A Response to Steven Urkowitz and Michael J. Hirrel

Scholars have long debated the relationship between the "received texts" of Shakespeare's longer plays (as they have come down through the editorial tradition) and the versions of these works that would have been performed in early modern playhouses. Specifically, we disagree as to whether or not these longer works would have been abridged in performance by Shakespeare's company. This question is of great importance to theatre companies who wish to emulate the staging conventions under which these plays were first performed (a movement often referred to as "Original Practice" or "OP"). For such practitioners, the question of "what they did then" when abridging (or not) the longer plays for performance informs their choice of "what we do now." Some Original Practice companies, including the Atlanta Shakespeare Tavern, stage nearly complete versions of the plays, because they believe that is what Shakespeare's company would have done. Other adherents of OP, including Brent Griffin's Resurgence Theatre Company, abridge the longer plays to a performance time of approximately two hours. They do this, in part, because they believe that the historical record establishes a precedent for such abridgment in Shakespeare's time.The most recent (to my knowledge) scholarly exchange on this topic occurred a few years ago. My article "'This is too long': A Historically-Based Argument for Aggressively Editing Shakespeare in Performance" asserted that the longer plays would have been abridged in outdoor performance at the Globe, and also possibly indoors at the Blackfriars.1 Steven Urkowitz's "Did Shakespeare's Company Cut Long Plays Down to Two Hours Playing Time?" argued instead that these works "primarily were played at full length (or approximate full length) for public, private and courtly performances in London and on tour."2 In "Duration of Performances and Lengths of Plays: How Shall We Beguile the Lazy Time?,"3 Michael J. Hirrel largely agreed with Urkowitz. This present article is a response to the contributions of my two fellow scholars on this topic.As Urkowitz notes, "Much of the same basic information appears in arguments both for and against" theatrical abridgment.4 Researchers therefore tend to interpret the scant historical record according to the prejudices and priorities of their respective disciplines. Urkowitz is a literary scholar and embraces the desire to maintain texts in a pristine form. For him, acting companies that would abridge Shakespeare's text are "purveyors of adulterated sausages or watered milk" who "regularly debase and reduce the artistic value of the plays they stage." He believes that companies should naturally prefer "the longer and better rather than the shorter but worse" version of any Shakespeare play.5 My own position, by contrast, is that of a theatre practitioner. I do not determine whether a playscript is "better" or "worse" by its absolute "artistic value" (if such a thing can be objectively determined), but rather by the extent to which it serves the needs of the performance in question. Theatre practitioners see playscripts as raw material, which is regularly abridged, adapted, and rearranged for a particular production or performance. Urkowitz accuses Andrew Gurr of lacking theatrical savvy by "passing over the difficulties" involved in revising plays to accommodate theatrical exigencies such as "accomodating a temporarily missing player." Such efforts, Urkowitz asserts, are "vastly more complex in practice than Gurr's hypothesis suggests."6 In fact, it is Urkowitz who betrays naivete in this matter. The kind of changes described by Gurr are commonplace in twenty-first-century theatrical production. Scripts are routinely cut, and parts reassigned, to suit the needs of the moment. Rather than being transcribed, as Urkowitz assumes,7 such alterations often remain undocumented. It is of course impossible to know for certain if the behavior of Shakespeare's company resembled that of today's practitioners. However, the evidence that Tiffany Stern has documented about the instability of dramatic texts on the early modern stage suggests that actors in the period did not share Urkowitz's reverence for the "longer and better" versions of Shakespeare's plays.8Urkowitz characterizes scholars in the opposing camp as contending that Shakespeare's longer plays "would never have been acted as written."9 This is not, however, a fair statement of our position. Urkowitz is probably right in chastising Stephen Orgel for dogmatically insisting that "every play . . . would normally have been cut for production," and that a period of "two to two-and-one-half hours" for all performances in this period is "universally accepted" (although, in the same paragraph, Orgel acknowledges that "private performances tended . . . to be longer than those in the public amphitheatres").10 There is, however, less rigidity in Andrew Gurr's notion of "Maximal and Minimal Texts," which asserts that performance scripts were "trimmed and modified, in varying degrees of substantiality, out of the maximal playscript."11 Ernst Honigmann suggests even more flexibility when he asserts that, because "the actors might perform their plays at the Globe or Blackfriars, at court, at private houses, or on provincial tours," any dramatic work of Shakespeare's had to be "infinitely adjustable."12 My own position is that, while the longer plays may have sometimes been performed indoors in their