Читать книгу The Lost History of "Piers Plowman" - Lawrence Warner - Страница 10

Chapter 2

Scribal Conflation, Convergent Variation, and the Invention of Piers Plowman B

The archetype of all surviving B copies, as we saw in the Preface, concludes with a rubric expecting a further, twenty-first passus: an odd error resulting from contamination by a C manuscript, in which the rubric worked perfectly well as an explicit. These few pen-strokes in manuscripts L and R thus have enormous consequences in that they identify the moment of Bx’s production as the era of conflation and conjoinment. The critical propensity for focusing only on authorial texts has led to an almost exclusive focus on the three versions, dated to the 1360s, 1370s, and 1380s respectively, but the landscape of Piers Plowman manuscripts in the final decade of the century was no respecter of that approach. This was when the desire for “completion” took hold: the energies that went into the compilations now found in the Ilchester Prologue, the BmBoCot group, and the TH²Ch group suggest that conflation and conjoinment might have been the normative modes of producing and reading the poem by this point in its history. Those who got their hands on a C version copy were set; the rest, though, sought out matter from here and there with the aim of compiling a complete Piers Plowman.

Might Bx itself have been the result of this desire for completion? If its very final inscription came from C, could the previous 850 lines, which are nearly identical in the surviving B and C traditions, have as well? And if that possibility is viable, as Chapter 4 will argue in detail, should we perhaps consider the idea that such conflation occurred in other locations? Say, those passages extant in only one of the B families? Over the 150-year editorial history of the poem, such ideas have never been mooted. As a result, many today might dismiss such speculations out of hand. Conflated texts always reveal their nature via signs of scribal officiousness, such skeptics might aver, citing George Kane’s thorough analysis of manuscripts W, N, and K. And more to the point, no potential sources of conflation, other than those final two passus, present themselves. Yet such rejoinders are either just wrong, or the product rather than the foundation of critical approaches to the textual state of Piers Plowman. A three-pronged attack, comprising lemmatization, versioning, and the invocation of convergent variation at any signs of rupture to the existing paradigm, has buried the evidence that shows, contrary to that paradigm, that Bx was indeed affected by the C tradition to a far greater degree than the LR rubric indicates. In conclusion this chapter presents the bulk of my positive evidence that Bx, at one stage in its career as exemplar, was supplemented by lines and paragraphs of new C material, the existence of which has been obscured by its atomization into versions and then lemmas. In sum, an early scribe, like so many others between the Era of Ilchester in the 1390s and Sir Adrian Fortescue in 1532,¹ was jealous for the completeness of his copy, and thus turned to C materials to fill out the seemingly deficient manuscript in front of him.

Officious Scribes, Inconsequential Transitions?

George Kane’s chapter defining the character of Piers Plowman A, titled “Manuscripts and Versions,” has never featured very prominently in Langland scholarship. C. L. Wrenn’s early review, in effect saying that we knew all this before but there was no harm in Kane’s rehashing of the material, is representative: “Though little that is positively new here emerges, there is real value in this thorough re-examination of the reasons which lead to the generally accepted conclusions.” The “most important part of a very important Introduction,” he continued, lay farther afield, in the following chapter on the classification of the manuscripts.² Yet insofar as it determined which manuscripts and which lections of those manuscripts would and would not be classified in any given volume of the Athlone edition, “Manuscripts and Versions” established the framework within which the Athlone enterprise, and all Piers Plowman criticism in its wake, would operate. In sum, it puts forth the idea that, once the obvious instances of scribal conflation among the versions are identified, the editor is left with the authorial versions: here, “Piers Plowman A,” but by implication the other two versions, B and C, as well.

The local problem is the inclusion in A-version witnesses of additional lines or passages corresponding to matter from the B and C versions. Kane argues “that these seven manuscripts received this B or C material, and their present shapes, through scribal compilation, and that they are actually contaminated and augmented copies of the A version.”³ The main offenders are the Duke of Westminster’s manuscript (W; 95 lines at twenty-eight points), National Library of Wales MS 733B (N; about 150 lines at twenty-one points, many shared with W), and Adrian Fortescue’s manuscript, Bodleian, MS Digby 145 (K; over 400 lines at thirteen points). Kane knows they are contaminated because they all bear the marks of “the officious copyist,” who “tended to leave signs of his activity in the character of the text produced by his conflation.” These signs “are happily unambiguous,” revealing themselves “in the presence of both primitive and revised forms of the same passage; inappropriate or misplaced augmentation; in broken syntax; and in broken or inferior sense.”⁴

Kane’s first exhibit, where an irritated Holy Church identifies Meed in WN, provides a good snapshot of both the phenomenon and Kane’s approach to it:

In þe popis palis she is pryvei as my selfe;

And so scholde sche nouȝt be for wrong was hir syre;

Out of wrong sche wer to wroþer haile many.

