To summarize, then: the only substantive edition of King John, that is, the only one derived without intermediary from the manuscript source, is that of the First Folio, and it is based on a transcript of the author’s foul papers, made some time between 1606 and 1622 by two scribes. It is now time to give some further details of the points summarized as 1–4 above, and indicate how they influence the editorial task.

  The practice of ‘improving’ Shakespeare editorially goes back, in fact, to the First Folio. Its compositors applied to the manuscripts which were their copy a relatively uniform and relatively modern (for 1623) form of punctuation, which was almost certainly unlike anything in Shakespeare’s own manuscripts, and probably not much like that of the professional scribes associated with the company.2 In doing so, and in regularizing other features of their manuscript copy so that the Folio would have a fairly uniform, consistent style and appearance, they engendered a so-far endless series of editors whose interventions have ranged from omission of large chunks of the plays (whether on the grounds of propriety or for some other reason), through the alteration of words and lines either because they thought them corrupt or otherwise embrangled, to the more routine and inevitable modernization and house-styling which all publishers impose on their products.

  In the late twentieth century and early twenty-first century these processes have been increasingly treated as anathema by a number of scholars, who resist and object to any editorial intervention between the reader and the original text. In many respects, however, the practices of editors and those of directors, dramaturges and script-editors in professional theatrical companies are very similar. Both editor and director are faced with a script which requires modification: the editor must make sense of what is confusing in the copy-text, either by emendation or by explanation in the commentary, and must also produce a version of the play that fits the physical and design parameters of the publisher. The purpose is to make the work as accessible as possible to a modern reader who may have no special knowledge of the period, or no special cultural connection with Shakespeare. Very similarly, the director, cast and aides are faced with a script whose cultural and historical distance from the audience requires it to be modified. The director must make sense of what is confusing in the copy-text, either by emendation or by explanation through additional ‘business’ on stage, and must also come up with a version of the play that fits the physical and design parameters of the theatre and the abilities of the acting company. The principal differences are that the great majority of modern professional productions of Shakespeare are cut, and that the theatrical company seldom gives chapter and verse for each of its modifications of the text in performance.1

  Happily, the various published texts of King John (excluding the bowdlerized and abridged editions) have had to make only minor changes to the fabric of the play’s dialogue: a word here or there, a conjecture modestly banished to the notes. Even so, many of the lines thought meaningless or corrupt by eighteenth-century editors have been found to make good sense, thanks to our ever-increasing philological knowledge of Early Modern English. One or two examples will suffice: in 5.7.15–16 our text reads, with F, ‘Death, having preyed upon the outward parts, / Leaves them invisible’. The word ‘invisible’ means ‘invisibly’, but it has bothered many editors, who have tried either fiddling with the punctuation, or supplying words like ‘insensible’ (Hanmer), ‘invisited’ (Collier), ‘ill-visited’ (Ingleby), ‘invincible’ (Smallwood) and ‘invasible’ (conjectured by Dover Wilson). A simpler example is the Dauphin’s scornful dismissal of the Bastard: ‘We hold our time too precious to be spent / With such a brabbler’ (5.2.161–2). Rowe emended to ‘babler’, which persisted in many eighteenth-century texts before falling out of use in favour of the F reading.

  The general editorial conservatism as regards the dialogue text of King John has not extended to its directions text. As noted above, the Folio text presents all sorts of problems to the performer, or to the reader attempting to envisage the performance from the text. The eighteenth-century editors sought to alleviate many of these problems, but in doing so wrote directions suited to the proscenium-arch stages and pictorial scenery with which they were familiar. Generally speaking, these directions proved adequate for nineteenth-century editors too, in the heyday of the ‘architecturally correct’ period of Shakespearean staging. Until recently, King John has so seldom been revived that theatrical representation has had little influence on the editorial tradition. Nonetheless, a heightened scholarly awareness of the nature of Elizabethan staging conventions and the physical structure of the theatres has meant the progressive elimination from modern editions of ‘pictorial’ or historical directions, and the substitution of more theatrically flexible ones. The present edition, while retaining as much of the wording of the F directions as possible, introduces additional directions, on the following general principle.

