University of Virginia Library

BECAUSE THE FIRST PRINTED TEXT OF THE TWO Noble Kinsmen, the quarto of 1634, preserves a number of manifest signs of prompt annotation, it has been assumed that the printer's copy was a prompt book, and specifically the prompt book that the King's company used for a revival of the play which apparently took place in the middle 1620's. In the notes to his edition of the play in The Shakespeare Apocrypha (1908), C. F. Tucker Brooke said that the "text is based upon the prompter's stage copy, as we know from the S.D.D. to I.iii.68, I.v.28, &c." (p. 434). In William Shakespeare (1930), Sir Edmund Chambers remarked that "evidently the manuscript had been used as prompt-copy" (I, 529), and in an article published in 1945, Sir Walter Greg stated that the quarto "shows unmistakable signs of having been set up from the very prompt-book used by the King's company."[1] In his latest work Greg adds that Q was "evidently printed from a prompt-copy written or at least annotated by Edward Knight, book-keeper in the King's company" (The Shakespeare First Folio [1955], p. 98).

But a close examination of the basic text of the play suggests that the case may not be quite so simple: whereas the printed text does show unmistakable signs of the prompter's hand in marginal warning notes and in actors' names, it also exhibits certain other textual peculiarities that point to foul papers. If so, then the printer's copy for Q would constitute annotated foul papers, a species of manuscript which Greg has elsewhere inferred from what appears to be the original draft of Heywood's The Captives. Heywood's autograph manuscript bears elaborate prompt annotations in another hand, but since both hands are too foul for prompt copy,

And in a note, Greg adds, "If the inference here made is correct, the important conclusion follows that a printed text that shows distinctive marks of the prompter may yet have been printed from foul papers."

The "distinctive marks of the prompter" shown in the Kinsmen do not seem to have been introduced for the original production. Scholars are now generally agreed that the play was written in 1613. The date is primarily established by the use in a morris dance in the play of the same characters who constitute the cast for the second anti-masque of Beaumont's Masque of the Inner Temple and Gray's Inn, presented at Whitehall on February 20, 1613. There is, however, no external reference to the play's first performance; the first which occurs appears on a scrap of paper from the Revels Office probably dating about 1619, and seems to indicate consideration for court performance.[2] The prompter's notes belong to a revival of 1625 or 1626; two actors he cast as minor characters were employed by the company together during those years only. The pertinent directions are printed as follows[3]:

"Curtis" has been identified as Curtis Greville, an actor who left the Lady Elizabeth's men for the Palsgrave's company in 1622 and had joined the King's men by the fall of 1626, when his name appeared in the cast of The Roman Actor, licensed October 11; Bentley suggests he probably joined the King's company in 1625 at the time of the reorganization of the companies.[4] "T. Tucke" is evidently an abbreviation for Thomas Tuckfield, whose name appears twelfth in the list of "Musitions and other necessary attendantes" of the company, whom Sir Henry Herbert exempted from arrest on December 27, 1624.[5] The Two Noble Kinsmen provides the only record of his appearance on the stage; and since his name is not in the full list of actors printed with The Roman Actor, it has been assumed that he was not with the company when that play was performed in 1626. Greville and Tuckfield were members of the company together, therefore, for no more than two years, and in view of the severity of the plague through most of 1625, a revival of the Kinsmen in the winter of 1625 or spring of 1626 seems likeliest. As shall be shown in a moment, the inferior limit is also confirmed by the marginal warnings.

The warning notes are printed in the outside margins of verso and recto pages, in small roman type; also so printed is one instance of what appears to be an authorial direction:

• 1. "2. Hearses rea-/dy with Pala-/mon: and Arci-/te: the 3. / Queenes. / Theseus: and / his Lordes / ready." (I.iii.58 ff., sig. C3v)
• 2. "3. Hearses rea-/dy." (I.iv.26, sig. C4v)
• 3. "This short flo-/rish of Cor-/nets and / Showtes with-/in." (II.v, opposite opening stage-direction, sig. E4v)
• 4. "Cornets in / sundry places. / Noise and / hallowing as / people a May-/ing." (III.i, opposite act division, sig. F2)
• 5. "Chaire and / stooles out." (III.v.64-65, sig. G2v)
• 6. "Knocke for / Schoole. Enter / The Dance." (III.v.137-39, sig. G3v)
• 7. "They bow se-/verall wayes: / then advance / and stand." (III.vi.93-96, sig. H1v)

In addition, one speech-prefix, "Daughter.", is printed in the left margin, at III.v.60, sig. G2v.

The first, second, and fifth of these are clearly the prompter's annotations. The specification of particular musical instruments and effects in the third and fourth sounds like the prompter, while the sixth appears to be the prompter's direction for a signal to bring on the morris dance. But the language as well as the substance of the seventh sounds authorial. That the direction appears in the left margin may be due to an absence of space in the white area to the right of the speeches, although elsewhere the compositors have set long directions between lines of speech, as, for example, on the next page. And there is another point to be noted: just opposite the first line of the direction, within a line of speech, there appears a star or asterisk, as follows:

• They bow severall wayes: then advance and stand.
• Arc. And me my love: * Is there ought else to say?
• Pal. This onely, and no more . . .