Talis pater talis filia.

For shal nevre brere bere bery as a vyne

Ne on a croked þorne kynde fygge waxe.

Arbor bona bonum fructum facit.

I ouȝt ben herre þan she for I come of a better.⁵

Kane identifies this as A 2.18–20, C 2.27α–29α (here in bold), A 2.21, and observes that in the C version A 2.20 “has been replaced by the point which C 2.28–29 are intended to drive home. Without this purpose, in WN, these lines are in the air. They can have been inserted only in ignorance of the local differences of meaning between the versions.”⁶ Since no poet could be ignorant of these differences, enter the officious scribe.

The chapter’s insistent and antiscribal tone and close attention to example after example might lead some readers to conclude that conflation is always marked by signs of officiousness. But in fact Kane discusses only about one-third of the conflated passages, enough to imply that all of them represent scribal intrusions from other versions.⁷ Many others are perfectly smooth. For example, Kane cites nine of MS N’s twenty-one sites of intrusion as officious. Of the remainder, five consist simply of the Latin found in the C version, and two are single lines in the list of companions who enter the pub during Gluttony’s confession (added as well by MSS VH(EA)MH³); none of these affects syntax.⁸ That leaves five sites of intrusion, totaling over sixty lines, about which Kane says nothing. All of them consist of English additions integrated seamlessly into received A.

Immediately after Holy Church’s enhanced identification of Meed in MS N, that document alone smoothly includes C 2.31–41α at the juncture of A 2.21–22. On one end line 21 is equivalent to C 2.30—indeed it would be just as accurate to say that N is simply continuing the conflation begun with line 27α⁹—and on the other, both A 2.22 and C 2.42 begin with the phrase “Tomorewe worþ.” Equally straightforward is the first conflation common to W and N:

“þu doted daff!” quod sche, “dul are þi wittis.

For litel lerestow I leve of latyn in þi ȝowþe:

Heu mihi quia sterilem duxi vitam Juvenilem.

It is a kynde knowyng þat kenneþ in þin herte …”¹⁰

Kane identifies this as A 1.129 and 130 into which are inserted (in bold) B 1.139–39α or C 1.140–40α. If these four lines were excerpted, though, they would simply be called either B 1.138–40 or C 1.139–41. Next, since lines 75 and 76 of A passus 3, “Ne bouhte none burgages, be ȝe ful certayn. / Ac Mede þe mayde þe mayre a bisowte,” are equivalent to C 3.85 and 115 respectively, it was very simple for N, or the text in its line of transmission that initiated the conflation, to add that version’s new intervening lines, 86–114. By the same token WN’s addition of C 2.246–51 at the juncture of A 2.194–95 (= C 2.245, 252), too, is seamless. Only a very willful scribe could have left signs of officiousness at any of these sites, given that Langland himself simply added new material at these junctures, which themselves survived the process of revision intact.

The most interesting such addition to MS N, where the friar-confessor asks Meed to engrave her name upon a window in the friary, shows that even where conflation is not a matter of simple addition, it could still be accomplished with no signs of officiousness. Here is the received A version of the episode, with the lines in question—to be replaced rather than added to in N—in bold:

“We han a wyndowe awurchynge wol stande us wel heye;

Wolde thow glase þe gable and grave ther thy name,

Sykir sholde thy soule be hevene to have.”

“Wiste I that,” quod þe womman, “þere nis wyndowe ne auter

That y ne sholde make or mende, & my name writen

That uch segg shal se y am suster of ȝoure hous.”

Ac god alle good folk suche gravynge defendeth

And saith Nesciat sinistra quid faciat dextera. (A 3.47–54)

The passage appears in National Library of Wales 733B thus:

“We have a wyndowe iwrouȝt stant us wel hiegh;

Woldestow glase þe gable & grave þere þi name,

Siker scholde þi soule be hevene to have.”

“Wist I þat,” quod þat womman, “I wolde nouȝt spare

For to be ȝour frende, frere, & faile ȝow nevre

Wil ȝe love lordis þat lecherie haunteþ

And lakkeþ nouȝt ladies þat loveþ wel þe same.

It is a freelte of flesche—ȝe fynde it in bokys—

And a course of kynde whereof we comyn alle;

Who may scape þe sklaunder, þe skaþe is sone amendid;

Hit is synne of þe sevene sonnest relest.

