Textual Editing in Early Printed Editions of Chaucer's Parliament

One consequence of the invention of printing was to help to yield the idea of 'standard text' for a literary work. How did, then, early printers/editors make a new edition once a text set in type appeared in multiple copies? To what extent did they still refer to manuscripts? In response to the enquiry made by Dr. Lotte Hellinga, this short article is going to examine the case of The Parliament of Fowls.

The Canterbury Tales provides a good illustration of the general development of editorial attitude in early book production. Caxton prepared the text from a single manuscript in his 1477 first edition (Cx1). The 'fixed' text of Cx1 was partially collated and amended in his 1483 second edition (Cx2) against a 'trewe' manuscript which a customer offered, claiming Cx1 was a corrupt edition. The corrected text was now used in two different directions: Cx2, in the capacity of a standard text, was faithfully reproduced in Pynson's ?1492 and 1526 editions, whereas Cx2 was extensively collated against a manuscript and a revised edition of Cx2 with some tales substituted by the manuscript versions was put out in 1498 by de Worde, who, as a helpmate in Caxton's print shop, may have heard of and have been concerned about the problem of 'correct text' in the Canterbury raised by a reader. In all these publications a printed edition served for the copy- or the base-text to its later edition, and, in the case of the base-text, it was revised (in varying degree according to the editor's textual respect) on collation with a manuscript source. But such editorial attitude with textual priority on a printed edition was counteracted by a fresh concept of 'authorial text' which was displayed in Thynne's 1532 edition: he put available sources on the same ground and collated them to reconstruct the author's intention.

By and large, the major poems Caxton printed for Chaucer were to undergo this pattern of editing. But the Parliament was an exception. The poem appeared first in Caxton's edition in 1477 (Cx), followed by Rastell in ?1525 (Rs), by Pynson in ?1526 (Pn), by de Worde in 1530 (dW), and by Thynne in 1532 (Th). It is unlikely, however, for the text established by Caxton to have been used as the exemplar for the three subsequent editions. It was Thynne that took up Cx along with some manuscript sources and, after collating them, depended on the printed text for his first 140 lines and line 524.

There are fourteen manuscripts extant for the poem, and the Riverside Chaucer postulates their textual affiliation in this way:

Group A: a Gg, Ff Group B: d F, B

b H, R, S, Hh, (Cx) e T, Lt, D.

g P, J, L

A textual examination of Rs, Pn, and dW against this chart indicates the nature of editing in their respective edition. Rastell's edition survives only in the text of two leaves (a1 (prologue) and a6 (ll. 225-80)), and as far as the surviving portion is concerned, Rs may be a modern version edited, actively and extensively at the editor's own discretion, on a manuscript source characteristic of A-group.

Pynson's text, which was published in one book together with the House of Fame and some Chaucerian minor poems, is interesting in the sense that it was a composite text edited out of two sources from different lines of textual transmission, one (B-group) as a base-text and the other (A-group) as a reference text. The primary source can be considered to have had a closer relation to the subgroup d in B-group, especially to the manuscript B (Bodley 638), because Pn agrees textually with all of the ten readings unique to B (lines 27, 63, 140 206, 313, 335, 394, 395, 637, and 644). Pynson may have prepared his text mainly from the base manuscript and at times have referred to the secondary source when he came across corrupt readings. The following variants will illustrate the way the editor went about establishing the text (Pynson's reading is cited as the lemma, followed by the readings of the other texts):

141 These verses of golde & asure / ywritten were

-2 Of whiche I gan / astonyed to beholde

astonyed] a stounde FBTCxdW, stonde LtD, so Others Th

There is a textual corruption involved with 'astonyed': the manuscripts F, B and T provide 'a stounde,' a decayed form of 'astonyed,' and the rest of B-group (i.e. Lt and D) use 'stonde,' a verb which was meant to have improved the corrupt reading. A likely explanation is that Pynson noticed the awkward phrase 'a stounde' in his base text and corrected it by incorporating 'astonyed' from another source of A-group.

A more decisive case to attest Pynson's use of multiple sources:

237 And on the temple / sawe I whyte and faire

-8 Of douues sitting / many a thousande payre

And . . . payre] so Th,

And on the temple saugh I white and faire

Of dowues whyte many a hundred paire

B-group (whyte] om D) dW,

And on the temple of dowis white & fayre

Saw I syttynge manye a thousand payre

A-group (thousand] hunderede Gg Ff)

And vppon the temple of douuys fayre

Sawe I syteyng many a thowsand payre Rs

The text in these lines of B-group is obviously awkward in its duplicated use of 'whyte,' and Pn has two textual variants from B-group: 'sitting' for 'whyte' and 'thousande' for 'hundred.' But the fact that both of Pynson's words can be found in the corresponding lines of A-group seems to point out the possibility of the editor's attempting, after referring to a second text in this corrupt line, to improve it by substituting 'sitting' for the second 'whyte.'

It is not certain whether the secondary source for Pn was a manuscript or a printed edition, but Rastell's extant text suggests that the printed edition is one of the strong candidates for it. As the second case quoted above shows, Rs, which is classified as A-group, can provide Pynson with 'sitting' and 'thousande,' and moreover Rs shares with Pn a curious description of the personified Beauty:

225 Than sawe I Beaute / with a nyce atyre

with a nyce atyre] so Rs Th, without ony atyre Others dW

If the printed text, Rs, was employed as an exemplar here, it stood merely as an additional reference text to complement a primary manuscript source which was posited as a 'standard' text.

The fourth printed edition, de Worde's version, also derives its text from a manuscript source. Copland's epilogue to the edition implies that the text was made from an imperfect mildewed manuscript, and this can be confirmed, though indirectly, by proving that the main text is not based on any previous printed edition. Like Pn, de Worde's text has readings characteristic of the subgroup d in B-group, especially the manuscript B (Bodley 638), and their respective source can have been textually so close to each other as to be regarded as one and the same manuscript. And like Pn again, dW seems not to have been edited from a single manuscript. The imperfect or illegible portion can have been restored from the text of Pynson's edition: Pn had eighty-one unique variants at the time of its publication, and of all these variants, dW shared twenty-three with Pn.

Every new edition of the Parliament was edited from a manuscript, and a printed edition was, if any, merely employed as a subsidiary text. This may be an unusual case of text-editing in early book production; rather, as is shown in Pynson's practice to make a better, composite text out of two sources, it should be interpreted as a forerunner to Thynne's editorial principle to restore the 'authorial text.'

Naruto University of Education Tsuyoshi Mukai

Notes (omitted henceforwards)