University of Virginia Library

The early editors of Marlowe's Doctor Faustus knew only the B-text, which had its first edition in 1616, until in 1850 Alexander Dyce printed both the A-text of 1604 and the B-text of 1616, a procedure followed by Peter Cunningham. However, the next editor, A. H. Bullen, in 1885 selected the A-version as his copy-text on the belief that it alone represented the pure original play whereas the 1616 text was a later revised version containing the additions for which Henslowe on 22 November 1602 paid £4 on behalf of the Admiral's men to Samuel Rowley and William Birde, or Borne. Bullen's hypothesis was followed by Tucker Brooke in his influential 1910 edition of Marlowe. This view prevailed until in 1946 Leo Kirschbaum exploded the bombshell of "The Good and Bad Quartos of Doctor Faustus" (The Library, n.s. 26 [1946], 272-294), which effectively argued that the A-text represents a so-called 'bad quarto,' that is to say a memorial reconstruction of some acting version contrived without direct reference to any manuscript in the authorial line. Because of the paraphrases of this A-version in the anonymous Taming of a Shrew (1594) and an apparent reference in The Merry Wives of Windsor (1600-1601) to a scene in the B-text not present in the A-text (combined with what Kirschbaum took to be internal cross-references in A to episodes present only in B), Kirschbaum concluded that the A-text was a memorial reconstruction of the B-text as we know it from the 1616 quarto and hence that neither text contained the Rowley-Birde additions.

Sir Walter Greg had been working on parallel lines, but he could not have escaped being seriously influenced by Kirschbaum's conclusions, which he made the basis for the extraordinarily intricate and detailed reconstruction of the history of both texts in the introduction to his parallel-text edition of Doctor Faustus in 1950. It was peculiarly


Page 2
unfortunate that Kirschbaum felt the demonstration of the memorial nature of the A-text by necessity involved the hypothesis that it was a redaction of the whole of the present B-text so that one in effect proved the other. It was also unfortunate that this linkage was accepted although not without question by Greg so that it became an article of faith: one could scarcely believe that the A1 edition was a bad quarto unless one simultaneously assumed that it derived memorially from the present B-text without the additions.[1] The two problems are in fact independent; and where the evidence appears to overlap in certain comic scenes, the hypothesis of B-text revision as part of the 1602 additions is as satisfactory as A-text memorial corruption and shortening where this evidence concerns the identifiable revisions.

In the opinion of the present writer Kirschbaum's argument needs re-examination, and a more scrupulous examination of the problem (removed from the question of the A-text as bad quarto) will suggest that the A-text — if it is a report as seems most probable — refers back to the original version of the play whereas the B-text of 1616 contains not only the 1602 Rowley-Birde additions but also their revisions, chiefly confined to the comic middle part of the play, conventionally assigned as Acts III and IV, but also affecting the tragic action in Act V.[2]


Page 3

Five parallel passages are commonly accepted as borrowings from Faustus in the anonymous Taming of A Shrew entered in the Stationers' Register on 2 May 1594 and often considered to be itself a bad quarto of Shakespeare's Taming of the Shrew instead of a source of that play. Two of these passages echo lines in which the A and B-texts agree.[3] One repeats the word from the B-text in a passage where both versions otherwise agree,[4] and one the word from the A-text in similar circumstances.[5] These have no evidential value in respect to the 1602 additions. On the other hand, the fifth case (if legitimate) may echo a passage in a scene of the play where the A-text is wanting and the B-text alone is available. This is A Shrew, IV.ii.60-61, 'This angrie sword should rip thy hatefull chest, | And hewd thee smaller than the Libian sands', generally taken to be derived from Faustus, 1397-98, 'And had you cut my body with your swords, | Or hew'd this flesh and bones as small as sand' (IV.ii). Moreover, although Greg (p. 28) admits that the parallel is not necessarily convincing, he and Kirschbaum agree that it can be strengthened by a reference in The Merry Wives of Windsor, IV.v.67ff. (TLN 2283-87) in Bardolph's speech, 'for so soone as I came beyond Eaton, they threw me off, from behinde one of them, in a slough of myre; and set spurres, and away; like three Germane-divels; three Doctor Faustasses.' This is taken to refer to the same scene, in which Benvolio, Frederick, and Martino are


Page 4
punished by the three devils Asteroth, Belimoth, and Mephistophilis, with especial reference to Martino. Faustus commands, 'Go Belimothe, and take this caitife hence, | And hurle him in some lake of mud and durt' (1408-9), and shortly in Scene xiii (IV.iii) Martino re-enters bewailing that he is 'Halfe smother'd in a Lake of mud and durt, | Through which the Furies drag'd me by the heeles' (1434-35).

