Collaboration (not necessarily unhappy) strikes me as somewhat closer to the truth than Dyce’s theory of consecutive composition. The play’s structure, with its almost complete separation of main plot and subplot and its large number of soliloquies, seems designed to facilitate collaboration between two people who did not expect to have much opportunity to talk about the work in progress. Hoy’s division of the play (89) gives Acts 1 and 5 (except 5.1.1–33 and 5.2) wholly to Shakespeare, as well as 2.1 and 3.1–2. Other scholars are doubtful about the authorship of 1.4–5, and 3.2, but Hope (86) claims both as Shakespeare’s. Thomas Horton’s work on function words, which otherwise confirms most of the traditional divisions, adds 2.3 and 4.3 to the scenes that Shakespeare might have written (Horton, 329–30). Most of these distinctions are made on a linguistic basis and (because the textual samples are so small) cannot easily take account of minor adjustments and revisions that one writer may have made to another’s work. A close study of the play from the dramaturgical point of view, however, suggests that a number of such adjustments were in fact made.

  Because of the state of the 1634 Quarto, we have unusually interesting evidence of the play in process – not only during the collaboration, but also in later revivals. For example, the first stage direction of 1.1, in the quarto, begins ‘Enter Hymen’, but then adds, between colons, ‘:a Boy, in a white Robe before singing, and strewing Flowres:’, which surely means that someone decided that it would be more effective to bring on the boy singer, rather than Hymen, at the head of the wedding procession. Perhaps it even means that the song itself was a late addition to the plans for the opening scene. While the insertion might have been made by the author himself in the process of composition, the quarto punctuation, with colons separating the different characters, corresponds to the manner of Edward Knight, the book-keeper for the King’s Men from 1625 to 1633. Not only is it unusual to begin a play with a song, but the elaborate procession, interrupted before anyone has spoken, must inevitably be mystifying to spectators unfamiliar with the Chaucerian source. The identity of the characters has to be deduced from references to Thebes, Creon and Hercules, until the Second Queen finally addresses Hippolyta by name at 1.1.77. There is similar evidence of the book-keeper’s hand at the beginning of 1.4, another scene which looks as if it had been slightly revised. Though the speech in which Theseus urges everyone to try to save Palamon and Arcite (1.4.28–47) is difficult in a characteristically ‘Shakespearean’ way, it also contains five examples of a ‘Fletcherian’ contraction, ’em for them. Perhaps these resulted from Fletcher’s having transcribed Shakespeare’s writing at this point.

  It is in discrepancies between the first two scenes of Act 2 that the change from one writer to the other shows most clearly. Scene 2 begins with the entrance of Palamon and Arcite, yet 2.1 gives them no exit from the upper level where they are first seen. In 2.1 the Jailer’s Daughter talks about what the kinsmen have been saying to each other; in 2.2 they seem to be meeting for the first time since the battle. These problems are easily solved, or not noticed, in the theatre. They are interesting because they indicate that Fletcher wrote at least some scenes before he saw what Shakespeare had written. Later in 2.2 Fletcher refers to the ‘Keeper’ of the prison, both in stage directions and in dialogue. In the rest of the play, he calls him ‘Jailer’ – presumably because he discovered that this was what Shakespeare had already called him in 2.1. Another example of possible Fletcherian revision occurs in 2.5, when the disguised Arcite has just been taken into service at court. Pirithous offers to supply him with a horse, but ‘a rough one’, for the afternoon; then Theseus invites Arcite to join the court in the next day’s May celebrations and hints that Emilia ought to provide a horse for her servant. Two examples of foreshadowing seem excessive. I suspect that Pirithous’ line was meant for deletion: Fletcher, having read the beginning of Act 3 where Arcite says that Emilia has given him two horses (and perhaps also the speech in Act 5 which shows that one of these horses was to kill Arcite), decided to alter the earlier scene. The two dramatists may, indeed, have had different ideas about the disguised Arcite’s relationship with Emilia, a subject ignored by Chaucer but slightly developed in his main source, Boccaccio’s Teseida. At the beginning of Shakespeare’s 3.1 Arcite is exulting in Emilia’s favour, but in Fletcher’s 2.5 she is much more detached and in 3.5 (where his presence is admittedly problematic – see the note on 3.5.93 SD) she pays no attention to him at all.

