Imlopttttt M&
«i
, 3
BY
JOHN S. P. TATLOCK, PH.D.
ASSISTANT PROFESSOR OF ENGLISH IN THE UNIVERSITY OF MICHIGAN.
PUBLISHT FOE THE CHAUCER SOCIETY
BY KEGAN PAUL, TRENCH, TRUBNER & CO., LIMITED
DRYDEN HOUSE, 43, GERRARD STREET, SOHO, W.
1907.
#01
A?
2 7
Strits, 37.
UILHARD CLAY & SONS, LIMITED, LONDON AND BBSGAY.
PREFACE.
THIS book was begun as an examination into the received chronology of all Chaucer's works, with a view to ascertaining how much of it is sound. There appeared on examination so many unworked corners, so much unused evidence, and so many of what seemed to the writer " vulgar and common errors," that weighing and expounding of old material had to give large place to destruction and construction. The work expanded so much that it became necessary to disregard most of the minor poems, and to discuss only those works a decision as to which was necessary to other decisions, and those on which the prevalent opinion seemed most erroneous or on which the most new light could be thrown. So the book does not profess to be exhaustive, or to afford anything except the essential elements of a general scheme. But neither has its purpose been merely to give the results of wholly original research ; rather by all means available, and to the extent of the writer's ability, to ascertain, advance, and present the status quo of our reliable knowledge as to when and how Chaucer's principal works" came to be written. This statement is made in order to anticipate the possible criticism that some sections contain little or nothing that is new. Some- times the purpose of the book has required long investigation and restatement of earlier opinion, with the result simply of confirming it ; yet this does not seem labour and space thrown away, for thorough con- firmation of earlier guesses or brief statements is of the nature of an addition to knowledge.
It may be proper to state here that the writer has nearly ready for publication another volume, on the evolution of the Canterbury Tales, and is only awaiting complete information as to the arrangement, and some readings, of the numerous manuscripts in private hands in various parts of Great Britain. In this second book he hopes to throw some new light on the supposed nine groups of tales, on the arrangement of the poem, and on the questions whether and how far Chaucer revised it, whether and how far he published it during his lifetime, and finally how it came into the shape in which we find it in the manuscripts. In particular he has made a thorough collation of MS. Harleian 7334 with ten of the most important others, with a view to discovering whether its peculiarities are due to corrections made by the poet himself.
In conclusion, the writer is glad to express a strong sense of obligation to others. First of all, to the two most distinguished living Chaucer scholars, the Kev. Professor Skeat and Dr. Frederick J. Furnivall. Without the great edition produced by the former, with its invaluable commentary, any sound work on Chaucer must be far more difficult and
v
VI PREFACE.
less extensive ; without Dr. Furnivall's prolonged and self-sacrificing labours on the manuscripts it must be impossible. All accurate philological and historical, which implies also all sound literary, study of Chaucer had its new birth, after Tyrwhitt, in two events, the publication in 1862 of Professor Child's Observations on the Language ofChattcer, and Dr. FurnivalPs launching of the Chaucer Society in 1863. To him, as representing it, I am bound again for its liberal dealing with this book. We may hope, after the recent second renascence of fruitful study of Chaucer, that oftener than ever the society dedicated to him will perform its trentals in his memory. With the passing away of the older generation of German Chaucerians, such as ten Brink and Zupitza, Anglo-Saxons on both sides of the Atlantic are showing more interest in the father of their poetry.
But I have other and still more personal obligations. The material on which is based my treatment of the two versions of the Trottus and 0)-iseyde is mainly the unpublished work of Professor W. S. McCormick, of Edinburgh ; which, since it was done for the Chaucer Society, I have had Dr. Furnivall's authorization to use. To Professor George Hempl, of Leland Stanford University, I return thanks for handing over to me unpublished work of his own on the same subject. I have often been indebted to Dr. George L. Hamilton, of the University of Michigan, and to his wide knowledge of mediaeval literature. To Professor George L. Kittredge, under whose supervision in Harvard University much of this work was done, my obligations are not easy to express ; in particular, whatever merit there may be in my manner of treating the two versions of the Troihis is due to him. I have been bound to him for proposing and making the way plain for an undertaking which seemed at first a trifle audacious ; for his keen insight and his inexhaustible liberality, with which all his pupils are so familiar ; and above all for what I can only call the education of my point of view.
INTRODUCTION.
THE early history of Chaucer criticism illustrates the pseudo-cla?sical indifference to everything except literary right and wrong, and later the curiosity about the past which came in with romanticism and the serious attempt to understand it which came with the beginnings of the modern scientific spirit. Till the last quarter of the eighteenth century there was scarcely any non-aesthetic Chaucer criticism, and since the reign of Elizabeth there had been a tendency not to take him very seriously under any aspect. Some of the early editors showed more discernment than others as to the works which they accepted as canonical, and that was all. Even Warton's treatment of the poet (1778), which marks the transition, was mainly descriptive and appreciative ; he did a large amount of re- search on the sources, but nothing on the chronology and development of Chaucer's literary work, and he wholly disregards them in his account of the poet. A worthy and thorough Chaucer criticism began with Thomas Tyrwhitt's edition of the Canterbury Tales (1775-8). Although his work was chiefly on them, and although he did little on the subject of chronology, his other results have frequently so important a bearing on it, and his taste and judgment were so admirable, that he deserves to head the list of critics. But he needs no praise of mine ; every later editor and critic who deserves the name has been glad to honour him.
For further noteworthy advances we have to wait nearly a century. But during the last forty years, to say nothing of numberless mono- graphs and articles, mentioned in their proper places in the present work, some dozen books have treated Chaucer's literary evolution and chronology with system and more or less independence, or have made other wide and general contributions to an understanding of the subject- Passing over the work of Henry Bradshaw,1 former librarian of Cam- bridge University, important in other directions than that of chron- ology, we come to the most influential book ever written on the subject, Bernhard ten Brink's Chaucer Studien,2 the starting-point of systematic work on Chaucer's development. To it are due the division of his literary life into periods and a good part of the dates usually accepted. It never reached the subject of the Canterbury Tales, and many of its results are unreliable ; but its value is permanent. Ten Brink's second book
1 See Memoir of Henry JBradshaw, by G. "W. Prothero (London, 1888), pp. 212-25, 346-59, etc. ; and Collected Papers of Henry Bradshaw (Cambridge, 1889), pp. 102-48.
2 Chaucer. Studien zur Geschichte seiner Eniwickluny und zur Chronologie seiner Schriften, von Bernhard ten Brink : erster Theil (Miinster, 1870) ; the second part never appeared.
Vlll INTRODUCTION.
which bears on the subject is his Gcschichte der Stujfachen Litterahir,1 where he deals fully with Chaucer, and (guardedly) with his chronology.
The work of Dr. Frederick J. Furnivall on Chaucer-chronology is less important than his work in other directions. The Temporary Preface to the Six-Text Edition of Chaucer's Canterbury Tales2 deals chiefly with the construction, interpretation, and manuscripts of the poem. The Trial Forewords to My " Parallel-Text Edition of Chaucer's Minor Poems"3 deals at some length with the dates of a few poems, and more summarily with those of all ; but Dr. Furnivall offered little new and reliable evidence.
Mr. F. G. Fleay, in his small Guide to Chaucer o,nd Spenser,* discusses chronology and the like. His manner is primitive and amateurish, but sometimes not a little suggestive. In particular he has some premature but laudable conjectures as to the development and arrangement of the Canterbury Tales, and also on the order in which they were written.
The next book to be mentioned is Dr. John Koch's Chronology of Chaucer's Writings.6 His most important contribution to Chaucer chronology is the date 1381-2 for the Parliament of Fowls, which he had announced many years before.6 The Chronology, though a convenient resume1 of earlier views, was less illuminating and judicious than might have been desired.
In 1892, in his Studies in Chaucer,'' Professor Thomas R. Lounsbury waged vivacious war against prevalent misapprehensions, endeavoured to put the poet in his proper relation to literary history, and incidentally collected a large amount of known facts and added not a few new ones. The value of the work, great though it is, cannot always be called proportionate to its bulk, and at times it shows a tendency to represent Chaucer as a modern exiled among barbarous ancients. It rarely deals with chronology directly, but it is indispensable to any student of Chaucer's literary evolution.
Mr. A, W. Pollard, in 1893, contributed to a series of Literature Primers8 one on Chaucer; in 1903 he republished it, with slight- changes. This is another, and especially convenient, summary of earlier work, treated with justifiable conservatism, but with many modifications and additions.
Professor W. W. Skeat, in the great Oxford CJiaucer,9 deals with the
1 Berlin, 1877 ; second edition, edited by Alois Brandl, Strasburg, 1893 English translations (respectively) 1883 and 1896.
2 Chaucer Society ; London, 1868.
3 Ibid., 1871.
4 London and Glasgow, 1877.
5 Chaucer Society ; dated 1890, but evidently not published till 1892, since he refers to ten Brink's death.
6 In Enylische Studien, I. 288 ; reprinted in the Chaucer Society Essays, pp.
7 Harper and Bros., New York ; 3 volumes.
8 Published by Macmillan and Company.
9 The Complete Works of Geoffrey Chaucer, Oxford, 1894; 6 vols., with supplementary one (1897).
INTRODUCTION. IX
chronology of all Chaucer's works. He accepts nearly all the results of ten Brink, Koch, and other writers, and in many cases simply repeats (without needed revisions) what he had himself published in his earlier editions of parts of Chaucer's writings. He is not always careful of consistency, and frequently draws conclusions without full examination of the evidence. The greatest value of his edition lies in its notes and indices. In his Chaucer Canon l he gives a conjectural chronological table.
Dr. F. J. Mather, in a school-edition of The Prologue, the Knight's Tale, and the Nun's Priest's Tale, from Chaucer's Canterbury Tales,2 has given us an interesting and valuable study of Chaucer's literary development. Since he embodies the latest results, some of them his own, and has gone at the whole matter afresh in a critical spirit, his book has more significance than its unpretending form would suggest.
The editors of the excellent Globe Chaucer 3 have generally expressed themselves on the subject of chronology. They are always judicious and sometimes original. The most important work is that by Mr. Pollard and Professor McCormick.
Among recent work most important of all, Professor J. L. Lowes has thrown much new light on the chronology of Chaucer's middle period in two long articles in the Publications of the Modern Language Association of America.4 I am able to accept by no means all his conclusions, which depart widely from earlier views, but his new facts, the product of careful and penetrating investigation, are an addition to Chaucer knowledge of high and permanent value.
Finally, Professor R. K. Root has just published, after most of the present work was in type, the best handbook on Chaucer yet written.5 It is an excellent guide to understanding and appreciation, and a good, though rather conservative, rationale of chronology and the like.
It may be convenient if I give a condensed summary of previous opinion as to the dates of Chaucer's principal works. It may strike a reader that later in this book some changes are suggested where hitherto there has been notable unanimity of opinion. But this unanimity sometimes ceases to be impressive when one sees on what slight grounds of evidence it has been based.
1 Oxford, 1900; see pp. 154-5.
2 Boston, 1899.
3 The Works of Geoffrey Chaucer, edited by Alfred W. Pollard, H. Frank Heath, Mark H. Liddeli, and W. S. McCormick ; London, 1901.
4 Vol. xix. pp. 593-683; vol. xx. 749-864. His conclusions are summarized on pp. 860-4 of the second article.
5 The Poetry of Chaucer, Boston, 1906.
INTRODUCTION.
Romance of the Rose.
Extant text not genuine ; Chaucer's
translation about 1377-80 (ten
Brink). Extant translation not genuine ;
Chaucer's translation 1366-7 (?)
(Koch). Not genuine, "with the possible
exception of" Fragment A
(Kittredge).1 Doubtful if any of the extant text
is genuine ; Chaucer's version
" early in life " (Mather). Fragment A (only) genuine ; very
early (Skeat).2 Fragment A may be genuine, B
not, C possibly (Liddell).3 Extant text probably not genuine ;
Chaucer's version early ( Pollard). Fragment A genuine, B not, 0
perhaps ; done in youth (Root).
ABC.
(1) 1367.(Furnivall). About 1373 (ten Brink). (?) 1368 (Koch). Before 1373 (Mather). Very early (Skeat and Eoot). 1369 or a little later (Heath). Before 1380 (Pollard).
Complaint to Pity.
(?) 1367-8 (Furnivall).
Probably 1370-2 (ten Brink).
(?) 1373-4 (Koch).
Before 1373 (Mather).
1372-3 or later (Skeat).
1369-71 (Heath).
Before 1380, perhaps after 1372
(Pollard). Very early (Root).
Book of the Duchess.
1369 (Furnivall, Mather, Skeat). 1369-70 (ten Brink, Koch, Pollard,
Root). Soon after 1369 (Heath).
Complaint of Mars.
(?) 1375 (Furnivall). 1379 (Koch). 1387-1400 (Mather). (?) 1379 (Skeat). After 1378-9 (Heath). "Probably towards 1380" (Pollard).
Boethius.
(?) 1376 (Furnivall). About 1281 (ten Brink). (?) 1377 (Koch). 1373-8 (Mather). 1377-81 (Skeat). Rather early (Liddell). 1380-3 (Pollard). About 1382-3 (Lowes). About 1380 (Root).
Troilus and Criseyde. (?) 1382 (Furnivall). 1380-81 (Koch). 1378-81 (Mather). 1379-83 (Skeat). 1380-3 (Pollard). Perhaps 1383-5 (Lowes). Not far from 1380 (Root).
House of Fame. (?) 1384 (Furnivall). 1384 (ten Brink, Mather). 1383-4 (Koch, Skeat, Pollard). Begun some years before 1383 ; finished after the Troilus (Heath). About 1379 (Lowes). 1378-85 (Root).
Parliament of Folds.
(?) 1374 (Furnivall).
1382 (ten Brink, Koch, Mather,
Skeat, Heath, Pollard, Lowes,
Root). About 1375 (Hales).4
Anelida and Arcite.
(?) 1375-6 (Furnivall).
Not long after 1390 (ten Brink).
1 Harvard Studies and Notes, I. 65.
2 I ordinarily quote Skeat from his Chaucer Canon, and others from their latest expression of opinion.
3 The views of Professors Liddell and Heath are quoted from the Globe Chaucer.
4 Diet. Nat. Bioyr., x. 164.
INTRODUCTION.
XI
(?) 1383-4 (Koch).
1378-81, after Troihis (Mather).
1372-7 (Skeat).
About 1380 (Pollard).
Before 1382 (Lowes).
Soon after 1380 (Root).
Palamon and Arcite.
After 1374 (ten Brink). (?) 1375-6 (Koch). About 1381 (Mather). Shortly before 1385 (Pollard). About 1382 (Lowes). About 1380-2 (Root).
Legend of Good Women.
(?) 1385, the Prologue; the rest
probably at various times
(Furnivall). 1385; G-Prologue "hardly before
1393 " (ten Brink).1 1384-5, the Prologue and some of
the Legends ; 2nd Prologue, 1385
(Koch). 1385 (Mather). 1385-6 (Skeat). 1384-5 (Pollard). F- Prologue, 1386 ; Legends about
1379 and later ; G-Prologue, 1394
(Lowes). 1385-6 (Root).
Canterbury Tales begun.
(?) 1386 (Furnivall). About 1390 (ten Brink).
1385 (Koch). Probably 1387 (Mather).
1386 (Skeat). After 1385 (Pollard).
1387 (Root).
General Prologue.
(?) 1388 (Furnivall). (?) 1385 (Koch). 1387 or later (Mather). 1386 or later (Skeat). After 1385 (Pollard).
Man of Law s Tale. 1390 or soon after (ten Brink). (?) 1386-7 (Koch). 1385-1400 ; possibly earlier
(Mather). 1373-7 (Skeat). 1370-80 (Pollard). Before 1390 (Root).
Melibeus.
1386-7 (Koch).
1373-8 (Mather).
1372-7, revised later (Skeat).
After 1385 (Pollard).
Monk's Tale.
(?) 1386-7 (Koch). 1373-8 (Mather). 1369-73 (Skeat). 1373-80 (Pollard). Early (Root).
Physician's Tale.
About 1388 (ten Brink). (?) 1386-7 (Koch). After 1382-5 (Skeat). After 1385 (Pollard). Before 1387 (Root).
Clerk's Tale.
About 1388 (ten Brink). (?) 1386-7 (Koch). 1385-1400 (Mather). About 1372-3 (Skeat). 1373-80 (Pollard).
Second Nun's Tale (" Life of St. Cecelia11).
(?) 1373 (Furnivall). About 1373 (ten Brink). (?) 1373-4 (Koch). Shortly after 1373 (Mather). 1369-73 (Skeat). 1370-4 (Pollard). 1373-4 (Root).
Enyl. Stud., xvii. 20.
CONTENTS.
PAGE
PREFACE v
INTRODUCTION vii
CHAPTER
v I. THE TROILUS AND GRISEYDE
§ 1. The Two Versions 1
§ 2. The Date (1377) 15
^ II. CERTAIN MINOR POEMS
§ 1. The House of Fame (about 1379) .... 34
§ 2. The Parliament of Fowls (1381) . . . . 41
III. POEMS DEPENDENT ON THE TESEIDE
§ 1. The Pcdamon and Arcite : Its Original Metre . 45
§ 2. The Knight's Tale : How Far Altered ... 66
§ 3. The Knight's Tale : The Date (1384-6) ... 70
§ 4. The Andida and Arcite (about 1383-4) ... 83
IV. THE LEGEND OF GOOD WOMEN
§ 1. The Two Prologues : The Question of Priority _ . 86
§ 2. Its Connection with the Queen .... 102 § 3. The Legends and the Date (1386-7) . . . .121
V. THE CANTERBURY TALES
§ 1. The Canterbury Tales as a Whole (1387-1400) . . 131
§ 2. The General Prologue (1387) 142
§ 3. The Physician's Tale (1388) 150
§ 4. The Clerk's Tale (after 1387) 156
§ 5. The Monies Tale (after 1387) 164
§ 6. The Man of LauSs Tale (1390-4) .... 172 § 7. The Tale of Melibeus (1388-94) . . . .188 § 8. The Wife of Bath's Prologue (1388-94) and Tale, the Shipman's Tale (1388-93), the Merchants Tale
(before 1395) 198
APPENDICES : —
A. The Date of Gower's Mirour de I'Omme (1375-81) . 220
B. The KnighVs Tale and the Teseide : Tables . .226
C. Chaucer's Treatment of the Teseide ... . 231
xiii
Ck Cbaurer
O iJ
NOTICE.
DURING the years 1903-6, the Society's Editors did not enable it to issue any Text except the short No. 36, the Four- Days Journey from London to Canterbury and back of the Aragonese Ambassadors in 1415. But several Subscribers generously continued to pay their Subscriptions, so that the Society has now rather more than £800 in hand to pay for its issues of 1903, 1904, 1905, 1906 and 1907, five years. These issues will be dated 1907 or 1908, &c., the year in which they are sent out, but about £200 worth of work will be assigned to each of the back years in which no Text was issued. The present volume, Prof. Tatlock's Development and Chronology of Chaucer's Works, will be taken as the second Text for 1903. It is hoped that Prof. McCormick will soon issue two vols. for 1904, and Miss Spurgeon and Miss Fox one — the Chaucer Allusions, 1360-1900, Pt. I— for 1905, with Prof. Syphard's work on The Hous of Fame, which has been for some months in the printers' hands. So far as is possible, the money paid in for every year will be spent on Texts for that year; and these Texts will be sent to the payers of the money.
The Announcements as to the issues for 1907 on the cover of Prof. Tatlock's volume will be alterd, in future Texts, so as to correspond with the Notice above.
F. J. FURNIVALL.
June 14, 1907.
an4
of Ottaitfep Itok
V CHAPTEE I.
THE TROILUS AND CRISEYDE. § 1. The Two Versions.
THE first suggestion that there are two genuine versions of Chaucer's Troilus and Criseyde was made by Dr. Furnivall, who in the Chaucer Society's Parallel-Text Print of Chaucer's Troilus and Criseyde (London, 1881, p. 195) indicates that a certain difference among the MSS. as to arrangement may be due to the poet himself. Professor W. S. McCormick, in a paper read before the London Philological Society, Dec. 6, 1895, and briefly reported in the Academy, Dec. 21 (no. 1233, p. 552), supported the idea with greater definiteness, and illustrated it by eleven printed pages containing five or six hundred various readings from the sixteen MSS. and Caxton's and Thynne's editions.1 When he wrote his introduction to the Troilus in the Globe Chaucer (London, 1901, pp. xli-xliii), he had come to believe in three versions, each represented by one of the three families into which he regards the MSS. as falling; the second containing "more -than one partial revision," and the third being "a later copy, either carelessly corrected by the author, or collated by some hand after Chaucer's death." His introduction to the Troilus MSS., announced by the Chaxicer Society in 1894 as at press, in which he may be expected to deal with the whole subject more authoritatively than any one else can do, has never appeared, and he has nowhere in print defended or even expressed his views in any detail.2
1 He indicates cases where one reading is nearer than another to the Italian or Latin original.
2 The probability of a revision is recognized also by Dr. F. J. Mather (Furnivall Miscellany, p. 309; Chaucer's Prologue, etc., Boston, 1899, p. xix) ; by Dr. G. L. Hamilton ( The Indebtedness of Chaucer to Guido, New York, 1903, p. 149); as well as by Dr. John Koch in his review of the Globe Chaucer, Engl. Stud, xxvii. 12 (cf. Chronology, p. 36),
P.EV. CH. B
2 THE TROILUS AKD CRISEYDE. [ciI. 1, i; 1
That Chaucer should at some time or other have revised the Troilus is far from being improbable a priori, even though revision was not his custom.1 Of his longer poems it is the most carefully studied and the only completed one, a work on which he must have spent some of his closest meditation, so mature that he could never have grown beyond it, as he grew beyond some of his other works. He shows solicitude about the purity of its text (as we say now) in book V. 1793-8, and in the lines to Adam. It seems highly natural that when it befell his scrivener to write Troilus anew, Chaucer should not always have allowed him quite to reproduce the old copy.
The question cannot be wholly settled till the relations of the MSS. are clearer than they are now, but a strong probability can be established by the use of Professor McCormick's table of variants already mentioned, and of the seven MSS. published by the Chaucer Society. I have been fortunate also in being able to refer to certain unpublished researches of Professor Kittredge's on the relations among the MSS.
It is necessary first to discuss the principal MSS. concerned. It is impossible to construct a genealogy for them, but their relations have been sufficiently determined to insure fairly reliable results. These MSS. are the following :
Ph— Phillipps 8252 Jo— St. John's Coll., Cambridge 4
H4— Harleian 2392 Cp— Corpus Christi Coll., Carnb.4
Gg— Cambridge, Gg, 4. 27 2 Hj— Harleian 2280 2
H2— Harleian^3943 3 Cl— Campsall2
Ph is a late MS., and (according to Skeat) not of much value ; 5 Hj is a late, not very correct, paper MS. ; H., does not seem to be very good. Jo is called by Skeat "a fair MS., perhaps earlier than 1450"; Cl, written on vellum before 1 41 3 for Henry V. while Prince of "Wales, he pronounces one of the best, derived from a still better; Cp is also an excellent MS., fairly early, and prob- ably once in the possession of John of Gaunt's granddaughter, Anne Neville, Duchess of Buckingham ; H: Skeat considers third
1 The revision of the Prologue to the Legend of Good Women I shall try to show later was due to a peculiar reason.
2 In A Parallel-Text Print of Chaucer's T. and C. (Ch. Soo., 1881).
8 In W. M. Rossetti's parallel-text edition of the T. C. and the Filostrato (Ch. Soc., 1873).
4 In Three Afore Parallel- Texts of C.'s T. and C. (Ch. Soc., 1894).
5 See his Chaucer, II. Ixvii. ff. ; and his Piers Plowman, II. Ixx.
CH. I, § 1] THE TWO VERSIONS. 3
best. The first four MSS. in the table Professor Kittredge says are proved to belong at least in part to the same family of MSS. by common corruptions which unite them by twos and threes. But the relations of all the MSS. are very complicated, and were frequently disturbed by contamination.1 The only sure footing in this quagmire, but a very satisfactory reliance, is the almost com- plete agreement of the three last excellent MSS., two of which seem to have been, once in the possession of members of the royal family. This group I shall call (2); the other five, though, singly or several at once, they often agree with (2), may be grouped as (1). It should be added that Ph is McCormick's main reliance for his first version ; the next four he .assigns partly to the first and partly to the second redaction ; and the last three wholly to the third.2
The most important various readings are those where one alternative is distinctly nearer than the other to the Italian original, of which I give some ten from McCormick's lists :
I. 111. With chere and vois ful pitous, and wepinge,' Ph, H2 E con voce e con vista assai pietosa
With pitous vois, and tendrely wepinge, Gg, Jo, (2)
IF. 734-5. Men loven women al this toun aboute ; Ph, H2, Og Be they the wers 1 why, nay, withouten doutei lo non conosco in qiiesta terra ancora Veruna senza amante, e la pih yente
s'innamora ...
E come gli altri far non e peccato.
Men loven wommen al biside hir leve,
And whan hem lyst namore lat hem leve. H4, Jo, (2)
IY. 57-9. To Priamus was yeye at gret requeste Ph, H2, (2) 3 A tyme of trewe, and tho they gonrien trete Hir prisoneres to chaungen, moste and leste.
Chiese Priamo triegua, efugli data;
E cominciossi a trattare infra loro
Di permutar prigioni quella fiata.
1 As is abundantly proved by McCormick's tables ; and cf. Globe Chaucsr, p. xli. The contamination, it seems to me, may sometimes have taken place as follows : a scribe with a good verbal memory, having already copied the poem once or more from one redaction, when he came to copy it again from another, might easily at times insert the older reading which he chanced to have in mind. In various Chaucer MSS. there is good evidence that the scribes did become familiar with Chaucer's poetry at large. But this sort of contamination would be quite impossible to trace.
a For full information on the MSS. see Skeat II. Ixvii. ff. and Globe, xli. f.
:f This is the only case where (2) has what looks like ail earlier reading. See p. 11 below.
4 THE TROILUS AND CBISEYDE. [CH. I, § 1
But natheles a trewe was ther take
At gret requests, and tho they gonnen trete
Of prisoners a chaunge for to make. H4, Jo
IV. 246-8. His eyen two, for piete of herte,
So wepen that they semen welles tweye ; Ph, Gg
The heighe sobbes .... H2, Jo, (2)
/ miseri occhi per pieta del core Forte piangeano, e parean due fontane . Gli alti singhiozzi . . .
His eyen two, for pitee of his herte,
Outstremeden as swifte welles tweye ; H0, Jo, (2)
Therwith the sobbes ... Gg l
IV. 258. That wel unnethe the body may suffyse Ph, Gg
Che'i capo e 'I petto append gli bastava That wonder is the body may suffyse H2, Jo, (2)
IV. 736-763. Lines 750-6 immediately after 735, and 750 reads: "The salte teres from hir yen tweyne," with other important variants m 747, 752, 757, 762-3,— Ph, Gg, Jo. Order and readings as in Skeat — H2, (2). (The first order is Boccaccio's, but the second agrees better with 735.)
IV. 882. As he that shortly shapeth him to deye. (1)
II qual del tutto in duol ne vuol morire
For verray wo his wit is al aweye. (2) 2
IV. 1214, And he answerde, " Herte myn, Criseyde," Ph, Gg, Jo A cui il disse : " Dolce mio disiro,"
And he answerde, " Lady myn, Criseyde," H2, H4, (2) 3
IV. 1218. And he bigan conforte hir as he niighte. Ph, Gg, Jo
Come potea . . . La conform,
And he bigan to glade hir as he mighte. H2, (2) 4
V. 923-4. I wil be he to serve yow myselve, H2, Gg, Jo
Ye, lever than be king of Greces twelve.
. . . assai degno amadore . . . io sarei desso, Pik volentier die re de' Greci adesso.
1 The other reading is probably the only genuine one, for this seems to occur in but one MS. ; Ph has " the sobbes," which may be the middle term between the two readings.
2 This line may be inferior in itself, but it greatly improves the gram- matical construction. Cf. the curious punctuation which McCormick, who keeps the first reading, finds necessary in the Globe edition.
"Herte myn" occurs in 1216 ; hence the change.
4 "Hir to glade "occurs in 1220, so the variant may possibly be a scribe's blunder.
CH. I, § 1] THE TWO VERSIONS. 5
I wol be he to serven yow myselve
Ye, lever than be lord of Greces twelve. (2)
These cases are only about half of those given by McCormick, though they are the most striking. It can hardly be doubted that at any rate most of these variations are due to Chaucer ; and there- fore that the second set of readings, in which he departs from his original, are the later.
A number of cases may be noted where a change seems to have been made in the interest of ancient and especially pagan colouring :
III, 188-9. Withouteu honde, me seraeth that in toune,
For this miracle, I here eche belle soune.1 H2, Gg, Jo
For this merveille ... (2)
III. 705, 712. Seynt Venus in one line or the other in every MS. of (1). Blisful Ventis in both lines in (2).
IV.299-301. Ne never wil I seen it shyne or reyne, Ph, Gg
Ne hevenes light ; and thus I in derknesse My woful lyf wil enden for distresse.
But ende I wil, as Edippe,2 in derknesse
My sorwful lyf, and dyen in distresse. H2, Jo, (2)
IV. 644. But any aungel tolde it in thyn ere. Ph, Gg, Jo
But-if that Jove tolde it in thyn ere. H2, (2)
These last two changes are certainly Chaucer's own, and if ver- sion (2) is later than (1), so .are the others, for the change from mediaeval to ancient colouring could hardly be due to a scribe.3
In a large number- of cases some stylistic reason is evident for the change from (1) to (2).
I. 640. NQ no man wot what gladnesse is, I trowe, (1)
Ne no man may be inly glad, I trowe, (2)
(Four other words in -esse occur just before and after.)
1 This is the earliest occurrence, so far as I can find, of an impressive cir- cumstance common later in ballads and folklore. 1'robably Chaucer derived it from some ballad or popular romance now lost. Cf. the ballad of Sir Hugh of Lincoln, Child's Ballads, III. 244 ; and ibid., I. 173, 231 ; III. 235, 519.
2 Troiltis is speaking ; but we may notice that one of Criseyde's favourite books was the Siege of Thebes (TI. 84).
3 There is just one case of the opposite kind. In II. 115 (1), except Jo, reads "Ye maken me by Joves sore adrad"; (2) and Jo read "By god, ye maken me right sore adrad." The change would be a strange one for Chaucer to make ; and since "god" occurs twice in the two preceding lines, it may be due to the scribe.
r, THK TROILUS AXD CJtKEYDE. [dl. 1, 3 1
II. 1210. Xow for tlie love of god, my nece deve, Gg, Jo
Xow for tlie love of me, my nece dere, HO, (2)
(The second is more Pandaresque ; not likely to be due to a scribe. Cf, II. 290.)
III. 256. Thou wost thysclven what I wolde mene. (1)
Al seye I noght, thou wost wel what I mene. (2)
III. 269. For never was ther wight". . . That ever wiste'H4,Gg, Jo For that man is unbore ... Ph, H2,(2)
III. 672. Than is it tyme for to gon to reste. H4, Jo
So go we slepe, I trowe it be the beste. H2, Gg, (2)
III. 677. And alwey in this mene whyle it ron. H4, Jo
And evere mo so sterneliche it ron. . H.2, Gg, (2)
IV. 638. Pandafe answerde — " Of that be as be may." Ph, Gg, Jo
"Why, so mene I," quod Pandare, "al this day." H2,(2)
IV. 1097. Canst thou not theuken thus in thy disese? Ph, Gg, Jo Lat be, and thenk right thus in thy disese. H2, (2) (No less than seven rhetorical questions have come just before.)
IV. 1138-9. So bittre teres weep not thurgh the rinde Ph, Gg, Jo The'woful Myrra, written as I finde,
So bittre teres weep not, as I finde,
The woful Myrra thurgh the bark and rinde. H2, (2)
Cf. also the following passages, which make in the same direc- tion: II. 1399, IV. 165-6, IV. 560 (cf. 567, 570), IV. '581 (cf. 580), IV. 696-8.
In other cases, though the motive for the change is less obvious, it is difficult not to attribute it to Chaucer.
II L 501-3. Som epistle . . . That wolde . . . wel contene H4, Jo An hundred vers . . .
Neigh half this book . . . H2, Gg, (2)
III. 568. And she on game gan him for to rowne H4, Jo
, Sone after this to him she gan to roune H2, Gg, (2)
III. 143G-42.
Thou dost, alias ! to shortly thyn oifyce, H4, Jo
Thou rakel night, ther god, maker of kynde,
OH. I, § 1] THE TWO VERSIONS 7
For thou dounward thee hastest of malyce,
[Thee for thyn haste and thyn unkinde vyce] H2 Gg, (2)
Thee curse and to our hemi-spere bynde,
[So faste ay to our hemi-spere bynde] H0, Gg, (2)
That never-more under the ground thou wynde !
For thurgh thy rakel hying out of Troye
[For now, for thou so hyest out of Troye,] H2, Gg, (2)
Have I forgon thus hastily my joye. H4, Jo
IV. 789-90. ... the feld of pitee . . . Ther Pluto regneth Gg, Jo „ „ „ „ That hight Elysos H,, (2)
IV. 828-9. Myn eem Pandare of Joyes mo than two Ph, Jo
Was cause causing first to me Criseyde. Pandare first of Joyes mo than two Was cause causing unto me Criseyde. H.,, (2)
Cf. also III. 543, 668; IV. 1093, 1113.
There are three important passages, the omission of which in some MSS. is strong additional evidence for more than one redac- tion. The first is Troilus' hymn to love at the end of book III. (1744-1771), from Boethius. The second is Troilus' long soliloquy (IV. 953-1085) on free-will, also mostly draAvn from Boethius. The third is the account of the ascent of Troilus' soul to heaven (V. 1807-1827), drawn from the Teseide of Boccaccio. A particu- larly significant fact is that they were all three omitted in MS. Ph, and in somewhat the same list of other MSS.
Troilus' hymn to love is absent from MS. Harl. 3943, and inserted later (which means the same thing) in MS. Ph ; in all other MSS. and early editions it seems to have been present from the first. Boccaccio at this point (III., st. 74-89) puts a very long hymn to love into Troilus' mouth, the first six stanzas of which Chaucer used to form the greater part of the proem to this book of the Troilus. Troilus' hymn to love in Chaucer, therefore, is not from Boccaccio, but is a versification, with a slight rearrangement, of Boethius, II., metre 8. The song is not at all likely to have been cut out by the scribe, and cannot possibly have been omitted accidentally. Its absence is a clear sign of incompleteness, for the context runs (in MS. HI. 3943,1 III., 11. 1743, 1772-3) :
And pan he wold syng in pis manere.
In al pe nedis for pe tounys werre he was & ay pe ferst in armys dight.
1 See W. M. Rossetti, parallel-text edition, 167-8.
8 THE TR01LUS AND CRISEYDE. [oil. I, § 1
The first of those lines translates the end of the stanza just pre- ceding the song in Boccaccio, and the second translates the first of the stanza just following it. It is clear, therefore, that Chaucer omitted Boccaccio's song for the ohvious reason that he had just nsed the first part of it ; and that he allowed some MSS. of the Troilus to go into circulation before he added the substitute. Such carelessness on his part is not unparalleled in the Canterbury Tales.1
The second of the three passages, IV. 953-1085, is wholly omitted in MSS. Harl. 2392 (H4) and Harl. 1239,2 omitted all but the last stanza3 in Gg, which hereabouts agrees with (1), and added later in Ph ; it is present in Harl. 3943 (H2), Johns and (2). In the first place, it is important to notice that the passage forms a complete unit; every stanza in it (except the last) is Boethian and scholastic, and its length and subtlety form a strange break in Troilus' passionate despair.4 It is hardly likely that so long and so unified a passage would have been omitted by a scribe. Secondly, the continuity of the context is better Avithout it. In line 947 Pandarus finds Troilus alone in the temple, yet seems to be stand-
1 It may be asked whether the present proem, or such part of it as is from Boccaccio, may not have originally stood as Troilus' song. Internal evidence is much against such a view. The first three lines of stanza 6 are fairly closely translated from the Italian, yet the last four lines are spoken by the author in his proper person, and cannot possibly have been in Troilus' mouth ; so also stanza 7, which is not, however, from the Italian. At first sight MS. Eawl. Poet. 163 ("not a very good copy," according to Skeat, II. Ixxiv.) seems to suggest that the proem was lacking in Chaucer's first copy, for this MS. (only) omits the proems to books II. -IV. (see W. S. McCormick, pp. 296-300 of An English Miscellany Presented to Dr. Ftirnivall, &c., Oxford, 1901). But this argument is quashed by two considerations. In the first place, this MS. has, in its proper place, Troilus' hymn from Boethius, so the absence of the proem is certainly not due to its use elsewhere. Secondly, the proem to book IV. we can hardly doubt was written continuously with books III. and IV., for all three correspond to consecutive parts of the Italian ; stanza 93 of Boc- caccio's third book is rendered at the end of T. C. III., the proem to IV. includes most of the 94th and last stanza, and book IV. of Chaucer begins with Boccaccio's next stanza (IV. 1). If the absence of the proem to IV. from MS. Rawl. cannot be due to its absence from Chaucer's first version neither can it well be argued that its omission of those to books II. and III. is. The absence of these proems, therefore, is a sign of lateness, not of earli- ness ; so much so, in fact, that it seems to me probably due to the scribe, not to Chaucer. But on this matter, as on so many others, we must defer to Professor McCormick's views, when they shall be expressed.
2 Printed in Three More Parallel- Texts. In this part of the poem it generally agrees with (1).
* This clearly belongs with what precedes, for 1080 ("wost of al this thing the sothfastnesse ") refers to the philosophical disquisition, not to the amorous lament in 950-2 : so also does 1084 (" Disputing with himself in this matere ").
4 Cf. Lounsbury, Studies, III. 374-5.
CH. I, § 1] THE TWO VERSIONS. 9
ing at the door during the whole of this discourse, for he does not come in till 1085 ; these two lines almost contradict each other.1 Another piece of evidence that the rest of the poem underwent revision, and that during it this passage was added, is that the only variants in it noted by McConnick are four trivial ones clearly due to the scribes (957, 958, 989, 1064) ; this makes 1 variant to 33 lines, but elsewhere in book IV. according to McCor- mick's tables there are about 1 to 11 lines, including some very significant changes.2
A wholly different consideration which distinguishes this passage from the context is the rhyme-usage, as to which, by the kindness of Professor George Hempl, I am able to present some information gained by him. Excluding this passage, the impure o : y 3 rhyme occurs, to 1000 lines, 3 times in book I., 2 in book II., £ in book III., and not at all in books IV. and V. ; but in these 133 lines it occurs twice (1035-6, 1072-4). He also points out that the cheap rhyme- words in accented -inge (or -ing, participle or verbal noun) occur, to 1000 lines, 18 times in book I., 11 in II., and 4 in III. - V.4 ; but in this passage they occur no less than 1 1 times (986-7, 989-91-2, 1014-15, 1016-18, 1075-6)— more than twenty times as often as they should according to what is usual in book IV. The force of this last argument is somewhat weakened, to be sure, by the fact that such a discourse as that of Troilus naturally contains an unusual number of abstract nouns in -inge. But the two points together certainly distinguish the passage sharply from its sur- roundings.5
1 Pandarus is named in both 1085 and 1086. If the lines had been written consecutively the repetition would probably have been avoided.
2 To this bit of evidence cf. a parallel in L. G. W.,p. 119 below.
3 Cf. ten Brink, Chaucers Sprache und Verskuiist (Leipzig, 1899), p. 191 ; but cf. p. 23.
4 I. e. only 4 times to every 1000 lines in book IV., excluding this passage.
5 Since all this evidence shows that Chaucer became more fastidious as to his rhymes daring the composition of T. C., it may suggest to some that this passage must have been written before the greater part of the poem ; it may seem as if we had here another example of Chaucer's "economy "(to use Professor Koeppel's word) in putting pieces of old cloth into a new garment. But this is more than doubtful. The first two and the last stanzas were certainly written for T. C. , an d the others, with their plentiful use of the pronoun /, have the appearance of being. I doubt very much if any thoroughly consistent uniformity or development in rhyme-usage or metrical-usage can be made out in Chaucer's poetry ; and there does not seem to be any a priori reason why it should be. I am not ignorant that others take a vehemently opposite view, and that Shakspere's practice has been pleaded as a parallel ; but Shakspere's metrical development was part of a wide- spread, traceable and easily explicable national evolution in versification. On all this cf. Lowes in Public. Mod. Lang. Assoc., xx. 811-12.
10 THE TROILUS AND CRISEYDE. [CH. I, § 1
Jt can hardly be doubted, therefore, that Chaucer added this passage when the poem had been some time in circulation. Most readers will agree that it was no great improvement. At times it is impressive and beautiful, and recalls part of the Complaint of Mars and Palamon's fine lament in the KniyUfs Tale, but enough has already been said of its unsuitability to the <l lowed " Troilus in a mood of despair. This is not the only case where Chaucer appears as a careless or injudicious reviser. It should be added that there is nothing surprising in the inclusion of this passage in one or two MSS., such as the Johns, which otherwise in this book follow the first redaction ; for the passage was one sure to interest the serious-minded reader, and therefore to be copied in where it did not belong (as we can see happened in MS. Phillipps). In a case like this, omission is more significant than insertion.
The passage from the Teseide (V. 1807-1827) is absent from MSS. Harl. 3943 (H2) and 2392 (H4), and added later in Phillipps j1 MS. Gg, which in this book generally agrees with Ph, breaks oil before this point. The passage is present in Johns and (2). It can hardly be doubted that this, too, is a later insertion. The passage contains unsympathetic erudite conceits, brought from afar, and forces apart two lines (1806, 1828) which are consecutive in the Filostrato ; we may wonder a little that Chaucer should put it in at any time, but his doing so is more intelligible when the poem had grown somewhat cold to him.2 It is true that other passages at the end indicate some sort of revulsion of feeling on Chaucer's part ; but a Christian transcending of a worldly poem, a sense of the futility of earthly happiness, which a mediaeval man might easily draw from the Troilus, is not the same thing as a rather meretricious piece of that paganism which Chaucer expressly disclaims a little later (1849-55). In the other cases the Middle Ages were simply calling back one of their children who was escaping from them. Without this passage the course of thought is decidedly better ; as things are, " SAvich fyn hath, lo, this Troilus" (1828), "in this
1 It is highly interesting to note that the later insertion of these three pas- sages in MS. Ph show that it belonged to a really scholarly admirer of the poet. We have here an example of something like collation in the fifteenth century.
2 Koch thought (Engl. Stud., I. 271) it was put in on first writing because Chaucer considered Boccaccio's account of Troilus' death too brief ; later, lie thought it first appeared in a second version (Chronology, p. 36). Cf. also A. W. Pollard, Chaucer: The Knight's Tale (London, 1903), p. 110. Lowes takes a more favourable view of the addition (Publ. Mod. Lung. Assoc., xx. 847).
CH. I, § 1] • THE TWO VERSIONS. 11
wyse he cleyde " (1834), have to go back twenty or thirty lines for their explanation, and after his cheerful flight and scorn of those who wept for him it is a little .odd to return to the pathos of his death.1 Considering, then, that the passage is a unit, of different source from that of its context, that it is lacking in at least three related MSS., some of which also lack the Boethins passages, and that such a passage is less likely to have been omitted by a scribe where it once was than to have been inserted where it was not,— 'this passage, too, is a strong argument for revision.
It may be taken as proved, I think, that we have at least two versions of the Troilus.2 And almost all the evidence that bears on the question of priority has indicated that the version con- tained in (2), MSS. Corpus, Campsall, and Harleian 2280, is the later; it is the 1'arther from the Italian, and the better. Disregard- ing the fact that this version omits I. 890-6 and IV. 708-M, Admirable and even essential stanzas which must have been omitted by oversight,3 I find just one case worth mentioning where the reading of (2) looks like the earlier, the striking one recorded above on pp. 3, 4.4 A few cases, like this and those in the note may be
1 Similarly ten Brink (Chaucer Studien, pp. 60-1) ; I agree with Koch that some of his other arguments are not so good (Eng. Stud., I. 270). I defer till later a discussion of the idea that this passage is part of the debris of a stanzaic Palamon and Arctic (see pp. 49-51).
2 It is suggestive to compare the clearly genuine character of these revisions with the insignificant various readings on which Prof. R. K. Koot bases his con- jecture that Chaucer revised the Parl. of Fowls (see Journ. of Germ. Philol., V. 189-193), and Prof. J. B. Bilderbeck that he revised the first six Legends of Good Women (Chaucer's L. G. W., London, 1902; pp. 34-42); or to. com- pare them even with the peculiarities of MS. Had. 7334 of the C. T., which I believe are not due to Chaucer. Cf. my preface, p. v. The genuineness of the revisions is further suggested by the fact that nearly all that I. have recorded (many of McCormick's variants may be scribal) are in books III. and IV. Evidently Chaucer took most interest in the more intense parts of the story.
3 The former passage is known only in three MSS., all belonging to the first version (McCormick, in the Furnivall Miscellany, p. 300).
. 4 There are three other possible cases. . "Or that the god ought spak " (Ph, H.,, Gg ; III. 543) introduces more variety than " Er that Apollo spak " (Jo, (2) ; cf. 541, 546). In V. 436, MSS. H2* Gg, Jo, and HI 1239 have it that Sarpedon was " ful of heigh largesse" ; (2) says he was "ful of heigh provv- esse" ; the Italian has "d'alto cuore," while the stanza dwells on his hospi- tality. But the first reading is doubtful English, and is very likely a scribe's blunder. In V. 1502-4, where the reference is to the Thebaid of Statius, IX. '497-539, 867-907, the reading of Gg and Jo is slightly more faithful to- the Latin (though it shows less familiarity with it), than that of Ph, H2, and (2). [Note here an important case where Ph agrees with (2) against others of (1). See below.] But when the reference is to another work than the general source of the poem accuracy is ambiguous in its testimony, and the second reading is hotter in other ways. Obviously nothing can be based on these cases.
12 THE TUOILUS AND CRISEYDE. [cH. I, § 1
accounted for in so many ways that they do not weaken perceptibly the conclusion that the version consistently represented by the second group of MSS. is the later.
Professor McCormick,1 as has been said above, believes that the versions which we have been discussing are the second and third in point of time, and that from the second may be extracted a first. This opinion is much more difficult to deal -with by evidence, so it is important to realize that the burden of proof is heavy upon one who holds it.2 The evidence accessible at present seems to me to be anything but favourable to the idea of a third version. In the first place, though it is quite true that we should not expect many cases of three genuine readings for one passage, it would be natural that in some cases Chaucer should not have satisfied himself even in his second version. Now there are no cases where a third read- ing carries conviction of its genuineness; and only twice can a third reading which occurs in more than one MS. possibly be considered.
II. 737-8. . . .he able is for to have . . . the thriftieste
That womman is, so she hir honour save. Ph, H.,, Gg
As ferforth as she may hir honour save. H4, Jo
To ben his love, so she hir honour save. (2)
III. 458-9. Lest any wight divynen or devyse
Wolde in this speche . . . Ph, H2, Gg
Wolde on this thing . . . H4, Jo
Wolde of hem two ... (2)
In the first passage the second reading is probably corrupt, and in the second the first. In neither is there any evidence for a third edition.
The only other satisfactory evidence would be a MS. which should consistently embody it, as group (2) constantly represents a version different from that best represented (according to McCor- mick) by MS. Camb. Gg; which should be nearest of all to the Italian, and which should sometimes agree with the second version and not with the third, and sometimes differ from both, and should never follow the third only. These demands are exacting, of course, but an approximation to them would be necessary in order to carry conviction. Some such MS. McCormick appears to think
1 Globe Chaucer, p. xli.
2 Many little slips in the C. T. and elsewhere show that Chaucer was not much in the habit of even reading his own poetry.
CH. I, § 1] THE TWO VERSIONS. 13
we have in Phillipps, and at the very beginning of his table we seem to find confirmation of his opinion. It can hardly be doubted, as we have seen already, that
" With chere and vois ful pitous, and wepinge " (I. Ill) was Chaucer's original translation of
" E con voce e con vista assai pietosa," and that he made a later improvement in
" With pitous vois and tendrely wepinge " ;
now the first reading occurs only in Ph. and H2, which agree closely throughout this book, and the other MSS. of (1) agree with (2). But this, so far as I can discover, is absolutely the last evi- dence of the sort ; there is no other significant case in which Ph is closer to the Italian than our last version, where the Gg MS. is not just as close.1 Moreover, the Ph MS. seems, on the testimony of Professor Skeat2 and Professor Kittredge, unlikely to deserve the importance which Professor McCormick attaches to it ; it is late and very corrupt, and appears to be at the end of a long descent ; it would be not a little strange if this MS. alone should preserve the first version intact But the most ruinous charge against MS. Ph is that several times during book III. and elsewhere (among others, in some of the passages quoted above) it switches over and agrees Avith (2), the Corpus-Campsall group, which throughout, McCormick says, represents the third version, while his second group (Johns, etc.) differs from both. This on his theory is absolutely inexplicable3; it can indicate just one thing — that in book III., at least, Ph is derived or corrected from some MS. of group (2). But if in practically all significant variations, Ph follows MSS. now of my group (1), now of (2), what becomes of its independence, of its testimony for a version different from both ? 4
1 The omissions or later insertions in Ph (already treated) are not peculiar to it.
2 See his Piers Plowman, II. Ixx.
3 The possible suggestion that Chaucer might have taken an uncorrected copy of the first version as a basis for the third, which would therefore at times follow the first and not the second, is negatived by the extraordinary solicitude which he shows for the text of the poem.
4 A further argument against the primatial position which McCormick assigns to MS. Ph is to be found in the peculiarities of MS, Eawl. Poet. 163, which he has thoroughly collated, and which his tables show to agree usually with (1), though it sometimes switches over to (2). In his article in the Furni- rall Miscellany (pp. 296-300) he shows that it contains at the very end of book II., between 1750 and 1751, a genuine stanza found nowhere else. Professor McCormick believes that it is misplaced ; but it seems to me that its insistence
14 THE TROILUfi AND CRISEYDE. [CH. I, § 1
The whole subject is immensely complicated ; to say that the poem underwent one thorough revision all at one time may possibly be too simple an explanation. All that I have said .must be, regarded as submissive to Professor McCormick's further communi- cations. But meanwhile it seems certain that Chaueer produced two versions, and fairly certain that he produced only two.1
As to the date of the revision, it is impossible to be very de- finite and certain, but it seems natural that some years should have elapsed between. There is one small, but perhaps respectable and certainly curious, piece of evidence in the two versions as to the date ' of the second. In book IV. , 596-7, MSS. Ph., Gg, Jo, Harl. 1 23£ and Harl. 4912 (all belonging, Apparently, to version 1.) make Pandaras say to Troilus, while urging forcible detention of Criseyde in Troy,
" It is no rape in my dom ne no vice, Hir to withholden that ye loven most," -
that Criseyde shall be merciful affords a perfectly logical connection with what precedes, and connects as well with what follows as 1750 does. It seems much less likely to have been added in this MS. than to have been omitted in the others, probably by a very early scribe. The MS. omits I. 890-6, no doubt by accident, and (as we have seen) the entire proems to books II., III. and IV. The presence of the unique stanza, and perhaps one or two of its other peculiarities, would put Rawl., and not Ph, in a peculiar position ; of which, again, it will be deprived by the fact that it agrees with three-quarters of the authorities in omitting the admirable (and indeed indispensable) lines I. 890-6, and in containing the song of love from Boethius. So we are farther than ever from having a MS. which consistently embodies Chaucer's first version. Is not the cruelly kind answer to the puzzle that which McCormick elsewhere shows must so often put the textual critic out of his misery : namely, contamination ? The more one studies the MSS. the clearer it becomes that Chaucer was not the only person who cared about the purity of his text In the fifteenth century there were more fastidious and critical readers than we always realize. In a graphic passage of the preface to the second edition of the C. T. (quoted by McCormick elsewhere) Caxtoii tells how one of his customers protested against the incorrectness of the first, and supplied him with a better copy.
1 The next thing we may hope for is a parallel-text edition of the two versions, which perhaps could be produced with a fair amount of accuracy.
2 It is worth noting that here is a clear case where rape means forcible de- tention or removed. It is high time that the more disagreeable interpretation of the incident to be mentioned were dismissed for good to the Limbo of Vanities. Chaucer's own father was abducted — "rapuerunt et abduxerunt " (Life Records, 1900, p. ix.) ; and in 1387 the thief was set to catch a thief — Chaucer was on a commission to inquire into the abduction of an heiress, of which exactly the same verbs are used (ibid., p. 270). On the frequency at the end of the fourteenth century of this sort of abduction and forced marriage, see S. Armitage-Smith, John of Gaunt, pp. 350-1. If the worse interpretation were the true one, is it conceivable that Chaucer would have adopted such a beginning to the Wife of Bath's Tale (D, 888), a beginning confined to his version of the story ? Cf. also Fumivall's Trial Forewords (Ch. Soc., 1881), pp. 136-44, for the law bearing on the subject.
CH. I, § 2] THE DATE. 15
for which the other and later authorities, (2) and Harl. 3943, read, " It is no shame unto you ne no vice " . . .,
certainly weaker and less appropriate. We ought to be able to discover some reason for the change. ISTow it will be remembered that on May 1, 1380, one Cecelia Chaumpaigne executed an instrument of release to Chaucer, " de raptu meo." l It may be not quite fanciful to suggest that when in the course of revision Chaucer came to this passage, a recent disagreeable in- cident sprang before his mind, and even at the cost of substitut- ing an inferior phrase he seized the opportunity of removing the reminder from his own and his friends' sight. He can hardly have been proud of the episode, and had probably suffered in his pocket.2 If this suggestion is allowed some weight, it indi- cates 1 380, or somewhat later, as the date of revision, which fits admirably (as will be seen later) with the evidence as to the date of first composition.3
§ 2. The Date.
The date of the original writing of the Troilus and Criseyde has always been a good deal of a problem, and it cannot be said to be settled yet. In 1903 I showed4 reason to believe that the poem was mentioned by Gower in his Mirour de I'Omme, in a passage (5245-56) which it seemed then could hardly have been written later than 1376, but which may probably date from about 1377.^ This early date has recently been argued against briefly by Professor John Koch,6 and more at large by Professor J. L. Lowes.7
1 See Life Records of Chaucer (Ch. Soc., 1900), pp. 225-7.
2 The force of this conjecture is not destroyed by the fact that he allowed the verb ravisshe to stand in IV. 530, 637 and 643, and in V. 895, and the noun ramsshynge in I. 62 and IV. 548 ; for the two forms of the word are so different in appearance and connotation that they would not necessarily be closely associated ; rape inevitably suggests the raptus, not so ravisshe. [Cf. snch a use of the verb as in T. C., IV. 1474 and N. P. T., 4514 (" So was he ravisshed with his flaterye ").] Moreover, Chaucer may not have been earnest enough in his antipathy to undertake so many further changes.
3 I shall show later that the insertion of the Teseide stanzas can hardly have been done later than the writing of the Knight's Tale (which I hold to be practically identical with the Palamon). The revision, therefore, must have considerably antedated the Prologue of the Legend, 1386. See pp. 74-5 below.
4 In Modern Philology, I. 317-324. I need hardly repeat the criticism of previous conjecture there given.
5 For a full discussion of the date of the Mirour I must refer to Appendix A, pp. 220-5.
6 Engl. Stud., xxxvi. 140-41.
7 Publications of the Mod. Lang. Assoc. of America, xx. 823-33.
16 THE TROILUS AND CRISEYDE. [CH. I, § 2
The objections of the former seem to me not difficult to meet. He thinks the period from 1373^0 1376 too crowded by the St. Cecelia, the Palamon and the Boethius. But there is not the least necessity for putting the first and last here,2 and the best possible reason, as we shall see later, for not putting the Palamon here and for not believing 'that any part of the Troilus Avas derived from that poem. The idea that the word comedie in Troilus, V. 1788, implies prevision of the House of Fame or the Parliament of Fowls I tried to show in this very article is groundless ; as also the gratuitousness of the idea that the epithet " moral " applied to Gower in V. 1856 must refer to any of his longer poems. "We may reasonably suppose that he was born about the year 1330 or possibly somewhat later ; " 3 are we to suppose that at the age of forty-five he had written nothing or shown no traits of character that would have earned him such an epithet from a personal associate as well then as ten years later ? Happily we are coming to realize Chaucer less as a literary phenomenon and more as a man ; were not his relations with Gower rather personal than literary ] Nor can I see that four or five years is too short a time for such modifications in the Troilus of Boccaccio's conception as Dr. Koch mentions. Altogether, therefore, he does little but reiterate, with- out developing, the arguments which I tried to refute at the begin- ning of my article. He suggests that Chaucer was writing the Troilus but had not yet finished it in 1376. But he does grant that Gower's reference is to Chaucer's poem.
Lowes' discussion demands more extended treatment. His argu- ments against my interpretation of the passage in Gower it will be more convenient to treat later ; first I shall consider his arguments in favour of a late date, that which he suggests being 1383-5.4 One matter which bears on the date of the Troilus is its relation with the Legend of Good Women. Lowes adopts 5 and develops ten Brink's view of a close chronological relation betAveen them. The matter can be discussed here only by anticipating some points in my discussion of the later poem. He declares (p. 821) that "the immediate occasion of the Prologue was manifestly the stir caused
1 May, not November (as Koch says, ignoring Mather's rediscovery of the date) ; see my article, p. 319.
2 On the Boethius, see p. 34 below.
3 G. C. Macaulay's Gower, IV. xxix.
4 Piiblic. of the Mod. Lang. Assoc,. xx, 861,
5 Ibid., pp. a 19-23,
CH. I, § 2] THE DATE. 17
by the publication of the Troilus," but I believe a very good case caa be made out for a different view. The God of Love reproaches the poet (322-35) *• with enmity to him and his servants, with hin- dering them by his " translacioun," and with having " translated the Eomaunce of the Eose," and having said as he " liste " of Criseyde. The Romance of tlie Rose, the translation par excellence, is at least as prominent in this passage as the Troilus, and so also in Alcestis' defence (362-72, 441). Therefore there is nothing in these references to make one suppose that the Troilus had just been published, anymore than that the Romance of the Rose had just been. But what is more important, I hope to show later strong reasons for believing, as Lowes does not, the orthodox view that not only is the F-Prologue an elaborate compliment to the queen, but that the whole Legend may have been written at her request. She landed in England in December, 1381, a girl of fifteen, who almost certainly knew no English, and it would be some years before she Avould be familiar with Chaucer's poetry. It seems to me that the language of the Prologue is at least as consistent Avith the view that she had just become familiar with his poetry and urged him to a more gallant manner towards women, as with the G view that it is the product of a supposed general sensation produced by the first publication of the Troilus.2
Of Lowes' arguments for a late date for the Troilus, there remain two — the fact (pp. 820-821) that the end of it seems to suggest prevision of the Legend of Good Women, and its excellence and maturity (833-840). As to the second, I have nothing to say against his fine analysis of some of the virtues of the poem ; assuredly, he says none too much of its vigour of characterization, its artistic mastery and its skill in dialogue and in episode. But I do deny his conclusion. In the first place, to an extent which is seldom realized, and which deserves much fuller treatment than this, the merits of the Troilus are due to the Filostrato. To my V mind the latter is quite as good a poem ; it is better proportioned, and its characterization, if less complex and attractive, is most natural. Again, I see no difficulty in believing that the powers evinced by the
1 I shall here assume that version F (" B ") of the Prologue is the earlier, a view which Lowes has done so much to establish. If G ( " A ") were the earlier, it would not matter in this connection.
2 The use in L. G. W. of three stanzas from the opening of the Filostrato (discovered earlier by Lowes ; cf. his article, pp. 822-3) of course is not sur- prising, since Chaucer owned a MS. of that poem, and implies no necessary chronological connection of T. 0. with L. G. W.
DEV. CH. G
18 THE TROILVS AND CRISEYDK. [CH. I, § 2
Troiliis were developed within a few years of Chaucer's introduction to Italian literature. It is possible to misunderstand the. Italian influence on Chaucer ; what it did for him, it seems to me, was to open the sluice rather than to fill the reservoir. He had long been a mature man, and, what we do not always remember, familiar with the greatest poets of the Romans. Till he went to Italy, what he lacked was a poetic form, and the ability to assimilate the influence of the ancients; he had had hitherto only the trouvere manner of the French. The Trecentisti were in part an inter- mediary between him and the ancient and higher ideal of poetic style, they performed (if so humble a metaphor may be allowed) the function of the plant between the mineral arid the animal. I see no reason why under a keen stimulus the poet should not have rapidly overtaken the man, why Chaucer could not do at thirty -five what he could do at forty-three. Any number of other poetic biographies will bear me out.1 As to the particular qualities which Lowes dwells on, it seems to me they would be almost as sudden in appearing at the latter age as the former, for I cannot possibly believe that the Palamon and Arcite and the Legends preceded the Troilus. Again and again a priori arguments of this kind have burst before a piece of evidence. May I say that I have become gradually but firmly convinced that Chaucer's literary manner after 1372 depended far less on the time of life when he was writing than on the character of his subject?! This is a highly important point, to which I shall have to return I repeatedly in treating the Canterbury Tales. It will account for the inferiority of the House of Fame and the Parliament of Folds to the Troilus. Therefore I cannot feel that the excellence of the Troilus is an argument against an early date.
The most striking point which Lowes makes, it seems to me, is the foreshadowing of the Legend in Troilus, V. 1772-85 ; Chaucer wishes he might write of Penelope and Alcestis, and warns women against false men.2 There is nothing surprising in the occurrence of this passage in the Troilus; even without the Legend it would not
1 Lowes at times well illustrates Chaucer's procedure by Tennyson's. May not the rapidity with which Chaucer responded to the Italian stimulus be paralleled by Coleridge's sudden poetic growth under the influence of Wordsworth ? Both he and Chaucer were impressionable poets, and it seems to me that their rapid growth was exquisitely natural.
2 Lowes, pp. 820-1. As to the camedie in line 1788, Lowes and I both show thoroughly that it cannot be made to imply any particular plan (P. M. L. A., xx. 855 ; Mod. PhiloL, i. 318).
CH. I, § 2] THE DATE. 19
"have seriously puzzled any one for a moment" (to borrow Lowes' own language from where it is less in place, p. 828, note) ; and there is nothing unlikely in Chaucer's having vaguely foreseen the Legend years before he wrote it. If it was written at the queen's sug- gestion, this passage at the end of the Troilus may have been what made her think of such a reparation for " the Eose and eek Criseyde." At any rate, I cannot think for an instant that this passage can be weighed against the evidence for an early date ; to which we may now turn.
Two considerations point to a fairly early date for the Troilus, earlier certainly than 1385, the date which Lowes assigns it. To begin with, it is well known that Chaucer is very fond of his own words, and constantly repeats favourite or convenient phrases or lines. I shall later have to point out very many cases of this. Now the present Knight's Tale is connected with the Troilus on j the one hand, and the Legend of Good Women on the other, by I a large number of such repetitions, as I shall show later, which! seem to indicate for the original Palamon and Arcite a position between the two.1 The absence of such parallels between the Troilus and \>\\Q Legend is very striking, considering their frequent parallels to other poems. Except for the passage in the Troilus Avhich foreshadows the Legend, and for one or two expressions which are paralleled in the Knight's Tale as well (which therefore was probably the transmitter), I find only two common to the Troilus and the Legend. T. C., IV. 15, is almost the same as L. G. W., Prologue G, 265 :
" For how (How that) Criseyde Troilus forsook."
But here, it will be seen, the parallel is in the prologue which we shall see is surely the later, dating from about 1394. T. 0., III. 733-4, is parallel to L. G. W., 2629-30 :
" 0 fatal sustren, which, er any clooth Me shapen was, my destene me sponne ;"
" Sin first that day that shapen was my sherte, Or by the fatal sustren had my dom."
But most of this is paralleled in the Knight's Tale.2 Considering,
1 See pp. 76-8 below, in my chapter on the Teseide poems. The value of this evidence is recognized by Skeat. though it makes against his chronology (iii. 394), and by Mather (Furnivall Miscellany, p. 308).
2 "That shapen was my deeth erst than my sherte " (1566). For the rest, T. (J., 111. 1282 = Kn. T., 3089 = L. G. W., Prol. F, 162. In the passage
20 THE TROILUS AX1> CIUSEYDE. [CH. I, $ '2
then, the closeness of the Kniglifs Tale in phraseology to the I1 roil us and to the whole Legend of Good Women, it is very strik- ing that there should be almost no parallels between the two latter. It certainly makes against the view1 that the Troilus was written close to and between most or all the individual legends and the Prologue of the Legend of Good Women; that the legends were written about 1380, the Palamon about 1382, Troilus 1383-5, and the Prologue of the Legend in 1386.2 So far as this evidence has value, it seems to indicate an order of things like that which I arrive at by other methods : Troilus (revised later),3 Palamon, Legend of Good Women.
But there is one more piece of evidence against Lowes' date for Troilue, and somewhat in favour of mine. Skeat points out that it is mentioned and frequently quoted in the Testament of Love,* once attributed to Chaucer, but really by Thomas Usk.5 I need not repeat all the instances of borrowing which Skeat mentions in his notes;0 the important passage is where Usk openly refers to Chaucer and the Troilus. The discourse between the author and Love (in close imitation of that between Boethius and Philosophy) has been on divine foreknowledge and human free-will.
" ' I wolde now (quod I) a litel understande, sithen that [god] al thing thus beforn wot, whether thilke wetinge be of tho thinges, or els thilke thinges ben to ben of goddes weting, and so of god nothing is ; and if every thing be thorow goddes weting, and therof take his being, than shulde god be maker and auctour of badde werkes, and so he shulde not rightfully punisshe yvel doinges of mankynde.' — Quod Love, ' I shal telle thee, this lesson to lerne.
which foreshadows L. G. jr., T. C., V. 1780-1 = L. G. Jr., Prol. F, 486 (G, 476) ; 1782 = 2546 ; 1785 = 2387.
1 Lowes, P. M. L. A., xx 860-62.
2 This date we may gladly accept.
3 I may also recall the date, 1380 or shortly after, which I have suggested for the revision, which will throw the original writing far back ; the earlier we put the latter, the more natural is the thorough revision. It must be recollected that revision was far from being Chaucer's custom. The only other known case, that of the Prologue of L. G. W. , was due to a very special cause, as I believe we shall see ; as we shall also see that P. A. was probably altered only at the beginning and the end.
4 See the Supplement to Skeat's CJuiucer, vii. 1-145. Practically all the knowledge we have of this work is due to Skeat, to whom my treatment of it is indebted at every step.
5 On the authorship, see Skeat, VII. xx. It may be remembered that the attribution of the T. L. to Chaucer, and a misinterpretation of it, were re- sponsible for a particularly absurd part of the " Chaucer Legend " (cf. c. g. Hales, in Did. Nat. Biogr., x. 162, and Lounsbury's Studies, i. 188-90).
6 As a possible addition, cf. III. 9, 89-90 with T. C., V. 1856-9, and see p. 23, note.
CH. I, § 2] THE DATE. 21
Myne owne trewe servaunt, the noble philosophical poete in Eng- lissh ... he (quod she), in a tretis that he made of my servant Troilus, hath this mater touched, and at the ful this question assoyled. Certaynly, his noble sayinges can I not amende; in goodnes of gentil manliche speche, without any maner of nycete of storiers imaginacion, in witte and in good reson of sentence he passeth al other makers. In the boke of Troilus, the answere to thy question niayst thou lerne.' " l
As Skeat points out (with less conviction than seems to me in place), the reference is to Troilus, IV. 953-1085, the passage already discussed at large, where Troilus soliloquizes on the ques- tion whether God's foreknowledge interferes or not with man's free-will. Now the interesting thing is that, as we have seen, this passage came in on the revision. Therefore Chaucer's revised version of the Troilus was known to Thomas Usk.
The question as to the date of the Testament of Love may be answered with certainty and exactness.2 Usk refers to events of 1384 in London in a manner much more certain and detailed even than Skeat points out. According to Malverne,3 John Northamp- ton, who in 1383 had been mayor for two years, was very severe toward the fishmongers, who had charged excessive prices, and thereby for a time he won popular applause ; but by extending the same austerity toward other trades he awoke discord and alienated his former friends, insomuch that, when he came up for re-election, after a stormy campaign Nicholas Brembre was put in his place. But the two factions so failed to agree, and the validity of Brembre's election was so doubtful, that the royal authority seems to have been necessary to secure the office to him. He at once undid the work of his stern predecessor, and restored their liberties to the fishmongers. Shortly after this Northampton caused disturbances in London, was accused of provoking sedition, and was arrested and imprisoned by the King in Corfe Castle. Brembre, however, laboured to calm the tumults against Northampton, and to promote peace. Usk was arrested about July 20, 1384, and induced to betray Northampton's secrets and bring accusations against him ; these Northampton denied, declared Usk a false ribald, and defied him to single combat. Subsequently other leading citizens were
1 III., ch. 4, 11. 241-9, 253-9.
2 Here I am simply enlarging on and confirming what Skeat has done, bee VII., xxii. ff.
3 Pp. 29-31, 45-51 (Malverne's continuation of Higden's PolycTiromcon, Rolls Series, vol, ;x.),
•22 THE TIWILUS AND CRISEYDE. [CH. I, § 2
arrested and accused ; of all this, Malverne says, the incensed fish- mongers were the cause. In October, 1384, when Brerabre came up for re-election, great precautions were taken to avoid a recurrence of such disturbances as those of his first election.
Usk's account, the vaguely expressed version of a personal enemy of Northampton, perfectly agrees with this.1 After dwelling on how much he has desired the peace of the city, he says he had been enticed into a faction which attempted to abate the evils of extortion, but really meant to make things disagreeable for leading citizens who disapproved of the present misgovernmeut. This faction and its " governour," after he had been put out in a " free eleccion," pre- tended that the latter had been invalid, and raised a great disturb- ance. Usk himself was imprisoned until he should reveal what he knew for the benefit of the commonweal, even if it involved betraying his "owne fere." He justifies himself for this action, but later he was accused of bearing false witness against his master,2 and offered to substantiate his statements by single com- bat.3 The neatness with which Usk's slightly cryptic account corresponds to the facts proves that it cannot have been written before 1384.
But we may go farther, and say that it must have been written later yet, after Chaucer's Legend of Good Women, with which it certainly shows familiarity. The following parallels, especially the first, seem conclusive.4
" Certes, I wot wel, ther shal be mad more scorne and jape of me, that I, so unworthily clothed al-togider in the cloudy cloude of unconninge, wil putten me in prees to speke of love, or els of the causes in that matter, sithen al the grettest clerkes han had y-nongh to don, and (as who sayth) 5 gadered up clene toforn hem, and with their sharpe sythes of conning al mowen, and mad therof grete rekes and noble, f ul of al plentees, to fede me and many another. . . . And al-tliough these noble repers, as good workmen and worthy their hyre, han al drawe and bouude up in the sheves, and mad many shockes, yet have I ensample to gadere the smale crommes. . . . Yet also have I leve of the noble husbande Boe'ce, al-though I be a straunger of conninge, to come after his doctrine,
1 Testament of Love, bk. I., ch. 6, especially 11. 53-6, 76-89, 93-107, 117, 130-50, 188-91.
2 He had been confidential secretary to Northampton. s I. 7, 10; II. 4, 116.
4 T. of L., I., Prol., 11. 94-114. Most of these parallels are pointed out by Skoat.
5 This phrase shows that the passage is a conscious reminiscence ; it will be seen how he plays with the idea (and mixes the metaphor).
CH. I, § 2] THE DATE. 23
and these grete workmen, and glene my handfuls of the shedinge after their handes ; and, if me faile ought of my ful, to encrese my porcion with that I shal drawe by privitees out of the shocke.".^
"Alias ! that I ne had English, ryrne or prose, Suffisant this flour to preyse aright ! But helpeth, ye that han conning and might, Ye lovers, that can make of sentement ; In this cas oghte ye be diligent To forthren me somwhat in my labour, Whether ye ben with the leef or with the flour. For wel I wot, that ye han her-biforn Of making ropen, and lad awey the com ; And I come after, glening here and there, And am ful glad if I may finde an ere Of any goodly word that ye han left " (F, 66-77).
" Hast thou not rad how kincle I was to Paris, Priamus sone of Troy] How Jason me falsed, foral his false behest?" (T. L., I. 2, 91-3 ; Love is speaking to the writer).
" Thou rote of false lovers, duk Jasoun ! . . . Ther other falsen oon, thou falsest two ! " (1368, 1377).
Jason swore to Medea that he
" Ne sholde her never falsen, night ne day." 1
' ' And nere it for comfort of your presence, right here wolde I sterve " (I, 3, 119-120; he is addressing Love). 2
" For, nadde comfort been of hir presence, I had been deed, withouten any defence" (F, 278-9).
We shall later see reason to agree with Professor Lowes that the Prologue of the Legend can hardly have been written before 1386, and to believe that the poem can hardly have been published till 1387. Hence the Testament of Love cannot have been written before that date.
On the other hand, it cannot have been written later than the early part of 1388, for the very good reason that in March of that year Usk was executed.3 The previous year seems to be indicated
1 Line 1640. These seein to be the only cases where falsen is use^d -in L. G. W.
2 Cf. also K. T., 1398. Test, of Low, III. 7, 36-9, affords a parallel to L. G. W., 735-6 ; but it is more closely paralleled in T, G'., 11. 538-9.
a Pointed out by Skeat, VII. xxiii. He was sentenced March 4 ; Mal- verne, p. 169. Yet Skeat "suspects " (p. 473) that Usk copies from Chaucer's Astrolabe, which Skeat himself (and everybody else) dates 1391 (cf. Chaucer, III. 352) ; and assures us (p. 458 ; cf. p. xxvii.) that Usk quotes the C-text of Piers Ploicman, which Skeat dates 1393. In neither case can I see the least internal probability of copying,
24 THE TRniLUS AND CPJSEYDE. [(TH. I, § 2
by the complete silence of the work as to Usk's final imprisonment and peril. At the end of 1387 the Duke of Gloucester and his party succeeded in turning the tables on Richard and his supporters, among whom were Sir Nicholas Brembre and Usk, now sub-sheriff of Middlesex.1 Though we hear nothing of Usk till February, some of his party were accused as early as November 14, 1387. Now Usk has his own affairs much on his mind ; in his Prologue he says, " this book shal be of love " (81-2), yet he has a great deal to say of the bygones of 1384, and seems greatly concerned as to what people think of his conduct in the Northampton affair, and very anxious to vindicate his reputation from the charge of false- hood and treachery. Is it credible that he should utterly ignore this new great danger ? 2 "Working backwards, therefore, as well as forwards, we arrive at 1387 as the date of the Testament of Love.