If the B-text contains the additions of 1602, these two scenes of Benvolio's attempted revenge and its punishment by Faustus must be among them, for they are wanting in the A-text and are intimately bound up, also, with the Emperor Carolus scene and its quite different treatment of the scoffing knight in the A-text. A third piece, one that may be called internal evidence, has also been adduced by Kirschbaum (pp. 285-286) and approved by Greg. In the A-text the Horse-courser, entering wet, cries out, 'he bade me I should ride him into no water; now, I thinking my horse had had some rare Qualitie that he would not have had me knowne of, I like a ventrous youth, rid him into the deepe pond' (Greg, A-text, 1180-83). In the B-text Scene xv (IV.v) which is absent in the A-text, the Horse-courser's narrative in a different context reads, 'Doctor Fauster bad me . . . in any case ride him not into the water. Now sir, I thinking the horse had had some quality that he would not have me know of, what did I but rid him into a great river' (1540-43), whereas the parallel B-text of the incident itself (Scene xiv) as quoted above in the A-text has simply, 'I riding my horse into the water, thinking some hidden mystery had beene in the horse, I had nothing under me but a little straw' (1484-86). Because the language of the A-text seems to repeat that of the B-text in Scene xv wanting in A, not the language of the parallel B-text of Scene xiv, Greg believed that it established the presence of the text of Scene xv when the A-text was memorially reconstructed.

On these three pieces of evidence, in the last analysis, rests the whole case for the absence in the B1 quarto of the 1602 additions. It is now my task to re-examine Kirschbaum's and Greg's postulate, deriving from this evidence, that the whole B-text as we have it is a unified one that represents substantially the original composition of the play.

The most obvious point of attack are those scenes, or major parts of scenes, that are present in the B-text but wanting in the A-text. These are in order: (1) lines [831-835], 836-980 of Scene viii (III.i) which represent all but the opening of the scene and concern themselves with Faustus' decision to forgo the sights of Rome in favor of disrupting the Pope's feast, the rescue of Bruno, and the exit of the Pope and his train to prepare for the banquet; (2) Scenes xii-xiii


Page 5
(IV.ii,iii), lines 1324.2-1456.1, comprising Benvolio's revenge and Faustus' retaliation;[6] (3) Scene xv (IV.v), lines 1504.2-1557.1, the meeting of the clowns in the tavern with the conclusion of this episode in lines 1588.1-1668 of expanded Scene xvi ( where they interrupt Faustus' exhibition of his art to the Duke of Vanholt and his lady and are humiliated; (4) the start of Scene xviii (V.ii), lines 1796.2-1819, in which Lucifer, Belzebub, and Mephostophilis exult over Faustus' imminent doom and Wagner thanks Faustus for the bequests in his will; and (5) lines 1879.1-1925 later in the scene, Mephostophilis' confession that it was he who tempted Faustus, followed by the visions of heaven and of hell under the direction of the Good and Bad Angels.[7]

In Greg's view (pp. 23-29) these are all A-text subtractions from the B-text original, not B's additions to the A-form of the text.[8] Nevertheless, the evidence for at least some of this material points in the contrary direction. The first of these scenes, in Rome, illustrates with superior clarity the case for addition and revision. Both the A-text and the B-text prefix Scene viii (III.i) with Chorus 2 (753.2-778.1) — the A-text, however, omitting 760-772, 774 — which recounts Faustus' eight days' trip through the heavens and then, on a second trip 'to prove Cosmography', introduces him as first arriving in Rome in order to take part in St. Peter's feast that is being celebrated on this day. Both texts then continue with Faustus' account of his travels through France and Italy, and then his question where they now are, followed by Mephostophilis' reply and his description of Rome. Both then agree in Faustus' impatience to see the sights of Rome and in Mephostophilis' advice to remain in the papal palace to 'take some part of holy Peters feast.'