Have mercy,” quod Mede, “of men þat it haunte

And I schal kevre ȝour kirke, ȝour closter do make,

Wowes do whiten & wyndowes glasen,

Do pointin & purtraye & pay for þe makyng

þat evry segge schal seen þat I am sistre of ȝour hous.”

But god to alle good folke suche gravyng defendeþ:

Nesciat sinister quid faciat dexter.¹¹

This instance appears to be N’s sole conflation from the B tradition, as this is equivalent to Kane and Donaldson’s B 3.51–63 (though only one of the two B families attests it, the other instead having a spurious version of lines that look like the ones N replaces, A 3.50–52). As such this will become very prominent later in the chapter, but for now the point is that N, whatever the source of his conflation, does not fall prey to the temptation of including both primitive and revised forms of the same passage. Since both the a-verse at the site of substitution and the entire line with which the passage ends are identical in both texts, the sin was not too difficult to avoid, but many readers would come away from Kane’s chapter assuming that scribes left signs of their incompetence whenever the opportunity presented itself.

This is not a criticism of Kane’s methodology or conclusion, simply a reminder that his point was local: where manuscripts that otherwise attest the A version have passages in common with C, and where the inclusion of many of those passages results in inconsequence, repetition, broken sense and the like, it is fair to conclude that these manuscripts’ shapes are the results of scribal rather than authorial behavior. He was not arguing that all conflations bear signs of officiousness. No student of the poem’s production can dismiss conflation from consideration just because a passage is smooth. After c. 1390, conflation is always a possibility. It comes to its fullest fruition in the appearance, a few decades later, of Huntington MS Hm 114 (Ht), which combines matter from all three major textual traditions,¹² but there is no reason to exclude Bx from consideration of its workings.

Shapes and Versions

Ever since Walter Skeat divided Piers Plowman’s manuscripts into three authorial “versions,” editors and critics have followed suit, assuming that each version comprised an integral body of manuscripts. It was clear that later texts combined matter from two or even three of these textual traditions, but no one seems to have considered the possibility that this could have occurred at the earlier stages as well. The versioning of the poem reached its apogee in the Athlone edition, which defined the A, B, and C texts by process of elimination followed by a leap of faith: once the editor identified the officious intrusions of the W, N, and K traditions, he would be left with “uniformity,” assumed to equate to authorial integrity: “Among the more than fifty manuscripts preserving the work called Piers Plowman two distinct versions are handsomely attested by substantial numbers of copies of uniform shape. The remaining manuscripts have various shapes, but seventeen of them are distinguished by the common possession of some 400 lines not found in the other two versions.”¹³ The “two distinct versions of uniform shape” are of course B and C, while the 400 lines shared by the remaining manuscripts identify them as witnesses to A, which take “various shapes” via acts of conflation and the additions of C conclusions in W, N, K, and the TH²Ch group.

Although the classification of any given witness to the poem by its “shape” appears to be merely preliminary to the more rigorous work of critical editing, it is in fact a compromised pre-determiner of that editing’s outcomes. The “uniform shape” attested by B and C manuscripts is very broadly defined: B manuscripts feature a Prologue and 20 passus; C manuscripts, a Prologue and 22 passus; and each, a series of passages not in the other. But given that the surviving B tradition did not assume its current “shape” until conflation had already begun, and that conflation did not always leave signs of its presence, it remains unclear on what grounds Kane assumes that the shape of B is solely the product of authorial activity. On the one hand, “shapes” are determined by the character of a manuscript’s passages; on the other, “shapes” predetermine the character of a manuscript’s passages. Here is a major dilemma that has not figured in Langland criticism precisely because that criticism’s methodologies bury the evidence that shows it to be so deeply misleading.

The prior classification of the evidence into three independent groups produced some decidedly odd results, which perhaps should have raised the alarm a while ago. For these “versions” interfered directly with the editors’ own presentation and analysis of the surviving evidence. They stress, for instance, that with regard to Piers Plowman B 19–20 and C 21–22 “the manuscripts of the two traditions were treated as constituting two great families with an exclusive common ancestor, a single scribal B copy.”¹⁴ Yet such an approach calls for a single edition of those passus, with full display of collations, in the B edition. As it stands, that edition does not indicate whether Kane and Donaldson weighed or even had access to all the C readings, or to what degree C-editor George Russell bore responsibility for the readings printed in their B text. The famous conclusion to their Introduction claims:

The apparatus supporting that text [sc. their B text] contains all material evidence for determining its original form afforded by the known B manuscripts other than S and Ht. How we have interpreted that evidence, and the evidence of the A and C versions bearing on it, has been laid wholly open to scrutiny in the preceding stages of this Introduction. Whether we have carried out our task efficiently must be assessed by reenacting it.¹⁵