A new English grammar, logical and historical

CO

PE

1101

S8

V.2

C.2

ROBARTS

NEW ENGLISH GRAMMAR

SWEET

VOL. II.

Oxford University Press

London Edinburgh Glasgow Copenhagen

New Tork Toronto Melbourne Cape Town

Bombay Calcutta Madras Shanghai

Humphrey Milford Publisher to the University

A NEW

ENGLISH GRAMMAR

LOGICAL AND HISTORICAL

BY

HENRY SWEET, M.A., Ph.D., LL.D.

MEMBER OF THE MUNICH, BERLIN, AND COPENHAGEN ACADEMY OF SCIENCES
UNIVERSITY READER IN PHONETICS AT OXFORD

Formerly President of the Philological Society

Editor of The Oldest English Texts^ Alfred' s "^ Cura Pastoralis' and'Orcsius'

Author of ^Ah Anglo-Saxon Reader '

"■A Firsf and ''A Second Middle- English Primer'

*A Primer of Spoken English,' "A History of English Sounds'

'A Primer of Phonetics,' 'Shelley's Nature- Poetry,'' etc.

PABT II— SYNTAX.

OXFORD
AT THE CLARENDON PRESS

Pt
■ ^^

Impression of 1924
First edition, i8g8

Printed in England

PREFACE

The first part of this grammar appeared in 1892. The
delay in bringing out the second part is due to a variety of
causes. For some years the whole of my time was given to
my Student's Dictionary of Anglo-Saxon, and in the interval
I have, among other work, made a thorough revision of my
Anglo-Saxon Reader. And after giving so much time to
promoting the study of English in this country, I felt that
I could, with a good conscience, return to those wider
studies in comparative philology to which I feel more and
more drawn.

On the other hand, there have been so many enquiries
after the Syntax that I did not like to delay it any longer.
I have therefore limited its scope by confining myself to
formal syntax (§ 582) and excluding what can be found in
the dictionary, such as the use of prepositions, and so have
been able to give all the more prominence to syntax proper,
especially those branches which have hitherto been neglected,
such as word-order.

It will be found by comparison with other grammars that
my syntax is fairly complete from this point of view. It
must be noted that there is a good deal of syntax in the

VI PREFACE.

introduction to the first part, where, for instance, the analysis
of sentences is fully dealt with.

Note the use of nominal as a common term for nouns
and adjectives.

The mark f is used to indicate literary as opposed to
colloquial. For the use of (;) as a stress-mark see § 1 88 1.

HENRY SWEET.

Oxford,
July II, 1898.

CONTENTS

WORD-ORDER.

Form I

General Pi*inciples 2

Emphasis § 1765. Convenience § 1767. Grammatical
Order § 1770. General Principles of English Order and its
Changes § 1772.

Adjeotives 7

More than One Adjective § 1789.

Genitives 10

Modifying Nouns 10

Pronouns ii

Verbs 12

Verb + Subject § 1806. After Front Words § 1807. Modifier
+ Verb § 1818. Several Verb-modifiers § 1823.
Verbals and Periphrastio Verbs . . . . 16

Adverbs 18

Adverbs + Nominals or Adverbs § 1834. Verb + Adverb
§ 1838. Adverb + Verb § 1846. Front Adverbs § 1854.

Sentence-Words 23

Conjunctions and Dependent Adverbs ... 24

Broken Order 24

Cross-order 25

Front-order 26

Vlil CONTENTS.

PAGK

Group-order 26

Verbal-groups § 1872.

Sentences 27

Elliptical Order 28

SENTENCE-STRESS 28

INTONATION 37

NOUNS.

Qender 42

Number 44

Singular Individual Noun = Collective Noun § 1966. Collective
Nouns with Plural Constructions § 1972. Changes of Mean-
ing in Plural § 1975. Used only in the Plural § 1979.

Cases 48

Common Case § 1985. Genitive § 1996.

ARTICLES 54

Definite Article 55

Referring Back § 2014. Identifying § 2019. Absolutely
Defining § 2026. Class- and Collective § 2028. Unique /
§ 2031. With Abstract Nouns § 2033. With Proper Names
§ 2035. With Absolute Adjectives § 2038.

Indefinite Article 61

Articles Omitted 63

ADJECTIVES.

Absolute , . . 65

Free 66

Comparison 67

Comparison of o/d § 2084.

PRONOUNS.

Personal 71

Indefinite § 2097. Pleonastic § 2099. Gender § 2101.
Prepositional Genitive § 2102.

Possessive 73

W^ith General Adjectives § an 2. Emphatic § 21 16.

CONTENTS. IX

PAGE

Interrogative 76

Kelative yy

NUMERALS . 81

VERBS.

Number 81

Person 83

Tenses 85

have-ioxms § 2165. do-ioxms, § 2169. will w^d shall § 2196.
Definite Tenses § 2203. Present § 2223. Preterite § 2233.
Perfect § 2240. Pluperfect § 2347. Future J 2249. Preterite
Future § 2252. Perfect Future § 2253. Immediate Future
§ 2255.

Moods 107

Subjunctive S 2259. Conditional § 2277. Compulsive {is
to . . .) % 2297. Permissive {may) § 3301.

Voice (Active and Passive) 117

Infinitive and Supine 118

Gerund 120

Participles 122

Present § 3337. Preterite § 2356.

SYNTAX
WORD-ORDER.

Form.

1759. As regards the relative order of two words, we
distinguish between pre-position and post-position. Thus
pre-adjunct or pre-adjective position means that the adjunct-
word precedes its head-word, or that the adjective precedes
its noun. We may call such an adjective a * pre-adjective,'
or, more definitely, a * noun-preceding adjective ' ; so also
we can define the noun as an * adjective-following noun.'

In groups or sentences composed of more than two words
we distinguish front-, mid-, and end- position, the last two
being included under non-initial position. Thus a verb at
the end of a sentence is said to have end-position : such
a verb may be called an * end- verb.' If such a verb were
put at the beginning of the sentence, it would be called
a ' front-shifted ' verb.

Position may be to some extent accidental. Thus the end-
verb order in such a sentence as // rains is merely the result of
the shortness of the sentence, so that it is a case only of what
may be called * negative * end-verb position.

We also have to distinguish between joined and broken
(1860), and between parallel and cross (1865) order. For
tag-order see § 1774.

VOL. II. B

2 SYNTAX. C§ 1760.

1760. We have lastly to distinguish between fixed and
free order. Some languages are freer in their order than
others. Very free order is possible only in inflected lan-
guages. Conversely, absolutely fixed order occurs only in
languages devoid of inflection. Even in one and the same
language some kinds of words may have freer order than
others : this is the case with the English adverbs. Hence
in most languages there is a distinction between normal
(regular) and exceptional order. This distinction is, of
course, most marked in highly inflectional languages.

Even in languages whose order is comparatively fixed there
are many devices for evading the restrictions of the normal order.

General Principles.

1761. The divergencies between the word-orders of
different languages, and the inconsistencies in the word-
order of one and the same language, are the result of the
conflict of various general principles.

1762. From a strictly logical point of view we should
expect connective words always to come between the words
they connect — we should expect prepositions always to pre-
cede the word they govern, relative words as conjunctions
always to have the front position in the sentences they
introduce, and so on. We should further expect subject
+ predicate order. In a less degree, we should expect
post-adjunct order to prevail — we should expect assumptive
adjectives to follow their nouns. But, as a matter of fact,
none of these general principles are carried out universally in
language.

1763. The most frequent deviation from purely logical
principles is the pre-adjunct order adjective + noun. This
order was probably originally emphatic (1765). From
a practical point of view the main distinction between the
pre-adjunct order big black dogs and the post-adjunct order

§1767.] WORD-ORDER: GENERAL PRINCIPLES. 3

*dogs big black is that the former is suspensive — it makes us
expect something to complete the sentence — and hence is
more connective than the post-adjunct order, and binds
adjunct and head-word more closely together. The looser
post-adjunct order is, on the other hand, naturally used in
apposition, even by languages which otherwise prefer pre-
adjunct order.

1764. As negation generally reverses the meaning of its
head-word, it is most convenient practically to let it precede
its head- word, so that the hearer's mind may be fully
prepared for the reversal of meaning. Hence languages
which otherwise have the order verb + adverb may have the
order negation-word -H verb, as in the Old-English ne cume
ge{lQ01).

Emphasis.

1765. The most general way of making a word prominent
is by putting it before the others — if possible, at the begin-
ning of the sentence. Thus in Latin the normal order in such
sentences as * Caesar conquered the Gauls * is to put the verb
at the end {Caesar Gallds devicit)^ but if the sentence were
meant to imply that Caesar conquered the Gauls and not
some other people, the word expressing the logically pro-
minent idea * Gauls' would have front-position {Gallds Caesar
devicit).

1766. But there is another more general principle of
position-emphasis — that of making a word conspicuous by
putting it in any abnormal — that is, unexpected — position.
Thus a word whose normal position is front or mid may be
made emphatic by end-position, as in the Latin sentence
aliud Her habemus nullum *we have no other road,' where
* none ' has emphatic end-position» Emphatic end-position is
suspensive (1763).

Convenience.

1767. It is evident that emphatic order often leads to
inconvenience, as in the last example, where we have the

B 2

4 SYNTAX, [§ 1768.

double inconvenience of the separation of nullum from its
head-word — broken order — and of suspensiveness, the
meaning of the three first words being completely reversed
by nullum,

1768. But a purely logical order may also lead to in-
convenience. Thus, as we have seen (1763), pre-adjunct
order has certain advantages over the more logical post-
adjunct order, especially in negation (1764), while in other
cases post-adjunct order may be more convenient. Indeed,
the best results are often obtained by a concurrent use of
both, as we see in the English order subject-adjunct +
subject -f- verb + verb-adjunct, which is the result of the
striving to avoid the suspensive end-verb order.

1769. In the Latin sentence last quoted broken order and
suspensiveness work together. But in some cases broken
order is a means of avoiding suspensiveness, as in good men
and true, where the inconsistent use of pre- and post- adjunct
order in the same word-group diminishes the suspensiveness
of the consistently pre-adjunct order in good and true mm.

Grammatical Order.

1770. We see, then, that in languages which have both
a normal and an exceptional order, the latter is due to a
variety of causes, the most important of which is emphasis.
In such languages the normal order is grammatical
(syntactic), serving to show the grammatical relation between
words. The fewer the inflections, the more important this
function becomes, but even highly inflected languages observe
general principles of syntactic order, however freely they may
disregard them in special cases.

1771. An order which is exceptional in one period may
become normal in another period. Thus the pre-adjunct
order of Old and Modern English was probably originally
emphatic (1765). In Old and Modern English as well as in
most other languages interrogative words generally have

§ I775-3 WORD-ORDER : GENERAL PRINCIPLES. 5

front-order, as in where ts he? compared with he is there;
this order, again, was probably at first only emphatic. The
front-shifting of the verb in interrogative sentences (will he?
compared with he will) was also probably at first simply the
result of emphasizing the predicate, as also the front-shifting
of the verb in imperative sentences (come ye /).

General Principles of English Order and its Changes.

1772. What appears to be the original Arian word-order
is preserved in the early Sanskrit prose.

1773. In a normal Arian declarative sentence the subject
is followed by its modifier the verb, but otherwise pre-adjunct
order prevails; thus genitives and adjectives precede their
nouns, and adverbs, accusatives etc. precede their verbs, the
result being that the verb comes at the end of the sentence,
as it continued to do in Latin (1765).

The same pre-adjunct order prevails in Arian compounds
(1546), which shows that this order must be very old in Arian.

1774. In careless speech it often happens that a speaker
finishes a sentence grammatically, and then adds one or
more words as an after-thought, to complete the meaning or
define it more clearly. Such tag-sentences are frequent in
Arian, so that a verb which would otherwise have end-
position loses it, just as in English we may say he came,
John instead of he, John, came =John came,

m5. As these tagged sentences were generally longer
than the normal end-verb sentences, and as it was found
inconvenient to put the verb at the end of long sentences
generally, whether tagged or not, a tendency might easily
develop to give up end-verb order altogether except in short,
familiar sentences.* Accordingly, we can observe in the
separate Arian languages a gradual retraction of the verb-
position towards the subject-word ; thus, already in Old
Greek the verb is generally put immediately after its subject-

6 SYNTAX. [§ X776.

word, the verb being thus followed, instead of preceded, by
its own modifiers, just as in Modern English — Caesar con-
quered the Gauls,

1776. Verb-position in Old-English tends to follow the
same general principles as in Modern German. In inde-
pendent declarative sentences, such as the one just given, the
order is the same as in Modern English ; but in dependent
sentences the verb has end-position : *when Caesar the Gauls
conquered had. In other words, the original Arian word-order
was preserved in dependent sentences because they are
generally shorter and more compact than independent ones,
till at last end-verb order came to be the grammatical mark
of dependence.

1777. But, as we see from the last example, this end-verb
order may often lead to illogical and clumsy collocations.
And when a more convenient order had already established
itself in independent sentences, it was natural to extend the
order of these to the dependent sentences as well, the result
being the Modern English order when Caesar had conquered
the Gauls parallel to Caesar had conquered the Gauls.

1778. Old-English, having a considerable number of
inflections, was able to preserve a good deal of the freedom
of Arian word-order, being in this respect intermediate be-
tween Latin and Modern German.

1779. In Middle and Modern English we observe the
same gradual restriction of the older freedom as in German
and the other Modern Germanic languages. But while in
Modern German the Parent Germanic order was, so to say,
fossilized, English agrees with Swedish and Danish in
developing a more natural and logical order, characterized
especially by the prevalence of mid- verb position.

In the following details of English word-order, principles
which are common to Old and Modern English are, as

§1786.] WORD-ORDER: ADJECTIVES. ^

a general rule, treated only from the Modern English point
of view.

Adjectives.

1780.^ Assumptive adjective- words precede their head-
words : young man, running water, settled weather, many
men, three men, my house, the earth.

1781. But post-order is frequent in Old-English with
quantitative adjectives: Sumorsmte ealle 'all the people of
Somerset ' | his suna twegen ' his two sons * | hie butu ' both
of them ' (the two armies) | wcBter genog, where Modern
English has both orders — enough {of) water, water enough,
the latter being less emphatic. In Old-English also the
postposition of these adjectives seems to be the result of
their want of emphasis. But in the Modern English i soldiers
three the numeral has full stress.

1782. In Modern English postposition is regular in the
case of cardinal numerals used as ordinals: chapter ten
[but the tenth chapter^ page three, number three, latitude 39**,
in the year 1000. This usage seems to be due to French
influence.

1783. Also with participles used as adjectives : the day
following [the following day], the time being, the money re-
quired [the required money]. This order is, of course, the
result of these words being still felt to be half verbs.

1784. In Old-English postposition is frequent in exclama-
tions, as in Hropgdr leofa I * dear Hro]?gar I ' bropor mm !
t * brother mine ! '

1785. God (Elmihtig 'God almighty' seems to be an
imitation of the Latin order {Deus omnipotens).

So also the Modern English, the body politic, the Staies-
General, heirs male seem due to French influence.

1786. the -f- adjectives follows proper names in such
groups as Edward the First, William the Silent, parallel
to William the Conqueror (1801). We find the same

8 SYNTAX. [§ 1787.

construction in Old-English — jElfsidn se hleria * iElfstan the
bald.'

1787. In such collocations as novels proper and novels
improper the postposition is emphatic. In the colloquial
whisky hot the adjective is tagged on because it has the
complex meaning 'made hot by the addition of boiling
water/

1788. Postposition is often necessary in the case of
assumptive groups : in a manner the most picturesque \ a man
wise in his own conceit \ names well known in literature. But
groups precede when pre- order involves no awkwardness of
construction, especially when the group is felt to be equivalent
to a single word, as in he plays a not very conspicuous part in
the story, or when the group may be regarded as a compound,
as in the now declining day^ his already wearied horse.

More than one Adjective.

1789. When a noun has more than one modifier, the
general principle is that the one most closely connected
with it in meaning comes next to it, as in the three wise
men, where ww^ men is equivalent to the single word sages.
Qualifiers come before such groups, the one that is the most
special in meaning (three) coming next to it. Hence there is
a gradation of increasing specialization from the beginning
to the end of such a group (the, three, wise). In this
example only one of the modifiers is attributive. In a series
of attributive modifiers the same principle is generally ob-
served, as in a tall black man = a tall negro. In bright blue
sky — brightly blue sky the position of the first adjective
is partly due to its being logically a modifier of the second
one.

1780. But very frequent collocations such as old man,
young man (=. youth) have become so fixed that no other
adjective, even if more special in meaning, is allowed to

§1795.] WORD^ORDER: ADJECTIVES, 9

come between the two words: a conceited young man.
Hence we cannot make old sage into *old wise man.

1791. When the modifiers are about equally balanced,
the order may vary, as in the two first weeks, the first two
weeks, and the Old-English on pam oprum prim dagum
compared with the Modern English in the course of the
three following days.

1792. We have seen that when the articles are associated
with another noun-modifier they normally precede the latter :
but in some constructions they come immediately before the
noun. The definite article does so when associated with
certain general adjectives of quality: all the books, all the
corn, half the day, treble the quantity, both the armies. Old
English sometimes has the same construction {ealne pone
dcBg), although it generally prefers postposition {pcEtfolc eall,
pa hei ealle, pa h§rgas begen (1781). Old-English also
has the construction adjective + genitive^ as in manige
(or feld) para manna * many of the men/ In Modern
English we feel all the day to be equivalent to the whole of
the day,

1793. The ind^nite article has the same position in
combination with half— half an hour [but a half loaf] —
and in other combinations : many a man, many a one, not
a moment to lose, ithe knight did bear no less a pack.
Also in combination with intensitive adjective-pronouns:
what a pity I \ I never knew such a man I These words
naturally precede the a + noun through being emphatic.
In such constructions as j^ long a time, as good a man as
any, too good a man the order is the result of avoiding the
awkward collocations *a so long time, etc

1794. The construction with possessive pronouns is
analogous : my old friend, but all my time, half their time,
both his eyes.

1795. We also have mid-possessive order in ^good my
liege I

lO SYNTAX. [§ 1796.

Genitives,

1796. Genitives always precede their nouns in Modern
English, which is also the normal Old-English position:
pees cyninges brobor 'the king's brother/ on Godes naman
* in the name of God.' The more intimate the connexion,
the more fixed this order is; hence it is absolttely fixed
in semi-compounds such as ^ngla-land ' land of the Angles,
England/

1797. But in Old-English post-genitives are also frequent,
especially in combination with quantitative words and groups,
as in manige para sekstena cyninges begna ' many of the best
king's thanes/ where the last two words form a semi-com-
pound, mtcel dctl para burgwara * a considerable portion
of the citizens/ Also in combination with preposition-
groups, as in on neawiste pare byrig *in the neighbourhood
of the city,' on twa healfa pmre te *on both sides of the
river'; here it is the result of avoiding such constructions
as on pare byrig neawiste^ where there is separation of the
elements of the preposition-group, although such construc-
tions do occur.

Modifying Nouns.

1798. In Old-English modifying nouns follow their head-
word, as in jElfred cyning ' king Alfred,* Eadgdr cepeling
'prince Edgar/ except when the modifier is emphatic, as
in midel pcBS folces ofer sa d-drafdony buton pam cyninge
jElfrede * (the Danes) drove many of the people over the
sea, except the king, Alfred/ It is to be observed that
in JE,lfred cyning the adjunct-noun is subordinated to the
proper name not only in position but also in stress.

1799. In Modern English the modifier comes first ; king
Alfred^ Mr. Smithy Dr. Tanner, Farmer Hughes, Brother
Jonathan, Friend Mill. This change of order seems to
be due simply to the analogy of the pre-position of ad-
jectives and other noun-modifiers, the adjunct-noun keeping,

§i8o4.] WORD-ORDER: PRONOUNS, II

however, its weak stress {-h'ng Alfred), which would be
impossible in Old-English.

1800. The articles may precede the adjunct-noun ex-
ceptionally in a Mr. Smith etc., and regularly in such
collocations as the angel Gabriel^ the emperor Maximilian ;
so also possessives, as in my friend Smith.

1801. But when the adjunct-noun has a strongly specia-
lizing function, it follows its head-word in Modern as well
as Old English, being preceded by the definite article,
as in William the Conqueror, fohn the Baptist compared
with the angel Gabriel, there being only one ' Baptist,'
while there are supposed to be many angels. In Mr. Smith
the bookseller we have both positions of adjunct-nouns.
In /, fohn Smith, he (the speaker) said . . the post-
position is inevitable; so also in Edward, Prince of Wales
compared with Prince Edward,

Pronouns.

The position of pronouns has been incidentally dealt with
under that of nouns and adjectives.

1802. The Old-English postposition of quantitative ad-
jectives (i 781) is still preserved in combination with pronouns :
are they all gone ? \ we thank you both \ the awkwardness of
our (or us) both addressing the same lady. But these words
necessarily precede possessives, except in constructions such
as that in the last example : both his eyes, \ to frustrate both
their hopes.

1803. Adjectives modifying indefinite noun-pronouns follow
them, as in something bad, anything good [compared with any
good thing'], nothing remarkable) similarly in anything else.
If the adjective precedes, they become pure nouns : in
her manner there was an indefinite something.

1804. myself etc. follow noun-pronouns in the same
way : he himself says so. Broken order is more emphatic :
/ will see about it myself

12 SYNTAX, [§ 1805.

Verbs.

1805. As regards the position of the verb in the sentence,
it is to be observed that some subordinate words — con-
junctions and adverbs — always take precedence, such as
and and the Old-English ne *not/ Hence in sentences
beginning with and or ne + verb, the verb is practically
initial.

Verb + Subject.

1806. (a) In interrogative sentences (1771) : are you
ready ? \ where is he ?

(l>) In imperative sentences, where, however, the pro-
noun is generally omitted. In Old-English it is added
regularly in negative sentences, where the verb necessarily
precedes the pronoun through being attracted by the ne
(1807): ne heo ge bitere I *be (ye) not bitter!' The
postposition of the pronoun in positive sentences, as in
cume gel for the more usual cumaj? ! *come (ye)!', seems
to be due partly to the analogy of the negative con-
struction, pardy to the feeling that the pronoun is a tag.
When the pronoun is added in Modern English, it follows
the verb in negative sentences — do not {you) do that 1 —
while in positive sentences it generally precedes the verb
in the spoken language : {you) let that dog alone ! \ never
{you) mind!

Pre- verb order occurs already in Old-English : /« soplice cyp
line gesihpe ! * do thou make known thy vision ! *

{c) In non-imperative sentences expressing wish or com-
mand: may I be hanged I \ perish India I In so be it I
the verb is attracted by the so (1810). This inversion is
avoided in the case of transitive verbs because of the
ambiguity that would arise : God forbid ! | God save the
Queen I Both orders are also found in Old-English. These
sentences evidently follow the analogy of imperative sentences.

§i8ia.] WORD-ORDER: VERBS 13

{d) In sentences of condition nare [= ne ware] sio ^dsf-
Hie getacnung * were it not for the spiritual meaning ' | were
Richard mine, his power were mine \ were he my brother =
* even if he were my brother.'

So also in sentences of alternative hypothesis : wyle he^ nyle
\=.ne wile] he * whether he will or not, willy-mlly.'

After Front Words.

1807. In Old-English certain adverbs draw the verb after
them, so that it precedes its subject. This is always the
case with ne, as in ne mceg id pcei don * I cannot do that '
compared with ic mceg pcet don.

1808. Generally also with such adverbs as /5 *then//<jr
'there,' which serve to connect the sentence they introduce
with what precedes : pa feng Alfred to ride ' then Alfred
succeeded to the throne ' | and par wearp se cyning ofslcBgen
' and there the king was killed ' | andpcssymd twegen monap
gefeahi se cyning wip pone h§re ' and two months after that
the king fought with the (Danish) army/

1809. The connective force of this order is shown by its
occasional occurrence after the conjunction and: and pat is
peah swipe cBdig ; and lie gap wilde moras wip eastan ' and it
(the country) is very rocky; and wild moors lie eastwards
(of it).'

1810. But even in Old-English there are many exceptions,
and Modern English generally has the normal order subject
+ verb, although inversion still occurs even in the spoken
language : nor do I \ so do I \ now comes the amusing part of
the story.

1811. Verb-inversion is sometimes caused by a preceding
dependent clause both in Old-English and Modern literary
English : not as the world gives, give I unto you. Here also
it is evidently connective.

1812. Verb-inversion in appended or parenthetic clauses
of statement is also connective ; yes^ said the boy \ yes, said

14 ' SYNTAX, [§1813.

he, I will I what, said he, do you want ? With t quoth he,
whose verb is used only in this construction, the order is
invariable, but with other verbs pre-subject order also occurs :
my son, my son I he cried, they have murdered him ! So also
in Old-English he cwcBp is used in the appended as well as
the front-position instead of cwcbJ> he.

In vulgar English the inversion occurs also in front clauses :
says he, what do you want f

1813. In Modern literary English connective verb-inversion
is frequent in dependent sentences as well : Death itself is
not so painful as is this sudden horror \ Ruppin, where lies the
greater part of the regiment \ a cottage in which lived the widow
of a former curate \ what were his thoughts I cannot tell. This
is a complete reversal of Old-English tendencies (1776).

1814. Another kind of verb-inversion is that caused by
emphatic or exceptional front-position of other classes of
words. This is frequent with adverbs and adverb-groups,
as in the following examples, where the inversion is obliga-
tory : scarcely had I sat down, when . . \ not till then did /. . |
to such straits were they reduced that . . Examples with other
parts of speech are : enclosed is a letter \ the greater their
power, the greater seems their revenge \ ^high sparks of honour
in thee have I found. Here also the inversion is often obliga-
tory ; such a construction as * enclosed a letter is would make
the is too emphatic (1766). In other cases the inversion is
exceptional, as in how foolish was It where Wd!j=(w9z),
not (woz), which would make the sentence into a question.
Even in Old-English front order does not necessarily cause
inversion unless connectiveness is implied at the same time.

1815. In such phrases as here is your hat I \ there goes the
richest man in England I, compared with here heist \ there he
goes I, the inversion is fixed, because the adverb is felt to
be a substitute for a subject-word, here is your hat being
equivalent to this is your hat. This construction is especially

§i82i.] WORD-ORDER: VERBS. 15

frequent with the weak there : there is no more bread \ there
lived a man.

1816. In the formal, elliptical style of public notices the
verb is put first for emphasis in such phrases as died on
the 14th. Similarly in periphrastic tenses, with omission of
the verb : to be sold, a desirable family residence.

1817. Otherwise front verbs occur only in poetry : smiled
then, well pleased, the aged man (Scott). With a verb which
is wholly, or partly, transitive, as in shook all the hollow caves^
this construction is hable to cause ambiguity.

Modifier + Verb.

1818. A verb regularly follows instead of preceding its
modifier when that modifier is a relative or conjunctive word,
as \nyou are the man whom I want \ I know where he is.

1819. Verbs also regularly follow exclamatory words : how
he boasts ! | what a strong man he is !

The original Arian post-verb order is also preserved in
some cases in Old-English in independent as well as de-
pendent sentences (1776).

1820. Thus pronouns generally precede the verb, as in
he hine gefeng 'he took him prisoner,' hie begeaton welan and
us lafdon * they acquired wealth and left (it) to us,' compared
with he gefeng pone cyning. So also with adverbs : he pa swa
dyde ' he then did so,' hie par wunodon ' they dwelt there.*
These words have evidently kept their original position
through being unemphatic. This order is only poetical in
Modern English : the serpent me beguiled.

1821. Even full nouns occasionally have the same position,
as in he par sige nam * he gained the victory there.' Here —
as also to some extent in the preceding examples — the two
words form a sort of compound, sige :nam being parallel to
bi'Standan, -hine gefeng to be'cuman (740). This order is
especially frequent in sentences introduced by and\ pa ridon
hie hider . . , and pone CBpeling on pare hyrig melton. If the

l6 SYNTAX. [§

second sentence came first in a group of sentences, the end-
verb order would be impossible, as being too abrupt. But
even in the earliest Old -English verbs of weak meaning such
as * have ' and ' be ' always follow the Modern order, keeping
as close as possible to their subjects, as in hcB/de stge ' had
the victory,' compared with stge nam \ he wees twelf niht mid
p&m cyninge^ and he hine miclum and his geferan mid feo
weorpode * he was twelve days with the king, and he honoured
him and his companion greatly with money/

1822. We still preserve post- verb order in ime thinks
through the analogy of / think.

Several Verb-Modifiers.

1823. Datives precede accusatives: he gesealde his hreper
heal/ his rice ' he gave his brother half his kingdom/ So
also in Modern English a datival noun or pronoun precedes
an accusatival noun, as in he showed me his pictures [in Old-
English a dative pronoun precedes the verb, § 1820] ; but if
both are pronouns, the accusatival pronoun precedes : give it
me I [Old-English : s^le me hit /].

1824. Object-complements naturally follow the object-
word : they made him king \ to call people bad names \ to paint
a house white \ \\his crimes make guilty all his sons\. So
also if the complement is a verbal : / saw him come \ I want
him to come \ I saw him coming \ I saw it done \ excuse me
interrupting you.

Verbals and Periphrastic Verbs.

1825. Verbals are followed by their modifiers in the same
way as the verbs they are formed from : i^e saw him) giving
a beggar some money.