At this point the two texts diverge. The A-text continues, normally,


Page 6
with Faustus' acquiescence to the plan and with his request to be made invisible. The Pope then enters to the feast, which Faustus and Mephostophilis disrupt, ending with the Dirge and the beating out of the Friars. This is a unified scene that follows the details of the source Historie of the Damnable Life. On the other hand, the B-text at the point of divergence very oddly repeats in expanded form the last two lines of the Chorus but for the sole purpose of changing the situation by yoking the annual feast of Peter with 'the Popes triumphant victory', which we later learn has been the defeat of the Emperor's forces and the capture of Saxon Bruno. After a boast about his life of pleasure, Faustus recounts once more the eight days' journey through the heavens already covered in the Chorus and then begs Mephostophilis to be allowed to stay in order to take part in the ceremonies. Since this was what Mephostophilis had originally suggested, he agrees, and then ensues the episode of Saxon Bruno and his rescue, subsequently given a special transition to the feast, at which point the two texts converge in their action. This material unique in the B-text cannot be a subtraction in the A-text, for all the evidence points towards its addition by a reviser in B. The abruptly altered purpose of the feast, different from the Historie, in order to insert material about Saxon Bruno from Foxe's Book of Martyrs is sufficient demonstration, with the confusion that results from this change and the resulting awkward join. Indeed, if more evidence were needed, it comes in 840-853, a complete non sequitur as a preface to Faustus' unnecessary plea to be allowed to stay, for the repetition of Chorus material suggests, as does the repetition in 833, that the writer envisaged the excision of the Chorus and was awkwardly providing within the scene itself the same information about the setting that had been the reason for the Chorus' existence. To reverse the situation by arguing for the A-text as a memorial redaction of B is quite impossible, for it would mean creating unity out of a thoroughly mixed-up situation.[9]


Page 7

Stylistic evidence confirms addition and revision. The reviser of this scene was free with rhyming couplets whereas none had appeared earlier in the play; moreover, couplets have no place in the original opening of the scene. Liberal use of rhyme, indeed, sets off the work of this particular reviser, as well as his occasional practice of completing a part line with the opening line of the next speech, a convention only sparingly used, if at all, in the original manuscript.[10] This partiality for couplets, it must be emphasized, is no part of the play where the A-text is present: couplets occur only in certain of the scenes, or parts of scenes, in which the B-text is the sole authority. Thus they constitute the best stylistic test available to isolate the work of this reviser, for they are unique to him.

Scene xii (IV.ii), 1324.2-1430.4, contains at least fourteen rhyming couplets, and Scene xiii (IV.iii), 1430.5-1456.1, in its short length has nine. These two linked scenes found only in B depict the attempted revenge of Benvolio and the punishment Faustus gives to Benvolio, Frederick, and Martino. Little doubt can exist that these two scenes are the work of the author of the additions and revisions in Scene viii and hence that they participate in the additions made to the B-text. Scene xii has one completed line between speeches, and Scene xiii has four. The elaboration of the material here away from the simple dramatization of the Historie in the A-text of Scene xi is obvious, and the same difficulty exists here as with Scene viii in accepting the A-text as a memorial reconstruction. That is, the A-text knows nothing of the elaborate byplay marking Scene xi between Benvolio and his friends, or between Benvolio and Faustus, that associates this new material with revised Scene viii through the introduction of Saxon Bruno, and that leads to the revenge and its punishment. Although the exhibition of Alexander to Carolus is recalled in its salient details, the reporter of the A-text would need to have suffered a complete blackout when it came to the extensive comic material that is interwoven in Scene xi, so complete indeed that not a single detail was remembered and he was forced to return to the Historie to rebuild in an elementary and quite different action the lost original. This is not a hypothesis that lends itself to ready belief when the stylistic evidence as well as the evidence for elaboration, not for subtraction, associates this extensive section with the additions in Scene viii.