We find, then, that Chaucer's revised version of the Troilus was known to Usk in 1387.3 If, as Lowes thinks, the first version was not finished till 1385, is not this rather quick work1? So
1 Malverne's continuation of Higden, IX., 106-8, 115-16, 118, 134, 150-1, 169; cf. also Walsingham's Historia, II. 173. The former of course was the ex-mayor, and Chaucer's former colleague at the custom-house. On his execution, cf. also Gower, Cron. Tripert., I. (Macaulay, IV. 318).
2 Skeat thinks (p. xxii.) that he was in prison while he was writing the latter part of the work, because in speaking of the events of 1384 he mentions being for the " firste tyme enprisoned" (II. 4, 103-5) ; but obviously he may have been in prison twice in the first connection, or once later for some un- known reason. His first chapter (e. g. 11. 14-17, 36-48) talks much of prison, but, as Skeat says, this is doubtless because he is imitating the prisoner Boethius, and is meant metaphorically ; for it is here that the allegorical fic- tion begins. In the Prologue, where he speaks directly in his own person, there is not a hint of such a thing. Nor can I see any reason to believe, with Skeat, that he was ever involved with the Lollards. His old associates, whom he has abandoned, were doubtless the Northampton faction, and the meaning of " Margaret " is too vague to be made to imply a recent reconciliation with the Church.
3 The Testament of Love borrows rather extensively also from the House of Fame. In a few passages it suggests Kn. T. , but that is not at all likely to have been seen by Usk, or to have been published before his death. T. L., bk. I. ch. 3, 11. 13-14 suggests Kn. T., 951 ; I. 3, 120 suggests 1398 (but cf. also L. G. W., F, 278, cited above). Other parallels to Kn. T. are paralleled also in T. G. or L. O. W. With T. //., I. 1, 70 (the sentence that follows shows it is meant as a quotation) and III. 1, 137 cf. Kn. T. 3089, T. C., III. 1282, L. G. W. (F) 162 ; and with T. L., III. 7, 50 cf. Kn. T. 1838 and T. C., V. 1433. In a good many other passages, some of which Skeat mentions and some not, T. L. recalls various other scattered parts of the C. T. But after considering every one, I am convinced that there is no evidence of borrowing, nothing like as much as there is in the case of L. G. W., or even Kn. T. Yet Skeat sometimes announces the borrowing without ever considering whether the thing is possible, or whether the borrowing may not have been on Chaucer's part.
CH. I, § 2] , THE DATE. 25
extensive and minute a revision of a poem originally so finished as the Troihis, it seems to me, implies the passage of a number of years. But all this agrees perfectly with the date 1377 for the original completion and 1380 or later for the revision.
A very early date for the Troilus and Criseyde is indicated by Lydgate's manner of speaking of it.1 In the Falls of Princes, in a long list of Chaucer's works which is roughly but rather strikingly chronological,2 the Troilus stands first and is attributed to the poet's youth :
" In youth e he made a translacion Of a boke whiche called is Trophe In Lumbarde tonge, as men may rede and se, •And in our vulgar, long or that ye [he] deyde, Gave it the name of Troylous and Cresseyde."
In the Troy-Book he speaks of Chaucer's
" book of Troylus and Cryseyde Which he made longe or that he deyde." 3
Fifteen years would not be so very long 'before he died, and youth in the fourteenth century certainly did not extend to the middle forties. The probabilities I think are distinctly in favour of the view that Lydgate knew Chaucer personally, and he certainly knew him well by hearsay.4 The list in the Falls of Princes shows very considerable intimacy with Chaucer's literary history, and I see no reason why a good deal of weight should not be attached to Lydgate's testimony. It is striking that he says nothing about the time when any other of Chaucer's -frorks was written. Perhaps the world had not even then got through marvelling at the precocity of such a work from an almost unknown poet.. It is certainly note- worthy that the evidence derived (as we shall see) from Chauc'Br's friend and contemporary Gower, and the direct testimony of his chief admirer and disciple Lydgate. should agree so perfectly on an early date for the Troilus.
1 The point developed here was first made (I believe) in my article in Modern Philology, i. 324, note.
2 See Lounsbury, Studies, i. 419-422 ; Morris' CJiauccr (London, 1891), i. 79. The list, in order, is T.C., Bocthius, Astrolabe, "Ceixand Alcion," B. D., 11.11., P.P., Origen upon the Magdalen, Book of the Lion, A. A., Mars, L. G. W., C. T., Melibeus, Cl. T., Monk's T., small lyrics. The Troilus is also first in the list of Chaucer's works in the certainly genuine Retractations at the end of the Pars. T. : T. C., H. F., L. G. W., B. D., P. F., C. T., Book of the Lion, small lyrics.
3 See Rossctti's edition of the T. C. and the Filostrato (Ch. Soc.), p. x.
4 Cf. Schick, Temple of Glas, xci. f. The Falls of Princes was written about 1430-8, and the Troy-Book about 1412-20 (ibid., cxii.).
26 THE TROILUS AND CRISEYDE. [CH. I, § 2
Up to tliis point it seems to me temperate to say that we have found no reliable evidence in favour of a late date for the T.roilu$, especially for so late a date as 1385 ; and evidence of no little value in favour of .an early, even a very early, date. It will be all clinched if we can be sure that the poem was referred to as early as 1377. The passage in Gower's Mirour de TOmme which seems to mention the Troilus is as follows (11. 5245-56) :
" Au Sompnolent trop fait moleste, Quant matin doit en haulte feste Ou a monster ou a chapelle Venir ; mais ja du riens s'apreste A dieu prier, ainz bass la teste Mettra tout suef sur 1'eschanielle, Et dort, et songe en sa cervelle Qu'il est au bout de la tonelle, U qu'il o'it chanter la geste De Trdylus et de la belle Creseide, et ensi se concelle A dieu d y faire sa requeste."
Koch admits that the reference here is to Chaucer's poem. This, however, Professor Lowes does not do, and with much thoroughness and ingenuity he tries to discover many loopholes of escape from the inferences which I have drawn from the passage.1
First of all, he thinks that the geste de Troylus et de la belle Creseide of which Sompnolent dreams may have been the Filo- strato. But consider that Gower knew no Italian, and was writing for people ignorant of both Italian and Boccaccio ; I do not ask what point there would have been in referring to the latter, but how could it ever have occurred to him, even if he had heard Chaucer speak of the poem, to make in so off-hand a manner a remark so unintelligible ? Is it impertinent to ask whether a modern preacher would rail at his parishioners for staying at home on Sunday to read the last Sherlock Holmes story or the works of a novelist of Paraguay? Obviousness and popularity are necessarily implied in Gower's remark. This and the apparently rather humble station of Sompnolent are what suggest that the poem is in English ; Lowes' suggestion that by the same token Cato and other ancients
1 Public. Mod. Lang. Assoc., xx. 823-33. The reason why Gower's editor, Mr. Alacaulay, did not recognize the allusion to Chaucer's poem is no doubt that by the received chronology it greatly antedated it ; the fact that he did not was my reason for ignoring his remark in my article (cf. Lowes, p. 824). The priority in recognizing the allusion rests with. Hamilton.
CH. I, § 2] THE DATE. 27
quoted in the Mirour, ly the author, should he in English rather surprisingly ignores the point. Moreover, Lowes does not seem to see the world of difference in naturalness between a poet referring, for a particular reason, to a little-known work by himself, as Chaucer and Froissart 1 do, and making a recondite allusion where a familiar one is to be expected. It seems to me the possibility that the reference is to Boccaccio ought to be eliminated.
Lowes' next attempt is to weaken the presumption that the allusion is to an independent poem of some length, rather than to a mere episode, by paralleling it with Froissart's references in La Prison Amoureuse to the " tretti^s," or " livret," " de Pynoteiis et Neptisphele." 2 But there is no parallel whatever ; not only is the latter work one by himself, but the poet as a character in the Prison writes the "livret," and the later references to it are by him and another character in the story. The "gest of Troylus and of the fair Creseyde," it still seems to me, certainly implies an independent work of some length.3
Lowes argues (p. 829) that Sompnolent's meditation should hardly be on so tragic a story as Chaucer's completed version. This seems a little fine-drawn, and at any rate will prove to be an argu- ment rather in favour of my view. No version of the story is known which is any less tragic than Chaucer's. Boccaccio's great innovation and success,4 in which of course Chaucer follows him, is the account of the courtship and happiness of Troilus. Benoit and Guido give no account of the story except reminiscently at the time when the exchange of Briseida for Antenor is arranged; in Benoit the prominent thing (though not treated very seriously) is the grief of Troilus and the fickleness of Briseida. Moreover, if there is any inappropriateness in Gower's allusion, does not this suggest that there was some special reason why he made it 1 And what more natural than that his friend had just been writing the story 1 Gower was not so sensitive an artist, and the allusion is not so much dwelt on, that this inappropriateness was very likely to strike him ; but it does seem that the story was not likely to have occurred to him unless for some special reason. If it cannot be called tactful to represent Chaucer's poem as a favourite with such a person as Sompnolent,
1 All that Froissart does in the Paradys is to mention, among a large num- ber of heroes and heroines, some of those of his own Mtticulor (cf. Engl. xxvi. 330). Lowes refers also to L. G. W., F, 420-1.
2 Scheler, I. 257-78, 286-340.
3 Cf. Lowes, p. 826.
4 Cf. Koerting, B.'s Leben und Werke (Leipzig, 1880), 584-5, 587.
28 THE TROILUS AND CRISEYDK. [oil. I, § 2
this hardly conflicts with the impression we get elsewhere of Gower's personality ; witness his remarks about Chaucer's '.' daies olde " at fifty or so. Perhaps the moral Gower somewhat dis- approved of the Troilus, even though it was dedicated to him. The fact that it was Boccaccio and Chaucer who made the story of Troilus and his lady-love prominent, and its insignificance all over Europe before or apart from their influence, must never be lost sight of, and is of high importance in weighing probabilities in this case. It was probably Boccaccio's relations with " Fiammetta " that led him to select this episode from the Troy story, enormously expand it, and in a measure make its heroine a warning to his own lady.1 While Troilus is very prominent all through Benoit's and Guide's works as a warrior, the mention of his lady and his amour are at very little length, and do not even form a unified episode ; 2 yet Lowes seems (p. 833) to entertain the idea that the geste which
1 Cf. Koerting, B.'s Leben und Werke, p. 585; "W. Hertzberg, Jahrbuch der deutschen Shakspcrc Gesellschaft, vi., 196, 199. Cf. also Schofield, Engl. Lit. from the Conquest to Chaiicer, 291-2.
2 In Benoit, Briseida is "termed 'la pucele' in verse 12977 The
loves of Troilus and Briseida are not described at length, nor the various vicissitudes of them notified : but, now that the lady is to leave Troy, Benoit informs us that she and Troilus are deeply enamoured. . . . Her monologue [as to the final capture of her heart by Diomed] . . . ends at verse 20330 ; and, though the poem goes on to the formidable number of 30108 lines, we hear henceforth no more of her, nor of Diomed as related to her, nor (save in one instance soon afterwards) of Troilus in the character of her deserted and incensed lover. It will thus be perceived that, in the Briseis narrative of Benoit, the more substantial subject-matter is the Briseida-Diomed amour, to which the Bnseida-Troilus amour forms rather the proem ; whereas, in the Chryseis narrative of Boccaccio and Chaucer, the main interest by far centres* in the Cryseyde-Troilus amour, to which the Cryseyde-Diomed amour forms but the sequel, and, even in that connection, is but little developed except in so far as it wedges the iron into the soul of Troilus" (W. M. Rossetti, Troilus and Filostrato, Ch. Soc., p. vi.). In both Benoit and Guido the account of Briseida is scattered in some four or five spots over the whole middle of the work.— In the Lcmd Troy-Book (E. E. T. S., 1902-3, ed. Wiilfing), which was probably written about 1400 (see Engl. Stud., xxix. 3-6, 377-8, 396), but which shows no knowledge of Chaucer's poem (the only, and a very insufficient, ground for dating it earlier), the episode is disposed of in about 60 lines (9060-90, 10365-6, 13437, 13543-64) out of 18664, much more brieflv than in Guido, the source. In the " Gest ffystoriale " of the Destruction of Troy (E. E. T. S., 1869 and 1874, ed. Panton and Donaldson), it occupies about 200 out of over 14,000 (7886-7905, 8026-8181, 8296-8317, 9942-1)959 ; cf. 10306) ; the author refers (8053-4) to Chaucer's poem for more particulars. In the Troy-Book the story is first mentioned when Diomed sends Troilus' horse to the heroine, and in the Gest when she is exchanged for Antenor. In the fourteenth-century Scege of Troye (ed. Wager, 1899), a greatly condensed poem of 1922 11., Troilus is frequently mentioned, but his lady and his amour never. It is the Seege and the Gest that Miss Kempe refers to in her rather vague statement in Engl. Stud., xxix. 3. There is not a single poem in English which mentions the love-story, and which can plausibly be dated before Chaucer's Troilus. i
CH. i, § 2] THE DATE. 29
Sompnolent dreams he hears sung l may have been a few scattered passages in Guide's Latin prose !
Furthermore, in works other than those which tell their story, though Troilus is not infrequently mentioned as a brave warrior, I find only one reference to him as a lover (by Froissart), and no reference at all to the heroine, earlier than Chaucer's Troilus, in any language.2 In the fifteenth century I find several references in French to the heroine and the amour; when we find that the Troilus and JBriseida,3 a French prose translation of the Filostrato, was written at the very end of the fourteenth or early in the fifteenth century,4 is not the inference obvious that the rise of the love-story to prominence was largely due to this 1 5 And when we find
1 Cf. Chaucer's address to his poem (V. 1797), "red wher-so thou be, or ellee songe " ; and (II. 56) "As I shal singe, on Mayes day the thridde."
2 Therefore to Lowes' question (p. 828, note), "Supposing Chaucer's Troilus never to have existed, would such a reference as Gower's, on the basis of known relations of the other versions of the story, have seriously puzzled any one for a moment ? " — to this question I answer Yes.
3 Or " Creseide " or " Brisaide."
4 See Moland et d'Hericault, Nouvelles Francoises, p. ci. There are six MSS. in the Bibliotheque Nationale alone (ibid., p. cxxxiv. ). Benoit's work was written about 1165, and Guide's in 1287.
8 The following are the only references I find to Troilus and Criseyde outside the works which tell their story. I. Troilus as a warrior is mentioned : —
I. In Partenopeiis de Blois, I. p. 6 (ed. Crapelet, twelfth century ; he is barely mentioned among five Trojan knights ; Hector is dwelt on). 2. In JFloriant et, Florete, p. 32 (ed. Francisque-Michel ; thirteenth century ; barely mentioned).
3. In Anse'is de Cartage, 1. 1653 (Stuttg. Lit. Ver., thirteenth century). 4. In Chroniqtic Rimte de Ph. Mouskes, I. 289 (ed. de Reiffenberg ; A.l>. 1243 ; Troilus with others used as a simile for bravery). 5. In 1249 the German Albertus Stadensis wrote a Latin poem in distichs, in the proem of which he says, "Liber est Troilus ob Troica bella vocatus." It does not, however, deal with Troilus particularly, and apparently never mentions his love affair, but is a mere para- phrase of Dares (see G. Koerting, Boccaccio's Leben u. Werke, p. 589 ; W. Hertzberg, Jahrbuch d. deut. Shaks. Gesellsch., vi. 181). 6. In Escanor,
II. 15698-9 (ed. Michelant : about 1285 ; he is barely mentioned). 7. By Deschamps, IX. 91 (Soc. Anc. Textes ; about 1381 ; barely mentioned among ancient warriors). 8. By Georges Chastellain, VII. 424 (ed. Kervyn de Lettenhove : C. died 1475 ; mentioned among many others). I may add Malory, Murte d' Arthur, XX. 17 ; bare mention among ancient heroes, prob- ably due to Malory himself; the numerous other fifteenth-century English references it is needless to collect. II. Troilus or Criseyde, or both, are men- tioned as lovers by:— 1. Froissart, I. 29 (ed. Scheler : before 1370 ; Troilus is barely mentioned, but heads a list of lovers ; cf. Lowes, p. 825, who makes too much of this single instance). 2. Jean le Seneschal, p. 203 (S. A. T. F. : about 1389 ; " Troyluz," lover of " Brisayda "). 3. Charles d'Orleans, p. 307 (ed. Champollion-Figeac ; lived 1391-1465, and often alludes to the Troy-story ; three lines on Troilus) ; p. 120 (not 126, as Dernedde says ; speaks of the beauty of " Criseis, de Yseud et Elaine " ; the editor says, p. 427, " lisez Briseis ").
4. Alain Chartier, p. 734 (Paris, 1617 ; lived about 1392-1429 ; Troilus barely mentioned ; " Briseyda," who broke faith with him, appears among a number of faithless ones). >5. Rene of Anjpu, III. 111-112 (ed. Hawke ; born 1408 ; T; is mentioned between Paris and Dionied, among many lovers ; loved " Grisadc "
30 THE TROILUS AND CRISEYDE. [cH. I, § 2
that Gower, after this first reference in the Mirour, makes many other such, in the Vox Clamantia, in a lalade and in the Confessio Amantis, never mentions Troilus but as a lover, and always spells the heroine's name with a C, is not the inference still more justifiable that the prominence of the story with Gower, as with fifteenth-cen- tury English Avriters, was due to Chaucer 1 The most interesting case in Gower is a reminiscence of the passage in the Mirour (C. A., IV. 2794-7) ; when Genius examines the Lover as to the sin of " Sompnolence," he proves himself innocent by showing his constant readiness to please his lady :
"Or elles that hir list comaunde To rede and here of Troilus, Riht as sche wole or so or thus, I am al redi to consents."
The reference here, of course, is to Chaucer's Troilus, which there cannot be a doubt that Gower knew well when he wrote the Confessio. Yet we are asked to believe that the precisely similar reference in the Mirour is to some poem unknown to Gower's readers (or else to us), or else to a few scattered bits lost in a long poem, or (worse yet) in a Latin prose work.1
I hardly think it can be said, then, that the argument " that Gower's reference is to Chaucer's Troilus, rests in the last analysis on a single letter, the initial C of the heroine's name" (Lowes, p. 826).
or " Grisayde "). Most of the references iu this note are derived from
Robert Dernedde's Ueber die den altfranz. Didvtern bekannten epischcn Stoffe atis dem Altertum (Erlangen, 1887), pp. 122-3 ; 1 owe the reference to Dr. G. L. Hamilton. No doubt the list could be extended, but after some search I find no more. Neither T. nor C. is mentioned, e. g. , in Petrarch's Trionfi (ed. Appel), where many such are ; she is never mentioned by Deschamps nor (so far as I can find) by Froissart, and he only once by each (as noted above). No other references are to be found in the ninety volumes of the Soc. des Anc. Tcxtes, or in Lauglois' Tabk des Noms Proprcs. I find Criseyde mentioned in no manner anywhere else ; but of course I have not collected fifteenth-century English references.
1 Troilus and Criseyde are frequently mentioned by Gower. The earliest case appears to be that in M. 0. The next, pointed out to me by Dr. Hamilton, is in Vox Cl. (soon after 1383), VI. 1325-8, where the faithful T. and Medea are paired off against the fickle Jason and " Crisaida." In one of the French balades (XX. 20 ; probably late) he speaks of T. as supplanted by Diomed hi the love "du fille au Calcas." In C. A. (finished 1390) among examples of supplantation he quotes the case of Agamemnon, Achilles, and "that swete wiht" " Brexeida," and then directly " Criseida," Troilus and Diomed (II. 2451-8); twice again, similarly, of "Criseide" and the other two (V. 7597-7602 ; VIII. 2531-5 ; the story of "Criseide doubter of Crisis," is told in V. 6433-75). It is worthy of remark that there is no significant change in his manner of mentioning the level's, which suggests that he had had no accession to his information since the first reference.
CH. I, § 2] TttE DATE. Si
But this is still a strong argument. We must not assume, it is true, that Chaucer was the innovator in this spelling; not only in one fourteenth- and two fifteenth-century MSS. of Guido does her name appear as Criseida or Griseida,1 biit in one fifteenth- century MS. C has replaced G even in the Filostrato.^ At the same time, it is not unlikely that Chaucer did substitute the less unfamiliar, and perhaps more agreeable, G for G. But the main point is that the form with G, and therefore that with C, is due only to Boccaccio ; without him, Troilns' mistress would every- where have been called Z?riseida,3. which seems to have been, outside Chaucer's and Gower's poems, the universal form in England in the fourteenth century.4 The spelling of Gower and Chaucer alike is due to the influence of Boccaccio. Through which of them is it more likely to have entered England ? 5 But it is not only the initial in Gower's spelling which indicates Chaucer's influence. Gower's form is French, Avith a final -e, Creseide. The name appears in Chaucer MSS. under various French forms, among which, though Criseyde is perhaps the commonest, Gower's form is often found I find the final -e nowhere else except in those who write under Chaucer's influence and two or so other post-Chaucerian writers.6
1 See Morf, in Romania, xxi. 101, note 1. See also G. L. Hamilton, Chaucer and Guido, 134-5.
2 Cryseida is the form in a MS. of the Marquis cle Santillane's library ; see Bibliotheque dc I'Ecole des Haides Etiides, Fascicule .153, p. 328 (pointed out to me by Dr. Hamilton). So in old printed editions (Hertzberg, Jahrb. d. deut, Sh. Ges., vi. 197); and cf. Koerting, Boccaccio, 569. The form with a C is the commonest, in the French translation (cf. Lowes, p. 827) ; otherwise with a B.
• 3 This is true, so far as I know, of all the documents. But the Italian Armannino, who ignores the love-story, in speaking of the Homeric " Brisseida " and " Crisseida," says that, according to some, the latter was the daughter of Calchas (see Gorra, Testi Inediti, p. 555).
4 To balance the substitution by a reader, under Chaucer's influence, of C for B in two passages in the Laud Troy-Book (cf. Lowes, 828), the Gest Hystorialc, which directly mentions Chaucer's Troilus, preserves only the form with B as I pointed out in my article (p. 323, note ; this will correct Miss Kempe, Enijl. Stud., xxix. 5) ; and the only MS. of the Filostrato in the British Museum (Addit. 21246, early fifteenth century) has Briscyda three times at the beginning, though elsewhere it always has Griseyda.
5 Lowes' suggestion (828-9) that the G may have been substituted by Gower's scribe after the poem was written of course cannot be disproved ; but, especially since the work was not a popular and much-copied one, and this MS. (as Lowes admits) was almost certainly ' ' written under the direction of the author," the suggestion can be allowed little weight. Lowes refers to my " tacit assumption" that the reading is Gower's, not the scribe's ; if we did not make the tacit assumption that the MSS. represent the author's words, where should we be in the study of mediaeval literature 1
6 The fifteenth-century, or post- Chaucerian fourteenth-century, French version of the Filostrato has commonly Creseide; probably influenced by
32 THE TROILtJS AND CRISJSYDE. fcH. I, § 2
Iii English, Germaii, Norse, French and Latin, in Benoit and Guido alike, the regular form is Brucida or Brisaida.1 Chaucer seems to have been the first to use the -e, of course for the sake of the rhyme. The combination of the initial C with the final e, apart from Chaucer and Gower and those who owe it directly to them, seems to be found only in some MSS. of the post-Chaucerian French translation of the Filostrato. Are we to look upon the occur- rence of one and the same very unusual form in the works of two friends within a very few years as a coincidence ]
So all the evidence looks in the same direction. "We have seen that if the reference is not to Chaucer's poem the spelling with
C e is surprising ; and that the occurrence of the reference at
all is more than surprising. Lowes must battle against the co- incidence of the two surprises. I must say that the more I investi- gate Trolls literature the more I am struck by the improbability that Gower's reference is to any work but Chaucer's.
But now Professor Lowes, who has as many holes to start to as the Wife of Bath's mouse, suggests that, even if Gower's reference is to Chaucer's poem, it may not have been made as early as 1376 (the extension of the limit to 1377 will not matter here).- He appeals (p. 830) to that forlorn hope, a possible interpolation in the Mirour. For this proposal I cannot see the slightest justification, or the slightest reason to believe that there are interpolations anywhere in the Mirour. Lowes' argument that the passage which mentions the Troilus looks like an interpolation is fallacious. Gower says earlier (5179-84), it is true, that Sompnolent will not get up in the morning, and leaves the labour of prayer to the nun and the friar ; yet he says here that he goes to sleep at the morning service in church. But Gower says expressly that this is when he has to go, on a high festival (5245-8). 3 Therefore there is no means of escape
this, the fifteenth-century Rene of Aujou has Grisndc, Grisaydc ; the Gcst Hystoriale, which appeals to Chaucer, has (rarely) Bresaide, Brci-saidc (usually — said) ; a late fifteenth-century MS. of an Old French version of Guido has brisade (Brit. Mus., Royal 16. F. ix. ). 1 find no other cases..
1 With the variants Breseida, Breyscyda, Brisayda, Briscida, Brisryiln, Brixeida, Brixeyda, Bryseida, Prixaida (quaintly, in a German version), Brciseida, Breiseidd, Breisida (these three in a Norse version). I find no other forms (except for Brisade in one late version, as noted above) after a thorough search in the British Museum, including seven early printed editions, twelve MSS. and five MS. translations of Guido, and one MS. and two modern editions of Benoit. The occasional occurrence of Briseide as a genitive or dative form in Guide's Latin seems to have misled Hertzberg (p. 210). A MS. of Guido in the Harvard Library has Briseida, Briscyda.
2 I refer again to Appendix A, pp. 220-5, on the date of Gower's Mirour. a For similar passages cf. 5557-68, 5617-28.
CH. I, § 2] THE DATE. 33
from the conclusion that the reference to the Troilus in the Mirour must have been written not later than soon after the death of Edward III. Altogether, therefore, Professor Lowes' whole long and ingenious argument seems like piling very numerous feathers into one scale to outweigh a lump of lead in the other. It is seldom in literary investigations that we have stronger evidence than we have here for the view that Chaucer's Troilus was mentioned not later than 1377.
After this exhaustive study of the evidence, the conclusion I reach, then, is that the Troilus and Cnseyde was written at the very beginning of the period when Chaucer was under Italian influence. And after all, why not ? He had returned from Italy by May 23, 1373,1 after an absence of six months, during which he doubtless read much Italian, including very likely the Filostrato and Teseide.2 On his return, having once learned Italian, is it not natural that he should plunge with zeal into the study of Italian literature 1 Not till over a year later, June 8, 1374,3 was he appointed Comptroller of Customs, and during the interim, adorned by several benefactions and payments from the king, he may probably have enjoyed much well-earned leisure at court,4 with his books and pen. After his re- sponsible mission to Italy, he would surely not be worked very hard as Esquire of the King's Chamber. In the office of Comptroller there is not the least reason to believe that Chaucer was overworked ; I have showed elsewhere, and Koch admits, that his supposed lamentations in the House of Fame are not such at all.5 As one turns over the pages of the Life Records for this period he sees indications that Chaucer's life was financially comfortable, and broken in upon by no public commissions until 1377 and the very end of 1376. With perhaps a year of leisure, followed by two years and a half or so of routine work, and another year only partly spent abroad, is there anything unreasonable in supposing the Troilus to have been produced at this time 1 °
1 Life Records, pp. 183-4.
2 Mr. Karl Young (Mod. PhiJol., iv. 169-77) makes out a good case for Chaucer's showing familiarity also with Boccaccio's Filocolo in the Troilus, especially book III.
3 Life Records, p. 191. 4 Cf. Life Records, p. 185.
5 Mod. Phil., i. 326-7 ; Engl. Stud., xxxvi. 142.
6 Koch believes the best conclusion to be that Chaucer was engaged on the T. G. in 1376, but did not finish it till several years later (Engl. Stud., xxxvi. 140-1). Grower's reference implies somewhat widespread familiarity with it. It is not quite impossible, perhaps, that familiarity might have
DEV. CH. D
34 CERTAIN MINOR POEMS. [CH. I, § 2
Chaucer chronology so haugs together that I can hardly avoid briefly discussing the position of the translation of Boethius. For previous opinion I may refer to my introduction. It is hardly necessary to say that it must have been written before 1387, since it is constantly used in Usk's Testament of Love, which we have found reason for dating in that year. No other evidence on the date has ever been found except its relation to Chaucer's original works, in which the philosopher is frequently borrowed from,1 especially and remarkably in Troilus. All ihe critics have there- fore put it immediately before the Troilus^ sometimes overlapping, a position which, since they all assign a rather late date to the latter, means that the date for the Boethius has ranged from 1373 to 1383, or even later, but has always been later than the first Italian journey. When we find that from the date of the Troilus to the end of Chaucer's life Boethius' views were never far from his thoughts, and influenced all his speculations and even his turns of phrase, and when we find that in the Book of the Duchess the influence is all second-hand, through Jean de Meun, we seem justified in conclud- ing that (in 1369) he was not yet familiar with the Eoman philo- sopher, and certainly that he had not translated his work. The extraordinary familiarity with it shown in the Troilus justifies the belief that, when he wrote that poem, he had already studied and translated it. But of course there is no reason in the world why this should not have been done before he went to Italy ; and since the Troilus itself is enough to fill the succeeding years, it very probably was. We may therefore with some confidence date the translation of Boethius about 1370-2.
CHAPTER II.
CERTAIN MINOR POEMS.
§ 1. The House of Fame.
I CAN hardly avoid briefly discussing the date of the House of Fame, since it has a necessary bearing on that of the Troilus. But
been gained for the poem in certain circles by author's readings and the like, before it was wholly finished. But Gower's remark is infinitely more naturally interpreted as implying that when it was made the completed poem was spreading abroad and exciting every one's interest. And the other arguments for an early date are not affected by this bare possibility. * Skeat very conveniently gives the cases : IJ. xxviii,-xxxvi,
OH. II, § 1] THE HOUSE OF FAME. 35
I have little that is new or certain to offer, and prefer to leave a detailed discussion of the problems connected with it to one or two other writers at present engaged upon it. Three much-debated points, in particular, I shall dismiss in a few words. All attempts to read a subtle personal or general allegory into the poem seem to me worse than futile. Subjective allegory is " wholly alien from Chaucer's realistic, unspecuiative genius " (to quote Professor Francis T. Palgrave).1 Renascence allegory is sometimes obscure, inconsistent and ambiguous, because frequently a side-issue and used (as Professor Courthorpe says) for purposes of decoration ; mediaeval allegory is clear and intellectually consistent. An alle- gory which does not fairly well explain itself I think had best be ignored ; and this no one can maintain the House of Fame does. As to the relations between the House of Fame and the Divine Comedy, it seems to me that while the relation of this poem to Dante is far closer than that of any other,2 it is entirely improper to call it an imitation of the Divine Comedy, and unlikely that Chaucer foresaw it in that light. Therefore Chaucer's aspiration, at the end of the Troilus, " to make in som comedie" (V. 1788), there is no reason to take as alluding to Dante's title, nor any reason to see here an allusion to the House of Fame. If an absolutely sufficient explanation of the " comedie " passage is the desire for a cheerful subject,3 what right have we to read anything else into it ] Therefore^ though the older inter- pretation would make in favour of my view that the House of Fame was written soon after the Troilus, all these points mentioned above may be rejected as chronological arguments.
My chief reason, of course, for dating Fame after Troilus is that the early date which I have assigned the latter makes it quite impossible to put between it and Chaucer's first Italian journey a poem so long and showing such familiarity with Dante as the House of Fame. There are other reasons also, and this order of things is the orthodox one, but I must first discuss the arguments
1 Nineteenth Century, xxiv. 345.
2 Besides the parallels pointed out by Rambeau (Engl. Stud., iii. 209-268) andCino Chiarini (Di una imitazione inglese della D. C., Bari, 1902; reviewed by F. N. Robinson, Journ. ofCompar. Lit., N. Y.,i. 292-7), 1063-81 may be suggested by the apparition, in the Paradiso, of the souls in the appropriate celestial spheres, though they are actually present in the Empyrean.
3 Cf. Lydgate's way of speaking of Chaucer's poems — "My mayster Chaucer with his fressh commedies" (Skeat, III. 431) ; cf. also my article in Modern Philology, i. 318, which seems slightly to understate the matter ; and Lowes, Publ. Mod. Lang. A$soc.t xx. 855,
36 CERTAIN MINOR POEMS. [CH. II, § 1
of Professor Lowes,1 the only writer who has defended the contrary opinion. His view is required by the fact that he believes the House, of Fame to have preceded the Legend of Ariadne, the latter to have preceded the Palamon, and that to have preceded the Ti-uilns. I have already tried to refute the last point, and shall later try to do the same for the second. It only remains to mention his auxiliary arguments.
He quotes the suggestion that H. F., 1391-2 looks as if it had preceded T. G., IV. 659-62. In attributing " partriches winges" to Fame, Chaucer clearly mistakes Virgil's "pernicibus alis" for " perdicibus " ; while in the Troilus Fame flies through Troy, more properly, with " preste winges." But Lowes seems to me to deprive the argument of all weight by showing that the latter passage closely translates the Filostrato, which has "prestissim" ale." If Chaucer thought Virgil wrote "perdicibus " or " perdicum," in quoting the passage why should he think of " pernicibus " or of "prestissime " or of "preste"] Nor will it do, especially in considering mediaeval literature, to assume that the incorrect impression always precedes the correct one.
Lowes seems also to believe that the fact that Chaucer here uses the 8-syllable couplet rather points to a time before the Troilm? though he fully grants the skill with which Chaucer uses it. His argument is not quite clear; but, aside from the fact that he himself believes Chaucer to have returned to the 7-line stanza from the 10-sy liable couplet, and that a somewhat similar return here is not surprising, he himself has also, with great justice, dwelt on the impropriety of drawing hard and fast lines as regards Chaucer's use of metres. I believe that many scholars, in their zeal for chrono- logical evidence, have been too much inclined to make Chaucer's style vary rather with epoch than with subject. For this humorous and almost jaunty tour de force, what verse could1 be more appro- priate than the 8-syllable couplet1! Why did not Chaucer write Sir Thopas in heroic couplets 1s
1 Pull. Mod. Lang. Assac., xx. 819, 854-60.
2 Professor Heath holds this view for the first part of the poem (Globe Chaucer, xliii. ).
3 One other argument for putting Fame before Troilus may be anticipated. The eagle in 630-40 commends Chaucer for his faithfulness to love :
"ever-mo of love endytest, In honour of him and preysinges, And in his folkes furtheringes, And in hir matere al devysest, And noght him nor his folk despysest. "
CH. H, § 1] fHE HOUSE OF FAME. 37
But to turn now to arguments in favour of putting the House of Fame after the Trbilus and Criseyde, One reason, it seems to me, as ten Brink points out, is the appearance of " LolliusJ" (1468) among the historians of the Trojans. It seems idle to discuss further Chaucer's reason for attributing Boccaccio's works to this shadow of a name, but twice in the Troilus he does so.1 If he invented an author named Lollius as a mere piece of mystification, he surely did not do so before he wrote the Troilus; and in any case it is certain that to his readers the reference would be quite unintelligible unless the Troilus was known to them. Another reason advanced by ten Brink2 is that certain parallel passages in the two poems suggest that the House of Fame followed the Troilus. H. F. 1-65 is surely a reminiscence of T. O., V- 358-85, rather than vice versa; Pandarus' discourse on dreams grows out of the situation in the Troilus, and is partly drawn from the Filostrato, while the passage in the House of Fame is a mere prelude, and looks greatly like an expansion of the other.3
As to arguments based on the style and subject of the House of Fame, they seem to me void of collusiveness; but they certainly do not make particularly in favour of an early date for the poem. Indeed, Lowes does not believe that they do, but rather devotes himself to dis- proving the contrary. The House of Fame may be, it is true, a poor piece of work, and show a lamentable falling off in design, substance
The Troilus is represented in the Legend of Good Women as utterly cynical and anti-amorous. Then does not this passage sound as if Chaucer had just been writing more conventional love-poetry rather than such a poem as this ? The answer to this possible objection is that the language of the God of Love in L. G. W. is greatly exaggerated for the purposes of the poem ; probably, as I shall try to show later, because Chaucer had been reproached for having represented the female sex in an unfavourable light. If any one wishes to see how far from cynicism Chaucer's Troilus really is, let him compare it with the Shaksperian treatment of the same theme. The faithfulness and sufferings of its hero are rather dwelt on than the pathetic inconstancy of its heroine ; and fully four-fifths of the poem are as amorous as possible. One little point more : at first sight one might expect a mention of Troilus and Criseyde among the seven faithless couples in 388-426 (cf. P. F., 291 ; Against Women Inconstant, 16) ; but among these it is always the man who is faithless.
1 I. 394 ; V. 1653. In the former case, of course, the fact that he really quotes a sonnet of Petrarch does not signify ; unless it strengthens the view that " Lollius " is a hoax.
2 Studien, p. 121.
3 It is natural that Pandarus should dwell only on ill causes of dreams ; in H. F. Chaucer does the same, though the dream which follows is pleasant. Cf. especially H. F. 30 with T. C. V. 360 (directly from Boccaccio), 15-17 with 362 and 371, 41-2 with 365-8, 21-22 with 369-70, 25 with 370.
38 CERTAIN MINOR POEMS. [CH. II, § 1
and style from the Troilus. Yet it lias one quality which indicates great maturity, especially in a mediaeval poet — freedom. It is not merely that as regards source it is among the most independently- imagined of Chaucer's narrative poems. It shows a general free- dom of self-expression, of informal and roguish humour, combined with remarkable composure and poise ; in this poem he has left the French house of bondage far behind. This seems to me more sug- gestive of a late date than the want of symmetry and method in the poem suggests an early one. It was in the Troilus, it seems to me, that Chaucer became emancipated. Boccaccio both stimulated his growth and was (if I may say so) the cocoon that protected it. The House of Fame followed his emancipation. My conclusion is that the necessity which I have found of following the usual view in putting the House of Fame after the Troilus is an easy, plausible, and well-supported necessity ; and that the auxiliary arguments in favour of this view are much more convincing than those on the opposite side.
On the exact date there is no very reliable evidence. Ten Brink suggests1 that since it is Jupiter that sends the eagle to Chaucer, the adventure may have taken place on a dies Jovw, a Thursday, which, since it was December 10 (11. 63, 111), indicates 1383. The v year 1377 would do as well, by the way. But from whom should an eagle come if not from Jupiter, and who is more likely to send a dream than the father of gods and men 1 And is not this a case where such a subtlety would be lost if not announced? The microscopic symbolism of Dante must not be attributed to Chaucer ; and even Dante, where particularly subtle, commonly gives a hint.
The extreme limits are June, 1374, and February, 1385, as ten Brink points out,2 the dates when Chaucer received his custom- house appointment and was relieved of his duties there by the appointment of a deputy ; though he cannot be said to complain of his clerical duties, it is clear that his life is one of routine office- work (652-60).3
There is a somewhat striking parallel between the House of Fame \ and Gower's Mir our de VOmme. Fame, according to Chaucer, is quite unaccountable in giving or withholding her favours ,
] Studif.n, pp. 150-1. 2 Ibid., 114.
3 If we needed any proof that the work was published by 1387, we should have it in the fact that it is extensively quoted in Thomas Usk's Testament of Love, written doubtless in that year. Cf. p. 24 above, and Skeat, VII. xx., xxvi. f. and notes. I add T. L. I. 6, 198 = H. F. 2088-2109 ; II. 3, 11 1-15 = H.F. 269-72.
CH. II, § 1] THE HOUSE OF FAME. 39
" Right as hir suster, dame Fortune, Is wont to serven in comune " (1547-8).
She sends for Eolus to be her trumpeter :
" And bid him bringe his clarioun, That is ful dyvers of his soun, And hit is cleped Clere Laude, With which he wont is to heraude Hem that me list y-preised be : And also bid him how that he Bringe his other clarioun, That highte Sclaundre in every toun, With which he wont is to diffame Hem that me list, and do hem shame " (1573-82).
Gower, after discoursing on emperors, Alexander especially, apos- trophises Fortune, and continues :
" Fortune, tu as deux ancelles Pour toy servir, si volent celles Plus q' arondelle vole au vent, Si portont de ta court novelles ; Mais s'au jour d'uy nous portent belles, Demein les changont laidement : L'une est que vole au noble gent, C'est Renomee que bell et gent D'onour les conte les favelles, Mais 1'autre un poy plus asprement Se vole, et ad noun proprement Desfame, plaine de querelles.
Cist duy par tout u sont volant Chascune entour son coll pendant Porte un grant corn, dont ton message Par les paiis s'en vont cornant. Mais entrechange nepourqant Sovent faisont de leur cornage, Car Renome, q'ier vassellage Cornoit, huy change son langage, Et d'autre corn s'en vait sufflant, Q'est de misere et de hontage : - Sique de toy puet estre sage Sur terre nul qui soit vivant" (22129-52).
No parallels to either passage have ever been pointed out, and the case for borrowing between the two poets is always especially strong because of their mutual relations. We know that Chaucer used the Mirour in the Prologue to the Canterbury Tales. As to which was the borrower here, the probability seems rather strong
40 CERTAIN MINOR POEMS. [CH. II, § 1
that it was Gower. Like Chaucer, he dwells on the capriciousness of the trumpeters, and departs from his original scheme and re- lapses into Chaucer's by forgetting all about Desfame and making Eenomee use both horns ; and by making the transaction consist not in conferring a good or ill lot in life, but in proclaiming good or ill moral fame. Gower's last six lines or so distinctly suggest that he was the borrower.
In Appendix A the date of the Mirour is discussed. This passage I should date about 1379. Is there not, then, some evidence here that the last part of House of Fame was somewhat known by 1379 1 In that case it and the whole middle part of the Mirour were in hand during the same time, and the House of Fame was finished about two years after the Troilus, a highly probable conclusion. This evidence should not be insisted on, but tenta- tively there is no objection to about 1379 as the date when the House of Fame was completed.1
1 We may notice then that there is no objection to believing it was begun directly after the Troilus was finished, and (if we wish) that the date 1377 for its beginning is as well indicated by ten Brink's astrological method as 1383. H. F. 130-9 seems to be the original of K. T., 1955-66, which is not from the Te-seide, as Skeat erroneously says in his note on the former pas- sage ; the source of the idea he shows may be in Albricus Philosophus, Koeppel thinks (Anglia, xiv. 233-8) that the Parliament of Fowls shows signs of borrowing from Boccaccio's Amoroso, Visione. This may be doubted, but the House of Fame certainly does, as is shown by him and also by Pro- fessor C. G. Child (Mod. Lang. Notes, x. 379-84). If one grants the influ- ence on P. F., since these are the only poems which do show this influence, a date not far from 1381 is suggested for H. F. Professor Heath believes he finds evidence that book III. was written some years after I. and II. (Globe Chaucer, xliii., f. ; cf. Lowes, P. M. L. A., xx. 860, n., who disposes of his argument as to a change of tone. ) Some of his arguments seem to me with- out value, but there are certainly some suggestions of an important change of plan during the composition of the work. In the invocation at the begin- ning of book III. Chaucer seems to be more conscious than before, as Heath points out, of the informality and sketchiness of his verse. But besides this, in the first part of the poem, Fame has the Virgilian sense of Humour, and what Qhaucer is to learn at her house is wholly about love (673-99, 701-6, 713-24, 782-6, 817-21, 848-52, 879-83, 1025-83). At the very beginning of book III. Fame acquires the. mediaeval and modern sense of Renown (1136-9) ; that she is " goddesse of renoun and of fame" is ex- pressed in 1312-3, 1320-3, 1405-6, and elsewhere, and this seems to be the point of introducing the harpers, trumpeters, and minstrels (1197-1258). The idea of Rumour does not recur till Chaucer has left the house of Fame and come to the revolving twig-house (1920 ff.) ; here, in a somewhat ex jwst facto way, Chaucer makes her a goddess of Fame in the sense of Rumour (2110 ff.) ; love, which was to have been the subject of Chaucer's news, is mentioned very casually in the house of Fame (1739-62) ; he does not seem to care very much about it (1889), though there is promise of some love-tales when the poem breaks off (2143). All this might be held to indicate a lapse of time between books II. and III. — A few analogues to one of the folk-lore elements in H. F. may be pointed out — the revolving-house. It is found in
CH. II, § 2] THE PARLIAMENT OF FOWLS. 41
§ 2. The Parliament of Fowls.
The greatest service which Dr. John Koch ever performed for Chaucer chronology was the identification, in 1877, of the eagles in the Parliament of Fowls with Anne of Bohemia and her various suitors.1 The date 1381-2 at which he arrived has been accepted, I believe, by every one who has written since that date, except Professor Hales,2 and has proved to be one of the two pivotal and unshaken dates in the chronology of all Chaucer's poems. It may be worth while, however, to give a few more particulars.
As early as the spring of 1377, when Bichard was ten years old, at the wish of Edward III. there were conferences between French and English commissioners, of whom Chaucer himself was one, regarding a marriage between the heir-apparent and Princess Marie of France.3 After the death of his grandfather the young king's guardians continued the matrimonial negotiations. Early in 1378 Chaucer was again a member of a commission for the same purpose.4 The negotiations, however, fell through. Early in 1379, Bernabo Visconti, Lord of Milan, anxious to ally himself by marriage with royal houses, sent to Bichard II. proposing a marriage between him and his own daughter Caterina ; 3 and may we not conjecture, by the way, that this was not unconnected with the visit to his court a few months before, in 1378, of Chaucer,
the O.F. romance of Perceval le Gallois (tr. by S. Evans, The High History of the Holy Graal, J. M. Dent, 1898 ; vol. ii. p. 21) ; in two Welsh works and one Latin (John Rhys, Stiidies in the Arthurian Legend, Oxford, 1891 ; pp. 301, 302, 326) ; and in the eleventh-century Voyayc de Charlemagne (ed. E. Koschwitz, 1880, in Foerster's Altfranz. Bibl., vol. ii.; 11. 354-91). In at least three of these cases, including the last, the turning-house is adorned with images which blow real trumpets, just as the house of Fame is in Chaucer (1193 ff.).
1 See Engl. Stud., i. 287-9 ; English version (enlarged) inEssays on Chaucer (Ch. Soc.), pp. 400-9.
~2 Diet. Nat. Biogr., x. 164 ; The Bibliographer, i. 37-9. Professor Hales argues that the poem is too poor to be so late. Few persons, I think, will agree with him in either point. He also argues, but unconvincingly, that Chaucer may have known Italian before his first journey to Italy.
3 Froissart, in Life Records (Ch. Soc.), pp. 203-4.
4 Life Records, pp. xxviii., 219, 230.
5 On March 18, 1379, Richard appointed a commission to treat on the subject (Rymer's Foedera, London, 1709 ; vol. vii., 213). See also Thomas Walsingham, II. 46, and Theodor Lindner, Geschichte des deutscJien Reiches unter K. Wcnzel (Braunschweig, 1875), i. 117 ; C. Hbfler, Anna von Luxem- burg, in Denkschriften d. Wiener Akad., Phil.-Hist. Cl., xx. 127-8. The latter long and valuable essay (pp. 89-240) deals rather with Richard II., Wyclif and general contemporary history than particularly with Queen Anne.
42 CERTAIN MINOR POEMS. [CH. II, § 2
who had so recently been a matrimonial commissioner, though his ostensible purpose this time was military1?1 The advances of the Visconti were finally rejected. The first intercourse between Kichard and the King of the Eomans, Wenceslas, was on May 20, 1379, when the latter took the initiative in treating with the former as to the recognition of Urban VI. as the legitimate pope.2 About the same time, according to Froissart,3 there was much deliberation about Richard's marriage. In June,
1380, commissioners were appointed to treat of a marriage between Richard and Anne, Wenceslas' sister, and December 20 Richard announced that he had chosen her.4 At Epiphany, 1381, pleni- potentiaries met in Flanders to arrange the conditions ; January 23,
1381, Anne in her own person appointed ambassadors to treat; and early in May it was agreed that she should be received by the English envoys on Michaelmas next.6 She actually landed at Dover about December 18, 1381, and was received with great enthusiasm; the marriage took place January 14, 1382.6
It was peculiarly natural that poetic notice should have been taken of all this by Chaucer, who had so recently been concerned in at least one of the earlier matrimonial negotiations of Richard. It is also interesting to observe how he manipulates the material. Anne's two tentative childish betrothals, which occurred when she was five and seven years old, are brought down and made contemporary with that to Richard, for literary reasons and in order to compliment him. Chaucer's courtiership and tact go further, in representing the affair as purely a matter of love on the part of the suitors, and the choice as purely on the part of the formel eagle ; in reality, we can hardly doubt that on the German side the choice was mainly on the part of Anne's king-brother and empress-mother, and that the chief exercise of choice was on the English side. The Parliament
1 Life Records, p. 218.
2 Hbfler, p. 127.
3 Chronicles, tr. by Johnes (N.Y., n.d.), p. 258 ; vol. ii., ch. 43.
4 See Lindner, I. 118 ; Hbfler, 128-9. She was born May 11, 1366, so was a few months older than Richard. According to Hbfler, she had already been twice betrothed ; once to Duke William of Baiern-Holland, in 1371 ; and in 1373 to a son of the Landgraf Frederick of Thiiringen. See also F. M. Pelzel, Lebensgeschichte des Konigs Wenzeslaus, 1788, which is Koch's authority and seems to be Hbfler's.
5 Hbfler, 131-2 ; Lindner, 118-9 ; Walsingham, i. 452 ; Rymer's Foedera, vii. 290, 295, 301. Later, her coming was deferred (Rymer, 334 ; October 28).
6 Hbfler, 136, 156; Wallon, Richard II., I 116; cf. J. L. Lowes, Mod. Lang. Notes, xix. 240-3.
CH. II, § 2] THE PARLIAMENT OF FOWLS. 43
of Folds was certainly an excellent beginning for friendly relations between the middle-aged poet and the girl-queen.
Can we narrow the date down any further than Koch has done ? Since the decision was not made till the very end of 1380, the early part of 1381 is the earliest probable date. But Pollard and Koch1 choose the early summer of 1382. The former believes that " royal marriages were too likely to be broken off for poets to hymn them prematurely"; but is this true of betrothals of fairly mature people, which had advanced as far as this had done by tho middle of 1381 1 The other argument is a highly ingenious one. Chaucer invokes Venus to aid him,
" As wisly as I saw thee north-north-west, When I began my sweven for to wryte " (117-8).
The planet Venus obviously can never be in quite that quarter, so Koch feels the need of emendation to ivest-north-tcest, though all the MSS. read north-north-west (or nvrili-west) ; the planet can be visible in such a position only in the summer or late spring, and with the assistance of two astronomers Koch finds that the only otherwise possible years which fulfil this condition are 1380 and 1382.2 Since he believes the former to be too early, he accepts the latter. But it seems to me that any argument which depends on an emendation and a slightly cryptic interpretation is to be very doubtfully received, if at all ; and there are arguments against this date which seem to me almost conclusive. The ending of the Parliament is a clever treatment of a somewhat flat situation : the other fowls mate, to be sure, and the poem ends with a beautiful lyric, but after all the main characters are left in suspense, "unto this yeer be doon." Chaucer missed a chance for a striking, com- plimentary and pompous climax, such as every mediaeval reader would have expected. A date before the wedding seems to me so clearly suggested that, if regard must be paid to this astronomical evidence, 1380 would seem to me more likely than the summer of 1382, when the pair had been married nearly six months. As to the time of year, the selection of St. Valentine's Day was made so inevitable by the conditions of the fable that there is no justification
1 Ch. Primer, pp. 50, 90 ; Chronology, p. 38.
2 Any mathematically-accomplished student who wishes to attempt the task of verifying these results will find the astronomical wherewithal in Astronomical Papers (Washington, 1898), vol. vi., pp. 271-382.
44 CERTAIN MINOR POEMS. [CH. II, § 2
for the assigning the poem to that date.1 We can hardly come nearer the truth, therefore, than that 1381 is the probahle date.2
1 As late as May 6, 1381, Anne was expected to arrive by Michaelmas, September 29 ; but the wedding actually occurred just about a year from the original decision. This may possibly suggest that the poem was written in the latter part of the year. Mather's opinion agrees pretty well with mine; he puts the poem between Anne's arrival in England and the wedding, i. e., between December 18 and January 14 (Furnirall Miscellany, p. 305, note).
2 An attempt has been made by Dr. R. K. Root to prove that Chaucer revised P. F. Miss E. P. Hammond, in her valuable paper on The Text of Chaucer's " Parlemcnt of Fowlcs" (Decennial Pvbl. of Univ. of Chicago, 1903, vol. vii., pp. 3-25 ; see pp. 8-9), gives fifty various readings by which the MSS. are divided into two main groups, which she calls A and C. She points out "the marked decrease in group divergences after line 250," and the fact that "the text of the A archetype was probably nearer to the ultimate original verbally." The latter point she bases partly on the fact that line 221 is nearer the Italian in A than in C. (She might have said the same of 238 ; see Skeat, I. 70, for the Italian.) Now Dr. Root, in his review of Miss Hammond's paper (Journ. Germ. Phil., v. 189-93), supports the view that these divergences show deliberate corrections made in the ancestor of A, stopping at 250 ; he marks with asterisks most of the fifty various readings, which seem to him ' ' reasonably clear examples of emenda- tion." He asks "who is this skilful reviser?"; and rather than believe that the falling off at line 250 in the number of changes is due to the fact that "an inventive and poetical" scribe passed on the task to "a sober, accurate " one, he would have it that Chaucer corrected at leisure this occasional poem which had perhaps been composed in a hurry. His a priori arguments seem hardly valuable. There is not the least evidence of haste in the ending of the poem, which can hardly be called abrupt, except in comparison with the dawdling start, and the character of which I have shown to be due probablj' to another cause. The appeal to Chaucer's revision of other poems is also unfortunate. I shall show later that the revision of Kn. T. was probably slight, and certainly was not a "complete reworking " ; and that that of the prologue to L. G. W. was due to a particular and unique cause ; and we shall see that Chaucer conspicuously neglected revision in some of the Cmiterbury Tales. I hope to show also elsewhere that the peculiarities of MS. Harl. 7334 of C. T., which look infinitely more like author's revisions than those here, are probably due to a scribe. The genuine revisions in the Troilus make these various readings in the Parliament of Fowls look like the. merest petty scribal variations. Of the 45 various readings before line 251 given by Miss Hammond and Dr. Root I find none that suggest to me revision by Chaucer, few or none that suggest anything like deliberate revision by any- body, and none that may not quite well be scribal blunders in C ; in the following 19 lines I think the C MSS. certainly are corrupt, and their readings cannot possibly be due to Chaucer.
3 69 135 221
5 70 178 229
43 72 194 234
55 107 209 238
64 110 215
I grant cheerfully Dr. Root's postulate that Chaucer was a conscious literary artist ; that is why I do not think the C readings can be due to him.
CH. Ill, § 1] THR PALAMON AND ARCITE : ITS ORIGINAL METRE. 45
CHAPTEE III.
POEMS DEPENDENT ON THE TESEIDE. § 1. The Palamon and Arcite: its Original Metre.
THE Teseide of Boccaccio is used in four of Chaucer's extant poems : the Knight's Tale in the Canterbury Tales is a condensed adaptation of it ; the Troilus derives from it (as we have seen) a passage at the end (V. 1807-27) and a small one earlier (V. 1, 8-11) ; most of the description of the temple of Venus in the Parliament of Fowls (183-294) is from the Teseide; and \nAnelida and Arcite a good deal of the first seventy lines is from it.1 It must also have formed the entire hasis of a work of which all our certain know- ledge is derived from the mention of it in the Legend of Good Women (Prol. F, 420-1 ; G, 408-9). Alcestis, it will be remem- bered, is mentioning those of Chaucer's works which speak well of women or of love, with a view to moderating the God of Love's indignation, and among these works, she says, is one on
"al the love of Palamon and Arcyte Of Thebes, thogh the story is knowen lyte."
It is impossible that this can refer to the Anelida, a mere abortive fragment which never mentions Palamon, and in which Arcite appears not as a lover but as a roving and heartless flirt.
That the passage cannot refer to the Knight's Tale exactly as it stands is equally clear, since in places it is directly adapted to the Canterbury Tales, which can hardly have been fully conceived when the passage was written ; but there has been a strong tendency among scholars to regard the Palamon and Arcite as having been widely different from the Tale, though for this view the evidence has always been, to say the least, very insufficient. Tyrwhitt (I. clxxii.) suggested that "it is not impossible that at first it was a mere translation of the Theseida of Boccace." Other early Chaucer scholars, down to 1870, were divided in opinion.2 It was ten Brink, in his distinguished Chaucer Studien,3
1 It has lately been proved that the Legend of Ariadne shows the same in- fluence ; but it hardly extends to verbal borrowings, and, I believe, comes through Kn. T. rather than directly (see J. L. Lowes, Pull. Mod. Lang. Assoc., xx. 802-18, and pp. 122-5, below).
2 On the views of Saiidras, Hertzberg, Ebert and Kissner, see ten Brink's Studien, pp. 39-48.
3 Munster, 1870 ; pp. 39-69.
46 POEMS DEPENDENT ON THE TESEIDE. [CH. Ill, § 1
who set up the highly ingenious theory that Chaucer not only largely altered the Palamon in adapting it to the Knight, but that he originally wrote it in the seven-line stanza ; and that the longer of the Teseide passages in the Troilus and those in the Anelida are fragments of this earlier version. The former he thinks (p. 61) may have been put in tentatively by Chaucer, or carelessly by his scribe, and the survival of the latter may be due to an attempt to preserve parts of the original poem which he did not require for the revised form of it (p. 56). The Parliament of Fowls he believes (p. 128) was written before the Palamon was finished, and therefore that the Teseide passage there was never in the Palamon.1 The almost universal acceptance which this theory has found must be due largely to the authority of the prominent scholar who proposed it. Dr. John Koch defended and developed, it in an article in the first volume of the Englisclie Studien ; 2 he regarded the Teseide passages in all three of the stanzaic poems as part of the debris of a Palamon and Arcite deliberately broken up before the Kniglifs Tale was conceived. Ten Brink's theory, and usually Koch's modification of it, was accepted by Dr. Eugen Kblbing,3 by Mr. A. W. Pollard,4 by Professor Skeat,5 and by many others, and to this day may be called the orthodox view.6 Only three writers, to the best of my knowledge, have expressed them- selves against it ; at the time of writing his introduction to the Canterbury Tales in the Globe Chaucer, Mr. Pollard had changed his views, and rejected the theory;7 Dr. F. J. Mather argues against it ; 8 and finally Professor J. L. Lowes agrees with them in rejecting it.9
1 Cf. also his Hist. Engl. Lit. (London, 1893), ii. 63-72 (German version, Strassburg, 1893, ii. 65-74).
2 Pp. 249-93 (1877) ; translated in the Chaucer Society's Essays, 357-400 ; cf. also his Chronology (1890), p. 30. 3 Engl. Stud., ii. 528-32.
4 Chaucer Primer (London, 1893), pp. 76-7.
6 III. 389-90 ; cf. his Prioress" Tale (Oxford, 1893), pp. xvi.-xvii. But in 1900 he seems to have held it with less conviction (Chaucer Canon, p. 57 ; yet cf. p. 154 !).
6 Assumed, e. g., by Koch (by implication, Engl. Stud., xxxvi. 140), Dr. R. K. Root, Journ. of Engl. and Germ. Philol., v. 193, and Professor W. P. Ker, Essays on Mediaeval Literature (1905) p. 96.
7 Though not decisively (pp. xxvi.-xxvii). He rejects it with horror in his Knight's Tale (1903), p. xviii., and in the 1903 edition of his Primer.
8 In his Chaucer's Prologue (1899), xvii.-xviii., and in a paper on The Date of the Knight's Tale in the Furnivall Miscellany (An English Miscellany Presented, etc., 1901), pp. 301-13.
9 Publ. Mod. Lang. Assoc., xx. 809; he refers to Mather, and to the present discussion (then of course unpublished).
CH. Ill, § 1] THE PALAMON AND ARCITE : ITS ORIGINAL METRE. 47
The capital errors in the stanza-theory seem to me its enormous a priori improbability, and its needlessness. The evidence for it, which might perhaps be respectable if the theory in itself were either plausible or serviceable, breaks down at once if these mistakes are recognized. In fact, the longer one ruminates on the theory and its consequences, and the more carefully he examines the evidence, the more inconceivable does it become. At the risk of being intricate and long-winded, it is worth while to try once and for all to destroy it.
Why should Chaucer have wished to transpose a poem of thou- sands of lines from stanzas into couplets ? The heroic couplet may be on the whole a finer and more useful form of verse than the rhyme-royal ; but Chaucer was very far from abandoning the latter even after he had begun writing in the former, and for a poem like the Knight's Tale the stanza would have been perfectly suitable. Such a proceeding would be strange in any one, and would require strong evidence for its proof. A writer might well wish to withdraw a short poem in order to develop it, but such destructive treatment as is postulated here is unparalleled, so far as I know, in ancient or modern times. But Chaucer especially was not a man to be easily brought to spend trouble on a detailed and vexatious task to gain no great advantage; on the whole, he took his poetry with a lack of constant seriousness that is in part characteristic of the Middle Ages and in part of himself, and his willingness to leave things unfinished and unrevised (even where revision would seem imperative) may be proved by a mere reference to the Anelida, the House of Fame, the Astrolabe, the Legend, and the Tales of the Shipman, the Squire, and the Second Nun. It is true that he revised the Troilus and the Prologue of the Legend; but these revisions must have been largely done by merely altering earlier MSS., and they left the metre untouched, which is the point at issue and makes all the difference.
The supporters of the stanza-theory of course have felt the need of discovering some motive for Chaxicer's supposed procedure. Ten Brink's explanation is very curious — that the poem was pub- lished and failed (Studien, p. 64). If the possibility of such a suggestion may ever be denied, it may here. It is difficult to imagine the failure of anything in the Middle Ages, and as nearly as we can reconstruct in fancy the supposed stanza-poem out of the Teseide, the Knight's Tale, and the Troilus, it would have had an
48 POEMS DEPENDENT ON THE TESEIDE. [CH. Ill, § 1
interest, a sweetness and a brilliance which would have been likely to make it one of the most successful of his works. Moreover, the changes which he would have made would have been far too subtle to weigh much with a mediaeval audience ; except a change in the direction of brevity, which in the Middle Ages was not deemed a virtue at all. Again, those who have criticized the theory 1 have raised the pertinent question whether, if the poem had circulated enough to have failed, it could have been withdrawn so completely that Chaucer would have cared to use it again in a new form, and that no MS. of it should have come down to us. This explanation is to be rejected without qualification.2
Dr. Koch3 gives a different reason for the suppression of the earlier form of the poem. He thinks it was never published, and was transformed not because Chaucer preferred the couplet to the stanza, but because it was written in the sentimental and pathetic tone of the Teseide and was inharmonious with the maturity of the Troilus, earlier than which he assumes that it was written. His only evidence for this belief is not a little curious — that there is no other discoverable reason why Chaucer should have rejected the earlier poem. This explanation is more extraordinary, if possible, than the other. The Teseide is not sentimental, and neither are
1 Pollard (Globe Ch., xxvi.) ; Mather (Miscellany, 307) ; and even Koch, who accepts it ; he thinks there was only one MS., which was never published (Engl. Stud., i. 281). Ten Brink thinks there were only a few (p. 65).
2 It has sometimes been thought to be countenanced by the fact that Alcestis says "the story is knowen lyte." This is not the view of ten Brink, Koch, or Skeat, but even Tyrwhitt made the suggestion (I., clxxii.). Another probably false explanation of her remark is ottered by ten Brink (p. 64) and Skeat (III. 306), and approved by Pollard (Kn. T., xiv. f., note)— that Chaucer is alluding to Boccaccio's statement in the introductory epistle to Fiammetta or in the opening of the poem (I. 2), that it is an ancient story known to few. But this was not true after Boccaccio had written his poem, still less if Chaucer's poem had been published. Why should Alcestis echo Boccaccio's language, and why shoxild she say thoiigh? The remark seems useless and senseless unless it means just one thing, that though she wished to make the most of all Chaucer's creditable performances, she doubted whether the fact that he had written such a poem had reached Love's ears (or those of Chaucer's readers). So thinks Koch (p. 282). The most natural reason for such a doubt is that the poem had not yet been published. If it had been written only shortly before, as we shall later find other reasons to believe (cf. pp. 76-80), there are several possible reasons for its still being withheld, much the most probable of which is that it was not quite finished. Any apparent strangeness in Alcestis' mention of an unpublished work is fully explained by the fact that she is raking in everything she can find to Chaucer's credit. Is not this more likely than that she should mention a work which Chaucer in vexation had been rending asunder ? (Cf. Pollard, Globe Ch., xxvi.)
3 Engl. Stud., i. 279-83.
GH. Ill, § 1] THE PALAMON AND ARCITE : ITS ORIGINAL METRE. 49
the passages from it in these two poems. But Chaucer was not in- disposed to sentiment at any time in his life, and the Knight's Tale itself contains affecting passages which are not in the Teseide (e.;/., 1281-1333).
Then if we cannot believe that Chaucer had the failure or thf immaturity of the Palamon and Arcite to induce him to recast it in a new form, the plausibility of the stanza-theory is reduced to the utmost tenuity.
The stanza-theory is no more useful than it is probable. There is not the slightest difficulty in accounting for the presence of stanzas from the Teseide in the Troilus, the Parliament, and tho Anelida. When Chaucer had become familiar with the poem, and before he had resolved to translate it, why should he not take from it a brilliant description for the second poem and an imposing open- ing for the third 1 l And even though we may regard the addition of
1 Pollard seems to feel some difficulty in the use of the same passages in the Anelida and the Kn. T. But there is none if the former was abandoned before the latter was begun. — Chaucer had been more or less familiar with the Teseide since shortly after his first journey to Italy. Besides the more important quotations discussed above, I am not the first, to point out the borrowing of T. C., V. 1, 8-11 from Tes., IX. 1 and II. 1. Parl. of F., 176-82 maybe another borrowing. One more, quite as clear, is more curious. In three passages Chaucer shows that he understood Helicon to be a spring.
"Ye sustren nyne eek, that by Elicoue In hil Parnaso listen for to abyde" (T. C., III. 1809-10).
(Rossetti, T. C. and FiL, p. 169, attributes this to Tes., I. 1, but he seems to be mistaken. He is clearly unjustified also in attributing the first part of the stanza to Tes., I. 3.)
' ' Be favorable eek, thou Polymnia, On Parnaso that, with thy sustres glade, By Elicon, not fer from Cirrea, Singest with vois memorial in the shade" (Anelida, 15-18).
(Cf. " Parnaso Cirreo," Tes., VIII. 57. In the passage of the Teseide which Skeat mentions as the source of this, Boccaccio refers to the "monte Elicona," but the passage presently to be quoted must also have been in Chaucer's mind. )
' ' And ye, me to endyte and ryme Helpeth, that on Parnaso dwelle By Elicon the clere welle " (H. F., 520-2).
This has always been explained as due to Dante :
"Or convien ch' Elicona per me versi" (Purg., XXIX. 40).
(So Skeat, III. 254. Scartazzini adduces a somewhat similarly ambiguous line from Virgil (Aen., VII. 641). Note that Chaucer always uses the Italian form for Parnassus; even in Frankl. Prol., 721, though that passage is supposed to be imitated from Persius. Cf. p. 165 below.) A comparison of these passages with one in the Teseide will show, I think, that Boccaccio's error was the source of Chaucer's ;
DBV. CH. E
50 POEMS DEPENDENT ON THE TESEfDE. [CH. Ill, § 1
the passage in the Troilus as no great improvement, it is no easier to explain it as a purple patch taken from an old garment.1 New or old, Chaucer would not have put it in unless he liked it, and it is a poor compliment to him and a very forced conclusion to say that he used the verses because he had them.2 To infer a thousand stanzas from two or even twenty seems very rash. The theory ex- plains just one real difficulty. The Prologue to the Legend of Good Women has usually been considered as Chaucer's first essay in the heroic couplet,3 and the God of Love's permission to " make the metres " 4 of the legends as Chaucer pleased (F 562) has been inter- preted as the proof of this and as the inauguration of a new metre, new not only to Chaucer but to all English poetry. Frankly, this is the most obvious explanation of a rather odd remark. But other explanations are possible. When Chaucer wrote the Prologue5 he may have intended to use various metres in the legends, as later
" Vedeasi appresso superar Pitone, E quindi sotto 1'ombre grazibse Sopra Parnaso presso all' Elicone Fonte seder con le nove amorose Muse, e cantar raaeatrevol canzone " (XI. 63).
The error was, first and last, somewhat widespread. It is explicit in the notes to the Dante passage by Dante's own son and by another Florentine of the fourteenth century (Petri Allegherii . . . Gommentarium, p. 503 ; Com- tnento . . . d'anonimo Fioreiitino, II. 475) ; perhaps both were misled by Dante. Deschamps, in his balade to Chaucer (No. 285 ; Soc. Ane. Textes, ii. 139), exhorts the English poet to give him an authentic draught "de la fon- taine Helye." The error occurs also in a letter of Boccaccio's (Corazzini, Le Lettere di Boccaccio, p. 195). Later, Skelton makes the same blunder (Philip Sparrow, 1. 610), and so does Spenser (Shep. CaL, April, 41-2) ; both, no doubt, were misled by Chaucer. In spite of the frequency of the error, Boccaccio was probably Chaucer's blind guide, as is partly shown by the fact that they two only (except the Dante commentators) represent Parnassus and Helicon as being close together. In reality they are some thirty miles apart.
1 Cf. ten Brink, Studien, pp. 60-2.
2 The idea that Chaucer used up fragments of cast-off poems in this manner has been advanced in another connection— to explain the presence of bits translated from Pope Innocent's De Contemptu Mundi in the Man of Law's Tale. Ten Brink, who held this view in the case of the Palamon, in this latter case rejected it with mockery (Engl. Stiid., xvii. 22). Dr. Emil Koeppel characterizes Chaucer's supposed procedure by the gentle word " economy," and pleads ten Brink's own example for holding the view in the second case (ibid., p. 197; cf. Herrig's Archiv, Ixxxiv. 410 ff.). Skeat holds the "economy" view in both cases (cf. pp. 181-2 below).
3 The very subversive views of Professor Lowes on this matter will be discussed later (see p. 125 below).
4 MSS. Pepys and Add. 9832 read "Make thy matere (the maters) of hem as the lest (ye liste)."
5 It will appear subsequently that version F (" B ") is the earlier ; the line is omitted from the other version.
CH. Ill, § 1] THE PALAMON AND ARCITE : ITS ORIGINAL METRE. 51
in the Canterbury Tales. Better still, if the Palamon was unfinished, and if it had not yet gone abroad among his readers, so far as they were concerned the decasyllabic couplet was a new metre, and it was by no means old to Chaucer. Therefore, though it is difficult not to regard this line as an allusion to a metrical innovation, the innovation need not have been made in this poem.1
This exhausts the a priori considerations, and it seems temperate to say that direct and weighty evidence would be required to over- throw the strong presumption against the stanza-theory. Such evidence does not exist; on the contrary, there is important evidence on the other side. And first let us examine the supposed Palamon passages in the three stanzaic poems for indications that they were not originally meant for their present positions, indica- tions which might be expected to appear on careful scrutiny. We may seek them, but we shall not find them. But suggestions that the two longest never occurred in any other English poems we shall find. Dr. Koch asks2 why, if Chaucer took these passages directly from the Teseide, he did not more completely fuse them with their present surroundings. As to the passage in the Troilus (V'. 1807- 27), whatever lack of harmony there may be between it and its context is entirely explained by the fact that it caine in on the revision.3 The passage in the Anelida is perfectly fused; if it seems to us partly superfluous, this may be because the poem is frag- mentary. As to that in the Parliament, it is difficult to imagine any fusion more perfect. In fact, examination will show, I think, that if Chaucer took these two longer passages from the Palamon he made a largely unnecessary revision of them.
In the Anelida and Arcite lines 1-21, 36-9, and 50-70 are (partially) from the Teseide, but ten Brink4 regards the whole of the first ten stanzas (lines 1-70) as derived from the Palamon, with certain changes. Now the first stanza contains a very warlike invocation of Mars, Bellona and Pallas, though the Teseide (I. 3) invokes Mars, Venus and Cupid. It is not at all likely that the more martial invocation stood in the Palamon, which if anything
1 So Dr. Mather suggests (Furnivall Miscellany, 312).
2 Engl. Stud., xxvii. 3.
3 If the Troilus passage ever stood in the Palamon, Chaucer must have rewritten the first line .(in the Italian, XI. 1, " Finite Arcita colei nominando, La qual nel mondo piu che altro amava ") ; no other change would have heen necessary, and there is no internal evidence for either opinion,
4 Pp. 56-8.
52 POEMS DEPENDENT ON THE TESEIDE. [CH. Ill, § 1
must have been less warlike than the Teseide ; l the Anelida , on the other hand, begins in a more warlike style than the Knight'* Tale, and since it breaks off with Anelida's vow of sacrifice and visit to the temple of Mars, the indications are that it was to continue in that style. Therefore the stanza must have been revised ; yet needlessly, for love is prominent enough in the Anelida to render Boccaccio's invocation perfectly suitable.2 As far as this evidence goes, then, it indicates that these stanzas were never in the Palamon. It may possibly be thought that the fact that some of them have no obvious connection with the story which follows suggests that they were not written for it ; but if the poem had been continued a connection presumably would have appeared, and we certainly ought to abandon the idea that Chaucer put them in for no better reason than to preserve them.3
The description of the temple of Venus in the Parliament (183- 294) is taken from a passage which -stands later in the Teseide (VII. 51-66) than in the Kniglifs Tale; the prayer which Palamon offers just before the tournament becomes personified as a kind of njTmph, who, before presenting herself to the goddess, visits and inspects the actual abode of Venus at Mount Cithaeron ; in the Teseide the oratories built by Theseus are not described at alL The first- personal verbs in the Parliament ("I saw," etc.) are therefore third-personal (" vide," etc.) in the Italian. Ten Brink (p. 128) thought this passage so thoroughly fitted' to the Parliament that it could not have stood in the Palamon. Koch,4 however, disagreed, and his view is accepted by Skeat (III. 390) ; it may therefore be
1 As ton Brink admits (p. 62). Kn. T. omits the first book of the Teseide, on the wars of Theseus with the Amazons. Ten Brink is not quite fair in saying, "wir haben nicht den geringsten grund zur vermuthung, dasz der kiiegsgottin in Anelida and Arcite ein groszerer spielraum zugedacht war als in Palamon and Arcite " (p. 57).
