University of Virginia Library

William Thynne's Printing of the Squire's Tale: Manuscripts and Printer's Copy
by
Donald C. Baker

That William Thynne used one or more manuscripts for consultation during his printing of the Canterbury Tales for his 1532 edition has long been assumed, but the evidence has been both vague and contradictory. W. W. Greg, in his analysis of the early printed editions using the first 116 lines of the Knight's Tale, concluded that Thynne had consulted a manuscript of

the large group that Manly was to call the cd* group.[1] Koch had suggested Phillipps 8137 (or Ph3, to give it the Manly-Rickert sigil), as a type of the manuscript used by Thynne.[2] But Pearsall, in his Variorum Chaucer edition of the Nun's Priest's Tale, concluded that for that tale there is no solid evidence of Thynne's use of manuscripts.[3] Ross, however, for his edition of the Miller's Tale in the same series, concluded that a manuscript very close to c/d in Manly's classification was used by Thynne for that tale.[4] I remarked in my edition of the Manciple's Tale that, though it was difficult to be conclusive in such a short poem, nevertheless the number of variants introduced by Thynne certainly suggested the use of a manuscript, or perhaps several.[5] The most intensive work upon Thynne's sources has been done by J. E. Blodgett, who in his essay upon Thynne in Editing Chaucer: The Great Tradition,[6] though he is quite specific about Thynne's use of manuscripts for other works, again concludes only that for the Canterbury Tales Thynne did use manuscripts, but does not speculate upon which manuscripts.

That Thynne was something of a scholar has long been recognized; we know that a good many Chaucerian manuscripts passed through his hands (Blodgett, p. 39). In the course of studying the printed tradition of the Squire's Tale, I have, I believe, uncovered evidence which allows us to be much more specific about the kinds of manuscripts which Thynne used, at least in the preparation of the text of that poem. I emphasize that I am restricting my conclusions to the Squire's Tale, for there is no reason to believe that, if

Thynne had had a number of manuscripts by him for the Canterbury Tales, he would have confined himself to one or two for the setting of the entire work. So, for the moment, my conclusions apply only to that fairly short poem.

First, it is spectacularly obvious that Thynne had recourse to texts outside the printed tradition. His text of the Squire's Tale introduces 365 readings not found in the early printed editions: the two Caxtons (CX1 and CX2), the two Pynsons (PN1 and PN2) and Wynkyn de Worde (WN). This in a poem of only 700 lines (counting the Merchant/Franklin Link)! The numbers of the lines in which Thynne's new readings occur are given below,[7] but I want to select a number of Thynne's variants which allow us to point rather clearly to one or more manuscripts, not now extant, which seem for the Squire's Tale to be very, very close, perhaps sister manuscripts, to Delamere (Dl), McCormick (Mc), and Rawlinson Poet. 141 (Ra1). McCormick and Rawlinson 141 are, of course, bracketed by Manly and Rickert very closely together as the subgroup Mc, and Manly concludes that Rawlinson 141 is very close to McCormick, ". . . quite possibly copied from it".[8] Delamere is described as being usually attached to the large group cd *. Manly (VI, 506) observes of Dl and Mc that until l. 558 they are together for the Squire's Tale. (Mc Ra1 are out for the following link.)

I shall argue further that, particularly in the latter stages of the poem, Thynne seems to have consulted one or more manuscripts from the Rawlinson Poet. 149 (Ra2) Hatton Donat (Ht) group, or its close relatives. Rawlinson 149 and Hatton are the subgroup Ra 2, and are together in the Squire's Tale according to Manly (II, 292-293).

