Internet Shakespeare Editions

About this text

  • Title: Textual Introduction
  • Author: Michael Best

  • Copyright Michael Best. This text may be freely used for educational, non-profit purposes; for all other uses contact the Editor.
    Author: Michael Best
    Not Peer Reviewed

    Textual Introduction

    Differences between the Quarto and Folio texts

    The nature of the changes between the two versions can be summarized thus:

    1. passages, phrases, and words that appear in the Quarto but not in the Folio;
    2. passages, phrases, and words that appear in the Folio but not in the Quarto;1. 1. changes in lineation of verse;
    3. speech assignments that are changed between the two versions;
    4. restructuring of speeches and staging;
    5. cumulative changes to roles;
    6. words and short phrases that are different in the two versions.

    I have phrased points 1 and 2 in such a way that I make no assumptions about the priority of the two texts. In practical terms, however, it is simpler to think of most passages missing from the Folio to be cuts, and passages appearing in the Folio only, in most cases, to be later additions by person or persons undefined—though there have been arguments that propose alternative reasons for the differences.

    251. Cuts to the Folio text

    The changes least amenable to the identification of the agent who caused them are the many cuts. By their nature, however significant their effect on the structure of the play and the nature of the actors' roles, cuts leave no stylistic fingerprints; thus it is unfortunately impossible to subject cuts to any kind of authorial test. Cuts may be made for many reasons: to placate the censor, to shorten performance time, to accommodate changes in casting, or to adapt to changing tastes in the audience. Some short omissions may be the result of simple error. Given the gap of roughly eighteen years between the first performance of the play and its publication in the Folio, it is certainly possible that the cuts were made at different times and for different reasons.

    There has long been a debate concerning who made the cuts. Before the claims made by those in The Division of the Kingdoms arguing that Shakespeare himself was responsible for many of them, the general assumption was that they were made by the players, the book-maker, or the prompter. The understanding was therefore that the play had been corrupted by ignorant interference. There is indeed good evidence that authors from the period were pleased when they were able to publish plays in a form that was more complete than the versions in which they had been staged. Ben Jonson and John Webster both published versions of their plays in which the title page trumpets the fact that the printed version contains more than was ever acted on stage (see Rasmussen 442). The preface to the collected works of Beaumont and Fletcher (1647) by the printer, Humphrey Moseley, is especially informative:

    When these Comedies and Tragedies were presented on Stage, the Actours omitted some Scenes and Passages (with the Authour's consent) as occasion led them; and when private friends desir'd a Copy, they then (and justly too) transcribed what they Acted. But now you have both All that was Acted, and all that was not; even the perfect full Originalls wihout the least mutilation. (A4r)

    What is particularly fascinating about this passage is that it not only reinforces the comments of Jonson and Webster concerning cuts to their works, but that it also includes the parenthetical comment that the omissions were made with the author's consent. The implication is that the author might well be involved in the choice of cuts as well as the players. Since Fletcher was a collaborator with Shakespeare in at least two of his last plays, there is a distinct possibility that within the company that performed their plays there was a culture that involved some level of consultation, even collaboration, between author and actor. It is very much human nature for authors to want the complete sweat of their labors to be made available to their readers, and it is entirely possible, even probable, that Shakespeare would have had some fondness for the sections that were cut, however effective the changes. I do think, however, that there is reason to believe that the agents who made the cuts would have been far from blind to literary and dramatic sensibilities. The actors who made up the Chamberlain's Men/King's Men were a generally stable group, would have performed with Shakespeare, and would have become accustomed to his style. While there is no evidence that Shakespeare in any way "directed" the performances of his plays, the members of the company will have absorbed, if only by a kind of osmosis, a sense of the dramatic structures and overall emotional directions he embedded in his plays. If members of the company were those responsible for the cuts, there is no reason to believe that they were not to a significant degree in harmony with Shakespeare's overall vision for the play even if they did not actively consult with him. I discuss this issue further in the General Introduction.

    The net effect of all cuts and additions is to shorten the play by some 200 lines, a significant but not major reduction of the time required for performance. Some minor cuts and modifications (especially of stage directions) refer to the French army. They may have been made in order to soften the references to invasion, with the intention of making the battle between Lear and his sons-in-law seem more a civil war than a war between nations. It is not clear whether this change was made as the result of actual censorship, was an act of self-censorship intended to bring the politics of the play more into line with changing attitudes, or was intended to focus on the dangers of internal rather than external threats to political stability (see Taylor "Monopolies," Halio Tragedy 83, Worden).

