Both Shakespeare and Fletcher also have a trick, possibly learned from Donne, of letting the stress of the verse line dictate its meaning. In Palamon’s speech to Venus, the metre forces the reading ‘I / Believed it was his, for she swore it was’ (5.1.116–17; my emphasis). Fletcher may have been doing the same thing at 3.5.78, where Gerald’s exchange with the Daughter is lineated like this in the quarto:

  SCHOOLMASTER

  And are you mad good woman?

  DAUGHTER

  I would be sorry else.

  Give me your hand.

  With a different lineation –

  SCHOOLMASTER

  And are

  You mad, good woman?

  DAUGHTER

  I would be sorry else.

  – the speaker would be forced to stress are. I suspect that this is how Fletcher actually heard the lines, but I have not felt sure enough to alter them.

  Seward’s 1750 edition was based on the belief that all the scenes printed as verse really were meant to be verse, and he did his best to make them so. He also attempted to regularize the metre of other scenes by a process of expansion, elision and contraction. For instance, this is his reading of the Wooer’s first speech in 5.2:

  Oh, very much: The maids that kept her company

  Have half persuaded her that I’m Palamon;

  Within this half-hour she came smiling to me,

  And ask’d me what I’d eat, and when I’d kiss her:

  I told her presently, and kiss’d her twice.

  He has been ridiculed for it ever since, though, as Bertram points out (27), his eighteenth- and early-nineteenth-century successors kept many of his changes and the lineation that resulted from them. Seward had a tin ear, but he was essentially following the practice of the 1679 and 1711 editors, who likewise modernized the contractions into a form familiar to them. Seventeenth-century texts frequently elide adjacent vowels, as in modern French and Italian. The speed of Fletcher’s dialogue suggests that many words were meant to be shortened (for example ‘I’d’ for ‘I would’), even where they have not been so indicated in the text. Modern British pronunciation elides many vowel sounds in words like ‘every’ and ‘flower’; in this edition such words are written in full, on the assumption that the reader does not need to be told that they are pronounced ‘ev’ry’ and ‘flow’r’. It is harder to be sure about pronunciations that add extra syllables to a line. For instance, ‘r’ is often rolled and treated like an extra syllable (bonefier is sometimes found as a spelling of bonfire); Fletcher apparently pronounces ‘sire’ with two syllables. Reaching a consensus on scansion, and hence on lineation, is important mainly because notes and concordances are keyed to line numbers. With computer analysis becoming increasingly important, the line will probably become less significant as a measurement; for instance, the most exhaustive Shakespeare concordances, those of Marvin Spevack, give not only the number of lines each character speaks but also the number of words and the number of speeches.

  Stage directions

  The Two Noble Kinsmen is unique among printed texts of Shakespeare in that it includes some ‘anticipatory directions’ (examples can be seen in the left margin on pp. 14 (Fig. 15) and 46 (Fig. 16)) of a type that are often found in theatre promptbooks. They are obviously directed to people backstage at the playhouse, not to a hypothetical reader. The one on p. 46 is in fact the last such direction to appear in the quarto, though it is unlikely to have been the last one to appear in the manuscript. This fact suggests that the compositors (perhaps inexperienced at setting playtexts) discovered that these directions were not meant for printing and discontinued the practice.

  The marginal directions are set in roman type rather than italics; the one opposite Emilia’s speech in 1.3 also makes use, rather oddly, of colons. Several scholars (Greg, 39 n.; Waller, ‘Printer’s’, 64–5) have noticed the resemblance of this punctuation to that of other manuscript promptbooks known to be in the handwriting of Edward Knight, the book-keeper for the King’s Men c. 1625–33. The unusual typeface is probably the compositors’ attempt to re-create the large print which characterizes such directions in play manuscripts. The distinctive colons also occur in non-marginal stage directions, like the one on p. 15 at the beginning of 1.4, which suggests that some of them may have been added by Knight (Werstine, 21). ‘Enter Theseus’, with its capital ‘E’, looks like the original beginning of this direction; the prompter would then have inserted specific directions for offstage sounds.