It is difficult to avoid the inference that there is a connection between the direction and the star—that the star is intended to mark the point where the action specified by the direction is to occur. If so, the direction is probably an authorial gloss.[6]

It is a fair a priori assumption that all the prompt annotations were made at the same time that the actors' names were introduced into the text, by the same hand; and there is some evidence, I think, that this is so in fact. The King's company's book-keeper at the time of the revival was probably Edward Knight; his name appears at the head of Herbert's protection of 1624, and he is specifically identified as the book-keeper by an entry in Herbert's Office Book, dated October 21, 1633. The extant prompt copy of The Honest Man's Fortune, still bearing Herbert's licence of February 8, 1625, is in his hand, and it has been suggested that he took up his duties as book-keeper not long before that date.[7] Now, all of Knight's manuscripts show a distinctive use of colons in name-sequences. His practice may be illustrated by a few examples from The Honest Man's Fortune and Believe as you List, the latter a manuscript in Massinger's hand but annotated and prepared for the stage by Knight.[8] At III.iii.37-39 in the former, an entry in the left margin reads, "Ent: Montagne: / bare: Lamyra: / Lady Orleance: / Charlot: Viram:". Or at the head of IV.ii, the following: "Ent: Dubois: Orleance: Longauile: Amience: / .2. Lacqueyes: A Page wth .2. Pistolls:". A typical example from Believe as you List is the entry in the left margin at ll. 342-344: "Ent: fflaminius: / Calistus: Demetri:". Massinger's own use of periods in name-sequences may be illustrated by the direction which Knight's replaces: "enter Titus flaminivs. Calistus. / Demetrius. 2 freedmen." The cancelled direction is some six lines below Knight's.

The sequence of names in the first of the Kinsmen's marginal notes is also punctuated with colons, and this feature, as Sir Walter Greg has suggested,[9] points to the hand of Edward Knight. It may be noted, too, that the single sequence of actors' names similarly shows a colon, in distinction from the sequence of character names to which it is attached; and in view of the probability arising from the date of the

revival that Knight introduced the actors' names, it seems to be a reasonable assumption that the colon here is his.

The punctuation in part of the direction at the head of I.iv may also be Knight's. The entire direction is printed in Q as follows:

The description of the action sounds authorial, but the specification of off-stage noises and music in the first line, where the colons appear, could very well be the prompter's.[10] The mid-sentence capitalization of Enter, not duplicated elsewhere in the text, suggests that the words preceding may have been an addition.

Otherwise, the stage-directions in Q are punctuated in what might be called conventional fashion, without an unusual use of colons in sequence. It would follow, therefore, that Knight did not transcribe the basic copy to which the annotations were made; in other words, the prompt annotations must be additions to the original manuscript and not integral parts of it; and if the clearly identifiable signs of prompt use are eliminated, it appears that the original manuscript may not have been a prompt copy.

A prompt copy was prepared in 1613, presumably, when the play was written and first produced. If this copy was not available when the play was revived, what kind of copy could the company archives have supplied Edward Knight when he made his annotations preparatory to the revival? Whatever it was, it was this copy with Knight's annotations which John Waterson, the publisher, obtained (from the playhouse, presumably[11]) for his quarto of 1634. Authors' fair copy may be eliminated because, first of all, the quarto exhibits too many irregularities for fair copy, either authorial or scribal. These will be surveyed in a moment. Secondly, Q shows two distinct linguistic habits which correspond to the authorial divisions indicated by other areas of evidence —metrical, stylistic, etc. The implication of this is that an authorial draft in the hands of the two authors lies behind the quarto; a faithful

copyist might preserve the authors' linguistic preferences, but an authorial fair copy would be likely to bear the impress of whichever of the collaborators made the copy, for it may be supposed that the responsibility of preparing the fair copy would fall to one of the authors, the better that any imperfections resulting from divided authorship might be removed. As will be seen, Q bears a number of just such marks of jointure. There remain, then, foul papers, probably the likeliest possibility on a priori grounds, or a form of intermediate scribal transcript, a class of manuscript Professor Fredson Bowers has recently postulated; such a copy, he suggests, might have been prepared from foul papers for "consideration, revision, submission to the censor, copying of the parts, or sometimes for marking and cutting in preparation for the final prompt copy."[12] The characteristics of foul papers have been summarized by Greg as follows: According to Professor Bowers, however, some of these marks might persist in an intermediate transcript: Bowers (following Hoy) cites the 1647 Folio text of Fletcher's Wife for a Month as perhaps exemplifying such a transcript, because on the one hand the textual tangles the play exhibits make it unlikely that a prompt copy stood behind the printed text, and on the other, "there is a considerable diminution in some classes, though not in all, of Fletcher's linguistic characteristics, which cannot be laid at the door of the printer," but which may be due to scribal intervention. Now, the quarto Kinsmen exemplifies all of the stigmata of foul papers listed by Greg except the "substitution of the name of an actor, when the part is written with a particular performer in view"; but it also shows the same kind of evidence for an intervening scribal transcript to which Bowers refers above—that is, in the scenes assigned to Fletcher there is a reduction from his usual practice in the number of ye's. The problem this represents will be discussed in a moment; first, however, the evidence for foul papers.