2 Other changes which must have been made are in 11. 11, 21 (probably), 48-9, and 67-70, as ten Brink admits (pp. 57-8) ; and it must be remembered that changing one line of a stanza may involve changing much more.
:{ Kolbing (Engl. Stud., ii. 528-32) points out verbal resemblances between A. A. and Kn. T. (to which I may add A. A. 182 = Kn. T. 2397), where there is little or nothing in the Teseide to correspond, and thinks they indicate that Kn. T. is done over from an earlier stanzaic version.- But if A. A. was written and abandoned before Kn. T. was begun, these reminiscences are natural enough. Mather bases on A. A. an argument different from mine against the stanza- theory (Miscellany, p. 307). He points out that "it stops abruptly with the promise of a description of the temple of Mars, a description which, according to the theory, lay ready in Palamon. It is strange that Anelida should end where it required only a little copying to carry the story scores of lines further." Cf. Pollard, Primer, p. 80.
4 Engl. Stud., i. 249, 261-2.
OH. Ill, § 1] THE PALAMON AND AECI'fE : ITS ORIGINAL METRE. 53
said that the prevalent form of the stanza-theory puts this passage in quite the same category as the other two. Now I think it may be shown by something like a redudio ad absurdum that if it ever was in the Palamon it must have been very much more extensively rewritten than Dr. Koch thinks, yet that this rewriting was largely needless ; and therefore that it never was in the Palamon.'
I agree with him that Chaucer is not at all likely to have adopted in the Palamon what he calls lt diese etwas gezwungene und unnatiirliche darstellungsweise " (p. 261) of personifying the prayers and conducting them to the actual dwellings of the deities. When we consider what liberties Chaucer takes with the Filostrato and the Teseide, how his treatment of his sources always tends to the rational and the simple, and how his sense of the incongruous was as much greater than Boccaccio's as his reverence for^ precedent was less, it becomes allowable to disbelieve that Chaucer would have adopted so frigid a conceit. Moreover, in the description of the temple of Mars in the Kniyhfs Tale, which belongs by hypothesis in the same category, there is not the slightest indication that from the first he did not appropriate Boccaccio's description to the shrine erected by Theseus in his theatre ; yet this passage is one of the few longer ones which follow the Teseide closely, with many lines almost literally translated, a fact which certainly makes against the supposition that such changes were made as would be involved in getting rid of the personified prayer. Therefore the indications are that in neither case was it personified. Finally, certain phrases in the description in the Parliament are distinctly inconsistent with the personification. We should have to believe that Chaucer attri- buted to the young woman who represents the prayer strong views on the subject of decorum, or else pleasure in beholding the thinly- veiled beauties of Venus :
" The remenant wel kevered to my pay Eight with a subtil kerchef of Valence, Ther was no thikker cloth of no defence " (271-3).1
After describing Venus, the poet says (departing from his original),
" thus I leet hir lye, And ferther in the temple I gan espye " (279-80),
singular conduct on the part of a prayer addressed to the goddess ;
1 " E faltra parte d'una Veste tanto sottil si ricopria, Che quasi nulla appeua nascondia " (Tes. VII. 65).
54 POEMS DEPENDENT ON THE TESEIDE. [CH. in, § 1
he then goes on with a part of the description which in the Teseide immediately precedes the description of the goddess. This change of order has no apparent motive, and certainly would not have been made if the prayer figured in the account. It is not legitimate to plead that in fitting the stanzas to the Parliament Chaucer made such gratuitous alterations as these in a passage that would have served perfectly well unaltered. It seems certain, then, that the prayers were not personified in the supposed stanzaic Palamon.
Yet if this passage was taken from that poem it is equally clear from internal evidence that the prayers were personified. Unless very extensive changes have been made, some one filled a prominent part in the description, and in a perfectly impersonal romance who could it have been if not the prayer? In the 112 lines such expressions as " herde I," " saw I," occur no less than fourteen times ; and, what is still more striking, five phrases imply motion on the part of the observer.1
Dr. Koch thinks the former set of phrases are entirely paralleled in such phrases as"ther saw I," which occur five times in the 84 lines of the temple of Mars passage in the Kniyld's Tale, and five times in the 38 lines of the temple of Diana passage. But if we take them in conjunction with the indications of motion (entirely absent from the passages in the Tale), it becomes clear that we have all the difference between a vividness due to poetic transport and a deliberate case of what rhetoricians sometimes call description by means of narrative. In order to make this quite unmistakable it is necessary to pause to account for these phrases in the Tale. Koch thinks that their absence2 from the temple of Venus passage in the Tale, and their presence in the Mars and Diana passages, and in the temple of Venus passage in the Parlia- ment, shows that the latter three passages are all derived from the Palamon. I think a perfectly satisfactory explanation is the following. Having written an original 3 description of the temple
1 " Ther I fond" (242), " as I wente " (253), " foiid I " (261), " thus I leet hir lye " (279), " ferther in the temple I gan espye " (280). All but the third contain rhyme-words.
2 He points out, however, " maystow se" (1918 ; see Koch, 260). Mather (Miscellany, 304) suggests that Chaucer may have been a poor enough Italian scholar to mistake vide for vidi. This explanation will hardly do, for no one could have failed to see that the observer of the temple was the personified prayer. He also makes the suggestion, given above, that the use of the first person is merely a licence for the sake of vividness.
3 For no other reason that I can conceive than that he had already written P. F. ; see p. 78 below.
CH. Ill, § 1] THE PALAMON AND ARC1TE : ITS ORIGINAL METRE. 55
of Venus for the Palamon (which I believe was substantially identical with the Knight's Tale), he returned to the Teseide for the temple of Mars, which in the Tale comes next. In reading the account in Boccaccio he felt the vivid effect of the repeated vide, and by a licence not unusual in poetic description he reproduced it, of course by verbs of the first person ; and then carried it through his original account of the temple of Diana, which follows. It appears, therefore, that the conditions in the Tale and in the Parliament are not parallel.
So we have accomplished a reductio ad absurdum. If the description in the Parliament of Fowls has not been considerably altered from its original form, the prayer at first must have been personified and- had the experiences indicated. But both proba- bility (as Koch admits), and evidence, oppose the idea that the prayer was personified. Therefore if this passage occurred in the Palamon it must have been considerably altered before it was put into the Parliament. But every one of the alterations was unnecessary — whether or not the personified prayer appeared in the first form, the passage would have served quite well unchanged (except for the person of the verbs) ; and, considering the extent to which Chaucer must have been affected by the sin of Accidia while he was using up fragments of this devoted poem, a sin in reality not unknown to him, it is very unlikely that he would have made these alterations. Some of them would have involved more trouble than recasting couplets into stanzas.
The irpshot of our examination of the supposed Palamon passages in the Troilus, the Anelida and the Parliament is about this : — There is not a shred of evidence that a single line of them ever appeared in an earlier English poem, and there is strong evidence that the two longest of them did not. If this conclusion does not destroy the stanza-theory, at the very least it disposes of the conjecture that there may be any evidence in these passages to favour it. Now since these passages are practically all the evidence which the theory has, and considering the enormous burden of proof which rests upon its advocates, it is not difficult to see where we are coming out.
But there is a whole set of evidence yet to be examined, which is to be derived from a minute comparison of the Knight's Tale and the Teseide, with a view to discovering if there is anything to show whether an English poem in stanzas intervened. It is
56 POEMS DEPENDENT ON THE TESEIDE. [cH. Ill, .5 1
necessary to begin with three postulates, supposing the Palamon and Arcite to have been written in stanzas. The first is that Chaucer used the Palamon as the basis for the Knight's Tale, and did not produce a quite new poem directly from the Italian; probably every one would grant this without cavil, as Kb'lbing,1 ten Brink,2 and Koch8 do. The second postu- late is that though it is not impossible that Chaucer might once in a while refer again to the Teseide, it is illegitimate to suppose that he did do so often or in any particular case, for once he had drawn his material from the Teseide and put his own interpretation on it, there is no reason why he should regard Boccaccio's form of the story, or keep it open before him during the revision. Finally it is ten Brink's opinion4 that in general "berechtigen uns die fragmente [des angeblichen stanzaischen Palamon und Arcite] wohl zu der annahme, dasz hinsichtlich der treue der iibersetzung und des ausseren umfanges derselben zwischen Palamon und Arcite und der Teseide ein ganz ahnliches Verhaltnisz bestanden habe wie rwischen Troylus und Cryseyde und dem Filostrato." If the stanza- theory is correct, this is an opinion which it is quite improper to deny,5 and it has always been granted. That it is therefore entirely fair to use the Troilus as a test for some of the evidence derived from comparing the Knight's Tale and the Teseide is my third postulate. It is highly important that these postulates should be clearly seen to be a necessary consequence of the stanza-theory, for it is by their means that I shall attempt to reduce it to an absurdity — to show that if they are a necessary consequence of it, the theory is wrong.
This evidence will show that there are no vestiges of stanzaic structure in the Knight's Tale at points where, if the theory is correct, it must necessarily appear on careful scrutiny. The evidence may be divided for convenience into three classes : that which deals with the actual number of lines taken from the Teseide ; that which deals with possible traces of the beginning and end of the stanza ; and that with passages where the Tale closely follows the Teseide for at least several lines. For the first two classes it is necessary to have a table of the lines in the Tale
1 Engl. Stud., ii. 529-31. 2 Studicn, pp. 56, 61, 65.
:i Engl. Stud., i. 277-8. 4 Stitdicn, p. 63.
8 Except that we can hardly suppose the Teseide to have been expanded as much as the Filostrato was. Teirfirink cannot have meant that.
CH. Ill, § 1] THE PALAMOX AND AKCITE : ITS ORIGINAL METRE. 57
derived from the Teseide, arranged according to the position in the Italian stanza of the lines from which they are translated. Such a table will he found in Appendix B.1
Much of my evidence is based on the following considerations. The last four lines of a seven-line stanza without change form two couplets, and a whole stanza may be resolved into couplets by the omission of the 2nd line, and perhaps some adaptation of the 1st and 3rd. This statement may easily be tested; of the 16 Teseide stanzas in the Parliament of Fowls, such a transformation would be perfectly easy in about 10, and of the 156 stanzas in book I. of the Troilus it would be perfectly easy in about 86. I include here only cases where the 2nd line is easily dispensable in sense as well as form. There is no reason to believe the supposed stanzaic Pcdamon to have been so much longer than the Knight's Tale that Chaucer must have generally used a more heroic treatment, and could not have used this device most of the time ; the Teseide is a very profuse poem without vivid psychological interest, and it is quite certain that from the first he would have greatly condensed it. And my representation of Chaucer's procedure should not be regarded as crude or trivial. No poet can escape the technical conditions of his art ; no poet would disdain a simple method of preserving his own good work. Especially would a sometimes impatient and easily-satisfied poet like Chaucer, who had a par- ticular fondness for his own words,2 have welcomed a device which would save time and trouble, and ensure the preservation of bits of this unhappy poem for which he would still cherish a certain tenderness.3 But whether, thus at close quarters, the stanza-theory begins to look absurd, let others decide.
First we will consider the frequency with which the various lines of the Italian stanza occur in the Knight's Tale.4 Here the important premise must be made that, when ottave rime are trans- lated into seven-line stanzas, in a general way an Italian line passes
1 Pp. 226-230 below. Book I. of the Troilus, which contains 1092 lines, is enough for purposes of comparison, for it follows the Filostrato closely.
2 As is shown by the very large number of phrases which appear in his works more than once.
3 There is a curious illustration of the ease with which the opposite change can be made (from couplets into stanzas) in the spurious prologue to the Franklin's Tale, composed out of the true Merchant's Epilogue and Squire's Prologue (see Six-Text, Introd., p. 54).
4 The total number of lines due to the Italian is valueless as evidence. According lo my count, it is 498 out of 2250 (22%); in the Troilus (according to Rossetti, p. iii.) it is 2583 out of 8246 (31%).
58 POEMS DEPENDENT ON THE TESEIDE. [cH. Ill, § I
over into a line of the same position in an English stanza ; in the Troilus, book I., f of the Italian 8th lines translated (35 oat of 47) correspond to English 7th lines, f of the Italian 2nd lines (56 out of 64) correspond to English 2nd lines, and ^f of the Italian 1st lines (64 out of 69) correspond to English 1st lines. A cursory inspection of Mr, Rossetti's edition of the Troilus and the Filostrato in parallel columns also will show at once that the first and second parts of an Italian stanza correspond respectively to the first and second parts of an English stanza. Now by hypothesis the same conditions should hold for the stanzaic Palamon and Arcite.
The 504 Italian lines which appear in the KnigWs Tale are distributed in the Italian stanza as follows :
1st lines, 98 (19%). 5th lines, 56 (11%).
2nd „ 80(16%). 6th „ 44(9%).
3rd „ 69(14%). 7th „ 55(11%).
4th „ 47(9%). 8th „ 55(11%).
For the Troilus the figures are these :
1st lines, 69 (16%). 5th lines, 47 (11%).
2nd „ 64(15%). 6th „ 45(10%).
3rd „ 59(14%). 7th „ 53(12%).
4th „ 51(12%). 8th „ 47(11%).
In degree of frequency the eight lines stand in the following order :
Knight's Tale 1, 2, 3, 5, 7-8, 4, 6 ;
Troilus and Griseyde . . . . 1, 2, 3, 7, 4, 8-5, 6.
The closeness with which these results agree is obvious ; but so far from favouring the stanza- theory, this fact makes strongly in favour of the view that the Tale was made directly from the Teseide. Here I take issue with Dr. Koch.1 I maintain that, if
1 Who says (Engl. Stud., I. 277-8) : "Die betreffenden stanzen d^r Teseide sind in der Knightes Tale gerade so behandelt wie in den siebenzeiligen strophen, in denen er Boccaccio's gedichte iibersetzt hat. Er ubertragt namlich mbglichst genau die ersten zeilen jeder stanze — insoweit er sich iiberhaupt an sein vorbild halten will — kann dies aber (einmal wegen der schwierigkeit des versmasses ; zweitens, weil er ja 8 zeilen des originals in 7 eigene zusammenzieht) nicht immer fiir den rest durchfuhren, und liisst daher hiiufig die mitte weg, um dann bfter wieder die letzten zeilen wbrtlicher wieder zu geben. Genau diese behandlungsweise finden wir in den stellen der Kn. T., welche mit strophen der Teseide correspondiren. In den von mir citirten versen ergibt sich das verhaltniss der nach art der siebenzeiligen strophen bearbeiteten stanzen zu denen, deren anfang unberiicksichtigt geblieben ist etwa fc 5 : 1 ; ganz so im Troilus, soweit er von Mr. Kossetti mit dem Filostrato verglicheu ist." Dr. Koch clearly misinterprets his facts.
CH. III,§ 1] THE PALAMON AND ARCITE : ITS ORIGINAL METRE. 59
Chaucer wished to translate a whole Italian stanza, it would make little difference whether he was putting it into stanzas or couplets, at least so far as concerns the presence in the English version of this or that Italian line. In either case, a falling off in frequency of occurrence is most natural at the middle of the stanza, for that is where the diffuseness of the ottava rima especially shows itself ; not only are the beginning and end of the stanza the strategic points,1 but the freshness of the rhymes there gives the poet a freedom which he has not in the 4th, 5th, and 6th lines. There- fore the agreement between the two poems does not favour the stanza-theory.
But further examination of these lists I think will reveal evidence which is absolutely destructive of the stanza-theory. We have seen that the last four lines of a 7 -line stanza are much easier to take over into a couplet-poem than the first three, and also that in general the last part of an Italian stanza corresponds to that of an English stanza. If the stanza-theory is correct we ought to find in the Knight's Tale the Italian lines 5-8 as compared with lines 1-4 much better repre- sented than they are in the Troilus. But the opposite is the case • in the Troilus they are 192 to 243 (44%), and in the Tale 210 to 294 (only 42%).
Most important of all, since it has been shown how easily and how often a 7-line stanza may be transformed into couplets by omitting the second line, it would be very surprising if this line should not suffer considerably during the transformation ; a fact which would be instantly betrayed in the Knight's Tale by a falling off in the number of Italian 2nd lines represented, about f of which we have inferred would have passed over into 2nd lines in the English stanzas. Yet on consulting the list we see not only that the Italian 2nd lines are the most numerous of all, except the 1st, but also that their percentage of the whole number of Italian lines represented is higher than in the Troilus (16% to 15%). Therefore the very closeness with which the figures for the Knight's Tale agree with those for the Troilus is a very strong argument against the. stanza-theory. If our postulates are sound, I think it disproves it.
I come now to the lines which are closely translated, and there- fore must have stood practically the same in the Palamon :
1 The Italian final couplet has a summary, epigrammatic character that tends to preserve it ; the very rhymes are sometimes carried over into the K. T. (1625-6, 2371-2, 2445-6).
60 POEMS DEPENDENT ON THE TESEWE. [cH. Ill, $ 1
1st lines, 26 (27%). 5th lines, 7 (7%).
2nd „ 20(21%). 6th „ 4(4%).
3rd „ 8(8%). 7th „ 17(18%).
4th „ 9(9%). 8th „ 5(5%).
The testimony of these figures is the same as that of the others. Though on the stanza-theory we should expect lines 5-8 to be much more numerous than 1-4, there are only 33 of them against 63 (34%) ; and Italian 2nd lines are more fully represented than any others except the 1st.1
The second class of evidence deals with possible traces of the beginnings and ends of the original stanzas. In the nature of the case, as will appear, it is less satisfactory and conclusive than that which has just been presented. But it bears out the other, and I trust will make my refutation of the stanza-theory more well- rounded.2
One characteristic of Chaucer's treatment of the couplet is the frequency with which a strong pause, marking a striking period in the thought, breaks the couplet at the end of the 1st line. If the stanza-theory is correct, we should find this characteristic rather unusual in the Kniylit's Tale; since in a stanzaic poem, as one may see by a glance at the Troilus, practically all such pauses come at the ends of stanzas, and we have seen reason to believe that the last four lines of the stanzas would in large measure have been carried over into the Tale unchanged. Therefore there should be a very striking contrast between the Knight's Tale and Chaucer's other poems in the heroic couplet as regards this break in the middle. Below will be found a table in which the comparison is
1 Another point connected with these may be mentioned here. The omis- sion of the 2nd line of an English stanza would leave lines 1 and 3 (perhaps slightly altered) as a couplet ; this ought in many cases to be betrayed by couplets in the Knight's Tale formed out of Italian 1st and 3rd lines. This actually happens four times out of a possible 80 or so. (See 1893-4, 2011-2, 2393-4, 2831-2. Four other cases are not to be counted, because the Italian 2nd line is fully included in one or both of the English lines.) But not only is it natural enough in any case to find this happening occasionally, but in every one of these passages a good reason is apparent ; the Italian 2nd line is unimportant, it sometimes partly survives in one of the English lines, and often the whole translation is exceedingly distant.
8 It may perhaps be asked if older stanzas show at all by the presence in the Talc of blocks of 6 or 8 lines ; there are not enough such to be significant. Another similar matter may be mentioned. Four times in the 1092 lines of T. C., I., one Italian stanza is expanded into two English ; in the Tale (more than twice as long) one Italian stanza makes 11 or 12 lines only three times, always where Chaucer is very closely following his original. This is not offered as evidence, but merely to dispose of a possible conjecture.
CH. Ill, § 1] THE PALAMON AND ARCITE : ITS ORIGINAL METRE. Gl
made. For the pauses which come at the end of the first line, and for those at the end of the couplet, I have borrowed from Shak- sperian metrical criticism the terms "run-on" and "end-stop." I have made two lists, one of which includes all the breaks in the sense which seem really considerable, the other (in order to secure as much objectivity as possible) only the ends of paragraphs.1 The poems selected for comparison are as miscellaneous as possible in character and probable date.2
|
ALL BREAKS IN SENSE. |
PARAGRAPHS. |
|||
|
' |
Run-on. |
End-stop. |
Run-on. |
End-stop. |
|
Squire's Tale |
•63 |
•37 |
•50 |
•50 |
|
Sumner's Tale |
•58 |
•42 |
•56 |
•44 |
|
Legend of G. W. |
•51 |
•49 |
•47 |
•53 |
|
L.G.W., Prol. G. (A)... |
•49 |
•51 |
•59 |
•41 |
|
Pard. Prol. and Tale ... |
•46 |
•54 |
•39 |
•61 |
|
L. G.W., Prol. F. (B)... |
•44 |
•56 |
•46 |
•54 |
|
Knight's Tale |
•40 |
•60 |
•30 |
•70 |
|
Physician's Tale |
•39 |
•61 |
•25 |
•75 |
|
Wife of Bath's Tale ... |
•34 |
•66 -31 |
•69 |
|
|
C. T. Prologue |
•31 |
•69 -31 |
•69 |
|
|
Franklin's Tale |
•27 |
•73 -17 |
•83 , |
|
|
Wife of Bath's Prol. ... |
•26 |
•74 -21 |
•79 |
1 As marked in Skeat's small edition; all his paragraphing is included in the second case, but a few cases in which he seemed to paragraph wrongly are disregarded in the first.
2 Some time after these tables were compiled, Mr. Pollard expressed the hope that some such study might be made (Knight's Tale, 1903, p. xx.). But he will see that the prospect is not very encouraging for "metrical tests" of chronology. Cf. also Lowes, Publ. Mod. Lang. Assoc., xr. 811-12. It will be easily seen by looking at these figures that they cannot be used, as at first one might have fancied, as a test of dates. Sq. T. and Sumn. T. on the one hand, and W. B. P. and Frankl. T. on the other, all of which are certainly late, stand at opposite ends of the list, and the Legend of Good
ifomen and the Prol. of the C. T., which I am convinced are nearly contem- poraneous, are also a long way apart. In Chaucer's case there is not the same reason as in S'hakspere's for a steady metrical development. Nor does the character of the poem seem to determine the proportion. It seemed at first as if it might, for in Pard. T. sindC. T. Prol., when the actual narrative begins and the more epigrammatic descriptions and moralizings stop, the propor- tion of run-on couplets rises. But this is at once outweighed when we notice that W. B. P., the most colloquial poem Chaucer ever wrote, is quite at the bottom of the list. In Romania, xxiii. 1-35, there is an important article by Paul Meyer which deals with this fracture in the Old Frencli couplet. According to him, Chretien de Troyes was the first important poet to break up the couplet in this way, and it was largely due to his influence that this great improvement in narrative verse became common.
62 POEMS DEPENDENT ON THE TESEIDE. [CH. Ill, § 1
It may seem at first as if this table contained but little evidence against the stanza-theory ; but certainly the position of the Knight's Tale is very far from being as peculiar as that theory requires. In the all-breaks list (the more important) it is the seventh of twelve, with 40% of run-on couplets as against extremes of 63% and 26% ; in the paragraphs list it is ninth of twelve, with 30% as against 59% and 17%. It is less than the average to the extent of 8% in the latter list and 2% in the former. When, once more, we "consider the conditions under which the transformation from stanzas into couplets would naturally have taken place, this is certainly a considerable argument against the stanza-theory.
We may look at this matter from the converse side, and inquire whether the English representatives of Italian 1st and 8th lines are more apt than we should otherwise expect to follow or precede (respectively) a full stop. It will be remembered that, to judge from the Troilus, \\ and f of these two lines (respectively) would have been the first and last in English stanzas. In the Troilus, book I., only three stanzas do not end in a full stop (numbers 25, 104, 106). It would be rather strange, therefore, if in a couplet poem reconstructed from a stanzaic a very large majority of these lines did not show this evidence of their earlier history. On the other hand, in a poem taken directly from one in the ottava rima, unless it were condensed more and quite otherwise than the Knight's Tale is, we should expect full-stops before and after these lines (respectively) considerably oftener than not, simply because the first and last lines of the Italian stanza introduce and conclude periods in the thought. The actual figures are these : of the representatives of Italian 1st lines, 53 follow a full stop and 45 do not ; of the representatives of Italian 8th lines, 30 precede a full stop and 25 do not. Considering the ease with which the last part of the stanza could have been transferred unchanged, these latter figures are amazing if the stanza-theory is right. In both cases the figures make distinctly against it.
Just one more such test may be given. Since the Italian first and last lines would almost always have become the first and last (respect- ively) in the English stanza, and since the last in an English stanza is the second in a couplet, and the first follows a complete couplet, and, therefore, would naturally form the first line in the next (if the stanzas were transformed as we have supposed), the representatives in the Knight's Tale of Italian 1st and 8th lines ought to be almost
OH. Ill, § 1] THE PALAMON AND ARCITE : ITS ORIGINAL METRE. 63
always respectively the first and second in their couplets. If the stanza-theory is incorrect, we should expect this to be so usually, for the reasons indicated in the last paragraph ; but to nothing like the same extent, since in a translation into couplets the Italian stanza is not transformed into a similar unit. A quick way of ascer- taining how often this is the case is to notice when the numbers of representatives of Italian 1st lines are odd, and those of represent- atives of Italian 8th lines are even.1 For the first line the figures are 33 even to 65 odd, and for the last 14 odd to 41 even. These figures are pretty much what we should expect if the stanza-theory is not correct, and harmonize well with the results of the last test.
So much for attempts at finding the supposed original stanzas in outline ; they certainly have not been successful. It is apparent that in the nature of the case this class of evidence could not be as striking as the first, but it has added some little weight to the negative argument.2
For the third class of evidence we may examine passages several lines long which are close to the Italian and therefore would have occurred, in much the same shape, in the Palamon and Arcite. Here, if anywhere, the supposed stanzaic form ought not to escape careful scrutiny and comparison with the original. First, I give the one passage which comes nearest to harmonizing with the stanza- theory, and then two or three which most strongly refute it. In a note I shall refer to about twenty more such cases, almost all of which are hostile to the theory.
And if ye wol nat so, my lady swete, E se t'e grave ci6 ch'io ti dimando Than praye I thee, to-monve with a Far, fa'che tu nel teatro la spada
spere
That Arcita me thurgh the herte Primaia prendi, ed al mio cor for- bere ; ando,
Costrigni che lo spirto fuor ne vada Con ogni vita il campo insanguin-
ando ; Che cotal morte troppo piu m'
aggrada, Thanne rekke I uoght, whan I have Chfe non sarebbe senza lei la vita,
lost my lyf,
Though that Arcite winne hir to his Vedendola non mia, ma si d' Arcita wyf (2254-8). (VII. 49).
1 If an Italian line is rendered by more than one English line, in the one case of course we should look at the first English line and in the other at the last.
2 One or two attempts, hardly worth describing in detail, to find vestiges of the rhyme-scheme of the ottava rima end in the same way.
64 POEMS DEPENDENT ON THE TESEIDE. [CH. Til, § 1
These five lines form a unit, and might easily have been trans- formed from a stanza by the omission of two lines, before and after the third ; yet it is difficult to fancy what the first of these two lines could have been.1
The following three passages are strongly opposed to the stanza- theory :
For from his feet up to his brest was
come The cold of deeth, that hadde him
overcome.
And yet more-over, in his armes two The vital strengthe is lost, and al
ago- Only the intellect, with-outen more, That dwelled in his herte syk and
sore, Gan faillen, when the herte felte
deeth, Dusked his eyen two, and failled
breeth (2799-2806).
La quale in ciascun membro era
venuta Da' piedi in su, venendo verso '1
petto,
Ed ancor nelle braccia era perduta La vital forza ; sol nello intelletto
E nel cuore era ancora sostenuta La poca vita, ma gia si ristretto
Eragli '1 tristo cor del mortal gelo,
Che-agli occhi fe' subitamente velo (X. 111).
Obviously this passage, if any, would have formed a stanza, but so far from its first part showing any signs of alteration, it is rather nearer the original than the latter part.
And after this, Theseus hath y-sent After a bere, and it al over-spradde With cloth of gold, the richest that
he hadde. And of the same suyte he cladde
Arcite ; Upon his hondes hadde he gloves
whyte ; Eek on his heed a croune of laurer
grene, And in his hond a swerd ful bright
and kene. He leyde him bare the visage, on the
bere, Therwith he weep that pitee was to
here (2870-8).
The case here is exactly the same. Considering that there are scarcely any other passages which correspond so nicely to an Italian stanza, these two are striking.
El fece poi un feretro venire Reale a se davanti, e tosto fello D'un drappo ad or bellissimo foniire,
E similmente ancor fece di quello II morto Arcita tutto rivestire, E poi il fece a giacer porre in ello Incoronato di frondi d'alloro,
Con ricco nastro rilegate d'oro (XI. 15).
Duk Theseus, with al his bisy cure, Caste now wher that the sepulture Of good Arcite may best y-maked be, And eek most honurable in his de-
Quinci Teseo con sollecita cura Con seco cerca per solenne onore Fare ad Arcita nella sepoltura ; Ne da ci6 '1 trasse angoscia ne dolore,
1 The other cases least inharmonious with the stanza-theory are 1999-2010 (= Teseide, VII. 34), 2011-6 (= VII. 35), 2334-8 (= VII. 91.)
CH. Ill, § 1] THE PALAMON AXD ARCITE : ITS ORIGINAL METRE. C5
And at the laste he took conclusioun, That ther as first Arcite and Palamoun Hadden for love the bataille hem
bitwene, That in that selve grove, swote and Ma pens6 che nel bosco, ove rancura
grene,
Ther as he hadde his amorous desires, Aver sovente soleva d'amore, His compleynt, and for love his hote
fires, He wolde make a fyr, in which Faria comporre il rogo dentro al
thoffice quale
Funeral he mighte al accomplice L* uficio si compiesse funerale
(2853-64). (XI. 13).
For several reasons this could hardly have formed two stanzas, or been expanded from one. Other similar cases I relegate to a note.1
The outcome of the examination of these passages is that in one or two of them there is nearly as much appearance of stanzaic form as could be expected or as there is against it ; but that in almost all of them, if the stanza-theory is correct, Chaucer must have taken the most extreme pains to obliterate vestiges which would have been apparent only on the minutest search. Whether this seems natural for any poet, especially a mediaeval poet, I will not say ; but I am quite certain that it is not like Chaucer. The result here harmonizes perfectly with those which have been gained by other methods.
This concludes my discussion of the stanza-theory. If it has been no less convincing than it has been tedious, I shall be satisfied. My tests, taken together, with all deference to the memory of ten Brink and to later scholars, I think show that the stanza-
1 905-15 = II. 26. 2314-21 = VII. 84.
1048-55 = III. 10. 2322-5 = VII. 85, 1-6.
1975-80 = VII. 31. 2349-57 = VII. 89.
1981-3 = VII. 32, 1-3. 2371-8 = VII. 23-24.
2221-6 = VII. 43, 2-8. 2410-18 = VII. 28.
2238-43 = VII. 46. 2423-34 = VII. 39-40.
2244-50 = VII. 47. 2561-4 = VII. 14, 1-4.
2275-80 = VII. 71. 2843-6 = XII. 6, 1-3,
2289-96 = VII. 74. 3017-26 = XII. 7.
2307-10 = VII. 81, 1-4.
(In line 2411, by the way, it looks as if Chaucer had mistaken compagnone for compagnia ; Arcite would be more likely to dedicate to Mars the arms of his former comrade and present enemy than those of his own followers.) Most of these passages might well be disregarded, but exhaustiveness makes the proof more conclusive. Every passage is given here or earlier in which the Italian offers guidance and which could reasonably be thought significant. I certainly have tried to treat the stanza-theory generously throughout, yet I have found nothing which is not perfectly natural if that theory is incorrect. The evidence which the passages in this note add to the main discussion is : — In favour of the theory, nothing ; against it, a considerable quantity of indications, singly small, but in the mass rather effective.
DEV. CH. F
66 POEMS DEPENDENT OX THE TESEIDE. [CH. Ill, § 2
theory is as destitute of evidence in its favour as it is of probability. Granted (and they do grant it) that Chaucer would have used his older version as the basis of the Knight's Tale, if not one sign of it can be discovered even where concealment would be equally difficult and unnecessary, I think the stanza-theory may be regarded as disproved.
§ 2. The Knight's Tale : How Far Altered.
We come back, therefore, to the position of Tyrwhitt and other early scholars,1 since if the Palamon was not written in stanzas it must have been written in its present metre. All we know is that a poem on the subject of " all the love of Palamon and Arcite of Thebes " was written before the Legend of Good Women, and that such a poem exists as the Knight's Tale. The question next arises whether in its first form it was practically the Knight's Tale as we have it, or whether it has undergone considerable revision and abbreviation. That some slight changes must have been made of course all are agreed. A passage near the beginning, lines 889-892, which allude to the pilgrims and the supper, must be new, and probably the whole paragraph 875-892. At the end there is nothing which must be new except the very last line, a benediction on the " fair company : " yet the ending is so brisk and succinct that it gives countenance to my belief that the poem was never finished in its original form and that the whole present ending was made for the Canterbury Tales. Elsewhere I find not the least indication of adaptation or alteration.
It will probably be felt, however, that there are some grounds for believing that the poem was originally much longer than now. There is a certain force in the analogy of the Troilus ; 2 if in translating one poem of Boccaccio's Chaucer made it half as long again, it might seem a little strange that he should reduce another to one-fifth. Yet on examination this argument loses most of its force. In the first place, consider the dissimilarity of the Troilus and the Knight's Tale. The one is a study of the human heart, with only so much incident as is necessary to keep it working and changing, a study on which Chaucer poured out all his interest and sympathy, and which
1 Cf. Ebert (1862), Jahrbuch fur rom. u. engl. Lift., \v. 95 ; Kissner (1867), Chaucer in seinen Beziehungen zur ital. Lift., p. 59.
2 So ten Brink, Studien, 43-4 ; Kissner, 1. c., p. 59. Cf. Mather (Furnivall Miscellany, p. 312, note), one or two of whose arguments against this objection agree with mine.
CFT. in, $2] THE KNIGHT'S TALE: HOW FAR ALTERED. 67
it is evident that he regarded as his great work. The Teseide, like the Knight's Tale, is a brilliant romance of picturesque incident, with little and weak emotional interest, the sort of thing which also appealed to Chaucer, in a more superficial way, Avhich he would be instantly moved to condense, and of which he would more readily tire. That he even tired somewhat of the Troilus I think there is evidence in the abruptness with which both Troilus and Criseyde disappear.1 What is more natural than that, after working long on one poem of Boccaccio's, within a few years he should turn to another by this same poet whose style he admired, a poem which he had known for years and had already quoted from ; but that from the start he should condense it 1 While the Filostrato has only about 5700 lines, the Teseide has nearly 10,000. As it is, the Knight's Tale is by a good deal Chaucer's longest single poem except the Troilus?
The general similarity in style between the Knight's Tale and most of the Canterbury Tales may appear to some readers a reason for thinking it largely transformed from its early form. But, to say nothing of what I believe to be the fact, that Chaucer's style after 1373 varied rather with his subject than with the date, we must remember that it is a question of only a few years in any case, and in general what a man can do at forty-five or fifty he can vdo at forty or forty-five. If it is a better poem than the Legend of Good Women, this is because Chaucer threw himself into it with greater
1 It clearly prolonged itself beyond his expectations. He meant to finish it in the fourth book, as he himself announces (IV. 26-8). The contrast in the ratio between the earlier and later parts of the Troilus and of the Filostrato is very striking. T. C., I. -IV. contain 6370 11. ; Fil., I. -IV. (which exactly correspond) contain 3688 11. T. C., V. has 1869 II., and Fil., V.-IX. (which correspond) have 2016 11.
2 One who compares the Knight's Tale with the Teseide will frequently wonder at the good passages which Chaucer omits (some of them are collected in Appendix C, pp. 231-3), and will perhaps wonder if their absence from the Tale is not due to revision. But if it has been shortened this was certainly done by small omissions all through ; the longest passage which he omits, book I., dealing with the war of Theseus with the Amazons, is so remotely connected with the rest of the ] oem that he doubtless omitted it from the first (as ten Brink and apparently Koch believe ; Sttidien, 62-3 ; Engl. Stud., i. 282). Chaucer must have had too much taste to cut out these good touches ; why did h« not reduce as well (or instead) a con- siderable number of needless and disproportionate couplets and longer passages which an attentive reading will discover in Kn. T. ? I may mention the first 150 11. or so, or such a speech as Theseus' at the end (2987-3066), which is of about the same length as its original in the Teseide ; such a couplet as 2087-8, or the passage where Theseus decides on a site for Arcite's funeral-pyre (2863- 64). There are some bits in the Knight's Tale which are distinctly verbose.
68 FORMS DEPENDENT ON THE TESEIDE. [CH. Ill, § 2
interest ; he really cared but little for the Romans and Greeks,1 and in a retelling of the stories of Ovid and the others, with their pale unintelligible background and the impossibility of making them over to suit himself, he had no chance to do his best work. Even as it is, not a trait of character or style appears in the Knight's Tale that is not also in the Legend of Good Women ; fresh love of nature, occasional levity, humour, satire,:.his own " favourite line,""Pitee renneth sone in gentil herte " (L. G. W., F, 503 ; K. T., 1761), and a remarkable number of other correspondences in expression (which will be mentioned later). It can hardly be supposed that Chaucer would have altered the Palamon quite completely; yet these characteristics run all through the Knight's Tale. The one legiti- mate deduction which I think we may draw from all these facts is that the poem in its first form was written only shortly before the Canterbury Tales and the Legend ; of which suggestion more hereafter.
Dr. Koch discovers another evidence of revision : " Die schil- derung des Marstempels tragt so sehr das geprage einer Uberarbei- tung, dass es kaum einemzweifel unterliegen kann, dass auch dieses stiick ein — wenn auch durch das verschiedene versmass mehrfach modificirter — theil der friiheren redaction des Palamon und Ar- citas ist." 2 Of the only two of his points which need be mentioned here, one (the use of the first person in the description) has been dealt with already ; the other is the fact that Chaucer seems to confound the portraiture on the wall of Theseus' oratory with the real temple of Mars in Thrace. But Dr. Koch seems to forget that the inconsistency is no greater in the KnigJit's Tale than it would have been in any form of the Palamon which we can postulate ; for he himself does not believe that Chaucer ever repre- sented the real temple as visited by Arcite's personified prayer. Furthermore, is not such an inconsistency far more likely to occur in an unrevisecl poem than to have survived revision ? It is easy enough to explain. Wishing to preserve as much as possible of Boccaccio's imposing and terrible description, he conceived on the walls of the oratory pictures of both the outside and the inside 3 of the Thracian temple, and even of the designs with which it was
1 Of course the people in Kn. T. are Greek only in name.
2 Engl. Stud., i. 258 ; cf. 258-61. This point is one of the more unintelli- gible parts of an unintelligible theory.
3 " Al peynted was the wal, in lengthe and brede, Lyk to the estres of the grisly jilace " (197Q-1),
CH. in, § 2] THE KNIGHT'S TALE : HOW FAB ALTERED. 69
" istoriato " (VII. 36). He even went so far in vividness as to describe the storms, the shaking of the temple and the shrieking,1 partly perhaps because he was carried away himself. This is not scientific but poetic description, and is simply carrying a little further the sort of imagination which we find in Keats' Ode to a Grecian Urn, and even in Chaucer's original description of the shrine of Venus, where broken sleeps, sighs and oaths are painted on the wall.2 The passage I think has no bearing on the question of revision.3
1 1979-80, 1985-6, 2004.
2 Kn. T., 1920-4. There is a much worse example of the same sort of thing in the House of Fame, mentioned earlier ; after describing the niches and statues on the outside of the palace, Chaucer goes on :
" Ther herde I pleyea on an harpe That souned bothe wel and sharpe, Orpheus ful craftely" (1201-3).
3 Dr. Koch sees evidence of a confused text in three small passages in this description. " Shippes hoppesteres " (2017) has been satisfactorily explained (see Skeat, V. 80-1). The connection with Mars of such undignified figures as "the barbour, and the bocher, and the smith" is due to the usual mediaeval identification of the pagan god with the planet and its astrological relations. Finally, Koch objects to the lines
" The sleere of him-self yet saugh I ther, His herte-blood hath bathed al his heer ; The nayl y-driven in the shode a-night" (2005-7) ;
he suggests an impossible emendation ("The housbond slain by his wif") which is modified and approved by Skeat (V. 80). It would be extraordinary to mention the driving of the nail after the flow of blood, and the emendation would destroy the force of the allusion in 2007, for Sisera was not the husband of Jael. The passage is perfectly simple if we divide it into two images, of which the first suggested the second. There are, however, certain real internal inconsistencies in the Kn. T., all due to careless treatment of the original (and more likely to occur in an unrevised than in a revised work). In 2355-7 Diana says to Emily :
" The fyres which that on myn auter brenne Shul thee declaren, er that thou go henne, Thyn aventure of love,"
although the performances of the fire have already been described (2334-40) ; here Chaucer has kept the future tense of the Italian, though he has reversed the order (Teseide, VII. 89, " vedrai" ; cf. 91-2). In 2858 ff. Theseus cuts down the grove and makes Arcite's pyre and tomb on the scene of the sylvan combat, regardless of the fact that he had previously erected a vast and sumptuous theatre on the same spot (1862); here Chaucer has followed the Italian in the later instance and not in the earlier. Finally, Theseus speaks of Arcite's " cosin and his wyf " (3062), though Emily had not married Arcite ; in the Teseide she had already done so. An apparent blunder in 2046 is due only to Skeat's punctuation ; the line looks not forward but back, and ohould be followed by a period. Professor Liddell (Chaucer's Prol., etc., N.Y. 1902, p. 169) says that the promise in line 2039, " Suffyceth oon ensample in stories
70 POBM8 DEPENDENT ON THE TESEWE. [CH. HI, §2
One bit of internal evidence tends to disprove the idea of exten- sive revision. After mentioning Theseus' following Pirithous to hell, Chaucer says (1201) :
" But of that story list me nat to wryte."
Nobody has doubted that the Second Nun's Tale, mentioned in the Legend of Good Women, and written, by an " unworthy son of Eve," to be read, was taken over unchanged into the Canterbury Talcs. Is not the case nearly as good for the story of Palamon and Arcite, also mentioned in the Legend of Good Women, also written for readers, and only known to have been adapted for the Canter- bury Tales at the beginning and the end 1 l
But the strongest argument against much alteration of the Palamon is that of probability and Chaucer's usual practice. In the Second Nun's, Shipman's and MercJiant's Tales (as we shall see later) Chaucer neglected very necessary revisions. The revision of the Prologue to the Legend I shall try to show was due to a very special cause. It is a fair presumption that Chaucer avoided need- less trouble in adapting the Palamon for the Knight. There is no reason or evidence for the belief that the original form of the poem was different from the present, or that if it had been Chaucer would have felt called on to alter it. The indications are therefore very strongly in favour of the practical identity of the Palamon and Arcite with the Knight's Tale.
§ 3. The Knight's Tale: The Date. All this is a long preamble to a discussion of the date of the
olde," is not fulfilled ; but Julius, Nero and Antonius, who have been spoken of a little way back, answer very well to the "slayn or elles deed for love," and the " oon emample " no doubt refers vaguely to them. Dr. Mather (Furnivall Miscellany, 303, n.) suggests "error or negligence" in 2914-5 ; but is not this a not veiy violent case of metonymy, paralleled only five lines below (2919) ? The peculiarity of the passage is accounted for by the fact that Chaucer has transferred to the pyre the language which Boccaccio uses of the grove which was cut down to make it (XI. 18, 19) ; an example of Chaucer's rather lax style of translation.
1 The point was first noticed by Holthausen (Anglia, viii. 453), who, however, did not see the full bearing of it, — "bei der umarbeitung hat der dichter unachtsamer weise dies iiberbleibsel der ersten redaktion stehen lassen;" and by Dr. Furnivall (Temp. Pref., " corr. and add."). It should be noted that in the prologue to Melibeus (2153-4) Chaucer speaks of the "tretis lyte After the which this mery tale I wryte," a similar oversight, which perhaps weakens the argument a little.
CH. in, § 3] THE KNIGHT'S TALE: THE DATE. 71
Kniyltt's IWe^but it is all essential to the subject, and has already thrown considerable light on it. I shall try to show that the tale was written between the Troilus and the Legend. A position after the Troilus has been assigned it hitherto only by Pollard2 and Mather,3 simply because almost everybody else has held the stanza- theory.4 A later date than that of the Troilus can hardly be denied if my date for the latter is accepted, since (among other reasons) it is impossible to put two such long and elaborate poems as the Troilus and the Knight's Tale between 1373 and 1377. The most important argument for the inverse order is that of Pro- fessor J. L. Lowes,5 whose opinion that the Troilus was written just before the Legend involves the priority of the Knight's Tale. I have already endeavoured to dispose of his arguments for this position for the Troilus. It remains to meet those for the priority of the Knight's Tale.
Lowes first points out the curious fact that it is on the 3rd of May that Pandarus6 has a particularly sharp attack of love, and that Palamon escapes from prison ; 7 and very naturally believes that one case must be due to the other. That the choice of this date was made first in the Knight's Tale he thinks is shown by the sup- posed facts that there is no reason for it in the other case, but that here it is "an essential part of the carefully calculated scheme of days and astrological hours on whose every step explicit emphasis is laid in the poem." Now Lowes' argument may be made to refute his own view. In the first place, the date in the Knight's Tale can be shown to be perfectly arbitrary, and not at all an essential part of the scheme. The essential parts are the hours, and the days of the week, which are wholly independent of the days of the month, and this is the only point where a day of the month is mentioned. Aside from the improbability that Chaucer's whole scheme was already devised at this point in the poem, where it first
1 Hereafter this term may be used interchangeably with Palamon and Arcite.
2 Apparently ; see Globe Clutucer, p. xxvii.
3 Furnivall Miscellany, p. 309 ; Chaucer's Prol., etc., p. xvii. He thinks, for no very clear reason, that Chaucer put the Teseide passages into the revised T. C., and into P. F., after " the plan of Palamon (Knight's Tale) was com- plete " (Prol., etc., xix., note ; cf. 102, n. Cf. also Misc., 309, 310, 312).
4 Cf., e.g., Koch, Chronology, p. 30.
5 Publ. Mod. Lang. Assoc., xx. 841-54.
6 Ibid., xx. 842-3 ; Lowes says Troilus, by a slip.
7 T. C., II. 56-63; Kn. T., 1462-8. The detail is in neither original. Cf. Pollard, Knight's Tale, p. 89.
72 POEMS DEPENDENT ON THE TESEIDE. [CH. Ill, § 3
begins, there is absolutely no reason whatever why he should have chosen this date unless it came into his head for some outside reason.1 But for the selection of this date in the Troilus there is a reason, though a homely one. The passages in the Tale and the Troilus run thus :
" It fel that in the seventhe yeer, in May, The thridde night" . . .
" it so betidde
As I shal singe, on Mayes day the thridde."
It should not be thought a criticism unworthy of a great poet if I suggest that Chaucer chose the word thridde for the sake of the rhyme. There are a large number of such cases in Chaucer's poetry, and some fairly important ones, as Lowes himself points out only thirteen pages later.2 If Professor Lowes will pardon me, I will sum up my argument in his own words ; "if in one of the poems the employment of the third of May is directly dependent upon certain exigencies of the treatment of the material itself, while in the other its relation to the story is wholly acci- dental, we may be practically certain that the instance which grows out of the requirements of the story came first, and that it naturally enough suggested the other."
Lowes argues (pp. 850-2) that the character of Boccaccio's two poems would- make it likely that Chaucer should translate the Teseide before the Filostrato. The former may well have been a part of his first introduction to Italian literature,3 but that he would translate it first does not at all follow. Lowes' argument that " an earlier attraction to the Teseide than to the Filostrato is what we should naturally expect," because the interest of the former is in superficial narrative and of the latter is in profoundly human feel- ing,— this argument, I say, seems to me a little odd. We must once more remember that the man Chaucer at his first going into Italy in 1373, at the age of thirty-three or so, must have been far more mature than the poet Chaucer who had written the Book of the Duchess only a few years before. Surely Dr. Lowes would not say that he who was capable not more than at most ten years later of writing the Troilus must have been at first more attracted to the lesser
1 See pp. 81, 82 below for a fuller treatment of this scheme and its value for dating K. T.
2 See p. 855, where he refers to T. C., V. 1788, 1797 ; L. Q. W., F, 328.
3 We have seen that he shows familiarity with it even in the first version of the Troilus ; cf. p. 49 above.
CH. in, § 3] THE KNIGHT'S TALE : THE DATE. 73
poem. It seems to me that to a man of his age and tastes — ^consider that his two greatest character-creations are of women, Criseyde and the Wife of Bath, — the Filostrato would have appealed especially and at once. Moreover, it would have seemed a less enormous task, and his experience with the Romance of the Rose had prob- ably already taught him the uncertainties in beginning on a poem of great length. He would have begun to work on the Filostrato with no intention of expanding (I have already pointed out that he meant to finish the Troilus in the fourth1 book). After his experience with the Troilus, it is not surprising that he greatly condensed the Teseide from the first.
Nor do I find any more convincingness in Lowes' argument (pp. 852-4) that the style and manner of the Troilus and Criseyde and of the Knight's Tale would make the latter the earlier. I must say again what I have said elsewhere, that Chaucer's style and manner, after his return from Italy, it seems to me depended very much more on the character of the poem he was writing than upon the period,1 though the former often depended on the latter. It seems to me that the argument from style is a very, very dan- gerous one. He treated the Teseide freely because he wished to condense that excellent but lengthy poem ; yet he made less change in its characterization 2 than in that of the Filostrato, because the characters are less important and naturally interested him less. Lowes himself lays great stress elsewhere on the fact that the centre of gravity in the Troilus is psychological; why should a brilliant romance of incident be -expected to compete with it in regard to characterization ? Was A Winter's Tale written before Hamlet ? Dr. Lowes thinks we should hesitate to put Emily and Arcite and Palamon and Theseus later than Criseyde and Pan- darus. But there is no reason why we should confine ourselves to comparing the Troilus with Chaucer's other Boccaccian poem. How about Dorigen and Aurelius and Arviragus? How about Canacee and Griselda and Constance ? 3 Lowes' argument, if carried to its logical conclusion, would make the Troilus the last of Chaucer's long poems. Lowes' comparison of the Troilus to the
1 Cf. what ten Brink has to say (Studien, p. 44) in reply to a remark of the usually judicious Kissner (Chaucer in seinen beziehungen zur ital. lit., p. 65).
2 But he did make a rather striking change in the characters of the cousins ; see Appendix C, pp. 231-2.
3 I hope to show later good reason for the belief that the Tales of the Clerk and the Man of Law are late poems.
74 POEMS DEPENDENT ON THE TESEIDE. [CH. Ill, § 3
Knight's Tale in regard to the idea of fate expressed in it, and to its greater suggestiveness, I think may be answered in the same way. How could the Knight's Tale have been treated in the same manner as the Troilus and Criseyde ? Which is more suggestive, Hamlet or A Winter's Tale ? These considerations which Pro- fessor Lowes adduces, it seems to me, have no argumentative value whatever.
Lowes l hardly does justice, I think, to the argument from the presence in book V. of the Troilus (11. 1807-27) of the stanzas which describe the flight of Troilus' soul to heaven, for which- in the Knight's Tale Chaucer makes a rather flippant substitution (2809-15). It is natural to see, as almost all critics do see, a parallel here to Chaucer's insertion in the Knight's Tale of an inferior and original description of the temple of Venus, because he had already used Boccaccio's description in the Parliament of Fowls. The best explanation of the peculiar character of the passage in the Knight's Tale about Arcite's soul, in which Chaucer professes utter ignorance as to what became of it, is that he is gently mocking at Boccaccio.2 It is hard to believe that he not only used but went out of his way to fit into a later poem a passage which he had rejected with some- thing like contumely from an earlier, unless there shall prove to be a very striking contrast in fitness between the two cases. This Lowes seems to think exists, but I cannot see it. The Troilus is a much more thoughtful and skeptical poem than the Knight's Tale ; why should this skeptical attitude toward the other world appear so spontaneously in the latter rather than in the former ? If this is why Chaucer omitted the passage from the Tale, it is doubly odd that he put it into the Troilus ; but if he had already used it in the Troilus, the gently joking manner of its analogue in the Knight's Tale seems quite intelligible. The striking thing is that he should omit the whole passage in the Knight's Tale, though, however inharmonious some parts of it might be with what precedes or follows,3 parts of it would do perfectly well, and though before and after it he is following the Teseide closely. The Knight's Tale is much less realistic and contains much more of the supernatural than the Troilus. I cannot but feel that the stanzas would be a little
1 Pp. 843-7.
2 For Lounsbury's strange opinion that Chaucer is here expressing ' ' agnostic " views, see his Studies, ii. 513-15. A still different interpretation is that of Dryden in the Palamon and Arcite.
3 Cf. Pollard, KnigMs Tale (Macmillan), p. 116.
CH. in, § 3] THE KNIGHT'S TALE : THE DATE. 75
out of place in either poem ; if lie had once weighed them and found them wholly wanting, it is passing strange that he used them later. Therefore the indications are that the Troilus was not only written, but also revised before the Knight's Tale was written.1
A more forcible argument for the priority of the Troilus seems to me to be that from metre. If Chaucer had been familiar with the possibilities of the couplet, it seems to me hard to believe that he would have written such a poem as the Troilus in the melodious, but difficult, wordy and languid stanza.2 Lowes thinks otherwise.3 But it is one thing that Chaucer should return later to this sweet, romantic and half-lyric form of verse for such poems as the Tales of the Prioress, the Clerk, and the Man of Law, and quite another to imagine his returning to it for one of his great realistic and dramatic creations, for which the simplest and most flexible of mediums would be the most suitable, for which he might well have used blank verse if he had known it; as well revive the seven-line stanza in the Wife of Bath's Prologue. Far be it from me to underestimate the skill with which he uses it in the Troilus, but I am sure that Chaucer would have felt at once that the other form would have been more suitable ; just as Shakspere and Dryden, though they may have been sensible that they could write good dramatic dialogue in the ten-syllable couplet, came to prefer the simpler and freer blank-verse. When Lowes argues that though Chaucer had already written the Knight's Tale in couplets, he had not shown its potentialities for presenting dialogue and shifting moods, and that therefore for the arduous task of the Troilus he returned to the more familiar instrument, I believe he is misled by a metaphor. We have already seen that it is merely the presence of the second line which distinguishes the stanza from three couplets. This line completely alters the effect of the stanza and adds very considerably to its difficulty ; but hardly makes it a different instrument. An accomplished pianist might well hesitate to perform in public on the organ, but why should
1 Cf. p. 15 above.
2 Koch(Engl. Stud.,-x*.vii. 3-4) uses the metrical form and free treatment of the Tale as an argument against putting it early in its present shape. Of course his conclusion is that the original form was very different ; if this is not so, he gives unintentional support to the view expressed above. (He is unjust to Pollard in implying that he puts it before T. C. ; cf. Globe Chaucer, p. xxvii.)
» Pp. 847-50.
76 POEMS DEPENDENT ON THE TESEIDE. [cH. Ill, § 3
a poet who felt himself thoroughly at home in the stanza distrust his own ability to manipulate it with the second line gone ? This seems to me to attribute extraordinary diffidence to Chaucer, especi- ally if he had already written over a thousand admirable couplets in the Knight's Tale. Even supposing he had written none, and supposing it might take him longer to produce a satisfactory passage in a new form of verse than in the old, with the same exacting taste and judgment the final result should be as satisfactory in the one as the other. But, more important yet, Lowes makes a rather curious oversight ; he tells us that when Chaucer began the Troilus the stanza was an instrument " whose stops he knew from its lowest note to the top of its compass," while the couplet was a " less tried medium." Yet, even if we accept Lowes' very late date for the Troilus, 1 383-5, a the only poems, • so far as we know, which Chaucer had then written in the stanza were the Parliament, the Second Nun's Tale, the Complaint to Pity, a part of the Com- plaint of Mars (perhaps), the Anelida, and a few short poems — at most perhaps 1800 lines; yet the Knights Tale, as it stands, contains 2250 lines, and surely nobody can deny that it shows far more mastery than these stanza-poems do, especially in the sort of manner required in the Troilus. Yet Dr. Lowes would have us believe that Chaucer felt very much more at home and self- confident in the more difficult and less-used 2 form of verse.
To the best of my belief this disposes of all the evidence which Professor Lowes adduces. It seems to me, therefore, even apart from the very early date for the Troilus which I have defended, that the probabilities are strong that the Knighfs Tale followed the Troilus. We may now consider certain other arguments on the date of the Knighfs Tale.
A clear indication that the Knight's Tale comes between the Troilus and the Legend may be found in the very large number of similar or identical phrases and lines in the Tale and one or the other of these two poems.3 It is well known that in almost every one of Chaucer's poems there are reminiscences of the phraseology of others ; it is clear that he had a vivid verbal memory, and had not
1 Pp. 860-1 ; a date later than that proposed by any other writer.
2 If we accept Lowes' opinion that most of the Legends were written before the Kn. T., the disparity is far greater. And even if we then should add the Clerk's and Man of Law's Tales to the opposite scale, the disparity is still almost exactly the same as at first.
3 Cf. Pollard, Kn. T. (1903), p. xii.
CH. in, § 3] THE KNIGHT'S TALE: THE DATE. 77
the least objection to using a good thing twice. In each of these poems there are such links to a number of Chaucer's other works, but those between the Knight's Tale and the other two are so much more numerous that it is fair to allow them considerable significance. It has been made plain, I trust, that they cannot be explained as having come in when the poem was being adapted to the Canterbury Tales. Since the passages are too numerous to quote in full, I merely give the references, first of those mentioned by Skeat,1 then of some which I add. Those in parentheses are the less important; those marked with a t are due to originals which are in the Teseide or the Filostrato but are not close enough to have suggested the English expression ; a J indicates that the Italian is very close.
Kn. T. and T. 0. Kn. T. and L. G. W.
(925 = 4, 2)t (1035-6 = 2425-6) tlOlO = 4, 627 (1196 = 2282)
(1047 = 2, 112) (1302 = 866) 1101 = 1, 425-6J 1502 = 1204 2 1133 = 1, 674 1566 = 2629
(1155 = 5, 382)t 1761 = 503 (F), 491 (G) 3 1167-8 = 4, 618t (2235 = 2132) (1401 = 4, 865)t f2602-20 = 637-53 (1500 = 2, 112) 1509 = 2, 920 1566 = 3, 733-4 1838 = 5, 1433 2449 = 4, 1456 1(3042 = 4, 1586)
873-4 = 1210-1
1057 = 937 1462-3 = 2, 56 4 t!060 = 1962
(1809 = 4, 1567) 1164-6 = 1186-7
J(2203 = 2, 503) (1333 = 2604)
+2406 = 1, 21J 5 fl403-6 = 2046-7
t2529 = 4, 1086 (also 1079) 1423-4 = 1070-1 2991-3 = 3, 1762-4 (1531 = 1167)
3089 = 3, 1282 J2506 = 1208
2565 = 635 3089 = 162 (F)
1 III. 394, and in the notes on the passages. Cf. Notes mid Queries, 4th Series, IV. 292. Only Kn. T. 1566 and 3089 are paralleled in both the other poems.
2 For a little note on this line, see Henry Hinckley in Mod. Lang. Notes, xiii. 461-2.
3 Skeat says 1163 (wrongly ; III. 394).
4 See p. 72 above.
8 The line in the Teseide is : " lo il diletto, e tu n' abbi 1' onore " (VII. 27) ; in the Filostrato: " Tuo sia 1' onore, e mio si sia 1'affanno" (I. 5). The latter looks like the original of both Chaucer's lines. It is worthy of remark that in the more striking cases above where the Italian has suggested a line in K. T. or T. G., it is in the Filostrato— an argument for the priority of the Troilus,
78 POEMS DEPENDENT ON THE TESEIDE. [CU. Ill, § 3
Some of these parallels are small, a few are due to Boccaccio or Le Roman de la Rose, or are proverbial, and one or two are (rather rare) idioms. But the important thing is their number,1 which is far greater than that of parallels between any others of Chaucer's poems. Another striking fact is that there are hardly any such parallels between the Troilus and the Legend ; of the few which exist, two are in the Tale as well.2 It seems fair to say that these parallels suggest for the Knight's Tale a position between the Troilus and the Legend.3
A date after the Troilus will also be necessarily involved by the early date which I have assigned the latter ; we can hardly crowd anything long between Chaucer's first return from Italy and the commencement of the Troilus. This gives 1377 as the earliest possible date for Chaucer's working on the Knight's Tale. But on other grounds it will be necessary to put it much later than this.
In the first place, it must come after the Parliament of Fowls, since there is no longer any possible reason for thinking that the Teseide passage there ever stood in the Palamon. However it may be with the passage about Arcite's death, it is quite inconceiv- able that in the Palamon Chaucer should have substituted an original description of the temple of Venus for the far superior imagery of Boccaccio, unless he had used that in an earlier poem.4 Hence we derive 1381 as the earliest possible date for the Knight's Tale.
For this there is some confirmation in the style of the poem,
1 It should be remarked also that two-thirds of them are in contexts which are fairly close to the Teseide ; this in further answer to the possible conjecture that they came in on revision, which I have shown other reasons for disbelieving.
2 See pp. 19, 24 above.
3 Skeat, who of course holds the stanza-theory, sees the force of some of these parallels between the Troilus and the Tale, and makes the rather curious comment: "This tends to shew that the Knightes Tale (rather than the original Palamon and Arcite) was written not very long after Troilus ; rather in 1386 or 1387 than in 1388 " (III. 394). Of. Notes and Queries, 4th Series, iv. 292, for his earlier view. Dr. Mather also (Furnivall Miscellany, p. 308) says: "Somewhere near Troilus it must surely go, for the two poems agree notably in thought and in expression." But neither of these two writers pays any attention to the correspondences between the Knight's Tale and the Legend of Good Women, which seem entitled to equal consideration.
4 On this point I must strongly disagree with Dr. Mather (Furnivall Miscel- lany, p. 310). It is striking that for this new description he turned in part to a passage in an earlier poem of his own, the House of Fame ; the description of Venus (1955-66) is expanded, but otherwise almost word for word, from H. F., 132-9. In his note to the latter, Skeat erroneously speaks of the former as from Boccaccio.
CH. in, § 3] THE KNIGHT'S TALE : THE DATE. 79
which instantly links it to the Legend of Good Women and especially the Canterbury Tales, rather than to Chaucer's earlier works; the good judgment, the keenness, the aptness, the rapid alternation of humour and pathos, the general certainty of touch. The poem contrasts even with the Troilus, and resembles most of the Canterbury Tales, in its condensation and vigour and speed. Though the Troilus is a greater poem, to me at least it seems less artistic and finished, and less marked by most of the qualities just mentioned than the Knight's Tale.1 There are also certain favourite phrases in the Tale which occur again and again in Chaucer's later poems, and seldom or never in the earlier. The phrase "gentil herte " (Kn. T., 1043, 1761, 1772) does occur in the Troilus (IV. 1674); but it is much commoner later.2 Chaucer's "favourite line,"
" Pitee renneth sone in gentil herte,"
occurs only in Knight's Tale, 1761; Legend, F, 503 (G, 491); Merchant's Tale, 1986 ; Squire's Tale, 479 ; and in a close variant in Man of Law's Tale, 660.3 Again, no locution is more charac- teristic of Chaucer's later style than such elaborate phrases as " by aventure or sort or cas," which I have elsewhere shown to be probably due to reminiscences of Dante.4 They occur only in the Canterbury Tales, and of the six cases which I have noted, two are in the Knight's Tale.
A similar date is indicated by two probable contemporary refer- ences in the Knight's Tale. Saturn, among the results of his rnale-
1 For a different view cf. Kissner, Chaucer in sebum beziehungen zur ital. literatur, p. 65 : and cf. ten Brink, Studien, p. 44.
2 L. G. W., 503 (F), 491 (G) ; M. L. T., 660 ; Mem., 2832 (the Latin has "ingenui animi," the French "gentil cuer "), Merck. T., 1986; Sq. T., 452, 479, 483.
3 Professor Liddell (Chaucer's Prol., etc., p. 167) says: "This seems to have been a proverbial expression "; but it seems more likely to be a favourite invention of Chaucer's own. Mr. Paget Toynbee (Jvurn. Compar. Lit., i. 351) announces the line as a translation of Dante's "Amor che a cor gentil ratto s'apprende " (Inf., V. 100). But the only phrase which the two lines both have is very common, in Italian, in French and (as we have just seen) in Chaucer. Professor Francis Palgrave had already announced this supposed borrowing in 1888 (Nineteenth Century, xxiv., 349). [In Mr. Toynbee's article just quoted, in which he conveniently collects most of Chaucer's borrowings from Dante, he attributes (as Cary had done) L. G. W., 2638 to Inf., VII. 64 ; but he exaggerates the similarity by reading gold for gode, the only read- ing in the nine printed MSS. On this line cf. W. B. T., 1064-5.]
4 See Modern Philology, iii. 372. Such cases asN. P. T., 4291, Kn. T., 1242, 1506, 1516 (not mentioned there), less striking and Dantesque, are certainly commoner in C. T. than elsewhere.
80 POEMS DEPENDENT ON THE TESEIDE. [CH. Ill, § 3
ficent influence, mentions " the cherles rebelling " (2459) ; we can hardly avoid seeing a reference to the peasant revolt of June, 1381,1 since the introduction of the item (founded on nothing in the Italian) before that date would be difficult to account for. Professor Lowes, in a thorough and judicious article,2 throws light on both the date and a puzzling line in the poem. " The tempest at " Hippolyta's " home-coming " (884) has never hitherto been at all satisfactorily explained. Lowes shows that it is probably an allusion to a strange and destructive upheaval of the sea just after Anne of Bohemia had landed, on her arrival in England in December, 138 1.3 This indicates 1382 as the earliest possible date.
Finally, an indication that the Knight's Tale was written only shortly before the Legend of Good Women is the often-quoted couplet which has caused all our pains :
" And al the love of Palamon and Arcyte Of Thebes, thogh the story is knowen lyte."
We can no longer explain the last clause and the utter disappear- ance of the supposed older form of the Palamon on the ground that it had been published and failed ; we can explain both on no ground so reasonably as that Chaucer had never published it at all. This will explain why he seems to imply that the Legend was written in a new kind of metre, though he had been using the same in the Palamon. When we consider Chaucer's position, and how simple a matter publication was in his day, it is hard to imagine any reason for withholding the poem, except that it was not yet finished.4 We shall see later that the form of prologue in which the couplet occurs dates almost certainly from 1386. The above argument seems to me so cogent that I have little hesita- tion in adopting the date about 1385 for the writing of the greater part of the Knight's Tale.
In this late date I differ from the only two writers who have as
1 Also alluded to in N. P. T., 4584-6. Cf. Skeat, I. Ivi., and Walsing- ham, Historia Anglicana, i. 458, 462.
2 Mod. Lang. Notes, xix. 240-3.
3 Cf. the Monk of Evesham, Hist. Vttfe, et Regni Ric. II., p. 129, for an odd coincidence when Richard brought home his second bride.
. 4 The Palamon was scarcely a poem to be voluntarily neglected. I shall show later good reason for the belief that L. G. W. was written in some sense at the command of the queen. The conjecture seems plausible that Chaucer broke off his work on P. A. in order to write L. G. W.
:
CH. in, § 3] THE KNIGHT'S TALE : THE DATE. 81
yet abandoned the stanza-theory and discussed the date at length, Mather and Lowes, who suggest 1381-2. But it will be seen, I think, that their possible objections to my date can easily be met. The former1 puts it very near the Troilus because of the verbal similarities already spoken of. But if the latter was finished in 1377 or so, and the Knight's Tale refers to events of 1381, it is impossible to put them close together. I have already shown that Chaucer's revision of the Troilus, perhaps in 1380 or later, will help to account for the two having been together in his mind ; and his permanent and intimate familiarity with the Troilus is accounted for by the fact that he had written it more carefully and valued it more highly than any other of his works. Mather's belief that the Teseide stanzas inserted in the Troilus during the revision were so inserted while Chaucer was writing the Knight's Tale I have tried to show is highly improbable. I must relegate to a foot-note what seems to me proof positive that Skeat's calendar method of dating the Knight's Tale, of which Mather and Lowes approve, cannot possibly work. Mather argues further that if we put the Knight's Tale in 1381-2, where we know the Parliament of Fowls belongs, " the whole preoccupation with the Teseide would have extended over only a year or so, and certainly this supposition is better than that of its gradual dismemberment." To say nothing of the inappropriateness of this last phrase, we know that Chaucer made some small use of the Teseide years before in the first version of the Troilus, so in spite of us his use of the Teseide extended over at least six years or so. This answers, I think, all of Mather's arguments. Lowes2 has no arguments not already dealt with except the reference to the " tempest " ; this obviously implies a date after 1381, but not necessarily just after. The incident may well have sprung vividly to mind two or three years later.3
1 Furnivall Miscellany, pp. 308-10.
2 Publ. Mod. Lang. Assoc., xx. 841, ff.
3 Professor Skeat has made an ingenious attempt to find the date of what he considers, the revised Knight's Tale (Notes and Queries, 4th series, ii. 243-4 ; reprinted with alterations in his Chaucer, V. 75-6). Falamon escapes from prison early in the morning of the 4th May (1462-7), and the woodland combat therefore occurs the 5th May (1610) ; that the first of these days was Friday, Skeat thinks is suggested by the fact (according to him, but Chaucer does not say so) that Arcite goes a-Maying in the first hour, which on Friday is dedicated to Venus, and by the fact that Chaucer uses Friday as a symbol for the moods of lovers ; and that the second day was a Saturday, presided over by the unlucky planet Saturn, by the fact that the duel is interrupted. ( But is not all this reasoning rather too much as if it were history ; would Chaucer have thought of all this ?) The assembly before the tournament is to
DEV. CH. G
82 POEMS DEPENDENT ON THE TESEIDE. [CH. Ill, § 3
The best conclusions as to the date of the first writing of the Knight's Tale seem to be these. It is later than the Troilus, and even than the revision of it — hence much later than 1377; later than the Parliament of Fowls — hence later than 1381, as is further indicated by two probable historical allusions. The manner in which it is mentioned in the Prologue of the Legend of Good Women points plainly to its having been written very recently. Everything seems to harmonize with the date 1384-6.
As to its completion and adaptation to the Canterbury Tales, this probably took place not many years afterwards. It is well known, or we shall see later, that in the Tales of 'the Second Nun, the Parson, the Shipman, and the Merchant Chaucer neglected to make even such revisions as appropriateness strongly demanded. Now at the beginning of the Knight's Tale Chaucer made such changes as were certainly not in the least necessary. This points to a time when the Canterbury Tales were fresh to him. It is also suggestive that the Knight's Tale stands first in the series, and that the Prologue directly
be that day fifty weeks (1850-3) ; no doubt, as Skeat says, a year (though it is odd that Theseus says, "fifty wykes, fer ne ner"), for Boccaccio has "un anno intero," and it actually occurs not in April but in May (2484). Sunday (2188), the 5th May if it is a year from the first fight, the knights assemble for the tournament ; Monday they amuse themselves (2486) ; and the tourna- ment occurs the following day (2491), Tuesday, the 7th. Skeat thinks it not unnatural to suppose that Chaucer took the scheme of the year in which he was writing; and finding (correctly) that the second set of dates fits 1387, concludes that this may have been the year of revision. The question for us of course is not the year of revision — that Chaucer should have made such an elaborate adaptation of course is not to be thought of — but the year of first writing ; although, risky as the scheme is and as Skeat admits it to be, it might have some value if it fitted in with the other evidence, ten Brink rejects it as too conjectural (Studien, 189), and I fear we must reject it on other grounds as well. The striking fact that Chaucer chooses such an unobvious date as 3rd May for Palamon's escape I have shown to be explained probably by a reminiscence from the Troilus. There are really no striking coincidences to indicate that Chaucer had in mind from the start an elaborate scheme cover- ing a year, and Pollard shows that he was quite indifferent to the larger time- scheme of the poem (Knight's Tale, 1903, pp. 81-2). The most striking defect in Skeat's scheme is that it is the second of the years in the poem which he identifies with a current year ; if the scheme is as elaborate as he whom I fear we must call its author believes, it would be strange that Chaucer should not have made the first year fit the current one. This would give 1380 or 1386. The former of course is impossible, and the latter would inadmissibly crowd the Legend of Good Women and the Canterbury Tales. Therefore Skeat's clever scheme cannot be accepted. This is only one of several cases in which more recent scholarship has come to see that in the past we have attributed to Chaucer more care and accuracy in insignificant matters than he really observed. Many of these tempting methods of dating poems must be abandoned. In regard to minute accuracy, Chaucer goes with Shakspere, not with Dante.
CH. Ill, § 4] THE ANELIDA AND ARC1TE. 83
introduces it. We shall see later that Chaucer was probably busied with the Prologue about 1387, and that it was perhaps the very first Avritten part of the whole work. There is much in favour of the view that the Knight's Tale was the first Canterbury Tale to be meant for such, and that it was put into its present position soon after the writing of the Prologue, about 1388-90.1
§ 4. The Anelida and Arcite.
As to the date and interpretation of that perplexing poem the Anelida and Arcite we have been left to rather vague conjecture. Dr. Furnivall2 dates it between 1374 and 1384. Dr. Koch3 suggests 1383, between the demolition of the supposed stanzaic Palamon and its reconstruction in the Knight's Tale. Ten Brink4 thinks it may have been begun before the recasting of the Palamon was finished ; he is quite certain that the opening was derived from the first form of that poem. Mr. A. W. Pollard, in 1893,5 put it about 1380, and suggested "that it represents Chaucer's first study of the Teseide before he turned to the Filostrato" Dr. Skeat merely puts it after 1373, and after the Palamon, from which he believes the opening to be taken ; with the added suggestion that " Chaucer's thoughts may have been turned towards Armenia by the curious fact that, in 1384, the King of Armenia came to England."6 Dr. Lowes dates the poem about 1380-2.7 The best treatment of its genesis is that by Dr. Mather,8 who denies that the opening was derived from the Palamon, and (rather extremely) regards the Anelida as " the necessary middle stage between " the Troilus and the original form of the Knight's Tale (p. 310, note) ; it must therefore have been begun before the Knight's Tale. He also suggests " that Chaucer having completed
1 Another of the earliest-written tales is probably the Physician's; see pp. 155-6 below. There- is evidence that Kn. T. was known to the world before 1392. Two lines of it (1785-6) are quoted in the Book of Cupid (Skeat, VII. Ivii. ff. , 347 ff.), which Professor Kittredge shows some reason to believe was written before that date (Mod. Philol., i. 13-15). This and one or two other things go to show that Chaucer allowed some parts of the C. T. to become known while he was still working on others.
