[in The Riverside Chaucer, ed. Larry D. Benson (3rd ed.) (Boston: Houghton Mifflin, 1987)] xlii INTRODUCTION gon henne (Tr 3.630). In some phrases the infini- with inversion of verb and subject, as Al speke he tive has a passive sense: to blame (Tales 1.3710), never so rudeliche (1.734), Al were he short to preyse (VI.42), to drede (VI1.3063 ),jor to chese (111.624), or in other ways such as hou soore that (Tr 2.470). This is related to the use of the me smerte (1.1394), wher-so she wepe or synge inherited "inflected infinitive" in ye woot what is (11.294). In all these the verb is usually subjunc- to doone (what is to be done; Tales 111.2194). A tive, but not invariably: e.g., Though in this toun concessive use of the jor to infinitive appears in is noon apothecarie (VII.2948), what so any wom- places like for to dyen in the peyne (though we man seith (VII.2912). were to die under torture; 1.1133), For to be deed (even if I were to die; IV.364). In sentences Negation The primary negative adverb is ne, expressing the terms of an agreement the infini- placed before the verb (or incorporated in it in tive often accompanies a nominative pronoun: those contracted forms noted above in the de- Andye, my lord, to doon right as yow leste (IV.105). scription ofverb forms, p. xxxviii), which is suffi- The subjunctive in main clauses may express cient to mark a sentence as negative: she ne wiste wish, as God yelde yow (may God reward you; what it signyfied (1.2343), I noot how men hym calle Tales 111.2177), an imperative ofthe first person (1.284). It is very often reinforced by another plural, as go we se (Tr 2.615), and concession, as negative, most frequently nat or nought: he ne Bityde what bityde (Tales VII.874), Be as be may le/ie nat (1.492), Ne studieth noght (1.841). Nat or (VII.2129). It is used in the apodosis (main nought alone, following the verb, may also clause) of a hypothetical conditional sentence: suffice, as in His arwes drouped noght (1.107), It is A clerk hadde litherly biset his whyle (a scholar nat honest; it may nat avaunce (1.246); exception- would have wasted his time; 1.3299). In depend- ally, it precedes the finite verb, as in Nat greveth ent clauses it is largely concerned with condition us youre glorie (1.91 7), Ye felen wei yourese/f that I or hypothesis, as ifgold ruste (1.500), ifthou telle nought lye (Tr 2.1283). Sometimes a series of it (1.3505), and especially in the frequent tag if words in a passage, pronouns as well as adverbs, so be / were, with a subordinate clause also using may all be given negative forms: He nevere yet no a subjunctive: if so be that thou my lady wynne vileynye ne sayde / unto no maner wight (Tales (1.161 7). It extends naturally to the negative 1.70-71). condition in clauses introduced by but if (un- less)-Who may been afool but ifhe love? (1.1799), Interrogation A pOSItive sentence is normally and the unfulfilled condition expressed by the converted into a question by inversion of verb past subjunctive nere (were [it] not), as Nere myn and subject: Wostow that wei? (Tr 1.775), se ye extorcioun (if it were not for my extortion; nought? (2.1465). When the question concerns 111.1439). Inversion of verb and subject, as in two alternatives, the interrogative may be inten- this example, often expresses condition: Had sified by beginning the sentence with the pro- thou nat toold (VII.2053), konne he letterure or noun whether, as Wheither seistow this in ernest or konne he noon (VIII.846). The subjunctive is fur- in pley? (Tales 1.1125). After interrogative ther used in hypothetical comparison, for in- words such as why, or what as object, inversion stance, As though he stongen were (as though he is also normal: Whycridestow? (1.1083), What do had been stabbed; 1.1079); it may express pur- ye? (1.3437), how thynke ye? (111.2204). pose, as Lest thee repente (111.2088); and anticipa- tion of a future event, as er that thou go henne (1.2356), til that it dye (111.1145). It implies the Versification uncertainty of information in Where that he be I Verse in English in the fourteenth century was kan nat soothly seyn (1.3670), and the unwilling- composed in two different traditions, which ness of a speaker to vouch for a report in I trowe were usually kept distinct, though some authors that he be went (1.3665), though it is likely that combined them in the same work. One system, in such dependent clauses the use ofthe subjunc- which evolved from Old English and the general tive was at least to some extent conventional. Germanic tradition, depended on the pattern of Another principal function of the subjunctive is stressed syllables in each line ofverse, linked by to mark concession, which may be expressed by alliteration of initial sounds; normally lines did the conjunction though (that), as Thogh thou heere not rhyme together. The other, which in Eng- walke (II.784), or by the introductory adverb al land began in the twelfth century and was imi- LANGUAGE AND VERSIFICATION xliii