  When it is absolutely clear from the dialogue that a certain action must take place on stage, even though there is no direction in F, we have added a minimalist direction. However, we have not inserted directions when the dialogue makes it obvious what actions the character must make. For example, consider 3.1.19–22:

  What dost thou mean by shaking of thy head?

  Why dost thou look so sadly on my son?

  What means that hand upon that breast of thine?

  Why holds thine eye that lamentable rheum …?

  Actors sometimes dislike and often ignore these in-text directions, but it seems evident to us that Shakespeare certainly intended them as directions for the performer. When it is clear that some action must take place, but it is not clear just what action, or when it happens, we have discussed the matter in the commentary notes. The most difficult crux is the handfast in which King John and King Philip must engage in 3.1: Philip relinquishes John’s hand, perhaps at 262 (‘I may disjoin my hand, but not my faith’), but more probably at 320 (‘England, I will fall from thee’) but there is at best only a vaguely plausible line or logical moment to suggest for them to join hands in the first place, that of 135.

  In accordance with the principles of the series, speech prefixes have been regularized. This has entailed a large number of changes, especially in 2.1. Certain scholars urge the unwisdom of this kind of regularization, on the grounds that it obscures the shifting attitudes of the author towards his characters.1 This raises an important point of principle.

  Until about thirty years ago or so, it was taken for granted that an edition of Shakespeare would attempt to reconstruct the author’s own manuscript as closely as the surviving documents permitted.1 More recently, editors have increasingly become aware that playscripts, even the sacred Shakespeare’s, are documents created in order to bring a play upon the stage. For this reason, the fact that the text of King John is based on a relatively unrefined transcript of the dramatist’s foul papers is less of a cause for celebration than it would have been a generation or two ago. It means that we are deprived of those developments of the script which would have occurred in the process of preparing it for production, developments which would have been at least in part recorded in the Book of the play.

  Just as it proved impossible to produce an edition which recovered Shakespeare’s manuscript, it is impossible to recover the production even from a text based on the Book, let alone one based on foul papers. Even so, it remains part of our editorial policy to remember that King John was indeed staged by the Chamberlain’s Men, and therefore that whatever in the foul papers was simply unworkable onstage must have been modified to make it stageable. These considerations have influenced the text we print only slightly, but the influence is there, and is found chiefly in the stage directions and speech prefixes. In this, our edition follows the general policy of the series, and extends the practices of the great majority of earlier editions of the play.

  Act and scene divisions

  The physical division of the play into acts and scenes must now be considered. First, the facts: in the Folio the first page of King John is headed ‘Actus Primus, Scæna Prima’, as is normal for the Folio. In terms of the editorial conventions of the seventeenth and eighteenth centuries, on which the act/scene division of modern editions is still based, a new scene begins when an exit direction leaves the stage empty. In this sense, Act 1 of King John is a single scene, since the stage is never empty.1 So the next scene, that before Angiers, is headed in F ‘Scæna Secunda’. Well, so it is, but by any reasonable estimate it is a new act: set in a different country, and with a brand-new cast of characters to get to know. Like Act 1, it contains a number of separable dramatic sequences,2 but the stage is never empty (it is an enormous scene, 599 lines long, including what might well be a rather spectacular battle scene). The next F heading is ‘Actus Secundus’ (without a full stop). Yet the matter of the ensuing scene is directly connected with what has just gone before: Lady Constance’s outrage at the perfidy of her French champions. It ends at 74 without a stage direction, whereupon F prints ‘Actus Tertius, Scæna prima’, and includes Constance in the entrance direction. F marks as ‘Scœna Secunda’ what most editors call 3.2, but ignores the clear stage at 10, before the second battle scene. Editors have usually, to be consistent, made scene 3 begin at 11, so the next F heading, ‘Scæna Tertia’, is changed editorially to 3.4. Act 4 requires no emendations, and neither does Act 5 apart from the misprint in the heading (‘Quartus’ for ‘Quintus’). The problem, then, is localized in Acts 2–3.