2 Trial Forewords, p. 16. 3 Chronology, pp. 46-8.
4 Geschichte, ii. 196-8 ; cf. Stiulicn, pp. 53-6.
5 Primer, p. 81 ; cf. his Knight's Tale (1903), p. xi.
6 Vol. i., p. 77. Skeat is mistaken as to the date, which was Christmas, 1385 (Walsingham, ii. 142, and cf. p. 151). This would put the poem at a time already crowded.
7 Publ. Mod. Lang. Assoc., xx. 861-2.
8 Furnivall Miscellany, pp. 307, 309-10 (note), 311.
84 POEMS DEPENDENT ON THE TESEIDE. [(3H. Ill, § 4
Troilus began Anelida as a pendant to it" (p. 311), since the plots of the two "are identical, only the main roles being reversed."
Professor Bilderbeck, in Notes and Queries^ suggests that the poem is an allegory on a contemporary incident. He quotes Thomas Walsingham's Historia Anglicana to show that in 1387 Robert de Vere, Earl of Oxford and Duke of Ireland, repudiated his wife Philippa, cousin of the king, and married a Bohemian lady, who had come to England in the train of Queen Anne. Obviously de Vere would be represented by the faithless Arcite, and the forsaken grand-daughter of Edward III. by Anelida, Queen of Armenia. He finds some confirmation for his conjecture in the King of Armenia's visit to England, which may have suggested the nationality attributed to Anelida. Prof. Bilderbeck's conjecture is rather attractive, but cannot possibly be accepted. I have shown elsewhere2 that only two years before the divorce episode, and a year before the date to which Bilderbeck assigns this expres- sion of reprehension, Chaucer fell under very considerable obligation to the Earl of Oxford ; the presumption is strong, therefore, that he would not have undertaken publicly to attack him.
But there is another and a much more conclusive argument against this date. Most of the light which we can expect on the date of the Anelida must be derived from its relations with the Troilus and the Knight's Tale. It must quite certainly have been written before the latter ; it was only the stanza-theory that required the reverse order.3
The first argument is the presence in it of passages from
1 Eighth Series, ix. 301-2. He might also have referred to the Evesham Hiit. Vitae et Regni Ric. II. (ed. Hearne, Oxford, 1729), p. 84 ; aiid to C. Hb'fler, in the Denkschriften of the Vienna Academy, xx. 188-91.
2 Mod. PhiloL, i. 328. It was de Vere that got Chaucer his deputy at the Custom-house.
3 Ten Brink's other arguments are nugatory. If it was written early he thinks it inexplicable that Chaucer should have permanently abandoned " ein iuit so groszem pomp eingeleitetes, mit so vielem aufwand dichterischer mittel begonnenes werk " ; and still more inexplicable that it should be preserved (Studien, p. 54). But why may not a poem lie in a chest twenty years as well as ten ? Its eventual publication is natural ; at his death Chaucer must have occupied much the same pre-eminent position as Dante at his, and somewhat as the last cantos of the Paradiso, according to Boccaccio's story, were sought and published, why not any interesting fragments of Chaucer's poetry that were found among his papers ? Nor is there any signifi- cance in the fact that Lydgate mentions the Anelida and not the House of Fame.
CH. Ill, § 4] THE ANELIDA AND ARCITE. 85
the Teseide some of which were used also in the other poem. It is natural to use parts of a poem and then decide to adapt the whole, and unnatural to use where they do not belong stanzas which had already been used where they do. Another consider- ation is that Chaucer is unlikely to have given to the heartless betrayer of Anelida the name and antecedents l of the chief hero of so important a poem as the Pcdamon and Arcite, if he had already written it. Moral indignation, to be sure, is not Chaucer's usual attitude, and he shows a certain tolerance for the faithless males of the Lef/end ; but the human emotion of his poems he took seriously, and the other Arcite embodies a high ideal.2 Such treatment of one of his own best poems would show an almost flippant lack of feeling. He would have been more likely to choose Palamon, whom he puts in a much worse light. Finally, Mather points out (p. 307) that the poem stops with a suggestion that Chaucer was about to describe a temple of Mars. Now, considering the intimate connection of this poem with the Teseide, and the imposing description in the latter, which so impressed Chaucer that he alludes to it in the invocation which heads the Anelida, if certainly looks as if a version of this were to follow ; otherwise, how could he have walked straight into such a no-thoroughfare 1 The feeling is hard to resist that the break in the Anelida just here is somehow connected with the presence of the description in the Knight's Tale. If the break can hardly come here because he had used the description, nothing remains except that he meant to use it.
It may be allowable to attempt a conjectural restoration of Chaucer's procedure. In the Parliament of Folds he had closely imitated Boccaccio's description of the temple of Venus, which almost immediately follows that of the temple of Mars. This use of the Teseide must have refreshed his memory of the poem, and he may then have undertaken to use larger portions of it, including this second fine description. It may also have occurred to him to sketch a poem in contrast to the Troilus, which he had probably been revising not long before ; a poem in which the tables should be completely turned on Arcite's sex.3 Whence he got the names
1 It is not quite accurate to say that this Arcite has only the name in common with the other ; cf. A. A. 85.
2 That Chaucer sketched him with strong liking is suggested by the changes he makes in Boccaccio's portraiture of the cousins. See Appendix C, pp. 231-2.
3 Cf. Troilus, V. 1779-85.
86 THE LEGEND OF GOOD WOMEN. [CH. IV, § 1
and material for the poem we do not know yet.1 But it did not proceed well, and the path ahead does not look very straight. At this point Chaucer brought up before the temple of Mars. He may have felt then that a much worthier use for that description and the admirable poem of which it is only one ornament would be a free but complete adaptation. Here therefore he permanently abandoned the Anelida.
As to the exact date, we cannot be sure. The above conjecture would put it about 1383-4, to which there are no objections. We cannot doubt that it comes between the Troilus, on the one hand, and the Legend"* and the Palamon, on the other, which gives the limits 1377 and 1385.
CHAPTER IV.
THE LEGEND OF GOOD WOMEN.
§ 1 . The Two Prologues : Ttie Question of Priority.
THE Prologue to the Legend of Good Women is extant, as is well known, in two versions, the shorter of which is found in only one MS., and is usually deemed the earlier. This I shall call G, and the other F.3 The existence of version G was not generally known
1 For Cowell's suggestion that Anelida was originally a Persian goddess, see Ch. Soc. Essays, 617-21 ; cf. also Samuel Dill, Roman Society from Nero to Harms Aurclius (London, 1904), p. 556. But there seems little doubt, as Professor J. Schick shows, that an Anelida was a character in the Matter of Britain, and the explanation of the Anelida and Arcite may lie in some voluminous Arthurian romance. In the old Italian Intelligenza (eel. by Cell- rich, Breslau, 1883, st. 75, 1. 2; cf. Schick, Temple of Glas, E.E.T.S., p. cxx.) she appears with Yvain among several pairs of lovers :
" La bella Analida e lo bono Ivano."
Froissart has the same couple (Dit dou bleu chevalier, 301 ; ed. Scheler, i. 357 ; cf. ten Brink, Studien, 213) :
" Je prenc Tristan pour Yseut le premier,
Et en apres Yewain le preu pour la belle Alydes. "
Just as she is here bella and belle, so Chaucer frequently calls her " faire Anelida. "
2 It is not mentioned in it, but Chaucer would hardly speak of an abortive fragment, which he had quite given over. Koch makes too much, I think, of what is no real difficulty (Chronology, pp. 46-7).
3 Prologue G is usually called A, and F is called B, designations which I reject because they imply what I believe to be a false view as to order ; this is also implied by the order iu which they are printed by Skeat in all his
CH. IV, § 1] THE TWO PROLOGUES: THE QUESTION OP PRIORITY. 87
of1 till it was printed by the Chaucer Society about 187 1.2 At first it was usually argued or assumed to be the earlier and rejected version,3 and it was not until 1 892 that a voice was heard on the other side. Since then the matter has been much debated, especially in Germany and lately in America, though something has come also from both sides of the English Channel ; and even now, in the view of some, the conclusive word has not been spoken, in spite of the fact that perhaps never has a scholarly question been settled so many times to the satisfaction of the settlers. In 1892, by a keen article in Enc/lische Stitdien? ten Brink supported the view that version G is the later, on the groiind mainly of its relation to Chaucer's life and later works. His opinion was promptly accepted by Dr. Emil Koeppel,5 by Dr. Max Kaluza,6 and by Dr. F. J. Mather ; 7 attacked by Dr. John Koch in an appendix to his Chronology oj Chaucer's Writings ;s defended again by Koeppel in a review of Koch's book ; 9 and attacked (on more purely aesthetic grounds, yet with a singularly cocksure manner) by M. Emile Legouis.10 This last paper was reviewed unfavourably in a valuable article by Gustaf Binz,11 and favourably by Koch.12 In England, up to this point, the whole controversy was ignored, and the older opinion supported by Skeat 13 and Pollard.14 But in 1902 Professor J. B. Bilderbeck
editions and by Pollard in the Globe Chaucer. I follow several other writers in calling the shorter G, after the unique MS. in which it is found, Camb. Gg. 4. 27, and the other F, after its best MS., the Fairfax, out of the eight or so which contain the Prologue.
1 Ten Brink states (Engl. Stud., xvii. 13) that in 1870 he had seen a tran- script of it, and then became convinced that it is the later version. It had been discovered by Mr. Henry Bradshaw and privately printed as early as 1864 (Trial Forewords, p. 104).
2 Odd Texts of Chaucer's Minor Poems, edited by F. J. Furnivall, 1868-1880.
3 As by Furnivall in 1871 (Athenceum, Oct. 21, p. 528 ; Trial Forewords, 106); by Skeat (Leg. of G. W., Oxford, 1889, p. xiii.); by Dr. Siegfried Kunz, Das Verhaltnis d. HSS. v. Ghaucers L. G. W. (Breslau dissertation, published in Berlin, n. d.), p. 12.
4 Vol. xvii. 13-23.
5 Engl. Stud., xvii., pp. 195-200. 6 Ibid., xxii. 281. — 7 Chaucer's Prologue, etc. (Boston, 1899), p. xxiii., note.
8 Published by the Chaucer Society and strangely dated 1890 ; see pp. 81-7 of the book.
9 Literaturblatt f. germ. u. rom. Philol. (1893), vol. xiv. 51-3.
10 Quclfut le premier compost par Chaucer des deux Prologues de la Legende des femmcs exemplaires ? In the Revue de I'enseignement des langues vivantes, Paris, April, 1900 ; pp. 58-71.
11 Anglia Beiblatt, xi. 231-7 (1900).
12 Engl. Stud., xxx. 456-8 (1902).
13 III., xxi.-xxv. (1894). » Globe Chaucer, xlv. f. (1901).
88 THE LEGEND OF GOOD WOMEN. [CH. IV, § 1
published a careful study of the Legend, in which he defended the older view on aesthetic and other grounds.1 In 1904 the most im- portant contribution to the subject ever made came from the pen of an American, Professor J. L. Lowes, who showed that version F contains borrowings from foreign poetry which prove its priority.2 In 1905 Dr. J. C. French supported the older view and attacked Lowes' position on aesthetic grounds ; 3 his book was reviewed, un- favourably to French's opinions, by the present writer,4 and was criticized by Lowes incidentally to a fuller discussion of the Legend.* Lowes' principal conclusions were accepted by Mr. A. W. Pollard.6 The fact that they are rejected in so good a book as Dr. E. K. Eoot's recent Poetry of CJiaucer will excuse my keeping • the subject open.
Although the succession of able articles by ten Brink, Koeppel and Binz, together with other evidence, had already thoroughly convinced me of the priority of F, the new evidence introduced by Lowes is particularly important and conclusive. The great service performed by him7 was the pointing out that Chaucer borrowed from a considerable number of French poems; by Machault, Deschamps and Froissart ; besides the verbal parallels in
1 Chaucer's Legend of Oood Women, London, 1902 (114 pp.).
2 P-ubl. of the Mod. Lang. Asst>c. of America, xix. 593-683.
3 The Problem of the Two Prologues, etc., a Johns Hopkins dissertation, Baltimore ; 100 pp.
4 Mod. Lang. Notes, xxi. 58-62.
6 Pull. Mod. Lang. Assoc., xx. 749-864 (on French, see pp. 749-51, note).
6 Academy, no. 1759, p. 62 (1906).
7 In his first article, PuU. Mod. Lang. Assoc., xix. 611-58. In my review of French I pointed out one or two other verbal parallels (see Mod. Lang. Notes, xxi. 59-60, notes 7 and 12). On the manner of introducing the Balade, cf. the following :
"So womanly, so benigne, and so nieke, . . . Half Mr beautee shulde men nat finde . . . And therfor may I seyn, as thinketh me, This song, in preysing of this lady fre " (F, 243-8).
"Son bel maintien, sa douce vois, . . . Me semout fort a ceste fois Qro une balade je die
En 1'ounour ma dame jolie " (Froissart's Lejoli mots de May, 11. 313-9 ; ed. Scheler, ii., 204).
While in G (89) May is almost past, in F (108)
' ' this was now the firste morwe of May ; "
so in Deschamps' Lay de franchise, which Chaucer used so much(<S>c. desanc. textes franc., ii. 204, line 14) :
' ' Le premier jour de ce mois de plaisance. "
CH. IV, § 1] THE TWO PROLOGUES : THE QUESTION OF PRIORITY. 89
F 40-65, he shows similarities of plan also to the Lay de fran- chise and the Paradys d'amours, by the two last (respectively). Not only is this highly interesting in itself; its chronological signifi- cance lies in the fact that though parallels exist in both F and G, there are far more in F. " The inevitable conclusion must be," to quote earlier-published words of my own, "that Chaucer read his French predecessors just before writing F. Now since their influ- ence on G is also unmistakable, a defender of the priority of G must ask us to believe that he went over these poems before each writing, and in F added to his mosaic with almost inconceivable \ care and ingenuity ; and, besides this, that he abandoned independ- ence in points where such a procedure was equally injurious and / uninotived. " l The priority of F, it seems to me, has been shown / by Lowes in a very demonstrative way.
But the question is a highly intricate and ambiguous one, more so, it seems to me, than Lowes altogether shows. The puzzle is that F, Avhich he proves to be the displaced version, seems to most readers the better and pleasanter. Legouis believes the aesthetic evidence speaks in favour of version F (p. 59). Even Koeppel characterizes the spirit of G as " ein ganz anderer, kraftigerer, aber auch etwas niichternerer Geist," with the personal feeling banished and the May scene relegated to the dream — he thinks G seldom improves over F, and more often shows signs of hasty revision.2 Lowes too is of much the same mind : " that the B [F] version has the note of freshness, of spontaneity, of composition eon amove to a greater degree than A — that it is even the more delightful version of the two — all will perhaps agree." 3 There are three more or less general and striking differences between F and G which Avill be thought at first to mark F as the better. These are its more genial and personal tone ; the pleasing suspense as to the identity of the lady of the Balade and the lady who enters with the God of Love, which is wholly given up in G ; and the fact that after she has been repeatedly named in his presence
1 M. L. N. , p. 60. Lowes' fuller and more authoritative discussion of this evidence will be found on his pp. 658 ff. French's unfair treatment of Lowes' arguments I pointed out in my review (see his pp. 32, 35-8, 65-6).
2 Literaturblatt, vol. xiv. (1893), col. 52. He attributes the change in spirit to the attempt to adapt part of the Legend for use as a Canterbury Tale. It is impossible to regard this suggestion with favour.
3 Publ. Mod. Lang. Assoc., xix. 683, note ; "but these," he continues, "are the very marks of a work written currcnte calamo, as against the firmer touch, the surer craftsmanship, the more compact unity of A " [G]. Why the latter merits should expel the former he does not tell us.
90 THE LEGEND OF GOOD WOMEN. [CH. IV, § 1
Chaucer in G affects not to know who she is, a blunder almost wholly absent from F. The first is particularly important, for it will be found that most of the detailed points of superiority in F are bound up with it. These matters nobody has adequately explained, especially no advocate of the priority of F. Lowes' attempt at some of them seems very slight and unconvincing,1 and his entire argument therefore lacking in finality. A perfectly satisfactory and rather illuminating explanation I believe is possible ; but must be deferred till the question of priority has been discussed on other grounds. Except for these three points I believe all of the important aesthetic considerations will indicate that G is the revised version.
All the thorough discussions of the aesthetic evidence, those of Legouis, Bilderbeck, and French, have been by the supporters of the priority of G,2 so it may be well to show that even on purely aesthetic grounds a good case can be made out for G as the revised ersion. These three writers have almost confined themselves to aesthetic arguments. But obviously, if others disagree with them as to the value of their arguments, and if Chaucer can be shown to have had a non-aesthetic motive for revision, which accounts for occasional inferiority in the later version, they have no case.
sgouis' argument seems much the best ; but it is not surprising that the accomplished critic of Wordsworth comes to Chaucer without the knowledge of the poet and his age requisite to a just estimate, and most of Legouis' points either prove to be connected with the omission of the personal feeling, which subject we are holding in reserve, or seem ambiguous or trivial. The other two writers, as I tried to show at more length, in the case of French, in my review, seem still more to select ambiguous or trivial details ; their standards are singularly arbitrary,3 and they never seem to see that many of their cases could be used as contrary arguments equally well. The fact that G exists in only one, and that a somewhat corrupt, MS.
1 Publ. Mod. Lang. Assoc., xix. 676, 681.
2 On the other side, of course, aesthetic considerations have not been wholly neglected. Lowes treats them more or less on pp. 661, 663, 665, 678-80 of his first article.
3 Notably as to alliteration and grammatical and logical structure. Nor do they seem to recognize how much of the broad and even careless style of mediaeval oral poetry still clung to Chaucer. The use which Bilderbeck makes of small peculiarities in G is particularly curious because he admits that it "has to some small extent been edited " by another than Chaucer (p. 47 ; cf. 71) ; and cf. French, p. 70.
CH. IV, § 1] THE TWO PROLOGUES : THE QUESTION OF PRIORITY. 91
vitiates minute points of evidence ; in particular, the small variations between F and G from F 426 to 495 are most probably due to a scribe. Besides this, the more important changes mentioned in the last paragraph may at times involve lesser changes which are not for the better. I am fully conscious of the difficulties and dangers of this kind of argiiment, and mean to notice every one of their argu- ments which does not fall under one of the condemnations which I have mentioned ; and mean to propose none myself which has not a large objective element. It might seem a priori that a thorough examination of the minuter differences of the two versions should clearly indicate which is the revised version. I can only state that after a very careful consideration of the two poems and of the attempts of the three writers just mentioned, I am convinced that it does not, partly, no doubt, because of the unsatisfactory MS. tradition.
> I am equally convinced thatKoeppel, defender though he is of G as the revised version, does injustice to the merits of G; and that apart from the three points held in reserve the more important and unambiguous aesthetic differences will speak in its favour.1 There is also evidence of a different character, which associates G with
* a later period in Chaucer's life than F.
— ^ We come now to the points in which G is the better. In the first place, it is more reasonably^ arranged — more methodical, though without stiffness. This is notably so in the proem and what leads
1 Two apparent important signs of the priority of G must be remarked on. The following couplet of F, 143-4, on the birds, is absent from G :
' ' Upon the braunches ful of blosmes softe, In hir delyt, they turned hem ful ofte."
That these admirable lines were deliberately omitted it is difficult to believe. But not only is there very considerable chance of accidental omission in a unique MS. , which has suffered serious damage immediately before and after the place where this couplet should be (as Binz points out, p. 236 ; and French admits, p. 70 ; Legouis does not see it, p. 67) ; but also, as even the hostile Bilderbeck shows (p. 45), some such couplet as this is needed to make gram- matical connection between lines 130 and 131 of G. So we may conclude that_ this omission was accidental. Secondly, in F 551 and G 541 Love declares that he shall " charge " Chaucer no more ; in G the Prologue ends in four more lines, but in F not for twenty-eight, which contain several instructions. At first sight it looks as if in F Chaucer had inserted a passage which makes 551 of none effect, as Koch thinks (Engl. Stud., xxx. 458). But the force of this argument is destroyed when we observe that in Fthe iastructions do not follow immediately on line 551 ; while inG, 541 is directly followed by the command to begin " at Cleopatre. " It seems quite as likely that G is the result of condensation as that F is of addition. To the best of my belief and judgment, no other signs of the priority of G can be mentioned without including the trivial and the still more debatable, and also multiplying instances on the other side.
92
THE LEGEND OF GOOD 11'OMEN.
[CH. IV, § 1
up to the dream and the entry of the God of Love, as may be made clear by a brief and bald analysis of F 40-213, and G 40-145.
G
His love of the daisy. He would fain praise it worthily, but " folk " have already done so better. Hopes he shall incur no ill-will for repeating their words, since he does all in hon- our of those who serve either leaf or flower. But he is no partisan of either. We should trust authorities. Means to de- clare old stories. After he has roamed the meadow, goes home to sleep. Dreams. Description of the meadow. The birds' mirth. A lark announces the God. He enters.
His love of the daisy. He visits it in the morning. None ever loved hotter than he loves the daisy. At evening he runs to see it close. It opens in the morn- ing. He would fain praise it worthily, and invokes lovers' aid, but they have already done so better. Hopes he shall incur no ill-will for repeating their words, since he does all in hon- our of love and in service of the flower. Again declares his love and reverence. Will tell later why he says that we should trust authorities. Love made him rise early to see the daisy ; he knelt to watch it unclose. Description of the meadoAv ; the birds' mirth. Allegorical digres- sion on the birds. Sank down to watch the flower all day. Praises it again. But he is no partisan of either flower or leaf. Toward night he goes home, meaning to rise early to see the daisy again. Dreams he is back in the meadow. Entrance of Love.
I do not mean, of course, that here or anywhere G is pleasanter than F on a casual reading ; rather the contrary, since it omits the passionate devotion of the other. But in a number of points here it is more reasonable and pleasing on examination, and closer to Chaucer's later work. A few of these points may be indicated.
U While in G he defends himself from the charge of partisanship immediately on mentioning the flower and the leaf (70), in F, though his devotion to the daisy is far more marked, he does not
\jlo so till over a hundred lines later (72, 188 ff.). If G is the earlier, there is no discoverable reason Avhy he should have made such a postponement in revising. Secondly, the analysis makes very clear the extraordinary skipping about in F between morning
CH. IV, § 1] THE TWO PROLOGUES : THE QUESTION OF PRIORITY. 93
and evening ; without motive Chaucer would hardly have made order into chaos. Thirdly, after asking in F why men should trust authorities, Chaucer says (101) " That shal I seyn, whan that I see my tyme," and then returns to dilate on his passion for the daisy, and never fulfils his promise. In G (81-8) he explains his attach- ing such importance to belief in authorities by the fact that he is about to relate tales drawn from them.1 The passage in F is a good example of the free-and-easy inconsequence of that version ; that in G, of its soberer forethought. Which of these characteristics may most naturally be attributed to a first version, and which to a second, is obvious enough. Next, the relation between the dream and the preparation for it seems better in G ; 2 after the essential introduction, his habitual affection for daisies, and the afternoon in the meadow which was the starting-point of the dream, he goes home and falls asleep, and the description of the meadow and the birds is a part of the dream.3 One advantage of the method of G is that it makes the entrance of the God less abrupt ; in F Chaucer begins to dream in line 210 and in 212 the God appears, when the poet has barely got his eyes shut. But for every reason I do not see how it can be denied that this shortening and clear-marking of the introduction, and this centring of the interest on the dream scenes- and incidents is an improvement. Nor, if G preceded F, is it likely that Chaucer would have made the contrary change, which would not have been in the least involved by the introduction of the personal feeling.
But more than this, version F in this point resembles Chaucer's earlier poetry, and G his later. In the Book of the Ducliess there is a preliminary ramble which forms nearly a quarter of the whole poem, and is not closely enough connected with the main trans- action to justify half that length ; in the Parliament of Foicls the introduction forms a sixth of the whole, and by no means justifies its length. In both he gives quite otiose accounts of what he had been doing. In the House of Fame the proem and invocation,
1 G 81-4 will.be seen to be not quite grammatical, a natural consequence of a not very careful change in the form of the sentence. An almost grotesque example of the rambling style of F will be found in the House-that-Jack-built sentence in 11. 103-114.
2 So Binz, p. 235. Skeat also points out (III. xxiii.) that the proem is more distinctly marked in G (1-88).
3 Legouis' reasons (see p. 62) for preferring the method of F are hardly intelligible, for the dream is quite sufficiently accounted for in G. Unfor- tunately the mediaeval court-poet needed little excuse for dreaming.
94 THE LEGEND OF GOOD WOMEN. [cH. IV, § 1
1 10 lines long, followed by hundreds of lines based on the Aeneid, make very little contribution to what follows. It is clear, there- fore, that in his earlier dream-narratives Chaucer, unlike his model Guillaume de Lorris, was in the habit of lingering in the world of actuality, even to the point of scattering, if not annihilating, the interest. But compare the fine rapidity with which he breaks into the main narrative in every one of the Canterbury Tales.1 Does not this comparison suggest that G was written not only after F, but long after ?
* One of the most striking points of superiority in G is in the entrance of the procession and the presentation of the Balade. In F, after the God and the lady have entered together and been described at great length, Chaucer introduces the Balade with the words :
"And therfor may I seyn, as thinketh me, This song, in preysing of this lady fre " (247-8).
After it he continues :
" This balade may ful wel y-songen be, As I have seyd erst, by my lady free " (270-1) ;
he praises her again, and finally (70 lines after the first two) intro- duces the rest of the procession, the nineteen ladies, followed by a great multitude, who kneel in honour of the daisy and sing a few lines to her. In G the God is announced by a lark :
" Til at the laste a larke song above : ' I see,' quod she, ' the mighty god of love ! Lo ! yond he cometh, I see his winges sprede ! '" (141-3).
After the God and the queen have been described, the rest of the procession enters, and the Balade is sung by the ladies. As to the lark, Dr. Skeat says (III., xxiv.) it "is left out, as being unnecessary. This is a clear improvement."2 I can only say that the lark seems to me just as necessary, and in the same sense, as the whole poem is. Again, the pause during the entrance of the pro- cession is only about half as long in G as in F, where the Balade intervenes. But the most striking point of superiority in G is the way in which the Balade is presented. In F it has no function in
1 Except the Pardoner's and Canon's Yeoman's, where the ramble is deliberate.
a Similarly French, p. 50 ; cf. my review, Mod. Lang. Notes, xxi. 61. Legouis (p. 62), however, says Chaucer had to sacrifice this pretty detail ; why ?
CH. IV, § 1] THE TWO PROLOGUES : THE QUESTION OF PRIORITY 95
the narrative,1 and even the ladies have little. I have shown else- where that the artistically unintelligible manner in which it is introduced is clearly one of the points in which at first Chaucer followed his French exemplars.2 A further disadvantage of tK state of things in F is that it makes Love refer (539 ff.) to a poem which he has not heard.3 Is there any comparison between the two methods as to art and grace ] Could Chaucer have changed the conditions in G to those in F ] 4
Among many small points in which on examination G appears superior to F, three may be especially mentioned. In F one of Chaucer's crimes is recorded thus :
" For in pleyn text, with-outen nede of glose, Thou hast translated the Romaunce of the Rose " (328-9) ;
in G thus, of course with the same meaning :
" For in pleyn text, hit nedeth nat to glose " (254)
The ambiguity of the F -reading is such that it misled Dr. Koch, who says of the Romance of the Rose that this line " implies, though not directly meant in that way, that his rendering was a literal one."5 Certainly there was no reason for change from the G to the F reading. Another change in the interest of lucidity occurs in G 343-6 :*"
" And takth non heed of what matere he take ', Therfor he wroot the Rose and eek Crisseyde Of innocence, and niste what he seyde ; Or him was boden make thilke tweye " ;
in F the passage is practically the same with the omission of the two middle lines. The naming of the two poems is necessary, for even in F ihilke must go back for its antecedent past thirty -four
1 So ten Brink (Engl. Stud., xvii. 16-17) ; Binz (Angl. Beibl., xi. 235) ; cf. also, on all this, Lowes in Publ. Mod. Lang. Assoc. , xix. 655-7.
2 See Mod. Lang. Notes, xxi. 59, and p. 88 above. Lowes did not remark 011 this point, which seems to me important.
3 It is curious also that in F Love reproaches Chaucer for not having put Alcestis into the Balade partly on the ground that he is "so gretly in hir dette " for the protection which she has only just given him. On the Balade ten Brink has some rather over-subtle criticisms (1. c., p. 16-17).
4 The one point of superiority in this part of F, the anonymity of the lady, I have asked to have held in suspense till later.
5 Chronology, p. 13. The first line obviously has the same meaning as the fourth line of the Wife of Bath's spurious "head-link " in the Lansdowne MS., " I will nouht glose, bot saye the text." On this couplet, see also Lowes in Publ. Mod. Lang. Assoc., xx. 855.
96 THE LEGEND OF GOOD WOMEX. [CH. IV, § 1
lines occupied with other matters.1 The third point is the ending ; after the God's final admonition,
" with that word my bokes gan I take, And right thus on my Legend gan I make " (578-9),
for which G has (544),
" With that word of sleep I gan a-awake " [sic].
That is, in F Chaucer passes from his dream-adventures in the meadow to working in his own library, without awaking.2 These instances of the superiority of G to F are by no means all ; there are many more in which most tastes would probably recognize improvement. In every case there is a clear motive for the change if G is the later ; in every case there is none if it is not.
We now come to the cases which show other than aesthetic evidence that G is the later version. The "old fool " passages are the first.3 In G (258-62) Love remarks that Chaucer's wit is full cool, and adds,
" Wei wot I ther-by thou beginnest dote As olde foles, whan hir spirit fayleth " ;
and later that (314-5)
" thou reneyed hast my lay, As othere olde foles many a day,"
' for which F has " other wrecches " (337). 4 Now, so far as I have been able to find, Chaucer's only other references to his own elderly
years are in the Complaint of Venus (76-8), where he says that age
•
1 Legouis thinks the greater clearness of G a mark of priority, and the obscurity of F a result of condensation (p. 64, note ; Koch agrees with him, Engl. Stiid., xxx. 457-8 ; cf. his.Chronol., 83, and Biriz, p. 236). But there is no indication in either prologue that Chaucer was trying to condense — certainly not at the expense of clearness. The notion that he was jests only on the supposition that Love is Richard 1 1., and that F 570-7 is an expression of the royal desire for brevity. It is unnecessary to dwell on this, especially since, instead of being shorter than G, F is 34 lines longer. S;
- Legouis curiously ignores this consideration, and thinks F 578 superior because it returns to the books mentioned at the beginning of the Prologue (p. 65). I may compare a similar change in the Confessio Amantis. Though, in both of Gower's versions of the end, the departure of Venus is mentioned, that of Genius is ignored except in the revised version (Macaulay, vol. iii., p. 467).
3 The point was first made by ten Brink (Engl. Stud., xvii. 14 ; and see Lowes, xx. 782-7).
4 Cf. G 400-1 (nothing corresponding in F) :
" Whyl he was yong, he kepte your estat, I not wher he be now a renegat. "
CH. IV, § 1] THE TWO PROLOGUES : THE QUESTION OP PRIORITY. 97
has dulled him and taken away his subtlety ; and in the Envoy to Scogan (27, 31-5, 36-42). l These two poems there are good reasons for dating between 1390 and 1400.2 Legouis (pp. 63-4) finds ground for change from the G-form in the fact that here Love falls below the dignity of a god ; but I think this reason would hardly have appealed to Chaucer, who enjoys nothing better than putting down the mighty from their seats as witness the colloquial dis- course of the eagle in the House of Fame, or of Pluto and Pro- serpina in the Merchant's Tale. Legouis also assumes with Skeat (III. xxii.) that the revision occurred very shortly after the first draft, an assumption which is made very unlikely by the extent of the alterations.3 Ten Brink points out that when Chaucer wrote the first version (whichever that is) he was not old enough to use such language even in joke.4 Of course the remarks are jocose ; but since the only conceivable reason for omitting such good and characteristic lines — sensitiveness — is negatived by all that we know of Chaucer's character and practice, the most reasonable inference is that G was written long enough after F for Chaucer to have come to make fun of his own advancing years.
F 537-40 and G 525-7 form a case where the superiority of F actually suggests that G is the later. F reads :
" Than seyde Love, ( a ful gret negligence Was hit to thee, that ilke tyme thou mad* " Hyd, Absolon, thy tresses," in balade, That thou forgete hir in thy song to sette ' " ;
for which G has :
" ' a ful gret negligence Was hit to thee, to write unstedfastnesse Of women, sith thou knowest hir goodnesse.' "
Negligence is as distinctly the right word in F as it is the wrong one in G. The line in which it occurs is the las't of a long passage in which probably only one of the differences between the versions is due to Chaucer ; to alter the word would have required recasting
1 The remark in the House of Fame (995) need hardly be considered.
2 On Venus, see Skeat I. 86 ; on Scogan, Skeat I. 556-7, and G. L. Kittredge, Harvard Studies and Notes, i. 116-7 ; on both in connection with the year of Chaucer's birth, Lounsbury's Studies, i. 36-42. In partial answer to Lounsbury, I may point out that January in the Merchant's Tale is regarded as an old man at sixty.
3 See p. 122 below.
4 Engl. Stud., xvii. 10 ; cf. Koeppel, Literaturblatt, 1893, p. 51 ; and Koch, Chronology, p. 82.
PEV. CH. H
98 THE LEGEND OF GOOD WOMEN. [CH. IV, § 1
of the whole couplet of which it is one of the rhymes. If Chaucer wrote G first, it is strange indeed that the change he made in the later lines exactly fitted the proper meaning of this word, while if F is the earlier it is not surprising that he failed to alter it.1
In F, among the parting injunctions of Love, is the line (562) :
" Make the metres of hem as the leste."
Is not this assuredly an allusion to the fact that Chaucer is using a metre new to English poetry 1 2 I have already shown that this cannot be the first poem in which Chaucer used the decasyllabic couplet, but no doubt it Avas the first one published. Such an allusion is certainly less surprising in a first version than in a second ; if a long interval elapsed between, this line almost proves F the earlier.
There are several passages which suggest that F is the earlier and G the later by certain points of connection with earlier or later works of Chaucer's. This has already been pointed out in the case of the introductory portion of the Prologue. But the most important cases of parallels to earlier and later works are the only two long passages that are confined each to one version.
In F 153-187, the digression on the birds, the first part is strongly in the style of the Romance of tlie Rose and the Parliament of Fowls,3 with its (quite superfluous) characterization of an indi- vidual bird, its vows of constancy, and its allegory ; it is a digression from a digression, with an impertinent quotation from Aristotle. The passage is so irrelevant that Mr. Bilderbeck 4 has found it necessary to fill it with political allegory. It alone would suggest that the version which omits it is the later, — it is surely not very likely to have been added on revision, especially if the revision was made a considerable time after the first writing, and most especially if in the Canterbury-period. But it is less significant than the second pas- sage,5 G 267-312, where Love asks Chaucer why he has not written of good women, and declares that he might have found many such
1 Cf. Lowes, in Publ. Mod. Lang. Assoc., xx. 799.
2 Cf. Lowes, p. 814, whose alternative suggestion seems to me hardly possible.
3 Cf. Skeat's notes, III. 295-6.
4 Pp. 101-3. His interpretation seems to me very unlikely ; it is vague, and touches the passage at only a few points. It is so easy to construct ex post facto allegories (as witness the procedure of the Shakspere-Bacon fanatics, and, in the sixteenth century, of Tasso and of the admirers of Ariosto) that it seems to me they should be submitted to a very austere criticism.
5 This argument is developed from ten Brink's, in Engl. Stud., xvii. 15-16.
CH. IV, § 1] THE TWO PROLOGUES : THE QUESTION OP PRIORITY. 99
in Valerius, Tigris, Clau/iian, Jerome, Oytcl, and Vincent. Only one reason worth mentioning why Chaucer should have omitted this passage is suggested by those who think F the later version ; Legouis (p. 63) thinks this passage a verbose pedantic sermon.1 We may like the passage or we may not, — in itself it is not much better than the verses on the birds which G lacks ; but it forms an integral part of the poem, which the other does not, by adding force to_Love^s_ rebuke. _As to the charge of pedantry, tnTMiddle Ages took a view different from ours of appeals to authority, even of a display of learning, and no criticism can do mediaeval literature justice which disregards this fact ; the greatest of all mediaeval poets is full of direct citation of Aristotle and the theologians. Ghaucer uses the practice with humorous effect in the Nun's Priest's Tale, but that he was far from meaning to ridicule it is shown by the discourse on ancient chaste heroines with which Dorigen assuages her grief.2 With this latter passage the one in question has much in common, in source, tone, and content — enough to link it rather to Chaucer's later work than to his earlier ; and it is certainly more in place.
But it is also important to observe the authors whom Chaucer names here. We may at once disregard Ovid, with whom he shows familiarity throughout his literary career ; Claudian, to whom he refers in the House of Fame; and Titus, no doubt Livy, with whom he had long been familiar through Le Roman de la Rose, and whom he quotes, not necessarily at first hand, in the Legend of Lucretia and in the Book of the Duchess. Vincent of Beauvais,3 it has been supposed, or else Jerome against Jovinian, is quoted on the use of a hyaena's gall to cure blindness in Fortune, 35-6, a poem of wholly uncertain date ; but it is impossible to be sure of* the source of an idea like this, and moreover this poem may be a translation.4 Chaucer possibly quotes Vincent in the Nun's Priest's Tale, 4354, and probably in the Wife of Bath's, 1195. So far as evidence goes, then, Vincent is associated with the period of the
1 See also Koch, Chronology, p. 83 ; Bilderbeck, p. 83.
2 Frankl. T., 1364-456. A similar list and discourse, under not dis- similar circumstances, is to be found in Boccaccio's Fiammetta (Moutier, Florence, 1829 ; vol. vi., pp. 181-99). The whole eighth chapter is occupied by a soliloquy, in which Fiammetta cites and dwells on two or three dozen antique heroines, in order to console herself for her disappointed love.
3 See Lounsbury's Shidies, ii. 379-80. * Ibid,, -p. 296.
100 THE LEGEND OF GOOD WOMEX. [CH. IV, § 1
'Canterbury Tales. As to Valerius, it is not quite certain who is meant, for Chaucer mentions three of the name. It is certainly not Valerius Flaccus, the author of the Argonauticoit.1 Skeat thinks it is Walter Map's Dissuasio Valerii ad Rufinum? which he mentions or quotes in the Wife of Bath's Prologue and the Merchant's Tale.3 Koeppel and Lounsbury 4 think it is Valerius Maximus, who elsewhere is quoted only in the- Wife's Tale and Prologue, and perhaps in the Nun's Priest's Tale and the Monk's Tale.5 Though the last has sometimes been thought earlier than the Legend, this has certainly not been proved, and later I hope to go very far toward disproving it.6 Certainly it is fair to say that Valerius Maximus7 is distinctly associated with a subsequent period. With the work of Map the case is still stronger. Jerome against Jovinian Chaucer uses or mentions only here and in the Cantei'bury Tales, so far as present information goes, and except for one or two possible cases. The first is the almost nugatory one mentioned already, in which in Fortune he may quote either this work or Vincent of Beauvais ; the second is a quotation from either Jerome or John of Salisbury 8 in The Former Age (33),
1 Cf. L. O. IV., 1457-8 ; T. and C., V. 8.
- In Map's De Nugis Curialium (Camden Society, 1850), pp. 142-52.
- See Louusbury, ii. 367-70 ; Koeppel, Anglia, xiii. 181-3.
4 Anglia, xiii. 182 ; Studies, ii. 276.
5 Lounsbury, Studies, ii. 273-6. Miss K. 0. Petersen (Sources of the Ar. P. T., Bosten, 1898 ; pp. 110, 117) shows that the two exempla in N. P. T. 4174-4294 may be from neither Cicero nor Valerius directly, but may come from the latter through Robert Holkot's Super Libros Sapientice. Professor Bright (Mod. Lang. Notes, ix. 241) has attempted to show that this Valerius is quoted in H. F. 516. The Eleanor to whom a marvellous dream is attributed he thinks is Hamilcar, whose dream as to the taking of Syracuse is narrated in a few lines by Valerius in book I. 7, 8. But such a monstrous corruption as this seems hardly probable in late written tradition. I fear that we must agree that this reference is still unexplained. The conjecture that there might be something to explain it in the romance of Escanor is negatived by an examin- ation of that poem kindly undertaken by Dr. G. L. Hamilton.
6 See pp. 164-172 below.
7 I agree with Koeppel and Louusbury that Chaucer probably refers to him. In Valerii Maximi factorum dictorumjue memorabilium libri ix. (Curiae Regnit, 1799), iii. 2, " De fortitudine " praises Portia, wife of Brutus ; iv. 6, " De amore conjugali " again praises her, and also Julia, daughter of Caesar, and others ; vi. 1, "De pudicitia" praises Lucretia and others, mainly severe- minded men; vi. 7, " De fide uxorum erga maritos" praises the wives of Scipio Africanus, Q. Lucretius and Lentulus. Cf. French, p. 57, whose treatment of the subject of these authors, however, is very unsatisfactory. The work of Map praises, to be sure, Lucretia, Penelope, and the Sabine women ; but immediately adds, ' ' Amice, nulla est Lucretia, nulla Penelope, nulla Sabina ; omnes time " (p. 145). An allusion to this book by the God of Love could be explained only as a mauvaise plaisanterie.
8 The Rev. W. W. Woollcombe can hardly be said to have proved Chaucer not to have known the Polycraticus (Ch, Soc. Essays, 295-8).
CH. IV, § 1] THE TWO PROLOGUES : THE QUESTION OF PRIORITY. lOl
a poein of unknown date, which cannot be assumed to be contem- poraneous with the Boethius. What has usually been deemed a third use of Jerome occurs in both forms of this very prologue — the mention of " Marcia Catoun " as a " Good Woman " in the Balade. But I have tried to show elsewhere that sucli is very unlikely to have been her source, and that she is most likely derived from the Divine Comedy, where the poets meet Marcia the wife of Cato in Limbo, and in Purgatory appeal to the husband in her name.1 There is no evidence, therefore, that Chaucer was familiar with any of St. Jerome's works before the time of the Canterbury Tales.'2' But then he quotes from Jerome against Jovinian fre- quently and extensively ; twice in the Pardoner's Tale (505, 527), once in the Manciple's Tale (148), largely in the Franklin's Tale (1367-456), and (asKoeppel shows) in the Wife of Bath's Prologue, the Sumner's Tale, and the Merchant's Tale, passim.3 Now in the G prologue Chaucer betrays great intimacy with the work ; otherwise it \ is the last thing which he would think of making Love quote, and f\?J while the other authors are barely mentioned he has twenty-four /^ lines on Jerome's work. Does not this fact point to a period when he was especially familiar with it? Therefore of the six authors mentioned we have found three to be more or less distinctly and emphatically aroociated with the Canterbury-period.4
The only two long passages, therefore, which are each found in only one version are unambiguous in their testimony. That in F is : likely to have been written relatively early in Chaucer's poetic career, because it resembles in tone several of his earlier works ; and might well be omitted on revision because it is a digression. That in G, on the other hand, performs a function in the narrative, and by its character and by the authors to whom it refers associates itself with Chaucer's later work.
A somewhat similar argument may be based on the mention in G only, among Chaucer's own works, of the book,
1 Inf., IV. 128 ; Purg., I. 78-81. See Mod. Philol, iii. 368-70.
2 He quotes him several times in the Pars. T., which may antedate most of the C. T. ; but it is certainly a translation.
3 See Skeat's index of authors ; Lounsbury's Studies, ii. 292-7 ; Koeppel, Anglia, xiii. 174-81. One cannot help fancying that Chaucer first became familiar with this work when he was planning and writing W. B. P. See also pp. 202, 209, 212 below.
4 Of course this is no proof that he did not know some of them earlier, but the inference is justifiable that he was riot familiar with all of them.
102 THE LEGEXD OF GOOD WOMKX. |_CH- IV> S -
" ' Of the Wreched Engendring of Mankinde,' As man may in pope Innocent y-finde." l
Professor Lowes 2 is no more assuredly right in rejecting the biographical reasons for dating this work in the late eighties than in deducing a late date from the use of it in the Man of Law's Prologue and Tale,3 and the Pardoners Tale, and also in dwelling on the improbability of Chaucer's mentioning it here unless he had just produced it.
This finishes the evidence on the question of priority, save for the three matters which I have been holding in solution. Aside from them, points of superiority in F are negligible. The indications that G is the later, on the other hand, are many and various, and by no possibility which I can conceive, even granted that individually they are sometimes small, can they be explained away. Considerations of merit and of literary relations both lend strong support to the crucial evidence supplied by Lowes' demon- stration of the closer connection of F with certain French models. It remains for me to attempt the rehabilitation and extension of the old and orthodox theory of a personal compliment to the queen paid through Prologue F, and removed from G ; which I believe will account for all the respects in which the latter seems inferior to the former.
§ 2. It 6- Connection icith the Queen.
The theory which I propose as to a connection between the Legend of Good Women and the queen is largely the old one ; but I can offer new evidence for it, and make a new application of it. I believe : — That Chaucer uses the daisy and Alcestis expressly as vehicles for his personal tribute to Queen Anne ; that accordingly the personal devotion expressed in F was meant and understood as a compliment to her ; that the writing of the whole Legend was a
1 It also seems odd that if F is the later, Chaucer should at once omit this work and substitute hotynesse for bcsinesse just before. Legouis takes an opposite view (p. 68).
2 Pabl. Mcd. Lang. Assoc., xx. 790-4. The force of the argument was admitted by even Koch (ChronoL, 86), but later he changed his mind (Engl. Stvd., xxx. 457). And cf. Legouis, 65.
3 Cf. pp. 181-2 below. I show on p. 214 that the passage in W. B. P. is perhaps due to Gower's Mirour. Another connection between G and M. L. T. (pointed out by Koeppel, Herrig's Archiv, Ixxxiv. 411) is that G 312, 529 parallel M. L. T. 701-2.
CH. IV, § 2] ITS CONNECTION WITS THE QUEEN. 103
task imposed, in a light vein, by her ; l that the revision of the Prologue was made after her death ; and that all the passages in it which definitely recalled the earlier connection with her were carefully excised, probably out of consideration to her husband's feelings. Put thus baldly, this may well sound rash and gratui- tous ; but I believe there is excellent evidence for all of it, and that thus alone can the facts be explained.
Tyrwhitt first showed the connection with Queen Anne made by the couplet (F, 496-7) in which Alcestis instructs the poet to present the finished work to the queen, "at Eltham, or at Shene."2 Ten Brink in 1870 suggested3 that the queen was celebrated by means of the daisy and Alcestis, and that the whole was a tribute of gratitude to her for having secured for Chaucer in February, 1385, permission to discharge his custom-house duties through a deputy. Till 1903 this view was accepted by everybody (I believe) who expressed himself in print on the subject ; by Dr. Furnivall (doubtfully) in 187 1,4 by Professor Skeat in 1894 and earlier,5 by Dr. Koch in 1890,6 by Mr. A. W. Pollard in 1901,7 and by Professor Bilderbeck in 1902.8 In one of the last articles he ever wrote,9 ten Brink kept this date for the first version of the Prologue and for the legends, and therefore evidently held to the theories on which the date rested. But in 1903 I showed10 that, since the petition that Chaucer might be allowed a deputy wair signed by Robert de Vere, Earl of Oxford, it was he and not the queen who was Chaucer's sponsor in this matter, and that therefore / there is no such external reason for connecting the Legend with , the date 1385, or with the queen. My conclusions have been almost universally accepted.11 But Professor Lowes,12 acting in part
1 In this point I slightly modify iny earlier article on L. G. W. (Mod. Philol., i. 326).
2 0. T. (1830), I. clxi. He pointed out that we must therefore date the poem not earlier than 1382, when Richard II. married.
3 Stud-ten, pp. 147 ff. A list of those who have accepted the identification of Alcestis and Anne is given by Lowes, Publ. Mod. Lang. Assoc., xix. 666 ; but add Legouis, p. 69. The daisy was always believed to mean some living woman ; Speght in 1602 (p. b. vi. bis) stated that it typified Princess Margaret.
4 Trial Forewords, pp. 25, 106. 5 III. xix.
6 Chronology, 44-5.
7 Globe Chaucer, xlv. ; cf. Chaucer Primer, 95-6.
8 Bilderbeck, Chaucer's L. G. W., p. 88, note.
9 Engl. Shid., xvii. 19. Cf. also his History of English Literature (1893), ii. 110-13. 10 Mod. Philol., i. 324-9.
11 By Lowes, Publ. Mod. Lang. Assoc., xix. 670 ; (apparently) by French, The Problem of the Two Prologues (Baltimore, 1905), p. 21 ; by Koch, Engl. Mud., xxxvi. 141-2. Root rejects them (p. 141). 12 L. c., 669-76.
104 THE LEGEND OF GOOD WOMKX. [cH. IV, § 2
on my evidence, goes so far as to reject all connection between the queen and the apparent symbolism of Prologue F. Herein I believe he goes too far, and that his and my opinion that F is the earlier version is greatly strengthened by the orthodox view as to the queen and the Prologue.
Let us first consider the surface appearance of the two versions. That some living woman is symbolized by the daisy and by Alcestis in F, and not in G, seems a plausible and almost inevitable conjecture. In their treatment of the daisy, the contrast, as to personal devotion, between the two versions can hardly be exaggerated. In G this devotion finds distinct expression only in lines 40-8, 55-60, 92, 511-12; in F, however, in 40-8, 50-9, 60-72, 82-96, 103-111, 115-17, 180-7, 201-2, 211, 523-4. Again, it is expressed in F with a warmth to which there is no parallel in the other version. Consider, among others, the follow- ing lines peculiar to F :
" Ther loved no wight hotter in his lyve " (59) ;
" The herte in- with my sorowful brest yow dredeth, And loveth so sore, that ye ben verrayly The maistresse of my wit, and nothing I " (86-8) ;
" My besy gost .... Constreyned me with so gledy desyr, That in my herte I fele yit the fyr" (103-6).
All this language, it must be remarked, is used toward the daisy. In lines 69-83 he appeals to lovers to help'.him praise the flower, and apologizes to them, instead of to the indefinite " folk " of G, for repeating their words. Three small points may be especially noted : in F he writes
" in the honour
Of love, and eek in service of the flour" (81-2), inG
" in forthering and honour Of hem that either serven leef or flour " (69-70) ;
he is kneeling by the daisy in F (308) when the procession enters and surrounds him — "faste by under a bente" in G (234); and only in the former does the God of Love call the daisy his flower (316, 318), or his relic (321). Finally, there is no mistaking in F the human symbolism of the daisy. This appears first in the pronouns used in speaking of it. In G, hit is used in 49, 52-3,
CH IV, § 2] ITS CONNECTION WITH THE QUEEN. 105
and she only in 95 (which reads practically the same in F) ; in F hit is used in 49, 52, 56, 62, 65, 111, 117, 183, but she (Mr) in 53, 63, 64, 84, 186-7, and yow (ye, your) in 86-7, 89, 92, 93, 94, 95. Although the change to the second person is due to the fact that Chaucer is translating here from the Filostrato,1 while Boccaccio uses the singular tu, Chaucer changes to the more reverent plural.2 The personal symbolism shows markedly in the use of such expressions toward the daisy as maistresse (88), lady sovereyne (94 ; cf. 271-5, where similar language is used of Alcestis, in F only), ertJdy yod3 (95 ; cf. the whole passage, 83- 96), this flour so yonyf so fresh of hewe (104), — all of which are unparalleled in G. In contrast to this reiteration, intensity and unquestionable inner meaning, we have in G only the minimum of devotion necessary to justify the introduction of the daisy at all.5 As to Alcestis, she is explicitly identified with the daisy in both versions (G 499-500, 506-7; F 511-12, 518-19), and in gratitude for her protection is highly extolled for her beauty and goodness ; but in F Chaucer's devotion to her is slightly ^
pronounced (cf. 270-5, not in G).° There can be no doubt that ^\ all these differences were deliberate ; either Chaucer introduced i / human symbolism and an appearance of warm feeling into a poem originally without a sign of either,7 or else he cut them out of a poem that had had both.
It has seemed worth while to sift out the reasons for the impres- sion of personal feeling which F gives as opposed to G, because it brings the issue to a head. But now how is it all to be interpreted ? Lowes says this feeling is all literary convention, and directed to Chaucer's i ideal- mistress Alcestis ^ "all these assumed allusions of CEaucer to the Queen are nothing whatever but translations-of such
1 See Lowes, Publ. Mod. Lang. Assoc., xix. 619.
2 Evidently he would not thou his queen.
3 Skeat (III. xxiii) glosses this phrase by line 387, where Alcestis says that lords (Skeat errs in saying kings) are " half-god des in this world here."
4 The queen was twenty at the probable date of the F prologue.
5 Chaucer expresses love for the daisy only once (42-4), and in the plural (these floures).
6 In F Alcestis seems to be the vehicle for Chaucer's veneration toward the queen, and the daisy for his "courtly love."
7 So Furnivall (Trial Foreivords, p. 106). Skeat (III. xxii.) thinks that even in G the queen was symbolized, but so inadequately that Chaucer at once rewrote it. Not only is such a procedure highly improbable, and not only does it represent Chaucer as singularly helpless and inept, but if we had only version G we should be unable to detect more symbolism than the relation between the daisy and Alcestis.
106 THE LEGEND OF GOOD WOMEX. [CH. IV, § 2
conventional expressions as form the very warp and woof of the French poems he was imitating;"1 all the personal language, including the lady sovereyne, he regards as " commonplaces taken over bodily from the originals " in French and Italian, and the use of she or her for the daisy is " simply the adoption of the convention of the type." As to this last point, Lowes disregards the obvious fact that while elle is required in French by the grammatical gender of flour and margherite, she in English is wholly personal. He is quite justified in saying that Chaucer's other personal and emotional language, and his celebration of the daisy, are paralleled in French poetry ; and he has made an important contribution to the subject by showing that this alone cannot prove a connection with the queen. But an examination of all the French poems in question will show that Chaucer altogether outdoes his French exemplars. These poems may be divided into two classes. Of long narratives there are two, Deschamps' Lay de Franchise, and Froissart's Paradys d'Amours, to which Lowes has shown that Chaucer is deeply indebted for his plan. The lyric poems comprise Machault's Dit de la Marguerite, Froissart's Dittie de la flour de la Margherite, Le joli mois de May, the 17th Pastourelle and the end of the Plaidoine, and a dozen or so of Deschamps' balades. Most of these poems fall far behind Chaucer's in intensity and insistency of feeling. Of all of them the warmest devotion and love is to be found in the first two of the lyrics ; elsewhere courtly compliment is paid to the flower, and devotion to the poet's lady, but the two are rarely combined, as in Chaucer. What Chaucer has really done is to combine the lyric warmth of Machault's Dit and Froissart's Dittie with the narrative schemes of Froissart and Deschamps, introducing also an intensely personal passage from Boccaccio's Filostrato ; so that he may indeed be said to have
, outdone his models in strength and personalness of feeling. These French and Italian poems are known to have been addressed to real ladies, and their strong language therefore had point; must we believe that Chaucer even went bejoud them, yet had
. nobody in view nearer than a mythical Greek lady1? With the conditions in the later version Lowes' view would perfectly agree ; but it attributes to F, it seems to me, tasteless and pointless extrava- gance. We may well agree with the God of Love that Chaucer's wit is full cool ; his manner here would seem very much out of character. 1 Publ. M. L. A., xix. 670-1 : cf. 620-1.
CH. IV, § 2] ITS CONNECTION WITH THE QUEEX. 107
But Lowes believes1 that some of this language would hardly fit the queen either ; that an identification of the daisy and Alcestis with the queen involves offences against taste and reason. The question is, of course, what AVC mean by identification ; it seems to me, though this is a charge that can rarely be brought against Professor Lowes' views, that his conception of it is rather bare and bald. I conceive that Chaucer wished to pay a gallant and delicate tribute to his queen ; that he adopted a well-recognized form, poetic praise of the daisy, which at once set people asking who was really meant ; his overt answer in the poem is — Alcestis ; an answer which, considering contemporary custom and the strength of his language, was hardly quite satisfying, yet took the crude edge off the identification with the queen ; the more subtle answer is indicated when Alcestis herself says at the end that the whole completed poem is to be laid as a tribute at the feet of Anne. He that had ears to hear, let him hear. A lady is ardently celebrated in the poem, which announces its own dedication and presentation to a lady ; must they not in some way be identified 1 Supposing Chaucer had wished to celebrate the queen in the Legend of Good Women, how could he have done it better 1 Obviously the daisy could not be made to speak, nor could he bring Queen Anne in person into the poem. I shall suggest presently that the poem was probably destined to be read at court ; what could be more tasteful and clever than Chaucer's method? There had to be a human understudy and intermediary, and what more suitable one could be chosen than Alcestis, the model queen and devoted wife, who had had for years such a charm for the poet?2 This tacit understanding secured delicacy, and gave him freedom ; he might express as much latria for the daisy as he pleased, and by the time it had passed through the hands of Alcestis to those of the queen it had become nothing more than a proper dulia. I do not think this is over-subtle, though of course what it makes explicit was in Chaucer's mind in part only implicit ; and it makes innocuous the warmth of the affection which Chaucer expresses. Considering that for years poets had applied similar language to ladies whom they had not always a right to address so,
1 Pp. 671-2, note.
2 Chaucer had already several times, while following a more or less common late mediaeval literary custom, foreshadowed his collection of the martyrs of love and his celebration of Alcestis. See B. D., 62-220, 330-1, 726-41, 1080-7 ; T. C., V. 1527-33, 1777-8 ; H. F., 239-382, 388-426.
108 THE LEGEND OF GOOD WOMEX. [cH. IV, § 2
and considering the free manners of the time, the customs of " courtly love," and the familiar sort of relations which we are coming to see more and more clearly had existed for years between Chaucer and the court, why should not the genial poet of forty- five or so have thus addressed his queen of twenty 1 It seems also to make the identification a little too strict and frank to see a I violation of good taste in the bare mention of Alcestis' death and going to hell instead of her husband ; was Chaucer to ignore the \ main element in her story ? And the fact that it is Alcestis who Vbids the poet present the book to the queen seems to me not in the slightest degree to contradict such a vague relation between the two as I have conceived, but rather to strengthen the probability of it.1 Finally, it seems to me that in one passage of the Prologue? quoted earlier for another reason, there is strong evidence for pre- cisely such an ill-defined but close connection of the queen with the poem, and its daisy and lady, 'as I had conceived before I noted this passage. After highly praising Alcestis,
" Than seyde Love, ' a ful gret negligence Was hit to thee, that ilke tyme thou made " Hyd, Absolon, thy tresses," in balade, That thou forgete hir in thy song to sette, Sin that thou art so gretly in hir dette ' " (537-41).
Love believes that the " my lady " of the Balade is another than Alcestis ; who is she if not the queen ? Chaucer may have landed himself in subtle difficulties3 by his hypostatic union, but some such union he clearly made. Here he sacrifices a little poetic i> propriety in order to make his compliment plainer. ^\ A somewhat close connection of the Legend of Good Women with the court circle and the queen is made particularly plausible by the close and familiar association with them which we are learning that Chaucer enjoyed. I need only recall his almost life-
1 The substitution in G of And in the "/ al foryeve" in F 450, of which Lowes (p. 672, note) makes much, I have little doubt is a scribal variation ; it comes in a long passage in which all the variants appear to be such, as we shall see presently.
2 Cf. ten Brink's . not very satisfactory discussion of this passage in Engl. Stud., xvii. 16-18.
3 Another one (it may be thought) is that while according to my suggestion it was the queen who had upbraided Chaucer for writing "the Rose and eek Criseyde," Alcestis apologizes for his having done so. Here it is the God of Love that plays the queen's part. But a critic must feel that this cold- blooded analysis rather spoils things. Chaucer's method here is not only intelligible enough artistically, but is notably delicate and clever.
CH. IV, § 2] ITS CONNECTION WITH THE QUEEN. 109
long connection with John of Gaunt, and the familiar relations which Professor Kittredge has .shown to have subsisted between Chaucer and other members of the court circle;1 the fact that pro- bably his wife was sister to Katherine Swynford, John's mistress and finally wife ; that exactly as he fell into misfortune in 1386, when Parliament began to object to the king's appointees, just so his prosperity revived in 1389, with the king's return to authority;2 that, to say nothing of many other appointments and pensions from the Crown, he had been sent to France in 1378 to negotiate Eichard's marriage, and (as I have said elsewhere)3 perhaps his later trip to Milan may have been not unconnected with the marriage- proposals of Eichard and Caterina Visconti ; finally, that he wrot/aT) the Parliament of Fowls to celebrate the betrothal of Eichard and ^ Anne, a poem written in such a light and at times even jocose vein as would have been very unsuitable as coming from the pen of any but a real friend. I should conceive Chaucer's relations with the royal family, allowing for personal differences, to have been some- thing like those between Sir David Lyndsay and the young James V. of Scotland, which account for the respectful familiarity which the former often expresses in his poems. There is a parallel to this in the admonitory tone of Chaucer's balade, Lack of Stead- fastness, obviously addressed to King Eichard.
For the connection of Chaucer and the Legend with the queen, and certainly with the court circle, there is some evidence in the allusions in the Prologue to the Flower and Leaf cult, which Professor Kittredge suggests imply some kind of a court club.4 In
1 Modern Philology, i. 1 ff.
2 Within two months in each case. Cf. Hales in Diet. Nat. Biogr. , x. 165 ; ibid., xlviii. 148. Chaucer's new appointment seems to have been connected rather with this than with John of Gaunt's return to England.
3 See pp. 41-2 above. See also Life Records, pp. xxviii. 203, 230. Other connections with the court are his intimacy with the courtier Bukton (pp. 210-11 below), and the fact that the Earl of Oxford got him his custom- house deputy.
* Mod. Philol., i. 1-2. The lines which I quote show, according to him, " that English court society, in the time of Richard II., entertained itself by dividing into two amorous orders — the Leaf and the Flower — and by discuss- ing . . . the comparative excellence of those two emblems or of the qualities they typified. If we call in Gower's testimony also, we are perhaps justified in supposing that the two orders sometimes appeared in force, each member bedecked with the symbol to which he or she had sworn allegiance." HR refers to Gower's Gonfessio Amantis (ed. by Macaulay, vol. iii., p. 453), and to L. G. W. (G), 69-70 ; T.C.,1., st. 3 ; Sq. T., 272. The Daisy cult, presumably at first independent, of course was readily absorbed by the other. It is a plausi- ble guess that the queen belonged to the order of the Flower, and therefore, celebrating her as the daisy, Chaucer is anxious to disclaim permanent partisan-
110 THE LEGEND OF GOOD WOMEN. [CH. IV, § 2
spite of his devotion to the daisy-blossom, he is anxious lest he shall be thought a partisan of the Flower against the Leaf, which he denies being (F, 191-6; G, 75-80):
' ' For, as to me, nis lever noon ne lother ; I nam with-holden yit with never nother. N&^Cnot who serveth leef , ne who the flour ; Wei brouken they hir service or labour ; For this thing is al of another tonne, Of olde story, er swich thing was begonne."
This sounds as if there were some jocose mystery about it, and (as Professor Kittredge points out to me) as if Chaucer had not yet become a member. Now the first literary expression of the Daisy and Flower and Leaf cults are in the works of Machault, Froissart and Deschamps, and further, one of these poems, Froissart's Prison Amoureuse, written in 1371, is addressed probably to Wenceslas of Brabant, Anne's own cousin, and Froissart's friend and patron.1 The second cult seemingly developed among royal ladies connected with France, and finally it involved one of John of Gaunt's daughters, in 1386 or earlier.2 May we not even conjecture that it was partly through Queen Anne that it was introduced into England ? There is some countenance for this suggestion in the way in which Gower mentions the Flower and the Leaf (VIII. 2407-72) ; the companies of lovers wore
" Garlandes noght of o colour,
Some of the lef, some of the flour,
And some of grete Perles were ;
The newe guise of Beawme there,
With sondri thinges wel devised,
I sih, wherof thei ben queintised."
He thus connects the Flower and Leaf cult with the new Bohemian fashions introduced by Queen Anne.
Two or three passages in the poem suggest that Chaucer had in mind to read it aloud in a circle of his friends, presumably at court.3 At the end of the Legend of Phyllis, he says (2559-61) :
"Be war, ye women, of your sotil fo, Sin yit this day men may ensample see ; And trusteth, as in love, no man but me."
ship. If the Leaf people wish to make him a member, he will not decline. On all this cf. an article by G. L. Marsh in Modern Philology, iv. 121-167.
1 See Lowes, in Publ. Mod. Lang. Assoc., xix. 600.
2 Kittredge in Mod. Philol., i. 4.
3 I find Mather makes the same suggestion (Chaucer's Prol., etc., xxiii. ).
CH. IV, § 2] ITS CONNECTION WITH THE QUEEN. Ill
In the Lee/end of Hypsipyle. he says of Jason (1554-5) :
" But in this hous if any fals lover be, Right as him-self now doth, right so dide he."
I take this as explaining a phrase in G 85 :
"For myn en tent is, or I fro yow fare, The naked text in English to declare Of many a story." x
Does not this also account for the informal, colloquial, jocose and even frivolous tone which is more striking in the Legend 2 than in almost any of Chaucer's poems, even than in the Canterbury Tales, which are represented as orally delivered ? Does it not especially account for Chaucer's jocosely classing himself, in G, among " old fools" (262,315)]
It will be recollected that according to Lydgate it was the queen that dictated the subject of the Legend:
" This poete wrote, at the request of the quene, A Legende, of perfite holynesse, Of Good Women, to fynd out nynetene." 3
1 This passage is Koeppel's chief argument for believing the G-prologue meant to be delivered as a Canterbury Tale (Literaturblatt, xiv., col. 52). Cf. p. 89 above. But allusions to the practice of reading aloud are not uncommon in Chaucer's works ; see A.A^l65=Q.-> T C! , T ..450, II. 30, 43, 1751 ; 01. T., 1163 ; even Pars. T., lQST~a^TAstrolabe, Prol., 48. Cf. also Lounsbury, Studies, i. 228.
2 Cf. the end of the Cleopatra (703-5) :
" Now, er I finde a man thus trewe and stable, And wol for love his deeth so freely take, I pray god lat our hedes never ake ! "
See also 863, 1076-7, 1383, 1557, 1887, 1893, 2177-80, 2227, 2490-3. I may ask, by the way, whether the intimacy with Minos as infernal judge which produced the rather superfluous apostrophe to him in one of the above passages, 1886-8, was not due to his prominence in the Divine Comedy (Inferno, V. 4-24, and elsewhere), rather than to the Aeneid, where he is barely mentioned (VI. 431-3).
3 See Skeat, III. xx. "Lydgate can hardly be correct," according to Skeat, for if Chaucer had done so, " he would have let us know it." Why, since by hypothesis he was writing for the queen and not for us ? Lydgate's testimony is also rejected by Pollard (Globe Chaucer, p. xlv.), but is accepted by Koch (Chronology, pp. 43-4), and Bilderbeck (p. 84 ; cf. 88, note). I have even suggested already the occasion of her (not very serious) request ; see the chapter on the Troilus, p. 17, in reply to Lowes' suggestion that L. G. W. is the response to a supposed sensation produced by the first appearance of T. C. Queen Anne, a foreigner, coming to England in December, 1381, would hardly have been able to read the Troilus and the Romance of the Rose much before the date of the Legend ; after she had done so, what more natural than that she should reproach the poet for Ids cynical taste, and tell him to write now on the other side — to accomplish his desire of writing on Alcestis
112 THE LEGEND OF GOOD WOMEN. [CH. IV, $ 2
I have shown earlier that Lydgate was in a position to know about the time when the Troiliis was written, and I see no reason why the above statement, which is very unlikely to have been made up groundlessly, should not be correct.1 To substantiate it there is very good internal evidence. For one thing, I have said earlier that Chaucer's manner of mentioning the Palamon suggests that it was unfinished. Why did he drop it and begin something else /'(returning to it later), unless on external pressure 1 But above all, / why did Chaucer, to whom poetry was an avocation, and who was / constantly leaving things unfinished, continue this poem long after 1 it^had become a burden to him? At times, as we shall see, the style is almost careless, and Chaucer expresses far more sense of haste and weariness than in any other of his works.2 At the end of Prologue F, 570-7, Love tells him to be brief,3 which is certainly more likely to be the poet's own excuse than the record of a com- mand by his patron ; so even at first he felt the task to be a large one. At times he seems to be spurred on only by a sense of duty, and shows a sense of the monotony attending his subject. He will not describe Cleopatra's wedding-celebration lest, having undertaken so much else, he should have to omit matters of more consequence (616-23) ; it would be loss of time to say why Dido came to Lybia, and he does not care to (996-7) ; he would to God he had leisure and time to rhyme all Jason's wooing (1552-3) ; Hypsipyle's and Medea's letters in Ovid would be too long to write
which he had expressed at the end of the Troilus ? He says so very much, with such iteration, about the faithlessness and dangerousness of men, that the whole poem is clearly, as Lowes points out, a rejoinder to comment produced by the Troilus and the Ease, yet I have tried to show that the Legend cannot have been written till ten years or so after the Troilus ; if it was a rejoinder to the general comment evoked by the latter, Chaucer certainly was, as-Lowes says, belated. But if it was the Qneen that chaffed him, all is explained. He felt, of course, in duty bound to carry out her suggestion ; the Prologue he wrote con amore, but the legends without enthusiasm. We may conjecture that after a time the queen "let him off," which accounts for the unfinished state of the work (tins in answer to Pollard, Globe Chaucer, p. xlv.).
1 Chaucer rather distinctly suggests in the Legend that the writing of the Troilus and of the Romance of the Ease was encouraged, at least, by some one of high station (F, 366-7 ; G, 346-7) :
" Or him was boden make thilke tweye Of som persone, and durste hit nat withseye."
2 Kn, T. show's a sort of conscious rapidity, because he was always aware of an original five times as long ; so does M. L. T., though without the same reason. But the manner of L. G. W. is quite different.
* This prosaic and gratuitous passage is omitted in G.
CH. IV, § 2] ITS CONNECTION WITH THE QUEEN. 113
(1565, 1679) ; in telling of Lucretia he will be brief and " touche but the grete " (1692-3) ; the tale of Minos and Nisus' daughter would be too long for him (1921) ; he is weary to tell of Tereus, and it is time he should make an end (2258, 2341, 2383) ; he says little of the reception of Demophon by Phyllis because he is sick of his subject, and must hasten him in his Legend, which he prays God to help him finish (2454-8); he will rehearse but a word or two of Phyllis' letter, for he will not vouchsafe to " swinke " on Demophou, " ne spende on him a penne ful of inke " (2490-1); he cannot write all of Phyllis' letter, " for it were to him a charge," but will repeat it only here and there where it is good (2513-17) ; and he fears that the tale of Hypermnestra may be too long (2675). Chaucer, the busy man of the world and of affairs, and in his leisure the easy and graceful poet, was not used to groaning over distasteful literary tasks, like the plodding Lydgate ; he simply dropped them. What was it here that aroused his sense of duty, unless somebody was urging him on whom he did not like to disappoint 1 l
Finally, the poem is dedicated, in a very graceful manner, to the queen.2 At the very end of the F- prologue (496-7) Alcestis bids him,
"Whan this book is maad, yive hit the queue On my behalf e, at Eltham, or at Shene."
This fact throws new light on his references to the Flower and
1 One more bit of a suggestion as to a close connection between L. G. W. and the qneen may be worth a foot-note. It must strike every one at once as odd that Cleopatra should appear as an estimable martyr to love. The account of her in L. Annaeus Floras i^on the sources, see Bech, Anglia, v. 314-18), and elsewhere, hardly explains this ; perhaps that in Orosius (VI. 19) is the least unfavourable, but Chaucer's high conception and praise of Antony and Cleo- patra are unparalleled anywhere, so far as I know. His account was clearly written from memoiy, but he cannot have been unaware of the changes he made. Furthermore, why does Love make it such a point that Chaucer shall begin with her ? Now it will be noted that of all the martyrs celebrated, Cleopatra is the only queen, and the only woman except Thisbe (the legend of whom comes second}, whose lover is quite blameless toward her. Just as Chaucer highly praises Antony, of Pyramus he says (917-19) :
"Of trewe men I finde but fewe mo In all my bokes, save this Piramus, And therfor have I spoken of him thus."
Chaucer may have felt a lack of delicacy in celebrating his own enamoured queen in the Prologue, and then immediately recounting the tales of other queens and women basely betrayed by their lovers.
2 Mr. Pollard (Acad. , no. 1766, p. 228) suggests that this no more constitutes a dedication to her than the allusions earlier to the French poets are a dedication to them. I fail to see the parallel, and can hardly conceive a method of indicat- ing a connection with the queen more worthy the name of dedication than this.
DEV. CH. I
114 THE LEGEND OF GOOD WOMEN. [cH. IV, § 2
Leaf cult, and his expectation of reading the poem, aloud at court, and lends countenance to the belief that he wrote it, as Lydgate says, at the request of the queen. After all this we shall surely not be unprepared to find evidence that the queen was definitely celebrated in the Prologue.
Lowes himself was the first to make the important observation that the omission from G of the dedicatory couplet, F 496-7, on the presentation of the poem to the queen, "at Eltham, or at Sheen," is probably due to consideration for the feelings of King Richard after the death of his dearly-loved wife.1 To most bereaved persons it would be a doubtful kindness to remove all references to the departed, but Eichard was emotionally eccentric. Chaucer's omis- sion of the reference to Anne as alive " at Eltham or at Sheen " is a perfect literary parallel to Richard's conduct.2 As Lowes points out, he caused the manor of Sheen, where she had died, to be destroyed, though it had been a.favourite royal resort. Further- more, for a whole year, according to the Monk of Evesham, he avoided every spot, except churches, associated with her.3 Will not this be paralleled if we find that Chaucer omits all reminders of the queen from the Prologue, and will it not explain his doing so? Is it not possible, for instance, that the Legend of Good
; "Women was a favourite poem of Richard's, but that he could not ndure in it specific reminders of his lost wife ? 4 Richard was an rratic member of an erratic family. •>>l\T
1 Pub. Mod. Lang. Assoc., xx. 780-1.
fl ? Yet Mr. Pollard (in his criticism of Lowes' article, Academy, no. 1766, p. 228) prefers to think Chaucer struck out the couplet because the queen had not prevented his loss of office in December, 1386, rather than because she was dead. He cannot believe that L. G. IV. was taken up again after the G. T. were begun, and thinks that in encouraging the royal grief Chaucer would have been childish. But was this anything like as childish, to say no more, as omitting the compliment for the reason which Mr. Pollard suggests ? Legouis (p. 63) agrees Avith Pollard in being unwilling to believe that Chaucer would concern himself with the abandoned Legend when once he was started on the G. T. But why should he not care to handle again that Prologue about which both these critics are so enthusiastic ?
