As is apparent from the illustrations, the text has, in a sense, been edited already. That is, someone has regularized its format by indenting the speech prefixes, placing stage directions flush with the right margin, and using roman letters for scene numbers but italics for proper names and foreign words in speech prefixes and dialogue. A few play manuscripts from this period have survived, so we know that much of this standardization (particularly the typographical differentiation) was probably done in the playhouse rather than at the printer’s shop. The normal playtext was written on a page folded to make three columns: one for speakers’ names, two to three for the dialogue, with stage directions on the right. The compositors who set up the type at the printing house followed much the same procedure except that they were generally concerned with saving paper; hence their speech prefixes are indented rather than placed in the margin. One always has to be aware that a line lacking final punctuation or using contracted word-forms may be the result of this space-saving instinct. On the two occasions where a line of dialogue was too long for the margin (Fig. 15, 1.4.4 and Fig. 16, 3.5.149), the compositors followed the standard practice of setting a ‘turned-over line’. (All line references are to the edited text.)

  On the other hand, some things have not been regularized. In 1.3, the name of Emilia’s childhood friend is spelled one way in line 54, another in line 84. Pirithous of line 55 is Pirothous in 95. A speech prefix for the Schoolmaster has been omitted at line 3.5.135, probably because of confusion over the positioning of the stage directions; it looks as if someone at the printing house, misled by the third and fourth lines of his next speech, took it for a song with a refrain and italicized it accordingly. In some places necessary punctuation appears to have been omitted: after Emilia’s ‘Yes’ at 1.3.54, at the end of 1.3.65, and probably after 3.5.117; a question mark also seems needed at the end of 3.5.146. Beside these comparatively minor points, what would most strike a reader accustomed to modern edited texts is the virtual absence of stage directions. In Figure 15, Theseus’ ‘What are those?’ (1.4.13) is unexplained; in Figure 16 there is no indication as to what is meant by the Schoolmaster’s references to ‘this Machine’ (3.5.112), what goes on after he has called for a dance, and who is addressing whom in the dialogue at the end of the scene. At the bottom of p. 47 (Fig. 16), it is obvious that the stage direction for Arcite’s entrance should precede rather than follow Palamon’s greeting of him (this may be a case of a compositor saving space by not breaking up a line of dialogue).

  In attempting to deal with the sorts of peculiarities I have just described, the editor has to draw on knowledge of the period in which the text was written and published. What follows will look first at some ways in which editors try to find out more about the kind of text they are dealing with. Then it will examine aspects of the text in more detail, taking examples mainly from the pages reproduced here, and summarize current opinion as to the nature of the manuscript given to the printer.

  Compositorial analysis

  In twentieth-century editing, the role of the compositor who set the type from which a page was printed has become a subject of great interest, because it is important to know whether differences (say, in spelling) in different parts of the basic text might be due to different hands in the original manuscript or only to the practice of different compositors. Compositor analysis has become less confident, and more complex, since it was first introduced in 1920 (Taylor, 41–7). The earliest work was done on the 1623 Shakespeare Folio, and distinctions were made chiefly on the basis of preferred spellings, but it has become evident that even these preferences are not consistent: they can be affected by printing-house conventions, the nature of the manuscript copy, and the need to fit the right number of lines to a page. As with the authorship tests described on pp. 22–27, it is now recognized that only the coincidence of many separate defining factors can really identify an individual compositor. A large project like the Folio would have involved a good many compositors, since it took longer to set up a sheet of copy than to print off a series of sheets. A small quarto like The Two Noble Kinsmen could have been set by only one compositor, but shared typesetting and shared printing were common practices even with small publications.

  In fact, despite the apparent similarity between the two openings illustrated here, Frederick Waller (TNK and ‘Printer’s’) and Paul Werstine arrived independently at the conclusion that the quarto was the work of two different compositors. Both scholars looked for small details of presentation which might result from automatic habits. For instance, on pp. 14 and 15, punctuation marks within the line are not normally preceded or followed by spaces, but there are exceptions at 1.3.73, 1.4.5 and 1.4.9. On pp. 46–7, on the other hand, it is easy to see that many of the colons are set off with spaces before and after them, while the commas are squeezed in without spacing except in one line (3.5.149) where the compositor must have realized that he was going to have to run over anyway and therefore needed to create a neat break. Waller found that one compositor used white spacing in this position 80 per cent of the time, the other only 30 per cent. There may be dispute as to what does and does not constitute a space in early printing technology; nevertheless, this difference is large enough to be significant.

  The Flavia/Flavina error mentioned above is almost certainly the compositor’s; the Pirithous confusion, on the other hand, recurs throughout the text (see notes to the List of Roles) and clearly derives from the manuscript.