tated from French and Latin models, depended Good Women and most of The Canterbury Tales. partly on the number ofsyllables in each line and No earlier model for this has been found; it partly on the linking of lines in couplets or may well have been his own extension to the groups by rhyming final sounds. Before the four- five-stress form of the pattern so familiar with teenth century, by far the most frequently used the shorter line. of the second type was a line that in principle Chaucer wrote rhymed verse in all his poetical contained eight syllables, but might in practice works. He clearly knew the alliterative form vary between seven and ten, arranged in rhym- (see the Parson's comment in Tales X.42-44), ing couplets. This form as used by French writ- and exploited features of it in a few passages ers regularly had eight or nine syllables, and the (such as 1.2605-16), but the alliterative form is syllable count determined the line. Syllable never essential to the structure of his verse. His stress is more prominent in English than in rhymes are in general careful, matching sounds French, and when this form was imitated in Eng- that must have corresponded exactly in the type lish syllables bearing stress usually alternated of English he wrote. In some words, or groups with syllables of less weight, so that the line was of words, he took advantage of the existence of characterized by four stresses, or beats. The alternative pronunciations (much as a modern number of unstressed syllables was less re- versifier may rhyme again either with main or stricted than in French, probably owing to the with men ). Thus the verb die often appears as dye, influence of the native type of line. This form rhyming with such words as/olye (e.g., 1.1797- was used in many English poems, some long and 98), but sometimes as deye, rhyming with words important, from the early thirteenth century on- like weye (e.g., 1.3033-34); the verb liste ward, so it was obviously well known long be- (please) often rhymes with such words as wiste fore Chaucer began to write. (e.g., Tr 1.678-79), but its variant leste, a pro- In spite of a good deal of irregularity of nunciation characteristic ofthe southeast ofEng- rhythm in the texts as preserved in the manu- land, rhymes with such words as beste (e.g., Tr scripts, there is no reason to doubt that this fa- 1.1028-29). Choice among optional variants miliar type of verse was what Chaucer intended such as these is one aspect of the flexibility of to write in his early poems The Book 0/ the Duch- Chaucer's use of language. ess and The House 0/ Fame. (Four-stress lines also The most significant difference between form part of the "tail-rhyme" stanzas of the Chaucer's verse and that of later centuries lies in Tale of Sir Thopas.) It has, of course, remained the greater number oflight syllables required or popular to this day. In his other poetry, how- permitted by the inflectional system of his lan- ever, Chaucer used a longer line, containing guage, described above. Some of these inflected five stresses, or beats, which before his time forms are comparatively familiar because related had appeared only rarely, in a few anonymous forms survive in particular circumstances in the poems. It is likely that this form was suggested modern language; for instance, participial forms by the decasyllabic line often used in French, such as aged and learned, when used as adjectives, especially arranged in groups to form stanzas, are pronounced as two syllables. In the four- but in part also by the eleven-syllable line used teenth century the ending -ed was pronounced as in Italian verse, which is of similar length but a separate syllable in other participial functions freer in rhythm. Chaucer took up the five-stress as well, for example, in perced and bathed (Tales line early in his writing career, in the eight-line 1.2,3). Other instances are unfamiliar to a mod- stanza of the ABC (a French form), and used it ern reader because the sound is no longer pro- again in The Monk's Tale. He used the same nounced. These are final unstressed syllables kind of line in seven-line stanzas (later called consisting of the neutral vowel written e (see p. cc rime royal") in The Parliament 0/ Fowls, Troilus xxxiii above). This vowel may function as a and Criseyde, and the tales told by the Man of grammatical inflection (indicating, for example, Law, the Clerk, the Prioress, and the Second case, number, mood, or tense) or it may be an Nun, as well as in much of Anelida and