  The most important recent thinking on this subject is to be found in a 1993 essay by Gary Taylor (‘Structure’). Taylor argues, persuasively, that the professional companies collectively adopted the practice of act-intervals after the King’s Men pioneered the concept in the Blackfriars, following their acquisition of the lease there in 1608. The evidence is that the published texts of all professional plays 1616–42 are so divided, as are the surviving Books from the period. (In all theatrical texts from the period, the word ‘act’ means the interval or the music played during it.) The children’s companies used act-intervals from 1599 (and possibly earlier – the evidence is not completely clear). The 75-odd adult plays surviving from 1591–1607 are undivided, again both in print1 and in manuscript.2 There are no eyewitness accounts of intervals before 1616 in adult plays; but act-divisions began appearing in published plays after 1609. The key variable is evidently the date of composition rather than the date of publication. The appearance of act-divisions coincides with the disappearance of the dramatic device of the chorus: some 10 of the 155 extant adult plays before 1610 ‘are divided into five natural acts’ by presenters; another 12 divide the play into more or fewer acts; but after 1610 presenters tend to appear ‘sporadically during, rather than between, the acts’ (Taylor, ‘Structure’, 24). The evidence is clear and convincing: ‘before c. 1607–10, the adult companies did not perform plays with intervals between the acts’.

  Taylor also discusses the retro-fitting of earlier plays for the new manner: any pre-1607 play that later shows up in print with act-divisions – most such are Shakespearean – must have been set in type from a Book (or transcript derived from a Book) that had been revised to incorporate the new fashion, or as a result of editorial intervention. This last is best demonstrated by the impossible division of Henry V;1 other plays in the Folio, especially in the Histories and Tragedies, were sensibly left undivided. We accept his conclusion as proven; but it raises a point. That Heminge and Condell or Jaggard and Blount might have wanted a nicely uniform volume, with all plays divided neatly into acts and scenes, is plausible; that they gave it up when it proved a bother is likewise reasonable; but the inference, that any Folio play badly or madly divided or not divided at all may well not have been revived after 1608 is startling. Taylor doesn’t draw this inference, but it seems worthy of consideration. At any event, it seems safe to say that Taylor’s evidence indicates that the transcript of King John’s foul papers which became copy for the Folio would not have been made before 1608.

  Taylor’s conclusions force a revaluation of the division problems in King John: the only one that survives as a genuine problem is that of Constance and the beginning of Act 3. If indeed Constance suits the action to the words of 3.1.71–4, she either sits down or throws herself onto the ground,2 declaring ‘Here is my throne, bid kings come bow to it’.3 With a Stoppard-like promptness, both kings and their trains come on, and she (presumably) gets up to abuse them. Did she remain seated onstage during the act-interval? Not in the 1590s; but that something like this could happen is confirmed by the F direction for the lovers in A Midsummer Night’s Dream between Acts 3 and 4: ‘They sleepe all the Act’. But recall that F adds Constance’s name at the end of the ‘Actus Tertius’ entrance direction. It is widely accepted among theatrical scholars that exit and immediate re-entry was regarded as undesirable stagecraft, and so avoided, in Elizabethan times; so we conclude (1) that in the Elizabethan production Constance remained on the ground or sat, and remained there until the royal procession was onstage; (2) that the scribe preparing Folio copy, marking the act-beginning here in accordance with post-1608 practice, found Constance talking, and added her name to the entrance, not noticing that she had not been given an exit; and (3) that if King John was indeed performed by the King’s Men after 1608 and if the act-division here is indeed correct, Constance must have remained onstage through the musical interlude. The other division difficulties cited arise mainly from compositorial slips.