3 "Set nee in loco [sic} aliquem, ubi sciebat illam perante fuisse, per totum annum sequentem introire dedignabatur, prseter in ecclesiam " (Hist. Vitas ct Itegni Me. II.. ed. Hearne, 1729 ; pp. 125-6). Richard's grief and demon- strativeness are illustrated by the fact that beside her recumbent effigy on her tomb in the Abbey he caused his own to be put, ' ' with their hands clasped together" (Gairdner, in Diet. Not. Biogr.,i. 422-3). As Clerk of the Works, 1389-91, Chaucer had had oversight of the manor of Sheen and the "lodge" at Eltham. See also Adam of Usk's CJironicon (ed. by Sir E. M. Thompson, 1994), p. 9.
4 Cardinal Manning, after the death of his much-loved wife, would never mention her to anybody (Purcell's Life, i. 123).
CH. IV, § 2] ITS CONNECTION WITH THE QUEEN. 115
What seems to me far the strongest argument in favour of the orthodox view that .Chaucer meant to celebrate the queen in the Legend of Good Women, and in some sort identifies her with the daisy and Alcestis, is the fact that this theory alone will account for the most puzzling peculiarities of the revision. We have already taken it as proved, especially by Lowes' parallels to the French poems, that version F is the earlier; but we have taken it as proved only on condition that we can account for certain diffi- culties. These are the abolition in version G of almost all the warm feeling, and with it many excellent passages ; secondly, the giving up of the suspense as to who is the lady of the Balade, and the lady who comes with the God of Love ; and thirdly, the fact that, although she is repeatedly named in G, Chaucer at the very end affects not to know who she is. Without the connection with
/ the queen, all these I believe to be quite inexplicable ; with it, all seems clear.1 . .
> Now, forgetting all this for a moment, let us examine the facts — the lines peculiar to F, which Chaucer deliberately omitted from G, if that is the later.2 These lines number 135^ Of these, 50 occur in two long passages; i.e. 152-77 contain the description of the birds, etc., in the vein of the Romance of the Hose, and 552-65 and 568-77 consist almost wholly of directions as to choice of subjects and brevity of treatment; for the omission of both we have seen that Chaucer had excellent reason. Of the remain- ing 85 lines, 153 are of miscellaneous and indeterminate character.
, The other 70 are connected more or less closely with the hearty personal feeling ; the poet repeatedly expresses his pleasure in the daisy, and warm love to it,4 calls on lovers to help him, describes his eagerness to see it and how he kneels to watch it and reclines thei'e all day, he praises the flower anew, introduces his Balade in
1 The principle on which Legouia based his discussion of the question of priority was aesthetic. "En 1'absence de temoignage direct qui tranche la question de prioiite, le bon sens dicte la regie suivante : si Chaucer a pris la peine de remanier son Prologue, c'est afin de le rendre plus parfait " (p. 59). This simple principle utterly breaks down under the failure of the critics to agree. The only way in which Lowes makes his study convincing to a reader is by almost ignoring it. To make Chaucer's procedure intelligible, I maintain that a different guiding clue is necessary.
- Skeat in his large edition marks them with an asterisk.
3 I put the following into this class : 101-2, 120, 143-4 (probably omitted by accident), 201, 229-31, 335, 348-9, 357, 368, 380.
4 Cf. the substitution of G 58, "As wel in winter as in somer newe," for F 56, " And I love hit, and ever y-lyke newe."
116 THE LEGEND OF GOOD WOMKX. [CH. IV, § 2
honour of his lady, whom he praises; records the women's song in praise of the daisy, calls it Love's "relik," says that the book shall be presented to the queen, and is reproached by Love for omitting Alcestis from his Balade, since she is the model ? of lovers. That is to say^over half of all the lines omitted are directly connected with the personal feeling in prologue F ; or, dis- regarding two unified passages, the omission of which has already been easily accounted for, about five-sixths.
Moreover, most of these passages are not only excellent in themselves, but leave the G-prologue noticeably poorer. Their omission is the reason why it is generally regarded as the inferior version ; even of those who believe it is the later, Koeppel, as we have seen, regards it as less rich and as injudiciously revised, and Lowes admits that F " is even the more delightful version of the two."1 We miss particularly the beautiful expression of the poet's love (83-96) which Lowes has shown to have been derived from the Fildstrato, the agreeable picture of him as reclining all day long in the meadows watching the daisy and kneeling by it when the procession enters,2 and the deliciously quaint line where Love says of the daisy :
" Hit is my relik, digne and delytable" (321).3
I am quite sure that a candid examination of the two versions will show that almost all the points of superiority in F, which are not trivial or debatable, are directly concerned with this matter. >
How is all this to be accounted for ] Koeppel thinks Chaucer
I revised carelessly and hastily ; Lowes_thinks that in F he had
"allowed himself to go on, adding for the sake of its beauty detail
after detail as one recalled another, until his lines are like the
' costume of the Squyer," that " the omissions in A [G] will then
be amply accounted for if we suppose Chaucer to have come
back to the Prologue, the spell of the marguerite songs no longer
upon him, with the unity of his plan the dominant motive in
his mind";4 that it was a "sterner sense of the subordination
of beauty of detail to the demands of the artistic whole that
1 Publ. Mod. Lang. Assoc., xix. 683, note.
2 G less picturesquely lias him " lening faste by under a bente " (234).
:1 But it had a much less rare poetic flavour in the fourteenth century than now, just as there was no conscious quaintuess in calling the Palladium a Trojan relic (T. C., I. 153). Cf. also L. G. W., 1310, 2375-6, etc.
4 P. M.L. A., xix. 676.
CH. IV, § 2] ITS CONNECTION WITH THE QUEEN. 117
seems to have underlain the excision "* and condensation. But
'. where else does Chaucer show any such austerity? Certainly not in the Canterbury Tales. Moreover, it seems to me the mere extent of the revisions indicates that Chaucer was not simply trying to improve things. The changes are beyond all comparison greater than those in the Troilus, which was perhaps not revised till after a number of years. Presumably Chaucer made his first version as good as he was able. Would it not be almost a self -stultification, a confession of weakness, so utterly to recast a carefully-studied poem ? If there is no personal bearing in F, the feeling which it expresses seems extravagant ; but if so, why did Chaucer put it in, we may ask, if his judgment later required him to omit it ? A long and elaborate poem, much more than a prose-work, must be a pro- duct of prolonged planning and workmanship ; however spontaneous ^ it may seem, it only has the art which conceals art. Are we to suppose that Chaucer's taste changed so extensively in a few years? I fully agree with Lowes that the plan of G- is improved, and that Chaucer did well to rearrange it ; but I do not believe that he was such a tasteless, hit-or-miss and unintelligent critic that, on one of the rare occasions when he revised an older poem, he impoverished it so much and so needlessly that posterity can hardly tell which is the revised version. To my mind all this is a convincing argument
, that he had a reason other than purely aesthetic to guide him in his
I revision.2
->> The second of the points of superiority in F,(of which I have
1 Publ. Mod. Lang. Assoc., xx. 799, note.
2 Besides showing consideration for Richard's feelings, Chaucer may have felt that there was no point in thus celebrating Anne after her death ; the Prologue to the Legend was hardly suitable to be turned into an In Memoriam. There is ample evidence that mediaeval poets sometimes rededicated their works. Gower transferred the complimentary notice in the Confessio from Richard II. to the future Henry IV. (cf. Macaulay, II. xxi., 11.), and Frois- sart that of the Mtliador from the Duke of Luxemburg to the Comte de Foix (see Kittmlge in Engl. Stud., xxvi. 323-4). But these cases lend no coun- tenance to the view that Chaucer may have abolished his laudation of the queen before her death, which I agree with Lowes is inconceivable. Gower, who transferred his compliment during the reign of Richard II., was a landed gentleman, independent of court favour, with uncompromising political and moral convictions ; Chaucer, on the other hand, was largely dependent on court favour, for which during the latter part of his life he was more or less suing, and, even had he been such stun" as martyrs are made of, can have had no adequate reason to inflict such a slight on the queen. These considerations do not seem to have struck ten Brink, who dates the poem 1393, or later—- possibly, that is, before the queen's death (Engl. Stud., xvii. 20 ; Koch justi- fiably objects, Chronology, p. 85) ; or Koeppel, who dates it even earlier (Engl. Stud., xvii. 198).
118 THE LEGEND OF GOOD im.VK.V. [cH. IV, § 2
postponed the discussion, is explained in the same way. In F, the lady escorted by Love, the heroine of the Balade, is anonymous; this, aside from the unintelligible and clumsy way in which the Balade is introduced, is clearly an advantageous bit of suspense.1
: Not till the end of the Prologue does Chaucer, with a rush of joyful surprise, learn that she is the lady who has been his ideal for years. In G this advantage is lacking ; directly she enters, her
". name is baldly and ungracefully announced — "Hir name was Alceste the debonayre" (179). Lowes does not try to explain this, and it seems, at first, evidence for the priority of G. But I have shown
'already that Love's belief that the lady of the Balade is another than Alcestis makes Queen Anne's presence in the poem particularly clear ; if Chaucer gives up the suspense, and makes it plain from the first that the lady who enters with Love, and she who is celebrated in the Balade, are both Alcestis and Alcestis only, he
I removes one of the clearest allusions to Queen Anne.2 -*> Connected with this is the third point, which may be called the main crux, of the whole poem ; namely, the fact that after Alcestis has been named in his presence, Chaucer affects not to know who
1 she is.3 The trouble exists in both versions, but is far worse in G. The passages involved are these :
Version F. Version G.
241 And by the hande he held this 173 And by the hande he held the
noble quene noble quene
179 Hir name was Alceste . . .
255 " My lady cometh " . . . 209 " Alceste is here " . . .
262 „ „ ,, 216 „ „ „
269 „ „ „ 223 „ ,, „
341 Tho spak this lady ... 317 Than spak Alceste . . .
432 " I, your Alceste, whylom quene 422 "I, your Alceste, whylom quene
of Trace " of Trace "
459-460 " And yeve me grace . . . 449-450 " And yeve me grace . . . That I may knowe soothly what That I may knowe soothly what
ye be." ye be."
499 "Wostow . . . wher this be wyf 487 " Wostow . . . wher this be wyf
or mayde ? " or mayde ? "
505 "Nay, sir" ... 493 "Nay, sir" . . .
510-11 " Hastow nat in a book . . . 498-9 " Hastow nat in a book . . . The gret goodnesse of the quene The grete goodnesse of the quene
Alceste ? " Alceste ? "
518 " Now knowe I hir ! And is this 506 "Now knowe I hir ! And is this
good Alceste?" goodf Alceste ?"
1 So Legouis, pp. 60-61.
2 I can see no other possible explanation for Chaucer's giving up what seems to me a great merit in F ; I certainly cannot attribute it, as Lowes does <Publ. M. L. A., xix. 681, note), to an "instinct for unity." Where, once more, does he show any such stringent (if not unintelligent) method as this ?
3 Of. Lowes, Publ, Mod. Lany. Assoc., xix. 653-5.
CH. IV, § 2] ITS CONNECTION WITH THE QUEEN. 119
The inconsistency in F (between 432 and the later passages) is hardly greater than many another slip in Chaucer's poetry,1 and may as easily have come in a first as in a second version. In G the blunders are so outrageous that whatever we do we cannot reason- ably believe that Chaucer made them in straightforward writing.
, But on my theory I think they can be thoroughly explained.2 The discords with the first part are all between lines 450 and 506, and there is hardly any variation between the two versions from G 416 to 525; between these limits the only difference between the two prologues for which Chaucer must be responsible is the absence in G of F 496-7 (the direct reference to the queen); of other variants there are nine, but none are greater than scores of
1 variations among the MSS. of version F.3 There is not the slightest evidence, therefore, that Chaucer made any change between F 426-537 (G 416-525), except to omit the single couplet which directly mentions the queen. This couplet he was sure to think of, on our theory ; and it is equally suggestive that the point where he returns to revising is at F 538 if., where he omits Love's upbraiding of the poet for " negligence " in omitting Alcestis from his Balade. Is not just one explanation of all this obvious and indeed irresistible] When he began to revise he made in the early part of the Proloyue the extensive changes required by his reason for revision, and took occasion also to make certain improvements; in the latter part his interest may have failed, and at any rate he believed that only one or two scattered
1 E. g., L. G. W., 2075, 2099 ; Melib. Prol., 2154 (the word wryte); and cf. several in Kn. T., pp. 69, 70 above, and in the Legend of Ariadne (Lowes, P. M. L. A., xx. 811). We moderns were not the first to notice the slip in F 432, for opposite it MS. Fairfax has nota. The fact that it survived for years and reappears in G is an illustration of Chaucer's habit of not reading his own poetry much.
2 Cf. Binz, in Anglia Beiblatt, xi. 233-4. Koeppel thinks the inconsist- ency in G due to haste in revising (Liter aturblatt, 1893, col. 51). Ten Brink curiously ignores the whole matter. Koch (Chronology, p. 84) thinks it indicates the priority of G ; so does Bilderbeck (p. 82).
3 The lines may easily be found, being the unmarked ones in Skeat's large • edition. At times some of the F MSS. agree with G, and several times their common reading looks like the only genuine one. It must always be remembered, also, that version G is in a unique MS. In the most important variant — " I al foryeve" (F 450), "And al foryeve" (G 440) — G is probably corrupt, for Chaucer had not offended Alcestis, and she needed no exhortation to forgive him (cf. French, p. 91 ; Lowes, xix. 672, note). For my view that in these 110 lines there are almost no genuine revisions there is a good parallel in T. O., IV. 953-1085; I showed (p. 9) that the absence of important variants here indicates that the passage came in during a revision, and was not revised itself.
120 THK LEGEND OF GOOD 11'OMEX. [CH. IV, § 2
changes were requisite, and neglected to read ifc through. He quite forgot that his having given up the lady's anonymity made some of the latter part of the poem nonsense. I hope I have shown, therefore, that the furious blunder of G, and its almost complete absence from F, so far from being an argument for the priority of G, is one for the contrary view.
And so we seem to find that the belief in the lateness of G and in the close connection in F of Anne with the daisy and Alcestis support each other. If we deny the second, there are such unanswerable arguments against the first that we are completely at sea; but if we believe both, everything connected with the Prologue falls logically into place, and nothing remains for us except a discussion of the dates.1
1 Starting with the identification of Alcestis with the queen, several writers* have identified the God of Love with Richard II. ; except by Bilder- beck, the point has scarcely been argued, — it has been assumed,! quite groundlessly, I am persuaded. Bilderbeck's arguments have been so thoroughly refuted by LowesJ that I may be brief, though I believe of course that Lowes errs in denying Bilderbeck's main argument, the connection ot Alcestis with Anne. In the first place, there is no presumption in favour of the idea ; because a wife is symbolically represented, there is no reason why her husband should be ; a symbolizing of real characters is under no obligation to be so complete. It may be noted that Love speaks of and to Alcestis in a distant and almost reverential manner. The sun-crown (see F 230) abeut his head, not only a sign of royalty, but also a source of brightness, is thoroughly paralleled elsewhere, as Professor W. A. Neilson kindly points out to me. In love-allegory the god is frequently spoken of as a king or prince (see Neilson, Court of Love, pp, 74, 84, 105) ; he always wears a crown in the illustrations to the 1493 edition of the Roman de la Rose (see nos. 13, 15-8, etc., at the end of vol. v. of Jules Croissandeau's edition ; .xOrle'ans, 1880). Much of the description of him is derived from the Roman / de la Rose, arm some details, possibly, (as Child points out, Mod. Lang. Notes, ( xi. 488-90) from Boccaccio's De Genealogia Deornm. — But nevertheless, 's in Alcestis' admonitions to the God of Love (F 373-402, G 353-88) it seems not at all unlikely that Chaucer had Richard partly in mind, somewhat as Bilderbeck believes (pp. 94 ff.) and somewhat as even Lowes admits (xx. 779), though I can hardly accept the former's specific suggestions or believe that Chaucer was so impertinent as to offer indirect advice to Richard through Anne. We can hardly hope to identify any particular incidents which Chaucer had in mind (though there may be something in those which Lowes rejects on pp. 778-9), nor can any chronological conclusions (I think) be based on these passages. But to one who was familiar with his character, even during the years when his government was going well, Richard must often have given occasion for anxiety. One particular point, however, may be mentioned. Two passages are added in G, 360-4 and 368-9, in which
* Skeat (III. xxiv. f.), Legouis (p. 69), Binz (Angl. Beibl, xi. 236), Koch (Engl. Stud., xxx. 457), Bilderbeck (85-7, 103).
t Binz, e. g. speaks of "den liebesgott, hinter dem sich offenbar der kbnig Richard selbst verbirgt."
£ P. M. L. A., xix. 674-5 ; xx. 773-9. He also disposes of Bilderbeck's arguments for 1385 and 1390 as the dates of the two prologues.
CH. IV, § 3] THE LEGENDS AND THE DATE. 121
§ 3. The Legends and the Date.
The date of the first or F version of the Prologue of the Legend I think Professor Lowes has settled definitively. One interesting argument he quotes from Hales. In F 203 Chaucer goes to sleep "in a litel herber that" he had, which implies a house in the country, or at any rate on terra firma. Now for many years he had lived in a house on the city wall over Aldgate ; hut in 1385 he almost certainly left this for Greenwich, where he lived probably till 1399.1 So unobvious and circumstantial a detail ar, this of his having a little arbor it is natural to connect with the facts, not only with the poetic fiction. This gives a date at least not earlier than 1 385. But in 1903 I showed that there is no reason whatever for connect-
(after urging that a lord or king should be righteous, not wilful and tyrannous and cruel, but benign and open-eared to his people, and should "kepe his liges in justyce ") the poet says,
"And therto is a king ful depe y-sworn, Ful many a hundred winter heer-biforn ;"
he then declares that the lords should be duly honoured but the poor treated with compassion. Did not Chaucer perhaps have in mind certain passages in Richard's coronation-oath ? According to Thomas Walsingham (I. 333), IIP swore: "Tertio, ut non esset personarum acceptor, sed judicium rectum inter virum et virum faceret, et praecipue misericordiam observaret, sicut sibi suam indulgeat misericordiam clemens et misericors Deus." Part of the coronation-oath, between 1307 and 1603, is given thus by L. G. W. Legg (Engl. Coron. Records, Westminster, 1901 ; p. xxxi.) : " Fades fieri in omnibus iudiciis tuis equam et rectam iusticiam et discrecionem in miseri- cordia et veritate secundum vires tuas. Respondebit, Faciam." Now on June 3, 1388, Richard had been compelled by Parliament to renew his coronation- oath that he would observe the laws of the realm, and follow the counsels of the lords and of parliament, not those of flatterers (see the Continuatio Eulogii Historiarum, ed. F. S. Haydon, Rolls Series, 1863 ; III. 367). It should not be supposed that a side glance at Richard Avould have been felt to be dangerous or in bad taste. I have already compared Chaucer's relations to the English court with those of Sir David Lyndsay to the Scottish, and Lyndsay was free-spoken enough ; Gower is frank enough to Richard in the Confessio, and treats Edward III.'s memory with scant respect in the Mirour ; I shall show later that the Physician's Talc, seems to contain clear references to two scandals in the family of John of Gaunt, and the bnladc Lack of Stead- fastness shows no fear of wounding the royal feelings. I cannot think that Lowes quite makes his point that this passage of the Legend is wholly accounted for by the situation in the poem ; a few lines on the " natural king or lord " might be used by Alcestis in admonishing the God of Love, but what was the poet's motive for putting in so long and detailed a discourse on the " Regiment of Princes," and even in adding two passages during revision, though this part of the poem is otherwise little changed ? I cannot but suspect an extra-sesthetic reason for this addition, as for the omissions early in the poem.
1 I treat this subject at length in the next chapter, and make some modifications of Hales' suggestion.
122 THE LEGEND OF GOOD WOMEN. [CH. IV, § 3
ing the Legend with the appointment of a deputy in the custom- house, and therefore with the year 1385; and now Lowes proves that it quotes Deschamps' Lay de franchise, which was written about May, 1385, and further that Chaucer can hardly have had an opportunity to see that poem before the spring or summer of 1386.1 A date much later than this we shall presently find to be still more unlikely; therefore we may accept 1386.
-^ On the date of the second prologue, G, I have little or nothing to add to Lowes' discussion,2 which shows on various grounds that it must have been • produced some years after the first. One reason (pp. 782-9) is the jocose references in G to
] Chaucer's old age ; another (pp. 790-6) is the mention among Chaucer's works of the (probably recent) translation of Pope Innocent's De Contemptu Mundi, with which work he shows such familiarity in the Man of Law's Tale, which I hope to show is late, and in the Pardoner's Tale, which certainly is ; a third3 (pp. 800-1) is the existence of G in but a single MS., since a revised version published immediately after the original would be likely to drive it out. Another may be added — the mere extent of the alterations, even apart from those in- volved by the moving cause of the revisions. We have seen also that in regard to structure, some of its contents and the reading of which it gives evidence, it seems to place itself in the period of the Canterbury Tales. As to the exact date, we have seen that it can hardly have been written before Queen Anne's death, June 7, 1394 ; and since the revisions seem to have been made out of con- sideration for Eichard's overwrought feelings, and since by the latter part of 1396 he had so far recovered that he was willing at any rate to go through the form of marriage again, it was probably , written soon after Anne's death. The date 1394-5 seems to be • clearly indicated.
Coming to the question of the time when the Legends were writ- ten, I find that I must wholly part company with Professor Lowes.4 It is in this connection, it is true, that he made one of his best
1 Publ. Mod. Lang. Assoc., xx. 753-71. He shows that the relations of France and England were prohibitively hostile, and that Chaucer's and Deschamps' common friends could hardly have served as intermediaries before 1386.
• Ibid., xx. 780-801.
3 And also a rather strong argument, I think, for the posteriority of G.
4 For his views, which are offered with the greatest opeu-mindedness, see Publ. Mod. Lang. Assoc., xx. 802-18.
CH. IV, § 3] THE LEGENDS AND THE DATE. 123
observations. He shows very convincingly that certain details in the Legend of Ariadne (1960-2122) are due to Boccaccio's Teseide. "The prison of Theseus is a tower, which is ' joyning in the walle to a foreyne ' belonging to the two daughters of King Minos, who dwell in their chambers above. The two young women hear Theseus complaining as they stand on the wall in the moonlight, and have compassion on the prisoner. When, their plan for his escape having been formulated, they disclose it to Theseus and the jailor, Theseus proposes to forsake his heritage at home and to become Ariadne's page, working for his sustenance. In order that neither Minos nor any one else ' shal [him] conne espye ' he declares he will disguise himself in lowly wise :
' So slyly and so wel I shal me gye, And me so well disfigure and so lowe, That in this world ther shal no man me knowe.'
The proposition is of course not carried out, and the remainder of the story follows more closely the classical sources " (pp. 804-5). The resemblance is unmistakable to the account in the Teseide and the Knight's Tale of the imprisonment of Palamon and Arcite, and of the disguise and service of the latter ; it even extends at times to verbal resemblances between the two English poems. But I cannot at all agree with the chronological inference which Lowes has drawn from it,1 that the Ariadne must have been Avritten before the Tale because it contains " a decidedly inferior and rather sketchy replica of two motives already fully and artistically worked out " (p. 809).
That Chaucer did not object to repeating motives, any more than Shakspere did, may be proved again and again ; as, for example, by the borrowings in the Merchant's Tale from Melibeus and the Troilus, which will be shown in a later chapter. We have also seen clearly how little he objects to repeating phrases and lines, a thing still less to be expected. Moreover, the parallels, though striking enough when pointed out, are so unobvious that it was five hundred years, so far as we know, before any one noticed them. As to the inferiority of the "replica," I do not at all see it — just the contrary, in fact. In contrast with the pretty but very com- monplace picture of Emily walking about the conventional garden,
1 Some of his secondary deductions I have already had to combat in my chapters on the Troilus and the Knight's Tale.
124 THE LEGEND OF GOOD 11'OMEX. [CH. IV, § 3
we have the two princesses upon the wall in the moonlight,1 look- ing across their courtyard 2 to the donjon whence issue the prisoner's groans, presumably through a loophole; a romantic picture which recalls that in the Troilus 3 which so charmed Shakspere's Lorenzo, where the deserted lover
" mounted the Troyan walls, And sigh'd his soul toward the Grecian tents, Where Cressid lay that night."
Lowes also thinks that Chaucer would not have superimposed, upon "the very noble and stately figure of Theseus in the Knight's Tale," " the despicable traitor of the Legend of Ariadne." But Theseus is not a central character in the Knighfs Tale, nor is there any sign that he regarded him there with such liking that he should shrink from repeating the very familiar story of his youth.4
I should go so far as to believe that internal evidence actually faA7ours the posteriority of the Ariadne. The intimate familiarity shown with the details of the story of Arcite and Palamon, which Lowes points out (pp. 805-9) more searchingly than I have done, is more likely to have followed than to have preceded the trans- lation of it. One or two of the details look like a reminiscence of the Knight's Tale rather than of the Teseide. Palamon has been in prison (and in love about) seven years ; Theseus declares he has
1 There is a very Chaucerian touch here :
' ' Hem leste nat to go to bedde sone. "
2 1 have no hesitation in accepting this meaning from Skeat. The question was, how were people in the same thick-walled building to hear the prisoner's lamentations ? The only possible way was across a courtyard, which corre- sponds to Boccaccio's giardino. The tower was "joining in the wall" to the "foreyne," which belonged to Minos' daughters ; they lived in the large rooms above the dungeon, but when they heard his groans they were outdoors on the wall, across the court from the tower. This meaning of the word is sufficiently supported by the N. E. D., which under the third definition of the noun (in the plural) gives : " The outer court of a monastery ; also, the space immediately outside the monastic precincts. Obs., but surviving as a proper name in various places where monasteries existed." Though the earliest quotation given is of 1668, this last sentence proves that it must have been common ; the extension from a monastery to a castle is easy enough. As to the extraordinary interpretation of the word in this passage offered by Matzner (in his M. E. Dictionary) and accepted by Lowes, it seems to me, though such is the commonest meaning of the word, no less repugnant to good sense than to good taste.
:! T. €'., V. 666-79.
4 Falstaff must have been a greater favourite with Shakspere than Theseus with Chaucer, yet the dramatist did not shrink from covering with ridicule in the Merry Wives and at the end of Henry IV. him who had always been so finely master of the situation earlier in the latter play.
CM. IV, § 3] THE LEGENDS AND THE DATE. 125
loved Ariadne seven years, though it is not clear how he has known of her; there is nothing of the sort in the Teseide (Lowes, 807 ; cf. 811, note). The curious blunder which LoAves (808, note) points out in 1966, where Chaucer (according to the MSS.) puts the prison where Theseus is confined in Athens instead of Crete, is more natural as a reminiscence of his own Knight's Tale than of the Teseide. On the other hand, the deliberate variations which Chaucer introduces, such as the substitution of the moonlit wall scene1 for the sunlit garden, show a natural unwillingness to repro- duce his earlier motifs quite identically. This is the chief variation from the original ; on the principle which Lowes uses in his treat- ment of the two forms of the Prologue, is not that of two versions which is farther from the original likely to be the later?
For an early date of the Ariadne Professor Lowes .believes he finds evidence also in its style. If it was written before the Knight's Tale, it was written also before the Prologue of the Legend, and for this he thinks there is evidence in the versification — a lack of flexibility and variety as compared with that of the Prologue. But in the nature of the «ase is not a semi-lyrical poem likely to have more melody and variety of verse than a rapid narrative1? So far as the Ariadne is needlessly inferior in this respect, I agree with Mr. Pollard2 that the fact is due, not to lack of skill in the Ariadne, but to lack of care. Chaucer makes repeatedly in the Legend, as we have seen, the plainest possible declarations that he is in haste. " Technique of that sort," says Professor Lowes (p. 813), "is scarcely a thing that can be put on and off at will." But is it not always rather a matter of pains'? Hasty writing at any date will make poor verse. The particular peculiarity of style on which Lowes dwells is so striking that I think it can hardly be due to inexperience ; when Chaucer began 21 out of 43 lines (2136-78) with and, was he unconscious of the fact or unable to remedy it 1 I hold that this is simply Chaucer's rapid narrative style. In the Knight's Tale, Lowes believes Chaucer had thoroughly learned the technique which he was practising here; yet in the Knight's Tale, 1399-450, out of 52 lines 21 begin with and and 7 with that. These awd-lines are
1 It may have been suggested by an earlier passage in the Ariadne (1908-11.)
" Academy, no. 1766 (1906), p. 228. For an earlier discussion of Chaucer's verse in its chronological bearings, see my chapter on the KnigMs Talc, \>. 61.
126 THE LEGEND OF GOOD WOMEN. [cH. IV, § 3
noticeably frequent ill the (late) Canon's Yeoman's Tale ; in 1026- 35 there are 6 out of 10, in 1102-15 there are 7 out of 14, in 1228- 35 there are 4 out of 8, and in 1308-26, 11 out of 19. In the lively scene at the end of the Reve's Tale, 4292-312, out of 21 lines 12 begin with and. Granted that the Ariadne passage is an extreme instance of what elsewhere is often employed with admirable effect,1 this seems to me due rather to excess than to defect of ease. I do not at all agree with Mr. Lowes (p. 813) that "it is a fair pre- sumption that the Ariadne is unmelodious becaiise the technical difficulties of a somewhat unfamiliar metre had not yet been surmounted." I have pointed out in an earlier chapter that the differences between the stanza and the couplet are hardly so great as to signify in this connection.2
So, as I read the matter, there is no evidence for the opinion that the Ariadne was written before the Prologue. A fortiori there is none for Professor .Lowes' opinion (p. 816) "that the Prologue was written after most, perhaps after all, of the narratives it introduces." For this view I fail to see the antecedent probability which he sees. It seems to me a prologue, which gives the plan of the ensuing poem, is likely to be written early, while the zeal is still keen ; in the next chapter I shall show very good reasons for the opinion that in the Canterbury Tales the Prologue was one of the very earliest-written parts. When Chaucer had become thoroughly weary of the Legends, it is hard to believe that he would have written the Prologue with such delight, unless for some external reason, and that which Lowes suggests, that it is a poetic retort to the criticism which the Troilus had evoked, I have tried to show on chronological and other grounds can hardly be accepted.
There seems to be evidence, as well as probability, for the
1 Cf. Kn. T., 2925-66; .V. P. T., 4565-72.
2 Rather I should find in the carelessness of the Ariadne and the other Legends (so far as it exists) an indication that Chaucer was kept at his task by an external motive after his pleasure iu it had evaporated. For this there is further evidence in the numerous inconsistencies and blunders in the Ariadne which Lowes points out (p. 811, uote), and which are much greater than those which I pointed out in the Kn. T. (pp. 69, 70). Cf. also the errors in the second part of the Sq. T., which Lounsbury (Studies, iii. 318) attributes to lack of revision. If any one should object that Chaucer would have put his best work into a poem written for and at the request of his royal mistress, I reply that the defects (to call them so) are such that nothing can be more unlikely than that she would ever have observed them, considering the kind of reading to which she was probably used. Compared with the extemporaneous style of most mediaeval poetry, Chaucer's style at its poorest is finished and polished. Besides, the duty-poems of later poets laureate are rarely among their best works.
CH. IV. § 3] THE LEGENDS AND THE DATE. 127
opinion that most or all of the Legends were written after the Prologue. In the Prologue we are told that nineteen ladies entered after Alcestis and the God of Love ; and in F 554-60 the latter clearly refers him to the Balade for their names, and appoints them to be the heroines of his legends :
" Thise other ladies sittinge here arowe Ben in thy balade, if thou canst hem knowe, And in thy bokes alle thou shalt hem finde ; Have hem now in thy Legend alle in minde, I mene of hem that been in thy knowinge. For heer ben twenty thousand mo sittinge Than thou knowest."
It is true that in the Balade there are only eighteen women, and that one or two of them would hardly have been suitable ; of course when the Balade was written Chaucer had no idea of making it a table of contents, and when he wrote the above passage he probably had not carefully considered the details.1 But if he had already written the Legends and introduced several persons not mentioned here, it is difficult to see why he should have introduced this per- fectly needless passage. Now he follows the list in the Balade till half- way through the fourth legend, in which, after treating Hypsipyle, the connection of Jason with Medea leads him to deviate for her ; and later he devotes the seventh legend to Philomela, also not in the Balade. When he wrote, in the Man of Laitfs Prologue, 63-75, the list of ladies whom he states there to have been treated in his Legend, he had entirely abandoned the list in the Balade; and finally, when he came to revise, the passage in question was omitted from the Prologue. How can we avoid attributing this omission to the fact that the passage did not agree with his changed plan,2 or
1 Cf. Lowes' sensible remarks (Publ. M. L. A., xx. 817-19); and French, p. 30. On this passage cf. also Legouis, p. 65.
'J So ten Brink, Enyl. Stud., xvii. 19. By "thise other ladies" Chaucer clearly means the 19 chief ones. Koch is surely not justified in saying that Love here gives him permission to write the lives of some of the 20,000 others. (This number is a mere convention for a vast quantity ; cf. H. F. 2119, Sumn. Prol., 1695.) Therefore this passage does not relax Chaucer's bonds, but puts them on. Dr. Koch's whole criticism of the matter is so confused as to be unanswerable (Chronology, p. 85). The same may be said of Bilderbeck's (pp. 82-3) ; he implies that the indefinite number in G must l>e larger than the number 20 in F. But x > 20 is not an axiom in algebra. It may be added that, just as is the case in the Canterbury Tales, the Prologue promises so much more than was ever performed, and than Chaucer must have seen before long was likely to be performed, that he is hardly likely except at the very beginning to have made a perfectly unnecessary announcement of his design.
12$ THE LEGEND OF GOOD WOMEN. [cH. IV, § 3
rather his desire not to be held to the original one 1 The natural conclusion is that the extant Legends were written between the two forms of the Prologue.
There are some indications that the Legends were written in about the order in which they stand in all the MSS.,1 which is —
Cleopatra, Hypsipyle and Medea, Philomela,
Thisbe, Lucretia, Phyllis,
Dido, Ariadne, Hypcrmnestra.
If, as we have seen is probable, most or all of the Legends were written after the F-prologue, Cleopatra must have been written among the first, since at the end of the Prologue (566) Love bids the poet begin with it. This is also suggested by lines 616-23 :
"The wedding and the feste to devyse, To me, that have y-take swiche empryse Of so many a storie for to make, Hit were to long, lest that I sholde slake Of thing that bereth more effect and charge." -
As to the later Legends, the only references which I find from one to another are from Phyllis, no. 8, to Ariadne, no. 6. If Chaucer rearranged the poems, Phyllis should directly follow Ariadne, since they are so closely and consciously connected in subject.3 As it is, the wholly irrelevant legend of Philomela is inter- jected. Again, certainly no method is discoverable running through the arrangement. Finally, the signs of haste and weariness which I have collected above (pp. 112-13) become noticeably more frequent and intense fcoAvard the end ; and it is the last legend that is unfinished. All the indications are that the present is the order of writing.
Indications of the chronological terminus ad quern of the Legend are to be found in the fact that two non-Chaucerian works seem to betray vestiges of its influence. One of these is Gower's Confessio Amantis. In book VIII., among lovers, the poet sees a company of unhappy women-lovers, namely (2550-96) :
Dido, Medea, Progne and Philomela,
Phyllis, Deidamia, Canace,
Ariadne, " Cleopatras," Polyxena.
Dejanira, Thisbe,
1 Cf. Bilderbeck, p. 74.
2 This passage contrasts with the other indications of hurry and distaste noted on pp. 112-13. Miss E. P. Hammond calls my attention to the parallel be- tween the above passage and Kn. T. 885-8, also at the beginning of a long task.
3 Just as Hypsipyle and Medea are, which form one Legend.
CH. IV. § 3] THE LEGENDS AND THE DATE. 129
Then, after the amorous sorceresses Circe and Calypso, come the best of women-lovers, Penelope, Lucretia, Alcestis and Alcyone. It is true that the tales of all of these except Cleopatra are more or less told in various scattered earlier parts of the poem ; but it is suggestive that here occur all but two l of the ten heroines treated by Chaucer in the Legends, and of the others some entered into Chaucer's announced plan. Cleopatra comes just before Thisbe, as in Chaucer ; but it is more important that Chaucer's " Cleopatras " has influenced Gower's in other things besides her name. All that Gower says is (2572-7) :
" I syh also the wofull queene Cleopatras, which in a Cave With Serpentz hath hirself begrave Alquik, and so sclie was totore, For sorwe of that sche hadde lore Antonye, which hir love hath be."
Chaucer, at the end of her Legend, says (696-702) :
" And with that word, naked, with ful good herte, Among the serpents in the pit she sterte, And ther she chees to ban hir buryinge. Anoon the neddres gonne hir for to stinge, And she hir deeth receyveth, with good chere, For love of Antony, that was hir so dere : — And this is storial sooth, hit is no fable."
The representation of Cleopatra as dying for the love of Antony by leaping into a pit filled with serpents, and as being buried there, is confined to these two accounts,2 and no one reading the above
1 Hypsipyle and Hypemmestra.
2 These points are not in any of Chaucer's probable sources as given by Skeat (III. xxxvii. ) and M. Bech (Anglia, v. 314-8), and are probably original with him. Macaulay (Gower, iii. 547) suggests that he may have derived his idea of Cleopatra's death from Vincent of Beauvais (by a very confused recollection). The passage mentioned above seems to be the only case of borrowing between the Confessio and the Legend, unless two details in their accounts of Ariadne show mutual influence (cf. Macaulay, iii. 503). Bech in one section of his essay on the Legend of Good Women attempts to prove a number of borrowings on Gower's part (Anglia, v. 365-71) ; Skeat in a rather confused passage (III. xl. ff.) reduces them to two, but his first seems hardly significant. The only one of Bech's cases rejected by Skeat which is worth mentioning is in Conf. Am., i. 93-202, where the striking thing is the similarity of the rdles played by Venus and Cupid to those of the God of Love and Alcestis in the Legend ; Cupid is stern to Gower (though without apparent reason), and Venus is kind to him. This evidence, however, is nullified by the fact that the situation is paralleled in other amoristic allegory ; as Professor Neilson points
might easily be derived by any poet DEV. CH.
130 THE LEGEXD OF GOOD H'OMKX. [CH. IY, § 3
passages can doubt that Gower was the borrower. As to the date, Macaulay (II. xxi.) has shown that- the Confessio was finished in 1390 ; we have evidence therefore as well as probability that the Legend was as much finished as it is now not later than 1390.
The date will be thrown still further back by the connection of the Legend with Thomas Usk's Testament of Love. I believe 1 have shown already (pp. 22, 23) that Usk certainly knew the Prologue of the poem, and probably one or two of the Legends. We have seen that the Testament cannot have been written later than 1387, and almost certainly dates from that year. The Lee/end of Good Women was therefore presumably brought to an end by the latter part of 1387.1 It may well have been not much earlier than that, for Usk's connection with Chaucer's associate Brembre would doubtless give him exceptional advantages for procuring Chaucer's works.
This throwing back of the date is further confirmed by what we shall find in the next chapter as to the date of the Canterbury Tales, the beginning of which we shall find reason, partly depending on what we have learned as to the date of the Legend, to put about 1 387. We can hardly believe that the Legends were continued after the Canterbury Tales were once under way. Nor is there need of thinking that the Legend was interrupted by the conception of the more promising poem ; - it has every appearance of having run down, as it were, of itself. We have seen good reason to think that it was written rapidly, and we may assume that no more of it was ever written than is now extant — that Chaucer never told us for just what " conclusion " the tale of Hypermnestra was said.3 Therefore
other during the Middle Ages by the conceptions of Venus and of the Virgin Mary. Both Skeat and Bech find a borrowing in Covf. Am., VIII., about 2440-2750, where Cupid comes with a vast train of lovers (2456-8) ; though this too is somewhat paralleled elsewhere (cf. Neilson, Romania, xxix. 87), the influence of Chaucer is not unlikely, but the passage is really part of the one I have cited. As to the mention of the flower and the leaf in Conf. Am., VIII. 2468, Kittredge has shown (Mod. Philol. i. 2) that this is an allusion rather to contemporary life than to literature.
1 If this view is correct, of course it disposes of Bilderbeck's suggestion that the Legends were produced at the rate of one a year (see his pp. 89-91, 108). But there are many other reasons to doubt this idea.
2 Cf. Pollard, Globe Chaucer, p. xxiv.
3 Lydgate's manner of speaking of the poem (quoted in Skeat, III.xx.), and the colophon put b}' the scribe of MS. Fairfax at the head of the Prologue, indicate that they at least believed the poem to have been not nearly finished. The unanimity of the MSS. is further confirmed by a spurious Cronycle made
CH. V, § 1] THE CANTERBURY TALES AS A WHOLE.
131
the whole period of Chaucer's occupation with the 2723 lines of the Legend may have been only a few months. We have learned from Lowes that the earlier part of 1386 is the earliest likely date for the Prologue; and we have just seen that the latter part of 1387 is the latest date possible for the publication of the whole work. The date 1386-7 for the Legend of Good Women may therefore be accepted.1
CHAPTER V.
THE CANTERBURY TALES. § 1. The Canterbury Tales as a Wliole.
SEVERAL attempts have been made to find a point of departure for dating the Canterbury Tales as a whole, but few of the results seem very reliable and some of them are worthless. The conjectures which have attracted most attention have started with the idea that the basis of the poem is an actual pilgrimage made by Chaucer. This idea seems to be wholly baseless.
<»*It is quite unnecessary, of course, in order to explain the exist- ence of the poem ; I need hardly recall the various mediaeval
by C'haucier, in one of Shirley's MSS. (see Odd Texts of C?taucer's Minor Poems, Ch. Soc., 1871, I. vi.-viii.), which treats of the " nyene worshipfullest Ladyes. "
1 Professor Bilderbeck (pp. 32-44) tries to prove that Chaucer revised the first six (but not the last three) legends, and that MS. Camb. Gg contains the earlier version of them as of the Prologue. Since it seems to be quite certain that this MS. contains the later version of the Prologue, his view as to priority between his versions of the Legends is hardly possible. When we come to examine the evidence, we find, I think, no reason to change our minds. Of the two or three dozen variants which Bilderbeck quotes, none compares in importance with those in the Troilus, or those in the Prologue to the Legend, even apart from the excision of allusions to the queen. Even the readings peculiar to MS. Harl. 7334 of the Canterbury Tales, which I am convinced cannot be attributed to Chaucer, look far more genuine than these. The most favourable of Bilderbeck's cases (2008-9) :
" . . .he shal at (on) him lepe> And (To) slen hym as (or) they comen . . .,"
is not in the least striking. In no case does the variation seem to me too great to have been produced by a scribe, even unconsciously. It is natural that the text of MS. Gg should be notably different from the others, since it probably parted from them very early in the MS.-tradition. If there were a striking contrast in the number of more important variants in legends 1-6 and 7-9, we might hesitate ; but according to Bilderbeck's account, in the former there is . 1 in 53 lines, and in the latter 1 in 100, and I cannot see that they are any less important. So probability and evidence alike seem to negative the idea of revision.
132 THE CANTERBURY TALEf!. [CH. V, § 1
collections of stories in a frame, of which of course the Decameron is only one, and other things may have contributed their hints.1 In the fourteenth century story-telling must have been common on pilgrimages.2 Nor is it necessary in order to explain the vividness of the narrative.3 Absolutely all the familiarity shown with the external circumstances of the pilgrimage Chaucer would have gained from the numerous times he had passed over the same road on his journeys to the continent,4 and the two observations on the position of the sun (M. L. Prol, 1-14; Pars. ProL, 1-11) might have been taken at home as well as on the road, or have been made \ip at any time of year by a little calculation.5
1 The assumption that the Decameron must have been Chaucer's model was the mere child of ignorance, and dates from the Dark Ages, the eighteenth cen- tury; it is one of the few things which we have to forgive Tyrwhitt (ed. of 1830; I. clix.). More recently it has been denied with patriotic vehemence ; e. g. by Skeat (III. 371 ; yet in V. 270 he seems to think Chaucer to have been familiar with the Decameron), and Pollard (Globe Chaucer, xxviii.). In Italy, naturally, it is still popular. Peter Borghesi argues very unconvincingly that Chaucer must have known the Decameron (Bocc. and Ch., Bologna, 1903 ; pp. 50 ff.). Professor Cino Chiarini inclines (though without bigotry) to believe it (Nuova Antologia, Ixxii. 334 ; on pp. 148-65, 325-43, he rather agreeably introduces the G. T. to Italian readers). It will be seen that none of these writers have any evidence ; the argument is alwaj's that he must have known the Decameron. It seems to me almost certain that he did not. Hales, who thinks he did, is misleading in his arguments (Diet. Nat. Biog., x. 163). Cf. also pp. 160-1 below. The germ of C. T. is in the house of Rumour in H. F. (lines 2121-36 ; cf. A. W. Ward, Chaucer, E. M. L. Series, 95-6) ; in Piers Plowman (Seeley, in Skeat, III. 372 ; ten Brink, Hist. E. L.t ii. 140-1— he rejects a real pilgrimage as unnecessary). Chaucer had already produced an approximation to the C. T. in the L. G. W. More than this, the frame-story might develop spontaneously at any moment out of the mediaeval fondness for anecdotes and exempla, as it did in Gower's Confessio Amantis. The H. F. illustrates the point, with its sketches of ancient heroines.
2 Only singing and piping are mentioned in the dialogue between Thorpe and Abp. Arundell (in 1407 ; cf. Littlehales' Road from London to Canterbury, 51-2) ; but during the halts, anyway, we may be sure there was " taling." That it was common on pilgrimages seems to be implied in all versions of Piers Plowman :
" Pilgrimes and palmers . . . Wenten forth in heore wey • with mony wyse tales. "
(A, ProL, 46-8 ; C has vn-wyse. )
3 Cf. Pollard (Globe Chaucer, xxvii.) : "No one who lias read the talks by the way can doubt that the poet himself had travelled over the ground. . . ". Chaucer's own pilgrimage, then, may have been made in 1385." Cf. Primer, p. 100.
4 Probably also in going to Canterbury on business connected with his wardship in 1375 ; see R. E. G. Kirk in Life Records (Ch. Soc., 1900), p. xxv.
5 Nobody pretends that Dante, from whom as well as from real life Chaucer may have imitated this way of telling time, must always have just made an exact observation when he mentions the positions of the heavenly bodies. This sort of attempt to extract chronological sunbeams from cucumbers is no more tempting than it is fallacious.
CH. V, § 1] THE CANTERBURY TALES AS A WHOLE. 133
Connected with the idea of an actual pilgrimage is the attempt to discover the year meant in the Canterbury Tales from the passage in the Parson's Prologue (1-11); at four1 o'clock
. . ." the mones exaltacioun, I mene Libra,2 alwey gan asceiide."
W. Hertzberg, in his German translation of the Canterbury Tales,3 follows Tyrwhitt (iv. 335) in thinking that exaltacioun cannot be used here in the strict astrological sense, since Taurus, not Libra, is the exaltation of the moon, and Libra is that of Saturn,4 but that it must mean simply rising ; and he thinks that Chaucer meant here to hint at the year of the pilgrimage 5 — apparently in a cabal- istic way. He assumes that the journey occupied but one day, and therefore, on the basis of Tyrwhitt's reading (also the Ellesmere) for M. L. Prol., 5, that the day here was April 28. With the assistance of his " verehrter Freund " Professor Scherk, ho an- nounces that on that day within the proper limits as to years the moon could have risen at four in Libra only in 1393. Therefore the date of the imaginary pilgrimage Avas April 28, 1393.
It is unnecessary to follow this ball as it was tossed back and forth in Germany, with an occasional kick from England. By various changes and corrections in the number of days of the pil- grimage, in the MS. reading and in the astronomy, Koch,6 Skeat,7 A. von During 8 and C. Ehrhart9 find (or accept or reject) the years 1393, 1391, 1388 and 1385. This last year was fully accepted by Pollard10 in 1893.
1 Most of the MSS. read ten, which is certainly wrong ; Chaucer cannot have blundered to this dxtent.
2 Harl. 7334, and also Laud 600 (in the Bodleian Library), read " In mena Libra," which gives no sense, and is one of the Harleian eccentricities which do not look as if they came from Chaucer. MS. Camb. li reads " I meen in libra."
3 Hildburghausen, 1866 ; pp. 666-7.
4 See, e. g., Wm. Lilly's Christian Astrology (London, 1647), pp. 57, 80.
8 Similarly A. E. Brae (The Treatise on the Astrolabe of Geoffrey Chaucer, London, 1870; p. 74). He deduces the year 1388, by emending "I mene Libra alwey" to "In Libra men alawai" (the name of a star, which he says could have risen with the moon at the proper time only in that year). In broad daylight !
6 Ch. Soc. Essays, 415-7 ; Ausgeicahlte Kl. Dicht. Chaucers (Leipzig, 1880), 65-6; Chronology, 49-50, 64-6. His opinion was the same in 1902 (Pard. T., xxii.).
7 Ch. Soc. Essays, 417.
8 See his German translation of Chancer, iii. 409.
9 Engl. Stud. , xii. 469-470.
10 Chaucer Primer, p. 100. In the Globe Chaucer, however, he ignores this argument (see p. xxvii. ).
134 THE CANTERBURY TALES. [CH. V, § 1
All this seems to me entirely out of the question. The idea that the passage concerned proves a real pilgrimage supposes that Chaucer either wrote the last link of all immediately on his return, or else that he made notes of such trivialities for use years after- wards. Even if the meaning assumed for exaltacioun were possible, it would be infinitely more likely that Chaucer inserted the remark simply as indicating the sort of thing which a star-wise person would have seen if he had been there, than because he remembered seeing it. But the whole idea is practically disposed of by the fact that Chaucer elsewhere uses exaltacioun and its adjective only in the correct astrological sense, which is what any fourteenth-century reader would have understood here ; and by the fact that the other interpretation really makes Chaucer say " the moon's rising con- tinued to rise," which is almost as bad as Dr. Johnson's "observa- tion with extensive view." The only possible explanation of the passage as it stands seems to be that either the scribe blundered,1 or Chaucer, through forgetfulness ; and I do not see the least im- probability in thinking that it was Chaucer, even if he did know his astrology fairly well.2 Therefore, whether in connection with the pilgrimage idea or not, conjectures founded on this passage are to be rejected3 without qualification.
1 Tyrwhitt (iv. 336) suggests that "themones" is an error for " Saturnes." An error would be the easier because the first 10° of Libra are "the moon's lace" (Skeat, V. 445); and the second 10° of Libra are "Saturn's face" (Lilly, op. cit., pp. 58, 81).
2 Surely Lounsbury has pointed out inaccuracies enougli in Chaucer's work, and the list can easily be enlarged. As another astronomical blunder, he puts Ariadne's Crown in the sign of Taurus, to which it is just opposite (L. G.W., 2223-4). But the curious thing about the passage under discussion, which apparently has never been remarked upon, is that what Chaucer seems to imply that he saw, the sign or constellation Libra rising, he could not possibly have seen at four o'clock of an April afternoon. The passage sounds much more like a conscious reminiscence of Dante than like an observation of nature. This manner of telling time occurs again and again in the Divine Comedy ; cf.,e. g., Inf. XI. 113-4, XX. 124-6, XXIX. 10; Purg. I., 19-21, II. 1-6, 55-7, XXV. 1-3, and E. Moore's Titm-Rffcrences in the Divina Commedia, Tables V. and VI. Purgatorio, II. 1-6 is particularly suggestive :
" Gia era'l Sole all' orizzonte giunto, Lo cui meridian cerchio coverchia Gerusalem col suo piu alto punto : E la notte, ch' opposita a lui cerchia, Uscia di Gange fuor con le bilance, Che le caggion di man quando soverchia."
Cf. also Man of Law's ProL, 1-12, with Purg., IV. 15-16 :
" Che ben cin quanta gracli salit' era Lo Sole, ed io non m'era accorto.'
3 So Skeat in 1894 (V. 445).
CH. V, § 1] THE CANTERBURY TALES AS A WHOLE. 135
That Chaucer did make a pilgrimage to Canterbury at some time or other is likely enough — religionis erya perhaps, or he may even have been so modern as to wish to know how a real pilgrimage felt while he was writing about an imaginary one. All that can be said is that there is not the slightest evidence for it in the Canterbury Tales.
Some scholars have advanced the idea that by the last decade of his life Chaucer had aged too much to have written many of the Canterbury Tales, or at any rate to have planned the whole. Dr. Furnivall in 1871 x took the years about 1386 as the central period of the work, when the best tales were written, and assigned the "dull ones" to times earlier and later; to the years 1390-1400 he definitely assigned only small and inferior works, and from a pas- sage in Venus and from the supposed inferiority of the minor poems known to have been written about then he deduces " a slow autumn of decay." 2 Dr. Mather agrees with his general idea,3 mainly because of the Retractations. Mr. Pollard says : " The short poems written towards the close of his life show that the not very advanced age to which he attained pressed heavily on him, and it would be unreasonable to assign the plan of the Tales to his last decade." 4 Similarly, Professor Hales believes that practically all the Canter- bury Tales which were not earlier work were written between 1387 and 1392.5
As to the Retractations, if the poem was never published by Chaucer as a whole, as I hope to show on a later occasion, they need imply nothing more than a few weeks of other-worldliness at the very end, and surely have nothing to say as to a whole decade. The remarks in Scogan and Venus seem to me to have little more signi- ticance. In the former he refers to his portly figure and to the fact
1 Trial Forewords, pp. 16, 25 ; cf. 6 (note), 99.
2 Ibid., 28, 99. 3 Chaucer's Prologue, etc., p. xxxiv.
. 4 Globe Chaucer, p. xxvii. So also Koch (Chronol., 51-2, 69), who thinks Chaucer was in poor circumstances during this period. He does hot deny that some of the tales may have been written then, but in his table (p. 79) he recognizes the possibility only for the Parson's Tale, Pollard regards "the scheme of the Canterbury Tales as taking form during" 1386-8 (p. xxvii.). Ten Brink denies (Studien, p. 153) that Chaucer has left any works which show failing powers. Kocli lias a quaint conjecture founded on Chaucer's age. He points out that Chaucer "in the wast is schape as wel as" the Host (B, 1890), who was "a semely man" (A, 751) ; therefore " we must figure the poet to ourselves as a stately man of s,ome forty years rather than as one who already feels old age approaching, and is ' hore and round of shape' ('Scogan ' 1. 31)" (Chronol., pp. 52-3). But the Prologue to Sir Thopas is a Selbstportrdt in hardly such a photographic sense as this. 5 Folia Litcraria (London. 1893), pp. 101-2.
136 THE VAXTERBURY TALES. [OH. V, § 1
that he is no longer young, and even says that he thinks never again to wake Jhis muse (1. 38) ; the date may well be 1393.1 Yet the whole tone of the poem is light,2 and any sense of discourage- ment which may lie beneath may be accounted for by the appeal for court-favour in the Envoi/. In Venus, which was probably written somewhere near this time,3 he complains that age1 has dulled his spirit and nearly bereft his subtlety, and that close translation of elaborate verse is difficult in English (76-82); yet it certainly cannot be said to show failing powers. The evidence of these poems, therefore, is almost negligible ; and since they may date from the same period, possibly one of trouble and ill-health, they cannot be used to characterize the whole decade. Buldon, on the other hand, certainly written at the end of 1396,4 obviously is with the Canterbury Tales in spirit, and we shall see later is closely associated with some of them ; it shows a gentle cynicism, somewhat recalls the Merchant's Tale, and refers to the Wife of Bath's Pro- logue. The Complaint to his Purse, one of the last things Chaucer wrote,5 is full of cheery punning, exaggeration and flippancy. Neither shows a spirit which was incapable of pro- ducing any part of the Canterbury Tales at the same time. The sharp contrast in tone among various parts of the Canterbury Tales and other works of this period simply shows what we may be very ready to believe of Chaucer, that he was a man of moods. It seems fair, then, to say that there is no evidence here against his having written any of the undated tales between 1390 and 1400, or even, if this were not unlikely on other grounds, against his having designed the poem then. It is rather satisfactory if we can feel under no necessity of believing Chaucer ever to have had a "decline."6
Several scholars have thought the time about 1386-8 so full of change and trouble as to have been unsuitable for projecting or even working much on the Canterbury Tales. Ten Brink,7 who did not commit himself as to exact dates for that work, believed
1 Skeat, I. 556-7: Lounsbury, Studies, i. 36-42; LoAves, Publ. M.L.A., xx. 787, 792.
2 Cf. G. L. Kittredge, Harvard Studies and Notes, i. 116-17.
3 See Skeat, I. 86. 4 See pp. 210-11 below ; and cf. Skeat, I. 85.
5 Skeat, I. 88.
6 Professor Kittredge has some wise remarks on the injudiciousness of taking these words of Chaucer's very seriously, in the New York Nation, liv. 214 ; he is answering Lounsbury, who is inclined to do so. See also Kittredge's article on Scogan, just mentioned.
7 Hist, of Engl. Literature, ii. 119-20 ; cf. also Koch, Pardoner's Tale (1902), p. xxiii.
CH. V, § 1] THE CANTERBURY TALES AS A WHOLE. 137
that the political unrest of this period must have produced a deep impression on Chaucer's mind, and that his personal troubles (financial and family) may well have produced a time of serious- ness ; to this time he accordingly assigns some of the more serious works which he thinks were not till later connected with the Can- terbury Tales. In 1389, however, he points out1 an improvement in Chaucer's circumstances, to which he attributes such spirited poems as the Wife of Bath's Prologue and the Merchants Tale — still unconnected with the poem as a whole. Not till about 1390, for no very clear reason, does he recognize the proper time for the conception of the whole work.2 Dr. Koeppel, similarly, in his review3 of the Chronology, thinks that Koch assigns "a feverish poetic activity " to years too engrossed with other things to be poetically productive ; he refers to Chaucer's parliamentary career in 1386 and the misfortunes of the succeeding years, and thinks that we may suppose him to have written then little besides Melibeus, the Parson's Tale and a partial translation of Pope Innocent's De Contemptu Mundi, and that the conception of the Canterbury Tales came later, beginning with an attempt to recast the Legend of Good Women for use in it.4
All such arguments as those of these two or three German Chaucerians seem to me such as we commonly use when we have no others. Caution here seems very necessary. We know so little of those details of Chaucer's life which may have had as much effect on his state of mind as Aveightier matters, so little even of the details of his personality, that it is unsafe to draw con- clusions. There is no necessary connection between ill circum- stances and solemn literary work, and the leisure perhaps implied by the former might make such a time peculiarly productive of poetry of all kinds. Was not Chaucer just the man to beguile a dreary time, and perhaps occupy his enforced leisure, by working on his art 1 5 So out of this whole mass of a priori conjecture we seem to have gained nothing reliable.
1 Hist, of Eng. Literature, ii. pp. 123 ff.
2 He therefore seems to put such tales as the Miller 's in the very period from which other scholars have excluded all but serious and dull works. He and Koch are diametrically opposed. Such disagreement indicates some- thing wrong with the method.
3 • Literaturblatt fur germ. u. roman. Philologie, xiv. 54.
4 Cf. p. Ill above.
s Cf. Lowes' wholesome remarks to this effjcct (Publ. Mod. Lang. Assoc., xx. 792).
138 THE CANTERBURY TALES. [CH. V, § 1
I shall begin my consideration of the more important arguments with one which justifies treatment at length rather because of its interest than its weight. Why did Chaucer select a Canterbury pilgrimage as the frame for his tales ] Even though such amuse- ments were common on pilgrimages, there is a certain lack of realism, even as Dean Stanley points out,1 in representing the tales as told during the ride, and heard by any considerable number of people amid the clattering and chunking of one hundred and twenty- eight hoofs. This he was willing to overlook for the sake of other points of fitness, a large and miscellaneous assemblage doing an everyday thing in common. But is not this selection especially natural if pilgrimages to Canterbury were daily under his eyes I "Where was Chaucer living during the planning and writing of the Canterbury Tales?
On May 10, 1374, Chaucer leased the house above Aldgate for his whole life, and without the power to sublet (" alicui dimittere ") ; four weeks later, June 8, he received his formal appointment as Comptroller of Customs of Wools, etc., in the Port of London. - February 17, 1385, he received the formal permission to discharge the duties of this office through a deputy which he already had for those of the Customs office received in 1382.3 October 12, 1385, he was appointed one of the Justices of the Peace for the county of Kent, apparently to take the place of one of those appointed the previous year who had died; June 28, 1386, he was one of sixteen (all but two of whom were in the previous list) to receive a new commission for the same office.4 In August or September, 1386, he was elected Knight of the Shire for Kent.5 October 5, 1386, the house above Aldgate was leased by the city to Richard Forster, probably a friend of Chaucer's.6 March 12, 1390, he was appointed, with five others, to survey and keep in repair the bank of the Thames between Greenwich and Woolwich.7 In the Canterbury
1 Historical Memorials of Canterbury (London, 1900), pp. 213-14.
2 Life Records, pp. 190-1 ; the two records are consecutive. :f Ibid., pp. 237, 251.
* Ibid., pp. 254, 259.
5 Ibid., pp. 261-2. The sheriff of the county who signed the return had been one of his colleagues as J.P. in 1385 (but not in 1386).
6 Ibid., p. 264 ; cf. p. 216.
7 Ibid., pp. 283-5. Among the commissioners were his friend Sir Richard Stury, and apparently one of the Culpepper family which had supplied one of his colleagues as J.P. in 1385 and 1386. Two of his present colleagues served also on a similar commission for Middlesex, but Chaucer did not.
CH. V, § 1] THE CANTERBURY TALES AS A WHOLE. 139
Tales (Reeve's Pro/., 3907) there is a curious and unexplained innuendo about Greenwich :
"Lo, Grenewich, ther many a shrewe is inne."
The last stanza of the Envoy to Scocjan (43-6) addresses the poet's friend thus :
" Scogan, that knelest at the stremes heed Of grace, of alle honour and worthinesse, In thende of which streme I am dul as deed, Foryete in solitarie wildernesse."
The MSS. gloss the first line "Windesore" and the last " Grone- wich." Finally, Chaucer did not hire his house near Westminster Abhey till 1399.1
The following explanation seems usually certain and always probable. In May, 1374, when he knew that he was to receive the appointment at the Custom-House,2 he took the house over Aldgate, ten minutes' walk from his office, a little over half-a-mile.s But the way in which Chaucer vivifies French conventions in the Prologue of the Legend of Good Women is alone enough to tell us that he was a lover of nature ; so as soon as his appointment of a deputy rendered his daily presence at the office unnecessary, he moved to an easily accessible spot in the country, Greenwich. The city did not, it is true, lease the house again till twenty months after the deputy was allowed ; but, especially since the new lease was by the city and not by Chaucer himself, he may have left the house long before.4 We can hardly doubt that he was a resident of Kent when he was appointed J.P.,5 eight months after the per- mission to have a deputy. It is almost equally certain that -as
1 Life Records, p. 329.
2 Cf. Hales, in Diet. Nat. Biogr., x. 161 ; cf. also his Folia Literaria (London, 1893), pp. 87-9 (reprinted from Acad., xvi. 410, December 6, 1879).
3 Aldgate is under half-a-niile north of the Tower. The Custom-House was very near the Tower (Life Records, 290 ; cf. xxxix. ) ; it obviously would be near London Bridge, the head of marine navigation. At the present day it is between the two, and was there in 1543 (cf. Van den Wyngaerde's Panorama ; e. <y.,-in Sir W. Besant's London in the Time of the Tudors, 350-1).
4 Cf. Skeat (I. xxvi., xxxviii.). Hales (Academy, Fol. Lit., I.e.] and Lowes (Publ. Mod. Lang. Assoc., xx. 772) rather assume the contrary, and think it may have been his entering Parliament which led him to move ; Lowes also suggests that his appointment as J. P. may have been the cause. But is not this putting it wrong-end to ?
5 Cf. Stubbs, Const. Hist., ii. 272-3, and especially D. J. Medley, Engl. Const. Hist. (Oxford, 1894), 351-60 ; cf. also Statutes of the Realm, I. 364. It is interesting to note that in 1388 J.P.s were required by a re-enacted statute to hold sessions four times a year, and during the session were to be paid 4s. a day (Medley, 354, 358).
140 THE CANTERBURY TALES. [c'H. V, § 1
Knight of the Shire he was a resident. In the latter part of the preceding century, those who held that office were clearly residents ; " the office was not coveted," and at times the sheriff may almost have had to compel service.1 By 1413, it is true, apparently non- residents sometimes had served, for in a statute of 1 Henry V. it is required that knights shall be residents of their shires; hut since the same statute requires that electors shall also he, it probably does not imply any frequent deviation from the obvious and original rule.2 "It may be said that, with here and there an exception, in the early days of the representative system the counties were repre- sented by men of landed wealth and social standing, and that the election of men not possessing land in the counties they represented was comparatively rare."3 Since Chaucer must have owned land in the county outside the cities and boroughs, which sent their own representatives to parliament, since he was not a rich man and can have had but little landed property, and since Greenwich was neither city nor borough,4 therefore his land was probably his home- stead in Greenwich. In his responsibility for the south bank of the Thames from Greenwich to Woolwich there is confirmatory evidence for his residence in the former place ; and although of course the Host's remark about the tough characters who lived in Greenwich might be a well-known local hit, it is natural to take it as a jocose dig by Chaucer at himself or his friends. Perhaps his friends and he, the genial man of the world and courtier, were regarded as fast by quiet suburban Greenwich, or he may be chaff- ing his unsophisticated neighbours. Finally, it is clear that when he wrote Scorjan he was living in some small place far down the river, and there is no reason to doubt that the scribe knew what he was talking about when he glossed the allusion as being to Greenwich. Skeat shows good reason to believe that the poem Avas Avritten in 1393.5
1 Stubbs, Const. Hist of Engl., ii. 68, 90, 221-3, 232, 483.
2 Statutes at Large, iii. 1 ; cf. Medley, p. 152, and Sir Harris Nicholas' Life of Chaucer, Note S (in Morris' Chaucer, 1883 ; I. 102).
3 E. Porritt, The Unreformed House of Commons (Cambridge, 1903), i. 511 ; cf. 21, 122, 512. Seats did not begin to be in demand till early in the fifteenth century. Non-residence first came in among the representatives of cities and boroughs.
4 See Life Records, p. 262 ; T. H. B. Oldfield, Representative History of Great Britain and Ireland (London, 1816), vi. 311, and History of Boroughs (London, 1792), iii. 42 (at the end). Before the nineteenth cen- tury, Greenwich itself was represented in Parliament only in 4 and 5 Philip and Mary. 5 I. 556-7.
CH. V. § 1] THE CANTERBURY TALES AS A WHOLE. 141
Chaucer's odd calamities of September 6, 1390, afford curious con- firmation of the belief that at this time he was living in Greenwich. It may be remembered that on this day he was robbed twice, once at Westminster, of £10, by one William Huntingfield, or Richard Brerelay, and others unknown, and again at " Hacchesham," Surrey, of over £9, by Brerelay with three others (being a gang of profes- sional robbers).1 Chaucer at this time was Clerk of the King's Works at Westminster, among other places ; Hatcham is between Peckham and New Cross, near the Old Kent Road, the direct route from Westminster to Greenwich, about two-thirds of the way. The obvious explanation of all this is that Brerelay, or whoever it was, after the first robbery, knowing or suspecting that Chaucer was to carry a large sum home with him the same night,2 therefore collected part of his gang and lay in wait for him on the way. If Chaucer was not going home, how did they know where to catch him 1 3
It may be taken as a certainty, then, that from 1385 till well into the nineties (probably till 1399) Chaucer lived in Greenwich. Not only has this some possible bearing on the date of the Legend of Good Women, as Professor Hales point? out ; 4 as Professor Skeat shows, it offers a bit of evidence for dating the Canterbury Tales. Since Canterbury pilgrims went past Greenwich, Chaucer's daily familiarity with them probably dated from his residence there ; living in Aldgate he would not see them at all. The inference, though by no means necessary, is natural, that the first conception of the Canterbury Tales dates from 1385 or later.
The most important element for ascertaining the terminus a quo of the Canterbury Tales is the date of the Legend of Good Women. Lowes has shown us that it cannot antedate 1386, and Skeat has shown reason (independent of the date of the Tales) to believe that it was known to the world in 1387. We can hardly doubt that the beginning of the greater work came after the Legend ; and it may be that impatience to be at it was one reason for the sense of
1 See Life-Records (1875), pp. 8, 9, 15, 19, 28, 30 ; also (1900) xl.-xlii. The accounts are not wholly consistent, but so much is certain from the indictments.
2 Possibly in order to pay wages, or the like, at some of the "King's Works " down the river.
3 For earlier and partial treatments of Chaucer's residence in connection with his poems, see J. W. Hales, Academy and Fol. Lit., I. c.; Skeat, Chaucer Society Essays, 670-1 (cf. Chaucer, I. xlii. ) ; and cf. J. L. Lowes, Publ. Mod. Lang. Assoc., xx. 771-3.
4 Academy, and Fol. Lit., 1. c. See also p. 121 above.
142 THE CANTERBURY TALES. [CH. V, § 1
haste betrayed in the other. There is the further consideration that as, with all proper deductions of parts earlier written, the Canterbury Tales compose nearly half of Chaucer's known literary work, so it is not injudicious to allow them nearly half his literary life. The date 13871 for the commencement of the Canterbury Tales harmonizes with all that we have found already ; and also with the results of our next deliberation, as to the date of the General Prologue.
• § 2. The General Prologue.
Was the Prologue written early or late in the Canterbury period 1 Dr. Furnivall believes that it and the links were written after most of the tales.2 Skeat says,3 " The Prologue, answering somewhat to a preface, is one of his very latest works, and in his best manner ; and before writing it, he had in some measure arranged a part of his
1 An opinion favoured by critics. Tyrwhitt (I. clxii., note) thought the poem could not have been " much advanced before 1389 " ; Mather (Chaucer's Prologue, etc., p. xxxiii.) thinks "the writing and arranging of the Canter- bury Tales must have proceeded intermittently from 1387 to 1400 " ; Skeat (III. 372) thinks the poem "was most likely in hand up to the time of his death, though he probably neglected it towards the last." Pollard, however, seems inclined rather to think that Chaucer dropped the Canterbury Tales soon after 1390 (Globe Chaucer, p. xxii.). There is possible confirmation for the date 1387 in a suggestion of Skeat's ; though I must say that by itself I should attach little value to it. Excluding all years except 1386-90, and starting with the date mentioned in the Man of Law's Prologue, April 18, the second day after the pilgrims assembled, he says (III. 373-4) that the year could not have been 1389, when that day was Easter: nor 1390, when April 17 was Sunday ; nor 1386, when the pilgrimage would have been in Holy Week ; nor 1388, when April 19 was Sunday, and something must have been said of the pilgrims hearing mass. (Skeat sometimes forgets the fragmentary state of the poem.) This leaves only 1387, when they would have assembled, Tuesday, April 16, "and had four clear days before them." (I should prefer to say three ; cf. my article in Publ. Mod. Lang. Assoc., vol. xxi., pp. 478-85, on the number of days of the pilgrimage. ) The confirmation which Skeat sees in the date 1387 which he had selected for the revised Knighfs Tale must vanish if that poem is practically identical with the Palamon and Arcite ; find I have tried to show earlier that his method of arriving at it is hardly trustworthy. In writing so protracted a poem as the Canterbury Tales, Chaucer would have involved himself in some inconveniences by choosing a definite year and carrying it all through, and nothing would have been gained by so doing. Dante did, to be sure, but in rigid consistency there is a vivid contrast between the two poets. Even if he laid his plan, and wrote the Man of Law's Prologue, at the season of year of which he writes, still more if he did not, there 'is no strong ground for thinking that he would have adapted his poem to the Sundays and movable feasts of the year in which he wrote, or of any year. But the coincidence between Skeat's date and that reached by other routes may perhaps suggest that he did.
2 Trial Forewords, p. 10.
* III. 374-5. Yet he quotes Hales' evidence as to the date (to be mentioned presently).
CH. V, § 2] THE GENERAL PROLOGUE. 143
materials." When Chaucer wrote the end, at least, of the Prologue, he had probably planned and perhaps written the first group or so ; the .Knight' s Tale was ready to hand, and Chaucer's apology (725- 42) seems to have reference to the Miller's and Reeve's Tales. But if a considerable time had passed since the whole work had been designed and begun, he would hardly have announced the immense plan which we find in lines 791-5, almost at the end ; and is not a prologue which lays a ground-plan likely to come early I1 I shall present evidence later that several parts of the poem were written after the Prologue ; most of the links palpably were, since they take for granted the characterizations presented in it. Therefore, quite apart from other evidence on the date, it certainly appears that the \vhole Prologue was among the earlier-written parts of the poem ; and there is nothing against putting it immediately after the conception of the whole, as I should do.2
For dating the Prologue exactly, only one piece of evidence has hitherto been found, but happily that, so far as it goes, is conclusive. The Merchant, says Chaucer (276-7),
" wolde the see were kept for any thing Bitwixe Middelburgh and Orewelle," 3
which makes it plain that those were the ports of entry and departure for the traffic in which he was engaged.
During the fourteenth century, as is well known, there was more or less legislation in England directed to the control of trade for the benefit of the royal exchequer and of English merchants, and one of the items in this legislation was the establishment of the staple. Though the exact history of this institution is not perfectly clear, it
1 This in answer to Lowes on L. G. W. (Pull. Mod. Lang. Assoc., xx. 816 ; and cf. p. 126 above).
2 The Prologue may not have been written quite continuously. As Miss E. P. Hammond suggested, I believe, in a paper read before the Modern Language Association of America, in Madison, Wisconsin, December, 1905, lines 542-4 look like a fresh start. No doubt Chaucer left for years a blank between the Prioress and the Monk, where the " Prestes thre " now stand (164) ; it is not impossible that at this point he cancelled descriptions of the Second Nun and the Nun's Priest ; it should never be forgotten, however, that all the evidence shows that cancellation was anything but Chaucer's practice. The " wel nyne and twenty in a companye " (1. 24) Chaucer must have put in after the Prologue was practically complete, since it is hardly to be supposed that he settled on this unobvious number at the start.
3 The former is in Holland, on the island of Walcheren, at the mouth of the Scheldt, and the latter is just across the river Orwell from Harwich. The route was therefore the same as that of the modern North Sea steamers from Harwich to Antwerp.