For the * split infinitive ' see § 1864.

1826: In Old-English a finite verb in combination with
verbals has the same freedom of position as a simple verb,

§ 1830.] WORD-ORDER : VERBALS, I7

being generally put at the end of a dependent sentence:
/fd hi geseah fxBt he oferwunnen beon wolde * when he saw
that he was going to be conquered/ In independent
sentences the tendency is to put the verbal at the end : pa
opre wcBron hungre d'cwolen ' the others had died of hunger ' |
vianige cupon fnglisd gewrit d'rcedan 'many could read
English writing/ It is to be noticed that in such sentences
as hie hcef don pone cyning o/slcBgenne (later: o/slcEgen) 'they
had killed the king ' the verbal comes after the noun because
it originally stood in apposition to it. In Old-English

end-verb order in independent sentences is frequent in
periphrastic forms composed of participles and 'have' or
'be/ but only in joined-on sentences in accordance with
§ 1822 : and hie pa ymb pa gatu feohtende war on oppcet . .
' and they fought round the gates until . /

1827. In Modern English, on the other hand, there is
a tendency to make the elements of a periphrastic form
into a compact group, whose modifiers follow it in the same
way as they follow a simple verb: I-shall-have-written my
letter. But many adverbs are freely inserted : / do not know \
I have never seen it.

1828. In Modern English the elements of a periphrastic
form follow each other in a fixed order, which is rarely
departed from even in poetry : a strong tyrant who invaded
has our land (Spenser). But emphatic front-position of the
uifinitive is not unfrequent: ifor die I shall!

1829. The order of the verbals in periphrastic verb-forms
is the natural result of their development. The finite peri-
phrastic verb shows the following orders : —

pres. partic. ( + pret. partic.)
infin. + - pret. partic. f + pres. partic. (-hpret. partic.)
auxil. + ^ I I -f- pret. partic.

pres. partic. + pret. partic.
pret. partic. -f- pres. partic. ( + pret. partic.)

1830. It will be observed that the infinitive occurs only
VOL. u. c

l8 SYNTAX. [§ 1831.

as the first of a group of verbals, the other places being
filled up exclusively by participles, which may also take the
place of the infinitive itself.

1831. The periphrastic infinitives always begin with the
supine ; they have the same forms as the finite verb as far
as their verbal-elements are concerned.

1832. The periphrastic (present) participle and gerund

has only the order pres. partic. + pret. partic. in common

with the finite forms ; it has in addition the following forms : —

f pjes. partic. ( + pret. partic.)
pres. partic+pret. partic. + ] ^^^^ ^^^.^

All periphrastic participles begin with pres. partic. + pret.
partic.

Adverbs.

1833. Adverbs modify so many different parts of speech —
besides modifying groups and sentences — that their position
varies greatly. They show, indeed, almost the last remains
of normal free order in Modern English.

For dependent adverbs see § 1858.

Adverbs + Nominals or Adverbs.

1834. In accordance with general principles, adverbs
precede adjectives, adverbs, and adverb-groups : very quick^ not
so, quite in-the-wrong. So also adverb-groups precede under
the same circumstances : not-at-all sorry, not-at-all in-the-
wrong.

1835. The adverb enough, like the adjective enough, can
either precede or follow its head-word in the earlier Modem
as well as Old English, but in the present English it can
only follow : good enough, quickly enough. In Old-English
it may follow adjectives, as in sweti genog 'sweet enough,'
but generally precedes adverbs, as in genog georne witan
* know well enough, know quite well/

1836. An adverb immediately preceding a noun is in-

§i84i.] WORD-ORDER: ADVERBS, IQ

distinguishable from an adjective, as in the then state of
affairs. But an adverb before a group-noun — even if the
group consists only of article + noun — still remains an adverb,
although, of course, it approaches in meaning to an adjective,
as in qiMe the gentleman^ not at all a lady, fully master of the
subject,

1837. In some constructions adverbs follow their nouns,
standing-to them in a kind of loose apposition, as in the man
there=* the man standing there' or 'the man who is there/
one more = * one in addition.' The connexion is still looser
in such constructions as to permeate space generally.

Verb + Adverb. ^

1838. In accordance with general principles adverbs and
adverb-groups generally follow verbs: come in \ let us go
now I I-shall'have- finished in-a few-minutes.

1839. not always follows a simple verb, but with peri-
phrastic forms and verbals it has the same position as verb-
preceding adverbs, that is, it follows immediately after the
auxiliary instead of coming at the end of the periphrastic
form — as in / shall not go \^:he will want not our help], do
not go! compared with I will not \ igo not! — and precedes
supines and other verbals, as in to be or not to be, that is the
question \ not knowing what else to do, I came home.

1840. If the verb has other modifiers more intimately
connected with it, these precede the verb-following adverb ;
this is especially the case if the other modifiers are logical
predicates (/ am ready now), or objects, as in ask him again !

1841. All adverbs — whatever their position may be in
other cases — necessarily follow interrogative verbs; but, as
a general rule, they do not follow the verb itself, but the
accompanying pronoun: is he here? j is he never ready in
time? If the verb is periphrastic, the adverb comes at the
end of the group: shall we go now? But not comes
immediately after either the verb itself — in which case there

C 2

20 SYNTAX, ^ [§1842.

is always contraction in the spoken language — or the accom-
panying pronoun, in which case there is no contraction : is
not (iznt) he here ? \ is he not here ? \ shall not (Jaant) we go
now? I shall we not go now? The second order is preferred
in writing, except m reproductions of colloquial speech,
where the contracted forms are generally written (w«7,
shan't). The second order occurs only occasionally in
speech.

1842. In a succession of adverbs and adverb-groups
those most intimately connected with the verb precede:
come up at once 1 \ we went to school together \ I want to
look about me a little.

1843. When a verb is followed by an object-word and an
adverb, the order of these is sometimes doubtful, as in
/ have brought back your umbrella or / have brought your
umbrella back. In such a sentence as bring in some more
coals! the adverb generally precedes. But the general
tendency is to put the object first ; in some cases, indeed,
no other order is allowable, as in let him in I \ I have left
my umbrella behind. The reason appears to be that the
adverb might be mistaken for a preposition, if put before
the noun-word.

1844. If several adverbs follow without being specially
connected with the verb, time-adverbs generally come first :
he has altered a good deal lately \ we expect him home again
to-morrow^ where, however, again seems to be specially
associated with the verb, forming with it a kind of com-
pound. If again is detached by end-position, it becomes
emphatic, as in to get back to civilization again compared
with to get back again to civilization. So also in he is there
still compared with he is still there,

1845. When one of two modifiers is a lengthy group, the
shorter verb-modifier is often allowed to precede even if
it would otherwise follow, as in he heard again the language
of his nursery [he heard it again] ^ I met him last night at

§1847.] WORD-ORDER: ADVERBS. Hi

a party at Mrs. Carter s [/ met him there last night]. In
such cases it is felt that the heavier modifier will easier
bear separation from the verb.

Adverb + Verb.

1846. An adverb which precedes a simple verb also pre-
cedes the supine and other verbals — misfortunes never come
single I I hope never to see his face again \ the thought of never
seeing him again — and follow auxiliaries and the unem-
phatic is : / have never spoken to him \ I should never have
thought of that | he is never ready in time. But if these verbs
are emphatic or detached, the adverb precedes them : he
never 'is ready in time! \ I never have spoken to him, and hope
I never shall.

Weak have (not auxiliary) and the link-verbs take the adverb
before them : he never has any money \ he never gets ready in
time. Note that *he has never would in speech be contracted
(hijz nevd), which would suggest he is.

1847. When an adverb is put before instead of after
a verb, it ceases to modify the verb exclusively, but modifies
also the whole group of words connected with the verb, as in
/ certainly think so compared with / think so certainly (364)
I / hardly think we want afire compared with to think hardly
of a person^ where the change of position completely changes
the meaning of the adverb. So also gladly in I gladly acceded
to his request means * I was glad to (accede to . .),' while in
I acceded to his request gladly it means simply ' with gladness.'
So also if we made he generally failed to explain his meaning
into he failed to explain his meaning generally, the adverb
would modify explain only, and the meaning would be ' he
succeeded only partially in explaining his meaning.' Other
examples are : Wednesday came, and luckily it was a fine day,
where luckily = * it was lucky that . .' | / cannot begin my
work again when I have once been interrupted compared with
interrupted once.

22 SYNTAX. [§1848.

1848. Another result of this is that a verb-preceding adverb
is often vaguer in meaning than a verb-following one. Thus
while the end-adverb in / understand you perfectly has its
literal meaning, it tends to become a mere expletive in such
sentences as I perfectly appreciate the delicacy oj your position.
So also in the last example of the preceding paragraph.

1849. As the pre-adverb order tends to distribute the mean-
ing of the adverb (1847), it sometimes has the same effect as
front-order often has (1854), that is, it tends to give it a depen-
dent meaning, so that it seems to refer back to what pre-
cedes. This is clearly seen in many adverbs of time, which,
when they precede their verbs, suggest the idea of sequence
as opposed to that of an isolated event, as in I now proceed to
explain . . compared with I proceed to explain now, where now
means * at the present moment.* So also in he afterwards
became a schoolmaster \ we then went on to Rome^ where then is
parallel in meaning to the initial connective adverb-group in
the next day we . . [contrast : he came on Monday ^ and went
away again the next day],

1850. Adverbs are often put before instead of after parti-
ciples in periphrastic forms as if the participle were an adjec-
tive, as in he has been very kindly treated 'mste3.d of the normal
he has been treated very kindly \ it is very well done [he will do
it very well].

1851. The adverb never always precedes its verb : never
mind! So also the synonymous hardly ever : / hardly ever
see him now. Several other adverbs of time also show a
strong preference for pre-order, especially the corresponding
always : he always dines at the same place [he dines there con-
stantly], I often see him at the theatre \ it rarely happens that. .
With all of these post- or end- order is much less frequent.

1862. The pre-order of the intensitive quite (I quite agree
with you) seems to be fixed.

1853. We have lastly to note the illogical pre-order of only
in such sentences as I only want sixpence-^ I want only sixpence

§ i857.] WORD-ORDER : SENTENCE-WORDS. 23

or / want sixpence only. The hesitation between these last
two orders was probably the reason for preferring pre-order,
which was of course originally emphatic, / want only sixpence
implying * all I want is — I want nothing more than —
sixpence/

Front Adverbs.

1864. Front-adverb position often suggests the idea of
dependence on what precedes, as in the next day zve went on
to Rome (1849). So also in here he stopped short in his
speech \ they used to be quite common, and now they are quite
rare. This order is not emphatic ; in fact, if we wished to
make the now of the last sentence emphatic, we should put it
at the end : they are quite rare now.

1855. But in some cases front-position seems to express —
or at any rate to accompany — emphasis on an adverb, and in
'now we are ojff at last! compared with / am going mow
[/ am going 'now with a different kind of emphasis from that
of the first sentence], 'certainly I think so \ of 'course I shall
\he said it of '.course without thinking\ after 'all, it does not
matter I The front-order in here we are ! there he goes! seems
also to be emphatic.

Sentence-words.

1856. Sentence-words and sentence-groups used as voca-
tives and imperatives are very free as regards their position
in the sentence, front-position being generally emphatic, as in
' sir f you are mistaken I compared mihyes,-sir / \ please do it
again I compared with half a cup, please! \ thank you, I would
rather not! compared with no, thank you. But An yes, please,
the please may be emphatic, because, 2Lsyes is also a sentence-
word, please itself may be regarded as beginning — or rather,
constituting — a new sentence.

1867. Interjections are generally emphatic, and therefore
prefer front-order.

24 SYNTAX, ' [§ 1858.

Conjunctions and Dependent Adverbs.

1858. Conjunctions, being connective words, naturally
come at the beginning of the sentence they join on.

1869. When a pure conjunction and a half-conjunction
or sentence-adverb come together, the conjunction naturally
precedes in most cases — always in the case of such conjunc-
tions as andy hut, which never occur except initially: hut,
nevertheless, . . With some conjunctions the order is
indifferent : t/, on the other hand, . . \ on the other hand^ if . .

Broken Order.

1860. Broken order is very frequent in Old-English. When
two or more co-ordinate words ought to precede a word which
they jointly modify or are modified by, there is a tendency to
avoid suspensiveness by putting only one of them before
this word, and letting the others follow in tag-order: swipe
midle m^ras fersce ' very large fresh-water lakes ' | Cyne-
wulf henam Sigehryht his rues and Wesiseaxna witan
' Cynewulf and the West-Saxon senators deprived Sigebryht
of his kingdom' \ges(Et poet land and ged&lde 'occupied the
country and divided it ' | het gewyrdan dne burg pckr on nea-
wiste and gemannian * ordered a fort to be built and garrisoned
in the neighbourhood.' Of course, if the whole group is very
short, or if the two words are closely connected in meaning,
or if one of them is of subordinate force, there is no break :
and he him/eoh andfeorh gehead ' he offered them money and
their lives ' | an berenfell ' one bear's skin.'

1861. In some cases Old-English broken order is the
result of the second word being in apposition, as if it were
tagged on by a relative sentence: hi hce/de tamra deora
unhehohtra siex hund ' he had six hundred tame reindeer un-
sold.' In such cases the tag- order is preserved in Modern
English as well.

1862. Old-English makes a free use of broken order in

§i866.] WORD-ORDER: CROSS-ORDER. 25

Other cases as well. In such a sentence as pa cyningas pe
pone anweald hce/don pCBs folces onpCBm dagum ' the kings who
had the rule of the people in those days ' the break between
anweald and its genitive is to avoid making hcefdon too
emphatic. In pat hira ne mihte nan to qprum * so that none
of them could come to the others ' it is the result of the front-
position of h'ra, which is put immediately before the verb
because it is logically equivalent to * they.' The end-position
of verbs and verbals often causes broken order, as in bd opre
waron hungre d'cwolene * the others had died of hunger.*

1863. Modern English is much more tolerant of suspen-
siveness, and the logical spirit of the language makes it
averse to broken order. The Old-English order adjective 4-
noun + adjective survives only in such isolated phrases as
good men and true. Breaks caused by emphatic end or
front order, such as bars and bolts we have none \ of fuel they
had plenty, belong only to the literary language.

1864. But there is one kind of break which is unknown
in Old-English, and is mainly of Modern English growth -
the so-called ' split-infinitive ' — that is, the supine with a word
or words coming between the to and the verbal, as in it is
necessary to clearly understand this point instead of the more
usual it is necessary to understand this point clearly.

Cross-order.

1865. When two word-groups or sentences of similar
construction follow each other, they may, in a language
which has free order, be either in parallel order (anaphora)
or cross order (chiasmus). Thus in Latin we have parallel
order in alio loco, alio tempore * in another place, at another
time,* cross-order in multos defendi laest neminem *I have
defended many, injured none.'

1866. Cross-order occurs in Old-English, as in pcet land
is eall wisie, butan on feawum stdwum styccemdlum wlciap
FinnaSf on huniope on wintra^ and on sumera on fisdope be

26 SYNTAX. [§1867.

p&re sa *the country is all desert, except that in a few
places Fins dwell piecemeal, (being engaged) in hunting in
winter and fishing by the sea in summer.' Here it is the
result of on sumera being attracted by the similar group on
wintra. This probably is the origin of the construction,
although in higher stages of development it was used for
emphasis and rhetorical effect.

Front-order.

1867. As we have seen, front-order is normal in the case
of conjunctions and other particles (1858), of interrogative
and exclamatory words (1771), and of verbs in interrogative
sentences (1806). In some of these cases the front-order is
emphatic, in others connective. We have also considered
more exceptional cases of front-verb order (1811 foil.) and
front-adverb order (1864). Other exceptional cases of front-
order are : —

1868. Front predicative adjectives: thoughtless he may be,
but not vicious \ victorious indeed they were, but at what a cost !
These are purely literary examples. But this order occurs
also as a vulgarism in such phrases as right you are 1

1869. Front object-word : this he owed partly to his father.

1870. Front object-complement : enclosed you will find
a letter. Compare enclosed is a letter and the other examples
in which front-order is accompanied by verb-inversion (1814).

The front adjective in such sentences as big as he is^ I know
a still bigger man \ black as he hcts been painted^ he is far worse
in reality is the necessary result of their grammatical structure.

For the front-order in of fuel they had plenty^ see §§1868,

1871.

Group-order.

1871. Preposition-groups normally follow their head-
words. Such inversions as ^of Corinth king on the
analogy of genitive + noun are rare even in poetry. But
preposition-groups often take emphatic front-position in

§i87S.] WORD-ORDER: SENTENCES. 27

prose as well as poetry : of fuel they had plenty \ of cowardice
he has never been accused; and with verb-inversion: | of
a noble race was he \ to such straits was I reduced. Such
constructions as three sons, of whom all died young are blend-
ings of . . all of whom died young and . . who all died young.

Verbal-groups.

1872. The position of verbal-groups in the sentence is
quite free : hearing you were in want of a secretary^ I have
come to tell you of one \ welU this being now settled^ let us go on
to something else \ they would not let him go home^ it being
a stormy night.

Sentences.

1873. In groups of sentences we can observe the same
distinction between post-adjunct and pre-adjunct order as in
groups of words (1871).

1874. Independent sentences follow post-adjunct order.
Indeed, no other order would be possible with most of them,
such as copulative and adversative sentences.

1875. Dependent sentences, on the other hand, oflen
follow pre-adjunct order, especially temporal, hypothetical,
and causal sentences (466): when he came, I was not at
home I if I can ^ I will do it \ as I saw it was no use arguing^
I said no more. The reverse order is nearly always allow-
able : / was at home when he came \ I will do it if I can.
The general principle is, of course, to put in front the
sentence which is most emphatic or most closely connected
with what precedes. But in some cases the order is fixed,
especially when there is a difference of meaning. Thus
causal clauses introduced by as always precede, for, if they
followed the head- sentence, they would be understood as
clauses of comparison i^do as you please /). This seems to
show that with dependent sentences the pre-adjunct order is
the normal one.

28 SYNTAX, [§1876.

1876. Front-position of conjunctive and relative clauses is
sometimes regular, as in what you say is true, sometimes
exceptional and rhetorical, as in what he wants 1 do not
know.

Elliptical Order.

1877. In the elliptical language used in titles of books,
headings of chapters, indexes, etc. there are many deviations
from the order of the ordinary language. Thus in an index
to a book of history v^re should find the French Revolution
given as Revolution^ The French in the strictly logical order,
while the order in William the Conqueror, being already
logical, is retained. So also with proper names the more
important surname would come first : Mill, John Stuart,

1878. This elliptical style also has an emphatic order
different from that of the ordinary language, as in Waterloo,
Battle of.

1879. In some cases this elliptical order has passed into
common speech, as when we say nine thirty instead of
half-past nine,

1880. Another kmd of elliptical order is preserved in
such phrases as /*/ costs five shillings a pound \ I paid twopence
each for them \ do you mean last Monday ^ or Monday week ? \
He lives at six^ Queens Road,

SENTENCE-STRESS.

1881. In addition to the three degrees of stress already
distinguished (659), we now require a fourth — extra stress,
marked (;).

1882. The general principle of sentence-stress is to put
strong stress on those words which are prominent in any
way, and to give weak stress to those which are subordinate
to them (736, 880).

1883. Hence the general tendency is to give strong stress

§i888.J . SENTENCE-STRESS. 29

to full words, weak stress to form-words. But if a full word
becomes subordinate m meaning, it can take weaker stress
(a :piece 0/ bread). Conversely, if a word which is usually
subordinate is made emphatic or becomes logically promi-
nent, it can receive strong stress.

1884. In connected speech all words that express new
ideas are more or less emphatic, and therefore take strong
stress; while words that express ideas that are already
familiar or can be taken for granted, are unemphatic and
have a weaker stress : a German :came to London . . the
.'German left : London, and .-went to Liverpool.

1885. In such a sentence as -I :got wet the first word is
understood from the context, and the second is a mere
connecting word, so that the stress necessarily falls on wet by
what may be called 'negative emphasis.' But in look how
;wet I am! there is increased stress on wet, which gives it
the meaning * very wet ' ; this is positive emphasis or
emphatic stress. If the word very is put in, there is no
necessity for emphatic stress : you are 'very 'wet.

1886. The example just given is one of what may be
called intensive stress. Another example is you will catch
your jdeath of cold. Intensive stress is often emotional, ex-
pressing excitement: ;what, do not you know? \ what on
;earth is the matter ? \ ;good jheavens, gentlemen !

1887. A less emotional form of intensive stress is that
which expresses the idea * even ' : enough to make a ;saint
swear ! [no extra stress in enough to make even a saint swear^ |
/ would not have it at a ;gift \ I did not do it: L never should
have /thought 0/ such a thing,

1888. Another form is that which expresses antithesis :
he did not remember your name, how could he ;help remem-
bering it, when he hears it twenty times a day ? \ we shall be
too late, after all. I ;told you so \ you have made a great
mistake. I jknow I have. In the first example the anti-
thesis is between not remembering and the impossibility of

30 SYNTAX [§ 1889.

not remembering ; the other two imply the antithesis ' you tell
me this as something new, but I told you — 1 knew about it —
before/ In all these examples the extra stress might be
accompanied by the compound-rise tone (1951), which would
suggest the emotional element of peevishness or impatience.
The second might be accompanied by the compound fall
(1952), with a different shade of meaning. In such an
example as the following, the extra stress seems to be neces-
sarily accompanied by the compound rise: can he do it? hi
;says he can,

1889. In the above examples the extra stress adds some-
thing to the meaning, but in some cases the extra stress falls
on words which are already intensive, without modifying their
meaning, as in I ; quite agree withyou^ where the substitution
of ordinary strong stress would not alter the meaning. In
fact, It is logically impossible to add anything to the force of
such words as quite, all, always^ never. But this kind of stress
may suggest something emotional, as in I will ;very soon
show you what I mean, where it may imply threat. The
practice of giving extra stress indiscriminately to all intensive
words is a common fault, especially among women.

1890. When two or more words are contrasted, either
they receive extra stress or the stress of the repeated word is
diminished : some English people have ;light :hairy some have
;dark :hair \ the town imouse and the country :mouse. So also
in counting, such numerals as 'thirteen, 'fourteen, 'fifteen have
the stress on the first syllable, while when isolated they have
even stress 'thirteen^ etc. (922). We may call this 'con-
trasting stress.'

1891. Modifying stress is similar : the earth is round, it
is not ; quite round, but a little flattened.

1892. The completion of a series is marked by what may
be called climax-stress : one, two^ three, and a;way I \ here
and there and ; everywhere \ the days of the week are Sunday,
Monday . . ; Saturday.

§ 1896.] SENTENCE-STRESS. 31

1893. In such a sentence as shall we smoke a ;pipe to-
igether ? compared with shall we smoke a pipe? the extra stress
is not the result of emphasis, but of the following weak-
stressed word, the effect of the increased stress being to bind
the two words together more closely, as in goodmatured man
compared with -good-natured (929). We can call this group-
ing stress. Other examples are : / am going to ;call on him
\I am going to call on some -friends'] \ one gets ;used to
things I a room with the ^-windows shut \ I forgot to wind my
;watch up [. . wind up my watch] \ I did nothing ;wrong^
'did I?

1894. So also formulae and traditional phrases are bound
together by one word in them having a predominant stress :
the house that ;fack built \ all is well that ;ends well \ this day
;week \ cut and come a;gain. The general tendency in such
combinations is to stress the modifying word.

1895. If a naturally weak-stressed connective word is
separated from the words which would otherwise follow it by
an inserted group or clause, it receives strong break-stress :
he 'is, physically speaking, a failure \ a man -who, if he had
the chance, would do great things \ 'if, as is most probable, . . |
-but, said he, . .

1896. There are some formal connecting words, such as
be^ which have little or no meaning in themselves, and are
therefore incapable of independent emphasis. . Hence a
strong — that is, in this case, an extra strong — stress on such
words is felt to be equivalent to emphasizing the whole
sentence: what ; are you doing? \ what ; does he know about
it? \you are late. I ;am rather late \ he will be angry. ;let
him be angry ! We may call this distributed stress. The
positive emphatic non-interrogative forms of the verb (2169)
always have distributed stress : I ;did say so.

The contrasting stress in man never /is, but always to jbe blest
applies logically not to the verb itself but to its distinctions of
tense : as such verbs as is have a meaning of their own from

32 SYNTAX, [§ 1897.

this point of view, strong stress on them is not necessarily
distributed.

The strong— not extra strong— stress in such sentences as
what is itt is only negative, being the result of the subordination
of «//^^/ (1915).

1897. As even stress tends to balance words against one
another (898), it is the rule in combinations of adjectives
(both assumptive and predicative) with nouns, and of adverbs
of marked meaning with other words : 'three 'big 'dogs \ I am
'quite 'ready \ it rains hard. This is especially the case with
negative words : I think not [I think -so] \ I cannot go ('kaant
•gou) 'now compared with / -can go.

1898. But negative words sometimes have weak stress, as
in have you heard whether the house is insured or -not?^ where
the not is not stressed because it is logically superfluous [but :
you will have to do it whether you like it or 'not\ and in some
cases where it forms a sort of compound with an intensive
word : the voyage is certainly -not a long one. So also in
/ would rather -not go. But when not precedes such words
it takes the stress ; not -very well \ not :quite ready \ it is not
-yet time.

So also such contractions as (kaant, wount) presuppose
can-notf will-not.

1899. When sentence-words and sentence-groups stand
in the unemphatic end-position, they undergo enclitic
stress- weakening : are you going jhome, :John ? ['John^ are
you going 'home .''] yeSy-sir \ sohie breads -please I \ it is late,

'isn't -it? I'zsn't it 'late ?\

1900. If three strong-stressed words come together —
especially in immediate succession, but also with intervening
weak-stressed words — the stress of the middle word is often
reduced, especially if it is a monosyllable, as in English .-plum-
'pudding compared with 'plum'pudding \ five .'minutes to nine
[^five -minutes] \ a great :big man \ a heavy .'round stone \ a
little :more room \ all :day long \ two:poundten \ a hard. 'day's

§ 1905.] SENTENCE-STRESS, 33

work. But this happens only when the middle word can be
regarded as subordinate to the preceding word or as being
specially connected with it, not when they are all of equal
weight and independence, as in a 'thick -gold 'chain \ it is
very hot in here with 'that 'big 'fire \ 'ten 'years a'go.

Hence also the distinction that might be made between he
'failed com'pletely to ex' plain his meaning and he 'failed
com.'pletely to ex'plain his meaning. In the former the adverb
modifies /<a!27<?^; in the latter it modifies explain^ with which it
is felt to form one stress-group.

1901. Rapid speech is unfavourable to even stress. If
two even-stressed syllables come together at the end of
a sentence, there is a tendency in rapid speech to throw the
stress forward, so that such a word as *Chi'nese becomes
:Chi'nesey unless there is some special reason for throwing
the stress on to the first syllable.

1902. In the colloquial combination nice and-\- adjective —
where the nice is practically a vague 'very' — the second
adjective takes the principal stress : the roads are mice and
*dry. Compare :cup and 'saucer (928).

1903. Combinations of verbs of full meaning with other
words may have even stress : to 'change ones 'mind \ it rains
hard \ he came running,

1904. But when a verb is intimately connected with its
object- word or object-complement etc., so as to form a kind
of compound with it, the stress of the verb is often subor-
dinated to that of its modifier : to dight a 'fire \ open the door I \
to run a race \ to feel too full.

But shall we have thejfire .'lighted t by grouping-stress.

1905. So also when a verb is followed by a preposition-
group with which it is intimately connected : to:gofor a 'walk \
to fly into a passion \ to cut it in two. Similarly in such com-
binations as :go -out 'hunting. But the verb may also take
the stress : he stood behind the door.

VOL. II. D

34 SYNTAX. [§1906

^1006. When a verb is intimately connected with another
verb or verbal, it tends to subordinate its stress, especially in
the combination verb + and •\- verb: shall I :go and look for
him ? I coine and try ! \ I heard him 'go \ did you see him
'Start ? I we have come to 'stay.

1907. But in the combination verb + adverb there is often
even stress, even when the two are intimately connected : to
'lie 'down on the 'sofa \ dressed up like a Turk \ put it down I \
my watch has not been wound up [wind -up a watch, § 1900].
Note the distinction between to carry about and to cry -about,

1908. In some cases, however, the adverb has the stress :
to see him 'off [but to 'go -off'^ \ to run a'way. Such sentences
as 'come 'in / may also be pronounced :come in / to express
impatience.

When the adverb is followed by a preposition-group, the
adverb necessarily takes grouping- stress : to kick joui at
him I when do you get jup in the morning f

1909. Prepositions of definite and marked meaning may
have full stress, as in be'hind the 'door \ since then \ he is a'bout
'my 'height \ without light we cannot distinguish colours^ while
those of indefinite and abstract meaning, such as at, of on,
are generally subordinated to other words, as in -at the door \
in life compared with 'through 'life \ -from, 'under the 'table.
Hence a preposition may have different stress according to
its meaning, as in -under an obligation compared with 'under
•age.

1910. But in combination with pronouns all prepositions
may take the chief stress, as in -will you :go with -him ? com-
pared with -will you go -with that man ? \ :run 'after -him,
with grouping-stress. Even of may take the stress, as in
niuhat -of -it!, where it is preceded by a word of subordinate
stress.

