper and obscurer. According to Mr. M. Bell there 
is the same distinction between {e) and (e), the latter 
being the wide form of the former, and he hears {e) 

^ The word pieree seems to hare re- ' Mr. Melyille Bell says in a private 

tained the spellmg ^$$, and the cor- letter^ that the sound of the short ** (i) 

respomti^ pronunciation to a later for 2 is very common, as in giTe=(g^), 

time, ^y^still write P^-cy, and P«rfl» gied, ^en, gie's [derivatives], whig. 

is called (Feis) or (Pis) in America, wig, hig [to build], build, -er, built 

In Love's Labour Lost, Act iv. sc. 2, [often i»lt] Idng-dom, wick, gig, 

L S6f 1623, Ck)medies p. 132, we find ginfham, widow, Britain, finish, whin, 

^ Master Person, guoM* Person ? And etc. ' In such words the Englishman 

if one should be perst, Which is the hear* the long (ii). This is a point 

one f* which indicates the pronuncia- which will have to be considered here- 

tion (Mas-ter Pers'on, kwaa-si "Pers- after. See especially the examnles of 

-oon" f And tf ** oon " shuuld bt Bcotoh pronunoiation in Chap. XL § 4. 
"pcwtj" whftsh h dhe "oon"P). 



106 T, I, IB — XVI TH CENTURY. Chap. III. j 8. 

in the French et, and English day, (drf, d^), and (e) 
in the Scotch ill, English ailment (el eeil'ment) and English 
air (eel), and also in my own pronunciation of the English 
ell, whereas he supposes the true sounds of English men^ 
man to be (mEn, maBn) and to differ precisely as (i, i). 
My own pronunciation of man he finds frequently the 
same as his pronimciation of men, so that to him 1 pro- 
nounce men, wan as (men, msn). To me (s) is a much 
deeper sound than {e, e) and is heard in the French mime, 
German sprdche (mEEm', shprEE^h*^). This discussion will 
serve to shew the nature of the difference (i, i), and the 
ease with which they mOT be confounded. Almost every 
Englishman pronounces French il as English ill (il), and 
almost every Frenchman pronounces English ill as French il 
(il), French tie, English eel being identically (iil). Now 
the true long sound of (i) is not an acknowledged sound 
in our language, although in frequent use among such 
sin^rs as refuse so sav happ^, st^l, eA, when they have 
to lengthen happy, still, ill.^ They say (msfiii, st2>l, nl) 
although some may prefer (stelll, till) which has a bad effect. 
Where the long sound of (i) might be expected, we get the 
long i, to be presently noticed. Hence most of those who 
examined sounds, as Wallis, naturally paired (ii), whose 
short sound was absent, and {i) which was without a long 
sound, and probably did not hear the difference,^ though Sir 
Thomas Smith could find no short sound for (ii) in the Eng- 
lish language.' What we have to conclude from this is, 
that because ee long and i short are represented generally by 
the same character, with or without a mark of prolongation, 
by orthoepists, it by no means follows that they nad the same 
sound. My own belief is that short i was (i) from the 

1 This was remarked bj Dr. Tonng, assertion that (t) was an independent 
Lectures on Natural Philosophy. 4to. vowel sound, ana resolutely paired (ii,f). 
Tol. ii, p. 277 : " When lip is length- This is by no means the only point in 
ened in singing it does not become phonetics concerning which the ex- 
Itap" Observe the singing of " s^tll perience of nearly a quarter of a century 
so gently o'er me st^lii^/ which be- has enlightened nim. He would, how- 
comes fstttl so dzheent'Iif* ooar mii ever, particularly notice the stepped 
siiil'uq.} Dryden's line, from his Vmi vowels, which on p. 63 of that work, 
Creator, " And make us temples yroxthy he found himself unable to separate 
thee,*' is well adapted to render the from their consonants, as in (p^ pet, 
difference of the vowels in (-dhi dhii) pset, i>pt, pat, put), but which he nas 
sensible. been in the habit of separating for 

* The present writer should be the many years, 
last to throw stones at those who^ do * See p. 112. Cooper, as we have 

not hear the difference between (i, %) seen (p. 83), forms an exception ; he 

for in his Alphabet of Nature, 1845, appears to pair {ee, Q, and certainly 

p. 65, the first work on phonetics whicli does not pair (ii, t). 
he published, he objected to Knowlee's 



Chap. ID. § S. Y, I, IB — XVI TH CENTURY. 107 

earliest times to the present day. Against this supposition 
must be placed the facts that^ as already pointed out, snort (i) 
is not at all unfrequent in Scotland, and was apparently 
recognized in English in 1701 by Jones, a Welshman, and 
1766 by Buchanan, a Scotchman, and also that in Ireland 
final -y, which is in England (-i), is invariably (-i). The 
Irish English ^nerally representing a xvn th century Eng- 
lish pronunciation, there is a possibility of (i) having been 
somewhat common in England during the end of the xvii th 
and beginning of the xviii th centuries, a period of English 
pronimciation remarkable for a tendency to thinness of 
sound. The true long vowel (ii) will come under consideration 
again in the next Chapter imder I, T, when the importance 
of the preceding discussion will more clearly appear. 

As to long i in English at present, it is without doubt, a 
diphthong, and has been generally recognized as such from 
early times. But orthoepists are not agreed as to the nature 
of its first element, and this becomes an important con- 
sideration. The Italians and French only approach the 
sound of our l«ng i very loosely, in the Italian words dome, 
loido, zaino, and the French paien, faience. These may be 
more properly written (daai*no, laai'do, tsaai'no; paiieA, 
faiiaAs), so that in the Italian the first element, m the 
French the second element is lengthened. In Germany the 
sound written ei, ey, ai, ay is intended to be (ai), although 
these diphthongs are ybtj^ variously pronounced. Rapp 
gives the literary high varieties (ai, oi, ei, ei) and Schmeller 
notices the Bavarian dialectic varieties (a, ai, ai, e, si, ei, ii).^ 
The different Scotch sounds of long i will be fully considered 
in Chapter IV. § 2, under I. In England we have only one 
recognized pronunciation of i long, but we have also two 
recognized sounds which may be heard in /satah, or in the 
ususi English pronunciation of %64> x^V** ^^^ ^^^ distinction 
is, or us^ to De, strongly insisted on at Eton. The second 
of these sounds, the English pronunciation of the Greek ai, 
is (ai). What is the first ? Knowles,' following Sheridan, 
says it is (a), the only difference between i long and oy con- 
sisting in the brevity with which the first element is dwelt 
upon in the first sound. This is an Irishism no doubt, 
although he is closely followed by Haldeman,' who makes 

^ Sapp, Fhysiologie der Sprachey and the Tarions properties of all its 

ToL It. pp. B6 et sqq. SchtmiUr, Man- simple and componna sounds, as com- 

darten Bayerns, p. 56. bined into syllables and words. Jjtm- 

' Jamet Knowkt, Pronouncing and don, 1847, 8to. 
Explanatory Dictionary of the English > Analytic Orthography, { 106, 400. 

Language, founded on a correct de- and examples § 602, 610. 
Telopement of the nature, the number, 



108 Y, I, IE — XVI TH CENTURY. Chap. III. } 8. 

the first element (a), and identifies English long i with the 
German eiy of which Schmeller makes the first element (a). 
Mr. Melville Bell identifies the first element of his pro- 
nunciation of English long i with (a). The first element of 
my pronunciation <^ the German ai he considers to be (ah), 
a soimd that I can only with difficulty disting^h from (a), 
as T am apt to labiaUse (a) in speaking. But in unaccented 
syllables ne makes the first element of his pronimoiation of 
long i to be (ah). This was the element he recognized in 
my own pronunciation of this diphthong in all cases. Many 
Londoners certainly use (s^) as the first element. Again, 
Wilkins and Franklin call tne first element (»). And Smart 
making the first element ur without soimdin^ the r must 
mean (ao). The second element is of course the glide, and 
the last element (or second as it is usually called) is the 
vowel (i) or (i), very often the latter I beheve in English. 
Mr. Bell only recomizes the glide, 5c (see p. 15), that is, the 
fflide to the (j) position. According to the mode of writine 
diphthongs which I adopt I must give (i) or (i) as the final 
element, leaving the ffhde to be denot^ by juxtaposition. 
Hence we have the following 

Analyses of English hng I— 

Sheridan aad Ejiowles (Ai) 

Haldeman (ai) 

Walker and Melyille Bell f ai) accented 

Melville Bell {^) tmaccented. 

Londoners (eei) 

Scotch (^, ei, si, ai, oi, ohi) 

"Wilkins and Franklin (eij 

WaUis and Smart (ooi) 

Now this being the sound of the personal pronoun, is 
heard every day and constantly ; but after competent orthoe- 
pists have carefully examined it, they are unable to agree as 
to its analysis. One reason is of course a real difierence of 
pronunciation, but another appears to be that the first ele- 
ment is pronounced with extreme brevity, so that in British 
speech it is not sufficiently heard as distmct from the follow- 
ing glide. In endeavouring; therefore to fix it, different 
ol^rvers either begin far back in the scale of distinct vowels, 
or catch the sound closer and closer to (i). Thus it may be 
that the whole series of sounds (o-ohoa-ahsDei) may be heard in 
this diphthong, all gliding into each other with inmiense 
rapiditv. Again the first element being so indistinct, others, 
as Wilkins and Franklin, or Wallis and Smart, take refuge in 
one of the colourless sounds as (e, oo). 



Chip. HI. § 8. Y, I, IB — XVI TH CENTUEY. 109 

Now I hear the vowel (a) very clearly in (ai) as in the 
Etonian pronunciation of x^ 9 ^^ ^ cannot hear it in the 
Etonian pronunciation of x^V^ ^^^ I do ^^^ <^ (g) there. 
I therefore prefer to represent the English % long, the Etonian 
pronunciation of Greek 6i by (oi), and the English aye, yes, 
the Etonian pronunciation of the Greek at by (ai). The pre- 
ceding discussion will apply, as to the first element, to &e 
present pronunciation of ^m? in nowy haro^ core. 

We are now better prepared to understand what our 
authorities say on the subject. The first one is sufficiently 
perplexing. 

1630. Palsgrave says : "/ in the frenche tong hath .ii. dyuerse 
maneis of souitdynges, the soundyng of t, whiche is most generally 
Ysed in the frenche tong, is like as the Italians sonnde i, and suche 
with ys as sonnde the latin tong aright, whiche is almost as we 
souiide e in these words ahee 9^ flie, a heere for a deed corps, a peere 
a felowe, afeev^ rew^rde, a little more sonndynge towards t, as we 
sound i with vs." 

Now du Guez says : " Ye shal pronounce . . . your i, as 
sharpe as can be," by which I understand, with the smallest 
lingual and pharyngal aperture, or as clearly (i) as possible. 
When Palsgrave says: ^* almost as we souwde ^," etc., the 
almost is merely one of those safeguards which orthoepists 
love to insert, and can scarcely avoid inserting, when they 
give the equivalent for a foreign sound which they seem to 
hear in their own tongue, but doubt the correctness of their 
hearing. But what does he mean by '^ a little more sound- 
ynge towards t, as we sound i with vs '' P A vowel cannot 
sound a little more towards a diphthong, and yet long i was 
certainly most generally recognized to be a diphthong in the 
XVI th centuiy, although it is probable that Palsgrave may 
have had an older pronunciation, rather of the xv th than of 
the XVI th century. Could he mean that the sound seemed 
between (i) and (e)P It would be difficult to insert one. 
Could he mean that as he pronoimced those. English words 
the sound had a tinge of (e) in it as it were (ii), and that the 
French pronounced a clearer (i) P The matter becomes still 
more enigmatical as he goes on to say : 

"If f be the first letter in a firenche worde or the laste, he shall 
in those two places be sounded lyke as we do this letter y, in these 
words with vs, hy and iy, a spye, a flye, awry, and suche other : in 
whiche places in those frenche bokes, as be diligently imprinted, 
they vse to writte this letter y : but whether the frenche worde be 
written with ♦ or y, in these two places he shal be sounded, as I have 
shewed here in this rule, as in ymaye, conuerty, ydole, eatourdyf in 
whiche the y hath suche sounde, as we wolde give him in our tong." 



110 Y, I, IB — XVI TH CENTUBY. Chap. III. § 3. 

This sound, whatever it was^ must be distinct from the 
other sound of i. Now as Falserave noways describes the 
sound, or hints at its being a dipnthong, we can do nothing 
but refer to Meiffret 1550, who writes : " je vi, oi, aosi, j'ey 
b&ti, je b&ti ou b&tis '' with precisely the same sign . as he 
uses in " Louis Meigrst, LionoEs.'' Perhaps Palsgrave 
would rejoin: "true, but he was a Lyonnais; I give the 
Parisian pronunciaticm/' In the mean time we are not 
assisted towards Palsgrave's own pronunciation of the English 
"by and by, a spye, a flye, awry."^ What follows is as 
perplexing : — 

"For as moche as r and i come often together in the frenche 
tonge, where as the v hath with them his distinct sounde, and the • 
is sounded shortly & confdsely, whiche is the proprete of a diph- 
thonge. I reken vi also among the diphthonges in the frenche 
tong, whiche whan they come together, sluiU haue suche a sounde in 
frenche wordes, as we gyue hym in these wordes in our tong, 
a »wyn$j I dwynej I twyne^ so that these wordes agvysh^ agvylUdny 
eondvyrej dedvyre, aviourdhvy, meih^if, and all suche shall sounde 
theyr t? and i shortly together, as we do in our tong in the words I 
have gyven example of, and nat eche of them distinctly by himself, 
as we of our tong be inclined to sound them, whiche wolde rather 
say aviourdhvy, dedvyt^ saufcondvyt, gyuynge both to t? and ♦ theyr 
distinct sounde, than to sounde them as the frenche men do in dede, 
which say <wtourdhvi/, dedvyt^ saufcondvyt, soundyng them both 
shortly together, and so of all suche other." 

It is a well-known modem English error to say (Iwii) for 
(lyi) ltd. Pabgrave, whose ears cannot have been very acute, 
here seems to authorize a similar use. At the same time the 
conversion of (y) into a consonant as (w), is directly opposed 
to the previous direction to give (y) its "distinct sound,'* and 
pronounce (i) "confusely." But can Palsgrave have also 
meant that the second element in m in the French words 
cited was the same as in stvyne, dwyne^ ttvyne ? The y in the 
French words is not even final or initial. It could have had 
no sound but (ii) even according to Palsgrave. Did Pals- 
grave say (swiin, dwiin, twiin) or (swim, awim, twim) ? It 
18 the only legitimate inference, and there is no slight proba- 
bility of its being correct. We shall see that Palsgrave pro- 
nounced ou as (uu), which was a xiv th century pronunciation 
continued archaically into the xvi th century, and although 

^ It deserves however to be recorded James the First's time has : " Lord 

that Gill writes (en*emai), not (en-emil, our Grod arise, Scatter his enemies," 

and has at least once (dim-adzhes], al- giving (en-emaiz), if ^e rhyme is to be 

though on another occasion he writes preserved, though in modem practice 

(tm-aadzh) so that the former may be we sacrifice the rhyme and often sing 

a misprint. The God savf the king of (enimttz). 



Chap. m. {3. Y, I, IE — XVITH CENTUKY. Ill 

the recognized pronunciation at tliat time was (ou), yet the 
example of Bullokar (pp. 94, 98,) shews that there were still 
many who preferred the (uu) sound. In the same way 
perliaps both Palsgraye and Bullokar preserved the (it) 
sound of long i, usual in the xiv th century, notwithstand- 
ing the genial adoption of (ei). The new (ei, ou) and the 
old {iif uu) stand precisely on the same ground, and therefore 
I am incHned to tnink that Palsgrave and Bullokar said (ie), 
as distinct from (ii). Further reference to this curious re- 
tention of an old sound wiU have to be made in the next 
chapter under I. 

1547. Salesbttry does not leave us in much doubt, for he 
writes (ei) for long i, thus : 

I ei (ei), vtne vein (vein), wtite toein (wein) ; dyches deiteyi 
(deitsh'tz) ; thtve ddein (dhein) ; siones Beine (scinz) ; Latin Dice 
deieu (dei'ku), tibi teibei (tei'bei), Dei Deei (Dee-i), qui quei (kwei). 

At the same time he reprobates this pronunciation of 
Latin, and says : 

''I in Welsh hath the mere pronunciation of i in Latine, as 
learned men in our time vse to sounde it, and not as they . . . with 
their lotacisme corrupting the pronunciation make a diphthong of 
it, saying veideiy teibei^ for vidi^ tihi.^^ **/ in their language is 
equivalent to the following two letters in oiu^ ei, but they are com- 
pressed so as to be pronounced in one sound or a diphthong, as in 
that word of theirs I, ei, (ei) ego." ** Y often has the sound of 
the diphthong ei as thyitb, ddein (dhein), tuus ; & its own soimd as 
in the word THTiwrE, ihifnn, (thtn), gracilis." 

That Salesbury's ei was different from his ai, and that he 
meant to indicate a different sound in such English words 
that have lonff i> from that in other words having ai in his 
transcription, is I think evident, because he never confounds 
the two sounds, and because in modem Welsh the sound ei 
sounds to me as (oi), and ai as (ai). I think, however, that 
his letters ei justify me in considering, or rather leave me no 
option but to consider that the EngUsh diphthong sounded 
(ei) to Salesbury. 

As to the short i, he identifies it with Welsh y, considering 
the latter the especial sound. He also says that Welsh u 
"soundeth as the vidgar English people sound it in these 
wordes of English, trust, bury, busy, Jluberden.^' I think 
that he cannot point to any other sound but {i), supposing the 
true Welsh to be (y), a sound which Mr. Melville Bell hears 
in the unaccented syllables : the hous^^, (dhy Hauz'yz) as he 
would write the sounds. The difference between (i, y) is 
very alight indeed. In practice Salesbury is not very precise. 



112 Y, I, IB — XVI TH CENTUEY. Chap. III. § 8. 

as may be seen by tbe following list of words in wbicb short 
i occurs, but his theorjr leads me to adopt (i) as the true 
sound of English short i in his time.^ . 

God be wtth you GM hiwio {Gtod bii'wi,©), oeactoitse groiiwB 
(graa'sijUs), oondictoit condisywn (kondi8*f,im), twtwcle twinkl 
(twiq'k'l), WKTNCLB wrinJU (wriq-k'I), ktvoss kings (kiqz), osldino 
gMing (geld'iq) ; Gylbbbt, Gilbert (Gil'bert), gykgee tsinMr 
(dzhin-dzhir), BEGeYves, legging (beg'iq); hoij, holy (hoo*1>, hoI**); 
EXHIBITION eesihiaiwn (eksibis'i^iin) ; PBomBinoy, proibisiwn (proo,- 
ibisiun) ; lyly lUi (lil'i), lady ladi (laa'di) ; VAsrikpapyr (paaptr), 
EYOHT rieht (rikht) ; thystle, thystl (thi's-fl) ; this ddys (dhis), 
BUSY husi (biz'i) ; wtnne wynn (win) ; thtntte thynn (thm) ; 
EarrzT hnieht (kn^t). 

1568. Sib T. Smith says: "I Latina, quae per se prolata, apud 
nos tantam yalet quantum Latine, ego, aut oetdus, aut etiam^^ 
by which I understand that the three words /, eye^ aye 
had the same sound, precisely as we are told by Shakspere, 
Romeo and Juliet, Act iii, Sc. 2, v. 45, (I quote from 
Steevens' reprint of the quarto of 1609, which agrees in this 
passage with the folio of 1623 ; the lines do not occur in the 
quarto of 1597) : 

Hath Bmmo slaine himselfe ? say thou but I 
And that bare vowell I shall poyson more 
Then the death-darting eye of cockatrice, 
I am not I, if there be such an I. 
Here aye is spelled /, and thoroughly identified with it, as 
"that bare vowell I," and with the suggested ^^ eye o£ 
cockatrice" in the next line. Although Smith identifies 
these three words, he spells them differently, introducing i as 
the sign for long i, and pairing it with short t. He thus 
deprives the Latin language of the soimd of (ii), for he pro- 
nounced Latin e as (ee). Hence when he comes to the sound 
of (ii) in English, he exclaims in perplexity : 

" Quid autem fiet ubi sonus invenitur quem neque Grseci, neque 
Latini habuenmt, praesertim cum omnes coram literae in simiHbus 
coram sonis faerunt absumptse ? Ecce autem sonum Angloram et 
Scotoram alium diversumque ab omnibus his,' qui nee i (ee) nee 
? (ei) reddit auribus, sed quoddam medium, et tamen simplex est, 
literaque debet dici : est autem semper fer^ longa." 

His examples are me, see, meet, deep, steep^ feel, feet, sheep, 
queen, mean? seek^ she, weeA, leeA, beef, neese, bee apes, 

1 So far as I could hear, the Welsh * That is, not one of the sounds 

dim was prononnoed bj scTeral Welsh which he had already considered, and 

gentlemen precisely as the English which were apparently (aa a, ee e, ei i^ 

dim^ that is (dtm)^ and they all ohjected oo o, nn n, yy). 

to ike pronunciation (dim). * ^ Intelligere." Qu. mim, Tultos f 



Chap. HI. } 3. Y, I, IB — XVI TH CBNTURT. 113 

wliraoe, through Salesbury and Palsgraye, we know that the 
sound was (ii). Smith therefore recognized no short (i) in 
English. The sound of his e short most therefore have been 
different from (i), that is, as I believe (i)> agreeing with 
Salesbury. 

Smith recognizes the two diphthon8;s (ei, ai) but finds 
scarcely any £fference between them, although he says that 
"muliercute" pronounce (ei) for (ai). This will be con- 
sidered under (ai), p. 122. In no case in which the or- 
thography uses long i does Smith write ei, so that but for 
his ratlier veiled identification of I with eye, we should have 
had no clue to the sound intended. 

1569. Habt says : ** Out of all doubt, no nation of the foresaide 
but we and the Scottish, doe at any time sound t, in the aforesayde 
sound of M ; wherefore that ETiglish Greek reader which shall giue 
the same sound to i which he doth to €^ doth ftirther this errour 
much amongst vs." 

He also writes (reid bei) for ride by. But he makes ee in 
Greeks the long soimd of t in in, that is (it), and is thus not 
80 accurate as Smith, who distinguishes the sound as (ii). 

1580. BuLLOKAR calls long i a vowel, and does seem to know 
that it has a different sound from short e. He says : '' I, 
hath two soundes, the one a^^eing to his olde & continued 
name, and is then a vowell, l£e other soimde agreeing to the 
olde name of ^, and of my c^ (dzh), and is then a consonant.'' 
He gives as examples : '^ I ly in my sisterz kitchen with a 
pillo*w besyd her peticot, and thy whyt pilion," where the 
accent denotes length, and o^w means (u). What " the old 
and continued name'' is, he does not write. He has no other 
distinction between long and short i but this accent, and 
never even hints at the possibility of their having two sounds. 
He uses the accent to indicate the long a, e, y, o only, and 
has a new sign ^ for (ii), on which he says, and it is the only 
clue I can find : 

« e hath two soundes, and vowels both, the one flat, agr6eing to 
his old and continued name : and the other sounde more sharpe and 
betwene the old sound of the old name oi :$: and the name of : • : 
for such difference the best writers did use :ea: for :e: flat and long: 
& 0a, ee, ie, eo fox 101 shaipe." 

This "flat ^," was undoubtedly (ee)> and the "sharpe «" 
was (ii). The "old name of e is therefore (ee), and the 
" ahiu^" sound of e, or (ii) is said to lie between (ee) and 
the name of i, that is, its long sound, whatever that may be. 
Now we have se^oi l^t Smith says that (ii) is "quoddam 
zaediom/' betweeiL (ee) and (ei), so that we need not expect 

s 



114 Y, I, IB — XVI TH CENTUEY. Chap. III. § 8. 

more precision in Bullokar, and although, it is really non- 
sense to say that (ii) lies between (ee) and (ei), since (ei) is 
compounded of (ee) and (ii), yet as Smith actually said so, 
Bullokar may have meant tne same. But BuUokar con- 
stantly neglects to write the acute accent, his sign of pro- 
longation, over i. Thus he has cqntryzy cgntriz in successive 
lines. Again he always writes wryth = written with a long 
y, and it would be difficult to beheve that even a pedantic 
theorist ever said (rtmt'n). GiU writes (writ'n). If how- 
ever we suppose that BuUokar, as well as Palsgrave, pro- 
nounced long i as (it) and short i as (i), all difficiidty arising 
from this source would disappear. And although tne state- 
ment that (ii) lies between (ee) and {it) is not so correct as 
that {ii) lies between (ee) and (ii), yet it is not at all ex- 
travagant for a phonetist of that time. K, as will appear in 
the next chapter, (it, uu) were probably the xiv th century 
pronunciations of long i and oUy then the retention of iji) by 
feullokar and Palsgrave wiU be precisely paraUel to their 
undoubted retention of (uu), and would have precisely the 
same archaic effect in the midst of the gener^ (ei, ou) as 
((7bliidzh', griit, briik) have at the present day amidst the 
usual (t^bloidzh*, greet, br^ek). The whole subject will be 
properly discussed in the next chapter, and in the mean 
time the only legitimate inference fix)m Bullokar's notation 
and practice seems to be that he pronounced long i as {ii). 

1621. QiLL uses also a simple sign for long e, namely / 
He says : 

'* Differentia significatioms (quoad fieri potest, & sonus permittit) 
orthografia discemitur. Sic J. ego. ei oculus, eft ita." — " Nee e, 
ssepiiis preeponitur i, dicimus enim hii (neei), adhortantes aut 
laudantes, & ei (ei) ete oculus, ^t (eei) etiam, ita: vbi tamen 
sonus Yocalis, exiguum distat ab illo qui auditur in ^n tnus, & 
mjn meus." — '^Communis dialectus aliquando est ambiguus. Audies 
enim ^ai aut ^ei (dhai, dhei) they, illi." — **/, est tenuis, aut 
crassa : tenuis est breuis, aut longa : breuis sic notatur t, vt in ein 
SDTKB peccatum : longa sic i. vt in Bin seene visus, a, um : crassa 
autem fere est diphthongus ei; sed quia sono exilior paul6 quam si 
dif^deremur in e, retinebimus antiquimi ilium et masculinum 
sonum .... eumque signabimus hoc charactere j, vt in ejn signs 
signum. Omnium differentia est in toin wjoshstr yinco, win weene 
opinor, ujn wrme vinum." 

The meaning of these passages is not very clear, and they 
have occasioned me considerable difficulty, as I felt it important 
to determine the precise signification of OiU's symbols. It 
is clear that his j was little, if at all, different from (ei), and 
that this difference consisted mainly in dwelling more upon 



Chap. III. { 3. Y, I, IE — XVI TH CENTUKY. 115 

the (e) sound in the diphthong which he writes (ei) than in 
that which he writes i ; this is the only sense I can attach to 
the expression that the sound oij ^^jere est diphthongus ei^ 
sed sono exilior quam si diffunderemur in ^, as it were, 
than if we were diffuse over the e. The distinction is then 
precisely similar to that which Sheridan and Ejiowles make 
between modem /, oy, where they suppose the first element 
in each case to be (a), but to be instantly lost in /, and retained 
long enoufi^h to be distinctly heard in oy, (p. 107). We seem 
to have omy to change (a) mto (e) to obtain Gill's distinction 
between /, eye. GiU frequently interchanges (ai, aai) and 
does not seem to be very particular about the distinction 
between (ei, eei), but he appears to have always attached 
great importance to the first element in (ei) and (ai). He 
says of diphthongs generally : 

''Nee tamen in ommum diphthongorom elatione, utrique yocali 
8onus integer ubique constabit. Etenim vocalis preecedens ssepe- 
numero acutiiis sonare videtor, & clariiis; in ai et ei, ita aures 
implcre, ut .t. subiungi sequius esset, quam ad latus adhserere," 

alluding evidently to the Greek forms c^ 17. The conclusion 
would appear to be that Gill's j', ei, az were more properly 
(*ei, e*i, a'i) where the apostrophe indicates for the moment 
the extremely unaccented or unimportant character of the 
element to which it is prefixed. For this we might write 
(ei, eei, aai) if Gill did not occasionally distinguish between 
(ei, ai) and (eei, aai). We must not forget however that 
Gill blames Hart for writing ei in place of /, where Gill 
prints / meaning, probably, j. In this case his j would 
appear to be considerably different from his (ei). 

Another hypothesis is possible. We shall see that at the 
time of Wallis, 1653, (ai) was a common form of long i. It 
is possible that this was one of the xviith century pro- 
nunciations which Gill adopted, and hence his J, ei, ai may 
mean (ai, ei, ai), and as this is the most convenient dis- 
tinction which I can draw between the sounds, and also 
a^^rees in making j but slightly different, and ^et decidedly 
different, from (ei), I shall adopt it in transcribmg QiU. 

But for the xvi th century generally, the positive assertion 
of Salesbury that long i was (ei), and the identification of the 
sounds of X ^69 <^y^ by Smith, leave me no choice but to 
use (ei) for long i. Shakspere was bom the same year as 
Gill, yet as he did not Uve so long into the xvii th century, 
he may have used the same pronunciation as Smith and 
Salesbury. Certainly his /, eye, aye must have had the 
same sound (p. 112). But perhaps long i was also often 



116 Y, I, IB — XVn TH CEKTUBT. Chap. HI. § 8. 

oalled (ai) as it still is, and as it probably was in the xivth 
oentory. 

If the bypotbesiB b^re adopted for the pronunciations of 
long i by Palsgraye and BnJlokar; Salesbury, Smith and 
Hart ; and QiU, niunely (U, ei, ©i) be correct, we have the 
ph^iomenon of the coexistence of two extreme sounds (it, ei) 
with their link (eiX during Ihe greater part of the xnth 
oentury, bringing the pronunciation of the xir th and xvu th 
centuries almost togedier upon one point. A curious ex- 
ample of the present coexist^ice of similar sounds in the 
various Scotch dialects will be given in the next chapter. 

The short sound of i, 1 take to be (f) and not (i)* notwith- 
standing that GHll and sub^u^it writ^^s consider (ii) to 
have been its lonff sound, l^is conclusion rests principally 
on the authority of Smith and Salesbury. 

Y, I, IE XVIITH CEKTtmT. 

Price's list of word» in w = (ii) has already been given, 
(p. 104,) and no Airther notice of this coonbination in tiie 
X vn th century is required. 

1640. Bbn Jonson, like BuUokar, entirely ignores the 
diphthong character of long i. His description answers 
to (i) or (t)p but certainly not to the diphthongs (ei, ei), one 
of which he most probably uttered for ms I. He says : 

'</, is of a naiTower sound then e, and uttered with lesse open- 
ing of the month; the tongue brou^t backe to the palate, and 
striking the teeth next the cheeke-teeth. It is a Letter of a double 
power. As a Vowell in the former, or single Syllabes, it hath 
sometimes the sharpe acc^it ; as in Hnding. minding, pining, 
whining, wiving, thriving, mine^ thine. Or, all words of one Syllabe 
qualified by e. But, the flat in more, as in these, hiU. hitter, giddy. 

UUle. incident, and ^e like In Syllabes, and words compos'd 

ci the same .Elemenis, it varieth the sound, now sharpe, now flat ; 
as in giv0^ give, alivej Uve. drive, driven, title, title. But these, use 
of speaking, and acquaintance in reading, will teach, rather then 
rule." 

1653* "Wallis says: "I vocalis quoties brevis est sonatur ple- 
rumque (ut apud Gallos aliosque) exili sono. ITt in hit morsns, 
totll volb, tiiu semper, u/in lucre, pin acicula, ^n peccatum, fill 
impleo. At quoties longa est plemmque profertur ut Oraecomm e^. 
ITt Mte mordeo, wHe BitreitsLgeni&, stile stilus, wine vinum, pine tabe 
eonsumor, etc., eodem fere modo quo Gallorum at in vocibus main 
manns, pain ]panis, etc. nempe sonum habet compositum ex GhJlo- 
rum ^ fbeminmo et » vel y." 

This should b^ (oi), or (osi), or {ooi), the diflTerence being 
slight, and all so like (ei) that we may take that as the souni^ 



Chap, IILiS. y, I, IE — XVniH CENTURY. 117 

especially as WiUdns adopts this {(xjxl Wallis also admits 
this sound in the first element of boil, toil, oil, bowl globus, 
owl, which he pronounces (bail, toil, oil, beul, eul). In 
another place he says that long i is ''idem omnino sonus cum 
Gradcorum et," 

1668. Welkins gives distinctly **(9i) our English i in 
biie," the first element bein^ identified with t# in '' but, full, 
fut^ mutt-on, pull, rudd-er,* which is meant for (o), as it is 
stated to be wholly guttural^ and to be represented by y in 
Welsh. 

1668. Price merely talks of long and short «V 

1669. HoLDEB says: '^Our vnlgar • as in Hih, seems to be 
such a diphthong (or rather syllable or part of a syllable) composed 
of 0, • or e, % (ai, ei), and not a simple original yoweL" 

1685. CooFXB says : '^ U in Cut et % (ai), dipthongum £Eunllim^ 
constituunt, quam •' hngam vocamus ; ut wine, yinum, hoc modo 
pronunciator ante nd finales ; ut blind csbcus, wind yentus : at 
pin^d pro pinned acicula subnexus; k yerbo to pin; breyis est; 
pined marcidus ; k to pine marceo ; dipthongos e^ Scribitor per 
m in beguile iaWo ; disguiee dissimulo ; guide dux ; guidon Impera- 
toris baculus : per oi in in-join in-jimgo, joint juictura ; jointwe 
doB, broU torreo, ointment tmguentum." 

1688. MiEOE says : '' L'autre • a un Son paiticulier, et qu'on ne 
■aurait mieux yous representer par la plume que par oes deux 
Yoyelles at ; comme dans les mots /, prtde, crime, H est vrai que 
oe Son paroit d' abord un peu rude et grossier ; mais les Anglois 
ku donnent un certain Adoudssement, doat les Etrangers se rendent 
bien tot capable. Cet Addoucissement consiste, en partie, ii ne 
faire qu'un Son d' ai, en sorte que ees deux Yoyellee ne sont pas 
tout-^-fiEdt distinctement prononc^." This expression seems to 
point to that extreme breyity of the first element which still pre- 
yails, and makes the analysis of this "Rw^liaVi sound so difficult. 
It must be also remembered that there is nothing approaching the 
compactness of English diph^ongs in ^French, where a looseness 
preyails similar to that in our oy. 

1701. Jones says in one place Ihat the sound of short « 
(o) is written o before i in boil, coil, coin, foil, mail, &c., and 
in another place that the sound of i is written oi in those 
words. It follows that he analyzed long i into (ei). 

It appears therefore that the long i of the xvn th century 
was the same as at present, and hence it must haye been so 
during the xviu th century, and indeed Franklin, 1768, 
writes (oi), and Sheridan, 1780 analyses long i into (aI) with 
veiy short (a), (p. 107,) and Walker into (aei) or (ai). 



118 EI, AI — XVI TH CENTURY. Chap. III. j 3. 



EI, AI — XVI TH Centuey. 

1530. Palsgeayb says: "jE!i* vniversally through out all the 
frenche tong shalbe sounded like as he is with vs in these wordes, 
obey, a sley, a grey, that is to say, the e shall have his distinct 
Bounde, and the t to be sounded shortly and confusely, as comeil^ 
uemmi, and so of all suche other." 

**^f in the frenche tong is sounded lyke as we sounde at/ in 
these wordes in onr tong rayne^ payne, fayne, disdayne, tiiat is to 
say, a, distinctly and the i shortly & concisely." 

The forms ey, ay, are mere varieties of ei, at, and need 
not be separatelv considered. Palsgrave's words ought to 
imply that the l^fflish and French ei, at, were pronounced 
(ei, ai) or else (eei, aai). This is very different from the 
present pronunciation in English, where they are generally 
(ii, €e)y or in Fr^ich, where they are generally (ee, ee) ; 
hence some confirmation is required. 

Meigbei says : '^ Considerons si ai, se treuue tousiours raysoun- 
ablement escrit, de sorte que les deux voyelles soient en la pronon- 
ciation comme nous les voyons en aymant, aydant, hair. H n 'y a 
point de doubte qu'en mais, maistre, aise, vous ny trouuerez aucimes 
nouuelles de la diphthongue ay, mais tant seulement d'yng e qui 
i'appelle 6 ouvert, conune ia i'ay diet. Parquoy telle maniere d'es- 
criture est vicieuse en cenx 1^, et en tons aulres semblables, es quelz 
la prononciation est autre que d' ai : comme tous pouirez cognoistre 
si Tous les paragonez k aydant, aymant, es quelz elle est veritable- 
ment prononc^e. le treuue d'auantage que nous faisons bien 
Bouuent vsuiper k la diphthongue ai la puissance de ei, comme en 
ces Tocables sainct, main, maintenir : es <^uelz sans point de doubte 
nous prononqons la diphthongue ei tout amsi qu'en ceint, ceinture, 
peindre, peinture, meine, emmeine. De sorte que si tu te ioues de 
Touloir. prononcer ai en ceux U, tu seras tronu^ lourd, et de mau- 
naise grace, et auecq aussi bonne rayson q'est le meiiu peuple de 
Paris quant il prononce * main, pain * par ai." 

Again in his phonetic cranmiar, he says — 

'* En comEuqant donq a Qmes qi ont a en t£te, nous En anons m 
En ai ou ay (car je ne fts point de differBn9', Entre 1* i e y GrEc) 
comme payant gajant J^gayant?] ayant .... Or comEUQ* En notre 
lange la £phthonge, ei, par e ouvErt, 8UC9eder a ^slle d' ai En 
aocuns vocables: tidlement qe nou' n'oyons plus dire aymer, si 
souutfnt q' Eymer. Ao regard d* am^, e amez dont no' Mtres de 
comissions sont pleines, V uzaje de V eloqEnqe Fran90Eze Ies a ja de 
si long tEUs cassez, qe ie ne psnse pas q'il se puiss' aoiouidhuy 
trouuer home qi Ies aye yu jam£s En aothorit^, pour Etre commune- 
msnt pronongez d'un bon courtizant." 

These extracts establish a French diphthong (ei, Ei),'it is 
impossible to say which ; and also a French diphthong (ai) 
or (aai), entirely diflferent fix)m the former, but gliding into 



Chap. m. i 3. 



EI, AI — XVI TH CENTURY. 



119 



it, 80 that the pronunciation was then beginning to change, 
and that in several words as mats, maiatre the diphthong (ai) 
had become the simple vowel (ee).^ 

1547. Saxesbury in no place gives an English word which 
he spells with ei^ ey, but as he explains the word vayne 
by the Welsh gtcythen ne tcac, i.e. vena vel mnm, it must be 
held to include both the words vein and vain. He pronoimces 
them both vain = (vain), and hence- makes no difference 
between ei and ai. But he distinguishes both &om long i, 
as he had immediately before written vyne, vein (vein) vitis. 
The following are all Salesbury's words containing ai with 
their pronunciation ; he has no special observations on the 
combination. Quayle has no pronunciation assigned ; naylb 
napl (natl) unguis vel clavus, nayles nayk (natLz) ; rayle 
aryl (rail) cancellus, rayles rat/la (railz), vayne vain (vain) 



^ The work of M. livet^ described 
on p. 33, enables us to confirm this 
Tiew bj the reiy objection which G. 
des Antels opposed to it. '^Anssi 
triomphes-ta de dire," said he to 
Meigret, according to p. 129 of M. 
lireVs book, **qae les diphthongues 
gardent toujonrs en nne syllabe le pro- 
pre et entier son de deux rojelles con- 
jointes; et sont encore pins gaillards 
tes exemples de pay ant et royal ... Je 
te dy done qu'il n'y ha point de diph- 
thongne en ces mots ayant, payant, 
roytUet loyal, mais senlement nne con- 
tracticm, qni encore ne se fait la oil tu 
prends la diphthongue, mais en la 
syllabe soivante, car en ayanty a est 
nne syllabe et yant nne autre imr con- 
traction de deux." On which M. 
Liret remarks : ** Ce passage montre 
assez la pronunciation de ayantypayant, 
qui s'est conserve dans le centre de la 
France et en Anjou. En Picardie, on 
dit aayoU pour giole (di^r^ de gedHjy 
et le oolosse d'osier qu*on prom^e 
dans les rues de Douai sous le nom de 
OayoHy k V 6poque de la Dveassej n'est 
autre que le G^ant, pris absolument. 
Cf. Escalier. Memarqu4a tur Is paioit, 
1 Tol. in-8o, 1866, p. 22." And Pierre 
Ramus (liTet p. 206) gives for ai the 
examples, ^in his orthography, using 
s, tf for his broad and mute e respec- 
tirely) ^ paiant gaiant, aidant,' and ^or 
it, ^fthMie, psindr^ crEindbr^, pnmtf, 
£DntBin#,' where the two last words 
have no suspicion of a nasal vowel. 
On paytr in the xvth century, see 
supra, p. 76. There is a fight between 
Meigret and his opponents respecting 



the mute e, Meigret only admits his 
B, e s= (b, e?) long and short, and 
identifies what G. d^ Autels, Pelletier, 
Ramus, and others, according to Livet's 
language, call the * mute e,' with his 
* short e' (e), livet (p. 133) con- 
cludes : " d'une part que les diff^rents 
sons de Ve ^taient alors ce qu'ils sont 
maintenant) et d' autre part qu*on ne 
s*entendait pas sur la mani^re de les 
noter ou de les nommer." But my 
German experience leads me to a di^• 
ferent conclusion. In the words : eine 
guU Oabe, the final e is pronounced in 
the greater part of Grermany very ob- 
scurely and more like (b), as most 
Englishmen pronounce their final a in 
ChmOy idea, and some their final -er in 
aaiter (which word they then speak 
like a common mid-German mispro- 
nunciation of Ooethe)y than like (#). 
Tet theoretically (#) is held to be the 
sound uttered, and in some parts of the 
Austrian dominions I have heard this 
distinct short final (e), which of course 
had an unpleasant effect on my un- 
accustomed ears. Now it is quite pos- 
sible that Meij^t may have, as an 
older and provincial man, retained the 
clear (0), that his younger opponents 
may have used the obscm^r («), which 
in course of time sank to thepresent {$) 
or entirely disappeared. This theory 
at least accounts for the confiiot of 
opinion, the decided retention of the 
final e in the phonetic writing of Pelle- 
tier and Ramus as well as of Meigret, 
and hence its continued use in the 
poetry of the xyii th century which set 
the rule for French versification. 



120 



EI, AI — XVI TH <:BKTUBT. 



Chap. IIL { 8. 



vena vel vanus. But it is to be obeerved iJiat be pronDinices 
ORANGES oremtsyB (or*eindzbtz), fund that be says that before 
ge, shy tch the sound of '' a is thought to dedine toward the 
sound of the diphthong m» and the wordes" damage^ heritage, 
language, ashe, lashe, waich are '^ to be read in thys wjse, 
domaige, heritaige, langttaige,^ aishe, waiiche.*' We have very 
little trace of this custom left. The unaccented syllables are 
apt to be pronounced with (t) or perhaps (j/), as (or'mdzhtz) 
daem'tdzh, Her'tttdzh, lasq'gwtdzh,) but ash, watch have be- 
come (sBsh, wAtsh), instead of (eesh, weetsh) as might have 
been exp^sted. Salesbury therefore only recognizes the 
diphthong (ai) and does not acknowledge a diphthong (ei) 
as distinct from the r^resentations of long f . Yet long 
i, ei, ai have in subsequent times traversed with different 
velocities three distinct paths ending in (ai, ii, ^) respectively. 
1568. Sib T. Smith says : ''Inter Ai & Ei diphthongos roiniTna 
differentia est, praesertim apud nostrates, apud nos tamen audiantur 
hi soni. (Fein) fingere, (deinti) delicatuB, (peint) pingere, (feint) 
languidos. Sed non hsec tantmn verba per ei prommtiantiir, sed 
ceetera omnia per ai scripta mnlierculse qusBdam delicatiores, et non- 
nulli qui volunt isto modo videri loqui vrbani^s per ei (ei, eei) sonant, 



^ Compare Palfgrare : ^'jUbo all 
wordes in the frenohe toiu^ whicfae in 
writtynf ende in «^ shall in redynff 
and mkyng sounde an i betwe^i a and 
/, as though that a were this diphthong 
mi: as for lamgdgej henUUfe^ tdfg, dmm* 
Widgey boequdge^ apf^rmtUMtdge^ they 
soimde langttaigt^ keriiaiffe, 9aig$y dmm" 
maigs^ boequaige^ mppr$nti$8mige, and so 
of all sudie fyke excepte rage. And 
note that many tymes I fpMie sadie 
nownes whiche hare the t in writdng 
betwene the a and g^ but) whether he be 
written or nat, in redyng or spekrng he 
shalbe sounded, aecordyng as 1 hare 
here shewed by example." M. Ed. Le 
H^richer (Histoire et Glossaire da 
Normand, de 1' Anglais, et de la langne 
Fran9aise, d'apr^ la m^thode hi^o- 
rioue naturelle et ^Umologique, 1862, 
Tol. i. p. 24) entirely misunderstands 
this passage, when he says : ^ C*6tait 
une r^gledu fran^ais, formula d'ail- 
leurs par Palsgrave dans ses BeUnr^ 
eitsement de la langtse frmi^ise, que la 
premidre lettre de Y Alphabet se pro- 
non^ait A et Ai." That M. Le H(hri- 
cher means that Palsgraye asserted 
French A to be (a) or {b\ and thai 
raierally, instead of generally (a), but 
(ai) in a yery limited class of words, 
appears by his next remark : *' Ce der« 



nier son pr^vant en anglais : il 6tatt 
anasi predominant en normand." I^ 
Tery row examples which he cites fair 
sucn an extraordinarr assertitm as the 
last, are fiur horn estaoUshinf the &et 
They are an assertion by Thierry that 
GranviUe was p rono un ced OraimviUe 
by the Normans : that in a MS. of tiw 
xiT th century at ATranches /aire dm- 
ekure rhyme, whereas they may be only 
an assonance as in moaem Spanish: 
that in Hbe xTih century a Caen hice 
has consecntiye lines ending in Imagg 
grtefie gUupe, and that aige, maige, etc. 
were finally written and printed, so 
that a sea song of 01. Basselin has a 
set of rhymes in -aige, the termination 
pointed out by Pahurraye. **C'est 
oette pronondation de P A qui fait une 
des prineipales differences eatre la 
langne des troubadours et oelle des 
trouy^res.*' This assertion must be 
receiyed witii due caution. Mr. W. 
Babington has kindly made inquiries 
formeoi inhabitants of yarious depart- 
ments in Normandy, and none were 
acquainted with an existinj^ pronuncia- 
tion of a as at in any part of the country. 
Hence it must be yery limited in ex- 
tent, and probably comparable to the 
— 1 mentioned aboye p. 76. 



Chap. III. § 8. BI, AI — XVI TH CENTURY. 121 

vt haec ipsa quae nos per ei (ei) Bcribimus, alij sonant et pronuwtiant 
per ttiy tarn aSidihopoi, siunas in Ms dnntaxat dnabus diphthongis 
Angli." 

" Est diphthongas omnis sonns i doabns yocalibns conflatns nt : 
AI, (jMii) solvere, (dai) dies, ^wai) via, (mai) possum, (lai) ponere, 
(sai) dicere, (esai) tentare, (tail) cauda, (Ml) deficere, (faain) libens 
ac volens, (pain) poena, (disdain) dedignor, (claim) vendico, (plai) 
Indere, (arai) vestire sen omare. In his est utraque litera brevis ' 
apnd Trbanius pronimciantes. Bustici utranque ant extremam' 
atdtem literam longam sonantes, pingaem quendam odiosum, et nimis 
adipatom sonum redduitt. (Paai) solvere, (daai) dies, (waai) via, 
(maai) possum, (laai) ponere. Sicut qui valde delicate voces has 
pronuntiant, muHerculse prsBsertim, explicant plan^ Eomanam dipb- 
thongum ae, AE dipbthongus Latina. Pae solvere, dae dies, tcae 
via, nuie possum, lae ponere" = (pee, dee, wee, mee, lee) I suppose, 
since tbe Latin ae bad long been pronounced (ee), as we know, 
among otber reasons from the frequency with which it is written e 
in works before this time. " Scoti et Transtrentani quidam Angli 
voces has per impropnam diphthongum Graecam a profemnt ut nee 
f nee e mm obscurissime' audiatur. A dipbthongus impropri^ Grseca 
(paa,^ daa, waa, maa, laa)." 

Again, in his De recta et emendata lingvee GrsecsB pronuntiatione 
.... ad Yintoniensem Episcopum Epistola, Paris, 1568: **Diph- 
thongi quo modo sonantur dicere in promptu est: Nam si duas 
vocales rect^ prius extuleris, & easdem coniunxeris, diphthongum 
babes, hoc est sonum quendam duplicem ex duobus commixtis inter 
se factum. Yt si nesciam mulsum quid sit, & audiam ex aqua & 
melle factum esse, potero fortassis commiscendo tale quid efficere, 
mel vt sentiatur & aqua ne dispareat. Aut si talem colorem habu- 
isse veteres, qualem viridem appellant, & hunc ex flauo luteove & 
ceruleo fuirae confectum, potero credo commiscendo videre, cuius- 
modi sit illud quod imitari cupiam, vt nee alteram ab altero colorem 
prorsus extinctum & obHteratum relinquam, & tamen vtrunque 
pariter in tertio conspici ac relucere faciam. Sed, diphthongi quo 
modo sonari debent, quivis etiam ex triuio puer qui literas didicent 
explicabit. Heus tu die sodes, a & i quid faciunt ? dicet cert^ at, 
04. 8i p pneponas, facit pai, irai, eolue. sin m, mat, fuu, MsCiiia 
mensis: sin tr, waiy oval, via; neque nunc pa i dicit, nee fna 
I, sed pai & mai, vt constituere diphthongos non dissoluere videator. 
Idem dicendum pnto & de ei, quod nos exprimimus cum hinnire, 
hoc est ney dicimus : & foeminffi qusedam delicatiores cuncta fere 
qusB per ay dicimtur per ei exprimunt : vt wey, dey, pei, vt eadem 
Eurosaxones populares mei rusticiores, nimis pingui et adipato 
s<wio, way, day, pay : vt etiam tinnitus illud i reddat in fine. Scoti 
& Borei quidem Angli per a, vix vt illud i audiatur, pa, da, wa, aut 

^ In one case (faain) be has marked examples be shews that the sonnd was 

the vowel as long ; peihaps a misprint, not heard at all. The present sound 

* Meaning the first element F is faa'), see chapter XI. 

* An or^oepical safegaarcL In his ^ Pay is now called (paa) in Norfolk. 



122 EI, AI — XVI TH CEKTUKY. Chap. III. § 8. 

potins per ae profenmt. mud obseraandum ne nimis Tideamur 
obes^ loqni propter exilissiinse liters prope latissimas ex breuibus 
TiimiTiTn tiimientis sontun, ci!mi ai & oi dictionem finiant, breuiter & 
corrept^ proferendas esse : quod Greeci Grammatiei notanmt, ne 
alioqni crassum ilium & adipatuin sontun rusticorum nostratixiin 
imitemiiry qui ciim a gay, hoy, ore pleno Uteris diductis in immen- 
sum dicunt, nimis prefects invrban^ loqui ab elegantionibus iu- 
dicemur." 

It would seem that Smith's fei) were precisely the same as 
his lonff 1, and that as a general rule, /, eye, aye were pro- 
nounced alike. Yet the two sounds (ei, ai) were recognized 
also as different, and (ei) was considered to be a dainty 
effeminate pronunciation of (ai), which when urged to excess, 
through (eei), merged into (ee), but of this mincing sound he 
decid^y disapproyed. This change makes it probable that 
eye and therefore long f was rather pronounced (ei) than (ai), 
because although (ei) could easily become (eei) and thence 
(ee), the course from (oi) to (ee) does not seem so straight. 
The sound of- (ai) has not yet disappeared in our provinces. 
I have frequently heard (dai, wai) or even (daai, waai) used 
by rustics. Smith seems decidedly to disapprove of this 
lengthening of the first vowel, which however is not un- 
common in GilL 

1569. Hart in the very next year after Smith had repro- 
bated the use of (ee) for (ai), published his treatise, in which 
he invariably uses (ee), and does not even give (ai) in his 
enumeration of diphthongs. In his French Lord's Prayer 
he transcribes faite as (feetan), which agrees with Meigret's 
(f£Et^). It was Hart's English use of (ee) for (ai) that 
especially excited the ire of Dr. (Jill. 

"lUe," says Dr. Gill speaking of Hart, "praeterquam qu6d 
nonntdlas literas ad vsum pemecessarias omisit, sermonem nosbimi 
cbaracteribus suis non sequi, sed ducere meditabatur. Multa 
omitto. Keque enim bene facta malign^ Detrectare, meum est: 
tamen haec paucula adnoto, ne me homini probo falsum crimen 
affinxisse putes. Emendate nostro charactere vtromque leges, quia 
de sono tanti!im certamen est.^ Sic igitur ille, folio 66, b. 
Pre\ /prai sed > /said iu \ 

ue 1 ( wai ei 1 I ei iuz* ) 

se f J sai ov > pro < of uii > pro 

dhe ( i dhei aunsuer i I answer uidh I 
bue 1 [ buoi riiding / V reeding knoon / 
me / Vmai 
Non nostras hie voces babes, sed Mopsarum fictitias." 

^ For the same reason, and also for ' 6iU has here mistaken Hart's sini 
greater ease to the reader, Gill's sym- which was meant for (yyz), as wiU oe 
Dols are here replaced hy palaeotype. shewn under U helow. 




Chap. III.*§ 3. El, AI — XVI TH CENTUKY. 123 

The withering character of this denimciatloii will be well 
understood by referring to the passac^ quoted above, p. 91, 
where he reproaches tne " Mopseys with saying (meedz, 

Slee) for (maidz, plai), although Gill himself writes (reseev, 
eseev) in place of (reseiv, deseiv)* receive^ deceive, which is 
a change in the same direction. After this expression of 
opinion by Dr. GiU it is impossible to accept Hart's pro- 
nunciation as that generally used in his time, though it is 
evidence of an existing pronunciation, then only patronized 
by a few, but becoming \dtimately dominant. 

1580. BuixoKAE says : " that there be seuen diphthongs of 
seuerall notes in voice, and differing from the notes of euery of the 
eight vowels aforesaide,^ may app^ere by the wordes following — 

a hay or net: in Latiae, Plaga, Italian, Hete da piff^iar animaU 

saluatichif French, Bouroettea a chasser. 
hey : in Latine, foenum, Italian, Fieno, French, Du foin. 
a boy : in Latine, Puer^ Italian, Oarzone, French, Oarson. 
a hw/* that is fastened to an anker with a rope to weigh the anker : 

in Italian, Anmnare. 
a hOfUf* in the eie : in Latine, Vnyuis, French, Faille, 
^Ua he,u smaller: in Latine, Concidere, Italian, Tagliare minuta- 

tnente, French, Hacher menu, 
a how : in Latine, Areue, Italian, Areo da saeiture, French, Are.** 

These diphthongs I read (ai, ei, oi, uui, au, eu, oou) of 
which the two last will be elsewhere considered, and (uui) 
is only a variety of (oi). Bullokar consistently uses (ei, 
ai) for ei, ai, thus (<Uiei konseiv) would be quite distinct 
from (dhai konsaiv*) which the modem English ear hears 
as (dhei konsaiv).* 

1621. Gill distinguishes (ei, eei, ai, aai), but he is not 
very certain in the use of (ai, aai). I find the following 
words in Gill's phonetic transcriptions. 

ei (ei) eye, (eiz) eyes, (eidher^ either, (valleiz)* valleys, — freseev) 
receive, (deseev, deeseev) deceive. — (dheei) they, (dheeir) 
their, (reeineth) reigneth.* 

1 See p. 64. derstood. Few English obserre the 

* The « is in Bnllokar a new letter pecnliar Scotch (ei^ for (ai). They at 
made by the union of the two oo, most take it for a Scotch way of saying 

* The comma before u and inyerted (ai), but recognize the latter diph- 
TOOBtrophe before t are printed under thonff. 

the letters in Bullokar, to indicate, ^ It is not to be supposed that 

first that u has the sound (u) or (u)f (yal'leiz) was meant, and not (yal'eiz}, 

and secondly that ,t« is the preposition, but in transcribing, I haye thought it 

* Falmeira Square at Brighton is al- best to giye Gill's own forms, howeyer 
ways called (FslmdiiTa), and thus careless and irre^iilar they may be at 
oomused with Palmyra, the original times. Corrections must be always 
Portuguese (Palmei-ra) not being un- theoretical. 



124 EI, Ai — xrn th gestubt. Chap. m. i 3. 

ai {wta^ waai) way, (nud, ntasi) mBy, (an, aaai) say, (pnd2> praaiz) 
praise, (alai) allay, (wait) wait, (slaia) alam, (sndain) iodam 
old fani ofnuUem; (daai) day, (kkai) day, (Tetaain) retam. 

1623. Butler says (uamg the odmnion orthography) : 

" The rig^t somid of mi, am, ei, em, oi^ ou; is the mixed sound of 
the two Towels, whereof they are made : as (hait, vaat, Hei, Heu, 
koi, koa) : no otherwise than it is in the Gredc" 

This might lead to (ai, aa, ai, en, oi, on), but it is im- 
possible to say exactly how Butler pronomioed Gred: av, u. 
Sir T. Smith's pronmiciatioii of the QreA diphthongs €u, u, 
Of, av, eu, ffUt ou, w, t» seems to hare been decidedly (ai, ei, 
oi, an, en, eeo, on, ooo, wi wei). 

'* Bnt ai in imitation of the French, is sometime cormptly sonnded 
like « ; as in suiy, fMy, jp%, jwnay, #ay, 9Unf,firap^ day : flpecially 
in words origioally French, as in jMijr, haili, tratmH : thon^ plaii 
have lost his natural orthogiBphy, and we wiite as we ^eak plsmd 
(pleed)." 

This implies that tiiongh some q»eaker8 insisted on pre- 
serving (ai) in these words, (ee) was the most general pro- 
nunciation, — ^which may seem a curious interpretation of 
^ sometimes corruptly," but allowance must be made for the 
mode in which orthoepists speak of common pronunciations 
which differ from their own, or from what they reconmiend, 
— ^by no means always the same thing. 

EI, AI — xm TH Century. 

1653. Wallis tells us that ei, ey, were (ei) or even simply 
(ee) without the (i), but adds ''Nonnulli tamen plenius 
efferunt, acsi per at scripta essent.'' The diphthong oi he 
upholds still as a dqihthong, *^ Ai vel ay sonum exprimunt 
compositum ex a Anglico (hoc est, exili) correpto, et y. lit 
in voce day dies, praise laus," which, if our interpretation 
of Wallis's a be correct is (dari, praeiz) v«y slightly different 
from (dassei, prseeeiz) and readily passing into (deei, preeiz) 
which is almost the sound of the present day. But the real 
transition was into (ee, ee), as we shall learn from Cooper. 

1668. WiLKTNS writes, (deei) day, (daeili) daily, (aeaeinst) 
against, (ssDints) saints, preserving the diphthong like Wallis, 
but has (kAuseevd) conceived, dropping the (t) entirely. 

1668. Price in the same year apparently agrees with the 
other two. He divides dqihthongs, or, as he spalls the word, 
" dipthongs,'^ into two classes, proper and improper : 

'* That is a proper dipthong wherein both vowels keep their 
sound. There are twelve proper dipthongs, ay ejf oy^ ai eif oi, 
aw ew oWf aueu w," 



Chap. HI. i 3. EI, AI — XVUTH CBNTUKY. 125 

wMch practically reduce to six, ai ei oi, au eu ou, and as we 
know that in ai both vowels kept their sounds, we should 
conclude that the vowels in the other two diphthongs did 
so too. 

''That is an improper dipthong that loseth the sound of one 
voweL There are eight improper dipthongs, ea ee ie eo, ea oo t*», 
ou obscure as in cousin.*^ 

Then, after giving a list of words in ai, comes the question, 
"Doth a-i always keep its sound?'* the hyphen seeming to 
imply separation. The answer is 

**Ai Boundes like e in bargain, chaplam, against, chamh^rlam, 
curtain, plaited^ raisin, travail, wainscot. ^^ 

This is therefore an exceptional list of words in which of 
= (ee), and hence implies that ^nerally, and in all other 
words ai = (aei), with the (sb) of the period. Again he says : 

" ^ sounds like, ay, in thsy, ohey, convey, conveyance, oheysanoe, 
jfvjK (or spoil), survey, surveyor, whey, but ey soundes ♦ (9i) in eye, 
0yM," and " M soundes like ay in heir, feign, weight, neigMnnw, 
deign, eight, f6rein, inveigh, to neigh, streight, streighten, veins,^* 

Now when it is remembered that these lists of words are 
opposed to those in which ey, ei have the sounds of (ee, e, i) 
it is evident that the general soimd of ai was still (sei), 
although it had become (ee) in a few words cited, and that 
ey in the above lists was (sei). 

" Ey soundes like ee (i) in valley, Turkey, barley, monkey, parsley, 
iaUey, Umsey.*^ " Ey sounds e (e) in countrey, attumey, abbey, alle^, 
Anglesey, causey, chimney, cockney, comfrey, lEackney, journey, a 
Grey, key, kidney, lamprey, money, pulleyJ^ 

It is doubtful for how long the short (e) in these words 
kept its place, and whether the final imaccented (e) and (i) in 
these two lists were ever kept very clearly separated. The 
long hey=^ (kee) remained for sometime, and (diould be con- 
sidered as belonging to the next list. 

" Ei soundes e long (ee) in receive, carreir, conceit, deceit, deceive, 
enterfeir, either, heifer, leisure, neighbour, purveigh, recent, seiu.^^ 

Man^ of these words are now sjpelt differently. Usage 
differs m leisure (lezhu, lii'zhj) and in either (ii'dhj, oi-dha). 

1685. CooFEB begins to recognizes ai as (ee) though he is 
not quite consistent with himself. After describing (b) he 
says: 

'^ Vera hujusce soni productio scnbitur per a, B,\x\ue a longum 
fidsb denominatur, ut in cane camia .... hic sonus, quando pur^ 
8(matur," that is when it is not (ee9), '^scnbitur per ai vel ay; 
ut pain dolor, day dies ; quae hoc mode in omnibus fere dictionibus 
plerumqiM pronunciantur : per ey in convey deporto, obey obedio, 



126 EI, AI — XVn TH CEKTURY. Chap. III. § 8. 

purvey rebus necessariis provideo, mrvey lustro, they illi, trey tmlla, 
tohey serum lactis: quandoque raro autem per ea ; ut pearl margaiita. 
Corripitiir in Prodncitur in 

sell vendo sail navigo 

sent missus saint sanctus 

teU nuncio tail cauda 

tent tentorium taint inficio." 

This makes ai (ee) except in a few words. But afterwards 
he says : 

^^Ai leniAs prolata sonatur ut a in cane ; fortius, plenum assumit 
sonum dipthongi ai ; ut hrain cerebrum, frail fiigilis ; ay finalis 
ut a, sic day dies ; ai ante r scribitur pro a in affairs res, air-y 
aereus, dairy lactarium, debonair candidus, despair despero, fair 
pulcher, fairy lamia, hair crinis, pair par, repair reparo, stairs 

scala; caetera cum are; ut are sunt,* dare audeo Aim 

bargain pactum, captain dux, certain certus, chaplain capellanus, 
curtain velum, forrain extraneus, fountain fons, mountain mons, 
villain furcifer, & prior ai in maintain sonatur ut a correptum 
sive e breve." Again he says : " Sonus a in I can possum ; I cast 
jacio ; conjunctus cum i sonum litersB ee exprimente ; oonstituit 
dipthongum in bait esca ; caitiff homo improbus ; ay pro / vel yea 
imo ; & eight quam vulgariter pronunciamus ait. Plures baud scio." 
This must be (aei) ; he seems to have thought of brain and frail 
afterwards. Then he adds : " JB in ken, vel a in Cane i praepositus 
diphthongum {m) priori (sei) subtiliorem constituit; ut praise laus: in 
paucis scribimus ei vel ey finalem ; ut height altitudo ; weight pondus, 
& convey deporto, aliaque qusB supra sub e ostendimus; quibus 
exceptis cffitera scribuntur cum ai vel ay ut hainous detestabilis, 
plerunq«« autem in coUoquio familiari, negligenter loquentes pro- 
nunciant ai prout a simplicem (ee) in CaneJ'^ 

Hence we may collect, that in the very few words baitf 
caitiffs ay, eight, brain, frail, Cooper still admitted the diph- 
thong (sei), and that ne also endeavoured to establish a 
diphthong (si) or (E£i), but that he was obliged to own that 
the generality of words written ai or d were then (be) 
or (ee). 

1688. MiEGE, writing nearly at the same time as Cooper, 
heard long a as French {ai), supri p. 71, and of Ai he says 

" cette diphthongue a le m^me son en' Anglois qu'en ces Mots 
FranQois, fai^-e, taire, &c. Exemple, fair, despair, hair, repair, 
airy, dairy. Ten excepte, 1. Les Mots finissans en ain, oA Vai se 
prononce ^ la Eran9oise, comme en ces Mots, villain, certain, &c. 
2. Raisins, qu'il faut prononcer Rhins^ 

Although his French ai seemed in the first place to 
imply English (sesB), it can be hardly other than (ee) in the 

1 This is peculiar, but still heard, in the form (eei). 



Chap. III. § 3. ET, Al — XVII TH CENTURY. 127 

present. Frenchmen do not generally distinguish these two 
related sounds, as they are unacquainted with English (aesD). 
Similarly Englishmen hear French (ee) as their own {ee). 
The meaning of the first exception is not very clear, because 
the French pronunciation of French final ^am is uncertain. 
Nothing can be clearer than that Englishmen never pro- 
noimced their final -ain as (-ca). Did the French sajr (-ein)P 
Miege says that n final is pronounced, " d'une maniSre plus 
forte en Anglois qu'en Fran9ois," and this is his only allusion 
to what is now the French nasal. Was the English (vil"Bn, 
si't^n), or (vil'yn, si'tyn), as at present ? We cannot learn 
from this passage, but it is probable that (vilen, ser'ten) 
represent the sounds with sufficient exactness. The ^ mascu- 
line in rdzinSj evidently implies (reez'inz) or (re^z'inz). The 
distinction here made between (ee) and (ee) or {ee), though 
real enough in French, is probably due only to insufficient 
observation or appreciation of the English sounds, and cannot 
be insisted on. 

"EI. Cette Diphtongne se prononce en Anglois coimne en 
firanqois. Exemple vein une veine, weight, un poids " (vein, weit ; 
veein, weeit) ? " Excepts 1. ces Mots oil elle soune. comine un e 
masculin, ou i, Savoir to conceive, deceive, perceive, receive, seize, 
inveigh, leisure, & leurs Derivatifs" (konseev, des^«v) &c.? "2. 
Ceux-ci, oA la Diphtongue prend le Son d'un e feminin. Savoir 
forfeit, foreign, surfeit, heifer, either, neither, '^^ (for'fat, for'an, sarfat, 
Hof-ar, adhar, nadh-ar)? "3. Ce Mot height, qui se prononce 
halt," (nait). This should be (HAit) according to Miege's custom 
of confttsing (a) with French a, and according to other authorities 
it should be (nseit). We have still a double pronunciation {neet, 
nait). 

1701. Jones seems not to have made up his mind entirely 
that ai was to be pronounced as (ee). Thus he says that the 
soimd of ai (whatever it may be) is written ei in 12 words, 
blein, conceit, deceit^ distrein, heifer, Aeinot^^ heir, reins, their, 
veil, vein, weif; and eipn in 5 words, darreign, deign, feign, 
reign, sovereign (" or soveraign") ; and eigh in 12 words, con" 
veigA, eight, freight, heigh I height, inveigh, neigh, neighbour, 
ptirveigh, straight, surveigh, weigh, and their derivatives, as 
eighteen, weight, etc., and eip "in receipt sounded resait,^* and 
es " in demesn sounded demain,^^ and eg in 12 monosyllables 
brey, Grey, grey, hey I key, prey, Sey, sey, they, trey, Wev 
(a River), wney, and their derivatives as breying, Weymouth, 
etc. It is to be observed that he never asks when is the 
sound of ai written e, that is (ee) ? 

He next says the sound of e is written ai, " when it may 
be Boimded ai,*^ which should imply that the soimd of e was 



128 EI, AI — XVnTH CENTUBY. Chap. IIL § 8. 

different from that of oi, "as in abipail^ afraid^ again^ 
against^ bargain^ capstain, captain, certain, chamberlain, chap- 
lain, complaisant^ curtain, debonair, hainaus, mountain, mur- 
rain, Prestain, raisin, said, Suis {?), suddain, vervain, mllain*' 
adding, '' see a — ai." He also says the sound of e may be 
written ay " when it may be soimded ay in the end of words 
or before a yowel ; as decay, decat/in^^ etc/' These expres- 
sions ought to imply that Jones distmguished the sounds of 
ai, e, but whether as (ei, ee) or (asi, ee) cannot be collected. 

But the above conclusion is not certain, for he says that 
the sound of d is written mg " in these six, darreign, deign, 
feign, reign. Seignior (sounded senior), sovereign,'* five of 
which darreign, deign, feign, reign, and sovereign are the five 
in which the sound of oi is said to be spelled eign. This 
would shew that these words were pronounced both ways, in 
accordance with Jones's custom of giving both ways, of 
pronoimcing. In reply to the question, when is the sound of 
e written eigh f he says, " see ai — eigh ; where jrou have all 
such," so that these words had also both pronunciations. 

Jones says the sound of e (e) is written « in 30 words atheist,^ 
atheism,^ conceit, conceive, counterfeit, deceit, deceive, ddty^ 
disseise, disseisin, either, forfeit, heifer, heinous, heir, inveigle, 
leisure, Marseilles, *neigh, ^neighbour, neither, perceive, re^ 
cme, receipt, seise, seisin, seive, surfeit, teirce, their. Those 
marked with * are in a previous list giving the soimd of ai, 
shewing again that the sounds of ai, e, if different, were at 
least frequently confrised. He also savs that Leicester was 
pronounced Lester, and ^ves a list of 32 proper names as 
Anglesey, Awbrey, etc., m which eg final had the soimd of 
e (e), and of 39 other words with ey final having the same 
soimd (e), some of which are words in which eigh was said to 
have the sound of ai, and others are words to which Price 
ffave the sound of (i) ; thev are abbey, alley ^ attumey, barley, 
orey^ causey^ chimney, cockney, coney, convey, cumfrey, grey, 
hackney, hey-dey I honey, journey ^ invey, key^ kidney, lackey^ 
lamprey, medley, money, monkey, obey^ parlev, parsley, prey^ 
pulley, purvey, sey, survey, tatley, tansey, they, trey, turkey, 
valley, whey. In answer to the question when is the soimd 
of ee (ii) written ei ? He replies, sternly, " Never.*' And 
adds, " Note then that it is ie not ei, which often sounds ee ; 
as in field, siege, etc.'* We may therefore conclude that ei, 
ey were always (ee) and never (ii^ ; although ai, being 
generally (aei) or (ei) was sometimes (ee). 

1 Theie nrast be meant 'to include << This diphthong #• is parted in tiheift, 
erroneoni pronunciations. Price says : athMm, dntii, polythntm* 



Chip. in. i a. m, AI — XVUITH CEHrtURY. 129 



EI, AI — xviHTH Centuky. 

Ei, (d fleem to liave remained at (ee) during most of the 
xvni th century ; at least at was fixed in that sound and has 
come down to us with the dight alteration into (ee), al&ough, 
in the south of England, (eei) is more commonly heard. 

1704. The Expebt Osthooeafhist says that " ai, ei, ay, 
ey are much the same sound, in many words, as pail, pay, 
ewht, they,** but gives a list of 11 words in which ** the sound 
of e is lengthened by ei** that is, in which ei is pronounced 
(ii) contrary to the express "never" of Dr. 'Jones; they are 
amceit s. and v., concewe, deceit, decdoe, dlher, inveigle, re- 
C^t, receive, weild now fvield. It is curious tiiat while he 
gives (ii) to conceit spejlei thus, he admits (ee), or raiher, 
^ the sormd of ai,** as the sound of d in " eon, de, re, eeivt 
or ceioe, heir, leisure, neither,^ rein, reign, their, vein, heiant, 
inveigh, neighbour, weight" He did not really distinguish ai 
from a long (ee) as may be seen under A, p. 74. 

1766. BucHAiTAK writes (faein) feign, (oobee*) obey, (slee) sleigh, 
(gree) grey, (leezvar) leisure, (nee'bar) neighbour, (tnvee*) inveigh 
fparvee*) purvey ; — (persiiv) perceive, (diisiiv) deceive, (siiz) seize, 
(invii-g'l) inveigle; (flet^r-ni) attorney, (kan-tn) counta^y, (ael't) 
alley, (kAA-8») causey causeway, (teen-si) tansey, (fai'fit) forfeit. 

Also- (reen) rain, (pee) pay, (eegeenst) against, (ree'stn) raisin, 
(ween'skot) wainscot, (baeaer'gm) bargain, (tshseaem'btrltn) chamber- 
lain, (kar'tm) curtain, (tnevil) travfuL 

Except then in very few words the usages are those of the 
present day. 

1768. FaiKKLiN has: (steens) stains, (reens) rains, (feer) fSedr, 
(asarteen) ascertain, (ateen) attain^ (ansarteen) uncertain. 

Also (dher, dheer, dhaer), their, (dhee) they; (aidher) either, 
and (f&renarz) foreigners. 

1780. Sheridan in his remarks on the Irish 'pronunciation (diseet*, 
riseev') deceii, receive, which belongs to the xvnth century, notes 
that ^'the Iri^ in attempting to pronounce like the Engli^," and 
to convert their ei, ey into (ii), oft^ overstrained the rule, and said 
(prii, kAnvii') prey^ convey ; this was simply an error of the same 
kind as that noticed above, p. 92. 

Hence in the xvith century we may assume ei, ai, to be 
(ei eei, ai aai) ; in the xvii th (ei eei ee, sei ee) and in 
the xvnith (ee ii ei, ee). But in the xviith century 
both ei, ai were apt to be confused with one another 
and wiUi long e under the common sound of (ee). Also 

^ Yet he writes (iidh*er). Tbii re- •oiLdoToaiay^imdh'er)or (iioUBi*er)f" 
nidi m of the questioii and answer ^* (Neeoh^, sir." 
^ai $mi lM b^ if not W4t{), " Dr. Mm- 



130 OI — XVI TH CBNTUBY. Chap. HI. i 8. 

even in the xvith century a krge section of people 
pronoonced (ai) as (ee), but thia, tliough adopted by 
Marty was thought effeminate by Sir T. omith and Dr. 
GHll. It however allowed Shakspere to pun on reasons 
and rcdsins and on here, heir (suprjt p. 80 note). 

01 — XVI TH Cbntury. 

1530. Fai^gbate says : ^^Ot in the frenche tonge hath .ii. dinerse 
soundes, for sometyme it is sounded lyke as we sounde oy in these 
words, a hoye^ a froyse, coye, aud suche lyke, and som^rme they 
sound the t of oy aknost like an a." 

1545. Meigbei says: '^En moms, royal, loyal, nous oyons eui- 
demment en la prolation la diphthongue commencer par o & finir 
par i. An contraire en moy, toy, soy, nous oyons la fin de la 
diphthongue, non sexdement en e, mais encore, en 6 ouuert, qui est 
moien entre a & e clos, & par consequence bieu estrange de la pro* 
nonciation de I'i, ou y grec. Nous escrirons doncq* lo6, ro6 et 
loyal, royal." And 1550, in his Grammar he says that " ao regard 
de I'o ouuErt 0. ne fet point de diphthonge preqedant Pa, pas qe 
j'aye decouuErt; ne parelLement auQq Fe clos: nix's ioint a Fe 
ouuert il Est fort frequEut eu la pronon9fa9fon Eran90Eze, qoE qe la 

Slume n*En neyt jamEs fet conte, vzant qelqefoES (come j'ey ja dit) 
e la diphthonge, oy, es aocuns dss vocables : come, moy, toy, soy, 
loy, foy : pour moE, toE, soe, Ioe, foE, qelqefoEs aosi pour fsa:^ Eucor 
p(s, il' luy ont ajout^ vne s, ; come, cognoistre pour conoEtre. b 
non contans de ^ete lourdeHe, qazi come tumbans de fieur' eu ^hi^)' 
mal, il' nous ont introduit oient pour oe', e' tierses pEisones plurieres 
du preterit impErfet: ecriuans estoyent, disoient, venoyent, pour 
etoie/t, dizoE^t, venoB't." 

It was this broad (e) which Palsgrave apparently con- 
founded with (a), and indeed we are told that in Parisian pro- 
nimciation it was already sometimes (a).^ Even now the ai is 

^^ Heigreffl analysifl of the French mant le son o, prononoent seulement 

diphthong ai s (ob) Ib oonfinned by «i : ainsi lea Normands 6criTent et pro- 

Pelletier, who writes (liyet, p. 174) nonoent /m, pour foi, et le peuple 

*FranQ0B8, disoBt, connoBlTa,' but parisien dit parlet^ alUtf venet poor 

* point, ToyBltf/ Bamns (ib. 206) parMt^ alloitf vmoit ; lea imitatenrs de 

writes *moB, Ion* for moi, hi. Beza de V italien prononoent de mdme 

fib. 622) is fuller and says : " cette AngU»^ Franekt^ Eeottka ponr AngUn%^ 

oipthongae fiedt entendre k la fois, mais JV-atif ot«, Beouoit, — Une (ante trds- 

rapidementy le son de Vo et de T i^ grande des Parisiens c'est de prononoer 

nod elle est soiyie de m, oomme Unn^ voirre (on verre)^ foirre palba, troitf 

in, issmoiuj mots que quelquesons oomme tfoarrej foarre, troat ou mSme 

tenninen<^ k tort, par un p. — Non suiyie trot,*' This last passage maj be com- 

de fi, la diphthongue oi prend une pro- pared with Gill's denunciation of the 

nondation yoisine de celle de la trip- Mopseys, p. 90. The two passages 

thongue oai on de la diphthongue ai shew how carefol we should be not to 

ou e ouyert; il a le son oai dans hi, stigmatize a pronunciation as fiiiultj, 

moiy /oi qu'on trouye souyent 6crity k when it differs from what we hold best, 

tori^ ayeo un y: quelques-uns, luppii- aa tiie firalts of one oentorj beocmia the 



Chap. III. § 3. OI — XVI TH CENTURY. 131 

acknowledged to be (oe) or (qb) by eminent French orthoepists, 
though it is generally admitted to be (na, ua). After a con- 
sonant the r^ effect of ot , at present, is generally to labialise 
that consonant and sulig'oin (a, a), as roi, lot (rtra, Itra), where 
the ordinary Englishman is apt to hear (rwAA, Iwaa), and in 
the cry fnve k rai, he often falls into (viiv lo taa). I have 
elsewhere given my reasons for supposing that the original 
diphthong from which the modem EngUsh (oi) descended, 
was (ui).^ In the French language, the intention of insert- 
ing before a Latin ^, as in rat, hi from rex, kx seems to 
have been to indicate a thickening or labialisation of the pre- 
ceding consonant, as opposed to the thinning or palatisauon, 
which would have been naturally occasioned by the following 
palatal vowel. Its use was much the same as the inserted 
u after g in French and Spanish before t, to prevent the 
palatisation of (g) into (zh) or (x), but whereas in the latter 
case,asintheuseof ^Aimder similar circumstances in Italian, 
the (r) was generally, not always, kept pure, in the former 
case uie labisd effect became finally constant. 

In Palsgrave's time the Enffliish oi must px>bablv be as- 
sumed as (oi) or (oe), the latter being a diphthong still found 
in Welsh oedd (oedh). The stress was, as usual, on the first 
element, and the apparent stress on I^q second element in 
modem French is due to the real absoiption of the first 
element by the labialized consonant. 

1647. Salesbury recognizes the diphthong cy solely by 
transcribing ioynt into Uioynt, meaning (dzhomt). 

1568. Snt T. Smeth says : " 01 per o hreuem (o) & i (i). Diph- 
thongoB Oi, vt GalHs frequentissima, ita nobis est rarissinia : habemus 
tamen & banc sonum (Coit) iacere discmn, (boi) puer, (toi) ludi- 
cnun, (toil, tannoil) laborare, (foil) bractea, (soil) solum, (koil) 
verbcnrare, (broil) assare in craticula, & (point) quse vox mncronem, 
et indice monstrare, et ligolam nobis notot, & (koi) qnibus ineptum 
et k faTniliaritate alienum signifieavimus. In IiiIb, propter breuitisitem 

reeetred mages of another. Beza's Spanish ne were mutations of the Latin 

r^rohation of the Parisian oa for oai, o, p. 138, note. It is worth noticing 

that isy pt, explains the last words of in reference to Heigref s oUf considerea 

Palsgraye, hut his supposition that the as o ela$j that Beza proceeds to say : 

Norman /ay resulted, like the nsnal " cette diphthongne on a nn son propre 

French at in the words cited, from the qui tient de To et de Yu, II mat se 

rejection of the prefixed o, does not garder de prononcer comme k Lyon pu 

i historicaU]^ correct, as this ortho- pour o (comme nout pour mm), et 

' 7» ^^■fiy^ ^ ^^ ^^^ ^ Norman comme dans le Danphinl et la Savoie 

11. We shaU have to consider o ponr ou : tels eop ponr am^, m pour 

this point in Chap. V, } 1, No. 8, where out etc." 

the Norman m and French i>t=(ei. ne) ^ Transactions of the Philological 

win he considered as mutations of the Society, 1867, Part J, On the JDtph^ 

Latin $, preeiiely as the French eu and thong OT, p. 69, hottom. 



132 OI — XVI TH CENTtJEY. Chap. IlL J 8. 

wmi, M qina hnms o mm nraltiUM ab u difiert, et propterea fait i 
Gi8EM3B dicta o fwcpop. Potent' fortaase k qmbnadam iudicaii hesc 
meliOB poflse per vi descrilxL Yidemns ei yeteres wMi & voiiris per 
acz^wusey qiue poatmoree p^ vuUis & vestrii acr^eemnt CetiA 
9om aepiaaim^ vanant. At flDqxnun est Bcnpturain sonoe aequi vt 
pictaram coipasi testenturqiie acriptuxe aonom flBtatia, yt aulaea 
Ibnnaa yeatxam. ^ JHphihomgu9 impropri^ Graea A:pxid noa in- 
cognita eat. Scoti tamen qiUB noa per oi acrihiiima per ^ promin- 
dant, yix yt t audiatnx." 

And in Ida Greek prontinciatioii be aaya : ** oi, Eefenmns noa 
cilon puemm a hoy dicimna, & cum hidicrom a toy^ & delicatiorem 
bominem yocant Borei noatratea nym & eoy, finequentator bsec dipb- 
tbongoa i Gallis plnrimilun, quorum lingua ci!km elegantiarnm 
atudiosia apud noabxM ysqueadeo placeat, miror ab bis qui bunc 
aonum tarn contemptim aspemantur n<m invrbauiHwmam iudicarL 
Hi chm, yolunt me^ te, iaeitum^ fidem dicere moy^ toy^ eoy, foy dicunt : 
e^que KcHinam Scytbica Danorum geits partem occuparint Galliae, 
ft quod in Grsecia Turci, iam in GalHa fecerunt, yt liugnam Galli- 
cam ynddiaceiefit, & peruera^ commutarent nunquam tamen potenuft 
effugere Normani, quin ai nunc quisquam eorum ruaticior pro moy, 
Uy^ eoy, /oy^ quod non rar6 ^lenit, my, ty, hy, fy, dicat, irri- 
d^EUtur k cfl^^eris GalHa, & non yrban^ ac ciyilit^ , aed inscit^ ac 
rustled loqui existimetur." 

We baye tberefore evidence tbat Sir T. Smitb beard little 
if any difference between (oi, ui), aa be doubted wbicb would 
be tbe beat ortbograpby. In tbe next cbapter fbrtber reaacma 
will be given for auppoaing (ui) to bave been tbe older form. 

1569. HABT^a views of dipbtbongs are ratber peculiar, 
owinp; to bia considering (j, w) aa tbe pure vowds (i, u) 
fonning a dipbtbong witb tbe following vowel, so tbat to 
understand bis account of oi it will be conveni^ to cite bis 
description of dipbtbongs at lengib. He says : 

'^ l^ow will I abew you examples of tbe Dipbtbongs made of two 
dunt yowels, and of otbera of one abort and ^ another long. And 
tlien of tiipbtbongs. Witb sbort vowels, as tbus, (ui uil reid bei 
ionder uel% nueer dbe uat uas uelneer taakn bei dbe iuq Hound) 
wbicb is written for [we wyll ride by yonder well wbere tbe Wat 
waa wel neare taken by tiie yong bound] wbicb doe come very 
oft^ in our speacb. Of dipbtbongs whereof one yowell is abort, 
and tbe other long as (iuu ueer uaakiq iu dbe fouurth touur, bueer 
az dbe buee did pouur uaater upon* dbe nueet flouur) which I 
write for [you were waking in the fowerth tower, when as the 
boye did poure water vppon the wheate flower] which also doe 
come verie often. And for triphthongs as (bi ueiz oy dbe Hueix 
buei) for, n)e wise of the hoyes bowyj. And (nark dbe kat duutb 
mieu Hueilz iuu milk dbe ieu) for [baik the Cat doth mewe, wbilea 

1 ETidai«l7«hci«iiaiii]fpimcteataonh«»,il ihould be ^ o /Mjcp^, potexft" 



Chap. HI. § 8. oi — XVH TH CBNTUBY. 133 

yon milke the yowe]. And a Basin and (eanr), for, [eawer]) and 
certaine others as will be s^ene hereafter. And for three vowels 
comming togither, and making two Billables as in example (dhe 
vyy,er seeth syy,er it is pyy,er) far [the vewer sayth sure it is 
pure]" where, as will be explained hereafter. Hart writes (in) for 
(yy), " and as in these wordes (dhis bei,er 12 Hei,er or pou,er dhen 
dhe dei,er bei mz fei,er). Eor [this bier is higher of power, than 
the dier by his fire J." 

He seems therefore to write (baee, Hueiz, buei) for boy, 
hoy's, buoy, though the precise yalne of the two last words 
is not very clear, and may be (wheiz hwei). Nautical men 
constantly call buoy (hum), and (bui, boi) are not unconmion 
proTincial forms of boy. Compare the Bavarian dialectic 
(baa) for (buu*be) bube, which leads to the notion that boy is 
a form of booby, a word of very doubtful origin. Although 
Hart thus confirms Smith's (m) ia one word, he differs from 
him in writing (vois'es). 

1580. BuLLOKAR, as we have seen, distinguishes boy, buoy 
as (boi, buui), and he gives no examples of oy as (ui, uui). 

1621. Gnx has the varieties (oi, ui, uui), as in the words : soil 
(soil, suuil), boil (boil, buuil), 8poii (spoil, spuuil), toil (toil, 
tumi), joint (dzhuuint), disappoint (disappuuint), buoy (buuil 
r^'oiee (redzhois), voice (vois), oil (oil). In these the double 
tendency is clear, and as the (ui) sounds .must have been the 
more ancient, they were no doubt in existence, though dis- 
regarded, when older orthoepists wrote. Thus Salesbury^s 
(d^oint) is really more modem than (jKll's (dzhuuint). 

1633. Butler says "01 in hoy we sound [as the Erench doj 
(woe), for whereas they write Jw, mtj droiet they say (bwoes, 
swoet, drwoet)."* 

01 — XVHTH CbKTUBY. 

1653. Wajjjb says : " In o» . . . vel oy . . . prseponitor aliquando 
6 apertum (ut in An^orum hdy puer, t6yi nug® ••«.), aliquando 
6 obscurmn, (ut in Angiorum hdil coqueo, idil labor, dil oleum . . .), 
quanquam non negem etiam horum nonnulla k quibusdam per 
apertum pronunciari." 

That is he said (bAi, tAi, boil, toil, oil) but admitted the 
pronunciation (bxil, tAil, Ail). It will be seen that WaUis is 
the first writer who acknowledges the vowel (o) and the 

1 The (w) in the two words is merelT fliaf the mnrndi were (Al'i0o*ee'), the 

ft eonnd dereloped by Butler himeeli. lyllables being lengthened ont, yet I 

Thniy when I waa nearinf Jilioa in the could not diveet myself of the feeling, 

steamer, the name of me place was that (Al'loo'wee') was really said, to 

called out in a slow measured tone by strongly was the sound of (w) dsTeloped 

the boatman, and altiiough I knew in tiie glide from 00 to (ee). 



134 01 — XVn TH CENTURY. Chap. III. § S. 

diphthong (oi). It is quite in conformity with this that he 
changes Uill's (buuil, tuuil) into (bail, toil), and his further 
pronunciation (oil) should imply that (uuil) as well as (oil) 
was prevalent in Gill's time. 

1668. WiLKiNS writes (bAt) for Joy. 

1668. Price says : 

" 0% never ends a word, but, oy, as ioy, eloyy " Oy sounds 
broader than, o», as moyd^ joiner^ joints boisterous, eloy, cloysters, 
mnbroyder, emroides [hemorrhoids], employ, expUit, foyl, tnoyst, 
noise, noysom, oyl, ointment, poise, quoif [coif], void,** 
It is possible that Price's broader oy may be (Ai) and the 
other (^), which would give (dzhaint, bai'staras, ekrolait*, 
naiz, aint'ment, paiz, kaif, vaid,) of which some are confirmed 
by subsequent writers. 

1685. Cooper generally gives oi as (aI), ''« in loss, lost, 
i praDpositus ... semper Grsece, ut woXKol," l^t he admits (wi) 
in boilf moil, point, poison, only, to which he says " oy in 
Qtdlico buoy supporto, quod nos scriberemus bwoy'' is equiva- 
lent, it is therefore to be presumed that he said (bwm). The 
most curious point is his remark that " boy puer dissyllabimi 
est, scilicet (be^Ai)," which is not confirmed by others. He 
likewise admits o» to be (ai) in in-join, joint, jointure, broil, 
ointment, see supr& p. 117, and also, '^ ut i diphthongus," in 
anoint, moil, toil, point. 

1701. Jones says that the sound of ooi was always written oi, 
** in the middle of words or before a consonant, as hoU, coil, join, 
&c.," which were therefore occasionally called (buuil, kuuil, 
dzhuuin), as in times past, and that the sound of i (oi) is written 
oi, ** when it may be sounded oi or ooi (oi, uui) in the beginning or 
middle of words ; as in boil, broil, eoil, foil, foist, froise, groin, hoise, 
join, loin, moil, oilet, poise, poison, soil, spoil, tortois, which some 
sound as with an »','' i-^* ^ (bail, brail, kail, fail) etc. ; and that 
(ai) is written oy ** when it may be sounded oy in the end of words, 
or before a vowel ; Chandois, decoy, &c. — loyal, royal, voyage ; some- 
times abusively sounded as with an «,'' i*^* (Shsen'dais, dekai', 
lai'891, rai-ael, vai'tdgh).* 

1688. MiEGE says nothing of the pronimciation of the 
English oi, but for the French oi he lays down rules some- 
what different from those now followed, saying : 

" The diphthong oi is pronounced oai (oe) as fot, hi, foire, toile. 
Except in some Cases, wherein *tis pronounced ai (e). And 1. In 
such Tenses of Verbs as these ; viz. JTaimois, tu aimois, il aimoit, 
tPaimerois, tu aimerois, il aimeroit. 2. In those Verbs whose In- 
finitive ends in oUre ; as oonoitre, paroitre. To which add the Verb 

1 Compare the sailor's spelling mg is t, ^ « (ai, dzh), according to the 
for (waidzb), i.e. voyag$, where ig, that alphabetic names of the letters. 



Chap. IIL i 8- OI — XVm THCENTURY. UI 135 

tffwif, and this tense of the Verb EirSf Je sou^ iu so%9f %l soti. 3. In 
these National Names, An^lou, FrangoU^ JScossoU, Irlandois, Sol- 
kmdois, MHanoii, Polonou ; with all their feminines in oise^ as 
Anghite^ Frangoue, &c. 4. In these Words, droit, (^dj.) endroit, 
0innt, etroitementf foible^ fr<^i s^d the Derivations of the two last 
Bnt before n, the • keeps its proper Sound ; as /oin, Ann, joindre^ 
ffoitU. Oigtum is pronounced, and begins to be spelt ognon» Die is 
a Triphthong, and is pronounced at in sucli Tenses of Yerbs as 
these are, ili aitnoientf ih aimeroieni, ih soienty wliere the n is left 
unpronounced. But it is no Triph^ong, where it ends a Word, 
the last e making a distinct Syllable of it self, thougli almost mute. 
As in these Words fine^ jne^ anehaie, where oi is pronounced oat ; 
m<moiey yvroie^ where it is sounded at." 

01 — xvmTH Centuky. 

1704. The Expert Orthographist admits (oi, Ai) in choice^ 
explaiif /raise, noise, poise, quoif, quoit, r^'oice, voice, void, but 
sa^s that " in the middle of most other words oi sounds t long 
(ei), as anoint, boil, broil, coin, loin, tnoil, toil, poison, poinf (n 
these (boil, lein, poiz'n, point) are still well-^own Tulgansms. 
1796. Buchanan admits (Ai, oi) only, to the exclusian of 
(ui, oi). 

1768. Franklin writes (distrAoid) destroyed, but unfor- 
tunately gives no other word in oi. 

We may conclude then that in the xvi th century (oi, ui, 
uui) all prevailed, (oi) being most in favour; in the 
xvn th century, most words had (oi, a!) and a few words 
(ui, oi) ; in the beginning of the xviii th centuir (oi, 
Ai, ei) were acknowledged, but at the latter end of that 
century only (oi, Ai) were admitted by orthoepists. 

TJI — XVI, xvn, xvin th Oenturies. 
The combination ui belongs to the xvnth and later cen- 
turies, except perhaps in one or two words, in which French 
spelling had an influence, as the following comparison of the 
orthography of the Promptorium 1440, Palsgrave 1530, 
Levins 1570, and Price 1668 will shew. 



iVw». 


Zmnt, 


beldyng 




build 


beald 


beeldynge 


circuit 


circuite 






conduit 


condit 


condyte 
conduyte 
, conduycte 




guild 






gyylde 
(gylde 


guilt 


giltie 


gylte 


gyity 



136 


tn. 


BU, Atr, 017. 


Chap. III. \ %. 




2tfPIM. 


^VM^)109f* 


XT'VlMyfiM*tMM» 


n. buy 


bye 


bjre 


byyn 


guide 


gyde 


fgyde 
Igoyde 
gyle 


gyde 


guile 


gyle 


gyle 


guise 


gyse 


(gyse 

Ipyse 

(Broose 
(brosyng 


gyse 


III. bruise 




brosyn 








fruit 


frute 


frato 


frute 


juice 


juoe 


juse 


(itiee 
(iowce 








suit 




sate 


sate 



Hence we must consider the combination as an inorganic t or 
u and it must follow tiie laws of those letters. In the above 
table the first group had short %}- the second long t» and the 
third the ti or oo of the period* 

EtT, AtT, OTT. 