Page 8

In Scene xviii (V.ii) the opening triumph of Lucifer, 1796.2-1819, has only one couplet (1811-12) but two completions of part lines. By common consent it stands out stylistically from what follows in the scene, it is wanting in the A-text, and thus it can be assigned with some confidence to the revision of the original play. In corroboration, a faulty join of the new material is revealed by the repetition in its brief interlude between Faustus and Wagner of information given in Wagner's soliloquy (present in both texts and therefore part of the original) beginning Scene xvii (V.i), a repetition reminiscent of the same overlapping of old and new material in Scene viii.[11] Finally, 1879.1-1925, another set piece later in Scene xviii, with its visions of heaven and of hell distinguishes itself from the surrounding action by containing at least ten couplets and two completed part lines. It is unknown to the A-text[12] and stylistically associates itself with the similar cases earlier of B-text additions.

The two crucial allied facts are that these noted scenes are unique to the B-text and, simultaneously, share distinctive stylistic characteristics not found elsewhere in the B-text (or in the A-text). The double association is critically significant in setting these scenes apart from the rest of the material that can be assigned as part of the original manuscript on the grounds that it is shared by both the A-and the B-texts. These scenes distinctive to the B-text are important in the comic action of Acts III and IV but they also touch up the tragic last scene. It now remains to examine the only other scene unique in the B-text, Scene xv (IV.v), 1504.2-1557.1, which instead is a part of the farcical action, with its conclusion in 1588.1-1668 attached to the comic action of Scene xvi (


Page 9

In these two prose sections involving the clowns' drinking bout and their interruption of the magic show at Vanholt, the touchstone of couplets and completed lines cannot be applied, and little can be done to associate the language with that in the verse scenes by the couplet writer.[13] General probability, then, constitutes the argument. That is, all other scenes unique in the B-text can be isolated as distinctive additions set apart from the style and matter of the original manuscript as shown by the agreement of the two texts; hence some special circumstance would need to be present to justify any hypothesis that, alone of all the scenes wanting in the A-text, the tavern action in Scene xv and the aftermath at Vanholt in Scene xvi were omitted in A1 because of total memorial failure or the exigencies of provincial production.[14] Two small hints may, instead, suggest that these sections are also additions, presumably by the same reviser(s) who supplied the other unique B-scenes. In the first place, the original scene at Vanholt— although it exhausted its source in the Historie insofar as the Historie account could be dramatized[15] — was brief and relatively undramatic, especially after the more extensive Carolus episode presented in the original play (as reconstructed by the A-text). Since in the Carolus episode the reviser chose to expand the action of the magic-show by an extensive and completely original reworking of the Historie's account of the knight and his horns (reproduced from the Historie with some faithfulness in the A-text), it may be suggested that he would have felt even more keenly the need to piece out the spare and static magic business at Vanholt by similar extraneous material. In contrast to the Carolus scene where the Historie and the original version of the play suggested the action to elaborate, the Historie was of no use for the additions to Vanholt and the grapes. It may be conjectured, therefore, that with some ingenuity the reviser chose to extend the Horse-courser action beyond its conclusion both in the Historie and in the A-text original. He took hints about the eating of the hay


Page 10
and a very bare account of the charming of some yokels from the Historie and combined these with a narrative recapitulation of the Horse-courser's theft of Faustus' leg that brought it to a new conclusion in the jokes about Faustus' wooden leg and the eventual revelation that Faustus was sound in both limbs.