1911. If, however, a strongly stressed word precedes, the
preposition loses its stress as Well as the pronoun : / will
think -of -it I do not talk about it!

§ 1915.] SENTENCE-STRESS. 35

1912. In some cases differences of stress — which are
often further accompanied by changes of sound — give rise
to doublets with special divergencies of meaning and
function. The rules which follow are of course liable to be
crossed by the general principles of emphasis already laid
down.

1913. this and that generally have strong stress both as
nouns and as adjectives : '/hts 'shows -that . .\is that you ? \
these young men. But if the demonstrative meaning is not
strong, they may have medium stress, as in :th{s hook has not
had its leaves cut compared with 'this book is one of those
1 want. So also when they are used as nouns in the sense
of * it * without any special demonstrative meaning : / saw
:that (or -that) quite clearly.

The adverb (^at) has of course only weak stress. So also the
relative pronoun (^3t).

1914. So also the corresponding adverbs here and there
generally have strong stress : here is your money ! \ there he
goes I I come here! But in the case of these words also,
weakened demonstrative meaning is accompanied by
weakening of stress, as in did you read as far as page ten ? we
left off 'there compared with we left off 'there, which would
imply pointing to the place. After verbs their stress is
often weakened : / saw him -there yesterday \ the sooner we
get -there the better!

In all the above cases there keeps the sound (^ea) : the weak
(^ar) has of course only weak stress.

1915. The interrogative what and which have weak or
medium stress when used as nouns, strong stress when used
as adjectives, as in -which 'boy did it P \ at what time? com-
pared with -which 'is it? \ -what is the time? The interroga-
tive who is used only as a noun, and has weak stress : -who

is -it? what and which when used as adjectives have strong

D 2

36 SYNTAX. [§1916.

Stress in all their other functions : submit with what grace he
can I which philosopher (relative).

The exclamation what/ has strong stress, as might be
expected.

1916. Interrogative adverbs have weak or medium
stress : where does he live ? \ when did he come ? \ why did he
do it?

1917. But in repeated questions all these interrogative
words take extra stress : jwhat did you say his name was?\
;where does he live ?

1918. Relative noun-pronouns and adverbs have weak
stress : a man -who could do great things \ the place where he
lives. When used as conjunctives these words also have
weak stress : / know -who he is \ doesiit he know -where to look
for it? (note that this sentence is negative in form but not in
meaning) | / see how it is done. But if the preceding clause
has a negative meaning, they have extra strong stress : / do
not know ;who he is \ I have not the slightest idea what he
means \ he will not know where to look for it | / cannot think
how it is done.

1919. enough has strong stress except when it follows the
word it modifies : enough breads enough of this but bread
-enough^ good -enough.

1920. some with strong stress is a quantitative word : with
'some 'difficulty \ some .'people think so. In the partitive
meaning * a little ' it is weakened into (sam) with weak stress :
-some water \ -some more bread.

1921. The weak forms of the auxiliaries occur only
before the nucleus (120) and — in the case of be — before the
predicate : compare / shall ( Jl) go with / think I shall (-Jael) |
he is (hijz) a good fellow with what a good fellow he is (-iz) !
In other constructions they often have strong stress, especially
be : and 'who 'is -he ? \ I wonder -where -he 'is \ -is -he ready ?
I expect he 'is.

§1926.] INTONATION. 37

1922. -havers' must/ -have (haev, 3ev)=' cause, have/ the
weak (h9v, av) being used only as an auxiliary, and occasion-
ally in the sense of ' possess ' : you will 'have to -have (hsev)
your hair cut \ to -have a party \ he -has a cold.

1923. must has strong stress, except in such phrases as
you -must (mas) know that . . , where it is almost unmeaning.

Note that compulsion is expressed by strong stress — / -must
'say I agree with him in that opinion — logical necessity by
extra stress : you ;must :know what I mean /

1924. The weak-stressed auxiliaries are all capable of
receiving emphatic stress: they said nothing, why jshould
they have said anything ? \ I ; should like you to see it I \ why
iwtll you persist in denying it? \ we ;can but try.

Intonation.

1925. The tones have both an emotional and a logical
significance.

The level tone is plaintive — especially the high level tone
— ^and suggests the idea 'of suspensiveness. The level tone
is not much used in English, where it has much the same
function as the rising tone.

The rising tone is associated with cheerfulness, animation,
surprise, expectation, hesitation, and suggests suspensiveness,
incompletion, and question.

The falling tone is the natural expression of dogmatism,
resolution, command, and suggests the ideas of completion,
finality, certainty, and of answer as opposed to question.

For the compound tones see § 1950 foil.

1926. Hence declarative sentences normally end with a
falling tone, the preceding portions of the sentence being
uttered with rising or level tones to show that the meaning is
not yet complete. l^hMsyou are quite' ox you are quite makes
us expect some word or words ending in a falling tone to
complete the meaning : you are quite right^. If, on the other
hand, the predicate comes first, it is uttered with a rising tone

38 SYNTAX. [§ 1937.

to show that the subject is being waited for to complete the
sentence : greai^ is Diana of the Ephesians^,

1927. Hence also general interrogative sentences have a
rising, special interrogative sentences a falling tone (503, 4) :
is he here' ? \ where is he'^? But special interrogative sen-
tences have a rising tone — together with extra stress on the
interrogative word (1017) — when the question is a repeated
one : ;mhat is his name — what did you say his name was'?

For the intonation of alternative questions — as in is he an
Oxford or a Cambridge man ? — see § 505.

1928. When the hearer of a statement repeats the whole
or part of it, he utters it with a rising tone, to show that he
is expecting confirmation : my name is Smith"". Smith': then
you are the man I am looking for^.

1929. Surprise or indignation is expressed in a single
word by a long rise, as in whatf /, in groups and sentences by
a fall, exclamative sentences (502) being regarded as em-
phatic affirmative ones : good heavens^ ! \ how well he looks'^ /

1930. Command is, of course, expressed by a fall.

1931. But statements and commands are often uttered
with a rising instead of a falling tone to mark them as
questionable or hesitating : is it fine'? yes', it is pretty finef
[compare ^^j^, // looks quite set/led^'].

1932. The idea of question thus implied is often used to
suggest that of appeal or remonstrance : / have done all
X can' = * I have done all I can, haven't I ? ' | / wish you
would let me alon/. So also all rights may imply * why don't
you start ? ' or ' we are waiting for you to start.'

1933. The rising tone also serves to soften a contradiction,
as m your friend is late, it is not late': it is only three 0^ clock! ^
a command, whether direct, as in don't forget to post that
letter'^ or implied, as in now you'll remember what I have said'y
or a refusal, as in will you have another cup of tea ? no' y thank
you'.

1934. Lastly, the rising tone often serves merely to give

§i94i] INTONATION. 39

a general character of cheerfulness or geniality to what is
said : well', good by/ ; hope to see you again soon\

1936. But the falling tone can be used in all the
examples given in the last three paragraphs, and would,
indeed, be preferred by many speakers, especially those who
wish to show a firm and decided character.

1936. The brevity and imperativeness of special interroga-
tive sentences such as what is his name'' ? is often avoided by
substituting a longer general interrogative form : can you tell
me what his name is' ?

1937. If a general interrogative sentence «is uttered with
a falling instead of a rising tone, it expresses command or
impatience ; will you do as you are told> I \ ;are you ready"" ?
In such cases the auxiliary often takes emphatic stress (1896).

1938. When a negative interrogative sentence is used
rhetorically to express affirmation, it necessarily takes the
falling tone : isn't it wonderful^ 1 \ he is very egotistical. yeSy
isn't he"" 1

1939. Sentence - intonation is generally continuous,
abrupt transitions from rising to falling tone and vice-versa
being avoided as much as possible.

1940. In such a sentence as it is fine, frosty weathef^ the
falling tone begins, not on the last syllable, but on the
preceding stressed one, the fall being continued downward
through the last syllable. If an enclitic word such as sir
were added, the fall would still begin on the first syllable of
weather, and would be continued through the enclitic. So
also in it ivill be fine'^ to-morrow, I hope. In the same way all
other modifiers of the predicate are subordinated to it in
tone, even if they have full stress: he is the most obstinate
child> I have ever had to deal with.

1941. If a complete sentence has a full-stressed tag added,
the tag is uttered with a separate fall of its own, instead of
merely continuing the preceding one: he is stupid", very^
stupid. Here the voice, after reaching a low pitch at the end

40 SYNTAX. [§ 194a.

of the first fall, leaps up and begins a fresh fall. Even if the
original sentence is not quite finished, it may have separate
falling stress in such examples as he is a iall>, a very tal1>
gentleman.

1942. Enclitic additions to a complete rising-tone sentence
simply continue the rise: are you read/y gentlemen? \ will

you make a little room\ please sir ?

1943. The following are examples of independent rises :
gentlemen^ are you ready' ? | are you ready' — all' of you ?

1944. When two or more clauses are intimately connected,
there is no tone-break till the end of the whole group : tell us
who your new friend"" is I \ it would be better if you were to do^
it yourself So also in such combinations as more haste' ^
worse speed>. This is of course still more necessary when
there is contraction, or in extended sentences (486, 8). In
sequences (482) — where there is no formal connexion — each
sentence keeps its own independent tone — whether rising or
falling — unless a rising tone is required to show the con-
nexion more clearly, as in / am sorry I could not come
before^ ; I had to finish writing a letter^.

1945. When the first clause introduces a statement etc.,
it takes rising tone if it is grammatically unfinished : he said'
he did not care^ \ the difficulty 'is', how are we to get back^ ?
Otherwise it takes the falling tone : what he said was this"" . .
I he speaks somewhat in this way^ . .

1946. Inserted or parenthetic groups or clauses (467)
naturally have rising tones : he is a man who', if he chose' ^
might do great things^.

1947. But, on the principle of the continuity of stress, if
the whole group shows a marked falling tendency, the
inserted words follow it.

1948. If the appended or inserted words have a marked
meaning of their own dependent on their intonation, that
intonation is kept, which often results in broken intonation :
which will you have^^ tea' or coffee^ ? When a tag keeps its

§1952.] INTONATION. 41

independent intonation (together with its full stress), it is
either emphatic, or else is felt as detached — as if it were
added with hesitation : which will you have'' ? tea>y please
[iea^y -please'] \ there is the bell again> : it is Frederick^, of
course y isn'f it? I will call to-morrow^ ^ if I can [. . to-morrow^ ^
if I -can\,

1949. Broken intonation is, of course, less frequent in the
case of insertion, as insertions are generally not emphatic:
t but thou\ if thou shouldest never see my face again^^ pray for
my soul> !

1960. In the compound tones the second element deter-
mines the general meaning of the whole tone, and the first
element only modifies this general meaning. These tones
are always accompanied by extra stress, because of their
emphatic meaning.

1951. The compound rise expresses doubt of some
implied statement, so that it expresses distrust, caution,
warning: "^ I will not try it; "^you may \ take ^ caret if you
"^do ity it will be at your own "^ risk. Hence it is used in
cautious contradiction: but for all ^ that ^ he is an exceptior^.
Also in contradiction or modification of the speaker's own
statements : / am sure he will come again"" : at least I ^ think
he will\ In its more logical uses it expresses contrast or
exception : / am what the world calls a woman-hater^ ; what
'^ I call a philosopher'^ \ the dinner was very good>^ wasiif^ it?
the ^ wine was bad^. It sometimes has an intensive meaning.
It sometimes contradicts the meaning of the word it falls on:
was it raining when you came in' ? '^ rather I [rather' would
imply ' only a little '].

1952. In the compound fall the relations between the
two elements are reversed. This tone hints at a doubt, and
disposes of it by a dogmatic assertion. Hence it expresses
contempt or sarcasm: ^/ can do if. you! It also ex-
presses remonstrance, contradiction, contrast, not cautiously,

42 SYNTAX. [§1953-

as the compound rise does, but confidently and dogmatically :
what\ are you going already ? you have only just ^ come /
I Sunday isn't the da/; it is ^ Monday I \ you say you are
sure of finishing it'; hut ^ when will you finish if" / — that is
the question"". Sometimes the dogmatic element disappearsj
and the intonation has simply an emphatic effect, with
perhaps a trace of impatience or contempt : you ought to have
done something to prevent it, what could I ^do? | shall we
have time'? '^oh^yes.

NOUNa

Gender.

106S. In Modem English the only gender-concord of
nouns is with some of the personal pronouns ; and accord-
ingly, the only general test of noun-gender is its association
with he^ she, it,

1964. Modern English has lost all traces of the gram-
matical genders of Old-English. Nevertheless, the genders
are not entirely natural.

1956. Words denoting the young of men and animals
without implying any special sex are often neuter — less
frequently however in the case of human beings: a simple
child , , what should it know of death ? \ the other twin fell
with its face in a furze-bush.

1956. it is sometimes applied to human beings to express
contempt : what a silly fellow it is/

1967. Names of animals when used without any personal
feeling towards individual animals are generally neuter. The
lower the animal in the scale, the more exclusively the neuter
is used.

1958. But in the spoken language there is a tendency to
give a personal gender to the higher animals without regard
to the sex of the individual animal. Thus dog, horse, fish.

§ 1963.] NOUNS : GENDER. 43

canary are generally masculine, cat^ hare, parrot are
generally feminine.

1959. Names of things are often personified in colloquial
speech. Such words as ship^ boat, balloon, steam-engine and
names of machines are made feminine, especially by those
who are constantly employed with such things, but also —
especially in the case of words for ships and boats — by the
majority of speakers, although it is always allowable to use
the neuter. This personification seems to have arisen from
a fanciful comparison with wife,

1960. The names of smaller objects arc made masculine,
such as watch, pipe (to smoke with). Some words, such as
kite (the plaything), may be either masculine or feminine.
But this kind of personification is less general than the
former, and is left more to the caprice of the speaker.

1961. It is to be observed that in the spoken language only
artificial objects are personified. It probably began with
such words as ship^ denoting objects capable of independent
movement, and therefore apparently alive.

1962. In the literary language we have a totally different
kind of personification. That this literary personification is
due partly to Latin influence is shown by the fact that it
makes moon feminine and sun masculine in direct opposition
to Old-English usage. But the usage of modern poets, though
partly founded on Latin and French traditions, is mainly
independent, and the same word may have different personal
genders assigned to it by different writers, while another
writer may, even in poetry, keep it in the neuter gender.

1963. The general principle is to give the masculine
gender to words suggesting such ideas as strength, fierceness,
terror, while the feminine gender is associated with the oppo-
site ideas of gentleness, delicacy, beauty, together with fertility.
Thus sun, summer, time, winter, death, rage, war are mascu-
line, and moon, spring (the season), dawn, mercy, peace, earth
are feminine.

44 SYNTAX. [§1964.

1964. The predominance of the feminine gender in Latin
abstract nouns has made the feminine the abstract gender in
Modern English also, even where there are otherwise no
specially feminine associations, as in truth, justice, silence.

1965. In Latin most names of countries are feminine.
Hence in Modern English not only poets but also newspaper-
writers make such words as Asia, Britain, France, Germany
feminine.

Number.

Singular Individual Noun = Collective Noun.

1966. By the analogy of the old unchanged plurals such as
sheep, a large number of names of animals have come to bfe
used in a collective sense without any plural inflection, as in
how many fish have you caught? two salmon and three trout.
The grammatical difference between three trout and three
sheep is that trout is formally singular, while sheep is, or may
be, formally a plural, for there is a plural trouts, although it
is not much used.

1967. This usage is confined to the names of wild animals.
Thusyi>Z£;/ if used collectively must be put in the plural, as in
to keep fowls = to keep poultry, where the individual plural

fowls is logically equivalent to the collective singular poultry ^
while wildfowl is regularly used as a collective without any
inflection : to shoot wildfowl. So also duck in to shoot duck
would imply that they were wild ducks.

The occasional use of the collective chicken for chickens {to
keep chicken) seems to be the result of the ending -en having
been regarded as a plural inflection; whence also the new
singular chick.

1968. It is also to be observed that these collective
singulars are used only when the animals are hunted because
of their usefulness to man, or are taken in considerable
numbers, but not when they are killed only in self-defence or

§1971.] NOUNS: NUMBER. 45

as vermin. Hence the strict adherence to this construction
in the case of fish-names [not, however, with eel. lobster and
some others], while the word/j^ itself, being more vague, is
often used in the plural, as in to catch three fishes. But if
fish is used in the special sense of salmon^ the singular form
is necessary : he killed three big fish.

1969. Hence, again, while this usage is freely extended
to unfamiliar foreign animals, as in a few antelope[s\ herds of
buffalo{es) and giraffe{s), to hunt pig (implying wild boars), it
is never used with such words as lion, wolf badger^ weasel)
but it is admissible with bear, because this animal is hunted
for its flesh.

1970. The use of a singular individual noun in a collective
sense is also found in other classes of words : the cannon(s)
were firing shot and shell — rings were used as a substitute for
coin(s) I written in blank verse{s) [but to write Latin verses\
In the last two examples the singular seems to be the result
of the analogy of the collective nouns cash (or currency), prose
{prose and verse). Other examples are : small craft = * small
ships,' horse and foot = * cavalry and infantry,' a fleet of twenty
sail, where sail = * ships.' These differ from the former in
using the noun in a special sense as well as collectively;
hence they cannot take the plural inflection without returning
to their original meaning, as in twenty sails compared with
twenty sail.

1971. In all the above examples the noun has a distinct
individuality. But when we talk of the hair of the head, we
do not think of each individual hair, but rather regard hair as
a material noun, like iron, so that we really have two distinct
words, ia) the material noun hair with no plural of its own
{two heads of hair), and {b) the individual noun hair ' filament
of hair ' with plural hairs : she has more hair (collective) than
wit, and more faults than hairs (Shakespeare). So also in to
plant maple{s) diXid forests of pine the singular seems to express
the idea of a mass of trees. We have the same fluctuation in to

46 SYNTAX. [§ 197a.

make brick{s) and tiles \ coals are (or coal t's) cheaper than
wood. So also va fruit and cakes, the ^\m2\ fruits (fruits of
the eartfi) suggesting the idea of various kinds of fruit rather
than that of a mass or plateful of fruit.

Collective Nouns with Plural Constructions.

1972. Collective nouns can always be regarded as logically
equivalent to plurals of individual nouns, and hence are often
ungrammatically associated with words that imply plurality :
twenty people, these vermin, many cattk, a few cavalry, the clergy
took their seats, two or three counsel who were never in any
cause (Dickens). We also find singular constructions : much
cattle, much people (Bible).

1073. There is often hesitation in joining such words to
numerals. This is especially the case when there are distinct
individual nouns by the side of the collective ones {clergyman
by the side of clergy), as in twenty clergy walking in pro-
cession ; here the collective is preferred because it implies that
it was not a fortuitous assemblage of clergymen, but that they
walked in the procession through being members of one
organization.

1974. The inconvenience of not having an individual noun
corresponding to such collectives as cattle and game has led
to the use of head ^ . . , as in « head of cattle, twenty head of
game, where head is the unchanged plural of measure.

The old collective /?/>& (Old-English neuter yi?/^ * nation ') was
at first used in the plural only in the sense of peoples— that is
'nations'— but in Modern English folks came to be used in the
sense of persons, as in the oldfolk{s) at home, folk{s) say. The
word is now almost obsolete.

Changes of Meaning in Plural.

1975. Some classes of words are rarely used in the plural
unless they at the same time undergo some change of
meaning.

§i98o.J NOUNS: NUMBER. 47

1976. Proper names can be used in the plural in their
literal meaning only when they have been applied to more
than one individual [the Tarquins)^ although they can always
have a plural when used figuratively : the Drydens of the
present day.

1977. Material nouns in their literal meaning are used in
the plural to denote difference of kind or quality : French
wines. But they are also used in the plural with change of
meaning to imply individual nouns denoting some definite
object, as in the leads of a house meaning * sheets of lead/
irons meaning * fetters of iron/ So also colours = * flag '
implies a combination of different colours. In such expres-
sions as sands = ' sandy beach/ the waters of the Nile, ^the
dews of heaven there is no suggestion of definite parts, and
the plural seems to suggest indefinite extension or repetition.

1978. Abstract nouns are not often used in the plural in
their literal meaning, although, of course, there is nothing to
prevent them from being so used when necessary : a man of
abilities, a thousand pities. But they often undergo changes
of meaning when so used, as in pay respects, do the honours,
in good spirits. They sometimes even become concrete, as
in spirits of wine, forces = * army,' effects, sweepings = * result
of sweeping, what is swept w^^ filings.

Used only in the Plural.

1979. The meaning of some words makes them more
used in the plural than in the singular. Such a word as
twins, indeed, hardly admits of a singular, although we do
not scruple to use twin in the sense of 'one of a pair of
twins.'

1980. There are other words which are used in the plural
to imply that they are made up of a pair of parts, such as
scissors, bellows, spectacles (also called glasses). These words
never drop the plural inflection. The resulting difBculties
are evaded by the paraphrase pair of . .: a pair of bellows^

48 SYNTAX. [§ 1981.

two pairs 0/ spectacles. When such plurals imply more than
two component parts, as in gallows, this periphrasis cannot
be used.

1981. Besides those plurals which cannot drop the plural
inflection without losing their distinctive meaning, such as
leads and spectacles^ there are many words used only in the
plural, without having any corresponding word in the singular,
such as hustings, sessions, mathematics, phonetics and the
other words in -ics, some of which, however, are now occa-
sionally used in the singular form, such as meiric(s), lecturer
on diplomatic.

1982. Most of these plurals are — occasionally at least —
used in singular constructions. Even with pair-plurals we
find such constructions as a silver scissors. The use of the
indefinite article is quite common with the other words :
a gallows, a hustings. So also we can say phonetics is . . as
well as phonetics are . .

1983. So also some of these words can be used in definite
plural constructions : two gallows, two hustings.

1984. In some cases when a plural noun has singular
meaning it is converted into a singular noun. The test of
complete conversion is that it can take a fresh plural inflec-
tion : two sixpences. In such examples as what is the news ? \
to take much pains there is only half conversion.

For apparent plurals see § 998 a.

Cases.

Common Case.

1985. The common case is used to express the subject,
the direct and indirect object, and the vocative relation :
John I, that man gave your brother a book. In Old-English
the first two nouns in this sentence would be in the nomina-
tive, the third in the dative, the fourth in the accusative. In

§1989.] NOUNS: CASES. 49

combination with prepositions it represents historically the
Old-English accusative, dative, instrumental, or genitive, and
has in Modern English a neutral function, being regarded
neither as a direct nor an indirect object, but rather as stand-
ing in a non-subject relation, or — more definitely — as being
converted into an adverb.

Common case = dt'rgcf object.

1086. Besides its purely accusatival function, the common
case is used to express

direotion : they went the same way.
space : three inches long, one storey high.
adverbial relations : he would be satisfied one way or
another,

1987. It also has in the spoken language a purely adjec-
tival function : it is no use \ you would not think those two
children were exactly the same age \ what colour shall I paint
your door ? [compare paint it white'] \ ivaier the colour of pea-
soup. This construction apparently began through dropping
a preposition, for we can still say in the above examples of
no use, of the same age, with what colour. It may then have
been extended to other constructions, such as he looks his
age.

Common case = indirect object.

1088. The dative is extensively used in Old-English not
only in company with accusatives after verbs of giving, etc.,
but also with intransitive verbs, not only verbs of addressing —
as in he him andwyrde ' he answered him ' — but also of
benefiting, helping, injuring and other personal relations, as
in mannum dorian ' to injure men/

1989. In Modern English the accusative and dative have
been merged into one case, and so the dative after originally
intransitive verbs has become indistinguishable from the old
accusative, and has come to be felt as a direct rather than

VOL. IL E

50 SYNTAX, [§ 1990.

an indirect object, as in he answered him, he answered the
question.

1990. In fact, in Modern English the common case cannot
be recognized as an indirect object unless accompanied by
another common case serving as a direct object, from which it
is distinguished solely by position (1823), so that the relation
is always liable to become ambiguous when it is not accom-
panied by an accusatival word, or when removed from its
normal position. In such cases the ' prepositional dative *
with to is substituted, as in / will write to him about it com-
pared with / will write him a letter \ to the devout believer the
Church promised pardons instead of the Church promised the
devout believer pardons.

1991. But in many cases the context enables us to recog-
nize an isolated or displaced common case as an indirect
object on purely logical, not formal, grounds, as in the
archaic woe is me ! \ me, poor man, my library was dukedom
large enough (Shakespeare). The unmistakably datival func-
tion of the pronoun in / will write you is probably the cause
of this construction — which occurs throughout the Modern
English period — being avoided by many who would not
object to it if it were possible to regard it as a direct object.

1992. So also the common case after adjectives and
adverbs of nearness and likeness — which is a direct descen-
dant of an Old-English dative — is still felt rather as an
indirect than as a direct object, as shown by the frequent
substitution of the prepositional dative : he sat near the fire
[Old-English: neah pcem fyre\ \ to wear flannel next the skin
[Old -English : niehstpcBre hyde\ \ unlike one another compared
with come nearer to the fire I \ I sat next {to) him at dinner \
like to or "^unto . .

1993. But, on the other hand, the adjectives and adverbs
of nearness are felt to be almost or quite identical with pre-
positions, near the fire, next him having the same meaning as
by the fire, beside him (or by his side), and this conception

§ 1996.] NOUNS : CASES, 5I

may be extended to like, although here there is no analogous
preposition.

1894. Even when two common-case words come together,
the construction is sometimes doubtful.

Thus in Old-English, verbs of teaching originally took
two accusatives, as in he Ictrde hine crceftas ' he taught him
arts.* In Modern English we can isolate the first object-
word {he taught him)^ which therefore ought to be regarded
as a direct object ; but, on the other hand, such transpositions
as ito him my tale I teach (Coleridge) show that we regard it
as an indirect object — at least, when associated with another
object-word. That this feeling is historically correct is shown
by the fact that in the later Old-English a dative of the
person is generally substituted for the original accusative : he
lardt him crce/tas. In he struck him a blow \ he struck the
desk a blow we do not feel the first object-word to be datival,
as we would in he gave him a blow^ but rather as a direct
object followed by what in inflected languages would be
a cognate accusative. In such constructions as /*/ will last
him his life-time \ to stare one in the face the datival character
of the first object-word is more marked, because neither of
the verbs is transitive, as in the two preceding examples.

1995. It is to be observed that many verbs take to just
as they might take any other preposition, not as a substitute
for the dative. Thus the to in speak to is derived from the
Old-English construction he cwcep to him, he sprcec to him
contrasted with the dative in he him andwyrde. But it was
of course through the analogy of such constructions that to
came to be used as a substitute for the older dative. The
non-datival is distinguished from the datival to by its never
allowing the substitution of the common case.

Genitive.

1996. In Old-English any noun could be made into an
adjunct-word by being put in the genitive ; but in Modern

£ 2

52 SYNTAX, [§

1997.

English the genitive is restricted mainly to nouns denoting
living beings, its place being otherwise taken by the 'pre-
positional genitive,' formed by the preposition of, as in at
the foot of a mountain compared with a maris foot.

1997. The introduction of the prepositional genitive was
no doubt prepared by such Old-English constructions as pa
m§nn of Lundenbyrig fetodon J?d siipu 'the men of (coming
from) London fetched the ships.'

1998. In Old-English the genitive was also used, like an
accusative, to denote the object of a verb such as gemunan
* remember,' giernan ' desire,' often with an accompanying
accusative or dative of the person : htm wcbs oftogen ctltes

fodan * they were deprived of all food.' This survives only
in the form of constructions with the prepositional genitive
(think of deprive of), which may however in these instances
have developed independently of Old-English traditions.

1999. The non-personal genitive is still frequent with
nouns of time, as in a days journey, a minutes notice, one
week's pay, and of space, as in he arrived at his journey's
end, a spear's length, and in a few isolated constructions such
as the sun's rays, the waters' meet, which almost constitute
compounds. Otherwise only in the higher style: Albion's
sons, an empire's dust, music's voice.

2000. The use of the personal genitive itself is limited.
Thus the genitive in combination with transitive phenomenon-
words is generally used only subjectively, not objectively —
that is, a mother s love, a father s care imply that the love and
care proceed from the parents, not that they are the objects
of it. The objective relation is expressed by the pre-
positional genitive, as in the love of God compared with
Gods love.

2001. In many instances both genitives may be used, the
prepositional genitive being sometimes preferred in order to
avoid awkward or ambiguous collocations, especially when
a genitive-group is preceded by other modifiers.

§2oo8.] NOUNS: CASES. 53

2002. When adjective + noun is preceded by general
adjectives, the latter necessarily modify, not the adjective,
but the noun — or, rather, the whole group. But if genitive +
noun is preceded by general adjectives, the latter are neces-
sarily associated with the genitive rather than the head-noun,
as in fhe king's son = i/ie son of the king compared with the
royal youth.

2003. Of course, if the genitive is so closely associated
with the head-noun as to form a sort of compound with
it (893), then the modifier is associated with the head-
word or the whole group, as in two lady^ s-maids^ a law-
yer's clerk, that butchers shop (:t53et butjaz /op). In the
last example the that may be associated with the geni-
tive, but in this case the stress is thrown on to it; (tJaet
ibutj^z Jop).

2004. It is also to be noted that the genitive seems to be
less freely used in the plural than in the singular, because
the genitive plural is indistinguishable in most words from
the common plural ; thus the meeting of the lovers is clearer
than the lovers^ meeting.