At line 21, Thynne has the new reading trewe for sooth. This is found, according to the Corpus of Variants in the Manly-Rickert edition, in only three manuscripts, Dl, Mc and Ra1. At line 61 Thynne has the reading so royal for solempne. Only Mc and Ra1 have this reading. Thynne has at line 96 com for were come; only Mc Ra1 and Ht have this. For the Thynne reading of instead of with in line 97 corroboration is found only in Dl. In line 107 Thynne reads my commun instead of commune; only Mc and Ra1 have this reading. Thynne introduces the writhyng in line 127, a reading found only in Dl Mc Ra1. In line 139, Thynne's the for his is found only in Mc Ra1. In line 176, Thynne's new reading al born was for and born anon is found only in Mc Ra1. Thynne's new reading bokes for gestes in line 211 is found only in Mc Ra1. Thynne's new reading at 228, certes for and seyde is found only in Mc Ra1. In line 244, Thynne's eke with al for ther withalle is found only in Dl Mc Ra1. In line 257 Thynne's they cessen for cesseth is found only in Dl Mc Ra1. In line 280, Thynne's her for his is found in only four manuscripts, three of them being Dl Mc Ra1. At line 298, Thynne's new reading hit to rehersen instead of yow rehercen is found in only four manuscripts, three of them being Dl Mc Ra1. In 303 Thynne's his for a is found in four manuscripts, three of them being Dl Mc Ra1. Thynne's odd reading merthe for much in line 349 is found in only five manuscripts, two of them Mc and Ra1 (and only those two of the five having drynk in the same phrase as does Thynne). In line 375 Thynne's omission of that is matched in only two manuscripts, Mc and Ra1. In line 388 Thynne's an easye for esily a is found in only four manuscripts, Dl Mc Ra2 Ht. Thynne's construction in line 421 that they wept is found only in Ra2 Ht. In line 458 Thynne's the loue for loue is found in seven manuscripts, among them Dl Mc Ra1. Thynne's certes or for er that is found in only Mc Ra1 for line 468. In line 504 Thynne's Ther for Tho is found in two manuscripts, one of them Ra1. For line 516 Thynne prints a spurious line: With his dissymilynge and fayre assemblaunces. This is found only in Dl Mc Ra1. In line 520 Thynne's omission of bothe is matched in four manuscripts, among them Mc Ra1. Thynne's crowel for crowned in 526 is only in Mc Ra1 (and Lansdowne 851). In line 545 Thynne's eke for as is found only in Mc Ra1. Thynne's as of for of in line 546 is found only in Mc Ra1. In line 572 Thynne's ne so for ne is found only in Mc Ra1. In line 610 Thynne's omission of propre is matched in four manuscripts, among them Mc Ra1. Thynne's leue for lete is found only in Mc Ra1. In line 657 Thynne's which I of for of which I is found in only five manuscripts, among them Mc Ra1. In line 660, Thynne's of so grete for so greet is found in four manuscripts, among them Mc. Thynne reverses lines 663-664, an arrangement found only in Mc Ra1. In line 680 Thynne's perseueraunce for continuaunce is found only in Ht. Thynne's spekyng for speche in line 681 is found only in Ht. In line 693 Thynne's comyn for commune is found only in Ht and Ra2. In line 698 Thynne's your for his is found in four manuscripts, among them Ht. In line 706 Thynne's may for wol is found only in Ra2.

Among the other Thynne variants from the earlier printed editions, one can point to the following lines of interest: 28, 29, 31, 39, 42, 51, 75, 91, 96, 109,

120, 129, 154, 173, 178, 220, 226, 277, 288, 300, 311, 317, 319, 325, 359, 370, 392, 397, 402, 405, 417, 418, 438, 443, 474, 475, 483, 498, 506, 515, 539, 547, 567, 594, 616, 632. These contain variants from the printed editions which are found in a number of manuscripts, sometimes even the majority, but in each case the manuscripts containing the Thynne variants include Dl Mc Ra1 (line 300 has only Dl of this group among the other manuscripts). From line 75 onwards Ra2 Ht play an increasing role. In addition to the instances given previously, one can point to line 75, 87 (three manuscripts only, including Ra2 Ht), 120, 172, 196, 215 (four manuscripts only, two of them being Ra2 Ht), 226 (only three manuscripts, Mc Ra1 Ht), 250 (six manuscripts, Dl Mc Ra1 Ra2 Ht and Ma), 288, 359 (only six manuscripts, including Ra2 Ht), 402, 405, 412, 417, 418, 432 (Thynne's doghter this Canace is found in Dl Ra2 Ht), 438, 443, 483, 498, 506, 508, 563 (Ra2 Ht only), 567, 632, 676. The reason for some repetition of line numbers is that Dl Mc Ra1 or a combination of them and Ra2 Ht are found with the same reading.

Of Thynne's 365 variations from the previous printed editions, many of which must in any case have been mere typesetter's errors, 43, which are pretty clearly not such errors, can be traced with some confidence to a manuscript or manuscripts of the Dl Mc Ra1 type, nine to a manuscript or manuscripts of the Ra2 Ht type, and 66 others which are found in a number of manuscripts, always including Dl Mc Ra1 Ra2 Ht or a combination of the first three and the second pair.

One manuscript, Ph3, suggested by Koch as a type of the possible source for Thynne's readings in his Canterbury Tales, does appear occasionally among those manuscripts giving Thynne's new readings, for example, lines 120, 173, 438, 459, 475, 690, 701—but never alone, and not with enough consistency to make it a credible candidate for the type of source. In a number of important instances Ph3 differs from the manuscripts which provide a new Thynne reading, particularly 161, 483, 498, 567, in addition, of course, to those many instances in which Dl Mc Ra1 or Ra2 Ht are, among them, the only instances in a manuscript of a new Thynne reading.