    The two most significant of these changes involve Kent, as he chats in each case to a Gentleman. In neither instance is Kent instigating or advancing plots; he is simply reporting on them. In the more substantial change—Q scene 17, omitted entirely from the Folio—Kent passes on news that the King of France has returned to his country, leaving a lieutenant in charge, and he instructs the Gentleman to head to Dover to pass on news about Lear's current state. The main emphasis in the scene, however, is on the description of Cordelia's response to news of her father's distress. In the earlier scene (Q 8, F 3.1, TLN 1614), the first in which the French presence is mentioned, Kent informs the Gentleman of the division between the dukes and the presence of French spies in the country. He is more informative in the Quarto, and gives the Gentleman instructions on travelling to Dover to pass on the information; in the Folio he is more circumspect and indirect. This sequence is the only instance where a significant number of lines in one version are replaced rather than cut in the other; it is discussed more fully below.

    30In presenting their case for Shakespearean revision, several contributors to The Division of the Kingdoms rely on arguments that the Folio version was revised by Shakespeare specifically to improve stage-worthiness and to make deliberate changes in characterization. The critical debate about the relative theatrical value of the Quarto and Folio texts illustrates the way that arguments of this kind can be made to work in both directions (compare Urkowitz, and Clare "Who is it"). It might be unkind to point out that the argument that cuts were made to improve the play's effectiveness on stage could readily become an argument in favor of "the players" as the agents who cut sections of the play since their concern would presumably be in favor of heightening the play's dramatic force.

    Productions of the play have varied enormously in the kinds of cuts and modifications that directors have historically made; one actor's/director's/critic's high point is another's candidate for the axe. The empirical test of production over many years argues rather against the clear superiority of the Folio text in the theater, since only those directors who have deliberately decided to produce a Folio version have followed its cuts. Michael Dobson, in a review of the Bate/Rasmussen Folio Lear in the Royal Shakespeare Company Works, commented that "Even now Ian McKellen is performing King Lear in Stratford using a text including the mock-trial scene, here relegated to an appendix. . . . Perhaps Bate and Rasmussen would have preferred McKellen to perform the mock-trial scene in King Lear only as an encore." In all fairness, someone should have directed Dobson to Rasmussen's more strictly bibliographical discussion on the nature of cuts in F Lear ("Revision of Scripts"). The cut that deletes the mock trial is one of the longest (30 lines). Other than reduciing performance time it is not easy to see why the lines were deleted. The shortened scene does still provide a suitably moving exploration of the nature of Lear's loss of mental control, but modern productions tend overwhelmingly to include the omitted passage, perhaps because our tastes favor its combination of black humor and pathos.

    Eric Rasmussen insightfully discusses the nature of cuts in Lear, comparing them to those in Hamlet and The Second Maiden's Tragedy, a play performed by the same actors at about the same time. The manuscript of The Second Maiden's Tragedy records cuts made for performance. Rasmussen points out that cuts in Hamlet and The Second Maiden's Tragedy are generally made in the middle of long speeches; from the point of view of the actors in the company, the advantage of this approach is that cue lines remain the same so that only one actor has to remember the change. In Lear, however, "lines are generally cut from the ends of speeches," and from the conclusions of scenes ("The Revision of Scripts" 445). One possibility is that the introduction of act breaks in the theater invited cuts at the end of two acts: nine lines of dialog between two servants commenting on Gloucester's blinding at the end of act 3 (3.7, TLN 2176.1-2176.9), and eleven lines at the end of act 4, when Kent and a Gentleman discuss the current state of affairs in the kingdom and the forthcoming battle (TLN 2843.1-2843.11). My analysis of 33 significant cuts/omissions in the Folio indicates that two-thirds (23) create new cue lines, two (possibly three) involving cues for entrances or exits. Only ten leave the cue lines unchanged. The equivalent figures for Hamlet are 20 cuts, half requiring cue changes and five involving an entry. Rasmussen rightly cautions that "we cannot conclude that one type of cut necessarily points to a revising author and that another points to a non-author" (446). A reviser conscious of actors' needs and aware of the relative convenience of leaving cue lines unchanged might focus more on cutting within speeches; an author might focus more on changes he wanted for dramatic or literary reasons; but Shakespeare was, or had been, actor as well as author.

    Longer cuts accumulate in the latter half of the play, principally in acts 4 and 5. Two significant cuts delete passages that provide glimmers of light in the growing darkness: the passage at the end of the scene of Gloucester's blinding where two servants react to the cruelty they have witnessed by deciding how best to help the old man (2176.1-2176.9), and the entirety of scene 17 (4.2a in the Extended Folio) between Kent and a Gentleman where they discuss the current state of affairs, focusing on the pathos of Cordelia's response to news of her father's predicament. In the General Introduction I suggest that some cuts follow the emotional trajectory Shakespeare began by creating a hard-edged tragedy from the romance plot he found in his sources.