  All of Act 2 after the first scene is attributed to Fletcher. Between 2.4 and 3.5 the scenes written for the subplot heroine, the Jailer’s Daughter, are too short to allow for successful tests of authorship. The sheer number of her soliloquies suggests that they were needed to separate scenes in the main plot which were expected to involve a large number of actors. Kristian Smidt has suggested of Hamlet that Shakespeare may have written the soliloquies ‘without committing himself at first to their exact positioning’ (83). Something similar may explain the misnumbering of scenes in the 1634 edition of The Two Noble Kinsmen; Act 2 has two scenes numbered 4, while in Act 3 scenes 5 and 6 are numbered 6 and 7. Perhaps the soliloquies existed on separate sheets and were added or transposed to suit the technical demands of a story which, as F. W. Brownlow points out (208–9), has a beginning and end but not much in the middle. Still, the dramatists have linked the action of this part of the play with references forward and backward in time which give an illusion of consistency (Palamon even anticipates objections when he comments in 3.6 on the surprising rapidity with which he has recovered from his half-starved state in 3.1).

  Closer examination shows some confusion. Giorgio Melchiori, who gives a chronology of the play’s action, points out that the Jailer’s Daughter takes several days to go through the same experiences that the kinsmen are having in one (6.975–6). Inconsistency with regard to the time-scheme is not unusual in drama, but it reinforces my suspicion that some of the writing may have been done without much sense of context. Both Shakespeare and Fletcher, in their unaided plays, use abrupt scene openings that imply a previous thought or discussion on the part of their characters, as in Brutus’ ‘It must be by his death’ (JC 2.1.1) or Fletcher’s ‘And as I told your worship’ (The Woman’s Prize, Bowers, 4: 4.2.1) or ‘We dare not hazard it’ (Bonduca, Bowers, 4: 3.2.1). In several scenes of The Two Noble Kinsmen, by contrast, the characters speak out of a dramatic vacuum, as if the writer was making a new beginning. As N. W. Bawcutt says (33), the author of the end of 3.6 seems deliberately to be following the pattern of repeated kneelings in Shakespeare’s 1.1. Yet he does not seem to have known what was going to happen in 3.5. Palamon’s words at the beginning of 3.6 recapitulate what was stated at the end of 3.3 and could have been taken for granted. Such vagueness would be surprising if this part of the scene were by Fletcher, the supposed author of the two previous ones. Perhaps he was uncertain about the length and nature of the morris-dance sequence that was going to occur in 3.5. Or perhaps he did not write 3.5 at all; some have wondered whether Beaumont, the author of the masque that was being alluded to in this scene, might have been invited to connect it himself to the play. Yet, though at first sight the morris-dance scene seems irrelevant (some modern productions have cut it altogether), Fletcher has taken trouble to integrate it into the rest of the play. The discussion of plans for the dance in 2.3 is closely linked with the villagers’ encounter with the newly released Arcite. The pursuit of the tanner’s daughter by Gerald, the schoolmaster organizer of the dance, is hinted at in both 2.5 and 3.5. Palamon threatens to shake his gyves and make a ‘new morris’ (2.2.276), while the Daughter in her madness talks of dancing ‘an antic’ in front of Theseus in order to beg Palamon’s life (4.1.75).

  Though both 3.6 and 4.1 are thought to be by Fletcher, there is some overlap and inconsistency between them. At the beginning of 4.1 the Jailer’s friends recapitulate some events from 3.6 but also mention others that did not happen there; these have to do with the Daughter’s part in the plot, and attempt to tie it more closely to the main one. The reference at 4.1.21–4 to Palamon’s giving money for the Daughter’s dowry seems inconsistent with his similar gift in 5.4, a Shakespearean scene, when Palamon appears to hear of her marriage for the first time. A possible explanation is that the author of 4.1 was still expecting someone else (Shakespeare?) to write 3.6 and ended up having to write it himself, possibly on the basis of Shakespearean material.

