The Properties of a Stemma: Relating the Manuscripts in Two Texts from The Canterbury Tales

The Properties of a Stemma: Relating the Manuscripts in Two Texts from The Canterbury Tales Alex Jones Time was, no study of a premodern text was complete without a stemma of its manuscripts. The distribution of variant readings was used to reason out the logically necessary relations between the manuscripts, and these could be expressed as a family tree in which the manuscripts were nodes and the copyings were branches, or as a bracketing which distinguished closer from more distant relations. The stemma once established could then be used (in a rather circular fashion) to guide the editor in the choice of readings; the stemma was logically entailed by the distribution of variants, thefinaltext by the stemma. The reason this method worked well for many classical texts (and it did work well) was that there were typically few manuscripts, or at least few lines of authority, to choose from. W h e n this was not the case, as so often with medieval texts it is not, the simplicities broke down. The approach taken by George Kane in his edition of the A text of Piers Plowman^ is typical of that of modern editors: after a lengthy review of the evidence for manuscript relations he elects to follow a single major manuscript except in the few cases where its reading is obviously deficient. The alternative for an editor is to rely on scholarly judgement and weigh each reading on its merits variant by variant, but perhaps 1 George Kane, Piers Plowman: the A version (London: The Athlone Press, 1960 36 Alex Jones few m o d e m editors wear their scholarship with the assurance ofA. E. Housman in his classic edition of Lucan.2 And yet there are questions that only a stemma can answer. While it may not go as far as providing authority for particular variants it will enable an editor to find the best manuscript, and to distinguish features that arise far from the archetype and are certainly not authorial; for any text that shows substantive variation, this is important information to have. For such reasons, the phylogeny of The Canterbury Tales, to borrow the title of a recent paper,3 must be of more than passing interest. While Chaucer's vast unfinished work is preserved in more than 80 manuscripts and fragments, no two of them are identical in text, and in the ordering of the tales themselves there are many groupings and subgroupings. The dominance of certain texts at different times - Caxton's in the early modem period, Skeat's for a large part of the twentieth century - has obscured the sometimes wild variety that lies behind the printed word to make a 'Chaucer' available to the unreflecting reader in every age; but scholars have never ceased to reflect on questions such as whether particular readings are authorial revisions or scribal editing, or what Chaucer's plan or plans for the work may have been. A n authoritative account of how the manuscripts of any tale are related to one another could at worst set some limits to what is worth talking about and at best might suggest further directions for enquiry. But of course there is such an account. Manly and Rickert in their magisterial collection of textual variants4 (hereafter M & R ) distinguish a number of groups of manuscripts whose text is consistently so similar that they clearly derive in each case from the one exemplar. Besides groups of two and three manuscripts, the large groups that M & R call a, b, c and d, accounting among them for up to 25 manuscripts, persist throughout The Canterbury Tales. 'Errors undoubtedly remain in our work', concede M & R , 'but w e hope that there are not many, and that none of major importance has escaped notice'.5 In view of the volume of work that led them to such consistent conclusions it would be 2 A. E. Housman (ed.), M. Annaei Lucani Belli Ciuilis Libri Decern (Oxford: B Blackwell, 1926). 3 Adrian C. Barbrook, Norman Blake and Peter Robinson, 'The phylogeny of The Canterbury Tales', Nature, 394 (27 August 1998), p. 839. 4 John M. Manly and Edith Rickert, The Text of the Canterbury Tales (Chicago and...