    2. Additions

    If cuts leave no clear sign of the agent who did the cutting, additions are more forthcoming and may leave potential stylistic fingerprints. The extent of the additions, however, is modest, so evidence of authorship will of necessity be tentative. By studying the possible sources and the vocabulary of rare words in the added passages, Gary Taylor has argued that Shakespeare added the passages "between the completion of Coriolanus and the beginning of Cymbeline" ("Date" 428); Richard Knowles suggests that the rare words in the additions "suggest a reviser with an idiosyncratic vocabulary, perhaps in some cases reflecting a familiarity with the contemporary theater, but probably not Sh[hakespeare]'s lexicon (King Lear Variorum "Appendix," 183). Studies of rare words have long been suggestive, but are far from conclusive, especially in the present instance where the size of the text samples is very small indeed. Studies that look at an author's use of common words using large samples have made some important contributions to authorship studies (see, for example, MacDonald P. Jackson Defining, Taylor and Egan Authorship Companion), but the whole point of looking at rare words is that they occur in small samples. The textual analysis I find more persuasive on the issue of the authorship of the additions to the Folio is Arthur F. Kinney's "Transforming King Lear" (Craig and Kinney 181-201). Working with Hugh Craig, Kinney has taken the additions as a block of 900 words (902, to be precise) and compared them with the word usage of other writers of the period who have been proposed as possible authors: Massinger, Fletcher, Middleton, Jonson, Chapman, and Webster—and Shakespeare's other works. Using tests that look at blocks of 900 words in the various candidates for authorship of the additions, Craig and Kinney applied two tests: patterns of usage within common "function" words (the basic building blocks of sentences that authors use unconsciously), and one test that looks for common words and word forms one author favours but the other does not. Kinney and Craig find that the additions accord closely with Shakespeare's usage on both tests, but differ significantly from all the other candidates. The important feature of this approach is that the test samples are statistically large enough to rule out chance as the reason for the differences.}}

    35As with the cuts, additions that appear in the Folio text are of varying lengths and importance, and are very possibly the result of differing circumstances. Some probably record instances where the Quarto text was deficient and omitted short passages from the manuscript, either because of the messy manuscript or as the result of eye-skip: Albany's speech, "My lord, I am guiltless as I am ignorant" makes sense, but gains in clarity with the Folio's addition ". . . Of what hath moved you" (TLN 786-7); in contrast, Regan's speech in the final scene reads well with or without the additional line at TLN 3019, here italicized:

    Lady, I am not well, else I should answer
    From a full-flowing stomach. [To the Bastard] General,
    Take thou my soldiers, prisoners, patrimony.
    Dispose of them, of me. The walls are thine.
    Witness the world that I create thee here
    My lord and master. (TLN 3016-21)

    The line may have been omitted from the Quarto in error or added later to enhance the passage with a suitably martial image. There are several other instances, especially where a short speech has been interpolated, that may have been added to the Folio by either original omission or later revision: TLN 94-5, 176, 774, 1281, 1798, and 2243. The Fool's last speech in the Folio (TLN 2043) could possibly be another omission from the Quarto, and there may be others.

    Two more substantial additions to the Fool's role have long been considered interpolations, possibly by someone other than Shakespeare. The first of these, at TLN 1322-1327, according to Doran (Text 66) "appear[s] to have been an afterthought given to the Fool to enhance his part." Certainly the added passage is interpolated between two intense speeches by Kent and Lear, as Lear learns of his servant's treatment by Cornwall and Regan; whether this is seen as an enhancement of the tension or an irritating interruption to it is both a matter of critical taste and of stage direction. It certainly fits well into both the Fool's role elsewhere and the general tenor of the play:

    Winter's not gone yet if the wild geese fly that way.
    Fathers that wear rags
    Do make their children blind,
    But fathers that bear bags
    Shall see their children kind.
    Fortune, that arrant whore,
    Ne'er turns the key to th'poor.
    (TLN 1322-1327)

    The later, longer addition (TLN 1734-1750) when the Fool makes an enigmatic and parodic prophecy also fits to some extent as a commentary on a world turned upside down, but could possibly be the result of a star actor panting for more stage time. Kerrigan and Taylor ("Date" 396) defend both passages as Shakespearean. The statistical tests of Craig and Kinney cannot provide evidence one way or the other, because of the small size of the sample. The lines are certainly not the kind of improvisational language associated with clowns from an earlier period (see Stern, 245-52), though Rasmussen provides evidence for one way that material of this kind could be added to a theatrical manuscript ("Setting Down").

    If some of the additions I have been discussing can be explained as the restoration of passages accidentally omitted from Q, or as non-Shakespearean interpolation, there are some crucial and eloquent additions that were either written by Shakespeare or by someone who was good at Shakespeare pastiche. In the opening scene there are three telling and eloquent additions to Lear's speeches, giving additional motivation for his crucial decision to divide the kingdom and enriching his commentary on the nature of the lands he is planning to bequeath to Cordelia (TLN 45-50, TLN 54-55, TLN 89-91). The tenor of these changes is so consistent that it takes real ingenuity to argue that they are somehow omissions from Q rather than additions to F—and therefore actual revisions. Further passages later in the play give Goneril an opportunity to establish stronger motivation for her actions (TLN 842-857), and heighten one of Edgar's soliloquies (TLN 2184-2187). Lear gains a powerful addition in the section where in his mental breakdown he achieves insight into the human condition (TLN 2608-2612); most strikingly, there is a brief addition to the scene of his death when he fleetingly believes, just before he dies, that he sees Cordelia alive.