So far as the sense of the spoken word is concerned, Q is generally fair. Still, there are a number of passages which can hardly represent the authors' final intentions. Many of the difficulties in Q are of course compositorial in origin; Clough hee (I.i.20), for example, is probably only a misreading of Chough hore. But there are some obscurities arising from the syntax, which must be, therefore, authorial. In I.iii.19-22, the essential meaning is clear enough, but the words must have sorely tried the art of the actor who had to read the passage as it stands:

The sentence would be less disturbing if the parenthetical element were transposed to the line below, following "killing 'em"; a misplaced marginal or interlinear addition or revision might be conjectured, although reconstruction is difficult.

Part of the trouble with a later passage in the same scene is probably compositorial misreading, but the passage remains defective even when the obvious corruptions are corrected. In Q, lines 72-82 read as follows:

"her careles, were," appears to be a misreading of "her careles weare," "humd on" must be a variant spelling for "humd one", and "misicall" is an obvious misprint. The emendation for "fury-innocent" now generally adopted is Lamb's "eu'ry innocent", although it does not satisfactorily answer all the questions which arise as to the origin of the error. Finally, "comes in Like old importments bastard" not only resists emendation but defies sense. The whole passage, which falls in one of the non-Fletcherian scenes, suggests that the author had not fully worked out his meaning, nor found adequate expression for it, and had written in an almost illegible hand.

The jumble at I.iv.40-45 is one of the cruxes of the play. It is also in a non-Fletcherian scene:

Another tangle appears at V.iii.85-89, again a non-Fletcherian scene:

As with the passage just above, some improvement may be effected by punctuational emendation, but even so the meaning, or a meaning, must be forced from the passage.

There are other passages nearly as obscure, but these are the worst. In addition to textual tangle, there are two evident hiatuses in the play, at V.i.50 and V.iv.77. The former appears in print as follows:

And the latter: In the former, an object for "prewarne", corresponding to "havocke" for "proclaime", seems necessary, and all editors following Seward have adopted "whose approach" arbitrarily. It is likely that merely illegible copy is at fault. Such may also be the case in the latter example; it may be conjectured, however, that the fragment of V.iv.77, "on end he stands", represents a revision of the four words immediately above in line 76, "on his hind hoofes".

It would seem significant that all the passages cited fall in scenes traditionally assigned to Fletcher's collaborator. It is not to be expected, of course, that Fletcher and his co-author, whether or not he be Shakespeare, should have necessarily produced scenes of equal verbal clarity; indeed, one of the features which is invariably mentioned in discussions of the non-Fletcherian parts is their denseness or "knottiness". But such denseness is spoken of as a positive characteristic of style and thought, while the trouble with the passages cited may almost be described as an absence of style and thought. It would appear, therefore, that the copy presented some differences as to the state of finish between the Fletcherian and non-Fletcherian parts.

Q also exhibits a number of defects in the stage-directions. An omission occurs in the opening entrance, wherein the names of Pirithous and Artesius are absent. Both characters are essential to the ensuing action. The discrimination against Pirithous may be laid to compositorial eye-skip, since Theseus's name appears twice in consecutive lines, once correctly and once where Pirithous should have been named. While the compositor may have overlooked Artesius, it is more likely, especially in view of another omission in a later scene, that the dramatist—the scene is one given to Shakespeare—had not fully planned the action when he wrote the initial stage-direction, or that in his concern with the pageantry of the wedding procession, he simply overlooked this minor functionary: Artesius is not involved in the action for some 160 lines, when he is directed to go out and gather an army.

The other omission occurs at I.iv, another non-Fletcherian scene. This time the missing characters are a Herald and the kinsmen themselves; the latter are apparently to be borne in unconscious in the hearses provided by the marginal warning note printed thirty lines earlier (see no. 1, above). There is no provision, either, for the attending Lords who, as the prompter recognized, should have accompanied Theseus on this triumphal occasion. The omissions are too many to make it likely that the compositor is to blame; it rather appears that the author was intent on the action which opens the scene, the meeting of Theseus and the three Queens (strictly speaking, there is no entry for them either—only their meeting Theseus is specified), and in the moment of composition left the other characters to shift for themselves.

