The Lost History of Piers Plowman: The Earliest Transmission of Langland's Work (review)

Using the lever of a simple question—"What is the date of the archetypal B manuscript?" (ix)—and the fixed point of established textual and codicological data, Lawrence Warner attempts to move, substantially, our understanding of what Piers Plowman is. That archetype (Bx), he argues, cannot have been compiled before 1390, and was therefore susceptible to contamination from the already-circulating C version of the poem. In fact, Warner believes that the exemplars of both the major B families, W-M and RF, were the products of conflation with C material created at different times. The logical conclusion of his argument is that Piers B as defined by Skeat, Kane-Donaldson, or Schmidt cannot be attributed to Langland: it is a scribal (and editorial) form of the poem (66-67).

A variety of arguments supports his conclusions. The ten extant fourteenth-century MSS that comprise the C tradition indicate circulation of C by 1390; and while the absence of early copies of Piers A is infamous, the dialectal layerings in and complex relationships among its extant witnesses imply that many of the MSS in the A family have been lost, perhaps because of heavy use in the 1370s and 1380s (4-6). However, we have no evidence that the MSS in the B family were circulating soon after its putative date of composition. Further, two MSS, L and R, from different branches from the B family contain a rubric common in and more appropriate to the C tradition: thus the C rubric must have existed before Bx was copied (xi).

Most forcefully, Warner observes that the three Athlone editions assume that the indisputably distinct manuscript shapes of the poem arose—however problematically—from authorial versions (22-24). Their presentation of textual evidence, segregated by volume, therefore obscures manuscript affiliations occurring across different versions; such affiliations could not be investigated until we had the Russell-Kane C-text in 1997. Warner has found eight or nine passages of varying length in which the two branches of B disagree with each other while consistently showing the same pattern of affiliation with other versions: at several moments in the Visio, RF agrees with A precisely when W-M agrees with a single C MS, N² (27-29). After B VII / C IX, the W-M / N² agreements continue sporadically, but again only where RF contrastingly aligns with Cx (46). In a famous textual crux in B XV, N² agrees with both parts of the B stemma against C (36-40) in numerous readings and in the absence of four lines.

Warner's explanation of these admittedly complex facts imagines N² as the descendent of an "ur-C" manuscript whose early C revisions were contained on loose sheets. Those sheets were borrowed to eke out an "ur-B" text in the W-M branch of B, but—perhaps recalled by Langland for further revision (46)—were not available to RF, whose early portions thus align with A. RF, however, had access to a later version of the C revisions for the rest of the poem. "Ur-B" must therefore have been less extensive than the B-text we read today (46). Much less extensive: Warner believes it ended after XVIII (52), and the hypothesis that both versions conclude with material composed specifically for C, added by scribes "jealous for the completeness of their copies" (51-52), explains the famous similarity of B XIX-XX to C XXI-XXII. Instances of RF-Cx agreement against W-M, far more common than W-M agreement with Cx, support that view (54-55): Warner again understands that an early draft of those passus was copied into W-M, the final version to RF.

Many readers will be skeptical: "This is, I acknowledge, a difficult narrative", Warner writes, countering that it is "not nearly as difficult as the...