    403. Lineation

    The Quarto

    The most obviously intrusive intervention required from an editor of Q1 Lear is the process of making sense of Q1's erratic typesetting of Shakespeare's verse. Many verse lines are simply set as prose, while others appear as a kind of "false" verse, where lines are set without justification and dutifully start with a capital letter, but create hypermetric lines with widely varying numbers of syllables. We can be confident that Shakespeare was writing in verse for most of the play; the Folio text and other plays from the period when he was writing King Lear are predominantly in verse. There is no simple answer as to why so many verse lines were set as prose or false verse. It is true that Okes's inexperienced compositors were working with a difficult manuscript, but it is also true that it was common for printers to try to save paper by scrimping on space in one way or another. The final page of the Quarto [[ No such copy 'BL_Q1_Lr_2' ]] is a good example; four short speeches are recorded in the middle of a line rather on a new line and a substantial section of verse is set as prose; on the other hand the final large-font "FINIS" is set off by ample space before and after. Ironically, it is likely that the reason for this particular instance of space-saving was to keep the last page blank—a common but not universal practice. Vickers bases his fundamental argument for the "One King Lear" on the assumption that Okes underestimated the amount of paper he would need and therefore cut lines that later turned up in the Folio. Syme ("The Text is Foolish," III A Printer Abridges) and Blayney ("Quadrat" 86) dismiss Vickers's arguments, but agree that Okes's compositors tried to save space and may not fully have understood the manuscript they were working with.

    The inescapable conclusion is that editors of the Quarto version must intervene to a substantial degree if they are to provide a text consistent with Shakespeare's work at the time he was writing King Lear. Because the level of editorial intervention in creating verse from prose is substantial, I provide a visual method of seeing the difference between the original and the edited verse: horizontal tabs permit readers quickly to switch back and forth. Where it is relevant, a third tab shows the difference between the edited modern text and both Quarto and Folio line breaks. While this approach is almost arrestingly simple as a method of recording and visualizing differences in lineation, there are some limitations. Divided lines are not rejoined, and I have not modified capitalization to accord with the alternative line breaks. Variant lineation is not recorded in a tab when the length of lines in the two texts is significantly different because of additional material in one of them.

    These are the principles I have followed in lineating the Quarto text:

    1. I follow Q1 wherever its lines create reasonably regular verse, whatever the Folio does.
    2. 45I follow the Folio where Q1 makes prose of passages that are verse in F, preferably avoiding further smoothing of the rhythm. Paul Werstine has convincingly implicated compositorial intervention in modifying lineation from the copy used for the Folio ("Lineation" 111); thus Folio lineation may itself be sophisticated.
    3. In some passages (notably those of Lear's madness when he is speaking to Edgar and Gloucester in scene Q 20, F 4.5, TLN 2554) the editorial tradition has, since Johnson, been to set the prose as irregular verse; I follow this tradition with some reservations. Any verse that is created from the prose will be somewhat irregular; the question is where to break irregular lines without forcing the passages into a rhythmic straitjacket. In all such cases, however, the horizontal tabs permit the reader to see the extent of editorial intervention.

    The Folio

    I have followed the Folio's original lineation unless there is a clear case for emending it; there are several places where the compositor set short lines to fill up space, for example. As with the Quarto text, wherever I depart from the Folio, horizontal tabs recreate Folio line breaks. The Folio text also provides horizontal tabs that allow the reader to switch to Quarto lineation wherever it is reasonably regular but differs from the Folio; the intention is both to underline the uncertainty of editorial decisions (mine and those of the Folio compositors) and to offer additional rhythmic alternatives for actors, critics, and readers.

    4. Changes to speech assignments

    Many of the changes to speech prefixes are relatively minor: a Servant in Q becomes a Knight in F, a Gentleman in Q becomes a Messenger in F, a Doctor in Q becomes a Gentleman in F, and so on. Changes of this kind carry meaning and have an influence on staging and costuming, but they make no change to the major roles in the play. There are a total of 45 changes involving minor characters of this kind. (Text 79).

    50Speeches for major characters are changed on twelve occasions. All but two involve speeches of one line; one is of two lines, and the longest (the last speech of the play) is four lines. Some changes have relatively little effect: in the first scene Kent and Lear swap speeches, each of one line directed at the Fool. The remaining changes cluster in two scenes, both of which show signs of revision. In Q scene 7, F 2.2, where Goneril's arrival at Gloucester's residence is significantly revised, the query addressed to Regan, "Who struck [Q] / stocked [F] my servant?" is spoken in Q by Goneril, who is enquiring about Oswald's treatment, and in F by Lear, asking about Kent. The combination of the speech reassignment and the very significant variant radically changes the stage dynamic at this point. The second cluster, of six changes, is in the final scene of the play; all make changes both to roles and to staging.