The prompter's specification of Theseus's Lords in the marginal

note at I.iii.58 incidentally reflects upon the adequacy of his manuscript with respect to the supernumeraries. In several other entrances, no attendants are indicated, when the circumstances clearly require them; for example, in addition to the omission noted above, the transport of the dying Arcite in his chair at V.iv.85 is left up to Theseus, Hipolita, and Emilia. However, Hipolita and Emilia were Amazons, and Theseus was the kinsman of Hercules. Otherwise, Q's stage-directions provide "attendants," "others," or most often, simply "&c." A permissive direction also involving a discrepancy between the number of characters indicated in the stage-direction and ultimately presented on the stage should be particularly noted. It occurs at the beginning of III.v, the scene in which the morris dance drawn from Beaumont's masque is performed. Beaumont's original anti-masque called for six women and six men, plus a presenter; all had specific roles to play, the women's corresponding to the men's, such as He-Fool and She-Fool, He-Baboon and She-Baboon, etc. There is some question whether the parts in the play are exactly parallel to those in the anti-masque,[15] but it is certain that in both there are supposed to be six couples. When the Schoolmaster, acting as presenter, asks, "where's their women?" five respond by name. When he commands, "Couple then And see what's wanting", the absence of a sixth, Cicely, is discovered. Her part is eventually taken by the Jailer's Daughter, who wanders on the scene a little later. At the very outset, then, there are five women on stage; but the stage-direction provides for "a Schoole master .4. Countrymen: and Baum [read Bavian] .2. or 3 wenches, with a Taborer." The discrepancy is especially curious because the source for the piece, Beaumont's anti-masque, clearly requires six; either the author—Fletcher—had not decided at the beginning how all the women were to be gathered on stage or he was simply uncertain as to the composition of the cast. At any rate, the discrepancies point to an authorial manuscript of a species closer to foul paper than fair copy. The inconsistencies in the casting of both female and male parts (see note 15) lead one to wonder, indeed, if the author even had a complete text of Beaumont's masque before him as he wrote. The masque was licensed on February 27, 1613, and printed, presumably, shortly after, but the cast of the morris dance is given in a narrative description, in language reflecting the viewpoint of a reporter or spectator, rather than in the text proper; the 1647 Folio text is more likely to be based on Beaumont's own papers, and it has only the bare direction, "The second Antimasque rusheth in, they dance their measure, and as rudely depart." It is quite improbable that the dance in the play was the genesis for that in the masque; Stowe's Annales (1631) speaks of the anti-masque as being of and as A. H. Thorndike wrote, "it is inconceivable that this antimasque should have been introduced into that notable court entertainment after having been staled at the public theatres."[17] Thorndike adds, however, that the company must have utilized the morris dance very shortly after the masque was presented, "while the novelty and success of this dance were common talk. A few years later, other antimasques would have been performed at the theatre." The quarto's description of the anti-masque especially notes the audience's delight with the "Countrey jollitie" and records that James himself called for an encore.

An inconsistency in the designation of a character occurs in II.i-ii and is notable because it appears to reveal the authors' division of labor. The facts are briefly these. II.i is a prose scene assigned to Fletcher's collaborator, with II.ii a verse scene clearly belonging to Fletcher. In the former, the Jailer makes his first appearance and is called Iailor in the stage-direction and in the speech-prefixes. But at both his appearances in II.ii, he is called Keeper, in the speeches, entrances, and prefixes. It is quite unlikely that the Jailer and Keeper are in fact two different persons; there is no indication elsewhere in the text that Palamon and Arcite are guarded by more than one man; there is only one Jailer in the play's source, Chaucer's Knight's Tale; and in II.iv Fletcher has the Jailer's Daughter, there so called, speak of her father as the "meane Keeper of his Prison". When the Jailer reappears in Acts IV and V, in both Fletcherian and non-Fletcherian scenes, he is

regularly Iailor (or Iaylor). The likeliest explanation is that Fletcher himself realized the discrepancy after writing II.ii.

The inconsistency in character-names is accompanied by a further irregularity in the articulation of scenes i and ii, as marked in Q. The dialogue establishes that II.i is played in a garden court just below the tower where Palamon and Arcite are imprisoned. It is to this court that the Jailer, his Daughter, and her Wooer enter; during the scene, Palamon and Arcite appear—without speaking—on their prison parapet above, upon which they are observed by the others below. The scene is closed with a bare "Exeunt". II.ii opens with another entrance for Palamon and Arcite, as follows: "Enter Palamon, and Arcite in prison." Now, on the Jacobean stage this would in fact have been a re-entrance and as such quite anomalous. There is no indication of an intended break in the continuity of action: Q's II.i-ii must have been played as a single scene, with Palamon and Arcite remaining on the upper stage after the departure of the Jailer and his company at the "Exeunt" of scene i.

But the irregularities are readily explained as consequences of divided authorship, with the shares of the two authors imperfectly joined. Had the authors composed their work in close collaboration, it is unlikely that the discrepancy in character-names or the re-entry of Palamon and Arcite would have occurred.[18] Nor would these irregularities be likely to have persisted in a fair copy made either by one of the authors or a competent theatrical scribe. The implication is that the printer's copy was the two authors' foul papers, or a very literal transcript of them, showing no editing or normalization in these scenes, at least. It also seems to be implied that the two authors wrote their shares simultaneously but without close consultation; had, for example, Shakespeare written his share (and to Fletcher's collaborator goes the initiation of all the major lines of action) from retirement in Stratford, leaving Fletcher to complete his fragments, as some scholars have postulated, we should expect Fletcher to have achieved a closer meshing of his scenes than appears here. The textual evidence thus supports the inference Sir Edmund Chambers drew from the

authorial divisions: "The distribution of the Shakespearean matter shows that it is a case of collaboration and not of the completion by Fletcher of a Shakespearean fragment" (W. S., I, 532).