Arcite integral part of a word-thus in the yonge sonne and some minor poems. His greatest contribu- (Tales I.7) the e of yonge is the mark of the tion to the technique of English verse was the definite form of the adjective, but that of sonne arrangement of this five-stress line in rhyming had been part of the word already in its Old couplets, which he adopted in The Legend 0/ English form sunne. Pronunciation of these end- xliv INTRODUCTION ings is, for the most part, necessary to the rhythm the line may be called "Lydgatian": of the lines, and Chaucer, in addition, nearly gretteste ooth was but by Loy (1.120) always avoids rhyming words ending in e with On the other hand, there are sometimes more others in which it would not be etymologically light syllables (mostly only one) than regular or grammatically in place. A telling example is rhythm requires. A light syllable may precede a found in Tr 5.260-64: since the structure of the natural pause. seven-line stanza requires that the rhyme of the For he was late ycom e from his viage (1.77) first and third lines should contrast with that of the second, fourth, and fifth, Chaucer uses the That no drope ne upon brest (1.131) rhyme of -ly in the adverbs pitously and myghtyly Andforth we riden a litel than paas (1.825) against -ie in the French-derived nouns fantasie But in numerous lines it clearly does not do so: and folie and the (English) infinitive drye to Of court, and to been estatlich of manere (1.140) achieve the necessary effect. Another indication that these final vowels were sounded appears in Men moote yeve silver to the povre freres (1.232) such rhymes as Rome: to me (Tales 1.671-72), With a thredbare cope, as is a povre scoler (1.260) where the to must be stressed and Rome must Pekke hem up right as they growe and ete hem yn have two syllables. But though the final' vowel (VII.2967) was evidently generally pronounced, there are many exceptions within the line. In particular, (Chaucer did not divide his lines by a regular when a final -e is followed by a word beginning caesura after a set number of syllables.) with a vowel or with an h- that is either silent (as The same features appear in the poems in the in French words like honour) or in weak stress four-stress line, and through them and other (as often in such words as he, his, hem, hire), it is means Chaucer gave his verse great freedom and normally elided. Thus in So hoote he lovede (Tales variety of movement. To illustrate using a char- 1.97), the -e of the adverb hoote, grammatically acteristic passage, the following extract from the correct, is not pronounced before he, The -e is General Prologue (1.285-308) is marked to also silent in many short unstressed words such show metrical stresses and the probable treat- as hadde, hire, oure, and it is sometimes slurred in ment of unstressed syllables. Stressed syllables polysyllables such as mesurable (1.435), benefice are marked by accents, but the degree of stress (1.507). Unstressed vowels next to certain con- must, of course, have varied with individual sonants within words may also be slurred, so that readers. Unstressed -e pronounced in final sylla- hevene (I.519) has two syllables, never (I.734) bles has a dieresis (..); when it is elided or has one. slurred, it is underdotted A regular line ofthe longer form contains five A Clerk ther was of 6xenford also, stresses, which may vary a good deal in strength That unto logyk hadde ygo, according to the sense, and indeed often be po- As leene was his hors as is a rake, tential rather than essential to a natural reading. And he nas nat right fat, I undertake, Usually a light syllable precedes each of the But lookea holwe, and therto sobrely, stressed syllables. In Ful was his courtepy, Biftl that in that seson on a day, For he hym yet no benefice, In Southwerk at the Tabard as I lay (1.19-20) Ne was so worldly for to office, a normal reading gives a regular rhythm. But the For hym was at his beddes heed order of stressed and light syllables is often re- Twenty bookes, clad in blak or reed, versed, as in the next line, Of and his philosophie, Redy to wenden on my pilgrymage (1.21) Than robes or or gay sautrie, and, more notably, there is sometimes no initial But al be that he was a philosophre, light syllable at all, and the line may be called Yet hadde he but litel gold in cofre:' "headless." But al that he myght of his freendes hente, Twenty bookes, clad in blak or reed (1.294) On bookes and on he it spente: A small number oflines lack a light syllable after And bisily gan for the soules preye the second stressed syllable, giving a "broken- Of hem that yaf hym wherwith to scoleye. backed" effect which some fifteenth-century Ofstudie' took he moost and moost heede', poets, especially Lydgate, much favored, so that Noght 0 word spak he moore' than was neede: THE TEXTS xlv