2005. The use of the two forms is sometimes a matter of
emphasis. Thus the answer to the question is he of good
family ? might be he is the son of a baronet, while if the
baronet had just been mentioned, we should speak of his son
as the baronefs (eldest) son.

2006. But there are many cases in which only the genitive
is admissible. Thus in such a sentence as where is fohns
hat? we could no more substitute of fohn than we could
substitute of me for my.

2007. Genitives can be used absolutely without any
prop-word: he neglects his own business to look after other
peoples.

2008. Hence has arisen the elliptical genitive, which
requires some such word as * house ' to be supplied, not from
the context — which would give simply an absolute genitive —

54 SYNTAX. [§ 2009.

but from the sense : «/ /he baker's {shop) \ to dine at JBrookes's
{club).

2009. The predicative use of the (absolute) genitive
is hardly colloquial: his purse and his heart were every-
body'Sy and his friends* as much as his own \ Heavens is the
quarrel/

2010. The pleonastic genitive, as in he is a friend 0/ my
brother* Sy is generally partitive = * one of the friends of my
brother*; he is a friend of my brother v/ould imply 'he is
friendly towards my brother.' But the construction is extended
to non-partitive instances, as in this fortune of my poor sisters
is old Radford's object. The main reason for keeping it up
is to avoid the prepositional genitive of personal nouns.
This applies still more to the parallel construction a friend
of mine (2114).

ARTICLES.

2011. The definite and indefinite articles are both mark-
words — they single out an individual (229). Hence they are
most used with concrete class-nouns. The absence of the
articles thus becomes a distinguishing feature of other classes
of nouns, such as proper names. The absence of the articles
has also more purely grammatical functions, as when it is
used to mark the vocative relation. We may of course
reverse the statement by saying that the presence of the
articles has the negative function of showing that the word
is not a proper name or a vocative.

2012. It follows from the definition of proper names and
material nouns 4;hat they are from a strictly logical point of
view incapable of being used in the plural. Hence the mere
fact of a noun being used in the plural is enough, as a
general rule, to stamp it as a class-noun ; so that, if a noun
in the plural has not the definite article, we naturally assume
it to be the plural of an undefined class-noun — that is, a

§aoi6.] ARTICLES: DEFINITE ARTICLE. 55

class-noun accompanied by the indefinite article. Hence
a noun in the plural without any article corresponds gram-
matically to a noun in the singular with an indefinite article,
not to a noun without an article: men is the plural not of
man but of a man.

Definite Article.

For its position — the whole day^ all the day — see § 1792.

2013. In Old-English the definite article se still has also
the function of a demonstrative, and in Modern English its
neuter kcet has become the purely demonstrative that, which
is now completely disassociated from the article.

Referring Back.

2014. In its purely grammatical function the definite
article is put before a noun to show that the idea expressed
by the noun has been already stated, and to refer back to
that statement. If, on the other hand, the idea is new, the
noun expressing it is accompanied by the indefinite article.
Thus in the fable of the wolf and the dog, the two animals
are introduced at first as a wolf, a dog, and are then spoken
of as the wolf, the dog : one night a wolf fell in with a dog.
the wolf was all skin and bones, while the dog was as fat as
he could be.

2016. The noun referred to is not necessarily marked by
the indefinite article ; thus the men may refer to some men or
three men. The noun referred to may also be a different
word altogether ; thus the man may refer to your friend or to
a proper name.

2016. In some cases the noun referred to — especially if
it is a material noun or a proper name — may be without any
assumptive : when earth is washed into a river by the rain,
the earth sinks to the bottom of the river and becomes mud.
It is to be noted that the meaning of the material noun is in.

56 SYNTAX. [§

2017.

this instance not modified in any way by the article, as it is
in such a sentence as the earth is round.

2017. In Old-English the repetition of a proper name may
be marked in the same way : he wolde ddr&fan dnne CBpelingy
se wcBS Cyneheard hdten^ and se Cyneheard wees pcBS Sige-
bryhtes brdpor * he wished to expel a noble who was called
Cyneheard, and (this) Cyneheard was the brother of (the above-
mentioned) Sigebryht/ In Modern English we have given
up this usage, probably because of the ambiguity that would
arise from such collocations as the baker {the Baker) — a more
serious ambiguity than that which arises from the use of the
definite article with material nouns {the earth). Hence if we
wish expressly to mark the repetition of a proper name,
we use some other demonstrative, or insert some adjective
{this, the above-mentioned).

2018. When used for back-reference, the definite article
requires another (preceding) sentence to complete the meaning.
In all the following usages it is merely a word-modifier.

Identifying.

2019. In the sentences given in § 2016 the in the river has
only the grammatical function of referring back to a river ^
which, so far from denoting one particular river, practically
expresses the idea * river in general.' But when London
people talk of the river, meaning * the most important river
near us,' the definite article identifies the river almost as
unmistakably as the proper name Thames does. So also
the king, the lord-mayor, the street may imply a proper name.
Similarly, the door, the window implying the door or window
of the room the speaker is in or is thinking of, and the horses
means * our two or more horses.'

2020. In have they a father ? no, the father is dead, but the
mother is alive the definite article does not refer back to the
preceding a father, but is purely identifying, being equivalent
to ' their.' In some cases the definite article may be regarded

§ ao25.] ARTICLES : DEFINITE ARTICLE. 57

as a direct substitute for a possessive pronoun : these animals
have the tail tipped with hair | shall I take a little off the beard,
sir ? (said by a hair-cutter).

2021. With collective plurals such as birds, trees, stars the
definite article has almost a demonstrative meaning. With-
out the article such a plural as birds expresses the idea of
* birds in general,' while the birds implies * the birds around us
or near us/ as in the birds are singing \ the stars are bright
to-night compared with birds do not sing in the winter \ stars
shine by night.

2022. Although we use the indefinite article in such
phrases as a story about a wolf and a dog, we substitute the
definite article in the title of the story— />i^ wolf and the dog —
meaning * the well-known story ' or ' the story you are about
to hear.'

2023. The presence of another modifier need not affect
the meaning of the identifying article. Thus the old horse
may mean ' our single old horse.* In this case the adjective
is purely descriptive. But this collocation may also imply
that the speaker has two horses, and that he adds old to show
which of the two he means (32). So also in the front door,
the hall door, the man who was here yesterday. In this case,
therefore, the article and the adjective share the function of
identifying between them, while in the meaning first given it
is the article alone that identifies.

2024. If we define the queen, the river as the reigning queen j
the nearest river ^ the article becomes superfluous as an identi-
fier, for there can be only one queen at a time, and only one
river can be nearest. So also mthe emperor Napoleon, the first
month of the year, the same man. In such cases the article
is added on the analogy of cases where the other modifier
identifies only partially.

2025. But the article is by no means grammatically
superfluous in these last examples : although it is not required
as an identifier, it still has the negative function of showing

58 SYNTAX. " [§aoa6.

that its nouns have not those meanings which are associated
with the absence of the definite article.

Absolutely Defining.

2026. In the examples last given the article has only a
Hmited sphere of identification. Thus it is only to those
who live near the Thames that ^/le river means that special
river; in other places *the river' may be the Severn, the
Rhine, the Nile.

2027. But in some cases the identifying article makes the
noun into what is practically a proper name, as is further
indicated in writing by the use of capitals : the Lord, the Scrip-
tures, the Flood, the City (that part of London so called), the
Holy Ghost. The following are still more marked examples,
in which the group becomes a full proper name : the Victory
(name of a ship), the Cape of Good Hope,

Class- and Collective.

2028. In the following examples the definite article is
used with a single class-noun to suggest the idea of belonging
to or representing a class : he looks quite the [or ci\ gentleman \
to play the fool = to behave like a fool \ the man in the street =
* the average human being ' | the lion is the king of beasts* As
we see, the definite article when used in this way has so little
distinctive meaning that the indefinite article may be substi-
tuted for it with hardly any change of meaning. But, never-
theless, the definite article is more emphatic. It emphasizes
a quality (the gentleman), or makes the individuals of a class
more prominent : the man, the lion make us think of indi-
vidual men and lions. In fact, the word king makes it impos-
sible for us to think of more than one lion, although we know
that the statement is meant to apply (figuratively) to lions in
general.

Such phrases as play the fool may have been originally
theatrical =* play the part of the fool.'

§203a.] ARTICLES: DEFINITE ARTICLE, 59

2029. In some cases the idea of collectivity predominates
over that of individuality, as in ilord of the fowl and the brute.

2030. With collective nouns and plurals the definite article
emphasizes the idea of collectiveness, suggesting that of ' the
whole body of . . ,' as in the nobility and gentry \ the dissenters
and catholics \ the Russians do not like the Germans.

Unique Article.

2031. When the definite article is used with nouns express-
ing persons or things which are unique in themselves, it has
of course only a negative function : —

{a) the Messiah, the Devil [a devil in a different sense],
the Bible. Some of these are almost proper nouns.

{b) the sun, the moon.

(c) the earth, built on the rock^ the sea, the sky [a bit of blue
sky], the air.; the north [to steer north, the wind is north].
Some of these are also material nouns, and the definite article
is peculiarly distinctive in the case of such words as the earth
compared with earth = ' mould.' iearth = the earth is by the
analogy of heaven : heaven and earth rejoiced. For on earth
see § 2063.

In the examples under {p) and (r) the article seems also to
have a demonstrative force : the sun = * the sun we see,' ihi
earth = *this earth/

In these cases it also expresses the idea of extension or
universality — especially when there is a corresponding material
word — and hence that of collectiveness, though in a different
way from its use in the examples in the last section.

(<f) which way is the wind? \ the rain fell and the wind
blew I the dew fell \ the tide is coming in. We can also use
the indefinite article in such phrases as there is a strong wind \
there is a heavy dew. These nouns can also be used without
any article in a more abstract sense, as if they were material
nouns : rain comes from the clouds.

2032. they came on the Sunday and went away on the

6o SYNTAX. [§ao33.

Monday has much the same meaning as they came on a Sunday
and went away on the Monday, although in the former the
first the seems to suggest the idea of ' the Sunday which was
agreed upon ' or something similar. The second the in both
examples refers back in a peculiar way to the preceding
Sunday, implying ' the immediately following . /

With Abstract Nouns.

2033. The definite article is often used pleonastically with
names of diseases : to have the scarlet fever \ I have got the
rheumatism very bad compared with to catch cold, to die of
consumption. The article might be omitted in the first two
examples as well. The indefinite article may also be used in
some cases (I have got a bad cold — afresh cold) to suggest
the idea of an isolated case. The definite article, on the
other hand, has a generalizing meaning, often implying that
there is an epidemic {the cholera, the influenza).

2034. The definite article in tell the truth, on the other
hand, suggests the idea of *the true details of the present
case.*

With Proper Names.

2035. The article in such collocations as the other Smithy
meaning ' the other man whose name is Smith,' requires no
comment.

2036. The collective article in the Germans has been
already considered (2030). It is similarly used with the
plural of family names, as in the Smiths, meaning ' the Smith
family.'

2037. Sometimes the definite article is added to proper
names without affecting their meaning or function in any
way — oftener with geographical than with personal names :
the Thames \i father Thames'], the Atlantic^ the Crimea; the
Douglas. We have an instinctive feeling that the Thames is
short for the river Thames, although in Old-English Tgmes is

§ ao42.] ARTICLES : INDEFINITE ARTICLE. 6l

used without any article. The definite article is regularly
added to plural proper names, where it is collective : the
(West) Indies, the Orkneys, the Alps, the Grampians; the
Perctes, the Gracchi.

This use of the definite article is the direct opposite of its
use to make an ordinary noun into the equivalent of a proper
name (2027).

With Absolute Adjectives.

2038. In Old-English the definite article is more freely
used than in Modern English in a class- and collective sense
and with abstract nouns, as in se mann *man (in general)/
}>d godan mptn * good men (in general).' In Old-English pa
gbdan by itself is used in the same sense, whence the Modern
English the good compared with good men,

2039. In such constructions as I will do the best I can\
the true and the beautiful the article is necessary to show that
besty true are nouns and not adjectives or adverbs. So also
in the Pacific {ocean).

2040. Although we do not use any article with proper
names denoting languages — to learn French, to translate

from German into English — we use the definite article in
such phrases as a book translated from the German^ where the
article seems to suggest * the German original.'

Indefinite Article.

For the position t)f the indefinite article see § 1793.

2041. In Old-English the numeral an * one ' is used both
as a noun and an adjective in the vague sense of * a certain
(one),' sum ' some ' being used in the same way : sum mann
or an mann * a certain man.' As an in such collocations is
less emphatic than sum^ its meaning is often weakened till it
has the function of the indefinite article, which, however, is
often not expressed at all, especially in the earlier period.

2042. Although in Modern English one and a have

62 SYNTAX. [§ 2043.

diverged so much through phonetic change that we no longer
feel any connexion between them, we still use a in its
original numerical sense in such phrases as a foot deep^ not
a wordy in a minute, in all of which we could substitute the
slightly more emphatic one without change of meaning. This
is the result of confusion between strong on and weak an in
Middle English before they had been completely isolated from
one other. As the indefinite article always implies oneness,
it is often doubtful whether a definite numerical meaning
ought to be assigned to it. Thus in seven days make a week
we could substitute one, but as there is no special contrast
implied between one week and two weeks, etc., it is safer to
regard the numerical meaning as negative or secondary.
The indefinite article proper has two distinct functions : —

2043. The introductory article singles out the idea ex-
pressed by its noun, and makes us expect further information
about it : once upon a time there was a king | they sailed on
till they came to an island. It may be made more emphatic
by adding certain : a certain friend of mine \ t a certain man
went up to ferusalem.

2044. The absolute article does not single out, and has
the purely indefinite sense of *any.' In it the numerical
meaning is reduced to a minimum; it simply picks out an
individual at random to serve as the representative of a class :
a poets eye \ the earth is like a hall \ a hill is the opposite of
a valley. A noun accompanied by this article may be put in
the plural without change of meaning: hills are the opposite
of valleys.

2045. In this last example a hill has plural hills (2012).
But if the indefinite article distinctly implies oneness, the
plural must be accompanied by some numerative word. In
the absence of such words as three, many, few, the vague
words some or any are used : they sailed on till they came to
some islands (or a group of islands) \ did they see any islands?
It will be observed that some in this usage has something of the

§2052.] ARTICLES : ARTICLES OMITTED. 63

defining force of the introducing article ; thus / have brought
you some flowers implies 'the flowers I have in my hand/
while the bare plural in he t's always bringing us flowers
implies ' flowers in general.'

2046. The distributive use is a special development of the
numerical : /'/ costs two shillings a pound^ where a = * each/
The definite article is also used : two shillings the pound.

2047. The indefinite article in combination with abstract
nouns has a very vague meaning, and hence can sometimes
be omitted or have the definite article substituted for it, as in
to catch {a) cold, in a (or the) state of vapour.

2048. Verbs of naming can sometimes take the name-
noun either with or without the indefinite article, the noun in
the latter case being a sentence-word : orange is the colour of
the fruit called (an) orange. The article could not be so well
omitted in a piece of land with water all round it is called an
island.

Articles omitted.

2049. The absence of the articles is the distinguishing
mark of proper names, such as fohn, Bakery Mr. Smith
compared with tlie baker , a smith.

2060. The articles are sometimes omitted from class-
nouns which, without being fully converted into proper
names, are regarded as such : Scripture or Holy Writ says
[but the Bible']. Smith the baker may be spoken of not only
as Mr. Smith, but also as Mr. baker, the name of the trade
being used as a kind of proper name. So also a cook may
be spoken of as cook as well as the cook.

2051. Also from some names of unique objects which are
not otherwise regarded as proper names : God [but the God
of mercy, the god of war~\, heaven, hell, paradise.

2052. The absence of the articles is also the mark of
material nouns, as in made of glass compared with a glass
of wine, ^

64 SYNTAX, [§ 2053.

2053. The omission of the articles from such words as
father^ mamma^ uncle {John) is the result of the analogy of
proper names and of the frequent use of these words as
vocatives (2056).

2054. The omission from such words as parltameni,
government may also be the result of the analogy of proper
names, or of personification.

2055. The articles are often omitted from names of meals,
as in is breakfast ready ? \ dinner will he ready soon compared
with / had a heavy breakfast \ they gave a dinner in his
honour.

2056. From a grammatical point of view the omission of
the articles is the mark of the vocative relation, as in father !,
baker!. Baker t As we see, this usage levels the distinction
between class-nouns and proper names.

2057. Some exclamations follow the analogy of the
vocative, as in good Godly poor fellow! [but what a shame!,
thefool!\

2058. The omission of the article in the literal use of
words, as in how do you spell receive r \ if is a conjunction^ is
necessary.

For the omission of the articles in the plural of undefined
nouns see § 2012.

2059. Nouns qualified by a genitive do not take the
articles (except when the combination is felt as a compound),
evidently because the preceding genitive is felt to define them
enough by itself, as in metis hearts, the-mans name compared
with the hearts of men, the name of the man.

2060. There is a tendency to use nouns predicatively and
in apposition without any article on the analogy of adjec-
tives : he became king [but he is a lawyer'], he turned dissenter \
Alfred, king of England. So also complementary nouns, as
in tliey made him prisoner.

2061. The absence of the articles is in most cases a
tradition of ^ time when they were more sparingly used, and

§2o68.] ADJECTIVES, 65

ultimately of a time — in Parent Germanic — when there were
no articles at all.

2062. Hence in some cases the articles are omitted
entirely through tradition and habit, against the general
tendencies of the language, but of course only in special
stereotyped phrases and imitations of them, which were pre-
served from change by their great frequency or development
of special meanings.

2063. Thus the articles are omitted in many preposition-
groups : on (dry) land, at Christmas, in early spring, in town,
come to dinner, up stairs, out of doors contrasted with in the
country, on the river.

2064. So also in many combinations of verb + direct
object : to leave school, to take root, to send word, to keep open
house compared with to send a message, to keep a school.

2065. In some cases, on the other hand. Modern English
has deliberately discarded the definite article.

2066. Thus in Old-English it was extended, as in Modern
French and German, to abstract nouns, such words as
* wisdom ' keeping it even when personified : seo Gesdead-
wisnes. In Modern English they are treated like material
nouns, such wor<^s as ivisdom, truth, mercy, victory taking
articles only when used in some special meaning.

2067. The use of such nouns as man, woman without an
article to express the idea of ' man or woman in general ' in
such a sentence as man is weaker than [the) animals is also
of later growth, for in Old-English the definite article would
be used : se mann is wdcraponne pa nietenu.

ADJECTIVES.

Absolute.

2068. When two assumptive adjectives are joined to one
noun, the three generally follow in immediate succession, as
\n your poor old father. But the first may be detached from

VOL. II. F

66 SYNTAX. [§ 2069.

the others, often by a conjunction (good and bad men,/air or
foul means), sometimes by a longer break : never do by foul,
what can he accomplished by fair means.

2069. In the above examples the first adjective is merely
detached from a noun which follows it. If the second adjec-
tive is separated from the noun by being put after it, giving
the order adjective + noun . . + adjective, it is no longer
merely detached, it is absolute : it is not just that new rules
should destroy the authority of the old (Dryden). In such cases
the prop-word one is generally added in the present English —
the old ones. In writing we often repeat the noun, to avoid
the colloquial one — the old rules. Sometimes the colloquial
association may be destroyed by transposition, as in many of
these rooms had doors which led into the one adjacent.

Free.

2070. The use of free adjectives is much restricted in the
present English.

2071. In good/ the adjective is used as a sentence- word,
that is, it is nearly converted into an interjection.

2072. So also in the literary language an adjective may
be used in the vocative relation : cruel, dost thou forsake us ? \
go forth, beloved of Heaven !

2073. Otherwise a free adjective must be preceded by the :
the good = ' good people,' * what is good.'

2074. In the personal sense this form is used now only
in plural constructions such as the good are happy \ the
mighty have fallen \ the English and the French. Formerly
it was used freely in singular constructions also, as in none
but the brave deserves the fair (Dryden). The few survivals
of this usage, such as the betrothed, the deceased, have been
converted into nouns : the deceased's relatives or the relatives
of the deceased.

2075* The the is occasionally dropped in the plural con-

§2o8i.] ADJECTIVES: COMPARISON. 6^

struction — especially when there are other qualifiers — by
the analogy of the construction with nouns, but only in the
literary language : older and abler passed you by (Gray) | ten
righteous would have saved a city once (Cowper).

2076. the + adjective in the singular neuter sense is used
not only in such phrases as the good and the beautiful,
but also in colloquial phrases such as to do the needful^ to
answer in the affirmative.

2077. It is occasionally used with names of languages, as
in to translate from the French (2040).

2078. Free adjectives often occur in combination with
possessive pronouns, and in singular constructions; but in
most cases where this construction is still preserved the
adjective is more or less converted into a noun: Wd my
sweet prepare to chide \ '\his distant fair \ her betrothed. In
the first two one could be added. This construction is
especially frequent with comparatives : he is my elder ^ my
junior \ his superior, my dear I is a sentence-group.

2079. In the above examples the adjective is in the
singular. In plural constructions -s is added — that is, the
adjective is converted into a noun : our better s, his superior s,
my dears !

2080. Most of those which are in colloquial use can also
take 'S in the genitive singular : his superior s orders.

Comparison.

2081. In Modern English there is a tendency to use the
superlative instead of the comparative. In the spoken
language we always naturally speak of the shortest of two
roads, the biggest of the two, although we use the comparative
in careful speech and in writing. But in such a sentence as
they are a bad pair ; but she is the worst of the two we could
not substitute the comparative without weakening the force
of the comparison, she is the worse of the two being equivalent

F 2

68 SYNTAX. [§

to she is rather the worse (or worst) of the two, where the
comparative may be used even in colloquial speech. It is
therefore probable that the superlative was at first used to
strengthen the force of the comparative, implying * very much
more/ In such a sentence as he came in first of the two the
superlative can hardly be avoided.

2082. In all the above instances the formal distinction
between comparative and superlative is superfluous, because
the number of objects to be compared is shown by the two.
But in such a construction as the younger Miss Pecksniff =
' the younger of the two Miss Pecksniffs ' a careful speaker
would keep the comparative, because the superlative would
admit the possibility of there being more than two Miss
Pecksniffs, although a careless speaker would not hesitate to
employ it.

2083. In some traditional phrases the comparative is used
instead of the superlative : the latter end, the latter day, utter
contempt.

Comparison of old.

2084. Of the two comparisons of old, the irregular
(original) forms elder, eldest have the more limited range —
being used chiefly to distinguish members of the same family
— and the more abstract meaning, for they imply that the
distinctions of age are made not for their own sake but as
a means of discrimination and identification. The new-
formations older, oldest, on the other hand, which keep the
vowel of their positive unchanged, directly call forth the ideas
of * age ' or * long duration.' Thus in the eldest son of his
oldest friend, the son may be a young boy, and the word
eldest makes us think of a connected series or gradation ;
while oldest deals with a shifting group of isolated individuals,
and makes us think more or less of the characteristics of old
age or long duration.

2085. The irregular comparison, in fact, makes old into
a mere qualifier— we might almost say a mark-word — the

§ao90.] ADJECTIVES: COMPARISON. 69

eldest son being almost equivalent to a proper name. We can
express the difference in other words by saying that elder is
practically not the comparative of old, but the opposite of
younger.

2086. Hence the irregular forms are used only assump-
tively — the eldest son, an elder son, her three eldest daughters —
not predicatively, except with the definite article in such
constructions as he is the elder of the two brothers \ she is the
eldest of the family, where the logical predicate is not elder,
eldest, but an understood noun of relationship.

2087. If no the precedes, the regular forms alone can be
used as predicates : he is five years older \ he is much older
than his brother \ my elder brother is five years older than I am
— than me.

2088. But with the definite article we can say she is about
four years the elder :=. . . the elder of the two sisters instead of

the more usual she is about four years older. The use of the
irregular forms in such cases has probably been kept up by
the analogy of the constructions given in § 2086.

In he is five years my elder the adjective is felt to be a noun,
not merely an absolute adjective.

2089. Even the assumptive use of irregular forms is
limited. Thus they could not be used in such sentences as
the following : fohn is not my eldest brother ; I have another
older brother = . . / have another brother who is older than
John. Here older does not state differences of age with
reference to a fixed, absolute series of brothers, but picks out
two brothers, and compares them as if they were strangers.
Hence the practical rule that the regular comparison is
necessary whenever the comparison is, or can be, defined by
a group or sentence introduced by than.

2090. The irregular comparison is not restricted to family
relations, but may be employed with reference to other
personal relations, especially those of a more permanent

70 SYNTAX. [§ ao9i.

character which imply gradations of rank or authority. Thus
in the elder of the two ladies the irregular comparative does
not necessarily imply that the ladies are sisters, or even
mother and daughter ; they may be friends living together, or
merely travelling companions. So also the elder boys are
expected to take care of the younger ones might be said of
schoolboys as well as brothers. The superlative eldest is rare
in this use.

2091. With names of animals we can speak of the elder of
the two sheepdogs^ implying a pair of dogs that are employed
together. When appHed to animals the irregular comparisons
seem not only not to imply relationship, but almost to
exclude it.

2092. In all the examples given in the last two paragraphs
the regular comparison may also be employed, although in
some cases it might be taken to involve a slight change of
meaning by emphasizing the idea of old age.

2093. The irregular comparisons may be applied to other
than personal nouns, if the noun suggests personal relations :
the elder generation \ the elder branches of the family. Here,
again, the regular comparisons may be substituted with hardly
any change of meaning.

2094. In such cases as the following the irregular com-
parison is rather literary or archaic than colloquial : our elder
writers \ the elder inhabitants \ the elder school of English verse.
It need hardly be said that as the irregular comparisons were
originally the only ones, and as the regular ones came into
use only slowly, the irregular comparisons were in the earlier
stages of Modern English used in many of the constructions
where we now use only the regular ones.

§ao97.] PRONOUNS: PERSONAL. 71

PRONOUNS.

Personal,

2005. we is used instead of / as a * pronoun (or plural)
of majesty ' to mark the supreme authority of kings, queens,
reigning dukes and other persons at the head of a state. We
see the beginnings of this usage in the Old-English laws,
where the king speaks of himself as tc, and then goes on to
say we bcheodap . . * we command . . ,' the we being meant
to include the witan or councillors.

2096. we is also used instead of /—though less frequently
now than formerly — by the author of a book in addressing
his readers in order to avoid the egotism of a singular pro-
noun. This ' plural of modesty ' is found already in Old-
English. It arose probably from using we in the indefinite
sense of * myself and the other authorities on the subject/
This usage occurs also in colloquial language, as when a boy
says give us some I meaning ' give me some ! '

Indefinite Personal Pronoun.

2007. In Modern English the place of the indefinite per-
sonal pronoun vian has been taken by one, but in the spoken
language we prefer to use the personal pronoun you, and
occasionally we : the right bank of a river is on your right
side when you stand with your face to its source \ when people
laugh, we know they are pleased. But these two pronouns can
be used only when the context allows us to take them more
or less in their literal meaning ; thus for the author of a book
abruptly to address his readers z.syou would be uncolloquial
as well as unliterary. They are, of course, also avoided in

72 SYNTAX. [§2098.

cases where taking them in their literal meaning would involve
some great change of meaning.

2098. In many cases where the first and second persons
are excluded by the context, the third person plural may be
used : fhey say we shall have a hard winter. Here neither
you, we nor one could be used without changing the meaning;
one would imply * people are in the habit of saying that . /

Pleonastic.

2099. The pleonastic insertion of a pronoun after a
noun in the subject-relation— /<?^« he says — occurs only as
a vulgarism in the present spoken English, or as the
occasional result of hesitation or carelessness, but is frequent
in the literary language, where it suggests picturesqueness
or quaintness : his coat it was all of the greenwood hue \
a frog he would a-wooing go.

2100. A pronoun may also be made pleonastic by tagging
on the equivalent noun. This is frequent in the spoken as
well as the literary language : he was a wonderful many that
uncle of yours \ fit ceased, the melancholy sound.

For the pleonastic pronoun with the imperative see § 1806.

Gender.

2101. In such a sentence as let every man or woman do as
he or she likes the group he-or-she is used as a sort of com-
pound to supply the want of a personal pronoun of the
common gender in the singular corresponding to the plural
they. This difficulty is evaded in various ways.

One is by using he only, leaving the application of the
statement to women as well as men to be taken for granted.

In the spoken language the difficulty is got over by the
use of the genderless plural they : let every one do just what
they like \ if any one comes, tell them to wait \ a person cannot
help their birth.

§aio7.] PRONOUNS: POSSESSIVE PRONOUNS. 73

Prepositional Genitive.

2102. The prepositional genitive (of me) is generally used
instead of the possessive {my) in the objective (2000) meaning:
/ cannot bear the sight of him— of it \ I will change my
(subjective genitive) treatment of him.

2103. Its subjective use in not for the life of me I is
probably due to the analogy of it will be the death — the
ruin — of him, where it is — or may be — felt to be objective.

2104. Its occasional threatening or contemptuous use in
such phrases as / will break the neck of you I seems to be
due to the influence of those dialects which distinguish
between the mans head and the head of a beast.

Possessive Pronouns.

2105. The possessive pronouns are used not only sub-
jectively but also objectively, as in they told me her history =
* what was told of her ' not * what was told by her.' But if
the noun is associated in meaning with a transitive verb, the
prepositional genitive is used, such constructions as his
murderer = ' the man who murdered him ' being now
exceptional.