  Scenes 2 and 3 of Act 4 most clearly indicate some confusion in the text. Scene 2 marks an entry for Emilia who has not previously exited; 4.3 gives an exit to the Jailer’s Daughter, but it is followed almost at once by a speech for her, with no direction for her reentry. Such errors are not in themselves unusual in dramatic texts, but 4.2 has other peculiar features. It opens with Emilia’s soliloquy over the portraits of Palamon and Arcite, and is at once followed by a direction for her entrance with an anonymous Gentleman. Both the Gentleman and Theseus (who enters ten lines later saying, ‘Bring ’em in’) seem to expect the immediate entrance of Palamon, Arcite and their knights. Instead, two messengers enter, though only one of them speaks. The idea of bringing in the knights is now forgotten, but, in answer to Theseus’ abrupt questions, the messenger and Pirithous describe three of the knights at length and Theseus, inspired by their speeches, rushes off to see them. The descriptions might perhaps have been appreciated as an act of homage to Chaucer (though Chaucer describes only two knights, this is otherwise the passage most obviously derived from The Knight’s Tale), but in the scene as a whole they seem a clumsy insertion. Emilia’s soliloquy, as Waith notes (Oxf1, 21–2), is very similar to the one Shakespeare gives her in 5.3. The latter is clearly essential, given the decision to place the tournament off stage, and the resemblance between the two speeches is not altogether surprising, since most of Emilia’s lines in the second half of the play are about her inability to choose between the two men and her grief at being the cause of so much bloodshed. The confusion over her entrance suggests that the first soliloquy was either cut or added at a late stage. Montgomery and Taylor, in the Textual Companion, consider the latter possibility slightly more likely. I agree; it seems to me that at some point 4.2 was meant to depict the return of Palamon and Arcite, entering in procession with their knights. This procession was later moved to the beginning of 5.1, and Emilia’s soliloquy was either moved from a position later in the play or written to help fill the gap between the men’s departure and return.

  The authorship of 4.3 has been debated; some scholars argue that the Jailer’s Daughter, after 3.2, is entirely Fletcher’s creation, others that Fletcher would never have written an entire scene, even a mad scene, in prose. A change of author at the start of 4.3 may explain why the Daughter knows something about a schoolmaster but calls him Giraldo rather than Gerald, his name in 3.5. If so, this might be the one point in the play to show Shakespeare accommodating to his colleague, rather than vice versa. However, there may be other reasons for the name change, including simple playfulness. Moreover, the author of 5.2 clearly knew the end of 4.3, to which the opening lines refer – but, again, with an air of recapitulation that suggests someone writing out of context. While a case may be made for either dramatist’s authorship of these scenes, it seems to me that both were probably written at a time when 5.1 was still not fully worked out.

  The series of addresses to the gods in 5.1 is probably the most obviously Shakespearean scene of the play, but it looks as if Shakespeare, when writing it, was unclear how it would be staged. The impressive exchange between Palamon and Arcite that starts at line 18 conveys no obvious sense of location. On the other hand, the lines given to Theseus at the beginning of the scene –

  Now let ’em enter…

  Let the temples

  Burn bright with sacred fires and the altars

  In hallowed clouds commend their swelling incense

  To those above us.

  (5.1.1, 1–5)

  – prescribe a spectacular processional entry, with candles, incense and an altar. I agree with those who think that these lines are by Fletcher; replacing whatever he had originally planned for 4.2, they provide a transition to and a setting for the speeches that follow.

  By contrast with the pageantry of 5.1 the rest of the play is visually rather disappointing. After stressing, in 4.2, 5.2 and 5.3, everyone’s breathless eagerness to see the fight, the dramatists deliberately deprive their audience of it, conveying the tournament through offstage sounds and Arcite’s accident through the messenger speech of Pirithous. There is neither funeral procession nor funeral pyre. Though the sense of anticlimax may be deliberate, something more elaborate may originally have been planned. The transition between scenes 3 and 4 is remarkably awkward. Theseus orders the execution of Palamon and his knights, adding ‘Let it here be done. / The scene’s not for our seeing; go we hence’ (5.3.133–4). As a way of dealing with the death of a major character, these lines are so perfunctory as to be an insult to the audience as well as to Palamon, though their very casualness may be a clue that no such death will occur. Their main purpose seems to be to maintain continuous action from scene 3 to scene 4. I suspect that they were interpolated; they contradict what seems to have been the original intention (presumably Shakespeare’s) of making a break between the two scenes. Why did he want the break? Perhaps in order to suggest a sufficient lapse of time to account for all the offstage events that precede Pirithous’ entrance in 5.4. Perhaps, also, he wanted an actual scene change. In 5.4 Theseus comments, ‘In this place first you fought’ (5.4.99), and at 3.6.292–3, supposedly located in that very place, he had declared his intention to ‘plant a pyramid’ (or pillar) at the tournament site – a locational pattern that derives from Chaucer. It seems likely that at some point Shakespeare envisaged the final scene taking place on the same spot as 3.6, possibly marked by the presence of the pyramid itself, but that he or someone else finally decided to avoid a scene change so late in the play.