The forms ew^ aw^ ow are identical in signification with eu^ 
au ou, and need not be separately considei^. 

The modem sounds of eu are (in) or (ju, Juu), and occa- 
sionally {oo)f of au (aa), and of ou (en) or (e), occasionally 
(oon, nu). But the diphthongal sound (ou) runs through all 
the varieties (ou, (m, an, ahn, sen, en, eu), and Franklin gives 
(au), while even {aa) may be occasionally heard, and, owinff 
to the orthography, this analysis is very commonly aeceptecu 
The Germans hear the diphthong always as their au = (au). 
The pronunciation (eu), a diphthong acknowledged in the 
Italian Europa = (enroo*pa), is heard in America for ou as 
(deun teun) for dcion town, and is said to be a common cock- 
neyism, although tibe cockney sound is, as Mr. M. Bell says, 
more probably (sen) as (dssun tsaun).* Many words now 

riled with u were written with ew in the xvi th centuiy. 
these, and some others still spelled with ew, were pro- 

> Br. Gill shimbles over buUd, gvring hear (tntte,imd*) or («rKB,imd*]t the first 

the three Bounds (boild, btld, byyld). element bem? lengthened and some- 

The more ancient sound mnst nare what nasalised. The Bey. Mr. D'Orsey 

been (beeld) or (beild) whence (btld) informed me that he found the nse of 

descends easily. Mr. Melrille Bell (en) for (an) rery common among 

says that built is often pronomced Londoners, even oi education, whose 

(b^lt) in Scotland, a yariety of (bylt). pronunciation he had to correct. In 

s In Mrs. Barney Williams's Yankee Norfolk ou is regularly pronounced 

tong " Bobbing around," which was so (su, »u). 
popular a few years ago, I teemed to 



Chap. III. } 3. 



EU — XVI TH CENTURY. 



137 



nounced with the long u of that time,^ which requires special 
ccmsideration, it will be most conyenient to postpone their 
consideration till afterwards. The sounds attributed to au, 
ON in the XYith century were also frequently attached to 
simpLe a, o before / or //, and Uieae will be considered under Ia 

EXJ — XVI TH Obntukt. 

1530. Palsgraye says : '^JEb in the frenche tong hath two dyuecse 
scmndynges, for sometjme they sound hym lyke as we do m our 
tonge, in these wordes a dewe, a sbrewe, a fewe," this is the sound 
whu^h will be oonsidered here, ''and somtyme like as we do in 
these wordes, ir$wef $Uwe^ rewe^ a mewe,^^ which will be considered 
under TJ. ''The soundyng of ev, whiche is most general in the 
frenche tong, is suche as I haue shewed by example in these 
wordes, a dewe, a shrewe, a fewe, that is to saye, lyke as the 
Italians sound ev, or they with ts, that pronounce the latine tonge 
aryght, as evrhx, irevxy liev, tU&v,^^ 

The reference to Italian completely establishes the sound, 
which is as sin^^ular and curious in French as in English. 
According to Meigret, however, the sound was (ey), for he says : 

" ^Bt e clos fist Encores vn' aotre diphthong' auEc u, come eu eur, 
pen, veu, eureus. Finablemmit il M vne taiphthonge se joNant a 
9Elle de ao ; come En veao, beao, moreao. Dont je m' emErvElLe de 
^eus qui premiers ont teimin^ qete triphthong' En u : yu qe la pro- 
non^fa^fon ne tient rien de Time m^es de 1' on clos qi a q^q' 
afflnit^ auiq Tu."* 



1 We find in LeyinB 1570, dtu?e 
debitam, ilhoe, fitwe, rtWy tp^w^ 
b l mp t f tr$wff *M#t0, rMk0w, reuemw^ 
9aUw [but 90riu$ although inBerted 
nader 'vE ante W,"] indewy eoHim4%o, 
pmnew, tUtte^ lr#iM«, knpf$, rmoU^ 
trtwtk [bat vntnOk although under 
the hea£ng ft^]. Words ttul written 
with M9. ttod promonnoed then as long 
If according to Sir T. Smith 166S, are 
M#w, shw, msw, hrewy bUw. 

' See the long extract fromHeigret 
concerning oo, aou, on p. 141-2 below. 
G. dee Antels obiects stroitfly to 
Meigref 8 analTne (#y) of the French 
m». Speaking of Meigrefe assertion that 
both sounds were h^urd in adiphthong, 
he asks (liTet, p. 130) : <* Je luj de- 
mande a la diphthongue frui^oise eu 
oes mots Jeu et feu garde le son 

tkr de l'«P" " 11 ne &ut done pw 

Ste les ToyeUes gardent aux diph- 
ongnes leur son propre et entaer, 
mais hien qu'elles servent toutes deux, 
amt en leur son propre on en nn autre 
Toisin, k faute de lettres plus idoines 
(oQnreiiables)." Felletier (ib. p. 183} 



is indistinct, at least as cited, but 
Bamus (ib. p. 189) says : '' Lasixiesme 
Yoyelle cost ui^ son que nous escripytms 
par deux ToyeUes, b et u, comme en ces 
motB peur, nrntr, Mtur^* and he pn^KMes 
a simple sign for it. Beza (ib. 521) as 
anal]f«ed by livet says: **Dans cette 
diphthongue bu ou n'^tend ni V$ ni 
Tm, mais un son qui tient de Tun et de 
Tautre : hemf, muf, peu paucttic, «#air 
BonoB, veu totvm, et un grand nombre 
d' autres que les Picards prononoent 
•ouvent u simple, disant Dw, ju pour 
Dieuyjm, Lee Fran^ais imitent quel- 
quefois les Picards, en ce qu' ils pro- 
noncent par u simple les mote teut 
SBCUBITB et ses d^riy^s ..» meur matu- 
BUS ... et en g^n^ral tons les noms en 
iwre long [now -ifr»] d^riv^ dee Terbes 
.... ; il en est de m&ne dans les parti- 
cipes pasi^ passim, masculins ou £^mi- 
nins, terminto en «m, #«# [now -m, -im] 
comme beu^ hem .... : c'est ft tort qu 
& Chartres et k Orleans on prononoe, 
BTeo une di6r^ eiiy et, d'autre part^ 
qu'on £ut rimer, hemr et dwry enffrttvture 
et/^NT^ hmre et twtute^ finite qu'on 



138 BU — XVI TH CBBITUKT. Cha». IIL J 8. 

But Englishmen heard tiiis (ey) as (en), as appears from 
Harty who in his French Lord's Prater, gives (sietiz, seoz) 
for deuXf eeux. As to the combination eau, which Meigret 
says was (eao), we have the word beautp, written b^ffte^ 
beawtye in the Promptorium, beautie in Palsg^ve, and bewtye 
in Levins. Hart gives (beau'ttft)^ Gill pronounces (beu'ti) 
and Butler (beau*tf) which may mean (beoai^), though 
some doubt attaches to the last pronunciation.^ 

1547. Salesbury does not notice the combination eu, and 
gives no English word in which it occurs. 

1568. Smith says : '^ Et ^ diphthongom Gnecam habent Angli, 
sed raiius, quae tamen apud Gallos est freqnens : (feu) pauci, (den) 
roB, (meu) vox catomin, (shen) monstrare, (streu) spargere." And 
in his Greek pronunciation he adds, " ev, vt eu, eiye, euge. Augli 
pauci few, <l>€u, ros. dew, Seu. rfv sonamus apertius, vt illud Gblli- 
cum beau, quod multi Augli heu : sonum etiam f elium quidam mew, 
alii msaUy quasi ijlcu, firjv exprimunt." 

Observe that mew for hawks had the sound of long u, 

1569. Hart, as shewn by the citation on p. 132, distin- 
guishes mew (mieu), ewe (jeu), you (juu). 

1580. BuLLOKAR recognized the diphthong (eu) distinctly 
by writing the word hew thus : he,u, the comma, which he 
wrote under the u, meaning that it had the sound of (u). 
In his list of synonymous signs he gives e,v e,u ew (where 
the comma should be subscribed to the r, u) as identical, and 
I find the word fiewed meaning (sheu'ed). 

retnmre en Onyeime." These last (ib. p. 138] : '*Rien de plus yagae, de 

examples point to a remnant of an (ij) plus ind^tennin^, que la prononciation 

diphthong, which is a real natural diph- de, u, ^ o, om an moyen age et encore 

tiumg, and was distinctly pronounced an zt« si^le. Nous ne pouTons mieux 



to me every morning at Konrich by a faire, au Ueu de donner a innombrables 

render of fish monotoning under my exemples de cette confusion, que de 

windows, (n^ bloo'tizs lii) = new reuToyer au Traiti de VeniJIeaium 

blotUeri here! The real mutations of firangoU$ d&J/i, Quicheratpp. 354>369. 

the Latin o, besides its natural change Cf. Observations etc. de Menage, 1. 1, 

into (uu), were howeyer two, closely re- p. 291, 324, 481. Olossaire p%e«rd par 

lated, first foe] falling into (ue), and rabb^ Corblet, p. 131. Sur la comu- 

secondly (eo) falling into (eu). The sion de mi et oti en particulier, Gf. 

form (ue) appears in yery early French, Quicherat, ouy. cit. p. 364-366." 
where it was probably soon discon- ^ Bamus (liyet p. 207) makes the 

tinued^ since (ue) was also used as a combination eau a diphthong, the first 



mutation of Latin «, but it remains the element being his mute $ and tiie 

regular Spanish mutation. The second second his simple yowel au. The dif- 

form (eo, eu) gradually prevailed in ference of Meigiet's sound and his may 

Freneh, and became replaced by foe) have been very slight («ao, tk»), but 

impttrently just about the time that the latter prevailed. Beza (ib. n. 523) 

Meieret wrote, so that he retained an analyses in the same way as fiamus. 

old (;#u) or (ep pronunciation (it is not These analyses at least snew the ex- 

ouite clear wnich) and his more youth- istence of an old s sound at tiie oom- 

ral opponents ignored the old sound mencement, and hence account for the 

altogether. The subject requires much English translation of ^e combination 

careful investigation. Livet observes into the fiimiliar diphthong (eu). 



Chap. HI. § 3. EU — XVUTH CENTURY. 139 

1621. G1LL9 in his anxiety to give prominence to the first 
element, lengthens it, tiius : '' E. saepiiis prsDcedit u^ yt, in 
(een) eawb ovicula, (feeu) fewb pauci, (seeu'er) sewer 
dapifer.'* 

1633. Butler distinctly recognizes (eu) in dew, ewe, few, 
hew, shew, rew, sew, strew, shew, shrew, pewter, see under U. 

It will be seen in the next chapter that Chaucer distin- 
guished the two sounds of eu by an etymological rule, the 
sound (eu) being reserred for tiiose which were not of French 
origin. This £stinction was lost during the xv th century, 
80 that in the xn th no general rule can be given, but each 
word must rest on its own independent authority. For lists 
of such words see Chapter lY, § 2, under £U. 

EU — xviiTH Century. 

1653. Wallis, says : '^ JSu, ew, eau sonantur per i claram et w, 
(eu). Tit in neuter neutralis, few pauci, beauty pulchritudo. Qui- 
dam tamen paulo acutius effenmt aesi scriberenter, niewter, fiew^ 
hiewiy, vel niwter, fiw, hiwty ; prsesertiiii in yocibus new novus, knew 
sciebam, enew ningebat. At prior pronunciatio rectior est." 

That is Wallis had heard some persons say (nieu'ter, fieu, 
bieu'tt) although many, perhaps most, at that time said dis- 
tinctly (niu'ter, fiu, biu'ti) ana he found this pronunciation 
particularly prevalent in new, which in the next century 
Franklin called (nuu) and which is still frequently so called.^ 
The sound (eu) was undoubtedly beginning to be imfrequent. 
The sound (iu) however cropped up chi^y in those words 
previously pronounced as long u. 

1668. WiLKiNs acknowledges (eu) in hew, and Price in 
the same year allows (eu), that is, says " ew keeps its sound" 
in brewess, few, lewd, ewe, feud, neuter, pleurisie, but gives 
(iu), that is, says '' ew hath now obtained the sound of itr" in 
ftfetr, brew, chew, crew, drew, embrew, eschew, hew, gewgaws, 
knew, sewer^ slew, stew, steward, vinew,^ monsieur, adieu, lieu, 

1685. Cooper hears onlv (lu), the same sound as long u. 
The diphthong is in America more frequently (tu) than (iu), 
and even (eu) remains there in some parts. 

1701. Jones seems still to have a lingering feeling of the 
difference between (eu) and (iu). He asks when may the 
sound of eu be written eu ? and answers : '' In the beginning 

^ In 1849 the present writer pub- newsrender, "we always call it (nnoz)." 
lished a newspaper (Milled the Phonetic * Probably in the sense of a waiter 

2few9y printed phonetically, and there- at table, 
fore bearing the title (Fonet'ik Niuz). * Probably, venue, 

" Why do you write (ninz) P" asked a 



140 BU — XVn TH CBNTUBT. Chap. IIL f 3. 

of all words, except ewy ewer, Eum,* and ^'in all foreign words 
firom the Latine, Greek &c as adieu, heuf, cavalUeur, DeucaUat^ 
Deuteronomy, feumet, geub^ grandeur, lieu, Meuae, Monsieur, 
neuter, pardieu, pleurisy, purlieu, Reuben, rheubarb, rheum, 
Theudas, Zeurin 8fc exc^t view." And he allows tlie same 
sound to be written ^m? '' in all English words as crewet, dew, 
pewter &c" But lie never asks, vni^i may the sound ot eu 
be written u ? On the other hand he does ask when may 
the sound of f« be written eu or ew ? And he answers, the 
first ** when it may be sounded eu in foraign words, as neuter 
&C,** referring to the list just given, and thus clearly dis- 
tinguishing the two sounds (eu) and (iu) ; and the second 
''^en it may be sounded ew in Englia^ words, that are 
purely^ such, as in askew, crewel, dewberries, eschew, ewer, geuh 
gaws, Hewet, jewel, nephew, pewet, sinew, vinew, and in olew, 
chew, clew, crew. Crew, drew, few, flew, Orew, grew, Jew, knew, 
mew, new, screw, shew, skew, slew, spew, stew, siews, strew, threw," 
Jones says that the sound of o and ou, evidently meaning 
{oo, oovl), is written ew when it may be sounded &f^ as in chew, 
eschew, shew, shrew, shrewd, Shrewsbury, pronounced ^*cho, 
shrode, Shrosbury &c.'* {Shoo, Shrooz-beri) arc the only 
sounds here remaining. But that (shroo) must have been 
known in Shakspere's time appears from the last couplet of 
Taming of the Shrew, fo. 1623, the preceding 14 lines being 
in rhyming couplets : 

JSorten. Now goe thy wayes, thou hast tam'd a curst Shrow. 

Lue. Tis a wonder, by your leaue, she wil be tam'd so. 

£we has still a provincial pronunciation (joo, jaa). 

Hau as is seen by the quotation from Wallis, follows the 
fortune of eu. Wallis has (beu'tt) admitting that some say 
(bieu*tO. Miege has (biu*tt). Jones says that beau is 
*' sounded beu in the beginning of all words,'' referring to 
e-ea, which shews that he considers ea in eau to be the digraph 
ea, that is, a mere representative of (ee), and satisfactorily 
determines his pronunciation. Even the word '' Beaw a 
name'' he writes beu. But he never allows the sound to be 
long u, that is, (iu). On the other hand he also says the 
sound of long o is written eau '* in the sound of beau in the 
beginning of all words," which should imply that (hoo'ti) was 
heard as well as (beu*tt). He also says that Bourdeaux is 
" sounded Boordo*^ (Buur'doo). 

The conclusion seems to be that some speakers still said 
(eu) and Jones recognized it as an admissible and theo* 

1 The following list would imply that Dr. Jonei did not know mueh of 
etymology. 



Chap.III.{S. EU — XVmTH. ATT — XVITH CENTUKY. 141 

Teticallj the best sounds but that he freqixently heard 
and admitted without any word of blame^ the newer 
Bound of (iu). 

ETJ — xvniTH Cbntuky. 

1704. The Exfebt OsTHoaBAFHiBn says: "it must be a very 
critical ear, that can distinguish the sound of eu in eucharist from the 
long u in unity j and the eu in rheuharh from the long u in rumour ^ 
wi&out an apparent and too affected c<mstraint, contrary to the 
usual pronunciation observed by the generality, which (in this case) 
would sound very pedantick." 

Here^ the confusion of thought and consequent nebulosity 
of espression^ which makes it difficult for an ear to dis- 
tingULsh sounds without a constraint which would sound 
pedaatiek, and which is contrary to the general pronunciation, 
18 a good example of the darkness in wmch we have to grope 
lor our results. It is to be presumed that the writer did not 
distrngmsh eu as (eu) from u as (iu), and found the utterance 
of those who still attempted to do so, affected and constrained. 
But did he pronoimce all his 32 words having ew filial^ with 
(iu), including '' sew or did sow with a needle, sewer a drain, 
shew or did Siow'' ? This is more than doubtful, and the 
distinctions here made between present sow, show, past sew, 
shew, are entirely without corroboration. 

1766. Buchanan generally makes eu, ew = long u or (iu), 
but writes seu^er {ahoor), shew (shoo) sew (soo). His ewe, 
monsieur, lieutenant are (iu, monsiur', liuten'tnt), chew (tshuu), 
beauty (Jbiwti), beau, beaux (boo, booz). 

1768. Franklin writes (nuu) for new. 

The usages of the xviu th century did not therefore sensi- 
bly differ from those of the xix th. But to shew how (eu) 
still lingers^ it is enough to cite the pronunciation (shtni), 
clearly a variety of (sheu), heard from a highly educated 
speaker, during the preparation of these pages. 

AU — XVI TH GBNTURY. 

ISSO. Palsgbave says : ^*Av in the frenche tonge shalbe sounded 
lyke as we sounded lyke as we somide faym in these wordes in our 
tonge, a dawe, a mawe, an hawe. Exce]{ft where a frenche worde 
begynneth with this diphthong av, as in these wordes, avlckn, 
d^Blire, av, wssi, avx, and aucthr, and all suche lyke : in whiche 
they sounde the a, almost lyke an o, and as for in avner, a and v be 
distinct syllables, as shal appere by his writtyug in the frenche 
vocabular." 

Now Meigret says : '' vn' aotr' eu ao, come aotant, aos, loyaos : 



142 AU — XVI TH CBNTURY. Chap. IIL } S. 

pour laqElle I'ecritture Fran90Bz' abnze de la diphthonge an, qe la 

Eronoii9{a9{oii ne conoEt point. Car com' aotrefoEs je tous ey dit, 
i diphthong' £t de tBlle nature q'slle requiert la prola9fon En vne 
msme syllabe d£' deu' yoyslles qi la compozet: come nou' le 
{e'zohs conmmnemEnt: s einsi obsErve I'Ecritture, sn moindre, 
peindre : e' qels non' pronon^ons 1e' diphthonges oi, s ei, sn vne 
msme syllabe. e ponrtant sont abuzes tons 9en8 qi se persuadet qe 
deu' YoyEllEs conioinctes Enssmble, caozet vn tiers son, qi ne tient 
ne de Tune, ne de I'aotre : come qant vons ecriuez mais, pour mE's, 
il dizet qe a, E, i, conjoins EusEmble, forjet la prola^fon de b, 
ouYErt: suyuant leur rEgle donq ie direy qe ayant, aora En sa 
pronon9ia9ion Eant ; payant, paye, pEant, p£e, je direy le semblable 
de toutes aotres diphthonges qe you' pronon9ez com' eUos sont 
ecrittesy q'Elles doEuet fcV vn son tiers, aotre qe ^eluy dE' deu* 
voyelles coniointes EnsEmble : e qe conseqEmmEnt vous ecriuez mal 
moins, eureus, eaje (on dit bien aosi aje, e st la diphthonge ea, 
bien rar' En Fran^oEs) vu qe vou' pronon9ez Ie mE mes voyElles qi 
sont ecrittes, e q'^es ne forjet point la vn tiers son. Yoyez donqes 
q'Elle opiniatret6 d'abus caoz' vn srreur inueter6 : tant Et difff9il' a 
I'home la reconoE88an9e d'une faote pour vne par trop grand' estim' 
E prezompsion de sa suffizan9e conioint' a vne meconoBS8an9e de I'im- 
bE9ilit^, E imperfec9fon de notr' EutEudemEnt: Ao reg^ml d'aou 
par on clos je ne I'ey point decouuErt, q'ao mot aou^ qe vous 
ecriuez Aoust, etant s, supErflue." 

This long quotation will serve to shew that Meigret's 
diphthongs must be accepted as such, with the exception 
of ou, of which he says ** aotrement ne Toze je noter, and 
which was the vowel (u) simply. Hence as Meigret only 
heard (au) in the one word aoUtt, now (uu), and heard (ao) in 
all other words, either the English must have been (ao), or, 
if it were (au). Palsgrave misheard the unfamiliar (ao) as the 
familiar (au). The latter is d priori more probable and agrees 
with all the other indications we possess.^ 

1 6. des Autels was yeiy yehement p. 133). It is erident then that Mei- 
Ugainst Meigret for using the diph- gret used and was fiimiliar with (ao). 
thong (ao). **Je Iny demande," says liTet (ib. p. 122) remarks: **il est 
he according to p. 130 of JAfe^ **oti certain qn*en Anjon Ton prononce de 
est le son^ non entier, mais demy on la eAoour, f ai ehaod, ehwaozj en appny- 
encore moms, de Va en la diphthongne ant snr la et glinant Ug^ment sor 
de sa nonvelle forge ao f To the Vo qni ne s'entend gndre plus qa*nn # 
first objection he had raised Meigret mnet;*' bnt this must be a recent de- 
had replied : ** si tous n*aTez le cerreau Telopment, the unstable (ao) becoming 
bien trouble d'opionastret^. tous trou- in this case (&o), while in ^e clasriou 
Terez qu'en introduisant la oipkthongue French it must haTc passed through an 
aOf je ne fais qu'aocorder I'^criture k (a6) form. That the a was origmally 
la prononciatiouj'* (ib. p. 122), and to pronounced there can of course be ety- 
the aboTe question he answered : ** le mologically no doubt, and the change 
plus opiniatre sourdaud du monde ne of (ao) to (oo) is preciselT similar 1o 
saurait nier qu*il n' oye (entende) en the chai^ of (au) into (aa), which will 
aoti (aussi) un a puis un o qui luy est be seen to haye taken mace in "Rw gH'h 
conjoint en une m^e syllabe," (ib. In Weldi we find Salesbury's 4hq be- 



Chap. HI. § 3. AU — XVI TH CBNTUEY. 143 

Palsgrave, speaking of French pronunciation, says : 
" If m or n folowe next after a, in a firenche worde, all in one 
syllable, than a shall be sounded lyke this diphthong av, and some- 
thyng in the noose, as these wordes dmhre, chdmhre, mand^f amdtU, 
Unvty quant, parldni^ regarddnt, shall in redynge and spekynge be 
sownded aunAer, chaumhre, maunder^ amaunt, taunt, quaunt, parlaunt, 
regardaunt, soundyng the & like au, and somethynge in the noose." 

Of this there is no trace in Meigret, but the observation is 
important as explaining the English pronunciation of words 
from the Frencn, and the nasaUsation of aun is remarkable 
when compared with Jacob Grimm's observation that modem 
English afi, which ca (aa), is pronounced " as a lengthened a, 
something in the nose'' (wie gedehntes a, ein wenig genaselt).^ 

1547. Salesburt has no special article on au, but he says : 

"tff English & w Welsh do not differ in sound, as wawe, wow 

unda, Also to is mute at the end of words in English, as in 

the following awe pronounced thus a (aa) terror." Also he 

says that ''sometimes a has the sound of the diphthong au> (au) 
especially when it precedes 7 or Z/, as may be more clearly seen in 
these words balde, hawld rbauld) calvus, ball, hawl (haul) pila, 
WALL, wawl (waul) mums/' And he writes "oalatjitt, gdawnt 
(galaunt)." 

The word (aa) for (aau) atioe is here singular, especially as 
it is adduced as an instance of the omitted (u). Smith pro- 
nounces this word (au) and GKll (aau). Salesbur^ is also incon- 
sistent with himself, for in his Welsh pronunciation he says : 

" All thoughe the Gkrmaynes vse w yet in some wordes sounde 
they it (to my hearing) as the forther u were a vowel, and the 
latter o (sic) consonant, where we Britons sounde both uu wholy 
together as one vowell, wythout anye seueraU distinction, but 
beynge alwayes eyther the forther or the latter parte of a dyph- 
thonge in Englyshe on thys wyse : wyth au>, and in Welshe as thus 
ufyth mcen,^^ 

ooming modem o. In Italian o apfio au ne diff^ pas sensiblement de la 

haa succeeded freqnentlj to Latin au, royeUe o/' to which he adds:/*leB 

and so on. The question of importance Kormands la prononcent en fiiisant en- 

here howeyer is, when did the change tendredistinctementa,o: disant a-o-^<m< 

take place P The testimony of Pau- ponr autant : peut-dtre est-ce la Traie 

mre to (an) and Meigret to (ao), and et ancienne prononciation comme la 

uie objections of des Antels and PeUe- Traie orthographe de cette dipthongne" 

tier— who says to Meigret (ib. p. 13S) —seem to shew that the change took 

''il t' eCLt antant vam mettre un o place in the first half of the xvith 

gtmple"^-€uid the assertion of Bamns centniy; that is, that about this time 

(ib. p. 1S6) that it is ** le son que nous the simple vowel (oo) prevailed over 

escripyons par deux royelles a et v, the diphthong (ao) or (au), although 

oomme en ces mots : aultrts^ auiUly on the latter did not absolutely die out. 
nous prononcons toutesfois une Toyelle ^ Deutsche Grammatik, yoL I, 3rd 

indifisibl&" together with the dictum ed., 1S40, p. 394. 
of Bess (lb. p. 520) **la diphthongoe 



144 AU — XVI TH CBNTUIIY. Chap. III. i S. 

It would seem impossible after the preceding remark to 
suppose that w were mute in aw. Indeed tcyth aw seems to 
be rather a W^h phonetic transcription than the usual 
orthography, in which, as in the other passage quoted above, 
we should expect awe. 

1568. Smith simply ^ves " AU seu ov. (Dan) monedula, (dan) 
unguis auium, (rau) cradus, (nauHt) nihil, (taunt) doctus, (laau) 
lex, (man) stomachus, (sau) serra, (an) terror, (lannter) risus, 
(faoHt)^ pugnayit, (strau) stramen.^' But in his Gbeek pronuncia- 
tion he adds : *' au, ev. Tfv* Eandem rationem seqnuntur, quam in 
reliquifi. Kam si fiiisset apud veteres tanta soni conunutatio, pro* 
fecto Gianunaticoram diligentia no#> hoc tarn insigne disciimen 
prsDteiitam reliquisset. Itaque eic avBdoo loquimur, vt audio nos- 
trates vnguem, claw, & scabere claw.** So that his au was cer- 
tainly (au). 

1669. BDlet says : " The Dutch" that is the Gennans, " doe vse 
also aUy eiy & w, rightly as I do hereafter.** 

Now the German soimds are, and probably were, (au, ai, 
jee) or (ii), but Hart clearly did not refer to this last sound* 
When then Hart writes (autours, auluaiz, aulso, tshaundzh, 
bikaus, radikaul) for authors, always, also, change, because, 
radical, he meant (au) to be sounded as in German. 

1580. BuLLOKAR distinctly writes ba,u, meaning (Hau), 
and uses (kaul, kau'st) for caul, causey = causeway. His 
notation at am ah he explains as = (aul, aum, aun).^ ^This 
agrees with the rest. 

Up to this time therefore, when Shakspere was 16, the 
pronimciation of au seems to have been mdisputably (au) 
the same as the modem German au. There can be little 
doubt that Shaks^e in his youthful days must have said 
(au), but during his lifetime the general pronunciation seems 
to have changed. Between BuUokar's and Gill's books, 41 
years elapsed, and although Gill had an old pronunciation, 
yet he seems to have followed the times somewhat in this 
combination. In determining the pronunciation of Shak- 
spere, we must remember that he ana Dr. Gill were bom in 
the same year, 1564, and that Shakspere died, 1616, eight 
years after Gill had been made master of St. Paul's school, 
and five years before the publication of (jell's book. Henoe 
Gill's pronunciation is the best authority which we have for 
Shakspere's, and certainly gives us the pronimciation of 
Shakq)ere's time. It is therefore singularly vexatious that 
we cannot mc^e out a very clear accoimt either of long i, 
(p. 114,) or of this diphthong au, fipom GiU. 

^ In the original (fooHt), which Sa dearlj a misprint. Poisibly (laaa) ht 
(Urn) was also a misprint. 



Chap. HI. § 3. AU — XVITH CENTUKY. 145 

1621. Onx says: "A, est tennis, ant lata; tenuis, ant breris est, 
vt in (taloou) taixowb sebum ; aut dedncta, ut in (taal) tale fabula 
ant computus: lata, Tt in tdl ialle procerus. Hunc sonum Germani 
exprimunt per oa. Tt in tnaal conuinium, hoar coma: nos Tnico 
cbaractere, circumflexo 4, contenti erimus." 

This ought to imply that a in tall was a simple vowel and 
not a diphthong,^ and that it was {aa, aah) or (aa). The 
Germans perhaps really said (aa) or at most {aah\ but (aa) 
was the sound which appears certainly to have been heard 
by the English in the xvu th centunr. But Gill, who is so 
particular in his phonetics, absolutely confuses the diph- 
thong (an) with his d, in the following curious paragraph, 
where I leave his symbols untranslated. 

'^ A pneponitur e^ ut in aerf aebeb aereus. o nunquam ; ssepius t, 
et tt , vt, in aid auxilium ; hait esca ; laun sindonis species ; & a paun 
pignus : vbi aduerte au nihil differre ab d. Eodem enim sono pro- 
ferimus a hdl, ball pila ; et tu hdl^ battle, voeiferari : at ubi ver^ 
diphthongns est, a, deducitur in 4 , vt 4tt awe imperium ; dug&r 
terebra." 

Here he admits that au in his own phonetic writing is 
sometimes the sound which he represents as a simple vowel, 
his *' broad <i" and sometimes ** truly a diphthong,^' but then 
becomes &u or A + u. I feel therefore bound to take his 
tfiu as = his & or (aa), and his dti as = (aau). In this point 
then (3ill must have given in to the xvri th oentury pronun- 
ciation. The pronunciation (aau) is not recognised by 
others. In Gill's first edition, 1619, he uses au instead of Ay 
for (a a) and in the case of 'Hhe true diphthong'' to make 
the u apparent, he considers the u and not the a to be 
lengthened. The meaning is evidently the same. 

1633. Btttler is still less explicit, for after saying that 
" the right sound is a mixed sound of two vowels whereof 
they (diphthongs) are made," and referring to the Greek, 
he merely tells us that '' au in PauFs and hia compounds, 
Pauk'CrasSy Pauls-cyrc-yard,' the Londoners pronounce after 
the French manner, as ow," 

We are therefore driven to Ben Jonson's grammar 1640, 
which was not published till two years after his death, and 
which has probably been tampered with. Jonson was bom 
in 1574, ten years after GKll and Shakspere, and his pro- 

1 But that it does not necoMarfly do with his admitting it afterwardfl to be 

80, appears from his calling long if " fere diphthongns au,** and, as it will 

which was *' fere diphthongns #i" the be seen, he aunost uses these Tery 

" thick f/' or " • crassa." So that his words. 

assertion that a in tall is ** a lata" or * The Greek c here repr^ents a 

" broad «" wonld not be inconsistent croased «, much resembling it in form. 

10 



146 AU — XVITH CBNTUET. Chaf. IIL | 8. 

nimciation at best belongs to the yery edge of the xvi th 
eentuiy. He says> 

when a '^ comes before /. in the end of a Syllabe, it obtaineth the 
full French sound, and is ntter'd with the month and throat wide 
open'dy the tongue bent backe from the teeth, as in a/, mm/, gal. 
fil. tal, eal. So in Syllabes, where a Consonant foUoweth the ^ as 
in $alt. maU. haUne. eahneJ^ 

BoUokar writes (baTm kal'm = banl'm kanl'm) for balm^ 
eahn. Salesbury gives ealme^ call in his Welsh pronuncia- 
tion, as words in which '' a is thought to decline toward the 
sound of the diphthong au, and the wordes to be read in 
thys wyse caul^ cavlme. Gill gives balm as (bAAm) accord- 
ing to our present interpretation of his A = au. Ben Jon- 
son's explanation of his a before / will really apply better 
to {aa) man to (aa), because he omits all mention of labial 
action, but I suspect that (aa) was fully developed in England 
at the latter end of his life, and that he intended to inoicate 
its soimd, but had not noticed its labial character. It is 
worthy of remark however that Jonson's account of this 
sound is almost translated from the description of Latin A 
in Terentianus Maurus whom he cites in a note : 

A, prima locum littera sic ab ore sumit| 

Immunia, rictu patulo, tenure labra ; 

Linguam<j«« necesse est it^ pandulam reduci, 

Ut niBUS m iUam valeat subire vocis, 

Nee partibus ullis aliqnos ferire dentes. 

and this renders his description altogether suspicious, as if 
it were the result of learning, not of observation. 

The result is that in the earlier part and middle of the 
XVI th century and at least to 1580 the sound of au was 
(au) or (aau) ; that at the close it may have passed 
into {aa) ready to Ml positivelv into (aa) in the next 
centurv. The modem contest between (aa) and (aa) 
in such words as gaunt, haunt. Jaunt = (gaant, Haant, 
dzhaant) or (gAAnt, HAAnt, dzhAAnt), while aunt has 
remained (aant), — seems to point to a time of (aa) or 
{aa) before (aa) was evolved. In ^ving the pronun- 
ciation of Shakspere, however, having regard to the 
archaic habits of the stage, I think it will be more cor- 
rect to write the full diphthong (au), see Chapter YIII. 
§ 8. The change of (a) by the action of (u) would 
naturally be to tne round form (o), for which in French, 
the narrower form (o) has prevailed. But if the (a) fell 
first into (a), the (u) would labialize it into (o), for 
which the narrower form (a) is frequently substituted. 



Chap. m. § 3. AU — XVII TH CENTURY. 147 

The distinction between primary^ or narrow, and wide 
forms, is seldom upheld in its parity, and the soimd 
varies frequently, imnoticed, from narrow to wide in 
different individuals, who believe themselves to be 
speaking alike. 

AU — xvHTH Century. 

1653. "Wallis says: '*Au vel aw, rect^ pronunciatum, sonum 
exhiberet compositiun ex Anglorom d brevi et tp, (sbu). Sed a 
plerisque nunc diemm effertur simpliciter ut Gennanoram d pingue 
(aa) ; 80110 nempe literse d dilatato, et Bono HttersB io prorsus sup- 
presso. Eodem nempe sono effenmt dU omnes, awl subula; eM 
voce, eaul, cawly omentum, vel etiam tiara muliebris." 

This is just the conclusion that Dr. Gill had arrived at, 
but he does not acknowledge the pair, fall folly , of Wallis = 
(fi^ fdi). 

1668. "WiLKiNS entirely agrees with Wallis. Price only 
says that '' aw soundes brcmder then au as dawby haunt^^ the 
meaning of which is not clear. 

1685. CiooPER, as usual,, is rather peculiar. He says : 
^^ Am ean^ east, cum u coalescens (seu) . . . nunquam occurrit in 
nostra lingua. Zanee hasta, lancet scalprum chirurgicum, k lanceola ; 
laneh navem solvere k G. laneer, Jaculaii, Qanch in sudes aoutas 
preecipitem dare, hant k G. hant&r frequento; haneh k G. hanehe 
femur ; 6 ant, macer quasi want ab A. 8. wana carens, gantUt chi- 
lotheca ferrea, hndrees k layando, nullo modo scribi debent cum ti ; 
contni enim suadent sonus et derivatio ;^ falsb itaqti^ seribuntur 
launce &c. Quasdam vocabula k latinis prsecipue derivata sciibimus 
per au pronunciamus prout au vel a (aa) audaeioua audax ; maunder 

murmurare ; k G. maudire maledicere in loss, lost con- 

junctus cum u semper scribimus per au (au), ut audible audibilis, 
audience audientia; audit-or-y auditorium, augment augeo, augury 
augnrium, august augustus, auricular auricularis, austerity austeritas, 
auihentiek authenticus, authority autboritas, cautious cautus, fravdw- 
lent dolosus, laudable laudabilis, laurel laurus, plausible plausibilis, 
negUgenter loquentes pronunciant prout a (aa) ; in ceteris vocibus 
au & aw semper prout a (aa) pronunciamus." 

This fancy for pronouncing au as (au) or (on) in certain 
words, seems peculiar to Cooper ; it may, however, have 
represented one of the transitional stages (au, au, au, aa) or 
(au, an, a\ aa, aa). We can readily conceive that the sound 
had passed through all these stages ; the (aa) often heard at 

^ As to sound, many eyen now say (m, n) when they represented what are 

(lAAntah lAAnsh, HAAnt, HAAntsn now the French nasals, was a regular 

HAAnsh, gAAnt'let, IxAn'dres^. As to indication of their origin, see snprii p. 

deriyatioii, the insertion of (u) before 143, and M, N below. 



148 ATT — XVnXH CEITTUBT. Ck4P. HL f I. 

present in haunt, gaunt, Jaunt, &TOim the notion that (aa) 
once existed. Cooper's ''negligent^ loqnentes" refers of 
course to the general pronunciation, whicui was opposed to 
his ideas of correctness. Whenever an orthoepist talks of a 
'' careless^' pronunciation, he means that which is most pre- 
TiJenty and which is therefore most valuable to the student 
of chaneesy while his '^ careful" pronunciation is that of 
Dr. CKlls ''docti interdum,'' seldom or never heard when 
speakers are thinking of the meaning, rather than the sound, 
01 what they saj. 

1686. Mnas says : '* La diphthonge am en An^ois se pnmonce 
oomme n6tre a en FrBn9oi8y Exemple, Cause, AiOhpr. IL en taxA 
excepter Auneient, & ses Dcnivatifs, o^ la Biphthongne se proaonee 
comme Va ample en ATiglais. De meme en est il des mots frniwans 
en aunt, comme aunt, to daunt, qu'il &at prononcer aint, ton darnt. 
To laugh, se pronom^ laiff. Paul suit la B^gle, hormis qoand on 
parle de 1' Eglise Cathedrale de 8. Paul k Londres. Alors on 
1' appeUe F6h .... La Bipththongae aw sonne comme un a long 
en Fran9oi8. Exemple, Law, flaw qu'il faxX prononc6 1&, fli. 
Mais il se prononce bref, dans awry" 

The difficulty experienced by the French in distinguishing 
(as) from (r), and (a) or (a) from (a) has been noticed on 
m, 71-2. The preceding indications lead me to suppose that 
Miege meant to express the sounds, (kxAz, AA'thar, aen'shent 
aDsen'shenty aeaent, daeaent, laaaef, PaaI Pooulz, Iaa Aaa). The 
sound of ancient is doubtfrd. The use of (aeae) in aunt, daunt 
is rather a thin pronunciation at the present day, which 
some ladies even still further thin to (sent, daent). The 
sound (Pooulz) is not now heard, but as Chaucer writes 
Powles, and as Butler gives the pronunciation (Pooulz) " in 
the French manner,^' we see that this pronunciation was 
very old, and was probably confined to this single word. 

1701. Jones simply identifies a, au, aw in aU, Paul, awl. 
But he gives the foUowing list of 

words in <Hi, '^ which many sound as with an o. Auburn, auction, 
audacious, audible, audience, audit, auditor, auf awf, augmont, 
augre, August, aumber, aumelet, aunt, auspicious, austere, authen- 
tick, author, Autumn, auxiliary, because, cautious, centaury, daunt, 
Dauphin, debauch, fault, flaunt, fraud, herauld, Henault, jaundice, 
laudable, maudlin, maugre, nauseous, Pauls, plausible, restauration, 
aausage, ribauldry, vault." 

He does not say whether the o is long (oo) or short (o). In 
sausage we now use (a), and frequently in because (bikAz*, 
bikoz'), but ai^ awf is now written 'oaf {oo{). Dauphin is 
frequently pronounced as French (DoofeA). The cases in 
which Jones finds al written for au will be considered under 



Chap. m. §3. AU — XVIUTH OU — XVI TH CENTURY. 149 

Xi ; and those in wMch au is written as a written before M. 
1^^ B will be considered under those letters. 

In the xvn th century, then, au was almost imiversally 
pronounced (aa), but there were a few exceptions, so 
that on the whole the rules resembled those now in use, 

AU — xvniTH Century. 

1704. The Expert Ortuographist take the soimd of au 
for granted, and must have im>nounced (aa). The following 
with the sound of (aa) are noteworthy, sausage, taunt, vaunt, 
launcet, launch. 

1766. Buchanan has (aa) in daw, maw, awe, vault, daunt, 
fault, taunt, but has (se®) in aunt, laugh, where Sheridan 
has (2d). 

1768. Franklin has (Iaz) meaning probably (Iaaz) laws. 

The usages with regard to au seem to have been nearly 
the same in the xvni th century as in the xix th century, but 
Hie orthoepists of the xvinth ignore the sound (aa) alto- 
gether, and consequently do not notice the sounds (aant, 
laaf), which are now extremely prevalent, and probably were 
frequently heard during the preceding century. Our pre- 
sent orthoepists reject the sounds also. 

OU — XVI th Century. 

1530. Palsoratb says : '' Ov in the frenche tong shalbe sounded 
lyke as the Italians sounde this vowel v, or they with vs that 
sounde the latine tong aright, that is to say, almost as we sounde 
hym in these wordes, a cowe, a mowe, a sowe, as Mtre, sovddyn, 
whlUr, and so ofsuche other." 

The OU in French is called " ou clos" and sometimes " o 
clos" by Meigret, which would lead to suppose it rather («h) 
than (u), see p. 131, note. There can be no doubt of the 
Italian u, which was certainly (uu). But it seems from 
other writers that this pronunciation of (kuu, muu, suu), 
although still heard in the North of England, was going out. 
Palsgrave's pronunciation is probably of the xvtn century 
in this point. We shall see that these words were so pro- 
nounced in the xiv th century, and it will hence be most con- 
venient to defer the consideration of the change of (uu) into 
(ou) to the next chapter. We are not to suppose that ou 
was universally pronounced as (uu), even by Palsgrave and 
older writers. In many words, ow derived from ags. aw, was 
called (oou). Palsgrave says in another place : 

" K m or n foUowe next jrfter o in a frenche worde both in one 
syllable, than shall the o be sounded almost lyke this diphthonge 



150 OU — XVI TH CENTURY. Chap. III. } 8. 

OY, and Bomethyng in the noose: as iliese wordes mon^ ton, Bon^ 
rendm, shalbe sownded movn, tovn, sovn, renovm and so of all snche 
other, and in like wyse shall o be sownded though the next syllable 
folowynge begynne with an other m or n, as in these wordes hdme^ 
sdmme, hdnne, idMrre, whiche they sonnde haumef houne, toumme^ 
taunner, and so of suche other." 

Meigret knows nothing of this, but the effect on English 
ears is important in the transference of French words to 
English, where on, when, at present, nasal, became otm, 
meaning (uun), which afterwards, as we shall see, fell into 
(oun). Thus Hart in giving the pronunciation of the 
French Lord s prayer, writes (tun, num, volunte', kum*aH, 
dun'e, pardun'aH, pardun*unz, imt), for, ton, nom, volenti, 
comme, donnez, pardonne, pardonnons, ont. 

1547. Salesbury gives no special article on ou, but he has 
the following words, involving this combination, which may 
be classified as follows, 
(oo) BOWE, ho (boo) arcus; cbowe hro (kroo) comix; tbowb tro 

(troo) opinor. 
(o) HONOURE onor (on'or) honos; — probably a mistake for omor 

(on-ur). 
(uu) wowB, w (uu) petere ut procus ; — a "Welshism for (wuu) now 

written ivoo, 
(u) KARBowE, tMtrw (naT'u) angustus; sparowe, sparw (spar'u) 

passer ; oractouse, grM%w$ (graa*si,iis) comis ; ehpeboure, em- 

perwr rem-perur) imperator ; double, dwhyl (dub**!), see also 

imder (ou). 
(ou) LOW hu) (lou) mugire; nowe now (nou) nunc; thott ddow 

(dhou) ; bottble u dowhyl uw (dou'btl yy), see also under (u). 
It is evident that ^' the (uuz) have it," but the (ouz) are 
in force. Those words marked (oo) by Salesbury were pro- 
bably (oou), as at present, but the (u) was possibly faint and 
disregarded. 

1555. Cheke says : " foule, boule, houle ij>ov\ fiovX 6v\ ful bul 
hul latinum u est. nam lumen nuntij acute argute XHfieu vouirru 
OKHre apyovre sic Greece transferuntur." 

Since Mekerch in taking the passage transfers it thus 
'^moule concha, douken panni, /ahX, h>ifc mul duk u Lati- 
num est," and we know that in the old Dutch words^ cited 
ou was (ou) or (ou), we see at once that these scholars were 
led awajr by their interpretation of the Greek ou as = (ou), 
to imagme that the Latin u had the same sound, instead of, 
conversely, from the known (uu) sound of Latin u conclud- 
ing the (uu) sound of Greek ov. In Cheke's time then the 
English "foule, boule, houle" were (foul, boul, Houl). 

1 The modem forms are mouwy moudy motU^ {ma% mcmd, mol*e), and dock (duuk). 



Chap. III. § 3. OU — XVI TH CENTURY. 151 

1668. Sir T. Sboth folly endorses Cheke's inference tha^t 
Hie Latin long u was pronounced as he pronounced Greek 
ov, that is, (ou), saying : 

*' or dipMhongua Qraca, (ou) et aov, (oou). Ex (o) breui & (u), 
diphthongum habebant Latini, quaB si non eadem, vicinissima cert^ 
est ov Graecte diphthongo, & proximo accedit ad sonum u Latinae. 
Ita quae Latin^ per u longum scribebant, Graeci exprimebant per ov. 
quaB per u breuem, per v, quasi sonos vicinissimos. At ex (oo) 
longa & (u) diphthongus apud nos firequene est, apud GraBcos rara, 
nisi apud Jonas : apud Latmos baud scio an f^t vnquam in vsu. 

(ou), (bou) flectere, (boul) spbasra, (kould) poteram, (mou) meta 

foeni, (sou) sus fasmina. 
aw. (boon) arcus, (booul) sinum aut scapbium, (koould) frigidus, 

(moou) metere, aut irridere os distorquendo, (soou) seminare, 

aut suere." ' 

And again in bis Greek pronunciation, be adds : " ov ab omnibus 
recti sonatur, & u facit Latinum quando producitur, vt aduertit 
Terentianus : diJSert tav granditate vocis, vt etiam rjv ab ev dis- 
tinguimus. 
ov, boWf /3oi>, flectere. a hay mow^ fjLov, foeni congeries, a gowne^ 

70W, toga. 
(OV. a bow, /Saw, arcus. to mow, fuov, metere, vel os torquere. yow, 

yaov, abeamus. 
V, T breue LatiQum. a buU taurus. u longum vel ov, a howl, 

fiovk, globus, dov, a howle /Saovk, Sinum ligneum, yas in quo 

lac seruatur, vel vnde ruri bibitur. 

Here Smith agrees with Salesbury in the close diphthong 
(ou), but distinguishes an (oou) where Salesbury only heard 
(00) as in bowe, arcus. In the same way at the present day, 
very few of those who say (boon) acknowledge the final (u), 
because most of them insert it in no, go, etc., saying (noon, 
goon) for {noo, goo), and hence consider that they pronounce 
simple (oo) in both cases. Very few would say (ei noon noo 
hoon BOO loon) for / know no bow so low, or would distinguish 
no beau as {noo hoo) from know bow (noon hoon). Smith at 
the same time absolutely disagrees with Palsgrave in mow, 
sow, saying (mou, sou) where the latter says (muu, suu). It 
is singular that this difference, to which we shall have to 
allude again presently, turns upon precisely the error con- 

^ At present it \s usoal to distrngiuBh clothys or o]>er sedys ' in Promptorinm, 

$ow seminare, tew snere, which would 'I sowe with a nedell* in PalseraTe, 

lead to saying (soon, sen). We find while Levins gives both sewe and «atr# 

for tow seminare ' sowyn come, or any for auertf and does not appear to give 

o^er sedys' in the Promptorinm, *■ 1 the English for seminare at all. Pro- 

Bowe come, or any other seedes* in bably Levins's sowe should hare been 

Palsgrave ; and for sew suere, ' sowe explained seminare. 



152 OU — XVI TH CENTUKY. Chap. III. § 8. 

oeming Greek ov. Although there were then liying persons 
who pronounced (uu) for ou, yet Cheke and Smith ^tn refer 
their sound (ou) to the Greek ov, and then infer the mon- 
strous conclusion that the Latins pronounced their long u in 
the same way. 

1569. Hart, in the passage already quoted, p. 132, writes 
fowerthy (fouurth) totter (touur),j90tire ('pounr), flower (flouur), 
marking the second element of the diphthong as long. 
There is no doubt that in prolonging a dipnthong the second 
element must be lengthened, because the first and the glide 
must pass in the usual time in order to preserve the character 
of the diphthong. As however the lengthening of the 
second element is accidental, it is not usually marked in 
palaeotype. In the course of his work, however. Hart does 
not mark the second element as lone ; for example I find, 
(nou, sound). Hart also leaves out me (u) occasionally as 
(vo,elz, knoon, thoH't, knoledzh,) for rowels, known, fhotight, 
knowledge. Hart also writes (dub'l) for double, thus agreeing 
with one of Salesbury's notations for this word. 

1580. BuLLOKAR in writing of the sounds of o (supr^l 
p. 93) says that the third sound is *' as, r, flat and short, that 
IS to say, as this sillable ou, short sounded." Again, imder 
u he talks of one of the vowel sounds of u being " of flat 
sound, agr^ing to the olde and continued sound of the diph- 
thong : ou : but always of short sounde.'' This he distin- 
guishes by writing a hook, like a comma below, which will 
be here, for convenience, printed as a comma before. He 
then identifies in his notation o,v o,u ,ow ,oow ,v ,u ,o ,oo, 
where the two o's are united into one sign like Greek oo, 
observing '' that no diphthong is of so short sounde as any 
short vowell, and that as well short vowels, as diphthongs 
ending a sillable, are of meane time, that is, betw^ene short 
and long, their time before shewed notwithstanding." The 
following are some of the words in the ordinary spelling in 
which he uses these notations sum, sound, doubty other, fully, 
some, such, without, precious, youth, good, much, under, colour, 
unwilling, comfort, double, vowels, come, but, word, our. With 
the exception of sound, doubt, without, vowels, our, which 
have now (eu) and youth which has (uu), all the above words 
have now (e), and it will be shewn under U that we may infer 
an elder (u) or {u) from a modem (e). There is therefore 
no doubt that BuUokar pronounced ou as fu) at times; at 
other times I think it must have been (uu), for he would not 
have used the phrase *^*ou short sounded" imless there had 
been an **ou long sounded." Thus it is probable that the word 



Chap. III. § 8. OU — XVI TH €ENTUKY. 153 

voweb was called by him (yuu'elz) rather than (yu'elz). We 
have here then a direct confirmation of Pakgrave and con- 
tradiction to Smith. Thus bow flectere = (bou) in Smith, 
and (buu) in Bullokar, both giving bow arcus as (boon). 
We are reminded here of the distinction between the Enff- 
lish (bdu) and the Scotch (buu). Again bowl sinum is (boom) 
in Salesbury, Smith, Bullokar ; but bowl sphsra, is (boul) in 
Smith and (buul) in Bullokar. The celebrated bowling 
greens at Nottingham are commonly called (bou'ltq) or 
(bou'liq griinz) to this day. Walker says on the word bowl 
sfdisera, which he calls (1x?^l) meaning (be^ul) : 

" Many respectable speakers pronoimce this word so as to rhyme 
with howl (Haul) the noise made by a dog. Dr. Johnson, Mr. Elphin- 
stone and Mr. Perry declare for it ; but Mr. Sheridan, Mr. Soott, 
Dr. Kemick aad Mr. Smith, pronoimce it as the vessel to hold 
liquor, rhyming with hole (bool, booul). I remember having been 
corrected by Mr. Gkurick for pronouncing it like howl\ aad am 
upon the whole of opinion, that pronouncing it as I have marked 
'% (bool), is the preferable mode, though the least analogicaL" 

Walker derived his knowledge entirely from observing 
the spelling and custom of his time. Hence his argument is 
perfectiy groundless. Bowly the cup, is connected with boU^ 
bole, and the soimd of (oo) is to be expected, the additional 
(u) arising merely from the following If as will be shewn 
under L. But bowl, the ball« was the French boule, correctiy 
written boul or bowl in older English, not only as we see 
from Bullokar, who caUs this sound of ou its *' old and con- 
tinued sound,^^ but as will appear from the study of Chaucer's 
orthography. The change of (uu) into (ou) in English, 
which occurred Mrtly perhaps in the xvth century, but 
which we see by Palsgrave and Bullokar, was not fully com- 
pleted in the xvith, and which the words through, youth, 
you, a wound some say (a wound), could, would, should, flowk 
(a flounder), soup, group, rot^e, route, occasionally called 
(rout) like rout, Coivper, only called (kou'per) by those who 
do not know the family, Brougham, (Bruum) as spoken 
1^ Lord Brougham, though the carriage is often called 
(Bro<?*9m), will convince us that the change is not yet com- 
plete. The nature and laws of this change will best be 
considered hereafter.^ 

^^ Walker contiiraes as follows, and ties of observing. " But as the yesiel 
it is worth while, perhaps, m a note, howl has indisputably this sound it is 



to draw attention to the extreme con- rendering the langua^ still more irre- 

ftision of ideas concerning language gular togiye the ban howl a different 

that possessed this respectable ortho- one." That is, because in early times 

episti because it is still widely preva- of our orthography, when the writer 

lent, as I have had frequent opportuni- did not know exaetly how to represent 



154 



OU — XVI TH CENTURY. 



Chap. III. { 8. 



1621. Gill agrees with Smith, and writes : (bound) bound, 
(sound) sound, (blooun) blown, (throoun) thrown, (bou) 
bough, (boon) arcus, (boul) bowl a ball, (booul) bowl a cup. 



tiie sound of (nn), but wandered be- 
tween and ouy ow, which last hap- 
pened to be also appropriated to soonds 
which were distinctly (oon), — and be- 
cause people following the tendencies 
of sound, ouite independently of spell- 
ing, alterea the sound of (uu) in many 
wordes to (ou, an), so as still to keep 
up a distinction in speech between 
words previously distinguished though 
in a different way, — all these tendencies 
are to be eiyen up for the sake of a 
casual similarity of spelling ; and it is 
to be deemed 1ms irregular, because the 
spelling is alike, to change the sound 
of one of the words, than to give a dif- 
ferent sound to two words spelled alike, 
or to change the spelling of one of 
them. Of course, then, know now should 
be pronounced alike, as also the latter 
paits of shoe, hoe^ changed hanged. The 
irregularity was not in the sound but 
in tne clumsy orthography. Walker 
proceeds thus, " The inconYenience of 
this irregularity is often perceived in the 
word bowj** the irregularity was spelling 
two words, i.e. two collections of sound 
in the same wa^ ; Walker assumes it 
to be, pronouncmg one word, i.e. one 
collection of letters, in two ways. 
The confusion of writing and sound 
could not be more complete. "To 
have the same word" i.e. sound, "sig- 
nify different things, is the fate of lul 
languages; but pronouncing the same 
word " i.e. written symbol, " differently 
to signify different things, is multiply- 
ing difficulties without necessity to 
the reader, not the listener, and the 
remedy is with the writer, not the 
speaker, " for though it may be alleged 
that a different pronunciation of the 
same word " i.e. written symbol, " to 
signify a different thing, is in some 
measure remedving the poverty and 
ambiguity of language*' i.e. written 
symbols, " it may be answered, that it 
is in reality increasing the ambiguity" 
of orthography, not of language, " "by 
setting the eye and ear at variance, 
and obliging the reader to understand 
the context before he can pronounce 
the word." A good aipiment against 
unphonetic spelling. But to conclude 
that pronunciation must follow the un- 
phonetic spelling, is to determine that 



every baby should learn to read before 
it speaks. This would almost beat 
those celebrated Irish infants of whom 
a native preacher is said, by Sir Jonah 
Barrington in his Memoirs, to have 
declare, inveighing against the pre- 
cocious wickedness of his times, tnat^ 
* little children who could neither walk 
nor talk, ran about the streets blas- 
pheming.' Walker continues: "It 
may be urged that the Greek and Latin 
languages had these ambiguities in 
words" written symbols, " which were 
only distinguished by their quantity or 
accent." That is, words differing in 
the accent given to the syllables, or in 
the length of vowel sounds were written 
alike — a defect in orthography, but 
certainly not in the language which 
distinguished the sounds. "But it is 
highly probable that the Greek lan- 
guage 1^ a written accent to ^stin- 
e-ui^ such words as were pronounced 
differently to signify different things," 
as the Greek accents were an invention 
of later grammarians chiefly to assist 
foreigners, it would have been more 
satisfactory if Walker had mentioned 
the grounos of this ' high probability,' 
" and this is equivalent to a different 
spelling," of course, when the accent 
points to a difference of sound, and is 
not merely, as old Bullokar often used 
it, and as we find in French a, d, * for 
the sake of equivocy,* just as we may 
imagine Walker would have looked on 
the diverse spellings n7^, writet rights 
Wright f or air, heir, eyre, ere, Ifer. 
Walker continues, "and though the 
Latin word lego signified either to read 
or to send, according to the quantity 
with which the first syllable was pro- 
nounced," that is, the word (leg'oo^ 
meant / gather or read, and the wora 
(leeg-oo) meant / send, and the two 
words were in this particular inflection 
written alike, " it was certainly an im- 
perfection in that language," read, or- 
thography, " which ought not to be 
imitated. Ideas and combinations of 
ideas wUl always be more numerous 
than words; and therefore the same 
word will often stand for very different 
ideas ;" and Walker has in this note 
strangely illustrated the danger of such 
results in bad writers and loose thinkers, 



Chap. III. § 3. OU — XVI TH CENTUEY. 155 

He has However some remBants of the Yuu, u) sounds, as 
(kuurts) courts, (kuuld) could, where Smitn has (kould), and 
admits (wound) as a Northern pronunciation of wound. 

1653. BuTLEB says (translating Ms symbols,) : ^^ouia the substan* 
tive termination ouTf as honour , labour^ succour , and in the adjective 
termination oua^ as glorious^ gracums, prosperous is sound as oo or m 
short" that is (u) or (w). "This being general, may be suffered 
as an Idiom : but in otner syllables of some few words, whereof 
there is no certain rale to be given, it is not so excusable : as when 
we write hloud^ floud, courage^ scourge, flourish^ nourish, young, youth^ 
wouif, double, trouble, &c., for blood, flood, coorage, scurge, floorish, 
nurnsh, yung, yuth, wulf, dubble, trubble, &c.," meaning (bltid,fltid, 
kur'adzh, skurdzh, flur'tsh, nin"wh, juq, juth. ?, wulf, dub'l, trubi), 
** for the same writing hath another sound in lottd, proud, cour,^ scour, 
mound, mouth, eoul, seoul, doubt, trout, aad the same sound hath an- 
other writing in good, stood, bud, mud, burge,* purge, furrow, murrain, 
bung, gulf, bubble, stubble, &c.," which had (w). "Neither is there 
any more reason why in would, could, should, roum, wouf? wound, ou 
should be written for oo long; than that for cool, pool, fool, tool, 
school, stdol, hoof, boom, moon, doom; we should write cotd, poul, 
foul,ioul, skotd, stoul, houf, bourn, moun, doum. The cause of this 
cacography which causeth such difficulty is a causeless affectation 
of the French dialect ; who for the sound of oo (which in their 
langaage is frequent) do sometimes write o and oftentimes ou ; as 
they write i, ai, oi, and sound (ii, e, woee),* or as they write en, an, 
aw, and sound an, aun, ow for entend, command, costeau, saying 
antand, coomaund, eoteow. But that they speak otherwise than they 

by confusing a n>okm and a written necked, pedantic, nnphilosophical, mi- 
word, /on^tM^^ ana orMd^ropAy ; "bnt Berably-infonned, and therefore su- 
altering the sound of a word, without premely certain, self-confident, and 
altering the spelling, is forming an self-conceited orthographers who make 
unwritten language. The orthoepist default, when they will not alter the 
the orthographer, the word-pedlar, is spelling ailer the sound has changed, 
here shewn to the life. It is a horror and maintain that though their niles 
to him, a monstrosity, this formation must be right, it is only the exceptions 
of an " unwritten language.** As if which prove them, — forgetting that as 
all languages were not formed un- some foreifper pithily said, ** English 
written, were not to the g^'eat majority orthographical rules are all exceptions.'* 
of present speakers, unwritten. As if * Meaning cowtr, written cowryn in 
aU those who made languages, who the Promptoriuna, ctnore in Fals^ve, 
altered their sounds, who Irought them and eoure in Lerins. 
to their present speech-form, knew or ' Query, borage, as written in the 
cared about writing ; as if even the Promptonum, the bourage of Palsgrave 
majority of those who speak, pause to and burrage of Lerins, exhibiting the 
consider in the rapidity of discourse, three common spellings for the same 
how the printers of the day choose to sound. 

print, and the writing-masters choose ^ Room, woof "of woven, as warp 

to order their pupils to write ! No, it because warped or wrapped round the 

is not the lanfftuage, or the speakers beam" adds Butler, 

that are in fault in obeyinj? ana carry- ^ Butler belongs to the latter part of 

ing out the organic laws of speech and the xyi th or to the xvn th centunr, in 

word formation. It is those word- his French, when the change oi the 

pedlars, those letter-^vers, those stiff- French at from (ai) to (b) was complete. 



166 OU — XVn TH CENTURY. Chap. HI. } 8. 

write, is no reason wliy we should write otherwise than we speak ; 
considering what an ease and certainty it would he hoth to readers 
and writers, that every letter were content with its own sound, and 
none did intrude upon the right of another. The termination our 
accented, is sounded in two syUahles : as in devour, deflour ; and in 
all monosyllahles, as our, hour, hour, Jhur, tour, sour, lour, scour, 
pour Verh fundo : the Noun is, for difference, written in two 
syUahles pouer potestar, and so are all the suhstantiyes in the 
plural numher ; as flouers, touera, Shouers : and sometinie in the 
singular not only m verse : hut in prose also." 



OU — xvn TH Century. 

1653. Wallis says: ** Oi* et ow duplicem sonum ' ohtinent ; 
alterum clariorem, alterum ohscuriorem. In quihusdam vocabulis 
effertur sono clariori per o apertum,^ et to, TJt in sdul anima, 96uld 
vendebam, venditum, mdw nix, kndw scio, s6w sero, suo, 6u>e deheo, 
hdwl poculum, etc., quo etisun sono et 6 simplex nonnunquam 
effertur nempe ante ^ ut in gdld aurum, $c6ld rixor, hdld teneo, 
c6ld frigidus, 6ld senex, antiquus, etc., et ante U in p6ll caput, 
rdll volvo, tdll vectigal, etc. Sed et haec omnia ah aliis efferuntur 
simpliciter per 6 rotundum acsi scripta essent adle, sdld, snd &c. In 
aliis vocabulis obscuriori sono efferuntur ; sono nempe composite ex 
d vel a obscuiis (o), et w (ou). Ut in hduse domus, mStMS mus, 
Idwse pediculus, hdul globulus, dur nostcr, dut ex, du?l bubo, tdwn 
oppidum, fdul immundus, fdwl volucris, hdw flecto, hdugh ramus, 
sdw sus, etc. At would vellem, should deberem, could possem, course 
cursus, court aula, curia, et pauca forsan alia, quamvis (ut proximo 
prsecedentia) per 6u pronunciari debeant, vulgo tamen negUgentius 
efferri solent per oo (uu)." 

Wallis seems to say that (soul, oavld, bucu) as well as {aool, 
Boold, Bjxoo) were heard, and that (govld, slo^uld, b^uld, k^uld, 
ovlA) were used, although he did not approve of them. This 
effect of L will be considered hereafter. The sound (hous, 
meus) &c. is the same as the modem English, and must be 
distinguished from the former. Wallis's dictum concerning 
would, etc., is only borne out by Smith's very peculiar 
(kould) could, suprii p. 151. We have seen that Gill said 
(kuurt); (kuurs) is still common in the North. Wallis 
wishes that the two sounds were distinguished in writing, 

^ This must mean '*^ apertum^** I suspect that this is a theoretical pro- 

that is f a), ^ving the diphthong (au) ; nunciation, arising from Wallis's con- 

althougn it is certainly very singular, sidering the Towel o short in the diph* 

as the words given were pronounced thong and his haying no notation for 

with (oou) in the xvi th century, and (o) The d apirtum he usually marks 5, 

he makes some of them have (oo), but here he has employed 6^ apparently 

This (au) is the diphthong recognized to connect the sound with his <$ s (oo), 

in a row words by Cooper, supr& p. U7. so that he may really mean (on). 



Chap. III. § 3. OU — XVn TH CENTURY. 157 

using Ati dtv OT du dw or simply ow for {on, oo) and du dw or 
simply OU for (eu). Yet how many would feel their eyes 
o£Penaed by seeing know, nou, hou, low, sou, sow, row, rou, 
notwithstanding the infinitesimal nature of the change. 

1668. WiiIkins speaks of (eu) only as the sound of ow in 
** owr, owle.'^ It is curious that, though (aeu) is the common 
Norfolkism now, Wilkins says that (sd) before (u) " will not 
coalesce into a plain sound." Writers on phonetics are too 
apt to measure the pronouncing powers of others by their 
own, although the extreme difficulty with which unfamiliar 
combinations of familiar elements become current to their 
organs, and the mistakes they make in hearing and imitating 
iiTifftmiliRr sounds and slight variations of fiEimiliar sounds, 
should teach them to be less confident. 

1668. Price makes several categories of ou, ow. 

1) oWy OU sound *' like o," that is, either {oo) or (o) in bestow, 
know, a bow, flow, low, window, throw, grow, glow; succour, 
brought, endeavour, although, armour, behaviour, clamour, colour, 
embassadour, emperour, errour, gourd, harbour, mannom*, nought, 
odour, ought, rigour, solicitour, soul, though, thought, wrought ; in 
some of which we have now (a, aa). 

2) Ow, OU keep their ** fall sound" (au) in how, to bow, fix)ward, 
allow, cow, cowMd, now, toward, devout, flout, fourth, our Saviour, 
stout. Although (tau'jd) may be occasionally heard, it is un- 
frequent ; (fr9u*jd\ I do not remember to have heard ; (fauith) is 
also strange, and (sseeB*vi,9Uj) the strangest of all. 

3) Ou sounds ''like short u," that is (o), in cousin, double, 
courage, adjourn, bloud, couple, courtesey, discourage, doubled, 
encourage, floud, flourish, journey, journal, nourish, ougly, scourge, 
touchstone, touchy, young. All these pronunciations remain in use 
although we no longer write hloud, floud, ougly. 

4) Ow, OU sound " like tr<w," that is (uu) in arrow, pillow, 
barrow, borrow, fallow, follow, haUow, morrow, shaddow, sorrow, 
swallow, widdow, willow, winnow, couch, course, discourse, court, 
courtier. 

5) ** Ou soundes like iw in youth," meaning (jiuth) ? This 
certainly ought to have formed part of the preceding list. 

1685. CooPEB says " in fuU, fole (w, oo) cum u (u) conjunctus 
constituit diphthongum in coulter yomiBffour quatuor, mould paniflco, 
mucesco, typus in quo res formatur ; moulter plumas exuere, poulterer 
avicularius, poultry alites villatici, should^ humerus, soul anima; 
in cseteris hunc s<mum scribimus per o ante // flnalem, vel /, quando 
pnecedit aliam consonantem; ut hold audax; quidam hoc mode 
pronunciant ow." 

** U gutturalem (o), ante u Germanicum oo anglic^ exprimentem 
(u) semper scribimus per ow, ui out en; about circa ; ou tamen 
aliquando, prseter sonum priorem, sonatur ut oo (uu) ; ut I could 
possem ; ut u gutturalis (a), couple copulo ; ut a (aa) houyht emptus." 



158 OU — XVn TH CENTURY. Chap. III. } 8. 

The first diphthong mast be written theoretically (tm), but 
it was probably meant to be the same as (on), coinciding 
with Wallis's diphthong, because Cooper does not distinguish 
{Uf o). The second diphthong was of course the modem (au). 

The words in ou which Cooper pronounces with the first 
diphthong (uu) or (ou), as above mentioned, all contain oul, 
and to these he adds the following with a simple o before l, 
behold, bold^ bolster, boU, eold, colt, dolt, droll, enroll, foU^ 
gold, hold, inholder hospes, jolt, knoll, manifold, motten, 
poll, roller, rolh, scold, sold, told, vpholster plumarius. He 
also says : ^^ Quidem scribunt troll yel trowt heyiter eo, ita 
controll controul, redargue, joU jole caput," jowl is common 
now, with the sound (dzhaul), " toll tale vectigal &c, mold 
yel mowld humus, at mould tjrpus," a distinction now lost, if 
it were eyer made by others beside Cooper, **bowl bole 
patera." 

The sound of the second diphthons^ (ou) is giyen by Cooper 
to all other words in ou, as** bout globulus, gout podagra, 
&c," some of which he allows to be written ow, as: ad- 
towson, allow, avow, bow torqueo, bowels, bower, brow, brown, 
browze, carowze, cow, coward, cower, crown, down, dowry, 
drown, frown, gown, how, howl, lower frontem capero, mow 
fsenile, now, owl, plow aro, rowel, rowin fcenum serotinum, 
shower, sow s., towel, tower, trowel, fx>w, trowel. He adds, 
** bounce crepo, bouser thesaurarius, down colonus, drousie 
somnolentus, loud sonorous, louse pedicular, renoun gloria, 
rouze excito, souse omasum, touze plurimiim yello ; etc., scri- 
buntur item cum ow. W quiescens adjungitur post o finale, 
(praeter in do facio, go eo, no non, so sic, to ad) ut bowe 
arcus, dowe farina subacta" i.e. dough, " owe debeo, sowe sero, 
towe lini floccus, &c, & in ovm assero, disown dene^, bellows 
follis, gallows patibulum, towardness indoles." 

Hence Cooper admits (on) but not (oou) making the latter 
purely (oo). He giyes no Ust of words with ou pronounced 
as (e) or {u, uu). 

1686. MiEOE*s lists are as follows: ou generally = o^Tt^, 
meaning (au), not (au), although Miege confuses French a 
with English (aa). 

1) ou = o, meaning (a), in adjourn, bloud, floud, coimtry, couple, 
courage, eourtesey, double, doublet, flourish, goumet, journey, 
Journal, noxuish, scourge, scoundrel, touch, trouble, young, in which 
(skan'drel) is new. 

2) <Hi = " un peu long," meaning (o) or (oo), or sometimes one 
and sometimes the other, or else (ou) which he was unable to ex- 
press in French letters : in coulteri moulter, poultice, poultry, ioxir, 



Chap. m. § 8. OU — XVII TH CENTURY. 159 

course, concourse, discourse, soul, sonldier, shoulder, mould, trough, 
dough, though, although. 

3) OU, value not named, and hence prohably French ou (u), see 
Jones, just beloWi in substantives ending in our as Saviour, factour, 
neighbour. 

4) oti, value not named, probably French ou (u), in adjectives ending 
in ouBf as vicious, malicious, righteous, monstrous, treacherous. 

5) ough = a long, that is (aa) in ought, nought, brought, bought, 
sought, thought, wrought = &t, nat, brat, b&t, &c., (AAt, uAAt) &c. 
except drought, doughty = draout, daouty (draut, dau't*) ; borough, 
thorough = horo, thoro (bor-o, thar'o) ; cough = c4^(kAAf ) ; rough, 
tough, enough = ro/, toff, enof{T9f, tof, enof). 

6) ou = ou French (uu) in would, could, should, you, your, 
source, youth, — Portsmouth, Plimouth, Yarmouth, Weymouth, 
Monmouth. 

1701. Jones says "that ou and ow have two very di£ferent sounds ; 
(1) that in soul, howl, old, told, &c., which is the true sound of o 
and 00 join'd together in one syllable (ou, oovl) ; (2^ that bi bough, 
eow, now, &c., which is the true sound of H short, m hut, cut, &c., 
and 00 join'd together in one syllable (ou)." 

But he characteristically seldom distinguishes which he 
means when he talks of the sound of ou, aw. He also says 
that ou is pronounced o, meaning either (oo) or (o), or even 
(aa) in " GHotAcester, sounded Oloster ; although, besought, 
borough, bough^^ bought, brought, cough, dough, doughty^ 
drought, enough? fought, hiccough, hough, lough, Lougher, 
mought, nought, ought, plough,* rough, slough,^ sought, though, 
thought, through, tougn,^ trough, whough, wrought; and "in 
souldier, sounded sodier,** the parent of the "sojer" of our 
plays and jest books. 

The soimd of ^ is also written ow, Jones savs : " When it 
may be sounded otv in the End of words, or Wore a vowel, 
as Ota, owing; follow, following, &c., otherwise it is always 
0, when it cannot be sounded ow (ou P), unless it be one of 
those above, that are written oughJ* 

Ou = (uu) is much more extended by Jones than by the 
preceding authorities, first to the terminations -our, '0U8 
"when it may be sounded ou," which seems very questionable, 
and then in the following words : ' couch, could, course, court, 
courtship, courteous, crouch, fourth, gouge, gourd, mouch, 
mourn, should, slouch, souse, touch, would; accoutre, amour, 

^ Sorely a mistake. ^ 4, 8, etc., which from this insertion by 

* (Dan'ti) not (d0o*ti) according to Jones would seem to imply a pronnn- 
Miege, and present use. ciation (ploon). But Cooper, suprii p. 

* Meaning (enoo*^ P 158, spells ji^otPf and yet pronounces 
^ The Authorizea Version has plow, (pleu). 

Dent 22, 10. 1 Sam. 14, 14. Job • Now (t9(, slef slen, taf). 



160 OU — XVniTH CENTUKT. V Chap. III. j 8. 

boute/eu, Bourdeaux, eapouch, eapouchine, coupee, courier, 

Courtney, courtrey, eourvee, enamour'd, gourmandise, Louvain, 

Lounre, rendezvous, rencountre, Toulon. For ow = (e), see p. 183. 

Hence in the xviith century ou, or ow had two sounds, 

the first {on) or (oon) corresponding to our present 

theoretical {oo) and secondly (au) where it is still so 

called. The sound o{ ou as (uu) was exceptional, and 

seems to have been used in a few more words than at 

present. 

on — xvin TH Century. 

1704. The Expert Orthooraphist seems to pronounce ou 
as (eu) in touch, Sou^h, gouge, rouge, coulter, boulter, poultry, 
moulter, shoulder, pouUice, wound, pour, bowl, cowcumber. 
But to distinguish botv flecto as (bau) from bow arcus as 
(hoou), and says that "All polysyllables ending in obscure o 
have w added for ornament s sake as arrow, bellows, &oJ* 

1766. Buchanan writes, (uAAt) nought, (mous) mouse, 
(foul) foul, (bou) bow Jlectere, (koun'ti) county, (koutsh) 
couch, (vou'tl) vowel, (sou) sow sus, (boul) bowl globus et 
crater ; (Shoo) though, (koors) course, (koovt) court, {noo) 
know, (bloo) blow, (\mioo') bestow, (sool) soul, (naer-oo) nar- 
row, (8B Xoovi) a low ; (suup) soup, (wud) would, (kud) could, 
(juu) you; (jeq) young, (trab'l) trouble, (kep-l) couple, 
/kor-fdzh) courage, (ken-tri) coimtry, (nar'tsh) nourish; 
(thAAt) thought, (bAAt) bought. 

1768. Franklin writes (fAul, Aur, dAun, thAuz'and, pUu*- 
meen ; \u)Ot%) ioxfoul, our, down, thousand, ploughman^ course, 
where if (au) is not a mistake, it is a singular form of the 
diphthong, agreeing however with the an^ysis of Sheridan 
and Knowles. 

Among the Irish uses noted bv Sheridan, 1780, we find 
(kuurt) court, (suurs) source, and (kAuld, bAuld) cold, bold, 
all of which clearly belong to the xvii th century. Sheridan 
pronounces (koort, soors, koold, boold). The Irish (druuth) 
drought, English (drAut) according to Sheridan, is very 
singular. 

XT — Round or Labialised Yowels. 

U has been reserved to the last, as in order to understand 
the relations of the various sounds which have been ex- 
pressed by u in our own and other languages, especial attention 
must be directed to the twofold manner in which the aper- 
ture of the mouth is varied. Speech soimds are essentially 
produced in the same manner as those in organ reed pi;>e6. 



GXAP. ni. {S. V — BOUND OB LABIAJLISBD VOWELS. 161 

In the larynx two highly elastic vocal ligaments, stretched 
to various degrees of tension at will, are put into vibration 
hj the rushing of wind from the lungs through the wind- 
pipe. The sound thus produced is highly complicated, 
cxmsisting, as Helmholtz has shewn,^ of a great number of 
simple tones, producing on the whole a buzzing, droning, 
imperfect effect, which would not be well heard. To make 
it penetrate as a clear distinct sound, a resonance tube must 
be added. This tube, according to its shape or length, will 
reinforce a greater or less numW of simple tones, which it 
selects out of the confused number produced by the unarmed 
elastic ligaments, thus generating, by the mere change of 
its shape and size a marked change in the soimd heard, even 
when the original mode of vibration remains xmaltered. 
Now above the larynx is situated a highly variable fleshy 
bag, the pharynx, communicating with two external aper- 
tures, the nose and the mouth, either or both of which can 
be opened or closed at will. The back nostrils are the 
entrance and the external nostrils the exit from the upper 
passage, where the sound passes through various fl;allerieB 
and encounters various membranes, whicn produce the well- 
known nasal modifications. The lower passage or mouth is 
principally modified by the tongue, which acts as a variable 
plug, and the lips, which form a variable diaphragm. By 
this means the volume of the mouth is divided into two bent 
tubes of which the first may be termed the Ivngiml passage 
as its front extremity is formed by the tongue, and the 
second, the labial passage. When the labial passage is large 
and imconstrained by rounding or narrowing of the labial 
orifice, the effects may be called simply lingual, and when 
the tongue is broug^ht so low as to remove the separation 
between the lingual and labial passages, the effects might 
be termed labial. Mr. Melville jBell has acutely preferred, 
however, to consider as lingttal all positions in which the 
labial aperture produces no sensible effect, and then to con- 
sider the labial effect to be superadded to the lingual, hj 
more or less rounding the lips while the lincual position is 
held. It was not generally noticed before tne publication 
of his Visible Speech, that the two labial vowels, as they have 
been called, (uu, oo) really required a distinct position of the 
tongue in order to produce them.^ This however may be 

^ The onlj satisfiMtory Moount of edition 1863, 2iid ed. 18S5. It has 
nrancal and Tocal tones which has yet been translated into French, bnt, HD- 
been published will be found in Heun- fortunately, not yet into English, 
holti 8 Lehre Ton den Tcmempfindnn- ' See howeyer the subsequent r^ 

gen, Bninswicl^ 8yo, pp. 600 fint foeoee to Holder, 1669, p. 178. 

11 



162 V — ROUND OR LABTALISKD VOWELS. Chap. III. } 8. 

practically felt by producing these sounds, and, while utter- 
ing them, seizing the upper and lower lips with the two 
hands and rapidly separating them. Two new sounds will be 
produced, of which the first (obcb) is a Gaelic vowel, which is 
the despair of most Englishmen, and the second is a sound 
(a) often given to our short u in but, and considered by Mr. 
M. Bell as its normal sound. On producing the effect, which 
after a little practice can be obtained without the use of the 
fingers, it will be found that the back of the tongue is much 
hi^er for (a) than it is for (a).^ Although both effects 
are different, and also different from the sound with which 
I pronoimce u in but, namely (a), few English ears would 
readily distinguish (ci? a a) in conversation. Hence we 
have this relation between (u) and (a), that (u) is almost (a) 
labialized or rounded.' 

Again, for the common vowels (ii, ee) the lingual passage 
is greatly reduced by means of the front of the tongue which 
for (ii) IS brought very near the palate, and very forward 
but not quite so forward for (ee), the lips being wide open. 
Now round the lips upon (ii, ee) and the effect is (i, 9), one 
a sound often heard m Germany for ii and in Sweden for 
y, and the other heard for the so-called French e mute when 
soimded and prolonged in singing, as heard in heur and the 
first syllable of heureux.^ 

It is now necessary to attend to a third, modification, 
principally in the pharynx. This consists in widening the 
Dag of the pharynx and all the lingual passage behind the 
narrowest aperture, and also increasing the volume of the 
labial passage. We are familiar with this in English in 
the passage from (i) to (t), and from (e) to (e). Applied to 
the rounded or labialised forms of these vowels, (i, 9) it con- 

^ In reading this discussion the dia- labialisation of (t) and assigns the latter 

grams of the vowel positions in the In- yalne to the French m^ which I have 

&odnction, p. 14, should be freqnently been in the habit of pronouncing as the 

consulted. wide of (#). Thus htureux according 

s The true sound of (9) has the back to Feline has the first syllable as in ^ 

of the tongue lower and the firont and the second as in Jeu, These I 

higher than for (a) ; the tongue is pronounce Tzhp, ihce), but M. Fayarger 

altogether raised, but is nearly parallel considers they shonid be (zh^h, z£»). 

to the palate throughout. The labial Undoubtedlj tbe sounds yaxy from indi- 

or* round' form of (a) is (oh), scarcely yidual to indiyidual, and hence the 

distinguishable from (0) by unpractised necessity of a diagrammatic yowel scale 

ears. like Mr. Melyille BeU's, which is inde- 

> Mr. M. Bell giyes it as the French pendent of key words. The Swedish u 

If in unsy but this is not my own pro- or (u) which is yery peculiar is closely 

nundation, nor does it a^ree with my related to (i), being produced in tiie 

own obeeryations. H. Fayaiger con- same way, wi^ rather a greater sepa- 

liders the French # mutt to be (#h) the ration l>etween the tongue and tht 

labialiaation of (b), rather than [t) the palate. 



Chap. III. i S. u — XVITH CENTUBY. 163 

verts them into (7, cb), which are the common forms, as I 
hear them, of the French u in une and eu in jeu. Hence (7) 
is the * wide' form of (i), and the * round' or labialised form 
of (»)• If we appl7 the widening to (u, 0) we produce (i#, o), 
and the Italian chitMO or (uh) appears to be the 'wide' 
form of the Swedish (u) alread7 described. 

We can then understand that (u, u) ma7 be readil7 oon<* 
fused, for no modification is so subtle as that produced b7 
the backward widening. Again, b7 merel7 neglecting to 
labialise, (u, u) are converted into (a?, v), both of which are 
confused with (a) b7 Englishmen. The last, {b), is indeed 
a ver7 common sound in English, but it is onl7 looked upon 
as unaccented or indistinct (e), in motion, ocean, etc. 

Again, if when we are pronouncing (u) or (u) we suddenl7 
throw the front of the tongue up to the (i) position without 
altering the form of the Ups, we obtain (i) or (j). There 
are some persons so used thus to throw up the iront of the 
tongue that the7 have great difficult7 in pronouncing (u) at 
all. To succeed the7 must exercise themselves in keeping 
down the front of the tongue b7 a muscular effort. 

Boughl7, we ma7 sa7 that (a) is (u) deprived of its labial 
character, and that (7) is (u) with a palatal character, 
or that (7) is an attempt to pronounce both (i) and (u) 
at the same instant. The further step, then, to pro- 
nouncing first (i) and then (u)^ producing (iu), is eas7, 
and since the (i) character predominates and gives the 
ke7 to the soimd, it would be natural in the absence of 
a proper sign for (7) to represent that sound b7 (iu), 

XJ — XVI TH Centukt. 

1530. Palsgrave says : '' 27, in the frenche tong, wheresoeuer he 
is a vowel by hymselfe, shall be sownded like as we sownde ew in 
these wordes in our tong, rew$ an herbe, a mew for a hauke, a clew 
ai threde, and such lyke restyng apon^ the pronoonsyng of hym: 
as for these wordes plue, nul, fue, ueir, Mimhle, uerii, tiiiey sound 
plevuB, nevul, fevue, evuser, hevumhle, uertevu, and so in aU other 
wordes, where 1; is a vowel by hymselfe alone; so that in the 
soundynge of this vowel, they dLffere both from the Latin tong and 
from vs." 

On referring to EU, p. 137, it will be seen that Palsgrave 
divided the English eu into two categories, trewe^ glewe, rewe^ 
mewe and clew having the sound of the French u, and dewe, 
shrewe,/etae having the sound of the Italian eu. The latter 
we have identified with (eu). There can be but little doubt 
^ Hispnnt forigMii. 



164 V — XVITH CENTUBY. Chap. HI. } a. 

that the former was (y), beoauae we know from Meigret that 
it was not (o) or (u). 

When Palsgrave here says that the sound of French u 
was different from that of Latin or English u, he must mean 
by the latter, English u short, because English u long was 
certainly not the same as the real Latin u long, even at a 
much earlier period than the xvith century. Hence cor- 
roboration, and contemporary explanations, are necessary. 

1547. Salesbury says : " xr vowel, answers to the power of the 
two Welsh letters «, w and its usual power is tnr, as shewn in the 
following words tsite truto yems, tebtue pertuw probitas. And 
sometimes they give it its own proper soimd and pronomice it like 
the Latins or Uke our own w (u), as in the words buckb hwek (bnk) 
dama mas, lttsi host (lust) libido. But it is seldom this vowel 
sound corresponds with the sound we give the same letter, but it 
does in some cases, as in bust bust, occupatus aut se immiscens." 
Again in his pronunciation of Welsh he says: *^u written afber 
this manner u,** that is, not as v which was at that time inter- 
changeable wiUi u in English and French but not in Welsh, '* is a 
vowel and soundeth as the vulgar English trusty ^tuy, hus^y Huher- 
dm. But know well that it is neuer sounded in Welsh, as it is 
done in any of these two Englyshe wordes (notwythstanding the 
diuersitie of their sound) iure, lucke. Also the sound of « in 
French, or U with two pricks over the heade in Duch, or the 
Scottish pronunciation of u alludeth somwhat nere vnto the soimd 
of it in Welshe, thoughe yet none of them all, doeth so exactly (as 
I thynk) expresse it, as the Hebraick Eubuts doeth. For ^e 
Weli^ u is none other thing, but a meane sounde betwyxte u and y 
beyng Latin vowels." 

The precise value of the Welsh u is considered ip a note 
on the above passage, chapter VIII, § 1, where it is shewn 
that it must be considered as the Welsh representative of (y), 
and that (i) or (p) is practically the sound it receives. If 
then Salesbury had to rq>re6ent the sound (yy), he could not 
have selected any more suggestive Welsh combination than 
tnc. To have written uu would have been to give too much of 
the (i) or (t) character, for when u was short he did not dis- 
tingiiish the sound from (0, a« shewn by busy which he writes 
busi, meaning (btz'i).^ If he had written tow he would have 
conveyed a completelv false notion, and iw would have led to 
the diphthong (lu) which he wished to distinguish from uw. 

1 Gennans who distmgnish their u and often so pronoonoed bj the Welsh 

from (ii) yery dearly when it is long, in familiar conversation. In the same 

readily pronounce snort u as (0 es* way StieU handles and Stii/Ue chairs, 

pecially when r follows, as (hhir'd^ for are identified in the common Dresden 

(bhyr'dj, bhn-d^^ umrde. The Welsh pronunciation of German. 
u long is heard by Englishmen as (ii) 



Cha». m. { 8. u — xviTH cxirruRY. 166 

Now my own Welsh master at Beaamaris told me that 
Welsh Duw and English due, dew were so distinct to a 
Welshman that he could tell an Englishman inmiediately 
by his faulty pronunciation* The dirorence may be (dm) 
Welsh and (diu) English. It is yery difficult to seize, ana 
some Welshmen themselves deny the difference.^ 

Ad(^tiDg then the hypothesis that Salesbury's uw meant 
(yy), but his u short meant (t), so far as the English sounds 
which he wished to imitate are concerned, — an hypothesis 
which agrees with Palsgrave'^s remarks and will be confirmed 
shortly — ^we may represent aU the EngUsh words containing 
u, (or ew pronounced as u, according to Palsgrave's intima- 
tion,) which are transcribed by Salesbury, as k)11ows. 

Chubchs UwrU tsiurtt (tshtrtsh) ecdesia; Dm duwk (dyyk) 
dux, suFFBE Bw^er (suf *er) srnere, gitttb gwt (gut) viscera ; Jbsit 
tnemw (Dzhee'zyy) ; bvckb hwck (buk) dama mas ; bull ho (buu) 
a rustic pronunciation, qttene hwin (kwiin) regina ; quabtsb hwarUr 
(kwar'ter) quarts pars ; xuse muwtos (myyz) meditari ; ibesubb 
iretuwr (tTez*yyr) thesaurus ; true truw (tryy) vems, this is one of 
the words cited by Palsgrave, under the form trewey as containing 
the sound of the rrench u (y) ; vbbtub v^rtmo (veriyy) probitas ; 
lust host (lust) Hbido ; bust tmsi (biz'i), much good no rr you my eh 
goditio (mttsh god'itio). This much contracted phrase is also given 
by Cotgrave, 1611, who writes it musktdiUi, meaning perhaps 
(mtM'ktdit't), and translates much good may doe unto you} 

1555. Cheke says : " Cum duke tuke lute rebuke Svk tvk 
\vr pePvK dicimus, Graecum v sonaremus." Of this Greek v 
he says " simplex est, nihil admixtum, nihil adjunctum 
habet, and it was therefore a pure vowel, with which he 
identifies the English long u, Mekerch in adopting Gheke's 
words changes his examj^les thus, ^'quum Gallic^ mule, id 
est mula, Belgicd duken, id est abscondere, iwK Svk dicimus, 
Graecum v sonamus.'' Mekerch, therefore, intending to give 
the sfune sound to Greek v as Cheke did, makes it (jj). 
This was the sound which Cheke identified with English 
long u and declared to be a simple sound, that is, not a 
diphthong. 

1 Dr. Benjamin Davies oonld see no JVur. I speake no treason, 

difference in ordinary conyersation, but Father, Oodi^oden, 

admitted that one was attempted to be which is transuterated in the Globe 

made in ** stilted utterance, and then editacm, act iM, sc. 6, y. 173, 

it seemed to me to be like (dnl)« Ifur* 1 speak no treason. 

* The same writer gives as tiie oon- Cap. 0, God ye god-den, 

traction for Ood give you food ovmrng, an erident mistake, as Oodi" is a oon- 

Godi^odin, meaning perhaps (Godi- traction for Ood g^you. The sentence 

gndiin*). In Romeo luid Juliet, Folio should be as much wrapped up into 

1628, Tragedies p. 70 ooL 1, we find one word, as the ordinary good byo. 



166 V — XVI TH CBHTUEY. Chap. m. { 8. 

1568. Sir Thomas Smith is still more precise and circum- 
stantial. He says : 

'* T yel V GrflBcum ant GaUicum, quod per se apnd nos taxnm 
arborem dgnificat. tazos v" meaning that yew := sound of Greek v ; 
Le. as he immediately proceeds to shew, and as I shall assnme in 
tranBcribing his characters, yew = (jj)y though perhaps this par- 
ticular word was (jyy). Tlie following are his examples : ** (snyy) 
ningebat, (slvy) occidit, (tryy) verum, (tyyn) tonus, (kyy) q. litera, 
{ryj) ruta, (myy) cavea in qua tenentur accipitres, (nyy) noTum ; 
(tyyUy yaletudinarius, (dyyk) dux, (myyl) mula, (flyyt) tibia 
Germanorum, (dyy) debitum, (lyyt) testudo, (bryy) ceruisia facere, 
(myy-let) mulus, (blyy) caeruleum, (akkyyz)* accusare." 

In this list we have true, rue, mew, which are the same as 
Palsgrave's examples of eto sounded as French u ; and duke, 
true, the same as Salesbary's examples of u sounded as 
Welsh uw. This would identify both sounds with (yy) if 
we could be satisfied of Smith's pronunciation. Now he says 
explicitly : — 

'* Quod genus pronunciatioms nos k Gallis accepisse arguit, qubd 
rarius quidem nos Angli in pronuntiando hac utimur litera. Dcoti 
autem qui Gallica lingua suam yeterem quasi obliterarant, et qui 
trans Trentam fluyium habitant, yieioioresque sunt Scotis, frequen- 
tissim^, adeo yt quod nos per Y Eomanum sonamus (u), illi libenter 
proferunt per v GrsBcum aut Gallicum (yy) ; nam et hie sonus tarn 
Gallis est peculiaris, ut omnia fere Eomane scripta per u et v pro- 
ferunt, yt pro Dominus (Dominyys) et lesys (Jes-yys),' intantum 
vt quae breyia sint natura, yt illud macrum v exprimant melius, 
sua pronunciatione longa faciunt. Hunc sonum Anglosaxones, de 
quibus postea mentionem faciemus, per y exprimebant, ut yerus 
Anglosaxonice tjxf. Angli (nuur) meretrix, (kuuk) coquus, (guud) 
bonum, (bluud) sanguis, (nuud) cucullus, (fluud) fluvius, (buuk) 
liber, (tuuk) cepit; Seoti {lajjr, kyyk, gyyd, blyyd, nyyd, flyyd, 
byyk, tyyk)." And again, " rotundo ore et robustiib quam 
priores effertur, u angustiore, castera similis r^ o. Bed v (yy) com- 
pressiB propemodxmi labris, multi> exilius tenuiusque resonat qphn. 

1 « TuLT, Poorly. * TWy-stomached.' been in conseopenoe often misled to 

< A well naaba, now de jeow fare P* write (tsh) for (k), thns he here prints 

*■ Wa' naaba, bnt tuly* ... T&o/y, vexed, aeeSz^ which shomd mean (atshtshyrz*) 

ill-tempered, Salop. ... tunly^ restless, an almost impossible combination, Irat 

wearisome, Somerset ; Uwly^ small really means (akyjE*^, though I haye 

and weakly, Dorset. Tmly^ qualmish, kept the incorrectly aoubled (k) in the 

in delicate health, Essex, [Sir T. text. 

Smith's connty] and Camb. Twally a ' The initial consonant must have 

whim, Suff." John Ore€tve$Nall, Chap- been (dzh) or (zh). Probably it was 

ters on the East Anglian Coast,^ 2 toIs., mere carelessness on Smith's part to 

8to, 1866, Tol. 2. Etymological and use (J), as when he wrote c for k. The 

ComparatiTe Glossaij of the Dialect first rowel, too, is accidentally short, so 

and rrovincialisms or East Anglia. that (Dzhee'syys) or (Dyhee'zyys), re- 

a Smith uses e for (tsh), but he has presents the real souna he intended. 