  From the direction on p. 14, we learn that while Emilia and Hippolyta were playing their intimate scene, six major characters and at least six minor ones were preparing for a carefully synchronized entrance – Theseus, in a triumphal procession, meeting the three queens. I have assumed that the marginal directions represent the point at which the stage manager began to assemble his actors and musicians (in this case, allowing some thirty-five lines, plus the time taken by the offstage sound effects, before their entrance). But it is also possible (as Richard Proudfoot suggests) that the word ‘ready’ means that by that point in the text the performers (and, as at 1.4.26–7 and 3.5.65–6, the hearses, chairs and stools) should actually be in position and ready to enter. The distinction is important for anyone trying to visualize the conditions of performance. An early line-up in 1.3 might have been necessary if, as Masefield suggested (193), the offstage actors were expected to join the company’s musicians in creating the sounds of the battle that is supposed to take place between 1.3 and 1.4. The direction for ‘Cornets’ at the end of 1.3 obviously belongs with the beginning of 1.4 and shows how the sound of the enemy’s retreat was to be created.

  Waller has argued that some features of the stage directions, like the absence of an entrance for Palamon and Arcite at the beginning of 1.4 and the vagueness about the number of ‘wenches’ at the opening of 3.5, indicate that the play was printed from an authorial manuscript, since a promptbook necessarily had to be more exact. But he himself notes that in one other play manuscript annotated by Knight, Massinger’s Believe as You List, the author’s original stage direction is deleted and replaced with a larger one in the margin six lines earlier (Waller, ‘Printer’s’, 64). It is possible that the same thing happened with the original direction for the entry of Palamon and Arcite (whether they were brought on at the beginning of 1.4 or just before Theseus notices their presence). The prompter’s point of view is a backstage one; he can get everyone on stage, but has to trust them to get themselves off again, which is one reason why exit directions are so often missing from printed plays of this period.

  Waller’s most convincing example of an authorial type of stage direction is the one at the beginning of 3.5, which asks for ‘2 or 3 wenches’. Certainly, no prompter would allow this kind of vagueness. But, as has been shown above, the dance was probably a special case: it was borrowed or adapted from Beaumont’s Masque of the Inner Temple and Gray’s Inn and may have been rehearsed separately. If the ‘wenches’ were engaged only for the dance, Fletcher may not have known what number would be involved.

  The odd wording of the marginal stage direction in Figure 16, 3.5.135 (the cue for the dance), must represent a misreading of whatever was crowded into the margin. The compositors omit a speech prefix for Gerald, giving the whole speech, improbably, to Pirithous. They seem not even to have realized that the first and second lines of Gerald’s speech would have been separated by a good deal of stage activity. Montgomery (Oxf) suggests that School, an abbreviated speech-prefix, got mixed up with the rest of the marginal note, which ought to read ‘Knock for the Dance’ (TxC, 632). But the direction remains puzzling: did knocking replace, or supplement, Gerald’s throwing up of his cap as an entrance cue? Leech suggests (179) that the prompter had forgotten the visual cue mentioned in the text and inserted a sound cue instead and Leslie Thomson has found evidence of ‘Knock Act’ being used in this way (188). From Beaumont’s descriptive pamphlet, we know that the original effect depended on synchronized spontaneity, ‘the Men issuing out of one side of the Boscage, and the Woemen from the other’. Perhaps the dancers (supposedly hidden in the bushes but really behind the doors at the back of the stage) were unable to see Gerald and had to be cued for their entrance by someone backstage. Or perhaps the knock was the cue for the musicians, who were dispersed (see 3.5.33) and might have had difficulty coordinating their first note. Some modern productions have created business to explain why Gerald is finally reduced to calling for his dancers to enter: either he discovers that he has lost his cap – stolen by the Bavian (Berkeley) or thrown away in a fit of exasperation (Ashland), or else the dancers are struck with stage-fright and need extra cajoling to bring them on. The ‘Enter’ which occurs in the marginal direction may refer to the entry of the dancers or to Gerald’s re-entry after the dance.

  I have already noted the difficulty of clarifying the direction of the dialogue at the end of this scene. Playtexts from this period rarely indicate the person to whom a line is addressed, and The Two Noble Kinsmen never does so, though a change in the addressee is sometimes indicated by parentheses. All stage directions of this type can thus be assumed to be editorial and I have not credited them in the notes.