    1. After Albany's strong response to Edmund, "I hold you but a subject of this war, / Not as a brother" (TLN 2999-3000), the line "That were the most, if he should husband you" (TLN 3012) is spoken by Goneril in the Quarto, by Albany in the Folio. The effect in the Folio is to lessen Goneril's antagonism and to give further strength to Albany.
    2. In Q Edmund's further response is to threaten military action: "Let the drum strike, and prove my title good" (TLN 3026). In the Folio it is Regan who threatens, necessitating a rewording of the passage: "Let the drum strike, and prove my title thine."
    3. When Albany produces the incriminating letter retrieved from Oswald by Edgar, in the Quarto he demands of Goneril, "Know'st thou this paper?" (TLN 3118); she replies "Ask me not what I know." In the Folio it is Edmund who is confronted, and who replies in the same words.
    4. Towards the end of the scene, two important speeches are reassigned from Albany to Edgar. When Edmund confesses his plot against Lear and Cordelia, it is the Folio Edgar who urges speed (by Q's 2 Captain, F's Gentleman) in rescuing them; and in a radical switch of speakers, Edgar, rather than Albany, is given the last lines of the play.
    5. Perhaps the most surprising switch involves the last moments of Lear's life where what is in the Quarto Lear's last speech, "Break heart, I prithee break," is given to Kent. This sequence is discussed in the next section.

    5. Restructuring of speeches and staging

    Five sequences of varying length present particularly interesting challenges in the construction of the two extended texts of King Lear. In each case, the two versions differ in ways that generate significantly different experiences on stage, and any attempt to combine them will only create yet a third version. For this reason, my solution has been to provide readers with horizontal tabs that allow them to switch between the two versions, or to view them in parallel.

    1. When Lear arrives at Gloucester's castle (Q 7, F 2.2, TLN 1466-1484) he suffers two shocks. He sees his servant Caius (Kent in disguise) in the stocks, and, as he seeks to find an answer to this extraordinary situation, Goneril arrives ready to team up with her sister. In the two versions Oswald's entrance is signaled at a slightly different point, and Goneril then enters; in the Quarto she immediately asks who "struck" her servant; in the Folio she is silent and it is Lear who asks who "stocked" his servant, without noticing Goneril. Both versions are dramatic, and the textual differences seem minor—a speech reassignment and one of those strikingly near-pun variants (struck/stocked), but in the Quarto Goneril is given an opportunity to be once again assertive by seizing the opportunity to attack, while in the Folio, Lear, ironically oblivious to the entry of his now disowned elder daughter, attempts to enforce his dwindling authority by once again demanding an answer to the humiliation Caius/Kent has suffered.
    2. The two versions of King Lear diverge in an especially interesting way as Kent imparts news of recent events to a Gentleman (Q 8, F 3.1, TLN 1626-1638, discussed above in the section on possible censorship). The Folio sequence is somewhat shorter than the Quarto, but rather than a straightforward cut (or omission), it replaces the longer discussion in the Quarto of thirteen lines with a different, shorter speech of eight lines. In the Quarto, Kent mentions that French spies have been gathering information about the dissention between Cornwall and Albany and that an invasion is imminent. He accordingly sends the Gentleman to Dover to communicate the ill-treatment Lear has received at the hands of his daughters. In the Folio, Kent, in rather indirect language, discusses spies reporting to France (the king rather than the country) rumors of the division between the dukes and their ill-treatment of Lear, but makes no mention of preparations for an invasion or of Dover. In both sequences, the general intention is to communicate that war between the sisters is threatened, and that France (and thus Cordelia) has been kept informed.
      Textual scholars have long debated the reasons for the change, which has presented editors wishing to conflate the two versions with a challenge, since simply to combine the two speeches in in some fashion leads to a rather awkward, somewhat repetitive speech. R.A. Foakes points out that the text "from the ending of 2.2 into 3.1 appears to have been subject to some deliberate reworking" (397); his solution is to replace some Quarto lines with those from the Folio, retaining the longer instructions to the Gentleman from the Quarto. Mowat and Werstine, on the other hand, combine both versions. In the process of summarizing the various arguments concerning the usefulness or vagueness of the two versions in terms of plot and language ("Revision Awry" 39-40), Richard Knowles (following Blayney) argues at length that the Folio version is "stylistically awkward if not unintelligible" (37). Arguments about the usefulness of Kent's speeches in either version in terms of plot seem curiously out of place in a play where both the passage of time and the landscape are more mythical than literal (see the General Introduction); the general purpose in each case is to provide the audience with both a sense of foreboding, in that conflicts are developing on two fronts, and a more comforting awareness that aid in some form is likely to be forthcoming. The Folio Kent's much-criticized indirection of language communicates, in addition, a deep sense of uncertainty and instability.
    3. In the final scene there are three occasions where the Folio text reworks the staging at especially dramatic moments. The first of these involves the trumpet calls to summon a challenger to answer Edgar's accusation that Edmund is a traitor (Q 4, F 5.3, TLN 3063-3067). In the Quarto, it is the Bastard who calls on the trumpet to sound, and there are no specific stage directions that indicate when the trumpet is to sound. The Folio is much clearer, with the Herald calling three times for the trumpet.
    4. The second reworking comes when a Gentleman (both texts) announces the offstage deaths of both Goneril and Regan (Q 4, F 5.3, TLN 3170-3178). The Folio version adds two short interjections from Edgar (TLN 3171, 3173), and heightens the suspense by momentarily delaying the revelation of the deaths.
    5. The final reworking of this scene is during the final moments of Lear's life (Q 4, F 5.3, TLN 3277-3286), in a passage discussed at the beginning of this introduction. The Quarto Lear asks for a bystander to undo a button, presumably to allow him to breathe more freely (a fascinating moment for the director, deciding who is the one to assist him), then groans; Edgar implores him to look up (another indirect stage direction), and Lear speaks again, "Break heart, I prithee break" (TLN 3285). Though the Quarto does not provide a stage direction, it is clear that the speech is followed by Lear's death. The Folio Lear, given additional, deeply moving lines where he believes he sees life in Cordelia, dies (there is a specific stage direction) immediately. Edgar's speech remains the same, as he hopes, no doubt, that Lear still lives, and then it is Kent who speaks the line "Break heart, I prithee break" (TLN 3285).