2106. Possessives are sometimes used as antecedents to
relative pronouns, as in inor better was their lot who fled.
But this construction hardly occurs in the spoken language,
which either avoids the relative construction or else sub-
stitutes the prepositional genitive: the lot of those who
fled.

2107. Although the possessive pronouns no more neces-
sarily imply possession than the genitive case does, yet it
is one of their most important functions to do so. If the
idea of possession is excluded by the context —so that there
is no possible ambiguity — they are freely used to express
a variety of relations, as in his /ear of his mastery where the'

74 SYNTAX. [§aio8.

relation implied by both possessives is the exact opposite of
that of subjective possession.

2108. But in such sentences as he fought three duels, and
killed his man each time \ that boy has just broken his fourth
window this week the freer use of the possessive is felt as
a licence, because at first sight we should assume his man
to mean 'his servant/ and his window to mean 'his own
window/ although the context suffices to suggest the freer
meaning. In the first sentence his adversary would be quite
normal, because the word adversary does not suggest in any
way the idea of possession.

2109. In some cases possessives are little more than
reference-words, and we are often at liberty to use either
a possessive or the definite article (cp. the heard^ § 2020).
Thus we may speak of the subject of a narrative or dis-
cussion as our hero as well as the hero, or, more familiarly,
as our young friend.

2110. The possessive >'^»r [not thy] is also used like our
in the preceding examples. But while our suggests the idea
of taking under one's protection or patronage— which may
degenerate into good-natured contempt — your generally
suggests antagonism to what is brought forward by the
person implied by the you, this antagonism often taking
the form of dislike or contempt, sometimes simply of in-
difference, implying 'say what you will,' 'these examples
will do as well as any others ' ; a smile — not one of your un-
meaning wooden grins — but a real . . smile \ I would teach
these nineteen the special rules^ as your punto, your reverso,
your . . (Ben Jonson). Hence it is often used to show that
the speaker is talking in a humorous or sportive vein.

2111. The use of my Lord, my Lady in the vocative
relation has led to these combinations becoming fixed that
they are freely used in the third person without any
suggestion of address, as in my Lord and my Lady quarrelled,
and abused each other ^ where my Lord =^ Lord A, This

§aii6.] PRONOUNS : POSSESSIVE PRONOUNS. 75

usage seems to have arisen from the desire to avoid con-
fusion with Lord {God) /, /he Lord,

POSSESSIVES WITH GENERAL ADJECTIVES.

2112. Possessives can take before them the same general
adjectives of quantity which can also precede the articles,
such as all, both, half: all my iimey both his eyes, half his time
[but also half of his lime].

2113. The other general adjectives follow the possessives :
his whole time, my three friends, fher every word a wasp.

2114. In most cases possessives when used assumptively
cannot take before them a general adjective without the
change of construction seen in a friend of his, he is no friend
of mine, three friends of mine. In all these instances the
construction has a partitive meaning: a friend of his — ^2^
friend from among his friends.' But there is a difference
between the purely partitive three of my friends and the
vaguer three friends of mine, which does not — as the pre-
ceding example does — necessarily imply that the speaker
has more than three friends; although, on the other hand,
it does not definitely limit them to three, as in my three

friends, where three is equivalent to a descriptive adjective.

2115. In the earlier Modern English possessives may be
preceded by the demonstratives this and that\ but in the
present English such constructions as' this our friendship,
these my children occur only in the higher literary style, the
colloquial language using the periphrastic constructions —
this friendship of ours — by the analogy of the partitive
constructions described above.

Emphatic Possessives.

2116. The emphatic possessives are — like the unemphatic
ones — used both assumptively {Jiis own house) and absolutely:
his house is his own.

They are also used as nouns : he knows how to hold his own.

76 SYNTAX. [§aii7.

2117. They take the prepositional construction in all
cases in which the unemphatic possessives do, as in he gave
it to a friend of his own \ he has no house of his own compared
with all his own houses are . .

2118. As own is never used except in combination with
possessives (or genitives) we cannot say in English *he has
own money = ' he has private means ' in an indefinite sense,
as we could in German. Hence, as the prepositional
possessive is in itself indefinite (2114), we always use it in
such cases' as he has money of his own compared with the
definite he has the control of his own money.

Interrogative Fronouns.

2119. In Old-English the neuter hwcet is always used
(conjunctively as well as relatively) instead of hwd when there
is an accompanying pronoun to show that persons are
meant: hwcBt sind ge? 'who are ye?' | he nyste hwa;t hie
wdron ' he did not know who they were,' that is, ' he did not
know what their nationality was.' In Modern Enghsh we
always use who of persons as an identifying pronoun, restrict-
ing what to the descriptive meaning : what is man ? \ what is
he — is he a lawyer ? compared with who is he ? he is the new
curate.

2120. which has a selective meaning : it assumes a group,
and asks for an individual of that group, whence its partitive
construction with of as in which is it to he? \ which is the
shortest way? \ which of these three will you have?

2121. The use of what in questions expressing surprise—^
whaf? and in rhetorical questions — what right have you
to interfere? — has led to an extended use of it in Modern
English as an intensitive word, and in exclamative sentences :
what {bad) manners 1 what a (ponderfut) man! \ what an eye
he has !

2122. In Old-English hwcet has the indefinite sense of
'something.' This survives only in the phrase I will tell you

§ai26.] PRONOUNS : RELATIVE PRONOUNS. 77

whai, which is often expanded into / will tell you what it is,
the shorter phrase being regarded as an elliptical form of the
longer one. This meaning survives partially in the correlative
adverbial what with . . what with.

Relative Pronouns.

2123. In the spoken language relative pronouns are
generally — not always— omitted in clauses dependent on
a word in the direct-object relation : the man I saw yesterday \
the book you told me of. In writing we generally put in a
relative pronoun : the man whom (or that) I saw \ the book of
which you told me,

2124. The omission of a relative in the subject-relation is
quite exceptional in the present spoken English, but was fre-
quent in the earlier Modern EngHsh. In who is that just
rang ? it is omitted in order to avoid the repetition involved
in who is that who . . , who is that that , . In it is not the fine
coat makes the fine gentleman we feel that the omission of the
relative is the result of confusion with the corresponding
positive statement the fine coat makes the fine gentleman.

2125. The oldest relative pronoun is that. It can refer
to persons as well as things, but its use is limited in other
respects. Like who, it is used only as a noun, not as an
adjective.

2126. It always stands at the head of its clause, and cannot
have a preposition before it. Thus in such literary sentences
as the book of which you told me \ the letter to which you refer
the change of which into that would necessitate putting the pre-
position at the end of the clause : the book that you told me of
This last construction is sometimes used in writing as a mean
between the stiff construction with which and the purely col-
loquial with omitted that {the book you told me of). But even
in colloquial speech the construction with preposition + J^^^VA
cannot always be avoided, as in observe the dignity with which

78 SYNTAX. [§ 3127.

he rises /, where we could not say the dignity he rises with^
which would, indeed, be unintelligible.

2127. It is important to observe that that is always very
closely connected with its antecedent both logically and for-
mally, and is never used when there is anything like a pause
between the relative clause and the principal clause. Hence
it is never used as a progressive relative.

2128. As it is always pronounced with a weak vowel (tSst),
it cannot take stress, and hence cannot be followed by a
pause. Thus we could not substitute it for who in he is
a man who, if . .

2129. that was formerly used with an antecedent that, as
in thai thou hadst seen that that this knight and I have seen !
(Shakespeare). We now use the condensed relative in such
constructions : / wish you had seen what we have seen.

2130. who is, of course, used only as a noun. It is used
mainly in reference to living beings, especially human beings.
It is occasionally used of the higher animals. In Early
Modern English it is sometimes applied to things, but gene-
rally with implied personification: a gentle flood who , .
(Shakespeare).

2131. But the possessive whose is still applied to lifeless
things, though with a certain hesitation, and only to avoid
the longer of which. In the spoken language we avoid such
constructions as a tree whose shade . . as much as possible.

2132. which diflfers from that and who in being used both
as an adjective as well as a noun. As a noun it now refers
only to lifeless things. In early Modern English it was freely
applied to persons as well, and this usage lasted into the pre-
ceding century. As which is the only relative adjective, it is
in this function necessarily applied to living as well as lifeless
objects.

In early Modern English which often takes the before it :

§2138.] PRONOUNS: RELATIVE PRONOUNS. 79

the cities in the which Lot dwelt (Bible) | she hath received your
letter^ for the which she thanks you a thousand times (Shake-
speare). As this usage is against all analogy, it is probable that
it is an imitation of the French lequel.

In some cases there is a certain amount of fluctuation in
the use of the 'different relative pronouns.

2133. We have seen that only which can be used as an
adjective. Also that only who and which can be used as
progressive noun-relatives.

2134. As adjective relatives and progressive relatives are
not natural to colloquial speech, which hardly ever occurs in
the spoken language except when it has a sentence for its
antecedent (217), which, again, is not a frequent construction
in the spoken language.

2135. Hence the only two relatives in general colloquial
use are that and who.

2136. The present spoken English shows a reaction against
the earlier colloquial tendency to favour that. The general
tendency now is to substitute who for that when persons are
referred to, that taking the place of the lost which.

2137. This tendency in favour of who is clearly seen in
many relative clauses which, although not fully progressive,
are more descriptive than defining ; in such sentences we
generally use who in preference to that, as in dinner-time then
came again, to the especial delight of the two children, who felt
rather empty | he had been well thrashed by a gentleman who
did not approve of his trespassing on his grounds, while in
parallel sentences dealing with things we can always substi-
tute that for which', the brook ran into a series of fishponds,
which (or that^ looked very old, for their sides were shaggy with
reeds, and . . \ the earth is a big ball that is always spinning
round like a top,

2138. In the above examples the who individualizes and
singles out the person it refers to. But if the relative describes
the person denoted hy its antecedent only by including him

8o SYNTAX, [§ ai39.

in a class, then that is obligatory : Newton ivas one of the
greatest men thai ever lived | he is a man that will never get on
in the world. In the first example the relative clause has so
little descriptive force that it might be omitted without sensible
loss to the meaning of the whole. With the second example
compare is that the man who is getting on so well at the bar y»,
where the relative is directly descriptive, and does not merely
include the man in a class. In such examples as the follow-
ing we may use either that or who, according as we wish to
show that we are thinking of the persons collectively or indi-
vidually : when the boat came near the shore, they thought they
recognized one of the convicts that were in her \ they give prizes
to the boys that have the best manners.

2189. Hence in those cases in which we can employ
either that or who with only a slight shade of difference of
meaning, the latter is more polite, as emphasizing the fact
that we are speaking of living beings and that we respect
their individuality: those {^embers) who are in favour of this
resolution will please hold up their hands \ I cannot understand
how any one who has once taken an interest in education can
ever lose it.

2140. Hence also the combination he that is now obsolete
in the spoken language, being preserved only in traditional
phrases such as he that fights and runs away may live to fight
another day. We now employ some such construction as
a man who . . , or, if absolutely necessary, a man that.

2141. Although a relative in the direct- object relation is
generally omitted, there are cases where this would lead to
obscurity or awkwardness ; in such cases the spoken language
seems always to use that, evidently in order to avoid the
' ungrammatical ' who = whom : Mrs. Carnahy was helped out
of the trap ; . . then the children were lifted out by the motherl-
and then the nurse, an awkward, plain girl that nobody helped,
tumbled out by herself. Here the that ought to imply that the
nurse as a general rule was not helped out by any one,

§ai45.] VERBS: NUMBER. 8l

whereas the meaning is that she was not helped out by
certain definite individuals on a certain definite occasion.

2142. In the written language the fluctuation between the
relatives is of course much greater, because of the greater
variety of constructions, and the necessity of putting in
relatives where they are omitted in natural speech, so that
the writer has no linguistic instinct to guide him.

NUMERALS.

2143. Cardinals are used instead of ordinals in some
constructions: in the year 1800, in 1800 | he lives at (number)
12, High Street.

2144. Numerals are used in a variety of elliptical con-
structions, such as a man 0/ thirty {years o/age), a carriage-
and-five (horses\ at ten (d clock) \ the tenth {day) of May, the
fifth {part) of . . , whence has developed the noun fifth in
four-fifths.

Although we feel cut in two to be elliptical, the Old-English
on twa to'd^lan shows that it cannot be a shortening of on
twegen dalas, and hence is historically not the result of
ellipse.

VERBS.

Number,

2145. Concord requires that a verb should agree in num-
ber with its subject. This principle is always adhered to
when the subject is a single word of unmistakable singular
or plural meaning as well as form ; but when the meaning is
in conflict with the number, logical considerations generally
prevail over grammatical, so that subject and verb are in
different numbers.

VOL. II. o

82 SYNTAX, [§2146.

2146. Collective nouns (compare § 1972) in the singular
are frequently joined to plural verbs whenever the statement
is meant to apply to the separate individuals, while a singular
verb implies that the speaker is not thinking of the individuals,
but rather of the whole collective body. Thus we can say
either the public is . . or the public are . . , the Council {the
Board) is or are of opinion that . . Both numbers may even
occur in the same sentence, as in the people is one^ and they
have {all) one language.

2147. Conversely, subjects in the plural which suggest
ideas of singularity may take singular verbs, as in forty yards
is too much \ thirty yards is a good distance.

2148. Special difficulties arise when the subject consists
not of a single noun- word, but of a group of words.

2149. Thus, although two or more nouns joined by and^
or standing in a copulative relation without any conjunction,
ought grammatically to have their verb in the plural, there
are many cases where the plural would be logically im-
possible, or at least unnatural, as with such groups as a needle
and thready the peace and good order of society, where the
combination expresses one idea.

2150. So also, on the other hand, two or more nouns
joined by or or nor do not necessarily take a verb in the
singular unless they stand in a strong alternative relation to
one another ; while two words standing logically in a strong
alternative or adversative relation to one another take a sin-
gular verb whatever their purely grammatical relations may
be : nor heaven nor earth have been at peace to-night (Shake-
speare) =>%^az'^» and earth have not , .\ my poverty, and not my
willy consents (Shakespeare).

2151. When an additional subject is tagged on, there is
a tendency to make the number of the verb depend ex-
clusively on the preceding subject- word, as in + the earth is
the Lordsy and the fullness thereof

§ai570 VERBS: PERSON, 83

2162. Hence the distinction we instinctively make between
/Ae captain was taken prisoner with three of his men and the
ungrammatical the captain with three of his men were taken
prisoner.

2153. When a verb is followed by more than one subject —
not by tagging, but by regular grammatical inversion — it
often agrees in number with the nearest one, especially in
earlier Modern English: where is Lysander and sweet
Hermia ? (Shakespeare).

2154. Anomalies sometimes arise through referring the
verb not to its subject, but to some word connected with the
subject, as in the opinion of several eminent lawyers were in his

favour^ which is at the same time a case of attraction | thai
is one of the most Valuable books that has appeared in any
language^ where the prominence of the logical subject one,
together with the want of any mark of plurality in that,
causes the plurality of books to be overlooked.

Person.

2165. In the older Arian languages the persons of verbs
were so clearly shown by their inflections that a verb could
stand alone without any subject-word, either noun or pro-
noun, and the pronouns were added only when emphatic.

2156. In Old-English, where the endings were less distinct
— there being, for instance, no distinction of persons in the
plural of any verb — the personal pronouns were regularly
added whenever the person of the verb was not shown
grammatically by some other word, and their omission is
exceptional.

2157. But Old-English still shows traces of the older
usage in the regular omission of the pronouns in explanatory
supplementary sentences closely connected with the preceding
sentence : J>d for he norpryhte be p&m lande : let him ealne-
weg pcBt weste land on pcet steorbord ' then he sailed direct

G 2

84 SYNTAX. [§2158.

north along the land, keeping the uninhabited country on
the starboard all the time' | hie hine pa bismorlide a'cwealdon :
of-torfodon mid bdnum and mid hrtpera heafdum ' then they
killed him ignominiously by pelting him with bones and
heads of oxen.' In this construction the verb always has
connective front-position (1809).

2158. In the imperative, where the distinction is not
required, the primitive omission of the pronouns is still kept
up, their addition being still emphatic (1806).

2159. In careless speech the pronouns are often omitted
when the context is clear, and in all persons : hg will not
have any help : {he) says he can do it all himself \ where has
he gone to ? (/) doiit know, {you) don't know, don't you .^

2160. When a verb refers to two or more pronouns of
different persons, there is difficulty in determining what
ought to be the person of the verb, as in either you or I am
or are in the wrong. It would seem most natural to make
the verb agree with the nearest pronoun. But am wrong
sounds unnatural, because it seems expressly to exclude the
you. On the other hand are wrong seems to suggest the
idea of the third rather than the second person plural. But
as it does not exclude the second person plural, it is pre-
ferred to am. We should not hesitate to employ i| in
a conjunctive collocation such as_y^« and I are one. In the
disjunctive either he or I is in the wrong we prefer the third
person singular.

2161. But we avoid such difficulties as much as possible
by using some verb which does not distinguish persons :
either you or I must be in the wrong.

2162. Early Modern English has similar difficulties in
some relative clauses. Thus the Prayer Book has thou art
the God that doeth wonders against the Bible's doest. In the
Present English we use the third person : are you the man
that has the key ?

1 63.]

VERBS: TENSES.

85

Tenses.

2163. The following paradigm gives a general view of
the tense-distinctions of the Present English verb.

ACTIVE.

Indicative.

PASSIVE.

Present

I see

I am seen

De/.pres,

I am seeing

I am being seen

Preterite

I saw

I was seen

Def.pret.

I was seeing

I was being seen

Perfect

I have seen

I have been seen

Def.perf.

I have been seeing

I have been being
seen

Pluperfect

I had seen

I had been seen

Defplup.

I had been seeing

I had been being
seen

Future

I shall see

I shall be seen

Deffut.

I shall be seeing

I shall be being seen

Future preterite

I should see

I should be seen

Deffut. pret.

I should be seeing

I should be being
seen

Future perfect

I shall have seen

I shall have been seen

Deffut. per f

I shall have been

I shall have been

seeing

being seen

Conditional.

Present

I should see

I should be seen

Defpres.

I should be seeing

I should be being
seen

Preterite

I should have seen

I should have been
seen

Def pret.

I should have been

I should have been

seeing being seen

Imperative.

Present

seel

be seen I

86

SYNTAX,

[§ a 16^

Supine {Infinitive).

Present

to see

to be seen

De/.pres,

to be seeing

to be being seen

Perfect

to have seen

to have been seen

ne/.perf.

to have been seeing
Participle.

to have been being
seen

Present

seeing

being seen

Preterite

—

seen

Perfect

having seen

having been seen

Defperf

having been seeing

having been being

seen

2164. Some of the longer forms — especially in the pas-
sive— seldom or never occur.

Have-Forms.

2165. In the periphrastic tenses formed with have Xht
preterite participle is generally indeclinable in Old-English,
but in the earlier period it is often put in the accusative, as
in pa hi hie ofslcegene hcefde * when he had killed them,' hie
hcefdon hira cyning d'worpenne ' they had deposed their king,'
showing that it was originally regarded as an adjective in
apposition to the noun-word governed by have — ' they had
their king in a state of being deposed.'

2166. It is evident that these forms were at first used only
with transitive verbs. Accordingly, in Old-English the corre-
sponding forms of intransitive verbs are generally formed
with be: he is hider cumen *he has come here' \hle waron
dfarene *they had departed.' Here the participle always
agrees with the noun-word with which it is connected. It
must, of course, be taken in an active sense — ' he is in
a state of having come.'

2167. But when the origin of the have-forms had been

§2171.] VERBS: TENSES (DO), 87

forgotten, they were gradually extended to intransitive verbs
as well, especially when stress was laid on the idea of inde-
pendent action, as in hte hce/don gegdn * they (had) marched.'
Even weorpan * become ' takes have in the special impersonal
construction hu hint hce/de geworden wip hie ' how he had
fared with her/

2168. In Modern English the use of have has been ex-
tended to all verbs, although we still use be in sbme cases to
imply a state or result rather than an action : is he gone ? [he
has gone on a journey] \ when he awoke, the hoys of the village
were gathered round him. In such constructions we feel the
participles to be equivalent to adjectives. Hence the collo-
quial construction / am done — I have done ' I have finished,'
on the analogy of / am ready.

Do-FoRMS.

2169. The simple forms of the finite verb — (indefinite)
present, preterite, and imperative — have special emphatic
and interrogative forms compounded with do: —

Unemphatic, Emphatic. Interrogative.

Present I see I ;do see -do I see

Preterite I saw I ;did see -did I see

Imperative see 1 ;do see ! —

2170. The remaining emphatic forms are made simply by
putting an emphatic stress on the auxiliary (of course in its
strong form); the interrogative forms by transposing the
pronoun and auxiliary.

2171. As all the interrogative forms can also be made
emphatic, we have in all eight forms. It will be enough to
give those of the indefinite present and future as examples: —

Affirmative I see I shall see

Affirm, emphatic I ;do see I ;shall see

Negative I don't see I shan't see

Neg. emph. I jdon't see I ;shan't see

88 SYNTAX. [§2173.

Affirm, interrogative do I see shall I see

Affirm, interr. emph. ;do I see ;shall I see

Negative interr. don't I see shan't I see

Neg. interr. emph. ;don*t I see ;shan't I see

2172. The first beginnings of the auxiliary use of do can
be traced back to Old-English. In Old-English the use
of do as a substitute for a preceding verb is fully developed,
as in Crist weox swd-swd dpre cild dop ' Christ grew as other
children do.' Allied to this is its peculiar anticipative use
in such sentences as se mona dip cegper^ ge wiext ge wanap
'the moon does both: both waxes and wanes/ where the
omission of the personal pronoun is quite regular, the second
clause being complementary or explanatory (2157). From
this half-auxiliary use was developed the full auxiliary use
with the second verb in the infinitive, which is however still
very rare in Old-English: swd dop nu pdpeostru wip'standan . .
'so now does darkness resist . .' This change of con-
struction was probably due to the analogy of the construction
of the other auxiliaries with the infinitive.

2178. A similar anticipative use of do in imperative
sentences is found in Transition and Early Middle English :
do^ gd and ne synga pU nafre md ' go and sin no more ' | do,
seie hwuil * say why ! ' It is very doubtful whether this has
any connexion with the Modern emphatic imperative.

2174. The other anticipative construction soon died out
in Middle English without leaving any traces in Modern
English, while the f/d? + infinitive construction extended more
and more, especially towards the end of the Middle English
period. By the beginning of the Modern period it had
become an integral part of the language.

2175. But even in the Modern period the do-ioim^ had
not at first any distinctive meaning, and were used promis-
cuously with the simple forms, according as caprice, con-
venience, and clearness of construction, or euphony suggested.

§ 2i8o.] VERBS : TENSES (DO). 89

2176. The auxiliary do was from the beginning capable
of taking strong stress when emphatic, just like any other
auxiliary. Hence Early Modern EngHsh was able to dis-
tinguish four forms : the simple (I see), which might be either
emphatic or unemphatic, the periphrastic unemphatic (/ -do
see), and the same emphatic (/ jdo see).

2177. By the middle of the sixteenth century the natural
tendency to regard the longer (periphrastic) forms as essen-
tially more emphatic than the shorter (simple) forms had
begun to prepare the way for the disuse of the un-
emphatic periphrastic forms, so that / ;do see came to
be regarded as the direct emphatic form of the simple
unemphatic I see.

2178. In the Present English the emphatic forms have
extra stress on the auxiliary, to which that of the nucleus is
subordinated, while in the unemphatic negative forms the
auxiliary and the nucleus have — or may have — equal stress :
ti 'doestit "matter! it jdoes matter/ But of course the
negative forms may be emphatic also : he ought not to have
told him of it. he ;didnt tell him 1

2179. As already remarked (1898), the emphasis of these
emphatic forms is always generalizing ; that is, such a form
as / do see does not emphasize anything in the meaning of
see itself, but implies some general antithesis such as that
between assertion and denial, the real and the unreal, present
and past time, so that I do see implies *I see as a fact,* *I see
nowl etc.

2180. If antithesis is the result of contrasting the special
meaning of a full verb with that of some other word, then the
emphatic stress necessarily falls on the nucleus in the peri-
phrastic forms, as did he ;ride? I thought he walked. In
Early Modern English this is also possible with afl&rmative
</(?-forms, as in for otherwise they -do pervert the faculties of
the mind, and not prepare them. This usage lasted longest
with verbs of requesting — I do entreat you — and asserting —

90 SYNTAX. [§ fli8i.

/ do assure you — but is now almost extinct in the spoken
language.

2181. The use of the ^(C?-periphrasis in questions was no
doubt prompted by the desire to avoid the inconvenience of
the old verb-inversion, especially the detaching of the verb
from its object {see you it?) and the placing of the subject
in what ought to be the object-position (cakh dogs mice?),
while the periphrastic forms do you see it? | do dogs catch
mice ? show practically no divergence from the order in the
corresponding affirmative sentences.

2182. That this was the real reason for the general
adoption of the periphrastic forms in questions is shown by
the fact that it is never used in those interrogative sentences
which have the same order as affirmative sentences, that is,
in special interrogative sentences beginning with an inter-
rogative pronoun in the nominative, as in who broke that
window ? I what brought you here /, or with an interrogative
word or word-group qualifying the subject, as in which boy
broke the window? \ how many people came?

2188. The occasional colloquial use of the simple inter-
rogative form in such phrases as what say you ? \ what think
you? is partly archaic, partly due to the influence of the
dialects.

2184. As might be expected, the do4orms are equally
obligatory in other cases of verb-inversion (1814), most of
which, however, do not occur in the spoken language : did
Nature act with full consciousness, these imperfect formations
were inexplicable \ no sooner did Boxer hear the gun than he
jumped up with a howl \ so high did political animosities run

that . .

2185. For the reasons given in § 2181 inverted transitive
verbs always take the periphrastic form, even in constructions
in which transitive verbs keep the simple form, either always,
as in the colloquial here comes . . , down fell . . , now comes . . ,
or occasionally, as in the more literary thus stood matters \

§ ai9i.] VERBS : TENSES (DO). 9I

wt/h ihe new life came new purpose, with which compare then
did I commit myself to the one physician of the soul \ thus did
the editor see himself All these last constructions are purely
literary.

But transitive verbs are inverted in parenthetic sentences
such as says he because there is no subject.

2186. The use of do in the negative forms of the verb
does not, as in the interrogative forms, make the expression
clearer or more convenient. It is therefore probably a
development of the emphatic use, all negative constructions
being essentially emphatic, because the negation reverses the
meaning. The analogy of the negative constructions of
the other auxiliaries must also have helped to bring them
into general use and fix their form.

2187. But many verbs in very frequent use in negative
constructions, such as know, doubt, care, still kept the simple
negative form long after it had been elsewhere lost, and we
still keep it in stereotyped adverbial or parenthetic phrases
such as / know not how, I doubt not, if I mistake not.

In such phrases as / hope 'noty I think 'not the not does
not negative the verb.

2188. As a general rule, do is, like other auxiliaries, used
only with verbs of full meaning. It would, indeed, be im-
possible to introduce it into such combinations as can he
come?t because can has not an infinitive.

2189. be and have do not take do in interrogative and
negative forms even when not used as auxiliaries : is he
ready ? \ has he any money ? But they take it in the negative
imperative, as in do not be afraid I =^ i\it literary be not afraid!

2190. But as have is a transitive verb (2181), there is a
greater tendency to use the periphrastic forms with it than
with be. Thus we can say either what sort of a passage did

you have ? or what sort of a passage had you ?

2191. But in British-English we avoid the American-

92 SYNTAX. [§ai92.

English periphrasis in he does not have to work by the use of
the construction he has not got to work, although we feel
instinctively that the strong meaning of the have in this
case justifies the American construction.

2192. The negative form of let us go / is do not let us go I
with an anomalous heaping of auxiliaries. The literary form
is, of course, let us not got \ let us not be selfish J

2193. do, like the other auxiliaries, is used absolutely.
In some cases the absolute corresponds with the full conjoint
use, as m you said so yourself I did I? \ did you tell him ?
no, I did not compared with did I say so ? \ I did not tell
him.

2194. But the imperative and affirmative absolute do does
not show anything of the emphatic meaning of the corre-
sponding full forms ; thus shall I ask him ? dol \ did you tell
him? yes, I did correspond in meaning to the simple ask
him I I I told him.

2195. These absolute forms are, in fact, remains of the
earlier unemphatic affirmative d0'ioivcis>, which have been pre-
served by the habit of repeating the auxiliary of the preceding
sentence, the sequence did you tell him? yes, I did being
mainly kept up by the analogy of such sequences as will you
tell him ? yes, I will.

Will and Shall.

2196. In the future (present and preterite) the first person
is formed with shall, the others with will : —

Singular i / shall see I should see

2 you will see you would see

3 he will see he would see
Plural I we shall see we should see

2 you will see you would see

3 they will see they would see

Examples : / suppose you will not wait later than six. no.

§ aaoi.] VERBS : WILL AND SHALL, 93

I certainly shall not \ we were afraid we should be late \ I knew
how it would turn out.

2197. The same rules apply also — but with important
exceptions — to the conditional: / should like a glass of
water, would not you rather have a cup of tea ?