And thdt tvas seyd in and reverence: ance of Manly and Rickert's edition, which pro- And short and quyk and /ul 0/ hy sentence'. vided the full collations that he had lacked in in moral vertu was his speche, editing the text for the 1933 edition, induced And glddly he and glddly tech e'. Robinson to reconsider his text of the Tales; his 1957 edition contained over 160 changed read- NORMAN DAVIS ings. That there were not more shows that Rob- inson was not convinced by Manly and Rickert's THE TEXTS theory that the Ellesmere is an "edited" manu- script whose scribe (or "editor") frequently The texts in this volume, with the exception of "corrected" the meter. That theory has now The Canterbury Tales, have been reedited on the been refuted by George Kane in his essay on the basis of a fresh examination of the textual evi- Manly-Rickert edition, which appears in Editing dence by direct consultation of the authorities, Chaucer (ed. Paul G. Ruggiers, 1984). The text including both manuscripts and early prints, printed here represents about the same degree though most often by means of facsimiles and of revision; when Ellesmere stands alone, its microfilms. These texts are thus at once revisions readings have been considered with somewhat of Robinson's texts and new editions in their more skepticism than Robinson showed in own right. Robinson's text was used as the base; 1957, and even more skepticism entered the that is, in most cases the editors use the same consideration of those cases where Robinson base manuscripts as Robinson chose and adopt chose a metrically smoother line from later his spellings, but each of his editorial decisions manuscripts against the testimony of both Elles- was reconsidered in the light of the evidence. mere and Hengwrt. But in those cases where it His textual judgments were treated with respect, seemed to the editors that the choice between but there are a good many changes; ideas about alternate readings was doubtful, Robinson's editing Middle English have developed consid- decisions were usually accepted. The text of the erably since Robinson's first edition appeared in Tales therefore contains no radical departures 1933, and the work ofscholars and editors in the from Robinson's 1957 edition, though it has intervening years has greatly advanced our been thoroughly revised. knowledge of Chaucer's texts. Most important, For The Book ofthe Duchess, The House 0/ Fame, Robinson did not have direct access to many of Anelida and Arcite, and many of the minor the authorities; for the poetic works he was poems, Robinson (like most editors) used the largely dependent on the Chaucer Society's tran- Fairfax manuscript as his base text, and the pres- scriptions, and for the Boece and Astrolabe he was ent editors do the same. This is a late manu- not able to take account of all the authorities. script, compared with Ellesmere or Hengwrt, That his texts emerge from revision with rela- and it is one of but four authorities for The Book tively few changes is a tribute to his sound edito- ofthe Duchess and five for The House ofFame, all rial judgment. ofwhich are late and, as Robinson noted, "agree The degree of change from Robinson's edi- in readings unsatisfactory in sense and meter. tion naturally varies from text to text. The text Earlier editors had therefore freely emended of The Canterbury Tales shows relatively few these texts. Robinson was more conservative changes, and it must be noted that this is not a than his predecessors, and his emendations were new edition in the same sense as are the other for the most part carefully considered. Larry texts in this volume. Only a few authorities were Benson's edition of The Book ofthe Duchess and consulted directly and those few only on occa- John Fyler's edition of The House ofFame do not sion. The editors, initially Robert A. Pratt and greatly differ from Robinson's texts. Many ofhis later Ralph Hanna III, in consultation with the emendations have been allowed to stand, General Editor, depended instead on work for though in many other cases the editors have re- Pratt's edition, The Tales of Canterbury, and on placed the emendation with the reading attested the materials contained in Manly and Rickert's by the manuscripts, with the result that in a few Text ofthe Canterbury Tales (1940). The text pre- places the text reads less smoothly than Robin- sented here is in effect, therefore, a revision of son's. Most of these involve "grammatical Robinson's work, with errors corrected and rela- forms apocopated in the manuscripts"-final -e tively few new readings introduced. The appear- and -n-which Robinson often silently restored.