If the printer's copy for Q were foul papers, it is improbable that the scene divisions, including the erroneous division at II.ii, were an original feature of the manuscript. The formal division merely reflects, of course, the apparent clearance of the stage at the end of II.i and the following entrance of Palamon and Arcite. It is also improbable that the divisions were introduced while the manuscript was in the playhouse. For one thing, there are several errors in the scene numbers,[19] whereas we would expect accuracy if the divisions were introduced for a functional purpose—for example, as a means of articulating the shares of the two authors, which, according to the pattern shown by the variant character-names for the Jailer, consisted of scenes (or groups of scenes) on separate sheets. Furthermore, it seems to have been Edward Knight's practice to delete scene divisions when they appeared in his copy, as he did when he prepared Massinger's manuscript of Believe as you List for the stage. Only in The Soddered Citizen did he allow the copy scene divisions to stand, and when parts of some scenes of that play were deleted, he was scrupulous in correcting the scene numbers accordingly. Possibly the divisions were introduced by the publisher, at the time of printing. The greater number of Waterson's plays are fully divided; of the ten plays he first published, seven are divided into acts and scenes and three into acts alone.[20] Greg suggests that the numbering of scenes was mainly a literary convention (First Folio, p. 144). It is fruitless to guess just how the errors in the scene numbers occurred, but it is safe to assume that accuracy would not be so pressing a concern in a text prepared for printing as for performance.[21]

Since playhouse manuscripts, that is, prompt books, are scrupulously accurate in the placing of speech-prefixes,[22] errors in this respect which appear in a printed text may point to foul papers as printer's

copy. Some of those shown by the Kinsmen may be compositorial in origin, but at least a few probably belonged to the copy. There is one misplacing: "Emil." is printed a line too late at II.ii.119, so that the first line of her speech is given to Arcite. There are three different repetitions of the same speech-prefix in consecutive lines, of "Per." at III.v.97-98, of "1. Fr." at IV.i.141-42, and of "1 . . . K." at V.iv.38-39. All three may represent compositorial eye-skip, but for the third there is also some evidence, to be discussed below, of copy error. There are two omissions, of the Schoolmaster's prefix at III.v.137 and of Palamon's at V.iv.1. Finally, there are two redundant prefixes (i.e., two in one speech), one for the Wooer at IV.i.46 and one for Palamon at V.i.130. The former appears in Q as follows: The Wooer's half line in 45 may represent a marginal addition, although, since it is metrically necessary, it is more likely that the compositor omitted it and in making the correction had to supply a new prefix in order to distinguish the Wooer's speech from the second friend's, on the same line. Palamon's extra prefix follows a stage direction which interrupts his long prayer to Venus.

Attention has already been called to the appearance in Q of indefinite and permissive stage directions—in entrances calling for "attendants, &c." or ".2. or 3 wenches"—and to the marginal direction in III.vi which amounts to an explanatory gloss on the text. A further word needs to be said on the stage directions for the spectacular episodes, all of which fall in non-Fletcherian scenes. These have all the tone of authorial directions for the producer:

The direction for music may be the prompter's addition—its location suggests interpolation—but the rest is clearly the author's.

It may be added to the evidence already presented that Q shows no general tendency to print entrances a few lines too early, a feature which is sometimes characteristic of prompt copy. It seems clear enough that foul papers are the ultimate basis for the quarto Kinsmen. As suggested above, however, there is some possibility of an intervening scribal transcript, in the light of recent suggestions made by Cyrus Hoy in Studies [23] and by Professor Bowers in On Editing Shakespeare, concerning the lesser incidence of Fletcher's ye's in certain printed and manuscript texts of his plays. Briefly, Hoy conjectures that when such plays as The Woman's Prize or A Wife for a Month show far fewer ye's than the average for Fletcher's other unaided plays, there is strong likelihood of scribal influence at work. The hypothesis is, as he points out, demonstrably true of Bonduca, which exists in two states: a private transcript from foul papers by Knight, showing 147 ye's, and the 1647 Folio text, from the prompt book, showing 352. The Folio text must more nearly represent Fletcher's usage. It has already been observed that the Kinsmen is marked by two sets of linguistic characteristics which may be presumed authorial; one is certainly Fletcher's. In the tabulation which follows, Fletcher is assigned lines 25-28 of I.i (the attribution will be discussed below); I.iv.25-38; II.ii-vi; III.iii-vi; IV.i-ii; V.i.1-17; V.ii; and V.iv.22-38; to Fletcher's collaborator go all of Act I except the passages noted above; II.i; III.i-ii; IV.iii; V.i.18-173; and V.iii-iv, except lines noted above. These divisions give Fletcher 1517 lines, according to Kittredge's lineation, and his collaborator 1257 lines.

               

	hath	has	'em	them	ye	y'	you	i'th	o'th
Flet.:	1	26	44	9	38	8	252	10	8
non-Flet.:	15	10	11	24		2	129	12	21
	other contractions in	s' (contractions of	let's [24]
	th' (e.g. to'th,	us & his; e.g., in's,
	th'offense, etc.)	told's, between's)
Flet.:	12	4	10
non-Flet.:	27	6	3

When projected against the average Hoy gives for Fletcher's unaided plays (322), the incidence of ye's in his parts of the Kinsmen is decidedly low. The occurrence of 'em's, i'th's, and o'th's in Fletcher's parts is about what should be expected, and according to W. E. Farnham's examination of colloquial contractions in The Winter's Tale, Henry VIII, and The Two Noble Kinsmen, the rate of occurrence of these and other contractions in the non-Fletcherian scenes of the Kinsmen identifies these scenes as Shakespeare's. If scribal intervention has been at work, it has manifested itself only in a diminution of Fletcher's ye's, as with The Woman's Prize, and even at the lower rate, ye serves to distinguish the shares of the two authors.