    6. Cumulative changes to roles

    The cumulative effect of cuts, additions, and speech reassignments has a significant impact on some of the actors' roles. I earlier expressed some doubt about the extent to which cuts do or do not improve the theatrical effectiveness of the play; in a similar vein, I am skeptical about arguments for consistent revision based on character rather than role. Character, as conceived in current criticism, is largely a construct of late nineteenth-century critics and twentieth-century acting traditions (see, for example the discussions by Slights and Holland). A number of arguments in favor of revision detail changes in characterization between the two versions. Many of these critics are subtle in their analyses, but I feel that their critical approach tends to over-value small changes in wording over the kinds of variation a good actor can instill into any passage by shifting emphasis. Lukas Erne astutely points out the profound influence that punctuation—of necessity very largely the responsibility of the compositor and/or editor—can have on meaning, and hence of the construction of character (Collaborators 18-19). In a modern production of Lear, a director can cut significant parts of a character's role without inducing changes in the interpretation of the character, since both the director and actor will have read the entire play (often in a conflated version) and can thus adduce motive to actions that are only obliquely referred to in the redacted stage version.

    Changes in roles, however, can be more reasonably explained as the result of revision. The roles of Kent, Albany, Goneril and the Fool are significantly changed in the Folio (Michael Warren, "Diminution"; Urkowitz, Kerrigan, Clayton; see also Hornback 337 on the possibility that the Fool's role was revised after Shakespeare's death to accommodate a new clown in the company); Cordelia's role is also sufficiently changed that Ioppolo finds evidence to suggest that she becomes a stronger character in F (Revising 167-83). The fundamental question remains: why were the changes made? Given (as I have suggested) that it is somewhat anachronistic to speak of character in its modern sense, how likely is it that Shakespeare, or someone else, decided specifically to make Cordelia a stronger character? Michael Warren provides a model in his essay on "The Diminution of Kent." He points out that even in the Quarto Kent's role becomes less important as others take over his role of supporting Lear in his distress—both Gloucester and Edgar care for him in the storm, then after his solitary madness he is sheltered and healed by Cordelia. Warren points out that Kent's diminution is almost wholly the result of the deletion of lines that provide "choric" (63), morally centered commentary on the events in the play. Warren's assessment of the rationale for the cuts is that they were made for theatrical reasons, to sharpen and speed up the action. So they do; but it is no less possible that audiences later in the seventeenth century lost their taste for this kind of meditative and poetic speech, and that the cuts were made more for the sake of this change in fashion.

    7. Changes to individual words

    Individual readings may provide more palpable clues to the question of revision. Many differing readings are trivial in their effect: changing words from singular to plural (kingdomes / Kingdome, TLN 7); changing similar words with similar functions (into / to, TLN 24); changing the order of words with little, if any effect on the meaning or meter (Sir a sonne / a Sonne, Sir, TLN 22), and many more. At the other extreme are individual words that are clearly distinct, modify the meaning of a passage in a significant way, and cannot reasonably be said to be the result of corruption: yonger yeares / yonger strengths (TLN 45). Between these two poles are a multitude of alterations that might be either the result of error or of deliberate change. Knowles sees no evidence of revision: "most of these changes are explainable as editorial, scribal, or compositorial error, correction, or sophistication," and he dismisses them as making "no practical difference. . . They are local improvements, not significant revisions" (Knowles, "Revision Awry" 32, 45). No doubt some of the non-trivial changes are indeed errors of the compositors, misreadings of the manuscript, or changes introduced by someone other than Shakespeare. But the sheer number of variants that carry distinct semantic meaning (anything from about 150 to 200, depending on the criteria) means that a great deal of ingenuity is needed in order to claim that the difficulty of the Q1 manuscript is a sufficient cause for the changes.

