the whole of Passus XIX and the beginning of XX, but at the end of XX R has ‘Passus iius de do best’. Remarkably, L has the same heading (but in its guide only) also at the end of the poem, suggesting perhaps expectation of a further, twenty-first passus. Agreement between L and R here may imply that this anomalous heading was present in the B archetype.”3 Burrow is pulling together various strands of discussion from the previous two decades-plus: MS R’s rubric came to light in an essay published in 1985 by Robert Adams;4 L’s in David Benson and Lynne Blanchfield’s 1997 catalogue of the B manuscripts;5 and the import of LR agreements (though of course voiced in ignorance of this instance) in another Adams study, of 2000: “When L agrees with R, even in the face of massive dissenting evidence from the other witnesses, the odds are always quite good that what R and L are mirroring is Bx (albeit the readings of that prototypical copy can never be simply equated with authorial text).”6
But Burrow omits to mention a crucial piece of information that would have answered the question of how this bizarre rubric got into Bx in the first place: “Almost certainly it was laterally imported, perhaps into an ancestor of R,” wrote Adams, “from a C manuscript (where an identical rubric occurs in the XU family at the beginning of the final passus and a very similar one at the end of the same passus in the majority of all C manuscripts).”7 In light of his later work and the newly apparent presence of MS L in the picture, we need now to substitute “into Bx” for Adams’s “perhaps into an ancestor of R.” The B archetype, in sum, was contaminated by the C tradition. Any other explanation for LR’s agreement in this error would have an extraordinarily difficult row to hoe.
This conclusion incidentally undermines the argument in whose service Burrow had mentioned it, for if the earliest possible stage of the B-tradition rubrics took shape under the influence of the C tradition, and was erroneous at that, it becomes quite difficult to argue for their authorial nature except on purely literary grounds (and in fact only Burrow and, more recently, Hanna have attempted to do so).8 Its implications, though, extend far beyond such arcane topics as the authority of the rubrics. For one, the only evidence for the B tradition’s existence before the later 1380s or early ’90s inheres in the indications of the existence of at most two copies, the C reviser’s B manuscript and possibly the document to which a scribe in MS F’s line of transmission had access.9 And it also means that Bx was the product of an era in which readers of Piers Plowman began to become “jealous for the completeness of their copies,” in George Kane’s words.10 This should immediately prompt us to wonder, for instance, whether this mistaken rubric was the only item that came into B from the C tradition: might, say, the entire final two passus have been part of this contamination? And to what extent should we assent to the editorial assumption that this could not have occurred?
The Lost History of “Piers Plowman” follows up on these questions, answering with a resounding “yes” and “not at all” respectively. It has two major aims: to lay out the evidence leading to this conclusion, which is far more extensive than what is found in LR’s final rubric, and to show that the previous silence about that possibility is not an index to its plausibility, but, rather, the product of the assumption that B was integral. That the only information regarding the LR situation offered by the Kane-Donaldson and Schmidt editions is that MS R has a distinctive rubric on its final folio—nothing about what the rubric says, or L’s role—shows that we need to go beyond their analyses and methodologies as we assess the earliest production and transmission of the poem.11 Chapter 1 examines in some detail the claims made in the service of the beliefs that B was widely available by c. 1380 and its converse, that Piers Plowman A did not achieve substantial circulation until well after its own composition, perhaps even until some decades into the fifteenth century. Some of these popular claims rely on a few codicological indicators, and others, on assumptions concerning supposed allusions to B around 1380. The chapter mines a body of evidence all but neglected in such discussions, that of the textual affiliations, which leads to the opposite conclusion: the A version was widely available, while its immediate successor, B, might not even have gone beyond the poet’s immediate circle till the 1390s or so.
Chapter 2 opens by querying the assumptions, first, that each of the three major textual traditions of the poem was at its origin integral and unaffected by the others, and second, that any indications that suggest otherwise are the result of convergent variation, the means by which unrelated manuscripts attest unauthorial variants not sourced from a mutually exclusive ancestor. Only slightly less powerful is the conviction that conflation—which is what I will argue was a major force in the production of our received B version—always manifests itself in scribal officiousness, such as inconsequence, bad sense, or repetition. None of these is well founded, as the pattern of affiliations summarized above reveals. This chapter examines the means by which MS N2’s “B” readings have been exiled from the B edition—or, to be more accurate, how W~M’s “C” ones have been kept from the C edition.
The primary mode of this “contamination,” Chapter 3 shows, was via the movement of passages on sheets of loose revision material that could go easily from Langland’s C papers to Bx as copied by the W~M subarchetypal scribe. The converse notion, that such loose papers went in the opposite direction, stumbles for a number of reasons, not least the fact that there is no evidence (unless this pattern is the sole exception) that the B continuation comprised such materials while the C additions, as E. Talbot Donaldson said, “were probably written on separate sheets and their position in the text indicated by one of those complicated systems of arrows and carets that every reviser finds himself adopting.”12 This chapter demonstrates the point via a focus upon the fortunes of a passage whose textual state serves as a litmus test for any belief in such an integral “B version”: the forty lines in which Langland inveighs against “the poison of possession” and calls for clerical disendowment (received B 15.533–69). (All references to passus, line numbers, and sigils are to the Athlone editions, whose conventions I apply silently to those quotations and references to other schemes—e.g., different line numbers in Schmidt or passus numbers in Skeat’s; see further below on my policies. I do not reproduce editorial brackets.13) It is already well known that these lines are misplaced in W~M and absent from RF. What no one has known before is that N2 attests them in precisely the “B” form, and the only viable solution to this mess is ur-C > Bx contamination.
Chapter 4 argues, primarily on the basis of textual affiliations—this time not so much N2/W~M as that pattern’s converse, RF/C—that the final two passus were among the C-tradition materials that intruded into B. Again, the assumption that the archetypal texts, Bx and Cx, could be established without regard to the evidence of the other tradition has hampered recognition that, whatever the reason, something very odd is going on in these passus. There is no way that accident could explain the extent to which RF and C agree in this portion of the poem alone. To date, though, no one has been able even to suggest a solution to the problem because our ways of thinking about Piers Plowman have prevented recognition of the problem in the first place. If my own solution—that these passus were not in any “B-version” manuscripts that might have circulated prior to the appearance of C—is viable, then some of the central objects of Middle English studies, ranging from the character of the C revision to the beginnings of “lollardy,” look much different from what we have assumed.
The patterns of affiliations that give rise to this argument confirm the interpretation of the LR “passus iius de Dobest” rubric given above, which already would be very difficult to explain otherwise. Yet this idea that Bx was contaminated by the C tradition, while it might seem surprising or even iconoclastic, has in fact already made its mark upon the received B version. For the Athlone editors themselves assumed Bx’s contamination by C, though they buried their statement to that effect in a subordinate clause, in support of