In view of these facts, and particularly of the spurious line and the reversed lines, can we be confident that we have located the exact manuscript or manuscripts that Thynne used? Unfortunately, no. The Ra2 Ht readings, though with Dl Mc Ra1 in the first half of the poem, tend to separate into another stream. And, of course, Ra2 Ht do not share the spurious line or the reversed lines, yet they would have provided Thynne with some readings that Dl Mc Ra1 could not have done. In addition, Thynne has a number of readings not found in any manuscript or printed edition, and although some may be errors, they cannot be simply dismissed as such. Thynne's openly se in line 136 may be a typo for openly, or it may be a version of openly telle in Mc Ra1. Line 175's and eek the for and the may again be a typo. But Thynne's courser of Poyle (line 195) seems to be a reading from a manuscript; there were many versions of the Apulian courser, for it caused trouble for the scribes from the beginning. Thynne's omission of that in line 231 is perhaps a typo. His hors for cowrser in line 310 may be an error, but it looks a bit

like a manuscript reading. In line 407 Thynne's same knotte for knotte seems very much like a manuscript reading. Thynne's in wode eyther in has got the syntax the wrong way about in line 420 and may, of course, be a printer's misreading, but could as easily be a scribal error. In line 472, Thynne's spitously for pitously could easily be either, as could his omission of is in line 519. However, Thynne's in otherwise for that ootherwise seems a reflection of a scribal reading rather than a printer's mistake in line 534. And Thynne's Adam for the firste man in line 552 seems a clear instance of a no-longer extant manuscript reading. Thynne was something of a pedant, but this kind of gratuitous assistance to the reader seems to have been beyond him (another instance of Thynne's presumed editorial intrusion, the reading of client for clerk or cherl in line 140 of the Physician's Tale is less likely the result of Thynne's consulting Livy to get the facts of the story straight than it is Thynne's picking up a marginal gloss—no longer extant—containing a brief phrase from Livy). Thynne's he flye for the slye in line 672 seems to have been a simple scribal misreading, though, of course, it could have originated in the printing room as well. And in line 684 Thynne's now were fallen for right now were falle is simply the dropping of a word either by a scribe or a typesetter. Of Thynne's unique readings, then, at least four, in lines 195, 407, 534 and 552 may well have originated in one of his manuscripts, no longer extant.

There are of course many, many instances in which Thynne's text differs from those of the manuscripts Dl Mc Ra1 Ra2 Ht. It is clear that he was using a printed copy, perhaps more than one, just as he was clearly using more than one manuscript. But even in the one spectacular instance of his dependence upon a type of Dl Mc Ra1, the spurious line at 516, Thynne's version is slightly different; he prints With his dissimulyng and fayre assemblawnce for With ille dissymelynge and fayre assemblawncis, found in Dl Mc Ra1. Thynne's his for ille does not seem an obvious sort of misreading. The more probable explanation is another manuscript with this variant of the line. And, of course, Thynne did not follow his manuscripts down every rabbitpath which they presented to him: at line 592 Mc Ra1 (and presumably a sister manuscript that Thynne might have been using) have another spurious line, at least as good as the first one, which Thynne does not print. The obvious answer, that Thynne was using Dl only, won't do, for the transposition of lines at 663-4 is found only in Mc Ra1, not Dl!

And there are readings in Thynne not found in the earlier printed editions which are found in only a few other manuscripts: Barlow 20 (Bw), British Library Additional 5140 (Ad1), Bodleian 686 (Bo2), Cardigan (Cn), Devonshire (Ds), Egerton 2726 (En1), Egerton 2863 (En3), Fitzwilliam Mc-Clean 181 (Fi), Cambridge University Ii. 3. 26 (Ii), Harleian 1758 (Ha2), Harleian 7334 (Ha4), Harleian 7335 (Ha5), Lincoln 110=A. 4. 18 (Ln), Manchester English 113 (Ma), Morgan 249 (Mg), Northumberland (Nl), Royal 17 D. XV (Ry1), Trinity College Cambridge R. 3. 3. (Tc1), and Trinity College Oxford Arch. 49 (To), all figure from time to time among very small groups of manuscripts or single manuscripts which provide the same readings as are

found in Thynne. Notably, there are En1 and Ds exclusively in line 10; Bo2 Ha5 Ry1 in 63; Fi with Ra2 and Ht in 87; Cn Fi To with Ra2 Ht in 114; Nl only in 122; To only in 136; Nl only in 147; Fi Ln with Ra2 Ht in 181; Bo2 and Ry1 in 305; Ry1 with four other manuscripts including Mc Ra1 in 340; Bo2 Fi only in 454; Ra1 and Tc1 in 504; Fi and To with Mc Ra1 in 520; Ry1 with Ha4 Mc Ra1 in 610; Ry Ha5 Bw with Mc Ra1 in 657; Ry1 Ii with Mc Ma in 660; To Ln Mg with Ht and Ra2 in 676; En3 Fi Ii Ph3 Ad1 in 690; Nl with Bw Ha2 and Ht in 698; Fi Ha2 Bw with Dl in 708. Three unique manuscript readings matching those of Thynne, two in Nl and one in To, are not much to draw conclusions from.