2198. As regards the origin of these forms, it is to be
observed that in Old-English the future is generally expressed
by the present, as in the other Old Germanic languages. But
the auxiliaries will and shall are used to express not only
futurity combined with the ideas of wish and compulsion re-
spectively, but also, in some instances, pure futurity : id wdt
pcEt pis folc miclum hlissian wile mines deapes * I know that
this nation will rejoice greatly at my death * | hie wendon pcet
hie sdolden mare onfon ' they expected to receive more.*

In Old-English the combination of the preterites would
and should with infinitives is frequently used like the modern
conditional as substitutes for the preterite subjunctive.

2199. In Middle English shall and will+ infinitive are used
as pure futures, shall being at first much more frequent than
will, will afterwards came into more general use, till at
last in many dialects— such as the Scotch — it has completely
banished shall.

2200. In Southern English, on the other hand, the origi-
nally unmeaning fluctuation between will and shall has
gradually developed into a fixed system of complicated rules,
which speakers of the other dialects have great diflficulty in
mastering.

2201. The present use of will and shall seems to be the
result of the desire to keep the original meanings of these
verbs as much as possible in the background. It is evident
that in the first person ' I must do it to-morrow ' suggests the
idea of futurity less ambiguously and more abstractly than
* I wish . . ,* because a mere expression of wish on the part
of a speaker does not necessarily imply any expectation of
its fulfilment; while his own statement of obligation or

94 SYNTAX, [§2309.

compulsion involves his belief that it will be carried out.
In the other persons everything is reversed. As we know the
wishes of others only by uncertain inference, we do not gene-
rally say * you wish . / or * he wishes . . / but prefer to put the
statement in a less direct form : ' I suppose you intend to . .*
Hence the bare unqualified statement j'<?« will go hardly sug-
gests the idea of volition at all, and so is excellently adapted
to express pure futurity.

2202. We will now consider the exceptions to these general
rules : —

{a) Unemphatic tutll and shall can be used in all persons
to express the idea of futurity combined with those of wish or
necessity respectively : I will come as soon as lean \you shall
see what I am going to do / The emphatic / ;will do it
expresses obstinacy, the emphatic / ;shall do it expresses
determination, as if the speaker meant to imply that his will
was so strong as to become a purely objective force. The
two may be combined : / ;shall and ;will do it. what ;shall
I do I expresses helplessness or perplexity.

{b) Such combinations as you and /, we two^ we three, we
all take will instead of shall: we shall get there first, but 1
expect you and I will get there first \ we two will he able to
manage it quite well \ I shall dream about those dogs to-night,
I am sure I shall, so shall I. so we all will. If we put the
all of the last example after the verb, the s?iall must be
restored : so shall we all. The explanation of this anomaly
is that you and the other words added to the we divert the
attention from the first person and make the idea of the
second person prominent enough to suggest the more frequent
will.

(r) In direct questions shall is used instead of will in the
second person, as in shall you be there ? compared with you
will be there, I suppose ? It is evident that our ignorance of
the will of others (2201) makes it perfectly natural for us to
ask questions about it, so that will you 6e there fo-nighi? is

§ aaofl.] VERBS : WILL AND SHALL. 95

easily taken in its literal sense * do you intend . . ?/ and hence
shall is substituted. Such questions must be direct: there
would be no motive in asking questions of B about C's will ;
hence the substitution of shall would be quite out of place in
the third person {will he be there to-night ?). In enclitic
questions — which, it must be remembered, are questions in
form only, not in meaning — a preceding would is always
repeated, or, in other words, the enclitic auxiliary is attracted
by the preceding dependent auxiliary : you will do it yourself^
will you ? I you would think sOj wouldn't you ?

(d) In dependent sentences of doubt shall is used in all
persons, as in ty he should come while I am out^ tell him to
wait compared with / wish he would come, because the idea
of doubt neutralizes that of compulsion. In the spoken lan-
guage this usage is preserved only in the conditional form,
but in the literary language it occurs in the future, where the
spoken language always has the present: if he shall call
while I am out^ ask him to wait-=iht colloquial if he calls . . |
whoever shall compare the country round Rome with the country
round Edinburgh, will be able to form some judgement as to the
tendency of Papal dominion^ where whoever shall compare . .=
' if any one compares . .' Note in this last example the
contrast between the shall of the sub-clause and the normal
will of the principal clause.

{e) In such a sentence as he says he hopes I will be there
compared with / told him I should be there the person of
* I ' is regarded from the point of view of * he,' as if the sen-
tence were in the form he said ' I hope you (or hi) will be
there! So also in {he said) he was afraid we would not {be
able to) come. In both of these instances shall {should) is
admissible, and would probably be substituted by many on
second thoughts, but the construction with will is the
genuinely colloquial one.

(/) In {he says) he wishes we would not keep the door open
compared mihyou seemed very anxious that we should go the

96 SYNTAX, [§ 2203.

would is obligatory because the will here denotes repetition
or habit, as in he will — would — sit for hours doing nothing.
This will is hardly ever used in the first person, but if so
used {he says I will sit for hours doing nothing)^ it would
necessarily be kept unchanged, for / shall would suggest the
idea of compulsion or futurity so strongly as to obscure
entirely that of habit.

Definite Tenses.

2203. The periphrastic forms corresponding to the Modern
English is writing, was writing are in frequent use in Old-
English, but are only vaguely differentiated from the simple
forms.

2204. They were no doubt originally formed on the
analogy of the combination of the verb * be ' with adjectives,
so that such a paraphrase as hie waron blissiende * they were
rejoicing ' was felt to be intermediate between hie hlissodon
' they rejoiced ' and hie waron blipe ' they were glad.'

^/ 2205. The most fundamental distinction between the
simple and periphrastic forms in Old-English appears to be
that the latter are associated with the idea of incompletion,
as in pa-pa he sprecende wees ' while he was speaking ' com-
pared with pd-pd he sprcBc, which may have the meaning ' was
speaking,' but may also have that of ' spoke,' or even ' had
spoken.'

2206. The natural result of this is that the periphrastic
forms occur very often in constructions which involve the
idea of continuity or progression. But that this idea is only
a secondary one is shown by those instances in which the
context excludes the idea of duration, as when the periphrase
is accompanied by the adverb s5na ' immediately,' as in pa
3ona on anginne pees gefeohtes wees se munt Garganus hifigende
mid orm&tre cwacunge.

2207. In this example, as in many others, the context
suggests — or at least admits — an inchoative meaning : * then

§ 3214.] VERBS : DEFINITE TENSES. 97

immediately at the beginning of the battle Mount Garganus
began to tremble with excessive quaking.'

2208. The analogy of the adjective construction (2204)
would make us expect to find the periphrastic forms used
mainly to express rest, and passive rather than active phe-
nomena. But, on the contrary, they are especially favoured
by verbs of motion and fighting — wcbs winnende, w&ron

feohtende — either with or without the idea of continuity.

2209. Here the periphrastic forms seem to be used mainly
to make the narrative more vivid and picturesque, so that
they have come to have what we may call a * descriptive '
force.

2210. In many cases they appear to have a purely stylistic
function, being introduced merely to round off a period, and
to avoid abruptness.

2211. In Modern — as in Old — English the definite tenses
always imply incompletion ; thus he is writing a letter implies
that the letter is not finished. This is still more marked in
the perfect, as in what have you been doing all day? compared
with what have you done to-day /, which really means * what
have you completed to-day ? '

2212. They also always imply a certain duration: they
are no longer used as point or inchoative tenses, as they
sometimes are in Old-English.

2213. But the expression of duration is not their primary
function in Modern any more than in Old English. Nor can
they be used to express unlimited duration or repetition :
this is expressed by the indefinite tenses, as in the moon shines
at night \ he goes to Germany once a year.

2214. The characteristic of these tenses is that they use
duration to define the time of a point-tense, as in when he
came, I was writing a letter. Here the action of writing is
supposed to be going on before the point of time indicated
by came, and to continue after it, the amount of the further

VOL. II. H

98 SYNTAX, [§ 2215.

duration of the action backwards and forwards being in-
different.

2215. A definite tense therefore often makes us expect
a clause containing a verb in the corresponding indefinite
tense to indicate the point of time which the definite tense
serves to define, as in / shall be writing when he comes
[= when he shall come], unless the clause precedes, as in
when he came, I was writing.

2216. But in the present the definite tense does not require
or admit of any accompanying point-tense, for the definite
present is, in the nature of things, self-defining : lam writing
a letter means ' I am writing a letter at the present moment
(of your coming).' So also with the definite perfect : / have
been writing a letter.

2217. Hence we may say that the definite present and
perfect are absolute tenses, while the definite preterite and
future are relative tenses, because they make us expect
another clause. This clause may of course take the form of
an independent sentence, if the connexion is clear : we were
expecting you yesterday ; why didntyou come ? Or the point-
tense may be inferred from the context — sometimes only in
a very vague way, as in / was coughing all night long, which
is almost as absolute as / coughed all night long. Here it
will be observed that / was coughing differs from / coughed
mainly in emphasizing the idea of duration.

2218. There are some verbs which occur only in the
indefinite tenses. This is especially the case with verbs
which express feelings, physical and mental perceptions etc.,
such 2i.s/eel, like, think : I /eel ill \ he likes being here \ I think
so. But as soon as the element of volition or action becomes
prominent, the definite tenses re-assert their rights : compare
it hurts with he is hurting him ; he doesn't see it with he is
seeing the sights ; / hear a noise with / am hearing lectures.
This seems to be a tradition of the Old-English descriptive
use of these tenses (2209).

§2222.] VERBS: DEFINITE TENSES. 99

2219. The tense defined by a definite tense may itself be
in the definite form : all the while I was writing there was
some noise or other going on : the children were having their
music-lessons, and the baby was crying next door. Here the
definite tenses imply that the writing and the noises were all
simultaneous.

2220. But when the connexion between the two sentences
as regards time is not specially intimate or important, we
prefer to put only one verb in the definite form, as in she
stood in an impatient silence while she was thus being talked
over^ where she was standing, although strictly correct, would
lay too much stress on the logically subordinate idea of
* standing.' So also in as he walked home, his heart danced
within him there is no special connexion of cause and effect
between the two clauses.

2221. When such words as always, constantly are added
to a definite tense, it necessarily loses its definiteness of
meaning as regards distinctions of time : she is a good woman :
she is always going to church ; she is ahvays doing things for
poor people. So also in the preterite and future : your mother
was a good woman : she was always going . . | she will grow
up to be a good ivoman : she will always be going . . The main
use of the definite form in such collocations is to make the
statement absolute ; thus he is always complaining \ he was
grumbling all the time we were there can stand by themselves,
while the corresponding indefinite forms make us expect
something to define the time or show that repetition is
implied : he always grumbles when he is at home.

2222. When a definite tense is used in a context implying
repetition, the definite tense does not share in this meaning
(for repetition is expressed by indefinite tenses, § 2213),
but keeps its own ; thus his temper only failed him when he
was being nursed means *on each occasion when he was
being nursed ' — that is, the definite tense applies to each of
the repeated phenomena singly.

H 2

lOO SYNTAX, ■ [§2223.

Tenses in Detail.

Present.

2223. The indefinite present is a neutral tense (289), im-
plying that a statement is of general application, and holds
good for all time {f/ig sun rises in the east), or that an action
or phenomenon is habitual, as in he gets up at six regularly
every morning \ I always get it at the same shop, or recurrent,
as in he goes to Germany twice a year \ whenever she sees him,
she begins to laugh.

2224. If the actual present is meantj the definite form is
used : he is getting up now \ where are you going ?

2225. The definite present is also used as a neutral
present to show that continuity and not repetition is meant
(2213). Thus if in such a sentence as the earth is a ball
that is always turning round, and at the same time it moves
round the sun in a circle we substituted turns, we should have
to answer the question ' when does it always turn ? '

2226. For the use of the indefinite moves in the last clause
see § 2220. But here we cannot say that the idea of revolving
round the sun is either subordinate to or not closely asso-
ciated with that of revolving on its own axis. It seems that
the shorter indefinite form is used because the context makes
the meaning ' is moving ' quite clear. Similarly in the wind
is rising: look how the smoke blows sideways/ Here we
might use the definite form is blowing. The indefinite form
seems to suggest 'smoke always blows sideways when the
wind rises.*

2227. The vagueness of the indefinite present makes it
possible to use it in constructions where we should expect
a non-present tense. Thus instead of saying / (have) heard
that you made a speech yesterday, we might say I hear you made
a speech yesterday, implying ' various people told me so and
others will probably tell me so afterwards.' Both construe-

§2331.] VERBS : PRESENT,

lOl

tions would be naturally followed by the question who told
you so ?

2228. Statements in narratives can be looked at from the
point of view either of the narrator or the hearer ; hence the
present in such constructions as Gibbon tells us in his History
that . . I what is the story about i^ it is about a young man who
goes to London and makes his fortune. In the last example
the present is extended from the act of narration to what is
narrated. This is the germs of the ' historical present,' used
to give greater vividness to a' narrative: he mounts the
scaffold . . the executioners approach. The fully developed
historical present seems to be due to Old French and Latin
influence.

2229. In describing the subject of a picture or piece of
sculpture — which appeals directly to the eye — the definite as
well as the indefinite present can be used : it is a represen-
tation of a lady, she is lying on a couch, at the side of the couch
sits a woman as in grief The definite" form could not be
used in stating the contents of a book, the plot of a story
etc. Observe also that we can say it is a picture of a lady
lying on a couch, while we should have to use a separate
clause in /'/ is a story of a lady who . .

2230. In such a construction as the moon halts opposite to the
window at which I sit — / write the indefinite present is used
to show that the speaker is making a statement which will
not reach those for whom it is intended till it has come to
refer to the past. The definite present might also be used,
but would imply a hearer present at the time, or else that the
narrator is, as it were, speaking to himself, not to others.

2231. The indefinite present is regularly used instead ol
the future in clauses dependent on a sentence which contains
a verb in the future, as in if it is fine, I will come early,
although the written language often substitutes the future
(2202 d). The present is also used instead of the future in
some independent sentences. In does the moon shine to-night i*

I02 SYNTAX, [§ 2230.

futurity is only indirectly implied by continuity or repetition,
the sense being ' is to-night one of the nights on which the
moon shines ? ' The full future meaning of the present in
independent sentences is most frequent when futurity is
clearly indicated by such a word as to-morrow : to-morrow ts
bank holiday [this falls, at the same time, under the same head
as the preceding example] | he star is for the Continent to-night.
In I bet you anything I'll do it! futurity is not marked in the
principal clause because of the necessity of keeping will in
the subordinate clause to' indicate the combination of will
with futurity.

2232. The definite present is also used in a future sense,
but only in combination with verbs of motion : where are you
going for your holiday this autumn? \ I am going home
to-morrow: it is (2231) my mother s birthday \ is anyone
coming to dinner ? = shall we have any guests to dinner? It is
evident that where are you going ? — that is, ' where are you
setting out for at this moment ? ' — implies * what place will
you arrive at ? ' Hence, while are you going to church ? is
present if addressed to one who is just starting, it is
necessarily future in arc you going to church to-day ?

Preterite.

2233. The preterite is used in the sense of the neutral
present in clauses dependent on a sentence whose verb is in
the preterite : people used to think the earth was flat \ I think
I once heard you say you liked it.

2234. In some cases it is used almost as the equivalent of
a full present : / have {got) a headache. I thought you were
not looking well \ I hardly ever see him now. I thought he
was an intimate friend of yours, oh no I In these examples
thought may be taken in the sense of * I thought when I first
saw you just now . . ,* *I thought, until you told me the
contrary, that . . ,' but it has really a present meaning — I

§2241.] VERBS: PERFECT. I03

thought you were not looking well =^ yes, you don't look well '
— and the preterite is used simply to avoid the suggestion of
contradiction ; for / think he is an intimate friend of yours
would imply ' I think so in spite of what you say/ while the
preterite implies that the belief is already a thing of the past.

2235. The preterite is used in many cases where we might
substitute the perfect with but slight change of meaning
(275). Other examples are: who took my book? \ I had
hardly any breakfast, but I do not feel at all hungry.

2236. It can be used in the same way as a substitute for
the pluperfect: one morning when they woke up, they saw
a ship at anchor in the bay \ the donkey never stopped till he
came to a tent of gypsies. Indeed the pluperfect could not be
used in either of the examples, the two events being regarded
as a simple sequence : * they woke up, and then they
saw . /

2237. The definite preterite is sometimes equivalent to an
indefinite preterite with duration implied (2217).

2238. It is sometimes equivalent to the definite perfect :
/ am afraid we have kept you waiting, oh no, not at all: we
were looking at these photographs. Here we were looking
refers to an implied while we were waiting for you, and so is
equivalent to we have been looking.

2239. It follows from § 2232 that the definite preterite of
some verbs is used in the sense of the future preterite : one
day he told his housekeeper that four gentlemen were coming to
dinner.

Perfect.

2240. The perfect, being intermediate between the
present and the preterite, is sometimes used in constructions
where we might expect one or other of these tenses.

2241. It is equivalent to the present in I have got a cold=
' I have caught a cold/ ' I have a cold.' In / have brought
back the book you lent me ; that is why I have called we could

I04 SYNTAX. [§ 224a.

not substitute the present for the last perfect. In a letter we
should say / enclose a receipt \ I herewith return the book
which you lent me because at the time of writing they have
not yet been received.

2242. The perfect is often used instead of the preterite to
express something which, although already detached from the
present, is connected with the present in thought : thank you

for the trouble you have taken \ you have not tied it tight
enough: it is sure to come undone again \ waiter I there is
something wrong in the bill: you have made this sixpence into
six shillings. But in the last two the connexion with the
present is shown to some extent by the accompanying
sentences.

2243. The perfect is used instead of the future perfect in
clauses dependent on a sentence with a verb in the future, as
in by the time you have washed and dressed, breakfast will be
ready, and in other cases where the future meaning is clear
from the context : when will you come again ? as soon as I
have finished my work \ I bet you half-a-crown that before
nightfall I have seen him 1

2244. The definite perfect (and pluperfect) emphasizes
the idea of duration up to the present moment (2211).

2245. But as the element of duration is not essential to the
definite tenses, it often implies something that has happened
immediately before the present time, either a succession of
detached events, as in where have you been meeting her ?
[compare : have you met her lately ?\ or merely a single event :
his bruised face and torn clothes showed that he had been fighting
— had just beeti fighting \ I hear you have been getting into
mischief again. But the last may also imply repetition.

2246. But this usage generally requires that the verb itself
implies — or at least admits of — the idea of duration. Thus
in I have just received a letter from him we cannot substitute
the definite form. But we can say / have been receiving
letters from him, which necessarily imphes repetition.

§2251.] VERBS: PLUPERFECT. 105

Pluperfect.

2247. We have seen (2236) that the pluperfect is often
expressed by the preterite. Conversely the pluperfect is some-
times used where the preterite would do as well : he had {got)
a cold I / did not think we had been so near Scotland \ he told
them he had gone for a little walk, and saw a donkey. . In the
second example the pluperfect is more graphic than the pre-
terite, as heightening the surprise by the reminder that it was
too late to take advantage of the knowledge. In the last
example the pluperfect is justified by the fact that the going
for a walk preceded seeing the donkey, and it is used here
because the seeing the donkey is the really important event,
to which the pluperfect makes it subordinate. In be/ore
breakfast they had settled the whole thing the pluperfect, if taken
literally, makes the exaggerated statement that at a point of
time before breakfast the thing had been settled, whereas all
that is necessarily meant is that it had been setded at break-
fast— that is, that they settled it before breakfast.

2248. The definite pluperfect is parallel to the definite
periect (2244).

Future.

For the use of the present instead of the future see § 2231.

2249. The future is sometimes used instead of the present
in such phrases as this will be the Tower 0/ London, I suppose?
= I suppose this is the . . , meaning, of course, ' this will turn
out on investigation to be . / This usage appears to be
dialectal (Scotch).

2250. The definite future has its normal meaning in the
following example : / shall not be at home much next summer:
I shall he travelling about on the Continent most of the time
parallel to the definite present in he is not at home now : he is
travelling about on the Continent.

2251. But it has often peculiar shades of meaning of its

Io6 SYNTAX, [§2253.

own : it generally gives the impression that the future event
is the result of causes with which the speakers have nothing
to do; and hence is often used to make the expression of
futurity more abstract, and especially to do away with any
associations with the special meanings of the auxiliaries.
Thus even in the preceding example, / shall travel may be
taken to imply that the travelling will be the result of a reso-
lution already formed, while / shall be travelling predicts it
as a purely objective phenomenon. So also / suppose you
will he going hack to England soon means 'you will probably
go back to England soon in the natural course of events/
Hence it often implies that the future action is the natural
result of something in the character of the person of whom it
is predicted : and novo I knovu you will be saying you cannot
afford it/\I expect some fine day he will he making off with
the money.

Preterite Futiire.

2252. The following are examples : // was settled thai we
should meet next morning at the same place \ I knew how it
ivould turn out.

Of the definite form : / knew that as soon as my father got
to Switzerland^ he would he wanting to push on to Italy,

Perfect Future.

2253. The following are examples : by this time to-morrow
I shall have crossed the Channel \ he wants you to post a letter;
he will have finished it by the time you are ready,

2254. Of the definite form : / shall have been writing for
six hours without stopping by the time you come hack.

Immediate Future (going to . .).

2255. In English we have an immediate future (284),
formed with the definite tenses of go and the supine, as in
I am afraid it is going to rain-=* . . it is about to rain,* * . . it
is on the point of raining,' compared with / am afraid it will

§3359.1 VERBS: SUBJUNCTIVE. 107

rain to-morrow. Other examples are : what are you going to
do noiv ? I am going to call on some ladies : will you come too ?
Sometimes it implies a less degree of immediateness : / am
going to call on him soon : I shall call on him as soon as I can
—in a/av days. The immediate future might be used in the
last sentence as well.

2256. This form is perhaps sometimes used — like the
definite future (2251) — to avoid the special associations of
will and shall. It certainly serves that purpose in the last
example but one.

2257. It is sometimes used not to disguise, but merely to
soften down the idea of will or compulsion : well, if you are
not going to have any more wine, we may as well go into the
garden— you are sure you will not have any more wine ? Here
will might imply an accusation of obstinacy.

2258. The above examples are all of the present imme-
diate. There is also a preterite : what were you going to do
when I came in ? I was going to take a walk. There do not
seem to be any other tenses in actual use.

Moods.

Subjunctive.

2259. In Old-English the subjunctive mood is in full
use as a thought-mood (298), very much as in Modern
German. Thus it is regularly used in indirect narration :
hie cwcedon pcet he ware god cyning * they said that he was
a good king.' But when the indirect statement is perfectly
certain in itself, and not merely accepted on the authority of
the speaker, it is put in the indicative : nU we willap s^cgan
bcBt we /use sind Urne eard to sedanne * now we wish to say that
we are starting to seek our country '=' we have come to take
leave.' In this sentence *say' has practically hardly any
meaning, the whole sentence being equivalent to we sind
fuse . . , so that the subjunctive would be too emphatic : it

io8 SYNTAX, [§2260.

would seem to imply that the speakers wished to hint that
their statement was false.

2260. Such exceptions as this helped the natural tendency
to get rid of superfluous distinctions (307), which, again, was
helped by the Late West-Saxon levelling of the distinction
between subjunctive and indicative in the preterite plural of
all verbs (1188).

2261. Hence in Middle English the subjunctive was soon
disused in many constructions — first of all in indirect narra-
tion ; and in the present English it is practically extinct as
a living form, surviving only in a few isolated constructions.

2262. Thus we still keep it in some independent sentences
of wish and command : God save the Queen / | God bless you ! \
bless me I \ so be it then I

2263. The phrase woe betide . . expresses not wish, but
fear of the future, in such a construction as woe betide us if
we are late!

2264. The only subjunctive form that is in regular
colloquial use is the preterite were. It is used in dependent
sentences, chiefly in clauses of rejected condition (305) : if
it were possible, I would do it \ J would not do it if I were
you.

2266. Also in clauses of hypothetical comparison : he is
always talking about honesty, as if he were the only honest man in
the world. Here, again, it implies rejection of the statement.

2266. Lastly, it is used after verbs of wishing to imply
rejected fulfilment — I wish I were at home — as opposed to
open fulfilment, as in / wish to go home, the first example
implying * I am not at home now.' So also in she says she
wishes she were dead \ would that I were free/

2267. The same implication of rejection is inherent in the
other colloquial uses of were : it is time we were gone — off.

2268. In the colloquial language of the last century there
was a tendency to substitute was for were, even in clauses of
rejection.

§2273] VERBS: SUBJUNCTIVE. 109

2269. In Early Modern English the subjunctive was still
in colloquial use in many constructions where it is now
obsolete.

2270. In the present literary language the subjunctive is
rapidly falling into disuse — except, of course, in those con-
structions where it is obligatory in the spoken language.

2271. It is otherwise obligatory in the written language
only in cases of verb-inversion in conditional and concessive
clauses, not only in such as were he my brother . . = ' even
if he were . . ,' but also such examples as were it scroll or
were it book, into it, knight, thou must not look (Scott) | come
what may, I will stand by him ! The last may perhaps be
still used colloquially.

2272. Otherwise the subjunctive is little used in ordinary
literature except in the case of be, and that in most cases
only in combination with if. But even within these limits
the usage varies greatly. Those who have been trained in the
use of the subjunctive by Latin prose composition and by
familiarity with its use in the earlier English literature would
instinctively avoid it in sentences which are only slightly
hypothetical, or hypothetical in form only, such as if there is
a thunderstorm some way off, we see the lightning some time
before we hear the thunder — ' when there is . . / especially in
familiar expressions such as she is thirty, if she is a day.
Such writers would, on the other hand, at least try to use it
consistently in such constructions as if it be necessary —
which I very much doubt . . Some of them will occasionally
use it with other verbs than be^ as in tfhe please one, he will
offend the others, and in other constructions than those with
if. not only in constructions similar in meaning to those with
if, such as you are in your sphere, humble though it be \ they
will not do it unless he bid them, but also in other reminiscences
of the earlier freedom : —

2273. / pray thee that thou assert my innocence \ look that
he hide no weapon \ see that there be no traitors in your camp !\

no SYNTAX, [§2274.

he /eels if the axe he sharp \ I know not whether it be true or
not I the tree will wither long before it fall \ wait till he come I

2274. In the above examples the spoken language simply
substitutes the indicative. In other constructions it substitutes
a periphrastic form. Thus in the following it requires the
conditional : it is better he die = ;'/ is better he should die \
I tremble lest he be discovered \ to do so were unnecessary. In
the following it uses the permissive : give me leave that I may
turn the key, that no man enter =. . . so that no one may come
in I to act that each to-morrow find us farther than to-day |
lest justice depart out of the land = so that justice may not
depart . .

2275. But the general literary tendency seems to be to
use be and is sporadically and almost at random after if the
uncolloquial subjunctive being regarded simply as a mark of
the higher style, which is therefore unconsciously dropped in
the more familiar passages.

2276. The artificial subjunctive is particularly objectionable
when the mechanical substitution of ivere for was after if
leads to confusion between open and rejected condition, as
when were is substituted for was in such sentences as if he
was there, I did not see him. In the natural spoken language
the indicative would always be kept here, even if the doubt-
fulness of the hypothesis were made as prominent as possible
by distributed stress (1896) — if he ;was there . . if he were
there ought, of course, to imply ' he is not there,' and must
then be followed by the present conditional (2280), so that
the sequence of tenses in if he were there ^ I did not see him
makes nonsense.

Conditional.

2277. The present conditional has the same form as the
preterite future, and the preterite conditional has the same
form as the perfect future. The two tenses of the conditional
have also definite tenses, like the parallel forms of the future.

§228i.] VERBS: CONDITIONAL. Ill

2278. The main practical distinction between the preterite
future and the conditional is, of course, that the latter is
dependent on a preterite, so that if that preterite is changed
into a present, the would of the preterite future necessarily
becomes will : he said he would come — he says he will come,
he would come when used as a conditional can, on the other
hand, be associated with a present.

2279. Under 'conditional' we include all combinations of
would and should -with infinitives which are not clearly futures,
even when their functions are not really conditional. But as
they all agree in being moods rather than tenses, the absolute
practical necessity of separating the mood- from the tense-
functions of these periphrases makes it all the more desirable
to avoid further subdivision.

2280. The most important function of the conditional is
in sentences of rejected condition. These have a tense-mood
(301) in the hypothetical clause, and a conditional in the
consequence-clause, the logical present being expressed by
the present conditional (with the preterite indicative — sub-
junctive in the case of were — in the hypothetical clause), the
logical preterite by the preterite conditional (with the pluper-
fect indicative in the hypothetical clause) : —

Logical present : if I were you, I would not do it \ he would
tell me if he knew — but he does not know anything about it \
I should like to try if I were not afraid \ if he were travelling
on the Continent, he would not be seeing the sights in London.
In if he knew it^ I do not know what he would do the con-
sequence-clause is not I do not know, but is contained in the
clause dependent on this one = he would do I know not
what.

Logical preterite : if we had started in proper time, we should
have been there by this time \ it would have been better if you
had written first, to find out whether he was at home or not.