  This example shows how editing a text can also involve imagining it in performance. One then has to decide whether to add stage directions to help the reader imagine it too. Editorial practice in this respect has changed over the years and at present the tendency is to be generous with such directions. The difficulty is that there are many points at which any one of several things might be happening. To choose only one possibility, and dignify it with the status of a stage direction, is to risk closing off interpretation for both readers and performers. I have therefore chosen the more cumbersome method of adding as few directions as possible in the text itself, using the notes instead for discussion of the most likely alternatives.

  Emendations of difficult passages

  Emendation was common in the earliest editions; editors with no knowledge of Renaissance handwriting or printing conventions recklessly suggested alternatives to anything they found at all difficult to understand. The reaction against this attitude was a conservative approach that has itself only recently been challenged, particularly in the new Oxford Complete Works of Shakespeare. On the whole, the practice of most editors is still to make their text look as much as possible like a modern one in everything but the actual words. But many critics are distinctly suspicious of the Shakespeare ‘not of an age but for all time’ who emerges from this treatment, especially when he is contrasted with his contemporaries, often available only in old-spelling texts that firmly anchor them in a particular historical period. The compromise made in the otherwise excellent Riverside Shakespeare, which retains a few archaic spellings with no special justification, has not been generally accepted. Now that modern technology makes the reproduction of original texts relatively easy, it is possible to show something of the process (as here) and thus avoid presenting an emendation as if it were self-evidently the correct reading.

  Both the openings reproduced here contain, along with the minor technical problems already mentioned, substantial textual ones. It is regrettable that Emilia’s speech in 1.3 about her friendship with Flavina, the most frequently anthologized in the play, is also one of the most difficult. Lines 72–3 (her affections … were); the hyphenated fury-innocent of line 79; the meaning of old importments bastard (80) and the question of whether to emend sex individual (82), do not spoil the effect as much as one might expect, since, as often in the play, the general sense is clearer than the precise grammatical construction. In attempting to clarify such a passage, the editor needs to decide whether its confusion results from misreading by the compositors or confused writing by the author. The various alternatives suggested in the textual notes mostly depend on the first assumption, while my very conservative reading of the text depends on the second. The various emendations normally ignore the hyphen in fury-innocent, yet this is not a feature likely to be misread; Bowers points out (8.270) that ‘the hyphen suggests some compound in the manuscript’. In looking again at Montaigne, an obvious influence on this scene, I became convinced that fury was not a misreading, however difficult it might be to make sense of its use in the construction as a whole. An eighteenth-century commentator on Seward’s edition, Benjamin Heath, decided, after analysing this passage in some detail, to leave it as it was: ‘Patching up a text, so as to give it the glimmering of a meaning, is so far from being of service, that it tends only, by removing all suspicion of a Mistake, to perpetuate the corruption’ (Heath, 88). I have followed his example.

  The difficulties with pp. 46–7 (Fig. 16) are of another kind. Without more information, it is impossible to know what is meant by the Schoolmaster’s references to Morr and Is and how these were glewd together at line 118. Bibliographical knowledge is valuable here chiefly in showing that the dash that follows both nouns is often used to indicate some kind of stage business. The nature of this business, the equally mysterious direction following line 135, and what exactly happens at the end when Theseus and Pirithous give presents to the Schoolmaster, are matters for the director rather than the bibliographer, and can benefit from research into the context of performance such as I have outlined in Appendices 4 and 5.

  The manuscript copy

  The Oxford editors in 1987 (TxC, 625–9) believed that the manuscript behind the quarto of The Two Noble Kinsmen was either ‘a revised holograph’ or ‘a scribal transcript to which revisions were made’; in either case, it ‘was annotated by a bookkeeper, c. 1613’, and ‘further annotated, probably by Edward Knight, in preparation for a revival, c. 1625–6’ (629). In subsequent work (Taylor & Jowett, 241) they have endorsed the second alternative, which also seems to me the more likely of the two. More work on the output of the printer, Thomas Cotes – how much did he normally revise his authors and impose a house style? – will perhaps settle the question for good.