  Such minor inconsistencies bear out the suggestion that there may have been a change of plan in the course of writing (Proudfoot, ‘New’, 260). The parallel wedding and victory processions in 1.1 and 1.4 and parallel songs of wedding and funeral in 1.1 and 1.5 (probably composed specially for the play) make Act 1 look, as Waith says (Oxf1, 63), like a prologue. The triumphal procession for Theseus in 1.4 and the ‘funeral solemnity’ of 1.5 (paralleling the marriage song at the beginning) create a strong sense of closure, particularly as the queens, who dominate the symmetrical opening and closing scenes, never appear again. Proudfoot suggests that Shakespeare had completed Act 1 at the time of the Globe fire and that the need for a new Blackfriars play for the autumn led to the co-opting of Fletcher and perhaps a third dramatist, who finished the play hurriedly (‘New’, 220).

  My own hypothesis is that the two dramatists began writing concurrently, but that Fletcher constructed the final draft. In 1.4, 2.2, possibly 2.5, and 5.1, he seems to have been working on, or in the light of, Shakespearean material; nothing suggests that Shakespeare was ever working on Fletcher’s. Whereas it is claimed that in Henry VIII the dramatists’ influence on each other’s style was reciprocal (Tarlinskaja, 217), in The Two Noble Kinsmen the influence appears to have been one-way. The most likely explanation is that Shakespeare was no longer in London when the play was put into its final form. Irwin Smith has suggested that Shakespeare’s purchase of property at the Blackfriars in March 1613 put him in a difficult financial situation when the Globe burned down in June and he, as a sharer, became liable for a substantial sum towards the rebuilding. Since he did not sell the Gatehouse, which is mentioned in his will, Smith argues that he must instead have sold his shares in the King’s Men, which are not mentioned there (Blackfriars, 252). This might have led to his finishing his part of the collaboration quickly – perhaps even before it was clear how soon the company would be able to replace any costumes and properties lost in the fire – leaving the play for Fletcher to put together once it had become clear how elaborately it could be staged. If a third hand was involved, as Richard Proudfoot suggests, the most likely candidate would be either Beaumont himself or Nathan Field, the actor-dramatist with whom both Beaumont and Fletcher had already collaborated. Though I do not agree with Donald Hedrick’s suggestion that Field was the author of the entire non-Fletcherian portion of the play, I am struck by a curious resemblance between Act 1 of The Two Noble Kinsmen and the first of the Four Plays in One that Fletcher wrote with Field, probably for the Lady Elizabeth’s Men, some time between 1612 and 1615. This sequence of short plays (the triumphs of honour, love, death and time respectively) is based loosely on the famous Triumphs of Petrarch, where processions in honour of love, chastity, death, fame and time succeed each other, each accompanied by historical and literary examples of the abstraction they commemorate. The first of the Four Plays, ‘The Triumph of Honour’, is ascribed to Field and based on the same story as Chaucer’s Franklin’s Tale. Like it, Act 1 of The Two Noble Kinsmen represents the triumph of continence over lust in a ruler. Theseus’ entry as ‘victor’ in 1.4 could have involved the same triumphal car as at the end of each of the Four Plays. The latter (at least, in the part written by Field) are also plays within a play, presented before a fictitious newly married king and queen of Portugal – a fact that, as Suzanne Gossett notes (103), makes it possible, though not provable, that they belong to the period following the royal wedding of 1613. Since Field had in any case worked with Fletcher on The Honest Man’s Fortune in that year, he would probably have known something about his colleague’s other project, and of course the resemblances just noted might result simply from the influence of The Two Noble Kinsmen on the Four Plays. But it seems to me likely that Field had something to do with the Kinsmen. Even if he was only one of the original actors, he would probably have had an important effect on its writing.