Dr. Hoy suggests that the reduction of ye's in the 1647 Folio text of The Woman's Prize may be attributed to Edward Knight; the Folio text appears to be based upon a prompt book which was very probably in Knight's hand. As noted above, Knight's manuscript of Bonduca also reflects a prejudice against Fletcher's ye's. If the low rate of ye's in the Kinsmen is indicative of scribal influence, Knight's hand might be suspected here also, were it not that the text provides other evidence—the restriction of his peculiar punctuation to the prompt annotations—indicating that Knight's work may be distinguished from the basic text. However, the manuscript Knight took in hand might well have been a transcript made by an assistant for the revival of the middle 1620's or by another book-keeper for even earlier stage use. We may recall that in addition to the original production of 1613 there appears to have been a court performance, or consideration therefor, in 1619 or 1620. A post-revival transcript is improbable; if a new prompt book were prepared for the revival in the 1620's, as seems probable, whatever manuscript Knight first took in hand would be available to serve as printer's copy and a further transcript would be unnecessary.

Whether it was the foul papers themselves or a very literal transcript of them, the manuscript from which Q was printed was utilized for some kind of theatrical purpose. The annotator showed himself implicitly aware of his text's inadequacy when he provided a warning note for the entrance of Palamon and Arcite at I.iv (see above), although he failed to correct the omission of their names from the stage-direction itself. This defect, or the confusion over the Jailer's title in II.i-ii, or the inconsistencies in the number of dancers required for the morris in III.v, would render the text inconvenient, to say the least, in actual performances. In advising that the Elizabethan theatre be approached with some consideration of present-day and past theatrical experience, Professor Bowers has a suggestion which might apply in the present case:

Thus, initial plans for production of the Kinsmen might have been laid with whatever manuscript was at hand, with the marginal annotations representing notes the prompter made during this planning stage, perhaps before the parts were copied out, since some of the speeches would need working on before the actors could begin memorizing their lines. The notes are manifestly incomplete, but they would certainly be supplemented when the regular book was eventually constructed.

To the above Professor Bowers adds, however, that

A transcript which reproduces as many of the original copy's errors as are shown in Q might seem uneconomical; on the other hand, the need for legibility may have been sufficiently great to justify an intermediate transcript, even one which turned out to be as literal as the printer's copy for Q. Such a transcript might have been particularly expedient for a collaborated play in which the authors appear to have written their parts on separate groups of sheets, with, furthermore, the parts of one author bearing interpolations by his colleague (see below).

At any rate, no very clear alternative to the intermediate transcript presents itself, to account for the diminished number of ye's in Fletcher's scenes. There is the possibility that Fletcher's collaborator gave the play its final form and in the process changed most of Fletcher's ye's to you's; but the evidence already shown indicates that the authors wrote their scenes separately, and there is further evidence, to be discussed below, that the play went through Fletcher's hands last, not for general revision but for the insertion of additional matter; there is no

evidence of the collaborator's hand in Fletcher's scenes. There is the second possibility that one of the compositors is to blame. An analysis of certain spellings and of the spacings around punctuation shows that the quarto was set by two compositors, Compositor A doing B1-C1, C2v-E4, F3v-G2v, G4v, I1-I4v, K3, and L4v-M2v, and Compositor B, C1v-C2, E4v-F3, G3-G4, H1-H4v, I4v-K2, K3v-L3v, M3v-M4v, and the Prologue and Epilogue on A1v and N1. According to this division, 25 of Fletcher's ye's and 122 of his you's in the text proper fall in A's stint, with 13 and 130 in B's; the ratio of ye's to you's in A's pages is thus about one to five, but only one to ten in B's, exclusive of Prologue and Epilogue. But even A's rate is far below Fletcher's for his unaided plays. And although there are three you's but no ye's in the Prologue, there are seven ye's but no you's in the Epilogue; the latter distribution does not show Compositor B's apparent preference for you in the play itself to have been constant. There is little doubt that the Epilogue is Fletcher's, and the Prologue is probably his as well; if the printer's copy for the dramatic text proper were a transcript, it is quite possible that the Epilogue, at least, was still printed from Fletcher's original draft. The you's in the Prologue, however, may denote that it was set from a transcript; perhaps there is other evidence of the textual homogeneity of the Prologue and Fletcher's scenes in the occurrence in both of a rather unusual spelling, take for tack, in the phrase, take about (see line 26 of the Prologue, and III.iv.10 and IV.i.152 in two Fletcherian scenes).

Some compositorial influence in the reduction of Fletcher's ye's may be present, but it would not seem to have been especially significant.[26]

There is one final possibility, which, although it rests upon an altogether undemonstrable hypothesis, does permit a rationalization of the apparently conflicting evidence of the linguistic criteria in Fletcher's scenes, on the one hand, and on the other, of the various tangles and irregularities which seem to mark foul papers. All the evidence, bibliographical as well as structural and stylistic, indicates that the two authors wrote their scenes separately, and in no very close collaboration; it is possible, therefore, that their separate sets of pages were handed in to be assembled at the playhouse, presumably by one of the authors, and that a transcript was made of Fletcher's scenes but not of his collaborator's. On the face of it, the transcription of the scenes of only one author in a collaborated play would seem somewhat unlikely, but it might be conjectured that Fletcher's pages were in such a messy state it was felt imperative to prepare clean copy of them, even though, for whatever reason, the original draft of his collaborator was accepted as it stood. It could well have been that some priority had to be established, with Fletcher's scenes being the less legible. The fact that the textual tangles and evidences of illegibility seem to be chiefly prevalent in the non-Fletcherian scenes would appear to support the hypothesis.[27] Evidence will be presented below that some of the non-Fletcherian scenes bear a few additional lines by Fletcher, to suggest that those scenes went through his hands last; but since his interpolations are relatively short, they are not necessarily inconsistent with the procedure sketched above; there is nothing to indicate any general revision upon Fletcher's part. The whole play would almost have to have been put together by one of the authors; Fletcher's treatment of his collaborator's scenes suggests that the responsibility of assemblage was his, and, as conjectured above, the task may have been performed at the play-house, after both parts had been submitted to the company. The interpolations could have been introduced at the same time.