144 THE CANTERBURY TALES. [CH. V, § 2
was an establishment in an English or continental port to which the chief products of England, wool, woolfells and leather, had to be taken before they could be sold to foreigners ; and it was connected not primarily with the collection of customs, but with the attempt to create a forced monopoly.1 Now Professor J. ~\V. Hales has pointed out,2 by a reference to Craik's History of British Commerce (cited below), that Middleburgh was the staple-port only between 1384 and 1388, and therefore concludes that the Prologue must have been written between those years. The matter may be confirmed by reference to more reliable sources of information, David Macpherson's Annals of Commerce 8 and contemporary documents.
In 1353 the staple was removed from the continent, where it had been for some time previously, and fixed " for ever " at ten places in England and several in Wales and Ireland; in 1363 the staple for wool, woolfells and hides was moved to Calais; in 1369, in consequence of the war with France, it was restored to much the same list of English towns as before; in 1376 it was restored to Calais; in 1378 merchants from countries in the extreme west of Europe were allowed to come to Southampton or elsewhere instead of Calais.4 In 1382-3 (6 Bic. II.) there was a prospect of its being moved from Calais, in consequence of a treaty with the Flemings; in 1383-4 it was arranged to be either at Calais or at some English port.5 That the staple was still at Calais on September 22, 1383, is probably indicated by the fact that on that date the King promised to repay a loan, which the mayor and com- monalty of London had made him, by abating their subsidies, etc., to him, " and by grant hereby made that when the 20001. for the safe keeping of Calais has been fully discharged by the subsidy of 23s. 4d. a sack of wool, the collectors of that subsidy shall deliver the same to the said mayor," etc.6 It was at Middleburgh April
1 See Hubert Hall in the Gentleman's Magazine, cclv. (1883), 255-75, especially p. 257 ; R. H. I. Palgrave, Dictionary of Political Economy (London, 1901), iii. 460-2 ; George L. Craik, History of British Commerce (London, 1844), i. 120 (the account here, however, is not quite accurate). For information and references on this whole subject I am much indebted to the kindness of Professor E. F. Gay, of Harvard University.
2 In a letter to the Athenaeum, April 8, 1893 (no. 3415, 443-4), reprinted in his Folia Literaria, 99-102. See Craik, i. 123.
3 London, 1805.
4 Macphersou, i. 546-7, 566, 576, 582, 587-8.
5 RotuliParl., iii. 136b, 159a.
8 Calendar of the Patent Rolls: Richard II. (London, 1895-1902), ii. 307.
CH. V, § 2] THE GENERAL PROLOGUE. 145
20, 1384, as is shown by the1 "appointment of William Brampton, of London, governor of the merchants of the staple of wools kept at Middelbufgh, to search " for money illegally exported ; and several references show that it was still there at least in 1386 and 1387.2 In February, 1388, the Commons prayed that the staple of wool might be moved f rom " Mideburgh " to Calais on or before the next Michaelmas (September 29) ; the king granted that it should be moved to Calais or to a port in England before the next Parliament (which was held in January, 1 390).3 According to other authorities, Parliament ordered that the staple should be moved from Middle- burgh to Calais by December 1, 1388.4 It had been moved to Calais before January, 1390.5 On December 3, 1390, certain wools, woolfells and hides were declared forfeit to the crown " because shipped in Newcastle on Tyne for the staple of Calais and taken to Middelburgh in Seland contrary to the king's prohibition thereof." 6 In November, 1390, it was ordered to be moved from Calais to England by the following January.7 In December, 1390, it was still in Calais; in November, 1391, it was ordered to be within the realm.8 From 1388 to 1390, according to Macpherson,9 the staple was at Calais, and during the remainder of the century it was sometimes at Calais and sometimes at English towns. It is certain, then, that during the latter part of the fourteenth century the staple for wools, etc., was at Middleburgh from late in 1383 or early in 1384 till 1388, and then only.10
1 Cal. Pal. R., Rich. II. , ii. 397. Of this says Macpherson (i. 596, note) : " This is probably the first establishment of the staple at Middleburg."
2 See note 10 below. 3 Rot. Parl., iii. 250b.
4 Knighton's Chronicle (Rolls Series, 1895), ii. 298, 308. Cf. also Walsing- ham, ii. 177 ; Statutes of the Realm, ii. 60 ; Macpherson, i. 600.
5 Rot. Parl., iii. 268b. The Monk of Eveshara is therefore clearly mistaken or misleading in implying that as late as 1392 the staple had been at Middle- burgh (Hist. Vita et Regni Ric. II., ed. Th. Hearne, p. 123).
6 Cal. Pat. JR., Rich. II., iv. 355.
7 Stat. of the Realm, ii. 76 ; Rat. Parl., iii. 278a.
8 Rot. Parl., iii. 279b, 285a.
9 i. 600, 602, 604, etc.
10 I give here certain further items about Middleburgh and Orwell as ports. Before the establishment of the staple at Middleburgh some persons had been allowed, by royal patent, though it was against the ordinance of Parliament, to carry wools, etc., to Middleburgh and elsewhere. The right was guaranteed by Parliament, with reference to Middleburgh, in 1372 (Rot. Parl., ii. 315b). But that both Middleburgh and Orwell were relatively unimportant for Eng- lish commerce except when the staple was at the former place is indicated by the fact that, while both appear frequently in the Cal. Pat. Rolls, beginning at the end of 1383, neither is mentioned during the years 1377-82 (see indexes), and each only twice between 1388 and 1392. The staple at Middleburgh is
DEV. CH. L
146 THE CANTERBURY TALES. [CH. V, g '2
From all this, two interesting deductions may be made. In the first place, it is natural to find that the Merchant was probably one of the merchants of the staple,1 and dealt in the commodity with which Chaucer was best acquainted — wool. But much more important is the fact that the description of the Merchant (and therefore, we may infer, most of the Prologue) cannot have been written earlier than 1384. That it was written some little time after this is suggested by the fact that the Merchant
" wolde the see were kept for any thing " ;
it had therefore proved to be dangerous. This is confirmed by actual incidents in 1385 and 1387.2 Therefore for two or three years there may have been agitation for a safer route, which may have been one reason for the petition of the Commons to have the staple transferred. It is not quite certain, perhaps, that the passage was written in 1388 or earlier, for Chaucer may have had in mind a definite year, or vaguer period, a little further back than the time when he wrote. But since there is not the slightest evidence
mentioned July 1, 1386, and January 15, 1387 (iii. 190, 253) ; on the latter date the king orders vessels of war to convoy certain ships laden with wool from the port of Orwell to the staple of Middleburgh. This aud the next item explain the Merchant's desire that the sea should be " kept for any thing." Under date October 2, 1385, we learn (iii. 86) that a ship belonging to Florentine merchants, laden at Middleburgh and bound for England, was chased by the king's enemies, beached and abandoned at Orwell, and her cargo plundered. (See also Essays on Chaucer, Ch. Soc., pp. 470-1. According to Walsingham, ii. 217, Danish pirates greatly harassed merchants and seamen, especially the men of Norfolk, in 1395.) The fact that several contemporary authorities mention the transfer of the staple back to Calais, and all ignore the previous change, suggests that the dangers of the North Sea passage had caused considerable agitation for removal. During the session of Parliament in October, 1385, there was agitation for restoring the staple to an English port ; where it was then is not stated, but it is implied that it was at some port, other than Calais, outside of England (Rot. Part., iii. 203a, 204b, 214a). For other references to Middleburgh and Orwell in these years see Gal. Pat. Rolls, Rich. II., vols. ii., iii., iv., indexes. Under date of February 20, 1388, Onvell is shown to have been a terminus of the wool-traffic (iii. 470). I may add here a little note on the 'Shipman, who for aught Chaucer knew was from Dartmouth, and was in the habit of stealing wine. On December 6, 1386 (Cal. Pat. R., Rich. II., iii. 247), the bailiff of Plymouth, John Hanley of Dartmouth and others were appointed to compel restitution by five men of Plymouth, Hugh de "Weston of Dartmouth, and three men of Kingswear (" Kyngeswere ") for the theft of four tuns of wine from the "Cristaven" of Middleburgh. But, unfortunately, we have no information that the master of the ship " Maudeleyne," hailing from Dartmouth, was ever named Hugh de Weston (cf. Ch. Soc. Essays, 484-5).
1 He even wears a " Flaundrish bever hat."
- See note 10 above.
CH. V, § 2] THE GENERAL PROLOGUE. 147
of this,1 we may conclude that the probabilities are strong for 1387-8 as the date of this passage, and therefore presumably of the entire Prologue.
Striking confirmation for a date no later than this is afforded by a probable other contemporary allusion. The Squire
" had been somtyme in chivachye, In Flaundres, in Artoys, and Picardye, And born hym wel, as of so litel space " (85-7).
On this Dr. Mather2 remarks: "The English under Edward III. made numerous descents upon the Low Countries. Chaucer may well be thinking particularly of the campaign of 1359-60, in which he himself was. taken prisoner." But this campaign did not take place in Flanders at all ; the English army went through Artois and Picardy, but only en route to Eheims, near which8 Chaucer was made prisoner, and to Paris ; the peace was signed at Bretigny, near Chartres.4 Chaucer no doubt did think of his own maiden campaign, but it can hardly have supplied him with his geography. Moreover I find in Walsingham no record whatever of an English campaign in Flanders between 1359 and 1383, or between 1383 and 1395.
But in 1383 there was one which exactly fits the conditions.5 In May of that year, Henry le Despenser, the militant Bishop of Norwich, with the benediction of Pope Urban VI., and to the indignation of John Wyclif, led from England an expedition, which he gave all the airs of a Crusade, against the schismatic
1 If there is internal evidence of adaptation to any year, that year, as we have seen, is 1387. 3 Chaucer's ProL, etc., p. 5.
3 At " Betters, " according to the contemporary account; i.e. at Rethel, as Lounsbury seems to have been the first to point out (Studies, iii., 452). J. W. Hales (Diet. Nat. Biogr., x. 157) says "Retiers in Brittany," which is certainly wrong. Rethel and Rethers were different forms for the same name (see Kervyn de Lettenhove, Froissart, xxv. 228).
4 See Walsingham's Historia (Rolls Series, 1863), i. 287-90.
5 The best early account is Walsingham's, ii. 71-104. Froissart gives a very detailed one in book II., chaps. 207-14 (ed. by J.-A. Buchon, Coll. des Chron. Nat. Franf., Paris, 1826; vol. xxxii. pp. 413-71; see also the translation by T. Johnes, pt. II. chapters 131-45). See also Eulogium Historiarum a Monacho Malmcsburiensi (ed. by F. S. Haydon, Rolls Series, 1863), iii. 356-7 ; Malverne, in Higden (Rolls Ser., 1886), ix. 15-26 ; Chronicon a Monacho S. Albani (R. Ser., 1874), pp. 355-7. The fullest modern accounts are The Crusade of 1383, by G. M. Wrong (London, 1892) ; and Der Krcuzziuj des Bischofs Heinrich von Norwich im Jahre 1383, a dissertation by Gerhard Skalweit (Konigsberg, 1898). For a fuller bibliography, see Diet. Nat. Biogr., xiv. 412 ; and Skalweit, pp. 5-7, 75-83.
148 THE CANTERBURY TALES. [CH. V, £ 2
French adherents of the antipope Clement. For political reasons, the greater part of the campaign was in Flanders, though the Flemings were as good Urbanists as the English, especially about Gravelines (G ravenynge) , Dunkirk, Ypres and Bourbourg ( Burburgh), all of them in that province.1 But in August the bishop, hearing that the King of France was como to Amiens with an army, entered Picardy 2 with a part of his force, and defied the king to battle ; his defiance not being accepted, he returned to Gravelines. He must have passed through Artois going each way. He ended his short " chivachye " by surrendering Bourbourg, retreating, and destroying Gravelines ; his reception in England shortly after- wards was not cordial.
In- discussing the characters in the Prologue there is always danger, of course, that we may attribute to Chaucer a more detailed and realistic conception than he had; but at any rate everything here tits with great nicety the strikingly circumstantial account given by Chaucer. The Squire had been on a "chivachye" 3 which had not lasted long (1. 87), in exactly the region which had been covered by the Bishop of Norwich's expedition, and which had not been the scene of such events for a generation or more. His father's campaigns had all been semi-religious, of the nature of Crusades, and the Knight was just the sort of man to be imposed on by the ecclesiastical zeal of the bishop into thinking his cause a sacred one. These events created a great deal of talk and scandal, and must have been fresh in every one's mind when Chaucer wrote the passage. The account of the Squire's experiences is as detailed and specific, so far as it goes, as that of his father's, which have all been identified with real events within the lifetime of such a man. Surely the inference is not forced that Chaucer meant the Squire to have been in this expedition.
But we have now ample grounds for believing that the Prologue cannot have been written before 1387. Professor Lowes has shown that the Prologue to the Legend of Good Women cannot
1 Cf. Atlas de Geographic Historiquc, ed. by F. Schrader (Paris, 1896), plate 28 ; or the Spruner-Mencke historical atlas. Nearly all the places mentioned in the sources may be found in the Atlas des Baillages en 1789, by Armand Brette (Paris, 1904).
2 See Wrong, p. 77, and Walsingham. Froissart passes over this episode, but Skalweit shows (pp. 71-4) that his account of the Crusade is neither complete nor very valuable.
3 Froissart constantly uses in his account the verb chcvauchcr. The campaign hardly outlasted the summer.
CH. V, § 2] THE GENERAL PROLOGUE. 149
well have been written before 1386, and Chaucer can hardly have been at leisure to begin a new poem before the following year. Yet the Squire at the time of the , pilgrimage was only about twenty, which would make him only sixteen or so on his campaign. It may be attributing too much exactness to Chaucer's conception to argue the matter thus, but at any rate this objection will prove to have no weight. It will be noticed that the Squire has not just returned — it was " somtyme," "at one time;" similarly, the Crusade was not over till the fall, and this is April. Moreover, if this is not pushing realism too far, while his father has just arrived, all travel-stained, from a journey, the Squire is in most exquisite order. But, most interesting of all, sixteen does not seem to have been an exceptionally early age for the fourteenth- century soldier to enter his profession.1 In the fourteenth century people certainly matured earlier than under modern social conditions, and at a time when the military class was not expected to have very much education, what should a squire be at, when once he had got his growth ? The hero of the romance of King Horn is ripe for warlike exploits and is knighted soon after he is fifteen.2 Much evidence to bear me out is also supplied by the royal families of England and France, the members of which are no more likely to have been precociously military than others of the fighting class. On Chaucer's campaign in 1359 Edward III. was accompanied by his four eldest sons, of whom Lionel was twenty-one, John nineteen and Edmund eighteen.3 At the age of seventeen Lionel was knighted, and went on a military expedition to France. At the Battle of Crecy (August, 1346) Edward the Black Prince com- manded one of the two main divisions of the English army, and was left quite independent by his father in order that he might win his spurs ; yet he was only sixteen.4 Most striking of all, at
1 Which should not be too surprising to a generation which has allowed preparatory-school boys to kill each other in playing university football.
2 MS. Laud, 1. 18 (Herrig's Archiv, vol. 1., 41) ; MS. Harl. (in Eitson) ; omitted in MS. Camb. (E. E. T. S., 1866). In the description of the Squire there are some possible reminiscences of the romance of King Horn; compare 83-4 with Horn 93-4, 899-900 (Cambridge MS., E. E. T. S.), and 100 with 233-4 and indeed the whole account of Horn and his education. (But cf. also Mill's History of Chivalry, i. 36 ; Life Records of Chaucer, Ch. Soc., 1876, p. xiii. ; Furnivall's Manners and Meals, E. E. T. S., 1868, pp. 137-9, 369). With Horn 133-4 (MS. Harl., in Eitson) compare also Nun's Priest's Tale, 4391-2.
:! Diet, of Nat. Biog., xxxiii. 336, xxix. 417, xxxii. 109.
4 Ibid., xvii. 90 ; Green's Short History (New York, 1890), p. 226,
150 THE CANTERBURY TALES. [CH. V, § 3
the Battle of Poitiers (September, 1356) King John of France was accompanied by his four sons, of whom Charles was nineteen, Louis was seventeen, Jean under sixteen and Philippe only four- teen ; and though the three older ones fled, Philippe stood by and aided his father in his last stand, and was taken prisoner with him.1 So there is nothing whatever against our believing that Chaucer deliberately represented his twenty-year old Squire as having campaigned in 1383. But to have done so much later than 1387 would have been a real oversight, and probably would never have occurred to him. We have here, I think, genuine confirmation for the belief that at least the first part of the Prologue was written in 1387.2
§ 3. The Physician's Tale.
The Physician's Tale has been little regarded in Chaucer criticism, for the obvious reason that it is short and not of the first merit. It has usually been put in the early part of the Canterbury period, but for almost valueless reasons.3 I hope to show others, conjectural but respectable, for the same opinion.
That it comes after the first Prologue of the Legend of Good Women may be inferred on several grounds.4 The argument, used more than once elsewhere, must be used here also, that it is pre- cisely such a story as Alcestis should have scored up to Chaucer's credit, had it existed. We have also seen that the more poems in
1 Michaud, Biographic Uniixrselle, vii. 531, xxv. 297, iv. 102, xxxiii. 1)8 ; Guizot's History of Frcmce (N. Y., 1885), ii. 104. In connection with Chaucer's own 1359 campaign, this opens the door to the belief that he may have been born later than 1340, for which I believe there is not a little to be said ; at least it would make the earlier part of his life somewhat more intelligible.
2 Possibly a little more evidence for a date in this neighbourhood may be found in the Prologue. In August or September, 1386, Chaucer was elected Knight of the Shire for Kent (Life Records, Ch. Soc., pp. 261-2) ; and of the Franklin he says :
" Ful ofte tyme he was knight of the shire " (356).
Of course this might be a coincidence, but Chaucer's own office makes the detail especially natural. Moreover ' ' at sessiouns ther was he lord and sire, " and his friend and companion the Man of Law was justice in assizes by patent and by " pleyn commissioun." Chaucer was appointed J.P. for Kent in 1385, and received a full commission in 1386.
3 Cf. my Introduction ; and also p. 155 below, and J. Koch, Pardoner's Tale, p. xxii.
4 Of course it followed the Troilus. For one thing, it may be worth noting that in T. C., IV. 414 Chaucer represents " Zanzis " as a writer ; and in Phys. T., 16, correctly, and following Le Roman de la Rose, as a painter. Cf. T. R. Lounsbury, Studies, ii. 411-12.
CH. v, § 3] THE PHYSICIAN'S TALE. 151
10-sy liable couplets we put before the Prologue, of the Legend, the stranger becomes the metrical allusion in F, 562. Doubtless, there- fore, it was written later than 1386.1
A date not far from this, near the Legends of Good Women, is suggested by its general similarity to them in treatment. Here, as in them, Chaucer is singularly bald in his account and slavish toward his source. There is none of the warmth and expansive- ness which characterizes most of the Canterbury Tales. Chaucer here observes the respectful and pupillary and frigid treatment of classical story which appears in the Legend, but which he wholly got over in such a poem as the Manciple's Tale. The Physician' 's Tale in every respect is harmonious with the Legend of Good Women.
On the other hand, it probably preceded the publication of Gower's Confessio Amantis, 1390.2 I hope to show strong reason to believe that Chaucer had read and remembered Gower's version of the story of Constance when he wrote the Man of Law's Tale ; 3 now there is not only no evidence that he had read Gower's story of Virginia4 — there is striking evidence that he had not,5 but
1 It may seem as if we should put Phys. T. after the translation of Innocent's De Contemptu Mundi. We shall isee in .the case of M. L. T. (see p. 182), the peculiarly strong probability that works which quote this book followed Chaucer's version of it, which there is reason to date (probably late) in the period 1387-94 ; Chaucer certainly shows no sign of having known the original before this time. Phys. T., 280,
"The worm of conscience may agryse,"
seems at first sight to repeat a phrase from the pope's work : " vermis enim conscientise nunquam moritur " (I. 19) ; "vermis conscientiae tripliciter lacer- abit" (III. 2). But the phrase "Livers de conscience" occurs in Jean de Meun's Testament, 1939, as is shown byKoeppel (Anglia, xiv. 266) ; he shows also that Chaucer certainly quotes this poem in W. B. Prol., and probably elsewhere in the C. T. Considering his intimacy with Le Roman de la Rose, he is likely at any time to have known Jean's Testament. On the possible connection of Chaucer's phrase with Innocent, see Skeat, V. 264 ; and K. C. M. Sills, Journ. C'ompar. Lit., i. 390-1 (1903-4, the only year it Avas published), in connection with a possible borrowing by Wyatt from Chaucer ; both Sills and Koeppel declare the phrase to be common, but each quotes only three passages. All I can find are the one in Chaucer, the two in Innocent, the one in Jean de Meun's Testament, and finally one in Richard III., I. iii. 222. Shakspere is no doubt quoting Chaucer ; of course the ulti- mate source of the phrase is "Vermis eorum non moritur," St. Mark ix. 43, 45, 47, perhaps through some patristic or scholastic allegorization.
2 The date is fixed by Macaulay, II. xxi.
3 See pp. 183-6. 4 C. A., VII. 5131-306.
5 So Skeat, III. 437 ; 0. Rumbaur, Die Geschichte von Appius u. Virginia in der engl. Litt. (Breslau, 1890), pp. 12-15. The latter is certainly correct in believing also that Gower's account shows no influence of Chaucer (p. 16). Lounsbury (Studies, ii. 281-4) shows very convincingly that Chaucer did not use Livy.
152 THE CANTERBURY TALF.?. [ciI. V, § 3
merely expanded the bald and crude account in Le Roman de la Rose.1 Had he read a better account, in a large number of points he could not have failed to show its influence.2 One small detail is especially significant. Chaucer always calls the judge " Apius,'' and the accomplice "Claudius," and is even very emphatic about the names.3 Gower makes particularly conspicuous the fact that the former was named "Apius Claudius," and the latter "Marchus Claudius."4 It is hardly a risky inference, therefore, that the Physician's Tale is not only later than 1386, but antedates 1390.
But this may be further confirmed. To begin with, Professor Kittredge points out5 a probable and very interesting contemporary allusion in the tale. " It is now generally admitted that Chaucer's wife was the sister6 of Katherine Swynford, who Avas for some time governess of John of Gaunt's daughters,7 and whose career as the Duke's mistress8 and subsequently his wife is well known. Is it possible that Chaucer put the following verses into the Doctor's mouth without thinking of his own sister-in-law ? " Then he quotes Physician's Tale, 72-85, verses in which the poet reminds the mistresses in charge of lord's daughters (note that he and Le Roman de la Rose call Virginius only a knight, and lay
1 5613-82 (Me"on) ; conveniently given by Skeat, III. 435-7.
2 E. g. , the fact that Virginia was betrothed (to Ilicius in Gower, Icilius in Livy, III. 44), heightens the pathos. All this is by no means parallel to such a bit of forgetfuliiess as is mentioned in connection with the Monk's T. (see p. 169 below).
:) . . . " The cherl, that high te Claudius. This false luge, that highte Apius, So was his name (for this is no fable) " (153-5).
Chaucer follows the error of R. R. For a similar blunder, cf. House of Fame, 177-8 ; and for the opposite kind, Monk's T., 3887.
4 " At Rome whan that Apius,
Whos other name is Claudius " (5131-2) ;
" Which Marchus Claudius was hate " (5167).
5 Modern Philology, i. 5, note.
6 In the Chaucer Society's Life Records (London, 1900 ; pp. xv.-xix. ) will be found all the evidence, which makes it practically certain that if Philippa was not Katherine's sister she was her sister-in-law. See also, among other references, J. W. Hales, Athenaeum, no. 3153 (1888), pp. 404-5, on Thomas, Philippa and Elizabeth Chaucer. Even if she was neither, Chaucer must have been so familiar with the affairs of the Lancaster family that the allusion, to be noted shortly, seems obvious.
7 From before 1369 to 1382, when she retired to the country with her illegitimate children, the Beauforts. See S. Amiitage-Smith, John of Gaunt, "Westminster, 1904 ; pp. 390-1.
8 Beginning about 1371-2 ; in 1396 they were married. See Armitage- Smith, pp. 196, 433.
CH. v, § 3] THE PHYSICIAN'S TALK. 153
no stress on his rank) that they owe their positions to the fact that either they have kept their virtue, or through having formerly lost it are peculiarly fitted to safeguard that of others.1
But the whole first part of the tale has a singularly actual effect, where, after a long and detailed account of Virginia's beauty and goodness, the poet addresses " maistresses " and parents, and recommends judicious strictness. Any one who carefully reads it, I think, will grant that it has every appearance of having been inspired by personal feeling or reminiscence ; it has much more the air of having been written with interest than anything which follows in the tale, it is not even remotely suggested by anything in Le Roman de la Rose, and is not a particularly obvious out- growth of the story itself. More than this, no such serioxis, overt and practical criticism of life is to be found anywhere else in the Canterbury Tales.2 We shall not be unprepared, therefore, if we de- tect another contemporary allusion, closely connected with the other.
One of Virginia's virtues was that she avoided company too old and too dissipated for a girl of fourteen :
" And of hir owene vertu, unconstreyned, She hath ful ofte tyme syk hir feyned, For that she wolde fleen the companye Wher lykly was to treten of folye, As is at festes, revels, and at daunces, That been occasions of daliaunces. Swich thinges maken children for to be To sone rype and bold, as men may see, Which is ful perilous, and hath ben yore. For al to sone may she lerne lore Of boldnesse, whan she woxen is a wyf " (61-71).
Then comes the warning to duennas. Now one of Katherine
1 Line 79 in this passage —
" (Or elles ye han falle in freletee,) And knowen wel y-nough the olde daunce,"
at once suggests the last line of the description of the Wife of Bath in the Prol. (476), and a line at the end of the account of her original in R. R., La Vieille, who acts as jailor over the imprisoned Bel-Acueil :
" Qu'el sect toute la vielle dance " (4078, ed. Marteau).
Clearly, the Wife of Bath and La Vieille were not absent from Chaucer's mind when he wrote this passage in Phys. T. ; but, clearly also, it was this passage which recalled them to him, not vice versa. Cf. pp. 209-210.
2 Except, perhaps, in the beautiful passage on connubial conduct in FranfcL T., 761-86 ; and possibly the ironical digression at the beginning of Merck. T.
154 THE CANTERBURY TALES. [CH. V, § 3
Swynford's young charges had been all that Virginia was not.1 Elizabeth, second daughter of John of Gaunt and Blanche of Lancaster, born about 1368,2 solemnly betrothed in 1380 to the young Earl of Pembroke, and of marriageable age in 1386,3 was then introduced to society and had her first taste of "chere of court." "Altera vero fuit desponsata," according to Malverne's continuation of Higden, "conriti Penbroke puero immaturae aetatis; sed ilia viripotens tune effecta, in regalem curiam est delata ad con- spicandum gestus aulicos et mores eorum. Quam ut aspexit domi- nus Johannes Holand, frater domini regis nunc ex parte materna, vehementer captus est ejus amore, propter quod die noctuque earn sollicitavit, tamen per temporum intervalla tandem tarn fatue illam allexit, sic quod tempore transitus domini ducis patris sui ad mare per eum extitit impregnata. Unde illam incontinenti postea duce acceptante, duxit in uxoreni ante prolis exortum transivitque in Hispaniam cum illo."4 Elizabeth and her husband returned to England in June, 1388, or earlier.5 The whole episode is the more
1 This is not the first time that the people concerned in this affair have been brought in to explain a Canterbury Tale. In 1889 Professor A. Brandl tried to show that the Squire's Tale is an elaborate historical allegory (Englische Studien, xii. 161-74). Professor G. L. Kittredge (E. S., xiii. 1-24 ; he gives a large amount of valuable information), as Brandl himself admitted, promptly and utterly overthrew his opinions. It will be seen that the incident to be quoted is the most complete possible confirmation of Kittredge's position (if such were needed) ; as he conjectures from Knighton's language, Elizabeth, instead of wearing the willow, conferred it ; this also explains why the Earl of Pembroke refused to confirm his marriage with her (see Kittredge, p. 21, and cf. p. 12). Brandl would have to change his sexes ; the tercelet would be Elizabeth, the peregrine falcon young Pembroke and the kite John Holland. One authority for the matter is Knighton's Chronicle (ed. J. R. Lumby, Rolls Series, 1895), ii. 208, but he omits the scandalous inner history, perhaps out of good feeling (see Kittredge, p. 13, n. 2) ; the authority for this is John Malverne's con- tinuation of Higden's Polychronicon (ed. Lumby, Rolls Series, 1886 ; who wrongly gives the date of the incident as 1387, p. xvii.), ix. 96-7. Armitage- Smith (p. 459) says that "he is usually so full and accurate that there can be little hesitation in accepting the story, especially as it squares with every- thing known of John Holland's character and the manners of the English court at the time." He was a contemporary (died about 1415, according to Gross, Sources and Literature of English History, p. 289). The only dis- crepancy between Knighton and Malverne is that according to the former the marriage of Elizabeth and Holland took place after her departure to Spain, and according to the latter just before, in which he is supported by Froissart (cf. Kittredge, pp. 12-14). An earlier intrigue of Holland's, according to Shirley's testimony, is the subject of Chaucer's Complaint of Mars (see Skeat, I. 65).
2 Cf. Kittredge, Engl. Stud., xiii. 19.
3 The marriageable age for women, according to canon-law, was twelve (see Kittredge, I. c., p. 20) ; of course she was long past this, but Malverne implies that she had been kept away from court.
4 Malverne's continuation of Higden's Polychronicon (Rolls Series, 1886), ix. 96-7 ; also quoted by Armitage-Smith, John of Gaunt, p. 459, and cf. 310.
5 Kittredge, I. c., pp. 14-15.
CH. v, § 3] THE PHYSICIAN'S TALE. 155
likely to have impressed itself on people because Elizabeth had been really married to Pembroke, in 1380,1 so far as legally she could be to so young a boy.3 He secured a divorce either just after, or shortly before, her departure for Spain ; it is pretty clear, there- fore, if we compare the accounts of Malverne and Knighton, that she was still bound to Pembroke when her liaison with Holland began, and the inference is obvious that it was at least one cause for the divorce. The suggestion of the passage in the Physician's Tale is clear; Virginia avoided just the dangers that had led Elizabeth (only four years older than she) into ruin. Everything fits so well the passage in the Physician's Tale that, considering Chaucer's relations with the Lancaster family, even if he had been writing such a passage ten years later, he could hardly have failed to think of his sister-in-law and the then Duchess of Exeter. Yet the reference is not direct enough to have given offence in a rather coarse age ; 3 besides which, directly after (or just before) the marriage, which would then have been thought to make everything right, all concerned went to Spain and France for two years.
The bearing of all this on the date of the Physician's Tale is plain. If the allusion is admitted, the date is 1386 or later. But such a long and serious discourse as Chaucer's is likely to have been written when the incident was fresh in his mind ; and perhaps he is a little more likely to have so delivered himself when the persons whom he had in mind were out of the country. A further suggestion of about this date is afforded by ten Brink, who believes that the strong interest which Chaucer shows here in the bringing up of young girls, and his warning to parents (93-104), indicates the time just after the death of his wife, between June and Michaelmas of 1387.4 This is not unlikely, though he is plainly thinking in the main of people of much higher station than his own, or even that of Virginius. But we certainly have tolerable grounds for dating the Physician's Tale between 1386 and 1390, probably about 1388. Since we have found reason to believe that the General Prologue was written about 1387, the Physician's
1 Artnitage-Smith, p. 459.
2 Kittredge, I. c., 18-23. He was born in 1372.
3 I have suggested already, in connection with L. G. W., that Chaucer did not have to be as careful as some have supposed about exciting royal resent- ment. Cf. also the balade, Lack of Steadfastness.
4 Hist, of Engl. Lit. (London, 1893), ii. 121, note. We know nothing, however, of any daughter of Chaucer ; Elizabeth Chausier, who entered religion in 1377, cannot have been, as Hales assumes, his daughter (see Life- llecords, 337-8) ; she was probably a sister, or niece, or cousin.
156 THE CANTERBURY TALES. [CH. V, § 4
Tale may very well be the' first story written expressly for the Canterbury Tales.
§ 4. The Clerk's Tale.
On the date of the Clerk's Tale opinions have varied rather widely. While Koch, ten Brink and Mather regard it as dating from the period of the Canterbury Tales, Pollard and Skeat date it immediately after Chaucer's first journey to Italy. The latter1 believes that Chaucer learned the tale directly from Petrarch. According to Pollard2 we have " Chaucer's distinct statement that he learnt the story of Grisilde at Padua of * Fraunceys Petrak,' " whom he "may have met on his Genoa mission of 1373, when Petrarch was living at Arqua, near Padua " ; Mr. Pollard is also somehow conscious of a " general agreement " that he wrote his Englishing of the Griselda story soon after his return.
The supposed early date of the Clerk's Tale partly rests on the idea of a meeting between Chaucer and Petrarch, though it should not be forgotten that the one by no means proves the other. On the other hand, if the meeting is disproved, a date for the tale earlier than 1378 or so goes with it, as will be seen later. Pleasant though the thought may be of an interview between the two most distinguished literary men of their time, it must (I think) be relegated to the Imaginary Conversations ; as Dr. F. J. Mather has done so much to show us in his admirably thorough and judicious article.3
It must be clearly recognized at the outset that there is not a shred of evidence for such a meeting. It is not in the least needed, of course, in order to account for Chaucer's obtaining the Latin ver- sion of the Griselda story. Considering the reputation both of the Decameron and of Petrarch, MSS. of his cultivated Latinization of its last tale are likely to have been speedily multiplied. His version and its authorship were known in France as early as 1392-4, for
1 III. 454-5.
2 Globe Chaucer, p. xxv. ; cf. Chaucer Primer, pp. 66-8, and Knight's Tale (1903), p. xvi.
3 Mod. Lang. Notes, xii., coll. 1-21. Besides other more obvious references see the following : Pro. : J. J. Jusserand, Nineteenth Century, xxxix. 923-1005 ; C. H. Bromby, Athenceum, no. 3699 (1898), pp. 388-9 (and later) ; Peter Borghesi, Boccaccio and Chaucer, Bologna, 1903, 71 pp. (thinks probably Boccaccio introduced Chaucer to Petrarch, pp. 17-18) ; J. W. Hales, Diet. Nat. Biogr., x. 160. Con : S. C. Baddeley, Atlienccum, no. 3710, p. 791 (and earlier) ; P. Bellezza, Engl. Studien, xxiii. 335-6"; Sir Harris Nicholas, in Morris's Chaucer, I. 7-17.
CH. v, § 1] THE CLERK'S TALE. 157
it is the avowed source of a part of the Menagier de Paris.1 Again, it is pleaded that the Clerk tells us he had learned the tale at Padua from Petrarch's mouth; we have equally strong evidence in the Canon's Yeoman's Tale that Chaucer had known a roguish canon who cheated chantry priests.2 It seems highly probable, as Mather suggests, that this dramatic touch is due to two passages, near the beginning and end of the letter to Boccaccio 3 which con- tains Petrarch's Latin version of the Griselda story, and which we know to have been Chaucer's source ; in these passages Petrarch tells how he had communicated the tale to many of his friends, how it had been praised and sought after, and how profoundly it had affected one of them. This anecdote, and the familiarity with Petrarch which Chaucer no doubt gained by hearsay, makes such a fiction as the Clerk's meeting him absolutely natural, and even obvious.4 At first sight a rather striking coincidence suggests intimate knowledge on Chaucer's part of Petrarch's movements ; during Chaucer's first visit in Italy, because of war between Padua and Venice, Petrarch was living at Padua, where the Clerk says he saw him, instead of at Arqua, his home. But the strangeness disappears when we reflect that to the western Italian Petrarch must have passed as a Paduan ; the two places are under twenty miles apart, and Petrarch had often lived in the larger, where he held ecclesiastical preferment. But, what is especially important, Mather has shown that it would have been no easy matter to crowd a long winter journey, across the Apennines, through districts full of wars and tumults, into the short time which Chaucer had in Italy, certainly less than four months,5 with king's business to attend to in Genoa and Florence.
-1 Ed. by Jerome Pichon, Paris, 1846; see vol. i., pp. 99-125, and on the date of the work, p. xxii. There is not the least evidence, as a brief com- parison will show, that Chaucer ever saw this version ; it is striking, however, that he uses the French forms of some of the proper names, such as Saluces. The Menagier contains also the French version of Melibeus, which was the source of Chaucer's tale.
2 Skeat's arguments seem singularly nugatory (III. 454, note), if not worse. Were not poor travelling clerks one of the most characteristic classes of the Middle Ages ? And how much realism does he feel justified in demanding of the Canterbury Tales?
3 Originals and Analogues (Ch. Soc.), pp. 152, 170-1.
4 Professor G. L. Hendrickson of Chicago shows in Modern Philology (iv. 179-88) that a similar method of making citations is a literary convention as old as the Ciceronian dialogue (though he does not show how Chaucer became acquainted with it).
J See Mod. Lang. Notes, xi. 419-25 ; and cf. my article in Mod. Phil., i. 319-21. On the duration of the journey to Italy, cf. Chronicon Adae de Usk (ed. by Sir E. M. Thompson, 1904), pp. xxii.-xxvi.
158 THE CANTERBURY TALES. [CH. V, § 4
One piece of negative evidence against the meeting I think has never been allowed sufficient weight. Can any Chaucerian doubt that Chaucer would have made a rather considerable impression on Petrarch? It was late in April, 1373, that Petrarch wrote the letter to Boccaccio which includes the Latin version of Griselda,1 and if Chaucer met Petrarch it must have been in February or March. Now in his letter Petrarch tells much, as we have seen, of
1 The most difficult point of all is the date when Petrarch composed his version of the Griselda story. Dr. F. J. Mather's reconstruction, following M. Jusserand's lead, of Petrarch's procedure in regard to his last three or four letters to Boccaccio is a clever and usually satisfactory treatment of a puzzling tangle (see Mod. Lang. Notes, xii., columns 1-21). Late in June, 1373, it seems quite certain, Petrarch wrote a short letter to Boccaccio in which he said that two months earlier he had written and begun to copy a long letter to him, evidently of rather impersonal character ; relieved of the labour of copy- ing by a friend, he wrote another, nearly as long as the first, and more personal. That one of the two letters contained the Griselda story we know from a letter of Boccaccio's. In the Epistolce de Rebus Senilibus,* which Petrarch himself edited, the first-mentioned of the above three letters stands first in the 17th book, followed by two longer ones ; of these the first has every appearance of being the second he mentions, and the second the first. This latter is the one which contains the Griselda story. We cannot doubt, therefore, that the Petrarch had the Griselda letter copied about the end of April, 1373. When the Griselda part was composed is a more difficult question. It is clear that the Decameron had not been long in Petrarch's hands : " Lib- rum tuum ... ad me delatum vidi. Nam si dicam legi, mentiar, siquidem . . . tempus angustum erat ; idque ipsum, ut nosti, bellicis undique motibus inquietum, a quibus et si animo procul absim, nequeo tamen fluctuante Republica non moveri."t The reference is to the war between Padua and Venice, which began about the middle of 1372 ; Venetian troops "penetrated into the Padovano ' (November, 1372), and spread desolation through the entire district." £ By November 14 Petrarch had taken refuge in Padua.§ It is clear that the Decameron came into his hands at a time of great anxiety over the war; not earlier, therefore, than the middle of 1372. He goes on to say that he had gone through the book in a cursory way, and read more par- ticularly the beginning and the end ; the last novel so charmed him that he learnt it by heart and used to repeat it to his friends. ' ' Quod cum brevi post- modum fecissem gratiamque audientibus cognovissem," it occurred to him that those who knew no Italian would also enjoy it. "Itaque die quodam . . . calamum arripiens, historiam ipsam tuam scribere sum aggressus. . . . Qnae licet a multis et laudata et expetita fuerit, ego rem tuam tibi non alteri dedi- candam censui." It is clear from all this, especially the last sentence, that the translation must have been made at least some weeks before the end of April, when the copy was made ; Mather seems to overlook this fact when he thinks April a possible date. The earliest possible date of Petrarch's Latin version of the Griselda story is therefore the end of 1372, and the latest is March, 1873. Since Chaucer, if he went to Padua, must have been there not later than March, nor earlier than February (cf. Mod. Lang. Notes, xii. 11), it is perfectly possible, so far as concerns the date of the story, that Chaucer got it immediately from Petrarch. But obviously there is no evidence here that he did ; and I have tried to show evidence that he did not.
* See the Italian translation by Giuseppe Fracassetti (Firenze, 1870), vol. ii. 523-566).
t Originals and Analogues (Chaucer So?.), p. 151. £ The Venetian Republic, W. C. Hazlitt, i. 653. § Petrarca's Leben und Wcrkc (Leipzig, 1878), by Gustav Koerting, p. 444.
OH. v, § 4] THE CLERK'S TALE. 159
how the story had been admired and sought after in both Boccaccio's and Petrarch's versions, and in particular how deeply it had affected a Paduan friend, and how differently a Veronese had taken it. If a month or two earlier Chaucer had heard it, and begged a copy, it is strange indeed that we know nothing of the fact ; that Petrarch says nothing of "quidam advena ultimse Thulse," or "viator a partibus barbarorum adhuc profectus." He missed an admirable chance to compliment his friend. And Petrarch's own vanity is sufficiently well known; he was surely not proof against such a compliment as Chaucer would have paid him by taking such a journey to see him, nor was he too modest to mention the fact. It seems to me this argument from silence is peculiarly strong.
A few other pieces of evidence may be given that Chaucer did not meet Petrarch. He was never at all familiar with his works. Besides this story he shows knowledge only of a single sonnet. In another point he shows strange ignorance. It is well known that Petrarch's father was named Petracco,1 and that the poet's name would naturally have been Francesco Petracchi ; but that for some unknown reason he changed it.2 The earlier form of the name is, however (even at times in autograph), often found in Latin and Italian MSS. of the fourteenth century,3 and must have been familiar. Now according to the great preponderance of MS. evidence,4 this is the form which Chaucer uses in the three passages in which the name occurs — Monk's T. 3515, CL Prol. 31, Cl. T. 1147.5 Petrak (with variants, once Patrikl), instead of Petrark, is the reading, in the first passage, of 12 MSS. out of 16, including the best (El, Hn, Cm) ; in the second, of 17 out of 24, including all the S.-T. MSS. ; in the third, of 14 out of 17, includ- ing all the S.-T. The spelling with two r's later of course became universal, in England6 as well as elsewhere, so its occurrence in late
1 Cf. G. Koerting, Petrarca's Lcben und Werke (Leipzig, 1878), p. 49.
2 There are facsimiles of three autograph signatures, "Petrarca" aud " Petrarcha," dated 1338-1341, in Ugo Foscolo's Essays on Petrarch (London, 1823), frontispiece.
3 Fracassetti, Letters Familiari, i. 216, note. "Petrac" is the form used in the introduction to the story of Griselidis in the Mtnagier de Paris, of date 1392-4 ; see vol. L, 99, 124.
4 As Pollard, in the Globe edition, seems to have been the only editor to recognize.
5 In the two latter cases fortunately we can consult sixteen MSS. ; see Spec, of all tlie Access. Unpr. MSS. of the " C. T." ; parts vi. and vii. (Ch. Soc., 1899, 1900). I have further supplemented by nine unprinted MSS., the four in Cambridge, those in the Lichfield and Lincoln Cathedral libraries, and MSS. Harl. 1239, 1758 and 7333.
6 Cf. the entry "Petrarchse qujedam" in Ritson's list of Lydgate's works
160 THE CANTKltBURY TALES. [cH. V, § 4
MSS. is natural. The older form would never have come into the earliest MSS., in all these passages, at the hands of scribes ; yet if Chaucer knew Petrarch, he certainly must have known his name.
Another rather strong piece of evidence, far stronger than his speaking of Petrarch, seems to me Chaucer's constant silence as to Boccaccio and his obligations to him. I do not propose to go into the Lollius problem, or record his various wrong attributions of Boccaccio's works, but I will recall the fact that in the passage of the Monk's Tale just mentioned he assigns to Petrarch the De Casibus, a work really by Boccaccio, just as in the Clerk's Prologue he implies as distinctly as possible that Petrarch was the author of the Griselda story.1 If he had met Petrarch, and obtained a copy of the Griselda story from him, he could not have failed to learn who was the author of it as well as of the De Casibus, and something of his personality and other works as well. It seems to me nearly certain that Chaucer did not know the Decameron,'1 and quite certain that he did not know it well or own it, yet Petrarch had just obtained a copy. If he met Petrarch, his attitude toward Boccaccio is utterly inconceivable,3 On the
(Bibl. Poet., p. 80). "Pctir Petrarke"is mentioned by Lydgate (Halliwell, Minor Poems, Percy Soc., p. ix.). L. Einstein (The Italian Renaissaticc in England ; see index) indicates that Petrarch was becoming fairly well known in England by the second quarter of the fifteenth century.
1 It has been pointed out before now that Pierre do Beauvau, author of the old French translation of Boccaccio's Filostrato, attributed the original with- out hesitation to " ung poethe Florentin nomme Petrearque" (Moland et d'Hericault, Nowo. Fran?, du xiv* Siecle, Paris, 1858, p. 120).
2 Professor Cino Chiarini (Nuova Antologia, Ixxii. 333) argues conversely that since Chaucer met Petrarch, he miist have known of the Decameron. In the complete absence of perfect evidence in regard to either matter, the a priori argument for the negative view of both is incomparably stronger than for the positive.
3 As to Chaucer's silence as to Boccaccio's name, I do not see how it can be attributed to any cause but ignorance. Probably all the works of Boccaccio which he possessed were in one or two MSS. , which lacked the author's name. Boccaccio was almost certainly not in Florence during Chaucer's short visit there in 1373. In the first part of 1373 Boccaccio seems to have been in Certaldo, where he had been extremely sick during the latter part of the preceding year. [See Gustav Koerting, Boccaccio's Leben u. Werke (Leipzig, 1880), p. 322.] He was not appointed to lecture in Florence on Dante till August, 1373, and did not begin till October 23. [See Paget Toynbee, Athenaeum, no. 4034 (1905), p. 210. Mr. Toynbee might have much more vehemently denied Hales" guess that Chaucer may have heard Boccaccio lecture, and must have become familiar with his name, if he had known of Mather's proof that Chaucer was back in England in May. I should point out that this fact was known to Furnivall as long ago as 1875-6 ; see Thynne's Animadversions (Ch. Soc.), p. 22, note.] Certaldo is twenty miles from Florence. Chaucer's short visit in Florence must have been in February
CH. v, § 4] THE CLERK'S TALE. 161
whole, therefore, these arguments, coupled with the total absence of evidence in its favour, perhaps warrant us in dismissing the idea of a meeting between Chaucer and Petrarch.
If they did not meet, of course it is absolutely out of the question that he learned the story on his first visit to Italy. The letter containing it, of which it is certain that Chaucer had a copy, cannot have got into general circulation before the middle of 1373, and probably not before a year later, after Petrarch's death. What can be plainer, therefore, than that Chaucer first came upon the story during his second visit to Italy, in 1378? So far, there- fore, this seems to be the earliest possible date for the Clerk's Tale.
But it is very difficult to believe that Chaucer made his trans- lation before the Prologue of the Legend of Good Women was written,1 that is, before 1386. There is not one of the works mentioned there which offsets the Romance of the Rose and the Troilus and Criseyde so capitally as the story of Griselda would have done. If he pleads the House of Fame, the Boethius, and (though it was unpublished and very probably unfinished) the Palamon and Arcite, could he conceivably have passed over such a story of feminine patience and devotion] This brings us to the very verge of the period of the Canterbury Tales, and I think we shall find that there is not the least evidence that it does not date from that period. Professor Skeat's discussion2 of the subject is one of the most unsatisfactory parts of his edition. The evidence of the metre as to date is wholly nugatory, and to plead it arrantly begs the question; that "the closeness of the translation also proves " the earliness of the tale is just as gratuitous. It is also
or March. Ho is therefore most unlikely to have seen the old invalid. Prof. Hales (D. N. B., x. 160), like Mr. Borghesi, suggests that Chaucer met Petrarch through Boccaccio; if he knew neither of them, how, asks Hales, did he obtain a copy of the Griselda story ? To this I reply — how did the French writer of the M&wgier de Paris get a copy of it (see pp. 156-7 above), no later than 1392-4 ? On Chaucer's second visit to Italy, both Boccaccio and Petrarch were dead. Chaucer's failure to learn more of the personality and works of the distinguished Italians illustrates vividly the degree to which he must have been preoccupied with business during his two very brief visits in Italy ; which makes it the more unlikely that he undertook the long and arduous journey to Padua. It must be remembered, too, that he went to Italy only as a diplomat, and at the time was still obscure as a "literary man ; he was little over thirty, and had written no poem of importance, and none of any length except the Book of the Ihtchess. Boccaccio was sixty and Petrarch sixty-nine, and Chaucer had no claim upon them. It defies chronology to picture two laureled forms rushing into each other's arms.
1 Cf. Mather, Mod. Lang. Notes, xii., col. 16. 2 III. 453 if.
DEV. CH. M
162 THE CANTERBURY TALES. [c'H. V, § 4
difficult to see how the excellence of verses 995-1008 indicates that they were written at a different time from the rest of the poem. The chief reason, no doubt, why many persons have felt disposed to jfut the Clerk's Tale comparatively early in Chaucer's career is its thoroughly mediaeval character — its want of harmony with the modern spirit and with that of the more advanced and realistic of the Canterbury Tales.1 May I he permitted to say here, as I say elsewhere, in the cases of the Man of Law's Tale, the Monk's Tale and Melibeus,2 that there seems to me something radically errone- ous in this, point of view ? Was Chaucer so far beyond the most modern of the Italians, Boccaccio and Petrarch ? Chaucer's feeling in the matter seems thoroughly intelligible and characteristic. He, like the two Italians, and like many men since, was profoundly touched by the ideal beauty of the story, and reproduced it with perfect sympathy ; then, like Petrarch, but unlike the less reflective Boccaccio, he disclaimed literal approval of Griselda's conduct, and drew an obvious mediaeval moral ; and finally, unlike either of his predecessors, he became somewhat frivolous and ironical in the Clerk's Envoy at the expense of the modern woman. The fact that the Greek tragedian completed his serious and elevated trilogy by a mocking and farcical satyr-play does not prove that it must have been written long after the trilogy, or that he had come to think lightly of the earlier plays. This is not the only time that Chaucer shows indisposition to take himself too seriously ; nor is he the last man who has covered sensibility by a little cynicism. His literary taste can hardly have changed so much between the ages of forty and fifty that he came to scoff at what had once affected him. I cannot believe that the Envoy implies any more aloofness from the tale than would have been as natural just after writing it as ten years later.3 It seems to me also
1 Professor Lounsbury (Studies, iii. 344) thinks Chaucer inserted it in the C. T. because he wished them to contain something which would appeal to all kinds of people. This is by no means the most striking case where that scholar seems greatly to exaggerate Chaucer's modemness.
2 I point out elsewhere that Cl. T., Monk's T. and J\Id. are the only ones of the C. T. which Lydgate thinks worth specific mention in his list of Chaucer's works in the Falls of Princes (see Lounsbury's Studies, i. 421). There is also evidence in the MSS. that Cl. T. was one of the most popular of the G. T. The popularity of the Griselda story in England lasted for centuries.
3 The Envoy is no doubt egregiously out of character for the Clerk ; but I cannot in the least see why this should indicate that it was written long after the tale (this in answer to Koch, Beitrdge zur neueren Philologie, Jakob Schipper dargebracht, Wien, 1902 ; p. 284). Chaucer would always rather
CH. v, § 4] THE CLERK'S TALE. '• 163
proper to deprecate the practice of regarding the Canterbury Tales as a dumping-ground for Chaucer's old outgrown literary work.1
A specific date for the Cleric's Tale seems impossible to arrive at for the present. Ten Brink 2 believes he finds internal evidence for a date after 1387, for he sees in 995-1008, where the narrator exclaims over the warm greeting given Walter's second wife by his people, a reference to the hearty reception given Richard II., on November 10, 1387, by the citizens of London, who both before and after sided with his opponents. But the conjecture carries no conviction.3 There was apparently a more gorgeous reception on a similar occasion in August, 1393.4 Or the reminiscence might equally well be of the events of 1381. But the closest parallel that can be found in contemporary history to the situation in the poem is the reception given the little Isabelle of France, in 1396, on her arrival in London to be Richard's second bride, by the citizens who had been so attached to the first ; " multi de civitate exierunt per pontem ad videndam earn," insomuch that in returning some were " oppress! et ad mortem conculcati." 5 But without further evidence
lose his dramatic propriety than his jest. Compare the self-exposure of the Pardoner.
1 It may be asked if any evidence for an early date for the Cl. T. is to be found in the MSS., especially in the two or three which contain it alone (and thus testify to its popularity). There is not the slightest evidence in any MS. that the tale ever existed in Chaucer's day apart from the C. T. Of printed MSS. there are sixteen, of which one (MS. Longleat) contains only Kn. T. and Cl. T., and two contain only 01. T. (MSS. Phillipps and Naples ; on this latter see Koch, in the Schipper Festschrift just mentioned, pp. 257-85). All these three naturally omit the Gl. ProL, and MSS. Longl. and Ph. omit a little at the end, including the reference to the Wife of Bath (Ph. omits 1163-76, L. omits 1170-6 ; MS. Petworth also omits 1170-6) ; but all three have the Envoy, and MS. Naples has the whole end, the Envoy and the " Host-Stanza " (given by Skeat in a note at the end of the tale). It is clear, therefore, that these MSS.. are all derived from more complete ones of the C. T., and are not survivors from an earlier version. 1 have also examined 32 unpublished MSS., being all that exist in public libraries in England and France, except MS. Sion (published). They com- pletely bear me out. To the above-mentioned three fragmentary MSS. , which testify to the especial popularity of Cl. T., I may add MS. Harl. 1239, which contains only Kn. T., M. L. T., W. B. T., Cl. T. and Frankl. T.
a Hist. o/Engl. Lit. (London, 1893), ii. 123.
3 See H. Wallon, Richard II. (Paris, 1864), i. 330 ; Knighton's Chronicle (Rolls Series, 1895), ii. 241. Walsiugham says nothing of the incident. The Monk of Evesham, to whom ten Brink refers, certainly does not dwell on it, and his account of the fickleness of the Londoners is in another connection (see Thomas Hearne's edition of the Hist. Vit. et Regni Ric. II. , p. 85).
4 Hofler dwells on this more than on the other ; see his monograph on Queen Anne, in the Denkschriften of the Vienna Academy (Phil.-Hist. Classe), xx., pp. 193, 218. See also Hist. Vit. et Regni Ric. II., p. 124.
5 Hist. Fit. ct Regni, p. 129.
164 THE CANTERBURY TALES. [CH. V, § 5
it will hardly do to use any of these possible references as arguments for dating the Clerk's Tale ; for the passage in the poem explains itself without them. For the present we must be contented with the certainty that it was written after 1378, and the strong pro- bability that it was written after the Legend and in the Canterbury period, after 1387.
§ 5. The Monk's Tale.
There cannot be the least doubt that the Monk's Tale dates from later than Chaucer's first journey to Italy. It is not merely that more or less of the poem is derived from Boccaccio's De Casibus Virorum Illustrium and De Claris Mulieribus,1 works which Chaucer is much more likely to have found in Italy than in Eng- land ; or that the account of " Hugelinus Comes de Pize " is derived avowedly from Dante,2 since some regard this as a later insertion. He quotes "Dante also in the account of Nero;3 and the Italian influence is also plain in the form of the names which he gives to
1 Cf. certain passages, especially, in the account of Zenobia (Skeat, V. 236- 238), and the rubric at the beginning of the poem in MSS. El, Cp, Ln, and Cm Dd. It seems at first as if we could prove that Chaucer could not have seen, and certainly not have secured a copy of the De Casibus until his second journey to Italy ; for Hortis (Studj suite opere latine del Boccaccio, Trieste, 1879 ; p. 134, note), Koerling (Boccaccio's Lebcnund Werke, Leipzig, 1880 ; p. 730), and others, declare with a good apparent show of reason that the work cannot have been published till 1373-4. But Henri Hauvette (Soc. des anciens Sieves de la Fac. des Lettrcs de I' Univ. de Paris, 1901 279-97) shows not only that this conclusion is not necessary, but also that there is strong evidence against it and in favour of the date 1356-9 for the composition of the De Casibus, and about 1363 for its publication (p. 296). One of the most promising Chaucerian subjects still to be investigated seems to be the sources of the Monk's Tale, which have been left somewhat at loose ends.
2 Possibly Chaucer used also Villani ; see Paget Toynbee, N. and Q., 8 ser., xi. 205 f. (and cf. S. C. Baddeley, ibid., 369 f.). But cf. M. T., 3651-2. J. W. Hales (The Bibliographer, i. 37-9) argues for a knowledge on Chaucer's part of Italian and Dante before his first journey to Italy ; he has no evidence, and his a priori considerations are not in the least convincing. For the same view cf. also, among other places, Lounsbury's edition of the Parliament of Fowls (Boston, 1877), p. 7, and Francesco Torraca (Journ. of Compar. Lit., i. 82-4) ; the latter's argument is completely disposed of by J. L. Lowes, Mod. Philol., iii. 1-46. At the same time it is wholesome to remember that the belief, on which so much Chaucer chronology is based, that the Italian influence cannot antedate 1372, supported though it is by probability and what evidence we have, is not quite a certainty.
3 " His lustes were al lawe in his decree " (3667) ; cf. " Che libito fe' licito in sua legge " (Inf., V. 56). This borrowing, along with many others, was first pointed out by Gary, in his translation of the Divine Comedy ; see vol. i. , p. 201 (London, 1831). Cf. my article in Mod. Philol., iii. 371, note. Skeat and Lounsbury curiously ignore this borrowing.
CH. v, § 5] THE MONK'S TALE. 165
Zenobia's sons (3535).1 So far we have 1373 as the earliest possible date, on which all will probably agree.
A date not earlier than 1379-81 2 is suggested by a probable quotation from Gower's Mirour de I'Omme. At the end of the account of Alexander, the Monk apostrophizes him thus :
" Thy sys fortune hath turned into as " (3851 ).3
Speaking at some length of the uncertainties in the life of potentates, Gower says :
" Fortune leur changa le dee4 Et desmontoit ce q'ot monte " (22024-5) ;
continuing, he apostrophizes Fortune for her instability, speaks of her wheel, compares her to the winds, tells of her two trumpets of fame,5 and relates the career of Alexander (22051-68), how Fortune made him king and then poisoned him. Addressing the goddess, he says :
" Le dee du quell tu jueras Ore est en sisz, ore est en as" (22102-3).6
An even closer parallel is to be found in 23399 : "Dieus changera tes sis en as."
Chaucer and other poets not infrequently derive figurative language from dicing, but no such cases as these have been found.
It is necessary now to examine carefully the prevalent view that the Monk's Tale was written not long after Chaucer's introduction to
1 Apparently Chaucer did not like the Latin fashion of the names (see Skeat, V. 236) ; he says they are Persian (3536), but has really changed them to an Italian form. It is remarkable that he has done the same thing in many other cases, either because the Italian form pleased his ear better or because it afforded more rhymes. In Skeat's index of names I find the follow- ing instances, omitting those in poems of directly Italian origin (i. e., T. C., K. T.), and H. F., 1229 :— Cambalo (Sq. T., 31, 667; but cf. 656); Danao (L. G. W., 2563, etc.); Hermanno (Monk's T., 3535); lulo (H. F., 177); Lino (L. G. W., 2569, etc. ; cf. Skeat, III. xl.) ; Myda ( W. B. T., 951, 953 ; possibly from T. C., III. 1389) ; Parnaso, Pernaso (passim) ; Sitheo(i. G. W., 1005); Thymalao (Monk's T., 3535) ; Vulcano (H. F., 138).
2 See Appendix A, pp. ^220-5, on the date of the Mirour.
3 I cannot deprive my readers of a "jewel five words long "afforded by the scribe of the Lansdowne MS. , who reads :
" pin suster fortune haj>e torne iii-to an as."
4 Cf. K. T., 1238.
5 In this part Gower seems to have borrowed from H. F. ; see pp. 38-40 above.
B Cf. also 11. 1 1 600-1.
166 THE CANTERBURY TALES. [CFI. V, § 5
things Italian, at any rate before the period of the Canterbury Tales.1 The ground on which this position rests appears to be that the tale to us is dull, and very inferior in merit to the tales which we know date from the period of the whole poem ; and further that in the Nun's Priest's Prologue the Host and Knight show somewhat the same opinion. It is presumed, therefore, to have been Chaucer's, as the Host's opinion of the Tale of Thopas no doubt is, and the history of the tale to have been parallel to that of the Second Nun's. The subject has never been thoroughly overhauled, however, nor the evidence all collected or carefully treated. By doing this I think I shall establish a strong probability that the Monk's Tale was written for the Canterbury Tales.
In the first place, I must protest against the slur on Chaucer's literary conscience cast by this opinion ; if the tale is too poor to date from his heyday, he must have been conscious of its inferiority, and could hardly have been so slack and slovenly as to embody it permanently in his masterpiece. Sir Thopas and St. Cecelia are not parallel cases ; the use of the former, an unmistakable parody, is doubtless due to Chaucer's tactful wish to avoid seriously competing among his pilgrims, and the latter is by no means so poor a poem, to our way of thinking, as the Monk's Tale, and the reasons for dating it early are quite different. Nor is there the least evidence that Chaucer thought ill of it. But I must also refer to what I have said elsewhere in connection with the Man of Law's Tale,2 as to the caution necessary in discussing Chaucer's taste. He was not wholly beyond his age, or beyond the sort of thing which appealed to men as advanced as Petrarch and Boccaccio. Though it was tempera- mentally impossible for him (if not for anybody) to write con amove a poem like the Monk's Tale, yet it is not without a certain impressiveness, even for us, and its subject, the mutability of fortune, had a peculiar interest for Chaucer to the end of his life. The stories are too brief to be interesting, and he was never good at vitalizing material derived from ancient sources. But is there any difficulty in putting the Monk's Tale not long after the Legend of Good Women ; or, allowing for the difference of plan, is it greatly inferior to the Physician's Tale, which perhaps dates from about 1388?3 And could a man who had quite grown beyond the
1 Cf. e. g.t Pollard, Knight's Tale (1903), p. xvi., andSkeat, III. 427-431. For a fuller statement of this view, see Lounsbury, Studies, iii. 332-4.
2 See p. 176. * See pp. 155-6.
OH. v, § 5] THE MONK'S TALE. 167
Monk's Tale have translated and inserted in his masterpiece the interminable dreariness of the Tale of Melibeus without a sign of emotion ] Moreover, sufficient stress has never been laid on the nicety with which the tale is adapted to the teller. The Monk, though a sportsman and a bon vivant, was a man of position and dignity ; of these he would be particularly conscious in a large and miscellaneous company, especially after the impudent familiarity of the Host in his prologue. Accordingly he searches his memory for something safe, monastic and improving ; if not the life of St. Edward, then tragedies, some biblical and all other-worldly in their tendency.
Professor Lounsbury, who is always, if I may be permitted to say so, much inclined to take Chaucer out of his age, with which he himself appears hardly to be in sympathy, it seems to me takes a very mistaken view of the Monk's Tale. It belongs to a "species of composition to which," he says, " the men of Chaucer's age were exceedingly addicted";1 he refers to Boccaccio's De Casibus, to Lydgate's Falls of Princes, and to the Mirrw for Magistrates. Though Chaucer, he thinks, "fell at first under the influence of the dominant taste," "his clear critical perception put him speedily in advance of his contemporaries " ; and in the Canterbury Tales "the Monk's tale is introduced as a specimen of these collections of stories, and largely and perhaps entirely for the sake of satirizing, or at least of censuring, the taste that created and enjoyed them." Now the first sentence which I have quoted is absolutely misleading. There is no question that the genre represented in the Monk's Tale was wholly the creation of Boccaccio,2 both in conception and form, though hints are of course traceable to other mediaeval works. If the De Casibus was the first work of the species, the "taste" was certainly not widely popular in Chaucer's age anywhere in Europe, and was doubtless wholly unknown in England. So Chaucer's procedure in introducing the species in order to censure it would be something like that of a prohibition agitator who should debauch an innocent community with strong drink in order to secure the diver- sion of preaching against it. The Monlf's Tale certainly could not be taken as a parody, and there cannot be the least question that
1 Studies in Chaucer (New York, 1892), iii. 332-4.
2 See Attilio Hortis, Studj sulle opere latiiw del Boccaccio (Trieste, 1879), pp. 117, 120. A forthcoming dissertation by Professor K. C. M. Sills will doubtless throw much new light on the subject. I am already bound to him for information and much generous assistance.
168 THE CANTERBURY TALES. [CH. V, § 5
it would have been thoroughly enjoyed by serious-minded readers. That the genre was likely to become popular in England is indicated by its harmony with medieval taste ; by its later vogue due to Lydgate's Falls of Princes (often printed in both the fifteenth and sixteenth centuries), and (nearly two centuries later than the Monies Tale) to the Mirror J 'or Magistrates ; and by the fact that the Monk's Tale, Melibeus, and the Clerk's Tale are the only individual Canterbury Tales deemed worthy of separate mention by Lydgate in his long list of Chaucer's works in the Falls of Princes.1 On the whole, therefore, instead of first following and then scorning a " dominant taste," it seems probable that Chaucer constantly shared it and was in the head and front of its creation.
The attitude of the Knight and Host toward the tale seems to me more worthy of attention than any other adverse argument ; it does seem at first a trifle odd that Chaucer should put into their mouths such disrespectful language toward the subject and even the phraseology of a poem seriously intended. But to this I reply that he may have had a revulsion of feeling when he wrote the Nun's Priest's Prologue, and have felt that a moderate amount of this sort of thing certainly was sufficient.2 After a time he may have wearied of its gloom and monotony, as he did in the Legend of Good Women ; but this does not mean that he came to regard the whole thing with alienation and scorn. A bantering manner is characteristic of Chaucer, even toward things which he really respected, and (if I maybe allowed to say so) it is perfectly possible to take his humour too seriously. Moreover, the attitude of the two critics is thoroughly good dramatically. Neither the Knight nor the Host was likely to care for such a tale. I can hardly grant Lounsbury that the Knight, who had passed his life campaigning, was representative of " the highest cultivation of the community " ; nor was he especially likely to welcome a recent literary departure. As for the Host, he was disappointed as well as bored. He deserved some reward for his patience through Melibeus ; it is evident, by his banter, that he expected something merrier from the Monk, and after the interruption he pleads for a tale of hunting (cf. 3114-5, 3995). But. once more, the presence of Melibeus just before,
1 Lydgate was surely uo unfavourable example of contemporary cultivation (cf. pp. 162, 190-1, and Lounsbury's Stiidies, i. 421).
2 Boccaccio and Lydgate, who were far more lengthy than Chaucer, express over and over again (as Mr. Sills points out to me) a sense of effort and exhaustion. But they certainly did not think ill of their work.
en. v, § 5] , THE MONK'S TALE. 169
uninterrupted and uncondemned, seems to me a sufficient refutation of the notion that the Host and Knight voice Chaucer's serious and permanent opinion.
So far there is nothing like proof of the earliness of the tale. Three other arguments adduced by Professor Skeat l seem to be of still less value. The canon that poems in stanzas, are early is useful for general classification, but has no weight in argument. The Prioress' Tale is universally granted to be late, and I have shown elsewhere reason to think that the Man of Law's and Clerk's are also. The Monk's Tale is Chaucer's only narrative poem in the 8-line stanza, but of the half-dozen other poems written in it two at least (Bukton and Venus) date from the last decade of his life. Skeat also deduces from Chaucer's confusion of Busiris with Diomedes (3293-4), who are properly distinguished by Boethius, that he had not yet produced his translation of that philosopher; it is hardly necessary to say that the lapse of fifteen or twenty years may produce forgetfulness of a trivial matter as dense as original ignorance. Dr. Skeat also tries to prove the greater part of the poem earlier than the so-called Modern Instances,2 which are known to date from 1386 or later ; " the difference in style between the tragedy of Ugolino and such a tragedy as that of Samson or Hercules, must strike the most careless reader." Skeat ignores the fairly obvious fact that in the Ugolino Chaucer is closely following one of the greatest poets of the world. The question of excellence is of course a purely subjective matter ; I can only say, however, that after many careful readings I can see no difference or superiority in the Modern Instances, except so far as the Ugolino is indebted to Dante. They seem as bald as any part of the poem, and even in the Ugolino the want of congruity and feeling at times (e. g. 3619-20, 3635-6) is the more striking because of the moving horror of the original.3
1 111.427,430. TfiB-Briok-also has another argument. He thinks (Sprache, p. 23) the imperfect rhyme of close with open o characteristic of Chaucer's earlier work, and points to ona in M. T., 3510-2-3-5. But, to say nothing of the excuse here in the number of rhyme-words required, ten Brink himself shows that the same rhyme (to, thg) occurs in the W. B. Prol., 369-70. Was this one reason for his extraordinary opinion that the Wife's Prologue was an early work ? We certainly need a more thorough chronological study of Chaucer's rhyme and verse usage ; it will be highly valuable negatively if riot positively. Where is the Quintus Curtius ?
2 The quaint and convenient term applied by Bradshaw to the tragedies of the two Pedros, Ugolino and Bernabo Visconti.
a I find that the late Professor Francis Palgrave expressed exactly the same opinion of the Ugolino passage (for his interesting essay on Chaucer and the
170 THE CANTERBURY TALES. [CH. V, § 5
The least significant of the contrary evidence seems to me weighty compared with this. We may note, in the first place, the manner of address in line 3429, at the end of the account of Balthasar :
" Lordinges, ensample heer-by may ye take" ;
lordings as a vocative Chaucer seems to use only in the Canterbury Tales and toward people physically present ; usually in the links, but sometimes in the tales1 as well it is used to the pilgrims. Here, clearly, Chaucer has in mind oral delivery. Again, Professor Koeppel2 has detected a probable borrowing from Pope Innocent's De Contemptu Mundi in the account of Adam (3199) :
' ' With goddes owene linger wroght was he, And nat bigeten of mannes sperme unclene" ;
compare " forrnatus est homo . . . de spurcissimo spermate." The pope's work is freely quoted, it will be remembered, in the Man of Law's and the Pardoner's Tales, and Chaucer's translation of it was produced between his two versions of the Prologue to the Legend of Good Women, 1387-1394. Considering its exceedingly uncongenial character, it is hard to doubt that his familiarity with it dates from the time of his own translation.3 Next, at the beginning and end of the Monk's Tale a colloquial style, an absence of formality, may be detected :
" I wol biwayle in maner of Tragedie" ;
the definition of tragedy, echoing that in the Monk's Prologue, is casually introduced in the third line of the last stanza, a strange place for it if the poem was originally independent and unconnected with the Monk's Prologue ; we should rather have expected it in the first stanza. Of course Chaucer might have made these changes in adapting the poem to the Tales ; but they are so unnecessary as to be wholly improbable. Elsewhere in reassigning tales he usually neglects highly necessary revisions.4
But the most important evidence relates to the position of the
Italian Renascence, see The Nineteenth Century for 1888 ; xxiv. 340-59). He even goes so far (p. 350) as to think such a passage as 3620, compared with its original, enough to make one doubt the authenticity of the whole Monk's Tale; which can hardly represent a mature conviction.
1 Pard. T., 573; Cl. T., 1163; Mane. T. 309. In Melibeus it is used occasionally by the characters in addressing each other (c. g. 2212, 2228).
2 Skeat, V. 228.
3 See pp. 181 -2. Cf. also Lowes, Publ. Mod. Lang. Assoc., xx. 794.
4 As in Kn. T., Sh. T., S. N. T.
CH. v, 8 5] THE MONK'S TALE. 171
Modern Instances. Bernabo Visconti died December 18, 1385,1 and thereby supplies us with the latest acknowledged allusion in any part of the Canterbury Tales. Clearly, this passage cannot have been written before 1386 at the earliest, so if the tale was early, this passage was a later addition ; this is generally assumed, and also that the three contiguous passages came in with it. But I think it can be shown conclusively that these passages were not a later addition. Out of 41 MSS. which I have examined, in 10 2 the Modern Instances come at the end of the tale ; in 22 3 they come about the middle. If they were added later, the natural place for them was at the end ; not only would this carry out the chronological order which is generally observed, but to put them in the middle would require MS.-readjustment, no small matter. But it is clear that when Chaucer put the Monk's Tale into the Canterbury Tales they were where they are now in the majority of MSS. The life of Croesus was clearly meant to come last, for it ends with a definition of tragedy, just as another precedes the whole poem at the end of the Monk's Prologue; and the last line of Croesus is alluded to in the Nun's Priest's Prologue, 3972.4 Another thing, the Monk in his prologue (3174-80) apologizes at some length for departing from the chronological order, of which to a mediaeval reader there is no violation worthy such apology if Bernabo and his associates are at the end.5 So when the tale was put in its place, it was certainly arranged in a strikingly incorrect order. The only way in which we can make Chaucer responsible also for the order in the Ellesmere group is to suppose that, presumably in preparing the poem for the Canterbury Tales, he first added the Modern Instances at the end, that the poem in this form got into independent circulation, that
1 Skeat, V. 240 ; Froissart, Chronicles (tr. by Thomas Johnes, London, 1839), ii. 32.
2 MSS. El, Hn, Cm, Hodson (Second Supplement to the S.-T., Pard. Prol. and T., Ch. Soc. 1900), Line, R. Coll. Phys., Ad 5140, Haist, Ch.Ch., Seld.
3 MSS. HI, Cp, Ft, Ln, Cm Dd, Lich., Cm Ii, Cm Mm, TO 3. 15, TC 3. 3, HI 1758, SI 1686, Roy. 17 D, Roy. 18 C, SI 1685, Rawl. 141, Laud 600, New Coll., T C 49, Bodl. 414, Hatt., Barl. 20. They (together with all or almost all the tale) are lacking in 8 MSS., HI 1239 and 7335, Ad. 25718, Paris, Rawl. 149 and 223, Laud 739, and Bodl. 686. In MS. HI 7333 they occur in both positions, a good example of scribal meddling.