    55An important advantage of the digital edition is that it is possible to create unambiguous and accessible visual signals of textual variation using well-understood conventions associated with linking protocols. The two extended texts of Lear provide an ideal canvas for the display: each variant is indicated by a dotted underline; hovering the mouse over it triggers a pop-up with the variant reading and its source. A useful side-effect of this design is that it shows clearly on the screen the way that individual variants tend to cluster. To use an appropriately digital analogy, in a standard collation of variants, online or in print, one sees the pixels; on the pages of one of the extended texts we can see the image the pixels create. Each variant of the kind Knowles identifies as the result of error (editorial, scribal, compositorial) can be explained in one way or another; but if there is cluster of variants requiring different explanations it may well become more reasonable to establish a collective explanation. Perhaps a section of the manuscript was exceptionally difficult for some reason, or a compositor was hungry and wanted to finish quickly, or perhaps a reviser was "tinkering" with the text in some fashion (Kerrigan 195; see also Jackson 331).

    At the point in the play where Lear has realized that both his elder daughters are determined to strip him of his retinue and he has stormed off stage into the night, Gloucester, alone on stage, speaks sympathetically of the king's plight (Quarto text, Folio variants in square brackets):

    Alack, the night comes on, and the bleak [high] winds
    Do sorely rustle [ruffle]. For many miles about
    There's not [scarce] a bush. (Q 7, F 2.2, TLN 1603-1605)

    In the Folio, rustle becomes ruffle. It is easy to see that this change could be the result of a simple misreading in the Quarto manuscript, as Stone (180) suggests. But the context calls this into question. In the previous line the Folio text also changes bleak to high, and in the following line not a bush becomes scarce a bush. It is difficult to see how either of these could be the result of misreading or compositorial error. Both passages are evocative. The Quarto passage provides support for the moody bleak wind with an image of sound, as it rustles through grasses or reeds, since the landscape is empty of bushes. The Folio passage is more intense, as the wind becomes stronger and the landscape, though less bare, is ruffled—disturbed and made uneven (OED I.1.a). Perhaps the most remarkable thing about variants between the two versions is that few of them are necessary in any obvious way. While some may be explained away as errors of some kind, there are clear cases where someone decided to make some changes, to tinker with the wording, perhaps, as I have suggested, to heighten the sympathy for Lear. And if that person happened to be a writer who was an inveterate player with words, perhaps the switch from rustle to ruffle was part of that process of tinkering (compare the change from fickle to sickly earlier in the scene at TLN 1473). Gloucester's speech follows soon after the reassignments of two speeches and another alteration difficult to explain as an accident (good/best, TLN 1601); somewhat earlier in this pivotal scene there is also some significant rearrangement of speeches, entrances, and individual variants as Goneril enters to join the fray.

    No single change in this cluster could be said to change the play in any dramatic way. Even when they are taken together the difference does not substantially change the direction of the play or the roles within it. The important point, however, is that it is difficult to come up with a coherent explanation for the variants other than that there was some degree of conscious revision, even if it was only on the level of tinkering. And it is equally difficult to imagine an actor or compositor making these kinds of modifications. They are sensitive, and granular at the level of the word, such that the most likely candidate is a writer confident enough to make "pixel-level" changes in Shakespeare's text. The passage I have been discussing comes at the end of the long, painful scene where Lear discovers the truth that both Goneril and Regan are determined to strip him of his remaining powers. Though changes within this scene are patchy, there are several extended passages where there are clusters of multiple variants.

    It is also true that not all clusters of variants make particularly interesting or consistent changes. It seems very likely that the compositor had problems with the copy he was working with. In scene Q 6, F 2.1, there is a series of interesting, but quite possibly accidental variants of this kind: pretense/practise (TLN 1010), spurs/spirits (TLN 1013), Strong/Oh, strange (TLN 1015), strange news/strangeness (TLN 1026). Some variants seem to update or simplify the language: bussing/kissing (TLN 936), warbling/mumbling (TLN 999), caitiff/coward (TLN 999), rash/stick (TLN 2130), dearn/stern (TLN 2135). A constant problem in separating the accidental from the deliberate is that the general messiness of the Quarto creates a level of noise that makes it harder to discern the underlying signal; however careful an editor may wish to be, the process of distinguishing between noise and signal is ultimately a critical rather than an objectively bibliographical act. With this caveat in mind, it remains true that some clusters have the cumulative effect of suggesting strongly that a revising hand was at work, playing, at times almost punning with wording. In the opening scene we may dismiss equalities/qualities (TLN 9) as accidental, but first/fast (TLN 43), of our state/from our age (TLN 44) are difficult to explain in the same way; Confirming/Conferring and years/strengths (TLN 45) follow hard upon, and there are at least twelve similar pairs later in the scene. The same can be said of the passage in scene Q 6, F 2.1 and of the final scene of the play.