Chaf. III. {3. V — XVI TH CENTURY. 167 

ant Uy (boot) scaplia, (bunt) ocrea, (^jyt) Scotic& prontinciatione, 
oereaJ*^ And again in hia Greek Pronnnciation : " v QraBcum Scoti 
ft Borei Angli turn exprimnnt ci!bn tanrnm sonant, & pro hul^ 
dicunt exiliter contractioribus labiis sono suppresso & qnasi prsBfo- 
cato inter i&uhil (byl)." 

It is scarcely possible to indicate the sound of (jy) more 
clearly and precisely in common language. 

Bespecting u short. Smith says : 

** y Latinam, apertiBsimam habemus Angli, quamvis illam non 
agnoscimns, jam longo tempore k GkJlis magistris decepti : at pro- 
nnnciatio sonnsque noster non potest non agnoscere. Brevis (but) sed, 
(Ink) fortuna, (buk) dama mas, (mud) limns, (fhl) plenns, (pnl^ 
deplumare, (tu) ad; longa (bunt) oerea, (luuk) aspicere, (buuk) 
liber, (mnud) ira aut affictus, (fdnl) stnltus, (punl) piscina, (tun) 
duo, etiam." 

(Buk) being in Salesbury's list serves to identify the two 
methods of symbolisation. Of course no such fine distinc- 
tions as (u, u) are to be expected, nor indeed are they gene- 
rally necessary to be insisted on. An attentive examination 
of tbe sounds of fool full in our present pronunciation will 
however shew that they contain dmerent vowels (fuul, fwll), 
each of which can be pronounced long or short (fuul ftd, ftttil 
ful) and that these cUffer as (i, f) by the pharyngal action 
already explained. As however short (u) rarely if ever 
occurs in closed syllables, and (uu) long never occurs in ac- 
cented syllables, except before r (j), it would be generally 
intelligible to make no distinction between (u) and (u) except 
in rare instances. One marked difference between the 
sounds (t, u) and (i, u) is that (t) may be easily sung to a deep 
note, but (i) cannot ; and on the contrary (u) may be sung to 
a very high note, but (u) cannot. 

1569. Hart calls u long a diphthong, but in his explana- 
tion he makes it arise from the attempt to pronounce (i) and 
(u) simultaneously, and he clearly points out that both the 
lingpial position of (i) and the labial position of (u) are held 
on steadily during tiie sound of long i#, so that if the (i) 
position be relaxed, the sound of (u) results, and if the (u) 
position be relaxed the sound of (i) results. This, as we 
have seen, amounts to a very accurate description of the 
simple sound (yy), which is therefore the sound which he 
means by the inaccurate tide and notation of " the diphthong 
III." His words are : 

"Now to come to the m. I sayde the French, Spanish, & Brutes,^ 

1 maye adde the Scottish, doe abuse it with ys in sounde and for 

^ Thatis, Welsh. 



168 V — XVITH C^KTUKY. Ckat. HI. | S. 

ooDBonaiit, except the Bnites as is sayd : the French doe neaer 
•onnd it ri^^ bat Tiorpe od, for it, the Spanymrd doth often yse it 
right as we doe, bnt often also abnae it with ys ;' the French and 
the ScottiBh in the sonnde of a Diphthong : which keeping the 
Towels in their due sounds, commeth of i & n, (or verie neare it) is 
made and pat togither vnder one breath, confounding the sonndes of 
i, & a, togither: which you may perceyoe in shaping thereof^ if yoa 
take away the inner part of the tongue, from the upper teeth or 
Gummes, then shall you sound the u ri^t, or in sounding the 
French and Scottish u, holding still your tongue to the vpper teeth 
or gums, & opening your lippes somewhat, you shall perccyue the 
ri^t sounde of »." Thus Hart writes : (ui did not mutsh abiuz 
dhem), meaning (wi did not mutsh abyyz dhem) as I shall here- 
after transliterate hb iu. 

1573. Babet says, after speaking of the sound of v con- 
sonant: 

'* And as for the sound of Y consonant' whether it be to be 
sounded more sharply as in spelling blue or more grosly like oo, as 
we sound Boohe^ it were long here to discusse. Some therefore think 
that this sharpe Scottish Y is rather a diphthong than a yoweU, 
being compounded of our English e and m, as indeed we may partly 
perceyue in pronouncing it, our tongue at the beginning lying flat 
in our mouth, and at the ende rising up with the lips also tliere- 
withall somewhat more drawen togither. 

This would certainly make a diphthong because there 
would be a change of position, but what is the initial sound P 
The tongue does not certainly ** lie flat in our mouth for ^." 
The nearest sounds answering to this description are (cb a, a 
0) and it is impossible to suppose any of these to be the 
initial of such a diphthong. The only interpretation I can 
put on this somewhat coniiised description is, that Baret was 
speaking of the position of the tongue before commencing to 
utter any sound, and that when the sound was uttering, the 
tongue rose and the lips rounded simultaneously, and this 
Agrees with the other descriptions, making the sound (yy). 

1580. BuLLOKAB says : "IT also hath thr^ soundes: The one of 
them a m^ere consonant, the other two soundes, are both yowels: the 
one of these yowels hath a sharpe sound, agreeing to his olde and 
continued name : the other is of flat sound, agreeing to the olde 
and continued sound of the diphthong :ou: but alwaies of short 
sounde." And farther, translating his phonetic into ordinary gel- 
ling : '^ and for our three sounds used m, y, the French do at this 
day use only two unto it : that is, the sound agreeing to his old 
and continued name, and the sound of the consonant, r." 

^ That IB, sometiinet say (u), and * Evidently a misprint for rowel, 
aometimet (yy), but thii ia not the I quote from the edition of 1680. 
oaae certainly m modem Cattilliaa. 



Chap. III. § 8. V — XVITH CENTURY. 169 

^ From these two passages it is clear that the "old and con* 
tinned name" of long u in English was the sound of the 
French a, that is (jy). The flat sound we shewed in treating 
of ou (p. 152), was probably {u). Bullokar adds, where I 
translate his phonetic examples into palaeotype : 

"17. aharpe, agreeing to the sound of his olde and continued 
name, is so sounded when it is a Billable by itself, or when it is the 
last letter in a Billable, or when it commeth before one consonant, 
&: e: ending next after the consonant, in one syllable thus : vnity, 
TuiuersaUy procureth yse to be occupied, and leisure allureth the 
vnruly to the lute : which I write, thus : (yyntt* yyntversaullt 
prokyyreth yys tuu bii okkyypiVed and leizyyr allyyreth the 
un-ryyli tuu dhe lyyt). 

"U flat is used alwaies after : a : e : or o : in diphthongs, or 
next before a single consonant in one Billable, hauing no : e : aiter 
that consonant, or before a double consonant, or two consonants 
next after it : though : e : followe that double consonant, or two 
consonants all in one or diuerse Billables, thus: the Tuiust are 
vnlucky, not worth a button or rush, vntrusty, ypholding trumpery 
at their foR lust : which I write, thus : (dhe un-dzhust aar un-luki, 
not wurth a but'n or rush up-Hoouldiq irumpen' at dheir ful lust). 

The word full is the same as one of Smith's examples of u 
short, and hence fixes the sound of Bullokar's u flat, which 
he does not otherwise explain. 

1611. CoTGRAVE says : " V is sounded as if you whistle it 
out, as in the word a lute" Now the French u (yy) has a 
very whistling eflect, both tongue and lip being diq)osed in 
a favourable position for the purpose. 

1621. Gill is again not so distinct as could be wished, he 
merely says, preserving his notation, and his italics : 

" V, est tenuis, aut erassa : tenuis v, est in Verho tu vz vse utar ; 
erassa hreuis est u. vt in pronomine usnos^; aut hnga ii : vt in verho 
tu iiz oosE seaturioy aut sensum exeo mori aqua vi expressa.** 

Gill never alludes to any diphthong (in). He uniformly 
uses a single sign, the Koman v, for the sound of long t/, 
employing the Italic v for (v). He also uses a single char 
racter for the diphthong long i, but then he admits that it 
is only slightly different from the diphthong (ei). There 
are very few mdications of the sound he really meant to 
express by his v. First we must assume that it was a simple 
sound ana *' thinner'' than (uu). This should mean that the 
entrance to the lingual aperture was diminished by bringing 
the tongue more into the (i) position. But this converts (u) 
ii^to (y), and hence leads us to Gill's v s (yy), as the sound 

1 Misprinted Not. 



170 V — XVITH CENTUKY. Chap. IH. J 8. 

is always lon^. Next in bis alpliabet he calls it tr^i^Xop, 
which should imply that it had the theoretical sound assigned 
to the Greek v. This we haye seen from Cheke and Smith 
was (yy). But then the example in the alphabet is ''st/r sure 
certus/' and Salesbur^ says that Welsh u is unlike the sound 
of English sure. This may mean that sure must haye been 
written suwr in the nearest "Welsh characters, because sur 
would haye sounded too like (stVr). Hart and Bullokar both 

S've (syyer). Lastly, in mentioning the words taken from 
e French he says: "Redvite nupera vox est d reduce,^ 
munimentum pro tempore aut occ<mone factum J* This should 
be the French rMuity with a wrong e added, and hence ought 
to establish the yalue (yy) for Gill's y. This therefore is 
the result to which all parts of the inyestic^tion tend, so 
that we must assume it to be correct. On the other hand 
there can be no doubt that the ii, u of Gill were (uu, u). 
1633. Butler is unsatisfactory, when he says that : 
"a, f, u differing from themselyes in quantity differ also in 
sound : having one sound when they are long, and another when 
they are short, as in mane and man^ shine and «Atn, tune and tun 
appeareth . . . Likewise oo and u long differ much in sound : as in 
fool and fule^ rood and rude, moot and mute^ but when they are 
short, they are all one ; for good and gtid, blood and hlud, woolfBud 
wulf have the same sound." 

From this we learn with certainty that short u was (u) or 
(u), and that long u was not (uu), but we cannot tell whether 
it was (yy) or (iu). As long i was (oi) at that time, and no 
allusion is made by Butler to its being a diphthong, we are 
unable to assume that lonff u was a simple sound. We 
might indeed be led by the following passage to suspect that 
Butler had begun to embrace the (iu) sound which must 
certainly haye widely prevailed, when his work was pub- 
lished, although it is not distinctly acknowledged : 

'^ / and u short have a manifest difference from the same long ; 
as in ride rid^ rude rudy dine din, dune dun, tine tin, tune tun ; for 
as t short hath the sound of ee short ; so has u short the sound of oo 
short. ... E and i short with w have the very sound of u long : 
as in hiwy kneew, true appeareth. But because u is the more simple 
and ready way ; and therefore is this sound rather to be expressed 
by it :" but he prefers eew for etymological reasons in ** hreew, 
kneete, hleew, greew, treew, sneew,'* where hreew, treew, sneew are in 
Smith's list of words having the sound (yy). Butler finally asks 
"But why are some of these written wilh the diphthong ewf 
whose sound is manifestly different, as in dew, ewe, few, hew, chew, 
reWf sew, strew, shew, shrew, pewter J^ 

1 Misprinted rtdneo. 



Chap. HI. § 3. v — XVn TH CBNTintY. 171 

Now deWjfetv, shrew are in Palsgrave's list of (eu) sounds; 
and the same, together with strew, are in Smith's (eu) list. 
Hence it is clear that Butler distinguished (eu) from the 
other sound of u long, and it is possible that his u long may 
have been (iu), but as Hart called (vy) a diphthong and 
represented it by (iu), while his careful description deter- 
mined it to be (yy), so Butler may have said (yy). 
At any rate it is clear that quite to the close of the xvi th 
century, (yy) was the universal pronunciation of long 
u in the best circles of English life, and that it remained 
into the xvii th century we shall shortly have further 
evidence. Provincially it is still common. In East 
Anglia, in Devonshire, in Cumberland, as well as in 
Scotland, (yy) and its related sounds are quite at home. 
The southerns are apt to look upon these dialectic forms 
as mispronunciations, as mistakes on the part of rustics 
or provincials. They are now seen to be remnants of 
an older pronunciation which was once general, or of a 
peculiar dialectic form of our language oi at least equal 
antiquity. The sound of short u was also always (u) or 
(u). There is no hint or allusion of any kind to such a 
sound as (o). The (u), still common in the provinces, 
was then universal. 

TJ — XVII TH Cbntury. 

1640. Ben Jonson says: "Y is sounded with a narrower, and 
meane compasse, and some depression of the middle of the tongue, 
and is, like our letter t. a letter of double power." 
By this he probably only means that it was both a vowel and 
a consonant (v). In his notes he gives quotations concern- 
Greek V, ov, the latter of which he identifies with (uu), 
though the cry of the owl, which is rendered tu tu in 
Plautus, Menechmi, act iv, sc. 2, v. 90. 

Me, Egon' dedi ? Pe, Tu, tu istic, inquam. vin' afferri noctuam, 
Quse, Tu, Tu, usque dicat tibi ? nam nos, jam nos defessi sumus. 

From these notes Jonson may have possibly distinguished 
long and short u as (yy, u). 

1653. Wallis clearly recognizes (yy) as long u and dis- 
tinguishes it carefully from the diphthong (iu). He says : 

" Ibidem etiam," that is, tn lahiisy " sed Mmori adhuc apertura" 
than (uu), '^ formatur it exile ; Anglis simul et Gallis notissimum. 
Hoc sono Angli suum u longum ubique proferunt (nonnunquam 
etiam eu ei ew quse tamen rectius pronunciantur retento etiam sono 
e masculi') : TJt muse, musa ; iuney modulatio ; lute, barbitum ; 

1 That is, as (eu). 



172 TJ — XVniH GENTURY. Chap. HI. { 8. 

dmr$^ diLro; mvte^ mntos; n$w^ noms; hreWy misoeo (cereyisiam 
coquo) ; hMWy novi ; view^ aspicio ; lieUy yice, etc. Hone Bonma 
extranei fere assequentury si diplitlioiigani iu conentar pronimciare ; 
nempe \ exile litterse u vel u; pneponentes, (nt in Hispanonun 
eiudad civitas,') non tamen idem est omiiinb sonus, quamvis ad 
illom proximo accedat ; est enim iu sonus compositasy at AngLorom 
et Galloram d sonus simplex. Cambro-Britanni hnnc fere sonum 
utcnnqne per itc, yto, uw describnnt, ut in Uiw color \ Uyw gaber- 
nacnlom navis ; JDmo Dens, aliisque innnmeris." 

Wallis therefore distinctly recognized tlie identity of the 
English and French sounds, and says that they are different 
from the diphthons" (iu) because they are simple and not 
compound sounds, out approach nearly to that diphthonfi", 
evidently because (yy) umtes the lingual position of (i) wim 
the labial position of (u). He also notices the proximity of 
the Welsh tir, yw^ uw to the sound of (yy), and thus explains 
how Salesbury came to hit upon uw as the best combination 
of Welsh letters to convey an approximate idea of the sound 
to his countrymen. Furmer on he says : 

"IT" longum effertur ut GaUorum it exile. TJt in litte barbitum, 
miite mutus, miue musa, eUre cura, etc. Sono nempe quasi com- 
posite ex I et «?," 
where he saves himself from the diphthong by a " quasL" 

As regards short u he says : 

" U vocalis quando corripitur effertur sono obscuro. ITt in hut 
sed, cut seco, hur lappa, hunt raptus, curst maledictus, etc. Sonum 
hunc GaUi profenmt in ultima syllaba vocis serviteur. Differt k 
GaUorum e feminino, non aJiter quam quod ore minus aperto 
efferatur. Discrimen hoc animadvertent Angli dum pronunciant 
voces Latinas itcTy itur; ter ter, turtur; cerdo surdo; temus Tur- 
nu8; terris turris ; refertumy furtumy &c." 

In his theoretical part he gives the following further 
particulars of the French e fcemininum and the & obscurum, 

" Eodem loco," that is, in summo gutture^ " sed apertura faucium 
mediocri," i.e. less than for (aa), ''formatur GaUorum e fcemininum ; 
sono nempe obscuro. Non aliter ipsius formatio differt ^ formatione 
preecedentis d aperti (aa), quam quod magis contrahantur fauces, 
miniis autem quam in formatione Vocalis sequentis (a). Hunc 
sonum AngH vix uspiam agnoscunt ; nisi cum vocalis c brevis im- 
mediate prsecedat Hteram r (atque hoc quidem non tarn quia debeat 
sic effern, sed quia vix commode possit aUter; Hcet enim, si citra 
molestiam fieri possit, etiam iUic sono vivido, hoc est, masculo, 
efferre ;) ut vertue virtus, liberty Hbertas &c. 

'' Ibidem etiam, sed Minori adhuc faucium apertura sonatur 6 
vel ft obscurum. Differt k Ckdlorum e foeminino non aliter qahsi 

^ The English usually call this word s Muz ^aaz ^*); the iu represents the pure 
(tliiadBad*), it is probably (cin^aa^* « (in) diphthong. 



Chap. m. { a. U — XVHTH CENTURY. 173 

Qubd are minus aperto, labia proprius accedant. Eandem soniim 
iet^ effenmt Oalli in postrema syllaba yocum 9&rviU%tr^ tacHficateur, 
etc. Angli pleramque exprimunt per ik breue, in tumy vcrto ; hwm, 
nio; duUf signis, obtusos; cut, seco, etc. Nonnunquam o et ou 
negHgentii!^ pronuntiantes eodem sono effenmt, ut in edms^ venio; 
Mdme, aliqnis; ddne, actum; cdmpanyy consortium; country ^ rus; 
€OupUy par; cdvet^ concupisco; Idve, amo, aliisque aliquot; quad 
alio tamen sono rectius effeiri deberent. Cambro-Britanni ubique 
per y scribunt ; nisi qu6d banc literam in ultimis syUabis plemmque 
ut « efferant." 

Wallis therefore heard the French feminine e in the laat 
syllable of serviteur, sacrificateur. In this he agrees with 
Feline, who draws a distinction between the first and second 
syllable of hetireux, making the first the same as the sound 
now considered.^ But Wallis makes the aperture of the 
lingual passage grow smaller at the back for d, e femimne, ft, 
the first being (aa) with the greatest depression, and he haa 
an action of the lips for &. This ought to give (aa, ar, u) 
for the three soundk But this cannot be right for H, bece^ise 
Wallis distinguished it from (u). Hence we must disregard 
the lip action of the last, and write (aa, x, oe). This how- 
ever, is scarcely probable. There is another difficulty. The 
sound of e in tertius is not at present formed with a wider 
opening of the mouth than the sound of u in Tumus. When 
any distinction at all is made it is rather the reverse.' The 

^ See supra, p. 162, note 3. Tarrer deux sons des premieres ToyeQes a et 4, 

nres the same rowel soimd to le, feu^ ^etiyOetS. Ce rapport est en effet n 

Swrope,n<nfd,p«Mt,anl,aiiteiff,bonli#Mr. bien marqn^, ^e, dans one foule de 

Feline makes the rowel sound in mots, comme jeune, p^ehewy on fait 

li, .fiffrope, p^ aril, antfMT, bonhfur entendre le son de Ve sourd et non 

the same ; but distmc:mshes it firom celni de Vm tel qn*il est donn6 par les 

that in U% ncmdu In M. Feline's Md* mots>£fi#, picheute.*' Now to my eaiB 

moire 9ur ia I^fonM tU r Alphabet i^T%' a d, ^^, o ^ are (a a, e b, o o). In the 

fixed to his Dtetionnaire de la pronon' first two pairs the circmnflexed rowd 

eiatum de la langue Fran^aise^ g^^ing expresses a deeper sound, formed by 

an account of t£e deliberations of a depressing the tongue ; in the last pair 

committee on French pronunciation, the uncircumflexea rowel is the wide 

formed at his request, he says : ** La soimd of the circumflexed. The re- 

•ondusion fdt que Ve muet proprement lations then being different do not lead 

dit existe dans r ortiiographe, mais non to the discorerr of the relations be- 

pas dans la langue ; que, dans tons les tween ^ eu. These mar be, that for 

mots, od il est n^cessaire de le pro- eu the tongue is more depressed than 

nonoer, il exprime un son r^l comme for ^ which would suit for «, «m =: (#, 

toiK les autres signes, et que oe son (s) ;^ or it may be that eu is the wide of 

derrait ^tre appel^ sourd et non pas «, this would suit 0, eu ca (0, oe), whic^ 

muei, cette demi^ denomination n* agrees with my own pronunciation. 

6tant qu'un non-sens. Apr^ Ve on ^ Mr. M. Bell who says (a>, ■) in 

passa au son eu. On recounut qu'il Umus^ Tumun respectirely, makes the 



existe bien dans la langue fran^aise, et opening for (a) wider than for (a>). I 

Ton remarqua qu'il pr^sente areo r# would rather write (t^Tias, Tj*n9s) re- 

que je riens d'appeler sourd le m§me spectirely, if any difference at all has 

n^port qu*on arait troar6 entre les to be recognized. 



174 V — XVII TH CENTURY. Chap. III. i S. 

peculiarity of the smaller lingual aperture and the action of 
the lips may howeyer brine us to (uh) as the last sound, and 
induce us to consider the three sounds as (aa, o, tih). So &r 
as the English passage of short u from (u) or (u) to (o), the 

S resent sound, is concerned (wh) forms a very appropriate 
nk, because Englishmen find it dif&cult to distinguish the 
Italian somma (st^hm'ma) from (sum*a) on the one hand and 
(sem'a) on the other. And we have seen (p. 94) that in 1611, 
the Italian Florio actually identified English (u) with Italian 
(uh), just as 1685, Cooper identified (u, 0)9 p. 101. But this 
sound hardly agrees with Wallis's identification of H with 
the Welsh y. On this sound, see the footnote on Y, in 
Chapter VIII, § 1, when it appears that the Welsh sound 
represents the vowel (a) but that in common discourse it 
passes into (9) on the one hand, and («) on the other, and 
may be always sounded (t). Wallis no doubt referred to the 
sound (9). 

Lastly, if we reflect that {oe) is the de-labialized (u), and 
that this would be a natural transition from (u) to (e), we 
might revert to the original deduction from Wallis's descrip- 
tion, and make his ti = (ce). 

On the whole I am inclined to think that the three sounds 
he meant were (aa, st, 9). Many English consider the 
French e muet, or sourd, to be deeper than (o), but of the 
same nature. The question however is impossible to decide, 
and I think it safest to transliterate d, e feminine, H by (aa, 
a>, 9), which indicates the modem pronunciation of the 
English vowels. 

The great peculiarity, the marked singularity, of Wallis's 
account, is the recognition and introduction of a sound re- 
sembling (9) into the English language in place of (u). Of 
this sound no trace appears in any former writer that I have 
consulted.^ But from this time forth it becomes the common 
sound. Wallis in this respect marks an era in English pro- 

1 In the passage cited from Gill to indicate the sounds (jd, jar skalen, 

inpr^ p. 90, in which he inreighs to), for which he had no symbols. ThiB 

against the thin utterance of affected is the closest allusion to tne sound that 

women,we find (bftsherz) for (butsherz). I have disooyered. For though the 

This is quite comparable to the Eastern account given by Florio^ 1611, p. 9i, 

English (kr?-er) for (kuver^, which which identifies short (u) wiA (nh), 

Gill had just mentioned, and appears might seem to indicate (9) as well as 

to have no connection with the sound (w), yet as the Italians confuse (9) 

(batsh-er) which is only heard from a rather with (a), which is nearly lU 

small number of people at the present wide form, thsm with (wh), and as («, 

day. But when he says that these uh) would probably be indistingmshable 

affected dames said (ja, jar skalerz, ta) to an Italian ear, the inference is rather 

for (jou, juur skolars, tu), it is just that the sound really uttered before 

poadble that he might haye intended Florio was («} and not (o). 



Chap. III. § 3. u — XVll TH CENTURY. 176 

nunciation, the tranrntion between the old and the new. 
This is more striking, because as he is the first to give u 
short as (&), so is he practically the last to give u long as 
(yy) except dialectically. 

At the present day (yy) has vanished from polite society, 
and is only heard as a provincialism, from Norfolk, Devon, 
or Cumberland, or as a Scotticism. No pronouncing dic- 
tionary admits the sound under any pretence. Indeed most 
English people find it very dif&cult to pronounce, either long 
or uiort, and consequently play sad tricks with French. But 
the case is different with (u, o). The two sounds coexist in 
many words. Several careM speakers say (tu pet, betsh'er), 
thoueh the majority say (tu pt/t, bMtsh'er). All talk of a 
put (p9t). Walker gives the following as the complete list 
of words in which u short is still (ii). 

bully pully full, and words compounded with -Jul; huUcck, huUy, 
huUety bulwark, fuUer^ fMingmiUy pulley , pullet, pu$h, bueh, bushely 
pulpit, puee, bullion, butcher, cuehion, cuckoo, pudding, sugar, [he 
makes sure = (shiui)], husear, huaa, and to put, with Fulham,^ but 
Bays that ** some speakers, iadeed, have attempted to give bulk and 
punish this obtuse sound of u, but luckily have not been followed. 
The words which have already adopted it are sufficiently numerous ; 
and we cannot be too careful to check the growth of so unmeaning 
an irregularity." 

Here the orthoepist unfortunately reverses the order of 
things, and esteems 'Hhe old and continued" sound of (u) an 
irregularity, and what is more, an "unmeaning irregularity," 
and is not aware that every change of (w) to (e) has been 
a modem encroachment. But if tne territories of (u) and 
(e) can be so strictly defined in the south of England, in the 
middle' and north the war is still raging, and though educa- 
tion has imported large quantities of (e) from the south, even 
magnates in the north often delight to use their old (u).' 

^ Smwrt 93dB,butiaee,fiilla^jfiiiiery, Hulk) they are not oommon, but may 

euthai, hurrah ! to the aboTe list It is be heard ; (pim*«Bh) was heard latelj 

emioiis that Walker (art. 177) speaks from an educated gentleman in Cornwall, 

of fuUome as a ** pnre English word,*' ' In the Midl^d counties the South- 

and Smart (art. 117) calls it a word ** of em usage is almost reyersed, (pat, fyt) 

classical deriTstion. Orthoepists are standing beside (kwt, kum). 

not always good in etymology, but * A Yorkshire country gentleman 

Walker appears to haye the best of it who wrote his name Button^ and whom 

here, and i^ as seems more than pro- all his friends called (u9t*n), always 

bable, fuUams is a derivatiTe of fuli, spoke of himself as {Hwt'n), and on 

(the I^mptorium has fulsunnest of one occasion spelled his name so to me 

mtt$j sacietas,) there would be a reason with phonetic letters. He would hare 

for retaining the sound (ful) in the first been about 90 years old now, were he 

syllable. At any rate the usa^ of stOl alire. All tne Yorkshire and Mid- 
land peasantry use («) as a matter of 
ooune. 



speakers with regard to (fi<l*s«mT and 
(iel*tim) Tariei greatly. As to (btdk, 



176 XT — XYHTH CKNTUKY. Chap. IIL { J. 

That there is nothing intrinsically pleasing in the sound 
of (q), may be seen at once by calling goodf stood (god stad), 
to rhyme with bloody flood, (blod^ flod). Those speakers, 
to whom {wu) presents a difficulty are apt to change it into 
(we) as (wddy wom'Bn) for (wtid, wum'tm), and the effect is 
anything but pleasing. In general the long Saxon (oo), 
which first became (uu) and tii^i fell into (u) or (u), has 
resisted the further change into (a). This difference of 
OTolution is similar to that which has befallen i, ei, ai, 
which Shakspere pronounced sufficiently alike to introduce 
a conceit upon them in one of his most tragic speeches, 
already cited (p. 112), but which have become three quite 
distinct sounds (ei, ii, eei), (p. 120). Both changes have 
occurred rather among the reading than the merely speaking 
section of our population. 

1668. WiLKiNS and "Wallis were contemporaries ; alihoufi^h 
the latter was the elder, and bom in Kent, and me 
former was bom in Oxford, they liyed as fellow collegians 
for some time in Oxford, and they mixed in the same society. 
Yet we have a striking difference in their pronunciation of 
long u. We have seen how Wallis identified the French 
and English u, how he considered the (yy) sound to be 
familiar to all Englishmen, and especially distinguished it 
from the diphthong (iu), and this he continued to do through 
all the editions of his grammar. Wilkins at the same 
moment can scarcely pronounce (yy) at all, denies that 
Englishmen use it, and makes every long u into (iu). 

'^ As for the u OaUicum or whUUing ti" says he, p. 363, '* though 
it camiot be denied to be a distinct simple vowel ; yet it is of so 
laborious and difficult pronunciation to all those Nations amongst 
whom it is not used, (as to the English) especially in the distinction 
of long and short, and framing of Dipthongs, that though I have 
enumerated it with the rest, and shall make provision for the ex- 
pression of it, yet shall I make less use of it, than of the others; 
and for that reason, not proceed to any further explication of it." 
And again, p. 382, ** u," which is his character for (yy), "is I think 
proper to the French and used by none else." 

This is a strong contradiction to Wallis, whose treatise 
Wilkins had read, and apparently studied.^ The only word 
which contains long u that Wilkins transliterates, is commu^ 
nion, and this he writes (kAmmiuuniAn), using (iuu) and not 
(yy) in the accented syllable. 

^ He says, p. 857, '* Dr. Wallis .... and sabtletj to hare eonsidered the 
amongM all tnat I have seen published, Philosophy of Artionlate soonds," 
seems to me, with greatest Accoratenass 



Chap. III. § 3. U — XVUTH CBNtURY. 177 

Short u is thus exemplified hj Wilkins and distinguiBhed 
from (uuy u), meaning (uu, u) most probably : 

(u) shori ML fdt pul 

Cnu) lon^ boote foole foote moote pools roods 
(a) 8h^i but Ml' fdtt* mutt-<>ii puU* rudd-^ 
(aa) lon^ amongst 

The sound, which he represents Inr y with a peculiar 
flourish added to its tail, and which I have translated into 
my (a), he describes as "a simple letter, apert, sonorous, 
guttural; being framed by a free emission of the breath 
from the throat/'* Again, p. 364, he says "the vowel (o) 
is wholly Ghittural^ being an emission of the breath from the 
throat without any particular motion of the tongue or lips. 
Tis expressed by this character," a variety of y, "which is 
already appropriated by the Welsh for the picture of this 
sound." As he here rejects both tongue and lips in the 
formation of (o) he differs considerably from Wallis in ex- 
plaining its formation. In another place he says that the 
Hebrew " Schevah" is rapidly pronounced " probably as our 
short (a)." He rives (ai, au) as the analysis of "our 
English i in bite, and of the sound in ^^ owv, ot£;le." And 
finally he says: "y" meaning (a) "is scarce acknowledged 
by any nation except the Wdsh." The words in which he 
employs this sign, omitting the combinations (ai, au) are : 
kingdom, came, done, but, Jesus, son, under, Pontius^ buried, 
third, judge, church, resurrection, which he writes (kiq*ddm, 
kam, dan, bat, Dzhesas, san, ander, PAnstas, barY,ed, thard, 
dzhadzh, tshartsh, reserreksioon), in which I give all his 
errors. I assume this soimd to be (a) both in Wallis and 
Wilkins, but what particular shade of this sound they pro- 
nounced, and whether they both used the same shade, it 
would be rash to assert. 

1668. Price does not help us to the sound of short u 
when he says : 

"The u is twofold, 1. short, as in hut, must, hurst, 2. long as in 
hUe, muse, refuse as if it were the compound of t'u?." 

This iw may mean (iu), agreeing with Wilkins, but it 
may also mean (yy) agreeing with Wallis. I am inclined to 
treat it as (iu). The short u I have, on the combined 

1 These words judging from futi^ are tIoiisIt written with one final consonant 
aU fancy words, (fel, mt, pal), intro- to indicate the sound («). If this 



daced to contrast with the (fwl, fut, theory be correct, the word full in the 
~ ^^ in a preceding line, and most first line, was a misprint for /W. 



probably the doubling of the final con- > This description is made up from 

sonant was intended to indicate the the different headings of the table 
sound (9), whereas fut, pul were pre- p. 860. 

12 



178 V — XVUTH CENTURY. Chap. m. { 3. 

authority of Wallis and Wilkins, been in the habit of con- 
sidering to be (a). The following notices agree with this : 

'^0 after w Boundes like short u as world, ncord, woman, won, . . . 
before m or n in the last syllable soundes like short u as fireedom, 
reckon, bacon, ... Ou soundes like short u in coucin, double, 
oourageJ*^ 

But there is one notice which^ thus interpreted, has a 
singular effect : ** Oo soundes like short u in good, wool, hood, 
wood, stood." The general use of (god, wel, nod, wod, sted) 
is difficult to believe in, though it is well known provincially, 
and is also mentioned by Jones, (p. 183). 

1669. Thouffh Holder's work was not published till this 
year, Wilkins had seen it in manuscript, and speaks highly 
of it.^ Yet in the letter u, both long and short. Holder 
differs from Wilkins. Holder has very acutely anticipated 
Mr. M. Bell's separation of the lingual and labial passages, 
and the possibility of adding a labial passage to every lingual 
one. He says : 

In the larynx is depressed, or rather drawn back by contraction 
of the aspera arteria. And the tongue likewise is drawn back and 
curved ; and the throat more open to make a round passage : and 
though the lips be not of necessity, yet the drawing them a little 
rounder, helps to accomplish the pronunciation of it, which is not 
enough to denominate it a labial vowel, because it receives not 
its articulation from the Hps. Oo seems to be made by a like 
posture of the tongue and throat with o but the larynx somewhat 
more depressed. And if at the same time the Hps be contracted, 
and borne stiffly near together, then is made V ; u with the tongue 
in the posture of i but not so stiff, and the Hp borne near the upper 
lip by a strong tension of the muscles, and bearing upon it at either 
comer of the mouth." 

<< b is made by the throat and tongue and lip ; in b the tongue 
being in the posture, which makes oo ; and in t< in the same posture, 
which makes i, and in this b and u are peculiar, that they are 
framed by a double motion of organs, that of the lip, added to that of 
the tongue ; and yet either of ^em is a single letter, and not two, 
because the motions are at the same time, and not successive, as are 

^ He Bays : *' Bat besides each," have had to penise from their privnU 

namely, ** in later times .... Erasmus, papers the distinct Theories of some 

both toe Scaligers, Lipsios, Salmasins, other Learned and Ingenious persons,'* 

VosBins, Jacobus Mathias, Adolphus Dr. William Holder and lir. Lodowiok 

Metkercbus, Bemardus Malinchot, etc., are named in the margin, **who hare 

besides seyeral of our countrymen, Sir with great judgment applyed their 

Thomas Smith. Bullokar, Alexuider thoughts to this enquiry ; in each of 

Gill, and Doctor Wallis," ''(whose con- whose Papers, there are seyeral sug- 

nderations upon this subject are made gestions that are new, out of the 

publick) I must not forget to acknow- mon lode, and Tery considerable." 
todge the &?onr and good hap I have 



Chap. UI. § 3. V — XTUTB CENTURY. 179 

ni, pla &c. Yet for this reascm they seem not to be absolutely so 
simple vowels as the rest, because the voice passeth successively 
from the throat to the lips in H and from the palate to the lips in u, 
being there first mould^ into the figures of oo and i, before it be 
fully articulated by the lips. And yet either these two, b and u, • 
are to be admitted for single vowels, or else we must exclude the 
lips from being the organs of any single vowel since that the mouth 
being necessary to conduct the voice to the Hps, will, according to 
the shape of its cavity, necessarily give the voice some particular 
affection of sound in its passage, before it come to the lips ; which 
will seem to make some such composition in any vowel which is 
labial. I have been inclined to think, that there is no labial vowel, 
but that the same affection from the lips may, somewhat in the nature 
of a consonant, be added to every of the vowels, but most subtlely 
and aptly to two of them, whose figures are in the extremes of 
aperture and situation, one being the closest and f(»wardest, which 
is f , and the other most open and backward ; there being reason to 
allow a vowel of Hke sound in the throat with Hy but distinct from 
it as not being labial, which will be more familiar to our eye if it 
be written oo; as m cut coot, full fool, tut toot, in which the lip 
does not concur; and this is that other. Thus u will be only i 
labial, and H will be oo labial, that is, by adding that motion of the 
under-lip, t will become u, and oo will become )^." He proceeds to 
use his », ti, 8 in the formation of diphthongs and concludes thus : 
'* Concerning H and ti, this may be observed, that in subjoining 
them to another vowel, H is apter to follow a and o, because of 
their resemblance in the posture of the tongue, as hath been said ; 
and for the like reason u is apter to follow a and e, as Hani wawl; 
euge etc. But generally if the vowels follow, then it is « precedes 
and not ti." 

No doubt the descriptions give very accurately oo = {(bcb\ 
M = (uu), u = (i) or (y). And the snoFt (a?) would then be 
Holder's sound in fuii. Now it is impossible to believe that 
fool was ever pronounced {{cecel), the sound being extremely 
difficult to any one but a Highlander (in whose word laogh it 
occurs), until the trick of removing the labial action from 
(uu) has been acquired. But if we remember that now full 
is rather (ful) than (ful) ; and that the widening of the back 
of the throat, by which (u) differs from (u) is so much the 
most essential part of the sound, that a very good imitation 
of it can be produced with the mouth wide open, it is very 
probable that Holder caXLed fool full at least when theorizing 
(fuul ful). The pairs of examples he gives are cut coot, full 
fool, tut toot, of which cut, tut would have been (k9t, tat) 
according to "Wallis and Wilkins, who would have perhaps 
preserved the old pronunciation (fwl) or (ful). Did Holder 
say or intend to say {kut kuui, ful (uul, tut tuut) ? In this 
case he must have altogether ignpred the vowel (a). Or did 



180 U — XVn TH CENTUBT. Chap. IIL § 3. 

he mean to say (kat kodt, fal fbol, tat taat) f or did he mean 
— what he has written — (koet kflwet, fol foal, ta?t taw?t) ? 
sounds which he may have imagined he said, but which other 
people are scarcely likely to have really pronounced. The 
• distinction which Holder makes between tne vowels in fool, 
ttco is peculiar to himself. Wilkins gives fool as an example 
of the long (uu), and /i4l as an example of both the short 
(u) or (w) and of (e), supri p. 177, note 1. This throws a 
doubt over the pronunciation of this particular word/w//, and 
renders Holder's explanations still more mysterious. Can 
it be that Holder's pronunciation was very peculiar so that 
he actually confused {u, a) at a time when the transition 
from old (u) to (a) was comins^ into vogue ? His {as) woidd 
not be a bad middle between the extremes o{(u, a). His long 
u in rule, which is usually now (uu), was manifestly (yy), if 
his explanation of superadding the labial to the lingual effect 
is to be trusted. His only notice of a diphthongal u is in the 
word etfffe, just cited, which must have been (eydzhe), if his 
explanation is to be relied on, but this is very doubtful. 

1685. Cooper pairs the vowels va/ull, fole, or as he some- 
times writes foale^ that is, in full he takes the vowel to be 
short (o). He may however have used (w) or (?ih). See 
the discussion on p. 84, and the passage quoted on p. 101. 
The observations in that passage serve to shew that a in 
full had at that time much of the (o) element in it; that 
some persons may have pronounced it quite as (o) ; and 
others as (u) the usual sound into which (o) degenerates, 
or {v), which is the more common English sound ; the true 
short (u) is so unusual to our organs, that when we hear it 
we take it for the long (uu), and we can hardly pronounce 
it except when long. The English (uu, u) as has been 
already mentioned, are related precisely like the English 
(ii, t). I shall, as already stated, p. 84, consider that Cooper 
pairs (oo, «). But Cooper also distinguished (uu, u) in food 
foot, see suprd p. 101. He illustrates this sound by German 
zufluch (misprint for zuflucht as shewn by the meaning re- 
fugium) and French coupe poculimi, now (tsuu'flukht, kup). 

Cooper is very copious upon short u which he clearly 
means to be (a) or one of those vowels, as (a, (e), which he 
would "scarcely distinguish from (a). The long u he makes 
(iu) and seems to have great difficulty in understanding the 
French u (yy). His words are : 

** U fonnatur tantAm in gutture, a larynge spiritum vibranto, 

^ Ab fool used to be written fohy the more common spellinf foalt oould 
nothing bat Cooper's haying once used hare shewn us what word he meant. 



Chap. III. { 8. 



13 — XVII TH C3ENTTTRT. 



181 



nndmn efficiente mnnnnr, quod idem est cum gemitu hominis ffigrita- 
dine yel doloie excraciati; quodqti^ infantes (priusquam loqni 
yaleant) primikm edunt : Et fandamentum est, k quo omnM eaUra 
voeales, YBiik modificatione constituuntur^ .... Hunc sonum cor- 
reptum yix unquam aliter pronunciant Angli qu&m in nut nux; 
prout etiam in Imgua latina, ni ubi consonans praecedens sit labialis, 
ut priik dixi, et labiis dat formam qui sonus plenior effertur, ut in 
puU yelloy inter bos minima' datnx, datur tamen specifica, diffe- 
rentia ; ille etenim sonus dilutior est. Me plenior, iUe formatur a 
larynge tantAm in gutture, bio k labiis contractis; dum itaque o 
labiis formatur in sono continuato, si recedant labia in oblongam 
formam formatur u gutturalis;' in quibusdam scribitur per o ut, 
to eome^ yenire ; Gallt boo modo, yel saltem persimili,^ olim sonarunt 

^ The natural voitelf should be the 
sound of the Toice, that is of tiie Tocal 
ligaments or glottal reed, without any 
resonance ta^ p. 161. This it is or 
course impossible to hear. But it must 
resemble the reed sound of the clarionet 
or hautboy, or the whistle of the flute 
or flageolet, and contain in itself all 
the tones which the yariously formed 
resonance tubes prefixed to it in speak- 
ing, by means of the pharynx, nose, 
tongue, mouth and lips, aeyelop or 
render audible. It is as the resonance 
tubes clearly separate the tones, or allow 
many nearly coincident to be heard to- 
gether, that we obtain distinct or con- 
fused, coloured or colourless, vowel 
qualities of tone. 

' This remark is important as shew- 
ing the ease with which (ti, &) were 
confiised by speakers at the time of the 
transition of short u from (m) to (d). 

' If the lips be mechanically opened 
by the hands while we are pronouncinjs; 



{oo) we shall pronounce (aa), which is 
the form that Mr. M. Bell adopto for 
the lon^ sound of u in up. Hence 
Cooper 18 quite consistent when he 
inakes « in full the short (o), and u 
in nut the delabialiBed short (o) or Qi). 
This is the most accurate description 
of the sound that I hare met with in 
any old book, and may be adyantage- 
ously compared with Holder's, given 
above p. 178. 

* Probably to is not intended as an 
example, but only eome. Both are 
italicized in the original. 

• As Mr. M. Bell hears (a) in 
English tq9 and (a) in French ouij and 
(a, 0^ only differ as back and mixed 
vowels of the same class. Cooper's ear 
was not far out. To me however now, 
the French e in pie sounds f^), which 
is a ' round' voweL English ears, 



however, readily confound (<k, a, cb; 
9y oe, 0h) with one another and with 
(e), and (j). What was however the 
old pronunciation of the present French 
mute e? Meigret, 15^, writes the 
same vowel in the first and last syllables 
of ^^merite, benite, perir, mere, pere," 
which Feline writes (mmt, l^nit, pmr, 
meer, peer) with two different vowels. 
I understand Meigret to mean {e) in 
both cases. But the lightly spoken 
unaccented (e) drifts very easily into 
(v, 0, «). From (#) therefore (») could 
have easily descended. In fact M is 
only the * round' or labialized {e). This 
recalls an apparently inexplicable re- 
mark by Pusffrave, 1530, who says: 
*^Jf e he the Iciste vowell in a frenche 
worde beynee of many syllables, eyther 
alone or wim an • folowynge him, the 
worde nat havynf his accent upon the 
same #, then shafl he in that place be 
sounded almost like an o and very moche 
in the noose, as these words homtM^ 
fAnmSy honiatSy pdrle. hdmmeSy fimmety 
hon^8tes, avicquet, shall have theyr laste 
$ sounded in maner lyke an o, as hommo^ 
fgmmo, honettOy parh, hommoa^ fammoty 
hoH0sto», aveeguos; so that, if the reder 
lyft up his voyce upon the syllable that 
commeth nexte before the same 0, and 
sodaynly depresse his voyce whan he 
Cometh to the soundynge of hym, and 
also sound hjrm very moche in the noose, 
he shall sounde e beyng written in this 
place accordyn^ as the Frenchmen do. 
Which upon this wamynge if the lemer 
wyll observe by the frenche mens 
spekynge, he shall easily perceyue." 
The luisality may be an erroneous 
observation, and the whole history may 
be a clumsy expression of the sound m 
{0), for which me rounding of the lips 
suggested (0). See suprl^ p. 119, note, 
ool 2. 



182 XT — XVHTH CKKTUBT. Chap. HI. { 3. 

fieminiimiii ^ at in providence, Gennoni flyHaiboB kam^ & herg* in 
piopriis nominibos. Nonqnam in proprio sono apnd nos prodactnm 
andiyiy ni in mosica moduladone,' yel inter pernios, pnecipn^ 
pneros cunctanter prononoiantes ; pro long& enim yocali assomit 
dipthongnm eu (tu) ; onde «tiam denominator ; nt mute mutus ; 
pront in Neuter , '^reSSo^y idem fere ctim Galloram u de quo inter 
dipthongos dicetor." 

**JSm wittf weal (t , ee) cnm u (n) coalescens nobis familiarismmus 
est, quem yocamns u Ion gum ; nt funeral fonns, huge inns ;^ juiee 
snccns, scribimus per ew ; ut ehew mastico ; knew cognovi ; aliisqtM 
temporibns yeil>onim prseteritis ; qnando syllabam ftialem cloudit, 
additur e, true yeros ; raih per eu, rheum rhenma ; sic semper pro- 
nnnciamns eu latinum, & €v Ghrsecum : et Gallf plemmqutf iUomm 
u, qnandoqtftf antem snbtiliiis qnasi sonns esset simplex, sed iuec 
difficilis & Gallie propiia." 

The last words shew that his conftision of (yy) with (lu) 
in French pronunciation was really fault of ear, and that he 
was quite ignorant of (yy) as am English sound. Cooper is 
Tery particmar in shewing how all yowels £edl into io) in un- 
accented syllables before r. These will be considered under R. 

1688. MiEOE of course hears the English long u as the 
French, but as the diphthong (iu) does not occur in French, 
this only shews the same defect of ear which makes him 
identify short u in cut with French o (o), and short u in us 
with French eu (ob). He says : 

"La Prononciation commune de I'lJ Voyelle en Anglois est la 
mSme qu'en Francois. Mais, entre deux Gonsonnes dans une m^me 
Syllabe, elle se prononce ordinairement en o; Comme hut, cut, ruh, 
up, humhle, under, run. Quequefois en ou ; Exemple chuee, puss, 
hull, pull, fuU, En eu, comme us, faculty, difficult, difficulty. Bury 
& husy se prononc^it bery, bisy. Et dans les Mots qui fimssent en 
ure, Vu semble reyetir le Son d'un e feminin, sur tout quand on 
parle yite. Comme nature, picture, fracture, qui se prononcent 
lamilierement naiter, picter, frecter," And again : ** TJ yowel, by 
it self, is pronounced in French according to the Sound it has in the 
Word Abuse in English. 

1701. Joirss says: "the Sound of it in hut, cut, &c. is the Sound 

^ Dr. Froembling, in IiIb Sltmmts of Froemblin? (who speaks English ex- 

the German Langwtge, 2nd edit. 1865, cellently) nit upon this oontriyance. 

p. 2, savs that tiie Gennan a ** is pro- Cooper naying heard ham as (iram) in 

nouncea like a in father^ if long ; and proper names only, mnst hare been 

like u in hut if short" This is the mistaken ; German proper names do 

only other instance I know in which not end in Juim bnt in heim. 

Gennan short a has been identified with > This mnst hare been a mere 

English (9^ ; it is usually conftised with Anglicism. 

English (a), which howerer would ^to * One of the best means of observ- 

a Tery broad Austrian pronunciation, ing the prolonged effect of short Towel 

and it was to aroid this on the one sounds, 

hand, and (as) on the other, that Dr. * Misprint for ingmt or immmiut t 



Chap. HI. i S. U — XVUTH CBHTURY. 183 

of the natural Immane Yoice, and therefore the easiest of all the 
Sounds that are made by the humane Yoice." 

And yet this easy sound is a stumbling block to all Eu- 
ropean nations, and is rarely heard except among Asiatics. 
It may be doubtful indeed whether the Asiatics pronounce 
the same variety of (a) as we do. Many Welshmen do not 
admit it as a proper Welsh sound, though their language is 
supposed to have an appropriate letter p to represent it. As, 
however, y in Welsh also represents another sound, it cannot 
be more properly considered the special representative of (a) 
than the English u, so that there is really no European means 
of representmg the sound, although, owing to its supposed 
relation to the French e mute, (d), so many writers have em- 
ployed an inverted e, that this has been adopted as the best 
unaerstood form in palaeotype. The sound of long u, Jones 
says, is compound, but he does not analyze it. 

Jones gives many lists for the representation of the sound 
of short u by various vowel forms, which need not be cited at 
length as they agree ^nerally with modern use. In the fol- 
lowing words the italic letter might be, or occasionally was 
soimded as (a) according to Jones. 

Christmas, WOluim, &c; centatoy, rest^iMration, &c; fasten, 
listen, &c ; aspen, burden, chicken, cozen, &c ; yeoman ; beztl, civtl, 
devtl, &c; basin, cabin, coffin, &c ; Westmmster "sounded JTesU 
muiter;^^ boil, coil &c «= (bail, kail) &c ; another, mother, pother 
&c ; boul, bout, font, lout, out, &c = (haul, baut, faut) &c ; dove, 
love, move — this is peculiar, shove &c ; cowl, howl, &c = (kaul. 
Haul) &c, voyage, &c ; = (vai'edzh) ; vouch, &c ; word, work, 
worth, &c ; yonder, yonker, &c ; colonel, colour, &c ; comfort, &c ; 
coney, conjure, &c ; money, monkey, &c ; mongcom, monger, &c ; 
ctiUy, &c; blomary, &c; (see under 0, p. 102), come, some, &c; 
bucksom, fulsom, &c ; kingdom, &c ; chibol, gambol, symbol ; 
son, does, reeo^rnisance "sounded recunnisance ;^* foot, forsooth, 
good, hood, look, soot, stood, took, " when it may be sounded oo 
rather than il;" wood, wooif, wool "which some sound as with 
u viz. wiid wiill &c" — adjofim, attowmment, attotmiey, hloud, 
Bourdeaux,' cotmtry, cotirage, cotirlass, cowrteous, cotirtesan, courtesy, 
cousin, doieble, doublet, floud, flourish, housewife, journey, mourn, 
nourish, scourge, sojourn, Southwark, touch, trouble, tmcouth, young, 
your, youth "and all the Names of Seaport Towns as Fabnouth 
Portsmouth Yarmouth" &c ; athu^ort, thu^ort "sounded athurt, 
thurty" answer, ttoopence "sounded tuppence," myrrh, pyramide 
&c ; comerade " soimded eufnrade,^^ hiccough " soimded hieeupy^^ 
frumenty " sounded /urme^y," construe "sotmded c<mitwr,^^ Catlui- 
rine " sounded Cattum,''^ 

> There is a place near Edinburgh Bourdeaux Souse, Jones also writes 
ealled (Bar*dt hous) from the dd (Bunr'doo), snprit p. 140. 



184 V — XVinXH CERTUBY. Y, W, WH Chap. III. | 4. 

In almost every instance (e) is seen to be a substitute for 
an older (u), or {u) as (an) was of an older (an). 



XJ — xviuTH Century. 

1704. The Expert Orthographist gives us no informa- 
tion on the nature of the soimds of u long and u short. 

1710. The Anonymous instructor of the Palatines says that 
u at the beginning is like the Qermanju, meaning that long 
ti=(iu). He also gives the pronunciation of the English 
words church, much, in German letters as tschurtsch, muUch 
=:(tshurtshy mutsh), so that he does not acknowledge (a) at 
alL This may have been designedly, because (a) wo^d have 
been so difficult to the Palatines, and because (u) would be 
intelligible to the English. 

1766. The following are a few words from Buchanan : 
(fill, push, shug'tr) sugar; (put ; batsh*«r, pas) butcher, puss ; 
(tu pat) to put ; (ber'i, hiz'%) bury, busy ; (triu, fiu'nas, liut, 
miuz) true, furious, lute, muse. 

1768. Franklin has (satsh, ranz, matsh) such, runs, much ; 
(fiu'rias, iu'sedzh, truu, ruulz, iuz'ed) furious, usage, true, 
rules, used. 

1780. Sheridan gives as peculiar Irish &ults, (bal, bash, 
pash, pal, pal*prt, pad'tn, kash*an, fat, pat) for (bt^l, btish, 
pt/sh, p2^1, pz^l'ptt, pt«d*tq, ke<sh'an, iui, put), all of which, as 
well as (drav, strav) for (droov, stroov) are, as is now mani- 
fest, remnants of the xvii th century. The other cases of 
Iridi mispronunciations which he cites, and which have been 
already noticed, (pp. 76, 92, 103, 129, 160), shew very 
clearly that the so-called Irish mispronunciations are merely 
fossil relics of the xvnth century, preserved in a com- 
munity separated by the sea from the mother country, see 
supri p. 20. 



§ 4. The Consonants. 
Y, W, WH. 

According to the present usages of Enc^lish speech, Y and 
W are the consonants (j, w) when preceding a vowel, as in 
ye woo (ji wu), and those who can pronounce these words 
differently from (ii uu) can generally pronounce these conso- 
nants. But there has been a p^reat dispute among orthoepists 
whether y, w shoidd be considered as vowels or consonants, 



Chap. III. } 4. y, W, WH — XVI TH TO XVIH TH CBNTURY. 185 

and yarious terms have been invented to suit the case. As they 
do not occur in French, Falsoravb of course does not notice 
them. Salbsbubt, with his Welsh habits always regards 
y,WB8 the vowels (i, u), and consequently writes (und'er, 
uu) for (wun'der, wuu). Smith has the same opinion, but 
writes (i-is, i-it, u-ul, u-ud) for (jis, jit, wul, wud), although 
these sounds cannot be distinguished from (iis, iit, uul, uud) 
unless either a distinction in the vowels be made, which he 
does not allow, as (its, i«t, uul, uud), or else the vowel be 
repeated as (i,is i,it u,ul u,ud). Hart carries the same 
prmciple to the extent of writing (iild uuld) for (jiild, wuuld) 
and even (ureit) for (wreit) meaning (rtreit) makine that 
word therefore dissyllabic. Gill has distinct alphabetical 
characters for (j, w), and says : 

*' Si (juis sonorom eequus sestimator vsum earom apud nos per- 
pendat, mTeniet esse consonas," 

but seems to consider that the principal test (" lapis Lydius") 
of the fact is that the indefinite article assumes the form a 
and not an before p, w. He adds : 

*' JF", aspiratum, consona est, quam scribunt per tch et tamen 
aspiratio prsecedit. IUsb* namqu^ voces quce per wh scribuntur; 
possont otque etiam ad exempla maiormn scribi debent per (hw) 
aut (hu) ; ita enim, nihil aliud inde coUigi queat, qakm quod ex 
ipso wh, intelligimus ; vt (wiil) sive (uiil) weele nassa,' (ffiml) sive 
THuiil) WHEELS rota. Tamen quia nostra experientia docet, (w) et 
(wh) veras esse simplicesqud consonas, in quarum elatione (u) sug- 
gmimit tantum, non clara vocalis auditur; ideo illud (w) ante 
Yocales aut diphtfaongos ius assignatum obtinebit; at (wh) mala 
tantum consuetudine' valebit in (what) quid, (whedher) uter & 
similibus." 

We have here the first distinct recognition of a consonant 
peculiar to the English langua^, which is seldom acknow- 
ledged even by recent orthoepists, most of whom consider 
(wh) as = (hw) or (hu). The preceding writers had all 
used (hu). It is to be observed that Gill had no (jh) ; this 
must have been because, as he used (yy) in place of (juu) 
initial, he said (nyym'ur) and not (jhuum'ur), for which 
most recent orthoepists have (HJuu'mai), a combination as 
objectionable as (nwiil) for (whiil). 

Gatakbr 1646, goes to the extreme of making y, w always 
consonants, considering ei, ew tohe (ej, ew). This, however, 

^ Misprinted tZfe. meant one (wh). ThU "bad onstom'* 

' Narrow necked basket for oatoh* is evaded by the palaeotynio nse of 

inff fish. (h) for the aspirate and (n) for the 

* The fault in Gill was that he wrote diacritic. 

two oonaonanti (wh) when he only 



186 T, W, WH — XVI TH TO XTHITH GKNTOKY. Chap. HI. { 4. 

depends upon a diphthongal theory, to which writers have 
been led by observing that (ai) is not merely (a, i), see 
p. 51. ^ Wallis inclines to Gataker's opinion, and says : 

"Diphthongi at, ei, ot, au, euy auy &c, recte pronunciataB com- 
pontintar ex vocalibus praex>06itivis et consonantibus y et ti^ qu» 
tamen pro vocalibus subjimctivis vulgo babentur." 

His contemporary Wilkixs, alluding to the opinion of 
Gataker and others says on his p. 370, that they 

''do earnestly contend that there are no such ihings as diptbongs. 
Their principal Arguments" be goes on to say, ** depend upon this 
Supposition that (i) and (u), which are necessary Ingredients to the 
framing of all usual Diptbongs, are Consonants the same with (j) 
and (w). Others would have tbem to be of a middle nature, be- 
twixt Vowels and Consonants ; according to which opinion I have 
abeady described tbem : From whence the Keason is clear, why 
these Vowels concur to the making of Diptbongs because being the 
most contract of Vowels, as is also the vowel (9) of wbicb more 
hereafter, They do therefore approach very near to the nature of 
Liters clausay or Consonants ; tiiere being no Transition amongst 
these, either from one another, or to the intermediate sounds, with- 
out such a kind of motion amongst the Instruments of speech, by 
reason of these different Apertionsy as doth somewhat resemble that 
kind of Collision required to the framing of Consonants." 

CoopBR recognizes (j, w) as consonants and also (jh, wh) 
under the form, (hj, hw), at the same time that he defines a 
diphthong as the '' conglutinatio duarum vocalium in e&dem 
syllab&." 

This theory of " conglutination," eflTected by the " glide," 
is that which I have adopted (p. 61), and, consequently, be- 
lieving that the sounds were in all cases the same, I shall, in 
transcribing the pronunciation of others, when they use (la) 
or (aj) consistently write (ja, ai), having precisely the same 
intention, and representing the same sound, on different 
theoretical principles. I consider the sounds of (j, w) to 
have been the same throughout the period now considered. 
Whether there may not be or have been a sound (bh), lead- 
ing to the confusion between (v) and (w), well marked in the 
South East of England, I leave unsettled. In Chapter Y, § 4, 
No. 1 , 1 shall adduce reasons for believing that the Anglosaxon 
w was not (bh). Although (wrait) can be pronounced, yet 
(vrait) or (bbroit) is much easier for the lips, and in Mr. 
Melville Bell's Scotch specimen Chapter XI, § 4, the initial 
(vr) will be found in (vraq) wrong, which may however pos- 
sibly have been (bhraq). As qu is now, and probably always 
was, (kw)t the labial modification of (k), produced by rounding 
the lips at the same time that the (k) contact is made, and 



Chap. III. { 4. Y, W, WH — XVI TH TO XVni TH CENTURY. 187 

releasing both contacts simultaneously, so (wr) probably 
always was (rw), the labidl modification of (r), produced by 
keeping the lips rounded during the whole time that (r) is 
trilled. It is similar to the sound in French roi, which 
Feline writes (rua), and which English now call (rwAA), the 
true sound being {tiv&), which produces a species of evanes- 
cent (u), but whether before (r) as Hart wrote (ureit), or 
qfter (r) as Feline writes, appears doubtful to the ear, simply 
because it is during (r), p. 131. Similarly (yy) is (tito) or (it) 
with a labial modification, and all the '' round " vowels might 
be written as ordinary vowels followed by the labial modi- 
fication {w), p. 161. At the same time, in transcribing the 
notation of others, I shall generallv use (wr), although this 
is probably as incorrect as (rw) would be, and is very diflicult 
to pronounce. The notation (wr) is similar to the notations 
(hw, hj) ; in all three cases succession (w + r, h + w, h + j) 
is written where simultaneity (w*r = rw, h**w = wh, h**j 
= jh,) is intended. See <nv^ wl, wr in Anglosaxon, Chapter 
V, § 4, No. 1. 

The interchange of the vowel (i) with the consonant (j), 
and the vowel (u) with one of the three consonants (w, bh, 
v) is an interesting phenomenon in all languages. In Europe 
(w) is thought to be peculiar to England ; Wales also claims 
it, but the claim is doubtful, as its (w), if it exists, is con- 
fused by its writers with (u). In Arabic however (w) is 
quite at home, and also serves to mark the vowels (o, u). In 
Sanscrit, if the native grammarians are correct, the (i) 
between two other vowels fell into (j) and the (u) into (v), 
and not (w) or (bh). In Germany (u) generates (bh) not 
(w). Similarly in modem Greek (ev, av) generated (ebb, 
abh) becoming (eph, aph) before mutes as (apntos*), although 
modem theory makes v a (v) or an (f) as (evris'koo, aftos*), 
evp^KO}, ovTo?. It seems probable that in precisely the same 
way, the original transition of the Sanscrit (u) was into (bh), 
and that the pronunciation (v), distinctly pointed out by the 
native grammarians, is a comparatively modem alteration, 
comparable with the chan^ of (k, ka, ff, ffH, q) into (tsh, 
tshH, dzh, dzhn, nj) and of (Ah) into (sh). The immediate 
change of (u) into (v) is difficult to conceive. 

The letter (w), or (u) forming a diphthong with a follow- 
ing (a), formerly kept the sound of (a) pure. Thus Bullokar 
writes (waar, war'm, waar'n, war'en, war, waa'ter) for ware, 
warm, warn, warren, war, water. As late as Wilkins we have 
(waez) for was. Price says that a is never sounded (aa) 
except before /, and hence he excludes the action of w. 



188 M, N, KG — XVI TH TO XVIII TH CENTURY. Chap. III. } 4. 

Cooper does not mention the effect of w, and Jones 1701 
only instances the word " water, sounded wauterJ* But the 
Expert Orthographist, 1704, says that a has its broad sound 
(aa) "between to and r as war, ward^en, toarm, toam^er, 
warren, watch, water, wrath.** It would appear then that this 
effect of te^ on a following a became prevalent at the begin- 
ning of the XVIII th century. It is by no means general in the 
frovinces, where (wat'er, waim, warm, war'm,) etc. still exist, 
have heard (waa'ti, ktrsel'ttt, Vwddwiiti,) from even educated 
speakers. Of course the effect of the (w) on the subsequent 
vowel arises irom beginning to pronounce it before the lips are 
sufficiently opened, so that the vowel becomes round, as (watr 
= wo), for which however either (wa), or (^wo) has obtained 
in practise. Although in London and the South of England 
(wh) is seldom pronoimced, so that (wAt) is the usual sound 
for both Wat and what, yet to write wot for what is thought 
to indicate a bad vulgar pronimciation. In the North of 
England (wh) is very well marked, and in Scotland it is 
often labialized to (kt^^h), owing probably to the intimate 
relation between (u) and (k). 

M, N, NG. 

These nasal sounds frequently disturb the pure sound of 
the preceding vowel, giving it more or less of a nasal twang, 
occasioned by allowing some of the breath to pass with more 
or less force through the nasal passages. We know that in 
modem French tw, an, on, un, represent four distinct ori- 
nasal vowels, palaeotypically written (ca, aA, oa, 9a) although 
their exact relation to the oral vowels is not pretended to 
be accurately determined.^ It is very difficult to determine 
how soon this change occurred. Palsgrave, who, it must be 
remembered, finds the French e feminine to be "sounded 
almoste like an o and very moche in the noose," ^ tells us that 
" if m or n folowe nexte after a in a frenche worde, all in 
one syllable, than a shall be sounded lyke this diphthong au 
and somethjmg in the noose," so that the nasality was not 
"very moche" as in the other case where no other writer 
recognizes any nasality at all, but only " somethyng." This 
would lead to am, an = (a^um a^un). Palsgrave notes the 
exception when " the syllable next folowynge of any suche 
wordes begynne also with a lyke consonant," such asflamme, 
where the sound of a is not changed — and we are left to 

^ See aboTo, p. 67, for a discussion > See p. 181, note, col. 2. 
of these sounds. 



!, 



Chap. HI. { 4. M, N, NO — XVI TH TO XVIII TH CBNTURY. 189 

euppose tliat the m and n have their normal sounds. As 
regards French e before tn and n Palsgrave says it '* shall 
be sounded Ivke an italian a and some thynge in the noose/' 
with a similar exception. See the passages cited for a on 

K 143, near the top, and for o, on p. 149, near the bottom. 

n the latter place, no distinction is made (except as regards 
the final e) between bon, bonne, which must be (bun, bun*e) 
putting (e) for Palsgrave e feminine, at a venture. He makes 
no mention of in, un, but in his transcription he writes " im- 
bevo, depainz, poant, insasiablo, inconsider^, uoazins, mayn, 
6vmblo, evnshemyn " for imbtie, depainctz, poynt, inaa/nable, 
inconsidere, voisins, maynt, humble, ung chemin, in which there 
is no apparent trace of nasality. 

On examining Meigret there is not so much evidence of 
nasality as in Palsgrave. From Meigret's notation, as may 
be seen in the numerous citations already given, there is no 
appearance of any nasal vowel. Indeed the following remark 
would seem to exclude the idea of any such nasals as now 
exist. He says : 

"Je ne veu* pas aosi oubKer qe la prola^fon FranqoB'ze n'uze 
pas fort sounsnt de dens mm, ne de deus mi, ensEmble, combien qe 
recritture ne les eparNe pas: come, En homme, comme, sommst, 
commsnt, commandemEnt, honneur, domier, somier, an9ieniie. II 
st yrey qe Ies mm ee rEDContret aos Auxrbes qi se tErminet En 
mEnt qaiit a, ou s ouuErt pre^edet: come pnidEmmxiit, suffizam- 
msnt. Notez aosi qe n finall' ayant En suyte, vn vocable comEn- 
qant par voyBUe (si qe ne sent qelqes aspirez) double sa pui8san9e : 
come En allant, En etant, qe nou' pronon9on8 come En nallant, En 
netant: tellement q'aotant sone I'un qe Taotre; e ny trouuons 
aocune diffEren^e." 

That is Meigret heard no difference between the final n in 
" Bn " and the initial n in " nallant," he must therefore be 
understood to have said (eu nalant) in lieu of the modem (a^ 
nalaA). See also John Hart's transcription of French, 
Chapter VIII, § 3, and suprA p. 150. There seems to be no 
intimation of the French nasal in Cotgrave, and Miege only 
says that English final m and n are sounded '' d'une mani^re 
plus forte en Anglais qu'en Frangais," which may mean 
almost anything. In his French part, he says nothing about 
an, on, but informs us that 

^* em m the same Syllable is pronounced am, the e taking the 
sound of a French a ; as embleme, ensemble. Except where the 
word ends in em, or emme ; as item, dilemme. And yet femme is 
pronounced /<im»w. .... So is ^ sotmded an. Except 1. after i or 
y, in which case the e retains its proper Pronunciation, but that it 
takes somewhat of the sound of an i; as in these Words bien, chim 



190 M, K, KO — XVITHTOXVinTHCENTUBY. Chap. IIL | 4. 

ftc." with other exceptions, thns antmne has '^e open" or at, bat 
tienne has *^e masculine." ''//t, making the first syllable of a 
Word is pronounced in French as in English, except the n, which 
is but gently sounded ; as incapable^ indivmbU, The same is to be 
understood of in at the end of a Word ; as fin, vin, vemn,*^ yery 
unlike the modem (eA, ea, sba). '< Before m and n in the same 
Syllable, it (m) takes the sound of the Dipthong ^ ; as humble, 
lundi."' 

The investigation of the time of commencement, and the 
origin of the French and Portuguese nasality, would be ex- 
tremely curious ; at present, however, we are only concerned 
with the effect of the French sound upon English ears. 

First then as regards aim, ain ; im, in ; urn, un, the English 
seem to have heard in the xvi th century and previously (aim, 
ain ; tm, m ; um, un), and to have pronounced accordingly. 
Thus Hart in his French Lord's praver writes (indui, point, 
peen) for indui, point, pain, where Efart's (ee) represents the 
contemporary English (ai). 

Next as to am, an the English generally heard an inserted 
(u), thus (aum, aun). This does not however appear in Hart, 
who writes (an, kotidian, ofanses, tantasion, pyys&nse, aman) 
for en, quotidien, offenses, tentation, puissance. Amen, The 
omission of the (u) may perhaps be due to his usual mincing 
utterance. Palsgrave however distinctly notices it, and to 
this must be due the orthographies aum, aun, which are fre- 
quent at this and an earlier date in English words taken 
mm tihe French. In Salesbury we have the example 
GALAUNT, galaumt (gal'aunt), and he particularly says that 
'' A in the British .... is never sounded like the diphthong 
au as the Frenchmen sounde it in commyng before m or n in 
their tongue." Levins, 1570, spells daunce, glaunce, launce, 
praunce, vaunt, but he is not fond of the orthography, which 
seldom occurs. The pronimciation of such words is still marked 
by many speakers, (p. 147,) and although some, especially 
ladies, say (daens, glaens, Ifiens, prsens, va^nt), others lengthen 
the vowel at least to (daesens) ete., while many say (dans, 
glans, lans, prans, vant), and others lengthening this vowel 
say (daans) ete., and the intermediate soimds (dahns, daahns,) 
are not unfrequent ; but although some say (vAAnt), no one 
perhaps will now be heard to say (dAAns, prAAns). 

In the combination ^nge, although we have the u inserted 
in Chaucer's time, a peculiar thinness seems to have been in- 
troduced by the -ge, for Salesbury gives oranges, oreintsys 
(or'aindzhiz), (p. 120,) and Butler says that before -nge, a is 

1 See alio the passage quoted wn^tk p. 126, and the obeerratioiis upon it 



Chap. HI. § 4. M, N, NO — XVI TH TO XVIII TH CENTURY. 191 

pronounced as at, (ai) or occasionally (ee), as in change, range, 
danger, stranger, words which retain the evidence of this pro- 
nunciation in the modem form (tsh^^dzh, r^nndzh, deeinr 
dzh'J, streHndzh'j). The last word is said to exist in America 
under the form (stra^i^ndzh'j). 

As to om, on, the English as we have seen, p. 150, heard (urn, 
im). In the older English, in which, as we see from Palsgrave 
and Bullokar, ou was pronounced (uu), we consequently find 
otim, oun = (uum, uun) for these sounds, and these became 
(oum, oun) in accented and (um, un) in unaccented svllables 
in the xvi th century. Hence the final (un) of Salesbury in 
ooNDiCYON, condisytcn (kondistun) ; exhibition, ecaibrnton 
(ek8ibisi,un) ; PROHiBmoN, proihisiwn (proo,ibisi,un). To the 
wav in which Palsgrave heard o pronounced in French even 
bemre wc, we may attribute Salesbury's (truim) for throne. 
We have also in the xvi th century a distinct recognition of 
the vocal ('m, 'n) constituting a syllable. Bullokar has even 
separate signs for them, an accented m, n\ 

The ^ttural nasal (q) seems to have been the regular pro- 
nunciation of ng in English, but it was not recognized as a 
simple sound by the older writers. There is a difficulty in 
pronouncing the true dental (n) before (k, g) so that nk was 
commonly written for (qk) or (qhk) as Mr. Melville Bell, 
among others, thinks the sound shoidd be more correctly 
written, and ng for either (q) or (qg), as in singer, linger 
(siqu, Itq'gj). This was observed by the Latin Grammarians. 
Nigidius, quoted by Aulus Gellius, lib. xix. cap. 14, says : 

'* Inter Uteram N et 6 est alia vis ; ut in nomine anguis et angaria 
et ancora et increpat et incurrit et ingenuus. In onmibus enim his 
non verom N, sad adulterinum ponitur. Nam N non esse, lingua 
indieio est. Nam si ea Htera esset ; lingua palatum tangeret." 

Nigidius appears to have considered this n to be ^, or 
perhaps only related to g. The Greeks wrote 77, 7/c, 7^ for 




(qg) was the older form in all cases. This would at any 
rate accoimt for no special symbol having been assigned to 
(q), in most languages. It exists in Sanscrit ^, but few 
Sanscrit transliterators think it necessary to provide a sepa- 
rate symbol for it. In recent English (q) occurs frequently 
as a final, did it so occur in early English? This is a 
difficult question to answer, when we consider the practice 
of modem Germany, because the present pronunciation of 
German and Dutch being less altered than English, repre- 
sents an earlier stage of English pronunciation. Now 



192 M| Ny NO — XVITH TO XVinTHCBNTURY. Chap. IIL J 4- 

according to Bapp ng is (qg) when final, and (q) when 
medial over the greater part, especially the North, of Ger- 
many. Hence ^mger Oesang would be (szeq'er sezaqg')* 
Practically, however, as final (g) is very difficult for Germans 
to pronounce, they use (qk) so that Oesang Dank rhyme as 
(ffezaqk* daqk).^ This is not the case in central Germany, 
where (q) final is common, and where therefore (gezaq* 
daqk) do not rhyme. Even in England many speakers con- 
fuse thing, think under (thtqk), but this seems to be an 
exceptional word. 

Gill appears to be the first writer who recognises (q) as a 
separate element. He says, leaving his notation unaltered : 

<'i^ in illis [Uteris] est quas nihil mutare diximus : at si >^, aut y, 
sequatur paulum minuendiet est nostra sententia : nequs enim (si 
accurate expendas) plan^ ita profertur in thank et think quemad- 
modum pronunciatur in hand manus, et ndn koke nnllus. Sed 
ne adeo nasutuli videamur ut nihil yetustate rancidum ferre possi- 
mus : quia k, ibi clard auditur, nee congruum esse reor quicquam 
yeritati propinquum immntare ; monuisse tantum volui, sed te in- 
vito non monuisse tamen. At si y subsequatur vt in thing res et 
sonp canticum; quia sonus UtersB ff ibi nullus est, at semivocalis 
pland alia quae ab n non minAs distat quibi m ; literse ng. una erit 
ex illis compositis, quibus fas esse volui sonum simpUcem indi- 
care, ut in sing canta, ct among inter, hue etiam refer iUa in quibus 
g, ab n, ratione sequentis liquidsB quodammodo distrahitur, a spangl 
nitella, tu intangl implicare.^' 

Hence he said (stq, amoq*, a spaq'g'l, tu intaq*g'l) according 
to the present usage of ng. It would appear therefore that 
we are justified in adopting this usage from at least the 
xvith century, and, in the uncertainty which cannot be 
dispelled, it will be safest to adopt it also from the earliest 
times that English became distinct from Anglosaxon, although 
the North German custom may have been that of Angh>- 
saxon itself, namely to call ng = (qg) when final, and (q) 
when medial. 

Gill names (q) as a bad pronunciation of the Hebrew ff, 
which is still heard, being replaced by (gn) when initial, as 
Europeans generally find a difficulty in initial (q), although 
it is not un&equent in extra-European languages.^ Sales- 

^ ThiiB VosB in his MinnelUd has " Sie trankte dich mit Reben/ronJt / 

" Der Holdseligen Und freudig tonte dein QetMng** 

Bonder Wank I have not noticed such rhymes in 

8inf * ich frdhlichen Schiller and Qoethe. 

Minnetaif^ : 

Denn die Reine, * The mlgar Parisian, howerer, sajs 

Die ich meine, (qja pa) for il Wy m jmm, and the Yien- 

V^nkt mir lieblichen Tbhedanh,** nese porters will call a gentleman («i 

And again in his address to Lather qa«d*n) or (oi q4Mhd*n) for mmt Onttim* 



Chap. HI. }4- L — XVI TH TO XVIIITH CENTURY. 193 

bury speaks of the ''Latine vocables agnua, magntM, iffnis, at 
what time tbey were thus barbarously sounded angnua, 
tnangnus, inffnia," meaning (aq*nus, maq'nus, tq'niiB). This 
nasalisation of (g) into (q) before the following nasal (n) 
seems to have been common in the middle ages, and has 
crept into the Latin orthography of the period. Gill in 
English gives both (bentg'n) and (bentq*n) for benign,^ 
This (qn) is the regular pronunciation of ffn in Modem 
Swedish, the poet TegrUr bemg (Teqneer*).* 

^6 (<lg» ^l'^) A^ heard in Itidian and Spanish, but they 
are unknown in French. The older orthography of French 
had ng in many cases where the nasal (a) is now heard. 
But Meigret does not recognise this, writing n simply in 
such cases. The French con^e our (q) with their gn = 
(nj) and some Englishmen seem to have fallen into the con- 
verse error. The Spanish n,' Portuguese nA, Italian and 
French ^, are all (nj), or nearly (nj). 



The great opening for the passage of the voice while L is 
pronounced and the very slight nature of the vibration of 
the sides of the tongue, tend to give it a strongly vocal 
character, and not unfrequently the L has been entirely lost 
in a vowel sound, produced simply by not bringing the tip 
of the tongue close enough to the palate to form a division 
of the passage and throw the voice out on both sides. Both 
French and English seem to have had a tendency to labialise 
(1) into (Itr) after (a, o), that is they rounded the lips either 
during the vowel or just as it glided into the consonant. 
The Latin alter thus became (alerter) or {bw\w\^v) felt as 
(aolfdre), till the (1) became absorbed, that is, neglected for 
convenience of utterance, thus (aotre), which is Meigret's 

^ Strange as the final combination modem Spanish for (Ij). The tilda 
'q*n) may seem, there is a well known oyer the k was merely the usnal ab- 



don Tnlgarism in which it is very breviation for the second n. " En los 

fiuniliar (I'q' nz) for (an-jimz) onions, tiempos mas antiguos de nuestra lengoa 

' In Sjoborg's Stoedisehe Spraehlehre, se explic6 con dos nn juntas esta pro- 

p. 10, this is the role laid down, but nunciacion, y algunos se ban persuadido 

mogna, tagne, ataane are said exception- & que la tilae sobre la n, como hoy se 

aUy to preserye the 1^) and in I'ogn the usa, se introduxo para denotar la otra 

sound IS (Icein). The irregularity of n que se omitia, al modo que la tilde 

Swedish ortho^aphy as compared with puesta sobre las yocales se us6 fre- 

pronunciation is considerable, shewing qiientemente en lugar de n.** Orto^ 

a great alteration of pronunciation in grafia ds la Lengtta CoMieliana^ com" 

the comparatiyely short period since pu$tta por la Real Academia JBspanola, 

the orthography was established. 7th ed. Madrid 16mo, 1792, p. 64. 

* In old Spanish nn, just as // is the 

13 



194 L — XVI TH TO XVIUTH CENTUKT. Chap. IIL § 4. 

form, and finally (ootr'), the modem form. In England 
^elto) became felt as (aid) or {sitoiw) and this degenerated 
mto (aaI), perhaps through (avl). Finally when a conso- 
nant followed, it was more convenient to leave out the (1), 
and the lazy or the nimble tcmgue, as usual, took the most 
convenient or shortest road, and (1) disappeared. The Scotch 
even lost it without a following consonant as (kxA aa) for 
(kAAl, aaI). The passage was perhaps (talk, talt£7k, taulu^k, 
tauk, tat^^k, tAAk). Whether (tAAlk) was ever said, except by 
Gill's ''docti interdum^' is more than doubtful. 

Similarly after (oo) we had (oolte?d, ooidt<7d, oovld) or 
(ooul). In this case the (1) was not generally absorbed, but 
we have provincially (ood) for old. 

Salesbury says that in the English calme, call, tiie a ** is 
thought to decUne toward the sound of the diphthong au.'* 
Again: ''o in Welsh going before //, soundeth nothing 
more boystous, that is to say, that it inclineth to the sound 
of the (Uphthong au (as it doth in English) no more than if 
it had gone before any other letter.'^ *' L hath no nother 
difference in sound in Welsh than in Englysh. And note 
that it neyther causeth a nor o when they come before it, to 
sound anye more fuller in the mouth, than they do else 
where sounde, commyng before anye oth^ letter.'' '' Some- 
times a has the soimde of the diphthong aw especially when 
it precedes I or II, aa may be more clearly seen in these 
words : balde bawld (bauld) calvus, ball, bawl, (baid) pila ; 
WALL watol (waul) murus." '* also before Id or U ia pro- 
nounced as thouffh w were inserted between them, thus 
oohDE, cowld (kould) frigidus, bolls bowl (boid), tolle towl 
(toul) vectigaL'' " In some districts of England // is sounded 
like w, thus bowd (booud) for bold, bw (buu) for bull, caw 
(kau) for call. But this pronunciation is merely a provin- 
cialism, and not to be imitated unless you wish to mince like 
these blimderers." But this did not arise from mincing, but 
from broadening. The mincer, so £Etr from droi)ping the 
front of the tongue from the palate, raises the middle part 
and produces (Ij) which de^nerates into (i), as in Modem 
French. The effect of / which Salesbury names is generally 
recognized and exists to this day in tne modified form of 
(aa) for (au) and (oou) for (oou) or (ou). The sound (ou) 
is however, heard in (ould) Ireland, either in its genuine 
form (ou) or its modified form (au) at the present day, 
Buchanan in the xvui th century wrote (sauld, keuld, bduld, 
skauld, teuld, Hoidd, sauld'Jtr) for sold, cold, bold, scold, told, 
hold, soldier. Sheridan did not imitate him, but scrupulously 



Chap. HI. f 4. L — XVITH TO XVm TH CEHTUBY. 195 

-used (ool) and notes (bAuld, kxuld) as Irishisms for (hoold, 
koold), in which again the Irish were only following the 
fiishion of the English in the xvii th century. 

Salesburj recognized ('1) or pndonged (1) as forming a syl- 
lable by itself in able, sable, twtkclb, wrynclb, writing 
abl, sabl, twinkle wrinkle (aa'b'l, saa'bl, twiq'k'l, wrtq'k'l). 
In this he is ftilly borne out by all subsequent writers^ Hart 
and Bullokar have special signs for ('!). Hart considers it 
to be the same as the Welsh il, (Ihh) which i» the reason 
why he j»x>Tides it with an especial character. He says 

''W^e haue further the 1, aspired lyke to the Spanishs and 
Walsh' often Tse of the 11, which maketh the .xij. dmabe or doll 
sounde, but we vse it not that I know of, at tiie begiomng of 
any wor^ as they do : bat often at thend of words, as in this 
sentence, the bedle is hable to fable. Where we wrest the e, 
which 13 but closely or (as it were) halfe sounded : wherfore we 
may with as smal cost and labour, as of the rest, rse a fit figure 
for it: and neuer n^ede to Tse the 11, or lb, and for the reasons 
abouesaid not to abuse the h.'' 

Smith says : 

<'Qui nescit q^ sit esse semivocalem ex nostra lingua fiacil^ 
poterit discere, ipsa enim litera L. quandam quasi yoealem in se 
ridetur continere, ita ut juncta mutsB sine yocali sonum &ciat,^ut (aabl) 
habilb, (staabl) stabilis, (faabl) fable, &c; alii n^t/ 9UM fabil, alii 
^ul 9tahd fuhul scribunt, sed ne quicquam i»x)nuDtiaiit ; nam con- 
sideratius auseultanti nee e nee i nee u est, sed tinnitus quidam 
TO^Hs naturam habens, quae naturaliter his liquidis ineel. In 
omnibus his quidam e addunt in fine, vt able, stable^ fable: sed 
certb illud e non tam sonat hie quam ftiscum illud et foemininum 
Francorum ^,' nam ne quicquam sonat." 

^ Like Saleflbory lia oonftiflee the to diaw the diftiiictioiL In the same 

Spanish (U) with the Welsh (Ihh). way I hare represented the final -# in 

s This 18 a recognition of an ob- Chancer by (e), as donbtfdl. Bappcon> 

fcorely sonnded fhud French #, the {Mre- tinnes : '* x et whare the srlkble MMwith 

sent {$), in the XTi th century, agreeing double n reenlts, (ikBDm) nermen is di*- 

with Palsmye bnt disagreeing with tinctly prononnced." Rapp writes fnBn- 

Meigret In the same way most Ger- n^n) owing i» his onstom of donoling^ 

mans call their e final in ^ne gute Oabg the consonant after a stepped Towet. 

a fine {e), and rery many Englishman ^ To exhaust what I have to say abont 

would odl it (#). Bapp, Physiologie the unaccented «, obserye that the first 

der Bpraehe, yol. iy. p. 16, says (trans- e is taken as the natural yowel in the 

lating the passage for oonyenience) : terndiuBiii<m tnm, (gtMmm) ae/aUmmt, 

" Short («) only occurs unaccented, as or else elided. The natural yowel is 

(b^ ge, sndtf), h$^e,0fide, doubtful, half- ctisiinet before M, R, S and T, (aatom, 

mute, or, when heard, with a faint nasal faatsr, guutei, bsBtet) athmn^ vater^ 

in #91 (jgKEheo) geben. On account of gut$i^ biut, foreign names as (moostfs) 

the uncertainty we generally prefer the of course excepted ; custom yaries in 

orthography (gnsben)." Rapp uses $ fjnupitar, mumtBr). The enoiitics (or, 

much as the palaeotypic (e), and repre- for, tsar; or, oar) m*, iw, ztr ; w^ dm- 

sents (b, «) by ^, ^, but (bb, m) by d, L must be mentioned amoi^ the (orj. 

Oenerally I haye used (e, ec) for his ^ The • is always mute before L, as in 

a, bnt in this passage it was neoeetaiy «U aUied langnages, as (mit*!, sq'l) 



196 B — XVITH TO XVmXH CENTURY. Chap. IIL § 4. 

In Bohemian the (1) is fbllj recognized, and forms the 
.only vocal element in some accented syllables, as wlky 
(bhl'ky) wolveSy sIsm (sl*za) a tear. It seems probable that 
it was the sound intended to be represented by Sanscrit 
^ 1( = ('1, "1) commonly called (Iri, mi), unless these were 
originally cerebral, as ('l, ''l). The modem French do not 
possess the sound, but pronounce (tabl^) or (tablh), some- 
times merely (tab'), althoufi:h their orthoepists write (tabl), 
and contend that (1) here lorms a syllable by itself! As we 
. have seen Hart indicates his own pronunciation of final -fe 
•^ to have been (-Ih.) 

R 

In English at the present day r has at least two sounds, 
the first, when preceding a vowel, is a scarcely perceptible 
trUl with the tip of the tongue (r) which in Scotlan<t and 
with some English speakers, as always in Italy, becomes a 
clear and strong trill (.r), but as this is otnly an accident of 
speech, it will not be further noticed, (r) being used indiffer- 
ently for both. The second English r is always final or pre- 
cedes a consonant. It is a vocal murmur, differing very 
slightly from (a). I seem to hear it occasionally m two 
forms, differing nearly as («, e) which I represent by {u, i). 
As however this distinction is, certainly, by no means always 
made, I do not usually mark it. This second (j) may diph- 
thonffise with any preceduig vowel. After (a, a, o) the enect 
is rauier to lengtnen the preceding vowel, than to produce a 
distinct diphthong. Thus fariher, lord^ scarcely differ from 
father, laud ; that is, the diphthongs (ai, oj) are heard almost 
as the long vowels (aa, aa). That a distinction is made by 
many, by more perhaps than are aware of it, is certain, but 
it is also certain that in the mouths of by far the greater 
number of speakers in the South of England the absorption 
of the (j) is as complete as the absorption of the (1) in talk, 

mitt^ mgel, and thii should be theo- " The tenninations (sq'lii, shmaUh'ln, 
retically the case eren when tennina- gaab'ln^ engein^ sehmeieheln, gabeln. 



tions are added on, although it is then are difficult to pronounce with purity 

certainly difficult to continue to make for foreigners and even for Germans, 

the Towelless L form a syllable by it- Finally ue natural Towel or mute $ is 

seli( as fshmaiih'l-ai, Bq*l-l£ndar, mifl- generated in popular specNch by n^- 

Isndish^ sehmeiehelei^ engellandery mit" lecting ancient terminations as in 

UUdnchseh." This theory is partly rnnkar, iizar, mnd'lshtat, iq'lshtat, 

wrong, for the yoeal ('1) being only a Qoktor, profesar) and among the un- 

lenethened (1) = (11) is nataraUy short- educatea even in (jeesus, jeena, goota)." 

euM before a rowel, as (ste^-b 1, stM*- This passage is mteresting as serving 

bliq; ftd*'!, fid'li^: so it should be in to shew the state of a language in 

C^ennan (shmaiArnlai), but in fact which the final ^ is in a transition state. 

(shmaiAhtflai) is said. Bapp continues : See saprk p. 119, note, col. 2. 



Chap. III. H- B — XVI TH TO XVIlITH CENTUKY. 197 

ufolk, pMlm, where it has also left its mark on the preceding 
voweL When Dickens wrote Count Smorl Torh he meant 
Small Talk, and no ordinary reader would distinguish be- 
tween them. But in (ai, oj) propel*, there is a slight change 
of ling^ position generating a glide, and consequently Mr. 
M. Bell represents the effect by a glide character especially 
invented for the purpose, which he terms the ''point glide *' 
and describes '' as a semivocaHzed sound of (r). The diph- 
thongs (ei, ai) are very difficult to separate from each other 
and nom (oe). But the slight raising of the point of the 
tongue will distinguish the diphthongs from the vowel 
in the mouth of a careful speaker, that is, one who trains 
his (MTgans to do so. No doubt the great majority of 
speakers do not make any difference, and I think that the 
best representation of these sounds is the simple (i) or (^i), 
which IS in this Feq>ect wholly comparable to the (^1) already 
discussed. It seems to be an indistinct murmur, differing 
from (1) by not having any contact between the tonme and the 
palate, but similar to it, in absorbiug a variety of other vowels. 
The following is a comparison of my notation of this mur- 
mur (i) and its various diphthongs, with Mr. M. BeU's. 
The (i) character will express Mr.lBell's glide, and (itc) its 
labialised form, as in Introduction, p. 15. The examples 
have been taken from Yisible Speech, pp. 113-116. 
Bbll Ellis. £xa)cplbs. 

word, jewmey, ftimish = (wjd, dzhjn't ) &c. 
pap^, ctrcuitous, answ^, martyr = (p^'i) &c 
iirey lyre, choir = (fail, laii, ktraii) 
wtry, fi>ry = (wainri, fainn* ) 
hour, ^wer = (eui, paui) 
ourselves = (aiuselvz') 

dowery, fLowery, showery = (dani'rt, floiUT* ) &c. 
luird, cbrk, h^ort, gourd = (naaid, klaaik) &c. 
barbarian = (b&ibeei'rrsn) 
altar, grammar, particuLir = (AAl'ti) &c, 
starry, tarry (adjective) = (staaTi, taa'r*) 
prefer, Mmest, firm, myrrh, guerdon^i (prifU*)&c 
near, he&r, hare, we're, "pter = (niii, biii) &c. 
aerie, er&, weary, "peereBa = (iiiTi ?, iiiTc) &c. 
car*, aer, pair, Ayr, prayer (petition), tlwre, 
b«ir, ne^er, iheir, eyre, mare = (keei) &c. 
mayor sb (meei) 

canary, fa«ry, therem, bearing =(kmieeiT»), &c. 
war, word, swarm, dwarf = (waaj, wAAjd) &c. 
poor, moor, tour, sure = (puiu, muiu, tuuj^ &c. 
poorer, Bt«rer, assuring, townst =^ (puoi'rj) &c 
cure, pure, endure, immur# = (kiiu, piui; &c 



XI 


1 


OT 


J 


ail 


ail 


ai'r 


ain* 


aoj 


auj 


tLlW 


am 


au'r, aur 


euir 


01 


aai 


ai 


ai 


ahj 


I 


a'r, oar 


aar 


ecu 


^,1 


iz 


ill 


i'r 


iiir 


ei 


eei 


eiw 


eei 


e'r 


ecir 


Al 


AAJ 


fa 


unj 


«'r 


uuir 


jm 


iui 



wr 


iiur 


iiu 


iuj 


iu'r 


iuir 


OJ 


001, 001 


o»r 


ooir, ooxr 


01 

olu 


901, AAI 

1 


VI 


J, im 


iVI 


iiu, Ji 



198 E — XVITHTO XVIUTH CEHTURY, Ohap. HI. { 4* 

BsLL. Ellis. Sxakplbs. 

fiiry, ptirer, enduring =» (fiiui^ piiuTj) &c 
Itir^, alltfTtf =s (liui, «liiu*) » 

Itfrid, alltiring = (lim'rtd, diiuTtq) 
hoar, o*er, door, floor, borne, torn, sor^, cof^, 
pouff tournament^ towarda = (booi booi) &c, 
glory, souring, pouring ^ {ghorri, gloaiit ^ &c. 
extmordinary, Qteorg^, order, bom =(booin) &<j. 
spectator, tailor, razor, orator = (spektM-ti) ftc. 
aztfr^ fiaaurtf, measKro, seiztiro «« (o^zh'j) Ac 
natwro, feature, staUrv as (no^iaiu nMvi) ftc. 

It will be observed that Mr. BeU has not marked a long 
Yowel iH man^ places where I hare marked one. His 
general habit is not to distinguish the length of the first 
element in diphth<mgB« Simple r is used in cnrdinary spell- 
ing, after long yowels, for tne combination (jt), or ('r) as 
Mr. Bell prefers writing. This combination is rery pecoliar 
in English; compare dear^ deary, mare, Mary, more, ghry, 
poor, poorer, with the IVench dire^ dirai^ mire, maitie, Maure, 
aurai, tour, Tomrame. 

The Scotch do not use (j) at all^ but only (r) or rather 
(.r), saying (word, serf, serf, karv) word, serf, eurf, carve. 

In itelf (.r) is constant, in France and a great part of 
Gennany (r) is prcmounoed in lieu of (r). Could it be to this 
sound that Fcdsgmve alluded when he said : 

'' ^ in the frenche tonge shalbe sounded as he is in latyn without 
any excepti<m, so that, where as they of Panrs do sounde somtyme 
r fyke s, sayeng piotys for parys, pMisieu for parmen, ehaiu for 
chayre, mawy for marVy and suche lyke, in that thyng I wolde not 
have them Mowed, albeit that in all this worke I moost folowe the 
Parisyens." 

Certainly % would be the nearest character b^ which, with- 
out explanation, he could haye giyen a conception of the true 
r grassey^ ou provengal, the French (r), which is not unlike 
the Arabic (grh),^ and l^e Northumberland burr. The last 
is often confused by southerns with (g), (Hafi;rh'iet) Harriet 
sounding to them like (Haeg'iet). The Spanisn r mmve is (r), 
with no more trill than in English, but the r fuerte is, ac- 
cording to Mr. M. BeU, the "usual (.r^, but according to M. 
Fayarger, (.r), a sharp uyula rattle without any moisture.' 