It will be observed that this hypothesis for the addition of the farcical scene in the tavern, with its transferred recapitulation of the Horse-courser's adventure with Faustus, answers Kirschbaum's and Greg's postulate (described above, earlier) according to which the A-text (1180-83 in Greg's numbering) in the original scene echoes language not of the parallel B-text but instead of the B-text recapitulation in the tavern scene, wanting in A, and hence provides evidence that this tavern scene had been present in the original manuscript where it affected the A-text memorial reconstruction of the earlier scene. Instead, if one hypothesizes the addition of the tavern episode, one may reasonably associate with this a corresponding revision in the earlier scene of the B-text Horse-courser's account of his adventure since it was to be repeated in the later tavern narratives of Faustus' mischiefmaking. Thus one need only take it that the supposed echo in A-text's Scene xiv (IV.iv) of B-text's Scene xv (IV.v), 1540-43, in fact represents no more than A's version of the original scene before its revision and hence before the B reviser's transfer of the fuller account in Scene xiv to the new Scene xv (with the corresponding reduction in B Scene xiv of the original details preserved in A). The narrative in Scene xv is thereby made no more repetitious than is effective in the retelling of a good joke, and the more precise detail transferred to the tavern scene leads naturally to the concluding jest of Faustus' wooden leg in expanded Scene xvi, which requires the tavern narrative as its motivation.[16] No technical objection on the evidence of the A-text


Page 11
can be made, therefore, to the hypothesis that these two farcical actions should be joined to the comic additions (and to the additions to and revision of the tragic action in V.ii) as material distinctive to the B-text not present in the manuscript from which the memorial reconstruction of the A-text derived.

The view that the scenes unique to the B-text represent the Rowley-Birde additions of 1602 has been objected to on the ground that these are not extensive enough to warrant the payment of £4, close to the price for a new play.[17] At this distance it is difficult to adjudicate the merits of such a payment, which presumably would be based more on the intrinsic importance of the new material for reviving the popularity of an old play than on mere length. If the number of editions of the B-text (six between 1616 and 1631) is any guide to its popularity on the stage — and editions may sometimes coincide with revivals — Rowley and Birde earned their money. It is possible, however, that they earned it more thoroughly than is usually supposed. That is, the question of less identifiable revision as well as of the readily isolated new additions must be raised.[18] That there was revision of Scene xiv as a consequence of the addition of Scene xv has been suggested above, with some evidence in its favor. The new material that expanded Scene xi at the Emperor's court as preparation for the added scenes xii and xiii has also been mentioned as a revision of the A-text's elementary (or original) treatment of the scoffing knight dramatized from the Historie. Although these two cases of revision were required by the additions that followed, it is possible that other scenes were worked over in consideration of the £4 payment. In this connection it may be no accident that the B-text's Scenes vii (II.iii) and x (III.ii) substitute


Page 12
Dick as Robin's companion for the A-text's Rafe of the combined scenes from A and that Dick is the name in the only other appearances of this character in added Scene xv (IV.v) and expanded Scene xvi ( This association leads to the conjecture that it was the reviser who split apart the unified A-text scene numbered by Greg as viii and ix although very likely it would be considered as one scene owing to the immediate re-entrance at A-text 985 of Robin and Rafe with the goblet in substantially a continuous action.[19] If so, he would be likely to rework the dialogue and not content himself merely with renaming the characters.[20] The reviser seems to have left the first farcical scene — between Wagner and Robin — more or less alone, for the A-text here is closer to the B-text than in any other farcical action. If so, this scene is the only trustworthy indicator among the farcical episodes of something like the true quality of the A-text memorial reconstruction of such material, the B-text's reviser having interfered in the rest to affect the evidence.

Since one result of the reviser's observed ministrations was the overlapping or the awkwardness of his joins to original material as in the beginning of the scene at Rome and in the bequest to Wagner,[21] other such difficulties may reveal his hand. One of these is the very odd presentation of two magic books to Faustus, the first by Mephistophilis in Scene v (II.i), 543-551, and the second by Lucifer in Scene vi (II.ii), 717-720, in both cases Faustus' thanks being cast in identical words. It is just possible that the reviser was working over the Seven Deadly Sins action and then, moving back, excised the weak prose continuation of


Page 13
the presentation of the first book (almost certainly an interpolation between two Marlovian sets of speeches[22]) and brought the scene to a close, in preparation for Scene vi, by repeating Faustus' thanks to Lucifer now applied to Mephistophilis. But this possibility is offered here with many reservations, for the matter is extremely obscure and an alternative speculation is almost as plausible.[23]