    60The question of revision: summary

    Richard Knowles, a deeply knowledgeable supporter of Q1 as the only reliable base text for King Lear, nonetheless comments: "The question is not whether there was revision—of course there was—but who did it, and when, and why" ("Two Lears?" 58). In the last forty years these three questions have elicited a wide variety of responses, and, as I have suggested above, a surprising level of passion. The truth, of course, is that none of the questions can be answered with any resounding confidence, and it takes a good deal of speculation and supposition to reach a consistent conclusion, especially concerning the why; any answer requires looking into the minds of those who might have been involved, all too often finding exactly what the enquirer is looking for. When the changes were made is interesting but not of overpowering importance: some could have been made almost immediately as Shakespeare prepared a fair copy for the company or read over a scribe's transcription of his draft, as Steven Urkowitz suggests (Revision 147); they could, as Taylor argues, have been made later when the play was being prepared for presentation at the indoors Blackfriars theater ("Date" 428); they could have been made when circumstances of some kind made it desirable that the French connection with the invasion of French forces headed by Cordelia be downplayed; or at some other time and for some other reason. The crucial question is who made the changes.

    Lying behind the whole discussion of the difference between the two versions of King Lear is a very simple question: was it likely that Shakespeare revised this, or any, of his plays? Two decades ago, Trevor Howard-Hill succinctly responded to this basic issue:

    . . . modern scholarship is disposed to entertain the conception of a Shakespeare whose compositional process was more usual than the conventional compliment of his colleagues Heminge and Condell suggests: if they "scarce receiued from him a blot in his papers" it was, we may assume, because having finished his rough draft . . . Shakespeare found it necessary to write out his play fairly (or, to have a transcript made). The preparation of a fair copy gave him an opportunity (or site) for second thoughts as well as to sort out certain unfinalized details in the first draft, that is, to tinker with forms of expression or rewrite certain speeches or parts of scenes. ("Two-Text" 32)

    If only some of the changes were made by Shakespeare, it means that the Folio in some measure records a later version of the play as he created it. My construction of a narrative to explain the variations between the two texts relies to a significant degree on a belief that the Folio, at least in some respects and in some passages, represents a conscious revision of the play by Shakespeare, though his is almost certainly not the only agency in the changes. The well-attested influence of a copy of Q2 on the Folio text sharpens the case for revision, since it would have been easier in the printing house simply to reproduce a version of Q2 than to work from manuscript or a heavily annotated print text; the motive for printing the modified text may have been that it was simply more current, but it is certainly possible that Hemminge and Condell chose it because they knew that it recorded more recent work by Shakespeare. It should be clear that I do not find it convincing or probable to assign all changes to the kinds of corruption and interference that Knowles lists ("editorial, scribal, or compositorial error, correction, or sophistication"); even he acknowledges that these errors cumulatively can only account for "most" of the changes.

    The evidence for revision rests on two kinds of differences between Quarto and Folio: a relatively small but arresting number of additions difficult to pass off as errors of omission or contributions by another hand; and demonstrable clusters of individual variants, with each group showing evidence of some variants that cannot readily be explained away and a multitude of others that could either be chronic error (not impossible, given the state of the Quarto text) or tinkering and touching-up by a writer who was habitually given to word play of one kind or another. The opening and closing scenes of the play combine both kinds of change—additions and changes to individual words—and there are sequences in scene Q 6, F 2.1, where there are significant changes to staging coupled with clusters of variants. The incidence of changes suggest that whoever made them did so in a kind of "hit and run" pattern, focusing interest on specific sections, leaving long sections of the play unchanged.

    It is my conclusion that the Folio version of King Lear records some second thoughts by Shakespeare as he revisited the play at some stage after he originally wrote it. He tinkered with wording in passages that caught his eye and added some generally short passages to enrich or modify key passages. In some places his may have been the hand that reworked the staging of some sequences; I also think it likely that some minor additions to the Folio text were made by other hands. Though I do not believe that the cuts can with confidence be assigned to any specific agent, many of them create fascinating critical and staging questions. Collectively they have a substantial effect on the play, in some ways intensifying the kinds of changes that Shakespeare wrought on the old story of King Leir. Although the revisions by themselves do not make the Folio King Lear a significantly different play, they do, as I discuss in the General Introduction, change its patterns of meaning and emphasis in critically and theatrically interesting ways.