"Learning, Taste and Judgment" in the Editorial Process: Vance Ramsey and Manly-Rickert

The title of Roy Vance Ramsey's book raises expectations that it will help to make sense of Manly and Rickert's daunting edition of the Canterbury Tales. To some extent, these expectations are fulfilled. Chapter 2 provides a fascinating and illuminating history of the Manly-Rickert project: much pathos is (justifiably) wrung from the account of the practical and financial difficulties under which the two scholars worked, culminating in Rickert's death a year before the publication of the edition and Manly's death a few months after it. It also helps to know that volumes 3 and 4, containing the text of the Canterbury Tales, were completed first, under pressure from the university administration to show some results. Ramsey summarizes the "fundamental discoveries" that Manly and Rickert made: "(1) the greater closeness of Hg to the original scribal copies of most of the links and tales than El or any other extant manuscript; (2) the fact that apart from the constant groups the manuscripts tend to shift their textual affiliations from one locus to another; and above all (3) the Stage 1 piecemeal disseminations of links and tale beginning in the midst of Chaucer's composition and revision of them and the consequent Stage 2 gathering of the links and tales to serve as exemplars for making full manuscripts of the Canterbury Tales."¹ In addition, Ramsey stresses the importance of the detailed descriptions of the Canterbury Tales manuscripts provided in volume 1, which pioneered the serious study of vernacular manuscripts.

As Ramsey's (2) suggests, the most intimidating part of the Manly-Rickert edition is the Classification of Manuscripts in volume 2, which presents a bewildering picture of shifting manuscript affiliation not only between the individual tales but also within them. The genealogical method, instead of presenting a tidy stemma that would indicate the relative authority of individual manuscripts by placing them on a line of transmission, seems instead to have produced a jostling throng in which significant relationships are hard to make out. In such circumstances, the principles on which the editors judge one reading rather than another as the more likely to be original are unclear (and the scanty nature of Manly-Rickert's textual commentary does not help). For this reason, they have in the past been accused of "counting heads" (that is, giving weight to the number of manuscripts that preserve a reading), rather than classifying manuscript relationships by means of agreement in error. It is here above all that the Chaucer scholar looks for aid from Ramsey's book, and to a certain extent receives it. Ramsey explains these shifts in affiliation not by contamination (ctm), which Manly and Rickert slowly discarded as an explanation, but as the surviving traces of the Stage 1 circulation of the tales, which resulted in strange amalgamations of material as scribes and would-be owners cobbled together tales and links (each with their own sets of scribal errors) that Chaucer was still revising. This confusion is (allegedly) the justification for Manly and Rickert's classification of manuscripts by shared readings (variants), despite the fact that many of these readings will undeniably have arisen by accident (Greg's "convergent variation," Manly-Rickert's "accidental convergence" or "acco"), because of the tendency for scribes to make the same kinds of error independently of one another. On pages 195-204, Ramsey takes the reader through a discussion of unique variants (those shared by no other manuscript), genetic variants (those that are transmitted by one manuscript to its descendants), and convergent variants (those that recur in different manuscripts by chance), and concludes that because of the last-named, only "the laws of probability" can reveal manuscript affiliations (though not right readings). For this reason, Ramsey has constant recourse to Charles Moorman's statistical analyses of the affiliations of Canterbury Tales manuscripts.² Using his own chart of variants in The Nun's Priest's Tale (199) as an example, Ramsey shows how Manly and Rickert established the existence of "constant groups" among the manuscripts, and why they suggested four...