Other anomalies in the play may refer back to the reviser. Some modern editors have found that the exhibition of Helen to the Scholars in Scene xvii (V.i) is inferior on the whole in the B-text to the A-text version.[24] In the same scene, moreover, the first speech of the Old Man is so completely different in both versions as to go beyond memorial corruption in the A-text and to require actual rewriting in one or the other form of the play. Similarly, it is perplexing whether the A-text's omission of the Scholars' discovery of Faustus' body represents Marlowe's original ending and a later production cut, or whether the


Page 14
B-text conclusion is a subsequent addition. As for the last, the terror of the great soliloquy provides a most effective ending, but there is something to be said for the traditional close of a tragedy in peace that is furnished by the Scholars' comments (Scene xix, 1982.3-2001.1). In favor of considering these closing lines as of a piece with the other passages unique to the B-text is the possibility that the final Chorus provides the assimilation of experience and commentary that is usually contained within the last lines of the action, as in the Scholars' comments, and indeed is repetitious if these comments are present. Yet opinions might well differ as to the effectiveness of either end and it is not beyond belief that at this point the acting version of the A-text could indeed have made a production cut. On the other hand, the lines themselves do not bear the unmistakable mark of Marlowe's hand (although he could have written them) and they would not seem to be beyond the capacities of the writer of the revised Old Man's first speech in the B-text.[25] The Scottish verdict of 'not proven' is perhaps the only one possible on the evidence, although in my own view we should have more pertinent evidence before assigning to Marlowe any such separate passage unique to the B-text where it automatically comes under suspicion of joining the other unique passages in the B-text that are the work of Rowley-Birde. For the rest of the possible tinkerings in the tragic action, one may say only that since the hand of the reviser has been seen in V.ii with some certainty, there is no bar to his also having worked over parts of V.i, specifically the sight of Helen and the Old Man's first speech.

The alterations made in the B-text at a later time than the formation of the A-text are not confined, then, to the additions, important as these are, but also include revision and expansion of comic and farcical scenes already in existence as well as some reworking of the tragic action in Act V. That these alterations are on the whole extensive enough to be identified with the additions by Rowley-Birde for which £4 were paid in 1602 is the proposition I wish to advance. They cannot be identified with anything else, certainly. Indeed, the characteristics of the verse in the additions to Scene viii in Rome, with special reference


Page 15
to the extensive use of adjectives in -al to end verse lines,[26] are so similar to Samuel Rowley's When You See Me You Know Me, printed in 1605, as to have led Greg (pp. 133-136) to conjecture that Rowley (and perhaps even Birde) was the original collaborator with Marlowe.[27] This is the reviser whose penchant for couplets in scenes unique with the B-text sets him apart from any of the remaining scenes that by their presence both in A1 and B1 may be thought of as original, whether or not by Marlowe.

The question at last arises whether the evidence herein adduced for the special characteristics that set apart the unique B-scenes and the expansions of other scenes, together with the revision, all of which points irresistibly to the Rowley-Birde additions of 1602, is strong enough to stand against the three pieces of contrary evidence — from The Taming of A Shrew, from The Merry Wives of Windsor, and from the supposed memorial echoes in the A-text of unique B-scenes. The 'echoes' can be discarded immediately on the ground that they are as reasonably to be assigned to revisions in the B-text as to memorial recollection in the A-text. The suggestion that A Shrew, IV.ii.60-61, 'This angrie sword should rip thy hatefull chest, | And hewd thee smaller than the Libian sands' is a paraphrase of Faustus, 1387-98, 'And had you cut my body with your swords, | Or hew'd this flesh and bones as small as sand' in the unique B-scene xii (IV.ii) is, in Greg's words (p. 28), 'one of the least convincing of the parallels, and cannot be taken by itself to prove anything at all.' But Greg continues, 'It is, however, supported by a curious point relating to another of the scenes peculiar to B,' and he describes the allusion in The Merry Wives of Windsor to the punishment of Benvolio, Frederick, and Martino, and approves of the authenticity of this latter reference. He concludes, 'Since, then, The Merry Wives was probably written in 1600 to 1601, it follows that the scene in question was at any rate no part of the Rowley-Birde additions of November 1602.'