4 Cf. Skeat, III. 429 ; his whole argument is unintelligible.
5 Omitting them, the order is Lucifer, Adam, Samson, Hercules, Nebuchad- nezzar, Belshazzar, Zenobia, Nero, Holofernes, Antiochus, Alexander, Caesar,
172 THE CANTERBURY TALES. [oil. V, § 5
here these ten MSS. go back to a MS. that originated thus, that he afterwards distorted the tale and placed it in the Canterbury Tales, and that most of the MSS. are descended from this form. It would require an enormous alternative difficulty to make one accept such an improbability as this ; yet there is no alternative difficulty whatever. Therefore I do not see how we can avoid the conclusion that the arrangement with the Modern Instances at the end is due to a stupid and pedantic scribe j1 that the other arrange- ment is the only genuine one, that therefore the whole second half of the poem was written not earlier than 1386.2 But that it was not written so immediately after Bernabb's death is suggested by the fact that his "tragedy" is preceded by those of the two Pedros, who died in 1369 ; we should expect that Chaucer would have begun with the modern potentate whom he had known, if he had just died. Finally, Professor Lowes has shown that Chaucer must have been occupied with the Legend in 1386, and we have seen that this and the following year were pretty well occupied with that and with the zealous beginning of the Canterbury Tales. Everything therefore indicates that the Monk's Tale was written when the Canterbury Tales were already well under way.
§ 6. T7ie Man of Laic's Tale.
The materials for dating the Man of Law's Prologue 3 and Tale, aside from their connection with the Canterbury Tales, are their relation, on the one hand, to Gower's Confessio Amantis — an allu- sion to it in the Man of Lato's Prologue, and the connection between Chaucer's and Gower's versions of the story of Constance; and, on the other, to the Legend of Good Women.
The Man of Law's Prologue, as I am not the first to point out, was certainly written after the Confessio Amantis. After giving a sort of programme of the Legend of Good Women, the Man of Law declares that Chaucer has written no word of the wicked example
1 It was pointed out long ago by Bradshaw that these MSS. show signs of "editing" (see FurnivalPs Temp. Pref., pp. 23-4). I am simply enlarging upon the opinion of these two scholars.
2 Of course it is open to any one to believe that the earlier part was written long before, but I do not see what will be gained by so doing.
3 I use this term for the Man of Law's Headlink (Furnivall), or Introduction to Man of Law's Prologue (Skeat) ; and for the stanzas on poverty the term Proem,.
CH. V, § 6] THE MAN OF LAW'S TALE. 173
of Canacee or of Apollonius of Tyre, and expresses the strongest abhorrence of such stories. When we find these two l the only really objectionable stories (and both related at length) in a con- temporary poem the author of which Chaucer knew well, and in the first of them the author's good taste so perverted that he throws blame on the father's violence and condones the corruptness of the children, it cannot be doubted that the reference is to that poem.2
1 Canacee is in the Confessio Amantis, III. 143-336, and. Apollonius in VIII. 271-2008 (Macaulay, vols. II. and III.).
2 Dr. Bech (Anglict, v. 375-6) offers the extraordinary explanation that it was the Man of Law's soul which was horrified by the illegality of the conduct of Canacee and Apollonius ; " bei dieser auffassung wird zugleich die annahme einer invective Chaucer's gegen Gower beseitigt." Could the force of perversity further go ? Dr. Root has an over- facile note on the subject (Poetry of Chaucer, p. 184). It is neither here nor there to urge that the pavement-detail (1. 85) is not in Gower ; neither is it in the half-dozen other versions of the Apol- lonius stoiy which I have examined, including Godfrey of Viterbo's, Gower's source. Chaucer must have had a confused recollection either of another horrible touch in the original Latin version (ed. Riese, p. 2), or of a passage in Gower's Canacee story ( I J I. 307-320). I do not see how we can deny the exist- enca of some ill-feeling, perhaps temporary and mild, between Chaucer and Gower ; who may be said, therefore, to supply us with one of the earliest bits of literary gossip in our history. Macaulay (I. xxvii. ) may be right in thinking that Chaucer, conscious of his own occasional lapses from decorum, could not resist the temptation to make a humorous dig at the moral Gower (cf. Karl Meyer, John Gower's Bezichungcn zu Chaucer u. K. Richard II., Bonn, 1889, p. 12). It is true that Professor Hales points out (Diet. Nat. Bioyr., x. 166) what looks like a complimentary reference (or is it sarcastic ?) to the Canter- bury Tales, of the character of which work Gower would be cognizant years before it was published, in the revised prologue of the Confessio (11. 81-2) :
" Bot for my wittes ben to smale To tellen every man his tale "
(cf. Pars. ProL, 25). This prologue dates from 1392-3. But the passage in the Man of Law's Prologue certainly gives an impression of perfect seriousness. Tt surely must also be more than a coincidence that the complimentary refer- ence to Chaucer which Gower had inserted in 1390 (probably) at the end of the Confessio he omitted before the middle of 1391. I must agree with Dr. Heinrich Spies (Engl. Stud., xxxv. 108) in rejecting Macaulay's suggestion (II. xxviii.) that Gower removed the lines merely in order to make room for something else. In 1387 Gower highly disapproved of the Earl of Oxford, to whom Chaucer was bound • by a great favour ; so the alienation of the two poets may possibly have had political connections (see Gower's Cronica Tripcrtita, I. 63-76; Wallon, Richard II., I. 484; Mod. Philol., i. 328). Very tentatively I will offer a further possible contribution to the evidence. There is certainly something a little odd about the Man of Law's Prologue. Almost half of it is quite irrelevant. After admitting his obligation to tell a tale, the Man of Law laments :
" ' I can right now no thrifty lale seyn, But Chaucer, though he can but lewedly On metres and on ryming craftily, Hath seyd hem in swich English as he can Of olde tyme, as knoweth many a man.
174 THE CANTERBURY TALES. [cH. V, § 6
And when we consider further that the story of Apollonius is the last in the whole of Gower's work, it is tolerably clear that 'that had
And if he have not seyd hem, leve brother, In o book, he hath seyd hem in another. For he hath told of loveres up and doun Mo than Ovyde made of mencioun In his Epistelles, that been ful olde. What sholde I tellen hem, sin they ben tolde ? In youthe he made of Ceys and Alciou, And sithen hath he spoke of everichon, Thise noble wy ves and thise loveres eek. ' "
Here follow seventeen lines describing the Legend, and then in thirteen lines he reprehends the tales of Canacee and Apollonius, ending :
" ' And therfor he, of ful avysement, Nolde never wryte in none of his sermouns Of swiche unkinde abhominaciouns, Ne wol I noon reherse, if that I may.' "
Chaucer nowhere else in the C. T. names himself, and he appears to be incognito when he tells his own tales. Why does he speak so'modestly of his own versification, one of the points in which everybody knew he was most in advance of contemporary standards ? If he wished to give a list of his earlier works, why does he mention those alone which relate classical love-stories, thereby naming the Book of the Dmhess only by a minor episode in it ? Why is this whole passage such an echo of the latter part of the Prologue to the Legend? I will venture to commit the following conjecture to fine print. Chaucer may have been more or less seriously nettled at a continuation or revival of the criticisms of him for misogyny and cynicism which had evoked the Legend of Good Women. These criticisms may have been echoed by Gower or accompanied by contrasting praise of him. Now he was the one contemporary poet with whose versification Chaucer had any reason to fear comparison ; much as we may prefer Chaucer's, Gower's is the most regular and accurate verse from Orm to Surrey (cf. Macaulay, II. xvi.-xix.), and some con- temporary taste may have preferred it, as Gascoigne and other mid-sixteenth century poets probably would have done, had they known how to read it. Chaucer declares that lewd though his metres and uncrafty though his rhymes may be, every one knows that he has done his best, in more books than one, to exalt lovers, and has written a whole large volume (here he stretches the truth) of legends of Cupid's saints ; but one thing he has not done, ' ' of ful avyse- ment," he has told no such tales as have defaced the Confessio, nor will the Man of Law do so. This explanation will account for his mentioning the Book of the Duchess as he does ; he needed its testimony in his favour, but perhaps did not care to recall the tears which he had shed for John of Gaunt's first wife after the bereaved husband's twenty years of domestic vicissitudes and his relations with Chaucer's own sister-in-Jaw. Though Chaucer was not very far from thirty when he wrote it, that was twenty or twenty-five years before, so that the phrase " in youthe" is not surprising, and we are not forced to the opinion that Ceyx and Alcyone was an independent work ; any- one who will read critically Professor Bilderbeck's note in his edition of the Minor Poems will see how little there is to be said for this view. One would hesitate to suggest such an explanation as this of the Man of Law's Pro- logue if it implied anything like pettiness or malice or ill-temper on Chaucer's part, which it is impossible to attribute to him ; but there is nothing here that is not perfectly just, and even delicate and good-humoured. It seems also to suggest rather vividly how much to the same "set" the two poets belonged.
CH. V, § 6] THE MAN OF LAW'S TALE. 175
been already finished. Professor Macaulay shows, on the clear evidence of dates in the MSS., that the second version of the epilogue to the Gonfessio was written in the last half of 1390 or the first half of 1391 ; and that the first form of prologue was written in 1390, therefore after the poem was finished.1 This plainly assigns to the Man of Law's Proloyue the date 1390 or later.2
As to the Tale of the Man of Law, it is necessary to notice first the view of Skeat, Pollard, Hales, -Professor W. P. Ker and others that it was written somewhat early in Chaucer's literary life.3 Skeat's belief that in his story of Constance Gower borrowed from Chaucer's will be noticed later, but its evidential value in this connection disappears at the same time with Pauli's early date for the Gonfessio. Xor can the fact that the Man of Law's Tale is in stanzas be used as evidence, for not only is the stanza particularly well adapted to a remote, lyrical and rhetorical poem like this, but Dr. Skeat himself admits that the stanzaic Prioress' Tale was written late.4 All that is left, then, is the subject, treatment and style.
1 Macaulay's Gower, vol. ii., pp. xxi., xxii. and 13, and iii. 468.
2 According to ten Brink (Engl. Stud., xvii. 19-20), M.L.P., 60-76 indicates that Chaucer was purposing a continuation of the Legend of Good
Women; since the list of heroines there said to be treated in the Legend is larger than the correct list, and otherwise different. He associates this project with the revision of the Prologue to the Legend, and attributes both to the year 1393 or a later time. Little can be built, I think, on this argument, simply because we cannot be sure that Chaucer had not intended all along to continue the Legend at some time ; as to the revision of the Prologue, we have seen that it was probably due to a special cause. Koeppel (Engl. Mud., xvii. 199) and Lounsbury (Studies, i.:418) drop a couple of other chronological hints which can hardly be taken up.
3 Hales (Folia Literarin, p. 101 ; cf. also Diet. Nat. Biogr., x. 161-2) dates M. L. T. , Cl. T. , Pri. T. , " and possibly other pieces," ' ' many years before " the C. T. Skeat says (III. 409 and cf. 413) : " We can easily see, from the style and by the metrical form, that this Tale is a piece of Chaucer's early work- manship, and was revised for insertion among the Tales, with the addition of a Prologue and four stanzas, about 1387." Mr. Pollard says: "There are many blots in the story: the monotony of the parts played by the two mothers-in-law — one in Syria, the other in Northumberland — the unreasoning prodigality of time, and the refusal of Constance to declare who she is, being the most obvious. Chaucer .... had not yet learnt to reconstruct a story for himself, or to clothe his characters in flesh and blood " (Primer, p. 69). "The Man of Lawcs Tale, once more a curiously inappropriate one, is cast in the same seven-line stanza as the Seint Gecylc and the Griselde, and from its subject, style, and tone appears to have been written towards the close of the same period" (1369-79; Pollard, Globe Chaucer, p. xxvi.). The remarks of Professor Ker are not dissimilar (see his discriminating Essays on Mediaeval Literature, London, 1905 ; pp. 96-7). Cf. also my Introduction, for other opinions.
4 The " quod she " of line 1644 of course shows that the proem at least was
176 THE CANTERBURY TALES. [cH. V, § 6
The idea that Chaucer is not quite as likely to have written a poeni on such a subject after 1390 as in 1380 seems in a measure to disregard two facts — that at earliest he was a middle-aged man when he wrote it, and that as a poet he was always a mediaeval as well as a modern. In the first place, we are not justified in assuming that a kind of subject which attracted Chaucer at thirty-five or forty he would have despised at fifty because mean- while he had begim writing on others which happen to please us better. I do not forget that Chaucer experienced a reaction against allegory, which is absolutely out of harmony with the concreteness which is his ideal in the Canterbury Tales, but there is no reason to suppose that in ten years of middle life his taste changed so completely that pleasure in the Nun's Priest's or Miller's Tale drove out pleasure in the story of Constance. He must have enjoyed reading both at forty ; at fifty, why not writing both ? Besides, is the tone of the Man of Laic's Tale so very different from that of the Franklin's Tale, for example ? Moreover, the fact that Chaucer turned his back on the Middle Ages in some respects cannot be held to show that he did so in all. Even the poet who, because his peculiar genius was for realism, was capable of so miraculously modern a touch as where the friar in the Sumner's Tale drives the cat off the bench — even he could express himself only in such ordinary mediaeval genres as religious and moral legends, or in Boccaccio's new invention, tales of fallen great ones, when he turned in a more serious mood to a subject which greatly interested him, the mutability of fortune, and to admiration of the Christian virtues. Chaucer's sympathy was catholic enough to embrace them all ; there were other Chaucers besides him of the May mornings or of the "merry tales." And is there any reason to suppose that he ever quite grew beyond the sort of thing which was written by the very Italians from whom he had learned so much ? l
written for the Prioress (and cf. also line 1653). The passionate indignation against the Jews is exquisitely in character for her ; and lines 1832-3,
"This abbot, which that was an holy man, As monkes ben, or-elles oghten be,"
suggests her disapproval of the worldly Monk.
*, Does not Professor Brandl commit the error of taking Chaucer too much out of his age by pairing the Prioress' Tale with Sir TJwpas (Paul's Grundriss, 1893; ii. 680): "Auf die unmittelbar vorhergehende Verspottung kindi-
CH. V, § 6] THE MAN OF LA^S TALK. 177
The element in the story and its conduct to which critics have especially objected is a certain crudity in the plot. We may admit that several motifs are a little overworked — the treacherous mother- in-law, the caitiff lover, and the divinely-guided voyage ; but the mediaeval reader, and writer, was used to such repetition of good things.1 As to what Mr. Pollard calls " the unreasoning prodigality of time," he himself has pointed out exactly the same thing as existing to an unusual extent in the Knighfis Tale,2 which is certainly neither crude nor early, and which Mr. Ker uses as a standard of comparison for the Man of Law's Tale. The lack of intelligible motivation in Constance's conduct I shall speak of later ; her refusal to declare her identity is more or less necessary to the plot, and, at any rate, is dismissed more briefly by Chaucer than by either Trivet,3 his source, or by Gower.4 It must not be forgotten that Chaucer was relating a story already made well known by Trivet and by Gower (as I hope to show shortly), and that in the Middle Ages history and fiction had not yet made the declar- ation of mutual independence which to their common advantage they have made since. He did not care by deviating markedly from the received version to make his readers open their eyes in amaze- ment; in one or two minute points among those where he does deviate, as we shall see, he comments on the fact. Wherever we can
scher Legenden setzt er die der vulgirisierten Romanzen " ? Professor Gum- inere has an admirable paper on the medireval and the modern in Chaucer, in Publ. Mod. Lang. Assoc., XVI. xxxvii.-xl., Appendix; and Professor Louns- bury makes some judicious remarks on the unwisdom of attempting to date poems merely according to their excellence, and illustrates his point from other poets (Parl. of Fowls, Boston, 1877 ; pp. 7-8).
1 It is much more striking in one of the gems of Middle English romance, King Horn ; which introduces a Saracen invasion three times, and twice Horn's coming to a foreign court and having a princess thrust on him, twice his arrival just in time to stop a fatal wedding, twice his entrance in lowly disguise and his slaughter of the guests, and twice a veridical dream. Ueynild is an understudy to Rymenhild, and Arnoldin to Athnlf. In that other Middle- English gem, Sir Gaiuain and the Green Knight, the unaccount- able and the unmotived are far more prominent than in the Man of Law's Tale. With how many mediaeval narratives would M. L. T. suffer by comparison ? If we compared Chaucer with our contemporaries less, and with his own more, we should get a truer estimate of him.
2 See his edition of it, pp. 81-2. Though one would hesitate to construct a time-table for a fairy ship, the allowance of several years (the same in Trivet and Gower) for drifting from Syria to Northumberland, and thence to Spain and Italy, seems rather a concession to realism than the reverse. A similar voyage in the lay of Emare takes only "a full seuene-nyght'and more" (1. 674).
3 See the passage from Nicholas Trivet's Anglo-French Chronicle, edited by Edmund Brock, in the Chaucer Society's Originals and Analogues, pp. iii.-53.
4 Confessio Amantis, II. 587-1598.
DEV. CH. N
178 THE CANTERBURY TALES. [CH. V, § 6
see Chaucer at work (especially on well-known originals) we never find him making such radical changes; he is well content with his data as he finds them, and confines himself, in the main, to adding, illuminating, and vivifying. Xor can I grant Mr. Ker that the story seriously lacks unity or is unwieldy. Few of the Canterbury Tales are more free from disproportionate and overgrown passages. If we compare some of the discourses and soliloquies in the Tales of the Knight, the Franklin, the Wife of Bath and the Nun's Priest, the Man of Law's Tale will not suffer greatly. It seems to me that Pollard and Ker make quite too much of a relatively small matter.
As to its conduct and style, the Man of Law's Tale seems quite harmonious with Chaucer's best period. In spite of the remote and fragile character of the subject, here and there are gleams of humour ; after esoteric discussion the soldan's counsellors can find no remedy for his woe but marriage (217), and the soldaness is of the opinion that if they are baptized (352),
"Cold water shal not greve us but a lyte."
The poet smiles again, for better or worse, in lines 272-3 (" Hous- bondes been alle gode"), 355-7, 709-14, and 789. In vividness and realism of detail (except as regards Constance herself) the poem compares not unfavourably with any of the non-humorous tales. We may note the conferences of the soldan (204-31) and the soldaness (326-57) with their councils (in neither Trivet1 nor Gower) ; in both the other writers the traitor knight is con- founded directly after his accusation of Constance for the murder, but in Chaucer there is a highly vivid judgment-hall scene (617-86); there is a lifelike and wholly original touch in the embarrassment of the pagan official at hearing Christianity openly taught by his wife (568-9). Xo one can miss what Mr. Ker calls the " nobility of temper " in the poem ; or its magnificently rhetorical character, especially in its use of astrology and in its occasional passages of melancholy, pathos and devoutness,2 a character which led ten Brink 3 for some reason to conjecture that Chaucer originally meant to deliver this tale himself on the pilgrimage. It is hardly just to pick out a few of these finer passages which are not found in Chaucer's source, as Dr. Skeat does
1 In whose version the soldaness merely hires seven hundred ruffians.
2 E. g., 11. 295-315, 421-7, 449-62, 631-58.
a History of Engl. Literature (London, 1893), ii. 157.
CH. V, § 6] THE MAX OF LAW'S TALE. 179
(III. 410), and explain them as later additions, especially since we have at least the tales of the Second Nun, the SMpman and the Parson to show that Chaucer's practice was not to revise works which he transferred to the Canterbury Tales or from one teller to another. Many of the best passages,1 sometimes contiguous with those which Skeat points out as possible additions, are so intimately connected with the rest of the story as to forbid the conjecture that they were written at a different time. And finally the Man ofLaio's Tale shows an ease, a mastery and an artistic aloofness in Chaucer's attitude toward his material which is far different from his earlier manner. The style of the poem is remarkably unified and har- monious ; 2 the original and splendid passages are not jewels stuck in a plain setting, but as it were flowers growing out of a plant which naturally produces them.3
This attitude toward his material may help to account for Chaucer's treatment of Constance, the chief puzzle of all and probably the main thing which has led some critics to put the tale early in Chaucer's literary life. In Chaucer she has, it is true, more human feeling than in Trivet ; she pities her child when they are about to be cast adrift (853-61), and is not without sense of her husband's cruelty (863, 1055-7), which accounts for her slowness to make herself known to him in Eome ; none of this is in Trivet.4 But though she says far more than in Trivet or Gower, she acts less ; except for her religious duties, she can be said to come out of her passiveness only three times, when she tells her son to stand before Alia, proposes to her husband a feast for the Emperor, and alights from her horse to make herself known to him (1013, 1079, 1104). Her concealment of her identity from the Constable and the Senator (524-7, 972-3) is more complete in Chaucer than in Trivet ; it is probably for the sake of brevity as well as mystery
1 E.g., 11. 211-7, 270-87, 351-7, 811-19, 1052-78.
2 With the sole exception of the position of Constance in it.
3 This air of mastery and aloofness shows especially in the religious and astrological passages, and is even the cause of some of the imperfections which strike a modern reader. Without being at all perfunctory, Chaucer greatly condenses, especially towards the end, and omits many minor circumstances. This accounts for the obscurity where the blind man appeals to Hermengild for his sight (561-2), without apparent reason or explanation ; in Trivet he is taught by the Holy Ghost to do so. The Tale is far shorter than Trivet's version, and (save for Chaucer's lyrical additions) even than Gower's. No other of the Canterbury Tales, unless it is the Knight's, has so many references to the fact that the poet was condensing.
4 Nor in Gower, except for her attentions to her child (C. A., II. 1061-83).
180 THE CANTERBURY TALES. [CH. V, § 6
that Chaucer refuses to dwell on it.1 Chaucer's Constance, as compared with Trivet's and Gower's more commonplace figure, is marked by vividness without intelligibility, and against Chaucer's far more realistic background she passes about, attended by miracles, like a being from another world (which perhaps she is). It is impossible to be sure of Chaucer's motive for the change which he made in her relation to the story, of which he can hardly have been unconscious, but the following suggestions may come somewhere near the truth. He was probably interested in the story chiefly for its possibilities of rhetorical poetry and impersonal feeling, and in its heroine chiefly as a decorative- figure, an embodiment of suffering and constancy. To rationalize her would have been to make yet more incongruous than it is a story which is incurably miraculous. Therefore, though giving her more human feeling than Trivet does, in order to enhance her pathos, he leaves her in the nimbus of conservatism which is the proper surrounding of a religious figure, while he draws forward the rest of the story into a more modern light. To all this there is a general parallel in the Clerk's Tale ; and just as by disclaiming an intention to hold up Griselda as a model to other wives he shows his consciousness of her remoteness, so here by affecting to attribute to all wives the sanctity of Constance (7 08-1 4).2
Whether all this was quite deliberate we cannot say, and it does not free Chaucer from the imputation of occasional bad art, but his method is the best possible with such an intractable subject. It is
1 In both cases, in Trivef s version, she does explain who she is, in very general terms ; but for no intelligible reason refuses to mention names, even to the Senator at Rome, whom she recognizes (pp. 13-15, 41). Gower's treatment of her reticence is odder than that of either of the others. Though she explains herself vaguely to the Senator (1148-69), she utterly refuses to do so to the Constable (738-9) ; and will not reveal her history to her husband either when they are married or when they are reunited in Rome (910-11, 1450-5 ; neither of these two points is in Trivet or in Chaucer). Constance's reticence is paralleled in the lay of Emare, which of course is nearly related to M. L. T. (11. 358-60 ; in Ritson, vol. ii., and edited by A. B. Gough, London, 1901), and the heroine of which changes her name. Dr. Gough shows that it is a primitive and wide-spread element in the story ; see The Constance Saga (Palaes- tra, xxiii. ; Berlin, 1902), pp. 13, 17. Is not this silence perhaps the relic of the tabu frequently found in tales of fairy-lovers, which doubtless Constance originally was ? (Cf. Schofield, English Literature from the Norman Conquest to Chaucer, pp. 191-2.)
2 On the similarity of Constance to Griselda, compare :
" And she sorwe as donib stant as a tree " (M. L. T., 1055). " And she ay sad and constant as a wal " (Cl. T., 1047).
CH. V, § 6] THE MAN OF LAW'S TALE. 181
a delicate matter to know just how much new wine can be safely poured into old bottles. Chaucer is a positive, not a negative realist ; that is, he constantly adds reality, but does not remove unreality. At times he becomes the more incongruous, therefore, by the very reason of his greatness. Though this may somewhat mar the perfection of his art, it adds greatly to its interest from a historical point of view. That his sense of congruity did not keep him from sounding, even in his best days, notes that jar on our ears, we shall see if we remember Troilus' long soliloquy (IV. 958-1078), Dorigen's long list of heroines on an agitated occasion (Frankl. T., 1367-456), and the introduction into the Canterbury Tales of Melibeus and the Parson's and Monk's Tales. Even if in a sense the Man of Law's Tale is more incongruous than the Second Nun's, chiefly because of its superior realism, it is certainly a far better poem and bears every mark of a much later period in Chaucer's development. In a word, can any one deny that Chaucer might choose such a subject late in his life ? And if he did, in what regard have we a right to expect the Man of Law's Tale to be different from what it is?1 It seems to me, therefore, that in the plot of the tale, still more in its style and subject, there is nothing whatever against putting it late, even in the last decade of his life. One piece of evidence that the Man of Law's Tale is later than the first Prologue to the Legend of Good Women, and perhaps not much earlier than the second, is to be found in the presence in it of five passages2 translated from the De Contemptu Mundi of
1 Emare is an example of a similar story completely rationalized, and thereby made (save for two or three life-like touches) utterly prosaic.
2 Lines 99-121 in the Proem and 421-7, 772-7, 925-7 and 1134-8. See Skeat, III. 407-8, or Koeppel in Herrig's Archiv, Ixxxiv. 405-18. Com- parison with the Latin will show that only the lines which I have indicated are taken from it. The passage in the Proem was first pointed out by A. von During in 1885 (see his translation of Chaucer, iii. 352) ; the others simul- taneously by Koeppel and Lounsbury. Skeat thinks (III. 307, 408) that all rive passages are fragments rescued from Chaucer's own poetic version of the Latin work, which he dates 1373-7 (Ghaitcer Canon, p. 154), and that they were inserted here on the revision of the Tale. Thus the evil communica- tions of ten Brink on the stanzaic Palamon continue to corrupt the world. Koeppel also thinks these passages derived from Chaucer's version of Innocent (Engl. Stud., xvii. 196-7, 199); which is the more odd because he (like ten Brink) believes that this work was in prose, and that, when Chaucer wrote the G-prologue, it had not advanced beyond the first few chapters of the pope's treatise ; while all the passages quoted in the Man of Law's Proem and Tale and in the Pardoner's Tale are from the last chapters of the first book or from the second. I find no evidence that these passages are in any sense quoted from Chaucer's translation. In the first place, the manner in which the work is mentioned in the Legend (G, 413-15) certainly seems to imply that
182 THE CANTERBURY TALES. [CH. V, § 6
Pope Innocent III. It will be remembered that Chaucer's lost translation of this is mentioned in the Prologue to the Legend, but only in the second version, and is the only addition which the latter makes to the list of Chaucer's works given in the earlier version. Disregarding the use of it here, it is quoted only in the Pardoner's Tale and perhaps the Monk's, both late. On this and other grounds, fully set forth by Professor Lowes, the probabilities are overwhelming that it was written not long before 1394, certainly later than 1386. a There is no impossibility, of course, in the idea that Chaucer made these quotations before he had made his translation ; yet one cannot but feel that the pope's work was so foreign to Chaucer's disposition that it could hardly have been one of his favourite books, and that he is not likely long before he translated it to have acquired such familiarity with it that he could readily have made these not very striking excerpts.2 It is not a forced inference, then, that these passages were written after (and probably a good while after) 1386.
But Skeat would have us believe " that the Prologue [Proem] and the four inserted stanzas were placed where they now are at the
it was in prose. Secondly, there is not the least reason to believe that it ever advanced very far ; if there ever was excuse for Chaucer's habit of dropping things in the midst, it was here. His way of speaking of it, —
" And Of the Wrecked Engendring of Mankinde, As man may in pope Innocent y-finde," —
strongly suggests that the translation included only the early part of the work. The second of the above lines seems to imply only a partial version ; and very much as Koeppel points out, while the title of the original is DC contemptu tnundi, sive de iniseria conditionis humancc, Chaucer's title corre- sponds only to the first five chapters of the first of the pope's three books ; they alone deal with conception and gestation, which, according to the pope, are very wretched indeed. This is an odd subject for Chaucer to treat, but so is the whole book, which may explain his getting no farther. Koeppel's only reason for thinking he did get farther is the presence of these quotations in the Man of Law's Tale and elsewhere ; which is amazingly like reasoning in a circle. Dr. Koeppel, in one of his admirable source-studies (Anglia, xiii. 175), affords us one more warning illustration of the orthodox view as to Chaucer's "economy" in cutting out purple patches from cast-off poems; Koeppel carries it to such a point that, in speaking of St. Jerome's "good women " mentioned in the G-prologue of the Legend, he says that Chaucer transferred them, after he had revised the Prologue, to Dorigen's lament in the Franklin's Tale.
1 See Lowes' discussion in Publ. Mod. Lang. Assoc., xx. 790-4 ; and pp. 101-2, 170 above.
2 Why did he translate it at all ? One cannot help guessing that Chaucer's rendering was done by request. As Lowes suggests, it may also have been not unconnected with Deschamps' version of a part of the De Contemptu; see Publ. Mod. Lang. Assoc., xx. 795, note.
CH. V, § 6] THE MAN OF LAW'S TALE. 183
time of the revision of what was once an independent tale " (III. 408). How it may have been with the proem we cannot tell, except that is far more closely connected with the Tale than with the Prologue ; but in the other four passages the evidence is all against Skeat's opinion. In only one case (421-7) do the lines from Innocent form a complete stanza. In the last case (1134-8) one line is in one stanza and four in another, both of which stanzas form an integral part of the narrative ; somewhat the same is true of lines 772-7. It is therefore incorrect to speak of " the four inserted stanzas." At worst, none of the passages shows any more sign of being a later addition than any of the exclamatory stanzas in the poem, and I have pointed out several times elsewhere that in reassigning tales it was Chaucer's practice rather to neglect necessary revisions than to make unnecessary ones. These passages therefore seem to have been written at the same time as the rest of the Tale, and hence to afford a respectable amount of evidence that the Man of Law's Tale was written well within the Canterbury period, certainly after the first Prologue to the Legend.^
But an almost conclusive argument against putting the Man of Law's Tale before the Legend of Good Women seems to me, as in the case of several others of the Canterbury Tales, the fact that it is not mentioned in the Prologue to that poem,- where Alcestis is dragging in everything to Chaucer's credit which she can find, and omits nothing of any length except the Anelida and Arcite, which was unfinished and doubtless unpublished. If the tale of Con- stance had been written as a separate work before the first version of the Prologue, where Boethius, the House of Fame and Origen on the Maydalen are duly recorded, nothing seems more unlikely than that Chaucer should have ignored it.
The relation between Chaucer's and Gower's versions of the story of Constance has been studied by Dr. Emil Liicke,3 incidentally
1 Koeppel (Herrig's Archiv, Ixxxiv. 411) points out another more trifling link between the Tale and the second Proloffue to the Legend; with M. L. T., 701-2 cf. L. G. W., prol.-G, 312, 529. I must say, however, that corn as a symbol for learning and poetry occurs also in Parl. of Fowls, 22-3, and L. G. W., prol.-F, 74-6 (G, 62-4).
2 Cf. Koeppel, Engl. Stud., xvii. 198, who thinks the omission shows that the Tale followed even the second version of the Prologue. But the
. reply will serve here, as in the case of other Tales of pious women, that Chaucer might not wish to mention a poem which he was reserving until the C. T. should appear as a whole.
3 Anglia, xiv. 183-5 ; whole article, pp. 77-122, 147-85.
184 THE rAXThHWKY TALE*. [CH. V, § 6
to an investigation of the obligations of both to Trivet's Anglo- French Chronicle. He proves beyond cavil that Trivet was the main source in ouch case, but also finds twenty-seven small resem- blances, founded on nothing in Trivet,1 which convince him that the two English versions cannot be mutually quite independent. Skeat agrees with him, and quotes (III. 415-17) the more striking parallels. Even though some of them are trivial, the cumulative effect is irresistible, especially when we consider the complete absence of parallels between the two poets' versions of the story of Virginia, and the almost complete absence of them between the Wife of Bath's Tale and the story of Florent in the Conf&ssio?
On the question which of the two was written first opinions differ Liicke, says Skeat (III. 413), "draws what is, in my opinion, the erroneous conclusion, that it was Chaucer who copied Gower ; which seems like suggesting that Tennyson was capable of borrowing from Martin Tupper." I cannot feel, however, that there is the slightest presumption one way or the other. Literary borrow- ing in the fourteenth century was quite a different matter from what it was in the nineteenth, and at any time a poet may " prendre son bien ou il le trouve." Chaucer frequently borrows from writers far inferior to Gower, and it is most unlikely that he had at all as • low an opinion as modern critics have of a poet whom their contemporaries and successors constantly put beside him. Fliigel has pointed out3 that in the best of all his works, the General Prologue, Chaucer was not above frequently drawing phraseology from Gower's Mirour de I'Omme.
As to evidence, it seems to me nearly convincing that Chaucer borrowed from Gower 4 — not that he wrote with the Confessio
1 A few others may be recorded. Constance's prayer and her pity for her child, as she goes aboard ship (825-68), resemble the episode in Gower after they are at sea (1055-83; in Trivet scarcely in germ). In both Chaucer (904 ff.) and Gower (1084, ff.), when Constance runs aground in Spain, she is not brought before the Spanish admiral (as in Trivet), but remains in the ship. There is an analogy between their ways of mentioning the death of Alia:
" Deeth, that taketh of heigh and low his rente " (J/. L. T., 1142) ;
" Bot he which hindreth every kinde And for no gold mai be forboght, The dethe comende er he be soght" (C. A., 1572-4).
2 Cf. pp. 151-2 and 217.
3 Anglia, xxiv. 437-508.
4 Ten Briuk favours this view (Hist. E. L., ii. 156).
CH. V, .§ 6] THE MAN OF LAlv's TALE. 185
Amantis open before him, but that he had read Gower's story attentively, and, perhaps not always knowing that he did so, repro- duced some of his ideas and phrases. It certainly does not look as if Gower borrowed from Chaucer. There is no point among the agreements of the two as against Trivet l which must have emanated from Chaucer, and which is beyond Gower's not inconsiderable abilities. Moreover, while in every point of any consequence2 where Gower differs from Trivet, Chaucer agrees with Gower, there are many other and more important places where Chaucer adds to, or otherwise differs from, Trivet, and where Gower does not follow him. I say nothing of Chaucer's rhetorical additions, which Gower might have wilfully disregarded, or of his more subtle touches, which he might have missed. But such matters as the following are worth attention. At the beginning, while both the English poets say nothing of Constance's learning, the only point on which Trivet dwells, Gower has none of Chaucer's eloquent praise of her beauty and goodness (155-68) ; he says nothing of her submissive grief at leaving home for the oriental marriage (264-87) ; nothing of the conferences of the soldan and his mother with their councils (204-31, 326-57); nor of Constance's prayer on being cast adrift the first time (449-62) ; nor of her mingled emotions toward her husband when they are reunited (1055-7S).3 Such omissions on Gower's part could not be explained by an effort at condensation, for which he shows in this tale (as usual) much less disposition than Chaucer does : nor by unwillingness to take hints, since if he was the latter he took many small points and one or two larger. To review the evidence here adduced, I say that since one of the English poets was so familiar with the work of the other as to reproduce even details of language, since where Gower departs (except for the worse) from their common source Chaucer departs also, and since in many more important points where Chaucer departs from or adds to their source Gower does not, the probable conclusion is that not Gower but Chaucer was the borrower.
1 See Orig. mid Anal., pp. vi.-x.
a Except for a few changes for the worse, as where the miraculous and unaccountable element is increased. See numbers 3, 5 and 7, Orig. and Anal., p. vi. ; and Conf. Am., II. 910-11, 1450-5, where Constance twice refuses to tell her husband who she is.
3 Cf. also the three writers' accounts of her rescue by the Roman Senator after her second solitary voyage (T., pp. 39, 41 ; G., 1126, if. ; C., 967-74). Here Gower reduces her toing and froing, just as he does when she runs aground in Spain, but not as much as Chaucer does.
186 THE CANTERBURY TALES. [CH. V, § 6
One thing more will clinch the matter. Speaking of Maurice at Alla's feast, Chaucer says (1009-10) :
" Soni men wolde seyn, at requeste of Custance, This senatour hath lad this child to feste " ;
and of the invitation to the emperor (1086-7) :
" Soni men wold seyn, how that the child Maurice Doth this message un-to this emperour ; But, as I gesse, Alia was nat so nyce." l
Tyrwhitt2 thought that Chaucer was alluding to Gower, from whom he believed him to have taken the whole story ; Skeat (V. 162-4) thinks the allusion only to Trivet. It is not Chaucer's practice, or that of mediaeval writers generally, to mention their departures from authority — rather to plead precedent where they have none.3 There was no reason Avhy Chaucer should call atten- tion to a deviation from Trivet, who was not an especially well- known writer. But if the reference is to Gower, all is explained ; conscious that he was differing from a poem which had (probably) but just appeared, and was being widely read in the very circles into which he expected his own poem to go, he suggests that his predecessor may have been mistaken. I find it impossible to doubt, therefore, that Chaucer had carefully read Gower's story of Con- stance, and therefore that the Man of Law's Tale was written after at least the early part of the Gonfessio Amantis.
The testimony which this conclusion bears as to the exact date
1 Both Trivet and Gower represent the invitation as being carried by Maurice, but Chaucer thinks it would hardly have been court-etiquette to send a boy, as Professor Child used to say, with the message, " Papa wants you to come to dinner." But in the former case it is curious that Chaucer's memory played him false, for neither of the others says that Constance asked that her son might go, but both merely say, as Chaucer does, that she instructed him to keep in the king's sight.
2 Edition of 1830, 1. clxxxvii. f.
3 Cf. an example of the usual attitude toward a source (at least as avowed) in lines 904-5 of this very poem :
" an hethen castel . . .
Of which the name in my text noght I finde."
We may notice also the definiteness of "som men" as compared with such more usual expressions as " but-if that bokes lye." The nearest parallel to these passages which I find in Chaucer is that in speaking of Criseyde and Diomed he says (T. and C., V. 1050) :
" Men seyn, I noot, that she yaf him hirherte," where the independence is less and the motive for it greater.
CH. V, § 6] THE MAN OF LAW'S TALE. 187
of the Man of Law's Tale unfortunately is not perfectly definite. It is quite possible that Chaucer read Gower's tale, on a " private view," soon after it was written, and when this was we have no means of knowing. Professor Macaulay thinks the plan of the Confessio was laid about 1386, "under the combined influence of Chaucer's Legend of Good Women and of the royal command ; " if we admit the influence of the Lee/end, we must bring the date a year or two later, but that influence is by no means clear (unless at the very end), and such a date would involve extraordinarily quick writing, since the poem was finished in 1390. The Confessio was certainly written after the Vox Glamantis. This was probably begun soon after the peasant rising in 1381, and a third of the way through the writer refers to an event of 1383.1 It may fairly be supposed that before beginning his elaborate English poem Gower would spend some time in planning and collecting materials. It is difficult to believe that he could have reached the second book of the Confessio before 1386-7, the verge of Chaucer's Canterbury period, and the earliest possible date, therefore, for the Man of Law's Tale. It seems much more likely, however, that Chaucer's knowledge of Gower's tale dates only from its publication, especially since his knowledge of it seems to have been so intimate. There- fore if an almost certain date is after 1386, a highly probable one is after 1390.
As to the meaning of the way in which the Man of Law's Prologue, Proem and Tale are put together, it is impossible to come to any certain conclusion. Ten Brink,2 Skeat 3 and others have more or less ingenious and unacceptable suggestions. We may be quite sure, however, that the tale of Constance was not written for the Man of Law — one of the most unworldly and poetic of tales for one of the shrewdest and most prosaic of the pilgrims. It is far more inappropriate to its teller than the Shipman's Tale, the only other one which is at all unsuitable, and that was certainly written for a different person.4 For whom the story of Constance was written it Avould be idle to guess. But it is certainly noteworthy that the manner of its assignment to the Man of Law is more ambiguous and clumsy than that of any other of the Canterbury Tales. In
1 The Bishop of Norwich's Crusade ; see bk. III., chap. vi.
2 Engl. Stud., xvii. 22 ; Hist, of Engl. Lit., ii. 156-9.
3 III. 406-8 ; cf. also Koch, Chronology, p. 68 ; Koeppel, Engl. Stud., xvii. 196. 4 Cf. pp. 205-6 below.
188 THE CANTERBURY TALK*. [CH. V, <j 6
fact, tlie unanimity of the MSS. in putting it after the M<m of Law's Prologue is the only thing that assigns it to him at all. Though the proem on poverty has no connection of content with the Tale, granted that Chaucer wished to present it he has effected a rather clever mechanical connection. But connection of any kind between the proem and the Prologue is totally lacking • more than this, they absolutely contradict each other. Though the Man of Law announces that he shall " speke in prose " (96), three lines later he begins his lyrical outburst. Moreover, as seems never to have been remarked, this derelict tale is no more anchored aft than forward. The following Link begins :
" Our hoste up-on his stiropes stood anon, And seyde, ' good men, herkneth everich on ; This was a thrifty tale for the nones ! ' "
and then proceeds to address the Parson. The only thing to which this passage is linked is the Prologue of the Man of Law (46), by this word thrifty, which the Host uses to assure the teller that he has been better than his word. Thrifty is surely a most non- committal, if not highly inappropriate, epithet to apply to this tale, and there is not a single other end-link in the whole of the Canter- bury Tales which is not indissolubly connected to the preceding tale or its teller. We are therefore forced to the conclusion that, when the Prologues of the Man of Law and the Shipman were written, the story of Constance had not yet been assigned to the Man of Law. As to Chaucer's original plan for him we may find some light when we come to consider the Tale of Melibeus.1
§ 7. TJie Tale of Mellbeus.
Before presenting evidence that Chaucer's Tale of Melibeus was written late, it is necessary to take up some a priori considerations. The dates of the individual Canterbury Tales have been so little discussed that one is sometimes compelled rather to anticipate than to answer objections. But there is one here which is quite certain to be raised. If the Tales of the Monk, the Clerk and the Man of
1 See pp. 195-7 below. The problems connected with " Group B " are more interesting and puzzling than any others involved by the growth of the Canterbury Talcs. The splitting of Group B in all the MSS. but one, the reassignment of the first two tales, and the variety of the readings in Shipm. Prol., 1179, are all elements in the puzzle. I must reserve further discussion for my book on the evolution of the Canterbury Tales, aoid for p. 218 below.
CH. V, § 7] THE TALK OF MELIBEUS. 189
Law have been thought to have been written early, before the period of the Canterbury Tales, because of their unmodern character, a fortiori such an opinion is sure to be advanced of Melibeus. Indeed it has already been advanced, casually and tenta- tively, even by so judicious a critic as Dr. Mather,1 who is " inclined to place" the composition of Melibeus between 1373 and 1378, "for it is difficult to believe that Chaucer would have included this rather stupid piece among the Tales were he not working in old material"; he even seems to suggest a motive for its inclusion — f< Chaucer, cut off in the middle of his Rime of Sir Tliopas, avenges himself by telling the very dull prose tale of Melibeus." May I be permitted again to deprecate what seems an unwise, though very natural, tendency to exaggerate Chaucer's modern side and take him too much out of his age ; and the still worse tendency to regard the Canterbury Tales as a kind of foundling asylum for the waifs and strays of his earlier begetting ] I shall endeavour to point out both probability and evidence that when Chaucer put Melibeus in the Canterbury Tales the value he set on it was such that he may perfectly well have just written it.
To us Melibeus is dull because its human element is thin and crude and its general truths are commonplaces. Is it impertinent to suggest that to the mediaeval reader neither was so] The interest of the earlier Middle Ages in creative literature had been chiefly for lyric feeling and for action ; they had produced little analysis of human motive and shown little knowledge of the human heart. At a certain stage in the intellectual development of a people, these become intelligible and attractive ; witness the rise of literary allegory into popularity in the thirteenth century. Now Melibeus offers both ; strange as the statement may seem at first, Melibeus really shows insight. We, the heirs of all the ages, do not readily perceive it ; but is not the case of Richardson's Pamela somewhat parallel, allowing for the fact that it is more than three times as near us as Melibeus is 1 Can most of us at present at all understand the furore which it excited, all over Europe? Again, though the sayings of dead wiseacres in Melibeus seem to us unspeakably trite and dry, all the literature of the Middle Ages proves that they took a different view of such things. There was a time when every commonplace was fresh and startling ; the
1 Chaucer's Prologue, etc. (Boston, 1899), xiv., xv., xxxi.
190 THE CANTERBURY TALKS. [CH. V, § 7
Middle Ages found mental stimulus in very obvious truths, and a perpetual relish in the gnomic style.1 Does not Chaucer's constant use of it, notably through the mouths of Pandarus and others in the Troilus, but everywhere else as well, prove that he could enjoy it? Moreover, we shall see later that Chaucer's extreme familiarity with the " plot " and contents of Melibeus during the middle of the Canterbury period is proved by its strong influence on the Merchant's Tale. Why did he become familiar with it unless he admired it?
But we are not wholly left to inference. In the Prologue to Melibeus, of course written after the tale, there is proof that Chaucer regarded it with no alienated eye. He alludes to the fact that more than one version of the work was already extant2 (2131-42), and apologizes for diverging from his original (2143-54). He thus shows solicitude as to the opinion of his readers. Can we believe, then, that as Mather seems to suggest, he deliberately afflicted his real readers in order to punish his imaginary auditors for their interruption of Sir Tltopas ? A prose tale of 16,000 words forms a pretty extensive practical joke. More than this, the Monk's Prologue does not show a sign that the pilgrims regarded Melibeus as a penance. In the insertion of it there was no doubt some irony and amused sense of contrast with the former attempt (cf. Mel. Prol., 2127-30, 2154); Chaucer in his own tales deliberately goes to the two extremes. But the fact that he apologizes, not for the tale, but for deviating from another version of it, and actually admits that he tells
" som-what more Of proverbes, than ye ban herd bifore" (2145-6),3
proves that he derived, and expected others to derive, serious pleasure from reading it.4 Is not the other view something like an
1 The scribes constantly call attention to adages or other pithy sayings in their texts ; MS. Harl. 7334 repeatedly has nota in the margin, MS. Arch. Seld. has A proverbe opposite Mill. T., 3391.
2 Therefore now, if not before, he knew both the French and Latin versions (cf. p. 216). His intimate familiarity, shown here and elsewhere, with two earlier versions of the work certainly proves that he regarded it with serious interest.
8 This is very odd, considering the character of Albertano's version, to which he is alluding.
4 The popularity of the French version seems to have been considerable from the end of the fourteenth to the early sixteenth century (see Skeat, III. 426-7). It is not a little striking, moreover, that of all the individual Canter- bury Tales the only ones which John Lydgate thinks worthy of mention in
CH. V, § 7] THE TALE OF MELIBEUS. 191
unconscious survival of the older view of Chaucer's relation to the English language ; and to represent him as a man aloof from his age, and taking a patronizing or improving attitude toward it? .Nor will it do to attribiite the insertion of such works as Melibeus and the Parson's Tale to a sudden whim or late aberration, for their prologues, which presuppose the decision, contain as good writing as there is in the whole poem. A frank recognition of Chaucer's mediaeval side, it seems to me, will promote both a more faithful estimate of him and also that intellectual breadth and that power of sympathetic insight which are among the best things one can gain from the study of mediaeval literature.
But to turn now to evidence as to the date. The Tale of Melibeus is in general translated very faithfully from its French original,1 as I find after a complete detailed comparison. Chaucer very seldom adds anything of consequence, and hardly evef omits anything, except a mere phrase or two, or a longer passage plainly skipped by accident : that is, when two neighbouring passages end in the same word, Chaucer or a scribe, glancing up after tran- scribing the first, confused the end of the second with it, and went on from that point.2 In the whole work I find just three passages
the Falls of Princes are Melibeus, Clerk's T. and Monk's T. (see the passage in Lounsbury's Studies, i. 421). Lydgate was no unfavourable specimen of the cultivated man of the next age.
1 Attributed sometimes to Jean de Meun, sometimes to Renaud de Louens. It is most accessible in Le Menagier de Paris (ed. by Jerome Pichon for the Socie'te' des Bibliophiles Fra^ais, Paris, 1846), vol. i., pp. 186-235. This work, which was written 1392-4 (ibid., p. xxii.), there is no evidence that Chaucer ever saw. The Latin original, Liber Consolationis et Consilii, by Albertano of Brescia, was edited by Thor Sundby, and issued by the Chaucer Society in 1873.
2 This is very frequent, of course, in MSS., especially those in prose. But of such passages omitted in Melibeus there are not more than half-a-dozen as long as two lines in the French. For one of the passages in Chaucer and not in the French text see pp. 193-4 below. Another is in 1. 2157 ; neither the Latin nor the French names the daughter. These are about the most important of twelve or so worth mentioning. Some of these passages were probably in the MS. which Chaucer used. The longest addition is at the end, 3074-8. Chaucer so constantly, however, adds unimportant or synonymous words and phrases (generally of an explanatory nature) that the translation is extremely verbose and dilatory ; the French contains about 12740 words, and the English about 16320. Matzner calls the translation " entschieden wb'rtlich " (Altengl. Sprachproben, ii. 373 ; his introduction and notes are excellent). Some idea of its character may be gained by examining Zupitza's quotations from the original in Koeppel's article in Herrig's Archiv, Ixxxvi. 30-8. Chaucer's MS. of the French was rather different from that published in the Menagier, and better ; see Mel. 2177, 2185, 2235-8, 2408-10, 2581-2 (Mtnagier, pp. 187, '188, 192, 203, 212), but in 2252-3 and 2515, e. g., Chaucer's readings are less good. The French version, on the other hand,
192 THE CANTERBURY TALES. [CH. V, § 7
of two lines or more in the French, the omission of which by Chaucer is not clearly due to this cause. Two of these (between 2702 and 2703, and 2776-7; see the Menai/ier, pp. 218, 222) are unimportant, and there is no visible reason for their omission. But the third is in quite a different category ; it is more than twice "ras long as any other omitted, there is not the least chance for such a skipping as I have described, and there is an obvious reason for its intentional omission. Prudence is instructing her helpless husband as to what sort of advisers he is to avoid, and ends in the English thus : " Thou shalt also eschewe the conseilling of yong folk ; for hir conseil is nat rype" (2389). But the French text continues: " De quoy Salemon dit : dolente est la terre qui a enfant a seigneur ! Et le philosophe dit que nous n'eslisons pas les jeunes en princes car communement ils n'ont point de prudence ; et dit encores Salemon : dolente est la terre de quoy le prince nc se lieve matin ! " * The meaning of this omission cannot be mistaken. Chaucer was thinking of Richard II., and was anxious not to annoy him and his family. Melibeus must, therefore, have been written after June 8, 1376,2 when Edward the Black Prince died, and Richard became heir-apparent. More definite than this we cannot be with equal certainty, except that in the later nineties, till the very end, Richard was neither so young nor so imprudent that the cap would have fitted. In the earlier nineties the memory of past unpleasantness would still be fresh. The fit would have been par- ticularly exact, of course, in the middle eighties, but at any time from 1376 to (say) 1395 a tactful and courtierlike person would not
departs 'videly from the Latin, and should really be called a paraphrase. It is much shorter, and makes important omissions, some additions, odd mistrans- lations, and other changes. In particular, on p. 202 (between Mel. 2389 and 2390) it omits almost two pages (Sundby, 53-5) ; and on p. 203 (between Mel. 2400 and 2401), about a page (Sundby, 57-8). In this latter passage is the quotation from the pseudo-Seneca on the virtue of prudence mentioned in my article on Chaucer and Dante in Mod. Philol. iii., p. 368; therefore this can hardly have been the source of T. C., V. 746-9. It is rather to be regretted that the Chaucer Society published the ultimate instead of the immediate source of Mel. We may hope that before long some one will give us a critical edition of the French version, perhaps in parallel-columns with Mel., at any rate with line for line references ; and with a discussion of the character of the MS. which Chaucer used.
1 Mtnagier, i. 202, and cf. the foot-note. Apparently one MS. substitutes for everything after seigneur the clause : " et de laquelle le prince se desjusne matin." This and the end of the alternative reading seem to be due to Albertano's "et cujus principea mane comedunt" (Sundby, p. 53).
2 See e. g. Armitage-Smith, John of Gaunt, p. 129 ; Richard was bom January 6, 1367 (ibid., p. 44).
CH. V, § 7] THE TALE OF MELIBEUS. 193
have hesitated to make the omission.1 Somewhere between these dates, therefore, Melibeus probably falls; certainly after 1376.
One argument for a late date for Melibeus is the fact that none of Chaucer's works which show its influence seem to be early, as is shown by Dr. Koeppel's2 article. Of the parallels which he quotes between the Troilus and Melibeus I shall speak in a moment. The parallels between Melibeus and the Wife of Bath's Prologue (itself late) we shall see are probably of no consequence. Besides these, the only works in which Koeppel finds parallels are Chaucer's Proverbs (of unknown date), the Nun's Priest's Proloyue and Tale, the Man of Law's Tale,3 the Pardoner's Tale and the Merchant's Tale. Considering the quotable character of the work, and Chaucer's fondness for pithy " sentence," this is a considerable argument.
That Melibeus was written after the Troilus is not only proved by the date 1376, or later, already arrived at, but is of course strongly probable a priori ; for one thing, the proverb-loving Pandarus4 would have been so particularly likely to show the influence of the work that probably, when he wrote the Troilus, Chaucer was not even familiar with the original.5 But there is strong positive evidence for the priority of the Troilus. In one of his characteristic sententious speeches (I. 956) Pandarus says :
" He hasteth wel that wysly can abyde."
As Skeat and Koeppel point out,6 the same words occur in Melibeus, 2244 : " The proverbe seith : ' he hasteth wel that wysely can abyde ' ; and in wikked haste is no profit." 7 Chaucer went
1 Walsingham frequently comments on Richard's youth and folly ; see, e. g., ii. 69, 70, 97, 113 (Rolls Series). He even quotes (under date 1383) the same words of Solomon which Chaucer omits: "Vse terrse, cujus rex puer est" (p. 97). One of the authors of Piers Plowman, also, who was restrained by no courtiership, quotes the same passage in the B-text in 1377 (Prol. 191 ; ed. Skeat, I. 16), and it remains in the C-text, about 1393.
2 Archiv, Ixxxvi. 30-9.
3 I have tried already to disprove the view that this is an early work.
4 It should be remarked that the use of proverbs is characteristic of the poem in general rather than of this particular person in it ; Troilus and Diomed use them as well.
5 Koeppel cites two parallel passages in the two works (Herrig'a Archiv, Ixxxvi. 30), but of course believes that T. C. antedates Mel. (p. 32). The first of them is so commonplace as to be nugatory (cf. W. Haeckel, Das Sprich- wort lei Chaucer, Erlangen, 1890 ; pp. 24-5). The other, quoted above, proves Mel. to be the later.
6 Oxford Chaucer, v. 206 ; Herrig's Archiv, Ixxxvi. 30.
7 With the last clause cf. Pars. T., 1003 ; Skeat, Haeckel and Koeppel also refer to T. 0., IV. 1567-8. A poem containing similar sentiments is attributed to Lydgate by Ritson (Bibl. Poet., p. 73 ; and is therefore probably not by him).
DEV. CH. O
194 TUB CANTERBURY TALES. [CH. V, § 7
out of his way to insert this passage, for it is in neither the Latin nor the French original.1 Moreover, though there are plenty of parallels for the sentiment of the proverb,2 none have been found for the form. But the striking thing is that not only are the words in Melibeus identical with those in the Troilus — they form a complete metrical line, which stands out as conspicuously from Chaucer's amorphous prose as a flint in a mass of clay. Can any one doubt that the poem which contains the proverb preceded the prose work '?
That Melibeus followed the Knight's Tale there is similar evi- dence. Not only is there not the least suggestion of the influence of Melibeus on it, but of this poem, too, there appears to be a line embedded in the prose work. In Arcite's farewell to Emily, he speaks of lying in the grave,
" Allone, with-outen any companye" (2779).
In her discourse on poverty, Prudence says : "And if thy fortune change that thou wexe povre, farewel freendshipe and felaweshipe ; for thou shalt be allone with-outen any companye,3 but-if it be the companye of povre folk " (2749-50). This would put Melibeus after 1384-6.4
Though I do not wish to use excessively the argument from the silence of the Legend of Good Women, I must point out that Melibeus, had it been written then, would have been a much more suitable work to mention in the first version of the Prologue than Boethius, the House of Fame, and perhaps some others. This would date it after 1386. Of course there is no reason why it should be mentioned in the revised Prologue, if it was destined for inclusion in the Canterbury Tales. That it comes from their period is clear from this date, and the busy fulness of the next year or two perhaps justifies us in putting it forward to 1388 at earliest.
Next may be noted a bit of evidence that Melibeus was written before the Man of Laic's Tale. There is no reason to doubt that the Man of Law's proem on poverty was written about the same
1 See Albertano in Thor Sundby's edition (Chaucer Society), p. 12 ; and Le Minagier de Paris, I. 192. I have said that Chaucer very rarely adds anything of importance.
* See Haeckel, Das Sprichivorl bei Chaucer, p. 25.
3 For the whole passage from fareivel to folk the French has only ' ' tu demoureras tout seul " (AUnagier, I. 221 ; or Zupitza's note in Koeppel's article, Archiv, Ixxxvi. 34). The Teseide has nothing corresponding.
4 See p. 82 above.
CH. V, § 7] THE TALE OF MELIBEUS. 195
time as the rest of the poem. Both draw largely on Pope Innocent's De Contemptu Mundi, and I have shown in my discussion of the tale that the Innocent passages within it cannot be a later addition.1 Besides this, the Innocent part of the proem grows into an apostrophe to rich merchants (120-6), which leads skilfully into the main narrative ; the proem has not all the air of having been added when the poem was assigned to the Man of Law. Now a part of the passage in the proem, lines 99-121, which is (somewhat freely) translated from Innocent's Latin,2 appears also in Melibeus, 2758-61, attributed to Innocent and still more closely translated from the free French version of Albertano's Latin, which quotes the pope fairly accurately.3 There is not the least suggestion of mutual influence between the two Chaucerian passages. If the Man of Law's proem had preceded Melibeus, we should expect that, in writing the latter, Chaucer would have recalled his former direct and much more extensive use of Innocent, and that at least a phrase or two of his neat and harmonious poetic version would have stuck in his memory, so retentive of words, and come forth in his prose.4 We have just seen that a few lines before in Melibeus he does quote the Knight's Tale verbatim, and elsewhere the Troilus, departing from his original in so doing.5 On the other hand, there is less probability that a prose version should affect one in verse, since verse requires more manipulation of the material, and Chaucer's prose is always less polished than his verse. Besides, in the poem he is translating from Innocent directly, and in Melibeus only from a small, and by no means literal, excerpt in French. If this evidence is allowed some weight, which I believe it deserves, a relatively early date is suggested for Melibeus.
Now can we form any plausible conjecture as to the original
1 Cf. pp. 181-3 above. Professor Lowes seems hardly to recognize the arguments for this view (Publ. Mod. Lang Assoc., xx. 796) ; see below.
2 For which conveniently see Skeat, III. 407.
3 For Albertano's Latin see Sundby's edition (Ch. Soc. ), p. 100; for the French see Menagier, I. 221-2, or Zupitza's note in Koeppel's article, Herrig's Archiv, Ixxxvi. 33. In a neighbouring passage there is a possible verbal reminiscence between Chaucer's two works ; cf. Mel., 2749 and M. L. P., 116 (not in either original).
4 E. g., Mel., 2761, and M. L. P., 114 : "bet it is to dye than for to have swich poverte," " bet is to dyen than have indigence."
5 See p. 213 for a possible similar case of reminiscence from Wife of BatKs Prol. to Mel. and then to Merck. T.
196 THE CANTERBURY TALES. [CH. V, § 7
purpose of Melibeus ? Lounsbury,1 I believe, first suggested that the Man of Law's Prologue, 96, indicates that Chaucer had in- tended a prose tale for the Man of Law.2 Skeat makes the same suggestion, though without conviction, and also the further one that this tale was Melibeus. He then proceeds to reject both ideas.3 Dr. Lowes4 takes up the first suggestion, and on the basis of the Man of Laic's Prologue, especially 11. 46, 90-6, seems to make it quite clear that Chaucer intended for the Man of Law not his present tale but something in prose, of a pedestrian character ; I need not rehearse his arguments, which of course are obvious enough.
It is impossible to regard with as much favour his very tentative suggestion that what Chaucer meant for the Man of Law was his prose translation of Pope Innocent's De Contemptu Mundi. The Man of Laic is nowhere represented as being of a " sombre " turn of mind, as Lowes seems to think. And could Chaucer conceivably have ever meant to have such a thing recited as a tale ] The only thing which even approaches it in character is the Parson's Tale, which is suitable to the teller, and for the insertion of which he fully accounts in its prologue. Certainly the Man of Law's Prologue does not prepare us for any such extraordinary selection as Innocent's work. Moreover, if he wrote Innocent for a Canterbury Tale, and just before the second Prologue of the Legend, as Lowes believes,5 how came he to mention it in that poem ? The obvious explanation of his omitting to mention such/ infinitely more appropriate works as Physician's Tale, Melibeus and perhaps others, is that he was holding them in
1 Studies in Chaucer (N. Y., 1892), iii. 436.
2 " ' But of my tale how shal I doon this day ? Me were looth be lykned, doutelees, To Muses that men clepe Pierides — Metamorphoscos wot what I mene : — But nathelees, I recche noght a bene Though I come after him with hawe-bake ; I speke in prose, and lat him rymes make. ' And with that word he, with a sobre chere, Bigan his tale, as ye shal after here " (M. L. P., 90-8).
3 Vol. III., 406. His idea that " I speke in prose" refers to the lawyer's pleading in the courts seems to me very unlikely ; for one thing, the Man of Law has been just speaking of the character of the tale he is about to tell, contrasting it with those which Chaucer habitually writes. Mr. A. W. Pollard (Primer, 123-4) also suggests Melibeus for the Man of Law, with more conviction than Skeat.
4 Publ. Mod. Lang. Assoc., xx. 794-6. 5 L. c., p. 793.
CH. V, § 7] THE TALE OF MELIBEUS. 197
reserve for the Canterbury Tales. Finally, Lowes points out that the actual Man of Law's Tale begins with a quotation from Innocent, and suggests that this and other hits of Innocent were derived from his own version and worked in when he was adapting the Constance-story to the Man of Law. But I point out elsewhere that the evidence is clear against any of those passages having been added on revision. Therefore, from the first, Innocent was quoted in the present Man of Law's Tale. Therefore the connection between Innocent and the Man of Law's Prologue is via the present Man of Law's Tale and not the earlier. Therefore all the evidence and an enormous weight of probability is against the opinion that Chaucer meant Innocent for the Man of Law.
We have seen, then, that at one time Chaucer probably meant a prose tale for the Man of Law, but that it was not his version of Innocent. If the tale was ever written, and has not disappeared without leaving the slightest trace, we must return to Skeat's suggestion and conclude that Chaucer originally meant Melibeus for the Man of Law.1 In this view I think there is great probability. From beginning to end Melibeus is one series of arguments, and formal ones at that, with constant appeal to precedent and authority. There is not a single other pilgrim to whom it would have been half so appropriate as to him of whom it is said :
" Discreet he was, and of greet reverence : He semed swich, his wordes weren so wyse.
In termes hadde he caas and domes alle,
That from the tyme of king William were falle."
It is perfectly prepared for by the talk of the Man of Law in his prologue, where, after answering the Host in legal phraseology, he deprecates comparison with Chaucer's mythological and poetic tales. It seems to me that therefore we have excellent reason to believe that Melibeus was at one time intended for the Man of Law, and was perhaps written for him.
1 It does not necessarily follow, of course, though it is very likely, that he composed it for him, nor is there the slightest sign that he composed it to recite himself.
198 THE CANTERBURY TALE*. [CH. V, £ 8
§ 8. The Wife of Bath's Prologue and Tale, the Shipment's Tale, the Merchant's Tale.
On the dates of the poems to be discussed in this section, the Wife of Bath's Prologue and Tale, the Shipman's Tale, and the Merchant's Tale, hardly anything has been written. Ten Brink * dates the first and the last, whose general resemblance he recognizes, about 1390, earlier, however, than the conception of the Canterbury Tales. The name of the Wife of Bath, he thinks, " had probably been a sort of proverb before the poet undertook to make it immortal" (p. 126). For these strange and unparalleled views lie gives no reasons which need be discussed here.2 We have therefore a clear field before us, for research and for conjecture.
There is something about the Merchant's Tale which more than calls for comment, which demands explanation. Every one knows that Chaucer was no cynic. We can throw ourselves heart and soul into accord with his moods of mockery and his flings of derision, as we cannot with those of such a man as Swift or Byron, because we can see that under his severity and contempt are inexhaustible stores of good-humour and tolerance and charity. But with the Merchant's Tale, if we read it with understanding, we cannot do so. Its spirit is anything but agreeable. Its satire on woman and on marriage is the bitterest that Chaucer has anywhere permitted himself, on this or any subject. The fact that it is somewhat covered3 only makes it the bitterer. The poem certainly does not strike one as an overflowing of jollity, or as a tour de force. The satire has a serious air, the emphasis is not at all on the brutal
1 History of English Literature (Engl. tr., London, 1893), ii. 126-32.
2 See Hist. E. L., iii. 267, and Ohaucers Sprache, § 31, but cf. p. 169 above. His only important points I treat later.
3 One or two writers on Chaucer have actually been misled into thinking the first part of the poem (1245-1392) sincere praise of woman and marriage. To say nothing of the caustic lines which are interspersed, it is astonishing that any one should imagine he finds sincere domestic sentiments in the preface to such a story. We must choose between bitter intentional sarcasm, or still bitterer and very stupid unintentional sarcasm. The ironical con- cessions which Chaucer makes in this passage, and which depend for their antidote on the tacit criticism supplied by the tale which follows, are wholly paralleled by the pillorying of men, ostensibly for the benefit of the female sex, in Mane. T., 187-95. Merck. T. is, however, well offset by some beautiful passages at the beginning of Frankl. T. ; note that M. T., 1260 = /'. T., 805, verbatim, and with M. T., 1379 cf. F. T., 751-2.
CH. V, § 8] THE WIFE OF BATIl's PROLOGUE AND TALE, ETC. 199
humour of the situations ; the teller even apologizes l for his indelicate speech :
" Ladies, I prey yow that ye be nat wrooth ; I can nat glose, I am a rude man " (2350-1); 2
" it may nat ben expressed, But if I wolde speke uncurteisly " (2362-3).
Not only is the coarseness less light-hearted and naturalistic than in the Miller's, Reeve's and Sumner's Tales, and the cynicism inherent in the story heightened in every way ;3 there is an occa- sional touch of earnestness and almost pathos, and the danger- ousness of woman and the folly of marriage, especially when the husband is old, are dwelt on at extraordinary length and with a notable air of feeling.4 The openly satirical flings are peculiarly frequent and keen ; the following passages may be especially noted :
" ' Wedlok is so esy and so clene, That in this world it is a paradys.' Thus seyde this olde knight, that was so wys " (1264-6) ;
" A wyf wol laste, and in thyn hous endure, Wei lenger than thee list, para venture " (1317-8);
"They been so knit, ther may noon harm bityde, And namely up-on the wyves syde" (1391-2);
" ' And elles god forbede, but he sente A wedded man him grace to repente Wei ofte rather than a sengle man," (1665-7) ;
' ' Whan tendre youthe hath wedded stouping age, Ther is swich mirthe that it may nat be writeu ; Assayeth it your-self, than may ye witen If that I lye or noon in this matere " (1738-41). 5
1 The only other apology in a coarse story is in Mane. T., 205-11. Cf. also, of course, Chaucer's own apology in ProL, 725-42 ; and Mill. ProL, 3167-86, and Reeve's Prol., 3917.
2 Of course the last phrase must not be taken too seriously. The impreca- tion on " the cursed monk dan Constantyn" (1810) is another suggestion of the refinement and seriousness of the teller.
3 Cf. especially 1967-76, where the narrator leaves God to decide why May fell so easily ; 1987-94, where he affects to praise her for her " franchyse" and soft-heartedness ; 2185-218, where in one breath she declares with tears her honour and fidelity, and coughs to Damian.
4 Cf. especially 1263-71, 1634-56, and the speeches of Justinus and Pluto.
5 Here and elsewhere one is almost inclined to feel that Chaucer was writing somehow from his own experience. If not, the intensity of Merck. T. is a little hard to account for, even with my explanation, to be mentioned later.
200 THE CANTERBURY TALES. [CH. V, § 8
Even Chaucer's own " favourite line" J appears in a connection which turns its inilk of human kindness sour : it is when May has resolved to grant her love to Damian that the narrator comments :
" Lo, pitee renneth sone in gentil herte " (1986).
It is specially noteworthy that when the poem is barely begun the narrator makes a long and quite independent discourse, unparalleled elsewhere in the Canterbury Tales,2 126 lines of veiled and grave irony. And as to the gist and upshot of the story, January's expectations and their outcome are a perfect commentary on the words of the Epistola Valerii: "Amice, nulla est Lucre tia, null a Penelope, nulla Sabina; omnes time." The anti-feminine quality of the tale is the more striking because the character of the " olde dotard holour " January (as the Parson would call him) has been such that we cannot but regard his cuckoldry as poetic justice ; 3 the emphasis with which Chaucer reads the story contradicts its natural emphasis.
Now how is all this to be explained ? An amount of it greatly less in quantity and intensity might be accounted for by the con- ventional misogyny of the Middle Ages, as a passing allusion to Chaucer's own experience or observation, as an excuse for the following story, or as mere wanton humour. But the discourse at the beginning and the tone all through suggest, it seems to me, if they do not imply, a definite purpose. We should not like to believe that it is to set forth Chaucer's own convictions,4 and we cannot if we read the first part of the Franklin's Tale. Nothing remains except that the tone of the Merchant's Tale is a dramatic device. Yet in the description of the Merchant in the General Prologue there is not a syllable to account for it. Nor should we seek an explanation in the Merchant's Prologue or Epilogue. The latter was certainly written after its tale, and the former, like
The poem comes to strike one as occupying somewhat the same puzzling and graceless position in Chaucer's works as the Troilus and Cressida does in Shakspere's, though of course the explanation must be quite different.
1 It occurs four or five times in Chaucer's poetry ; see Skeat, V. 383.
2 The only abstract digressions elsewhere are of about a fourth the length ; seePhys. T., 67-104, and Frankl. T., 761-86.
3 No doubt because Chaucer wished to keep the story well within the limits of comedy.
4 In one of Chaucer's poems he does set forth, with every appearance of seriousness, an unfavourable view of wedlock — Lenvoy a Bukton ; it refers to
JV. B. P., and is full of parallels to its phraseology, but it is utterly unlike Merch. T.
CH. V, § 8] THE WIFE OF HATIl's PROLOGUE AND TALE, ETC. 201
almost all the prologues, most likely was. With its improbable exaggeration, its account of his wife's " passing crueltee " after only two months of marriage, it has every appearance of having been written to account for the extravagant animosity of the tale which follows. So we are still left with an inviting field for conjecture to run riot in.
We may get some light on the subject if we observe the affilia- tions of the Merchant's Tale. The parallel passages in it and the Wife of Bath's Prologue are numerous and important ; especially is the precept of the former constantly supported (or refuted) by the Wife's example in the latter :
' ' For she wol clayrae half part al hir But tel me this, why hydestow, with lyf" (M. T., 1300; cf. 1343). sorwe,
The keyes of thy cheste awey fro me ? It is my good as wel as thyn, pardee (W. B. P., 308-10).
". . . She that waiteth ay They had me yeven hir gold and hir
After thy good, and hath don many a tresoor ;
day " 1 (1303-4 ; cf. 1270). Me neded nat do lenger diligence
To winne hir love, or doon hem reverence (204-6 ; cf. 197, 526).
She seith not ones "nay "whan he For by my trouthe, I quitte hem seith "ye" (1345). word for word
(422 ; cf. 425 and 379-92).
Suffre thy wyves tonge, as Caton bit ; Suffreth alwey, sin ye so wel can She shal comande, and thou shalt preche (437).
suffren it (1377-8). . . . Sith a man is more resonable
Than womman is, ye moste been suf- frable (441-2 ; cf. 434).
For sondry scoles maken sotil clerkes; Diverse scoles maken parfit clerkes, Womman of manye scoles half a clerk Divers praktik, in many sondry is (1427-8). werkes,
Maketh the workman parfit sekirly. Of fyve husbondes scolering am I (between 44 and 45).2
1 The first two passages in Aferch T. are from Theophrastus.
2 These lines, with two more, are to be found in only a few MSS. ; besides the three mentioned by Skeat, they are in MSS. Tiin. Coll. 3.15, Royal 17 D, Christ Ch., New Coll. and Arch. Seld. They are in no other MS. in any public library in England or Paris (but I have not examined MS. Sion). No one can doubt their genuineness, but Tyrwhitt and Skeat regard them as rejected on revision. The fact that they resemble the lines in Merck. T. seems to be no reason whatever for this opinion ; Chaucer is particularly unlikely to have rejected the more for the less elaborate version. The connection between the including lines, 44-5, is so perfect that we may well believe the extra lines to have been inserted later, and their presence in some of the MSS. to be due to contamination with a separate copy of W. B. P. There is other evidence that it circulated, somewhat, apart from the rest of the G. T. These lines should certainly, I think, be restored to the text.
202 THE CANTERBURY TALES. [CH. V, § 8
But sires, by your leve, that am nat I Aud lonlinges, by your leve, that am (1456).1 nat I (112).
But I wot best wher wringeth me my Whan that his shoo ful bitterly him sho (1553). 2 wrong (492).
Paraunter she may be your purga- By god, in erthe I was his purgatorie, torie ! (1670 ; cf. 1647). For which I hope his soule be in glorie
(489-90).
He was al coltish, ful of ragerye, And I was yong and ful of ragerye,
And ful of largon as a flekked pye Stiborn and strong, and loly as a pye (1847-8). (455-6).
And whan he wolde paye his wyf hir That man shal yelde to his wyf hir dette (2048 3; cf. 1452). dette (130).
Though such reminiscences are frequent in Chaucer, we have found some significance in such numerous parallels as connect the Troilus, the Knight's Tale, and the Legend, and are justified in find- ing it here.