longer versions, they were normally shortened for performance at the public amphitheatres. My logic does not rely, as Urkowitz suggests, on "the existence of a hitherto unacknowledged police power—or perhaps a self-regulating discipline among acting companies, or some other market pressure imposed by impatient audiences at the professional theatres in London."13 Rather it depends on a simple and banal fact: without artificial lighting, one cannot successfully stage outdoor drama after dark.Beginning in October 1594,14 companies were ordered by the Lord Chamberlain to have their performances "begin at two and have done between fower and five.'"15 Scholars have understandably questioned whether these companies, left to their own devices, would have chosen to stop performances precisely at 5 p.m., as this order requires. The 2 p.m. starting time, however, is not in dispute. Furthermore, Hirrel suggests that "no one expected the plays themselves to begin promptly at two"16 because "preliminary entertainment took a half hour."17 If Hirrel is correct, the featured comedy or tragedy would have thus begun no earlier than 2:30 p.m. Even if Hirrel overstates the prevalence of pre-show entertainment, one must consider that for much of the early modern period plays were typically followed in performance by a "jig."18 More than a simple dance, jigs were bawdy musical sketches that David Wiles describes as "a form of soft commercial pornography."19 These popular pieces were up to six hundred lines in length.20 With that much dialogue and the added time required for song, dance, and comic shtick, jigs would easily require thirty minutes of the maximum three-hour performance window proscribed by the Lord Chamberlain. The pre-show entertainment described by Hirrel, for which there is less documentation, would have eaten up even more time.A table produced by R. B. Graves indicates "Sunset Times in London, 1598" as ranging between 3:48 p.m. and 5:40 p.m. between October and March, the height of the theatrical season.21 Urkowitz nevertheless claims, "There were no obligatory constraints imposed by the physical phenomenon of gathering night-time darkness" during outdoor performance.22 He justifies this assertion by noting that Graves has determined by "close observation of lighting conditions at the rebuilt London Globe"23 that "an audience watching a play lit by natural light will gradually grow accustomed to the decreasing light and may easily find viewing comfortable well past sunset."24 In Graves's original text, however, this sentence is a more general description of outdoor performance and not specifically tied to the Globe. Of this reconstructed amphitheatre, Graves notes instead, "the waning daylight at the new Globe is entirely adequate for most modern spectators half an hour past sunset."25 Acted at the rate of 1,200 lines per hour, which Hirrel considers "conservative and achievable"26 but which contemporary practitioners have struggled to replicate,27 the 3,674 lines of Q2 Hamlet would take approximately three hours and four minutes to perform. Beginning at the 2:30 start time suggested by Hirrel, such a performance would have concluded at 5:34 p.m., an hour that from October 14 to February 7 occurred more than "half an hour past sunset" in London.28 Such a performance would have ended, as James Shapiro puts it, "with Hamlet and Laertes sword-fighting in the dark."29In an effort to overcome this obstacle, Hirrel claims that "outdoor theatres had artificial lighting." This is the most important assertion in his article. Perhaps due to limitations of space, Hirrel does not give this topic the attention it merits, confining discussion of artificial lighting to a single paragraph.30 Graves writes of outdoor playhouses that "the evidence for artificial light is slim at best"31 and depends upon Randle Cotgrave's 1611 definition of the French word "Falot" as "A Cresset light (such as they use in playhouses)."32 While Graves believes that cressets may have been used "as property lights or as lights to illuminate dark passageways toward the end of winter performances" he doubts that they "furnished the entire illumination of the theatre."33 Hirrel dismisses Graves's skepticism, claiming, "In any event, the part of the theatre the cressets were used to light is not the issue."34 A twentieth-century analogy reveals this statement to be rash. As longtime residents of Chicago will remember, there were no night games at Wrigley Field until light towers were installed in 1988. For decades before this, electric lights were used to illuminate passageways leading to the grandstands. No one would confuse these dim bulbs with the massive towers required to illuminate the playing field. Graves concludes, "Considering the size of the area to be lit, the desirability of using as many lights as possible, and the problems English weather might have caused in keeping them lit, I am skeptical that the artificial lights at the actors' disposal played a major role in the general illumination of the amphitheatres."35 Hirrel produces no new evidence to challenge Graves's expert opinion.The lengths of extant early modern play texts offer clues as to the duration of performances in this era. Hirrel is skeptical of Alfred Hart's attempts to determine an average length for these documents, and of Hart's conclusion that acting editions of approximately 2,400 lines would have existed for most plays. Hirrel writes, "Hart's fascination with averages seems to have blinded