  The quarto is a well-printed text by the standards of the period, but with one or two anomalies, which may be evidence of revision. Other errors were probably the result of difficulty in reading the original manuscript. The fact that there is only one authoritative edition limits editorial disagreement to relatively minor emendations and the addition of stage directions, with the further debate as to whether to treat the prison scene in Act 2 as continuous with 2.1. Clearly, the task for a modern editor is to regularize the presentation of the text, so that variations of format are not mistaken for significant symbols; to correct obvious errors, identify others that may not be immediately apparent, and suggest alternative possibilities where there is no clear solution; and to supplement incomplete information. The resulting text must be as clear as possible, but should retain original features even where these are confusing. Each editor has been able to make some contribution to the play, but there is still much to do. In particular, the increasing availability of computer databases should make it possible to write with much more certainty about the authorship question.

  Note on personifications

  The following have been silently capitalized because they are clearly envisaged as personifications in the text: Prologue 7 Modesty; 1.1.13 Nature; 1.2.19, 23 Peace, 67 Chance; 1.3.40 Death; 1.5.8. Pleasure; 2.1.27 Fame; 2.2.38, 57 Fortune, 104 Time; 2.5.29 Honour; 2.6.8 Love; 3.1.15 Lady Fortune; 4.2.7 Nature, 14, 42 Love, 21 Fame, 21 Honour, 52 Fancy, 108 Victory; 5.4.20, 112 Fortune.

  For a note on the editions cited.

  ADDITIONS AND RECONSIDERATIONS

  It has taken roughly a generation, but collaborative plays like The Two Noble Kinsmen seem finally to have become a focus of interest in their own right, instead of an embarrassment to studies of an author’s ‘development’. This section will trace a process that had not yet become apparent when, some twenty years ago, I finished editing this play. That process had already begun ten years earlier. The year of Barry Kyle’s important RSC production at Stratford, 1986, also saw the publication of the revolutionary Oxford Shakespeare (ed. Stanley Wells, Gary Taylor, John Jowett and William Montgomery), which argued that several plays once thought to be by Shakespeare alone were the products of collaboration and/or revision. The play’s London transfer in 1987 coincided with the appearance of the Textual Companion to that edition, including the tables of computer-assisted analysis on which the editors had based their hypotheses about attribution.

  What followed was increased scholarly interest in editing, while the recognition that some textual difficulties might be explicable by the presence of more writers than one led to a burgeoning of authorship studies, a field once almost monopolized by those wishing to prove that someone other than Shakespeare wrote Shakespeare. Then came new publications attempting to theorize the subject of collaboration in the special case of the theatre, which by its nature is usually a complex operation involving many hands, even when only one writer is involved. Just how broadly the concept of collaboration can be stretched has itself become the subject of debate. Jeffrey Masten’s Textual Intercourse (1997) used it as a metaphor with which to query the value of the traditional focus on the individual author’s personality as a way of understanding a literary work. Douglas Brooks has argued that the practice of designating collaborative authorship by placing brackets around the authors’ names, which he traces from its first use in 1617 to the Restoration, through the 1634 Two Noble Kinsmen and the Beaumont and Fletcher Folio of 1647, reflects publishers’ desire to contain and simplify a process that escaped the definition of authorship as the seventeenth century increasingly came to understand it (157–72). Brian Vickers’ Shakespeare, Co-Author (2002), a comprehensive survey of the evidence and scholarship on the collaborative plays, argues against the ‘death of the author’ approach, insisting in an appendix that authors from classical times onward had a clear sense of their works as their own property. (For a fuller discussion of this controversy, see McMullan’s excellent account, Late, 233–45.) Given the current fascination with the ‘authorship question’, John Jowett felt it necessary to point out, in his chapter on ‘Shakespeare as Collaborator’, in Shakespeare Beyond Doubt, that to say that Shakespeare sometimes collaborated, and thus did not write every word of a play, is not the same thing as saying that he did not write any of the plays attributed to him. In fact, ‘the evidence of dramatists other than Shakespeare within the Shakespeare canon is also, by the same token, evidence that the residual body of writing (maybe ninety per cent of the canon) is by Shakespeare’ (97).