If some such procedure as that described above were followed, the chief desideratum would appear to have been simply clean copy, in accord with Professor Bowers's hypothesis, for some discrepancies remain in Fletcher's scenes—i.e., the variant character-name for the Jailer in II.ii (although it may have been the scribe who later regularized his title), or the confusion over the number of wenches for the morris dance in III.v. So, while it cannot be demonstrated, the transcription of Fletcher's scenes provides an explanation for the diminished ye's and for the signs of foul papers: the preservation in Q of various

marks of jointure between the scenes of the two collaborators and the tangles and irregularities which are more numerous in the scenes given to Shakespeare. But whether the printer's copy was an intermediate transcript, foul papers, or a mixture of the two, it seems clear enough that foul papers are the ultimate basis for Q.

Differences of style or metre in scattered passages in certain of the non-Fletcherian scenes have led a number of critics to suppose that such scenes had been reworked by Fletcher. Thus Chambers (W. S., I, 532) found no evidence of revision but suggested that "in fitting the scenes together, Fletcher may possibly have added a few lines to Shakespeare's." For at least one of the interpolated passages Chambers cited, some visible "seams" remain in Q, to provide further evidence of the author's method of collaboration. The passage is V.iv.22-38;[28] it presents no conclusive linguistic or metrical evidence for or against Fletcher's authorship, but the exaggerated courage Palamon displays in it is strongly suggestive of Fletcher, and it incidentally resolves the sub-plot involving the Jailer's Daughter, the development of most of which had been in Fletcher's hands. There is no mark of textual disarrangement at line 22, but at the other end there are two significant features. Q prints lines 21-23 and 37-40 as follows:

Line 39 is not only metrically defective but the speech-prefixes show a manifest error; the dramatist hardly intended that the first knight should answer himself. But if lines 22-38 are excised, as they may be without destroying the sense of what is left, then the first and second knights' declaration in line 39, "Wee'l follow cheerefully", becomes a natural answer to the third knight's question in line 21, "Come? who begins?" Significantly, the question and answer make up one full line.

An alternative explanation for the confusion in speech-prefixes based on compositorial misreading must ascribe the apparent relation between divided authorship and textual irregularity to sheer coincidence. It is more probable that the repetition of prefixes is an inadvertent seam resulting from an imperfect fusing of Fletcher's addition with an originally non-Fletcherian scene. It would appear that after the initial draft was complete, Fletcher went through his collaborator's scenes, or some of them, not to effect a thoroughgoing revision (unless he subsequently prepared a fair copy, for which there is no evidence), but here and there to insert slight additions, making little or no attempt to remove the marks of jointure. One of the other passages Chambers cites—I.i.25-28—is also marked by defective lineation, as is a third if its line divisions are slightly adjusted. Chambers put the interpolation at I.iv.32-38; line 24, however, is a short line followed by a full line beginning a new speech, and if we allow the addition to have begun at line 25, it can then include a passage which I think has a Fletcherian ring: "Yet they breathe And haue the name of men" (ll. 27-28). Thorndike thought Fletcher had retouched 28-49 of this scene—that is, to the end of the scene. If Fletcher's work extends that far, the textual confusion at 40-45 may be related to his intervention, but in what fashion it is impossible to say, except that in this case Fletcher must have cancelled the original ending of his collaborator or worked his additions into his collaborator's lines so that they cannot now be distinguished. The division I have adopted, 25-38, is a somewhat unsatisfactory compromise. Finally, Chambers and every recent commentator has assigned V.i.1-17 to Fletcher; the number of 'em's and ye's is evidence of his work, although there is no disarrangement of text or lineation here to mark off the interpolation.

The quarto provides a little other evidence of alteration, but it is probably not authorial. It lies in two questions of plus or minus, with probability pointing to the latter, at least for the second instance. The first occurs in IV.ii, which opens with a long soliloquy by Emilia. The initial entrance reads, "Enter Emilia alone, with 2. Pictures", but her speech is closed with a second entrance, "Enter Emil. and Gent." The second example occurs in IV.iii, where two exits are marked for the Jailer's Daughter, at lines 43 and 63, both on sig. K2v. The

exits just bracket her vehement and no doubt censorable denunciation of adulterous courtiers and her cynical comment on proud ladies. In each case we can only guess whether the duplications represent addition or deletion. The whole scene in which Emilia's soliloquy appears is unquestionably Fletcher's, so if there has been addition it was Fletcher's; on the other hand, in 54 lines of soliloquy Emilia protests too much, so that deletion is equally possible, by Knight or any one else through whose hands the manuscript passed in the playhouse. On stylistic and linguistic grounds IV.iii appears to be homogeneous, and it is probable that the first exit for the Jailer's Daughter shows deletion to avert censorship.