^ The Frenoh rama (razia) is a cor- remo^ rieo, romo^ rueda; after /, n, « 

mption of the Arabic iV^ ferhazaat-)- ^^^J' malrotar, mriquec^, homra, 

'^ J^ ^ ' i«r<wis»<M^'<><^<o/uicompoimdfl, where 

* See Ortografia de la lenffoa Cag- the second part b^^ with r ; and 

tellana compoesta por la real Academia where rr is written as harra, earro. In 

Espafiola, 7th ed. Madrid, 1792, p. 70, other oases the soft r (r) is to be pto- 

where the strong r (.r) is said to ocoor, nonnoed. 
at the beg:iniii]ig of words as mom. 



Chap. m. $4, E — ' XVI TH TO XVm TH CEKTUBT. 199 

No allusioii to more than one ioand of r is ibimd in any of 
the older writers except Ben Jonson, yet it oan hardly be sup* 
posed that even if the northerners have retained (r), the com- 
plicated (r, J, Jr) svstem could have grown up in a single 
century in the South. For the old wr=^ (rw)^ see p. 187. 

1647. Salisbury has the fc^owing words which are now 
pronounced with (j), the old spelling being in small capitals 
and the phonetic Welsh in italics. 

PAPTB papyTy QUABTEs kworter, sra syr, tbesuse tretuwr, vertub 
verhtw, chubchb Uurts, laddeb lad-dr, bladd' hlad-dr, eicpeboube 
$mperwr^ euebxobb efermww^ thonpbs thumdr^ wondbe wndr, suffbe 
•wj^er, Gtlbsbt Ghibert, otngeb t$inUir, hoitoubb onor. 

Here we find the unaccented syllable er or ir represented 
by the Welsh er, yr, ir, and finally simple r. This points out 
to an indistinct murmur, where the writer tries first one 
vowel sound and then another and finally gives them all up 
in despair, and trusts to the simple consonant (r) as best re- 
presenting the sound. Now in Bohemian (r) is recognized 
as sufficient to form even an accented syllable, as srfM a roe, 
%mo kernel, tm thorn, dm turf, chrt greyhound. I do not 
know whether the sound is here (j) or ('r), but as Ziak (Boh- 
mische Sprachlehre) compares it with the German termination 
-er, which Rapp (supri p. 194, note) declares to be (er), it 
will be safest to consider it as ('r) or ('.r), though even the 
Germans are apt to fall into the convenient (j) final. The 
examples from oalesbury would therefore lead us to conclude 
that {^t) was sufficiently common in English of the xvi th 
century, but would not allow us to assume either that the 
syllables he writes er, yr, ir^ r were (j), or that every final r 
was (i) and middle r (jr). 

1569. Hart says of / «fi n r that they are '^ rightly vsed 
in sounde when they be single.'' 

1680. BuLLOKAKy who has ee^pecial signs for ('1, 'm, 'n), 
has none for ('r) or (j), writing (foormer, dheer, aar, 
severawl, letterz, figyyrz,) for former, there, are, several, 
letters, figures, 

1521. Gnxsays: " ^^ fere trissyilabmn est ; «ir/ mobilis ; apud 
alios enim diphthongas valet, hie irl auditur, illic trV^ 

Here some tinge of ('r) or (j) seems to come into play, 
(a'ri, e'rl, eerl). Gill also writes (fei-er) fire, and complains 
that they say (fir) in place of (fei-er) in the East of England. 
But the Germans also write /^wer (fayer, foyr, foir), and this 
does not imply (i). 

1653, Wallis and 1668, Wilkins have no allusion to (j). 



200 B — XYiTH TOXvniTHCKvnnnr. CHAP.IILf4. 

If it was then heard it was poaaibly considered to be an 
OToneoiis nttermce not worth naming. 

16S5. CooPKR aojB : "Yerba Angticana ft latina de ri v a liva qaas in 
Ofriguie flciibuiitar cam er scribinnis item fr^ pronn nfuwmifl antem 
vr (ar), mm qam ae pnifer ri debet, sed quia propter Uterc r Tibra- 
tionem tix aliter efferri potest; nt tMnr coluber, jtrefrr prsefero, 
$Under temiiB." 

Here the mention of the vibration exclodee (x) and insists 
on (or) or (V). Cooper proceeds to give lists of soch words 
with final (or) spelled -ar, -er, -tr^ "Cr^ and eren -urf, shewing 
that he prononnced "lure as (-tor) in adventure. Juncture, 
lecturef nature, pasture, picture, rapture, ser^ture, etc, which 
are ndgarisms at present under the form (-ti), although in 
figure, injure, measure the (i) is common (f«g*x, tn'dzhi, 
mezh'i). Cooper also says: '^r sonator post o in apron 
eremiale, citron citremn, environ circondo, gridiron craticnla, 
iron ferrom, saffron croons; quasi scrib^'entur apum, &c,** 
almost as at present. 

1688. MiseE also says of r, '' en certains mots la voyelle qui la 
suit se prononce derant, comme en here, sire, spire, Jmndrei, apron, 
citron, sajfron, iran;^^ 

but this can only point to (or) or ('x) after what Cooper has 
said, Jones identifies the sounds of er, ur, referring firom 
the latter to the former, and making both co-extensiTe with 
the modem (x), but he does not help us to determine the 
double power of r. 

1640. Ben Joksoit says : '' B is the Ikgs letter, and huireth in 
the sound ; the tongue striking the inner palate, with a trembling 
about the teeth. It is sounded firme in the beginning of the words, 
and more liquid in the middle, and ends : as in rarer, viper, and so 
in the Latine." 

This seems to imply that a difference was made so early as 
the end of the xvi th and beginning of the xvii th century. 
The precise meaning of the vague terms ^rm and more liquid 
cannot of course be assigned. But probably firm meant 
more consonantal and liquid more vocal, so tnat something 
like the difference between (r) and (x) is indicated. The 
reference to the Latin is of no value, as it was only to its 
English pronunciation. 

Walkbr, 150 years later, refers to this passage and says : 

'' The rough r is formed by jarring the tip of the tongue against 
the roof of the mouth near the fore teeth : l^e smooth r is a vibra- 
tion of the lower part of the tongue, near the root, against the 
inward region of the palate near the entrance of the throat This 
latter r is that which marks the pronunciation of England, and the 
former that of Ireland.'' 



Chap. III. j 4« R — XVI TH TO XVIII TH CElHrtrRY. 201^ 

But he does not proceed to point out where the rough and 
smooth r were pronounced, and his description of the smooth 
r better agrees with a gentiy pronounced (r) or (grh), the 
uvula trill, than with (j). Tne theory of a vibration of the 
back or lower part of the tongue is untenable ; that part of 
the tongue is too firm to vibrate in the manner conceived. 
And in England we do not perceptibly vibrate the uvula. 

Smart, who has entered into the consideration of (j) more 
than any preceding writer, calls (j) a " guttural vowel 
sound." He says of (r) that " it is formed by a Btrong trill 
of the tongue against the upper gtun,'' to wmch it may be 
objected, first, that the tiill is gentle in English, and, secondly, 
that the tongue vibrates freely, near, but not striking the 
upper gum. For (j) he says, "there is no trill, but the 
tongue being curled back during the progress of the vowel 
preceding it, the sound becomes guttural, while a slight 
vibration of the back part of the tongue is perceptible in the 
sound." Now I do not find the tongue to be "curled back," 
although it passes from the preceding vowel to the (i) position, 
and I find no vibration of the back of the tongue, though 
vibration of the velimi mav occasionally be felt, and some 
persons may more or less vibrate the uviua. 

On account of the resemblance of (j) to (e), a sound to 
which all unaccented vowels approximate in the mouths of 
of many southern speakers, and also because when (j) is 
followed by a vowel, it is usual to interpose (r) thus (Heei'rt, 
Hiij'nq), hairy^ hearing^ illiterate speakers — ^those who either 
do not know how to spell, or ignore the rules of spelling in 
their speech — ^usually interpose an (r) between any back 
vowel, as (a, a, a) and a subsequent vowel, thus (drAA'rtq, 
Iaat e-dhe-lcend, winder e Shi ,8eus) for (drAA'tq, Iaa ov dhe 
IsBud, wm'do ov dhe hous) drawing^ law of the land, window of 
the house. From this habit, a very singular conclusion has 
been commonly drawn by a great many people, namely, 
that such persons habitually say (drAAr, Iaat, wth'dar) 
when not before a vowel, — a feat which thev are mostly 
incapable of performing. They will indeed rnyme window, 
cinder, not b^use they say (wth'dar sm'dar) as generally 
assumed, with the trilled (r), but because they say (wmda 
stnda) or (wiwAi stn'di), omitting to trill the r in both cases. 

Another point on which Smart insists is the distinction 
between serf, surf, which Mr. M. BeU writes (saoif, saof ), and 
I write either (s.*f, sjf ) by preference, or (serf, saif ), or else, 
sinking the distinction, as is far the commonest practice, write 
(sif ) for both words. A distinction of course can be made, 



203 B — 3tVT TH TO XVin TH CEMTUBT. Chap, IIL {4* 

and without much difficulty, by those who think of it, and is 
made by those who have formed a habit of doing bo ; but the 
distinction is bo rarely made as to amount almost to pedantry 
when carefully carried out, like so many other di^uictions 
insisted on by orthoepists, but ignored by speakers whose 
heart iB in me thought they wish to convey, not in the 
yehide they are using. Smart, notwithstancQng the pains 
he has bestowed on this subject, finds that the wordB payer, 
player, slayer, which are dissyllables = (pe^V pl^ V> ^^'9^)$ 
rhyme perfectly with care, fair, hair, share, which are mono- 
syllables =(keei, feel, Heei, sheej) with a different yowel. 

The action of the ('1) in altering the preceding (a) into (au) 
and thence into (aa) has already been noticed. It is always 
the tendency of two sounds combined in n^d succession, to 
&;enerate some alterations in one or both, or to fuse themselyes 
mto some new sound (p. 52). This is yery marked with (jl). 
It is now not customary to pronoimce (ee) or (00) before (Jt). 
Such words as (meeJi^ mooi) haye a yery peculiar effect, either 
antiquated <»r illiterate, and are replaced by (meeJ, mooi) 
tnare, mare. Mr. M. Bell considers that (uu) is in like 
manner altered to (uu). This is certainly often the case, but 
(puuj) for (puui) haa no singularity in it We certainly do 
not change (ii) into (u) and say (iii) for (iij) ear.^ It is pro- 
bably this acticm of the (j) wnich has presenred the sound 
of (a) so that art, part are not (flDit, psBit) but (ait, pait) or 
(aait^ paaJt) or simply (aat, paat). Indeed, in ordinary 
spellinK» many writers now habitually use ar to indicate the 
sound (aa), in the same way as they use ar to represent (aa) ; 
(p. 197). At the same time (sdr, s&bbt) were certainly preyiedent 
in the xyii th century, and are fossilized in America. 

How far all these effects are modem, or how &r th^ were 
heard eyen in Ben Jonson's time, I haye been quite 
nnaUe to determine. But as (r) may still be said, and 
is still used by Irishm^i and Scotchmen ^implying an 
older form of English) and, carefully inserting ( ) or (a), 
is eyen now used by many Englishmen without giying 
offence to the ear (u'r, iiar), it is certainly s^er to 
assume that there was formerly only one sound of (r), 
but that a murmur (') was generally inserted before it 
when following a yoweL In my transcriptions, howeyer, 
I haye been obliged to omit this theoretical (') for which 
I haye no proper authority. 

^ But olMerre the Norwich street C17, p. 138, note, ooL 1. 



Chap. IIL {4- P> B. T, D. C^ K, Q. O. CH. J. 208 



P,B. T,D. C,K,a G. CH. J. 

The prontoiciatioii of P,B does not seem to have yaried in 
any respect. 

T,D nave now a tendency^ ignored by most orthoepists, 
under particular circumstances to pass into (tsh, dzh) ; thus 
nature, verdure are, perhaps most frequently, pronounced 
(ne^'tshi, yrdzhj), the last word beinff in that case identSed 
with verger. This alteration takes place generally through 
the action of a palatal sound, originally (yy), then (iu, ju) so 
that the transition was (-tyyr, -tiur, -troi, -tjj, -tshi), I 
have not found traces of the change however, but the pro- 
nunciation (nee'ti) or its equivalent given by Jones seems to 
shew an effort to avoid it by omitting the pcifatal element (j). 
In the xvni th century Sheridan carried this still further and 
allowed for such pronunciations as (tshuut'Oj) for tu^. The 
palatals (i, j) have always had a great effect upon preceding 
consonants of the dental and guttural class, as they tend to 
materially alter the position of the tongue, in order to facili- 
tate the transition to a following vowel. The languages de- 
rived from the Latin are full of mstances. It is a fashion in 
modem English to resist, or to believe that we resist, tlus ten- 
dency in the especial case of -ture and 'dure, but we have 
given into it completely in -tion, where the t, hesitating in 
olasacal times between e and t, underwent a change which 
gave (-sioa) in Fr^ich, whence in English, first (-srun) and 
then (-shen), — tkeyer, except in orthoepical fsmcies, (-shon), — 
and in Italian ]^x>duoed (-tsitftih*ne). A similar change is re- 
oognieed in '-ekms, ^euU. And it is in vain to protest agaiiuft 
^iure, 'dure becoming (-tshj, -dzhi), at « time when even 
(-tjta, -dJ2a), though far less pedantic tiian (-tiuj^ -diuj), 
have a angularly orthoepistic effect. 

C, G also underwent a ^milar change, not from the action 
of an (i) sound, but paradoxically, as it might i^pear, through 
ikkQ action of a following (a) sound. The letter k is not mnm 
used as an initial in EngUsh and hence the observation refers 
in spelling to c but in sound to (k). It would be interesting 
to know when the English began to introduce an (i) sound 
between (k, g) and an (a) sound. There is no trace of it in 
orthoepists, but there are traces of it in a very early stage of 
our language, in the Anglosaxon orthography, and there are 
traces <h final (k, g), especially after (1, n, r) naving been also 
palatalized to {k, g). The word ckurch, now (teSiJtsh), but 
previously (tshwiah) if we may trust Salesbuiy's Welsh tran- 



204 P,B. T,D. C^K^Q. G. CH. J. Chap. HI. } 4. 

scription tsiurts, is an excellent example. The Anglosaxon 
forms are circ^ ciric, cyric^ circe, ct/rieea, the Gh-eek being 
Kvpuucov, which in the present Gh-eek pronunciation, pre- 
yalent certainly in all its main points when the word was 
transplanted into Anglosaxon, is called (kiriakon*), and the 
word (kirik) or (kirk) probably arose ^ from omitting one or 
two of the intermediate vowels. Ormin's kirrke = (kirk'e) 
and the Scotch kirk (kerk, ke.rk), shew the impalatalized 
fortti. That the initial consonant should haye yielded to the 
following (i) was to be expected, and although in modem 
high German we have kirche Qdrkh'e), the old high German 
offcen shewed an initial ch = (kh) or perhaps {kh), a palatal, 
although it possibly meant the upper German initial (ka). 
The final k in this word is palatalised in modem German, for 
it is (A;h) and not (kh), and it is to be remarked that the 
Germans always use (kh) and not (kh) after (1 n r) shewing 
the tendency of Germanic lanraages to this palatalisation. 
The transitional form between (xirk) and (tshirteh) was (kiik). 
From (k) to (tsh) seems a great stride. Yet there is no doubt 
that the passage was accomplished in Italian, where every 
(tsh) results from a palatal {k), and every (sh) from a palatal 
(fik) precisely as in English. In modem Greek xal, properly 
(ke), becomes {ke, ki, tiSii) in various dialectic pronunciations. 
In Sanscrit also there can be no doubt that the palatal series 
tr 1[ ^ U ^ were originally (k kn g gn q) auhough they 
are said to be now (tsh tshn dzh dzhH nj).' This is not 
the only change of the palatised (k). The older French seem 
to have generally palatalized the Latin c before a, as (A»mp) 
fr^m camptis, whence afterwards (shamp, shaA), (p. 53). But 
the change was often first into (s), whence (sh) became evolved 
by a furmer action of an (i) sound, so oceanus, oc^an, ocean 
(oAe'anus, os^aA, oo'sh^n). 

In pronoimcing (j) the middle of the tongue is arched up 
against the palate ; while for (k) the back, and for (t) the 
tip of the tongue only come in contact with the palate. 
When then (kj) or (tr) come together rapidly, the first 
change is to produce (kj) and (tj). By (kj) is meant pre- 
cisely the same as (k). The latter is generally the more 
convenient notation, but the former seems more suitable for 
the present discussion. For (kj) there is an attempt to pro- 

^ There is a possibility that eirc is remain ; few Englishmen would detect 

not of Greek origin, see Graff, iy, 481, the difference between (nj) and {g) that 

Dieffenbach's Goth. Wort. ii. 460. is (qj), and some mispronounce the 

This howeyer will not affect ine de- French ^ bb (q). The sound (ni) be- 

riyatiyes of the Anglosaxon. longs to a series (tj ^h dj dJH nj), not 

* It is yery poodble that (^) may deyeloped in Sanscrit. 



Chap. m. H- P|B. T^D. C,K^Q. G. CH. J. 205 

nounce (k) and (j) simultaneously. Hence the back of the 
ton^e still remaining in contact with the palate, the middle 
of uie tongue is also raised, so that both back and middle 
lie against the palate. This is rather a constrained position, 
and consequently the back of the tongue readily drops. 
The result is the exact position for (tj) which, originating in 
an attempt to sound (t) and (j) simultaneously, brought the 
tip and middle of the tongue to the palate, and this being 
almost an impossible position dropped the tip. The two 
consonants (kj, tj) are therefore ready to interchange. The 
passage from (tj) to (tsh) is yery short and swift, so much so 
that many writers, as Wallis, have considered (tsh) to be 
really (tj).^ But the organs of different speakers have dif- 
ferent tendencies, and in some (s) or (sh) are more readily 
evolved than (tsh) from (tj). It must be remembered that 
when the sound is thus spoken of as changing, it is not meant 
that it changes in the mouth of a siagle man from perfect (k) 
to perfect (tsh). Quite the contrary. It probably required 
many generations to complete the change, and the transi- 
tional forms were possibly in use by intermediate genera- 
tions. From these must be excluded aU intentional, that is, 
artificial inorganic changes, such as those induced by modem 
orthoepists. The (s, sh, tsh) were all imperfect attempts at 
imitating (tj), a sound which is said to have remained stable 
in the Hungarian language where it is written ty, while its 
congener (dj) is written gy^ Magyar being called (Modj-ar). 

Tne reason why (k) should have been palatalized to (kj) 
after (l,n,r) is not so clear, but the example of the modem 
high (German milch^ manch, dutch (miUh, manA;h, durA;h) 
shews that the tendency is a reality not an hypothesis, and 
enables us to imderstand milch as well as milk ags. milcy 
meolc ; bench as weU as bank, ags. banc ; drench ags. drencan 
as well as drink ags. drincan, stark and starch ags. stearc, 
mark and march a border, ags. mearc. Chaucer interchanges 
toerk, werch, etc,, to suit his rhyme. It would seem there- 
fore that about this time there was a great tendency in the 
two sounds to fall into one another. The close connection 
also of the sounds of (k, tsh) naturally suggested the related 
signs c, ch, a notation early adopted. And as (sk) became 

* Wallis says : " Anglomm eh Tel tyan-eyer, at si pneposaerit /, d for- 

teh sonat /y ... Si Toci Anglicanae yew mabit Angloram ekanger, hoc est, tyan» 

taxus sigillatim pneponantur </, ty <, z dyer.*' There is no doubt of the 

fiunt dyewy tyewt eyew, tyew^ hoc est, readiness with which the first sounds 

Anglorum Jew Judseus, chew mastico, generate the second, but the two are 

thew ostendo, et Oallorum jeu lusus. quite distinct, and a very little practice 

Qui syUabis yan^ yer prseposuerit «, s enables any one to distinguish tnem. 
fofinaDit Oallorum changer, hoc est, 



206 P,B. T,D. C^KyQ. G. CH. J. Chap. IIL § 4. 

(skj, Bt], eh), the earliest ngn for the new sound was sch. 
This has been adopted in (^rman where ch by itself has a 
different meaning. See also Chap. Y, § 4, No. 1. 

Bat the phenomenon which suggested these remarks, 
namely, the palatalisation of (k) before an (a) sound, is dif- 
fer^it. Generally the consonant follows the tendency of the 
▼oweL A German is so imbued with the tendency of ch to 
become (kh, kjh, ktch) according to the preceding yowel^ 
so used to say (akh, ikjh» auktrh), that lus organs would find 
(akjh, ikh) an impossibility. But different speakers seem to 
have been affected with the yery opposite tendency; some 
striving to render the consonant thinner, or more palatal, by 
inserting an (i) effect, between it and a following (a) sound ; 
others avoiding the palatalisation of a consonant before an 
(i) sound by the introduction of an (u) sound. The first 
would convert (ka) into (kia), whence Qua, kja), the common 
Italian schiacciato (skjatt8haa*to) effect; the second change 
(ki, ke) into (km, ktce) or (kwi, kwe). These tendencies are 
carried far beyond these limits in the Sclavonic palatalisation 
and the French labialisation of consonants. They are not 
widely developed in our own language, and, beine inor- 

ric, may prevail only partially both in time and place, 
modem Italian both chi and cut (ki, cuui) occur, the 
French qui though written with the mark of thickening or 
labialisation, is palatalised into (kji) and similarly in all 
words where qu precedes a (i, e) sound in French. 

As respects the particular usage, (A-aJt, A^ind, sXai'let, 
sA»i; ^aid, ^id) for cart, kind, scarlet, sky; guard, guide, it 
is now antiquated in English. But in Walker's time it was 
so much the custom that he found it '^ impossible '' to pro- 
nounce garrison and carriage with the pure (g, k), without 
any inserted (i) sound. I have however not been able to 
find any allusion to this practice in the older writers. The 
custom is now dying rapidly out. But we find the same 
tendency in other languages. Thus in Modem Greek, I 
have b^n told, that % is always (Jkh) even before q>, a, and 
it seems that the Sanscrit i| had the same sound. 

What has been said of k applies directly to g, substituting 
sonants for mutes, and as {k) produced (tsh), so did (g) pro- 
duce fdzh). The Anglosaxon g has however usually re- 
mainea (g), and even in several cases, as edge, bridge in which 
the change to (dzh) has been made, the (g) is found as a 
dialectic form. The alteration of the Anglosaxon g has 
generally taken other directions, which will be considered 
under gh. 



Chap. in. § 4. p,B. T,D. C^K^Q. G. CH. J. 207 

CH and J, G are also (tsh, dzk) when oorresponding to 
the present French sounds (sh, zh). Palsgraye admits that 
French ch is English (sh), but he makes the French and 
English y identical. It is not easy to determine whether in 
Tery old French eh,j were read (tsh, dzh) or (sh, zh). Hart 
makes eight pairs of consonants (b Pi y f> g k^ dzh tsh, d t, 
dh th, z s) and two breaths (sh h'). The letters here trans- 
cribed (dzh, tsh), he identifies with Italian (gi, ci) and the last 
with the " High Dutch'' tseh^ by which their sounds are deter- 
mined. Then he sajrs, translating his phonetic orthograj^y, 

" The French do use the / consonant in a sound which we use 
not in our speech, whereof this (sh) seryeth for the sister thereof, 
with us, as cA doth with them, haying no inward sound, and are 
botL firamed with keeping of the tongue from the palate and bring- 
ing the teeth together, or the one or other lip to lus counter teeth, 
and thrusting the breath through them with the inward sound for 
the French j consonant ; which if we had in use, should make us 
the eighth pair. For want whereof the (sh) doth remain to us, a 
brea^ without fellow, which the other seven pairs haye. But for want 
of that sound, we haye four others which the French neyer use, to 
wit of (dzh, tsh) and (dh, th) which are yery hard for any natural 
French to pronounce : other than such as are brought up amongst 
us somwhat in youth." And again in the theoretical part of his 
work, after an elaborate description of (sh) he adds : '' For the 
felowe of which sh, the French do sounde their g, before e, and i, 
and the L consonant before a, o, and u, and sometimes before e, 
and doe neuer soimd perfitely our sounds beforesaid for (dzh) & 
(tsh), in all their speach." 

Hence the French j is fixed as the yoiced form of (sh), 
that is (zh), as Hart heard it in 1569. Yet Palsgraye, 
whose ear was unfortunately by no means delicate, ccmfiised 
(zh) with (dzh). The Welsn haye no (sh, zh, tsh, dzh), and 
are forced to transcribe the two first by st and the two last 
by t9i, while they sometimes use si for all four. Thus Sales- 
bury transcribes Jesu, John, joynt by tsieauto, tsion^ tsiof/nt, 
and makes a Jack ape into a (siak ab) in his dictionary. 
He admits that the Welsh tai is as like the English (tan) 
** as brass is to gold,'' and says of the English " ch, o and i" 
(tsh, dzh), that there is " the same likeness between these 
three English letters as exists between pewter and silyer, 
that at first sight they appear yery like each other, but on 
dose examination they differ." 

The letters ch when transcribing the Oreek x ^^ called 
(k), and in the word ache which the Promptorium also writes 
ake, ch has generally the sound of (k). But Hart says: 
''We abuse the name of h, calling it ache, which sounde 



208 P^B. T,I>. C,K,Q. O. CH. J. Chap. III. § 4. 

semetli very well to expresse an headache or some bone ache," 
80 that as the name of the letter could only have been 
(aatsh)y the words imply that ache was also so pronounced. 
Bullokar also notes it as (aatsh), and thus, by the very same 
collocation bone ache, is confirmed a &ncy of John Kemble's, 
in pronounciDg the line (Tempest, act i, sc. 2, y. 370) : 

Fill all thy bones with Aches, make thee rore. 
It is true Kemble said (^d»h'ez), and therefore erred in 
the vowel, though right in the consonant ; and the feeling of 
the O. P. rioters in placarding, "Silence! Mr. KemUe's 
head aitshes,'' was in so far correct, that it was absurd to 
retain a single antique pronunciation in the midst of his 
modem sounds. 

The initial k according to all the authorities was still 
heard in the xvith century before n, as (knoou, knot, 
knuk*l) and hence probably mitial ffn was (gn), as both are 
used in present German knochen, gnade (knokh*ni, gnoa'de), 
but I have not met with an instance of gn. Jones mokes 
initial gn always (n), but says that initial kn "may be 
sounded An," which was therefore imusual at that time. 
Wallis however fifty years before allowed (knou, knyy) 
know, knew, and Cooper, strangely enough says : " Kn 
sonatur ut hn ; knave nebulo .... quasi hnave &o.,'* meaning 
(nh), but perhaps really simple (n), the aspiration being a 
theoretical difference to distinguish initial kn from simple n. 

Labialised / or (Iw) has alr^idy been shewn to have existed 
in our language, (p. 193,) but it has died out. liabialised k or 
(ktc), the hps being opened simultaneously with the release 
of the k contact and not after it, is an ancient element of our 
own and probably of many other languages. In Anglosaxon 
it is written cw, in Latin qu, which is the form adopted in 
English. It is needless to say that no orthoepist has dis- 
tinguished (kw, ktr). Ou properly bears the same relation 
to ^ as gu to A;, but as the form of the^ remained imchanged, 
litde attention was paid to it. It does not exist as part of the 
Saxon element of our language. Initially it is generally 
used superfluously for g. Occasionally it has the sound (gtr) 
as in language, itself a modem form, anguish, distinguish, &c. 
Usage, however, varies, some saying (laeq'gtfydzh, asq'guniBh) 
and others (laeq'wydzh, aeq'wwh). The ItSian guaie, guanto 
are apparently (ktruaa'le, gtruan'to). The final -gue for ^g 
as in tongue, plague is quite a modernism. Ague, also spelled 
agwe in the Promptorium, was probably (aa'gyy) or (aa'guu) 
from aigue, and hence does not belong to this category. 

As we have (kj gj, ku? gu^), so also to our unacmowledged 



Chap. III. § 4* GH — XVI TH TO XVIH TH CENTURY. 209 

(tj dj) correspond an equally unacknowledged (tto dw,) 
wnicli, written tw dw as in between^ twam, twang, twist, twelve, 
twirl; dwindle, dwell, dwarf, have been generally considered 
as (tw, dw), but many of those who have thought on pho- 
netics have been more perplexed to decide whether w was 
here really a vowel (u) or a consonant (w), than in the cor- 
responding words wean, wain, wist, well, war. The difficulty 
is resolved by observing that the opening of the lips is really 
simultaneous with the release of the (t, d) contact. 

The termination -age is represented as having the soimd 
(-aidzh) in Salesbury, in domaae, heritage, language, all 
French words, and this a^ees with Palsgrave, suprd, p. 120, 
note. Smith, Bullokar, Gill, and Butler, however, do not re- 
cognize this tendency in English, although Butler notes the 
similar change of (a) to (ai) before -nge (-ndzh), and both 
are confirmea by the modem sounds (-^dzh, -^endzh), of 
which the first is a degeneration of (-edzh, -e^h). 

GH 

The Anglosaxon alphabet having no especial letter to repre- 
sent the guttural (kh), the single letter h was used, as in old 
High German the double letter hh was employed. As g often 
interchanges with h in Anglosaxon, as lagu, lah, law, it is pos- 
sible that there was a tendency in those times to pronounce 
g final or medial as (gh), just as the Upper Germans now do, 
and as the Dutch pronounce their ^ in all positions. At a 
later period the Anglosaxon g seems to have become (^h) and 
then (j), sounds even now confused by German phoneticians. 
Hence ;, which was also written }, and occasionally printed 
%, became the regular sign for (j) till it was supplanted by y. 
When, therefore, it was desirable to shew that g retained the 
soimd of h, that is, (kh), it was natural to write gh in its 
place. In the Orrmulum we have all varieties; fulluhht 
/fohhtesat, mihhte are instances of h, doubled merely to shew 
that the preceding vowel is short ; ma}}, e}}whcer, a}}, twi}}e88 
illustrate the use of }, doubtful whether (^h) or (j), while 
re}heU'boc, foll}henn shew the use of }A. As in Dutch the g 
often sounds (kh) as well as (gh),^ and as the Scotch adopted 
the orthography ch, it seems probable that (gh) early ac- 

1 Recent opportunities of hearing (^''M^) ^^^ than (skrh^, Bkh^p). 

Dutch pronunciation hare oonyincea The Dutch themselves consider the 

me that the Dutch eh^ g are rather sound yerv soft. The Dutch final and 

(krh, grh) than simple (kh, gh). But medial ten is pronounced as simple «, 

the sounds are so ughtly and gentlj thus vleeseh (yiees), a modem example 

pronounced that thej rather resemble of an omitted guttural, 
(rh, r) than (krh, grh), thus aehip^ 

14 



210 GH — XVI TH TO XVniTH CENTURY. Chap. III. { 4. 

quired the sound of (kh) only. But it is by no means 
certain. The two sounds (kh, gh) are so easily confused by 
those not familiar with mem^ and may so readily inter- 
change owing to the nature of the adjoining consonant, and 
so few languages have provided for their discrimination, that 
we cannot be certain of their not having both existed even 
though onlv one is named. It is the same with (sh, zh), the 
latter of which is scarcely ever noticed, so that it is not easy 
to say when it first came into use. Even (s, z) are constantly 
confused. They both exist in Italian, and have only one 
sign 8. But only one of them (s) exists in Spanish and 
Welsh, having the same sign 8. Hence it is impossible to 
tell from the orthography ffh whether it represented onlv 
(kh), only (gh), or occasionally (kh) and (gh), nor would it 
be certain u a Welsh writer, for example, who onlv knew 
(kh) and was not acquainted with (gh), asserted that the 
English ffh was (kh). Now Salesbury says : *^Gh has the 
same sound as our ch (kh), except that we sound ch deeper 
in the throat and more harshly." The two expressions 
" deeper " and ** more harshly ** might be applied m Sales- 
bury's popular language in two ways. For example, (kh) is 
deeper than (A:h) and harsher. Ana (kh) being called 'hard' 
in contrast to (gh) 'soft,' fkh) mifi;ht be esteemed harsher 
than (gh) ; or the reverse, wnen (kh) is a familiar and (gh) a 
strange sound. But certainly (kh) would be felt to be much 
deeper and harsher than (ffh). There is another supposition, 
namely, that ffh was merely (h*), the simple jerk of the 
aspirated breath. In most cases (h, h') are confused, and 
the aspirate is considered to be (h*). In my own opinion 
(h*) is much less frequent than (h), but (h*) is occasionally 
said when only (h) is intended. Sir T. Smith writes h for 
either sound, and this is the general custom of orthoepists. 
He also represents ffh by h only, saying : 

" Scio tauhtf nihtf fiht & csetera ejusmodi scribi etiam ff adjuncta, 
vt taught^ nighty fght, sed sonum ilHus g quaerant, quibus ita libet 
scribere, aures profecto mess nunquam in illis vocibus sonitum rov 
g poteraat haurire." 

This ought to imply that the sound was (h') and that 
(tauH't, nin't, fin't) was at that time the pronimciation of 
tauffht, niffht, fiyht Hart at the same time writes lauht^ oht 
= (lauH't, oH't) for lauffht, ouffht Bullokar has also (Kht^ 
bowht s= (ltH*tboouH*t). But then Gill finds it necessary to 
introduce a new sign, namely, h with its stem crossed like a 
if to represent the sound of ffh in bouffht^ and says : 

''X. ch. Graecorom in initio nunquam vsurpamus, in medio, et 



Chap. HI. } 4. GH — XVI TH TO XVIIITH CENTURY. 211 

fine saepe ; et per gh, male exprimimxis : posthac sic (kh)' scribemus : 
Yt in (warkht eni^) weight enough satis pondeiis." 

Now those who do not possess a symbol for (kh) often 
write h for it, as we have seen in Anglosaxon finals, and as 
Bapp considers to have been the case in the Anglosaxon 
initial M^ hr, hw, which I rather suppose to have been (Ih, 
rh, wh). The sound of (khw) is very harsh, and in Scotland 
and North Wales it is modified into (ktrh), corresponding to 
the English and South Welsh (wh). Those who wish to 
acquire the sound of (akh) may be led to it by endeavouring 
to say (au'), and at the same time slightly raising the back 
of the tongue. Hence it is possible that Salesbury's eh, 
(which is not so "deep" and "harsh" as the Welsh ch,) 
Smith's, Hart's, Bullokar's h, and Gill's Xf ™^y ^ ^ ^^® 
and the same sound, either (h') or (kh). But it is certain 
that when Gill wrote, the sound (kh) was disappearing in 
the south of England, for Butler, who uses a g wim a 
crossed stem, to represent gh, says that "the Northern 
Dialect doth yet rightly sound " it, implying of course that 
it had gone out in tne South by 1633. 

The safest conclusion seems to be that the sound in the 
XVI th century was really (kh), but was generally pronounced 
very lightly ;* it might, however, have been (Arn) after (i,e). 
This is still Uie custom in Scotland. 

By the middle of the xvn th century the rule had become 
to omit the sound, after changing the preceding vowel, or to 
change it into some other sibilant, generally (f ), in one or 
two cases provincially (th). Wallis, 1653, after noticing 
that initial ffh is simply (g), adds : 

" alias vero nunc dierum prorsus omittitur ; syllabam tamen pro- 
ducendam innuit. A quibusdam tamen (prsBsertim Septentrion- 
alibus) per moUiorem saltern aspirationem h effertor, ut might 
potestas, light lux, night nox, right rectus, sight yisus, sigh singul- 
tus, weigh pondero, weight pondus, though quamvis, thought cogi- 
tatio, wrdught operatus est, hrdught attulit, taught docuit, sought 
quaesiyit, fraught refertus, nattght nihil, naught malus, &c. In 
paucis YocabuHs effertur plerumque per ff; nempe eough tnssis, 

^ Gill misprints 1$, wbich be uses for bour voealur nebonr ; neigb abreniated 

(db) and in Ids errata endeayonrine to ne." Tbis seems to sbew that both 

eorrect this tniafaVft and also (iniULb) (neekh) and (nee) were heard in the 

for (ennkh), he has accidentally re- first syllable of tnis word, and would 

peated the error instead of making the imply that (neekh) was rather pedantic 

correction, as has been done here in Indeed if it were to be obissed witii the 

tiie text. other pronunciations which the pedant 

> The Pedant in Lore's Labour Lost, recommends, as (doubt, debt, kail, Htdf) 

Act ▼. So. 1. 1623 comedies p. 136 it might be oonndeied as obsolete. 
C(»nplainfl of the pronunciation '* neigh- 



212 GH — XVI TH TO XVmXH CKHTUEY. Chap. III. { 4. 

trough alveolus, tmigh tenax, rough asper, Umgh rideo profenmtar, 
ceffi trdffi tuff^ ruffy laff, Inough (smgulare) sat multum, sonatur 
inuff: at inough (plurale) sat multa, sonatur enowJ^ 

WiLKiNS, 1668, after 8a3ring that ffh might have been 
(gh) adds : 'Hhis kind of sound is now by disuse lost among 
us." Price, however, in the same year, says : " Gh sounds 
now like h in Almighty, althauffh" etc., adding in the margin 
" but the Ancients did, as the Welch & Scots do still pro- 
nounce ffh thorow the throat." He notes that ffh sounds as 
(f ) in couffh, laughter^ enough, rouffh. Cooper, 1685, says : 
** hodid apud nos desuevit pronunciatio ffh, retinetur tamen 
in scriptur&," but he makes it (f ) in couffh, laugh, rough, 
iouah, trough, and makes Wallis's distinction between enough 
and enow, Miege, 1688, says also that gh is generally mute, 
but is (f ) in lauffh, drauffht, rough, tough, enough (not distin- 
guishing enow,) but adds " sigh, un Soupir, et le Verbe to 
Sigh soupirer, ont un son particulier qui approche fort de 
celui du th en Anglois." Jones, 1701, extends both the (f ) 
and the (th) list. According to him (f ) is heard regularly 
in draught, draughts, laugh, cough, enough, hough, rough, 
lough, trough; and he adds ''some also sound daughter, 
bought, nought, taught, &c., as with an f, saying daufter, bqft, 
&c." And he states, that gh, ght are th '' in sigh, sounded 
sith ; in drought, height sounded drouth, heith,** but in other 
parts of his book he also admits the sounds (soi, drAAt, Heet). 
In the xvin th century we may notice that Fielding in his 
Tom Jones, book vii, chap. 13, makes his landlady say oft, 
thqft, for ought, thought, and Mrs. Honour write soft for 
sought, book xv, chap. 10. These are meant to be West of 
England vulgarisms, but they sufficiently shew the tendency. 

It would be vain to consider the changes thus indicated, 
without proceeding at once to the fountain head. In Anglo- 
saxon itself g became A before t very firequently, and was 
often omitted. Let us therefore consider the sound as some- 
times (kh, gh) and sometimes (kh., gh). Let these sounds be 
kept as widely apart as possible. Then (gh) must be rounded, 
that is, there must be a rounding of the ups while the gut- 
tural is uttered, producing (kt<^h, gt^h), thus German auch, 
auge are, as already mentioned, in reality (avkwh, SLUgwh'e), 
The Scotch sough is (suukti^h), and ffenerally the juu) sound 
before (Ui) has a tendency to produce (kirh). This would 
then have a natural tendency towards (wh, w). On the 
other hand (kjh, gjh) are in themselves the closest allies of 
(jh, j). Hence an effort to keep the two sounds of {sh, gih) 
weU apart would result in producing (w, j), which, i3ter 



Chap. HI. § 4. gH — XVI TH TO XVIII TH CENTURY. 213 

vowels, would diphthongise as (u, i), and after consonants 
would form the syllables (u, t). Now this is precisely what 
has happened in the passage from Anglosaxon mto English. 

First the (u) change. From lagu^ lah comes law (laau, 
Iaa) ; from draff an comes draw (draau, drAA) ; from boga comes 
first bough (boouktrh) and then bow (boon) or (bovikwh, bou, 
beu). From halffian comes hallow (Hal*u, Haloou, Hael'o) 
from tcdff comes (tal'u, tal'oou, tael'o). In JEdinburffh, Mussel' 
burffh, etc., although gh is written, (o) is regularly sounded. 

Next the (i) cnange. From tccBffn comes wain (wain, 
wtfdn) ; from /(Bffer comes fair (fair, feei), from r^gn comes 
rain (rain, r^ein). From bieiff come bulffe (buldzh, baldzh), 
bellows (bel'uz, bel'oouz), and belly (bel'«), shewing three 
changes of ff. 

If instead of falling to (u), the (k«rh), remained at (wh), 
this would after a vowel rapidiy become (f). In Aberdeen- 
shire (f ) is the regular substitute for (wh) or rather the 
Scotch quh, which looks like an attempt to write (ktt?h) imder 
the form of (kwh). Dwarf from dweorh is an instructive 
example. The old English forms dwerghe^ durwe and the 
dialectic durgan are found; a dialectic Swedish dwetf and 
Dutch dwarf dorfare said to exist (E. Mueller, Etym. Wort, 
d. Eng. Spr., i. 327). The Dutch agler, kragt and English 
after, craft, Anglosaxon (efler, crceft, are examples of the 
correspondence of (f ) and (gh) in different forms of the same 
low German word. The chief English examples have been 
already cited, and it has been shewn that the chancre pre- 
vails dialectically much further than it has been admitted 
into the received forms of speech. Some words have even in 
English both forms, as hough (naf, Hok), trough (trof, troou), 
slough of a snake (slaf), slough a quagmire (slau), tough (taf, 
tooviYenough (enof*, eneu*), the grammatical distinction made 
by Wallis and Cooper that the first is singular, sat multum 
and the second plural sat multa, although conformable to 
Scotch usage, does not seem to be historically justified. 

The change of gh into (p) in hiccough (Hik'cm) is mentioned 
by Jones 1701, and must be considered to be of the same 
nature as the change to (f), as (wh, w, p) are even more 
closely related than (wh, f ). The curious but not admitted 
change to (th) seems to rest merely on the conftision of the 
(f, th) hisses.^ When these are pronoimced without any 
vowel it is venr difficult to distinguish them at a little dis- 
tance, as is well known to those who teach to spell by means 
of the powers of the letters. 

^ Sigh, which Jones and Miege giTe as (sdith) is called (saif) in Deyonshire. 



214 g, c ; z. SH. X — XVI to xvni centuky. Chap, HI. { 4. 

When^A falls into (u) it naturally alters the preceding 
Towely with which it diphthongises, hence (a) becomes (au, 
aau, aa). Similarly (o) should become (ou) and thence (ou), 
but in this case the tendency has been rather to (ou, oo, aa), 
as in ought^ btmgM, etc. When gh falls into (i) we have 
alterations in the ot^er direction, as (ai, eei^ ee). 

After the vowel (i), the (i) change of gh, which is the only 
natural one that could be expected, would simply prolong the 
(i), and hence, from hih, niht we might have (mi, niit), 
forms which really exist dialectically for high, night; and 
from the termination -ig we might expect (-t), the com- 
monest form in present use. 

We shall see in the next chapter that such were probably 
the original forms of transition. In Cumberland and West- 
moreland igh is regxdarly replaced by (ii), and the chance to 
(ei), which is constantly attributed to the omission of the 
guttural, seems to have no real connection with it, but forms 
part of the general change of long i from (it) through (ei) to 
(oi), which wiU be minutely considered in Chap. IV, § 2, 
imder I. If we are to trust Gill, the sound of (ei) and the 
guttural coexisted, as he always prints (naikht) and neither 
(nikht), the pronunciation of Salesbury, nor (noit) as became 
prevalent during the xvii th century. 

With this gh proper must not be confounded gh written 
for g, in comparatively recent times, at the beginning of 
words. Jones tells ub that the soimd of ^ is written gh in 
gherkin, ghess, gheus, ghittem, ghost, where gheas is foimd in 
Spenser for guess. 

S, C;Z. SH. X. 

The use of c for (s) follows the same rules as at present, 
throughout the period under consideration. The letter s seems 
also to have been (s) or (z) under the same circumstances as 
at present, but as the sound of (z) doep not exist in Welsh, 
Salesbury had no means of indicating it by Welsh letters, 
and he therefore writes « in all cases, although he names the 
% sound. Smith, Hart, and Gill all use £, but none of them 
are sufficiently- careful. Still there can be no reasonable 
doubt that s was pronounced (z) under the same circum- 
stances as it is at present. The letter s is now used for (sh), 
where the change has been generated by a subsequent (i) 
sound, and the same remark applies to o, ^, as in mission, 
pressure, special, motion; and s passes in certain cases into 

gi) under similar circimistances, as vision, excision, measure. 
ere is no trace of this in the xvi th century, Salesbury 



Chap. HI. J 4- 8, C ; Z. 8H. X — XVI TO XVIH CENTURY. 216 

has GRACTOTJSE, grositcs (g^raa*8i,us), gondicton, condisyum 
(kondis'twn), exhibition embisiwn (eksibis'iun), prohibition 
proibisiwn (proo,ibis'i,iin^, trbsurb tremtor (tree'zyyr). 
Bt7Llokar has (abrevias'ion^ komposiz*ion, naa'sion, syyor, 
svygar) for abbreviation, campositian, nation, auer, mgar. 
And Gill writes (ekspektas'ion, Habitaa'sion, naa-sion, 
okaa'zion, pas'ion) for eocpectation, habitation, nation, occasion, 
passion. In the xvii th century Wallis generates (sh) from 
(sj), but WiLKiNS writes (reserreksion) for resurrection. 
Price, 1668, only recognizes ** hard 8 in passion ; soft s in 
concision, and sh in cushion, fashion/* Cooper, 1685, does not 
name the use of (sh) in such cases, but admits shure, shuffar, 
which may have been, (shuur shog'er), "facilitatis causA," 
although he places such words immediately after his '^ vitanda 
barbara dialectus.'' Miege, 1688, writes chiire, pennchoun in 
French letters for sure, pension, states that in the termination 
'ision, s sounds as French ff or j Tzh) and writes (tjuai, train- 
gient, lijeur, ^'er, hdjer, crdjer lor usual, transient, leisure, 
osier, hosier, crosier. Jones, 1701, says : " Tho' you have the 
Sound of sh ver^ often in the Beginning of the last Syllable 
of Words, as m action, nation, &c. sounded, acshon, nashon, 
&c. yet is sh never written there in Words of two or more 
Syllables ; except in cushion, fashion, hogshead^ lushious. 
Marshal.'* He admits that s is commonly sounded sh (sh) 
in assume, assure, assurance, censure, consume, desume, ensue, 
ensure, fissure, issue, leisure, measure, pleasure, pressure, pursue, 
pursuer, pursuit, sue, suet, sugar, suit, sure, sute, tissue, 
treasure, and says that ocean is " sounded oshan." He does 
not recognize (zh), but says that sh is written z '4n azure, 
sounded ashure.'* The change was therefore fully estab- 
lished at the end of the xvn th century. 

Though the orthoepists of the xvii th century were slow 
to recognize this change, and those of the xvni th and xix th 
evcD acbnit it rather grudgingly, while those of the xvi th 
do not seem to be even aware of such a *' slovenly habit," 
yet we have at least two early traces of the degeneration of 
suit into shoot, in Shakspere and in Rowley, for a notice of 
which I am indebted to Mr. Aldis Wright. In Love's 
Labour Lost, Act vs. Sc. 1, written before 1598, the folio 1623, 
Comedies, p. 130,^ there is apparently a play on suitor and 

^ ^ Qu, Wlio gave thee this Letter P Clo. From my Lord Beroume, a good 

Clo. 1 told you, my Lord. master of mine. 

Qu. Towhomshomd'stthougiueitP To a Lady of France^ that he called 
Clo. From my Lord to my Ltuly. £o$aline, 

Qu. From which Lord, to which Qtt. Thou hast mistaken his letter. 

Lady f Come Lords away. 



216 8, C ; Z. SH. * X — XVI to XVin century. Chap. HI. § 4. 



shooter^ deer and dear. The two latter words were pronounced 
alike by Smith. Were the two former really pronounced alike 
by Shakspere, as they were by Jones, 1701, and Buchanan, 
1766, though Cooper, 1685, gives (stut) and Sheridan, 1780, 
(sunt) for suit? Gill, 1621, only allows (syyt), Bullokar, 
1680, has (sj^'gar). Hart has (syv'er).^ But some persons 
must have said (shuut), or such jokes would have been lost, 
and, whatever was the case in Shakspere,* we have this pun 
in Eowley's Match in the Dark, 1633, Act ii. Sc. 1 : 

Moll. Out upon him, what a miter have I got. I am Bony you 
are so bad an Archer, sir. 

JSare, Why Bird, why Bird? 

Moll, Why to sitoote at Buts, when you should use prick-shafts. 

In the present day we have a joke of an Irish shopman 
telling his customer to shoot himself, meaning suit himself. 



Here sweete, pot yp this, 'twill he 
thine another day. 

£x(mnt. 

Boy, WTio is the shooter P Who is 
the shooter P 

Rosa. Shall I teach yon to know. 

Boy, I my continent of heautie. 

Boia. Why she that beares the Bow. 
Finely pnt off. 

^oy.Jiiy Lady goes to kill homes. ••• 

Bom, Well then, I am the shooter. 

Boy. And who is yonr DeareP' 

In Boyet's first speech, Steerens, at 
the suggestion of warmer, altered the 
•hooter of all the quartos and folios, to 
tuitor^ which is tne reading usually 
adopted. The preceding dialogue, 
which has been given for the purpose 
of comparison, seems at first sight to 
point to suitor as Boyet's meaning, 
which Bosaline perversely takes as 
shooter. But the connection is not 
evident. There is no allusion to euitor, 
but much to shooter in what follows. 
Boyei knew both the suitor (whether we 
take him as Biron or Armado), and the 
shooter (the Princess, apparently, who 
is represented as going to shoot a deer 
at the opening of the scene), but 
Bosaline's reply, and her remark that 
it is a "put off," look as if she was 
purposely misunderstanding him. In 
the absence of a tenable hypothesis 
for the introduction of the new word, 
suitor, we may suppose that Boyet, 
looking off after the shooting party 
which has just left, sees an arrow sped, 
and inquires of Bosaline who shot 
it, whereupon she puts him off with 



the truism that it was she (one of the 
Princess's company) who bore the bow. 

^ John Halt, in his first treatise, as 
cited in Chap. VIII, { 3, note I, classes 
the three words ** suer, shut, and bruer," 
as he spells them, together, and pro- 
nounces (syy-er, shyyt, bryyer). The 
first may lie suer or sewer, the last is, 
of course brewery is the second suit, 
or shoot intended to be written shute 
(Scotch, schute s shoot), as Hart in 
that treatise constantly omits the 
final e? It is the only indication of 
. such a change in the xvi th century, 
and the word suer renders it very 
doubtfuL We can hardly suppose the 
word to have been shut, Stratman 
gives the old Bnglish forms for shut, 
schutten, schitte, schettin, shette; for 
shoot, sceoten, schetin, sheten, scheete, 
ssete, schete, scuten, soten, shoten, 
schoten. The original difference of the 
words is difficult to determine; £tt- 
miiUer does not give any ags. word 
seyttan, to shut, as different from 
sceotan, to shoot; E. Miiller refers 
shut to shoot from shooting the bolt 
of the door. 

' Steevens quotes an equivoque of 
suters and shooters, miscalled archers 
by a servant, from "The Puritan, 
1607,*' and Malone a similar play upon 
archers and suitors in "£^ys and 
Characters of a Prison and Prisoners, 
by G. M., 1618,** and also Antony and 
Cleopatra Act v. Sc. 2, where Pope reads 
" a gfrief that shoots My very heart at 
root," and Capell reads smites for the 
folio, 1623, suites. 



Chap. III. { 4. 



Z. 8H. X — XVI TO XVIII CBNTURY. 217 



The Irish pronunciation however only shews an English 
pronunciation of the xviith century. In England at the 
present day, shoot for mit would be vulgar, but the joke 
would be readily understood, though few persons use, or 
have even heard, the pronunciation. Might not this have 
been the case in Shakspere's time P At any rate there is no 
authority for supposing that such a pronunciation could have 
been used seriously by Shakspere himself.^ But the sound 

two different editions, but between an 
uncorrected cop^ and a corrected copy 
of the sufne edition. The later qnartos 
foUow the corrected copy but their tes- 
timony is of no value, because their 
reading is merely a reprint." Hurried 
corrections, whether of print or manu- 
script, frequently introouce additional 
errors, and hence there is no guarantee 
in thiis curious history that the com- 
positor who substituted ehewted for 
tnytedy did not himself put ehewted 
when he meant to have inserted eewted. 
More instances are certainly required 
to decide the point. The Scotch 
wrote eehute for ehoot. Palsgrave 
writes euU for suit. In Henry V., 
Act iii. sc. 6, fol. 1623, p. 81, we find 
" what a beard of the Generalls Cut. 
and a horride Sute of the Campe, will 
doe among fomins^ Bottles and Ale- 
washt Wits, \a wonaerfull to be thought 
on." In the Chronicle History of 
Henry the Fifth, printed in the fourth 
vol. of the Cambridge edition ehout 
stands for sute. If we take Bullokar's 
old pronunciation, shout would be 
(shunt). Mr. Aldis Wright observes 
that tnis was **an instance of a play 
apparently taken down at the time of 
acting, and whether shout or suit be 
the true reading, one of them could not 
have been substituted for the other 
unless the pronunciation was some- 
thing similar," and he thinks that 
these instances lead to the conclusion 
that the pronunciation (shuutj "was 
in existence at the beginning of the 
XTTith century. The jokes upon 
shooter and suitor certainly establish 
that a sufficiently similar pronunciation 
of the words was in existence to make 
the joke appreciable. The various 
spellings, I fear, prove nothing, be- 
cause, considering tne frei^uency of the 
word — suit occurs 163 times, suitable 
once, suited T, suiting 1, suitor 38 times 
in Mrs. Cowden Clarke's Concordance, 
— ^the rare variations can only pass for 



^ Mr. Aldis Wright seems to sup- 
pose that the compositors might have 
had that pronunciation, and that it 
therefore might have crept into the 
text. In L^, Act ii. Sc. 2, the word 
three-suited of the fo. 1623, is spelled 
three shewted in all the quartos but 
one, where it is three snyted, an evident 
misprint for three suyted. Now shewted 
would probably have been written for 
(shyyted), and ma^r indicate the tran- 
sitional pronunciation ; on the other 
hand it may be itself a mere misprint 
for sewted, which would be a legiti- 
mate orthography for suited. This 
hypothesis is questioned by Mr. Aldis 
Wright, who says : " in books printed 
in the time of Shakespeare and Bacon 
variations occur in different copies of 
the same edition. I have never seen 
two copies of the 1625 edition of 
Bacon's Essays which were exactly 
alike. A list of the variations is given 
at the end of my edition. Now there 
are six copies or the quarto of King 
Lear printed in 1608. which we [Mr. 
W. 6. Clark and himself editors of 
Uie Cambridge Shakespeare] have in 
our notes erroneouslv (as we confess in 
the Preface) called Qs, whereas we are 
now convinced that this edition was 
earlier than the one in the same year 
which we have called Q< . These 
copies of Qa (so called) differ from 
each other in having some of them 
been corrected while passing through 
the press. The earliest of these which 
we nave met with is one of the two 
copies in the Bodleian, and we call it 
for distinction sake Qa (Bodl. 1). This 
has the reading three snyted: but cUl 
the other copies of the same edition 
read three shewted, I suppose therefore 
that while the edition was in course of 
printing the error was discovered, and 
the correction communicated verbally 
to the compositor, who inserted it accord- 
ing to his own notions of spelling. It is 
not a question between the reaoings of 



218 s, c ; z. 8H. X — XVI to xvni century. Chaf. m. i 4. 

may well have existed unrecognized, preciaely as the sound 
of (sh) is supposed to be unknown in Welsh, although cei9io 
is now generally called (kei'sho), and not (koi'sio). Simi- 
larly in Dutch (sh) has been developed firom (si-, (sj-) in 
several words, but it is not orthoepistically acknowledged. 
In the xviii th century there was a decided tendency towards 
(sh). Thus sue, suet, sugar, suicide, suit, suitable, suitor, 
sure, suture, all commence with (sh) in Buchanan, sue, suit, 
suitable, suitor, have (s) in Sheridan, but the rest have (sh), 
which Sheridan also uses in sudorific, sudorous, super-, 
superable, superb, superior, supernal, supine, supinity, supra-, 
supremacy, supreme, sural, where Buchanan has (s). 

The sound of (sh) was well known in the xvi th centuiy. 
Salesburt says : 

*' Sh when coining before a vowel is equivalent to this combina- 
tion »9%, thus SHAPPE ssiapp (shapp), shepe Mtip (shiip). Sh coming 
after a vowel is pronounced m«, thus asshe aisB (ash, aish?), wasshe 
wa%9$ (wash, waish ?). And wherever it is met with, it hisses Hke 
a roused serpent, not unlike the Hebrew letter called schin. And 
if you wish further information respecting this soimd, you should 
listen to the hissing voice of shellfish when they begin to boiL" 

We learn from Hart, supr4 p. 207, that (zh) was im- 
known in the xvi th century. Wilkins, 1668, says that (zh) 
is " facil and common .... amongst the French, who express 
it by e/, as in the word Jean, &c., and is easily imitable by 
us,'' implying that it was not in use in England. But 
Miege, 1688, being a Frenchman, heard it, as we have seen, 
p. 215, in the words where we now use it. He is the only 
writer in the xvii th centurv who notices it, and, as he is a 
foreigner, his testimony is suspicious. Franklin, 1768, 
seems only to know it in French, as he has no special sign 
for it, and even in French writes (zshaeme) for Jamais, 
Just as Hiirt writes (ozdzhuurdwi) for aujourdHhui, for want 
of an appropriate sign, although he had reco^ized the sound. 
Sheridan, 1780, fully acknowledges it. It is always written 
(s) or (z), and arises in English from palatisation as (z*j). 
In French it seems to be a degeneration of (dzh) formed 
from a palatalised (g*j); or else to have arisen from (j)S pre- 

muprinti. The absence of any notice > The Dutch at the Cape of Good 

of such a practice in orthoepists of the Hope say (dzhoo, Dzhan) etc., for (ja«, 

XTith century (if we except the yery Jan),ya, Jan, This is an alteration of 

doubtful passage from Hart in the last precisely the same character, and is 

note), tocher with the depreciating comparable with the Italian GiuffnOy 

manner in which similar usages are Giunonef Giuglio (Dzhuunjo, Dzhu- 

mentioned in Cooper, shew that any nMwh'ne, Dzhuu Ijo) from the Latin 

such pronunciation was considered not Junitun^ Jutumem^ Julium, 
worth mentioning. 



Chap. HI. § 4. P, V. TH — XVI TH TO XVIHTH CENTTTRY. 219 

oiBely in the same way as (sh) derives in some parts of 
Germany, and still more frequently to English ears, from 
(kh) as (ish) for (i*h). 

A was usually (ks). Salesbury gives flaxb Jflaca (flaks), 
EXHTBiTJOM embmum (ek8ibis'i,tm), oxB ocs (oks), but, ap- 
parently by a misprint, axe ags (agz). 

F, V 

2^ and v seem to have retained their sounds throughout, 
but in the earlier times v and u were interchangeable, and 
either could be used as a vowel or consonant. This was not 
the case in Welsh, where u was the vowel, and v the conso- 
nant. The consonant has been generally replaced by / in 
Welsh, jf being used for (f). Salesbury notices as a dialectic 
variety in " some countries of England" the use of (v) for 
(f), but he does not particularize the districts. Gill attri- 
butes it to East Anglia, " (v) pro (f), ut (vel'oou), pro 
(fel-oou)." 

TH 

The double sound of th as (th, dh) is fixed by Salesbury 
as the Welsh th^ dd, and the two uses were distinguished 
almost exactly as at present; mth seems however to have 
been always (wtth), though (wwih) is now more common. 
Salesbury gives (th) to through, thystle, thynne, wyth, thanke, 
thorowe, thyck ; and (dh) to this, thyne, the, that, thou. He 
also notices that th sounds (t) in Thomas, threasiire and throne, 
which he writes trwn (truun) ; and (d) in Thavies Inn. 
Smith, Hart, Bullokar, Gill, Butler, have all diflFerent signs 
for (th, dh) and use them according to our present custom 
of speech. Jones makes th = (t) m antheme, or anthymn, 
Anthony, apothecary, asthma, Author, authority, authorize, 
Catharine, Canthartdes, Esther, Isthmus, Lithuania, posthu* 
mus, priesthood, Thames, Thannet, thea, Thomas, Thomson, 
Thamasin, Thuscany, thyme. 

It is difficult to determine when these uses were settled. 
The two Anrfosaxon letters ]> ^ are usually taken to be 
(th, dh) but their emplovment is almost exactly opposite to 
modem use. In later Anglosaxon and Early English only 
one, either ^ or, more usually, ]> was employed, and even 
Omnin makes no distinction. This might have been a 
pecidiaritv in writing names. It seems safest to infer the 
old use nrom the modem, which is found to hold for the 
XVI th century. 



220 H — XVI TH TO XVIUTH CENTURY. Chap. III. § 4. 



The question concerning h is simply when was it mute P 
for its sound, or rather its action on the following vowel was 
always the same as (h) or (h*). Palsgrave says h is mute in 
honest, honour, habundaunce, habitation. Gill does not agree 
in the last word, and the h has now disappeared, even in 
writing, from the last but one. Salesbury says h is mute 
in honest, habitation, humble, habite, honeste^ honoure, exhi" 
bition, prohibition. Modem orthoepists will not admit the 
two last, though custom sanctions them, but habite and 
habitation have recovered their A, and himible is still 
doubtful. Gill adds the words hour, hyssop, which he 
writes (ai'zop). Abhominable was a common orthography in 
the XVI th century, and the h seems to have been occasion- 
ally pronoimced or not pronounced, for the Pedant in Love's 
Labour Lost (1623, Comedies, p. 136) says : " neighbour 
rocatur nebour ; neigh abreuiated ne ; this is abhominable, 
which he would call abhominable.*' ^ It is usual to print the 
second abhominable without the h and the first with it, but 
it seems more proper to reverse this, and write "this is 
abominable, which he would call abhominable," for the 
Pedant ought certainly to have known that there was no 
h in the Latin, although in the Latin of that time h was 
used, as we see from the Promptorium, 1450, " Abhominable, 
abhominabilis, abhominacyon abhominacio," and Levins 1570, 
" abhominate, abhominari," as if the words referred to ai- 
homine instead of ab-omine. 

In the XVII th century. Price 1668, says that h is mute in 
ghost, rhetorick, catarrh^ dunghill, host, hour, John, impos* 
thume, myrrh, Rhene, rheum, rhode, Wadham. Miege, a bad 
authority, because Frenchmen cannot rightly appreciate the 
English aspirate, having no such element of their own, de- 
clares that hour, hourly are the only two words in which 
h is mute, and especifdly instances honour as having an 
aspirated h. 

1701. Jones says h "may be soimded in halleluiah, harber- 
geon, habiliment, haver-du-pois,*' &c., but seems to imply that 
it is generally mute in these words, and says that 'ham in 
names of places in England is -am as in Broxham, Bucking' 
ham. He also makes h mute in cowherd, Nehemiah, shepherd, 
swine-herd, and in Heber, Hebraism, Hebrew, hecatomb, hectical. 
Hector, hedge, Hellen, hemorrhoids, herb, heriot, hermit, &c., 
" which h may be found by putting a Vowel before them." 

^ The quarto 1631 also prints abhonUndble in both places. 



Chap. III. § 4. H — XVI TH TO XVIII TH CENTURY. 221 

He allows unaccented his to lose the A, ''as in told his man, 
sounded told is man, &c." He says o is written ho " when 
it may be sounded ho, as in homage, holster, homo, in the 
beginning of all words, hosannah, host, hostage, hostess, 
hostler, hostile, houlet, hour, so^ho, inkhom, &c., often 
sounded as with o only.'* Also he says oo is written 
hoo, "when it may be sounded hoo after a vowel, as 
hood, hoof, hoohy hoop, hoord, and in hood in the End of 
Words as in likelihood, manhood. Priesthood, &c." Finally 
he says u is written hu " when it may be sounded hu, espe- 
cially after a Vowel, as in humble, humility, humour^ Hum- 
phrey** This frequent reference to the vowel depends on 
the following remark: "That h is hardly sounded before or 
after consonants ; but more easily before and after Vowels, 
therefore the best Way to discover on A, is to sound the 
Word that be^s with it after a vowel; as a hat, &c." Un- 
fortunatelv this rule would make a vast number of A's to be 
heard in London, as (a H'oi, a H'ass), an eye, an ass. 

At the present dav great strictness in pronouncing h is 
demanded as a test of education and position in society, and 
consequently most of the words mentioned in Jones are now 
aspirated. Smart, 1836, reduces the list of words with mute 
h to heir, honest, honour, hostler, (in which the h is now 
commonly not written) hour, humble, and humour. It is 
certainly at present very usual to say (nam'bl, jhuu'mj), so 
that the list is reduced to Jive words, which it would be 
considered social suicide to aspirate. But in practice, even 
of the most esteemed speakers, -ham in names of places has 
no aspirate, exhaust, exhibit, exhibition, lose h, and his, him, 
her, etc., after an accented consonant when perfectly un- 
accented, drop their A. Tt is extremely common in London 
to say (« too'trm) for at home. A vast majority of the less 
educated and refined in London, and a still greater majori^ 
in the Midland Counties, never use the h, pronouncing their 
words as if they never had had an h at all. The insertion of 
the h, generally in the form of a very strong (h*), is also a 
remarkable phenomenon, not so common, and still more 
illiterate. 

(H) is properly only a jerk of the voice, and as such forms 
part of the Sanscrit post aspirates (kH gH) etc., and is fre- 
quent as a post aspirate in the Irish brogue. It also occurs 
before every o in Tuscan pronunciation, in which dialect (k) 
is also changed into a strong (.h') thus (.n'onfrnon'tHo) for 
confronto. I have heard lAvomo pronounced in the place 
itself, almost like (livH'or'nn'o) so that a foreigner might 



222 H — XVITH TO XVniTH CENTURY. Chap. HI. § 4. 

easfly persuade liiinself that he heard (lighor*no),^ whence 
an Englishman's Leghorn is but a step. As an initial letter 
however (h) is not common. Thus Sanscrit has no initial 
(h), the letter ^ being (gh). Precisely the same thing 
occurs in Russian, where the (gh) has also to be used for a 
foreign (h). The Gothic h may have been occasionally (h), 
but seems to have been frequently (kh), in place of which 

gi*) as a milder form, became gradually prevalent in the 
ermanic languages. No German at present leaves out or 
puts in an initial h contrary to the orthography ; but final h 
after a vowel, which is dialectically pronounced (kh) or 
(ktrh) as (shuuktrh) Schuh shoe, has disappeared in the re- 
ceived pronunciation. No Scotsmen omit the aspirate. Hie 
old Greeks had an aspirate, the exact nature of which cannot 
be accurately known, as every trace of it has disappeared 
from the lang^ge, and its old relations were rather singular. 
It is a matter of dispute how far the Latins pronounced their 
A, but the Italians, Spaniards, and French have nothing 
resembling the true sound of (h), although the French have 
a trace of its former existence, asserted by Palsgrave but not 
recognized by Meigret, in that hiatus which they call an 
h a^rS, The French and Italian also have no (kh), whidi 
has been retained in the form (kh) by both the Sanscrit and 
Greek. The so-called (kh) x,j\ of the Spaniards seems to be 
a Moorish importation, and is possibly an alteration of (A). 
In Spanish America it is said to be replaced by (h). The 
Spaniards used it to replace a foreign (sh), as in Mexico; the 
French transliterate it by ch = (sh), and the English have 
made Xerez (xee'reec) into sherry. The (h*) is abundant in 
Arabic. 

In England the use of the (h) among the illiterate seems 
to depend upon emphatic utterance. Many persons when 
speaking quietly will never introduce the (h), but when 
rendered nervous or excited, or when desiring to speak par- 
ticularly well, they abound in strong and unusual aspirations. 
It is also singular how difficult it is for those accustomed to 
omit the A, to recover it, and how provokingly they sacrifice 
themselves on the most undesired occasions by this social 
shibboleth. In endeavouring to pronounce the fatal letter 
they generally give themselves great trouble, and conse- 

1 Eear-Adm. W. H. Smyth. The the name of livorno. This would be 

Mediterranean, London. 1864, p. 831, prononnced (leghor'no), and is a sinffa- 

mentions that a map belonging to a far testimony to the antiquity of uis 

Greek Pilot in 1550, now in Srit. Mns. custom of speech. 
Add. MS. 10,134, contains \tyopwo aa 



Chap. III. i 5. PRONUNCIATION — ^XVl TO XVniTH CBNTUEY. 223 

quently produce a harslmess, quite unknown to those who 
pronounce (h) naturally. An English author, S. Hirst, 
writing an English Grammar in German,^ in which 50 
quarto pages are devoted to a minute account of the pro- 
nunciation of English, actually bestows 167 quarto lines of 
German, measuring about 90 feet, upon attempting to shew 
that formerly h was not pronounced in English, and that it 
was altogether an orthoepistic fancy to pronounce it, saying 
that almost all non-linguists would admit that h was gene- 
rally mute, or at most scarcely audible, and that linguists 
who denied this in theory gave into the practice.* The 
division of the people is not exactly into linguists and 
non-linguists, but it must be owned that very large masses 
of the people, even of those tolerably educated and dressed 
in silk and broad cloth, agree with the French, Italians, 
Spaniards, and Greeks, in not pronouncing the letter H. 

§ 6. Eealisation of the Pronunciation of English in the 
XVI ^A, XVII th^ and xvm th centuries. 

Thb results of the two preceding sections are sufficiently 
minute to give an indication of the pronunciation of English 
during the xvi th century, but it is not easy from this mass 
of details respecting incuvidual words, to arrive at a con- 
ception of the actual sounds of sentences. Hart, BuUokar, 
Gill and Butler have however given specimens of connected 
speech, and in Chapter VIII, §§3-6, sufficiently extensive 
extracts will be given from their works, and translated into 
palaeotype, to enable a reader to form an accurate conception 
of Ae sound of our langua^ in the xvi th century. After 
these, follows, § 7, a vocabmary of the principal words pro- 
nounced by the authorities of this period, which will be very 
useful in endeavouring to read any other work of that time, 
because, even if the unknown word is not there found, some 
analogue will almost certainly present itself, which will suf- 
fice to determine the sound within the requisite limits.' 
Finally, applying all the results of previous investigations, 

1 Kritisches Lelirgebandes der en- him. If, however, he had heen aware 

eliichen Sprache yon S, Hirsts Mitolied of the loose manner in which A is in- 

der Universitat zu Cambridge, 2nd ed., serted and omitted in Layamon, GenesiB 

Leipzig, 1847. and Exodns, Prisoner's Prayer, and 

» His nrincipal argument is the re- other writings of the xiii th centnry, 

tention or an, mine, thine, etc., before he would doubtless have considered his 

words b^inning witii h, in the author- point established. In practice I under- 

ized Tersion 1611. The lists of words stood from a gentleman who conversed 

with mute A given by Palsgrave, Sales- with him, he omitted the h altogetiier. 

bury, etc, were of course unknown to * See also the Index of Wot&, 



224 PEONTJNCIATION — ^XVI TO XVm TH CBNTUBY. Chap. HI. { 6. 

I have in ^ 8, endeayoured to realise the pronunciation of 
Shakspere, and have reduced my conception to palaeotypic 
spelling, which will enable a reader of moderate perseverance 
to reproduce it orally. The result is peculiar, and has been 
generally well received by those to whom I have had an 
opportunity of communicating it pivd voce. There can be 
no reasonable doubt, after the preceding discussions^ of its 
very closely representing the pronunciation actually in use 
by the actors who performed Shakspere's plays in his lifetime. 

In Chapters IX and X, I have endeavoured to give a 
similar realisation of the pronunciations which mark the 
XVII th and xviii th centuries. The only connected phonetic 
writing of the xvn th century which I have foimd, is Bishop 
Wilkins's transcription of the Lord's Prayer and Creed, but 
this very inadequate spencimen is eked out by a vocabulaiy 
collected from the principal authorities of the time. It is 
with considerable hesitation, that in the midst of such di- 
versities of sound attached to the same symbols, and such 
numerous lists of rules and exceptions, relating to different 
parts of words and not furnishing the complete representation 
of entire words, that I have endeavoured to restore Dryden's 
pronunciation, or rather the pronunciation of some contem- 
porary reader. It is impossible to feel the same certainty 
respecting his sounds as respecting Shakspere's, and the 
attempt should be viewed with indulgence. 

For the xviiith century, the complete vocabulary of 
Buchanan has enabled me to give his pronunciation of a 
passage of Shakspere, and Dr. Franklin's interesting letter 
furnishes a contemporary piece of phonetic writing, uncor- 
rected certainly, but sufficiently suggestive. A vocabulary 
of the principal words in which Buchanan, Sheridan, and 
other authorities, differ &om the received pronunciations of 
to-day, or anticipate them, will complete the account of this 
century. 

It has not formed any part of the plan of this work to 
enter into detail upon the pronunciation now prevalent, 
although incidental allusions to it perpetually occur. This 
is a very difficult and very complex subject, which has been 
taken up by many other writers, but requires entirely new 
treatment, in reference not only to the results of the present 
investigation, but to those abnormal, cacoepistic, rare, vulgar, 
and dialectic forms, which the history of the past shews mat 
we ought to collect for the benefit of the future, and for the 
thorough appreciation of the real state and possible develop- 
ment of our language, which is principally unwritten. ALr. 



Chap. m. { 6. DIBSCnON OF CHANGE. 225 

Melville Bell's Visible Speeohi op my own Palaeotype, now 
give a means of writing all such forms with great accuracy, 
and the rougher Glossotype (p. 13 and Chapter YI, § 3), will 
enable those who do not wish to enter into minuter distinc- 
tions of sound, to write our dialects much more intelligibly 
than the generality of systems hitherto pursued. I^iose 
therefore who wish to assist in forming a written picture of 
our language for the first time, should neglect no opportunity 
of immediately noting diversities of pronunciation whenever 
heard, after some of these comprehensive systems, of which 
Palaeotype possesses the fi;reat advantage of requiring none 
but ordinary type. To shew the nature of the process re- 
quired, I have in Chapter XI contrasted Mr. Melville Bell's 
and my own pronunciation of the parable of the Prodigal 
Son, and transliterated many specimens of Scotch dialectic 
pronunciation which he has ftunished, both into palaeotype 
and glossotype, while the noliteness of several correspondents 
in the provinces, has enabled me to give a first instalment of 
a greatly needed comparative phonology of the English 
duuects. 

% Q. The Direction of Change. 

For determining older pronunciation than that of the 
XVI th century, it is important to consider the direction in 
which sounds have changed since that period, because we 
can then by continuing the line backwards, arrive at some 
conception of the sounds from which those in the xvith 
century were derived. It is for this reason that so much 
space has been devoted to a consideration of the pronuncia- 
of the XVII th and xviu th centuries. 



TABLE OF CHANGES IN THE VALUE OF THE LETTEBS. 
1. Short VoweU. 

A short, in inth century decidedly (a), became (cb) in the course 
of the zmth and has so remained except in a small class of 
words, where the various sounds (aa, a, aah, ah, sbab, se) are 
heard. 

E short, has remained (e) throughout, but is locally (e) and 
may have been (b) at any period. 

I short, has remained (t) throughout. 

short, seems to have been generally (o) and often {u) in the 
xn th century. The (o) sounds became (o) or (a), it is impos- 
sible to determine which, in the xvnth century, and have so 
remained, the present sounds being generally (o) in dosed and 

16 



226 



BIRECriON OF CHANGE. 



Chap. III. } 6. 



(o) in open syllables. In a few words (o) remains, as cro»$^ 
gone. The {u) sounds, as in the case of short w, became (a) in 
the xvn th century and have so remained. 
TJ short, was either (u) or (w), probably the latter, in the xvith 
century, but during the xvnth become decidedly (o)', which 
has remained to the present day, with the exception of a few 
words which retain the old (w) sound, but some of these are 
^ occasionally pronounced (a), and more of them probably were 
so pronounced in the XYin th century. 



2. The I/mg Vowek. 

A long, was (aa) in the xn th century, but inclined already to a 
very fine and thin pronunciation, nearly (aah), quite different 
from (<w).* In the xvnth century this seems to have become 
decidedly (aeaB), advancing at the close of that century or the 
beginning of the xvmth to (ee), which in the xixth century, 
if not earlier, became (ee) and even (wi). 



^ In an unknown treatise on the 
pronunciation of French, of which two 
quarto leaves with the signatures B i, 
B ii, bearing date 1628, (two years prior 
to Palsgrave's book,) are preserred and 
described in Rev. S. B. Maitland's 
List of some of the Earl^ Printed Books 
in the Archiepiscopal library at Lam- 
beth, 1843, p. 291 (but which did not 
fall under my notice till the preceding 
pages were printed), we read of the 
French A and E, " A. ought to be pro- 
nounced fro the botom of the stomak 
and all openly. £. a lyteU hyer in 
the throte there properly where the 
englysshe man soundeth bis a." This 
would imply that the French sound 
was (aa)y unless it was rounded into 
(aa), as we know that it sounded to 
Englishmen in the xvuth century. 
The English a was quite distinct from 
this and sounded more like (eb) to 
French ears, than (aa). The sound 
could certainly not haye been (eb), or 
Palsgrave woiUd not have found it like 
the French 0, and Salesbury like the 
"Welch a. If we suppose the English 
«, were (aa, ee) and the French were 
(rto, be) we shall be probably very near 
the truth which underlay this and simi- 
lar statements. Compare Gilles du 
Guez, supr^ p. 61. ^ce the above 
was written, Mr. Payiie has obligingly 
brouj^ht under my notice : " The French 
Garden : for English Ladyes and Gen- 
tlewomen to walke in. Or, A Sommer 
dayes labour. Being an instruction for 
the attayning vnto the knowledge of 



the French tongue .... By Peter Enm' 
delly Professor of the same Language, 
London, 1605, Svo., the English in 
black letter, the French in Boman 
type, unpaged, signatures extending to 
P 3, witn two more leaves. The au- 
thor has taken considerable pains, but 
not always successfully, to indicate the 
French sounds, and occasionally refers 
to the English, in passages which will 
be quoted as footnotes to this table. It 
must be remembered that as in the two 
cases just cited, the author was French. 
** Our A is not sounded altogether, as 
this english word awe as some haue 
written, but as the first voice of this 
word Augustine or After opening some- 
what the mouth, as for example, jS<9^ 
tiete^ taeitement, e^auoir : and not after 
the rate of the english word ale^ for if 
a Frenchman should write it according 
to the English sound, hee would write 
it in this wise eel and sound it as if 
there were no i ." This passa^ seems 
to indicate clearly that French a was 
rather {aa) than (aa). It also infers 
that this {aa) was heard in the English 
after, where we retain (aa, aah), but 
that in ale and other wonis of that class 
the Frenchman heard (be). I may 
mention in illustration that Padre 
8ecchi. the astronomer, when speakiof 
English at the meeting of the British 
As^ciation at Norwich, 1868, said 
(mBEd) for made, which to Engli^ ears 
sounded very nearly as (msesed), and 
very unlike (m^). It must be borne 
in mind that &x)ndeU'B etl was quite 



CHikP. III. i 6. 



DIRECTION OP •CHANGE. 



227 



£ long was (ee) during the xvi th and xynth centuries, except in a 
very few words, as he, she, me, etc., because in the xvi th century 
the spelling ee was introduced for those words in which the 
sound has actually altered to (ii), but no such alteration of 
spelling was afterwards admitted, and in the beginning of the 
xvrn th century the sound of (ii) began to prevail, and became 
general by the close of that century, as it now remains. 

I long was a diphthong in the xvith century, probably (ei) but 
occasionally (ai). In the xrn th century, and perhaps during 
the latter part of the xn th, the sound of (ai) was introduced, 
which has remained. Even at the present day, however, (ei, 
ai, ai) and other varieties may still be heard: 

O long was apparently (oo) in the xvith century, a sound which is 
still g^ierally heard before r, in more, glory, &c; but in the 
xvnth century, {oo) was introduced, and still remains, though 
frequently called {po^w) or (oou), and dialectically (ou). Some 
words containing o long were pronounced (uu) but in the xvith 
century these were mostly written with oo, and hence o long is 
sounded (uu) in only a very few words, as move, prove, 

JJ long does not occur in any Saxon words, and intiie xvith and 
down to the middle of the xvnth century had the sound of 
(yy) or some closely allied sound as (n, w, 99) which may be 
still heard dialectically both in the East and West of England.^ 
After the middle of the xvnth century the long u became (iu) 
after a consonant in the same syllable, and (juu) at the be- 
ginning of a syllable, and this sound has remained; in the 
xvmth century, as at present, after (r) it is pronounced (uu). 



dirtinct from 4U our present ail (m1). 
Kb in 1605 there must have been a 
*lam class of speakers who called long 

(aah> or (sBffi), which could hare 
floondea nothing but (bb) to a French- 
man, we may suppose that this was the 
sound with which Brondell, with his 
limited experience as a foreiener, was 
familiar. In: The French Littelton. 
A most easie, perfect and absolvte wav 
to leame the French tongue, Set foortn 
by Clavdivt Holyband, Gentil-homme 
Bourbonnois, London, 1609, 32mo., 
pp. 223, for a knowled^ of which I 
am also indebted to ^r. Payne, the 
author says, p. 184 : ^* At, and ay, have 
three diners sounds : for the first per- 
son 6in^:ular of the future tense of the 
Indicatiue moode, and these three 
yerbes ay, and his compounds : Je s^ay, 

1 know, nay, I am borne, be fiilly 
pronounced as, ^, masculine : b&j then 
for ay, fay 1 have,^V diray I will say, 
Je liray I will reade, faimeray I will 
loue, &o., as if it were written 4, j4, J9 
diri^ je Uri, &e. But the first person 



singular of the first perfect tense of the 
Inmcatiue moode, is sounded as it is 
written, as faimay I loued,^> trouvay 
I sound, je parlay I q>oke, &c. (ai ?). 
As for the rest, wheresoeuer you shall 
find at, sound it as gaye \gay in p. 185] 
gaping.*^ He means of course (bb), 
and he seems to agree with Hart partly 
in gay, and with the xvnth century 
pronunciation generally in gaping. The 
only English writer who would make 
^ay = (gBB) is Cooper, samk p. 125. 
Most probably the Frenchman heard 
an English (gaei) as bis (gss), and 
found the first syllable of aaping^ 
(gaese), more like his gai than nis ga, 

* Erondell says of French ui " t> Is 
sounded without any help of the tongue 
but ioyning of the ups as if you would 
whistle, say «, which «, maketh a Billa- 
ble by it selfe, as vnir, pniquement as if 
it were written v-neer, pronounce then 
muii^, punir^ ntbvenir not after the 
English pronounciatioft, not as if it were 
written mueaique, pueniry euevenirj but 
rather as the m in this word, mMrtheret, 



228 PIBECnON OF change. Chap. m. § 0. 



3. Cofnhtnaiumi with A flnai. 

AA was not used in English words in the xnth or subsequent 
centoriesy except in Hebrew names, as liaae. 

EA, which had been used occasionally without any strictness for 
long tf, was established towards the close of the xn th century 
as (ee)y and remained so throughout the xvnth century, witib. 
the exception of about 30 words. In the xvinth century 
however it rapidly altered its sound to (ii), only a few words 
finally resisting tne change, after having yielded to it for a 
time. Several words with (e) short, were fix)m the middle of 
the XVI th century, and etill are spelled with ea. 

lA had no particular value separate fix)m (ta), and has followed the 
fortunes of its components, one or the other letter being 
frequently omitted. 

OA was introduced at the close of the xnth century for the long 
(oo) in closed syllables, after oo had been appropriated to (uu). 
In the xvn th century it became (oo), except in broad, groat, 
where it was (aa). It has retained these sounds. 

U A is not an English combination. 

4. ComUfHUum with E final. 

AE was so to speak, not msed, in the xvith century ; even in Latin 
words $ was often employed* When a was introduced into 
English it was always pronounced as the long d of the period. 
This a is one of Bullokar's signs for (ee). 

EE was introduced in the middle of the xvith century for the 
sound of (ii), which it has since retained. In the earlier part 
of the century no distinction was made between e$ and long $, 

IE was a combination having the same meaning as long $ until the 
xvnth century, when it was considered the same as ee, 

not miking the II too loiiff.^' It is very whenEnffliafamendoprofSnryTj&ejiaj, 

difficult to imderstand we meaning of you : ana for, q, we snppoae they say, 

thifl nasMge. It is possible that as kion: but we sonnd^ t, without any 

Erondell may have met with those who helpe of the tongue, loyning the lips as 

said (8B»\ he might haye heard (iu), if you would whistle ; and after the 

which or course most hare been fre- manner that the Soots do sound Oud." 

quently used at this date, though it was Here we hare the first distinct reoog- 



not reoeiyed, and as this sound did not nition of the English long u as (iiO 
satisfy him he took refuge in (u) or (ii^ distinct from &e Scotch and French 
as coni^ised by a following (r), ana (yy). Hart, who in his first treatise 



yidual pronunciation, which he had not idenlifies English long u and ycu, 

satiafiictorily apmneciated, but conceiTed makes both the same as the French and 

to be generaL Holyband also (French Scotch, and in his second treatise, suprli 

Littelton, 1609, p. 162^ seems to hare p. 167, distinctly describes (yy) and not 

recognized (in) in English and not (yy), 7iu) for this sound. Wilkms, 1668, is 

for Ee says: "Where you must take the next author who distinctly reoog- 

paine to pronounce our, t, otherwise nizes (iu), Wallis, 1653, being the last 

^en in English : for we do thinke that who as custinctly insists on (j^). 



Chap, m.} 6. DIBECnON OF GHAKGB. 229 

OE was not an Engliah combination ; when it was introdnced as ce, 
it followed the sound of the long e of the period. 

U£ was only used at the end of words in the xmth oentory and 
later, for the long u, which had in this situation been pre- 
Tiottsly written ew. 

5. Camhinatuma with I or Y final, 

AI was (ai, aai) in the xnth century and possibly (aai, sesBi) in the 
xm th ; but towards the close o€ that century, and in the pro- 
nunciation of a minority even as early as the middle of the 
xvi th century, ax was called (ee).^ Becoming thus identical 
with long a, it shared its fortunes and fell into (^, ^i). 

EI was (ai) or (ei, eei) in the xyith century, and seems to have 
retained the sound of (eei) or (ee) till a late period in the 
xvin th 'century, when many, but by no means all the ei fell 
into (ii). In either^ neither, the old (ei) developed (ei) as well 
as (ii), and both sounds are yet heard from the same speaker at 
different times. 

n was never used. 

01 was (oi) and nearly (ui) in the xn th century, in some words 
(oi, uui) were heard indifferently. In the xvnth centuiy 
though (Ai) or (oi) was the rule, (oi) was frequently hearo. 
In the xvmth and nxth centuries only (oi) was recognized, 
although some speakers still say (oi), now considered a vul- 
garism. 

UI was not a genuine English combination, and was only a sub- 
stitute for long tt, or long and short i , and followed their laws. 

6. Combinatiom with O findt, 

AO is only accidentally an English combination in extraordinary , 
where it is usually pronounced (aa). 

EO when used at an earlier period seems to have been considered 
identical with long e, and has been generally so treated. In 
pigMm, dungeon, the combinaticm eo is only apparent, for the e 
belongs to the preceding y, 

10 is not found. 

00 was used in the beginning of the xvi th century indifferently 

t with long 0, but was introduced towards the close of that 
century to indicate those long o which had come to be pro- 
nounced (uu), and it has retaiued this value. 

UO IB not used. 

1 Erondell says in the French nonnced as these english words dS(iy,fay, 
Garden, 1605, speaking of French 0^ t^oy" which he therefore id^tifies 
which was then certainly (s) : " Also with long a. No Enelish writer of the 
if f doe follow tU, it maketh the word period makes this concision. Bnt corn- 
long, and the a ynsonnded, as Mautre, pare Holyband's gay, gaping, sapr^, 
paisire, where the ui or ay he pro- p. 227, note, coL 2. 



230 DIBBCnON OF CHANGE. Chap. HI. i 6. 

7. Combinations with Uor Jf^ final. 

ATI was (an, aau) in the xvith century, and seems to liave passed 
by the absorption of (u) into (tc\ or simple labial modification, 
into (aa) in the xvnth century, which sound it generally 
retaros although there is still a contest between (aa, aa) in 
a few words. 

ETJ had in the xvith century two sounds (yy) and (eu) which 
were not distinguished by any orthographical expedient. In 
the xvn th century the (yy) sounds became (iu, Juu), and the 
(eu) soxmds either remained (eu), or became (oo). In the xvui th 
century those which had become (oo) remained so, the rest fell 
into (iu, juu) where they have since remained. 

in is not used. 

OU in the earlier part of the xnth century, and in the pronuncia- 
tion of some writers even down to the latter part of that 
century, had the sound of (uu, u) ; by the middle of the 
XVI th century it was generally pronounced (ou), but occa- 
sionally (uu). A class of words in ow, however, derived from 
the Anglosaxon aw, ow, was by both set of speakers pro- 
nounced (oou). In the xvn th century the (oou) sounds be- 
came (oon) as they have since remained, though theoretically 
considered as simple (oo). The (ou, u) sounds at the same 
time became (ou, a) and have since retained these forms. 

TJU is not used. 

8, ComonanU, 

B invariably (b). 

C invariably (k) before a, o, u and (s) before (e,i), except that 
in the xvnith century, and perhaps earlier, e before a became 
(Jc) ; and ci- before a vowel became (sh). 

CH sometimes (k) in Greek works, generally (tsh) throughout the 
period. 

D mvariably (d) except that, in the xvmth century, d in the 
termination -dtire, -diet became (dj) or (dzh). 

F invariably (f). 

G invariably (g) before a, o, m, and almost invariably (g) in 
Saxon words before e, i; otherwise invariably (dzh) before 
e, f. In the xvmth century and perhaps earher, p before a, 
and OH before i long became (p), 

GH in the beginning of the xvitii century, fidl (kh) or {kh); 
towards the middle and close, very gently pronounced, almost 
(h*); and in the xvnth century and subsequently entirely 
lost. In a few words of the xvi th century and more after- 
wards, gh was sounded as (f ). In one word, siffhy in the 
xm th and xvm th centuries ah was called (th), and in one 
word, hiccough, (p). When gh was omitted in speech after i , 
the sound of that letter was changed from (• ) to (oi) ; the 
sound of augh with silent gh was either (aa) or Taa) ; of ough 
with silent gh, {oovl) or (aa), sometimes (ou) and (uu). 



Chap. III. { 6. DIRECTION OF CHANGE. 231 

H in many words in the xvntli century, where it is now never 
omitted, was not sounded. 

J or " I consonant" had invariably the sound of (dzh). 

K was (k) before all vowels, perhaps inclined to the palatalised 
{k) before the sound of (ii), and in the xvin th century fre- 
quently became {k) before a (ae, aa), and long » (ai). 

L invariably (1) or ('1). In the xvi th century it was beginning 
to disappear after a, after becoming labialised to (Iw) and thus 
changing the sound of a from (a) into (au, aa), the latter pre- 
vailing in the xvn th century ; (aa) is now commonly heard in 
the termination -aim, 

M invariably (m) or ('m). 

N invariably (n) or f 'n). 

NG invariably (q) or (qg), except in the combination -nge when it 
became (-ndzh) and had a tendency to change preceding (a) 
into (ai) which became subsequently (ee). 

P invariably (p). 

PH invariably (f ), except perhaps in such combinations as Clapham^ 
in which the h was omitted in the xvnth century. 

QTJ invariably (ku?), or labialised (k). 

R preceding a vowel, invariably (r), following but not preceding 
a vowel, it was most probably (i) as early as the xvmth 
century, and possibly in the xvn th. 

RH was the same as simple r. 

S initially, invariably (s), medially and finally either (s) or (z) 
according to present usage. In the xvmth century s before 
long ti, and «i- before a vowel became (sh), and -isi- became 
(-izh-) ; in the termination -surej s became (sh) or (zh). None 
of these changes seem to have been acknowledged before the 
middle of the xvn th century. 

T invariably (t), except that ti- in the terminations -^fb», -tiouSf 
was (st) in the xvitii and xvnth centuries, and became (sh) in 
the xvnth. In the termination -ture in the xvmth century, 
t feU into (tj) or (tsh). 

TH either (th) or (dh) according to the present laws, except that 
in the xvith century it was (t) in Thomas as now, and also in 
throMy and (d) in Thavies Inn ; and generally (th) in with 
instead of (dh) as now. 

V br "TJ consonant" invariably (v). 

W as a consonant, whether confused with an initial (u) or not, 

invariably (w). 

WH, whether confused with (hu) or (nw), was probably always (wh). 
X invariably (ks), the present use as (gz) seems to have been 
unknown previously. 

Y as a consonant, whether confused with an initial (i) or not, 
invariably {j\ 

Z invariably (z). 



mmam 



232 



DIRBCTION OF CHAIVGE. 



Chap. HI. i 6. 



On examining this table of changes, it would appear that 
the consonants haye been subject to little or no alteration, 
except under the action of an (i) or (u) sound. The action 
of an (i) sound changes (t, d, s, z,) to (ti tsh, dj dzh, sh, 
zh), but this action did not materially anect the English 
pronunciation of the xvi th and earlier part of the xvn th 
centuries. The (u) soimd was generated through the labiali- 
sation of (1) which gradually di^ppeared, labialising the pre- 
ceding vowel. 

The consonant gh, originally (kh)^ became gradually dis- 
agreeable and harsh to the l^uthem English and passing 
through (h') soon ceased to be appreciable, and was therefore 
neglected, although it was probably theoretically maintcdned 
long after it had practically disappeared. On examining 
the oldest forms of words, howeyer, this sound appears to 
haye passed through (i, u), and in its disappearance to have 
acted by palatisation and labialisation on the preceding 
yowel. The change of igh to long t is th^ only one that 
presents a difficidty, and this depends upon the same cause 
which changed long t generally from (ii) to (ei), p. 234. 

For the yowels the following changes occur, taking the 
sounds only, independent of the spellings. 



Short VoweU, 


Long Vowels, 


D^lhomgs. 


a, » 


aa, 8B8B| ee, 


M, 061 


ai, 8Bi, ei, eei, ee, m, en 
au, aa\ aa 




ee, u 




61, 91 

ei, eei, ee, ii 
eu, iu 
eu, 00^ oou 


0, 


00, UU 




OU, 9U 




00, 00 oou 




oou, ooy oou 


U, 9 


UU, OU, 9U 




Ul, 01, Al, 91 



The directions of change are here seen to be three, — ^towards 
(i), towards ^u), towards (e). But the two last are not 
essentially different, as (u) may be considered as a labial- 
ised (e), p. 162. 