Another link is to be found in the works which are quoted in the two poems. The Parson's Tale is quoted in the Merchant's more frequently than in any other of Chaucer's works except the Pardoner's Prologue and Tale, and next most frequently in the Wife's Prologue.4 He uses St. Jerome's work against Jovinian, and the extract from Theophrastus which it contains, more ex- tensively in the Wife's Prologue than anywhere else ; 5 next to this and the Franklin's and Sumner's Tales he uses them oftenest in the Merchant's Tale, where he also refers explicitly to Theophrastus (1294-5, 1310).6 Walter Map's Epistola Valerii ad Rufinum he mentions and quotes in the Wife's Prologue ; elsewhere he quotes it only in the Merchants Tale.7
Certain other points of contact between the Wife of Bath's
1 Cf. Melibeus, 2278. See p. 213 below. The order of composition is perhaps W. B. P., Mel., Merch. T.
2 From Jerome against Jovinian.
3 Cf. Pars. T., 940.
4 See Koeppel in Herrig's Archiv, Ixxxvii. 39-46. Some of the passages are biblical, but the more one investigates ; Chaucer's reading, the more convinced one becomes that his familiarity with the Bible (and other quotable literature, like Cato and Seneca) was largely at second-hand.
8 See Koeppel in Herrig's Archiv, Ixxxiv, 414-15 ; and in Anglia, xiii. 175-6 ; W. W. Woollcombe in Chaucer Society Essays, 297-304. The use of Theophrastus is the reason for the resemblance between some of the remarks at the beginning of the Merch. T. and those quoted by the "Wife from her old husband, as is noted above ; cf. 1294-310 with 248-378.
6 Anglia, xiii. 178-80.
7 The "Valerie" referred to in L. G. W., G-prol. 280, is doubtless Valerius Maximus. See Anglia, xiii. 181-3 ; and also, on all this, Skeat'a index and notes.
C!I. V, § 8] THE WIFE OF BATH'S PROLOGUE AND TALE, ETC. 203
Prologue and the Merchant's Tale make it difficult or impossible to doubt not only that the two were written near together, but also that the latter was written after the former ; further, not only that Chaucer had the Wife of Bath and her prologue in mind when he wrote the Merchant's Tale, but also that he meant his readers to have them in mind. January is remarkably like the Wife of Bath's old husband.1 It is noteworthy that of the dozen or so of analogues to the story2 none seem to have anything about difference of age between the husband and wife except Boccaccio's, which barely mentions it (Decameron^ VII. 9) ; it is natural that the only two great mediaeval writers who treated the story should develop this dramatic contrast, but Chaucer lays much stress on it. May has striking points of similarity to the Wife of Bath (with M. T., 2187-206, 2368-415, cf. W. B. P., 443-50, 226-34); she certainly follows the Wife's principles, and does rather more than bear her husband " on hond the cow is wood."3 Again, just as Pluto's talk is suggestive of Jankin's,4 Proserpina's is a curious reminiscence of the Wife of Bath's ; women, she says, shall never lack the power of facing out their offences,5 and she flouts the authority of Solomon.6 Another suggestion of the Wife's Prologue is that January will have none of an elderly wife :
" And eek thise olcle widwes, god it woot, They conne so muchel craft on Wades boot, So muchel broken harm, whan that hem leste, That with hem sholde I never live in reste. For sondry scoles maken sotil clerkis ; Womman of manye scoles half a clerk is" 7 (1423-8).
1 Cf. one external touch :
" The slakke skin aboute his nekke shaketh " (1849) ;
" Mote thy welked nekke be to-broke ! " ( W. B. P., 277).
2 See Originals and Analogues (Ch. Soc.), pp. 177ff., 341 ff. ; and Varnhagen in Anglia vii., Anzeiger, p. 163.
3 Unless this was a by-word, it shows that Chaucer knew the version of the Tell-tale Bird story which occurs in the romance of the Seven Sages ; cf. par- ticularly W. B. P., 233-4. If he did, it is odd that he used for the Manciple s Tale the vastly inferior and less Chaucerian version found in Ovid. See Skeat's note, and Academy, vol. xxxvii. p. 239.
4 With Merch. T., 2237-53, cf. W. B. P., 641-785.
6 It is true, of course, that the way in which May allays her husband's indignation is one of the traditional elements in the story.
6 With Merch. T., 2264-2310, cf. W. B. P., 226-234,35-43, and the whole early part. It is striking that whenever Chaucer portrays a sceptic, it is as a woman. His four sceptics are the Wife of Bath, Proserpina, Partlet, and Criseyde.
7 Cf. W. B. P., 601-6, 44c-44f, and passim.
204 THE CANTERBURY TALKS. [CIT. V, § 8
It is true that Chaucer along here is using Albertano's Liber de Amore, but all that the latter says is : " Et uxorem accipias potius . . . puellam quam viduam ; dixit enim quidam philosophus : ' Accipe puellam in uxorem, quam vis sit vetula.' " l January's remarks sound very much like a deliberate dig at the Wife of Bath, and certainly reflect her language. Clearly, then, the Merchant's Tale was written with one eye on the Wife of Bath's Prologue, and Chaucer must have known that his readers would be aware of the fact.
But finally the allusions to the Wife of Bath became explicit. Justinus, at the end of his temperate and comparatively optimistic advice, some of which contains reminiscences of her prologue, openly appeals to her. Skeat tries to make the lines an interpola- tion of the narrator's, and prints the passage thus :
" ' My tale is doon : — for my wit is thinne. Beth nat agast her-of, my brother dere.' — (But lat us waden out of this matere. The Wyf of Bathe, if ye han understonde, Of mariage, which we have on honde, Declared hath ful wel in litel space). — ' Fareth now wel, god have yow in his grace '" (1682-8).
In this endeavour to save Chaucer from himself I think the editor makes two mistakes. For we (1686) all the eight published MSS. except the Camb. Dd and the Hengwrt read ye;2 and, therefore, though Dr. Skeat would doubtless explain it as caught by a scribe from the line above, I think we should accept it. Secondly, at a time when there was no such paraphernalia of dashes, parentheses and quotation-marks as Dr. Skeat needs to bolster up his inter- pretation, it is certain that any reader would have understood these lines to be a part of Justinus' speech, as any one will be convinced who will glance at the passage in the Six-Text ; what the reader would have understood we may be sure Chaucer meant, even if he had been capable otherwise of such a piece of monstrously and gratuitously bad style as the editor attributes to him. Chaucer therefore deliberately perpetrates so gross a dramatic impropriety as
1 Koeppel, in ArcJiiv, Ixxxvi. 42.
2 Ye is the reading of eight others, in London and Oxford (Laud 600 and 739, Harl. 1758 and 7333, Royal 17D and 18C, and Sloane 1685 and 1686) ; we, of six others (Bodley 686, Arch. Seld., Barlow 20, Rawl. 149, Egerton, Addit. 5140 ; passage imperfect in Harl. 7335). Mr. George Stevenson has kindly given me this information.
CH. V, § 8] THE WIFE OF BATHES PROLOGUE AND TALE, ETC. 205
to make a character in one of the tales refer to one of the people on . the pilgrimage ; why, unless she had been in his mind all along, and he wished to make the connection explicit 1 l
The Shipmaris Tale must now be brought into the discussion. To begin with, two verbal parallels may be noted between it and the Merchant's Tale (1199 and 1315, apparently taken from Parson's Tale, 1068 ; and 1559 and 2322, apparently from Le Roman de la Rose).2 The two plots in outline are also more alike than either is to any other except the Miller's Tale.3 They stand together and quite apart from any other of the coarse tales4 in their higher literary and (if I may so put it) social tone. They are more refined, and more cynical. Between the Shipman's Tale and the Wife of Bath's Prologue there are one or two rather striking parallel passages ; compare Sh. T., 1194-209 with W. B. P., 337-56 (on the extrava- gance of wives in dressing, and its perils), and 1363-7 with 257-62 (on the 'six good points of a husband and those of a wife).5 Besides these there is the general congeniality between the woman in the tale and the Wife of Bath in her prologue.
If these points of contact between the Shipment's Tale and the other two poems do not seem very significant, there is another which is quite conclusive. The Shipman's Tale was certainly written not for the Shipman but for a woman ; six times the speaker classes himself among wives (1202-9, 1364). 6 The Shipman no
1 Another reference to the Wife of Bath, which is almost as plain, and which in a manner makes a connection between Merch. T. and W. B. P., is in Merck. Epil. , 2433-40. The Host regrets that he is bound to a shrewish wife, but will say no more of her, for fear his words should be reported to her by some woman in the company. It is as plain as possible that 2437-8 mean the Wife of Bath.
2 Cf. Koeppel in Anglia, xiv. 257.
3 As long ago as 1877 Mr. Fleay noted their resemblance to each other, and to the Wife of Bath's Talc (apparently), and suggested that they were written in the order, W. B. T., Merch. T., Sh. T. (Guide to Chaucer and Spenser, pp. 56, 62). In general, however, Fleay's little book is a blind guide.
4 Unless perhaps the rather slight Manciple's Tale.
5 For similar or complementary passages, see W. B. T., 925-50 ; N. P. T., 4102-7. With Sh. T., 1417, also compare W. B. P. 312 (also Reeve's T., 4264, and Sumn. T., 1943).
6 First pointed out by Tyrwhitt (London, 1830 ; iv. 280) : " Which would lead one to suspect that this Tale was originally intended for a female character." The matter was noted also by A. J. Ellis (Early English Pro- nuneiation, i. 244), Hertzberg (German translation of the C. T., p. 644), Furnivall (Temp. Pref., p. 10), Fleay (Guide to Chaucer and Spenser, 54), Louusbury (Studies, iii. 435), and Skeat. Skeat seems to be referring to this confusion of sexes in Sh. T. , but in a manner still more confused, when he mentions the Wife of Bath's Tale in The Chaucer Canon, p. 110. Furnivall and Skeat suggest that Sh. T. may have been meant for the Wife of Bath's second tale (Temp. Pref., 10, note ; v. 168).
206 THE CANTERBURY TALES. [en V, § 8
doubt had his faults, but muliebrity was not oiie of them. Nor is the subject, drawn from trivial social life, appropriate to him. And there cannot be the smallest doubt that the woman is the Wife of Bath, since the only other women in the party are nuns. Two or three passages in the tale, already mentioned, are especially appropriate to her — those on dress and " society," and on the six good points of a husband. Considering the way in which the Canterbury Tales grew, it seems to me much less likely to have been meant for her second, than to have been displaced from the position of her first tale.
But to recapitulate. We have seen that the Shipment's Tale was certainly written for the Wife of Bath. We have found many points of connection between it, the Merchant's Tale and the Wife of Bath's Prologue ; strong probability that the Merchant's Tale was written near the Wife's Prologue, and irrefragable proof that it was written after it. Next, it is plain that the Merchant's attitude toward the Wife of Bath, and " al hir secte," is by no means an amicable one ; that he betrays a deep-seated and cynical animosity toward the latter, and pretty clearly also toward the Wife herself, which is by no means accounted for.
]N"ow we must observe that the victim in the Shipman's Tale is a merchant, who has considerable points of resemblance to him of the Prologue. We may note three things especially. The French merchant has business in Flanders (1245, 1490, etc.); so has Chaucer's (272, 277 ; cf. p. 146 above). The former says (1479) :
' ' We may creaunce whyl we have a name " ; of the latter Chaucer says (279-82) :
" This worthy man ful wel his wit bisette ; Ther wiste no wight that he was in dette, So estatly was he of his governaunce, With his bargaynes, and with his chevisaunce."
This statement that, in spite of appearances, Chaucer's Merchant was in debt, and his presence on the pilgrimage, are illustrated by some of the other merchant's remarks to his wife (1420-4) :
" We may wel make chere and good visage, And dryve forth the world as it may be, And kepen our estaat in privitee, Till we be deed, or elles that we pleye A pilgrimage, or goon out of the weye."
CH. V, § 8] THE WIFE OF SATH's PROLOGUE AND TALE, ETC. 207
Though it would be too much to say, perhaps, that the personality of the merchant in the tale is imitated from him of the Prologue, in the former Chaucer certainly followed the type of the latter.1 My main point, however, is that in the Sliipman's Tale it is a merchant who is put in a pitiable and ridiculous situation by being cuckolded and cheated of his money. The Merchant on the pil- grimage therefore had as much and the same reason to take offence at this tale as the Eeeve had to take offence at the Miller's Tale, and nearly as much as the Sumner had to take offence at the Friars.
What I propose is that here we have the vestiges of Chaucer's original design for an exchange of hostilities, a polite quarrel, between the Wife of Bath and the Merchant, somewhat like those which we actually have between the Miller and Eeeve, and between the Friar and Sumner. If the two tales were a part of the same design — i. e., if Chaucer had not changed the assignment of one before he wrote the other, some such explanation seems inevitable. If such a tiff was intended, there is point in the direct reference to the Wife of Bath in the Merchant's Tale, 1685; what is lost in dramatic propriety within the tale is gained if we consider it as a part of a larger whole; the im- propriety of the reference in Justinus' mouth vanishes before its exquisite appropriateness in the Merchant's. If we reject such an explanation the passage becomes an extraordinary aberration.2 And I think also that my suggestion helps to account for the earnest, disagreeable and cynical character of the Merchant's Tale.
Chaucer's procedure, I think, can be restored with both plausi- bility and completeness. He first wrote the Shipman's Tale for the Wife of Bath, following out more or less the characterization of her which he had given in the Prologue, and perhaps without intending any particular allusion to the Merchant. He then went on to write a prologue for the tale ; and, becoming more interested in
1 On the whole, Chaucer deals throughout his Avorks in vivid types rather than individuals. As another illustration of the fact, in some points there is a resemblance also between the Monk on the pilgrimage, and him of the Sh. T. Though the former is the older man, stress is laid on the good looks of both of them (A, 165, 167; B, 1215, 1218); both are "outriders" (A, 166; B, 1255-6), and high-livers (A, 200, 205-6; B, 1260-4); both are masculine, prudent and worldly.
2 As Lounsbury deems it (Studies, iii. 435); but is it not a little too extra- ordinary, like the blunders as to Alcestis' identity in the G-prologue of L. G. W. , to be a mere slip produced in straightforward writing ?
208 THE CANTERBURY TALE*. [CH. V, § 8
her personality, proceeded to far greater length and elaboration than he had intended.1 It then occurred to him, perhaps not im- mediately, to write a sort of masculine rejoinder to her prologue ; and the Merchant's Tale is the result. The whole gist of the poem, when it is read after the Wife's Prologue, is : " Now just look at it from the man's point of view ; not only are elderly widows untrust- worthy— even young girls are."2 And into whose mouth should the retort be put but his who had suffered most from her tale ?
But why did Chaucer change his plan? It is natural that, in the course of time, he should have come to see that the Shipman's Tale was not wholly suitable to put into the Wife of Bath's mouth after she had recited her prologue. Her tone in the latter is one of bold self-vindication, it is true, but she is a little on the defensive,3 and was by no means so bad a woman as the wile in the Shipman's Tale. To tell such a story would have exposed her to damaging retorts. Chaucer's change of plan may have been hastened by the striking appropriateness of the story which he has used in her actual tale, the gist of which, the sovereignty of woman, has often been pointed out as exactly that of her prologue. There was now no longer any occasion for the Merchant to take personal umbrage against her, and for some reason Chaucer gave up the idea of any direct answer to her prologue ; therefore the Merchant-Wife- of-Bath unpleasantness was cancelled. But the idea of an exchange of hostilities, beginning after the Wife's Prologue or Tale, being still in Chaucer's mind, he transferred it from her and the Merchant to the Friar and the Sumner. The separately-rubricated part of her prologue (829-56), containing this quarrel, would therefore be much later than the rest of it.4 After cancelling the original
1 The self-revelation of the Wife of Bath comes near, at times, to being as impudent as the Pardoner's ; or that of Placebo in Merck. T., 1491-505, or the Friar in Sumn. T.t 2074-8. May we not regard this sort of thing almost as a conventional device to show the speaker's state of mind and character, like the stage soliloquy, as, c. g., those of lago and Richard III. ; and there- fore not to be tested too strictly by realism ? The source of Chaucer's con- ception of the Wife is discussed by Professor Mead, in Publ. Mod. Lang. Assoc.y xvi. 388-404. I find that he makes a remark similar to the one above, that W. B. P. and Pard. P. belong to a well-marked literary form ; they "are alike in that they are, in a sense, confessions — a popular mediaeval type, by the way " (p. 388).
2 Cf. 1393-1468. 3 Cf., e. g., 229-30, 485, 825.
4 It will be observed that the Friar's tolerant attitude toward the Wife of Bath and her prologue is admirably characteristic, and wholly different from that which I have postulated of the Merchant. At the beginning of the actual W. B. T. (865-81) she gets in a little dig at the friars.
OH. V, § 8] THE WIFE OF BATH*S PROLOGUE AND TALK, ETC. 209
assignment, he had to transfer the Shipman's Tale to some other of the less refined pilgrims ; the Shipman is by no means appro- priate, but for the more suitable persons probably he had made or planned other arrangements.1 In order to account for the feeling with which the Merchant speaks of woman and marriage, ex post facto domestic infelicity was manufactured for him, of which there is not a hint in the General Prologue. Chaucer's failure to adapt the tales to the new conditions2 of course agrees with his general carelessness of such things in the Canterbury Tales.
This whole theory I advance quite tentatively, as a conjecture. But it seems to me natural, to contradict no facts, and to explain some things which call for explanation.
And now what light have we on the dates of these poems 1 The early limit is fixed, Avith a fair amount of positiveness and exactness, by the certainty that the Wife of Bath's Prologue was written after the General Prologue? Whatever antecedent probability there may be in the case is decidedly in favour of this view, but there is good evidence as well. The Pardoner's interruption, 164-8, is a clear allusion to a passage in the General Prologue.* But besides this, the Wife of Bath's Prologue was surely developed and modified from her description in the General Prologue.^ It is. rather suggestive that of St. Jerome's treatise against Jovinian, to which so much of the Wife's Prologue is due, there is not a trace in the General Prologue. One of the bits derived from Le Roman de la Rose 6 is also suggestive ; in the General Prologue we are told :
1 There is good reason (in the so-called Shipman' s Prologue in at least five MSS.) to believe tliat he meant at first to reassign it to the Sumner, before the Friar-Sumner quarrel was arranged. See p. 218 below.
2 So the Shipman classes himself among women, and Justinus still makes his strange reference to the Wife of Bath. Another revision neglected in Merch. T. is in 1305-6. Chaucer probably wrote of this couplet only the words, "And if thou take a wyf," and the MS. readings for the rest are all spurious. Some of the MS. readings are given by Skeat, V. 354-5 ; H large number are to be found in some copies of the Six- Text (Introd., pp. 70 ff. , between F and G), but (oddly) not in others. Chaucer's neglect here is another illustration of his habit of rarely reading his own works.
3 I have already mentioned ten Brink's wholly unsupported opinion that it was written before it.
4 LI. 688-91.
5 One or two points, it is true, may seem to suggest the opposite conclusion. Prol., 446, on her deafness, may seem to be an allusion to the incident narrated in W. B. P. , 634-6, 788-810. But it may equally well be a casual and arbitrary detail, like the Cook's mormal, and not have been developed till later; if Chaucer had already written W. B. P. this is hardly the point with which he would have begun his second description.
6 The use of Lc Roman de la Rose in the two poems is interesting (see on this
DEV. CH. P
210 THE CAX'I'KRBUlir /M /./.-. [CH. V, § 8
" Housbondes at chirche-dore she haddc fyvc, Withouten other companye in youthe ; But therof nedeth nat to speke as nouthe" (460-2),
but in her own discourse she pretty much contradicts this :
" For lordinges, sith I twelf yeer was of age,1 Thonked be God that is eterne on lyve, Housbondes at chirche-dore I have had fyve ; For I so ofte have y- wedded be " (W. B. P., 4-7).
Obviously she could not have had much " companye in youthe " before she was married ; so probably the passages were written in this order.2 Finally, all the proportion and emphasis of her per- sonality are other and lighter in the General Prologue than in the Wife's Prologue. In the former she is merely a capable and ambitious housewife, who excels in making cloth ; there is no suggestion what- ever of her relations with her husbands, and almost nothing about her character. This description is a little more individual and less typical than most of those in the Prologue, but otherwise does not differ from them. Can we doubt, then, that it was written before the most vivid and detailed piece of character-drawing that Chaucer ever did 1 This gives, as the earliest possible date for the Wife of Bath's Prologue, 1388.
For the later limit of the Wife of Bath's Prologue the most reliable evidence has been pointed out by Skeat. In Lenvoy a Bukton, after warning his friend3 of the risks of matrimony, Chaucer says, in words which strikingly resemble those in the Merchant's Tale :
Skeat's notes, Mead inPubl. Mod. Lang. Assoc. of America, xvi. 391-404, and Koeppel in Anglia, xiv. 250-5) ; in W. B. P. it is veiy extensive, and in Gen. Prol. two lines are due to its account of La Vieille, who guards Bel-Acueil. Line 461, on her "other companye in youthe," is due to "Car j'avoie autre compaignie " (13369, ed. Pierre Marteau, Orleans, 1878 ; she is recalling her youth) ; and 476 is translated from " Qu'el scet toute la vielle dance " (4078 ; the phrase also occurs in the Troilns, III. 695, and inPhys. T., 79). Chaucer's first conception of the Wife of Bath was partly due to La Vieille. It was the last two lines of her description in the General Prologue that he took as his point of departure, almost his motto, for the later and fuller portraiture ; they may have led him back to Le Roman de la Piosc, whence he now drew also largely on its account of Le Jaloux (9697 if., Marteau's edition ; cf. Mead, I.e., pp. 398-403).
1 Twelve was the marriageable age for females according to canon law. Cf. p. 154 above, note.
2 On her pilgrimages, cf. Prol. 463-7 with W. B. P., 495, 557 ; and on her teeth, 468 with 602-4. The tone of easy allusion in the passages in W. B. P. rather suggests that they were the later.
3 Tynvhitt was mistaken in identify ing him (edition of 1830, I. xlviii.)with Peter de Buketon ; among several Buktons of whom there is word in the re-
CH. V, § 8] THE WIFE OF BATlfs PROLOGUE AND TALE, ETC. 211
" The Wyf of Bathe I pray you that ye rede Of this matere that we have on honde" (29-30).
In his introduction, Skeat declares the allusion to be to her tale; in his note, to her prologue.1 We cannot doubt that the latter is the case ; but it will not matter, for the tale we shall see must be later than the prologue. The date of Bukton may be fixed with great exactness and certainty. The reference in line 23 to the dis- advantages in being taken prisoner " in Fryse " is amply explained by Froissart's 2 account of the expedition against Friesland between August 24 and the end of September, 1396; therefore the poem cannot have been written before October, 1396. Nor later than January, 1397, since, as I have shown in my note, Robert Bukton must have been married by that time. So the date which Skeat assigns to Sufcton, " about the end of the year 1396," is absolutely and exactly established. At latest, then, by the end of 1396, a copy of the Wife's Prologue was in the hands of Chaucer's friend Bukton, and may have been sent as a gift with the Envoy.
cords, it is easy to make a choice. Queen Anne, by letters patent of December 1 , 139 1, granted for her lifetime ' ' to her esquire Eobert Bucton " " a quantity of pasture and wood called ' Gosewold' in her lordship of Eye"; October 6, 1393, this benefaction was enlarged ' ' into a grant of the same to him and his heirs by the yearly service of the rent of a rose as of the honor of Eye ; " and Sep- tember 29, 1394, a few months after her death, grant was made, "for life, to the king's esquire Robert Bucton of the constableship of the castle of Eye, co. Suffolk" (Cal. Pat. Rolls, 1391-6, pp. 324, 495). He may have been the same Robert who had been appointed in October, 1390, one of four king's justices for South Wales (ibid. 1388-92, p. 435). The queen's grant of Goosewold was confirmed by Henry IV. in 1399, and Bukton was still constable of Eye in September, 1401 (ibid., 1399-1401, pp. 16, 540). In July, 1402, and Sep- tember, 1403, he was given a Commission of Array for the county of Suffolk, but his militia glories seem not to have prevented his being sued for debt in 1402 or 1403 (ibid., 1401-5, pp. 114, 149, 288, 291). Chaucer himself was still called "king's esquire " in 1394, and may have frequently seen Bukton before the latter retired into the provinces. But for our most valuable intelli- gence we must go to the Calendar of Papal Registers ; Papal Letters, vol. v. (pp. 57, 63). March 14, 1397, indults were granted in Rome to "Robert Bukton, donsel, nobleman, and Anne his wife, noble woman, of the diocese of Norwich," in which Eye was, and is, situated, to have a portable altar and to have mass celebrated before daybreak. Obviously the young man cannot have been married later than January, 1397 ; nor earlier than October, 1396, since the Envoy was written not earlier than that time, and shows that he was still unmarried then. It is curious to see that, like Lord January, he lost no time in flouting Justinus-Chaucer's advice. In spite of the intense piety of the Lady Anne Bukton, we can imagine what kind of a welcome Chaucer would receive in the castle of Eye.
1 I. 85, 559.
2 Ed. byJ.-A. Buchon (Chroniques franfaises, vol. xxxvii. ; Paris, 1825), vol. xiii., 376-7 (book iv., ch. 50); tr. by Thomas Johnes (J. Winchester, N. Y., n.d.), p. 585 (bk. iv., ch. 79).
212 THE CANTERBURY TALES. [cH. V, § 8
But there are also some grounds, rather ticklish it is true, for dating it before the G-prologue of the Legend of Good Women, which we have found good reason for putting about the latter part of 1394. We have seen how extensively Chaucer uses Jerome against Jovinian in the Wife of Bath's Prologue, far more than in any other of his works ; it is not unnatural to infer that his great familiarity with it dates from his writing of that poem. JSTow the manner in which he refers to and uses it in the Legend, G-proI., 281-304, implies great familiarity; the other five authors men- tioned in the G-prologue are dismissed with a word, and Jerome's work is hardly one to be referred to for laudation of women except by one who knew it well, a fact which is illustrated by the surprise of some of Chaucer's critics at its occurrence here.1 We have already seen that there is no evidence, at any rate, that he even knew the work before he wrote these poems.2 In the absence of contrary evidence, this may perhaps justify us in tentatively putting the Wife of Bath's Prologue not later than 1394.
But we may be able to push it still further back, for I think the indications are that it was written before Melibeus, which there is some slight reason for dating earlier than 1394.3 At first sight the arguments seem weak, for they are mainly ex silentio, and not even, some one may at first think, dead silence. Between the Wife's Prologue and Melibeus there are two, and only t\vo, parallel passages.
1 Skeat, Legend of Good Women (Oxford, 1889), p. 141 (but cf., of course, his larger edition, III. 302-3) ; Koch, Chronology, p. 83. Koch does not seem to see any difference between writing against women and against marriage. Neither did most of Chaucer's contemporaries, but Chaucer and the Fathers did. It is worth noticing, however, that here in praising women Chaucer praises virginity, and that the Legend of Good Women in general is rather against men than in favour of women ; so that its general tendency is at least as much against love and marriage as in favour of them. If the comparison will not be thought impious, Chaucer here is not free from the fluctuating point and purpose which is the main fault of Gower's Confessio. That characteristic mediaeval quality, incongruity, is frequently present in Chaucer, as in other mediaeval poets ; the more, many times, for the very reason of his literary greatness. Often enough, too, it adds more to his interest than it detracts from his perfection. This incongruity is sometimes due to his failure quite to unify material of diverse origins, as here in the Legend and in Dorigen's lament in the Franklin's Tale.
2 See pp. 100-1 above. The belief that he knew it rather early depends on the belief that Prologue G is the earlier ; cf. c. g. , Mead in Publ. Mod. Lanj. Assoc., xvi. 401. See also my article in Modern Philology, iii. 368-70.
:i Besides its omission of the reference to young kings, we may perhaps put Mel. before M. L. Proem and Tale, these about the same time as Innocent, and that certainly no later than 1394, the probable date of L. G. W., G-Prol. Of course all this is exceedingly risky.
CH. V, § 8] THE WIFE OF BATH'S PROLOGUE AND TALE, ETC. 213
In speaking of Christ's precept of virginity, the Wife of Bath
says :
" He spak to hem that wolde live parfitly ; And lordings, by your leve, that am nat I " (111-12).
Prudence, in her self-defence, is speaking of " jangleresses " ; "of vvhiche wommen, men seyn that 'three thinges dryven a man out of his hous ; that is to seyn, smoke, dropping of reyn, and wikked wyves ' ; and of swiche wommen seith Salomon, that ' it were hettre dwelle in desert, than with a womman that is riotous.' And sir, by your leve, that am nat I; for ye han ful ofte assayed my grete silence and my grot pacience " (2276-9). The French original runs : " femmes jengleresses desquelles on dit : trois choses sont qui gettent homme hors de sa maison, c'est assavoir la fumee, la goutiere et la femme mauvaise. Et de telles femmes parle Salemon quant il dit : il vauldroit mieulx habiter en terre deserte que avec femme rioteuse et courrouceuse. Or scez-tu Men que tu ne m'as pas trouvee telle, ains as souvent esprouve" ma grant silence et ma grant souffrance. . . ."1 Novf the indications are that the phrase which Chaucer repeats was used for the first time in the Wife's Prologue.2 One noticeable 'point is that in Melibeus it is as nearly metrical as it can be, and we have seen that twice elsewhere in Melibeus he quotes lines from his own poetry.3 Secondly, Chaucer quite wantonly departs from the French in using it, a thing which I have said he rarely does; the timid and literal character of his translation, of which any one may soon convince himself by a comparison, is well illustrated by the remainder of the passages quoted above. His independence at this point is natural enough. The phrase which he repeats is a neat, forcible, and striking one ; that it stuck for some time in his memory is shown by its recurrence in the Merchant's Tale,4 and I have known modern students of Chaucer in whose memory it has also strangely stuck.
The other parallel passage in the Wife of Bath's Prologue corre-
1 Le Md'iKtgier de Paris, p. 195 ; the italics, of course, are mine.
2 It is true that it is not quite as strictly grammatical there as in Mel., but it is perfectly good Chaucerian style.
3 See pp. 193-4 above.
4 Where it is less apt :
" Or for that ech of hem sholde helpen other In meschief, as a suster shal the brother ; And live in chastitee ful holily. But sires, by your leve, that am nat I " (1453-6).
214 THE CANTERBURY TALE?. [CH. V, § 8
spends, curiously enough, to the first part of the passage just quoted from Melibeus:1
" Thow seyst that dropping houses, and eek smoke, And chyding wyves, maken men to flee Out of hir owene hous ; a ! benedicite ! " (278-80).
This cannot be held to prove a connection simply because it is an extremely common saying. "Without the " smoke " it occurs in Parson's Tale, 631, and many other places; smoke and all, it can be found at least in four Latin works (including Innocent's De Contemptu),2 two French and one non-Chaucerian English work.3 Of these the most likely source is Gower's Mir our de I'Omme, 4117-22 :
" Trois choses sont, ce dist ly sage, Que 1'omme boutent du cotage Par fine force et par destresce : Ce sont fume'e et goute eauage, Mais plus encore fait le rage Du male femme tenceresse."
We know that Chaucer often quotes the Mirour in the General Prologue and elsewhere.4 Therefore of the two passages common to the Melibeus and the Wife of Bath's Prologue, one proves nothing, and the other rather indicates that the latter is the earlier.
This view is confirmed by the complete absence of other parallels. The argument from silence is strong because of the frequency with which Chaucer borrows elsewhere from Melibeus and the works of Albertano,5 especially in the Merchant's Tale, which we have seen was written about the same time as the Wife's Prologue ; and by the obviousness of quoting such a work as Melibeus in such a work as the Wife of Bath's Prologue. Would not Jankin have been
1 Possibly his recollection of the first passage in W. B. P. suggested the use of the second.
2 Which is suggested as Chaucer's source by Koeppel, in Herrig's Archiv, Ixxxiv. 414 ; but cf. Ixxxvi. 31.
3 Piers Plmcman, B, xvii. 315-22 ; C, xx. 297-304. See Skeat, V. 207 ; cf. also How the Wise Man TaugM his Son, Ritson's Anc. Pop. Poetry (1833), p. 94. It is composed out of several passages in the Book of Proverbs. There is obviously not the same reason for expecting the influence of Chaucer's poetic on his prose version (if the second is the later) which we found in the case of the Man of Laic's Tale and Melibeus ; the passage is much shorter, and in the poem he is not quoting the original sources.
4 See Fliigel in Anglia, xxiv. 437-508. W. B. P. 727-32 is doubtless from St. Jerome ; but cf. also Mirour, 4165-88.
5 Koeppel in Herrig's Archiv, Ixxxvi. 29-46,
CH. V, § 8] THE WIFE OF BATrf$ PROLOGUE AND TALE, ETC. 215
likely in such a passage as 775-87 to have shown the influence of such a passage as Melibetia 2245-301, which is several times quoted in the Merchant's Tale ; and the Wife herself to have borrowed at times from Prudence? The two works must have been written near together ; I find it hard to believe that the more original fails to show the influence of the translation for any other reason than that it was written earlier. Therefore it may have come even some years before 1396, when Bukton had a copy.
As to the date of the Merchant's Tale, the only evidence aside from its connection with the Wife's Prologue is the certainty that it was written after Melibeus,1 probably just after. Of the many parallel passages in the two works, all in the Merchant's Tale are poetic paraphrases of Chaucer's own prose, as Koeppel's article makes perfectly clear.2 If he had borrowed in the poem directly and only from the original, his language in the prose translation would hardly agree with the poetic passages so closely, yet always without rhythm, except where it is inevitable; in contrast with the cases already shown where he embalmed a bit of rhythm in the very cloudy amber of his prose. But more than this, though curiously enough it seems never to have been observed before, Melibeus has even affected the plot and characterization of the Merchant's Tale. It can hardly be doubted that the whole first part of the Merchant's Tale is Chaucer's own addition to the story ; there is not the least suggestion of it in any of the analogues which have been found, and Professor Varnhagen, who has investigated the history of the story, attributes to Chaucer all but the pear-tree episode, the bare kernel.3 Now when January has resolved to marry he sends for his friends (by no means an obvious thing to do), states the case, and calls for their advice, but in such a way that they know what advice he desires (1397-468). Just so, after his family misfortunes, Melibeus called a conclave (2194 ff.), "shewed hem his cas" and then "axed he hir conseil upon this
1 If any evidence were needed that Merck. T. was written after 1378, we should have some little in the fact that in lines 1245-6 Chaucer tells us that January was a worthy knight born in Pavia and living in Lombardy, local details which were probably not in his source. His first Italian journey had not led him at all into those parts ; but his second took him to Milan, the capital of Lombardy, and only twenty miles from Pavia.
2 Herrig's Archiv, Ixxxvi. 34-43. It is plain, from his article, that the connection of Mel. with Merck. T. is far closer than with any other of Chaucer's works.
* Anglia, viL, Anzeiger, p. 163.
216 THE CAXTERBURY TALES. [CH. V, § 8
matere," though "by the manere of his speche" he showed what counsel he wished (2198-200). In the Merchant'* Tale, after other speeches, the flattering Placebo1 advises Lord January to follow his own wishes, discoursing on the wisdom of "working by counsel" and, very undramatically, on the folly of giving lords unwelcome advice. Complaisant advice similar to Placebo's is given, also after others have spoken, by Melibeus' flatterers (2208-10). Placebo's two specific points just mentioned are based on Melibeus 2193 and 2340-3, and even the idea of his character is drawn from the latter passage. The indebtedness in its plot of the Merchant's Tale to Melibeus is unmistakable.2 Therefore, considering the strong influence of Melibeus in general and in detail, the conclusion is irresistible that when he wrote the Merchant's Tale he had made his translation, and probably just made it. Koeppel finds no evidence, it is true, that Chaucer used Albertano's Latin at all when he wrote Melibeus ; while one or two passages in the Merchant's Tale which are taken from the Latin, and are in neither the French nor Melibeus, show that by that time he had procured a copy.3 This does not necessarily imply that any considerable time elapsed between Melibeus and the Mercliant's Tale* He may have owned the Latin all the time ; when he had elected to translate the shorter French version, there was no reason why he should consult the original ; or it may have been the admiration which led him to translate the French version that finally brought
1 The name seems drawn from Pars. T. 617, but Skeat gives other parallels. Of course it is a joke on the vespers for the dead, and may be proverbial. Placebo's discourse recalls the similarly undramatic self-revelations of Chaucer's Pardoner and of the friar in the Sumners Talc.
a And is so extensive that the latter deserves to be called one of its sources. Another bit of the plot apparently borrowed from an earlier work of Chaucer's own is where lovelorn Damian takes to his bed and May pays him a visit (1932-5). This strikingly recalls the scene where Criseyde makes a similar visit to Troilus (III. 64-75), which is Chaucer's own addition to the Filostrato. There is a suggestion of irony in making January play the part of Pandarus.
3 £ee Herrig's Archiv, Ixxxvi. 29, 38-9. I have shown conclusive evidence that when he wrote Prol. to Mel. he knew the Latin of Albertano, and expected that his readers also would know it. In lines 2131-42 he alludes to the fact that two versions were extant already, and in 2143-54 apologizes for diverging from his original — Albertano, since he does not diverge from the French version. In the same volume with the Liber Cmisttii was very likely Albertano's Liber de Amore Dei, which he also quotes in Merch. T. and probably only there (Koeppel, 1. c., pp. 40-4 ; the parallel passages in T. C. are from Solomon and Seneca) ; and possibly also Albertano's DC Artc Loquendi, used in Mane. T. This may have a bearing on the date of that poem.
4 Skeat, for no very visible or good reason, puts several years between (V. 353).
CH. V, § 8] THE WIFE OF KATH's PROLOGUE AND TALE, ETC. 217
the original to hand, too late to be used in the translation. There- fore there is nothing to contradict the obvious conclusion that the Merchant's Tale was written shortly after Melibeus, very probably not later than 1394.
The theory which I have advanced of course implies that the Wife of Bath's Tale was written for the "Wife of Bath and after her prologue. This nobody doubts, and the evidence for it is quite conclusive. Lines 925-50 and 1258-64 are full of parallels to the Wife of Bath's Prologue. Certain passages (it is true) in the Shipman's Tale, 1194-209 and 1363-7, which also parallel the Wife of Bath's Prologue? yet were written earlier, are a natural enough comment on the ensuing tale and development from the characterization of the Wife of Bath in the General Prologue. But these lines in the Wife's Tale are in quite a different category ; they contain very numerous detailed resemblances which show that her character was already fully developed.2 As to her Tale and the Merchant's, there is no internal evidence to show which came first. The Wife's Tale contains, it is true, no parallels to Melibeus ; but the abstract topics on which there is discourse in it, gentility and the advantages of poverty and of old and homely wives, are not treated in the prose work. Therefore, especially if it was written some time after Melileus, we should expect no influence. There is nothing, accordingly, against the requirement of my theory that the Wife of Bath's Tale shall have been written after not only her prologue but also the Merchant's Tale.3
1 Cf. p. 205 above.
2 We should observe particularly 928, 929-34, 937 (cf. W. B. P., 662-3), 950 (W. B. P., 531-42), 1027, and all of 1258-64.
3 It may be asked whether there is any visible relation between Gower's tale of Florent (C. A., I. 1407-861 ; published in 1390) and W. B. Tale, such as we have found between the two poets' stories of Constance, which might aid us to date the Wife of Bath's Tale. It is quite clear that neither of the poems was the source of the other (see Dr. G. H. Maynardier, The Wife of Bath's Talc, London, 1901 ; pp. 128-46). The only verbal resemblance which I find describes the knight's distress over his ill-looking bride :
" Bot as an oule fleth be nyhte Out of alle othre briddes syhte, Riht so this knyht on daies brode In clos him hield, and schop his rode On nyhtes time" (G. A., 1727-31) ;
''For prively he wedded hir on a morwe, And al day after hidde him as an oule " ( W. B. T., 1080-1).
It seems to be the general view that Chaucer's poem followed Gower's, which is confirmed by the probabilities as to its date. But here there is no reliable
218 THE CANTERBURY TALES. [cH. V, § 8
The question may now arise as to the temporal relation of these poems to some of the minor parts of the Canterbury Tales. We can squeeze out a few more inferences, though they do not all depend on the foregoing. The Prioress' Prologue was not written till after the Shipment's and Merchant's Tales and the Wife of Bath's Prologue, and after the change of plan in regard to them, since it refers to the tale of the Shipman as already assigned to him. Lines 829-56 of the Wife's Prologue, containing the beginning of the Friar-Sumner squabble, I have shown would be later than the change. So, no doubt, with the present Shipman' s Prologue ; l so also with the Merchant's Prologue and Epilogue.2 The same is pro- bably true of the Monk's Prologue, since it seems to have been written' after fhe Merchant's Epilogue. In the latter the Host says (2427-30) :
' I have a Avyf, though that she povre be ;3 But of hir tonge a* labbing shrewe is she, And yet she hath an heep of vyces mo ; Ther-of no fors, lat alle swiche thinges go."
evidence. W. B. T. offers another parallel to Gower. On the ubiquity of the friars, cf. W. B. T., 865-81 with Vox Clamantis, IV., cap. xxiii. Note especially 867-8 (but cf. the whole contexts) :
"That serchen every lond and every streem, As thikke as motes in the sonne-beem " ;
" ludeos spersos fratrum dispersio signat.
Nescio si supera sibi clauserit ostia celum ;
Dat mare, dant ampnes, totaque terra viam" (1113, 1123-4).
The Protestant Pilgrim's Talc (ed. Furnivall, in Appendix to Thynne's Animadversions, Ch. Soc., 1875 ; see 11. 88-100, pp. 79-80) makes interesting quotations from this part of the tale. A partial analogue to W. B. T. is suggested by a passage in Miss Edgeworth's Modern Grisclda, chap. ix. : "... the Princess Rhezzia, in the Persian Tales; who was blooming and charming, except when her husband entered the room . . . doomed to this fate by a vile enchanter."
1 Hence it was probably later yet (see p. 188) that M. L. T. was assigned to the Man of Law. M odern editions obscure the puzzling problems connected with Sh. P. , which I hope to treat more fully at another time. The indica- tions are, I think, that Chaucer meant at first to reassign Sh. T. to the Sumner, to wliom it would have been far more appropriate than to the Shipman, and that he wrote the present Sh. Prol. for the former. The read- ing " Sompnour " in line 1179 found in five MSS. would therefore be the original one. The unification of Group B made by Bradshaw, modern editions and MS. Arch. Seld., I believe was intended by Chaucer, but never actually accomplished. To treat this subject further would anticipate a future book on the evolution of the C. T.
2 The latter still maintains the allusions to W. B. Prol., a work which is striking enough, even without the intended tiff, to be in mind during the later part of the C. T.
'•'• Did Touchstone remember this line (As You Like It, V, iv, 55) ?
CH. V, § 8] THE WIFE OF BATH'S PROLOGUE AND TALE, ETC. 219
This conspicuously ignores the far more detailed and vivacious account in the Monk's Prologue of the manners and customs of Mistress Bailey. Surely, therefore, it must have been written before it.1
This concludes the present discussion of the chronology and development of the Canterbury Talcs. It is hardly necessary to say that the evidence presented has differed in value, and the con- clusions accordingly in certainty. They are presented for Avhat they are worth, because the publication of plausible conjecture, founded on investigation and recognized as conjecture, leads in the long run to the most fruitful and reliable results. Up to the present time surprisingly little investigation has been done on the Canterbury Tales, considering that they have been recognized for five centuries as the greatest work of our first great poet. The reason, no doubt, is the complexity of the problem, and the in- accessibility of much of the evidence. Chaucer students await with deep interest the publication by Mr. George Stevenson of a full description and analysis of all the sixty-odd MSS. of the Canter- bury Tales. We may then be in a position to show that the very puzzles which make the study of the work perplexing, such as the different readings in the Shopman's Prologue, 1179, and the presence in some MSS. of the "Host-stanza" after the Clerk's Envoy, help to provide the solution of the whole problem. In putting the poem together, Chaucer did not cover his own tracks. By painstaking examination of all the evidence, and by harmony among reasonable guesses as to separate problems, we may hope to arrive at something like certainty as to the way in which the Canterbury Tales came into being and into their present form. But the time has not quite come yet for putting results together.
1 Several other parts of 0. T. are more or less closely connected with 8h. T., W. £. P.. or Merch. T., either by parallel passages, by showing the influence of the same reading or by some striking correspondence in subject ; these are Pard. Prol. and T., Mill. Prol. and T., Reeve's ProL and T., Frankl. T., Sumn. T. , Mane. T. , Pars. T. It must be said, however, that the evidences of connection are much slighter than in the cases which I have discussed. I advance nothing as to the dates of these poems ; I simply raise the query whether the connection means anything. Mr. George Shipley (H. L. N., x. 275-6) shows some reason to believe that Frankl. T. directly alludes to W. B. T., and perhaps even Cl. T. ; cf. F 745-7, 751-2, 764-6, 792-3 with (e. g.) D 1038-41.
APPENDICES.
APPENDIX A.
The Date of Gower's Mir our de TOmme.
THE date of Gower's Mirour de lOmme is of no little importance in Chaucer investigation, for it will aid in ascertaining the dates of more than one of Chaucer's poems, especially the Troilus. Professor Macaulay determines it to be about 1 376-9 ,l but the matter is so important that it is worth while to discuss and strengthen his evidence.
Macaulay points out that lines 22801-24 refer to the conditions at the end of Edward IIT.'s reign, especially to the domination of Alice Perrers:
" Voir dist qui dist femme est puissant, Et ce voit om du meintenant. . . . Qe femme in terre soit regnant Et Rois soubgit pour luy servir. Hois est des femmes trop de£u, . . . Dont laist honour pour foldelit."
This implies a date some time later than August, 1369, when Queen Philippa died, after which the liaison became more open than before;2 and very likely a good deal later, since things gradually became worse. In 1376, Parliament had to legislate against the Perrers woman.3 But the passage may quite well have been written after Edward's death, June, 1377. It may reasonably be doubted whether Gower would have cared to express himself so fully and frankly on the king's shortcomings, before the king's death, in a poem meant for publication ; his other two great works were clearly meant to reach the royal eye. The passage simply expressed generally contemporary conditions, and may well denote a foregone conclusion. .
1 Complete Works, I. xlii.-xliii.
2 Diet, of Nat. Biogr., xvii. 66.
:1 Stubbs, Const. Hist. ii. 431 ; Th. Walsingham, i. 320. 220
APP. A] THE DATE OF GOWER's MIROVR DE L'OMMK. 221
Lines 22225-359 were almost certainly written by or after 1371. At the very beginning of Gower's discourse addressed to kings, lie devotes himself to a king's duties to the Church and to the prelates and dwells with especial disapproval on excessive taxation or pillag- ing of the Church (22242-5, 22276-8, 22297-359 ; here comes a long lacuna). Now in 1371 the action of Parliament was more strongly anti-clerical than for many years before ; very heavy taxes were laid on the Church, there was even talk of confiscation, and several bishops were ousted from civil offices, a movement which, according to Stubbs, King Edward may even have instigated.1 There can be little doubt that this passage was written not earlier than 1371, and since Gower was a strong and conservative supporter of the Church, his feelings may well have been keen for some little time after that date.
In several passages it is difficult not to see the influence of Gower's friend Chaucer's journey to Italy, which gives 1373 as the earliest date possible. All are agreed that Gower knew no Italian. Yet lines 3831-4 ran :
" Sicomme ly sages la repute, Envie est celle peccatrice, Qes nobles courtz de son office Demoert et est commune pute,"
which cannot be independent of Dante's words on envy :
" La meretrice, che mai dall' ospizio Di Cesare non torse gli occhi putti, Morte comune, e delle corti vizio "
(Inf., XIII.2 64-6).
The phrase ly sages Gower frequently uses to introduce quotations from various sources. We can hardly avoid believing that Chaucer read or repeated the passage to Gower.3 Secondly, the reference of
1 Stubbs, Const. Hist., ii. 420-4; Green, Short History (N.Y., 1890), 234 (chap, v., sec. 2).
2 Ufizio, curiously, occurs in line 62, and peccatrici (the only time in the D. C.) in XIV. 80. Chaucer of course quoted the same passage, less exactly, in L. G. W., Prol. F, 358-60. Gower quotes it again, far less exactly and many years later, in Conf. Am., II. 3095 ff. By this time he had forgotten its origin, and attributed it to "Senec," but Haase's exhaustive index shows, as indeed we should expect, that it is not in Seneca's writings ; nor do Fraticelli, Scartazzini, Moore, or Paget Toynbce attribute Dante's words to Seneca or to any one else.
3 This is the only clear case of Dante's influence on Gower. But M. 0. 11953-6 sounds Dantesque. The anecdote of Dante in C. A., VII. 2329* ff., seems more likely to have come through Chaucer than through a work of Petrarch's or otherwise.
222 THE DATE OF GOWEn's MIROUR DE i'o.V.l//;. [APP. A
"la geste deTroylus et de la belle Creseide" (5253-5) seems to me quite certainly to Chaucer's poem, and therefore postdates 1373, presumably by some years.1
For lines 2142-8 the date of 1377 or earlier is quite certain, as Macaulay shows ; 2 for as an example of the sin of Inobedience Gower speaks of the French as in rebellion
" A celluy qui de sa nescance Le droit depar sa mere prent."
This can only refer to Edward III. and the Hundred Years' War, and must have been written before June, 1377, when he died. But it may well have been written only shortly before ; from 1360, just after the Treaty of Bretigny, to 1369 "peace was fairly preserved," but during 1374 Aquitaine revolted from England and joined France, and during the ensuing years hostile relations were only partly interrupted.3
For lines 18817-40 the date 1378 or later is equally certain, as Macaulay also shows ; Gower, in addressing the Court of Rome, speaks of the monstrous birth in the Church of one body with two heads, obviously referring to the Great Schism, which began in September, 1378.4
This date may seem inconsistent with a somewhat later passage. In discoursing on emperors, and addressing Borne and speaking of her spiritual head, Gower says :
1 Two doubtful points may be added. In 23233-68 Gower discourses on the excesses and outrages of tyrants of Lombardy, in a rather hearsay style (" ascuns diont," " om solait dire "). One thinks immediately of L. G. W., Prol. F, 374, and wonders if Chaucer was not the reporter, in 1373 or perhaps after his mission to the Milanese Visconti in 1378. In 18697-732 the Court of Rome is reproached for neglecting to make peace between England and France ; the reasons for its neglect are said to be lack of charity and of impartiality, and the fact that it has wars of its own in Romagna. The two latter reasons would apply to pretty much any of the Avignonese popes ; but Milman tells us (Lat. Christianity, N.Y., 1862, vii. 201, 218, 219-220) that about 1352, 1370, and 1370-77 Innocent VI., Urban V., and Gregory XL (especially the last) did tiy more or less sincerely to make peace. But the humiliating failure of the negotiations of 1374-5 might well be attributed to the Roman Court by an Englishman irritated by the diplomatic victories of the Papacy over England in "the negotiations which were carried on at Bruges fora concordat with the pope, "under the shadow of which, according to &ivh\)&(Const.Hist.,\\. 427), the peace-conference met. There were also unsuccessful negotiations for peace in 1376 and 1377, but the war of course lasted on for years.
2 I. xlii.
:i Thomas Walsingham, i. 317-18 ; Green, Short History, pp. 233-4.
4 Macaulay (p. xlii. ) suggests that this passage may be a later addition, not seeing that the passage about Alice Ferrers may well have been written after Edward's death.
APP. A] THE DATE OF GOWER*S MIROUR DE L'OMME. 22$
" S'il avient qu'il t'est prochein, Lors tolt de toy le flour et grein, Et laist la paile deinz ta bonde, Et puis se tient de toy forein " (22195-8).
This seems to be an equally certain allusion to an earlier state of things, the Babylonish Captivity, 1305-1377, but stands more than 3000 lines later than the reference to the Schism. Yet it is not at all necessary to believe that it was written before the discourse on the Eoman Court. Urban V., elected in 1362, had been zealous to restore the Papal See to Rome, but did not do so till 1367, and returned to Avignon in 1370. Gregory XI., elected in 1370, permanently restored the Papacy to Eome in January, 1377, but at his death in 1378 he also was meditating a return to Avignon.1 For some years the permanency of the Papacy in Eome must have seemed highly doubtful, and such words as Gower's quite natural- Tins passage is the only suggestion that the Papacy had ever been anywhere but in Eome, though over 600 lines are devoted to the Curia, and though the abuses of the Avignonese court were particu- larly obvious to the English. On the contrary, it is implied or stated again and again that the seat of the Papacy is Eome, and Avignon is never mentioned in the poem. A bull comes " du Eomanie" (18995) ; of the upright clerk in contrast with the simoniacal " provisour " it is said :
" N'a Eome s'en vait pas serchant" (16109) ; 2
cf. 3330, 7360, 18450, 18502, 18421-19056 passim, 20349,21445. In many of these cases the mention of Avignon would have barbed the shaft. And in many cases Gower's omission to mention the domination of the Papacy by the Crown of France would be strange indeed if it was'* so dominated. So there is nothing against the opinion that most of the Mirour was written after the termination of the Babylonish Captivity.
At first sight, the mention of " Innocent" 3 in 18783 suggests the pontificate of Innocent VI. (1352-62). Yet this passage stands but
1 Milman's Latin Christianity, vii. 209-26.
2 Cf. Vox Clam., III. 1551, 1575 ; he calls the provisors "Romipetre."
3 " L'estat du pape en sa nature Ne porra faire forsfaiture En tant comme pape, ainz Innocent . . . Oil puet mesfaire d'a venture."
224 THE DATE OF OOWEB's MIROUR DE L'oMMK. [APP. A
thirty-four lines before the mention of the Great Schism. Macaulay is doubtless correct in thinking the name " only a representative one," and this allusion may as well denote a foregone conclusion as that to Edward and Alice Ferrers, both alike being used to illustrate general truths. Gower may have meant some irony in the use of the name Innocent, he may have thought it less disrespectful to use the name of a dead rather than of a living pontiff, and was doubtless glad of the rhyme in a stanza which requires six of each.
So early a date as 1362 is contradicted, and the other evidence confirmed, by the fact that, as no one who has toiled through the Mirour needs to be told, it cannot be the work of a very young man. This is also clear from specific internal evidence ; l formerly, he says (27337 if.), he abandoned himself to " foldelit," and wrote " fols ditz d'amours," but now all is changed. He has come late to repentance (27300). I3ut he lived till 1408, and about 1390 pro- duced the Confessio Amantis, the liveliest of his works. Macaulay conjectures that he was born not far from 1332,2 which fits all the conditions. So for the Mirour some time about the seventies would seem to be indicated.
Finally, it seems fairly certain that the poem was finished before or by 1381. As Macaulay points out, the Peasant Revolt of 1381 produced a profound effect on Gower's mind ; in the Vox Clamantis it forms the subject of the whole first book, and the rest of the work is devoted to ascertaining the causes of it. That Gower dimly fore- saw such troubles is shown by a number of remarkable passages in the Mirour, which Macaulay points out (p. xlii.) ; but of the events of 1381 there is not a word. The aryumentum ex, silentio seems to me convincing.3
Everything indicates that the composition of the Mirour must have fallen wholly or almost wholly in the seventies. The effect of all this may be more convincing if put in tabular form :
1 Cf. Macaulay, I. Ixii.
2 I. Ixii. Cf. Macaulay's note in the Athenaeum, no. 3856, p. 385 (September 21, 1901).
3 It is worthy of remark also that there is not a word of the Bishop of Norwich's Crusade against the Flemings, 1383, which is dwelt on in the Vox, III., cap. vi. ; yet a mention of it would have given peculiar point to some of the remarks in the Mirour on the bishops and the regular clergy. Another matter ignored which is prominent in Gower's later work is the growth of the Lollards ; it was in 1382 that Archbishop Courtenay began his campaign against Wyclif, and that the Lollards may be said to have tirst risen to a bad
APP. A] THE DATE OF GOWER's MIROUR DE L'OMME. 225
Rebellious French, 1. 2142 1377 or earlier.
Dante quoted, 3831 1373 or later.
Troilus mentioned, 5254 after 1373.
? (Pope's neglect to make peace, 18700 ... 1374-5 or later.)
Great Schism, 18817 1378 or later.
(Bab. Capt. alluded to, 22195 1377 or thereabouts.)
Anti-clerical movements, 22297 1371 or later.
Alice Ferrers, 22807 1369 or later.
Gower not a young man hardly before 1365.
Bab. Capt. almost ignored 1377 or later.
Peasant Revolt ignored 1381 or earlier.
If my arguments are accepted, it will be seen how perfectly satisfactory and consistent this all is. Everything supports the two crucial arguments adduced by Macaulay — that 2142 ff. was written in 1377 or earlier, and 18817 ff. in 1378 or later; and the practical ignoring of the Babylonish Captivity indicates that the middle of the poem must have been reached not earlier than 1377. There is not the smallest reason for believing that the poem was not written in about the order in which we have it, or for postulating interpolations. My conclusion reaffirms Macaulay's, except that, considering the great length of the poem, I should extend the limits to about 1375-81.
The passage (5245-56) in which the Troilus is mentioned must have been written, in all probability, by 1377, since 18817-30000 was written between 1378 and 1381. If we accept the reference as being to Chaucer's poem, we may, without reasoning in a circle, declare that it cannot have been made much earlier. So it seems to me that for Gower's reference to the Troilus the date about 1377 may be accepted with considerable confidence.1
1 A recent and rather extensive thesis on French and Latin sources of Gower's treatment of the vices and virtues (unimportant for my purposes) is by Miss R. E. Fowler, submitted for the doctorate of the U niversity of Paris (Macon, 1905). In a MS. in the British Museum (Addit. 15606, ff. 61-351) is an allegory on the Seven Deadly Sins and the contrary virtues, which may be worth mentioning as a parallel to Gower's. The allegory is military, based on the siege of Jerusalem by Nabugodonosor.
DEV. CH.
226 THE KNIGHT'S TALE AND THE TESEIDE. [APP. B
APPENDIX B.
The Knights Tale and the Teseide : A Table of Parallels.
ON this table is based much of the reasoning in chapter III., section 1, which will explain its peculiar form; but it is believed that the contents of it may prove otherwise useful to Chaucer students. Each column corresponds to one line of the ottava rima. The number before the sign of equality always indicates a line of the Kniglifs Tale, according to the Chaucer Society's numbering ; and the numbers after represent the book and stanza of the Teseide, the column indicating the line of the stanza. Parentheses show that one English line corresponds to two or more Italian lines, or a considerable part of them, but it has not always seemed necessary to take account of a very trivial part of a line. Therefoi-e a line-number in parentheses always appears in at least two columns. I have meant to include every Italian line to which an English line is clearly due, even though there may be no verbal agreement. Italics indicate that the translation is very close. Where two lines in the English answer to one in the Italian, occasionally one is a close translation and the other not ; this is indicated thus— 2385-Q.
In preparing this table I have been very materially aided by Mr. Henry Ward's marginal references to the Italian in the Chaucer Society's Six-Text edition. It is proper to say that these notes are somewhat deficient in both extent and accuracy ; but, so far from wishing to reflect on the labours of their author, I must add that my own may not be faultless, though they have been prepared with the utmost care, and thoroughly verified. Very nice cases constantly arise, and not only would the tables of no two men agree perfectly, but any man would possibly revise his own every time he reviewed them. What errors there may be, however, cannot be serious, for it will be seen that the question at issue is always one of proportion.
APP. B]
THE KNIOHT'S TALE AND THE TESEIDE.
JNCOOOOi
II II [I II II II II II II II II II II II II
tOCOrlOOCO^IO rHTj(l>.
i-tO^COt^t-'tDt^. O»O<O
a>O5OOOO<MCOCO COm>O
OiOOO'-HIN
,a
'S.
-0
|
kft |
CN" <M" «<T oq" of co" eo" |
T* HTrt* HTHTi-CtCvcr t-TjCco" |
|
H |
II II II II II II II |
II II II II II II II II II II II |
|
s |
CO <M rH OO rH CO <N O CO | rH (N »O OS OS OS O O O O <N |
§|T11S22 i?3 |
|
rH |
r^ •"" |
|
|
,-, |
CD '— .CO ^^-— v |
|
|
<MC^ |
i— 1 rH OS OS rH CO CO l--» <N |
|
|
H g |
II II II |
ii ii ii ii ii nil n n |
|
l-q |
OS !>. ?>» |
TjfOOSrHOO COCO»O(N IO-*CO»OCO OOOOTjl |
|
•^CDt^COOO OSOrHrH |
||
|
^ ^^^ |
||
|
^_^ f s |
^^00 OS ^. ^ ^~. ^~, |
|
|
COr~-rHOlOOrHf- |
OO C*l CD f~* O rH C^I CO "^ kO rH H^ OJ |-~- |
|
|
(MCNCO»OCOrHrHTj( |
COCOOSrH rHCOCOCOCOCOt^rH(NrH |
|
|
H |
II II II II II II II II |
II II II || II II II II II II II II II II |
|
|
||
|
1-3 |
rHrHlCDO^ICOOS |
OO<MOO OSl>.COOOOrHOOi|^J<cb |
|
OSOSlOOSOOOi— i |
^ iO OO OO CO OS OS CTj ^_» O rH r-l rH r—i |
|
|
OS v — |
|
|
|
£i § 2 ii^S S?S |
SO ^-~. O OO O ^-v • — . |
|
|
(M S . |
n n n n ii n n n |
II II II II II II II II II II II II II II |
|
3 |
OCO^OSOSrHScO |
IO ^O ^O C*^ CD l^» OO OS 1 OS d Tt* CD OO |
|
OSOSOrHrHINCOCO |
Tjt^OCocOOOOOCOCO (MOSOi— IrHrH |
|
|
v_x , -^-^_- |
-' " ^ S " |
|
|
COOCCI>.J>.'*COt-~OOC |
^ OO ^--l>. rH J^ (N CD CO rH O ffTcO •*»( "d |> rH rH CM CD |
|
|
i— i H ' |
II II II II II II II II II 1 |
1 II II II II II II II II II II II II H II H II H H H |
|
» |
||
|
J |
§£oo2SS££S |
j.J;*^f»ls»2I.ss§s.ggs2|.s |
|
X -^^ |
i— . rH |
228
THE KNIGHTS TALK AND THE TESKIDE.
[APP. B
|
i-H ift CO OS •'J' CO Tt< -*• |
co I-H os <^> t- OS CO ®* |
in i^, c; o ?» |
|
|
oo |
co" co iCtC |
jCt^Tt-Tz^. |
l^ i^ ?^ l^ >> |
|
H g 3 |
II II II II »n o o co t^ os m m i-H I-H (M CM (M (M CM_CM |
II II II II os co I JN. CM CO CO «O CM (M in ©* CO CM |
II II II II II. OS Ci) Q^ CO *Sfr CM O} <^> CM S} |
|
fsss |
^SScSSI? |
C^ ^-O Cl ?^» GO |
|
|
a |
II II II II |
II II II II II II |
II II II II II |
|
a |
221§ |
S £ CM "Y In £ |
25S-55 |
|
is ~ |
~"~^ CO |
CM o |
|
|
*rt O Oi CO »™* |
E^cfcSSS |
m co oo o~f- oo |
|
|
CO H |
H II II II II |
II II II II II II |
II II II II II II |
|
M |
C-l CM CM CM (M |
CM CM CO CO CO CO |
CO 05 1 CO 1 I-H — ^' — ••& <&t CM |
|
o»nososeococo»-H |
i-~ t» t-~ OS OS |
'-'J CO do O I— CO |
|
|
in H |
II II II II II II II II |
II II II II II |
II II II II II II |
|
^ |
i-H r-H i-H CM 4} CM CM CM |
^5 CO CO CO CO |
CD OS 0^ I-H OS Tl« 1 1 >-< CO 1 l-H |
|
^l CM CM |
|||
|
Tji |
^5 in os co co t-— co |
os I-H m *-H os o t-~ CO CO OS CO OS |
co'o'co oo" |
|
fc |
II II II II II II II |
II II II II II II |
II II II II |
IM<M(M(N<MC<I(M
o «^ >o to t~- os
H H H II II II
OSOOi-l<M>O CNCOCOCOMW <M CM CM CMIN O
II II II II II
- CM CM CM (N ^J
to tr>T?x tCt-C
II II II II II II
CO CM CO to 0 OS OCMtN^JOt--
CMCMCM^CMCM
II II II II II II II II
t^ OS 00 CO CO t-H Ci 10 C5OS<2i,-l<N»I5Ki<O
<s,j(N<5ococoe^eoco
II II II II II II
•* 0 -* V|- »-H ?>. C
l--OOOsC>*-l&J»
co co co -*. •^vn
APP. B]
THE KNIGHT'S TALE AND THE TESEIDE.
229
U^ OO t-~ CO SQ 1O kO 5O CO J^ <M <N !N IN <M
I-H i-H 00
1-1 r-l *-
- - _-<N
t-~OOOCCO
rH »1 II II
•* CO •-"
O5 CO <N i-H SO OO r-l CO 00 OO O5 Oi <M <M 71 (N
CO ,_ i-H <M <N 10
*>• t-- OO Oi i— I
II II II II II
|
"i £_ 1 oo I-H l '<M (N |
1 |
OO O5 |
|
|
O5 (N OO !>• rH i-H Tjl |
IT, rH CO ^ I-l r-l |
1O O O O m I-" ^ r-l CO CO -tf (jq (N cq |
|
|
M K |
t^l-. 00 O5 II II II II -# OO OO 1 <N |
I— 1 rH I— t 11 II II oo oo oo |
II II II II || II II t^- OO 00 rH °p. (M CO OO OO CO O3 gj O5 O5 |
|
l-H l-l _ |
o^Qj |
rH rS CO C^ <N S |
|
|
g |
1-- t~- I-H II II II |
T-H *~H 'M II II II |
II II II II II II CO O3 CO O »O O CO OO (N O5 <3i O5 <N <M O5 <N <N <M ^-XQO |
|
CO O C<1 ^-05 CO CO OO i— 1 CO |
o 3 ^ ® to S ;* |
r^S S%8g |
|
|
CO H |
_ - . . - .. <M~ CM 0 O" t^t^J-~?^r^C5i-li-HrHrH II II II II II II II II II II |
cr^Ti-Ti-H e<! i-TrH i— 1 i— | r- 1 i— 1 ^ i— 1 i-H ii ii ir ii ii ii H l-~ Oo OO CO '-^ OO OO (N i?J (M (N ^-(N 0^ |
r-l I-H i— 1 >-) r- 1 rH »-H »H rH '-H i— 1 i-H II II II II 11 II 2>. OO O5 ^> 1 ^ Oo | OO Cl 5O C5 " l>- OS ' — ' OO O1 2- |
|
2~O ^}- tO C5 OO CO i— i |
•t^^ ^^^ tO ^ r-H to i-l i-H |
•^ IO CO O ^ ^ CO i-H i— i i— t CO ^ CO CO |
|
|
M |
II II II II II II II II II II •Jf (N S\> (M C^ <?> <N <M (M -* CM |
II II II II II II 1- ?^ 00 Oo OO OO <M •?•> c-1 ^; !M (M |
II II II II II II II tO rH O 1— 1 Lc^ O O co oo oo oo oo oo O5 (N OJ !M <M ®^ CM <M |
|
s?*^..a «««»sas |
O lO i— i ^ •• ^* |
Tjl 1O CO O IO OO t-- Oo rH rH rH CO CO CO CO *^H |
|
|
H g |
»«•»,»•»••». *(M O O t>» t>» t^ 'CO OOOi^OiCiOif— *i— ii— 1 II II II II II II II II II II II II II (N |
O O O r* '**'-* ^> ^ rHr-lr-lr-ir^rHr^r^ II II II II II II II II |
II II II II II II II II |
230
THE KNIGHT'S TALK AND THE TKSKIDE.
APP. B
O CO Oi ^P C> OO OO C^J CO "^ OS Oi O O O
CS <N CO CO CO
|
J |
«O CO ^! CO CO Oi O5 O O O <>} (>l <3O CO CO |
|
1A X>» OO CO |
|
|
50 |
rH i— I rH i— 1 |
|
H g |
II II II II »fl> CO O <M CO (M CO O 05 O O i-c (M CO CO CO |
|
CO ^-**~* iO ^5 ^ J>. OO Ci |
|
|
«v |
I-H ^> <N cT (N~ <>> |
|
H 2 |
II II II II II II O ^O !>• CM CM v> C35 Oi O5 O O "ii <N ^ CN CO CO 615 |
(M <Sj <N (M CN (N
Ift O5 ?^ CO O <-l CO JO CO ?^ I (M CO CD 05 05 0} <N O O O
C* (N ^} OO CO CO CO
li-*i-ii-lf-ii-l>«|i-4C43tG4G49)C4<N
.2 a
Tc ^
I
d
1
E-i co
APP. c] CHAUCER'S TREATMENT OF THE TESEIDE. 231
APPENDIX C.
CJiaiteer's Treatment of the Teseide.
BEADING the Knight's Tale with the Teseide, one is frequently struck with the preternatural condensation of the Tale, and wonders how Chaucer could have left out so much, sometimes such fine touches and passages, of Boccaccio's admirable poem. In the Teseide the portraiture of Emily is far more vivid and complete than Chaucer's. In book III., for example, she hears Palamon's " Ome ! ", becomes perfectly aware that the prisoners are in the habit of watching her from the window, and shows a little harm- less coquetry; when Arcite has returned to Athens (book IV.) Emily recognizes him at a feast, wonders about him but discreetly holds her peace. Emily is the first to see the cousins fighting in the wood, stands stordita, then cries to Theseus (V. 81). In her attitude towards her suitors there are many nice touches ; in spite of her desire to remain single, she feels a certain attraction towards them (VII. 85), and her soliloquy during the tournament is extremely good — she is not worthy, she says, of the courage which is being dis- played (VIII. 97). Elsewhere, too, Chaucer omits good passages. He manages less plausibly Palamon's escape and meeting with Arcite, in books IV. and V. of the Teseide, where the detail " by helping of a frend " is fully accounted for ; he says not a word of Palamon's devoted squire Panfilo, and wholly omits Arcite's triumphal procession after the tournament, which Boccaccio strikingly portrays, and which we should suppose would at least be mentioned — Arcite has to be carried, and for his pleasure the vanquished knights follow voluntarily (IX. 30-4).
One characteristic change, however, Chaucer does make, in the treatment of the characters of Palamon and Arcite. In the Teseide, though Arcite cuts slightly the better figure,1 they are hardly distinguished, and both are valorous and honourable young knights
1 When he is released from prison (III. 74-6), Arcite takes a tender farewell of Palamon ; and is very pathetic and regretful (V. 45 ff.) when Palamon demands a combat.
232 CHAUCER'S TREATMENT OF THE TESEIDE. [AI-P. c
full of all worthy emotions.1 But Chaucer draws such a sharp distinction between them that it can hardly have been unconscious. Arcite is very highly praised by the poet and by his own associates for his physical development (1422-5) and for his character (1429-32). He is an agreeable spectacle, like Chaucer's own Squire, in his cheeriness and youthfulness over his ramble on May-day (1500 ff.), and in his falling into lover's dumps (1530). He is judicious in his retorts to Palamon's reproaches (1162-86, 1606) ; honourable and generous (1608-16), modest and manly (2393-9) ; and shows the last magnanimity in commending Palamon to Emily (2783-97). Palamon, on the other hand, is more jealous and less affectionate toward his cousin (1281-1333) ; despairs readily when discovered by Theseus (17 1 5 ff . ) ; is ungenerous towards Arcite (1722-31, 1740 2); has no desire for victory, but only cares for the possession of Emily (2234 ff.) ; and cares no more what becomes of her after his death than the Pardoner about the souls of his dupes after theirs. With this passage it is curious to compare the Italian lines from which it is altered :
" Thanne rekke I noght, whan I have lost my lyf, Though that Arcita winne hir to his wyf " (2257-8) ;
" Che non sarebbe senza lei la vita, Vedendola non mia, ma si d' Arcita" (VII. 49).
Such a change as this must be the result of a purpose. Palamon's only amiable traits are his courage (1591-5), and his grief after Arcite's death, and that seems to be only conventional (2882-4, 3062).
Yet it is he who after all wins Emily. Here is an example of the subtle and perhaps only half-deliberate satire which runs through the Knight's Tale. Poetic justice itself shows in a rather ironical light where Chaucer reverses Boccaccio's order and gives Palamon the better claim to Emily by letting him fall in love with her half-a-minute before Arcite. A tone of levity and even of gentle ridicule rises here and there all through above the state and pathos of the poem, and contrasts with Boccaccio's perfect
1 Kissner, in his excellent dissertation, Chaucer in seinen Heziehungen zur italienischen Literatur (p. 63), says that Chaucer misses the point of the Teseide, which is the conflict between friendship and love.
2 In the Teseide (V. 86) it is Arcite ("Penteo") who first speaks when Theseus discovers them ; he does not accuse his cousin, as Palamou does in Kn. T.
A PP. c] CHAUCER'S TREATMENT OF THE TESEIDE, 233
gravity. Satire is easier to suspect than to prove, especially in a poem written when ideas of what is ludicrous and the connotations of words were so different from what they are now. It may perhaps be purely modern to see a slight lack of seriousness in the account of Palamon's tears wetting the fetters on his great shins, and of the tower resounding "of his yowling and clamour;" and in the description, condensed yet exaggerated, of the sylvan combat. But levity is impossible to mistake in the passage about Emily's bathing (2282-8). l In what Chaucer says of the experiences of Arcite's soul there is certainly a lack of seriousness; after the piercing pathos of Arcite's death there came a reaction. The remarks of the old Egeus, who Professor Child used to say was " delicious," read like a satire on commonplace consolation, and after his beautiful thought that just as nobody dies without first living, so nobody lives without dying, Chaucer adds (it would seem with a little irony) "over al this yet seyde he niuchel more to this effect." These are the most striking passages, but they are not the only ones.
The light and satirical tone x>f the Knight's Tale seems to me to favour the view that it was written near the time of the Canterbury Tales; just as the omission of so many of Boccaccio's good touches suggests that he greatly condensed from the first, and that, therefore, the Knight's Tale, as we have it, is practically identical with the Palamon and Arrite.
1 Dryden attributes the narrator's hesitation about speaking freely to the fact that the rites were pagan but sacred. This is only one of many cases where Dryden misses Chaucer's point, and tries to make more imposing what Chaucer meant to be light. Two others are his expansions of the two last passages mentioned in the above paragraph. See Palamon and Arcite, III. 197-206, 844-53, 883-90.
THE END
RICHARD CLAY & SONS, LIMITED,
BREAD. STREET HILL, B.C., AND
BUNGAY, SUFFOLK.
, -
MAY 15 1353
Chaucer Society, London Publications.,
CIRCULATE AS MONOGRAPH
PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET
UNIVERSITY OF TORONTO LIBRARY