Except for the passage cited above, the play shows no evidence of general censorship, although, as Greg warned, "the presence or absence of profanity cannot be regarded as affording altogether reliable evidence respecting the nature and date of the manuscript that served as copy for a printed text" (First Folio, p. 152). The Two Noble Kinsmen is not a heavily profane play, but it does contain a few oaths with which Sir Henry Herbert might have taken exception—e.g., "Marry," "god's lyd," "Faith," numerous invocations of God and Gods, and even one affirmation by "wine and bread"—and if we may judge from his practice in Barnavelt, Buc would have objected to the courtier reference in IV.iii. For what it is worth, then, such evidence points to an unrevised draft or at least an unexpurgated transcript; it also points back to the date of original composition rather than revision or transcription for the revival of 1625-26. One might also note that Knight's copy of The Honest Man's Fortune, made within a year or so of the Kinsmen's revival, shows fairly extensive purgation, particularly of such obvious expletives as "God", in which the Kinsmen abounds.

There is one other area of evidence which may have some bearing on the nature of the copy. In The Shakespeare First Folio, Sir Walter Greg wrote, "if we find any considerable number of eccentric or archaic spellings in a print, the likelihood is great that it was set from the author's own manuscript and not from a scribal copy" (p. 148). I should not want to say that a "considerable number" of eccentric spellings occur in The Two Noble Kinsmen, but I have the impression that the texture is more archaic in the scenes ascribed to Shakespeare. The list which follows is drawn from his scenes and also represents spellings which are paralleled in various texts of his acknowledged plays, thought to have been printed from authorial manuscripts: angle (i.e., angel), Asprayes, boudge, cease (i.e., seize), Cizard, cizd, mervaile, right (i.e., rite), and Wrinching (i.e., rinsing). In addition, I

conjecture the anomalous form, a eleven, at I.iii.54, to be a compositorial misreading (or "normalization") of an original copy spelling, a leven. Spellings of eleven with a occur, of course, in a number of the Shakespeare good quartos and in Addition D of Sir Thomas More. In the Fletcherian scenes I note only stoa (i.e., stow), aborne (i.e., auburn), and take (i.e., tack, in both Prologue and text). I had hoped that a tabulation of various morphological groups according to the authorial divisions might be possible, but any significant variation which might have once appeared seems to have been almost hopelessly obscured by the compositorial pattern. In several classes the distribution of variant forms is disproportionate for one author or another but not sufficiently so to permit certainty that one of the compositors has not intervened. For example, the ratio of spellings in -nck and -nk in the non-Fletcherian scenes is 8:16, but in the Fletcherian scenes, 3:48; however, in the second half of the text only one -nck spelling occurs. For various classes of elisions, the patterns may be authorial; of elisions of e in verbals ending -en, the distribution between non-Fletcherian and Fletcherian scenes is six to three; of e in verbals ending -er, eight to zero; of e in various nouns and adjectives with -er-, nine to three, all instances in Fletcherian scenes being wondrous; and finally, the elided spellings medcine and heav'nly both appear in non-Fletcherian scenes.

Save for the diminished ye's in Fletcher's scenes, the accumulation of evidence is, it seems to me, strong for foul papers. The hypothesis of mixed copy described above offers one resolution to the apparent dilemma; otherwise, we may choose between annotated foul papers and an annotated intermediate transcript when we pay our money. There are arguments on both sides. But even if it be concluded that a scribal transcript has intervened between the author's original draft and the printed text, the preservation in the quarto of so many signs of jointure and divided authorship would still warrant the inferences which have been drawn concerning the authors' method of collaboration. Not only the authorial tests but bibliographical evidence indicates that they divided their work by scenes (or structural dramatic units, since Q's II.i-ii really constitute a single scene) and wrote their shares concurrently, for, as suggested above, Fletcher would hardly have called his tower guard "Keeper" if he had just read his collaborator's scenes and there found the title "Iailor". Presumably they worked from a scenario or some sort of outline; analysis of the plot shows that each must have understood the essential outlines of the whole play from the very beginning. The Jailer-Keeper confusion is not overly serious, but

it is just the kind of discrepancy which might have occurred had the authors divided the scenes between themselves and then begun writing their respective shares simultaneously but in no very close collaboration. Then, while Fletcher seems to have given the text a final reworking, his revisory attentions appear haphazard and incomplete: they amounted only to the addition of some new matter, with a number of irregularities, some quite obvious, left untouched. It is quite possible that the play had to be written and put together under some pressure, so that it might be on the stage before the festivities attendant upon the wedding of Elizabeth and Frederick were over—very likely while the morris dance it borrowed from Beaumont's Masque was still fresh. The great difficulty in the acceptance of Shakespeare's authorship is the inferiority of the scenes given to him, especially in characterization and thought, in comparison with his known work. Attention has already been called to the rather marked difference as to state of finish between the Shakespearean and Fletcherian parts of the play. The textual tangles and general impression of carelessness, the evidence that the authors did not work in especially close collaboration, and the evidence that it was Fletcher who gave Shakespeare's scenes their final touches, all suggest that Shakespeare simply did not give the play the thought and attention which might have satisfied those who reject the attribution of the quarto title-page.[29] "Foul papers," therefore, may lie at the heart of the authorial problem.