The long yowels haye altered more than the short yowels. 
The yoice being sustained there was more time for the yowel 
sound to be considered, and hence the fancy of the speaker 
may haye come more into play. This has generally giyen 
rise to a refining process, consisting in diminishing the lin- 
gual or the labial aperture. The lingual aperture is materi- 



Chip.iil§6. tanasuncfs of chanov. 233 

ally dmunidied in the passages (aa, wsb, ee, ee) and (ee, ii). 
It seems carious that the firat was not continued as £ur as 
the second. In the name James, however^ which became 
(Dzheemz) in the xvii th century, and has passed to (Dzhiimz) 
in flunkey English, and to (Dzhtm) as a common abbrevi- 
ation, the series of changes is complete. Fashion and refine- 
ment have nearly banished (aa), but have not yet confounded 
in one (ii) all the words formerly disting^sh^ by (aa, ee). 

The change of (oo) to (uu) was a similar refinement, con- 
sisting first in the elevation of the tongue, and correponding 
narrowing of the labial passage, producing (uu), and secondly 
in the narrowing of the pharynx. The change from (oo) to 
(po) consisted simply in narrowing the pharyngal cavity. 

One of the most remarkable changes is that from (uu) a 
simple vowel, into (ou) a diphthong. Both sounds held 
their own side by side for some years. Palsgrave in 1530 
and BuUokar in 1580 both upholding (uu), while Salesbury, 
Smith, and Hart declared for (ou), which finally prevailed. 
Although the change is certain, there is no trace of any 
reason being given, and as the sound (uu) had been repre- 
sented by the letters otc in those cases where it changed mto 
(ou), whereas when (uu) was a change of (oo), it did not 
further change into (ou), and the orthography also did not 
give ou, — the mere accident of the spelling naturally presents 
itself as a cause. This hypothesis is strengthened by ob- 
serving that in the north of England, where reading was 
perhaps less common than in the S>uth, the sound of (uu) in 
these words still remains unaltered. But such a supposition 
can hardly be correct, because the change of (uu) into (ou) 
is precisely analogous to the change of (ii) into (ei), a change 
which must certainly have occurred in passing from the 
Anglosaxon period to the xvi th centuir^ althougn it has not 
yet come distinctly before us, and had no connection with 
the orthography. In each case the change simply consists in 
commencing the vowel with a sound which is too open, (that 
is, with the tongue not sufficiently raised), and, as it were, 
correcting that error in the course of utterance. This variety 
of speech might easily be generated and become fashionable 
in one part of the coimt^ and not in another, and as it 
penetrated far beyond the classes whom orthography could 
affect at a time when books were rare, and readers rarer in 
proportion to the speakers, the physiological hypothesis 
seems more deserving of adoption than the orthographical. 
On further examination it will be found that this hypothesis 
has an analogue in a well known custom of the South of 



234 DlBECnON OP change. Chap. m. J 6. 

England. In the North of Eneland, in France^ and Ger- 
many, no difficulty is felt in prolonging the pure sounds of 
(ee) and {oo), but in the Soudi of England persons have in 
general such a habit of raising the tongue slightly after the 
sound of (ee), and both raising the tongue and partly closing 
the lips after the sound of (oo), that these sounds are con- 
verted into the diphthongs (ee'j, oo'w), or (eei, oon) where 
the (ee, oo) parts are long and strongly marked, and the (i, u) 
terminals are very brief and lightly touched but still per- 
ceptible, so that a complete diphthong results, which how- 
ever is disowned by many orthoepists and is not intended by 
the speaker. Now we have only to suppose a habit growing 
up of beginning the (ii, uu) sound with a tongue somewhat 
too depressed, and in the latter case with the lips also too open, 
but passing instantly and rapidly from these initial sounds 
to the true (ii, uu), and (di, ouu; would result. From the 
habit of accenting the first element of a diphthong, the 
initial touch of (e, o) would come to have the accent, and 
being very short and indistinct might readily vary in dif- 
ferent mouths into (a, a, e). We should thus obtain the 
diphthongs (ei, on ; ei, ou ; ai, au ; ei, eu) in which also the 
second element may be, and at present in the South of En&^ 
land seems to be (i, u) rather than (i, u). Thus on length- 
ening out the terminal sounds of nigh, now, I seem to hear 
in my own pronimciation (notu, nduuu). 

The generation of (eei, oon) from (ee, oo) consists then in 
subjoining brief (i, u) to long (ee, oo) ; while the generation 
of (di, ^uu) from (ii, uu) consists in prefixing brief (e, o) to 
long (ii, uu). The elements in both cases are the same (eei^ 
en; oon, onn) and the accessary sounds are in both cases 
brief, but when terminal they are unaccented, when initial 
accented, just like an appoggiatura in music. 

We might therefore expect to hear (ei, on) developed 
either from (ii, uu) or from (ee, oo). Further reasons for 
supposing the first to have actually occurred will be given in 
Chap. I V, § 2, under I. For the second, it is not uncom- 
mon at present to hear (a) for (ee), and (^u) for (oo), although 
these changes have not been generally recognized. 

This change of (ii) into (ei, ai, oi), and (uu) into (ou, au, 
9u) is etymologically interesting because it is by no means 
confined to our own country. The Gothic (ii) corresponded 
to (ii) in Icelandic, Anglosaxon, Friesic, Old Saxon, Low 
German, and Upper German, and is still (ii) in Danish and 
Swedish, but is now (ei) in English and Swabian, and (ai) in 
Dutch, High German, Frankish, East Frankish and Bavarian, 



Cbap. m. i 6. DIRECTION OF CHANOE. 235 

according to Bapp (Fhys. d. Spr. iy.^ •144) and the same 
writer says that (uu) in Uothic was (uu) in Icelandic, Anglo- 
saxon, Friesic, Old Saxon, Low German, Upper German, 
and is still (uu) in Danish, but it has become (au) in English 
and Swabian, (au) in High German, Frankish, East Framcish 
and Bavarian, (ay) in Dutch, and (uu) in Swedish. Except 
the two last changes, the phenomena must be all referable to 
local habits of the kind named. The Dutch sound (ay), written 
tdy would appear to be an alteration of (au), but whether there 
is any historical as well as phonetical ground for supposing 
such a form to have existed, I cannot say.^* It is impossible 
not to be reminded in this historical change of (ii, uu) into 
(ei, ou) of the (guNa) chanffes in Sanscrit, because they are 
phonetically the same, although they arise in a difiterent 
manner. 

^ We haye then briefly the following changes of the prin- 
cipal yowel sounds, of which the change (ii) to (ei) was 
anterior to the xvith centuiy, unless, as seems to be the 
only legitimate inference, Palsgraye's and Bullokar's state- 
ments (pp. 109, 114) are held to imply that long t was still 
pronounced as ii in some words by them : — 

From (aa) through (sbab) to (ee, ee^ eei) 

From (ee) to (ii) 

From (ii) through (ri) to (ei, ai, oi) 

From Too) to (uu), or to (oo, oon) 

From (uu) through (ou) to (ou, eu) 

Proceeding backwards, then, we must, if there was any change, 
look for it in the same series. Thus (aa, aa) may haye 
preceded (aa). Perhaps (be) may haye preceded (ee). The 
sounds {ee, oo) may haye preceded (ii, uu), and it is possible 
that (aa) may haye preceded (oo), as the latter is only 
the roxmded form of the former. 

The yowel (yy) can hardly haye been an original yowel 
sound. Its relations to (i, u) and (iu) are so close, that it 
might haye arisen from any one of the three, but it has 
principally the appearance of being an alteration of (u) 
caused by making the narrowest part of the lingual channel 
with the middle instead of the back of the tongue. This 

^ In the actual Dutch prommciation make on hearing the sound, not (ij) as 

of huist muiSf it is yery difficult to Br. Bapp remarks. The Dutch oon- 

di8tin^:ui8h the sound from (au), and sider it to be the sound of the German 

the difference seems mainly produced eu, which Dr. Bapp also says is sounded 

by altering the form of the lip into (^y) in the North-East of Oermany, 

that for fyy), which is slightly flatter Berlin, Brandenburg, and on the Baltic 

than for (uu), rather than oy bringing coast from Mecklenburg to Bussia ; 

the tongue into the (i) position. StiU the general sounds being (ay, oy, oi) 

(ay) was the best analysis I was able to and eyen (oi) in Hamburg. 



236 DntscnoN of chakob. ghap. m. f 6. 

d priori physioloflnoal ccmoeptioa is oanfinned by finding diAt 
dialectically, in oootland and in Devondiire, (yy) or some 
form of it as (n, uu), occurs as a sabstitnte for (an), as tiia 
Devonshire (myyr, my3m), or more properiy (munv, muun) 
for (muay, mnmi). In G^erman we find thieit (yy) has also 
been generated from (an) by the retroactive «nect <rf an (i) 
or (e) sound in an added pliable. In French, the snbstitn- 
tion of (yy) for the Latin (uu) can only be traced to a 
national habit. The same seems to have occurred in Greek, 
where v was at a very early period changed firom (uu) into 
(yy). There is nd hiatorical evidence that (yy) can be con- 
sidered in any case as an alteration of (in), idthoogh we have 
in English the proof that (in) may be an alteration a£ (jj), 
and we know by the Welsh um and Hart's iu, that the use 
of w as a representative of (yy), was naturaL In fisict the 
second vowel u in both iu, em naturally suggests a labiafiaation 
of the preceding, which would give mi, oti = {iw, av) = (i, 
o), whence (y, a) readily derive. This se^ns to have been 
the case with Ulphilas, ^dio certainly uses am for (a) and 
probably iu for (yy).^ 

In such languages as the English, French, and Ghredt, 
where the natural sound of s had beoi replaced by (yy), 
the only device left for marking the (uu) sound was to use 
the m>m which it was derivra, as in the Swedish, or to 
put an before, after, or over the u to indicate more dis- 
tinctly that the combination was to have tiie modified o 
sound. This seems to be the origin of the use of ^m in 
older English, French, and Greek for the sound of (uu). 
Similarly in dd High German no, in Italian no^ in Bohe- 
mian i are employed to indicate rdations between u and o.^ 

^ Weingfirtner (Die Aoi^iTache dea seqiieitt (w) by tiie fip aetieB of (jy), 

Oo«hiiGhenxiirZeitdeiXJlfilM,Leipd|^, which is ntn^j tbt Mine m thai of (u), 

1858, 8to. pp. 68) sqjdb up all the on the following TowelSf precisely as ia 

argomentB bearing on the pnmanda- ^e case noticed <m p. 133 note. The 

tion of Gothic m in fiiTonr of (n). cofflbination m is me most Jt<B«tH 

The actual En^ish change of (jy) into to appredato ia ^b Grothic and old 

Qn), and the common German change high German orthographies, 

of (yy) into (ii), seem sufficiently to * The Dutch use o§ for (uu) or (u), 

account for the rarious farms, which tiieir long and short u being (yr, xl 

the Gothic iu received, or rather to tiiat is, nearly precisely the same as Wal- 

wfaifih it eorrespooded in rariovs Ger^ bs*s En^x^ sotBoda. The okkr Dufieii 

manic dialects. The alteration of in writna seem to hnve used • as a simpk 

into w before rowels, as in ibmi, ibiirM, ngn of prolongation in as^ m, «•» so 

may be explained as perhaps (hnyy, tiiat ss can only be regarded as « used 

knywis) the full written form kmrnit for (uu) witii a saedal mark of pr»- 

haying been contracted into kninsy as longation. In mooem Dutch Hut souad 

the single letter 9 seemed most neatiy b frequently diort, as there is no otihor 

to express first the labialisation <^ the means of repreeentiiw (u, «). Stsgea- 

I, and secondly the gvuratioii of a sob- Utk (yedgrdnitochaSpdlia& /Iwrtg- 



GkAP. HL { e. vtMEcaoax or ghahob. 237 

In EngKA the change of (jj) has been into (in), Imt in 
German it changeB into (ii), tliat is, in Engliah the Upe were 
not rounded at Uie b^;iniiing of the sonnd bat were rounded 
at the end of the aoond, producing first (iy) and afterwards 
(iyn, in), while in Ghnnan the lips are firequently not 
roondedat alL 

For the long Towels^ then, anterior to the xvi tb oentniy 
we may possiUy have (aa) for (aa) ; (eb) for (ee) ; (ee) for 
(p) 9 (oo) for (nu), and (nu) for (yy) ; (oo) is not likely to 
haye been changed. 

For the short Towels we find no change in (i, e), which we 
tberefore most suppose to haye existed anteriorly in this 
fbnn. The chanjge (a ) to (se) could only give {a) for an 
anterior sound. The changes (o, o) and (u, e) could lead to 
no conclusions respecting any anterior sound. The first 
change (o, o) consists merely in depressing the tongue, the 
second change (u, e), as has been shewn, ms^ consist only in 
n^lecting to close the lips suflSciently. These changes do 
not giye sufficient indication of direction. It would be safest 
to conclude that (a) or (a) and (e, t, o, u) were the sounds of 
the fiye vowels before the xvi th century,^ but the words busy, 
bury (btz't, beri) and the pronunciation (trist) for trust, 
leads us to suppose that u in writing may often indicate a 
short (y) which would be taken as {i). 

We find then that there was probably an older pronuncia* 
tion of the English yowels than that of the xvi th century, 

dan, 1804, p. 139V denies that it should alitj, in Bdginm. This left m free for 

be considered as long i. although it is (nu, n) without any danger of oonfti- 

now pronounced (ii), because lonf t sion, and eren the belgians admit the 

used to be written iu ih and says &at distinction oo, oo. 

in the protinoe of Zeeland to is still ^ Hart expresslj says : "And|to per- 

heaid as a distinctly mixed sound swade you the better, that their auncient 

** dnidelijk een gemengd gelmd," pro- sounds are as I bane sayde," that is 

bshly (iia). The same author (p. 82) (a, e, i, o, u), " I report me to all 

aocoonts ror the use of o as a mark of Musitians of what nations soeuer they 

prolongation in oo, oo, n^j on the ground be, for a, e, i, and o ; and for u, also, 

that when words anciently written except the French, Soottish and Brutes 

wmU kopty mure, came to be pronounced as is sayd : for namely all English 

»imf, hojfy Mwr*, without the final o^ Musitians (as I can mderstande) doe 

the was transposed in writing, thus sounde theuL teaching vi^ re^ mi, fiu 

wmtt, koep, tniior, precisely as Lane pro- oo/, to ; Ana so do all speakers and 

posed to write English, supr^ p. 44, readers often and much in our speach, 

L 3. The orthographies oo, tio for as in this sentence: The jpratling 

(oo» 77) 1^ l>cc^ replaced by oo, uu Hosteler hath dressed, cumed, and 

lor more than two centuries liefore he rubbed our horses welL Where none 

wrote, and he proposed and prevailed of the fine rowels is missounded, but 

on the Dutch to use 00 for oo, an kept in their proper and auncient 

orthogra^y jealously retained with tio, soundes : and so we mare Tse them, 

yfiur MS y> as marks of distinot nation- to our great ease and profite/' 



238 DIBECnON OF CHANGE. Chap. HI. i 6. 

and that we may not unnaturally expect to find in it {aa^ ee, 
ii, 00, uu) for (aa, ii, ei, uu, ou) of the xvi th century. 

As to the diphthongs they have followed two courses, ac- 
cording as the first or second element became the most con- 
spicuous. In (ai) the (a) has been gradually made closer, 
chanfiing in the diphthong (asi, ei), as in the simple sound 
(sB, e) , and then the first element being lengthened (eei), the 
second gradually disappeared (ee), onljr to reappear as a faint 
aftersoxmd in the present century {eei). Hence, before the 
XVI th century we can only expect the (ai) to have been the 
same, or at most to have been preceded by (ai). On the 
other hand (ei) may have had an antecedent (ai). It is a 
remarkable circumstance that (ai) in French also gave place 
to (ei) and then to (ee), p. 118. In Modem High .German 
we also find a dialectic substitution of (ee) for (ai), afi (een) 
for (ain) one, but it remains to be proved which is the older 
form, the old high German ei answering to the Gothic ai = 
(ee), and the modem high German ei often answering to an old 
high German I = (ii), of which (ee) may be a first degradation. 
In Latin (aaii) as in piclai appears to have generated (ai, ee) 
as in pictie (pik'tee). In Greek <u, which could hardly have 
been originally anything but (ai), is now (ee) and was so ap- 
parently at the time of Ulphilas. In Sanscrit the (guNa) 
combination (ai) resulted in the present (ee) or (ee). 

In (au) the (a) has been gradually made opener (a), and 
the (u) has acted more and more to produce a labialisation 
of this open (a), thus (a'it) till it disappeared altogether ; 
leaving (aa) only. We cannot, therefore, well suppose (au) 
to have preceded (au). The soimd may have had an ante- 
cedent (eu), but was most probably original. It is remark- 
able that (au) in Welsh generated (oo), uiat is (a) was labial- 
ised to (o = aw), without being previously broaaened to (a), 
in quite recent times, pob, patch = (poob, paub) being still 
co-existent. In French (au) produced {oo). In German (au) 
is often dialectically (oo). In Latin (au) became It^ian 
(oo), as paucus poco (poo'ko). In Sanscrit the (guNa) com- 
bination (au) has become (oo) or (oo). In Greek the vowel 
(u) fell into the consonants (bh, ph) and hence the vowel 
was preserved. But Ulphilas used tie combination (au) for 
the Greek 6 fiiKpov. 

The change (ei, ai) hardly indicates a direction. But as 
(ou) had an antecedent (uu), so (ei) may have had an an- 
tecedent (ii). 

The change of (eu) to (iu) on the one hand and {oo) on the 
other is recent. One or the other seems to have occurred 



Chap. IH. § 6. 



DIRECTION OF CHANGE. 



239 



according as the first element (e) or second (u) preyailed. 
The number of words in which llie soimd of (eu) remained 
is so small that it is difficult to form any conclusions on the 
change.^ 

The change (ou, eu) would have been insufficient, if we 
had not known that (uu) generally preceded (ou). 

As far as the xvi th century is concerned (oou) is original, 
but as (aa) may have preceded (oo) so (aau) may have pre- 
ceded (oou). 

There seems every reason to suppose that (ui) was the 
original form of the diphthong which is now (oi), and that 
the form (uui) which we find in the xvith century, and 
which, altered to (ei), appeared in the xvii th century, and 
crops up even now, is not an alteration of (oi), but is rather 
a renmant of the older form. It does not appear possible to 
suggest an antecedent for (ui). 

Combining the above observations on the direction of 
change, with the orthographical representation of sound, we 
shoidd be led to expect that previous to the xvi th century 
the sounds attributable to the various letters in alphabeticsu 
order might possibly be as follows : — 





P088IBLB Sounds 




P088IBLB SOTTICDS 


MODBBN SPKTiLUKO. 


BEPORB XVI TH 


MoDBBir Spbllino. 


BEPORB XVITH 




Cbntubt. 




Cbntubt. 


a short 


a, a 


f short 


f 


along 


aa, aa 


t long 


ei, ii 


at 


ai, oi 


ie 


ee 


au 


au 


short 


0, a; u 


e short 


e, B 


long 


00, aa; uu 


e long 


ee 


oa 


00, aa 


ea 


ee 


0% 


oi, ui 


ee 


ee 


00 


00; uu 


ei 


ei, ai 


ou 


oou, oou ; uu, u 


eu 


yy, eu 


u short 


^; hj 






«long 


yy, uu 



But at what time any such combinations were prevalent, 
and how early the xvi th century pronunciation had prevailed, 
we must seek other evidence to shew. In the meanwhile, by 

habit of *' separating the labio-linciial 
vowels (n, 0) into their ling^ & labial 



> The pronunciation cited on p. 141, 
(shvn) for shtWj must be some dialectic 
remnant of (sheu), and suggests an 
intermediate between (sheu) and (shoo). 
Hart in his phonetic writing uses bow 
(shio) and (sheu) for thno. Mr. M. 
Bell notices that there is a * Cockney" 



components, & pronouncing the latter 
successively instead of simultaneously," 
one result of which is saying (au) for 
{oo), Yiaible Speech, p. 117. 



240 DIBXCnOH OF CHABOB. Ckaf. m. { 6. 

ocmipttrmg this porely ihecyretical tftUe, Cnmd^ 
of an J kmd, pot purely deduced firom a comBidenitioa of the 
directioa of cnange, and not limited to any paiticular period 
of time preceding the xvi tH century, with the taUe giren 
by anticipation on p. 28, as an expression of the general 
general results of the following inyestigaticm respecting the 
xivth century, it will be seen that there is a remarxable 
agreement between the two, so that all the results there 
obtained may be pronounced theoretically probable, howerer 
strange they would have appeared if the direction of change 
had not been previously ascertained. At the same time the 
great difference between the sounds here considered as pos- 
sible, and those which, based upon present habits, are usually 
assumed, will serve to shew the value and importance of the 
preceding investigation. The subject has hitherto been 
considered from mr too modem a point of sight, and with 
fSEtr too limited a range of vision. The changes in the last 
three centuries, of which we have contemporary evidence, 
not having been generally known, and the changes in the 
cognate Oermanic dialects, although recorded by Bapp and 
Grimm, not having been duly weighed, and the habit of 
reading Spenser and Shakspere in our modem pronunciation 
having become ingrained, we were prepared to regard the 
sounds of our language as something fixed and settled in 
point of time, at most admitting a dimectic difference which 
we perhaps attributed solely to geographical causes. This 
must now be given up, and we must proceed to investigate 
pronunciation with a knowledge that it has changed, and 
must change chronologically, that at any time there must 
be, even at the same place, diversities of coexistent forms ; 
and at different places, even when the language has been 
derived, at no very ^"eat interval, from the same sources, 
there must also be differences arising from want of commu- 
nication, which will therefore be the more striking, the 
earlier the period and therefore the more imperfect the 
means of transit, and especially that any cause which will 
occasion the intercommunication of districts usually isolated, 
must have a great effect on pronunciation. Our endeavour 
therefore will be to discover, not what earlier English pro- 
nunciation was generally, but as definitely as possibly what 
it was at different particular times and places. Of course 
this can only be done by means of determining the value 
attributed to the ali)hal)etic symbols by writers^ of known 
time and place. Tnis is the object of the investigations 
contained in the two next chapters. 



241 



CHAPTER IV. 

On the Pronunciation of English during the Four- 
teenth Century as Deduced from an Examination of 
THE Bhtmes in Chaucer and GtowER. 

§ 1. Principles of the Investigation. 

The War of the Roses raged from 1466 to 1486. The 
Long Parliament met in 1640, and Charles II. returned in 
1660. Hence the xv th and xvn th centuries were memor- 
able in English history for two long continued civil wars, 
causing imprecedented communication between all parts of 
the country, and withdrawing the minds of men from litera- 
ture to fix them upon the events of the day. This ^'commyxs- 
tion & mellynge,'' as Treuisa hath it, of men from the 
Tarious counties of England necessarily produced an effect 
both on the structure and pronunciation of the language. 
The whole style of English at the close of the xvii th cen- 
tury is diHBJTnilar from that at the close of the xvi th. A 
different mind reigned in the people and required a different 
instroment to express itself. And that this was not confined 
to an alteration of words, idiom, and composition of sentences, 
but extended itself also to pronimciation in a most distinctly 
characterised manner, we nave already seen. The xvii th 
century produced a number of writers who paid attention to 
pronimciation, who sought either to investigate the relations 
of spoken soimds, or to supplement the deficiencies of ortho- 
graphy by Usts of words and rules, by which the pronuncia- 
tion coula be tolerably ascertained. These lists and rules 
became so fiill towards the close of the xvii th century, that 
we have been able to trace the successive phases of alteration 
which words imderwent, and to see how the sounds of the 
XVI th century gave place to those with which we are more 
fiEoniliar. 

If then the civil commotions of the xvii th century pro- 
duced such important changes in our language and pronun- 
ciation, what must we expect from the still longer and ruder 

16 



242 EFFECTS OF CIVIL WAKS ON SPEECH. Chap. IV. { !• 

distorbances of the xvtli centaiy, when the language was 
in a more inchoate stage, when the French element was 
fusing with the Saxon into the familiar alloy of the xvi th 
century, when no printing had as yet called forth an abund- 
ance of readers,^ so that the language altered organically 
from mouth to mouth untrammeled by literary fetters, and 
men of the north, middle, and south, jostling with each, wore 
down the angles of their dialectic differences, and gradually 
produced an English of England ? Practically we know 
that the xv th century was a period of great change in the 
whole character of our language ; the laist remnants of our 
inflexional system were abandoned, the sharp distinction 
between the "gentilmans" French and the " vplondische- 
mens" English, disappeared, and a "common dialect" was 
acknowledged by all writers.^ The distinction between the 
English of Chaucer, writing down to the close of the xivth 
century, and that of Spenser, the next great poet on our roll, 
who wrote after the country had well settled from its 
troubles, and printing had formed a reading public, is so 
sharp, that we seem to have fallen upon another language 
rather than upon a form of speech differing only by five 
generations. 

As then the language altered so markedly, must we not 
look for similar changes in the pronunciation P The exam- 
ple of the xvii th century irresistibly forces this conclusion 
upon us, and we also feel that if there had only been a 
succession of writers to chronicle them, we should have had 
a continual list of changes, comparable to those furnished 
while the xvii th passed its meridian and drew to its termi- 
nation, only more complex, more striking, more characteristic. 
Unfortunately we have no such writers, no such rules and 
lists to refer to; only a certainty of chaos and no guide. 
In shewing the development of the spellings ee^ ea (p. 77) 
and 00, oa (p. 96) in the xnth century, to mark distinc- 
tions in the sounds of long e and long o, familiar to the 
speaker, but iepiored by ihe writer, and, without such a 
guide, impossible to discriminate by an iffnorant reader, as 
one of the xix th century must naturally be in this re^>eot, 
we foreshadowed the confusion in the orthography of the 
latter end of the xvth and commencement of the xvith 

' Caxton set np bis press in 147t ; loauor, ad rostioos tantiun periinere 

the effect on the masses did not make yeiim intelligas ; nam mitionbns in- 

itself felt till the next centniT. tn^ijs & cnltitM enntritis, nnna est 

* Gill, after distin^uiBning the nbiqne sermo & sono, & significatn," 

Northern, Eastern, and W estern dia- and this he terms tbe '* dialectua oom- 

lects, saya **qnod hie de dialeetia mnnia." 



Chap. IV. j 1. THB XVTH AND XIV TH CENTURIES. 243 

oentuiy, a confusion which it is as yet impossible to dissipate. 
We caUi as in the estimate made at the end of the preceding 
chapter, be tolerably sure that a given written vowel or 
combination of vowels, was pronounced in one of two or three 
ways, but there does not appear to be, at present, any means 
of deciding which of those ways should be chosen in any 
particular case. After we have arrived at a more definite 
notion of the pronunciation of the xrv th century, the range 
of diversity will be somewhat narrowed, and by comparing 
the xrvth with the xvith century pronunciation of any 
word, noticing the direction of change, and, theoretically 
estimating the time necessary to effect it — an estimate which 
must be always hazardous — we may feel somewhat more 
confident. As however it is advisable in a preliminary 
investigation like the pres^it, to reduce theory to the nar- 
rowest possible limits, and to base results upon evidence, or 
a wide induction, I have thought it necessary to exclude the 
XV th century altogether from my researches, and to proceed 
by one step from the settled period of the xvith to the 
settled period of the xrv th century. In § 7 of this chapter, 
however, I shall indicate a rough practical method wnich 
may be adopted for reading works of the xvth century, 
founded upon the comparison already indicated. 

The manuscripts of the xrvth century poems, which the 
name of Chaucer points out as the principal subject of in- 
vestigation, though all belonging to the xvth century were 
fortunately written in its early part, and the Harleian MS. 
of the Canterbury Tales, No. 7334, which will be here 
generally followed, was probably written before the Hose 
troubles had commenced, so that although it labours under 
the disadvantage of being a generation after time,^ yet it 
was not subject to those more violent changes which raider 
the earlier printed editions of Caxton and others useless for 
our present purpose. This manuscript has, in addition to 
its careful execution, early date, and accessibility in the 
British Museum, the advantage of having been twice re- 
cently printed, by Mr. Wright,' and by Mr. Morris.' In 

^ Mr. Morris in bis Cliancer Ex- * Mr. Morris's edition forms the 

tracts, (see note 3, below), p. xliv, calls second and third Tolumes of his com- 

ihiBa**MS.,not later perhaps than the plete edition of Chauoer*s noetioal 

year of Chancer's death." works in six volmnes, pnblisned by 

' Mr. Wrighfs edition has been re- Bell and Daldy, London, 1866, at fiye 

printed in doable columns large octaro, shillings a Tolume, the only edition of 

■nd is published by Richanl Qriffin Chanoer*s works taken wholly from 

and Co., London and Glasgow, for half- MS. authority where MSS. exisi In 

a-crown. It is the most conTenient the Clarendon Press series Mr. Morris 

working edition. has reprinted the Prologue and two 



244 caxvcER and ooweb. 



Chap. IV. { 1. 



both editions the punctuation and capitals and tlie uses of 
th, f/, u, Vy are modem, and the contractions are all extended. 
In Mr. Morrk^s edition, the Lansdowne MS. 851 has been 
collated throughout, but every word not in the Harleian is 
printed in italics, and many final e's have been also added in 
italics when considered to be grammatically necessary.^ The 
long and tediously writt^i Cot^essio Amantia of Oower, has 
not be^i properly edited. Dr. Beinhold Pauli*s text, like 
Tjrrwhitt's Chaucer, ^diibits the text and orthography of no 
particular manuscript or time. But three good MSS. in 
the British Museum, and one at the Society of Antiquaries, 
are readily accessible, and Pauli's edition serves as a guide 
through the ponderous mass. The great regularity of 
Gower's verse and rhymes, renders his works a convenient 
supplement to Ohaucer^s, and I have found it necessary to 
make a complete examination of his rhvmes. Hie mode of 
referring to Chaucer's and Gower's works will be explained 
at the end of this section. 

The principles of the investigation on which I am about 
to enter, as to the soimds intended to be conveyed by the 
orthography used by the scribe of the Harleian MS. 7334 in 
particular, which may be assumed as the received Court pro- 
nunciation towards the close of the xrvth century, and 
will be bri^y termed the pronunciation of Chaucer, are the 
following. 

tales in a cheap form from ihiB MS. ample, in tha Seammde Nmnea DiU, 

This will be referred to as liis Chancer supposed to be ioid bj a womanj not 

Extracts. written by a man, we have — 

i In the nnmerons citations which I And though that I, unworthy §on0 

shall have to make I hare generally of Eve. 

followed Wright's edition, but in all Be syniul, yet accepte my bileye. 

important or doubtful cases I haye re- 11990. 

ferred to Morris's. One reason for Yet pray I you that nden that I 

using Wrighfs edition, besides con- ionte, 12006. 

T^ce, was that ^e lin^ axe num. ^^^^^ ^ the Sehipmannei TuU, sup. 

^^r^^^^^'^l^^}^^^^. pSedtobetoldbyTmL^in^e^ 
the CMi Tale o/Gcnu^n^vflich is 5f>iVee ^e find-^ ^ *^^ 

"''^'^"S^t^J ^1 «*; Theselyhousbondalgatmostepay, 

omitted by ^hitt as ceitemly not He mort •« clothe ^ good alri^^, 

Chaucers. ifr. ^o™ s editaon has ^ ^^ ^ ^^ worsch^Hchely; 

fresh s^ of numbers for erery pro- j^ ^^^^^ ^ daunoe>lily ; 

logue,tak, and part of talethoughout. ^^ ^^^^ ^/ ^^^ piiaventuiB, 

^•''♦lT!rtJi\'ri]^W OreUeswilnotsuchdil^endW 
S!^ w ^ ^Tk. tn^^S ^* ^y^^^ i* ^ wasteland i-los? 

l±TZ^e^i:is^h^& Or lene«. gold, thatTperflous. 14422 

no order as yet adopted is that into These expressions are in both cases ir- 

which Chaucer would have cast the reconcilable with the supposed speaker, 

poems had he liTed to give them the so that there must hare been some 

esteision originally designed. For ex- jolting or OTersight in the editing. 



Chap. IV. { 1. FIKOT PEINC3IPLE OP llfVBSTIGATION. 245 

1.) When few pecple can read, rhymes to be intelUffibk must 
be perfect. 

Owing probably to a change of sound which has not been accom- 
panied by a change of spelling, English poets of the ^Tinth and 
XIX th centnries tiie the liberty of considering such words as love 
nufvs, puU euUy eternity ly pass was, none stone, etc., to be rhymes, 
and readers are accustomed to pass them over as *' licenses," 
although they always produce a disagreable efltect upon children 
and unlettered adults. On the other hand words of which the 
final parts are pronounced almost identically, at any rate with a 
much nearer coincidence of sound than those cited above, are abso- 
lutely tabooed as rhymes. A xix th century poet would be much 
sooner allowed to rhyme whelk, with talk, than harm with psalm, 
or fork with hawk, although an unlettered Southern makes no 
dijSerence in the sound, and a lettered Southern rather imagines 
that he makes than really makes any distinction (p. 196). It is 
different with Northerns, Irish, or Scotch. It would be, perhaps, 
incorrect to push the theory too far, and say that in the very earliest 
attempts at rhyme an untutored audience would be satisfied with 
nothing less than that perfection which they could not possibly 
appreciate. But even then the general tendency becomes a suffi- 
cient guidor In finished and careful writers like Chaucer and 
GK)wer, such imperfections are not k priori likely to occur, and, as 
we shall see, are in fact unknown. 

The various kinds of rhyme which are actually found are as 
foUows. Let BAG, DEP represent two syllables, A, E being 
any vowels, and B, C ; D, F any consonants. Then if B = D but 
AC is not =3 EF, as in Bae, Bef, we have initial rhyme or alliiera* 
tion, which was used in the earliest form of EngHsh poetry, the 
Vision of William concerning Piera Plowman, 1362, being a com- 
paratively modem instance. !Next let A=E, but B-C not equal 
B-E, as hAc, dAf, the result is middle rhyme or assonance, which 
prevails in Spanish ballad poetry, where the same vowel occurs in 
the final syllable of alternate lines throughout the whole ballad, 
and the consonants must vary.* Thirdly let C = P but BA not = 
D£, 03 haC deC we have fnal rhyme, the English ^'rhymes to the 

> This is the theory; in practice bow- derecho, fecfao, medio, alojamimto, 
erer the difficulty of keeping the con- fecho, mensa^eros, stornunimto, man- 
sonants always distinct mis occasioned oebos, acumcf, arreoy redro, heredero, 
rhymes to be occasionally mixed up contesto, easamt^tos. In ^Despnes 
with assonances. If a oiphthong is que ret6 k Zamora;' among otners 
introduced in place of a simple Towel, occur : Lara^ haya, contrartos, causa, 
the assonance refers only to the ac- In * Considerando los condes,' amonj^^ 
eented Towel, e,ff, in Spanish at, au are others : rale, paces, batles. In * Monr 
assonant wid a, to, ua, and «t, eu with tos queredes, padre ;' Taiada, precioda, 
e, ie, ue. Thus in the Cid romance catga. See also the Cid baUads * Con 
*£hi las c6rte8 de Toledo,' the asso- el cuerpo que agoniza,* *Fablando 
nant words are : Sesto, sentimtimto, estaba en el claustro/ * Si atendeis ^ue 
mtifrto, d^udo, dello, proptt^sto, ptMsto, de los brazos,' * De palacio sale el Cid,' 
BUtflo, ast#nto, dentiffitos, rWno, teneos, * Desterrado estaba el Cid,' ' Aquese 
condeno, oonsejo, pUito, reto, escuderos, famoeo Cid,' * Non quisiera, yemos 



246 SEOom) primciple of ikvestioatiok. Chap. iv. } i. 

eye," like love^ mow; (the words was, pau form no rhyme at all). 
I am not aware that BA = DE, but C not = F, as BAe, BAf 
that is iMihU initial rhyme, or B-C = D-F but A not = E, as BaC, 
BeC, that is extreme rhyme, are recognized as rhymes under any 
system. But AC ==EF, and B not = D, as bAC, dAC or double 
final rhyme, is the ideal of a perfect rhyme in modem English and 
most European languages, and is the normal rhyme of Chaucer. 
Nevertheless modem French writers, as well as Chaucer, admit the 
identical rhyme BAC = DEF, that is BA C^ BA C, which under the 
name of r?tyme riche is constantly used in French versification. 
Either perfect rhyme bA C, dA C, or identical rhyme BA C, BA C, 
and even tiie assonance bAe, dAf would obviously serve to deter- 
mine either one of A and E &om a knowledge of the other. This 
leads to the second principle — 

2). When a word containing a knotcn rotcel sound rhymes 
with a word containing an unknown towel sound, the sound of 
the latter may generally be assumed to be the same as the former 
before xv th century. 

The difficulty consists in finding words whose vowel sounds are 
known. These are supplied in Chaucer from three sources, Latin, 
French, and those known sounds of the xnth century which we 
haVe a right to suppose, according to the results of the last chapter, 
came down to that period in an mialtered form. 

As regards the Latin words we may assume a Boman Catholic 
pronunciation, which will give a, e, i, o as certainly (a, e, i, o) 
long or short, and short u as (u). There may be a doubt whether 
long u had its general sound (uu), or its occasional Latin and 
general French soimd (yy). I am rather disposed to think that 
Chaucer, to whom French was familiar, used the French sound 
(yy) for Latin long ti. Even in 1580 we leam from Bullokar that 
Latin as pronounced in England did not possess the sounds of (ch, 
ii, uu, sh, dh, w, wh, j), so that long u was pronounced by him 
in Latin as in English and French, namely as (yy).^ We are 

mios,* ^Despnes que el Cid Campeador,' cent English they are aToided, or oocur 

' En Yalencia eotaba el Cid,' * De Cas- only from ignorance or carelessness, as 

tilla ran marchando/ &c In *Ciiando in we Nnrsery Rhyme ** Sit on a Wn 

d rejo T claro Apolo/ we find Idaiima And keep himself warm,'* and in the 

qnan Idtfma, assonancing with : estaba old catch ^* Cinnamon and ginger, nut- 

pasan. In the oldest Bomance poema, megs and cIotcs, And that gare me this 

assonances occur mixed with rhymes ; jolly red nose," or as Ben^ck (Moch 

the following are instances of diph- Ado^ v. 2) **can finde out no rime to 

thongal assonances: EtUalia (Diez: Ladie but babie, an innocent rime." 

Altrom. Sprachdenkmale 1846, p. 21]J In Ooethe's son^ in Faust: 

tost coist ▼. 19, Leodegar (Diez : Zwei "£s war einmalein Ednig 

Altrom. Gedichte, 1852, pp. 39-46) fiet Der hatt' einen ^oszen Floh, 

rei ffoffsa 9, mesfait ralat 15, advuat Den liebt' er gar nicht weniff, 

estrai 16, mors toit 20, preier deu 25 Als wie sein eignen Sohn, 

and 31, talier quen 27, deus eel 40. In the apparent assonance: Floh Sohn, may 

English poems of the xui th century, have only been a reminiscence of his old 

assonances are well marked, see Chap. Frankftirt pronunciation Sob for Sohn. 

y, \ 1, and especially No. 5, HareloK, ^ See the example of Bullokar's pho- 

and No. 6, King Horn. In more re- netic writing Chap. YIII, } 4. 



Chap. IV. § 1. SECOND PRINCIFLB OF INVESTIGATION. 247 

therefore hardly justified in awnmiTig a different pronunciation for 
the Latin long u in Chaucer's time, as the English long u had most 
probably the same sound. The case is different with respect to 
long f which was (ei) or (ai) in the xvi th century both in English 
and the English pronunciation of Latin, but was I believe (tt) in 
both during the xivth century. 

The French of the xivth century would, on this hypothesis, 
have the same set of vowels as the Latin. It would be useless 
attempting to distinguish in the French pronunciation of that time 
two sounds of e and two of o ; we caimot even be sure that they 
existed at that early period, as we know from Meigret that they did 
in the xvith century. The combination ou in French was in 
Chaucer's time (uu, u) and eu was probably (eu) or (ey) and oc- 
casionally (yy) as in the xvith century ; (ob) the modem sound of 
French eu appears not to have been developed in Chaucer's time, or 
Meigret would have been familiar with it. The French diphthongs 
at, au could not have differed from (ai, au) or (ai, ao), since we find 
them in the latter form in Meigret. The syllables an, in, on, un 
now pronounced as the nasal vowels (aA., eA, oa, oa), seem to have 
been received in England as (aan, aun, en, oon nun, un), without 
any nasality, and it is difficult, if not impossible, to discover any 
trace of vowel nasality in the notices which exist of early French 
pronunciation; Beza, 1584, the earliest I have found, seems to con- 
fuse (a) with (q). This tolerable certainty with regard to the sounds 
of French letters will be found extremely useful, expecially when it 
is remembered that Chaucer not only used French phrases, but in- 
troduced a large number of French words into his poetry, and as 
tiiese were familiar to the gentry in the pronunciation of the time, 
he could not have ventured to give them a different form in poetry 
intended especially for the delight of that gentry. We have modem 
examples of the same kind. Old French words we ruthlessly angli- 
cize ; we talk of a feat (f iit) of arms, as if it were feet, but we 
refuse the same soimd to fSte, We speak of recoup (rikuup-) and 
estate (est^*) but of coup d'Stat (kudeta) not (kuup destM*^. We 
do not scruple to say annoy (senoi*) but we try to say ennui (amyi), 
and even if the trial results in (onwii*), it has not the true Englif^ 
ring with it like (senoi*). The old words aid (eed) and camp 
(kaemp) will not allow us to call an aide de camp an {eed. di 
ktemp;, although our (d^-di-kaA) is not the French (eed dp kaA). 
Envirom, envelope are words in a transition state (envoi'renz, en'- 
vel(wp) and (on-viron, on-vilop) being both heard. Chignon and 
crinoline, constantly spoken of, remain French (shinjoA, krinolin) 
or as nearly so as the speaker can contrive. ^ 

For old English words we shall have to lay most stress on the 
pronunciations of those now written with ai, ea, and pronounced in 
the XVI th century as (ai, ee). We might safely assume that these 
soimds must have been the same in the older periods, but we shall 
be generally able to establish the feu^t by the otiier two sources. 

1 This sabjeet will have to be specially noticed in the next section, under I, T. 



248 TEUM) PBIUCIFLE OF DTVESTIGATIOlf. Chat. IV. f 1. 

In case of any marked peculiarity, the imperfectioii of manu- 
scripts will make it necessary not to draw c<mchisi<His from isolated 
examples, bnt to collect as many examples as possible, and to search 
as carefiilly for exceptions as for corrobonfliTe instances. The 
exceptions will then have to be separately examined, and carefdlly 
investigated to see whether they are mere mistakes of the scribe, 
which other known orthographies would explain, whether they are 
simply solecisms not borne out by other instances and therefore 
incorrigible errors, or whether they really indicate a double pro- 
nunciation. 

Having thus obtained an insight into the system of orthography 
used by the writer, having learned to estimate his various oontri- 
ranees to represent sound, at their true worth, we may venture to 
assume as a third principle, — 

3.) Orthographies shewn hy rhymes to have certain values, 
may he assumed to have those values even where they are not 
confirmed by rhymes. 

This assumes that the intention of the writer was to represent 
the sounds of the words, and that his variants arose, not from 
simple ignorance, but from the fact that he had to make his ortho- 
graphy, as he proceeded, after the usages which he had been taught 
m 'youth, and he naturally hesitated as to which usage was most 
appropriate at any time. Other variants of course occur from care- 
lessness, for which the scribe who writes many hours a day is 
scarcely to be blamed, — he that is without such carelessness among 
us, let him throw the first stone, I cannot.* That the writers 
anterior to printing had any intention of representing the histories 
of words by means of the orthography, in place of the mere sounds, 
it is impossible to believe. Not only do the variants we meet with 
exclude this notion, but there was the all-sufficient reason that they 
could not indicate what they did not know. New French won& 
would be written, of course, in the French way, but then this 
accorded so closely with the English way, that the scribe would 
hardly note the difference.* 

^ In reading orer the first draft of But natheles, pas over, this is no fon, 

this chapter, I found I had written I pray to Grod to save thi gentil eorpt, 
coHiequenee for confident^ to such utter 13718. 

destruction of the meaning of the sen- Where the /» is written altiionpfa not 

tence, that I had some difficulty in re- ^nounced, as in the French rashiom. 

coTering the original word. Similar Yet we have now hoth eoru and eormy 

examples will occur to every author, and it may have been mere accident 

and his own difficulties in correcting that the copyist wrote eorpt for eorMy 

his own errors will lead him to appre- just aa i^ bcK^ause eorpu is the more 

date the difficulty and danger of a usual word, we made it in writing 

critical restoration of any corrupt text. rhyme with remorse. In the middle 

* So far as I can recall, there are of a line we find tempt 12803. The 
Terr few decided examples of a French use of ^n in French words where we 
filing being retained which did not have reason to think only n was pro- 
represent the English sound. The nounoed in Fngliwh may oe also oon- 
omy example I haye noted where the sidered as a case in point, as digne 
rhyme pointed it oat, is 619, atteigne 8323. 



Chap. IV. § 1. NO RBAL FAULTY RHYMES IN CHAUCEB. 249 

These are the principles on which I shall endeavour to 
determine Chaucer s pronunciation. The question naturally 
arises, how far is the first and most important principle, to 
which the two others are only subsidiary, justified by the 
manuscripts P A carefiil examination of all the rhymes, in 
the 17368 lines which compose the Canterbuir Tales as 
exhibited in Wright's edition, has resulted in finding less 
than fifty rhymes m which the spelling indicates a difference 
of pronunciation. Of these a liu*ge number consist in one of 
the two words cited havin? a final e added or omitted, while 
there are constant examples in other places of an ortho- 
graphy which would render the rhyme perfect. 

The principal instances are : — ^bom bifome 1225, trace alias 1953, 
here messager 6142, eeke leek 6153, potestate estaat 7599, wolde 
brynge, for her lyvyng 8101, of hew, at newe 8253, withoute youre 
witjnge, in this thing, in your wirching 8368, mighte, to sight 
8556, solace alias 9149, atte laste, it cast 9827, est beste 10773, 
her witte, it 8303, rest, he keste 10663, hert smerte 10793, kepyng 
rynge 10965, hoste west 11007, ever dissevere 12802, Galiene 
Egipciene Arrabiene sleen 15822, matere gramer 14946, tresor 
l^abugodonosore 15629, gold olde 15645, may aye 17105, leye 
pray way 8753. 

These cases are often mere slips of the pen and can easily be 
corrected. The considerations in §^ 4 & 5, will be sufficient to 
explain them aU, and they must be all reckoned as errors of writing, 
not of rhyme. Poor Chaucer is very pathetic in reference to the 
damage done to his verse by scribes. In Troylus and Cryseyde 
5*74 he says, addressing his " litel boke," 

And for tber is so grete dyrersite 

In Englissh, and in writynge of onr tonge, 

So preye I to Grod, that non myswrite the 

Ne the mys-metere, for defante of tonge ! 

And red wher so thow be, or eUes songe, 

That tiion be nnderstonde, GK)d I beseche ! 

But yet to pnrpos of my rather speche. 

And what he suffered from the carelessness of scribes is well ex- 
hibited in his address to his own scrivener, which by the bye has 
itself been much injured in transcribing.! He is made to say : 6*307 

Adam Scriyener, if ever it the be£edl 

Boece or Troilus for to write new. 

Under tiiy long locks maist thon haye the scaU, 

But after my making thou write more trew! 

So oft a day I mote thy werke renew, 

It to correct and eke to mbbe and scrape ; 

And all is thorow thy necligence and rape. 

"Wonld that we had a text corrected by Chaucer's hand ! 

^ Mr. Morris had added seyeral $'% proyed" to suit the xvi th century 

required by the language. But the pronunciation. It is a wonder we do 

lines are quoted from Thynne's edition not find anew in the second line ; for 

of 1532, and were eyidenUy **im- in the second, long in the third, and 



250 NO REAL FAULTY BHTMBS IN CHAUCEB. Chap. IV. { 1. 

The cases in which short or long « rhyme with short or long 0^ 
may either belong to the class of accommodation rhymes, to be im- 
mediately noticed, or are explicable on the principles laid down in 
the next section under «. The following are the chief instances noted : 

geven lyven 917, list best 6819, 7567, list rest 9299, 16559, 
abrigge alegge 9531, swere hire = A^ 11101, 12076, pulpit iset 
13806, shitte = shut lette 14660. 

There remain only nine instances of other classes to be considex^d, 
and some of these are patent clerical errors. Thus since hye is con- 
stantly foimd for highj it follows that in : charged hem in hyghe^ 
some remedye 4629, the yA is a mere error of the writer. In : 
tyrant Buserus, serpent vmenem 15589, there is little doubt that 
-neuB is a clerical error for -motu^ which would give a perfect rhyme 
and be a correct form, as Mr. Morris reads and as is found in 
16063. The common yen for eyes, shows that the initial e^ in : 
thin outer eyen^ may well aspien 12426, is a mere slip of the pen. 
The rhymes : aUe thastates, of debates, desolat 4548 are manifestly 
clerical errors, and we have probably to read : thastat (= the 
estate) debat, desolat. The lines 

There saw he hartes with her homes hee 

The gretest that were ever seen with eye, 11503 

given in "Wright and Tyrwhitt (who has hie eie) are not in Morris, 
and correspond to a gap in the Harleian MS. K genuine, the 
rhyming words should clearly be the common pair hye ye or heighe 
eyghe. In : more and lasse, marquisesse 8816, lasse is evidently a 
clerical error for lesse, which is the reading of the MS. Dd. 4. 24, 
University Library, Cambridge. 

The rhyme: i-eased, y-preisedy 6511, is given as: y-eased 
y-presed 2*234 by Morris, and : esed ypreised by Tyrwhitt, but the 
Harl. 7334 reads: I eased, y pleased, and the Landsd. 851 esede 
yplesede.* These are usual rhymes. Lastly : jelousye me 1809, 

more in the fourth line are eyident in- cessarily added in moU, werke, eke; and 
sertions ; e final was omitted in befaUe, thorow should be thurgh. The lines 
tmotf scalUf trewe, rettewe, and onne- may then hare possibly sonnded thus : 
(Aadaam Skrtmeer*, if eer it dhee be£Ed*e 
^,ees* or Troo'ilns to rtrtt*te neu'e, 
Un'der dht lok'es maist dhu Han dhe skal*e 
But aft'er 'mii maak'iq- dhu nrii-te treu-e ! 
So oft a dai tV moot dht werk reneu-e, 
It to korekt* and eek to rub and skraa-pe, — 
And al is thurkirh *dhti neglidzhens* and raa*pe !) 
^ Wright says in a footnote : ** The y-pleased, for flattery and pleasing, 
Harl. MS. reads y^pUaudi but the named at first, are repeated 9b Jlattery 
reading I haTe adopted seems to give and a^tetMtoitctf,dimt}«M, afterwards. The 
the best sense." Tlie context as well whole passage, insertingthe bracketed 
as the rhyme declares in fevour of words, runs thus in the Harl. 7334 : — 
Some fayden [)>at] oure herte is moft I eafed 
Whan [^at] we ben y flaterid and y pleafsed 
He go)» ful neigh )>e foth I wil not lye 
A man fchal wynne ts beft wi^ flaterye 
And with attendaunce and [wibl bufyneflb 
Ben we y limed boj^e more ana lefle. 



Chip. IV. § I. KO EBAL FATJLTT BHTME8 IN CHAUCER. 



251 



is not even an approacli to rhjme and is manifestl j corrupt. I 
find on examination that all the other MSS. in the British Museum 
read j'olite, which is Tyrwhitt's reading, and is no douht correct. 
The rhyme: mercy sey 13308, will he specially examined in the 
next section, under I, when it will he shewn from other MSS. 
that the proper reading is : mercy sy. 

This examination is calculated to make us feel confident in the 
correctness of our first principle as applied to the Canterhury Tales. 
On extending the examination oVer the whole of Chaucer's poems, 
the following faulty rhymes are all that I have noted, which do not 
admit of an immediate correction. Except in certain pieces, of 
which the originals are therehy proved to he of very douhtftil 
authority, and of comparatively recent date, the faulty rhymes will 
he found exceedingly rare. The citations refer to the volume and 
page of Mr. Morris's edition, and the references to the original MSS. 
or editions, are all given. 



Vols. H. & HI. 

1. The Canterhury TaUsj from the 
Harl. MS. 7334, collated with Lans- 
downe MS. 861. After the prerious 
examination this may be said to hare 
no faulty rhymes. 

Vol. rV. 

2. 7^e (hurt of Love, pp. 1-60 : from 
Trin. ColL Cam. MS. R. lii 20 : write 
aright 1, discriTe high 4, wonderly 
signifie 4, decree jQ—eye 6, white de- 
lite hight 6, hie crye whye 10, 1 eepye 
je^eye 10, hie besyly je=eye 11, fan- 
tasye merily 16, ye =» eye pretily 16, white 
dehte sight 16, eschewe newe due 17, 
ben engyne 19, ye =eye wonderly hie 24, 
9eye=eye 27, shewe hewe 34, by nye = 
netir 34, modifie truly 36, aTOwe trowe 
=woo howe 42, I flye sodenly 46, 
trewe dewe porsne 48. 

3. The Forlement of Briddee, or th$ 
AasewMy of Fouies^ pp. 61-74, from 
Bodleian MS. Fairfax 16, colkted with 
Harleian MS. 7333, and Bodleian MS. 
Seld. B. 24. None. 

4. The Bohe of Ci^nde, Ood of Love^ 
or the Ouekow and the NightingaUt 
pp. 76-86, from Bodleian MS. Fair&x 
16, collated with Harl. MS. 7333, and 
Bodleian MS. Seld. B 24. None. 

6. The Flower and the Leaf pp. 87- 
107, from Speghf 8 edition of Cnaucer 
1697 and 1602, no manuscript copy 
being known : hie = high certainely 87, 
truly company 93, melody soothly 93, 
company lady richely 98, sautry craftely 
98, womanly daisie 99, company friendly 
103, properly company 103, chivalry 
worthy 104, victory mightily 104, com* 
pany humbly \XQ^haete 107. 



6. Troylus and Oryteyde, p. 108, 
from Harl. MS. 2280 collated with 
Harl. MSS. 1239, 2392, 3943, and 
Additional MS. 12044. Troye, joye, 
fro the 108, oontrarie debonaire staire 
116. 

Vol. V. 
Troylui and Oryteyde continued, pp. 
1-77. None. 

7. Chaueeree A, B, C, called Za 
Friere de Noetre Dame, pp. 78-86, from 
the Bodl. MS. Fairfax 16, collated 
with a MS. ia the Hunterian Museum, 
Olasgow, medycine resygne 81, this 
rhyme is probaoly correct. 

8. Chaueeree Dream^ pp. 86-164, 
from Speght's edition of Chaucer 1697 
and 1602, no manuscript copy being 
known : tesuB^eyen kene 87, was glasse 
88, paire here Tthis word seems to have 
been supplied by the editor) 88, hies 
high 8ie=sM 88, be companie 89-90, 
come some 92, undertaketh scapeth 96, 
grene jeue^eyen 96, place was 100, 
named attained 104, een=^Mi queen 
106, joyously harmony 107^ gentilnesse 
peace (r) 107, be companie 108, de- 
stroid conclude 108, vertuous use 
110, signe encline(P) 113, resigne 
mne(?) 120, found bond 126, re- 
member tender 129, fiftene, an even 
132, ligne compane 132, sidfety com- 
pany 133-4, greene eene=#ym 138, cry 
company 138, softely harmony 141, 
nine greene (P) 142, vertuouse use 143, 
company by 147. 

9. The Soke of the Bueheeee^ or the 
Dethe of Blanche^ pp. 166-196, from 
the Bodl MS. Fairfax. 16 : Pythagoras 
ehes 176. 



252 



NO RBAL FAULTY RHYMES IN CHAUCER. Chap. IV. } 1. 



10, Of Quint Anelyda and False 
AreyU. pp. 196-208, from the Bodl. 
MS. Faimx, 16. None. 

U. The House ^ Fame^ pp. 209-275, 
from the Bodl. MS. Fairfax, 16. None. 

12. The Leaende of Qoode Women^ 
pp. 276-361, from the BodL MS. Fair- 
fi^ 16, collated with Bodl. MS. Seld. 
B. 24, MSS. Karl. 9832, Addit 12S24 
^ritish Museum) and Grg. 4. 27, in the 
Universitjr Lihrary, Cambridge, pri- 
vately prmted by H. Bradflhaw, Cam- 
bridge, 1864. None. 

Vol. VI. 

18. The Romauni of the A>m, pp. 
1-234, frt)m the unique MS. in the 
Hunterian Museum, Glasgow : be 
nycetie 1, samet delit(P) 27, loreyea 
o&Teris 41, 1 maladie 67, hastily com- 
pany 67, generaly vilanye 67, worthy 
cortesie 68, more are 68, abrode for- 
weriede 78, annoy away (P) 82, escape 
make 84, joye conveyefP) 89, curtcsie 

fladly 91, folv utterly 97, laste barste 
7, foly hastily 99, 100, werye scye 99, 
redily maistrie 101, flaterie uttirly 103, 
affere debonaire 106, bothom salTacioun 
106, angerly rillanye 107, espie sikirlye 
116, foUlye jelousye 116-7, jelousie I 
1 19, 1 26, 1 lechery 1 19, bothoms sesouns 
122, high delyrerly 123, certeynly 
jelousie 123, glotouns bothoms 131, 
storme come 132, sikirly foly 136, 
bittirly foly 138, 1 curtesie 139, lorde 
rewarde 141, eeignorie I 142, eyer 
fer(?) 146, engendrure plesjiig 147, 
oompanye disrewlilye 149, servise preise 
^praise 161, worthy drurie 164, yice 
wys 164, to bye hastily 171, sy^part 
of the second syllable of fysie, foly 176, 
coyertly ipocrisie 186, company onterly 
192, whye tregetrie = trickery 194, com- 
panye I 209, mekely trechery 228, 
Bobrely, je yous die 226. 

14. Complaynte of a Loveres Lyfe, or 
the Complaint of the Black Knight, 
pp. 236-269, from the Bodl. MS. Fair- 
rax, 16: white bryght nyght 236, 
greyously petously mamdy 240, felyngly 
malady 242. 

16. The Complaynt of Mars and 
r#ww«, pp. 260-274, from the Bodl. 
MS. Fairfax, 16, collated with MS. 
Ff. 1, 6, in the Uniyersity Library, 
Cambridge, edition of H. Bradshaw, 
1864. None. 

16. A goodly Ballade of Chaucer, 



pp. 276-277, from Thynne's edition of 
1632 : supposeth ryseth 277. 

17. A rraise of Women, pp. 278- 
284, from Thynne's edition of 1632. 
None. 

18. The Oompleynte of the Dethe of 
Pile, pp. 286-286, from Bodl. MS. 
Fairfax, 16, collated with Harl. MS . 78. 
None. 

19. Ballade de Vilage 8auns Fleyn^ 
ture, pp. 289-292, from Bodl. MS. 
Fairfax, 16. None. 

20. Ballade sent to King Richard, 
pp. 292-293, from Bodl. MS. Fairf&x, 
16. None. 

21. The OompleynU of Chaucer to 
his Purse, p. 294, from BodL MS. 
Fairfax, 16, collated with Harl. MS. 
7333 and Bodl. Seld. B. 24. None. 

22. Oood Counseil of Chaucer, ^.296, 
from Bodl. MS. Fair&x, 16, collated 
with Cotton MS. Otho A. xviii., and 
MS. Gg. 4, 27, in Uniy. Lib. Cam. 
And A&. MS. 10340, see Athenaum, 
14 Sept. 1867, p. 333. None. 

23. Prosperity, p. 296, from Bodl. 
MS, Seld. B. 24. None. 

24. A Ballade, pp. 296-7, from 
Harl. MS. 7338. None. 

26. V Envoy de Chaucer a Seogan, 
pp. 297-8, from Bodl. MS. Fair&x, 
16. None. 

26. r Envoy de Chaucer a Bukton, 
pp. 299-300, from Bodl. MS. Fairfiix, 
16. None. 

27. JS^M Prima, pp. 300-302, from 
MS, Hh. 4. 12. 2, late MS. Moore 947, 
in the Uniy. Lib. Cam. None. 

28. Leaulte vault Riehesse, pp. 302- 
308, from Bodl. MS. Seld, B. 24. None. 

29. Proverbes of Chaucer, p. 303, 
from Bodl. MS. Fairfax, 16. None. 

80. Roundel, pp. 304-6, reprinted 
from Percy's iteliques of Ancient 
English Poetry. None. 

31. Virelai, pp. 306-6, from MS. 
R. iii. 20, Trin. Coll. Cam. : infortunate 
&te whate 306, hate desperate estate 
306, oertayn payn 306. 

32. Chaucer's Prophecy, p. 307, from 
Sir Harris Nicolas's edition of a MS. 
belonging to Mr. Singer. None. 

33. Chaucer's Words unto his own 
Scrivener, p. 807, from Thynne's 
edition, 1632. See supr^ p. 260, note. 
None. 

34. Orisoune to the Holy Virgin, 
pp. 808-812, from BodL MS. Seld. B. 
24 : honour cure 810. 



Chap. IV. f 1. NO REAL FAULTY RHYMES IN GOWER. 253 

In examining Gower's rhymes through the medium of Pauli's 
edition, I have put aside his orthography as of no value, and have 
reckoned as faulty rhymes only such as I could not immediately 
correct by means of the results obtained from an examination of 
Chaucer, and exhibited in the following sections. The citations 
refer to the volume and page of Pauli's edition. 

Vol. i. sely privete 225, er =i formerly ware 231, 

Vol. ii. named proclaimed 84, joy money 147, Troy monaie 188, 
nine peine 261, enemy michery 355, 

Vol. iii. accompteth amounteth 54, straught sought 374. 

Nine faulty rhymes out of more than 33000 verses would not be 
much. But in fact the editor Dr. Pauli, and not the author, is the 
person reaUy answerable for them, as the following examination 
will shew. 

The reading : sely privete i 225, is wrong on the face of it, for 
uly makes no sense ; the word is cdee or eeU as in Harl. 3490, 3869, 
7184, and Soc. Ant. MS. 134, meaning secret, a purely French 
word. The passage runs thus in Harl. 3869. 
As who faij'. I am so celee 
Ther mai no maimes prinete 
Ben heled half fo wel as myn. 

The reading: er ware i 231, is: er war in HarL 7184, but : ar 
war in Harl. 3490 and 3869, the passage in the last being 
Of snch enfamples as wer ar 
Him oghte be )>e more war. 

The rhyme: named proclaimed ii 84, is given: named, pro- 
clamed, by the three Harl. MSS, and: naimd proclaimd, by the 
Soc. Ant. MS. The first reading is evidently correct from the 
French proclatrUj and even Pauli in another place writes : named 
proclamed i 6. 

For: joy money ii 147, Troy monaie ii 188, the Harl. MS. 3869, 
reads : ioye monoie, Troie monoie. These rhymes will be farther 
considered in the next section under 01. 

The rhyme : nine peine ii 261, is written : nyne peyne in Harl. 
3869, but this is an evident slip for: nyne pyne, the reading of 
Harl. 3490 and 7184. 

For: enemy michery ii 355, both Harl. 3490 and Harl. 3869 
read : enemie micherie.* The enemy is Venus, and the word re- 
ceives the French feminine form, thus, according to HarL 3869 
For Venus which was enemie 
Of )>ilke loues micherie. 

The words : accompteth amounteth iii 54, are so spelled in the 
three Harl. MS., but as it is certain that the two French words 
from which they have been taken, had the same soimd, the rhyme 
was really perfect. This then is an example in Oower of the 
retention of a French spelling, which did not represent the English 
sound, supdi, p. 248, note 2. The orthography aecompt is even yet 

^ Harl. 7184 is illegible ; the word they mean it is hard to say ; probably 
is like enme, that is, there are fire we should restore missing letton thus : 
strokes between the two e's, and what en#mte. 



254 AOOOMMODATIOH KHTMSS. Chap. IV. ) 1. 

letamed in our written language, tliongh generaUj snpereeded by 
aecaunt. 

The words : stranght sou^t iii 374, were wnmgiy tranBcribed 
by Panli from the HarL 3490, which he professed to follow in this 
passage, and which reads : stranht caoht. 

This examination most be held to establish the correctness 
of the first principle for all the writings of Chancer and 
Gower. The exceptions are clearly due to some error of the 
editor or the scribe, or to certain varieties of pronunciation 
which will meet with an explanation hereafter. In Chaucer's 
time many words certainly existed in two or more forms 
either entirely different, as tho for those, say for saw, they 
for though, mo for more, etc., or only differing in a vowel as 
kess for kiss, lest for list lust, stree for straw, etc. We find 
instances of this double use even in prose, and in places 
where the use was optional,^ but it was evidently a most 
convenient instrument in the rhymester's hand, and Chaucer, 
who, notwithstanding the far greater facilities for rhyme at 
his time than at the present, seems to have been frequentlv 
"hard up,"* to judge by those numerous little tags which 
appear in his poetry and are absent from his prose, has ex- 
tensively availed himself of them. The following are a few 
examples of these Accommodation Rhymes, as I propose to t^m 
them: — 

rood upon a mere (= a mare), and a mellere 543, gan the child to 
blesse, gan it kesse 8428, holde champartye, may sche gye 1951, 
Then pray I the, to morwe with a spere That Arcita me thurg^ 
the herte here = bore 2257, unto oon of tho, moche care and wo 
2353, that on myn auter bren, that thou go hen^=^henee 2357, 
stree » straw three 2935, Paternoster soster « sister 3485, eampttme^ 

> A cook thei hadde with hem for 851, Harl. 1758, MS. Reg. 18. C. iL; 

the nones, and Sloane MSS. ) 686, 1686, all arne 

To boyle chiknet and the mtay m reading: eompame hiamg, urL 

bones 381. 7835ha8foiNrteffM^Harl.73^andM8. 

Hence nmrrybcnes for marrow bonet Re^. 17 D. zt. hare boUi com pome, 

fpossibly a reference to St. Mary lebon) which Wright prints compamt in one 

18 not a recent Tulgarism, but can boast word, and Morris misprints compaim, 

a high antiquity. and it should be observed that there ia 

« Compare Chaucer's own admission, • ^^***^,^. i^^'SoPf^?^* !" ?"i 
0.274 . ^ MS. 738 i, fol. 49 4, which looks at 

Andeketomehitisagretepenannoe, ^^ «^^* fj^^!*^ *?^ "''* 5^ 

Byth ryme in EnglS»h ^ iuS ^«" "^Jfi^ i°* ^t^ t^^ 

skarsete, ^^^^ '"■^^ throughout the MS. for tiie 

To folowe worde by woide the dot over an i; which is always rwre- 

curiosite sented, when written, as it would be in 

Of Graunson, floure of hem that ^J ^^^1 ^^^\^^i ^^ '^ 

maken in Fraunce. ^*1Li i ?*?* i^** /?v ^v^** -S? 

puzzled that he left out the line with 

s This reading isdoubtful. Lansdown eompame, altered the next line to 



Chap. IV. { 1. THE ORTHOGRAPHY SHEWS THE SOUND. 255 

=. eompagne blame 8709, beete ^heeU = bTiooU 3927, day lay = law 
4795, wirehe=:work chirche 9267, Eve pr&ve= prove 9203, feste 
meiU^tnosU 10613, est almest = almost 15168, aU = aUo falB 
4315, speche neehe = seke 4939, beech, theech := the ich = prosper I 
12856, sein = seen agayn 5177, time emenyme 6055, n^leyOj 
preye, seye 8704, therto, is do=:tdon 10313, glayre of an ^y, dey 
12734, seye aheye 13514, mystrist wist 13784, the mery orgon, in 
the chirche goon 16337. 

These instances, which are only a few out of many, are 
abniidantly sufficient to shew that the scribe was not content 
with continuing to write one form of a word, and allowing its 
di£Ferent soimds to be elicited from the rhyme (as we should 
now write a tear, to tear) but that he altered the spelling 
when he wished to shew a difference of sound. Hence 
although we have detected him tripping at times, from mere 
carelessness, we can feel confident that when varieties of 
spelling as eyen yen, hye hxhe, deyde dyde^ etc. constantly 
occur, they really indicate diiferent soimds, such as for ex- 
ample we shall learn to attribute to ey, y, ih, in other com- 
binations, so that the words just cited should be read (ai'en, 
ii'en, nii-e nikh'e, daid*e diid'e), and we are thus led to a 
corroboration of our third principle as well as of our first. 

Having thus established the trustworthiness of my instru- 
ment of mvestigation, not merely for the particular instance 
of this Harleian manuscript 7334, but for all good MSS. of 
the period, I shall proceed to apply it to discover a complete 
^stem of pronimciation, so as to allow us to declaim Chaucer's 
(Wterbury Tales as they might have been read during his 
lifetime, although doubtless with a modem accent which 
would have failed to satisfy the poet's ear. Still this pro- 
nunciation would have probably been perfectly intelligible, 
while our modem English method ot reading must have 
sounded as mere gibberish.^ 

rhyme, omitted the following which which has been scored ont, as it was 

was then withont a rhyme, and read : thns left without a rhyme, bat is per- 

Go from be wyndowe, Jacke fole fectly legible. 

fhee fayde 
I love bette ofer and eXlea I were to ^ This opinion I entertain so strongly, 
blame ^ that I retain its expression in the text, 
WeUe more ^an ^e by Jhefa and his notwithstanding that I have been in- 
dame formed, since it was written, that many 
8o lette me slepe a twenty denlweye. Early English scholars adopt systems 
The words : and his dame, in the last of pronunciation agreeing in the main 
line but one, are in another ink, and wim onr barbarous method of reading 
are apparently written over an oblitera- Latin and Greek. While this sheet 
tion. The last line was originally pre- was passing through the press I re- 
ceded by : ceived the following : ** As to O.E. and 
Go forth thy weye or elks I wolle A.S. Pronunciation, my scheme is icsi 
easte a stone, of shme^ 6see of JM, a=a Ot father^ 



256 



REFERENCES TO CHAUCER Ain> GOWER. Chap. IT. { 1. 



Mode of Reference to Chaucer and Qower. 

The lines of the Canterbury Tales will be cited by their numbers 
in Wright's single volume edition (p. 243 note), the number refers 
to the first line or word cited. The lines in any of Chaucer's other 
poetical works will be cited by the volume and page (not number of 
line) in which they occur in Morris's edition, a turned period being 
placed after the number of the volume ; thus, 4*87 means vol. 4, 
p. 87. As final words are usually cited, hardly any difficulty will 
be thus experienced in finding the passage. The list of Chaucer's 
poems on pp. 251-2, will show at once from the reference the par- 
ticular poem in which the passage occurs. The lines in (}ower will 
be cited by the volume and page in Fauli's edition,* the number 
of the volume being in small roman letters and the number of the 
pages following without an intervening comma, thus ii 84 is voL 2, 
p. 84. By ti^ means the form of the reference distinguishes 
the book cited, which will therefore not be named. 

As Mr. Morris's edition of the Canterbury Tales is not numbered 
throughout, and as Tyrwhitt's order of the Tales is not entirely the 
same as Wright's, the following comparison will be found usefuL 
The numbers refer to the volume and page in Morris and the line in 
Wright and Tyrwhitt. Occasionally some lines are inserted in one 
of these editions and omitted in the others, hence it will not always 
be possible to refer from one to the other by the numbers with 
certainty, but the difference is always very small, and if allowed for, 
will create no confusion. In order to correspond as far as possible 
with Tyrwhitt's system, Mr. Wright's first lino of a piece is not 
always numbered consecutively to the last line of the preceding 
piece, and his number 6440 is a misprint for 6439. The roman 
titles of the pieces in the following table follow Mr. Morris's edition ; 
the italic titles of the tales have been added by the author in ac- 
cordance with the text of the poems, for convenience of reference. 

Haskoxtt of the Eefekences to Morbis's, Wright's, Am) 
Ttbwhitt's Editioks of the Caittebbttrt Tales. 



Name of Piece. 



Morris. 



Wright. 



J^frwhitt. 



1. The Prologae 

2. The KnighteB Tale. Palamon and Arcite 
8. The Prologue of the Myller - - - - 

4. The Milleres Tale. NieholaSy Absolon, and 

the Garpenteres Wyf 

5. The Prologae of the Keeye . - - - 

6. The Beeyes Tale. The Miller of Tromp- 

yngtoun --------- 

7. The Cokea Prologue 



21 

227 

2-96 

2-98 
2120 

2*122 
2186 



1 

861 

8111 

3187 
8853 

3919 
4328 



1 

861 

8111 

3187 
8853 

8919 
4823 



lisso of bone^ &t=a of fate^ tl=ou of 
htmesy ftc,*' a scheme utterly irrecon- 
oilable with the direct evidence of the 
last chapter. See also Benjamin Thorpe 
on the pronunciation of Omnin, (Ana- 
lecta J^lo-Saxonica, 1846, 8yo, pre- 



&oe, p. zi) quoted below Chap. Y, { 2, 
No. 1. 

^ Confeasio Amantis of John Gower, 
edited and collated with the beat manu- 
scripts bj Dr. B«inhold Pauli, London, 
Bell and Daldy, 1857, 8yo, 8 yob. 



Chap. IV. } 1. RKFEBENCES TO CHAUCER AMD OOWBR. 



257 



NameofFiece, 

The Cokes Tale. Thi Prtniyt - . - 
The Cokes Tale of Oamelyn .... 
The Man of Lawes Prolo^e . - - - 
The Man of Lawes Tale. Qmstance . . 
The Prologne of the Wyf of Bathe - . 
The Wyf of Bathes Tale. The Knight 

mid the Foul Wfff. 

The Prologue of the Frere .... 
The Freres Tale. The Sompnour and the 

Bevyl 

The Sompnonres Prologae ..... 
The Sompnoures Tale. The Frvre and 

the Housbond man ...... 

The Clerk of Oxenfordes Prologne . . 
The Clerkes Tale. Grieildee. - . - 

Pars Secnnda 

Incipit tertia pars ...... 

Inoipit qnarta pars ...... 

Incipit pars qmnta - 

Parssexta ........ 

L*EnToye de Chancer ..... 
Proloeneofthe MarchaondesTale - - 
TheMarchanndes Tale. January and May 
The Sqnyeree Prologue . . . . - 
The S^njeres Tale. Camhynekan. . > 

Incipit secnnda WB ..... 
The Frankeleynes Frologe .... 
The Frankeleynes Tale. Arveragm and 

Lorygen ----...-- 
The Seconnde Nonnes Tale. Oeem$, . 
The Prologe of the Chanonnes Teman - 
The Chanonnes Temannes Tale. The 

False Chanoun and the Freet . - - 
The Dootonres Prologe ^i . . . . 
Tale of the Doctor of Phisik. Virgimue, 
The Prologue of the Pardoner - . - 
The Pardoneres Tale. The Thre Riot- 

touree -...--.--- 
The Sohipnuuines Prologue .... 
The Schipmannes Tale. Dan Johan and 

the Marehaunt ....... 

The PrioreBsee Prologe ..... 

The Prioresses Tale. The litel Olerpeom 

and the Jewes, ....... 

Prologe to Sire Thopas 

The Tale of Sir Thopas 

Prologe to Meliheus ...... 

The Tale of Melibens, proee . . . - 
The Prologe of the Monkes Tale ... 
The Menkes Tale. The harm of hem that 

stood in heigh degre ...... 

The Prologe of the Nonne Prestes Tale. 
The Nonne Prest his Tale. Chaunteelere, 
The Prologue of the Maundples Tale - 
The Maunciples Tale. Fhebtse and the 

While Crow 

The Prologe of the Persones Tale . . 
The Persones Tale, j^rvM ..... 



Morris. 


Wright, 1 


2136 


4363 


2138 




2170 


4421 


2173 


4519 


2*206 


6683 


2-232 


6439 


2-246 


6847 


2-246 


6881 


2-268 


7247 


2-259 


7291 


2-278 


7877 


2280 


7933 


2-284 


8073 


2-292 


8326 


2-297 


8486 


2-302 


8661 


2-307 


8816 


2-316 


9063 


2-317 


9089 


2-318 


9121 


2-354 


10-293 


2-366 


103-23 


2-366 


10661 


8-1 


10986 


8-2 


11041 


8-29 


11929 


3*46 


12482 


3-60 


12940 


376 


13410 


3-76 


13416 


3-86 


13702 


3-90 


13878 


3106 


14384 


8-107 


14412 


3-121 


14846 


3-122 


14864 


8-130 


16102 


3131 


16123 


3-138 


16327 


3139 




3198 


16376 


3-201 


16477 


3-227 


16263 


3-229 


16307 


3-249 


16933 


3262 


17037 


3-261 


17296 


3-263 





Tyrwhitt. 



8. 

9. 
10. 
11. 
12. 

13. 
14. 

16. 
16. 

17. 
18. 



19. 
20. 
21. 
22. 

23. 
24. 

26. 

26. 
27. 

28. 
29. 
30. 
31. 

32. 
33. 

S4. 

36. 

36. 
87. 
88. 
39. 
40. 
41. 

42. 
43. 
44. 

46. 

46. 
47. 



4363 

4421 
4619 
6683 

6439 
6847 

6883 
7247 

7291 

7877 

7933 

8073 

83-26 

8486 

8661 

8816 

9063 

9089 

9121 

10293 

10323 

10661 

10986 

11041 
16469 
16022 

16188 
11929 
11936 
12221 

12263 
12903 

12981 
13366 

13383 
18621 
13642 
13847 

13896 

13997 
14778 
14827 
16960 

17064 
17812 



17 



258 THE VOWELS — XIV TH CKNTURT. Chap. IV. § 2. 

§ 2.— The Voweh. 
Long and Short Vowels. 

The orthograpliic custom of the Germanic languages is to 
consider a mial vowel in an accented syllable long, and a 
vowel in a syllable closed by a consonant short. The physio- 
logical cause for the duplication of a consonant between two 
vowels to indicate the snortening of the first vowel has been 
already explained, p. 55. But long vowels also occur in 
syllables closed by a consonant, and here the writers have 
generally been put to great straits. Omnin by simply leav- 
mg the consonant single after a long vowel, and alwavs 
doubling it after a short one, escaped the difficulty. In the 
oldest Germanic monument, XJphilaB's Gospels, the Gbeek 
custom of using different signs for long and short (e, o) was 
usually followed, thus e at, o au were generally, = {ee e, oo o) 
Long i was represented by ei, following the Greek custom of 
pronouncing ev at that and the present time. Long a, u, 
were not distinguished from short, even if the real long 
(aa, uu) existed in Gothic.^ 

In Anglosaxon an accent is occasionally placed over the 
long vowel, but it is frequently omitted. Li modem high 
German and Dutch aa, ee, oo are often used for the long 
vowels, but this system of reduplication does not extend to 
long i and long u. When the • was not dotted, it would have 
been difficult to distinguish ii from u, and the combination 
uu mi^ht be read nu, un^ itn, mi, ini, which seems sufficiently 
to explain the non-use of reduplication to express these pro- 
longations. Still I find reduplication sufficiently distinct 
even in these cases, provided that the * is properly dotted, 
and hence I have employed it consistently in palaeotype. 

In Chaucer, as represented by our MS., reduplication is 
not unfrequently resorted to in the case of aa, ee, oo, but as 
the writer often neglects to mark the distinction (compare : 
in such a caas 657, arwes in a cas 2081), and sometimes 
employs ee where we expect to find a short vowel (as tceel for 
toel 2125), not much reliance can be placed upon this ortho- 
graphy. The fact, however, that both short ana long a, e, i, o 
rhyme with each other, but that long u and short u never 
rhyme, leads at once to the conclusion that the sounds of the 
long and short a, e, i, o differed only in quantity, but the 
sounds of long and short u differed also in quality. This 
general conclusion, will be abundantly confirm^ 

^ See an account of the yalnes of the Gothic letters, Chapter Y, { 4, No. 3. 



Chap. IV. j 2. A — XIV TH CBNTUBY, 259 



A — xivTH Cbntury. 

That long and short a could not be yery different finom (aa, a) 
we have already seen. It is not possible to distingaish after such 
a lapse of time between (a, a) and it is safer probably to consider 
(aa, a) as the real sounds. The effect of a preceding w does not 
appear to haye been felt ; that is, a in waSf warm would not haye 
differed from a in has^ harm. 

Latin ehtxbb. ... as assoillyng sayeth, a iignifieacit 663, where 
the old habit of reading the Latin termination -it as (-ith) may 
haye been alluded to ; ^ the Fsahn of Bayid, eor meum eruetavit 7515 ; 
Tet spak this child, when spreynde was the water, 
And song, alma redimptoria mater. 16061 

My teeme is alwaj oon, and ever was, 
£adix malorum at eupiditas, 13748 

On which was first i-writen a crowned A, 
And after that, Amor vineit omnia. 161 

These examples lead at once to the conclusion that a was called 
(aa), and that savethy David, water, was were pronounced (saayeth, 
I)aa*yid, waa'ter, was). Hence tdso the woi^ rhyming with wat 
will haye (-as) or (-aas), e.g. hire statue clothed was, arwes in a 
cat 2081, tlierto chosen was, such a caa9 2111, he walketh forth a 
jww, ther hir temple was 2219, this hors of hras, siege of Troye was 
10619, of Macedon he was, alat, such a caat, thyn sis fortune is 
tomed into an aas 16142, where «m, aa« are Hx, ace. These words 
giye the key to many others, thus : in this caas, of solas 799, and 
all words of that kind now usually spelled -oc^, as : paas Thomas 
827. We should also conclude that in: caught in his laoe, this 
trespace 1819, we ought to read laas^ trespas, as in: a dagger 
hang3rng on a laas 394 &c. 

I^^CH BHTMES. . . . haddc thei ben to blame, to be clept madams 
377, hadde hosen of the same, no wight clepe hir but madams 3953, 
fy for shame, sayde thus Madame 16377, it happed him par eaSy 
ther the poysoun was. 14300 

This last example confirms one of the Latin rhymes. In the 
other examples observe that Madame is a word which has preseryed 
its French sound (or what is meant to be such) down to the present 
day, and hence the rhymes with it are conclusiye. 

ShOBI AlTD LONG A BHYHES. 

A long snrcote of pers nppon he hadde 

And by his side he bar a msty bladde, 619 

Here, judging by the modem use. Blade is spelled Bladde simply 
to secure the rhyme, that is the long yowel is, for the occasion, 
treated as a short one. This of course could not be done if the 
quality of the yowels changed with the length, as in the present 
hadf Blade. In the following example — 

Each after other clad in clothes blaks 

But such a cry and such a woo they mais. 901 

1 See Salesbnry, mtk, Chap. YIII, { 1, under T. 



260 B, BE, EA, EO, OE, IB — XIV TH CENTURY. Chaf, IV. J 2. 

we have exactly the eonverse, the vowel in hlacke being lengthened 
to rhyme with make. This is also the case in : I may no lenger 
tarry, lady seinte Mary 7185, where the correct reading wonld 
probably be tarter Marie, In ag«. both hlad and hk^ had short 
vowels. 

The pronunciation of a in Chancer, which scarcely admitted of 
doubt before, is so clearly indicated by these three classes of ex- 
amples, that it is unnecessary to accumulate passages of the last 
kind, those cited in the first two cases are all that I have observed 
of that description in the Canterbury Tales. We must, there- 
fore, conclude that 

A in the xiv th century was always either (aa» a) or (aa, a). 



E, EE, EA, EO, OE, IE ~ xivth Century. 

Final e presents peculiar difficulties, and will therefore be treated 
separately in the fourth section of this chapter after the oiher 
vowels and the consonants have been fully considered. At present 
it may be assumed to be pronounced as the inflexional German final 
e (p. 196, note) in all cases where it ends a line or seems to be re- 
quired by the metre, and to be otherwise omitted in pronunciation, 
leaving me precise discrimination of these cases to fature investigation. 
The combination ee is used so frequently in place of e long, that 
it oannot be considered as a different letter. The combination ea 
is rare, but occurs most frequently in ease, pleaee, which are also 
found without a. ^, oe are occasionally used instead of #, when 
an e usurps ihe place of o, but there does not appear to have been 
any variation of sound. le and e alternate in some words, especially 
matiere matere, hiere here, but ie does not appear to have had any 
roeoial signification distinet fix>m e. The modem pronunciation of 
the 0, and the separation of its long sound into (ee, ii) which was 
oonfimed in t^e xvi th century, does not appear to have commenced. 
Laxik Ehtxes. — The only Latin word ending in e which con- 
cludes lines in Chaucer is henediciUy and this was almost always 
pronounced in three syllables, but whether (ben*diste) or (ben*aite, 
ben'ete),-— compare Seint Beneyt 173, and the modem Bennet — ^I 
am not able to say, I incline however to (ben*ete).^ The following 
are all the passages in which I have observed the occurrence of this 
word, and as m^ of them illustrate the sound of #, ee, it may be 
best to cite them all at length. 

The god of loye, a ! henedieite (5 sjllables) 
How mightj and how gret a lord u he. 1787 
TofighteforaladT; bentdieiUl 
It were a lusty dghte for to see. 81 17 

What? Absolon, what P Cristes swete tree! 
Why ryse ye so rathe f hiMdieUe, 3765 

1 Plrof. Child (infri, { 6, art 90) has fire svllahlee. The word has alwap 
suggests beneitk as the contraction and fire syllaoles in Gower. 
sni^peots a lacuna in ?. 1787, where it 



Chap. IV. § 2. B, EE, EA, BO, OB, IB — XIV TH CBNTURY. 261 

Et, heneditiUX than bad I foole i-sped. 4218 

What roune ye with hir maydenes ? btnediete, 

Sir olde lecchonr, let thi japes be. 5833 

And chyding wyres mak^i men to fle 

Out of here onghne bouB ; Aj benedieiis, 6861 

And sayd, O deere housbond, benedicite^ 

Fareth every knight with his wyf as ye. 6669 

I trowe thou hast som frere or prest with the. 

Who clappith ther P sayd this widow, bmediciU, 7106 

Til atte last he sayde, God yow se I 

This lord gan loke, and sayoe, Bensdieite, 7761 

A wyf ? a 1 seinte Mary, Unedidtt^ 

How might a man have eny adyersite 

ThathaUiawyfP 9211 

Unto onre oost, he seyde, BenedieiU I 

This thing is wonder merreylons to me. 12566 

I see wel that ye lemed men in lore 

Can mochel ^ood, by Goddes dignitee. 

The Person him answerd^ : BetudieiU ! 14389 

Of seinte Mary, henedieUe (3 syllables) 

What eyUth this loye at me 

To bynde me so sore f 16196 

8o hidons was the noyse, a hmedieiee I 
Certes he Jakke Straw, and his meyne, 
Ne maden schoutes neyer half so sohrille. 16879 

These examples establish the pronunciation of, in modem spelling, 
h$^ see, treSf bee, fiee^ ye, thee, me, as (nee, see, tree, be, flee, jee, 
dhee, mee), so far as the Towel is concerned. The other rhyming 
words, advereity, dignity, meny, will be considered under I, Y. The 
words thns established suffice to prove the pronunciation of many 
others and shew that the personal pronouns, he, she, we, ye, which 
were exceptionally pronounced with (ii) in the xn th century, (p. 
77), and the combination ee which was conflned to (ii) at the latter 
end of the same century (p. 79), had in Chaucer's tune, exohisiyely 
the sound of (ee). 
It might seem proper to reckon among these Latin rhymes 
Yet schal I sayen hir, and the, and me, 
Hastow nat herd how sayed was Koe, 3633 

But certeynlv no worde writeth he 
Of thilke wikked ensample of Oanaet. 4497 

But the preceding examples will also shew that Ifoe Canace must 

have had a flnal (ee). 
Frb27Ch bhtmes ... a sop in fyn clarre;\hB3i sittith he, 9717 away 

firo me, as well as thin parde 5891, the lasse light ^ari^.^ the thar 

not pleyne the 5917. 

For oosynage, and eek for heU cheer 

That he hath had fnl ofte time beer. 14820 

Long ajstd Shobt Bhthes . . . trapped in steel, dyapredt^^/ 2159, 
here the long pronunciation of wel is not noted as it is in 

Som wol been armed on here legges taeel, 

And haye an ax, and eek, a mace of Heel, 2126 

Thanked be fortune, and hire false wheel. 

That noon estat assnreth to ben weeL 927 

His eyen steep, and rollyin^ in his heed 

That Btemed as a fomeys of a leed, 201 



262 B, EB, BA, BO, OB, IE — XIV TH CENTUBT. Chap. IV. J 2, 

Here head, lead are now both short (ned, led). They may have 
been both long occasionally, as bread, dead spcUcd breed, deed 147. 
In : Jerusalem, a straunge etreem 465, both words may have been 
prononnced with (eem). But in : I holde my pees, al the prees 5096, 
we have either short and long rhyming, or else a short lengthened 
to rhyme with the long. In either case the sound of long e is 
shewn to be (ee). 

In the following examples we have words written in the xnth 
centoiy with ee and then prononnced (ii), rhyming with words then 
written ea and prononnced (ee). Those afterwaids written with ee 
will be italicised for distinction : fal lene, no calf y-«^iie 593, this 
cost (coast) so clene, that ther nys no ston j-eene 11307, his epeeMe, 
gladly teche 309, it needeth nat the teche, I the byeeehe 3599, 
wolde ban caught a eleep, Johan the clerk up leep 4225, in this 
drede, at thy grete neede 5077, at Ynsfeet, and of a man he eet 2049, 
a child that is i-bete, went he over the etrete 3757, in word and 
dede, repentaunce and drcde 1777, bodyes dede, of hemeys and of 
wede 1007, glorious for to eee, fletyng in the large see 1957, with 
leyghen etepe, noon in chepe 755. 

In the next examples we find ee rhyming with words which the 
Latin rhymes have established to be sounded with (ee): so as it 
semed me, of what degre 39, so ofte of his degre, hadde he be 55. 

The following are examples of words written with ee or simple e, 
which were afterwards written with ea. The ea words are 
italicised : humble cheer, ye schal heer 2221, piled berd, sore a/erd 
629, hem to wreke, scholde ^ke 963, breeth, heetk 5, as of the dttkj 
upon an keih 608, agreved with here, to a bere 2059, pite to heere^ 
Dyane gan appeere 2347, quod sche, in the salte eee 5527, in the 
Qreete eee, hadde he be 59, or foige or bete, to counterfete, 13432. 
These examples might be greatly multiplied. Ea occurs in: for 
eaee, nought diepiUaee 5709, sche wolde vertu pleaee, noon ydel earn 
8092, his spirit was at eaee, nothing may me diepleaee 9507. 

The use of eo and oe is shewn by the spellings : theof 13498, 
theres 13499; eorthe 8557, boef 9295, poepel 9241, pepul 2536, 
reproof 10078, 10137, preef 5829, reproeve 17002, repreve 6759, 
these latter words having generally simple e. 

The following shew the pronunciation of t# as (ee) : with evel 
preef^ a great meeehief b%29, al your $reef, an odious wkeechiefTnX^ 
a theef, wueeheef 1327, me repreve, we heliete 6759, ere that it was 
eve, made him bUete 4993, and eek a firere, disshe and «ui<Mr»6418, 
in this wuftere, quod the Frere 6421. 

The following are some instances of words now spelled with ie 
but apparently only written with e in Chaucer. See the table, 
p. 104. I sawh no man him grete, Osewald the Beeve 3857, be 
agreted, be rdeeted 4179, by youre leve; ye yow not greeve 7395, 
a frend, as a femd 5825, loth or Uef, an ivy leef 1839, longen unto 
eelde, mowen be tmweelde 3883, oon bar his ecJkeeld, in his hondes 
keeld 2895. We also find MerU 5978 fix- chereie, and w]M 15482 
for 9cheei, 



Chap. IV. { 2. EI BY, AI AT, AU AW — XIVTH CBNTTJBY. 263 

These rhymes lead irresistibly to the conclusion that the 
one general sound of e, ee, ea, eo, oe, ie ia Chaucer was (ee) 
long or (e) short, and they leave no room to conclude that e 
was ever pronounced as (i) except in the prefix be which 
we find written indifferently be bL The double forms lesse 
Uisse, left lafty seem however to indicate that e short wa 
occasionally pronounced as broadly as (a). In the xiiith 
century this was certainly dialectic, and the various forms 
may have remained in use during the xiv th. Perhaps the e 
was generally broad, as (b) rather than (e). In the same way 
we shall find % short to have been occasionally pronounced as 
(e), and this might be rather held to indicate the broader 
sound of (0> for t, or the finer sound of (e) for e. Such 
delicate distinctions, difficult to appreciate in actual livinc^ 
speech, are quite beyond our grasp at such a remote period 
and we must be content with one form (e) for the, possibly, 
three forms (^, e, b). It is indeed very probable that w 
three coexisted, and were not discriminated by the speakers 
themselves. Practically this is the case at present. 



EI ET, AI AT, Atr AW— xtvth Century. 

It is needless to shew that at, ay were generally (ai) and om, aw 
generally (au). They could not have had any other sound, as we 
saw at the conclusion of the last chapter, p. 238. But whether any 
distinction was made between ei and ai may be doubtful. In the 
greater part of modem Germany, ^, ai are both (ai), and they seem 
to have both had the same sotmd in Chaucer. Thus we have them 
rhyming together in 

That we with pitons hert unto yow playn$ 

And let yoore eeris my Tois no4 d\»dtyn$» 7973 

But playne is written pleyne in 

He was ont cast to wo and into peyne. 

glotony, wel ought ns on the pl^ne, 18926 

Again: I wot it well certeyn, I dar well sayn 8185, may be 
compared with : myn harmes not bewreye, I may not seye 2231. 
In 13336, 13511 thay occurs for they. And generally tiie same 
words constantly vary from ey to ay, and conversely, so that the 
phonetic identity of ey, ay is the only legitimate inference. Thus : 
for sche was feir, to maken hir his heir 3975, what so men jape or 
pleye, holden the righte toeye 9263, companyes tweye, that cowthe 
#^0 = 8ay 2591, 

Kepeth this child, al he it fool or fair, ...« 
Cnst whan him lust may sende me an hair 
More a^;reahle than this. 5184 

"Well wiste he hv the drought, and by the rei^, 
The yeeldyng or his seed and of his §ir$jfn, 597 



264 EI BY, AI AT, AU AW XIV TH CENTURY. Chap. IV. } 2. 

And Venus fayliih wher Mercury ii rented 

Therfor no womman of clerkes is preised. 6287 

Ben thaff ns seely men for to detetyv^ 

And from a soth ener wd tkmif weyve^ikey wai90, 10297 

The assumption that ai was pronounced as (ai) is confirmed by 
the French riiyme : how lasteth hir vitattte^ no wight but Crist 
Bauwifaile^ it was a gret mervaile 4919, and the Latin rhyme, as all 
rhymes with Scripture names must be considered: the mount 
of Synay, fasting many a day, 7469. 

It would appear that (ai) was sometimes lengthened and divided 
into (aa,i) forming a disyllabic. Thus seynt is a monosyllable 
(saint) in 

For by that lord that cleped is $eyn( Jame. 4262 

But when prefixed to the same name it becomes a dissyllable 
(saa,int) in 

Wei be we met, by God and aemt Jame. 7025 
Where, however, by may have been omitted after and. On the same 
principle I would explain 

Hire grettesfc ooth nas but by aeynt Loy. 120 

That is (saa.int Luu'i), St. Louis, as Meigret writes his first name 
Lays in his TraiU touohant U common vsage^ etc., but Lou\» in his 
phonetic French Grammar. Prof. Child would read othe^ but this 
form is not well established. 

