 | |  |
Appendix 1
Charlotte Brewer believes that Kane's entire approach to editing A is vitiated by his having based it on two defective premises: (1) that only one variant in a given set can be authorial and (2) that an accurate description of scribal usus scribendi can be formulated by generalizing from cases where originality is not in doubt (i.e., in practice, those instances where one variant's correctness is vouched for by a sizeable majority of genetically unrelated witnesses). Concerning (1), Brewer notes the intrinsic likelihood that some of the "better" rejected variants in extant A manuscripts may reflect minute authorial revision within the A tradition. Regarding (2), she objects that Kane's method is circular since he begins by appealing to attestation but then uses the principles derived from these examples to repudiate attestation in other cases. Her second objection (discussed on p. 15 and in accompanying n. 14) seems, to me, to misconceive the relationship between data and hypothesis. As for the first, Kane was surely ill-advised to dismiss the possibility of authorial variation within the A tradition but I cannot concur that this (1) undermines his entire method or (2) necessitates a complete re-editing of A according to the principles applied in Athlone B (these two assertions cohere rhetorically but are probably incompatible logically).
Clearly, Brewer's most forceful argument for authorial variants having survived in extant A manuscripts lies in the fact that certain minority A readings are also widely supported in B and/or C. There are, then, perhaps slightly more than 300 separate instances (out of 2,441 lines) where Kane's editing of A would have benefited theoretically from collating the lectional evidence of the other two versions in the way that he and Donaldson did for B. Nevertheless, in only a comparatively small number of these cases does such a comparison reveal strong likelihood that Kane's choice of readings in A was wrong. In many, the manuscript evidence suggests instead that the author of B either incorporated scribal readings from an imperfect A copy that he was using or coincidentally revised away from his original A reading in the same direction already chosen by one or more A scribes. In such examples, therefore, minority A readings would indeed be witnessing to authorial variation, but not to authorial variation within the A tradition. In other instances, Kane's preferred explanation for all of these phenomena is probably correct, viz., that some A scribes coincidentally erred in the same way that the archetypal B and/or C scribes erred and thus the appearance of authorial variation is an illusion. In still others, isolated lectional contamination of A manuscripts by hypercorrection from B/C seems the likeliest explanation (see p. 32 ff.). Finally, some cases (especially, pairs of roughly equivalent A variants in lines unparalleled in B/C) appear to invite Brewer's hypothesis of authorial indecision (or "retouching") within the A tradition. However, positing the existence of an extensively (albeit sporadically) revised exemplar of A as the primary means to account for "good" minority variants in surviving A manuscripts appears to violate the principle of economy; after all, these variants themselves are the only real evidence that such a document ever existed and yet most of them can be accounted for on other grounds (including later authorial production in known, extant revisions of A, e.g., B and C).
Since Brewer has conceded Kane's judgment that the surviving manuscripts of A all derive from a single basic draft of that version (78), and differs from him only in suggesting that certain rejected variants may have originated from intermittent authorial "retouching" of that draft, one must ask what set of editorial results she would have preferred to Kane's? Do the manuscripts in which these rejected variants most commonly occur (viz., the EAMH3, WN, and VH families) faithfully reproduce the poet's "revised" choices? That seems indeterminable, for the occurrence of these variants follows no consistent pattern such as might imply that these manuscripts all derived from a single "retouched" document. Furthermore, the only common features to these variants are (1) that George Kane rejected them for Athlone A and (2) that many of them also appear in Bodley 851 (Z) and B. The notion of authorial variation
| | |