him to a crucial point. As his own data show, Elizabethan playtexts range widely in length, and those varying lengths are distributed normally (in statistical terms) within the range. Twenty-four printed plays written or performed between 1594 and 1616 hold fewer than two thousand lines. Twenty-seven have more than three thousand."36 What Hirrel doesn't say is that, of the 185 plays under consideration in this sample, 134 had lengths between 2,000 and 3,000 lines. Plays over 3,000 lines and those under 2,000 lines were therefore equally rare. When the range of dates is broadened to encompass 1590 to 1616, only 29 of 233 plays (approximately 12.5%) were over 3,000 lines long, as opposed to 48 plays with texts under 2,000 lines (approximately 20.5%).37 Long plays were therefore the exception rather than the rule.The most common figure cited for length of performance in early modern play texts is two hours, most famously stated in the Prologue to Romeo and Juliet, but also in The Two Noble Kinsmen, The Alchemist, Hengist (another King's Men play), The Duke's Mistress, and Love's Pilgrimage.38 These quotations, among other evidence,39 lead Andrew Gurr to conclude, "Two hours was the standard time of performance."40 Urkowitz acknowledges the many references in prologues to two-hour performance times, but notes that this evidence is sometimes contradicted by title pages, such as that of Q2 Romeo and Juliet, which announces that nearly 3,000-line edition to be "As it hath bene sundry times publiquely acted."41 Urkowitz gives equal weight to the "contradictory testimony" of title pages and prologues, and therefore "questions their reliability as evidence for duration."42 The problem with this logic is that, while Tiffany Stern cautions that prologues were among the most volatile and frequently cut pieces of early modern playtexts,43 they were nevertheless written to be performed (at least occasionally) and therefore figure into a discussion of performance scripts. Title pages, on the other hand, are exclusively literary documents. Q2 Romeo and Juliet may have been "publiquely acted" in its entirety at some point (most likely at court or some other indoor venue), or the claim may have been just a marketing ploy—comparable to the "Now a Major Motion Picture" blurbs that sometimes appear on paperbacks. Just as such novels contain much that does not make it onto the screen, so Q2 Romeo and Juliet included more than was acted at the outdoor amphitheatres.If one is willing to consider evidence from title pages, several of these suggest that published plays in this period were frequently longer than the versions that were publicly performed. The title page of John Webster's 1623 edition of The Duchess of Malfi announces that it is the "perfect and exact Coppy, with diverse things printed that the length of the Play would not beare in the Presentment."44 Humphrey Moseley, in his address "The Stationer to the Readers," which prefaces the 1647 collected Comedies and Tragedies of Francis Beaumont and John Fletcher, notes that "When these Comedies and Tragedies were presented on the Stage, the Actours omitted some Scenes and Passages (with the Authour's consent) as occasion led them," but that in this volume "now you have both All that was Acted, and all that was not; even the perfect full Originals without the least mutilation."45 Similarly, the title page of the 1600 edition of Ben Jonson's Every Man Out of His Humor claims that this text is "AS IT WAS FIRST COMPOSED by the Author B. I. Containing more than hath been publikely Spoken or Acted."46 Hirrel asserts that Jonson "was not likely to offer scripts to acting companies that were too long for them to use,"47 but the title page of Every Man Out announces that this is exactly what Jonson did. These quotations, along a with comparable example from Richard Brome's The Antipodes (1640), lead Eric Rasmussen to conclude that in the early modern era "scripts were often deemed to be too long and were cut in order to reduce overall playing time."48Further evidence for abridgment comes from within the texts of several plays by Jonson and Shakespeare. The Induction to Bartholomew Fair, for instance, claims that this play will occupy "the space of two houres and an halfe, and somewhat more."49 At 4,344 prose lines,50 Jonson's Induction underestimates the performance time required for the full text of this comedy. Even at the aggressive rate of 1,200 lines per hour proposed by Hirrel,51 an unabridged Bartholomew Fair would last over three hours and forty minutes. David Klein's claim that Jonson's "play could very well have been done in the time stated, for its natural tempo would call for rapid action and dialogue"52 is merely wishful thinking. Similarly, the 3,066 lines of The Alchemist53 are too many to be acted in the "two short hours" promised by Jonson's prologue to that play.54 Most famously, the 3,185 lines of Romeo and Juliet cannot be presented in the "two hours traffic of the stage" mentioned in its opening speech.55 The most likely explanation is that the published texts of all these plays are longer than what was publicly performed.Many scholars, including Gurr and Lukas Erne, see evidence for theatrical abridgement in what were long known as the "bad" quartos of several Shakespeare's plays. While there is never total unanimity in Shakespeare Studies, a fairly strong consensus now exists that these quartos represent, however imperfectly, versions of these plays as they would have been