I bad the printe of »efmt Venus sel. 6186 

lliat teynt Peter hadde, when that be wente. 699 

So alsoybtr in 

To lede bim forth into a/atV mede. 7621 

And maUtrye in 

Bacbns bad of hir montii no mairirys. 13472 

In the four last cases there is no simple means of altering 
the reading,^ and on repeating the lines it will be readily perceived 
that this pronunciation is not at all strained, and immediately solves 
their metrical difficulties. In the Prisoner's Prayer, Chap. Y, § 1, 
No. 2, it will be seen that the French diphthongs in: ueine 17, 
mayn 36, are given to two musical notes each, though they are 
frequently given to single notes, and other examples from Korman 
poems will be found near the end of Chap. Y, § 1, No. 3. 

As compared with Salesbury's observation that a in atih$ is 
'' thought to decline toward the sound of the diphthong ai" it is 
interesting to note aisshen 3880, oMsohes 12735. Four words now 
written ai were either always or occasionally written with «, Be and 
hence pronounced (ee). They are tuBtain, hair, day, strain, and I 
have not observed more. Thus for sustain : to sttsteene, bright and 

' And sayede twyes, Seynt Mary^ ! In : a goune cloth, by God, by seini 

Thou arte noyonse for to cary^. 5*226 Johan. 7833 

we should probably read : Seynie Marye. the word wtd has been probably omitted 

Compare before the second by. 
Tweli pens P quod sche, now lady 
seinte Mary«. 7186 



CUAT, IV. { 2. EI ET, AI AY, AU AW — XIV TH CENTURY. 265 

Bchene 1995, ache myhte nouht hir Hutene, dt adonn upon the 
grene 11173, o blisfol queene, in my wyt ausUene 14892, 
Then nys ther noon oomparisoon bitwene 
Thy wo, and any woo may man mstens, 5265 

For hair (ags. haer) : a tufte of heres^ a souwes eeres 557, h&er 
677, keres 1390, kempt bis heerey a trewe love he boere 3691, myn 
olde yeeres, so moulyd as myn heeres 3867, Sampson left his heri»f 
knt hem with hir schens 6303, under his lange herisy tuo asses 
eeiis 6535. On the other hand as we have seen that heir is spelled 
heir and hair. But we have heire 12061, for hair shirt. 

For slai/ (ags. slan, slean, sleahan) : or elles sle his make 2558, 
the freisshe beaute eleeth me sodeynly 1120, for curs wol slee 663, 
hir self to «^, as it thenketh me 1 1709. 

The iUtr of himself yet sauf h I there, 

His herte-blood hath bathea al his h^rt. 2007 

For strain^ in the sense of race (which is derived from ags. streon, 
streonan, strynan, and has nothing to do with the o^er word 
itrain), we have 

For God it woot, that childer ofte been 

Unlik her worthy eldris hem bifore : 

Bonnte oometh al of Ood^ nought of the strten 

Of which thay been engendrid and i-bore. 8031 

Strain, hair, slay, are clearly not proper instances of ai pronounced 
as (ee), but rather examples of a subsequently inserted •'. But 
iusteene would have naturally appeared as susteigne, as we have 
atteigns 8323. 

Connected with this is the converse use of (ai) for (ee) or (e), 
thus: fleissh 147 ior Jlessh, have ye not seye 5065 for seen; and 
%0€^ke ben the oxen 889, this weyhe woman 5352, to arreyse, at eyse 
7683 for ease. That the word was then really pronounced (aiz'e) 
and not (eez*e), appears not only frx>m this rhyme, but from the 
following lines in Gower, where Pauli incorrectly prints $se : the 
orthography is that of the Soc. Ant. MS. 134 : 

Whyche hadde be femant to Thaife 

So bat fehe was ^e worfe at ayfe. iii 820 

Anmere)> and faye)> my name is Tha^e 

That was fom tyme wel at ayfe. lii 332 

The use of fleissh, tcayk ' is not so easy to explain, but eyse, fireissh 
367, 1120, hurgeys 371, paleys 2201, 2697, 9585, 10374, hertusys 
2498, hartleys 3760 are rather direct representatives of ai, oi in 
French, the latter being changed into ei in Norman French, so we 
have in the rhymes to the two last instances palfreys 2497, Oemeys 
3759 and deys 9585 = dais. T\ns is an argument in favour of the 
Norman pronunciation (ai) for ei. 

We find say for saw 8543, 9810, 13642, 16600 and elsewhere, 
and in the same way we now have a saw for a saying. 

The sound of au is of course generally (au), as is confirmed by 

^ It is remarkable that both words Compare Jleyes Bel. Ant i, 22, Jleits, 
hare $i in Modem Gennan./Crt«e^ toeich, ib. 57, and veikr in Icelandic. 



266 O, 00, OA — XIV TH CENTUBY. Chap. IV. { 2. 

the French rhyme : to make hir alliaimce, him happede par chaunee 
14020, but the name of St. Paul, especially when applied to the 
cathedral church, was pronounced with (oou) as we have found for 
this particular case in the xn th century (p. 145). The orthography 
by seint Paules belle 16266 is yerj unusuil and probably erroneous, 
we have : seynte PouUsj chauntene for soules 511, in Petres wordes 
and in PouleSy cristen mennes soules 7401, with Fowles wyndowes 
corven on his schoos 3318, after the text of Crist, and Poivel and 
Jon 7229, 

Of this matier, Foul, wel camtow trete. 
Mete unto wombe, and wombe unto mete, 
Schal God destroyen bothe, as Fii>wel saith. 13938 

The most singular interchange, however, is that of (au) with (ee). 
GiU complained of his Mopsae saying (leen) for (laun) (p. 91), but 
200 years before that time we find : for keftd is with force force to 
schowve 3910, in mullok or in stree, so fare we 3871, of the stree, 
of the realite 5121 and elsewhere. The two forms straw ^ sire are 
due of course to ags. straw^ strea. But lee must be a form oi lay, 
OB ese of ayse. The form lay for law occurs, for the rhyme, in : 
on a day, that sche wold reney hir lay 4795, and must be due to 
the French /oi, Uiy while law must come from the ags. lah. The 
interchange was therefore not phonetic, but etymologic. 

Hence we conclude that EI, AI were always (ai), and AU 
was always (au) in the xiv th century. 

O, 00, OA — XIV TH Century. 

long and oo must be considered as the same letter in Chaucer. 
The regular sound was (oo), as shewn by the Latin rhyme, 
For though a widewe hadde but oo schoo 
So plesaunt was his In principio 
Tet wolde he haue a ferthing or he wente, 253 

whether the sound was {oo) or (oo) is of course open to the same 
difficulty as in the xvi th century, but the perfect agreement of long 
and short vowels, turns the balance strongly in favour of (oo), 
which seems to have been the original Latin sound. 

The sound of Bcho gives that oi do hy\ may nought do, is not 
worth a scho 6289, which gives to, iherto, a hoo, by : oon hole to 
sterte to, than is al i-do 6155, he addid yit therto, what schulde 
yren doo 501, 

An herowd on a skaffold made a hoo 

Til al the noyse of the pepul was i-doo. 2585 

After this we may feel tolerably certain of the sound of long o 
and its identity with that of oo = (oo). The following examples 
are however worth attention : of symony also, did he grettest woo 
6892, never the mo, tel me who 6273, for he saith us soth, that so 
doth 6523, ever in oon, thought anoon 1773, as stille as stoon, 
for ther ascapith noon, as we £iowe everychon 7997, al ther sche 
goth, I have no thrifty cloth 5819, a fan right large and brood, 



Chap. IV. j 2. O, OQ, OA — XIV TH CENTUET. 267 

lay his jolly schood 8315, his eyghen grey as goos, corven on 
his schoos 3317, God amend it soone, ye wot what is to doone 7775, 
whan he awook, he the lettre took 5226, 

Tel, quod the lord, and thon schalt have anoon 
A gonne cloth, hj Ood, hy seint Johan.^ 7833 
And every statute couthe ne pleyn hy roote 
He rood out hoomly in a mealed coote. 329 
Wei may men knowe, hut it he a fool, 
That every partye dyryveth from his hool. 3007 

As then oo seems to be always (oo) we must assume wood := mad, 
often spelt w^orf, toodef to have had (oo) and hence conclude the 
same of bloody stood, good from the rhymes: upon a carte stood, 
grym as we were wood 2043, jalous and eke wood, wel neyh al the 
blood 1331, that is so good, of blood 2565. The change of long o 
into (uu), developed in the xvith century, had therefore not yet 
occurred. 

But did short o always represent (o) ? Generally it did so, but 
there must have been exceptions. It would be difficult to imagine 
an interregnum of (o) between two reigns of (u). It will be shewn 
soon that ou represented long (uu) and but rarely short (u) for 
which certainly u was available, but nevertheless o seems to have 
been often employed. Thus we have 

Outher for ye han kept your honeste, 

Othtr elles for ye hau falle in frelete. 13492 

So that in two consecutive lines oti, o are used in the same word ; 
in the Knightes Tale Falamon seems to have had either (o) or (u) 
to suit the rhyme, as : oon, Palamon 1015, doun, Palamon 1072, 
prisoun, Palamon 1453, 1469, Palamon, opynyoun 1481, while we 
have the orthography: doun, Palamoun 1517. Again : he might 
not lenger B<^'oumey homward most he tome 6569, had I not done 
a frendes torn to the 14230, for fere of beres or of holes blake = 
hdls 16421, i-lyk to the stremes of homed hete = humed 13453, 
hokeler 1 12, asonder, thonder 493. 

The fact is that short (u) is comparatively rarely represented by 
t^, perhaps among other reasons because short u was as we shall see, 
frequently called {%) or (e), as in our modem words husy, hury, so 
that except in certain very well known words there might be more 
error induced by writing u than by writing o. Under these cir- 
cumstances I have been compelled to adopt a theory, indicated at 
the commencement of the last paragraph, and I consider short o to be 
(u) in all those words where it replaces a former w, and was in the 
XVI th century pronounced (u) ; that is, as a practical rule where it 
is now called (a). There will be exceptions to this practical rule, 
thus word is now (waid) and Bullokar makes it (wurd) but in 
Chaucer it was (woord) as we see from 

But al for nought, he herde nat o word^ 

An hole he foira right lowe upon the boord. 3439 

There might seem to have been another sound of short o in a few 
^ Johan, written Jon, 7229, is regularly a monosyllable. 



268 01, OY — XIV TH CENTURY. Chap. IV. § 2. 

words, compare the hbos : hadde we on honde, my fourth houshonds 
6033, to withatandey thral and honde 7241, in londe, to telle it wol 
I fonde 15295, as liked Cristes sonde, ap{»x>ched unto hndu 5322.* 
In comparing tliis o in place of in land^ withstand^ hvshand, with 
oa in loande in the Proclamation of Henry III., and with the inter- 
change of a and in northern and southern dialects, the use of not 
for not frequently by Chaucer, and later by Palsgrave, it was easy 
to imagine the pronunciation (a) or (ah) as an intermediate sound, 
which the scribe did not know whether to represent by or a. 
Thus Englishmen now confuse Scotch (man) or (mahn), and Irish 
(sahr) with their (mon, sor), and write them mon, sorr = man, sir. 
But this conjecture will not explain such rhymes as the above. As 
handef sonde must have had (o) and howhonde ought to have it, we 
must read (o) in londe, stonde, and in stronde and elsewhere, compare : 
straunge strondes, sondry londes 13. 

I have not noted any instance of the combination oa, but some 
cases may have escaped me. The modem oa is replaced regularly 
by 00 or as: goot 690, hoot 9298, hrode 2919, loode 2920, ook 
10473 for goaty boat, broad, load, oak. 

The conclusion seems to be that long or 00 in Chaucer 
was (00), that short was generally (o), but occasionally (u), 
the latter cases being those in which there was a previous 
Anglo-Saxon (u), and a xvith century (u), now become (e). 



01, OT — XIV TH Century. 

This is a rare diphthong and its sound cannot be satisfactorily 
established by the rhyme. If the identification of Zoy 120 with 
Logs, that is, Louis, be correct, then: fill symple and coy, by seint 
Loy 119, should give (kuui) as the sound of cog. In my article on 
the Diphthong OY (Trans, of PMl. Soc, 1867, Supp. part I.), I 
have given reasons for supposing (ui) or (uui) to have been the 
original sound of this diphthong, which we have seen was fre- 
quently so pronounced in the xvi th century. Thus Hart gives the 
sound (buee) for bog (p. 133), and if we interpret this as (bui) or 
(buui), the above pronunciation of Log is confirmed by the rhyme. 
That was wel twight, mjrn oughne lyard, boy, 
I pray God save tny body and seint Loy. 7143 

The word boist 13722 is merely the French boiste now boite, box, 
which historically would have the sound (buiste), and in our bushel, 
Fr. boisseau, which Chaucer writes buisshel 4310, we have preserved 
the (u) of the original. The two spellings boist, buisshel seem to 
shew two ways of writing the same sound, the writer, accustomed to 
use either or m for short (u) hesitating between them. This is still 
more plainly shewn by the double orthography of the word destrog. 
It doth no good, to my wit, but anoyeth 
See ye nouht, lord, how mankind it deUrogeth f 11187 

^ 8<md$ 5245 rhymes with grotmde, indicating the pronnnciation (sund*e}. 



Chap. IV. § 2. 01, OY — XTVTH C3ENTURT. 269 

Where anoyeth most probably had the old sonnd (annui^eth), and 
deUroyeth is used to make the spelling agree with its rhyming word. 
Bat where this motive did not act we find uy written, as 
That hath deitruptd wel neyh al the blood. 1032 
How he destruyed the ryoer of Oysen. 7662 

And in the prose tale of Melibeus (Wright's ed., p. 159, col. 2, 
L 32, Morris's ed. 3-172, 1. 13) : by vengeaimce tak]^ge be wikked 
men destruyed. 

The words: fruit destroy i 137 are written in Harl. 3869 and 
3490 fruit destruie, in Harl. 7184 fruit destroie, and in Soc. Ant. 
MS. 134, frute destriue, the last being clearly a mistake for destruie. 
It cannot be supposed that the combination ui was pronounced in 
the same way in both words. The last is the more common 
spelling oi fruity viz. frute = (fryyt). The same MS8. in the same 
order read in i 140 despuiled, oespoiled, despuiled, despuiled. 
From these readings, it would seem to follow that (ui) was the sound 
meant, but that the writing oy was preferred, short o having as we 
have seen (p. 267), very commonly the sound (u) or (w), because 
ui rather suggested the sound (yy). Probably oui was not employed, 
because ou rather suggested tne long sound (uu).^ Thus achyeth 
anoyeth 4*68, enchned annoied ii 47, must refer to a French aelouiy 
eneUmi, and hence ought to have been written oui and to have 
had the sound (ui), which they therefore lead us to infer in armoy. 
See also the sound of (ui) cropping up even in the xnth century 
(pp. 131 sqq.). But this was probably not the onlv sound of words 
generally written oy in the xnrth century. The French oi was as 
we have seen (p. 130), pronounced (ob, ue) with the stress on the 
second element, which was generally converted into English as (ue, 
ui) with the stress on the first element, but Gtower probably retained 
the French pronunciation when he invented the rhymes: ioye 
monoie ii 147, Troie monoie ii 188, (p. 253). On the other hand, 
the Norman ei, pronounced originally perhaps (ei), but, on account 
of its interchange with ai in the xivth century, pronounced in the 
same way (ai) at that time, see Chap. Y, § 1, No. 3, regularly 
replaced the French o», so that many French oi appear as ^ in 
Chaucer. In: Gregois vois iii 188, the o« was probably the usual 
(ui), just as in : chois vois ii 181, 206. But Harl. 3869 writes: 
gregeis curteis ii 238, and considering that the latter was the usual 
form of this word, the reading is probably correct. If any depend- 
enoe can be placed on the readings of the Huntenan MS. oi the 
Bomaunt of the Rose (p. 252), this must be the explanation of : 
joynt queynt 6*62-3, annoy away 6*82, joye conveye 6-89, but the 
passages are probably corrupt.* In the Canterbury Tales there 

^ It mi^ht hare suggested a division et ne sont pas, par consequent, nne 

of the diphthong into two syllables, triphthongue." 
Beza (livet, p. 523) says of oui : ' It most not be assumed that this 

** Qnand ees trots lettres sont placto is the origin of (oi) in a well known 

derant //, Ft sert senlement k pr^enir vnlgarism, as (boil, point, dshdnt) for 

le lectenr qn'il faut monillec U\ par- boil^ pointy joint, becaose this was a 

tout ailleors M«i fonnent denx syUaoee, mere regular xvnth eentoiy trans- 



270 I, Y — XIVTH CENTURY. Chap. IV. { 2. 

seems to be no iostance where (ui, ne) might not be used, with 
the stress on the first element, and the modem English (Ai, oi) is so 
limited geographically, and appears to be so modem, that it wonld 
be merely truckling to present habits to introduce it into Chauc^. 

We must therefore conclude that the most general pro- 
nunciation of 01 in the xiv th Century, was (ui). 



I, T — XIVTH Century. 

It will probably prove the most difficult conclusion for the 
reader to admit, that long t in Chaucer's time had not that diph- 
thongal sound (ai) with which we are so familiar, and which we 
have since the xvi th century at least, recklessly introduced into our 
pronunciation of Latin and Greek, and into our method of reading 
Anglo-Saxon and Gothic. The belief that long t was anciently (oi) 
or (ai) is not simply shared by those only acquainted with modem 
English, it is adopted by men eminent for their knowledge of the 
older languages. To assert, then, that so recently as the xivth 
century this sound (oi), now so uniyersal, in different modifications, 
was never associated with the letter », is a thesis which will require 
ample justification. As regards the present writer it may be men- 
tioned that before he began his researches he simply wondered 
whether it was possible to establish any conclusion whatever, that 
he inclined to the supposition of (ai) or (ai),* and that, even after 
having established the general pronunciation (ii), misled as he now 
believes by an isolated instance, he for a long time imagined that 
he could point to a whole class of words in which long t had the 
sound of (ai). A rhyme in Gower first induced him to reconsider 
this conclusion, and he then undertook the examination of the 
rhymes in the whole poetical works of Chaucer, in addition to the 
Canterbury Tales, and in Gower's Canfessio Amantisy with the sole 
view of discovering something which might help to decide the point, 
and he examined or caused to be examined all the available manus- 
cripts containing the passage in question,' seventeen in number, to 
see whether there were not sufficient orthographic variants to render 
it doubtfuL He also made inquiries into various existing dialectic 

formatioii of the xti th centiaj (bnil, the pronmiciation of the account of the 

puint, dzhuint), see p. 134. The Dor- Prioresse, 117-162 in my £ts$HtuiU 

setehire (pwaint, bwail) etc. is probably of J^honettes, 1848, 1 find (ai) riyen in : 

a descendmt of (puttnt, hnttl) etc. the smiling, by, wiped, eyen, I, and (ii) in : 



stress falling on the second element, prioresse, hire, Eglentine, serrice, de- 

which then became transformed from vine, swetely etc., Paris, curtesie, 

(it) to (ei, ai, ai) as almost all other digne, tretis. 
accented long (it) m that dialect, eheentf 

the$H for ehimey thine being the only ' Qnoth the chanonn, and £eur wel, 

exceptions noted by Mr. Barnes {Foems praunt mercy, 

of Mural Life^ 1848, p. 28. He went his way, and never the priest 

himsey. 

^ In a theoretical attempt to assign After this day, 18308. 



Chap. IV. § 2. I, Y — XIVTH CENTUBY. 271 

pronunciations, of long i and the prononn /in England and Scotland* 
to see what corroboration there was for any theory on the subject. 
These various researches have led to one conclusion, already antici- 
pated as the only possible explanation of Palsgrave's and Bullokar's 
otherwise enigmatical treatment of the letter t (pp. 110, 114), 
namely that 

The vowel t in the xrv th century was probably called (m) 
when long, and (i) when short. 

The sounds of (tV, f) as distinguished £rom (ii, i), the true Italian 
vowels, have been already careftdly considered (p. 106). The first 
point which strikes an Englishman in endeavouring to teach the 
common short sound (t) to a foreigner, is that the latter most 
generally conftise it with (tf, e), p. 83. The words in French 
final -^, the representatives of the Latin -tas, and similar words, 
Chaucer still distinctly pronounced (-te, -tee), etc., rhyming them 
with he, me, toe, he, see, three, degree, as : be chastite 2237, charite 
me 1723, we felicite 1267, he faculte 243, vanite thre 3833, degre 
destyne 1843, destene be 1467, possibiUte free 1293, subtilitees 
bees 10295, citee iniquite 941, adversite parde 1313, thentre see 
1985. In aU these cases we now use (-»), and it is curious to trace 
the change in the spelling. Promptcrium 1440, chastyte, charyte, 
faculte, vanite, desteyne destenye,' cyte, entre. Palsgrave 1530, 
chastyte, charjrte, vanyte, desteny, cytie, entre = entree, entrye = 
auant portail, entry = introite. Levins, 1570, chastitie, facultie, 
vanitie, destenie, citie, entrie, and he classes -ie, -ye, -y as identical 
endings. We have here then an example of the change of (-e) into 
(-•) while any living Frenchmen will prove that the best way to 
teach him to pronounce pity (pttt) is to tell him to consider it as 
written, in French letters, p^ti§ (p^). Again in Scotland the short 
f in closed syllables is almost invariably pronounced (e), our words 
ill, pit, bid, hit becoming (el, pet, bed, bet), but are saved from any 
confusion with eU, pet, bed, bet because a Scotchman ccdls the latter 
(eI, pEt, bM, bEt). In Scotland moreover (tV) is considered to occur. 
But when Mr. Murray pronounced some words to me in which he 
thought he said (it), and which he writes weade, heate, keate, I 
seemed to hear rather {ee) than («). In examining Cooper's vowel 
system, 1685 (p. 83), we were led to consider his pair will, weal to 
mean (wtl, wmI) rather than (wtl wtVl), that is. Cooper classed as 
(»V) a sftund which in the general opinion of other writers was {ee) 
or (ee). 

These facts serve to shew that (•», i) are now often confused with 

* He is particularly indebted to the ingly answered a general inritation in 

elaborate observations of Mr James the Athenaum to give the author in- 

A. H. Murray, F.E.I.S., of the Philo- formation on this point, by which 

logical Society^, on the Scotch dialects traces of the older pronunciation, as 

which were kindly placed at his dis- he believes, have been unexpectedly 

poeal, and had their value enhanced by brought to light, 

oral explanation and pronunciation of > This is the reading of one MS., 

the difficulties. One lady and several and is probably erroneous, as indeed 

eentlemen from different parts of Eng- desteyne Tor tUttene would appear to be. 
bnd (p. 277, n. 1) have also most oblig- 



272 I, Y — XIV TH CENTURY. Chap. IV. } 2. 

{esy e, e\ and hence we shonld be led to expect, if there he any 
truth in the theory advanced that we ^ould not nnfreqnently find 
«, e confdBed by the scribe, and allowed to rhyme by the poet, both 
when long and short. Cases of the short vowel are not uncommon, 
for example: list best 6819, list rest 9299, abrigge alegge 9531, 
abregge tallegge 3001, pulpit i-set 13806, shitte lette 14660, blesse 
kesse 8428, schert, hert 9757, yett witt 4-117. Cases of the long 
vowel also occur, as: swere hire 11101, 12076, geven* lyven 917, 
enquere lere 5049, there requere 6633, enquere were 8646, afered 
requered 4-244,' matere desire 4-333, desire manere 6 85, lere desire 
6-143, and in GK)wer, her sir i 161, here spire i 198, yere fir©, 
i 802. These rhymes are not only reconcilable with Ihe theory 
that (tV, t) were the usual and proper sounds of », but are exactly 
what we should expect fix>m the mistakes which occur at the 
present day. If indeed Icmg % had been pronounced (ei) and the 
first element had been slighUy lengthened, as (eei), we should get 
a sound almost identical with a pronunciation of long a now much 
in use in London.' In this case the riiyme might also appear to be 
explained. But this theory would not account for writing a simple 
e ioT long t; we should rather expect to find ey, and this never 
occurs except in a few words, as eye, high, dte, dry, sly, etc. 
to be especially considered presently, in which there is every 
reason to conclude that there was a double pronunciation. 
Hence the q)ecimens of long t rhyming to long e, and being fre- 
quently replaced by long e, throw great difficulty in assuming any 
d^hthongal sound for long t, and tend greatly to confirm the 
hypothesis that the sound was not pure (ii), but such a modification 
of it, as would easily fall into (ee), namely (it). Add to which 
there is the negative evidence that long t does not rhyme to ey, ay 
and that, except in the few cases of a double pronunciation, long t is 
never written fy by an error of the scribe in any decent manuscript 
There are a number of words of French origin which have now 
the accent on the penultim or antepenultim, but which were used 
as if with an accent on the last or penultim respectively, in Chaucer's 
verses. In the French language when these syllables, which are 
now imaccented, had the vowel t, it was pronounced (i) or (ii), and 
it would be difficult to suppose that Chaucer, who was familiar 
with French, and, in the spirit of the times as shewn by the con- 
temporary practice of GK)wer, was introducing it into English, could 
have changed the French sound and have pronounced the words 
with (ai). Still more difficult would it be to suppose, that at a 
time when the (ai) or (ei) or (ai) pronunciation of long t was 

^ This is from the ags. form ffeofan, when Mr. Matthew Arnold visited a 

and is thOTefore not an instance of $ school at Tenby, Pembrokeshire, where 

written for t, bat of # long rhyming an ancient Flemish colony seems to 

with t long. have materially affected the language 



« The French forms sufficiently ex- and pronnnciation of the people, the 
lain the termmation -quen, childrp" ^-^ — * ^.flR.«if- ,- a..^^ 

' A correspondent informs me that (feit). 



Chap. IV. § 2. I, Y — XIV TH CENTURY. 273 

oommon, as at the close of tlie xvfh and begiimiiig of the xnth 
century, it should have been deliberately rejected fix)m these words, 
and replaced by (t) when the accent was thrown back permanently. 
But we know that such words had (t ) in the xn th century, and 
that this sound has continued to the present day. For my own 
part I cannot force myself to suppose that t in the last syllable of 
the following words ever had any other sound but (ii, i, »V, t ) : 
Venise, lycorise, coveytise, servyse, justise, merite, Evaungiles, 
malice, sangwyn, famyn, Latyn, Jankyn, opposit, superlatif, motif, 
Phisik, ypocrite, practike, riche, cherice, office, Cupide, visite, 
avarice, cowardyse, Ovide, authentik, sybil, retorike, magike, cubit, 
Virgile, famyne, ruyne, apprentys, relyke, doctrine, profit, positife, 
peril, musike, chronique, inquisitife, mechanique, elixir, olive, etc., 
etc. ; or that the t was ever diphthongal in the penultim of: possible, 
digestible, fusible, etc., etc. Now if we admit that » in these 
words was (i) or (iV), or if we even allow it to have had the purer 
French sounds (ii, i), — and there is absolutely no ground whatever 
for any other conjecture, and great reason for this, — ^we have gone 
a long way to prove that long i in Chaucer was (tV) or (ii), and was 
not (ei, ai, oi). For in the first place these words rhyme as 
having long vowels, and rhyme with words which are by no means 
always French, and which in modem pronunciation have (ai), and 
had generally received (ei) by the xvi th century. That is, from 
undoubted cases of long (tV) or (ii), we are led to infer that the 
rhyming words had also long (n) and not (ai, ei, ai). If at present 
we saw machine rhyming with seenj we certainly should rather 
conclude that the i in the first word was (ii), than that the ee in 
the second word was (ai), and we should never dream of rhyming 
mine, seen, even in these lax rhyming times. Perhaps even 
Butler has not such a rhyme in his Hudibras.^ Hence it is of great 
importance to study and weigh the rhymes to the words just cited. 
They are as follows : and to Venise, were to devyse 7927, at point 
devys, cheweth greyn and lycoris 3689, which I shall devyse, augur 
coveytise 3881, ther any profyt should arise j lowe of servyse 249, 
for that thay ben toy Be, sittyng as as a justise 6609, so wel to tDrite, 
do me endiUy thurgh hire merite 11958, i- write with evaungiles, in 
the mene whiles 5085, to pitous and to nyce, of his crouned malice 
10838, he was sangwyn, a sop of wyn 335, sterve for famyn, licour 
of wyn 13866, wel dronken hadde the wyn, he speke no word but 
Latyn 639, oure apprentys Jankyn, schynyng as gold oofyn 5885, 
a gate of marbul whit, another in opposit 1895,' in gre superlatif, an 
humble wyf 9249, of me tak this motif, a court man al my lyf 9365, 
Doctour of Phisik, he was ther non him lyk 413, to byte, ypocrite 
10826, of youre practike, syns it may yow like 5769, solempne and 
so riche, was there noon it liche 10375, cherice vice 4*148, nyce 
vyee cherice 4*182, office vice 4*283, cupide tabide gyde 4*298-9, cryede 

^ On p. 16 of the Grammar of 1713, Bonse of rhythm, accent, qnantitj or 
snprii p. 47. we find incline rhyming for rhyme seems to yanish, p. 275, note 3. 
the nonce with magazine and Join, bnt > Compare the modem names Whitby 

when memorial lines are attempted, all and Whittunday, hoth from whii$. 

18 



274 1, Y — XIV TH CENTURY. Chap. IV. § 2. 

glide Cupide 4*349, Cupide »yie 5-25, beside Cupide Mde 6*238, 
Cupide aide i 160, Cupide ^«iW« i328, (hide Cupide iii 351, vysite 
wyte 4-227, visite delite myte 4*328, avarice vice 4*298, ^mprtM 
cowardyse 4*324, alyde Ovide 5*172, Ovide wide 5*254, ^*« 
autentyke 5*188, Sybile yU 5*22, retonke ^A:^ 5*235, magike syke 
5*248, lyte cubite 5*251, Viigile whiU 5*254, femyne ruyne 5-269, 
apprentys wy« 6*22, relyke lyke 6*82, doctrine discipline 6*146, 
profite »ty^ 6*176, positife strife i 12, t^i/^ peril i33, musike heswike 
1 58, cronicjue like i 145, inquisitife life i 226, mechanique /iifetf iii 142, 
fire elixir li 86, WtW olive ii 266. 

If that were possible, and more penyhle 8589, digestible, (m the 
Bible 439, in the Bible, it is an impossible 6269, on the Bible, so 
redy and so penyble 7427, metal fusible, wold passen eny bible 
12784. 

The last cited rhymes to Bible were the first which gave me any 
hope of being able to discover the pronunciation of Chaucer, ap- 
proximately, by a study of his rhymes. The above list does not con- 
tain by any means all the rhymes of this sort which I have noted 
as important ; but it is obviously sufficient to establish that in the 
words : devyse, devys, arise, wyse, write, endite, whiles, nyce, wyn, 
fyn, whit, wyf, lyf, lyk, byte, vice, abide, gyde, cryede, glide, side, 
beside, delyte, myte, wide, yle, while, stnfe, vile, fire, &c, all of 
which have now (ai), the i could not have been diphthongal in 
Chaucer's time. And these words admitted, determine so many 
others, that the proposition might almost be considered proved ; 
but it is one which many will find so difficiilt to believe that it 
is worth while accumulating proofs. 

Besides the French words already dealt with, in which the accent 
has been thrown back and the sound (») preserved, there are many 
others which have either not become paart of our modem languEige, 
or have not been left without at least a secondary accent on the t . 
We may divide them into three categories, which however do 
not include aU, such words as sacrifice, &c. being omitted. The 
first class comprehends those French words in which the i is fol- 
lowed by a simple consonant, the second those in which • ends the 
word, and the last those in which i is immediately followed by an 
e final. Now we have at present in our language a series of French, 
Italian, and other foreign words containing », of comparatively 
recent introduction, which we may therefore properly compare with 
the words then recently introduced into English by Chaucer, Gower, 
and others. The following list is taken from Walker, into which a 
few wordfi in [] have been introduced ; the f marks words which 
have become obsolete since Walker's time, and the italics words in 
which the French (ii) has become (t) ; in all other cases the sound 
(ii) has been retained in modem English, notwithstanding our pre- 
dilection for (oi) and our association of (ai) with long t. 

Ambergris, verdegris, antique, becafico, bombasin, brasil, eapivi, 
capuchin, fcolbertine, chioppine or chopin, caprice, chagrin, chevaux- 
de-frise, [chignon, crinoline,] critique, -(festucine, fiize, gabar^ei 



Cbatw IV. 4 2. T, Y — XIVTH CENTURY. 276 

haberdine, sordine, f rugine, trephine, quarantine, nnitmey fascine, 
fatigue, intrigue, glacis, invalid^ machine, magazine, marine, pa- 
lanquin, pique, police, profile, recitative, mandarine, ftabourine, 
tambourine, tontme, Isransmarine, ultramarine. 

JN^ow if it would sound hideous in our ears to talk of (Luoi'zaz 
shainvan send krainoloin,) notwithstanding our acknowledging 
(/lai'za send £ser*oldin), can we imagine Chaucer having called lys 
(lais),' partes (parvais), a^riee (agraiz*), $ophime (sofaim*), dmr 
(dezair-), avys devps (avais- devais'), assise (asaiz*), ievyne, (de- 
vain*), &c. ? Such a suppositicm appears to be monstrous, unless we 
also adopt the theory that French in England in that day was pro- 
nounced with (ai, ai, ei) for (ii) as now used. Of this there seems 
to be no shadow of proof, nor even a germ of |Hrobability.* Since the 
present habit of Englishmen is to ms^e long t into (ai) in all words 
not of recent introduction, it would be necessary to establish that 
the Normans so pronounced and that that ^nronunciation of French 
was general in England during the xmth and xivth centuries, in 
order to use this hypothesis in opposition to the usually accepted 
theory that the French sound was (ii). We shall find however 
tiiat any doubt of this kind affects the present argmnent very 
slightly, because most of the words rhyming with those just cited, 
are also found rhyming to words of the xn-eceding class, in which 
there can be no reasonable doubt of the old sound having been pre- 
served by the throwing back of the accent. The following are some 
of the rhymes which belong to this class : — 

he bar utterly the prys, the flour-de-lys 237, war and try*, atte 
parvys 311, might agrise, may devyse 7231, som sophime, hath time 
7881, to Wilde fuyr, it hath desir 5955, to aryse, I you devyse 33, 
make it wySy more avys 787, ne non novys, wily and wye 
15425, so wise, in assise 315, madfuoie Englentyne, service devyse 
121, lord and sire, knight of the shire 357,' Arcyte quyte 1033, 

* For conTenience the modern (oi) is into notice. Walker quotes the foUow- 
written for whaterer diphthongal form ing lines from " the Grammar called 
(ei, ai, ai) etc. the reader may choose Bickerstaff's, recommended by Steel,'* 
to adopt which this quotation i^ntifies with the 

> M. Le H6richer*8 opinion to the Anonyinous Grammar of 1713, suprik 

Cfmtrary will be considered in Chap. p. 47, in which they occur, p. 16. — 

y, § 1, No. 3, at the end, together with fiickerstaflTs recommendatten is quoted 

^e ralne of the Old Norman French opposite the title page — 

aiy ei, and some other matters relating ** To sound like doable e^ i does incline, 

to modem Norman French pronnn- As in JfocAiiw, and «ft«r«, and Jr<v«»ii«." 

dation. Walker adds: "If may likewise be 

• The pronunciation (sheii) is very observed, that this word, when imac- 
recent and by no means general, oented at the end of words, as Nottin§' 
Valker gives (shiir), and says that hamsMre^ Wiltthirt^ &c., is always pro- 
this "irregularity," as it appeared to nounced with « like #•." Smart says: 
him, " is so fixed as to give the regular " Letter t or y under tiie accent, and 
sound a pedantic stiflfness." Even his final in a syllable, or followed by a con- 
recent emtor Smart, 1836, gives (shiii). sonant and * mute, is irregular m no 
Webster has (sheir). This is an excel- word purely English except the verbs to 
lent example of the change of soimd, live and iogive^ and the noun^Air*; but 
and the dijficulty with which a new there are several semi-Frenoh and other 
£uhion of pronunciation forces its way foreign words in which the Frenek 



276 1, Y — XIV TH CENTURY, Chaf. IV. § 2. 

Arcite endiU 1381, Arcyte, a liU 1335, litej qnyte 3861, delyte 
lyte 4*52, vyne devyne 4*57, devyse gyse 4*64, suffice nyae devyse 
viie agiise 4*75, desire fire 4^76, enjclyne pyne myne 4*180, arise 
forhiee^ empryse 4*209, affile wMle 4*221, vcefire 4*225, desire ^^rd 
enspire 4*254, myne Froserpyne pyne 4*319, ile tpile 5*321, rye 
(= rice) tretys 6*32, ile while i 95, Cecile while i 104. 

The word lyte, which seems shewn to have been (liit) or (liVt) by 
some rhymes above, being the origin of our little, can hardly be 
conceived as (lait).' The following among other rhymes to this 
word, however, not only establish the sound as (liit, ItVt), but settle 
many other words as well. 

Lite rhymes with delyte 4*52, quyte 4*55, kyte 4'63, white 4*76, 
white delite 4*94, 6*237, wyte = /bMW 4141, delite endyte 4*163, 
plite 4*202, write 4*202, 5*269, 6*256, wjte = wit 4*255,' myte = 
mite 4*259, white 4*289, 5*195, 5^*282, Arcite wite =punish 5*200, 
smyte 5*232, cubite 5*251. 

The word Inde must be considered French, and most probably 
had the sound (ind*^) which the Englidi heard (ind^e). The pre- 
sent nasal pronunciation of Fr^ich in is certainly not at all indi- 
cated in any of the numerous words beginning with i«, which we 
have taken from the French, and without any intimation of this 
nasality or any trace of it in English derivation we have no right to 
assume it. The vowel in India is short in the original language, 
and in the Greek and Latin derivatives. It is still so pronounced 
in English, and although I have heard some persons read (aind), 
for the sake of a modem rhyme, I doubt whether they would 
venture to talk of (ain*dia). It seems therefore just to conclude 
that the Saxon words which rhymed with it, most or all of which 
had acquired the sound (eind) in the xvith century had also the 
sound (f'nd). Thus we have kynde Inde 6405, and fynde kynde 
mynde Inde bynde lynde 9057, 9063, 9069, 9075, 9081, 9087, 
rhyming together in bEnvoye de Chaucer, at the end of the Clerkes 
Tale. The last worde /ynrftf = linden or lime tree, still has the 
sound (fhd) and confirms the other conclusions. The use of mende 

soundof tiB retained; as fMartfitf,jM>/tM, yation of quality by shortening of 

profile, &c. : .... The word oblig$, quantity, as in p. 273. Shire, an. 

wluch formerly classed with marine, scire is said to nave a long vowel oy 

&e., is now pronounced regularly." Bosworth, and a short vowel by £tt- 

Live, gothic liSan, ags. libban, Orrmin miiller. But the vowel became de- 

libbmn, had from the first a short cidedly long, and, as we have seen, it 

Towel, with which, however a long has preserveid the (ii) sound. The 

vowel alternated in Orrmin in li/e^^. cognate word sheer, a^s. eeir with long 

lifenn, and a long vowel seems general i, which has preservea its sound in aU 

in Chaucer, and hence we have simply Germanic dialects, will be especially 

the usual continuation of the short noted in Chap. Y, } 1, No. 6, at the 

voweL CHve, gothic giban, ags. gifan, end, as a rhyme iojlre, 
geofan, also had a short vowel, but in 

Oirmin, all parts except the imperative ^ "Set an example to," from ags. 

Tty, and preterit ^a/; have long vowels. bieen, example. 

From geofan, we have the frequent » Lite, however, the Danish lilU for 

form %eve in Chaucer. In this case we little, is called (lail) in the North of 

have then perhaps rather the preser- England. 



Chap. IV. § 2. I, Y — Xrv TH CENTURY. 277 

for minde to rhyme with ends in the carefully spelled Harl. MS. 
8869 of Gower, ii 23, ii 67, and kmde for kinde also to rhyme with 
$nde iii 120, is scarcely reconcilahle with the present diphthongal 
sound of f in mind. 

Through the kindness of several gentlemen^ I am enahled to say 
that in South Shields, Eendal, Westmoreland, and Cumherland 
generally, and parts of Lancashire, the short vowel (») is still heard 
in the words hind, blind a., hehindy hinder a., hindmost f Jind, grind, 
icind* V. = (bind, blind, biHint*, Hind'j, Hin'mast, find, grind, 
wind). See also the Scotch pronunciation infr^ p, 289. With 
these analogies it would be considerably more difficult to imagine 
the diphthongal sound than the short vowel in such words. 

The French Words of the next class are those which end in i or y, 
and which are referred to in that paragraph of Palsgrave which 
occasioned so much difficulty in the last chapter (p. 109), and they 
are also remarkable for the English words which rhyme with them 
in Chaucer. The French words are themselves not numerous. In 
the Canterbury Tales, there seem to be only mercy, fy, enemy, fool- 
hardy, cry, ^uirhoily, to which perhaps yifory, vicory, although the 
final y is difficult to account for.* These words rhyme, first with 
each other, next and very frequently with the termination -ly, and 
these words and this termination rhyme with the Dutch (?) cowrUpy, 
and with the Anglosaxon /, why, by, thereby, sty. The only words 
among these which could have a plural, enemy, sty, do not occur in 
the plural in rhymes in the Canterbury Tales. It was with special 
reference to this investigation that I enlarged the field of enquiry, 
extending it over the rest of Chaucer's poems and Gower. Some of 
these poems, as we have seen, are not in a trustworthy form, especially 
the Court of Love (p. 251), Fhwer and Leaf (p. 251), Chaucer's 
Dream (p. 251), and Romaunt of the Rose (p. 252), because they 
admit of rhymes which belong to a later period. The best manus- 
cripts are altogether free from such rhymes. The spelling in Fauli's 
Gk)wer must always be corrected by the manuscripts. Allowance 
must be made also for those words which had a twofold pronunci- 
ation, as (ai) and (iV), not always marked with sufficient care in the 

^ Rev. C. T. Potts, of Ledbnir, for great trouble to themselyes enabled me 

South Shields ; Mr. Brown, or St. to supply these illustrations. Messrs. 

Peter's College Peterborough, for Ken- Potts, Brown, Hetherington, and Shelly 

dal ; Mr. J. I^. Hetherint^n. Clifton hare been pa^cularly uberal with the 

Parsonage, Worldngton, for Cumber- time they nave bestowed on me. I 

land ; Messrs. Jackson, Fielding, and shall term these assistants generally 

Axon, for Lancashire, — ^have supplied my dialectic corr^pondents. 

me with information firom personal ' The substantive toind is generally 

knowledge on tMs and other points ; (wind), but in Cumberland it seems to 

and Mr. Shelly, of Plymouth, for be always (wahind, waind), so that 

Devonshire; Messrs. Atkinson and wind s. wind y. have precisely the 

Moore, for Yorkshire; Mr. Hallam, opposite pronunciation to what they 

for Derbyshire : and a lady near generally receive in the south. 

Norwich, have also supplied much in- * Diez says that avori, ivori are 

formation on dialectic pronunciation. Provencal forms, which it is singular 

I beg to express mv thanks to these to encounter in English. For vicory I 

and other corresponoents who have at know no authority. 



278 T, Y — xnr th century, Chap. rv. i a. 

spelling, to be carefully considered presently. With the exception 
of such words no case has yet come before my notice in which -« or 
-y final rhymes with -fy or -ay. In the following list of rhymes 
fdl cases of -/y rhyming to -ty, which are veiy frequent and convey 
no information, are omitted; and by no means all the rhymes, 
except in the Canterbury Tales, of 7 with -/y, -^, forthi, etc., are 
given. 

Soburly courtepy 291, pitously mercy 951, enemy I 1645, ryally 
by 1689, fy mercy 1775, ryally enemy 1795, synfoUy fy 4499, 
mercy solempnely 5110, pitously, mercy I 5479, by specially 5544, 
therby I 6597, prively therby 6925, yvory fetisly 7323, sty I 
7411, comunly why 7839, stedefestly mercy tenderly 8970, why 
I 9315, uncurteisly cry 10237, cry pitously 10727, therby I 
12650, mercy sey 13308, therby ydelly 13860, subtily by 13980, 
redily forthby 14082, pitously, ther by 15011, quirboily yvory 
15283, I fool-hardy 15401, trewely by 15411, sodeinly enemy 
16889, lustily vicory 17315. 

I mercy 4*65-6, truly unlusty I 4*76, by prively 4*77, by I cry' 
4-78, cry ocy 4-79, ny cry I' 4-81, wrongfdly 1 4125, redy I 4*148, 
trewely I by 4*175, tyme, hi me, pryme 4*193, by hertely 4*205, 
whi by bisily 4*272, I fynaly 4*336, pitously by hastily 4*337, I 
certeinly therby 4*341, y why 5*173, why comelely 5*180, trewely 
lady 5*190, hooly mercy 5*193, I why 5*239, I mercy 5*266, by, 
domus Dedaly = Dadali 5 '267 ^ y by 5*269, by and by, curteysly 
5*285, y by 5*34L 

I openly i 44, why I i 47, forthy pleinly i 51, forthy therby i 53, 
cry unhappily i 54, redily by i 93, sodenily by i 102, I, graunt 
mercy i 103, forthy mercy i 106, I forthy i 107, worthy mercy 
i 107, sky sodeinly i 109, why forthy i 114, openly cry i 115, 
mercy why i 116, why prively i 148, communly why i 172, 
why forthy i 173, comely awry i 174, redely forthy i 200, 
kindely why i 205, sely privete i 225, time, by me i 227, 309, 
370, ii 41, 49, 114, iii 6, 869, I truely i 227, bodely why 
i 259, why forthy i 280, lady thereby i 292, cry buxomly 
i 297, by lady i 298, cry therby i 314, forthy enemy i 330, I forthy 
i 332, enemy why i 347, why forthy ii 20, I by ii 24, 41, sky 
by ii 29, bodely tberby ii 34, forthy therby ii 50, openly forthy 
ii 51, traely sky ii 59, why I ii 69, besily enemy ii 75, I forthy li 
95, why cry ii 122, bodely forthy ii 133, redely by ii 137, why 
sky ii 158 forthy Eoly = uEoli ii 160, forthy by ii 161, forthy why 
ii 163, sky wt^ ii 167, Satiry= Satyri properly ii 171, forthy 
proprely ii 187, by I ii 219, why buxomly ii 228, by mercy 
li 278, esely mercy ii 295, why therby ii 301, mercy redy ii 314, 
mercy therby ii 373, I worthy ii 379, sodeinly askry ii 386, 
mercy rudely ii 396, why almighty iii 61, mercy thereby iii 82, 
forthy mightily iii 92, lugh sky iii 93, by and by sky iii 116, 
Gemini redely iii 119, Gemini forthy iii 119, Gemini proprely iii 
127, I by iii 168, I forthy iii 185, mercy redely iii 198, sodemly 

^ Erroneously spelled ^#, crtV. ' Erroneously spoiled nye^ erii. 



Chap. IV. } 2. I, Y — XIV TH CBNTURY. 279 

aakry iii 217, why pitonsly iii 260, wliy Genesy iii 276, by and by, 
prively iii 305, pitously I iii 315, enemy envy iii 320, cry by iii 
321, lady prively iii 325, forthy by iii 348, redely wby iii 368, 
I mercy iii 372, sodeinly sky iii 375. 

It is impossible to glance over the above list without feeling that 
whatever was the pronunciation of this final -y in any one word, it 
must have been the same in all the words, and hence if there is a 
certain clue to any one word, we have a clue to all the rest. Two 
rhymes are very noteworthy: mercy sey 13308, and sely privete 
i 225, but their very peculiarity and the absence of any corrobora- 
tive instance whatever, render them suspicious. Yet, as the first 
of these was the only clue which I could obtain for some time, I 
was misled by it to suppose that this termination -y had like sey the 
sound (sai). This shews the danger of trusting to single instances. 
Even in tiie HarL 7334, which is followed by Wright and Morris, 
we find: an hihe, sihe 11161, which should be: hih, sih, probably 
(n^kh, stkh). But an examination of seventeen MS. which con- 
tain V. 13308, shews the following variants. 

In the BriUth Mutmm.^ Bawl. MS. Poet 149 mercy sej 

Harl7333 mercy sey ?^^° on ""^^ "*? 

Harl7834 mer^ sey ^^^V?a t> ,^ "^^ 'y'^^ 

, V or« lu^v^ OCT Arch. Seld. B 14 mercy sy 

Unsdowe 861 mercie nhe ^ ^ ColL MS. 198 ^ ^ 

Sloane 1685 mercyesay ^ v/ ^.vix. i,x«.*,7o 

Eeg.lTDxT . mercy sy ^'^'^ "^^"^y ^y 

Beg. 18 C ii mercy sey At Camhridge^ 

jsn^^^% ^' -*• 27 (No. 1) ley 

At Oxford.^ ii»3 26 ^ ' ge^ 

Land 600 mercy sie Mm. 2. 5. seye 

Laiid 739 mercy sey Trin. Coll. B. 3. 8. mercy sigh. 

It is clear that the passage has much exercised the scribes who 
have occasionally ventured to add an « to mercy ^ which is quite 
illegitimate, and the majority have inclined to the more usual form 
in Chaucer, Bey. The usual form, however, in Gower is «tA, written 
sigh by Pauli. The above 17 instances may be divided into an 
(ai) class and an (n) class, thus — 
(ai) sey sey say sey sey sey say sey sey seye . .10 

(n) sihe sy sie syhe sy se sigh 7 

The word clearly belongs to those doubly sounded and doubly 
spelled words to be presently examined, and we must conclude that 
those scribes who used the (ai) class of forms were misled by habit, 
and should have used an (it) class, and, since the guttural could not 
have been pronounced in French, the scribes ought to have omitted 
it in the English word. It will be seen that when eye^ high are 
pronounced with («) the guttural is frequently omitted. This leads 
us to prefer «y, given by two MS, of which w, m are mere accidental 
varieties. The preterite (sii) as : I see him do it yesterday, is not 
yet obsolete among the uneducated, while (sai) is unknown. 

^ Examined by myself. * Examined by Messrs. H. Bradshaw 

« Examined by Mr. G. Parker. and Aldis Wright. 



280 I, Y — XIV TH CBNTURY, Chap. TV, } 2. 

The second instance : sely privete i 225, although unparalleled 
among these rhymes, would not be unprecedented, for we saw at 
the beginning of this investigation that long t and long e occauon- 
ally interchange, but we already know that the proper reading is : 
cele privete, (p. 253). 

Eejecting these isolated instances, we are struck by the rhyme : 
tyme, hi me, pryme 4*193 in Chaucer, and the eight times repeated 
rhyme : time, by me, in Gower. The rhyme : sophime, time 7881, 
has already (p. 275) led us to consider (tirme) a probable pro- 
nunciation, and hence these repeated rhymes lead to calling by (hit). 
More than this, by is often spelled bey be thy trouthe 5*227, 
alle be hemselve 5*246, be Gk>d 5*256, and indeed be^ by occur in 
the same line: be strengthe and by his might. 5*348, from the 
Legende of Good Womm^ following the Bodleian MS. Fairfax 16, a 
good manuscript. These variants strongly confirm the hypothesis 
that by = (biV). 

It is certainly fair to conclude that the purely French words in 
these rhymes had the sound (ii) or (iV), the latter probably in 
England, and the former in France. We were driven to this sup- 
position on comparing Palsgrave with Meigret in the xvi th century 
(p. 110). We might therefore assume that: mercy, enemy, fy, cry, 
quirboily, fool-hardy, envy, had the sound (ii) or (tV), and these 
would be ftdly sufficient to determine all the rest. But as this 
assumption in fact involves the whole question, it will be better not 
to lay great stress upon it. 

The cry ocy attributed by the cuckow to the nightingale 4*79 — 
For thon hast mony B,fiyn$d qaeint cry, 
I have herd the seye, * ocy, ocy ; ' 
Bat who myghttf wete what that shulde be ? 

leaves us in the same ignorance as the cuckow, and can be of no 
assistance if we go' to the real cry of the bird; but if we take it as a 
French spelling of an imitation of that cry,* then we have simply 
two French sounds cry, ocy rhyming. 

There are several instcmces of Latin final -f, one in Chaucer: 
Dedaly 5*267, and several in GK)wer: Eoly ii 160, Satiry ii 171, 
Gemini iii 119, twice, and iii 127, and it is difficult to suppose that 
Latin was at that time so mispronounced as to have i called (ai). 
The Eoman Catholic tradition must have saved this heresy, which 
seems to have only crept in with the xvi th century, and was even 
then reprobated by many, as by Salesbury. At least these rhymes 
must be considered to add to the probability of the (ii) or (n) 
pronunciation. 

With regard to the termination -ly which plays so great a part in 
all these rhymes, it is to this day generally pronounced (If) in con- 
versation, although declaimers will sometunes permit themselves to 

' " FiSB, FiBB, OCT, OCT : SoDS nng chesne sur lec^nel avoit one ros- 

onomatop^ repr^ntant le chant du signol qni chantoit tr^ melocueiiBe- 

rossignol (r^pet^s plus has dans one ment et cryoit ainsy qne tont endear^ 

chanson)." Koqoelort, eub./w, where et Jler, JUr, ocy, ocy,** from Soman d4 

he cites : **il y avoit au-dessus de Iny F9rc$'Fare9U 



Chap. IV, §2. I, T — XIV TH CENTURY, 281 

say (lai), and we find GKll in his transcript of the Psahns con- 
stantiy using this sound, apparently to add dignity. He also says 
(madzh'estai), and, at least in one place (mer'sei), but the latter is 
probably a misprint, for he generally writes (mer'st ). Modem poets, 
working upon an old foundation, permit themselves to consider -y, 
under a secondary accent, as either ^-ai) or (-ii). This belongs to 
the licentiousness of modem rhyming, superinduced by an un- 
phonetic orthography. I cannot consider this early usage of Gill 
to indicate in any way the old pronunciation. It was undoubtedly 
wrong in words which had formerly -w, -e, and was probably fanci- 
ful in other cases. Dr. Qill had a notion that the (oi) added to the 
beauty and strength of the English language,^ and hence his employ- 
ment of it is suspiciouB unless weU corroborated. As to the practice 
of modem times, it is sufficient to cite Walker and Smart, who, not 
recognizing the difference between (i, »') identify this termination 
with (-li), but that is properly an Irishism. As, then, there seems 
no reason to suppose that this termination -ly ever had, in natural 
speech, the sound of (-lai) but only (-lii, -ItV, -li, -h), the conclusion 
in favour of the (ii, ii) pronunciation of the other words seems 
inevitable. But tiiose who have made up their minds to the (ai) 
pronunciation of long », and especially of the pronoun 7, will object 
that we have in Qill an actual example of the (ai) sound, and ^at 
we hear occasionally, under peculiar circumstances perhaps, and by 
no means uniformly in the same speaker, but still we do hear (-lai) 
now and then, and that it is possible that (-h) may be a "cor- 
ruption" of (-lai), rather than (-lai) a mistaken intensification of 
(-It). It is therefore necessary to try some other words, which 
are fi-ee from Gill's imputed (ai). Enemy is not such a word, for 
he writes (en-emaiz), supri p. 110, note. But lady 5*190, i 292, 
298, iii 325 ; almighty iii 61, worthy i 107, seem unexceptional. 
The words do not occur in Gill, but lady does occur in Salesbury, 
who transcribes it in Welsh letters ferfi = (laa'di). In modem 
ballad poetry we have constantly to read (b^dii*),' but the pro- 
nunciations (Iwdai, leediQv) are utterly unbiown. As this word 
determines -ly -by, by its rhymes, and these are sufficient to de- 
termine all the rest, the difficulty may be considered as solved. 

But there are still important considerations which lead the same 
way, and which must therefore still be adduced. It is difficult to 
suppose that a cry and the verb to crye^ had their y diflferently 
pronounced. This y would probably retain its sound in the in- 
flected form cfyede, often a dissyllable as cry^de, Now we find : 
cryede glide Cupide 4*349 in Troilus and Cryseyde from a good 
manuscript, and Cupide is one of those words in which we have 
already recognized the persistence of the {ii) sound. Again : criede 
Cupide Cipride 5*9 occur in the same poem. Gower has: cride 
hide i 149, cride wide iii 213. All this points to the pronunciation 
(cnV'de) and hence (citV) for the substantive. But there is one 

^ '* Retinebiiims antiqunm iUmn et Hazeldean, in which the first stanza is 
mascnliim sonmn." Logonomia^ p. 7. said to be ancient: "Why weep ye by 
« As in Sir W. Scott's Jodt of the tide, ladie P" 



282 I, Y — XIVTH CENTURY. Chap. IV. J 2. 

word which seems at first sight to ran counter to this conclusion : 
reneye 4796, 12196, 12376, 16047 etc, always meaning to renounce, 
abjure, in modem French rmier^ so that ey seems to answer to 
French ». But Eoquefort (Gloss, de la lang. rom. ii, 463) gives 
the old forms renoier^ reniier, and Kelham (Diet, of the Norman or 
old French language 1779) has reney$e renegade, r$neign refdse. 
So that the t is a modem French development, which does not 
affect the present investigation. 

Perhaps the strongest evidence of all is fumiahed by the very 
word enemies^ which was lately rejected on account of Gill's (en*e- 
maiz). Of course there is no doubt whatever of the sound of % in 
the words «, kis. These words never could have been (ais, nais) 
at any time. No champion of (oi) could ever entertain such a 
notion as this. Now in Gt)wer we have : pris is ii 341, wis is iii 
226, which may be taken to settle the pronunciation of prii^ wis 
i.e. price, wise, in the xiv th century, and strongly corroborate the 
method by which we have already arrived at this result. Bearing 
this in mind, the rhymes : enemies pris ii 67, iii 199, enemies is 
ii 342, enemis his iii 214, enemies wis iii 216, leave no doubt that 
Gower said (eu'emtz) or (en*emts), and that he therefore must have 
said (eu'emt) as the natural pronunciation of his time, (^ have 
occasionally lengthened the final vowel into (it, ii). But if so, all 
the rest follows fix>m the rhymes: enemy I 1645, ryally enemy 
1796, sodeinly enemy 16889, forthy enemy i 330, enemy why i 347, 
besily enemj ii 75, enemy envy iii 820. 

It seems mipossible to form a stronger chain of evidence in favour 
of an unknown pronunciation, but the strength is rendered more 
evident by the circumstance that there is no instance of -» rhyming 
with -ey, except such as are explicable by the fact that the word 
had several soimds and several modes of writing, often used in 
other places, and that the scribe accidentally employed a wrong 
orthography, as in the instance : mercy sey 13308, already con- 
sidered. Everything is therefore so far reconcilable witii the 
hypothesis t = (tV, «), and many circumstances are irreconcilable 
with the hypothesis t = (ai, ♦). Hence I feel compelled to admit 
that even the personal pronoun /was called (u) by Chaucer. This 
personal pronoun had three forms, / most commonly, ie, ieh, rarely. 
That in these latter forms the % was (• ) short, seems proved by sudi 
contractions as theek 3862, theech 12857, 14362, =thee ik, thee 
ich. The diphthong could hardly have been so lost. Again the 
change t>, ichy would be unusual, though possible, if i' were (ai). 
But / seems formed from iV, «VA, just as a is from an. The original 
pronunciation of the indefinite article was of course (a), and it is 
now frequently (a, «), but the emphatic pronunciation {ee) is of 
modem growth, and seems precisely comparable to the emphatic 
use of (ai) for (• ) in /. 

Further corroboration of the above conclusion will be afforded by 
considering the termination -ie, -ye. In two instances Chaucer uses 
the French words par compaignye, at the end of a line, not as 
Anglicised, but as a real French phrase. There may be some doubt 



Chap. IV. § 2. !# Y — XIVTH CENTURY. 283 

as to the sound of gn, ^dietiior (nj, n/) or siiiiply (n), as will be 
hereafter considered, btit as it is also written as a simple n, it will 
be sufficient to consider it here as (n). The two last letters mnst 
haye had the French sonnd, which cannot well be conceived as 
anything but (ir«), or the English modification (n'e), a change so 
alight that the KngliwhTHRTi would have thought he was exactly 
correct. Hence: par compaignye, £antasye 8837, par companye, 
molodye 4165, must be considered as establishing the Engli^ pro- 
nunciation (fantastV'e, melodiV'e) of these Anglo-'French words. The 
following rhymes strongly confirm this conclusion : 

hosteMe companye 23, multiplie Marie 15100, Emelye melodye 
873, Emelye, gan to crie 2343, signified, sche cryed* 2345, philo- 
sophie, wolde he crye 647, envye' crie 909, tyrrannye espye 1113, 
chyvairye curteaie 45, I made him Me, jalousie 6069, ragerie, as a 
pye 6037, maladye manye = mania 1375. 

I schal not lyej companye 765, curtesye lye 7251, vilonye, nat a 
flye 4189, Emelye, gan sche hye = Ai>, hasten 221 bj harlotries, 
toUen thries 563, boiUe and Mc, bake a pye 385, melodie, my body 
^e 12062, curtesie, for to gye 7950, maladye, moist or drye 421. 

The first list consist entirely of Anglo-French words, the second 
gives rhymes of such with other words. Now throughout Harl. 
7334 this termination -ye never rhymes' with any other termination, 
such as -y, -tf, which has now received the same sound (-t). But 
during the zvth century the final e was thrown off, and then these 
words fell into (mel'odi, fan'tast) etc, and became rhymes to -/y. 
These rhymes therefore not only shew a later date, but indicate an 
identity in the pronunciation of i in the two sets of words. As 
then we have no conception of there having been an (ai) sound in 
the -ye endings, (except in such words as signify y where of course 
it is due to the accent), we have a corroboration of our former 
conclusion that long « was (n, ii). Whenever we see in any 
manuscript of Chaucer or GK)wer such rhymes as -y, -ytf, or as'-d, 
-ye, we may be sure either that there has been some accidental 
orthographical error of the scribe, or that some words of a more 
recent period have been substituted. The error is often very 
obvious and easy to remedy, thus : high testifie 4*1, tnajesWe dignyte 
kne 4-3, see ryaltie 4*5, lihertie degree 4*10, crueltie pyte 4*12, 
should have: hye, majeste, ryalte, liberte, cruelte. But degree 
ye = eye 4*5, I dye hi^ 4*8, hie crye whye 4*10, I espye ye = eye 
4*10, hie besyly je=eye 4-11, fantasye merily 4*15, ye = eye pretily 
4*15, se JO ^= eye 4*27 etc., are certainly erroneous, and could not 
have been written by a xiv th century writer. They serve there- 
fore to discredit the MS. (R. iii. 20, Trinity College, Cambridge,) 
of the Court of Love, 

^ Probably ngnijiedef eryede are fhe ' The mistakes hyghe remedye 4629, 

proper forms. gym aspien 12426. hee <y# 11503, 

» Both French forms envi, $nvi$ jeUnuye me 1809, naTe already been 

occor, old and recent, and both envy, noticed (p. 260) ; the proper readings 

•nvie are £rand in old English. are hye, yen, hye ye, jolite. 



284 I, Y — XIVTH CEirrUBY. Chap. IV. f 2. 

Three other corroboratiTe drcmnsUiicee may be mentioiied. 
First, if long t had been (ai) in the xrrth century and earlier, 
Kngliah would have presented the extraordinary spectacle of a 
language without a long (ii, tt), one of the primitive vowel forma. 
Sir Thomas Smith had indeed reduced Latin to such a condition, 
but this was a purely artificial formation, due to a mistaVeB theory, 
and we may safely say could never occur in practice. Secondly, 
if long % had been (ai), we should have to account for its common 
unaccented form (t ). There is a dispute among orthoepists as to 
whether (ai) or (t) should be pnmoimced in certain unaccented 
syllables, such as (srvtlfz^'shen) or (srvtlaizM'shtm), or (didzhest% 
daidzhest), (in*finit, th-fainait). These disputes at least serve to shew 
that there is no difficulty whatever in using (ai) in an imaccented 
syllable, and hence make the employment of (t) inexplicable, except 
on the theory that it was the original normal sound. The change 
of (ai) into (t) is of course possible, but it is generally throu^ 
(ei, ee, ii). We have this very transition in deceive, which was 
(desaiv) in the xiv th and even xvi th centuries, became (deseiv) 
and passed into (deseev) in the xvn th, and fell into (disiiv) in the 
xvni th century. But the transition took a long time. Tlus was 
probably the course by which the old Oreek u reached the modem 
Greek (ii). We have no trace of such a change in the words con- 
sidered. The third circumstance is, that the scribes of the xivth 
and early part of the xv th centuries seem to have had no hesitation 
in writing • and ei or y and ey according as they wished to indicate 
a difference of pronimciation. This is especially the case with the 
words die, dry, eye, high, lie, eih, tie, pine, which must therefore be 
considered individually. 

Die = (dai'e, dii*e). This common old English word is not 
An^Losaxon. The old Norse is deyja, ek dey, do (dei'ja, ek dei, 
doo), and desenn in Omun, dei^en m La^amon, deyin' in the 
Promptorium, point out (dai'e) or (dei*e) as the older pronunciation. 
The same sound is indicated by : seye deye 4944, 7207, waye deye 
5010, 5238, 11649, disobeye deye 8239, deyth seith 7623, seyde 
deyde 2847, preyde deyde 8424, sayde abrayde deyde 8935, and 
generally. In: brayde prayde dyde 16022, we have therefore a 
clerical error for deyde. But we have a different spelling and a 
different set of rhymes in : Marie dye 5261, Emelye dye 1569, 
1589, 1595, dye, folye 1799, ye = eye- dye 7913, Lombardye hye 
allie dye 15886, die Galaxye 4*53. Hence in: deyevilonye 11715, 
deye bigamye 5667, deye sloggardye 1 1943, deye is a clerical error for 
dye. Whether this double pronunciation was of a much older date 
or not, it is difficult to say. The point to note here is, that there 
was a double method of spelling, and that, except from mere 
carelessness of the scribe, each method answered to its own rhymes, 
which we had previously recognised as (ai, ii). At present (dai) is 
the conmion form, but (dii) is more usual in South Shields, Kendal, 
Westmoreland, and Lancadiire. 

> MS. UniT. Lib. Cam. Dd. 4. 24, icMk eye deje, which is dio legitiiMte. 



Chap. IV. { 2. I, Y — XIV TH CENTURY. 285 

Buy == (biV*e, bai'e). The first seems the older form as an 
alteration of biggen, the second b not so frequent : to byen 14467, 
bye housbondrie 5869, preye beye 12564. 

Dry = (drtfe, drai'e). Here (iV) seems to have been the 
original form corresponding to ags. (yy), and (ai) the derived. Ags. 
dryge drige drege dry, Orrm. drijje. Hence : maladye drye 422, 
diye remedye 4*56, drye dye drie crie guye 5*208, where the first 
drye means to suffer, still found in Scotch as dree (drii). On the 
other hand : weye dreye 8773, drye seye preye 4*64, where drye is 
evidently an error for dreye, aweie drey(e) i 220, but : drie deie 
iii 93 might be : drye dye, or : dreye deye, probably the former. The 
form dreye seems proved, but it is not so common, and what is most 
important for the present purpose, it was a derived, not an original 
form, which the scribe was not content to leave under the old 
spelling drye. The legitimate inference is, therefore, that if in 
other words (ai) had been pronounced, ey would have been written. 
At present (drai, drai) are the common sounds, but (drii) is known 
in South Shields. 

Eye = (ai*e, tre). The older sound seems to have been (aijth'e, 
eiilh'e). The more usual orthography is eygJie, eyghen, or eyhen 
when the word does not occur final. I have not noted it in a 
rhjrme in Chaucer, but we have : eie seie i 72, eye awey(e) i 127, 
and Pauli constantly writes eie when the MSS. have yhe. The 
guttural (ifeh) seems to have been often entirely lost, passing 
probably through (jh), and then becoming absorbed in the 
preceding (i) ; or more properly the diphthong (ei) grew out 
of (ejh). The value (tre) results from : melodic yhe 9, companye 
dayesye = daisy = day^s eye 333, (for dayse hie 4*77, read daysye 
hye,) crye yhe 1097, ye = eye plye 9044, yen wryen 17193. For: 
specifie eye i 3, highe eye i 106, sigh eye i 116, as Pauli writes, 
read: specifye ye, hye ye, syhe yhe. Although (oi) is very 
general, yet (ii) is almost the only form known in Newcastle, 
Cumberland, and Lancashire, and is even used in Devon. 

Migh = (nai, mi). The older form is here (nei, nai) the (i) 
being generated from (jh), the representative of (A;h). The usual 
forms when the rhyme does not require the others, are heih, height 
frequently with an added e. Possibly, as in eye, the guttural was 
early lost in developing the diphthong, compare Ornnin*s heh, 
he^e. In rhymes this older form b not common, and is often 
doubtful, thus : heye eyghe 3243, heyghe eyghe 10587, might 
have been : hye ye. More certain seems : heyghe piggesneyghe 
3268, on heigh seigh = saw 1067, which may have been : 
on hih sih, compare 11162. This form often occurs in Gower, 
where Pauli writes : high sigh i 2, i 24, i 137. On the other 
hand the form (h«) is very common : hye crye 10725, hye 
prye 7319, hihe eye, read yhe 11347, eyen read yen, prion 9985; 
prye hye compaignie 4*222, hye gye compaignye 4*296, hye navye 
5*215, hye jurye 5*253, hye skye 5*258, high read hye, poesie ii 36. 
(Hii) is used in Cumberlcmd and Scotland. 



286 I, T XIVIHCBWTUKY. Chap. IV. } ti. 

Sfy = (slai, sliV). The first is the old form, in Omnin sleh, and 
(sliV) is more recent. The rhyme slye, lye mmtiHy ye oeuUu 6'87-8 
is ambignous ; but if: high testifie sly 4*1 shoidd be hye, testifie, 
slye, this is a rhyme in point. Sleigh occurs 3201, 4*339 v. 944. 
(Slii) is still found in Cumberland and South Shields. 

^>= (tai'e, tile). The first is the old form, from ags. tegan, the 
second seems to have come from a second form ags. tygan ; seyd teyd 
10305, gives the first distinctly, the form : ty^ed, Allit, Foetm by 
Morris A. 464, suggests the second sound, for which I have noted 
no rhymes. (Tii) is found in Kendal, Cumberland, and Lancashire. 

Pine, pain = (piiTie, pai-ne), are really two separate words, but 
they are used so much in the same sense that they might be easily 
supposed to be different forms of the same word. The first is 
Anglosaxon, the second French, but both apparently come from 
Latin poena. They have come down to the present day also with 
different pronunciations (pain, p^fn), and different meanings. The 
following passages will shew how the words are confused by Chaucer 
as the exigencies of the rhyme require. 

And whan a beste is deed, he ne hath no peytii, 

Bat man aft^ his deth moot wepe and pleyne. 

Though in this world he have care and woo : 

Withonten donte it may stonde so. 

The answer of this I lete to divinis, 

Bat wel I woot, that in this world gret pyne is. 1321 

In which ther be som merthe or doctrine. 

Gladly, qnod I, by Goddes swete p^ne. 15343 

That telleth ns the peyne of Jhesa Crist 16352 

And sythen that 1 knewe of loves peyns 

And wot how sore it can a man destreyne. 1817 

Fnl golteles, by Goddes swete pf/ne, 

For as an hors, I conthe bothe bite and whyne. 5967 

who wold* suppose 
The wo that in my herte was and pyne ? 
And whan 1 saugh he nolde nerer fyne 
To reden on this cursed book. 6369 

In Armorik, that clepid is Bretaigne 
Ther was a knyght, that loved and dide his p0yn4 
To serven a lady 1 1041 

We thus see that in the xiv th century there was a tendency to 
two forms in certain words, and that in general the original form 
has (ai) and the secondary form («i). In one case, however, at 
least, drt/f the (n) form appears to be the older. In every case, 
however, except from mere carelessness of the scribe, the two 
sounds were carefully distinguished as ei, i or fy, y. There can 
therefore be very little doubt that when only one form i or y, wae 
employed, there was only one pronunciation, (n), because the 
Bcribe, who was hampered by no historical associations, must have 
many a time and oft written ^ if he had ever heard the sound (ai). 
In all of these cases the (i ») sound has been dialectically preserved. 

This completes the argument in favour of the proposition with 
which I stwted, viz., that the sound of • in Chaucer's time was 
(tV, •) and not (ai, •). But the result admits of iUustration by 



Chap. IV. § 2. I, Y — XIV TH CENTURY. 287 

dialectic peculiarities in addition to those just adduced. Isolated 
and small societies necessarily preserre idiomatic expressions, pecu- 
liar words and peculiar pronunciations. Of course the so-called 
Anglosaxon which established itself in England was not uniform. 
The languages with which our dialects began, so to speak, were 
remarkably different in many respects. It is not merely the pro- 
nunciation of a few words which now distinguishes the men of the 
North, North-west, North-east, West, East, Midland, South-west, 
and South-east, from each other and from those who speak literary 
English. The whole intonation, many of the words, the idioms, 
the grammatical constructions, are different. The effects of isolation 
are shewn strongly among the scanty population that speaks what 
we call Scotch, and consider it as a single language. Mr. Murray 
has been able to distinguish eight Scotch dialects so sharply as to 
translate the book of Ruth into each of them. In some of these 
dicdects the differences of pronunciation are as great as those which 
separate English utterances in distant centuries.' Nevertheless 
we feel that all these dialects have one common origin with the 
literary English, and that an examination of their peculiarities, 
as respects this vowel t, will be of some assistance in conceiving 
the former existence of a pronunciation so extremely different 
from our own. It was with this view that I requested the 
cooperation of those personally acquainted with these modes of 
speech — ^which every one must regret to see at present so impCT- 
fectly writtei, that the spelling conveys but little knowledge 
to a reader who is ignorant of the dialed;, and whom the writing 
ought principally to aim at instructing. 

Mr. James A. H. Murray's native dialect was that of Teviotdale, 
and this possesses a very remarkable peculiarity. The following words 
which are pronounced with (ii) in all other Scotoh dialects, are in this 
dialect, which extends over Roxburghshire, Selkirkshire, and part 
of Dumfries, pronounced with (ei) : eye, be, bee, die, dree entktre, fee 
madf a fly, to fly, fi-ee, gi' ye ffive you, glee squint, gree a^ee, he, 
key, lie falsehood, me, knee, pea, plea, pree try, see, stee ste^, 
spree, tea, ti' ye to you, tree, thigh, three, wi' ye with you, agee 
aslant. That is where other Scots say : (ii, bii, dii, drii, f ii) ete, 
the Borderers say (ei, bei, dei, drei, fei) ete. This one pecu- 
liarity is very striking. Some of these woitls as: eye, fly, lie, thigh, 
are pronounced with (ai) in the South, but what iiglishman would 
say (bai) for hee, (frai) for free and so on ? Conjoined with this 
curious correspondence of (ei) with the (ii) of other dialects is 
another of precisely the same charactCT. The sentence : Tou and 
me will go over the dyhe and pull a pea, is a perfect shibboleth 
in this dialect. Alone, in all Scotland, it says : ( Jau an mei al geq 
«ur dhi deik an pau b pei).' On the other hand, the Edinbur^er 

^ See' Mr. Murray's paper on the « Mr. M. Bell writes (myi pyi) for 

Lowland Scotch Dialect, read before (mei pei). The latter were the sounds 

the Philological Society on the 4th as I appreciated them when Mr. Mnnmy 

and 18th Dec., 1868. pronounced them. 



288 I, Y — XIVTH CENTURY. Chap. IV. { 2. 

would Bay: (Juu en mii'al gjsq but dht dsik on puu s pii). Obeeire 
the (jau pan) for (juu puu) corresponding with (mei pei) for (mil 
pii). We have here, then, two sets of words in a living dialect 
coiresponding in precisely the same way as the ivith century 
(ei ou) with the xivth century («• uu),* and similarly in the 
Netherlands, we shall find (ei, ii) coexisting in adjacent provinces, 
as pronunciations of the written •;*. The phenomenon, then, of the 
change of (m uu) to (ei ou) ought not to present any very serious 
difficulties. Nor ought we to feel any great suiprise at Palsgrave 
and Bullokar having retained (tV uu), while their fellow countrymen 
generally said (ei ou). 

The sound (ii) for long ♦ is hy no means extinct, and the douhle 
use of (ii) and one of the (ai) sounds is, as we have seen, familiar 
in the very words which have heen noted above. Mr. Murray, 
notwithstanding his residence in England, and his critical know- 
ledge of our language, confesses that he is ** continually discovering 
words which he has all his life pronounced with (ii) which English- 
men pronounce (oi)." "In fact," says he, **long (ii) is the sound 
we instinctively associate with the letter * unless we have been 
taught to pronounce it as in English." The following is taken 
from some remarks which Mr. Murray obligingly commimicated in 
writing. 

Fly s. and v. general Scotch (flii), but Teviotdale (flei). Cleve- 
land (flii) a fly, but (flig) to fly, compare lie. 

Lie (mentiri), general Scotch, Westmoreland, and Cumberland 
(lii), Teviotdale and Dumfriesshire (lei). 

Lie (procumbere), Westm. Cumb. Lane, and Cleveland (lig, leg) ; 
this does not seem to cross the border where the word is (lai, la', 
lohi), although the older Scotch always wrote lig^ lyg. 

By preposition of the agent, (bi). Teviotdale (Hei waz sin bi 
sfverelz) = he was seen by several. 

By of place is always (bai, bohi). 

Thigh Scotch, Westm. Cumb. and Cleveland (thii), Tev. and 
Dumf. (thei). 

Friar = (friir), thus a part of Jedburgh is called the Freirs} 

Briar = (briir), Cleveland (briir) and (brii), inquire (enkwiir*), 
choir (kwiir) and (kwwr) (?), squire (skwiir). 

Site, old people pronounce (sit, zit). 

Neighbour = (nib'er), with a short vowel, not (nii'ber) as Eng- 
lishmen hear. 

Like = (lek, leik), the latter more common, but (lek'hz) is used 
for likely ; in Cleveland also, Hke = (lah'k), but likely = (lek'l*, 
lik'hy 

^ The difference between (au on) is to attempt to diBcriminate between (on 

very slight, the latter haying simply ou) in an ancient form of speech, when 

lahialised the first element of the former, it would be difficult to do so in liring 

which effect readily produced hy the pronunciation. 



action of the subsequent (u). The ' A well of yery fine water at Work- 

difference between (©u ou) is merely ington, Ci 
that the first element of the latter is the (frii-jV 




difference between (ou ou) is merely ington, Cumberland, is always called 

that the first element of the latter is the (frii'j). 

widened, and it would be presumptuous * An old Scotch jeweller, who had 



Chap. IV. § 2. I, Y — XIVTH CBNTURT, 289 

Oblige^ o^%M^=:(obliidzh% c^bluist'y and siinilarly in nmneroiis 
Prench words, as invite^ polite^ and words of classical origin as idol 
(iid'l) type (tip), baptiu, ehoBtiwe^ ewiUud (sivaliizt), advertUe-mmt, 

Eyey general Scotch (ii), Teviotdale (ei), plural in both (in) with 
shoit (i). Cumb.y Wes^., Lane, and North Yorkshire (ii, iin) 
with long (ii). Bamsley, South Yorkshire (ii, iiz). 

^^h Tev. (Hei^h, nei, nai), other Scotch (nekh, sdkh, mi), as 
(as Hi'lahnt az dhe mi rood) «= as highland as the high road.* The 
guttural form is common but is passing away, and (mi) is used 
instead in Centre, West, and North of Scotland, as also in Cuinb., 
and Westm., (nai, Hohi) are the common recent forms in Teyiotdale. 

i)w, general Scotch, Cumb., Westm., Lane, (dii). Teviotdale, 
Eskdale, Annandale (dei). 

Dree (drii) endure, and so in Cleveland ; but dry (drai drai dra' 
drohi), and so with hu/y. 

Sly follows the analogy of hiyh, but the guttural form seems only 
to occur in eleight (sleAht) like height (neiEiht). The usual Scotch^ 
Cumb., Westm., and Lane, is (slii), Tev. (slei), or more commonly 
(alai, dohi). 

Hie is not known to Mr. Murray in living speech, in reading 
ballads it is called (nai Hohi) in Tev. La Westm. dialects it is 
sometimes written hii} 

'lyht, words of this class, as right, might, light, sight, which 
in Scotland are (lekht, leibht) are in Cumb., Westm., Lane, and 
Yorkshire, (riit, niit, lut, siit) etc.* Li cases where -iyht does not 
represent ags. 'iht, the pronunciation is different, ao Jiyht ags. feoht, 
Tev. (foBiifcht), Lane, (feit) not (fiit).» 

Sigh (sekh), 

China, the ware or the country (tshin-e, tshin-t), as in (Whuht 
est vts «t Jens uut b tshiu'i tm en'B tshin't? Tei) = Wnat is-it 
that 'is at once out of China and in -of China? Tea. Walker 

lired from youth in London, always men in the Dales sounded sucli words 

said (lek) for like, in all senses. He as siffh, niffhtj ligtUy &c.j with a gentle 

was constantly using the word, and guttural breathing/' which, he adds in 

neyer seemed to hear that other persons a footnote, " seemed partly to come 

pronounced it differently. from the palate," ana was therefore 

^ Obserre the form of the past tense. (Arh). See : A Memorial by the Trus- 

I quite lately heard (obliidzh-, obliitsht*) tees of Gowe^ (Koo'gtl) ChapeL with a 

from a noble lord at a public meeting. Preface and Appendix, on the Climate, 

* Perthshire simile m describing one History, and Dialects of Dent, by 
who is ultra Celtic. Observe here Adam Sedgwick, LL.D,, senior fellow 
the different use of (as, az). of Trinity College, and professor of 

' A gentleman in Derby informed Geology m the Uniyersi^ of Cam- 

me that in North Derbyshire the bridge. Cambridge, 1868, 8yo. pri- 

peasantry say (mak mi) for make haste, yat^y printed, p. 103 — a book of affec- 

Compare : I se where come a messengere tionate and interesting reminiscences of 

tn fie^in haeie 4*10. ags. higian «. manners and speech, extending oyer 

higS 9. Omnin hih e. Fromptorium nearly 120 years, through Prof. Sedg- 

hyyn' p. 229. wick's father, the honoured clergyman 

* Prof. Sedfi^^ck, a natiye of the of Dent, who was dO years older than 
dale of Dent, Yorkshire, writing at up- his son. 

wards of eighty years of age, says : " I * Several correspondents have con- 

remember the day when all the old firmed this rule, and the exception. 

19 



290 1, Y — XIV TH CBNTUEY. Chap. IV. { «. 

gives (tBhee'nt) for china ware or orange, but (Tshai-nse) for the 
country, and has a long note on it. 

Bind, find, hind, blind, grind = (b^d, fend, Hmt, hlmd, grrnd), 
wind V. and s. := (wand), but kind, mind, wynd = (kaind, maind, 
waind), and little is often (lait'l) especially as a proper name. 

Why .' as an exclamation, not why f the interrogative, is (wt !) in 
Scotch, and (wiia !) in Cumb., Westm., Lane, and Cleveland. 
(Wiia ! sez ai) = Why! eaye 7, is a common formula in the 
Northern counties. 

Can this existence of the (ii) sound, and its general association 
with i in Scotland, be considered a modem development ? Has it 
not rather the appearance of an ancient form ? The latter view 
seems confirmed by seeing that numerous words are pronounced 
with one of the (ei) forms as (ei, ei, sei, ai, ai, ohi, ai), and that 
these various forms are differently distributed in different localities, 
whereas the (ii) form when it occurs is almost generaL Mr. Murray 
gives the two following lists of words which have (ei, ei) in Teviot- 
dale, but (ai) in Western Scotch, the first element of these diph- 
thongs being more distinctly heard than in'English (ai, au). 

Tev. (ei), west Scotch (ai) : bike wasp^i nest, dyke, fike to irk, 
like, pike pick, sike wet hollow, spike, strike, tike ; bite, clyte clot, 
dite doit, flite scold, gite ermy, kite a belly, mite, knite (kneit) rap 
the knuckles, quite, white (ktrheit), spite, suite blow the nose, wite 
blame, write (w'reit),* yite (jeit) yellow hammer, gjrpe (geip) im- 
pudent fellow, (Hei'pelt) awkward clown, pipe, ripe, sipe ooze, snipe, 
tripe, wipe ; — ^bice, Brice, Christ, dice, grice, Hce, mice, nice, price, 
rice, spice, sklice slice, trice, wise (weis), twice, thrice, fife I^fe, 
five, life, knife (kneif ), rife, strife ; — ^pint (peint), ninth (neint). 

Tev. (ri), West Scotch (ai) : bide, bride, guide, hide, pride, ride, 
side, slide, tidy, wide ; — jibe, kibe, siba (sri'ba) onion Lat. cepa ; 
— guize, prize, rise, stays (stdiz) ; — ^kithe shew, lithe, writh ; — dive, 
drive, hive, alive, lives^ knives, deprive, schive slice, strives, thrives, 
wives ; — ^tiogs (t«qz) tonys, whings (wh«iqz) shoe-strinys ; — ^brine, 
cryne dry in, fine, Ime, mine, nine, pine, sine since, swine, shine, 
tine lose, twine, wine, vine; — crime, dime, glime glimpse, lime, 
prime, rime, stime indistinct form, tune; — ^bile, file befoul, guile, 
kile hay-cock, mile, pile, sile strain milk, tile, vile, wile, stile, 
smiLe ; — ^bire cowshed, chair (tshrir), fire, hire, mire, sire sewer, swire 
tire, wire ; — ^wild, mild ; — ^mind, hmd, kind, rind, sind rinse. 

In the second list the consonant is a Uquid, nasal, or voiced 
letter, which distinguishes it from the first. Generally in Scotland 
when English long « or y is final in monosyllables, as cry, dye, or a 
long • occurs in underived words, as dial, trial, the sound is (ai), 
and in Teviotdale (ai, ohi). Derivatives follow their root sounds. 

The two sounds, that is the (ei, ei, ai, ai) series, and the (sei, ai, 
ai, ohi) series, attributed to the Scotch long •', are strongly insisted 
on by Scotchmen, and in 1848 when I was printing much English 
in a phonetic form, the Scotch always exclaimed against the use of 

^ In Aberdeen (triit) or (bbriit). 



•■ aT6' - 



Chap. IY. } 2. 



I, ¥ — XIVTH CEKTURT. 



291 



one fflgn for the two forms. The late Professor "W. Gregory, of 
Edinburgh, divided the sounds into (ai) and (ai),^ in which case 
they answer to the two sounds heard in haiah in England. Mr. 
MeiYille Bell in a private letter says that: ''in different districts 
you hear {a\ a,\ ahi), but the rejH^sentative sound is (cei). This is 
heard regularly when the sound is £nal, before a vowel, or before 
final r, and generally when it occurs before (z) or (v). This (aei) 
is the ' genteel' form of i. I hear it from all my educated Scotch 
pupils ; though they come from widely separated districts they give 
(aei) for *!' etc., with absolute uniformity.' The other soui^ (ei) 
is the regular one for t in other syllables, and in a few words for 
d," as aye, pay, clay, Tay, May, way, plague, etc. In Teviotdale, 
«ytf, may, are called (^ mn) to distmgnish them from (ei, mei) ^^ 

My dialectic correspondents (p. 277 note), and Mr. Murray have 
frimished me with the following words in which (ii) or (»V)* re- 
mains in the provinces. Abbreviations — C. Cumberland, I). Devon, 
Db. Derbyshire, K. Kendal, L. Lancashire, N. Norfolk, 8. Shields, 
generally South Shields, sometimes North Shields, and occasionally 
Newcastle, Sc. general Scotch, W. Westmoreland, Y. Yorkdiire, 
Yc. Cleveland, Yorkshire The list is of course very incomplete, 
both in words and localities. The numerous French and classical 
words pronounced in Scotland with (ii), p. 289, are omitted. 

Words spelled with I, ttsvallt soxtkded (oi), but Pbovikgiallt 

FBQNOITirGED (ii). 

alike D fly «. CESScWT liar 8 ai^t CW8 

briar CTc fly 9, CKLSScTTo lie 8, CBXSScW aly CLSScW 

bright CKLSW fiiar CSc hgbt CDWSY atile C 

hj preposition of fright 8 lightning 8 thigh CS8cWYc 

apeni 8c hie Db mice DN thy LW 

child D lugh C might «. D tie r. CKL 

die CKL8ScW hind $. mind D whj I CLSoWYc 

dry 8 Ide D my paatim wnght 8T 

dyke N I'll C night CDELST write 8 

eye CDL88cWY kindly D nighest (niist) D 

^eaight Y kite Y right C8WY 

It would be difficult to suppose that in all these cases, widely 
differing from ordinary use, and extending over several counties, 
the (ii) should have been a recent transformation of (ai). The 
probabOities are all the other way. 

The personal pronoun / is one of the greatest difficulties. In the 
Aryan languages its changes have been great. The original word 
seems to have been (a) to which a strengthening termination (gham) 



^ 8ee my E9$mtialt of PhoMtietf p. 
172, notCf where (ai) is used when not 
followed hv a consonant and before the 
inflectional (d, z), and also before (v, z), 
but otherwiae (oi) is more common. 

> Mr. Morra^r accounts for this ab- 
normal uniformity, by sayine that (sei) 
ii not a Scotch soun4 but tne Scotch 



conception of the proper prommeistion 
of the English long t. In England 
(sBi) is rather cockn^ed. 

* It is impossible to tnut Hie unac- 
customed ear to distinguish these 
sounds, though they have separate 
letters I, •', in Icelandic. 



Chaucer as we have seen (p. 21 
sometiiaes palatalized to tek (it 
and Gt>wer is /.* By Shakspep 



292 I, Y — XIV TH CENTUEY. Chap. IV. §2. 

was affixed, producing (agham) as in Sanscrit.^ The vowel (a) was 
retained, and the following gattoial altered to a sibilant in Zend, 
Lithuanian, and old Sclayonic. In Greek, Latin, and Gothic, the 
euttnral was retained, but the vowel palatalized, into (e) in Greek 
eya>v (eghoon*), and Latin e^^o (eg'oo, eg'o) which rctcuned por- 
tions of the following syllable, and into (i) in Gothic (ik), which 
dropped the following letters. This low Geiman form (ik) was the 
noraud Baxon form, probably («k), and the orthography iec in 
Orrmin, guarantees the shorbiess of the vowel. Li Icelandic we 
find ecy eky ig^ where the vowel seems to have become long, and (j) 
was prefixed in speaking. The Modem Danish is^V^ (jei, jai). In 
Chaucer as we have seen (p. 282), the form ie still occurs, and is 

ytsh), but the usual form in Chaucer 
Shakspere ^e words /, ey$^ ay$ were identi- 
fied in jound (pu 1 12). The foequent phrase qMoth-ay may some- 
times mean, qttoth /, but is often interpreted quoth hs, and the well- 
known passage in Henry Y, act ii, sc. 3, describing the death of 
Falstaff, is fall of a for he. Now as hs was certs^y generally 
pronounced (nii), as it was frequently written hee^ at that time, the 
provincial, or vulgar, or dialectic correspondence of (a) with (nii), 
would be precisely similar to a dialectic use of (a) for (tV), sup- 
posing the last to have been Chaucer's personal pronoun. At the 
same time the .acknowledged form ^mi) for hs^ would lead us to 
expect some acknowledged forms (li) or (ft) for /, existing in 
difdects. 

Now both of the forms (a) and (ii) exist in the provinces for I, 
though the traces of (ii) are very few and very slight, but few as 
the^ are, it would be difficult to account for tiiem except by the 
action of an old tradition, and as in some cases the pronunciation is 
only known among very old p^ple and is fast gomg out, it may 
have been much more common as lately as one or two hundred 
years ago. 

**£edtrsl had: If eed done soa, it wad sartainly hev been 
better."' "/, offs, eigh. Yes. I is sometimes pronounced like E, 
particularbr when the pronoun follows the verb, as ' do E,' for I 
do." * " 1 is often sounded like E, in «n," • probably (t) as a con- 
tracted form of (m). 

^ F. a August lick, Wdrterbvfih * Bsv. W, Carr, CnT«n Gloaraiy, 

der Indogennanischeii Omndsprache in toI. i. p. 127, 2iid ed. 

ihiem Bestande tot der YtflkertreB- * Ibid., p. 241. The author cites as 

nung, 1868, p. 4. C. F, Koch, His- an illnstration, what looks like a conp- 

torische Granunatflc der Englischen let, from Cant Tales, 12630, by which 

Sprache, toI. 3, p. 3. it seems as if iim, / rhymed. Of conrse 

* The onussion of the gattnral ii this was not the case. The author has 

quite similar to the (ai, i, mi. di, si, taken together two Unes belonging to 

aa, do, no) for euch, ich, mich, dich, different couplets, and the whole rhymes 

sich, auch, doch, noch, in the neigh- taejolitc mc, IthrifiUy, 

bourhood of the Danube, Bayaria. ^ Ibid. The author has unfortunately 

SehmcOer, Orammatik art. 427. So in not followed any strict orthography, and 

old high German, and old English we has not attempted to explain that which 

find ine for ih ne, ic nc, GrafL 1, 118, he has used. 
Ed. Ant. I, 235. 



Ohap.IY. {2. I, Y — XIVTH CBNTUBY. 293 

In Lancashire (t ) is used when unemphatic, as (mon i tel dhe ?) 
must I tell you.^ 

In Blackburn " the old fashioned way'' oi pronouncing I" is 
(♦) ve^ short."* 

"I have freouently heard old peo^e pronoimce /like our own 
ee (ii), especially in the interrogatiye form, did ee do it? will ee 
go? must ee do it ? etc. This is very common, in fact about twenty 
years ago it was the invariable pronimciation. In the phrase : (aiz 
gaa*an mam, at tz ii !) = 7 am going home, that am /, ee (ii) is 
as decidedly emphatic as I ordinarily is. The contraction Pll for 
/ ihaH^ is frequently given ee^ll. Ee is also used occasionally 
but very seldom in eveiy tense and form. This pronunciation is 
only used by old people nere, but in central Cumberland it is more 
general lAie same people use the form (aa) and sometimes (a), 
but never in questions or in the direct ftiture."' 

Scarcely 1^ convincing as respects the vowel in English ieh are 
the contractions eham^ ehae, ehU (tsham, tshas, tshtl) for ich am, 
ich was, ich will, mentioned by Gill {Zo^onomta p. 17) as a Southern 
mronunciation, in Eev. W. Barnes's edition of the Glossary of the 
Dialect of Forth and Bargy, and in the Glossary to his Poems in 
the DOTset dnlect, 1858, p. ISO. See also J. Jenninge^ Dialects of 
the West of England.* 

The dialectic pronunciations Jm, ^eh are preserved in Shakspere, 
King Lear^ act iv, sc. 6, L 240, Globe ed., Tragedies p. 304, col. 2, 
folio 1623, which reads : 

Edg. Chm> not let go Zir, 
Withoat Turtber 'cuion. 