  What is best to be done? First, let us reiterate that in the Elizabethan period, performance was broken neither at act nor at scene ends: act-division came later to the professional stage, and scenes were important chiefly for planning the doubling, that is, the playing of two or more roles by the same actor. When King John was first staged, the performance was continuous. There is no concrete evidence that it was revived in the Jacobean period, and if it was, we have very little to go on in trying to reconstruct how the act-divisions worked in the theatre. Since the act/scene-division in modern editions of the play therefore reflects first, the work of the scribes, and second the desire of the Folio’s publishers to have a house-styling that dignified Shakespeare’s works in the way Jonson’s 1616 Folio dignified his, there seems little literary or theatrical justification for worrying too much about the acts and scenes. We have decided, therefore, that our primary concern should be ease of use for the reader. Since the most widely used and accessible concordance of Shakespeare is Marvin Spevack’s one-volume Harvard Concordance, which we ourselves use and cite, and since that work is keyed to the text of the Riverside Shakespeare, we have for purely pragmatic reasons adopted the Riverside’s act and scene divisions.

  Let us turn now to the activities in Jaggard’s printing shop and how these may have influenced the text. The details of the methods used by the workmen in setting type for and printing the First Folio have been known in considerable detail for some time (Hinman). The fact that the book was a large folio has all sorts of practical consequences, and therefore we must explain its bibliographical make-up. A sheet of paper imposed for a folio book will end up having four pages, two on one side of the paper, two on the other. After printing, the sheet of paper is folded down the middle for binding. A moment’s thought will show that pages 1 and 4 must be on one side of the sheet (the ‘outer forme’) and pages 2 and 3 on the other (the ‘inner forme’) so that when folded the pages will appear in the right order. However, like most folio books, the First Folio was sewn into its binding not in single sheets (which would have made for a very thick spine) but in groups of three sheets folded together. This is called a gathering in sixes (that is, the folded gathering will have six leaves, and each leaf two pages: the recto (r), the right-hand page when a book is opened, and the verso (v), the other side of the recto). To recapitulate: a First Folio gathering is made up of three sheets of paper, six leaves and twelve pages. Remember how these must be printed so as to come out right when folded and bound, and you will see that the first sheet must have pages 1 and 12 on the outer forme, and pages 2 and 11 on the inner forme. The middle sheet of the gathering will have pages 5 and 8 on the outer forme and 6 and 7 on the inner – the only consecutive pages in the gathering.

  English printers were fond of an economy measure that enabled them to print big books without using too much type (the printer’s most important investment). Supposing you were to start setting type at page 1 of a six-leaf folio gathering, and keep on going through 2, 3 and so on. When would you be ready to impose and print a forme? Not until you had set page 7. Five big pages of type would be sitting around idle; and you could not print the forme containing page 1 until page 12 had been set. To save type, then, Jaggard adopted the expedient of casting-off the copy so that printing could begin with the inner forme of the third sheet, that containing pages 6 and 7. One compositor would work on this forme; then when it was ready it would be imposed and printed while he was working on the outer forme (pages 5 and 8); the other compositor (for they usually worked in pairs) would be doing the same thing for pages 4 and 9 (inner forme of the second sheet) and 3 and 10. So they would go, working simultaneously forward and backward through the gathering and on to the next until they were done. The advantage was that no more than four pages were usually in type at any time, since as soon as the forme was printed the types would be distributed back into the cases by the compositor.1

  To make this technique possible, the printers’ copy had to be cast off (counted off in page-lengths) so that the compositor, finishing work on page 12 of a gathering, would know where to begin setting page 6 of the next. Naturally, casting-off was not an easy process, especially in prose, and there are plenty of places in the Folio where the arithmetic went off the rails. Obviously, it was easier to cast off verse, since you could count lines rather than words, but even in an all-verse play like King John there are some signs of trouble.2