2281. The conditional is here a substitute for the original
subjunctive, which was still preserved in Early Modern

112 SYNTAX, [§3282.

English ; thus the Old-English gtf pit wkrc her, nctre rnin
hropor dead appears in the Bible as if thou hadst been here,
my brother had not died. We see from this example that it
was the ambiguity of the consequence-clauses which led to
the general use of the conditional, the germs of which may
be found in Old-English itself. The original subjunctive
construction is still used in the higher literature wherever it
can be done without ambiguity, especially, of course, in the
case of were: it were a pity . . But even in the highest
literary style such a construction as ^he told me if he knew
would be impossible.

2282. But the archaism is still necessarily preserved — in
colloquial as well as literary English — in the case of those
anomalous verbs which have no infinitives, and consequently
no conditionals : he could do ii if he liked — he would be able
to do . .\ if we had started in proper time, we might have been
there by this time.

2283. In many cases the present conditional does not
imply rejection of the hypothesis, but simply shifts its conse-
quences into the future, as in if we missed the train, we
should have to wait an hour at the station, which means if we
miss the train — which I hope we shall not — we shall have to
wait . . The only way of definitely rejecting the hypothesis
is by putting the pluperfect in the hypothetical clause: if we
had missed the train, we should {now) have to wait an hour.
Here we have a case in which a pluperfect in the one clause
is accompanied by a present instead of a preterite conditional
in the other, because the action of state denoted by the verb
in the conditional is not yet completed ; compare if we had
missed the train, it would have been rather awkward [at the
time when we missed it).

2284. We see from this example that hypothetical non-
continuous phenomena require the pluperfect to show dis-
tinctly that they are unreal. It is indeed difficult to think of
them in the present at all : it is at any rate more important

§2290.] VERBS: MOODS. TI3

for us to know whether or not we have caught a train than
to realize that we are engaged in trying to catch it at the
present moment.

2285. The consequence-clause of conditional sentences
is often used absolutely with a variety of meanings. It is so
used to express a modest wish, request, or question, some such
hypothetical clause as ' if it were possible,' * if you will allow
me — give me ' being understood : / should like a glass of
water, wouldtityou rather have a cup of tea? \ he says he
would like to go for a walk. I should like to go top.

2286. In some phrases this construction has come to
express indignant assertion — at first, probably, ironically:
I should {rather^ think not!

2287. In some cases the absolute clause has almost a
future meaning : (/ think) you would like it {if you were to
try if). {I am sure) I should like it — // would suit me exactly \
I am going to call on some ladies : will you come too ? (/ have
no doubt) they would be very happy to make your acquaintance.

2288. The absolute conditional — which, of course, does
not imply rejection of any kind — is necessarily kept un-
changed in preterite constructions as regards tense, but with
a tendency to substitute should for would', he said he should
(or would) like to go for a walk \ he said he was sure he
should like it. But would must be kept in / said I was sure
he would like it.

2289. In a preterite context the absolute conditional
sometimes expresses wliat is to be expected, what happens
as a matter of course : men began with ready-made tools,
they would soon learn how to cut and scrape with a sharp
piece of flint, and make holes with a sharp tooth.

2290. The absolute conditional is sometimes used as
a cautiously expressed or ironical present : in short you must
know very well that Mr. and Mrs. Boffin are just absolute
perfection, truly it would seem that we are required to think
so. We can al^o use it should seem in the same sense.

VOL. II. I

114 SYNTAX. [§2291.

2291. In some hypothetical clauses the conditional form
is used to express indefinite futurity, as in if he should call
while I am out, tell him to wait \ if you should happen to see
him, tell him to expect me this evening about eight, the same
form being used in a preterite context as well : she said, if ht
should call while she was out, I was to tell him to wait. Here
the conditional is used not in the consequence-clause, as in
sentences of rejected condition, but in the hypothetical clause.
If the conditional is used in both clauses, the general char-
acter of the hypothesis is determined by the hypothetical
clause; that is, the whole sentence implies hypothetical
futurity, not rejected hypothesis : if he should see me, he woula
know me ■=z if by chance he sees me, he will recognize me.

2292. In such a phrase as you should not make personam
remarks the preterite should is substituted for the present
shall in order to soften down the imperativeness oiyou shah
not make . . Here there is no conditional meaning; the
should keeps its original meaning, and is not even an
auxiliary.

2293. In the following examples we can also observe the
original meaning of shall, but softened down so that the shall
becomes a pure auxiliary: why should you suspect him ? \ ii
there anyone with him / no ; who should there be ? \ as Iwem
down the street, who should I meet hut our friend himself i
We can still see the influence of the original meaning in the
first example ('what obliges you to suspect me?'), furthei
softened down in the second, till in the last should meet
becomes simply a periphrastic preterite.

2294. It will be observed that all these examples are
interrogative in form only. In the following we have a
similar vague use of the periphrasis in declarative sentences ;
it seems odd {that) we should meet here \ it is strange they
should have met in the very same place.

2296. In the following examples we can again observe
clearly the original meaning of shall', it is not fair thai

§3301.] VERBS: MOODS. I15

/ should suffer for other people's misconduct \ it is quite right
that he should bear the expense.

2296. But in the following parallel construction the
original meaning of the auxiliary is quite obscured : / am
sorry to think {that) a son of mine should behave so badly at
other peoples houses — should have behaved so badly {on that
occasion).

Compulsive (is to . .).

2297. This periphrasis primarily expresses necessity and
obligation : when am I to come again? \ do not forget that you
are to be there exactly at five. So also in the passive : you
know what is to be done \ the doctor says he is not to be
worried.

2298. The passive form is also used to express possi-
bility : where is it to be found/* \ he is not to be found any-
where.

2299. In a hypothetical clause the preterite was to . . does
not, as the simple preterite would do, imply rejection of the
hypothesis so much as its remoteness or improbability, as in
what should we do if it were to rain = * if it by chance comes
on to rain.' So also compare tf I were to see more of him,
I could speak to him about it with tf I saw more of him . .
While the second implies ' I do not see much of him,' the
first implies * I may see more of him in the future.'

2300. There is also a peculiar traditional use of this form
in the phrase that is to say : he was very eccentric, that is to
say, he did odd things that made people laugh.

Permissive (may).

2801. The verb may in its full meaning implies possibility
as the result of the absence of external hindrance, especially
through the interference of others, whence its frequent
meaning of ' have permission,' * be allowed to ' ; while can

I 2

Ii6 SYm-AX. [§2302.

implies possibility as the result of something in the subject
of the statement, such as strength, capacity, or knowledge :
may I climb that tree ? yes : you may, if you can.

2302. In the above example may is a full verb ; but in
many cases the combination may or »2z^^/+ infinitive is used
as a true mood having much the same function as the old
subjunctive.

2303. Thus the present permissive is used in independent
sentences to express wish : may you succeed! \ may it please
your Grace to hear me I In this construction there is now
always verb-inversion, which is not always the case in Early
Modern English. This construction is not much used in the
third person (singular), because the simple subjunctive is here
distinctive enough (2262) — [may) God bless you /

2304. The combination wz^ 4- infinitive is sometimes used
as a kind of future: this place is not safe: the roof may come
down any day. it is safe enough for the present : the roof may
not come down yet. It will be observed that in the negative
consti-uction of the last sentence the original meaning of may
is almost lost — may not come down here = * will probably not
come down.' In itself it might, of course, also mean * is not
allowed to come down.'

2305. In the same way the combination »z/^/^/+ infinitive
in independent sentences is used to soften a request by making
it more indirect, as in inight I ask . . ? \ might I be allowed to
ask . .? — ' may I ask . . ? ' Here may is still a full verb.
Compare the parallel use of should (2285).

2306. In dependent sentences the permissive is used to
imply various degrees of impossibility and uncertainty.

2307. In the following concessive sentences the indicative
forms may be substituted without much change of meaning,
and yet the original full meaning of the may is still felt:
although it may seem incredible, it is nevertheless true \ incredible
as it may seem, there is no provision for the teaching of phonetics \
whatever the reason may be, the fact remains \ whatever hts

§3313.] VERBS: VOICE. 117

former conduct may have been, his circumstances should exempt
him from censure now.

2308. In the following object-clauses may keeps much
of its original meaning, and cannot be omitted : / beg thai
I may not be interrupted \ oh that I might recall him from the
grave I

2309. In the following object-clauses the indicative may
be substituted without much change of meaning: I hope it
may be so \ they were a/raid— they thought he might have been
carried off by gypsies = they were a/raid he had been carried off.

2310. The use of the permissive is especially important in
clauses of result and purpose : let the dog loose that he may
have a run \ we let the dog loose that he might have a run \
let us hide the brandy for fear he may drink it all up \ we put
the milk on the shelf for fear the cat might get at it \ speak, {so)
that I may hear you / In the present spoken language we
can no longer substitute the indicative or subjunctive — such
constructions as lest he drink are literary and archaic.

2311. In the preterite we can substitute should for might :
for fear he should drink it all up. Here should suggests the

idea of the result being in itself inevitable, or as probable,
while might suggests it only as a possibility. It is to be
observed that even in the literary language the periphrastic
forms must be used in the preterite ; we cannot write */^x/ he
drank.

Voice (Active and Passive).

2312. The definite active forms are occasionally used in
a passive sense : that house has been building a long time \ there
is an answer waiting = * . . being waited for.' This is the
result of the Modern English gerunds having originally been
abstract nouns (1257), which, of course, are neutral as regards
the distinctions of voice.

2313. In Old-English only transitive verbs could be used
in the passive. Verbs which governed any other case than

Il8 SYNTAX. [§2314.

the accusative could not be put into the passive. Thus there
is no passive form corresponding to he pancode hire ' he
thanked her.' But as soon as the distinction between dative
and accusative was lost, it was inevitable that from the active
he thanked her should be formed the passive she was thanked.
To us, thank is as much a transitive verb as praise. But we
still hesitate over and try to evade such passive constructions
as she was given a watch \ he was granted an audience htc^MSt
we still feel that she and he are in the dative, not the accusative
relation.

Infinitive and Supine.

2314. Of the large number of verbs which take the infini-
tive in Old-English the greater number are now followed by
the supine.

2315. The substitution of the supine for the infinitive
began in Old-English itself. Thus the supine of purpose, as
in hte comonpcet land to sceawienne ' they came to spy out the
land,' gradually supplanted the older infinitive with many
verbs of desiring, intending, attempting, etc., so that while
such a verb as willan ' will * continued — as it still does in
Modern English— to take the infinitive only, other verbs of
similar meaning, such as wilnian * desire,' together with such
verbs as on-ginnan * undertake, begin,' began to take the
supine as well as the infinitive. In Middle and Modern
English the gradual loss of the inflections of the infinitive
contributed further to extend the use of the more distinct
supine.

2316. In Modern English the auxiliary verbs always take
the infinitive. So also do most of the defective and analogous
verbs, such as can^ must, dare [but / dared him to do it\. But
need has both infinitive and supine, and ought has only the
supine, have takes the infinitive not only when an auxiliary,
but also in its other transitive uses {^hat would you have him

§2325-] VERBS: INFINITIVE AND SUPINE. II9

do?), but takes the supine in the sense of 'must': j/(7« will
have to do it.

2317. The full verbs that take the infinitive are mostly verbs
of feeling and perception, such ?is/eel {I felt my heart heat\
see (I saw him go out), hear, find [find pleasure end in pain).
But many verbs of this kind take the supine : I perceived him
to be . .

2318. The three verbs bid (now obsolete in the spoken
language), make {;we made him come in), let still keep the
infinitive, while the others of similar meaning have the supine.

2319. The infinitive is also kept after the groups had better,
had (now would) rather : we had better go home now. I woula
rather stay a little longer.

2320. It is to be observed that the supine is regularly used
after passive verbs — including those' which in the active take
the infinitive : he was heard to say . . \ he was made to come
[but it need not be seen^.

2321. The infinitive is sometimes used absolutely : what,
not know me I \ why not go there yourself i* | why complain ?
But as the infinitive has no distinctive form of its own, it is
a question whether we do not rather feel it to be the mere
neutral verb-base in such constructions. We feel the vague-
ness of the form still more in such groups as cough-no-more
lozenges, where the verb may also be felt to be an imperative.

2322. In rather than — sooner than yield he resolved to die
the infinitive is not absolute, but is dependent on the follow-
ing verb, the to of the supine being omitted on the analogy of
such constructions as he would rather die than yield.

2323. The supine is used absolutely in the subject-relation :
to be good is to be happy \ it is pleasant to see oneself in print,

2324. It is also used elliptically : but how to get in ! {that
is the question). Hence it is so used in adverbial phrases : to
tell the truth, I do not know much about it \ to be sure {it is) /
= 'certainly.'

2325. In Old-English the supine is used in a passive sense

I20 SYNTAX, [§2326.

to express what must be or ought to be done : paping pe to
donne sind ' the things which are to be done/ We still keep
up this passival use in the phrase a house to let ; but as we
cannot do this with other verbs, we have to use the passive
form in such constructions as this house is to he let or sold^
whence there is a tendency to say a house to be let. Originally
these passival uses were probably simply ambiguous : td
donne meant indifferently ' for some one to do ' or ' to be done
by some one/

Gerund.

2326. When the supine is substituted for the gerund in
the subject-relation, it seems to bring out more strongly the
attributes of phenomenality — action and quickness; thus to
see is to believe means * seeing is immediately followed by be-
lieving,' while seeing is believing means * seeing as a general
rule is followed by belief/ We could hardly substitute the
gerund for the infinitive in to know him is to love him without
weakening the sense — still less in to be or not to be, that is the
question. In there is no getting rid 0/ him the gerund must be
used because the supine cannot take such an adjunct as no.

2327. Some verbs, such as like, prefer, can take either the
supine or the gerund in the object-relation. Here, again, the
general difference between / like to get up early and / like
getting up early seems to be that the latter implies duration
and habit. But it is often difficult to see any distinction.

2328. In the combination possessive -|- gerund, as in / do
not like his coming here so often, the oblique case may be sub-
stituted for the possessive, so that the gerund becomes a
present participle : / do not like him coming here so often.
The difference — if any — appears to be that in the former
construction the logical emphasis is on the possessive, in the
latter on the verb. But there seems also to be a tendency
to give up the latter construction altogether, as if it were
a mere variation of / do not like him to come here so often. In

§8334-] VERBS: GERUND. 121

the following examples we could hardly alter the possessives :
in honour of its being Christmas day \ when metal came into
use, men were able to make their knives much longer, without
their being afraid of their breaking. In the last sentence the
their could be omitted, but not changed into them.

2329. So also the genitive in who toldyou of your wife^s
being there ? may be made into the common case — of your
wife being there. In such constructions as / cannot accept the
notion of school-life affecting the poet to this extent the common
case is preferred to the genitive.

2330. Although the ing-ioxm after the objective or common
case is formally a participle, we certainly do not feel that
coming in / do not like him coming here modifies him in the
same way as it does in / saw him coming : coming in the
former sentence is, in fact, a half-gerund.

2331. As we have seen, we recognize the gerund element
in the former sentence by our instinctive tendency to regard
him coming as a substitute for his coming. It is important to
note that the absence of a distinction between common case
and genitive in the plural often makes it impossible in the
spoken language to distinguish between gerund and half-
gerund, as in to prevent the ladies leaving us, I generally
ordered the table to be removed (Goldsmith), where the purely
orthographic alteration of ladies into ladies* would make
leaving into a full gerund.

2332. But leaving in this sentence could also be made
into a full gerund by making it inio from leaving. In pardon
me blushing we could in the same way either change me into
my or insert/br.

2333. Indeed, there seems little doubt that the colloquial
half-gerunds in such causal constructions as she caught cold
sitting on' the damp grass | he tears his clothes climbing trees
have arisen through dropping a preposition.

2334. The half-gerund in these last two examples can
easily be made into a full participle by a mere change of

122 SYNTAX. [§ 3335.

order, though the result will be a very stilted literary form —
she, sifting (or having sat) on the damp grass, caught cold.

2335. In several of the other half-gerund constructions the
participle can be substituted by a change of construction.
Thus I enjoy being here suggests I /eel enjoyment while beiftg
here.

2336. The constructions which most resist this change are
those which also allow the substitution of a possessive or
genitive for the preceding objective or common case, for the
change of / do not like him coniing here into / do not like him
when coming here — when he comes here involves a distinct
change of meaning.

Participles.

Present.

2387. The present participle is sometimes used in a passive
sense in some definite tenses (2312).

2338. Allied to this is the peculiar use of the adjective
participle in falling sickness = * illness in which the patient
falls/ dying day, parting glass^ waking vision. That the first
element is really a participle is shown by the even stress, and
also by the Old-High- German vallandiu suht ; if it were
a noun, as in dining-table, sleeping-apartment, there would be
uneven stress (900).

2339. When the present participle is added to an intran-
sitive verb, it is logically partly in a kind of apposition to the
verb, and at the same time qualifies the subject : he came
running = * he ran up ' \ the fog came pouring in at the
window I / cannot go on doing nothing. In the first two
examples came is so subordinated in meaning to the par-
ticiple that it is felt almost as an auxiliary. This is often still
more the case when the participle is joined to a verb of rest,
as in he lay sleeping compared with he was sleeping \ he stood
looking on.

§2347.] VERBS: PARTICIPLES. 123

2340. When it is added to a direct-object word, it is
associated sometimes more closely with the preceding transi-
tive verb, as in / saw htm coming up the road^ sometimes
rather with the object- word, as in / have kept you waiting.

2341. The difference between / j^zw him coming and I saw
him come is that the former is more descriptive, the latter more
a statement of a bare fact.

2342. In the preceding constructions the participle is run
on to the words it modifies without a break. In the follow-
ing constructions there is generally a more or less distinct
pause or change of intonation, the result of which is that the
participle or participle-group is felt to be equivalent to
a dependent clause.

2343. When equivalent to a relative clause, such a group
may be regarded simply as a post-adjective-group (1788), so
that in such a sentence as here are my letters announcing my
intention to start a pause after letters^ though allowable, is not
necessary.

2344. Such groups may be used to modify sentences in
various ways, which are often difficult to define exactly.
Thus in she^ dyings gave it me the participle is evidently
equivalent to a temporal clause, while in seeing a crowd,
I stopped the group can be expanded into when I saw . . or
because 1 saw . . , although the causal meaning is not so clear
as in not having received an answer, I wrote again,

2345. This vagueness is, of course, one of the reasons'
why this construction is often preferred to that with a de-
pendent clause, especially in the literary language. Compare
composition (1560).

2346. On the other hand, these participle-groups, through
having the same function as dependent sentences, have come
to adopt some of the grammatical peculiarities of the latter.

2347. Thus they can take conjunctions whenever clearness
seems to make it desirable, as in Mac Ian, while putting on
his clothes and calling to his servants to bring some refreshmefit

124 SYNTAX, [§ 2348.

/or his visitors, was shot through the head \ I wrote a similar
epitaph/or my wife, though still living compared with dogood^
hoping for nothing.

2348. A participle-group introduced by a conjunction no
longer requires to be placed next to the word it modifies, as
in the preceding examples, but may be detached from it:
t nor ever did. I love thee less, though mourning for thy
wickedness.

2849. The groups we have hitherto been considering all
modify some definite word, but there are others which are,
logically speaking, complete sentences, having their own
subject and predicate. When grammatically disconnected
from any one word in the accompanying sentence, they
modify that sentence as a whole, as in t we sitting, as I said,
the cock crew loud=while we were sitting, . . \ soon afterwards
the truth revealed itself, the real criminal confessing the crime.

2350. This absolute construction is found already in Old-
English, in which it appears, however, not as a nominative,
but as a dative absolute {Us sittendum), being an imitation of
the Latin ablative absolute. In Middle English we have an
oblique absolute, which is due rather to the influence of the
Old-French accusative absolute than to any tradition of
the Old-English construction. It lasted into the Modern
period {us sitting), where it was kept up mainly by the vague
analogy of the Latin ablative absolute. The change of us
sitting into we sitting — which began already in Middle
English — was of course the inevitable result of the desire of
getting rid of the oblique forms in their anomalous position-
at the head of a logical sentence.

2361. The absolute participle-construction is not only un-
coUoquial, but is by many felt to be un-English, and to be
avoided in writing as well.

2352. The logical subject of such a group is sometimes
repeated in the accompanying sentence in the form of
a pronoun, as in our guest offering his assistance, he was

§2356.1 VERBS: PARTICIPLES, 125

accepted among the number; or a pronoun in the group may
refer to a noun in the sentence : neither could he suspect that
he had missed his way, it being so broad and plain.

2353. By the analogy of the absolute construction,
participle-groups without a logical subject are sometimes
used absolutely. This harsh construction is quite a man-
nerism with some writers, such as Richard Jefferies, as in the
two following examples, taken from a single page : crossing
to the other side 0/ the bridge, and looking over, the current had
scooped away the sand . .=' when I had crossed . . I found
that the current . / | carefully looking over that side again, the
moorhen who had been out rushed back. In the last example
it is only the context that tells us it was not the moorhen who
looked over the side of the bridge, but the author himself.

2354. This is an extreme development of constructions
such as besides, being rebels, all their acts are illegal, where
the group is an adjunct to the pronoun implied by the
possessive.

2355. The omission of the logical subject is natural
enough in adverbial phrases : talking of subscriptions, here is
one to which your lordship may affix your name \ my father
had, generally speaking, his temper under complete control. In
the first example talking of is evidently a shortening of while
we are talking of Such phrases may be worn down till the
participle becomes a preposition such as concerning.

Preterite.

2356. The preterite participle is sometimes used in an
active sense, as in a learned man. The Old-English form
was gelared, literally 'taught,* being a translation of the
Latin doctus ; when l(sran became obsolete, the participle of
learn was substituted, necessarily in the sense of * having
learnt.' Other examples are drunk{en), mistaken, the
obsolete drawn in why are you drawn i* = * why have you
drawn your sword?', which, like learned, are, logically,

126 SYNTAX. [§ 2357.

perfect participles. In the obsolete well spQken = ' eloquent *
* plausible/ spoken is equivalent to an indefinite present
participle. The preterite participle of intransitive verbs has
an active sense in periphrastic forms such as /le is gone.

2357. The preterite participle is also used in constructions
analogous to /ailing sickness (2338), as in the vulgar tn all
my born days I never saw such a rascal, where born days
= ' time since I was born, life-time.*

2858. The preterite participle after intransitive verbs
stands in the same relation to verb and subject as the present
participle does under similar circumstances, as in he sat
buried in thought \ but the predicative function of the preterite
participle is much more marked than that of the present,
especially when the verb is weakened to a mere link-verb :
to stand convicted \ to get married \ to become distinguished \ to
be known.

2359. It is added to a direct-object word in the same way
as the present participle : / saw it done \ I saw it being done \
I will see it done = / will have it done \ he declared himself
satisfied \ do you not wish him gone? In the last example
the participle has an active sense.

2360. The preterite participle may be added to a pre-
dicative noun: that is a good thing done parallel to I saw
a good thing done (2359). Here the participle is connected
in thought directly with the preceding verb — ' I saw the doing
of a good thing.' The similar addition of the participle to
a subject- word is only literary : a Deity believed is joy begun
— * the belief in a Deity is the beginning of joy.'

2361. After a preposition-group it generally has quite
a different function, as in people often fight without any mis-
chief done (Sheridan). Here done is felt to be a substitute
for being done — which would, indeed, be the more general
form of expression — so that it is really a kind of half-gerund
(2330). So also in they set him free without his ransom paid
(Shakespeare).

§2367] VERBS: PARTICIPLES. 1 27

2362. But in some cases the participle remains a pure
adjective after a preposition-group, as in ^fortune is famous
for her numbers slain.

2363. A group headed by a preterite participle when
equivalent to a relative clause is generally run on without
pause, as in <7 Ufe wasted in the vain pursuit of pleasure^
though a pause is also admissible, as in I am the sister of one
Claudio, condemned to lose his head.

2364. A variety of other relations are expressed by joining
on (with a pause) a preterite-participle group : once seen, it
can never be forgotten \ planned merely, 'tis a common felony ;
accomplished, an immortal undertaking \ which, testified or not,
does verily remain the fact.

2365. Here, again, the relation may be made clearer by
conjunctions : when once seen \ if planned merely \ whether
testified or not \ if deceived, I have been my own dupe.

2366. There is an absolute preterite particle, as in this
done, find the councillor I \ which said, he sat down, but its
use is more hmited than that of the absolute present participle.
It is very little used with pronouns in the subject-relation
(«j dispossessed, he once passed), because a preterite participle
after a nominative is generally not distinguishable from
a finite preterite.

2367. The following examples contain constructions
parallel to present-participle ones : ithus repulsed, our final
hope is flat despair (2352) | thus saying, he took up his hat . .
when gone, we all regarded each other for some minutes with
some confusion (2358).

INDEX

The references are to the paragraphs.

a article 229.

a- 1569.

absolute 55 ; a. adjectives 2068 ;

a. pronouns 203.
abstract 30 ; a. nouns 165.
accidence 581, 945.
accusative 131.
acoustic qualities of vowels 673 ;

a. sound-changes 717.
active 311, 2312.
'Od 1 700.
-ade 1702.
adjective 96, 176, 1023, 2068;

a.-clause 498 ; a.-word 103.
adjunct-word 40, 90.
adverb 102, 336, 1496; a.-clause

499.
adverbial object 251.
adversative conjunctions 417.
ag' 1571.
affirmation 1518.
affirmative conjunctions 414; a.

sentences 501.
after- clause 466.
■age 1703.
agreement 91.
-a/ 1730.

all 211 \ all the 234.
alternative conjunctions 415; a.

questions 505.
although 418.
-an 1735.
-ana 1707.
anacoluthia 118.
analogy 535.
analysis, phonetic 650.

VOL. II. K

-ance 1708.

-ancy 1709.

and /{o\, 414.

-ane 1735.

Angles 598.

Anglian 602.

Anglo-Frisian 596.

Anglo-Saxon 603.

-ant 1742.

antecedent 217 ; a. verb 221.

antigrammatical constructions

116.
antilogical 119.
any 230.

appended clauses 473.
apposition 90, 47 note,
-an 683, 1744.
-ard 1687.

Arian languages 595.
-art 1687.

articles 201 1 ; a. omitted 2049.
-ary 1745.
Aryan = Arian.
as 370.
aspirate 708.

assimilation, phonetic 720.
association 20 ; a.-group 523.
assumptive 44.
-ate 1716, 1751.
attraction 117.
attribute 28, 44,
auxiliary ^242.

back-consonants 699 ; b. vowels

671.
back-pointing pronouns 225.

I30

INDEX,

base 74.

be 1488.

^e- 1572, 1575-

bear vb. 1439.

beat vb. 1403.

because 423.

/^<f^« 1385.

^««ar 1326.

bereave 131 2.

^/af 1394.

bide 1 41 9.

/^/W 1364.

^///f 1409.

blade consonants 702 ; b.- point

consonants 702.
-ble 1719, 1720.
blend 1327.
blending 125.
blow 1452.
-bond 1 721.
^<?M 235.
break 1431.

breath sounds 651 ; b.-group 453.
bring 1339.
broad vowels 672.
broken intonation 1948 ; b. order

i860 ; b. quantity 233.
huild 1324.
■bund 1 721.
bum 1304.
burst 1354.
but 1^11 .
buy 1 32 1.

can 1478.

cardinal numerals 239.

case 949, 1985.

cast 1344.

catch 1336.

categories, grammatical and logi-
cal 24, 25.

causal conjunctions 423.

causative verbs 250.

Celtic influence 607.

changes in language 511.

chide 141 o.

choose 1437.

class-nouns 150.

clauses 49, 463 ; c. of purpose
396.

cleave 1313.

climax-stress 1893.

climb 1 41 8.

cling \i^\.

close vowels 672.

clothe 1335.

Cocicney dialect 647.

co-clauses 463.

co-complexes 465.

co-conjunctions 426.

cognate languages (to English)
312 ; c. object 253.

collective nouns, 150, 151, 158;
c. pronouns 235.

colloquial 576.

combinative sound-changes 719.

come 1406.

common case 140, 1985 ; c. nouns
150.

comparative degree 1 76 ; c. gram-
mar 5.

comparison of adjectives 176,
1036 ; c. of adverbs 1523, 2081.

complete tenses 281.

complex 462 ; c sentences 49 ; c.
words 62.

composition 62, 1545.

compound conjunctions 400 ; c-
falling tone 661 ; c. preposi-
tions 385 ; c.-rising tone 661 ;
c. tenses 274 ; c. words 62.

compulsive mood 300, 2297.

concessive clause 295 ; c. conjunc-
tions 418.

concord 91.

concrete 30 ; c. nouns 1 50.

condensation 112.

condensed relative 220.

conditional 300, 2277; c. sen-
tences 295.

conjoint 55.

conjugation 242.

conjunctions 102, 399.

conjunctive adverbs 373 ; c. pro-
nouns 221.

connected names 160.

connective 88.

consonant-influence 754.

consonantal verbs 1285.

consonants 653, 691.

continuous quantity 232 ; c. tenses
283.