8t$w, Let go Skoe, or tiion dj^A. 

E4g. Good Qentleman goe your gate, and let poore volke paflb : and 'child* 
ha'bin zwaggerd oat of my life, twonld not ba'bin so long as 'tu, by a 
Tortni^ht. jNay, come not neere th'old man : keepe ont cbe Tor'ye,^ or ice * 
try whither your Coetaid, or my Ballow be the harder ; chill^ be plaine with yon. 

8t$w, Out Dunf hill. 

Bdg, Chill* picke yonr teeth Zir : come, no matter tot your foynes. 

About thirty years ago utehy (otsh't ?) was in use for / in the 
Eastern border of Devon^iire and in Dorset, and examples of dum^ 
ehould = I am, I would, occur in the ** Exmoor Scolding," which 
dates from the beginning of the last century.* 

The prevailing dialectic forms of the pronoun are however (a, a, a, 
oh) occasionally (a, «), and (ai, ai, ohi, aI, oi). In Derbyshire I 
generally heard (a), but in the northern parts it is said to be (aI^. 
Mr. Murray writes : '^ / in the Northern dialects of England is 

1 Letter from Mr. John J. L. Jack- • I will, 

•on, teacher of languages, Manchester. * I wonld. 

s Letter from ^.T. Fielding, Man- ^ Printed eA«iwf»y« in i&e4to, 1608. 

Chester. **Autt rale 9 ( Tshi toot n\ pro (ai 

* Letter from Mr. J. N. Hethering- war-ant Jon) e$rtum d&,*' Gill, Zo^ 
ton, Clifton Parsonage, Workington, nemia^ p. 17. 

Cumberland. * Ioe«Ise»I; printed iZ^sru, in 

* For these references to Glossaries the 4to. 1608. 

I am indebted to Mr. W.Aldis Wright, * Letter from Mr. John Shelly, 
Trin. ColL Cambridge. Plymovtit, 



294 I, Y — XIV TH cmrruRY. Chap. lY. { 2. 

nsuallj a simide vowel of the (a, a, oh) series. In some dialects it 
iSf when accented, a diphthong composed of the same first element 
and (i,'). In Scotch (oh, aa), even when emphatic ('oh wohd'ne 
gohq) = I wonld not go. In Ayrshire it would probably be (aai, 
aa'j) in such a case, so also in dumb, and Westm. In Lancashire 
it is (aa) even when emphatic, in Bamsley, Yorkshire, (aa). When 
nnemphatic it is in all the dialects an obscure (a, s, «), it is hard 
to say what." Unemphatic syllables have always a tend^icy to 
fall into this colourless (e, «) sound. Even in Germany, where 
there is no tendency to pronounce ieh (ikh) with an (ai), rapid 
speaking will generate (9), as (nab'adi, laa'omi, taa'tado, deqk-emo) 
s= habe ich dich, lasse ich mich, thate ich dur, denke ich mir, in 
Bavaria.^ 

The confusion of (•*) with (e) penetrated, as we have seen, into 
orthography, p. 272. But during the xv th century there also arose 
a tendency to thin (ee) into (ii), whereby so many (ee) of the 
xiYth century became (ii) by the XTith. This tendency was pre- 
cisely the same as that which converted so many of the remaining 
(ee) into (ii) at the beginning of the xnnth century, p. 88. Kow 
if we suppose these two tendencies to act together, which is no 
extravagant hypotheais, since they certainly co-existed, the result 
would be that (n ) would be begun as (ee) and ended as (ii), that is 
that (tV) would become first (eei) and then (ei). During the same 
time we know also that (00) was in many instances refined to (uu). 
We might therefore suppose that there was the converse tendiency 
to take (uu) as {uu), and then as (00), which is by no means un- 
common, and then that the joint action of these two tendencies pro- 
duced first {oau), then (<?u) or (ou) as it would have been certainly 
accepted. This supposition as to the mode of generating (ei, ou) 
from (ft, uu), has the advantage of being based upon known &cts. 
But tiie considerations adduced on p. 233, are quite sufficient to 
account for the change. At the present moment the {ee, 00) of the 
South of England are actually changing into (ei, ou), and these 
sounds have been developed by the less educated, and therefore 
more advanced speakers, the more educated and therefore less ad- 
vanced having only reached {eei, oou)' although many of them are 
not conscious of saying anything by (ee, 00). 

^ SehmiUer, Hnnd. Bay. art 384. thongs. This is iUnstnited .... in tiie 

* **The English alphabetic aooented^i, regular pronvniciation of the Towels in 

in tiie mouth of a well-edncated Lon- M<, ail, aim, ache, &c. (a), ode^ oak, 

doner.... is not qnite simple, but finishes alobi, &c. (ou). The same tendencj 

more slenderly than it begins, tapequg leads to ihe * Cockney* peculiarity of 

BO to speak, towards the sound (i) .... separating the labio - linfi:ual vowek 

o in a Londoner*s mouth is not always (u, o) into their Ungual ana labial com- 

quite simple, but is apt to contract ponents, and pronouncing the latter 

towards the end, finishing almost as oo suooessively in^bead of simultaneously, 

in too,** B. H. Smart, Walker Be- Thus we hear (oh, tra, yu) for (u), and 

moddled, 1886, Principles, arts. 1 and (o*w, o*io, ah*w) for (o),*' Visible 

7. Mr. M. Bell, among <' English Speech, p. 117. As Mr. Bell marks 

Characteristics" reckons : " The ten- the second element by the glide sign 

denoy of long vowels to become diph- he does not distingnian the length of 



Chap. IV. { 2. 



I, Y — XIV TH CENTURY. 



295 



As has been already remarked, p. 234, the change from (ii, uu) 
to sounds of the (ai, an) order has not been confined to England, 
but took place in the literary language of the other Germanic 
countries, nearly at the same time, that is, during the xy th and 
XTith centuries; and in these countries as well as in England 
traces of the original pronimciation remain in the provinces. 

Siegenbeek, whose work on Butch Spelling originated the ortho- 
graphy now in use, tells us that old Dutch manuscripts employed 
t, t«, for their long «, which, partly for distinctness and paitly for 
ornament, became i;*, and hence that the inhabitants of Eriesland, 
Zeeland, Guelders, Overyssel, and Groningen, who still pronounce 
(ii), evidently preserve the ancient sound ; but that the inhabitants 
of the province of Holland had at an early period changed the 
sound into one very like (ei) ^ and that after the Spanish disturb- 
ances, that is, about the end of the xvith century, this province 
having become the seat of learning and civilisation, its pronunciation 
necessarily became prevalent, and is now the literary pronunciation 
of the country.* Hence we have an indubitably ancient (ii), pre- 
served in those provinces of the Netherlands whose dialect most 
resembles ancient English, and passing into an (oi) in other pro- 
vinces which by a political accident was able to set the fashion of 
pronimciation. 



the first element, so that with him (ee^ 
00^ haye alreadj in appearance become 
(a, 0a), bnt this does not represent his 
actual pronandation, which is rather 

1 The Bntch if, #» differ slightly, 
if at all. Sir Hendrik Gehle, D.D., 
minister of the Dutch Reformed Chorch 
in Aofttn Friars, London, who kindly 
pointed ont to me the passage in Sie- 
genbeek (Sii'ghenbeek) refeired to in 
the text, and confirmed what is there 
said of tiie prorincial (ii), said that he 
fdt more of the e in pronouncing #» 
than ^\ reminding me much of 6ill*s 
remark (suprii p. 114), of beinff diffuse 
oyer the e. At first he seemed to call 
both fei), bnt afterwards he recognized 
my (9\f ei) as the two sounds, and, as- 
suming tne EngUsh as (ai), he said he 
considered the Dutch a neater sound. 
The distinction (ai, ei) is precisely that 
which I had to make in Gill, an<( con- 
sidering the close connection between 
Dutch and English, the coincidence is 
remarkable. 

* " Doch deze enkele t kon geene 
plaats hebben in lettergrepen, op eenen 
medeklinker stuitende, als mi^n, s(^ 
bin/ en soortgeliike ; maar moest luer 
noodzakeUjk yerdubbeld worden. — Men 
schreef dus oudtijds, met eene dobbele 
♦, bHif, iw tn, $ehriiff Yon welke schrijf- 



wijze, in oude handschriften, nog vele 
iporen yoorhanden zijn. Doch, om de 
gelijkheid der dubbele t met de u, 
waaruit ligtelijk rerwarring kon ont- 
staan, en miaschien ook sieraadshalye, 
begon men de tweede t reeds vroc^ 
met een' lan^n staart te schrijyen^ H 
welk man, bij hare platnng y66r eine 
Tokaal aan het begin der woorden, ins- 
gelijks in zwan^ bragt Wij kunnen 
niet Toorbij, hier te doen opmerken, 
dat zij, die, in de woorden blifven, 
tehruven, mifn, t^fn, bij de uitspraak 
den klank der enkele en dubbele t doen 
hooren, als de Vriezen, Zeeuwen, Qel- 
derschen, OTerijselschen en Oroningers, 
blijkens het yoor^estelde, de echte en 
oorspronkelijke uitspraak dezer woor- 
den behouden hebben. Doch op de 
tong der Hollanders is deze echte 
klank reeds yroeg verloren geraakt, en 
Toor eenen anderen, eenigzins zwe- 
mende naar den klank d; yerwisseld 
geworden. Nadat nu Holland, wer- 
waards, na de Spaansche beroerineen, 
de Toomame zetel der beschaaidheia en 
wetenschappen word oyergebragt, door 
middel van dit uitstekend Toorregt, 
zijne uitspraak meer en meer als de 
auremeene en heerschende heeft doen 
geiden, is ook die yerbastering in de 
meest beschaafde uitspraak en daarop 
gebouwde schrijfwijze ingevoerd, en 



296 I, Y — XIV TH CENTURY, Chap. IV. { 2 

We haye predsely the same phenomena in the less closely related 
High German dialects. An old and middle high Oerman t (ii) 
he^une a modem High German ei (ai). All these latter ei are how- 
ever not derived from I (ii), but some come from a middle and old 
High Gennan 0% (ei), answering to the Gothic at (ee).^ Moreover 
we have the same phenomenon of a persistence of the sound of (ii) 
in the provinces, notwithstanding tiie real change of orthography 
fix>m f to eiy whereas in Dutch tiie change is only apparent, from 
ft to f;\ and hence resembles the English retention of t through a 
change of sound. Schmeller says : ^^ei sounds, conformably with 
its origin, like a long (ii) b^ the lake of Constanz, i.e. on the Upper 
Rhine, and by the tnbutanes to the W^ser from the Bhon-chain of 
billa J* (miin, Hiin^ siin,— bii, drii, lis, Fliis, Uim, liiib, bhiis, Tsiit 
— bis'e^, blii-be^, grif-e,, ii'le,, lii'de^, shnii*def, shrii*be^, trii^bej, 
szs mein, dein, sein, — ^bei, drei, Eis, Fleiss, Leim, Leib, weiss, Zeit, — 
beissen, bleiben, greifen, eilen, leiden, schneiden, schreiben, treiben. 
Also on the Lauter (siin) for seyn^ on the Hz (n^) for ein, as in 
(iifSpan*e J «=> einspannen ; on the east of the Lech, (drii)-£Etch, 
(drii).ftiesz, (shliif )stain."» 

Dr. Bapp in the passage previously cited (supr& p. 235) has 
endeavourod to ^ve the relations of all the long vowels throughout 
the Germanic languages, and it seems worth while to reproduce his 
table here, although it is only a sketch, and requires much filling 
in to make it at all complete. The first line gives what Dr. Bapp 
imagines to have been the seven primary vowels in this system of 
languages. The lines 2 to 6, refer to the older, the lines 7 and 8 
to &e intermediate, and the following lines to modem forms. The 
pronunciations assigned may be occasionally disputed, but they are 
near enough for the present purposes, and wiUiout attempting to 
make any change, I have translated the phonetic symbols as well as 
I could understand them. The unifonnity with which the Ger- 
manic, as distinguished from the Scandinavian, branches have in 
recent times adopted the (ai, au) forms in place of (ii, uu) is very 
striking. Many persons may feel that it is an argument in mvour of 
the pronunciation of • long as (ii) in Anglosaxon, and therefore in 
Early English, that the Scandiniavians certainly ciedled their long t 
(ii), as their descendants in Iceland, Norway, Sweden and Denmark 
continue to do. But that conterminous districts may differ precisely 
upon this point we have already seen in the case of Scotland (p. 287) 
and Holland (p. 294), and another instance may be cited from the 

daarin reeds eoo Tast gewortcdd, dat het Spnohe, iii, 267. (Ttmum, DentMhe 

thans Tolstrekt onmogeliik is, deeeWe Gram., 8rd ed. i, 285, 317. 

«iit te roeijen." Verlmnaeling orer de ^ SfipPi Phys. d. Spr. ir, 11. Orimmy 

Nederdnitsche Spelluig ter bevordering ib. 96, 106, 176, 182, 226. Grimm 

Tan eenparigheid in deaelTe, door aasamea Gothic #•; dt « (ei, ai) appa* 

M*tthM» 8%sg$nb0ekj hooffleerar in de rentlv ; in Chap. Y, } 4, No. 3, the 

Kederduitsche LetterkundbB te Leyden : aonnos (ii, ee) are prefored. 

uitgegeren in naam en op last ran bet * In the same oistriot, au sounds as 

Staats-Bewind der Bataa&che Bepub- (uu) conformably with its origin, 

liek. Amsterdam (1804, 8yo., pp. > MMmdtartm jB^pmet Art 244. 
380), p. 66. Bee also Sofp, Fhys. der 



Chap. IV. { 2. 



I, Y — XIV TH CEKTUBT. 



297 



Norman peninsnla oontaining Cherbourg. At Montebourg, only 
fifteen miles SSE of Cherbourg, the pronunciation of t as (ai) is 
very common, whereas at Beaumont Hague, on the same penmsula 
and only twenty-five miles NW of Montebourg, this pronunciation 
is unknown.^ Such examples shew the necessity of examining 
ftxiHtiug phases of pronunciation before attempting to decide upon 
extinct usages. 



BxLAnoirs of the Setxn Loire Yowels dt thx Oxbicaitic 
Lanouaoxs aocobdino to Db. M. Kapp. 



Long Vowels. 


I 


n 


m 


IT 


V 


TI 


Tn 


1. Primary - - - 


aa 


ee 


ee 


ii 


AA 


00 


uu 


2. Gothic - - - 


ss 


ee 


iu 


ii 


AA 


00 


uu 


8. Icelandic • - - 


AA 


ei 


iu 


ii 


OVL 


00 


uu 


4. Anglosaxon - - 


ee 


AA 


So 


ii 


^ 


00 


uu 


5. Friesian . - - 


ee 


ee 


ia 


ii 


AA 


00 


uu 


6. Old Saxon - - 


aa 


ee 


iu 


ii 


AA 


00 


uu 


7. Middle Saxon - 


AA 


ee 


ee 


ii 


00 


00 


uu 


8. Middle German - 


aa 


ei 


ie 


ii 


OVL 


no 


uu 


9. English - - - 


ii 


00 


ii 


ai 


ii 


uu 


au 


10. Banish - - - 


AA. 


ee 


yy 


ii 


OBOB 


00 


uu 


11. Swedish - . - 


00 


ee 


JJJU 


ii 


oeoe 


uu 


ru 


12. Dutch - . . - 


aa 


ee 


ii 


ai 


00 


UH 


py 


18. High German 


aa 


ai 


ii 


ai 


au 


uu 


au 


14. Suabian - - - 


AA 


oi 


ia 


ai 


au 


U9 


au 


16. Frankish - - - 


00 


ee 


ii 


ai 


aa 


uu 


au 


16. East Frankish -' 


8U 


aa 


oi 


ai 


aa 


au 


au 


17. Bavarian - - - 


AA. 


09 


ia 


ai 


aa 


ua 


au 


^Bomploi. - . - 


Jahr 


breit 


Dieh 


weii 


Lauh 


gui 


Maui 


English - . . 


year 


broad 


thief 


wide 


leaf 


good 


house 



Although the subject is far from exhausted, as we are thus 
led into an examination of the cognate dialects, sufi&cient has 
been adduced to shew the antecedent probability of the 
theory that in the xiy th century long i was Dronounced as 
(if), and as all the facts which we have been able to discover, 
agree with and are explicable by this theoir, whereas the 
uffoal hypothesis that lon^ t was one of the (oi) diphthongs 
during all periods of our language, is not reconcilable wim 
many of the £Etcts adduced, and is opposed to the general 
t^idency of the cognate dialects on the continent, it seems to 
be the only legitimate inference that in Chaucer's time long i 
was («*) and ^ort (t) was (t). 



^ This corioiu £Mt ii siTen on the 
autiioritj of Dr. Le TaiUu, mayor of 
Beaumont Hague, Imt a natiye of 



Montebonrg. See the note on M. Le 
H^richer and Norman k at the doee of 
Ch^. y. } 1, No. 8. 



298 U — XIV TH CENTURY. Chap. IV. { 2. 



TJ — XIV TH Cbntuet. 

After the lengthened proof which has been given that long u in 
the XYi th century had ^e French sound (yy), it follows almost as 
a matter of course, that those words in Chaucer which have long 
u, and which are as a general rule all taken &om the French or 
Latin, had also the sound of (yy),^ and this will be Airther con- 
firmed when we find that (uu) tiLe only other sound it was likely 
to represent had a different symbolisation, ou. We may, how- 
ever, notice the pure French rhyme— 

Another day he wil par adv$nturg 

Beclayme the, and bring the to Inre. 17008 

compare by aventure 25, the English phrase. With this French 
sound there was also a tendency to dwell on the syllable ure with 
more accentual stress, so (naa'tyyr*) 11, and 
Venus, if it be youre wil 
Tow in this gardyn thus to transfigur$ 
Biforn me sorwfiu wrecched creature. 1106 

Short u was properly (w) or (u) as in the xvi th century, and as 
in the Anglosaxon times. This we see from the Latin rhymes — 

Sayde Plato. Ye, sire, and is it thue $ 

Tms is ignotumper ignotiue, 13384 

In which I pleyne npon Virginius, 

And if he wile seyn it is nought thut, 13582 

At the same time we find u short occasionally used as a substitute, 
apparentlv, for e and « short, where we cannot imagine that a dif- 
ference or pronunciation was intended, as for example in the verbal 
termination -^, hathud 3, enspirud 6, eiud 29, while in the same 
passage occur perced 2, engmdred 4, temed 39. Li connection with 
the common forms list^ lest should Ittst 102 be taken as different, or 
as another way of writing the same soimd ? Suster 1835, 8465, 
seems to have some claim to be called (sus'ter) on account of the 
form »08ter 3486 rhyming with Pater-nosUry and the Anglosaxon 
form suster as well as swsosUr, swyster^ but it may have been like- 
wise generally called (sfs'ter). 

In fithfd 298 = fiddle, fadur 100 = father, gult 10142 = guilt, 

^ Mr. Mnrraj informs me that u stdll when ue is final, and where 0w is pro- 
retains its French sound in Scotch in nounced (iu) in English, whether de- 
words taken from the French, as : tune, riyed horn French or Anglosaxon 
lute, cure, sure, Bruce, reduce, conduce, sources, it is sounded (ra) or rather 
consume, assume, hruise, jud^, endure, (yn) with the accent on the first element, 
rude, mute, secure, use, abuse, suit, as in : blue, due, duty, sue, ensue, hue, 
mule, ftde, just, fis the Cockney (dzhist) few, dew, rue, crew, blew, flew, grew, 
a corruption of (dzhyst) P it looiu very threw, brew, drew, view, new, clew, 
Kke it,] justice, humour (ymar), ulzie Jew, rule (nul, ryul), sew, skew, 
(yli^ ylit) oil, and similarly &, 912 are beauty, feu, feud, feudiu, queue (kyu), 
representatiyes of Qi, nj), changed in lewd, ruin (ryu'tn), Euen (Tu'en) not 
some districts into (li, nn in : assuilzie (Ju'en). But the mew of the cat, and 
acquit^ tuilzie a quarrel, ndlzie contente toew of the kitten are in Teviotdale 
of the pariah duet eart, the toon^e fuileie, called (mffiu, wen), 
gaberluinzie walletf cuinzie eoin. But 



Chap. IV. § 2- ^ — XIV TH CEKTURT. 299 

fwr%t 1920 » first, compare ferti 530, KM 16699 :» held, hMm 
15802 = helden, hMe%1^2\ =s hills, pui 14962 »= pit, and many 
other cases there seems to be no doubt that u must be read as i or e, 
C?ompare Cantwrhwry 16, with : from Canturherf/, the more m^ry 803, 
and this again with the three rhymes — 

And thuB I lete him sitte in the piris 

And January and May rompge miiye, 10091 

thow pHoete Marcian, 
That writest ns that ilke weddynf merye 
Of hir Philologie and he Mereuru. 9606 

Him thought uat how the wenged god M$ret»rie 
Byfom him stood, and bad him to he murye, 1387 

Here we have all three spellings miryey merye, murye of the same 
word, the first rhyming distinctly with t short or long, (•*) or {if), 
and the two last rhyming with u long which we must consider as 
(yy). Now in the Schipmannes Tale there is occasion to mention 
the town of Bruges, and we find it spelled Bruges 14466, but 
Briggee 14472, 14669, 14712, which must have been intended for 
the same sound. EecoUecting that the soimd of (y) short is in 
Sweden, Denmark, and most of Germany scarcely ' distinguished 
from (t) short, into which it very often entirely falls, it occurred 
to me that the explanation of this use of u short as • might be a 
similar vagueness or indistinctness of pronunciation, and that the 
scribe, writing from dictation, either actual or internal, (for it will 
be found that the copyist usually pronounces the words to himself 
as he writes, with a mental effort which reproduces the sound to 
his consciousness although it is externally inaudible, and although 
the organs of speech are not even put into the corresponding posi- 
tions), feeling doubtful, ocaasionally wrote «, but generally % or e. 
This theory supposes that the (y) was a known English sound, and 
that the u represented the Anglosaxon y. In the words busy, bury 
where the old u spelling has clung to the words notwithstanding the 
(t, e) sounds, we have y in Anglosaxon hysig, hyrigean. Trust is 
marked by Salesbury as having the sound («), and so it has in 
Scotch, where (pit) or (pet) is also said occasionally ior put. This 
again caUs to mind the East Anglian (ktver)* for (kuver), now 
(k9v*i) mn ewer, mentioned in Gill, and also his denimciation of the 
Mopsey transformation of (butsh'erz meet) into (bitsh'erz miit). 
There would seem therefore to be some physiological connection 
between u short, and t short, which must be sought for in the eleva- 
tion of the tongue, both being high wide vowels, although (t*) is back 
and (») fit)nt, (w) round and (») primary. 

This theory that, when short u stood for short t or e, it was in 
fieust meant for the short sound of the French u (y), of which the 
long sound was at that time represented also by u, wiU receive ad- 
ditional corroboration in the next chapter. 

^ The East Anglian Fromptorium spellings fydyU fiddle, fadyr fother, 

writes euvtrynge, and, in connection g^^lte goilt, yW-^^ first, hyUi/8 hills, pyt 

with the words we haye heen preTiously pit, j^t put, lytty lusty lusty, cyityr 

considering, it is interesting to note the lister, Mereurye Mercury, myry merry. 



300 



U — XIV TH CENTURY. 



Chap. IV. { 2. 



In Treuisa's Higden, taking the chapter 59, Be Incolarum IdnguU 
and comparing the text in Mr. Morris's Specimens of Early EngHsh, 
p. 338, taken from the Brit. Mns. MS. Tiberius, D. yii., with the 
Harleian MS. 1900, and Caxton's edition (Brit Mus. 0. 21. d) I 
find the following spellings : 

Tib$rmt 2>. vii, SttrleMH, 1900. CaxUm, 

bu]? be^ ben 

furste first first 

btir]?etonge birVetonge langage 

snathe si]^pe syn, syth 

lumede lemed lemed 

wondnr wond^ wonder 

nndurstonde]? vnd^rstonde^ ynderstande 

This comparison at any rate shews that different scribes had a 
different feeling as to me vowel that should be employed, and 
proves the practical identity of this short u with short i or e. If 
any one will resolutely say,' (byth, fyrst, byrth'etuq, syth-e, 
lyr'nede, wun'dyr, un'dyrstondeth), and then compare his pronun- 
ciation with provincial utterances of the same words, which are the 
best living representatives of the ancient, he will be better able to 
appreciate the trouble of the scribe in selecting the proper letter, on 
the theory here advanced. It must be borne in mind that the 
scribe was quite fGuniliar with long (yy) and had a letter for it, u, 
and that he had no other letter for short (y) but the same ti, 
although he had three signs for short («), viz. ti, o, ou. In such a 
case he most probably felt it to be a greater liberty to use ♦, or e, 
than u in many words, although, to avoid the ambiguity of sound 
(y, u) in the letter i«, he often employed ♦, e. 

Although it is of course possible that there was a dialectic West of 
England pronunciation (i#) which replaced (y) or (♦),* it is at least 
extremely doubtM, and certainly cannot apply to the indifferent 
use by the same writer of ti and e in similar situations in the same 
sentence as already pointed out (p. 298). 

pronimoiafion especially in the 

distinction of long and short" See 
BUpr& p. 176. 

> Mr. Barnes, in his Poems of Eoral 
Life in the Dorset Dialect, 1848, p. 81, 
says 1**17 in umU^ will, is rather un- 
settled, being mostly sounded in the 
Yale of Bladcmore as « in kuU (« ) ; 
but in some parts will is tm/, « in luU 
fa), and sometimes vmll with the ti of 
German mmller (y). ... In the Yale 
of Blackmoor wul is at different times 
ipddlf ipull and taull (wid, wal, wyl) 
even in the same month." In tiie in- 
troductory letter to Nathan Hoge's 
Letters in the Deronshire Dialect^y 
Mr. Henry Baird, of Exeter, 1847, 
12mo, pp. 51, 1 find the following or- 
thographies Idndly interpreted for me 



^ Without eoxffiiderable praotioer an 
Englishman may find the distinct enun- 
ciation of these words Texr troublesome, 
especially when he feels bound to keep 
himself clear of (u, t^ e). The true 
short (j) in a dosed syllable is an 
especial stumbling block to English- 
men. Prof. Max Miiller, gets so often 
called (Mal*i) and (Midu), that it is a 
pity EngUsh people do not know that 
these sounds would be unintelligible in 
Oermany, where their own (MtVi) 
would be readily understood. Even 
Wilkins, who Uved at a time when we 
know from Wallis that (yj) was a 
common sound in England, and who 
must have constantly heard the sound 
from Wallis himself says that this 
Towel is of ** laborious and difficult 



Chap. IV. § 2. BIT, EW — XTVTH CBKTtJBY. 301 

The oonolusion is that XT in the xiy th century was gene- 
rally (yy, u), but short XJ was occasionally employed for (t, e), 
which were generally sounds into which a more ancient, ori- 
ginally Angh>saxon (y), had fallen, although through errors 
of the scribe XJ was employed in many words for I, E simply. 

EXJ, EW — XIV TH Century. 

In the xvith century there were two pronunciations of this com- 
bination, as there were also in the French language, (yy, eu). The 
following lists may be collected from Chap. III., under Ibe headings 
0U (p. 137) and u (p. 163), where the italicised words in &w are 
now spelled with ¥$. 
En = (yy) ; hhw, brew, glewe^ knew, mew (of hawks), new, r0w$ 

(a plant), slew, anew, irmoe 
Eu = (eu) ; dewe (moisture), ewe, fewe, to hew, mew (of cats), 
sewer (a waiter), shew, shrewe, strew 

Bhymes in ^t^ are necessarily few in number. I have noted 
rather more than thirty in the Canterbury Tales. For the purposes 
of comparison an alphabetical list of all the words in these rhymes, 
indudmg one Latin word, and a few words whose spellings seemed 
of importance, though they do not occur in rhyming ^Uables, has 
been annexed. Against each word its pronunciation m the xnth 
century has been written, when it could be ascertained, on the au- 
thority of Bull. (BuUokar), But. (Butler), G. (GiU), P. (Palsgrave), 
Sa. (Salesbury), 8m. (Smith). The immediate ags. (Anglosaxon), 
or fr. (French, often old French), origin follows, together with the 
ortho^phy, when it could be found, in the Pr. (Promptorium), the 
first being the reading in Mr. Albert Way's text, and the sub- 
sequent ones those which he adds from other MS. Next follow the 
rhymes in which the word occurs, with its orthography in the place 
and the reference number. By this means a complete compaiatiye 
yiew of all the words is fdniished, which will enable us to draw a 
satisfSactory conclusion. 

by Mr. J. Shelly, of Plymouth, in may not be the case, for Ttel, spel) may 

Wnich M is apparently nsed for (a, o, u, be representatiTes of (t?!, sp^I). The 

Ji yy» *» ^) ; ^^^"'^ C^*") f**', vury (vwri) DevooBhire (y) is here seen to be nn- 

Tery, gude (gMd) good, du (dyjr, dy) do, certain and to admit U) as weU. The 

purmotmg (peimoot'tn) promoting, dude same is the case in iforfolk. Mr. M. 

(d^) did, yu*w {Tjyr) yonWe, mv (ev) BeU hears French u as (#). In Nathan 

0^ ku»e Hl^m) course, ttdl (tal) tell, Hogg's New Series of poems, including 

tpuU (spel) spell, betoiwul (biu'ttytil) 'Macksy Lane' a ghost story in the 



beautiful, uhe (els) else, abul Mly Deyonshire Dialect, dedicated by per- 

«b'9l) able, uny (out) onl]r, ihur (dha) mission to H.I.B. Prince Louis Lucien 

thee, wuUing (wdl'tn) ynlling, huket Bonaparte, London, 1864,1 2mo, pp. 62, 

(b^^ks) books, adu (adyy*) adieu. Here Mr. fiaird uses an italic u for the (yy, 

we have dude (dpd) precisely as in the 90) sound, resenring roman u for the 



xni th century, in Robert or Gloucester others, and similarly uses a for (a), and 
etc, but iuUy epuU (tal, spal) seem to the whole orthography is much im- 
indicate an ancient (ttil, spul) ; yet this proyed. 



302 EXT, BW XIV TH CENTUBT. Ohap. IV. { «. 

A careful examiiiation of this list would shew that if attention is 
confined only to the words for which we have xnth century autho- 
rity, the old classes would remain undisturbed, because no (y) word 
rhymes with an (eu) word or conversely. But if we remark that 
hw rhymes with true, knew, and also rue, and that rue, which rhymes 
with hue, also rhymes with true and with shrew, we are led to con- 
clude that true and shrew would have rhymed in the xivth, as they 
do in the xixth century. But this breaks up the old classification 
altogether. On examining the etymological relations, it will be seen 
that the old classification is at variance with them, but taking them 
as a basis we can divide the words into two classes, French and 
Anglosaxon, — ^including in the latter, words certainly Germanic, 
though not accurately traced, — as foUows : 
French — blue, due, eschew, glue, mew, remew, stew, sue, 
Ajigloeaxoji--drunkelew, few, hew to hack, hew servant, hue, knew, 
new, rew row, rue, shno, shrew, threw, true. 

The following table then shews that words of the first class 
rhyme together, but no word of the first class rhymes with any 
word of the second class. The first class corresponds to a French u, 
the second to an Anglosaxon iw, sow. Taking into consideration the 
Latin rhyme : de coitu, eschieu 9685, as well as the derivation of 
these words, there can be littie doubt that in Chaucer's time the 
first class had (y) and the second (eu). This distinction, then so 
carefully kept, was not understood in the xvith century in which 
several of the (eu) words, as knew, new, true, had fallen into the (y) 
class. At present all the (y) class, and most of the (eu) class have 
formed an (iu) class,' except when, through the infiuence of a pre- 
ceding (r), the modem English organs naturally change (iu) into 
(uu), but some of the (eu) class have become (oo) as shew, now 
more firequentiy written show. In such a word as Theseus 862, 
there is no diphthong, and we have to read (Thee*se,us). 

In the XIV th century then it will be safest to call ETJ, 
EW, (yy), in words of French origin, and (eu) in all 
other words. 

Alphabetical List of EW Ehticbs, etc. 

««w/y(beu*ti)G.,fr.beaut6,Pr.bewte, due (dyy) Sm. G., fr. d<l, Pr. duly 

beawtve decor, bewte 2387 tUhtU, dae eeohiewe 9826, aschewe 

biue (biyy; Sm. ags. bleoh, bleow, dewe 3046 

bleo, blio, Pr. bU)o lividut ; blewe esehewy fr. escbiver, escherer, eschvir, 

mewe (for bawks) 10967 oi^mYer, Pr. acbwyn vito; eschieu 

eoUuy Lt. de ooito, eschieu 9686. As coitu 9686, eechiewe due 9326, 

the practical identity of the spelling eechewe dewe 3046, eechiewed 

M with t has almdy been estab- aewed=fottowed 16823 

liflh^ no weight can be laid on few (feu) P. Sm. 0., ags. feawa; Pr. 

the rariant ieu as distinct from eu. fewe paueut ; fewe schewe 7431, 

drmkeUw, Pr. dmnkelew (see Mr. 12646, 13768, fewe schrewe 14234 

Albert Way's note there) ebrio9u$, glue (glyy) P., fr. glu birdiime, gluyer 

dronkelew© schrewe 7627, 9407, ttiek together, Pr. glwyn vieeo, 

13910 iwglewed remewed 10496 

^ For the Scotch sounds, see p. 298, note 1, at the end. 



Chap. IV. § 2. 



0X7, OW — XIV TH CENTURY. 



303 



Mho (uea) Bull., ags. heawan, heawian, 
ft. hewyn mw, hakke and hewe, 
lay hem on a Tewe=rotr, 2867 

hsw a hind, domestic servant, ags. 
hiwa; hewe untrewe 9659. 

hue, ags. hiw, hiw, heow ; hiewe trewe 
13836, hewe trewe 10901, 17207, 
hewe newe 1039, 10963, 11327, 
bewerewe 3= AiN^i eomptttaion 12666 

knew (knyy) Bnl, ags. cneow perf. from 
endican ; knewe newe 14996, 
knewerewe=nrpen/, 3061 

wmo, for hawks, (myy) P. Sm, fr. mne 
place for putimg poultry to fatten ; 
P. mne ror haukes meve ; Pr. mi 
of hawkys, faleonarium, mwe or 
oowle, mv, eaginarium; mewe 
(for pool try) stewe 361, mewe 
(for hawks) blewe 10967 

new (nyy) Sm. G., ags. neowe, niwe, 
nywe ; Pr. nwe, ney, novue ; newe 
hewe, 1039, 10963, 11327, newe 
trewe 14344, 16636, newe nntrewe 
787, 12970, 16614, newe knewe 
14996, newe threw (error for 
tKrewi) 14983 

rtmew, fr. remner ; Pr. remown or re- 
mevyM, amoveo; remewed i-glewed 
10496. 

row, ags. rawa, Pr. rowe eeriea ; lay 
hem on a rewesroM^, hakke and 
hewe 2867 

rue, pain, repentance, repent ; ags. 
hreowe, hreowan; Pr. mwjnpoe^ 
niteo eotnpatwr ; rewe = pain 
schrewe 6087, rewe = have eom- 
paeeion trewe 1866, rewe^repent 
trewe 3629, rewe = have eompae* 
eion hewe = Aim 12666, rewe=r^ 
pent knewe 3081 

rule, fr. rinle monaetw rule, Pr. rewle 
of teohynge, regula, norma ; reule 
173, reuled 1674 

ruth, see rue, quasi hreow}>e Pr. mthe 
eompaeeio ; renthe = compaeeion 



6074, reuihe^eompaeeion trenthe 
14608, ronthe = eompaeeiony 
trowthe slouthe^f^M 4949 

ehew (shen) Sm. Q. Bull, ags. scawian 
sceawian; Pr. schewe or schew- 
ynge monttraeio ; schewe schrewe 
6866, 12844, schewe fewe 7431, 
12646, 13768 

ekrew (shren) P., etymology unknown, 
see Wedgewood 3. 176. Pr. 
schrewe pravue, schrewyd prO' 
value, schrewyd hertyd pravicore, 
schrewdenesse pravitae, schrewe 
rewe t=pat>t 6087; schrewe shewe 
6866, 12844, schrewe dronkelewe 
7627, 9407, 13910, schrewe fewe 
14234 

etew, fr. estuve, Pr. stuwyn mete, stuyn, 
etupho; stuwyn menn or bathyn, 
stuyn in a stw, balneo; stwe fysche 
pond, stewe, vivarium; stwe bathe, 
etupha ; stewe =.;Ei^ pond mewe 
(for poultry) 361, styTes=^^AW;t 
Ivres 6914 

eue, u, suir, sivire, siyre, sewir ; Pr. 
svyii or pursryM pereeguor, suwynge 
eequela, sringe euceeeeue; sewed 
escbiewed 16823 

eurety (syyr) Sa. Bull., fr. seur ; seurte 
1606, sewerte 6486 

threw ags. )>reow ; threw (error for 
12970, threwe) newe 14983 

true (tryy) P. Sa. Bull. G, ags. treowe, 
trywe ; Pr. trwe verue, truwe 
mann verax, trewe hewe = hue 
10901, 17207, trewe hiewe =Ati0 
13836, trewe rewe 1866, 3629, 
trewe newe 14844, 16636. 

truth, ags. treowS, Pr. trowthe veritae, 
treuth reuth 14608, trowthe routhe 
8loutheB«^M4949 

untrue, see trtie, untrewe hewe^eer-- 
vant 9669, untrewe newe 737, 
16614 

value, fr. Talue ; valieu 14682 



OXJ, OW — XIV TH Century. 

Ab we have already bad occasion to remark (p. 236), when the 
letter «, which is the natural representative of the (uu) sound in 
all languages that have adopted the Boman alphabet, has come to 
lose its proper sound, as in French, Dutch, Swedish, English, but 
that sound remains in the language, it becomes necessary to adopt 
some other notation for (uu). The (uu) sound in these cases bias 
been generally a transformed (oo). Hence it lay ready at hand to 
use simply for this sound, as we have seen was occasionally done 
in Chaucer (p. 267), and is still done in move, etc., and as the Swedes 
have been content to do. The Dutch employ oe for (uu), as they 



304 OU, OW — XIV TH CENTUBY. Chap. IV. { 2. 

1186 00 and o for (oo), but, as appears from the history of this ortiio- 
graphy (p. 236, note 3), ne was in &ct long o used as (uu), precisely 
as in the last case. The French used oti, in the earli^ existing 
documents/ though the Normans used u for both (yy) and (uu) ap- 
parently, as may be seen in the French original of Henry III. 'ids 
English proclamation, Chap. Y, § 3, No. 1. On an examination of 
the documents of the zm th century it will be found that the use of 
u for f , tf, representing the y, that is (y), of the Anglosaxon, greatly 
increased towards the end of the period, so that confusions between 
the values of ti as (uu, yy) became aimo3ring. Writers then appear 
to have introduced the spelling tm towards the close of that period, 
in conjunction with ti, to represent (uu), but, the oonyenience being 
manifest, ou became general by the early part of the ziYth century. 
These facts will be established in the next cluster, and are here only 
stated by way of anticipation. There was one disadvantage in the 
use of OU, namely that it had also to be employed for (oou), but this 
occasions very slight inconvenience. In llie present place we have 
only to establish that <m really represented (uu) generally, and con- 
sequently (u) occasionally, in Chaucer. 

As the use of u for short (u, u) was already well fixed, and its use 
for f , $ was rapidly going out, <m was of course not so frequently 
employed for short (u) as for long (uu). Examples however occur, 
thus: o\u 5729 stands for tM, (mUrly 6245 for utterly, and the 
orthographies Arrious 6344 for Arrius, Caukasous 6722 for Caueasui, 
leave no doubt of the use of on as short (u). Curiously enough the 
sound of (uu) fell into (ou) about the xvith century (p. 150), and au 
served then to represent that sound without change of spelling. But 
after this it became important to distinguish the (uu) and (oo) sounds 
of long 0, and the orthography oo, adopted for the former (p. 96), 
has remained in use to the present day. In the unaccented syllables 
-our, representing -(uur), the orthography was left unchanged as 
well as the pronunciation. In the xvn th century these syllables fell 
into (-or), and either the o or tf in -our was felt to be superfluous. 
In quite recent times factions have been formed, one requiring -or to 
be used universally, others maintaining that -our should be preserved 
to distinguish the words that come from the French, which now ex- 
hibits -eur, corresponding to a later development of that language. 
In Chaucer's time however -our was used, simply because the pro- 
nunciation was (-uur), as -oun was used for the present common 
termination -on, compare corrupcioun 13950, confessioun 1735, 
regioun 2083, visioun 7259, leoun 6377, etc., which were pro- 
nounced (un) or (uun) even in the xvith century (p. 99). We 
have retained -ous unaltered, and this was also (-ms) in the xvi th 
century (p. 150). 

^ Diegf Gram. d. Rom. Spr. 1, 429, vowel, as naybbous s nopihtu, obsenr- 

2nd ed , where he quotes Benarr Bom. ing that Mommsen (XJnterit Dialects, 

Lantlehre, 82, to shew that the Old 217) and Ritschl (De milliario PopU- 

Romans occasionally used on as a mere lano, p. 34) are of a different opimon, 

ortho^;raphical sign for u, and remarks and consider that in really old inscrip- 

that It was even employed for a short tions ou « ovy and not u. 



Chap. IV. § 2. OU, OW — XIV TH CENTURY. 305 

Ab Palsgrave (p. 149), and Bnllokar (p. 152), in the xvith oentory 
recognized this (un) sound of ou, it will only be neoessary to intro- 
duce a few examples. 

Bhthes with Latdt Names: — ^Theseus, desirous 1675, curious, 
Darius 6079, Venus, contrarious 6279, Apius, leccherous 13680, 
Claudius, coitageous 15821, yicious, Swethoneus =3 Suetonius 15949, 
Antiochius, venemous 16061. 

Khthxs with Pbsnch Wobds : — 

What will ye dine ? I will go there abeute. 

Now, dame, quod lie, jeo voua dy saunz deuU. 7419 

Full many majde bright in hour 

They monine lor him, par amour, I51ff3 

Compare — 

And bat thou do my norice hononre 

And to my chamberer withinne my heure, 5882 

Natubal Sotjwi). — The cry of the euohoo was certainly intended 
to be (kuk'kuu*), and this determines oto in 

This crowe song, Cncko^, cuckow, euekow ! 

What brid, qu^ Phebns, what song syngistow now f 17175 

Perfectly Saxon words as hour, now, ahoute, having thus the 
sound of (uu) established, we may feel sure of it in other cases, as : 
hous Caukasous 6721, thus vicious 7629, dowte aboute 489, tour 
honour 2029, Arthour honour 6440, dortour hour 7437, powre 
laboure 185, flour odour 2939, hour schour 3519, emperour 
honour flour 5507, in an hour (error for hours), to honoure 14954, 
houres schoures 3195, 10431, and hence schowres 1 = (shuur'es); 
yow how 7982, youthe nouthe 463, to give the child to souke, all 
in the crouke 4155, colours (error for eohures) floures 10824, licour 
flour 3, adoun broun 394, licorous mous 3345, pitous mous 143, 
houndes stoundes 5867, stounde founde 5441, vertuous hous 251, 
for to touche, in his couche 5669, untrouthe routhe 5107. Whence 
also we conclude that: cowde 110, flowtynge 91, drowpud 107, 
embrowdid 88, so woweth hire 3372, thay blew and powped, thay 
schryked and thay howped 16885, facound 13465, and numerous 
other words in <w, have also (uu) or (u). 

As examples of those cases in which ou, ow, had the sound (oou) 
maintained in the zn th century as (oou) practically, but (00) theo- 
retically, we may take : anoon the souleSf with fleischhok or with 
oulee = awh, ags. sawl, awul 731 1, Bowe, unknowe 125, lowe 
knowe 2301, I trowe, undurgrowe 155. 

In the provinces two sounds of ou, ow are also common. One of 
these is (uu) in almost all districts, but the others varies as (aa, aa, 
an, iau, ou, iou), and even (su, ou), and there is great difficulty in 
obtaining a satisfactory account of what the sounds really are, and 
consequently in classifying them. The following lists referring to 
the dialect of South Shields,^ will serve as a specimen. For the 

^ Obligingly communicated by the Ber. C. Y. Potts, of Ledbnry. 

20 



306 OV, OW — XIV TH CENTURY. Chaf. IV. § 2. 

present pnrpose the most important point to dwell on is the per- 
& .;ence of the (un) sotind. 

aw = (uu) in : down, town, crown, tower, now, trowsers, how, 
flower, power, drowned, cow, sow, bow «. & v. fleeterej bow 
areut =» (ban). 

ou » (nu) in : plough, round, sound, mound, hound, doubt, thou, 
about, count, out, house, sour, flour; — found, bound, ground, 
these three words are also pronounced with (o), but this is for 
the dialect even, very vulgar ; — our, which is vulgarly (wor). 

ou = (au) in : brou^t, soug^^ fought, bought, thought, ought s, 
& v.f nought, soul, four, loup «. ft r. = leap, coup := exeha/nge, 

aw = (aa) in: blow, snow, low adj\, row #., crow, slow, below, 
know, callow, arrow, barrow; — owe, own, another and less 
vulgar pronunciation of these words would be (au, aun), and 
in tibese words generally (au) not (oo) would be 1^ alternative 
pronunciation. 

a 1= (au) in : old, cold, also (aad, kaad) ; — sold, told, also (sold, 
teld) ; — old, bold, fold ;— stroll toll, roll ; — over (au'er). 

(au) is heard in : daughter, neither, either, loose, sew, chew, mew, 
row r. ft «., low = flame, bow arcus. 

Mr. Murray has been kind enough to furnish the following in- 
teresting account of the Scotch usages : 

'' In all the Scottish dialects the Anglosaxon long u, and French 
au, retain their old sound (uu, u) before a consonant as: hour 
(buur) bower, clour a eweUing cawed hy a blow, dour, stubborn, 
flower (fluur), hour (uur), power (puur), tour (ets juur tuur ta 
pW) its your turn to play, tower, sour, stour^ loose dust, shower, 
scour, devour (di-vuur), our (uur), your, pour (puur), cower 
(kuur), spout (spuut), shout, lout (luut) A.S. litian, to stoop, rouse, 
bouse (ruuz, buuz). 

'' In the following the vowel is shortened in quantity but un- 
changed in quality: brown (brun), crown, doun (dun), drown 
Bi), gown, loun, town (tim), bowl Fr. boule (bul), foul, fowl 
, swim (sum), sum (sum), howl, yowl, scowl, owl, howlet Fr. 
tie (Hul'ot), mouldy, course, court (kurs, kurt), source, douce, 
croose (krus) sprighthy, house, mouse, louse, mouth (muth), drouth 
drought, south, Soutra,* souter, snout, out, about, (ut, obut'), doubt, 
clout, bout (« dr^qk'in but) a drinking bout, stout, scout, pouch, 
vouch, crouch, often (kruutsh), couch, bulk (buk), duck verb— 

^ The first stanza of Bums's addrefls well illustrates these (nn) sounds. The 

''to a Mountain Daisy, on turning one nrononciation is that heard by Mr. 

down with the plough, in April, 1786,*' Morray from a townsman of the poet 

Wee, modest, crimson-tipped flower, (Wii, mod'ost, kremzn-tep'tt flanr, 

Thon's met me in an evil nour ; Dhan -z mst me en m iivl nnr / 

For I maun crush amang the stonre For aai man. krssh amoq* dhs stuor 

Thy slender stem ; Dhai sl£na*dr stsm ; 

To spare thee now is past my pow*r, Te sp«rr dhi nun ez post mo puur, 

Thou bonnie gem, Dhuu Don't dzuEm.) 

9 The YaStj ridge which separates the Lothians from the south country. 



I 



Chap. IV. { 2. OU, OW — XIVTH GRNTUBT. 307 

the noun is (dyk, doek), — dronk to drmeh^ jonk to elude, louk, p'xuk 
to picky pilfer , ploock to plucky sack, touc o'drom, stonk a ehobd of 
oom. 

'' The combination -ouHd is, like -ind, in a transition state ; the 
past participles: bound, found, ground, wound, are usually (bsm, 
&nd, gran, wsn), and ground «. (grand), but I consider this to be 
recent, for I have heard (u) in some of tiiese from old people, and 
we always hear it in : Where are ye (btm) or (bund) for, to beat 
the (bunds), boondit, boondarie, boim'tree: and the sound is 
always used in round Trund), sound, to found, founded, foundation, 
stound a fit or * epelr as (« stund o dha toeth'ak) »a fU of tk§ 
toothaehe. Mound is occasionally (Hand), usually (nund). 

" Ang^osaxon u final is also (uu) in most of the Scottish dialects, 
but in that of the Southern counties, the same law which has de- 
veloped long i into (ei), here developB (uu) into (am). The follow- 
ing words therefore pronounced in ^e other dialects with (uu) are 
pronounced in Teviotdale and Dumfriesshire with (au) : cow, sow, 
how, you, now, bow to bend, through, doo dove,^ loe to love, brow, 
^' fuU, tipsy, go(it, on after taste (guu), Tev. (gau), as (it hass « 
kwiir gau abut* tt) =^itima queer flavour about it, pu' pull, (suprii 
p. 287,) mou' mouth. 

" The Borderers thus pronouncing (eu) where the other Scots say 

(uu), — ^where the others say .(au) they advance a step and say Ton), 

so that the following words are in the Lothians pronounced (au), in 

Teviotdale (ou), in English (oo) or (oon) : bow arcue, grow, dow 

to avail, howe a hollow, knowe a knoU, bowe a boU,^ lowe ' a flame, 

powe a poU, rowe roll, row, stow, tow, trow, thowe to thaw, drow 

a Scotch mist, a dritsUe, bowl, soul, four, glower to stare, ower over. 

" The two pronunciations may be shewn thus : 

Central Scotch : (£nir baulz fuu e njun malk fe dha kuu) 

Teviotdale : (four boulz £iu 9 niu melk thrsB* dhe kau) 

English : four bowls fall of new milk from the oow.^' 

The conclusion seems therefore to be that OU, OW in 
the xiv th century should. be read as (uu, u) except in 
those cases where aw, or simple o was used in Angk>- 
saxon. 

1 A school inspector wishing to ^ * Compare — 

the Bound of (nn) out of a Hawick girl, (pharz let*l wxt en dhe poit 

and unaware of this peculiarity of pro- Dhat leArhts dhe kon*! at dhe Ion) 

nnndation, asked her what she called a = There's Httle wit in the poll or head, 

pigeon, (A dan) replied she, and posed That lights the candle at the low or 

nim as much as the child posea the flame; 

teacher, who, wanting to obtain from and the pnn on the names of Mesm. 

him the word take^ asked him : *' What Lowe and Bright at the Edinburgh 

would YOU do, if I gave you a piece of Beform Demonstration : ** The Lowe 

cake? andreceiTM the Tery natural that'll never bum Bright" (Dhe Ion 

reply : ** Eat it." dhef 1 never bam breAht). 

* (Compare Sir T. Smith's /3«v, fimX, * So likewise in the Bamde^ dialed 

saprk p. 151. ihroo is used for from. 



308 B, C, CH, D, F. O, GN — XIVTH CBNTUKT. Chap. IV. § 8. 

% 3. ne Cimsonants. 

Yery little is to be learned from the rhymes respecting the 
consonants. With our knowledge of the xvi th century con- 
sonante^ however, there can be but little doubt as to the 
values of any one of them. 

B, 0, CH, D, R 

B when ealesxt as in daubt^ debt^ was not written thus : dowU 489, 
dette 282. It was otherwise (b) of course^ 

C was (s) or (k), according to the same rales as at present, but 
ci- remained (si-) and had not become (sh). In the termination 
'tian^ we find e^ «, t interchanging, shewing the identity of sound, 
but it always formed two syllables. Compare 

liO, heer bath kynd his dominadoMM, 

And appetit flemeth ^^actetiotm, 17114 

wantrust) fdl .of fids suspeo^ioim 

Where was thy wit and thy discreabtMu 17214 

And eke he was of sach discresMotifi. 16795 

CH was generally (tsh), see J, K. 

D was (d) of course. 

F seems to have been always (f), so that of must be called (of) 
not (ov). Judging from other writing, as Robert of Gloucester and 
Trevisa, fi or i^ would have been used had (v) been pronounced. 
Mr. Murray says that cf\a still pronounced (of) in the North, when 
the consonant is retained before a vowel, as (dha md of « bist) th$ 
head of a beast, 

G, GN. 

G followed the same rule as at present, and was (g) in an 8axon 
words, but in French words (g) before a, o, u, and (dzh) before 
(e, i). See J. 

GN occasionally represented simple it, as in the couplet 

Sche may unto a Imare child aiteigne 

By Uklihed, sith sche nys not bareigne, 8823 

where an represents an old French gn, in baraigne, which was pro- 
bably (nj) as now, so that (atain* barain*) would be the natural 
EngUsh representatives. Accordingly the MS. Univ. Cam. Dd. 4. 
24, here writes atteyne^ bareyne ; a spelling found also in Harl. 
7334, in 

ThoQ maist to tiiy desir aomtyme atteifne 

But I that am eaoled, and bwreyne 

Of alle grace. 1245 

while gn and n Tkjme in 

And of his oughne rerta nnconstrei^ed 

Sche hath M ofte tyme hire seek y-feyned, 13476 

where we should have expected gn in the second line as much as 
in the first. Companye 24, was also commonly written for : com- 
paignye 3837. 



Ohap. IV. § 8. G, GN — XIV TH CENTURT. 309 

How were diffne^ henigne 519, pronounced? Ab Airglo-French 
rdim'e, benwii'e) ?* Or after the custom of Latin pronimciation 
(maq-nns, iq'nis) in the middle ages — testified by the medieval 
Latin orthography, and still existing in Salesbury's time, — as 
(diq*ne, beniq'ne) ? The question affects also such words as dianiU^ 
signifies sign. Here the modem use condign dignity, henian oenig- 
nifg, sign signify (kmdain dtg-nrU*, bindin* bintg-n»tf, soin sig*n»- 
fai) would seem to lead to an anterior (dun d4g-ntte, bernVn* be- 
mg'ntte, sim sfgn«f»V'e). But the old example of i-^eined for signed 
in Henry Ill.'rds English proclamation, throws a doubt over this. 
As however the special woid sign, had assumed a th<»t>ughly Saxon 
form, segnian to sign or bless, segnung a signing with the cross or 
blessing, the (ai) sound would be developed naturally by the 
passage of the guttural g into (j). 

Can we consider the fonns: deynous 3939, 6*114, deyne 3961, 
6*204, deyneth 5*288 as conclusive. The French digne, daigner, 
shew a double form in these words, and hence leave us still 
in doubt. The word: dyne 4*200, 4*201, = rftVw, was in French 
disgner, dispner, and is considered by Boquefort to be derived from 
the commencement of the grace dignare^ dominsy but the etymology 
is so doubtful' that no weight can be attached to this. The termina- 
tion 'ign6 is not found rhyming either with -syne or -yiw, and this 
would k priori lead us to conclude that the sound was different 
from either, that is, neither (-ain-e) nor (-tVnie). But we find : digne 
benigne resigne 4*125, 4*225, sygne benygne 5*183, digne signe 
5*330, so that the old and proved (sain) and the occasional (dion) 
would seem to imply also (benain*, resain*)^ On the other hand 
Gill writes (benig*n; or (bemq*n) for henign^ and this ought to im- 
ply that he did not know the pronunciation (benain*), which may 
nevertheless have existed, and been ignored. «rones, however, 1701, 
gives only (binig'on), though he admits (sain, rezain*),. and Sales- 
bury and Smith give (sein), Gill (sain), Buchanan and Sheridan in 
the xvnith century give ^binain* biiuAin*). Sinular difficulties 
have existed in the pronunciations of impugn^ impregn. 

If the sound (ain) had prevailed in Chaucer's time, we should 
have expected (ain), not (ein) in the xvith century. Bullokar 
seems to write (stin), and the (sein) of the xvi th and (sain) of the 
XIX th century are in harmony with this, which would imply (stVir) 
in Chaucer aliso. In this doubt the safest plan seems to be to adopt 
(fin) for Chaucer's pronunciation, admitting the secondary form 
(ain) when eyn is written. TMs will be eonaistent with the present 
and intermediate pronunciation, with the general use of « in Chaucer, 

^ Diez (Ch*.- de E.S. i, 439 note, 2nd oonque n seqmtnr t in media diccione, 

ed^ 8ayithat<%9i««ccan in old French in mTersis siUabis g debet interponi, 

with silent t, as bjyana dign$9 rhymed nt eertaignttnmi^ bmignement; sed g 

in^brigandinesdimg'Ducaa^BaiYoce non debet sonari." 
hriffa. And the MS. 188 of Mag. CoU. 'Among the etymons giren are 

Oi^ord, cited by M. G^nin (Introduce itnrrHy, decoenare, decima (hora), 

tion to the French reprint of Palsgrave, sdigiimare, d^jeftner sdisjejiBiflre. See 

p. 29) sayB, nde 92 : item, qiumdo- Donkin's Diez, sub dssmare. 



310 



OH, T, Z — XrV TH CBNTUBT. Chap. IV. { 8. 



ttod with hiB use of -^m in other words, and as regards the word 
^ign would imply that he took it from the French with the other 
words, or designedly adopted a French in preference to the an- 
tiquated pronunciation (sain). The question is one of extreme dif- 
ficulty and the conclusion is doobtfuL 

GH, T, Z 

The modem editors usually represent s or rather ^^ \fj gh when 
medial and final, and by ^ or 5f when initiaL In Mr. Morris's 
Chaucer Extracts he purposed to shew where the manuscript ex- 
hibited f for his printed ghy y, by italicising these letters. He has 
not carried out his plan with sufficient accuracy to make an examin- 
ation of the MS. unnecessary.* Assuming, however, that where he 
has used the italics, } was employed in the MS., we obtain the fol^ 
lowing results for the Prologue, Knightes Tale, and Nonne Frestes 
Tale, in which I have here used a common s in place of ^ or |. 
The numbers annexed to the words indicate the obsenred number 
of occurrences of this orthography. 

asens 

brouzt 

deyzen 

douzter 

drauzt 

eyzen 

fiztyug 

forzete 

forzeye 

heyz 

heizer 

knizt 

nozt 

nouzt 

perfizt 

lizt 

But the orthography is not consistent, for gh is often employed in 

the MS. Thus, acc^ting Mr. Morrises edition as correct, except in 

the words yof«, etc, we ^d in the Frdogue only 

brought 1 canghte 1 foughte 1 herbergh 2 

bythoughtl draught 1 foughten 1 heye 1 

caught 1 drought 2 heih 1 heyg^ 



1 


sauz 




zelwe 


1 


zolden 1 


1 


thouzte 




zemen 


3 


zoUyng 1 


1 


unzolden 




zerd 


6 


zolo 1 


3 


upzaf 




zerde 


1 


zolow 1 


1 


weyzede 




zeros 


I 


zolw 1 


9 


wizt 




zet 


8 


zolwe 3 


1 


ynowz 




zette 


1 




1 


yze 




zero 


5 


zonder 1 


2 


zaf 




zeven 


3 


zong 3 


1 


zalwe 




zevest 


1 


zonge 6 


1 


zate 




zeveth 


1 


zore 2 


2 


zeddynges 1 


zjf 


3 


zou 2 


8 


zeeldyng 


1 


ziftes 


2 


zoung 1 


9 


zeer 


14 


zit 


18 


zouthe 5 


4 


zeldehalle I 


zive 


3 


zou 1 


1 


zelleden 


1 


ziyen 


1 





1 



^ Thii cbaraeter in the MSS. jm 
generally indistrngaiahable from i, bo 
mat when an editor prints tome words 
wilh ) and others with s he is making 
«n arbitnffj distinction like that m 
sraanting«,9. In Mr. Morris's edition 
of ^ Gmomm for the Earif EngUth 
Tttet SodHfy ) is printed for both % 
and I. It would hare been more eon- 
sistent with the anployment of Boman 



trpes to nse s instead of ) in botii eases. 
This is the plan I haTe panned in the 
following listB, and it ii one followed 
hy older printers and embalmed in the 
Scotch Mensies, Dalid, Mackoudt, 
which are often called (Meq*tK, Btil, 
DrsL, Maken'Jt) in Scotland, aee p. 
29S,n. 

> Thns in t. 84 and 88 he prints 
'yow' in place of 'jrow' that is *}oy.' 



Chap. IV. § I 


1. 


GH, T, Z — XIVTH CENTURY. 




81 


high 

highte 

inough 

knight 

might 

mighte 

mighten 


1 
1 
1 
2 
4 
1 
1 


neigh 2 oug^te 
neighe 1 raught 
night 1 right 
nightertale 1 seigh 
nightyngale 1 deight 
nought 1 streight 


1 
1 
4 
1 
1 
1 


taughte 

though 

thought 

wight 

wnght 

WTOug^te 


1 
2 
2 

1 
1 
1 



It may be doubtful whether y is ever used initially , in the modem 
sense. I have not observed any instance in the MS., but I have not 
examined it thoroughly with this view. The use of y was quite 
established however before the time of printing. 

The reader is requested to refer to the remarks on y^ in Chap. m. 
(pp. 209-214). As ffh still retained its ^ttural sounds in the xnth 
century,^ we cannot but beliere that it had these sounds in the 
xiyth, whatever may have been the Anglosaxon original sounds. 
The divarications of (kh) into (kjh, kirh) pointed out in the remarks 
referred to, so that it sank to (j, i) on the one hand, and (wh, u) 
on the other, are well shewn. Thus, to the first class belong thei^h 
= (dhaij^h) for though. 

For theiffh thou night and daj take of hem heede. 10926 
which becomes simply th&y (dhai) in 

That Chancer, theff he can but lewedly 
On metres and on rhyming craftdv. 4467 

and similarly sei^h 9605, setf 18807 lor saw. 



^ The Boond is hardly lost yet in the 
prorinces, thns Prof. Sedgwick in the 
work dted above, p. 289, note 4, says : 
^ The ti^nfresiion of the guttural mmndt 
is, I think, the greatest of all the mo- 
dem ehanges in the spoken language oi 
the northern counties. Erery sylhible 
which has a Towel or diphthong fol- 

egh was once the symlxu of a 
sound: and I remember the 
I all the old men in tiie Dales 
sonnded such words as tighy nighty 
aghif {sikhf mkhif sij^ht), &c., with a 
gentle gnttnral breathing, and many 
other words, snch as trouehy rough, 
tough (trookh, ronktoh, tnnktdO, had 
their utterance, each in a grand sono- 
rous gutturaL The former of these 
guttural sounds seemed partly to come 
from the palate ; the latter from the 
chest. Both were aspirated and articu- 
late; and diflfered entirely from the 
natural and simple rocal sounds of the 
guttural Towels d, 6 (aa, aa). AU the 
old people who remember the con- 
tested elections of Westmoreland, must 
have [p. 1041 heard in the Dales of 
that county ine deep guttural thunder 
in which the name — Marry Brougham 
(Bniktt^'Bm) — was rererbOTated among 



the mountains. But we no longer 
hear the first syllable of Brougham 
sounded from the caverns of the chest, 
— ^thereby at once reminding us of 
our grand northern ancestrr, and of 
an ancient fortress of which Brough 
(Bmkieh) was the written symboL 
The sound first fell down to Bruff ham 
(Bnif'tnn, Broftnn), but was too vigor- 
ous fbr the nerves of modem ears ; and 
then fell lower still into the mono- 
syllabic broom (Bruum, p. 153) — an 
implement of s^rile use. We maj 
pofish and soften our language by this 
smoothinff process ; yet in so doine we 
we are forgetting the tongue of our 
fathers ; anc^ like degenerate children, 
we are cutting ourselves off from true 
sympathy with our ^reat northern pro- 
gemtors, and depriving our spoken 
language of a goodly part of its variety 
of form and grandeur of expression." — 

fel0a-4, pfQaeotvpe introduced. Mr. 
urray notes that the Southern (a) is 
always (if) in Cumberland and West- 
moreland, and that (ru( Ud, Bnif^) are 
the present pronunciations of roughs 
ioughy Brough, in those counties, and 
(Btuum) for Brougham in Cumberland. 



312 GH, Y, Z — XIV TH CBHTURY. Chap. IV. § 8. 

To the second class belong Imoghe 476, hwh 3117 = laugh, Mugh 
5268, 9726, sawh 5265 = saw.^ Compare also herbergh 767, her- 
berwh 4117, herberw 4143. Sometunes the transition is complete 
aain 

For, ai I trowe, I hare yow told piow$ 

To leyee a feend, al loke he nerer to rowe. 12788 

where y-notM, row$ (inuu*, ran) stand for enough^ rough, in which 
the modem sound of (f), as already suggested in p. 213, has arisen 
from (wh). So frequent was this change in the word eiumgh^ that 
it is sometimes neglected in writing as 

For had we him. than were we ayker y-noiyA, 

Bat onto God of hoTen I make avcw. 12792 

only a couplet beyond the last example quoted, where we must read 
(inuu*, ayuu*). Similarly gnough, now 12946, where gnow should 
be read as in gou, g-now 11019. Plough which rhymes with inough 
889, 3159, had generally the pronunciation (pluukh), and this re- 
duced to (pluu), (shewn in the spelling plow, which I have noticed 
elsewhere, but not in HarL 7334, an ortiiography found also in the 
authorized version of the Bible in the ivn th century,)' generated 
the modem (plou).* The following rhymes may also be noted : 

When that he sangh that al the peple iou^h. 

No more of thia, for it is right y-nough, 14876 

He alio hath to do more tlum y^nough 

To kepe him & his ci^il out of the $lough, 16995 

Compare 

Now is my cart out of the iko parde. 7147 

In which ther ran a swymbnl in a iwough 

As it were a storme schiild berst erery bough, 1981 

He siketh with fdl many a sory tuxmgh 

And goth, and geteth hun a kneedyng trough, 8619 

The regular pronunciation of all these ough words seems to have 
been (uukirh), whence f uuwh, uu), which afterwards changed to 
(uf, ou), and finally to (of, ou). That gh was occasionally written 
wilJiout being pronounced, we see by tiie rhymes: at his retenue. 
Sir Hughe 6937, melodic yhe 9, etc. We shall see that this is the 
case also in Shakspere, whenever it was convenient for the rhyme. 

The form augh may have had similar varieties of sound, as the 
spellings already cited indicate. In both cases we cannot do better 
i^an follow the spelling of the moment, except the rhyme requires 



^ There is a similar resolution of Pror. 21, 4, Lnke 17, 7; 

medial a in Icelandic. Thns IHiga to Isa. 28, 24, Amos 9, 13 ; piowmm Isa. 

tell a falsehood, is theoretically (baa*- 61, 5, Jer. 14, 4 ; plowharet Isa. 2, 4, 

fftrha), and practically (bnu'wa). See Joel 8, 10. Supra p. 169, note 4. 
Chap^V. § 4, No. 2. » Mr. Murray obserres : '^ynough 

* The passages are : plough Ps. 87, and ynow (anikirii*) and (dnra*) or rather 

1 2 ; plow Dent. 22, 10, i Sam 14, 14, (envktrh*, enyu*) are both used in Scotch 

Job 4, 8, ProT. 20, 4, Isa. 28, 24, with a difference of application. FUmgh 

Hos. 10, 11, Amos 6, 12, i Cor. 9, 10 ; and flow are synonymous for the noun 

plowed Judg. 14, 18, Ps. 129, 8, Jer. (plyku'h, pl^;, the former the more 

26, 18, Hos. 10, 13, Micah 3, 12 ; common : for &e verb Ihe latter alone 

plowort Ps. 129, 8 ; plow$th i Cor. 9, is used as (« plyud fild, « plyain 

10 ; ploufitig i Kings 19, 19, Job 1, 14, motah.}*' 



Chap. IV. { 8. OH, T, Z — XIV TH CENTUKT, 313 

one of two forms to be altered, and then the jwit should generally 
be accomodated to the second, as there is a probability of its having 
been written down without consideration of what was to follow, 
and of its having been then left uncorrected, as being of sHght im- 
portance. Thus augh, auwh, auh, aw = (aukt^h, auwh, auH', au), 
where (aukh) may be used for (aukirh). 

When the letter i follows fresh difficulty arises. How should 
drought f fougJUen, daughter ^ nouht, be pronounced? There seems 
notlung but theory to guide us. At present we say (draut, drAAt, 
fiA*t'n, dAAij, UAAt), but these are aU quite recent developments. 
We find fought = (fauH^) in Smith, daughter =s (daukh'ter) in GiU, 
nought = (nouH*t, nauH*t) in Smith, and (nooukht) in Gill. There 
is no XVI th century authority for drought. Taking into considera- 
tion the double use of ou (uu, oou), it seems probable that when the 
original vowel was u in ags. as drugo^, the sound should be (uu) as 
(druukht, druuktrht) of which the modem (draut) would be a legi- 
timate descendant ; and that when the original vowel was o as ags. 
dohtor, the sound was (oou) or perhaps simply (ou), the (u) having 
been developed by a (ku^h) sound of ^A. This would give (druukht, 
fooukh-t'n,dooukh-ter, nooukht) or (druktt^ht, foukw7h*t'n,douku7h'ter, 
nouktrht). It will probably be as near the truth as we are able to 
get to write (drukht, foukh'ten, doukh-ter, noukht). The spelling 
nouht, however, indicates a very light sound of the guttural, as 
(nouH*t), which rapidly disappeared in (not, nat).* 

What the initial sound of 5 or j might have been, it is more 
difficult to say. Probably the sound of the ags. letter became 
{kh) or (yh) at an early period. Now in modem Germany (j^h) is 
often considered to be tiie hiss of (j), that is (ih), and the difference 
is certainly very slight. The ease virith which initial (Ah) will pass 
into (j) may be well studied in modem German pronunciation. 
During the xv th century when initial j was replaced by y, the 
transition was certainly complete. In the next chapter (§ 2) reasons 
will be given for thinking that this transition may have been pre- 
valent in the time of La^amon and Orrmin, the proceeding (i^h, yh) 
stage being relegated to the Old Anglosaxon period. It will there- 
fore be safest to pronounce the initial j as (j) where it corresponds 
to the modem y. 

We shall have an opportunity of seeing g in every stage of tran- 
sition, from (g) through (^, yh, j) to (i) on the one hand, and 
through (gt^h) to (w) on the o&er, and even absolutely disappear- 
ing through a scarcely pronounced (gh, gu^h), in the living Ice- 
landic tongue, the veiy interesting phonetic phenomena of which 
will be considered in Chap. Y. § 4, No. 2. 

1 Mr. Mmraysays thatinTeyiotdale In the other dialect they are (fokht, 

dnmghi U (dmth) doughtery foughtm, bokht, sokht, w'rokht), Aberdeen 

nmgkt, bought f brouahiy thoughty naughty (vrokht) with simple (0) and (kh). So 

wrought are (donktrhtar, fonku^ht'n, also with loeh, houg\ eough, trough, 

bouktrht Vronku'ht], &c., or perhaps &c. Tev. (lonktrh, loonkirn), Oentral 

(doonkumteTy foonktrht), he prefers the Scotch (loui, lookh). 
fozmer, though the is aheolutely long. 



314 H. J — XIVTH CBNTITBT. Chap. IV. } 8» 



H, by its snbstitation for gh^ is shewn to have been pronounced 
when final distinctly as (h*). In what cases, when initial, it became 
(h) or vanished, it is now impossible to say. It appears by many 
^dMSS. that there was often great confosion as to tiie use of initial 
A in many words, indicating local and partial peculiarities of pro- 
nunciation, similar to those now found. But the MS. under con- 
sideration seems to be q^uite consistent in the use of initial A,^ and 
there is therefore nothing to shew that it was not pronounced in 
honour J honesty hour, as well as other woids. However, in this 
doubt, I have thought it safest in my transcriptions, to follow the 
modem use. In the words he, his, him, hir$, hem, before which, 
especially when enclitic, the final is, as we shall see, generally 
elided as freely as before a vowel, it is extremely probable that Uie 
h was silent under the same circumstances. It is known to be con- 
stantly so in modem English, and some orthoepists even admit that 
it should be silent.* The apostrophe in eateh em indicates the ab- 
sent h, not an omitted ih. When hath, have, hadde, were similarly 
placed they also probably lost the A, as they also admitted the elision 
of the voweL lie modem contractions I ve, toe ^ve, they ^d, and the 
old nadde = ne hadde 3751, point to the same conclusion. Hence 
when those words beginning with A stand in such a position that a 
final e might be elided before them, I omit the A in my traoscriptions, 
but indicate the omission by a hyphen in the usual way, thus : (w^ 
kuud -e sit on Hors) 94. 

J 

J when representing the French consonant /, is now called (dzh) 
and was so in the xvi th century. Was the old French sound (dsh) 
or (zh) ? Diez (Gr. d. E. S. i. 400, 402) shews good reason to sup- 
pose that the Provenqal pronunciation of eh, j, was (tsh, dzh), as for 
example Petrarch's ciant for Provenqal ehant, and Dante's giaueen 
for Tr.jamen. Again (ib. p. 448, 451) Diez c^ews reason for sup- 
posing (tsh) to be an old French sound of eh, although in Falsgrave^s 
time it had sunk to (sh), and observes that in middle Greek, the 
French Jean, Geojroi, are rendered TJai/, T(f€^pi, which are the pre- 
sent combinations for (tshan, tshefree'V Considering that the Greek 
had no means of representing (dzh),* this would stand for an original 
(dzh) rather than for (zh), which would have been best rendered by 

^ Ebit and ott, hotUMe and oMrie, World, the italics are mine) : ^Xns « 

both occur. feia undurwen't nn inTil'imiorj ablw'- 

t Thus in : Fhonotypy by Modifica- j^ und t> Aiwnd imaelf lidTW^st tw 

tion, a means by which nnnsoal types it prim'itiT cumplek'j'^UL and in'di- 



be dupenaed with on a plan pro- ^ena ; that is : Thus hia &oe under- 
posed by T, W, Hill (the &ther of Sir went an involuntary ablution and he 
Bowland Hill, and a well known or- found hhuelf reduced to hia primitiTe 
ihoepist and edncationaUst) printed in complexion and indigence. 
1848 for private circulation only, the * In the most recent Greek vrC is 
last sentence runs thus gt is a quota- used initially for (dzh), as rr{9fU 
tion from Qoldsmith's Citisen of the (dahami*) a mosque. 



Chaf. 17. § 8. K. L, M, N, NO — XIVTH CENTUKY. 315 

(or ^. The middle Greeks accoiding to Diez also ifiote r^for ek, 
BB* PtT^dpSo^== (ritshard-os) for Eichard. These transcriptions are 
precisely similar to Salesbuiy's UiurUf tsiff, tstMuWj tston, for ehurehef 
chefe, Jeiu, John^ and shonld eyidently be interpreted in the same 
way. Even in PalsgraTe's time he makes Prench / =: English /, 
which we know (p. 207) was then (dzh), but this certainly only 
implies a rooted mispronnnciation, because we know that although 
(zh) had not then been developed, in English, it existed in French 
(p. 207). But it implies the traditional pronunciation in English, 
because PalsgraTe was decidedly archaic in his tendencies, as we 
have seen in his retention of {%%) for long % (p. 110), and (uu) for <m, 
ow (p. 149), out of the xv th into the zyi th century. This mispro- 
nunciation therefore is in itself a strong proof of the old pronun- 
ciation of j as (dzh). If to this we add that in the present pronun- 
ciation of the iM'onnan peasantry (tsh, dzh) are occasionally used for 
(ah, zh),^ it will be dMcult to suppose that ehj j\ in Chaucer had 
any other meaning than (tsh, dzh). 

K 

E in Anglosaxon constantly generated tsh in English, as already 
explained (p. 205). The orthography of our MS. and the alterations 
of words to suit the rhyme, shew that although in many cases the 
custom was firmly established, in others there was a fluctuation of 
use similar to that in the present day between hreeks, hreeehes, Scotch 
hri^p, kirkf English bridge, ehureh. The termination -/fs ^^ -^*^ ^^^ 
become generally -ly = (-1»V) in Chaucer, but traces of the original 
form remain as 4ik, Itch ; thus we have : sikurly 137, 154, against : 
sikirlik 3889, and: smoterlich, dich 3961 = (smoo'terlitsh, ditsh), 
=s dirty, ditch. Against : the holy bHsful martir for to seeke 17, 
we have: withoute more speche, not longe for to seche 785, I 
schuld yow seeehe, in softe speche 6993, and we may compare our 
modem words seek, beseech. Against the common form toerk, as in : 
that was a clerk, al this werk, 11417, we have the altered forms : 
wirche, 2761, 7559, 9535, werche 4986, and so on. Such changes, 
which have been shewn to be common to other languages, confirm 
the yalue of ^A as (tsh) even in Saxon words. The pronunciation of 
ieh as (itsh), in the phrase : so theech 12857, for example, =» so the 
ich (soo thee-tsh) is singularly corroborated by QiU's observation 
that in the East of England '' pro (s) substituunt (z\ ut (ztq) pro 
(stq) cano ; et (ttsh) pro (ai) ego : (tsham) pro (oi am) sum : (tshtl) 
pro 691 wil) volo : (tdi» voor Ji) pro (ei war*ant jou) certnm do," see 
supiu, p. 293. 

L, M, N, NG 

L, M, N must have been (1, m, n) as in all languages. The ter- 
mination 'le froim the French is occasionally written -tJ, -i7, -y/. It 

^ "Comme en anglau, D se fiiit TCH ; Ibhim^ chieii, Tehidbomy, 

sentir devant G et J, comme dans Gerct, Cherbourg." Le H^richer, Glossaire 

brebiB [Bgeroe], .... CH se pro- Normand, yoL L pp. 30 and 32. 
nounoe sonyent conune en anglais 



316 P, PH, QU. K — nVTH CEI9TUBY. Chap. IV. J 3. 

will be best to call it ('1) as in modem English. Before a following 
Towel it probably became (1) as : simple and coy 119 = (sfmpl-and 
cui) jnst as in modem English we have double, douhling not douhle- 
ing, Le. (dab'l ddb'liq) not (dab-'liq). As there is a difficulty in 
establishmg a nasal value of n in Old French,^ there can be no 
thought of its occurrence in Chaucer. 

NG was either (q) or (qg) (v occasionally one and occasionally 
the other as in mod^ English . Modem use can be our only guide. 

P, PH, QTJ 

There is no reason ist supposing p, ph, qu to have been anything 
but (p, f^ kt(7), but of course it is impossible to determine whether 
qu was not (kw, ku) instead of {kw). In Chap. V, § 4, No. 1 & 3, 
the £eu^ of the Bmuc and Gothic alphabets haying a single sign for 
this sound, has led me to suppose that it was r^dly simple Chu^), 
and not doable (kw, ku), even at that early epoch. The use of two 
letters ew m Anglosaxon would not decide anything, as (Jew, ku) 
would be a sufficient approximation for all purposes of writing. 



B presents the same difficulties as m the mth century, yet we 
cannot allow it to have *any yalue but (r). It must however have 
affected the preceding vowel,' as we could otherwise scarcely account 
for the use of or, ^, tr in the same words, as warehe 9231, werh 
481, wircMng 8371. In one case at least we find or where the 
modem form is ^, as : thurgh the cite large, with doth of gold and 
not with Barge 2569, but bofii eerge, earge are old French forms. It 
is also observable that many words in which the sound was (ar) in 
the XVI th century appear as (er), thus, yerde, emertej herte 149, 
werre,ferre 47 ; serve, sterve 1145, prive and pert 6696, pryvy and 
aperi 10845, deere, eteere 4867, 5252, stere, here 2151. Against 
wors 9183, we have : wers, ers 3731 ; I moot reherse, al be they 
better or werse 3173, it needeth nat to reherse, who can do toerse 

^ The ehief reasons assigned by tiiat t^ «, were prononnoed as nasals 
Diez (Gram, der rom. Spradi., 2 eo. eren in the xti th centoir^ Bappreads 
^^''-s^ voLl. p. 437), for considering the nse nasal »&= ^q). See Chap. Y, { 4, note 1. 
^^TiB^^Tgf^ nasals to be om are the * Mr. IdSrray says : ** R affects pre- 
identity'of^^ assonances on and en ; ceding vowel in Scotch eren while re- 
and the oonstanlj^nfasion of the forms mainmg (r). A simple vowel, diort 
androit endroiL^^ Bnt the modem before other consonants is long before 
^MmiM rhymes withYsffM, and yet there final r: heat hear, bat ba>, not nor, 
is no trace of nasali^^ere. I)iez also stout stoor, (mt mir, bat boor, not 
names the ancient rhwee of Salomon noor, stut stunr). And a before a con- 
fermlmny tabulon eonvimum; but these sonant foUowed by # mute is in the 
may have been due rataer to a peculiar South of Scotland ea (ie) but before r 
(-om) pronunciation od the Latin, the it remains (ee) so mam and mans are 
m and n beinf allowed xo rhymes as in distinguishea (men, mtai) bnt/itr, /ors 
many English popularUongs. At any are both {(eer, Uer) not (feety fi€t) £he r 
rate these fonns are nvt incompatible preyenting the closing of the sound." 
with non-nasality, whien was tne rule Compare Cooper's obsenrations, si^ri 
in Provencal, and Walloon, and there p. 70, where his (aaa) is the counter- 
are absolutely no grounds for st^posing part of (i^). 



Chap. IV. {8- »* 8CH T, TH, p V, W, WH, X Y, Z, J 317 

10913. Since the xrnth centory there has been a great tendency 
to pronounce er as (ar) or (ai), as in elerk, Derby ^ sergeant, and 
formerly servant, but the contrary tendency to use (er) for (ar) does 
not seem to have been at all developed except at tiiis earlier time.^ 
The concision of (ur, er) as in wors, were, is very like the modem 
concision of (ai, ej) with ('j). By a change of re into er the 
rhyme: ers, kers 3753 is obtained. The terminations -re, -er 
alternate, as : mordre 16538, morder 16539, at the commencement 
of two consecutive lines. It would seem then that we should 
always sound {-er), as (mur'der). The metathesis of r is frequent. 
§ 5, art 98, d. 

S, SCH 

S =^ (s) also represented (z) in plu]:al terminations, but never 
had the sound of (sh), which was always represented by 

SCH a combination derived from the Saxon se, in the same way 
as eh from Saxon e, to shew t&e effect of palatisation. In later 

times the e was omitted. 

T, TH, p 

T seems to have been g^ieraUy (t), but it became (s) in the ter- 
mination 'tian, see examples under C. 

TH, which is used promiscuously with ^ ia the MS., had pro- 
bably the same sounds as at present, and distributed in the same 
manner. Occasionally we meet with d in places where we should 
have expected th = (dh), as in fadur 100 = father, hider 674, 
thider, slider 1265, where the rhyme shews that the sound was 
really (d) and not f dh), but the (d) seems to guarantee the pronun- 
ciation of th as (^) when written in these words. 

V, W, WH, X 

These letters as consonants seem to have had precisely the same 
sounds as at present, but w was also used occasioniBdly as a vowel, as 
herberw 4143. In anoes 104, haUoes 14, which had arwe, hakoe in 
the singular, there seems no reason for not giving w its usual sound. 

WB was probably pronounced (jw) as in ags. and down to the 
XVI th century (p. 186). 

T, Z, ? 

The Y consonant is always represented by ) which is the same 
form as the letter used for s. The meanings of this letter must be 
disentangled by a consideration of modem usage, see suprii under 
GH (p. 310). 

The consonants seem to call for no frirther remark, and the mles 
laid down ia this and the preceeding section are sufficiently general 
to permit the reader to read any line in this edition of Chaucer with 
tolerable certainty, except as regards the use of the E final, which 
has now to be considered. 

1 For the xvn th century see p. 86. (ar) or (aa') in : clerg^jr, person, mercj. 
The Bey. C. T. Potts remarks that in eternal, universal, learning, the last word 
Sooth Shields er is nsoally pronounced being also called (leer'ntq). 



318 E FINAL — XIV TH ClUri'UJfcY. Chap. IY. f 4. 



§ 4. £7n the Pronunciation of E Final in the xiY th Century.^ 

That e final was at least occasionally pronoanced, and that its 
Bonnd did not differ, except in accent, from that of me, the = {mee, 
dhee) is conclnsively proTed by the following rhymes. It must be 
remembered that to me, to the, when the accent is thrown on to the 
preposition, become (too-me, too*dhe), with brief and indistinct (e), 
that is nearly (tocmB, too'dhB), or as in modem High Oerman 
(p. 321, n. 1). Hence the following rhymes shew that Borne, cyna- 
mome, sothe mnst have been (Boo*me, sinamoo'me, soo'dhe), although 
there may have been, as frequently at present, a little liberty taken 
with donble rhymes, and (soo'dhe) may have been used for (sooilie), 
and similarly (jnu'dhe) for (jnu^the), (swii'dhe) for (swith'e)* in 
the following couplets : 

That streyt was oomen from the oonrt of Seme. 

Fal lowde he nng, Come hider, lore, to me, 673 

Hj fayre bryd, my swete eynumome, 

Awake, lemman myn, and speketh to me, 3699 

8o faren we, if I schal my the 9othe, 

Now. ouod onre oet, yit let me talke te the, 12590 

Qnod toe Frankeleyn. comidering thin fouthe 

8o felingly thou spekest, sire, I aloue the, 10987 

EUes go bye som, and that as twithe. 

Now eood sire, go forth thy way and Ay the, 18222 

Al enly now, for the lore of Mute, 

Quod Fandaroa, for erery thynge hath tfme; 

So long abid til that the nyght departe, 

For aljM) siker as thow list nere ^t m«, 

And God tofbme I wol be thare tiiprjfme.* 4*198 

Bot&der, if it fo betide 

That I aproche at eny fide 

The place wher my ladi is 

And Jeanne ^at hire like ywyfl* 

To fp«ke a goodly word fmtime, 

For al ^e gold bat is in Some 

Ne cow]^e. I. after that bewro^, 

Bot all myn Ang^r outfTgo^.* i 282 

Here hy the stands for hye the, bnt the final e of hye is not pro- 
noimced, as also it is not pronoimced in aloue the, so that we read 
(alun* dhe, mi dhe). This omission will be considered afterwards. 

The middle e in Dertemouthe holds the position of a final e in : 
For ought I woot he was of Dertemouthe 391, where it is necessary 
for the metre, and it is observable that the e is here pronounced to 
this day by the peasantry in the neighbourhood of Dartemouth and 
Dartemoor.^ 

^ This section was written before I * Jnst as /, 9 rhyme in theyya, gref 

had had an opportunity of seeing Prof, is 7756. 

F. J. Child's admiraole ObseirationB * The rhyme time, by me, oocort 

on the Language of Chaucer and Gower. eight times in Grower, i 227, 309, 370, 

I have thought it best to leave my in- ii 41, 49, 114, iii 6. 369. 

Testigation almost in its original state, * Printed 6t)m the Harl. MS. 8869. 

and to give a complete account of these ' Priyate letter from Mr. Shelly, 

obsenrations in the following section. of Plymouth. 



Chap. IV. § 4. B FINAL — XIV TH CBNTUKY, 319 

• 

In the Man of Lawes Tale, there is a king called AUa, whose 
name on one occasion is reduced to AUe, which mnst have been 
pronounced (Al'e), so that ealle and hifalle which rhyme with it 
mnst have also been (kal*e, bifal'e) in — 

ManriciiiB atte ftrntstone men him eatt$. 

This coDstabil doth come forth a meesager. 

And WTot to his kyng that cleped was Alk, 

How that this bli^tydyng is In/alU. 6143 

Scarcely less convincing than the above instances is the case of 
the plurals in -m, where they do not at present form a distinct 
syllable.^ Not only are these frequently spelled -m,* as is the case 
still in Scotch,* but they also often rhyme with the verb w. Thus, 
taking first those spelled with es : — 

For sondry scolis maken sabtil eleries ; 

Womman of many a scole half a clerk »*• 9301 

How schuld I thanne, that lire in such pleasannoe 

As alle weddid men doon with their uyvM, 

Gome to blisse ther Crist eteme on l^$ i$f 9625 

Him wolde he snybbe scharply for the wmety 

A bettre preest I trowe ther nowher non it, 626 

Crist, which that is to every harm triacle. 

By certejn menes ofte, as Imowen elerkes, 

Doth thing for oerteyn ende, that feel derk is, 4900 

Thy wyf eek and thy wenohe sinfully 

Dronke of the same yessel sondry wynu; 

And heriest false goddee cursedly ; 

Therefore to the schapen f\il gret pyn» $$, 16718 

Withinne the doyster of thi blisful iffdet 

Took mannee schap the eternal love and pees, 

That of the trine compas lord and guyde %$, 1 197 1 

And nyl himselye doo no gentil d^iet 

Ne fblw his gentil aunceter, that deed ie, 6737 

In the following the plural is written -m, but it rhymes with is 
in precisely the same way. 

Of catapus, or of gaytre beriit 

Of erbe yye that groweth in our yerd, ther merep it A 16461 
Ther schuln ye se expresse, that no dred it, 
That he is gentil that doth gentQ dedit. 6761 

Te loke as though the woode were M of ihepyt, 
Sit doun anoon, and tel me what your grrf it, 7766 

After the opynyoun of oerteyn elerkit, 
Witnesse on him, that eny parfit derk it. 16721 

And for that faith is deth withouten werkie^ 
8o for to werken give me witt and space, 
That I be quit fro thennes that most derk it, 11992 

• Which gift of God had he for all his wypit f 

No man hath such, that in the worid on lyve it, 6621 

^ In the difficult combinations writit, -« in familiar Tersification, and in prose, 

priettty we hear ^erally in the pro- even in the xsy th and xyth century, 

yinces, (rtst'CE, priist-tz). as shewn in Mr. Murray's paper, sagik 

* Sometimes Misused, with the same p. 287, note 1. 

pronunciation as -if or -et, (p. 298). ^ These lines are eyidentiy corrupt 

* This Scotch final -t«, generally as they stand. Morris reads 3*233, Of 

fbrmed a distinct syllable in serious erbe yre growftrf^# in our yezd, ther 

poetry, but was practically reduced to mery is. 



S FINAL — XIVTH CBNTUBY. Chap. IV. § 4. 

• 

So made he eek a temple of fiUi ffodit, 

Mom nof^ihd do A mmgihAt more forbod is f 10169 

Bnt me was taught, non^ longe tyme foon is. 

That sYimes Cnrt went nerer but <wy« 

To weddyng. 6591 

Alias! andcanyehenagastofMMMfifff/ 

Nought^ God wot, hot Tanite in ticwwi sir. 16407 

Since in plods, place is 7349, the final -m must of necessity be 
prononncedy it is not reckoned among these examples, which are all 
that I have noted in the Canterbuiy Tales. To these, however, 
should be added, as equally convincing, — 

Take yonre disport : I njl Here no talii; 

I know yow for a trewe wif^ dame AUi, 5901 

From hons to hons, to here sondry talis, 

That Jankyn derk, and my gossib dame Alis, 6129 

It would be impossible to read many lines in Chaucer without 
finding that the number of syllables in a line would be constantly 
in default, if the final ^'s were not reckoned. At the same time 
the number of syllables in a line would often be in excess, if every 
e final were reckoned. Again, the slightest examination shews us 
words which are at present identical, difiering in different places 
by having and not having a final e. That this insertion or omission 
of the e final is not due simply to carelessness or option of the 
scribe,^ is apparent from the presence or absence of the e being 
generally essential to the metre, or the rhyme, and a notion seems 
t:> have possessed some persons, that lines could be made to scan by 
omitting or inserting these ^'s at pleasure. The examination of the 
prose tales, where these final ^'s are also found, ought to disabuse 
us of this absurd notion. We must admit that these final ^'s formed 
a part of the language of the time, and that there must have been 
some reasons for their insertion and omission. These we have, if 
possible, to discover, and the first step is to examine two modem 
languages, German and French, in which final ^'s also occur, and 
which are the living representatives of the Saxon and Norman 
elements of which Chaucer's poems were composed. 

Final e in Oerman, which is always pronounced where written, 
arises in several ways : 

1) it is a natural final of many words as Ituhe, Weise, JReise, 
Matte, Babe, K&se, Knabe, Seerde, Rerherge, toeise, leise, saehte, 

^ This refers to the Harleian, No. sundre (for sondry), 19 sesone daie, 

7334; other manuscripts are much 20 laie, 22 devoute, 23 ni^hte, 24 twente 

less strict, and the conftision in the (for twenty), 25 sondne folke be (for 

use of the final e seems to indicate a by), 26 pilgrimes, 27 towarde, 29 esede, 

date of writing about the middle of 31 euenrcnone, 32 anone, 34 ]mre 

XT th century or later, or else a scribe jowe, 3/ resnone, 38 condicionne, 40 

of Northern origin. In the first 42 whiche whate 41 eke whatte araie, 42 

lines of the prologue in the Lansdowne knighte, where the Harleian shews no 

MS. No. 851, with which Wright ff,and: 8 half; 9 smal, llher, 30 sonn, 

and Morris collated the Harleian 7334 31 had, 82 felawschep, where the Har- 

to form their texts, we find : 1 wy>e, leian has the final ^ It is obrious 

2 ha^e, 3 suche lycoure, 4 whiche that no conclusions respecting e final 

floure, 5 eke bre>e, 6 ha>e hethe, 7 could be deduced from such an ortho- 

ha^e ramme, 12 one, 13 straungere, 14 graphy. 



Chap. IV. § 4. K FINAL — XIV TH CENTURY. 321 

lange = (ruu'tf,* hhai^ze^ rai'z^, mytsy, rewb^, ksE'z^, kaaa'he, 
Heerdtf, neer'd^r, Her-ber'^h^, hhai'ze^ hi'ze, szakht'^, laq'e), and so 
fort<h, mostly representing some other vowel in old high German. 

2J it is inflexional, frequently expressing — 

a) plurals as der JFind die Wtnde, der Zug dU Zikge^ der Hertog 
die Henoge, &c. = (der bhind dii bhind'^, der tsuugu^h die teyjgli'e, 
der Herts'og dii nerts-oglw). 

h) dative cases singular, as dem Winde, dem Zuge^ dem Henoge = 
(deem bhind**, deem tsuugtrh**, deem Herts'oghtf). 

e) the plural of the indefinite adjective, as gute GutUfy aUe Men- 
sehe/i, lange JReisen = (guut'tf goet'er, al-* mensh'tfn, laq'e raiz-fli). 

d) the feminine singular of the indefinite adjective, as gute Mutter, 
arme Frau, keine Fhteht = (guut** mut'er, aim'* frau, kain*« 
frukw^ht). 

e) the nominative singular of the definite adjective in all genders, 
and accusative feminine and neuter, as der gute Mann die gute Frau, 
doe gute Weib, ich ehre die gute Frau und das gute Weih = (der 
guut'* man, du guuty frau, das guut'* bhaib, iAh ee*r*), &c. 

/) the imperative singular of verbs, as liehe Qott, ehre den Ebnig 
= (liib** got, ee're deen koBOB'ni^h). 

g) the first person singular of the indicative mood present tense 
of verbs, as ich liehe t An, ich fange an = {ikh liiby iin, ikh faq;e an). 

