especially MS H2. These affiliations are most probably explained, as Carl Grindley has shown, by positing an exclusive ancestor (18) of 6 and 7, with MS T’s correct readings against H2BmBoCot representing scribal improvements to a faulty exemplar’s readings.18 And Russell and Kane characterize the Ilchester MS’s A frame text, copied from 5, as “ordinary: in about twenty departures from the adopted A text it enters more than a dozen different agreements.”19 But this is misleading, for its mere fifty-two lines share six errors with the mid-fifteenth-century MS W alone, and another three with W and one other witness or group.20 This is about the densest rate of error in the surviving manuscript record.21 Unless W descends directly from the very exemplar behind MS J, and unless that exemplar was a direct copy of Ax, by the 1390s at least one now-lost generator of the family to which MSS W and J belong, number (19), must have been extant.
So wild are the A-text affiliations that nearly every member tends to follow the pattern represented so far by MS V. Hanna, contradicting his earlier stern judgment, has pointed out that the “impenetrable dialectal mixtures” of the A manuscripts constitute a state of affairs that “suggests that many manuscripts are the surviving product of several generations of copyings in diverse locales.”22 And while these “generations” are stages of copying in lines of transmission rather than fixed periods of time, it takes time for manuscripts to be copied, travel, and find new scribes and audiences. All of which is to say that the figure of nineteen now-lost A manuscripts is the result of the most efficient explanation of textual affiliations; only once, in the case of MS V, did we even begin to make use of the dialectal data here cited by Hanna. The wildness of the A tradition’s textual record alone, not to mention its impenetrable dialectal mixtures, should have put the restraints upon all the rhetoric about its supposedly late circulation.
It remains unclear how long before 1400 or so the A tradition achieved wide circulation. MS T’s distance from Ax—some six generations of copying—does not necessitate, even if it might imply, a very lengthy period of time. The earliest surviving manuscripts with A material are probably Vernon and Ilchester, but they might be only three or so generations from the archetype. It is clear, though, that since 1960 plenty of evidence for Piers Plowman A’s substantial fourteenth-century readership has been available. Readers have on the whole ignored this remarkable body of data, focusing their energies and anxieties instead on the established Athlone text. But it enables a much more thorough mapping of the shady terrain in which Piers Plowman first circulated than do the dates of the extant witnesses.
Lost: Piers Plowman B
The corollary to the belief that A had no public life in the fourteenth century is the assumption that B, by contrast, “quickly achieved something like a canonical status,” as Robert Adams has put it.23 Yet the earliest extant B manuscripts, like those of A, date to the 1390s.24 More problematic is that the B tradition’s “tightly bifid stemma” below, which Adams produced on the basis of work done by the editors of the Piers Plowman Electronic Archive, is suggestive, as he says, of “a tradition that never consisted of many manuscripts.”25 A total of eight now-lost manuscripts (in bold: Bx, two in the α family, five in β) are necessary to explain the relationships among the survivors; see chart.26
All eight lost B-tradition manuscripts, and at most about five extant ones, are likely or possible products of the 1390s. Given the tenor of most discussions of the early transmission of Piers Plowman, the fact that the number of pre-1400 manuscripts necessitated by the A tradition’s affiliations is about 50 percent greater than this figure might seem surprising in itself. But most striking here are how well defined the B tradition is (quite opposed to A), and how late it came into being. Two of the extant fourteenth-century productions, plus one dated to the beginning of the fifteenth century (B.L. MS Add. 35287 [M]), are only two generations of copying removed from the archetype.
The notion that B achieved something like a canonical status between its postulated date of completion (c. 1377–78) and, say, the Rising of 1381, then, is an assumption rather than a postulate based on textual evidence. Schmidt is the only critic to have explained his belief that Bx was a product of the late 1370s in positive terms, and he carefully presents the case as a matter of probabilities and presumptions: “it is likely to be the longer version that the peasant rebel leaders alluded to in 1381,” he claims, on which basis he later says that B’s “archetypal manuscript was presumably generated only a couple of years before the Rising [of 1381].”27 The substance of this proposal, Ball’s presumed knowledge of B, will occupy us below; but the fact remains that the only pre-archetypal manuscripts inferable on textual grounds are the C reviser’s B manuscript and the copy to which the F or α1 scribe had access, which the Athlone account necessitates. And some critics have cast doubt upon the existence of even these two, particularly with regard to F.28 Robert Adams has interpreted the “multilayered complexity of dialects” in the B tradition as “indicative of a wider circulation (and more extensive recopying) than that achieved by the C version,” but in fact there is no compelling reason to attribute these layers to any more than the three-plus post-archetypal generations represented above.29 Anyone eager to use this evidence to bolster the count of early B copies, if playing fair, will need to follow suit with regard to the “impenetrable dialectal mixtures” of the A manuscripts.30
In sum, if the messy state of affairs in which the A version survives looks exactly like the result of a tradition whose ancestors had disappeared, the cleanness of B’s makes it look as though it never had such ancestors.31 This cleanness, and not the absence of early A copies, presents the most pressing question about the earliest transmission of A and B. Almost fifty years ago in Sydney, George Russell spoke to the heart of the matter: “Why is it,” he wondered,
that that version of the poem which almost all modern readers and critics agree in judging to be the most impressive, was the version whose lines of survival were most tenuous; which, in fact, seems to have escaped extinction only by the chance survival of a single manuscript? It may be that this was a mere matter of accident: but the facts of the A- and C-descent suggest that this is unlikely. If the short and essentially incomplete A-text finds readers and copyists in numbers from the beginning, and if the same is also true of C-, why then should B- not find them?32
By way of explanation Russell proposed that B, “for various reasons, mostly political, religious and ideological, was either called in by the author or was suppressed by others,” of which destructive program Bx was a chance survivor.33 But what mechanisms could have enabled so thorough a suppression? If this version was circulating widely, how did Langland or these other censors know how many copies were extant, where to find them, and how to prevent further copying?34 A more efficient explanation, if one less conducive to most narratives of the history of Piers Plowman, is that Langland’s own B version achieved absolutely minimal circulation, if any at all, before c. 1395, and not very much between then and 1550 either.
The Early Proliferation of the C Version
The most likely scenario is that readers’ quick embrace of Piers Plowman C led to the release of the dormant B version.35 This claim follows from the fact that C’s transmission history differs so starkly from those of A and B. Far more of its manuscripts predate 1400, and four manuscripts from Schmidt’s collective sigil y (XYHJ) have excited particularly strong interest. These texts exhibit both dialectal features very close to those presumed to be Langland’s, which many critics have taken as evidence of his whereabouts when he composed C, and codicological and paleographical characteristics very similar to those of the London-based productions associated with Gower and Chaucer.36 In two of these, San Marino, Huntington Library MS Hm 143 (X) and the Ilchester manuscript (J), critics have found evidence of “some direct connection to the author.”37 Such judgments, however, have not taken account of these surviving manuscripts’ positions within the C stemma. Ten surviving witnesses—three of these four (XHJ), plus MSS UTPEVMK—predate 1400, as do some fourteen more now-lost copies in Schmidt’s chart.38 MS X alone, like MS T in the A tradition,