INDEX,

131

contracted complexes and sen-
tences 488.
contrasting stress 1890.
convergent sound-changes 730.
conversion of parts of speech 105.
converted verb 255.
co-ordinate sentences 460.
co-ordination 44, 89.
co-ordinative 426.
copulative conjunctions 4 1 6.
correlative 370, 15 14.
creep 1295,
cross-order 1865.
'Cy 1699.

dare 1480.

dative 132.

dead phenomena 588^

declarative sentences 501.

declension 150.

declinable 100.

definite article 224, 2013; d. pro-
nouns 201, 224, 1 144 ; d. tenses
288, 2203.

demonstrative pronouns 225, 11 15.

dependent adverbs 369; d. pro-
nouns 199 ; d. sentences 459.

deponent verbs 316.

derivation 69, 1562.

derivative words 62.

descriptive grammar 2, 586.

detached clauses 491 ; d. conjunc-
tions 431 ; d. prepositions 396.

dialects 574.

dictionary 583.

dig 1370.

dight 1345.

diphthongs ^d^i^ 686.

direct narration 297 ; d. object 131.

discrete quantity 232.

distract 1.^37.

distributed stress 1896.

divergent sound- changes 721.

do 1493 ; do-iQxxas, 2169.

-dom 1601.

doublets 54, 520, 1912,

draw 1456.

drink 1386.

drive 1420.

dual 127.

dwell 1305,

each 235 ; e. other 210.

-ean 1738.

Pearly (period) 594.

East-Anglia 601.

East-Arian 595.

East-Germanic 596.

eat 1402.

■ed adj. ending 1606.

-ee 1682.

-eer 1683.

either 235 ; cither . . . or 415.

ellipse III, 534.

emphasis 1765.

emphatic pronouns 205 ; e. stress

1885; e. verb-forms 317; e.

word-order 554.
empty words 58 ; e. sentences 450.
-en adj. 1607 ; vb. 1616.
-ettce 1708.

enclitic stress- weakening 1899.
-ency 1709.
-end I S)^ J.
end-position 1759.
endings 70.
English 603.
enough 234.
-ent 1742.
-er 1592, 1683.
-ese 1747.
-esque 1750.
-ess 1688, 1698.
-et 1693.
even stress 1897.
every 235.
exclamation-words and -groups

438.

exclamative sentences 502.

explanatory grammar 3.

extended complexes 473; e. sen-
tences 486.

external sound-changes 512, 718.

extra stress 1881.

-ey 1694.

fact-form and -statement 294, 303.
/all 1404.

falling intonation 66 1«

familiarization 541.
feel 1316.

feminine 145.
few 235.

K 2

132

INDEX.

fight 1438.

Jind 1365.

finite verb 103.

first XI 70.

fixed order 1760,

fiee 1294.

fiing 137a.

fiy vb. 1450.

-fold 161 1.

for conj. 423.

for- 1573-

form 53.

formal 16 ; f. analogy 540; f. syn-
tax 58 a.

formulae, stress of 1894.

form-words 58.

forsake 1447.

forward consonants 703.

forward-pointing pronouns 325.

free adjectives 178, 2070; f. groups
440 ; f. order 1 760.

freeze 1432.

freight vb. 1443.

French influence 610, 622.

front-clause 466 ; f. consonants
700 ; f.-modified consonants
706; f.-order 1867; f.-position
1759 ; f.-vowels 671.

ful 161 2.

full word 58; f. diphthongs 687.

function of parts of speech 97.

future 273, 2249; f. perfect 277;
f preterite 278.

•fy 1756.

^^-1574-

gender 142, 947; g. of nouns

1953.
genealogical 86 note,
general grammar 6; g. interrogative

sentences 503 ; g. words 37.
generalizing use of words 39.
genitive 133, 1996.
Germanic 596.

gerund loi, 324, 1257, 2326.
get 1414-
give 1407,
glides 664.
go 1458.
government 94.
gradation 770.

grammar, definitKon of f ; g. and

dictionary 583.
grammatical difficulties 589 ; g.

analysis 591 ; g. gender 146.
grave vb. 1459.
grind 12,66.
group-association 20 ; g.-com-

pound 440 ; g.-influence 535 ;

g. -prepositions 386 ; g.- verbs

251, 394.
grozv 1454.

half-conjunctions 409 ; h.-diph-
thongs 686 ; h.-long 658 ;
h.-strong 659.

hang 1 369.

have 1452; h.-iorms of verbs
2165.

he 200.

head-word 40, 90.

hear 1300.

heave 1433.

her 203.

here-in etc. 379.

hew 1460.

hide vb. 141 1.

high vowel 673.

hight 1474.

his 203.

historical grammar 4, 586.

history of English 594 ; h. of lan-
guage 511.

hold 1403.

homonyms 55, 55a.

-hood 1602.

hortation 308.

hypothetical conjunctions 419.

/200.

-iac 1727.
-ial 1730,
-ian 1739.
-ic 1722.
-ical 1 731.
-ice 1698.
-icism 1 715.
iclept 1475.
-id 1700, 1728
idioms 448.
-ier 1683.
*/4i9-

INDEX.

133

-He 1734.

imitative sound-changes 512, 717;
i. words 544.

immediate future 284.

imperative 308 ; i. sentences 506 ;
i.-interrogative sentences 508.

impersonal verbs 257.

-in 1740.

incapsulation 470.

inchoative tenses 305.

incomplete tenses 381.

indeclinable 100.

indefinite article 239, 3041 ; i. pro-
nouns 329, 1 146 ; i. tenses 288.

independent adverbs 358, 409 ;
i. pronouns 199 ; i. tenses 459,
280 note; i. verbals 330.

indicative 398.

individual nouns 150 ; i. name-
words 158.

indirect government 134; i. in-
terrogation 314; i. narration
397 ; i. object 133.

'ine 1740.

infinitive loi, 331, 3314.

inflection 73; i. of word-groups
1016.

inflectional languages 86.

•ing 1594, 1600.

inner position 707.

inserted clauses 467.

instrumental 134.

intensive stress 1886.

interjectional words 544.

interjections 103, 433, 154a.

internal sound-changes 716.

interrogation 314.

interrogative pronouns an, 11 38,
21 19 ; i. sentences 503 ; i. verb-
forms 317.

interval 663.

intonation 661, 1925.

intransitive 348.

inverted object, subject 313.

-ion 1706.

-ior 1 746.

irregularity 23, 526.

-ise 1698, 1758.

-ish 1609, 1757.

•ism 1714.

K

isolating languages 86.

isolation 21, 525.

isolativ« sound-changes 719.

-ist 1689.

it 300.

-ite 1690, 1 75 1.

its 303.

■itiide I'ji'j.

•ive 1755,

-i%e 1758.

Jutes 598.

-kin 1590.
knit 1357.
know 1455.

-Ictcan 1618.

language, definition of 16,

Late (period) 594.

Latin influence 606, 625.

leady\i. 1461.

leap 1296.

learn 1306.

learned words 625.

least 234.

leave 13 19.

lend 1328.

-lent 1743.

less 234.

-less 1613.

lest 435.

let 1349.

'let 1693.

level intonation 661.

lexical side of language 1 8.

lie 1405.

light vb. 1408.

-//«^I595.

link-verbs 263.

lip consonants 704.

literary language 576.

little, al. 2^j^

living phenomena 588.

load yh. 1461.

locative 135.

-lock 1603.

logical control of sound-changes

533 ; 1. side of language 16 ;

1. syntax 583; 1. word-order

553.
London dialect 619.

134

INDEX,

long quantity 658 ; 1. tenses 283.

loose 1323.

loss of sounds 723.

Low German 596.

low vowels 673.

-ly 1614.

make vb. 1334.

many 235.

mark-words 35, 195.

masculine 145.

material nouns 150.

may 1481.

mean vb. 1320.

meanings of words 5 1 3.

medium quantity 658; m. stress
659.

melt 1462.

-ment 1704.

Mercian 601.

metaphor 514.

mid position 1759 ; m. vowel 672.

middle clause 468 ; m. voice 316.

Middle English 594, 613, 771.

mine 203.

mixed complexes 493 ; m. sequen-
ces 493; m. verbs 1457; m.
vowels 671.

Modem English 594, 629, 823.

modification 115,

modifiers 40.

modifying form-word 88 ; m. stress
1891.

monophthongs 667.

moods 293, 2259.

more 234, 235.

morphological 86 note.

most 234, 235.

mote vb. 1482.

mow vb. 1463.

much 234.

murmur-diphthongs 688,

must vb. 1482.

mutation 751 ; m.-plurals 1002.

my 203.

name-words 157.

nasal consonants 696 ; n. sounds

652 ; n. vowels 685.
natural gender 146.
-nd 1591.

needvh. 1487.

negation 1518.

negative conjunctions 416 ; n. in-
flection 77 ; n. pronouns 236 ;
n. sentences 501 ; n. verb-forms

317.
neither 2,^5, 336, 416.
-ness 1597.
neuter 145.
neutral tenses 289.
no adverb 368 ; pronoun 230, 336.
nobody 236.
nominative 129.
nor 416.
Northern 601,
not 365.
nothing 236.
noun 96, 126, 945 ; n.-clause 497 ;

n.-word 103.
number of nouns 127, 1966 ; n. of

verbs 269, 2145.
numerals 101, 237, 11 59, 3143.

object-complement 267.

objective case 141.

oblique cases 1 39.

of- 1576.

-oid 1729.

Old-English 594, 598.

on- 1577.

one noun 201 ; numeral 237 ; pro-
noun 210.

open condition 305 ; o. consonants
693 ; o. vowels 673,

optative 398.

or conj. 415.

-or 1684, 1710.

order of words 1 760.

ordinal numerals 239.

organic sound-change 512, 716.

origin of language 544 ; o. of parts
of speech 560.

-ory 1 71 1.

-ose 1748.

other 231, 235.

ought vb. 1483.

our 203.

-our 1684, 1 7 10.

-ous 1748.

outer position 707.

own pronoun 206, 209.

INDEX.

135

parenthesis (kind of sentence) 471.
parenthetic sentences 471.
participles loi, 330, 2337.
particles loa, 1494.
parts of speech 95, 98.
passival supine 322 ; p. verbs

249.
passive voice 312, 2312,
patronymics 162.
perfect 275, 2240,
periods of English 594.
periphrase-verbs 320.
periphrastic verb-forms 24a.
permissive mood 300, 2301.
person of pronouns 200; p. of

verbs 270, 2155.
personal gender 211 ; p. names

162 ; p. pronouns 200, 1058,

2095.
personification 1962.
phenomena 29.
philosophical grammar 6.
phonetics 648 ; phonetic analogy

540-

phonology 648.

pitch 662.

pluperfect 276, 3247.

plural 127; p. nouns 152.

point consonants 701.

popular words 625.

positive degree 176.

possessive pronouns 203, 1093,
2105.

post-position 1759.

predicate 42.

prefixes 70.

prepared sentences 403, 458.

pre-position 1759.

prepositional complements 251.

prepositions 102, 383, 1528 ; pre-
position-groups 387.

Present English 639, 879.

present tense 273, 2223.

preterite 273, 2233; p.-present
verbs 121 1.

primary clause 475 ; p. com-
pounds 67 ; p. conjunctions 399 ;
p. prepositions 383; p. tenses
379.

principal clause 463.

progressive relative 218.

pronominal adverbs 1508; p. con-
junctions 1 5 16.
pronouns loi, 189, 1053, 2095.
prop- words 180.
proper names 150.
purpose, conjunctions of 425.

qualifiers 34.

qualifying attribute-words 36.

quantitative pronouns 232, 1152.

quantity 658.

quit 1358.

quoth 1473.

r in English 71 1.

-rceden 1604.

reach vb. 1342.

reciprocal pronouns 210; r. verbs

356.
recurrent tenses 283.
redundance no.
reference-pronouns 325.
reflexive pronouns 207 ; r. verbs

254; r. voice 316.
rejected condition 305..
relative adverbs 373 ; r. pronouns

217,1138,2123.
retained object 315,
ridvh. 1359.
ride 142 1.
rise 1423.

rising intonation 661.
rive 1464.

rounded consonants 705.
rounding 669.
ru7t 1383.
-ry 1712.

same 228.

saw vb. 1465.

Saxons 598.

say 1293.

Scandinavian 596, 608.

second ii*io.

secondary clause 4.75 ; s. com-
pounds ; s. conjunctions 399 ;
s. prepositions 384; s. tenses
279.

see vb. 1445.

seek 1 341.

seethe 1416.

136

INDEX,

self 2ot^i 207, 1 104.

sill 1302.

sentences 446 ; sentence-adverbs
368; s.-antecedent 217; s.-con-
necting conjunctions 403 ; s.-
group 452, 453 ; s.-introducing
adverbs 369 ; s.-modifying ad-
verbs 364; s.-stress 84 note,
1881 ; sentences of wish 296 ;
sentence-vt^ords 49, 452, 545.

separative pronouns 235.

sequences 482.

set vb. 1350.

several 235.

shake 1448.

shall 1484, 2196.

shape vb. 1466.

shave 1467.

she 200.

shear 1444.

shed vh. 1351.

shifting 120.

-ship 1605.

shoe vb. 1 301,

shoot 1417.

short quantity 658 ; s. tenses 283.

show 1468.

shred I7,e^2,

shrive 1423.

shut 1347.

"Stan 161 7.

side consonants 694.

simple conjunctions 400 ; s. tenses
274 ; s. words 62.

singular nouns 152 ; s. number
127.

sink 1390.

'Sion 1706.

sit 1383.

slay 145 1.

sleep 1297.

sling \h. 1373.

slink 1374.

slit vb. 1360.

smite 1424.

smoothing 720.

j^ 377, 424; so ,. .as 370.

some 230, 235.

•some 1610.

sound-change 512, 715.

Southern dialects 602 .

sow vb. 1469.
speak 1434.

special words 38, 39 ; s. interroga-
tive sentences 504.
spin 1375.
sptt vb. 1384.
split infinitive 1864.
spoken language 576,

spread i^f^Z-

stand vh. 1446.

standard dialect 573.

stave vb. 1430.

steal 1435.

'Ster 1593.

stick yh. 1376.

stink 1392.

stopped consonants 695.

strata of language 646.

stress 659 ; s. -doublets 521.

strew 1470.

stride vb. 1425.

strike 1368.

string vh. 1378,

j/r»W 1426.

strong inflection 948; s. stress

659 ; s. verbs 1190.
sub-conjunctions 437.
subject 42 ; s. -clause 463 ; s.-com-

plex 465.
subjunctive 298, 2259.
subordinate 44, 89; s. sentence

460.
subordinative conjunctions 427.
substances 28.
such 227.
suffixes 70.
superlative 176.
supine 243, 321, 2314.
swear 1440.
sweep 1298. (
swell 1 47 1.
-sy 1699.
syllables 666.
symbolic words 544.
synonyms 522.
syntax 581, 1759.
synthesis, phonetic 656.

./ 1751.
take 1449.
-te I7j5i.

INDEX.

137

teach 1343.

tearvh. 1441.

te/l 1303.

temporal conjunctions 421.

tenses 271, 2163, 2223; tense-
mood 301.

-M 1599.

Maw 380.

fhai adverb 374; conjunction of
purpose 425 ; dem. pronoun
225 ; relative 216.

Mtf 225.

ihee 200.

fAeir 303.

rkere 344 note ; therein etc. 379.

therefore 424.

thmk 1340.

Mw 225.

M^« 200.

though 418.

thought-form 294, 302, 303.

thrive 1427.

throat-sounds 651.

thy 204.

-//^;/ 1706.

to- 1579.

tones 661.

transition periods 594.

transitive 248.

tread 141 5.

trilled consonants 697.

triphthongs 667.

-trix 1691.

-ty 1 718.

•u 1598.
•m/; 169a.
«»- 1580.

unconnected names 160.
ungrammatical 12.
unless 419.

unprepared sentences 403.
-ure 1 713.
-«/^ 1 75 1.

verbals loi, 319; verbal-group
445.

verbs 96, 240, 1177, 2145; ano-
malous V. 1287; iiregular v.
1292.

vocalic verbs 1285.

vocative 130.

voice 311, 651, 2312.

vowels 654, 668, 674; vowel-like
consonants 655.

vulgar 575.

wake 1429.

wan- 1581.

-ward 1615.

we 2 GO.

weak sounds 660 ; w. stress 659 ;
w. verbs 1202,

wear vb. 1442.

weave 1436.

weep 1299.

wendvh. 1333.

West-Arian 595.

West-Germanic 596.

what 211, relative 216.

wherein etc. 379.

which 211, relative 216.

Wi4<> 211, relative 216.

xvhole 234.

ivillwh. 1485, 2196.

win 1380.

wind\h. 1367.

wist vb. i486.

wont, be 1476.

words 51, 81, 83, 84; word-con-
necting conjunctions 403 ; w.-
formation 62 ; w.-groups 50,
440 ; w.-introducing adverbs
369 ; w.-modifying adverbs 358;
w.-order 1759.

workvh. 1338.

wot vb. i486.
write 1428.

•y 1608, 1694.
yes 368.
you 200,
your 203.

PRINTED IN ENGLAND
AT THE OXFORD UNIVERSITY TRESS

MIDDLE ENGLISH

SIR GAWAIN AND THE GREEN
KNIGHT. Edited by J. R. R.
Tolkien and E. V. Gordon. Crown
8vo. Pp. xxviii + 212. 7s.6d.net.
At the Clarendon Press.

A STUDY OF GAWAIN AND
THE GREEN KNIGHT, by G. L.

KiTTREDGE. 8vO. Pp. 332. lOS. 6d,

net. Harvard University Press.

SELECT EARLY ENGLISH
POEMS, edited by Sir I. Gollancz.
Fcap 4to.

I. Patience, an alliterative version
of Jonah by the Poet of Pearl.
Pp. 58. 5s. net.

II. The Parlenunt of the Thre Ages.
An alliterative poem on the Nine
Worthies and the Heroes of Ro-
mance. Pp. 116. 5s. net.

III. A Good Short Debate between
Winner and Waster. An alliterative
poem on Social and Economic Pro-
blems in England in the year 1352,
with Modern English Rendering.
5s. net.

IV. St. El kemvald{)i\s\\oTp oihon-
don 675-693). An alliterative poem,
written about 1386, narrating a
miracle wrought by the Bishop in
St. Paul's Cathedral. Pp. 112.
5s. net.

VII. Cleanness. An alliterative
tripartite poem on the Deluge, the
Destruction of Sodom, and the Death
of Belshazzar, by the Poet of Pearl.
Pp. 114. 5s. net.

VIII. Pearl. An English poem of
the Fourteenth Century. Pp. 348.
25s. net

Oxford University Press, London.

^ Olker Romances,

KING HORN, a Romance of the
Thirteenth Century, with text of
Home Childe, edited by Joseph Hall.
Pp. Ivi + 238. 145. net.

At the Clarendon Press.

THE LAY OF HAVELOK THE
DANE. Second edition, revised by
K. SisAM. Fcap 8vo. Pp. 212.
4s. 6d. net. At the Clarendon Press.

SIR EGLAMOUR, a Middle English
Romance, edited by A. S. Cook.
Pp. 148. 5s. net.

Yule University Press,

August ig2j

IWAIN, by A. C. L. Brown;
ARTHUR AND GORLAGON, by
G. L. Kitthedge. 8vo. Pp. vi +
276. los. 6d. net.

Harvard University Press.

STUDIES ON THE LIBEAUS
DESCONUS, by W. H. Schofield.
Pp. iv + 246. 8vo. los. 6d. net.

Harvard University Press.

MEDIAEVAL ROMANCE IN
ENGLAND. Astudy of the Sources
and Analogues on the Non-Cyclic
Metrical Romances, by Laura A.
HiBBARD. Med 8vo. Pp. viii + 342.
17s. 6d. net.

U.xford University Press,

A merican Branch.

^ Tiers Tlowman.

THE VISION OF WILLIAM CON-
CERNING PIERS THE PLOW-
MAN, in three parallel texts, together
with Richard the Redeless, by William
Langland. Edited with preface,
notes, and glossary by W.W. Skeat.
Two volumes. Pp. 636, xcvi + 484.
50s. net. Ai the Clarendon Press.

WILLIAM LANGLAND'S PIERS
THE PLOUGHMAN (the B text,
about A. D. 1377). Edited by W. W.
Skeat. Ninth edition. FcapSvo. Pp.
xlviii + 216. 4s. 6d. net.

At the Clarendon Press.

PIERCE THE PLOUGHMAN'S
CREDE (about a.d. 1394^. Edited
with introduction and notes by W.W.
Skeat. Fcap 8vo. Pp. xxxii + 74.
2S. 6d. net. At the Clarendon Press.

5 John Gower.

THE COMPLETE WORKS OF
JOHN GOWER, edited with intro-
ductions, notes, glossaries, and fac-
similes by G. C. Macaulay. Four
volumes 48s. net, or each volume
i6s. net.

Vol. I. The French Works. Pp.
Ixxxviii + 564, with facsimile.

Vols. II & III. The English Works.
Pp. clxxvi + 520, 656, with facsimile.

Pp.

Vol. IV. The Latin Works.
Ixxviii + 430, with facsimile.

At the Clarendon Press.

GOWER'S CONFESSIO AMAN-
TIS, selections edited with intro-
duction, notes, glossary, and index,
by G. C. Macaulay. Fcap 8vo.
Pp.lii + 252. 5s.

At the Clarendon Press,

^ Qhaucer.

THE WORKS OF GEOFFREY
CHAUCER and Chaucerian and
other Pieces, edited with introduc-
tions and notes by W.W. Skeat. Vols.
I-VII, with portrait and facsimiles ;
the set of seven volumes 105s. net,
or each volume i6s. net.

At the Clarendon Press.

THE CLERKES TALE, illustrated
edition, with introduction, notes, and
glossary by K. Sisam. 1923. Pp.
xxvi + 78. 2S. 3d. net.

At the Clarendon Press.

THE COMPLETE WORKS OF
GEOFFREY CHAUCER IN POE-
TRY AND PROSE, together with
the two fragments of the Romatint of
the Rose not by Chaucer, with intro-
duction, appendix of variations and
emendations, and two glossarial in-
dexes, by the late VValti- r W. Skh at.
8v'o, Pp. xxiv -t- 882. 6s. net.

Ai tie Clirendon Press.

THE CANTERBURY TALES from
the text of W. W. Skeat, edition
with glossary and brief notes on the
language and metre (World's Clas-
sics). 2S. net.

Oxford Ura'vtrsity Press, London.

THE CHAUCER CANON, with a
discussion of the works associate. 1
with the name of Geoffrey Chaucer, by
W.W. Skeat. Crown 8vo. Pp. xii -r
168. 3s. 6d. net.

At the Clarendon Press.

CHAUCER AND HIS POETRY,
by George Lyman Kittredge. 8vo.
Pp. 236. 6s. 6cl. net.

Harvard University Press.

THE TALE OF GAMELYN.
Second edition. Edited by W. W.
Skeat. Fcap 8vo. Pp. xl + 64.
Paper boaid-;. is 9H.

At the Clarendon Press.

^ 1^/igious PForks,

THE ORMULUM, edited by R. M.
White and R. Holt. Two volumes.
Fcap 8vo. Pp. Ixxxviii + 598. 18s.
net. At the Clarendon Press,

lACOB AND lOSEP, a Middle
English Poem of the Thirteenth Cen-
tury, edited with introduction, notes,
andglo?saryby A. S. Napier. Crown
Svo. Pp. xxxii + 42. 3s. net.

At the Clarendon Press.

THE PSALTER, or Psalms of David,
and Certain Canticles, with a trans-
lation and exposition in English, by
Richard Rolleof Hampole, edited
by H. R. Bramley, with an intro-
duction and glossary. 8vo. Pp.xxiv-i-
556. 2is.net. At the Clarendon Press.

THE MIRROUR OF THE BLES-
SED LYF OF JESU CHRIST, a
translation from the Latin work en-
titled ' Meditationes Vitae Christi ',
attributed to Cardinal Bonaventura,
made before the year 14 10 by Nicho-
las Love, Prior of the Carthusian
Monastery of Mount Grace. Edited
with introductory note and glossary
by L. F. Powell. Crown 4to. Pp.
330. ais.net. At the Clarendon Press.

THE EXEMPLUM IN THE
EARLY RELIGIOUS AND DI-
DACTIC LITERATURE OF
ENGLAND, byj. A. Mosher. Svo.
Pp. xii + 150. 6s. 6d. net.

Columbia University Press.

^ 7{eligious Lyrics,

RELIGIOUS LYRICS OF THE
FOURTEENTH CENTURY.
Edited by Carleton Brown. Crown
8vo, pp. xxiv + 358. IDS. 6d. net.

At the Clarendon Press.

THE PENITENTIAL LYRIC IN
MIDDLE ENGLISH. By F. A.
Patterson. Svo, pp. x -1- 207. 7s 6d.
net. Columbia University Press.

^ "Readers^ ^c.

SPECIMENS OF EARLY
ENGLISH, edited with introduction,
notes, and glossarial index. Thiee
volumes. Fcap 8vo.

Part I. From Old English Homi-
lies to King Horn (a.d. 1150 to
a. D. 1300), by R. Morris. Second
edition revised. Pp. c -1- 554. 7s. 6d.
net.

Part II. From Robert of Gloucester
to Gower (a.d. 1298 to a.d. 1393),
by R. Morris and W. W. Skeat.
Fourth edition. Pp. xl -I- 490. 7s. 6d.
net.

Part III. From the Ploughman's
Crede to the Shepheardes Calender
(a.d. i394-i579)> by W. W. Skeat.
Sixth edition. Pp. xxxii + 550. 7s. 6d.
net. At the Clarendon Press.

SELECTIONS FROM EARLY
MIDDLE ENGLISH, 1130-1250,
edited with introductions and notes
by Joseph Hall. Crown Svo, cloth,
21s. net. Separately : Part I, Texts,
pp. viii + 222, 7s. 6d. net; Part II,
Notes, pp. iv + 676, 15s. net.

At the Clarendon Press.

FOURTEENTH CENTURY
VERSE AND PROSE, edited with
introduction and notes by Kennkth
SiSAM. Crown 8vo. pp. xlviii -t- 292.
With a Glossary by J. R. R. Tolkien,
pp. 16S. los. 6d. net.

Separately, Text, Introduction
and Notes, 7s. 6d. net. Glossary,
4s. 6d. net. At the Clarendon Press.

5 Miracle Tlays.

ENGLISH MIRACLE PLAYS,
MORALITIES, AND INTER-
LUDES, being specimens of the
pre-Elizabethan drama, by A. W.
Pollard. Seventh revised edition,
8s. 6d. net. At the Clarendon Press.

YORK PLAYS. The Plays per-
formed by the Crafts or Mysteries of
York on the day of Corpus Christi.

% Miscellaneous Texts.

Edited with introduction and glossary
by L. TouLMiN Smith. 3 plates. 8vo,
pp. Ixxviii-f 558. 2ts. net.

At the Clarendon Press.

THE ASSUMPTION OF THE
VIRGIN, from the N-Town Cycle,
edited by W. W. Greg. 8vo, pp. 76.
Paper cover, 4s. 6d. net.

At the Clarendon Press.

THE PROVERBS OF ALFRED,
edited with introduction, notes, and
glossarial index, by W, W. Skeat.
Fcap 8vo, pp. xlvi + 94. 3s.

At the Clarendon Press.

EARLY ENGLISH PROVERBS,
chiefly of the thirteenth and fourteenth
centuries, with illustrative quotations,
collected and edited with notes by
W. W. Skkat. Fcap 8vo, pp.
xxiv+148, and portrait of John
Wyclif. 5s. net.

At the Clarendon Press.

POEMS OF LAURENCE MINOT,
edited with introduction and notes
by Joseph Hai l. Third edition,
revised and reset. Fcap 8vo, pp.
176. 4s. 6d. net.

At the Clarendon Press.

THE SERPENT OF DIVISION,
by John Lydgate, the Monk of
Bury, edited with introduction,
notes, and glossary, by H. N
MacCracken, with three full-page
reproductions from contemporary
illuminations accompanying the text.
4to, pp. 76. los. net.

Yale University Press.

^ Some Latin Texts.

OPERA HACTENUS INEDITA
ROGERI BACONI, edited by
Robert Steele. 8vo, stiff paper
covers. Ease. I, 7s. 6d. net ; II-IV,
ics. 6d. net each ; V, 28s, net.

At the Clarendon Press.

WALTER MAP'S DE NUGIS
CURIALIUM, edited by M. R.
James. AnecdotaOxoniensia, Mediae-
val Series, XIV. Crown 4to, rag-
made paper, pp. 326. i8s. 6d. net.
At the Clarendon Press.

lOANNIS SARESBERIENSIS
POLICRATICI Libri VIII recog-
novit commentario instruxit C. C. J.
Webb. Two volumes. 8vo, pp. 1 -»-
368 and viii-f-5TO. 36s.net.

At the Clarendon Press.

NOVA LEGENDA ANGLIAE, as
collected by John of Tynemouth,
John Capgrave, and others, and first
printed, with New Lives, by Wynkyn
de Worde, 1516. Re edited with
fresh material from MS. and printed
sources by C. Horstman. Two
volumes. 8vo, pp. Ixviii + 506 + 732.
36s. net. At the Clarendon Press.

^ 'Bibliography.

A MANUAL OF THE WRITINGS
IN MIDDLE ENGLISH, 1050-1400,
a source book for the history of
English Literature, by John E.
Wells. Med. 8vo, pp. 956, with

two Supplements : I (1919^, pp. 90;
II (1923), pp. 114; 23s. 6d. net.
Supplements separately, each 6s. 6d.
net. Yale University Press,

PE Sv/eet, Henry

1101 A new English grammar

SB cist ed.j

V.2

PLEASE DO NOT REMOVE
SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO
LIBRARY