A) the first and third person singular of the present and past 
tenses of the subjunctive mood of verbs, as er sagt, sie komme; tie 
sagten er knme = (er zaaght, szii kom**, szii zaaght*«n, er ksEm**). 

t) the first and third person singular of the past tense of weak 
verbs, as ich liehte und er liebte dieseibe Freutidin = (iAh liibt'« und 
eer liibt** dii'zelb-tf froynd'in).* 

j) it is frequently added on to numbers in familiar counting, as 
eine^ zweie, dreie, viere, funfe, &c. = (ain*^, tsbhai'^, draiv, fii'r^, 
fynf-e). 

With all these reasons for adding on e, and the very sicnilar syl- 
lable tffi, (which on the Ehine is constantly called e\ the language 
is necessarily full to overflowing with this termination, whidi 
is consequently very often dropped or slurred over with great 
rapidity in conversation. But tiat poets with perfect sensations 
of rhyUmi, and immense power of expression, accept this flnal e and 
even multiply it in a single line, may be collected from this one 
example in (>oethe*s most finished drama, Taeso, Act I., So. 1. 

Ich brine' ihm seinen Sohn .... (I^h briq iim zain'Ai zoon .... 

Und ^me mne vaterlich^ Freods unt tail'e zaia'e fee*terlUli*« trojd'e,) * 

^ The final German ^ ^ in these * In these transcriptions the German 
transcriptions have been generally re- #w has been represented bj (oy), thi 
presented bv {t, en) as they are theo- sound preferred by Dr. Rapp, but (c^ 
retically held to represent these sounds, oi) are n^uent in the North, and (oi) 
but the reader should consult p. 11^, in the South of Germany. Some theo- 
note 1, col. 2, and p. 195, note 2, wliere reticians prefer (py), ana others (ay), 
tiiese cases are Mij discussed. 
* There are as many final e*s in Chaucer's — 

Him thoughte that his herte wolde broke 956 

(Htm tho&tffh'te dhat Hts Heer'te wol-de bree'ke), 
where the repeated e giyes a melancholy softness to the line. 

21 



322 B FINAL — XIV TH CEKTimY. Chap. IV. } 4. 

At ijtie same time the first line gives an example of the elision of 
an $ — ^ich brings ihm — ^before a following vowel. This is not a role, 
or a necessity, it is merely a matter of feeling. In snch a verse as 

Wie brennt mdne alto Wnnda.— (iMitf't Did Or mHU i i i r$) 

(Bhii brent main*# alt'# bhniid'#) 
the elision mein^ would have been impossible, or acconnt of the 
concord, although it would have avoided a trisyllabic measure and 
improved the metre. But throughout the first act of Tasso I have 
only noticed one instance in which Goethe has not avoided the 
necessity of an open vowel which he could not elide, namely 

Fiir holde Friioht# Hnet wahren Liebe 

(Fyr Hold'# frjJIAi^$ ainer Uiaar«ii liibv). 
where the natural pause at the ceesura assist the reader. Thus when 
iehj er, ihn, es follows a verbal -^, the e is always elided, as : gar oft 
beneid' icb^ irr* ich mich nicht, besser war's =s ware es, ich geb' 
ihm oft a= (gar oft benaid* i^h, i.r iArh miiUi niitht, bes'er bhiOBrz, 
iih geeb iim oft), and so on. The feeling is strongly shewn in 

ErwaehM Enrach^I Lass mu nicht empflnden, • 
Dast dn daa Gegenwarfge gam Terkennst. 
nSrbhakh*, erbhakh'tf ! Ltm nns nUht empfind*#n 
urn dan daa gecyh'e&bhart*^ gants ferkenat'), 

Where there are two other disions one marked in : Gegenwartfjge, the 
other unmarked in : veikemust, both similar to what might occur in 
Old Eng^dsh as $emde for semkle = seemed, iingst for singM. 

But Goethe does not hesitate to add on his « to an open vowel, 
as : ich thue was ich kann sa {ikh tuu*^ bhiis ikh k^n). 

The e of the dative case is frequently omitted, as after the itali- 
cized words in — 

Vtti laaa midi der Gelegcoheit, dem Oiiek^ 
Hir ist an dieaem AufmUuk genng — 
Ach I aie ▼cnagt mir eben jetzt ! Im OKek — 
• Doch war an WiaKnschaft, an rechtem Sirm — 

nind Un JDikh. der gelecj^h'dnHait, deem glyk — 
liiir ist an dii'zem aa'g|«7henbUk gennugirb' — 
Akh ! azii forzaaeht* miir eeb-#n jetst ! Im glyk — 
Dokh bhaar an bnifenalurfk, an reiUit*em sin — ) 

The imperative e is frequently omitted even when no vowel 
feUows, as 

Und litbt«r nicht — vtruUC daai ioh ea sage! 

(Und liibt er niiht — fertni* das iiHi es azaagh'tf.) 
The &ial $ is omitted in many other cases where the feeling of 
the poet requires it, even before a consonant, or at the end of a une 
whi^ the elifiton is not absolutely necessary to the metre, as 

Fast bleibt dem Sinn, nnd richtig dein Geschmack, 

Dein Urtheil /rod; stets ist dein Antheil gross 

AmGrosaen. — 

Una fiir dan Schata erkennte, den er Um^ 

Veigebens in der weitea WcAt gesncht — 
heiligt er 

Den PfiBud, den Mi ibr schSner Fuss betrat — 

Ioh sab ihn hmt* Ton fern ; er bielt ein Bnch — 

Und bist dn in gaimd\ so will ich treiben — 



Chap. IV. § 4. E FINAL — XIV TH CENTURY. 323 

Die Menge macht den EttnBtler trr* imd scheu — 
Von fremden Heerden Witi und Bosch erfiillt — 

g'est blaipt dain zin, nnd rUhtn^h dain eeshnuik, 
ain ur-tail graod, diteets ist dam an'taU grooe 
Am CTooB'fli — 

UnB Vft deen shotB erkent*«, deen er laq 
Feigeeb'enz in der bhait*en bhdt geznnktrht — 

Hai'lLbnt eer 
Deen pfkad, deen laiz iir shoecen'er funs betraat* — 
Vita zaa iin Hoyt fon fern : er milt ain bunkt^h 
Und bist dnu tea eelind', zoo bhil iJErh traib*en-^ 
Bii nieq*« makht den kynst'ler ijr not sboT — 
Fon fremden Hoerd'en ohiiz and bosh erfylt' — } 

All these examples are taken from the first act of Tasso. In 
lyrical poems we find similar omissions, not merely for the sake of 

rhythm or force, but also for the sake of rhyme^ Thus in the 
Maylied. 

Zwifiohen Waizen and Kom, (Tsbhiah'tfn bhaiU'Mi nnt korn, 

Zwiscben Hecken and Dom Tsbhish'At Hek'en nnd dom, 

Zwiachen Baumen and Qraa Tsbfaiah'^n bo^ en and graaBf 

V^o geht *8 liebcben P Bhoo eeet -s hib'ArhAi ? 

Sag mir das ! Szoa^ mir das ( 

An dem Felsen beim Flusi^ An deem fels'^n baim flos, 

V^o sie reicbte den Eass, Bhoo zii raiJirht'« deen ka% 

Jenen ersten im Qrtui^ Jeen en erst-m im graas^ 

Seh* ich etwas ! Szee iib et'bhos* ! 

Ist sie das P 1st seii das P) 

Here Oras (graas)'for Grose (graaz'^), and Flwa (flus) for Flusse 
(flus*^) are necessary for the rhyme. The most common omission 
is that of the dative e^ but even the essential final e is occasionally 
left out, thus in the lines An Luna, we have Buhe (ruu'tf) abbre- 
viated to Buh^ (ruu) for the rhyme. 

Und in woUnstroUer Jiuh* (Unt in bboMostfol'er ran 

Sdh* der WellvertcAlagi'ne Bitter Szes der bhelt*fershlatfgh'n« rit'er 

Darcb das ^laseme Gegitter DorArb das glEsz'emtf geffit*er 

Seines Macfebens Nacbten zn. Szaines nuBsd'Arhens nEiirnt*^ tson.) 

Less common and, no doubt intentionally, very harsh, is Schiller's 
DormerBpracK (don'er,shpratfkh') to rhyme with nach (noakh), in 
his Kinde$-mdrderin, st. 9. 

On the other hand in Goethe's GlUek der JEntfemung (Glyk der 
Entfemuq) we have an e apparently added in Olikeke for Gliieky — 
really an archaism from the middle high German OelUeke^ — also for 
the rhyme and metre. 

Trmk\ o Jiingling ! beil'ges Olueke (Tri^l!^ ^ jyq-liq ! Hail'^bes glyk-f, 

Taglang ans der Oebsten Blicke. Taogb'laq aos d^ liib'st^ bl3L*«.) 

All poets do not avoid the open final e with the same scrupuloiis- 
ness as Gk>ethe, thus Wilhelm Miiller in his Alexander Tpetlanti has 
An des Mittags Horizonte bing sein An^^ Nnverwandt. 
(An des mit*taakbs HOO'iitson-td mq szain aagu7by un'ferblumt*) 

Such examples are however rare. On the other hand the omis- 
sion of final e for rhyme or metre is very frequent. Thus for rhyme 
in Euckert's D^r Betrogene Teufel (der betrcwghen^ toyfel), EiV 



324 E FINAL — XIV TH CENTURY. Chap. IV. § 4. 

(ail) is used for EiU (ail'tf) to rhvme with Theil (tail). In Heine's 
Die Orenadierey already quoted lor non-elision, we have Grenadier* 
twice to rhyme with Quartier, mir (kbhortiir-, miir), and hitV 
(bit) to rhyme with mit (mit), and for metre 

Und gurf mix urn den Begen. (Und gyH mir umn den d«0gh'Ai.) 

These examples, which could easily be greatly multiplied, will 
seire to shew how a living language deals with its final ^'s, and 
Germans know that this treatment of e final is not a mere license 
taken by the poet to help him out of difficulties, but is on the con- 
trary a source of great power of expression, giving force and cha- 
racter to many passages by omission, and soffiess and delicacy to 
tiie others by the frequent use of the final e. Hence we are led to 
look upon the use and disuse of this letter, (the feeling for which 
has been entirely lost by Englishmen,) as a great resource for the 
poet, and a great beauty in the language. To those whom long 
custom has made familiar with the German language and the music 
of its poetry, the idea of constantly clipping off these final d's in the 
English fadiion would be distasteful and barbarous to the last de- 
gree, and their frequency conveys no feeling of trailiness or weak- 
ness, as it does to the mere English reader. 

Proceeding to French we meet with a new phenomenon, an 
existing system of versification founded upon an obsolete system of 
pronunciation (p. 119, note). In looking at French songs when 
set to music, we see that all final ^'s are pronoimced, except before 
a following vowel or a mute A, and that the -ent of the plural of 
verbs is cdso pronoimced as «, (except in the combination -aient 
where it is absolutely mute), although it is not elided before a fol- 
lowing vowel. But in common French discourse this final e and 
many medial ^'s may be said to be entirely elided.* The consequence 
is that there is a great schism between tiie language of poetry and 
that of common life. When singing, the French not merely pro- 
noimce these ^'s, but dwell upon iiem, and give them long and ac- 
cented notes in the music. This recognition is absolutely necessary 
to the measure of the verse, which, depending solely upon the num- 
ber of the syllables in a line, and having no relation to the position 
of accent, is entirely broken up and destroyed when these syllables 
are omitted. And yet when they declaim, the French omit these 
final tf*s without mercy, producing, to English ears, a hideous rough 
shapeless unmusical result, which nothing but a consciousness of lie 
existence of the omitted syllables can mass into rhythm.' 

1 In M. Jobert*B CoUoquial French poetrjr (in tragedies especially, and 

giondon, Whittaker, 1854), M. and principally in those wmch are con- 
lie. Th^riat's Phonoeraphe and Tour- sidered as standards of classic pnrity,) 
rier^B Model Book (4tn ed. 1851, Lon- is seldom pleasant to English ears; 
don, Nutt), will be found excellent but in the complaint whicn is gene- 
roles for shewing when this « ir or is rally made of the want of harmony of 
not to be pronounced. the French verse, there is not sufficient 
* TheutteM.Taryer, of Etca,inhi8 allowance made. One is too apt to 
Choix en Prou et m Vert TLondon, forget that the Ear^ accustomed to the 
1833), says : " The reading / r French rhyme and peculiar intonations of one*8 



Chap. IV. § 4. 



E FINAL — XIV TH CENTUBY. 



325 



M. Feline, who endeavoured to introduce a phonetic system of 
printing French as an assistance in teaching ignorant adults to read, 
has, at the end of his JSzm-cise de lecture PhorUtique^ Aventuree de 
Rohiruon CruaoS (Paris, Didot, 1854), given an Exemple de Licla- 
mation, consisting of a fragment of Lafontaine's Fahle (xi, 7), Le 
payean du Danube, which he has printed phonetically. We are thus 
presented with a Frenchman's views of how French poetry should 
be read, and as this is important in relation to the use of the final e, 
I think it worth while to give the greater portion of it in ordinary 
spelling and in a palaeotypic transcription of M. Feline's characters. 
The lines are supposed to be spoken by a German peasant to the 
Roman Senate. They are introduced by the following remarks : 

" Get exemple nous montre que, m^me dans la declamation, il est 
des e muets qui ne se prononcent pas, quoique leur presence soit 
n^cessaire ik la mesure syUabique des vers. Cette suppression a lieu, 
soit parce que les deux consonnes s^par^es par 1' e muet s'unissent 
fieuiilement en raison de leur douceur, soit parce que le sens est inter- 
Tompu. n importe aussi de faire observer que, presque toutes les 
fois que Ve muet est supprim^, la syllabe qui le pr^c^de en acquiert 
plus d'intensit6 ou de longueur.^ A la fin des rimes feminines, 
quand il est prSc^d^ d*une voyelle, cette voyelle devient plus longue.' 
On remarquera, en outre, que, lorsqus le sens unit la fin d*un vers 
au commencement du suivant, la liaison doit avoir lieu." 



lan^^uage, is not easily pleased by 
forei^ sounds ;— that want of habit of 
hearing French read renders it a bad 
judge in point of harmony j that the 
ftill and rapid comprehension of the 
meaning of the author greatly influ- 
ences our finding the words harmonious 
or harsh ; and now few there are who 
can boast of so iamiliaran acquaint- 
ance with a foreign language ! The 
following brief rdnimd of tne laws of 
French yeraification given by M. Tarver 
^.) may be useM. ** Measure and 
Bhyme constitute French Terse. Mea- 
sure is determined by the number of 
syUablee contained in the Terse. The 
longest French Tersee haTe tweWe syl- 
labfos, commonly called feet When, in 
the body of a Terse, a word ends with 
an e mu$L that is, an « not accented, 
and is followed by a word beginning 
with a Towel, the e mvet is blendea 
with that Towel, so as to form one 
sound, and consequently one foot only, 
instead of two. When the e muet is 
followed by an «, there is no elision. 
The termination mt, of the third per- 
son of Terbs, which, in prose, is gene- 
rally blended with the following syl- 
lable, if it begin with a Towel, must in 
Terse, be souMed as a distinct syllable 
or foot, but, in the third person plural 
of the imperfect and conditional of 



Terbs, such as parlaientf parUraimt, 
the mt of aient does not form one dis- 
tinct syllable, because there is but one 
sound uttered, jMir-foim/, par-U-raimt. 
Some diphthong form two syllables, 
and some one, at the option of the 
author. The e^attre is a rest which 
comes after the sixth foot or syllable in 
heroie Terse, and after the fourth syl- 
lable in Terses of ten syllables. — ^There 
are no blank Terses in French ; they 
always rhyme. There are two sorts of 
rhymes, tne mateuline which ends with 
a consonant or combination of letters 
forming one foil sound, such as, lan- 
guissan/, Tanit^, &c., ihe/#inmtiM with 
an e muet. In heroic Terses, the rhymes 
must be regularly and alternately, two 
masculine and two feminine. If a 
stanza end with a masculine rhyme, 
the following must begin With a femi- 
nine, and yice yers^.*' *^£ty'ambement, 
the running on of the sense firom the 
end of one yerse to the beginning of 
the following. It is a fault and to be 
ayoided," but is often designedly com- 
mitted by Victor Hugo and recent poets. 

1 This Mr. Feline has not marked 
particularly, I shall therefore place two 
dots (..) in place of the suppressed **# 
muet,*' in order to guide the reader. 

* This he has marked, and henoe I 



326 E FINAL — XIY TH CBNTHRT. Chap. IV. $ 4. 

Le PATSAJr J)v Daiotbb. — ^Fbaoxsht. 

Oraignez, Bomains, craignez que le ciel quelque jour 
Ne transporte chez yous lea plenrs et la misdre ; 
Et mettant en nos mains, par nn juste retour, 
Les aimes dont se sert sa vengeance a^v^re, 

U ne yous fasse ^i sa colore 

Nob esclares a voire tour. 
Et pourquoi sommes nous les votres ? Qu*on me die 
En quoi yous valez mieux que cent peuples divers. 
Quel droit vous a rendus mattres de Tunivers ? 
Pourquoi venir troubler une innocente vie ? 
Nous cultivions en paix d'heureux champs ; et nos mains 
£taient propres aux arts, ainsi qu'au labourage. 

Qu'avez vous appris aux Germains ? 

lis ont I'adresse et le courage ; 

8'ils avaient eu Tavidit^ 

Comme vous, et la violence, 
Peut-^tre en voire place ils auraient la puissance, 
Et sauraient en user sans inhumanity. 
Celle que vos pr^teurs ont sur nous exerc6e 

N'enire qa^k peine en la pens6e. 

La majesty de vos auiels 

EUe m^me en est offens^e ; 

Car sachez que les immortels 
Ont les regards sur nous. Gr&ces & vos exemples 
lis n'ont devant les yeux que des objets d'horreur, 

De m6pris d'eux et de leurs temples, 
D'avarice qui va jusques k la fdreur. 
Bien ne suffit aux gens qui nous viennent de Bome, 

La terre et le travail de Thomme 
Font pour les assouvir des efforts superflus. 

Betirez-les: onne veut.plus 

Cultiver pour eux les compagnes. 
Nous quittons les cit^s, nous fhyons aux montagnes, 

Nous laissons nos chores compagnes ; 
Nous ne conversons plus qu'avec des ours afi&eux, 
D^urag^s de metire au jour des malheureux, 
Et de peupler pour Bome un pays qu'elle opprime. 

mark the prolongation by reduplication long YoweU in French, and that I have 
as usnal. It is to be obaerred that M. strictly followed his system of notation, 
Feline seldom admits the existence of except in his employment of ti^e hyphen, 



Chap. IV. § 4« ^ FINAL — XIV TH CENTUBT. 327 

Lp ptf,izaA dy Danyb. — ^FragmaA. 

GrsRJej BomeA, cr«nj> kp h siel kelkp zhur 
Np traAsporttf sh^ vu le plprz ^ la mizer.. ; 
JE metaAt aA no meA, par 9a zhystp ivtor 
Jjez armp doA sp ser sa vaAzhaAw s^er.. , 

n np Yu f as aA sa koler.. 

"Noz eskl0v..z a votiv tor. 
JE porkoa som.. nn le votr.. ? X-oa iii9 dii.. 
Aa koa Yu Ydle mioe kp saA ppplp diver. 
Kel dma vaz a r&idy metiv dp l-yniver ? 
Purkaa ypnir tnibW yn inosaAtp yii.. ? 
Nn kyltiYioAz aA pe d-proe shaA ; e no meAz 
Ete propipz oz ar, eAsi k-o labnrazh.. 

K- ave vnz apriz o ZbermeA? 

Hz OA 1- adres e h kuiazb.. ; 

S- ilz avet y 1- avidity 

Kom.. YU, p la violaAs.., 
Ppt etr- aA votrp plas ilz ore la pyisaAs.. , 
J? soret aAn- yze saAz inymanitp. 
SeL. kp Yo prptPT oa syr nnz pgzerspp.. 

N- aAtiv k- a pen- aA la paAspp.. 

La mazhestp dp voz otel 

El., mem- aAn- et ofaASPP.. ; 

Kar Bashp kp lez immortelz 
Oa le ipgar syr nu. Grospz a voz pgzaApL. , 
In n- OA dpvaA lez job kp dez obzbe d-oripr, 

Dp mppri d- odz p dp Ur taApL. , 
D- avaiis.. ki va zhyBkPz a la fyrpr. 
BioA np syfit o zhaA ki nu yien.. da Bom.. : 

La ter p b trayalj dp 1- om.. 
EoA pur lez assuvir dez efor syperfly. 

Eptdrp le : oa np voe ply 

Kyltivp pur od le kaApai^.. . 
Ku kitoA le sitp, nu fyijoAz o moAtanj.. ; 

Nu lesoA no sher.. koApanj*. ; 
Ku np koAveisoA ply k- ayek dez urz afirae, 
Bpkurazbp dp metr- o zhur de malproe, 
JE dp ppplp pur Bom oa pp,i k- el oprim.. . 

wbicli he places bifon a pronounced and which I employ in the nsual pa- 
final **$ mnet,*' or a consonant that laeotypio manner, 
which runs on to the following vowel, 



328 E FINAL — XIV TH CENTURY. Chap. IV. { 4. 

Notwithstanding that this passage does not offer nnmerous 
examples of the disarrangement produced by modem speech in 
French verse, yet it is evident that had French verse arisen in the 
present day, or had it followed the usages of pronunciation, it could 
not have taken such a form. Thus the distinction between the 
masculine and feminine rhymes, which is so important in the con- 
struction of French verse, has entirely disappeared, sh^e, coUre, 
becoming (s^er, koler), do not differ firom divers^ univers (diver, 
yniver), though a French poet who attempted to make the first 
rhyme with the second would be laughed from Parnassus. The 
rhyme fnains, Oermains, has disappeared in (moAz, zhermeA), owing 
to a '^ liaison" preserving the » in one case, while it was lost in 
another. The open vowels, which are jso strictly forbidden, crop 
up, as in 

Comme tovs, et la Tiolenoe. 

(kom Yu # la yiolaAB.) 

This Hue also wants two syllables, which the singer would have 
added as — 

(komp TQz # la TiolaAw). 

Observe also how the lines 

Elltf mSme en est offens60 — 
B'aTaric# qui Ta josques h. la fnreur — 

suffer from the want of the italicized syllables. 

The composition of French verse is as purely regulated by rule in 
France as i^t of ancient Latin and Greek verse is at modem English 
schools ; it is thoroughly artificial. The French have got to feel a 
sort of rhythm in it as Etonians feel a rhythm in their own hexa- 
meters; but that the former at all resembled the rhythm known 
to the old French poets, can as little be imagined, as that the latter 
resembled the rhythm tiiat guided Virgil. Even the popular rhymes 
of Stranger connot always imitate the speech of the people, witness 
the italicized «'s in the foUowiug first stanza of PaiUa»$e ^ — 

J'snis n^ Paillaase, et mon papa, 

Pour m'lancer but la plac«, 
D'un coup d' pied queuqu' part m'attrapa, 
£t m' dit : Sauttf, PaillaBM! 
T'fUB r iarret dispos, 
Quoiqu^ f ay* 1* Tentri gros 
£t la Deu^' rubicond^. 
N' sauf point-z k d^mi 
Paillass' mon ami : 
Sau^ pour tout \e mond« ! 

From the French we learn then this lesson, that it is possible to 
have a versification which requires the pronunciation of e final, 
although it has disappeared from the language. Hence Chaucer 
may have used an e final in poetry, which was unknown in common 
speech. But the French e final, which has now disappeared, wm 
pronounced in general conversation as late as the xvi th centuiy, as 

^ (Euvres completes de P. J. de Paris, 1835, 2 toIb. 82mo., yoI. i. p. 
Bdranger, Edition reyue par Tauteur. 232, written in 1816. 



Chap. IV. § 4. E FINAL — XIV TH CENTURY. 329 

we know both from Palsgrave, and from Meigret, and hence it must 
have been so pronounced in Chaucer's time, and must have formed 
part of the rhythm of the Prench verses with which he was well 
acquainted. 

This examination of German and Prench versification has led us 
to two very different results. In German the final * is a living part 
of the language and metre, affecting the music of speech, a real 
element in prose and verse, in the loftiest and the homehest discourse. 
In French the final e, although the representative of other original 
vowels, the note of feminine and of many parts of verbs, and of con- 
stant occurrence in writing, has died out as utterly in French as it 
has in English speech, but forms an element of ^e commonest as 
well as loftiest versification of the present day, any attempt to build 
verses upon the theory of its disappearance, as in English, being 
scouted as low and vulgar. What was the case with Clmucer ? 

The foundation of our language is Saxon. The construction of 
our sentences, the expressions of the relations of ideas by the order 
of words, has undergone little or no change frx>m a period when 
French words were still unused. The only effect of the introduction 
of French words was to enlarge our vocabulary, not to alter our 
grammar. Hence it would seem more likely that while the Ger- 
manic e final was still in use in our language, it was employed by 
English poets much in the same way that it is now used by German 
poets. That is, we have every reason to suppose that it was gene- 
rally, as we have proved that it was occasionally pronounced, 
whetiier it was a substitute fpr some other original vowel or was 
merely inflexional, but that in both cases it was omitted, ^ when not 
destructive to the sense, before another vowel, or whenever its 
omission gave dignity, force or precision.' 

In French venifieation the rule for the elision of final e before a 
subsequent vowel or h mute was absolute. We should therefore 
expect to find this rule absolute in Chaucer at least for French 
words. But it may have been only partially adopted. In this case 
however we have no occasion to go to a French model. In Chap. V, 
§§ 1 and 2, we shall see that this was the rule of English versifica- 
tion, even in the xm th century. 

It is quite possible that^ as the inflexional condition of our lan- 

1 In (}eniiaii and French poetry the altogether even in reading Latin Terse, 

omisnon of the Towel is complete and Except in a few instances, as T, f^ &c., 

ahfiolnte. It is not in any way slurred the French do not mark the dision of 

oyer or rapidly pronounc^ in connec- a final e hefore a foUowing vowel, and 

tion with the following Towel, as is in old English the yowel was written 

the case in Italian and Spanish poetry, even when elided, 
and eyen in Italian singing. The * Occasionally, hnt less frequently, 

Oermans, like the Greeks, do not even the final e may naye heen also omitted 

write the elided yoweL The Latins for the sake of the rhyme or the metre, 

wrote the elided yowel as the Italians but in such cases the poet must haye 

doj and may therefore haye touched it felt that the sacrifice would haye heen 

^ " as in the English custom of greater to turn hisyerse so as to render 



reading Latin yerse, whereas it is the tiie elision unnecessary. 
German custom to omit such yowels 



330 B ^NAL — XIV TH CEKTURY. Chap. IV. { 4. 

goage miderwent a rapid degradation in the xy th centniy, and was 
certainly much inferior in tiie xivth to what it was in the xm th, 
(several of the inflexional #'b having perhaps disappeared even in 
Chaucer's time), and as most of the manuscripts belong to a period 
of at least a generation after Chaucer's death, this disuse of the final 
# may have considerably advanced before the best copies of his writ- 
ings, which we possess, had come into existence. It may therefore 
well be that the scribe has frequently introduced or omitted final «'& 
with rather an indistinct and uncertain feeling as to where they 
ought or ought not to be pronounced.^ 

We know indeed that even in the xnth century, when the final 
e^s had altogether disappeared from speech, they were considered an 
indispensable ornament in writing, and were added on without any 
knowledge on the writer's part whether their addition was or was 
not historically justifiable.* 

Before judging from the inner part of a line in Chaucer, whether 
the final 0's that are written should be pronounced or mute, it is 
necessaryto obtain some feeling as to the style and character of his 
verse. We have no occasion to consider the shorter lines of Bir 
Thopaz, nor the grouping of the lines into stanzas. The question is 
only, of how many syllables did one of Chaucer's longer lines consist, 
and where did the stress fall? 

The last question requires the position of the accent ' in Chaucer's 
words to be considered. Or rather the two questions must be con- 
sidered together, for there is no means of determining the position 
of the accent but by the metre. We .may assume that the rhyming 
syllables had sufficient stress to make the rhyme frdly audible, but 
we must be aware of concluding that therefore they had the chief 
stress. This rule would be gencurally true in German verse, — where 
however it is sometimes transgressed,^ — ^but it is not at all true of 
French verse. Many writers assert that French words have a fixed 
accent. In the xvith century Palsgrave marks the position of the 
French accent and lays down rules for it. Bo does the very high 
phonetic authority, Bapp, in the xix th century. Nevertheless one 
of the great pecuHarities of French, as distinguished from Italian on 
the one hand, (representing its Latin element,) and German on the 
other, (representing its Frankish element,) is the absence of deter- 
mxnate stress upon any syllable in a word. French speakers do fre- 
quently put a stress, but that stress varies with the feeling of the 
moment, and without affecting the intelligibility of a word. I have 

^ See suprl^ p. 320, note. eyer, the present inyestigationB make 

* See the latter part of SaleBbmyB it requisite to reconsider. In these 
obeerrations on in his Welsh pronnn- pages I haye strictly confined myself to 
tiation, infrl^ Chap. YIII. {1. the smallest amount of discussion which 

* The following remarks on the yery my object allowed. 

difficult subject of acorat and metre, ' Compare etutae in the Maylied, 

make no pretension to completeness, suprii p. 323, a word which generally 

The two Tolumes of Mr. Guest's JETw- has the stress on the #<, as in ouier oom- 

iory of BnglUh Msfthms, 1838, shew pounds of 4t, but there has nearly an 

the extent of the subject, which, how- eren stress on both syllables. 



Chap. IV. } 4- ^ FINAL — XIV TH CENTUEY. 331 

heard the last word in les champs JBlyiiss pronounced with a distinct 
stress on the first syllable on one occasion, on the second on another, 
and on the third on another. A German speaker is apt to accent 
the final syllable in French words, an English speaker the first. It 
is the 0vmnes8 with which a Frenchmen pronounces the syllables 
that gives so much peculiarity to his pronunciation of English, and 
reflects his national habit of speech, a habit also shared, as I am 
informed, by the Turks. A simple example of the effect of this 
evenneis is that most Englishmen feel the French Alexandrine to 
consist of four measures, of three syllables each, accented more or 
less distinctly on the last syllable, whereas the English and German 
Alexandrine founded upon it consists of six measures of two syllables 
each, more or less distinctly accented on the last. That the French 
allowed very evanescent syllables, as for example the final e, to fall 
on the even places, may be seen from the italicised syllables in 
Comeille's lines {Limitation de lews- Christ) : 

Lee tenebr«t iamais n'approchent qui me rait ; 

£t paitont rar mes pas il tromie Tn ionr latis irait, 

Qui porte iuMjue an coeur la Imniertf de Tie. — 1, 1, 1 

Ne lui B^anroit of&ir d*afi:reab/«f victimes — 1, 1, 3 

£t la Terta saiui eux est de telltf valenr, 

Qa*il Taut mievx bien sentir la douleor de tes ftntea, 

Que s^anoir definir oe qu'est oett^ dooleur.^ If 1» 3 

"We also find the same word differently placed in a verse with 
respect to the odd and even places, which wiould therefore be dif- 
ferently accented according to any accentual theory. For example 
(Comeille, Imitation) : 

£t ta Terras qu'eiifiii tout n'est que ponite, 1, 1, 8 

Vanity d'entaaser richesses sur richesses. Ii 1> ^ 

Le d^nr de B9auoir est natnrel aux homines. 1, 2, 1 

Bone tous tes deairt h ee qu*il U fant faire. I, 2, 2 

Les Sfttuam d' ordinaire aymen^ qu*on les regaide. 1) 2, 2 

Qui puissent d*Tn Sgauant faire Tn homm« de bien. 1, 2, 2 

And so on, shewing that in the year 1651, when this was published, 
there was no proper determinate stress on any French woids. From 
this to the xivth century is a great leap, but the very fact that 
Chaucer employs his French words in the same way, leads us to 
infer that he was accustomed to the same practice in his French 
originals, thus: 

Tronthe and honour^ freedom and cnrtesie. 46 

And eTere honoured for his worthinesse. 50 

Sche was so charitable and so pitom, 148 

Thejr fillen gmf and criden pitoutly, 951 

Tathenes, for to dwellen in priaoun. 1025 

Oure pritoun for it may non othir be. 1087 

Fairest of faire, o lady min Venus, 2223 

And ye be Vtnus, the goddess of loTe. 2251 * 

^ If ^e text be correct we find precisely similar cases in Chancer — 
Ful wel sche sang the serried dcTyne. 122 

That often hadde been att0 parrys. 312 

As seyde himself mor« than a curat. 219 



332 E FIVAL — HVTH CENTUKT. Chap. IV. } 4. 

It is needlefls to heap np examples as tibe fact is well known. It 
is dwelled upon by Mr. Skeat,' but although he names the equable 
French pronnnciationy he seems to tbfnk the final stress in l^^n ^iab 
words to be dne to the French and the change of accent to be en- 
tirely F<nglish. It is more probable that the words were always 
pnmonnced with an equable stress, which allowed of their s^pearing 
in either position, and this was altogetiier French. 

There is at least one Finglish termination which could be placed 
either in an odd or eren place, namely -ifn^e^ thus in 
8jfngyng§ he wai mJUiwiyng$ al ^ day. 91 

-ynge occurs both in an even and odd place. This termination, as a 
true participial form, is difficult to derive from Anglosaxon, where 
the termination was -ende^ -^nde. In the Bomaunt of the Bose we 
have -ands in an even place — 

Poyntifl and sleeyea be welle niUmdi 

Bight and stregfat on ihe hande. 6*09 

They ihid hir telle boa they thee ftnde 

Cmteii and wya, and welle dotmde 6*83 

And in the Canterbury Tales, 

Tbuehand the cherl, ^ey aayd that sabtQte 7872 
But it occurs in an odd place apparently in — 

The God of Lone dehrrerhr 

Come lepande to me nastily. 6*59 

and in the Canterbury Tales, 

Ther ia M many an eyghe and many an eere 
Awaytand on a lord, anid he not where. 7635 

His meyne, which that herd of tiiis affiray, 
Com l^and in, and chased oat the frere. 7738 

and by the analogy of all Germanic inflexional syllables it ought to 
be unaccented.' 

As a verbal noun the -ynge came directly from Anglosaxon, and 
it occurs in an even place so early as Genesis and Exodus. 
pride and giteinge of lonerd-hed. t. 832 

Chaucer therefore apparently took the liberty of placing French 
words, foreign names, and English words with heavy terminations, 
as -ynge, -nesse^ and some others,' in any part of his line which 

^ In the additions to Tjrwhitt's pre- The change of form of the present par- 

Uminary Essay, Mr. Morris's edition of tidple is careftilly noted in Koeh, His- 

Chaucer, vol. 1, 172-196. Bell and tonsche Grammatik der Engbschen 

Daldy, London, 1866. See the list of Sprache, toI. 1, p. 342. to which I am 

words giyen by Prot Child in his indebted for the references to the 

Essay, reproduced in the next section, Romannt of the Bose, the text of which 

art. 99. Prof. Child cites as *'Ex- however, is unfortunately very doubt- 

amples of the French accent,'' which ful (p. 252). The form -efnie is very 

he evidently regards as lying on the common in Gower, and is generally 

last syllable— accented. See Prof. Child's obsenra- 

tiier was discord', rancour^, ne heyy- tions in the next section, art. 64. 

nes'se. 8308 > Prof. Child loc. cit. art 99, also 

glori and honou]<, regn'e. tresor' and notices felaw'e 2550, &c., fel'aw 650, 

rentfe) 15697 melle're mylle're 544, 8167; mel'er 

> Mr. Skeat accents it (ib. p. 185). 3923, &c., yeman' 6962, ye'man 101. 



Chap. IV. }4- ^ FINAL — XIV TH CENTURY. 333 

snited his convenience, most 'probably pronouncing them with an 
even stress on each syllable, which in process of time became trans- 
formed into a double method of accentuating. For English words 
generally the usual Oermanic rale of the stress on the radical syl- 
lable apparently prevailed. 

Chaucer's verse seems to consist generally of jwt measures, with 
or without a final unaccented syllable, forming a ''feminine rhyme," 
added at the pleasure of the poet. There is no trace of the strict 
alternation of couplets with masculine and feminine rhymes which 
distinguishes French verse of the classical period. Each measure 
properly consisted of two syllables, with more or less stress on the 
last, but each syllable might also have nearly the same stress. In 
the first measure the chief stress was often on the first syllable, as 

Bright was the day and bliew the firmament 10093 
Mr. Skeat has pointed out (ib. 174) that the first measure might 
consist of a single syllable, which then ought to have a certain 
stress, or at least be followed by a decidedly unaccented syllable, as 
Jfay with all thjn floures and thy greene. 1612 
Thsr by ayentore this Palamoun. 1618 

Now it schyneth, now it reyneth fastb. 1637 

His example 

I make pleynly my confessioun, 

That 1 am the woM Palamoun. 1737 

can scarcely be correct, as such a reading would be quite destrac- 
tive of the sense, for that, am, must be without stress, and /must 
have the stress. The line is therefore corrupt. Tyrwhitt reads 
ihilke for they another mode of correction would be 

That I am he^ the wofnl Palamoun, 

That hath thy prisoun broke wikkedly. 

Probably Mr. Skeat is right in admitting- a monosyllabic first 
measure, but it should not be accepted in any particular case, 
unless the single syllable it contains has a decided sb^ess.^ 

In the modem verse of five measures, there must be a principal 
stress on the last syllable 

of the second and fourth measures 
or of the first and fourth measures 
or of the third and some other measure. 

^ The first line of the Canterbury The Harleian 7333 has [swoote 

Tales seems to belong to this category. Whanne ])^ Aperyll w^ his shoures 

The Harleian 7334 reads [swoote where whanne is an Anglosaxon form. 

Whan that aprill^ with his schowres Caxton's first edition reads [sote. 

where the itaUcised e has no authority, ^^ *»»** ^VVria with his shouris 

compare Averil 6128, but is also found And Pynson's edition 1493, has [sote 

in tne Corpus MS. Oxford. The Whan that Aprille with his shoures 

Hengwit MS. reads— [soote Marking the monosyllabic first measure 

Whan that AueryU wiM his shoures i, j^Ucs, I would read fswote 

The Haxleiwi 1 768 reads- ^„ ^^ ^ ril with his schoures 

Whan that Apnll.w* his schouresswote «. .. , 

The Lansdowne 861 has [soote Similarly 

Whan J>at April wyj?e his schoures it/ bysmoterud with his haburgeon. 77 



334 B FIKAL — HVTH CKMTUKT. Chap. IV. } 4. 

There is alBO generallj a stress upon tibe last syllable of tiie fifQi 
measure, but if any one of the three conditions above stated are 
satisfied, the Terse, so far as stress is concerned, is complete, no 
matter what other syllables have a greater or less stress or l^igth.^ 
It is a mistake to sappose that there are 0(munonly or r^;ularl7, 
fiye stresses, one to each measure.' 

This role of stress is necessarily not so strictly carried out in 
Chancer, who was prorided with a nnmber of words hajing eren 
syllabic stress. Bnt on examination it will be found to hold toler- 
ably welL There are however many lines in which so many syl- 
lables come together, with little or no stress, that unless they are 
read somewhat $ffUahiealiy rat^r than by measures, or stress, we 
foil to feel their rhythm. Thus 

Thai^reryofyoasehalgotrAirhimlest 1860 

may be accented on the italicised syllables, (first and fourth mea- 
sures), in which case ofy<m schal go would be passed over lightly, or 
else tiie whole line may be read with an even stress like a French 
verse, and this seems the more probably correct method. 

Any measure may occasionally consist of three syllables, but in 
this case the two fint are always very light. In 

Wyd was his paruM, mnd ko%ueB fer aaondor. 493 
Biforn me mrwfiU wrtedisA creatiire. 1 108 

the third italicised measure has three syllables. In such cases it 
will be generally found that the first syllable is merely an in- 
flexional or derivative «, «i, er. 

It is not usual in modem verse to have two trissyllabic measures 
in the same line, or if they do so occur they must be widely sepa- 
rated. It is also not customary in modem verse, but it is not un- 
frequent in Chaucer, to give three syllables to the fifth measure, as 
Than with an angry woman donn in a Mou9. 6361 
As wel o9er Mr housbond as ot^t kU lovt, 6621 

^ The length of syllahles has much lines of Lord Byron's Oor$tur, marlring 
to do with the force and character of the eren measures by italics and the 
a Terse, but does not form part of its rektire amount of stress by 0, 1, 2, 
rhythmical laws. we hare — 

* Take for example the first six 

loitooot It 
O'er the glad Mooters of ih€ dark blue sea 

1 1 otoootst 

Oar thoughts at hotmdleMj and our tonlt as free, 

too lotoiot 
Far as tke breeu can bear, tke di/lows foam, 

• lotoooto t 
Surrey our empire, and bekold our home ! 

«oo I tiooo t 
These are our realnuy no limt^t to their sway — 

iiotoiitot 
Our flag tk$ teei^xe uHicko me$t obey. 

The distributioa of stress is seen to and others might think that it would 

be Tery yaried, but the action of the be sufllcient to mark stress and no 

rules giTen in the text is well marked, stress. The last line most nearly ap- 

Different readers would probably differ proaohes to haring fire principal 

as to the ratios 1 and 2, in some lines, -* 



Chap. lY. }4- ^ FINAL — XIV TH CBNTURY. 335 

If gentileB were plaunted nfUunUp, 6716 

For vilevn synfm deedes maketh a eherl, 6740 

That will nought he govenied af^ h&r wyvet.^ 6844 

Besides the stxess, the caesura plays an importajit part in modem 
verse. This consists in terminating a word, at the end of the 
second measure or in the middle of the third, or else more rarely at 
the end of the third or middle of the fourth measure. "Words 
forming a logical whole must in this case be considered as parts 
of the same word. Thus Chaucer's 

That sleptfM ai the night — with open yhe. 10 

(where the even measures are italicised) has the caesura (marked by 
a dash) after night, the end of the third measure, not at a/, or the, 
because al the night has the effect of a single word. 

If we now read Chaucer's lines with the pronunciation obtained 
in our previous investigations, we shall find it very difficult to say 
in general where the final «, when written, may not be sounded.' 
But the principle of economy would lead us to avoid the use of 
trissyllabic measures where they are not agreeable, or where they 
would be too frequent. 

Final e arises in Chaucer' from nearly the same sources as in 
G^erman: 

1^ as a substitute fit)m some original final vowel — eisential E 

2) as a mark of plural, oblique case, or definite adjective— rn/foc- 
tiondl, oblique, definite JS 

3) as a mark of adverbs — adverbial JS 

4) as a mark of the infinitive mood and gerund, past tense of 
weak verbs, and imperative mood — verbal JS 

5) as a representative of the French final e — French E, 

^ The triflflrllabic measures in 6621 precisely the same rhythm in a line in 

are avoided oy reading o*er for over, Goethe's Tasso, act 1 : 
as in modem times, and in 6740 by ein nen Hesperien 

reading maWth, Uns dustend bild^ erktnnsi da sie 

nicbt alle 

> "It is difficnlt to point out in- FUrholdeFriichteeinerwahrenLiebeP 
stances where the -e final is not sounded (ain noy Hespee-rim 

but it appears to be silent in dore 2424, ^^ dust'md bild'en, erkenst* du di 
fnU 886, regne 879, and beil$ 1328." rniht tl'e 

Skeat, ibid. p. 183. The reference Fyr Hold-e frykti-e ain-er bhaaT« 
numbers hare been adapted. Now on lii'be ?) 

ftTftitiining these lines — ^^ f*ct when the csraura occurs in this 

The rynges on the tempul dc^^Ao. pj^^^^^'^^^l 

^'^^^T/^r^^i^^-'^r^ &--rrprytXoro?nsi 

comparable to the above instances before a foUo^g vowel. Hence S 

where it is formed without a fimd #. ^^ -^^^^^ J^^^ , Mr Ske!lt 

And of the fest^ that ioat at hire wed- finom the whole of the ^ligfates Tale, 

dynffe. 885 come to nothhig. 
Ther as a best« may al his lust iulfille. 

1320 * Prof. Child's minute examination 

have trisyllabic tiiird measures, which of the final E's in Chaucer, is given in 

have never a bad effect, indeed we have the next section. 



336 E FINAL — XIV TH CENTURY. Chap. IV. } 4. 

The use of the final e seems to have been more regular in poetiy 
than prose, to judge by the prose tales in this manuscript, but this 
may be erroneous ; the reason may only be that the scribe, to whom 
many of the uses of e final had become obsolete, had no guide, when 
writing prose, to correct his more modem spelling, or, as is more 
likely still, at once used the orthography corresponding to his more 
recent pronunciation. 

The question now arises, was final e ever added on by the poet for 
the sake of metre or rhyme, as Ooethe apparently added on in 
OlUeke as shewn above (p. 323) ? It is possible, but not probable, 
as it would have been instantly detected as a weakness, unless it 
could be justified as an archaism, like Groethe's, or a colloquialism, 
as when zweiey drete, is said in Grerman.^ But the scribe certainly 
not unfrequently added on an « when it was not required, shewing 
that the value and meaning of the final e was disappearing in his 
time. Mr. Skeat calls this '* orthoepic" and considers that it has 
" solely to do with the length of the preceding vowel" (Ibid. p. 
189). I am more inclined to consider it " ignorant," and as point- 
ing out a later date for the writing of the MS. 8ee the observations 
on the Lansdowne MS. 851, supi^ p. 320, note. It would be im- 
possible to suppose that the writer of that MS. added on an ^ in : 
wyjTC, haJTc, suche, whiche,— examples which occur in the first four 
lines, — ^to shew the lengthening of a vowel which was not lengthened. 

The following examination of words with final E in the first 100 
lines of the Canterbury Tales will give a clearer notion of their 
origin and use. To each word is added the number of the line, with 
an accent after it when the word is final. From the metre alone it 
is of course geiierally impossible to determine whether the final E 
at the end of a line is to be pronounced. Therefore we may, for 
the moment, reject all such from consideration. When an apos- 
trophe is substituted for a final E, it shews that the e is written, 
but not pronounced, and is followed by a vowel or enclitic beginning 
with A. A double apostrophe shews that the e was written, but 
should apparently be omitted for the sake of the metre. When the 
word is in itcdics, it is essential to the metre in the middle of a 
verse. Prof. Child's remarks in the next section should be con- 
sulted by means of the list of Forms of Words in Chaucer and Gotcer 
re/erred to in Frof, ChihPs memoirs there appended. 

1. Superfluous final JF, that is, a final E not required by grammar 
or by Anglosaxon usage. Aprille 1, vertu* 4, nyn' 24, wey' 34, 
all' 38, fiftene 61', hethen' 66, mek' 69. Here Aprille 1, is really 
not essential to the metre, if we allow of a monosyllabic first mea- 
sure. Nyne 24, bjh^ fiftene 61', may have assumed the e as numerals, 
5 5, art. 39. Weye 34, is written wejje in Omnin, so that the e 
was no more an addition of Chaucer's than the e of Gliicke was an 
addition of Goethe's. The word occurs frequently without the ^, 

See Prof. Child on the cases where infr^ } 5, art. 13, 14, 16, 17, 30 ; and 
final e is found in Chaucer in words my footnote on art. 13. 
where it does not exist in Anglosaxon, 



Chap. IV. § 4. E FINAL — XIV TH CENTUKY. 337 . 

and should be so written here. Meke 69, frequently requires to 
have a final e pronounced, but Omnin writes fMoe, mso without 
a fmal e. 

2. French final E^ veyn* 8, melodie 9', natur* 11, %trawnge 13, 
pilgrimage 21', .78', corage 22', hostelrie 23', companye 24', i^ventur* 
25, space 35', 87', chyvalrye 45', curtesie 46', %%ege 56, yiage 77', 
statur* 83, chivachie 85', grace 88', seryysable 99', table 100'. 

3. JEsaential final E, that is, already existing in Anglosaxon or 
used as a substitute for some other vowel or syllable in Anglosaxon ; 
the Anglosaxon form is given immediately after the word : swoote 
swete r, swete swete 5, 8(mM sonna 7', ende ende 15', her" hira 32, 
tym* tima 35, tale talu 36, inne innan 41', trouth' treow^e 46, werre 
werre 47', ferre feorra 48', mayde mseden 69', son' sunu 79, hop" 
hopa 88, mede, medu 89*, goun" (old Mesic gone) 93, nightmgale 
nihtegale 98'. In here =s their 32, the e seems to Imve been scarcely 
ever pronounced. Though hope 88 may have been merely (noop), 
the e may have been sounded (Hoop*e) producing a trissyllabic second 
measure • 

In hop0 to ttondea in his lady grace. 88 

In goune there is no Anglosaxon authority, the e was not required 
and perhaps not pronounced. 

4. Verbal final E, that is a final E which arises from the inflec- 
tions of the verb: they wende 16', to seeke 17', wer" thei 26, 
wolden ryde 27', hadd' I 31', made 38, to arvse 33', I yow devyse 
34', I pace 36', to telle 38, wol I begynne 42 , he lovede 45, it was 
wonne 61', he hadd" the bord bygonne 52*, hadd* he be 56, he 
sayde 70', he wente 78, I gesse 82', syngyng*, flowtyng* 91, wel 
cowd' he sitt', ride 94', eowde mak', endite 96', justn', daunc', 
write 96', he lovede 97. Were 26, hadde 56, were frequently, or 
generally monosyllabic ; portray 96 should be portraye^ but ttie e 
would be elided ; lovede 46, 97 had the first e elided Mde (luvde), 
and similarly frequently. 

5. Oblique final E, that is, e added to form a case or plural of 
substantives : to the roote 2', in every holt* 6, in felaschip* 26, 32, 
atte beste 29', to reste 30', of ech' 39, in hethenesse 49*, for his 
worthinesse 50', in presse 81', of lengthe 83', of strengthe 84', by 
nightertale 97'. 

6. Adjectival final E, that is, an added to form the plural or 
feminine of adjectives, or to make adjectives definite : the yonge 
Sonne 7', his hal/e cours 8, emale fowles 9, feme halwes kouthe' 14, 
whan that they wer" seeke 18, thei alle 26 , weren weyde 28', our" 
34, ful ofte tyme 52, alle naciouns 53, the grete see 59 ; this tike 64, 
lokkescrull' 81, evene lengthe 83, freeehe floures white and reede 90', 
sleeves wyde 93'. Ofte 52 seems here used as an adjective, for manye. 
In owe 34 the e does not seem to have been ever pronounced. 

7. Adverbial final E^ used to form the adverb : oft' 55, evere- 
mor' 67, late 77. 

8. Contracted article, atte beste = at the beste, 29', 56. 

22 



338 B FINAL — XIV TH CENTURY. Cbap. IV. } 4. 

It is thus Been that if we omit the consideration of final e at the 
end of a line, and allow final ^ to be elided before a subsequent 
Towel, we have only 23 cases in the first 100 lines in which the 
fiaal e was essential to the metre. These ere distributed as follows : 

1. 8up0r/h(nu final ^{douhtial) 1 

2. French final E ^.*.*2 

5. Essential final E 3 

4. Verbal final E 6 

. 6. OhUque final E 

6. Adjeotwal final J^*. .-*-*-- 10 

7. Adverbial final E ' - 1 

—28 
Shewing that the verbal and adjectival final E's were the most 
important. When the final E was so seldom required to satisfy 
the ear of a scribe who had ceased to use it in speech, we nrast not 
be surprised if he often treated it as an ornament to be added or 
omitted at pleasure. This seems to have been the case with all i&e 
later manuscripts. 

Kow turning from verse, let us examine the use of the final e in 
prose, as in the Tale of Melibeus. Here we do not find by any 
means so many ^'s, or such regularity in their use. I refer to the 
words by the number of the paragraph containing them, and give 
two or three words together to facilitate reference, italicisLng the 
word under consideration, 
mighty and riehe 1 has the French e^ 
upon a day 1 for daye, 
him to play 1, for ^ phye, 
dores were fast i-shitte 1, pi. part. 
olde foos 1, plural adj. 
here feet, here, &c. 1, as usual. 
no$e 1, ags. nasu. 
rendyng 2 for rendynge, the final e is here constantly omitted, and it 

is not always inserted in verse, 
gan wepe and crie 2, infinitive «, this is generally correctly inserted, 

but the gerund e is often omitted, 
as she dorste 2, verbal e, 

of his wepyng to stynte 2, the gerund e is correct, the oblique e is 
omitted, so again, of here wepytiy to stinte 3 : but, what man 
schulde of his wepynge stynte 4. The oblique e of the dative 
we found most frequently omitted in GFerman, and it is clear 
that after a preposition which shewed the connection sufficiently, 
the inflection could be readily dispensed with. 
Remedy of Love 3 for remedye. We have already noticed in the 
poetry many cases in which y final had been written for ye in 
French words. It is very possible that in these words the use 
of the final e rapidly dropped from speech, and that then the 
words had final long (n). See p. 283. Lwe^ ags. lufti, has 
always retained its «, although the o may have been short (u) 
in the xiY th century ; it is long in Omnin. 



Chap. IV. j 4- ^ FINAL — XIVTH CBMTX7EY. 389 

of bir ehilde 3, oblique #, but 4fhild$ is oonstfintly found with ^ er^n 

when not oblique. 
l^Jille 3, this seems a supeifluons e^ ags. fyU pUmtuio. 
diligmue amydbU 3, bave the French tenaination. 
bir houshontU 3, ags. busbonda, is regular, 
in this wiu 3, ags. wise. 
youre self 3, usual form, but e not pxonoumced. 
fonothe S, adv. #, or ehe/or iotha^ oblique #. 
to a 49y8 man 3, ags. wis, distinot from the former wm. The 

oblique 4 is here omitted, 
such iorwe 3. Orrmiu has »errgh$^ but there is no « in ags. waxgj 
sorb, which should only form «onr, from ^lonirA »= (sorkiph), 
compare 8orwful 4. 
ye ne oughte nought 3, past tense, 
youre silf destroye 3, infinitive e^ 
The trwtf man 3, definite adjective, con^>are the indefinite a wyi 

man above, 
his ownep&raone 3, (mme feminine e, Budpenone French e, 
amwerde anoon and tayde 4, past tenses. 

And whan thou hast for-gon i^jfrend^ do diligence to geU another 
frende, and this is more wisedom than to wepe for ihj frmd, 
which thou hast lorn, for therein is no hooie 4. The spemng of 
/rend is very careless, the first time it is right, the two following 
times it is reversed, frende /rend for frend firende. To gete, to 
ioepe are gerunds. Wisedom is an error for wUdom. Boote, old 
norse hyti. 
out of youre hert . . . glad in herte 4, ags. heoite, hence t^ first 
spelling is incorrect. Orrmin has heorrte, herrte ; hert, would be 
a stag. It is singular that heart, hart are now distinguidied by 
an 0, but the e is put in the wrong part of the word. In<}eimaB 
her% is a contracted form, and h&me is occasi(mally used in poetry, 
o.h.g. herza, goth. bairto (ner'too). 

It is not necessary to continue this examination. Suffideiit has 
been adduced to shew that the system of final e is the same in prose 
as in verse, so that it has not been invented by the poet or his scribe 
to patch up a line where necessary. If an editor of Chaucer would 
carefully examine all the final «'s, restoring all those grammatici^y 
necessaiy, and ruthlessly omitting, or at least typographically in- 
dicating, all those which neither grammar nor derivation allow, 
when diey were not necessary for the metre or rhyme, and liien 
submit the others to a eareM consideration, he would do the study 
of English great service. The elaborate researches of Prof. Child, 
described in the next section, have smoothed the way for such an 
edition, and in Chapter VII I have endeavoured to carry oirt this 
suggestion for the Prologue to the Canterbury Tales, in a method 
there explained, and in an orthography which the present examina- 
tion has suggested. The careful examination of every verse thus 
rendered necessary has resulted in convincing me that Chaucer and 
Goethe used the final e in precisely the same way, with the solitaiy 
exception of the consistent elision of e before a vowel and silent K 



340 E FINAL — XIVTH CENTURY. Chaf. IV. § 4. 

This conclusion is in karmony with the historical position of 
Chaucer. He was not the first or the only writer of smooth verses 
in English. Omnin's are as regular as any written at the present 
day, and he treated his final e in precisely the same manner as 
Chaucer, making the same Visions. We shall find the same prin- 
ciple marked in the other versifiers of the xm th century. Gbwer, 
Chaucer's contemporary, carries out the use of the final e even to a 
greater extent than Chaucer. As Gower wrote also in French, this 
greater regularity may he attributed to French influence, hut we 
must remember that the French fiual $ at that time must have been 
regularly and distinctly pronounced in common conversation as well 
as in verse, or it would not have formed a part of Mdgret's phonetic 
prose in the middle of the xvi th century. 

Although Chaucer, by the mere force of his genius, became the 
apparent founder of our English poetry, — ^few ever thinking of the 
equally smooth but insufferably tedious Gower, — ^he was in fact the 
last, not the first of a period. The wave of civil war passed over 
the country after his death, and when poetry again rose under 
Spenser, the language was altered in idiom and in sound, and 
Chaucer could only be * translated,* * not imitated. A new versi- 
fication suited to the new form of language rose to majesty in 
Spenser, Shak^ere, Milton. Hence we must not look upon 
Chaucer as an innovator, and the justification of his final must 
not be sought for in an imitation of the French, but in the custom 
of all the versifiers which preceded and accompanied him. 

Acting upon this feeling I have examined what would be the 
result of this theory upon the pronunciation of Chaucer's lines, and 
the mode in which I have printed the Prologue to the Canterbury 
Tales in Chap. Vli, having given great facilities for performing 
the calculation, I have drawn up the following table. It must be 
remembered that the text in Chap. YII does not precisely accord 
with any manuscript, a few simple alterations having b(3en made where 
the metre seemed to require it, but the general results will not be 
at all affected by these changes. The enumeration is by no means 
easy to make, as different opinions may be entertained of the cate- 
gories under which elisions or retentions should be classed, and it 
is not possible to dieck it without taking far more trouble than the 
results deserve. In the present case the enumeration has been made 
twice, at considerable intervals, and the text was corrected between 
the two enimierations. The results differed, but not in any way to 
affect the conclusions to be drawn ^m them. The second series of 
numbers are here given because they refer to the text as it stands, but 
I would by no means guarantee their absolute correctness, although 
they were obtained with care. 

1 Dryden'i and Pope's 'traiiBlatioiifi' BoitofMy blesse thee; thou are trans- 
of Chancer, remind one irresiBtibly of lated.'* — ^Mid. N. Dream, act 3, sc. 1, 
Qoince's exclamation: "Blesse uiee speech 41. 



Chap. IV. j 4. B FINAL — XIV TH CENTURY. 341 

Final E waa pronounced — Times. 

Before a vowel, doubtful : th'old^ Esculapius 429 * - 1 
Before a consonant ---.--. 238 
At the end of a line, that is. it is consonant with strictly 
preserving the grammatical inflection, and the essen- 
tial flnal E, wi& the rhyme, and with the cases last 
numbered, to suppose that it was pronounced in this 
position --..----- 420 

Final ES was pronounced— 

In the middle of a line ---.-- 87 
At the end of a line *------ 87 

Final E was elided — 

Before a following vowel, always, with only one doubt- 
ful exception, v. 429 316 

Before he 92, his 22, him 13, AiV' %, her* A, hem 1, hadde 
7, have 1, how 1, with one doubtful exception before 
he : that on his schyne a mormal hadde he 388, and 
none for the other words, except hadde^ how, have, 
which have not been noted, total - - ^ . 147 
Final ES was treated as simple S--- 

In the middle of a line ------ 18 

Final E was regularly elided — 

In hadd' (with 12 exceptions: v. 253, 286, 310, 373, 
379, 386, 447, 464, 654, 677, 700, 760, as num- 
bered in Chap. YII, where the nimibers sometimes 

differ by 2 from Wright's) 18 

In MY = her, without exception . - - - 25 
Atfr* = their, without exception - - - - 12 
iper^ = were, one exception noted: woo was his cook, 

but if his sauce were 351 ----- 14 
our' = our, without exception - - - - 19 
yotff'ss your, without exception - - - - 5 
Final E was arbitrarily elidedr^ 

as in modem German poetry, for the sake adding force to 
the expression, for the metre or for the rhyme, either 
at the end of a line or before a consonant—- 

when the mark of the oblique case - - . 37 
when the mark of verbal inflexion - - - 17 
when essential, or representing a final vowel in an 
anterior stage of tlie language -' - - 13 
Final E was arbitrarily added — 

for the sake of rhyme or metre, in no case noted* 
These enumerations enable us to lay down the following rules for 
the pronunciation of final E, which would have to be verified by a 
wider field of research, and as they agree essentially with the 
results of Prof. Child's more elaborate examination, — see the next 
section, arts. 74 to 92, — ^they probably represent the practice of 
the court dialect in the xiv th century as nearly as we can hope to 
attain. There is reason to suppose that the e final had been long 
much neglected in the Korthem dialect. 



342 PROF. CHILD OK CHAXTCBR AKB OOWER. Chap. IV. ) 6. 

Final nnacoented e, when essential or inflectional was re- 
gulai'ly pronounced, except in the following cases : 

1. It was regularly elided before a following vowel. 

2. It was regularly elided before a following he^ his, him, 

hir^, her, hem, and occasionally before hadde, have, 
how, to which Prof. Child adds hath and her = here. 

3. In the following words, e though generally written was 

never sounded^ Air* = her, Aw^ = their, our* = our, 
your* = your. 

4. Final e was frequ^itly not sounded in hadd, um^, tim\ 

mor^ 

5. Occasionally, but rarely in comparison to the other 

cases of elisicm, essential or inflectional final e was 
elided to render the expression terser, or to assist the 
metre or rhyme, precisely as in. modem German 
poetry, but not so frequently as in German^ The 
oblique e and essential e were moat frequently dropped, 
as IB also the case in German ; the e of verbal inflec- 
tion was seldom omitted. 
By the elision of final e is meant its absolute suppression 
as in German, Gh^ek, and French, not its rapid or slurred 
utterance as in Italian and Spanish. But there may be 
many cases of the fifth exception in which the elision may be 
saved by introducing a trissyllabic measure, without material 
harshness, and it must remain an undecided question whether 
Chaucer would or would not have elided the vowel in such 
cases. Judging from the practice in German, the elision 
seems most probable. For the efiect of the action of these 
rules in declaiming Chaucer and Gower, reference must be 
made to the examples in Chap. YII. 



§ 5. Professor F. J. ChikTi Obiervatiofia on the Language 
of Chaucer and Gower. 

In the Memoirs of the American Academy, New Series, 
Vol. viii, pp. 445-502, 3 June 1862, and Vol. ix. pp. 266- 
314, 9 January 1866 (subsequently revised so that it ma^ 
be considered as dating from Nov. 1867), Professor Francis 
James Child, of Harvaixl College, Cambridge, Massachusetts, 
XJ.S., has given the results of an elaborate and searching 
examination into the language of Chaucer's Canterbury 
Tales as exhibited in Wnght's edition of the Harl. MS. 
7334, and Gower's Confessio Amantis as edited, from no one 
manuscript in particular, and with an arbitrary system of 



Chap. IY. § (Xl PBOF. CHILD ON OHAUOSE AND GOWEIU 348 

speUing justified by no single manuscript, by Dr. Beinbold 
Pauli.^ As a large portion of these investigations tend to- 
wards the discoTcry of the number of syUables in words, by 
determining when the final e was or was not pronounced, or 
should or should not be written^ the present work would be 
incomplete without a full accoimt of them, more especially 
as the memoirs themselves are not readily accessible.' 

Art. 1. Nouns which in Anglosaxon end ia a vowel terminate ii^ 
Chaucer and Gower umformly in e.' 
2.** First declension of Anglosaxon nouns. Neuters. (1. 1. Bask.)' 
Ex. CAauMT— eere, yhe, ye. Chwer — ere, eye, eie. 

^ Sapri^ p. 256) note 1. qnentiy omitted altogether. The words 

' In the Memoir on Gower, for of the aathor haye generaUy been re- 
{i 20, 21, 22, 2a, 24, 25, 26, 27, 28^ tained. This re-arrangement is made 
as printed, read 21, 22, 23, 24, 25, 26, with the kind permission of Prof. Child. 
28, 29, 80 re8i>ectiTely, as pointed oat * In Prof. Child's papers e means 
in the corrections to that paper. The an e pronoimced, ^ an « elided, S and # 
corrected nnmbers only are' nsed here, written and not elided bnt not forming 
The memoirs haye been slightly a sjdlable in the editions nsed, [e] an # 
abridged, chiefly by omission, and added by himself (e) an e which occurs 
amalgamated. The long lists of words in Wright's edition, bat which he con- 
ap^ded withoat references to certain siders snonld be omitted. The grare 
articles, are giyen at length in a com- accent (*) marks the accented sylkble. 
mon index at the end, for convenience * The asterisk appended to the nnm- 
of eaanal eonsoltation. When they do ber of an article shews that the ftill 
not appear in this index references are r^srenees and explanations of the ex- 
generally appended, bat the whole of emplifloatiye words are in fVfoiD. the 
the references are not always giyen, iDsXtahleof Forma of Words mC^iaucir 
and those to Paoli's Gower are fre- and Oower. 

' The following extract from B. Thorpe's Translation of E. Rask's Grammar of 
the Anfirlo-Saxon Tongne, C(^>enhagen, 1830, p. 28, will explain these references. 
" 49. The following tables will serre as a synopsis of all the regalar declensions : 
Thb Simplb Osdbb, ob Ist DacLBifsioir. 
1. Niui. 2. Ma»e. 3. Fern. 



Ding* liom. 
Ace 

Abl. Sc Dat. 
Gen. 


• 
. Declensi 


-e 

-e 

-an 

-an 


-a 

-an 

-an 

. -an 


-a 

-an 
-an 
-an 


nnral Nob. and A 
AbL and D. 
G. 

2nd 


Thb Co] 
on. 


-an 

-am 

-ena 

UPLEX ( 

* 
1 

>1 


DnPBB. 

3d. Declension. 


l.N0Ht. 

Sing. Kom. „ 
Ace. „ 
Abl. & Dai -e 
Gen, -es 

Ploral N. k A. „ 
Abl. & D. -am 
Gen. -a 


2.Mato, 

>, W 
-e 
-es 

-as 

-am 
-a 


3. JSnm, 

ft 
-e 

-9 

-a 

-nm 
-a 


.NmU 2. Mate. S. Fm. 
(e) -a -a 
(e) -a -e 
-e -a -e 
-ea -a -e 

-a -a -a 
-om -am -am 
-a (ena) -a (ena) -eaa." 



344 PROF. CHILD ON CHAUCER Ain> GOWER. Chap. IY. { 5. 

3 * Masculines. (I. 2.) 

Ex. Chauar—upey aaM, balke, bane, wone, wright*. ^oirer— «pe, aaM, here, 

bere, bi-leere, bowe, difte, crede, cronkey be-lere, bonde-man, bowe, crede, cappe, 

oappe, drope, dwale, fane, feere, foode, drope, dwale, fere, flete, fode, role, 

galle, grame, hare, harre, hawe, biwe, pile, gere, grame, bare, herre, bewe, 

nouflboiide botuebonde, bope, bunte, nope wan-bope, base- (boiiBe-)bonde, 

byne, knave, knotte, kyie, lappe, leere, knape knaye, lappe, like, lippe, make, 

Iqipe, make, mawe, moone, mouthe, mone, name, necke, onde, oxe, pese, 

name, nekke, oxe, poke, po^ pride, pope, pricke, pride, see, ahrewe, anake, 

prikke, reeye, scbrewe, spearwe, stake, BjMirke, spore, stake, steede, store, stenv, 

steede, steere, sterre, stikke, tene, tyme, itikke, swere, tene, thombe, time, wane, 

wele, welle, wete, wille, wreccbe. — wele, welle, wille, wone, wreocbe. — 

Uosme, gere, scboppe, stele, webbe, cope, bunte, like, wan[e], wrenne. 

4.* Feminines. (I. 3.) 

Ex. CftoiMer— almeese, arwe, belle, wise— 4>ircb*,tapetere. O o tt e r a lmewe, 

berye, cappe, cheeke. cbircbe, cloote, arwe, belle, blase, cbeke, cbircbe, crowe, 

oote, crowe, deepe, dowfe, erthe, flje, oromme, deepe, erthe, barpe, berte, 

flose, barpe, beepe, beire, berte, hose, bitte, kerse, lilie, longe, masse, mite, 

owre, larke, lilie, mare, mane, mjte, molde, nettle, nigbtingale, nonne, oole, 

nigbtyngale, oole, panne, pipe, pirie, panne, pipe, ressbe rissbe reissbe, se 

pisse-myre^ pose, rake, rose, scbeete, see, sbete, side, sive, sonne, swalwe^ 

aohere, scbire, sebyne, side, snare, Sonne, throte, tonne, ton^e, waccbe, weke, 

swalwe, targe, tnrote, ton^ tonne, wiccbe - craft, wenSie, wise. — lappe- 

trappe, wake, wencbe, wiccbe-craft, winke, more, sale. 

5.* In the following the final e has been absorbed by yoTW, In the 
following the final e seems to have been transposed from after / (as is 
often the case after rV. Oower ' — The following may or may not be 
correctly written. The combination of a liquid with e is unstable, 
the Towel easily slipping from one side to the other of the consonant. 

Ex. Cftoueer— play, lady, sty : ber- fitbnl, ladel, wesil, wbistel. — nerew. — 
berw berberwb berbergb, widow widw : ^otrer^-throstel, nedder, adder. 

6.* Exceptions to art. 3, 4. 

Ex. Chancer— pith, beech, kers, stot G'oiffer— laTerock, to(e), roo. 

7.* Second declension of An^osaxon nouns. Masculines. (II. 2.) 

Ex. Chawer — awe, bale, cheese, mayd, mete, see. Omotr — ^bale, breche, 

ende, hate, hefifge, herde, byre, ire, brimme, ohele, cbese, ende, hate, herde, 

leche, lye, mede, myre, pilwe - beer, ire, leche, love -dronke, mede, mele, 

reye, skathe, tete, whete— come, mere- mete, shipe, slitte, stede, tete, tie, whete. 

8.* Exceptions to art. 7. Termination -ichipe. The length of the 
words compounded with this termination may perhaps account for 
the final e being soon dropped. Termination -ere in Saxon nouns 
signifying for the most part an agent. It h quite as likely as not 
that in 544, 3167,' the final e of meUere was pronounced. Oower — 
Such representatiyes as occur of the Saxon noun in -ere, denoting 
an agent, seem to want the final Towel. Nouns of this kind were 
by no means as common in the old language as in the modem. I 
have noticed but three fair cases in Gower. There are other in- 

^ Paraffraphs introduced by the word * The simple nnmbers refer to ^e 
Oower feUowed by (— ), are taken from lines in Wright's edition, as throngh- 
the memoir on Gower, the other being out this chapter, suprik p. 256. 
from the memoir on Chaucer, but 
occasionally paragraphs are headed 



monallY paragraphs are h 
ttiMT— lor greater distinctness. 



Chap. IV. } 5. PBOF. CHILD ON CHAUCEB AND GOWER, 346 

stances without the final e, but in these oases the succeeding word 
begins with a vowel, and it is supposable that the e may have been 
elided. It is doubtM whether tiiese words should be called excep- 
tions to art. 7 ; for, in the first place, the metre does not settle the 
question of their form, inasmuch as cl^ppSre, for instance, would 
suit the verse (which hath no clapper for to chime, ii 13 *), as well 
as cUpper ; and secondly, for few, if for any of them, can we show 
a form in -ere in the Saxon dictionary. 

Ex. Chaucer — felawschip^, fiiend- founder, Both(e)-iaier, Bpekerin, fiBher 
8chip£, lordschipS, worschip : carter, in, fdrtherer of^ maker o^ tedier of^ 
hopper, loTer^, mellere miller, ontry- keper nnarraied. 
dere, deper, wonger. ^wrei^-clapper, 

9.* Third declension of Anglosaxon nouns. K'euters. (HI. 1.) 

Ex. Chaucer — ale, ancle, mele, spere, riche worlde-riohe, skille skill, spere, 
WOTre, wyte — stree. Gower — ale, werre, wile, wite— kne, stre, tre — ^be- 
chinne, inne, -riche heyen-riche kinges- yete, winge. 

10.* Masculines. (III. 2.) 

Ex. Chaucer— ]ake, Gower — eone 8on8, wode. 

II* Feminines. (III. 3.) 

"Ex. Chaucer — breede, care, elde, dore, highte, mayne. Gower — answere, 

fiire, gappe, hele, hete, lawe, nave, brede, care, dore, elde. fare, heighte, 

nose, sake, sawe, schame, schonde, hele, hete, lawe, leese, lode, Iotc, nase, 

scbadwe schawe, scole, sowe, spade, nntte-tre, sake, sawe, sohame, shawe, 

tale, talS yit, trouthe tronth^ ware — scole, spade, tale, tronthe. 

12.* Exceptions. It will be noted that the nouns sone and love 
have the final e regularly in GFower, contrary to the apparent rule 
in. Chaucer. The same is true of the importcmt word titne, art. 3. 

Ex. Chaucer — sone sonS, woode woodS ; answar, love 1ot£. 

13. Many nouns which in Anglosaxon end in a consonant have in 
Chaucer and Gower the termination e, derived fix)m an oblique case 
the old inflection. A few familiar parallel formations in other 
modem languages may be mentioned* LaU radix, Itai, radice; 
animal, animale ; cupido, cupidine ; imago, immagine ; nix, neve ; 
latro, ladrone ; honor, onore ; libertas, Hbertate ; voluptas, volut- 
tate. So in colloquial Romaic, as compared with Greek : — Qreek^ 
Xafiira^, Romaky Xafji/rrdSa; yfivy 'xfiva; wJf, vvKra. 

Two forms not unfrequently occur ; one with, and the other with- 
out the vowel. By the dropping of this vowel in later English, the 
primitive form is restord. Tliough this secondary, transitional form 
in e is found in Layamon and the Ormulum (quite frequently with 
Feminines of the second Saxon declension), yet it is by no means so 
common as in Chaucer. 

As it is possible that some may think the forms in e of the Mas- 
culine and Keuter nouns to be oblique cases of a nominative, which 
(if it occured) would be found to end in a consonant, the gramma- 
tical relations of these words are always indicated, but tMs (pro- 
bably superfluous) trouble has not been taken with tiie Feminines.' 

^ This mode of citation refers to * On examining Prof. Child's lists in 
Panli's edition of Qower, toL ii, p. 18. \\ 14, 16, 17| 30, I have obtained the 



346 



PROF. CHILD OK CHArCBB AKD GOWBR. Chap. IV. { 6. 



14* MaacuHnes and neaters of the second and third declensions 
(11. 1, 2y lU. 1, 2). ^ot&ar-^Most, if not all, of the following, and 
jnany other nonns of the same declensions, are fonnd in the primi- 
tire form without the voweL In many instances the tenmnal o 
might be explained as the Saxon dative inflection, bnt it will be 
found on inspection that about half of the nouns in the list occur in 
the BominatiTe or accusatiye case. 



Ex. Chaucer — bedde, berne, bisie- 
mare, bladde, IxMire, botme, brede, 
brembre, brtmde, carte, childe, come, 
croppe. culire, dale, donge^ drynke^ 



fidshede maTihedo maydenbede wom- 
manhede, heede, benM, beyene, bewe, 
bole, -bolme, bouse, kynne, leeke, 
licbe, londe, loode, loone, lyste, lyre, 
morne morwe, mynde, othe, scbippe, 
sitbe, Bleepe, smoke, sore, sotbe, speUe, 
Btalle, style, swyne, temple, totbe, 
towne, wawe, wedde, werke, weye, 
wbelpe, wbippe, wife, wronge, yere— - 
arme, bore, cole, derke, dom^, fissbe, 
keepe, mele, scbepe, sigbbe, swoune, 
teere, walle, wyne wyn. Chwer — 
bedde, bede, berae, benie, bore, bore, 
borwe, bote, bo^e, browe, carte, 



ebOde, clerke, eole, cope, dale, dawo^ 
dele, detbe, dome, dnnke, &e. fere, fire, 
flesshe, flete, folde, fiolka, tote, nte, 
ffolde, gniTO, Rovnde^ -bede -aods 
nisbeda ffodbe£ battibede kingbeda 
knigbtbode knigbtlibede ladybeda Uk- 
lybede maidenbede manbede sosterbede 
inf(e)bode wmaaabede, bewe, borne— 
tbe adyerb sbould be spelt Ami, WBk 
bam, and not kom$; at horn is also &» 
coiTect form, an. at bam-^orse, boise, 
kinge, kinne, leefe, licbe,^ limme,^ life 
Htc, lode, londe, lope, middle, minde, 
montbe, mordre, monre, mote, monthe, 
mule, rede, rore, soome, sete, sbape, 
sbippe, shotte, sbrifte, sitbe, slepe, 
smoke, sore, sotbe, stronde, temme, 
tbewe, tbinge, towne, wawe, weada, 
weie, weigbte, wbippe, wisdome, wive, 
worde, wortbe, wronge, yere — ebo, fo 
— ^bie, kepe, lette, leye, swoune, were. 



15.* The following merely drop a final n (compare Lat. and Ital. 
acumen, acume ; certamen, certame ; yimen^ yime). 

Ex. Chaucer & Ootoer — ere, game, mayde. 

16.* Feminines of the second declension. (11. 3.) These nouns 
have in Anglosaxon all the oblique cases of the singidar in e. 

Ex. Chaucer — beere, bene, boone, balle ballS, beede, belle bell^, belpe 
boote, brigge, bryde, bnrtbe, bynne, belp(<, besto best, byre bnyre, keye, 
dede, donne, diede, fllle, gifte, gleede, kynde man-kynde, lengtbe, lere, line, 



following results for Cbaucer — 1 bare 
not examined tbe instances in Gower 
because of the great uncertainty of 
Faults text. If we reject tbose nouns 
wbicb are only found m oblique cases, 
those whose mud e is elided before a 
yowel, those which occur at tbe end of 
a line where the final e of tbe rhyniag 
word may baye been omitted for the 
rhyme, tnose in which '•re may baye 
been written for -^, those in wnich e 
may baye been a connecting yowel in 
eompounds as in licb#wake 2000, and 
those in which the authority of Omnin 
shews that a final e had Icmg been as- 
sumed, the lonff list of masculines 
reduces to the following : childe 6339, 
14980, foote 11489, bewe 1366, Ipte 
1864 (which should apparently bt 



lyetet as in 1861), morwe 14710, sotbe 
12690 (probably an adyerb) swyne 
16972, wawe 4888, wife 6648. In the 
feminines we should also omit tbe ac- 
cusatiye which had an « in Anglosaxon. 
They reduce to: bryde 9764, gifte 
9167, lengthe, 17302, mertiie 768 
(plural P), achipne 2002 (probably an 
error for sehtpm), spanne 166 (the ao- 
cusatiye of dimension P), tyle 7687 
probably accusatiyeV youthe 2381 and 
frequently. The aqectiyes reduce in 
the same way to : bare 8766 (feminine f), 
blewe 666, eebe 1184 (ech would only 

g'ye a monosyllabic first meature>i 
nge 1676, lowde 10682 (feminine F), 
merye 208 (Bosworth gives an ags. 
formmirige), sborto 6206 (not in Harl, 
7384), tame 2188^ wete 2340. 



Okaf. IV. i 6. PROr. CHILD ON CHACCEB AS1> GOWSR. 



847 



loode-iftene, lore, iTdne, Iff ere, meede^ 
meUe xnyllS, merke, merthe, mjle^ 
■eede, -nesBe bwynease boldenesse 
br^htneue deimesse cune^eBae dnm- 
KeuKB00e fairneeee falsnesse goodneeae 
liardyneflse hethenesse heTyBMse hoU- 
nesse homlynesae lewednease newefan- 
g;£hi6Me scharafSutnesee seeknesse siker- 
■ene stedfiutaeaae wannnesee wikked- 
aeaw witneese worthineese wrecched- 
nesM ydelnesBe^ besynes denness^ 
goodneft lewednes lushes wortbines 
iroodnes, ore, plyte, pyne, qnerne, rtme, 
loode, adupne, aleeye, doufthe, aonde, 
sorwe, foiue tonl^S, spanne, ^peehe, 
ipech^ stonnde, streete, Btrenghte, 
Bynne, throwe, tyde, tyle, npriste, 
wede, werte, while, wolle, wtnnbe, 
womide, yerde, yoQtbe — asp, booke, 
drowhthe, lynde, rewtbe, vcberte, 
rieigfite, sterene, wilw, wreehe. 6fouf€r 



boDde, bone, bote, brigge, cheate, dede, 
d^e, effge, fille felle ftille, filthe. for- 
gifte, elMe, gloTe, halle, halfe nalye, 
bade, bde, belle, belpe belp, bette, 
binde, bire, keie, kiaae, kiste, kitbe, 
lengtbe, lore, liade, lore,Tiiarche,mede, 
mertbe mirtbe, mile, nede, -nesse be- 
sinesse buxomnesse baUnene idelneaae 
rigbtwianeBse sik(e)ne88e sikerneaBe we- 
rinease wildemease witneaae, ore, qnene 
quen^, reste, rewe rowe, rinde, rode, 
roode, aalye, score, sbelle, sigbte siime, 
slere, slontiie, sonde, sorwe, sonle, 
^anne, specbe, stempne, stoonde, 
strete, strengtbe, tbefte, throwe, tiltbe, 
tide, warde, wede, wene, wbile, wombe, 
wonnde, wratbe, wrecbe, wnlle woUe, 
yerde, yifteyefke, yon^e^ — ariat, fiste, 
iigbt, irbt, ladder, lefte, liyer, nedel, 
rontbe, skerte, staogbt, sleigbte, steltbe, 
weltbe, wierd, wente. 



— Mnke, bene, bertbe birthe, bliase, 

17.* Exceptions to art 16. Oow^ — Hand^ mighty nighty wight ^ 
aie exceptional in Anglosaxcm, having the accvsative singolar like 
the ncnninatiye : so worlds more commonly : hok (constantiy mi&- 
^elt hoke) i 2, 5 : ii 58 : iii 65, 133, etc. ; hwrgh, ii 232 ; iii 
292 ; fwghy ii 245, all feminines, are also irregiQar in Saxon, and 
have the accusative singular like the nominative. ClMuar — ^Nouns 
derived from Saxon feminine nouns in -ung, -ing, or formed in imi* 
tation of such, terminate in Layamon mostly in -inge, rarely in -ing. 
In the Ormulxmi the termination is almost invariably -inng, hut one 
or two have'the nominative, and three or four an accusative in -innge. 
The more usual ending in CSiauoer is certainly -yng. The termina- 
tion -ynge occurs frequentiy at the end of a verse, and in most cases 
rhymed with an infinitive. Qower — ^Nouns derived firom Saxon 
feminines in -ung, -ing, or formed in imitation of such, generally 
have in Gower tiie termination -bige, less frequentiy -ing : in the 
latter case the accent is sometimes thrown back. 

Ex. CSIauMT^aldir, ax, benob, bliss werk^nge all rhymed with tnfinitiyea 



blisse, box, chest, ours, fum, ftst f(B8t, 
fltt, flight, floor, hand bond, beeth, 
hen, mark, n)ighi milk, niebt, ok ook, 
qneen, sight, rest, soken, tow, wight, 
world, nonns in -yng ixyng begynnvng 
ddthing com^g cdnnyng or^ellyng 
flght;^ h^tngyng bUrpynjer bilntyng 
Idldng long^g making onr^ng r^n- 
nyng sm^lyng tdching w^dryng w^ 
yngwdnyng writ^^wynn^g, lem^ge 
tume^ge, yanysscb^ge walk^ge, ca- 
rol^ge connjiige dawen^ge enren- 
ym jnge fd^ge lyr^nge monren^nge 
offir^nge rejois^nge sem^ge taryinge 



synge brynge stynge spryng^e f^d with 
the exception m fd^nge 16779 all ob- 
liqnel Oowtr — axd Mnob bridfi flight 
flor(e} ben bond les might milk night 
p]it§ 8ped(^ tow wis^ht world, nouns 
m -inge axlnge bakbilinge caroUnge 
ehiMlnge cominee oompleignlnffe 
gmochinge knonlediinee leslnge lik- 
inre lokinfe mishandunge spekinge 
ti£nge wemilllnge wepin^ writinge, 
begfinning knonlechlng tecnlng, bidit- 
faig liking waning writing(e) ; excus- 
ing of, hunting as, sheding of are ap- 
parently cases of elision — steyen. 

18.* The following nouns, of etymons more or less uncertain, hut 
mostiy of undoubted Gbthio origin, are found in Chaucer and Oower 
terminating in e. 



348 



PEOF. CHILD ON CHAUCER AND GOWER. Chap. IV. § 6. 



Ex. Chaucer— hnnke, cake, chaffare, 
doke, dowde, cope, damre, deynt^ 
dogge^ drake, fel&we lelaw (Akwi 
fhuiwi, gable, jade, kname, know-leche, 
kjn-rede, marie, roteroote, senile, slynge, 
snowte, stalke, tare, wynddwe wyn- 



d^w[e ?]. G'oirir— babe, boihe, brmke, 
bulle bolle, cake, cbaffare, clowde, 
creple, deinte, fel^we fel&w f^law 
f^low^ fiinke, geese gnesse, mone^ 
packe, rote, senile, snowte, tacle, were, 
wicke, ^ndow. 



19.* The unaccented final e of nouns of French origin is sounded 
in Chaucer as it is in French verse. Exceptions, however, are fire- 
quent. Gower — ^Exceptions are by no means so common as in 
(Wright's text of) the Canterbury Tales ; a few exceptions, after 
the sounds r aud s, are cited under arts. 84, 91/. So in adjectiveB. 
Chaucer — ^It is scarcely necessary to mention that an internal e in 
French words is also pronounced, as, comaundement 2871, jug- 
gement 780, etc. 



Ex. C^otiMT— Arclte Arclt^, aimte, 
best, biUe, cause, centre, cbambre 
chambrS, concbe, cynamome, dame 
madame ma-damS, donte, Dy&ne 
Dykn^ eese, e^le, entente entent, 
e^rience expenens, fiice faafl, feste 
fest, force forc2 fors, fortune, grace 
grac^ gras 15242!, hannche, berbe, 
neriti^ bomidde, hoste oste bost est, 
joye, juge jngge, male, manere manor, 
medecine, nece, persdne persdn, peyre, 
pbisik, place piac£, plante, pompe, 
regno rcgnS, remembraunce, requeste 
request, Bome Rom^, sauce, sege, ser- 
rise, signe, spouse, tente, tnimpe. Adj. 



chaste, ezcellente, nice, pore, ricbe, 
Bolempne. Oower — abb^se, adyentnre, 
avarice, baptisme, beste, borde, bounde 
bonde, bowele, chore, Constance, de- 
faulte, deseite, egle, entente,' enrie, 
fiBste, fortune, grace,^ haste, homicide, 
houre, joie, justice, madame, magique, 
manere, mappemounde, numiage, ma- 
tere, medicme, menreille, messa^ 
mewe, mule, multitude, nature, navie, 
offrende, oile, pacience, passlige, per- 
sdne, pestilence, phisique, place, pompe, 
Rome, spume, rice, vireine, ymage. 
Adj. chaste, double, hugne, inrislUe, 
nice=:foolii^ riche, solempne. 



20. The accented final e of French nouns (in modem English, y) 
is of course preserved in Chaucer. 

Ex. Chaucer — adversity, bounty, on 6 is due to the editors, and is not in 
cherts, clarr^, contr6, liberty, perr^, the MS.] 
plenty piy^yt^) renom^. [This accent 

21. The Genitive case, Singular, ends in -es. 



Ex. Chaucer — schires 15, dierles 
7788, lordes 47, Cristes 480, pigges 
702, reeyes 601, modres metes kyng^ 
5433-5. Oower— losen iii 85, mannes 



iii 86, Cpoddes iii 88, worldes iii 90, 
nightes iii 96, daies iii 111, bnlles iii 
119, kinges iii 146, wiyes iii 73. 



The following have, at least sometimes, no termination : 



Ex. Dec. I. Chaueer—\io\j chirche 
good 3981, holy chirche blood 3982, 
holy chirches feith 11445; his lady 
grace 88, oure lady yeyl 697, his ladys 
grace 9892; the sonne upriste 1053, 
the Sonne stremes 16240, myn herte 
blood 10221, a widow sone 14918. 
Otnoer — ^the chirche kei i 10, mone 
light iii 109 r|>erhaps compounds), the 
mones cercle iii 109 ; my lady side i 
160, this lady name ii 157, my ladv 
chere ii 21 3, my lady kith[e] iii 5, 
my ladj good iii 30, ladies loyeiB i 228} 
i<ne8ti84, «8elyei228, .idoughterii 



227, /• mercy ii lis. So, Chaucer— i^et 
9239, 9012, 15670, 8772, 4036, 9389, 
12757, 15423, but &dres 5883, 8738, 
8685, 8747, 13626, 783 P, 10175 P, 
14883 P brothir 3086, 13360P, brothers 
11478, modres 15004, philosophre 
12790, heyen 6763, 10281, 12470, 
16282, 13017. 6W«r— hoise i 40, 
119, heyen ii 187, helle ii 97, sonle 
i 89; fader i 209, faders i 157, 
brother i 199, brothers i 214, mothefr 
i 289, moders ii 354, doughter i 208, 
doughten i 150. 



Chap. TV. § 5. PROF. CHILD ON CHATTCBR AND GOWEK. 349 

So, many proper noims in s, as in Anglosaxon and Modem English 
Chaucer—'E^icoTUB 388, Penens 2066» phemus i 166, Bachus ii 368, Phebns 
Yenus 10586, Melibeus 16882, Phebns lii 260, etc. 
17170, Marquys 8870. Goirer— Poly- 

22. Plural of nouns. Nominative. The Nominative Plural is 
formed for the most part in -es ; occasionally in -us or -is, a dialectic 

variety. Gower * only is frequently added, especially to nouns 

terminating in a liquid or in -t; sometimes when -es is added 
(rightly or wrongly), only -s is pronounced. 

£x. Ohaucer ^ hudleB 900, bodres pens 7168, lazan 246, sellen 248, 

1007, knees 1106, 1877, deerees 17298; achatours 610, pilonis 1009, loTers 

fowles 9, domes 326, cbiknes 382, 1633 — scboos 369, dys 1240; bisschops 

bones 702, fyneres 129; croppes 7, 4673, keverchefii 466, oaytifs 926; 

robes 319, knoobes 636, wyieB 234, reliks 13764, lordyngs lordvng^s 7260, 

knyfes 368, kaytyres 1719, lewes 1498 ; 16726| yeddyng^s 237, precnings 6139 ; 

loues 76, Bonges 96, branncbes 1069 ; seryantte 101, contraciSs 6890, yesti- 

bootes 203, ar^omentes 4632, oma- mentz 2960, marchaiintz 4668, 4691, 

mentee 8134, nonndes 146, swerdes arguments 4648, mannderoentz 6866, 

2028; stremes greyes dropes leeves instmmentz 9687; greybonnd^ 190, 

1497-8, brawnes scbnldres armes stiwardSs 681, bnsbonds 2826. Gower 

2137-8. Chwer — weies, tiranniei, — anngels, cardinals, nations; courts, 

thewes, sonles, hilles, formes, pbilo- points i 149, pointes i 161, elements, 

sdpbres, fires, lores, sterres, droppes, jugements, arg^oments, tiraonts, Sara- 

berbes, leyes, liyes, wiyes, toryes, zios, complexions, masons ; saints, 

bokes, derkes, beinges, tbinges, notes, estat(e)8, crafU^s, climats, berts bertes 

frostes, bestes, flodes, clondes, beyedes i 326, loyers, flatronrs, fetbers ; words 

=3 heads, monthes, monthes. Chaucer i 176, wordes i 161, Grek^s ii J 71, 

— ^pilgiyms 2860, nacionns 63, bar- Grekes ii 166, knes kne&, tres tre^s. 
gayns 284, sesomis 349, sessions 367, 

23. The following have -en, -n, derived from the Saxon plural in 
-an of the 1st Declension : asschen 1304, assen 5867, aissches 12735, 
heen 10518, bees 7275, eyen yen 152, fleen 16949, hosen 458, 
oxen 5867, schoon 15143, schoos 459, ton 16348, toos 16817. 

24. The following have -n, -en, by imitation, being of various 
declensions in Saxon. Gower — The following, which have the 
termination -u in Saxon, have superadded the -en of the 1st Declen- 
sion to a weakened form of the Saxon plural. 

Ex. Chaucer — bretberen 13831, 16317. Ootaer — ^brethren, bretberen, 

14192; dongbteren 11741, doughtres brethern, brethemS, children, [dongbt- 

16316, sistren 1021, snstres 16363, eren sistren, do not occur] dongbteres, 

children 1196, 14908, childer 8031, donghter ii 172 P susters. 
14912, foon 16192, foes 16816, kyn 

25. The following have no termination in the plural, according to 
the rule of the Saxon neuters of the 2nd Declension : deer, folk, 
hors, neet, scheep, swin, thing, yer. (The word good added in 
Chaucer is corrected in Gower). So night 7467, wynter 10357, and 
probably freend 3052, 3053. 

26. The plurals formed by change of vowel are the same in 
Chaucer and Gk)wer as in English : feet, gees, men, teeth. 

27. The following plurals of French words are remarkable : caas 
325, paas 1892, degre 1892, secre 6923 (?), orgon 16337, vessel 
15634, but vessealx vesseals 15680, 15687, richesses and riches. 



350 PBOF. CHILP ON CaAVCBSL AND OOWER. Qkap. IV. § £. 

28. The G^nitiye Plural in Ohauoer and Gower is macli the saaie 
as in Knglish, saving, of course, the use of es instead of s. 

Ex. Cfhaueer — lordeB hestes ^iOl, 101 S4. &0tMr-^ie Grekef lawe, aUt 
lordes doughtres 13488, foxes tailes mennes ipecha, mennes goodei ii 8St« 
16519, bestes dennes 16749, jeintes ont of all other briddea sight i 100, 
lyres 6272, meimes wittes 4622, wym- princes heredes, of the goddet par- 
mens oonnseilei 16742, hit eyghen nght Teinmoe. 

Adjecciybs. 

29.* Adjectiyes wbich end in e in Saxon end in e in Chauoer and 
Gbwer. Several other adjectives might probably be inserted in ^ns 
list, but as they are found in the Canterbury Tales only in the 
'' d^nite form " (see art. 82), they have not been noticed. 

Ex. Cfhaue$r — bUf^e blithi, dene bHthe, a-cale, clene, dere, dertte,drie, 
clenS, dere, derne, drye, elenee, ^mde, fre, grene, kinde nnkinde, mete nnmete, 
grene, heende, kene, kynde, lene, newe, mMs munilde, neisriie* newe, soAe, 
pronde prowd, ripe, scheene, softe, sterne, stilk, swete, thicke, thimw, 
stille, sterne, swete swote, thenne, trewe nntrewe, nn-wjdde, yare