What we know of the provenance of Dl Mc Ra1 Ra2 Ht is not very helpful in determining whether precisely these manuscripts might have been available for Thynne.[9] We do know, of course, that Urry and the Thomas brothers made use of Dl for the Urry edition.[10] It is within the bounds of possibility that Thynne actually had one of Mc Ra1 and one of Ra2 Ht and perhaps even Dl to call upon, the other eccentricities being explained by errors on the part of typesetters, corrections by Thynne, and the consultation with still further manuscripts for the Squire's Tale. But for the reasons given above I think it not proven that Thynne made extended use of these extant manuscripts. Nevertheless it seems clear to me that he made extended use for the Squire's Tale of at least two manuscripts, one very, very close to Dl Mc Ra1 and the other equally close to Ra2 Ht. And, I may add that my study of Thynne's printing of the Squire's Tale has convinced me that the description of manuscript affiliations given by Manly and Rickert (II, 288-297) for that tale are in the main correct.

That Thynne used manuscripts similar to those I have discussed seems now clear. A remaining question is: what text or texts did he use? Tyrwhitt (I, xl) suggested the second Caxton; Skeat suggested the first;[11] W. W. Greg (pp. 757-759) argued for WN; and Blodgett (p. 47) has suggested either Pynson edition—these suggestions generally for the Canterbury Tales as a whole, based upon quite select studies as of Greg (Knight's Tale) and Blodgett (Canon's Yeoman's Tale), or upon impressions as in the cases of Tyrwhitt and Skeat. So we are left with a free field, all earlier editions having been suggested by one or another prominent scholar. For the Squire's Tale, however, I believe that the first Caxton (CX1) can be safely ruled out as Thynne's copy. Although the first Caxton alone among the early printed editions gives the Squire's Tale-Merchant's Tale link and arrangement that Thynne used, I do not believe that that is enough to indicate any other than a very general use

of the first Caxton. Thynne certainly knew it, of course, but the arrangement of the Squire's Tale-Merchant's Tale could as well come from Thynne's manuscript or manuscripts: as it happens, Mc and Ra1, incomplete manuscripts, do not have the Merchant's Tale and are out for the link, but Dl has the link with the reading "Merchaunt" for "Frankeleyn" and the Merchant's Tale following, as do both Ra2 and Ht (Manly and Rickert, VI, 502, 570). But the first Caxton commits 136 errors which are corrected by the second edition; it is out for lines 554-557, and transposes lines 183-184, none of which had any effect upon Thynne's edition. Thynne prints a CX1 reading (which had been corrected by CX2) on only eleven occasions, and all but one of these readings are also found in manuscripts Dl Mc Ra1 Ra2 Ht. The CX2 in addition to its corrections of CX1 contributes 14 new variants of its own, but is an excellent text for the Squire's Tale. The first Pynson contributes 51 variants of its own, 31 of which do not appear in any other printed edition. There are 19 which appear in no manuscript or earlier printed edition, but which are corrected by PN2: 73, 177, 211, 220, 253, 262, 275, 277, 331, 334, 339, 340, 568, 601, 605, 609, 640, 650, 689. The high number of oddities in PN1, especially the readings not found in manuscripts, and the fact they did not affect Thynne, would seem to argue strongly against consideration of PN1 as one of Thynne's texts. WN adds 15 variants of its own not found in the earlier editions, including five not found in manuscripts: 216, 319, 435, 461, 487. WN is also out for line 23. PN2, on the other hand, adds only one variant of its own, at line 44, where the only manuscript support is found in Cambridge University Library Gg. 4. 27. It corrects PN1 as we have seen, and is very close to WN, but ignores 12 of WN's new variants. But where it follows WN in a variant, as in line 689, TH usually follows.

In conclusion, since Thynne introduces 365 new readings to the printed editions, it is obviously difficult to be certain of a copy for the text of a poem that runs only to 700 lines. But we have seen that CX1 is almost certainly out of consideration, as is PN1. CX2, WN and PN2 remain as candidates, my own instinct favoring PN2 with consultation with both CX2 and WN.[12] By no means all of Thynne's 365 new readings are now accepted as "correct"; but by the simple expedient of looking at several manuscripts in addition to his one or more printed texts, he consistently improved the text of the Squire's Tale, and the evidence solidifies Thynne's claim to be the first true editor of the Canterbury Tales, as opposed to being a printer of them.

Notes