  Press correction

  Page 15 (Fig. 15) contains one of the relatively few press corrections in the quarto, the substitution of smeard for succard in 1.4.18. Press correction at the seventeenth-century printing house went on while sheets were still being run off the press, with the result that different copies of the same book will show different mixtures of corrected and uncorrected sheets. Of the forty-five copies of The Two Noble Kinsmen known to exist, Waller examined thirty-one, Bowers sixteen. They found only two corrections on sheet C, two on E, and one each on K and M. Only one forme of each of these sheets was proofread – in other words, only four out of the twenty-six formes that would have been used to set up the entire play. These figures may give a misleading impression of printing-house sloppiness. We can tell that a page was corrected only if we find an uncorrected page with which to compare it, but the number of corrected formes would normally have been much greater than the number of uncorrected ones. Thus in many cases we probably have no evidence that correction took place at all. The significant corrections so far discovered are:

  Uncorrected Corrected

  C1v (1.2.70) on; on

  C4 (1.4.18) succard smeard

  E2v (2.3.6) fins sins

  E3 (2.3.24 SD) Garlon garlond

  K2v (4.3.54) behold behind

  K2v (5.1.76) his her

  M4 (5.4.79) victoros victors

  A small number of copies also contain what Richard Proudfoot describes as an intermediate state of correction on sheet C, with succard uncorrected but with on (the dramatist’s spelling of one) at line 70 instead of on;. The probable explanation is that someone realized the need for punctuation in line 70, but put the semicolon in the wrong place, after instead of before on. The absence of authorial proofreading is obvious. No author, for instance, would have missed the Flauia–Flauina confusion; the compositor did not make a connection between the word set at line 84 and the one already set in line 54. These corrections are interesting not only because they sometimes (as on sheet C) supply readings that would be hard to arrive at otherwise. The initial misreading can also show something about the handwriting that was misread (for instance, her in 5.1.76 may have been misread as his because it was spelled hir) and hence about the provenance of the manuscript; the extent of correction can also give some indication of the general reliability of the printed text.

  Punctuation

  By 1634, nouns were beginning to be capitalized more heavily than in earlier years; commas still tended to appear between subject and verb and before conjunctions; full stops, or periods, occurred more rarely than in modern texts, and ‘?’ was sometimes used where one would expect ‘!’ (for instance, Emilia’s speech in 4.2 contains ‘what an eye? / Of what a fyry sparkle, and quick sweetnes, / Has this young Prince?’ (4.2.12–14)). As Bowers points out (‘Readability’, 214–15), not all Elizabethan compositors even had exclamation marks in their cases of type; many questions are rhetorical – halfway between exclamations and questions – like Palamon’s ‘Oh you heavens, dares any / So nobly bear a guilty business?’ (3.1.89–90). Conversely, the absence of a question mark after ‘How does my sweetheart’ (3.5.146), though it may be an error, may also be a clue to Theseus’ delivery of the line, as the author imagined it. In manuscripts, writers often do not bother to punctuate at the end of a line. They may have felt that the edge of the page itself constituted a break.

  Pages 14–15 (Fig. 15) contain a remarkable number of parentheses. These have a variety of functions. Some are used where modern convention might prefer commas. Sometimes, as in modern punctuation, they indicate an aside, not quite important enough to be part of the sentence; in both Emilia’s and Hippolyta’s speeches, seen here, they also indicate a change of tone. Sometimes they indicate a change in the person addressed. For instance, Theseus in 2.5.48–9 obviously begins by speaking to Hippolyta, but then goes on, ‘Sweet, you must be readie, / And you Emilia, and you (Friend) and all’; the ‘Friend’, clearly, is addressed to Pirithous. For some reason, there are many more parentheses in the first part of the play, and in scenes attributed to Shakespeare. One reason may be that they represent additions to the original manuscript. But, as noted above, a parenthetical structure is important in the play, whether or not it is designated by the punctuation marks known as parentheses or brackets.

  There has been debate for some time about the value of following seventeenth-century punctuation. As is often said, it is rhetorical; that is, a comma is likely to come at a point where a speaker might be expected to pause (even if this is between a subject and a verb) rather than where it is logically needed. The punctuation of The Two Noble Kinsmen is very heavy, in the mid-seventeenth-century mode rather than the lighter one of the late sixteenth century. I suspect that at times it may convey something of the original tone, as in the nudge-nudge pauses of the Prologue:

  New Playes, and Maydenheads, are neare akin,

  Much follow’d both, for both much mony g’yn,

  If they stand sound, and well?…

  (1–3)

  One can imagine an actor making the most of all these short pauses, looking around for a laugh on Maydenheads and the double meaning of sound. While it may look as though Playes and Maydenheads were meant to be extra-emphatic, Figures 15 and 16 show no clear pattern of capitalization. Most modern editors use lower case even for nouns that are used as personifications (like Peace, 1.2.23). I have capitalized the latter even where the quarto did not (see list on p. 145), to alert the modern reader to the special usage. Conversely, the complete lack of full stops or periods in the whole of Emilia’s long speech on Q’s p. 14 (Fig. 15) helps to suggest the ‘high-speeded pace’ on which her sister comments. But in most cases the punctuation was probably the work of the printer. Adding punctuation (especially in the form of exclamation marks) often has much the same effect as adding stage directions. Thus it is important to find a way of making the meaning as clear as possible without preventing the reader or actor from recognizing intentional ambiguity.