Although not so positive as Kirschbaum and Greg of the precise application of this reference to the B-text's Scenes xii-xiii,[28] I am not


Page 16
prepared to deny it out of hand; but even if it were to be accepted — and this is not altogether certain — the conclusion does not follow by necessity that the Rowley-Birde additions are automatically disproved. Although The Merry Wives was probably composed about 1600-1601, the text we have comes to us from the Folio of 1623, printed from a transcript by Ralph Crane of the promptbook current at that date. A bad-quarto text registered on 18 January 1602 (almost a year before the payment to Rowley and Birde) and published in the same year may offer only doubtful testimony about the form of the original play but at least it does not preserve the allusion. A number of anomalies in the Folio text, including the change of Ford's alias from Brooke to Broome, and a clearcut rewriting of parts of the fifth act, constitute a warning that the Folio text cannot be relied on in all its details to provide us with an exact copy of the 1600-1601 text. The differences between the Folio and the bad quarto in the matter of the Germans, moreover, suggest some caution in assigning to 1600-1 anything to do with this puzzling German material in its Folio form. In general, the Folio material seems to allude to Frederick Count of Mömpelgart's attempts first to be honored with the Garter and then to secure its insignia. In 1592 and 1595 he or his representatives had some obscure trouble about the purchase of horses that excited considerable comment. He was elected, finally, in 1597 but despite protests in 1599 and an embassy in 1600 he did not receive the insignia until November 1603 under James. The possibility holds, therefore, that the allusion to Faustus, whether or not specifically to B-text's Scene xiii, was a part of the rewriting that produced the altered Folio text as we have it, just possibly for the court performance of November 1604 when the award of the insignia was a year old and the joke was less touchy and more reminiscent.

In the midst of these uncertainties about the exact form of the Shakespeare text, therefore, no one can say for sure that the allusion to Faustus was present in The Merry Wives on its initial writing in 1600-1, given the known facts that suggest Folio alterations in the play.


Page 17
Under these circumstances, in my view, the keystone evidence that demonstrated for Greg the impossibility of Rowley-Birde's additions being present in the B-text of Faustus is too uncertain, and even suspect, to bear the major weight of excluding these additions. Greg's interpretation is contrary to the general run of probability[29] and also contrary to the specific evidence for the distinctive quality as well as the unity of the added B-text material, the evidence that it does not always fit smoothly into the original context, and the generally accepted relation of its style to that of Samuel Rowley. Once the additions and revisions herein isolated are accepted as the 1602 reworking, Greg's extreme complexity of hypothesis is materially simplified, for analysis shows that almost every one of his arching conjectures that pile problematic speculation on speculation refers ultimately back to the difficulties he faced in arguing for Kirschbaum's postulate that the B-text could not contain the Rowley-Birde additions largely because of the allusion in The Merry Wives of Windsor. Instead, we have a relatively simple case with no necessity for the various expedients that — among others — give us Marlowe's reworking the play in its promptbook to explain why the A-text sometimes seems more authentic than the B-text 1602 revisions, or Greg's acceptance (pp. 95, 132) of Bullen's suggestion (made at a time when the memorial nature of the A-text was not understood) that the passages peculiar to B in Act V 'are original drafts discarded in the final make-up for performance.' I suggest that Kirschbaum's premise was faulty because it denied the 1602 additions in the B-text. Hence when Greg accepted the premise without sufficient scrutiny, his supremely logical intellect was channeled into constructing the complex series of hypotheses, each one more speculative, that seemed necessary to solve the massive difficulties of what was, in fact, an essentially unnatural problem. Once the narrowness and the ambiguity of the evidence behind Kirschbaum's premise is recognized, the simpler logic


Page 18
of accepting the 1602 additions as present in the B-text reduces the problem of Doctor Faustus to relatively manageable terms.[30] The numerous textual problems are not automatically solved by this hypothesis, but various of them are materially clarified; and except for the contamination of the manuscript text from the 1613 quarto and the ministrations of the editor-inscriber of the 1616 copy we now know pretty well where we stand. The critical shift in our view of the play that will result from the exclusion of the unique B-text material and the return to the identification of the A-text as representing the form of the original play even though in a memorially corrupted text, should also clarify the position of Doctor Faustus in the history of the drama.