  Lineation and scansion

  Many Renaissance plays are written in a mixture of verse and prose. The Two Noble Kinsmen is largely in verse, with two prose scenes (2.1 and 4.3), and two scenes (2.3 and 3.5) in which the verse is so loose that it is hard to tell it from prose. The distinction was not always any clearer to Renaissance compositors than it is to us. In the 1634 quarto, for instance, the prose of 2.1 and much of 4.3 is printed, like verse, with unjustified right margins, while in 2.3 the compositor seems to have a change of mind as to whether to set verse or prose. The reason may have been that, as Hans Gabler (Bowers, 6.6) has noted, dramatists after about 1620 used quarto- rather than folio-sized paper for writing playscripts. As a result, verse lines often extended to the right margin and became indistinguishable from prose. Whereas the dramatic blank verse of Shakespeare’s earlier plays was generally regular, his later writing relaxes the rhythm by allowing extra unstressed syllables and running on from one line to the next. Fletcher’s writing carries the first practice still further.

  The result is that at times, for instance in Gerald’s first speech (3.5.1–22), the rhythms of blank verse are so loose that it is difficult to know where lines begin and end. At lines 15 and 18, for example, several possible lineations will result in a good blank verse line:

  And unto him I utter learned things

  or

  I utter learned things and many figures

  At length, I fling my cap up – mark there! – then

  or

  I fling my cap up – mark there! – then do you

  The problem becomes even more acute when a verse line is divided among two or more speakers. Seventeenth-century printers were not concerned to preserve the shape of a blank verse line in these circumstances, and editors at present are divided as to whether such lines should be treated as ‘short lines’ (which means printing them like prose) or as ‘shared lines’ (which means deciding where they begin and end). Both Shakespeare and Fletcher were so used to writing blank verse that they almost automatically produced sets of speeches that added up to complete lines. But often there is a spare half line which can join itself to either the preceding or the following line of text. A good example is the exchange at 1.3.85–6. I have divided the lines as follows:

  [HIPPOLYTA]

  Love any that’s called man.

  EMILIA

  I am sure I shall not.

  HIPPOLYTA

  Now, alack, weak sister,

  But there is an equally possible alternative (which Waith adopts in his edition for the Oxford Shakespeare):

  [HIPPOLYTA]

  Love any that’s called man.

  EMILIA

  I am sure I shall not.

  HIPPOLYTA

  Now, alack, weak sister,

  My scansion might result (depending on elision and stress) in either a masculine or a feminine ending; Waith’s is definitely feminine. This is one reason why metrical tests based on line endings are difficult to use effectively (see Bertram, 21–34). The editorial decision makes a difference to the actors who speak the lines, since the speaker who ends a line of verse can seem both to enter into the rhythm of the previous speaker and to cap what has just been said. For instance, in the quarrel between Palamon and Arcite at 2.2.166–70, this exchange occurs:

  [ARCITE]

  So both may love.

  PALAMON

  You shall not love at all.

  ARCITE

  Not love at all,

  Who shall deny me?

  PALAMON

  I that first saw her?…

  Here, the decision between two kinds of lineation is also, to some extent, an interpretative one. Leigh Hunt, in 1855, suggested:

  [ARCITE]

  So both may love.

  PALAMON

  You shall not love at all.

  ARCITE

  Not love at all! Who shall deny me?

  PALAMON

  I.

  By contrast, this is my version:

  [ARCITE]

  So both may love.

  PALAMON

  You shall not love at all.

  ARCITE

  Not love at all!

  Who shall deny me?

  PALAMON

  I that first saw her?…

  Hunt’s reading gives Palamon literally the last word. In mine, the metre breaks down after Arcite’s rebuttal; the rhythm, originally showing the almost instinctive rapport of the two men, comes to reflect the breach between them. I preferred this effect, but readers and actors who study the notes will be aware that it is the result of a choice which could have been made differently.

  A possible solution to lineation difficulties is to abandon the attempt at differentiating verse which can only be scanned with effort. Waith (Oxf1) prints Gerald’s first speech as prose, while acknowledging that many of its lines have blank-verse rhythms. With regard to divided lines and short lines, it can be argued (as it has been by Bertram (29)) that, since the question is so insoluble, it is better not to force a series of short lines into a blank-verse unit (see also Bevington, 266–70). Nevertheless I decided to retain a basic blank-verse shape, because this edition is intended for readers as well as actors. Both Shakespeare and Fletcher, at this stage of their careers, were blank-verse virtuosos who seem to have enjoyed creating smooth lines out of apparently disparate elements. For instance, at 3.5.91, if the Countryman’s ‘Do, do!’ is lost in the general response to the Daughter’s dancing, the result is a line that scans, even though there is no particular need for it to do so. I should add that my textual notes do not include details of the line endings of passages printed as verse in the quarto and as prose in this edition, or vice-versa.