publicly performed during the early modern era. Support for the "bad" quartos as records of performance comes from another Shakespeare play, which paradoxically survives only in its Folio version. Macbeth at 2,084 lines is the shortest of Shakespeare's tragedies.56 According to Nicholas Brooke, the received text likely represents "a version cut for performance."57 Grace Ioppolo concurs that the play's "numerous plot inconsistencies do suggest cutting."58 More subjectively, the characters of Macduff and Lady Macduff are underdeveloped during the early acts of Macbeth. We will never know if Shakespeare's original text included earlier moments featuring this couple, but William Davenant (who claimed intimate familiarity with Shakespeare's intentions) included such scenes in his Restoration adaptation.59 The brevity of the received text of Macbeth suggests that Shakespeare's other long plays may have been reduced to a similar length on the public stage. As Nicholas Brooke observes, "If the 'bad quarto' of Hamlet, which certainly derives from performance, were fleshed out with more perfectly remembered words it would be only slightly longer than Macbeth and few would notice the omissions."60Urkowitz claims (with some justification) that the arguments of those who argue for the "bad" quartos as performance texts are circular.61 Having already decided (Urkowitz would say arbitrarily; I would claim on the basis of the other evidence cited in this essay) that performance scripts for early modern amphitheatres ran no longer than two hours, we then deny this status to any text that does not conform to this restriction. Urkowitz points out that one documented instance of theatrical adaptation in this era, the conflation by Sir Edward Dering of the two parts of Henry IV, created a single "3100-line cut down script that would require a playing time in excess of anything that could be played in two to two-and-a-half hours."62 He neglects to mention, however, that Dering was (in the best sense of the word) an amateur. A cast list discovered along with the manuscript in 1844 lists "the names of relatives, friends, and neighbors of Sir Edward's," suggesting that Dering's theatrical abridgment was intended for a private performance at his estate.63 Samuel Burdett Hemingway allows that Dering's adaptation also could "have been prepared for one of the performances at court in 1613,"64 which is possible. Graves notes that performances lit solely by candlelight, which could extend beyond nightfall or even begin after dark, were common at court. Such spectacles operated under no time constraint and could have presented all of Dering's 3,100 lines. But candles were expensive, and "the court had resources of which most early Tudor acting companies never dreamed."65 The Dering manuscript is therefore in no way connected to outdoor performance at the public amphitheatres. Urkowitz's citation of a 3,200-line conflation of the two parts of Heywood's Edward IV, "marked for cutting in preparation for a professional production as a single-play version," bolsters his argument better than his analysis of Dering.66 There is not, however, enough information in the text or notes of Urkowitz's Shakespeare Bulletin article to allow readers to evaluate this evidence.Hirrel notes that, while early modern writers "do conventionally describe play times as two hours, they also frequently describe the span as three hours."67 Among other evidence, he quotes Thomas Dekker's promise of "three howres of mirth" in If it be not Good, the Devil is in It.68 Although this does indicate that some plays may have lasted as long as three hours, it also raises an important point about pace. Urkowitz claims that we can assume a pace of 1,200 lines per hour for early modern performance, because this speed has been "achieved today by troupes unencumbered by modern production values of scenery, variable lighting and intermissions."69 More specifically, Hirrel notes that the American Shakespeare Center in Staunton, Virginia, has long pursued this rate of performance.70 However, the ASC's longtime Artistic Director Jim Warren has acknowledged that a pace of 1,200 lines per hour is not feasible, and that "2,200 lines is a better bet for us . . . to hit at/under the two hour mark," thus defining an actual maximum pace of 1,100 lines per hour.71 Patrick Tucker's Original Shakespeare Company, doctrinaire in their efforts to recreate the conditions of early modern performance, has not been able to exceed 1,000 lines per hour.72 Ironically, evidence of a slower performance pace in the early modern era comes from the Dekker play cited by Hirrel. If the 2,700 lines of If It Be Not Good did take three hours to perform, this would indicate a rate of only 900 lines per hour. This speed of delivery would produce a Second Quarto Hamlet of over four hours. This is too long for performances that had to "begin at two and have done between fower and five.'"73A similar question about pace arises from internal references within The Tempest. Prospero states near the beginning of this play that the hour is "At least two glasses" past mid-day (1.2.240–41), meaning after 2 p.m. Near The Tempest's conclusion, Ariel announces that the time is now "on the sixth hour" (5.1.4). These two quotations define a possible length of performance for the play, which Hirrel calculates at four hours.74 Gurr uses these same references to clock the play's duration at three hours. The difference hinges on the phrase "On the sixth hour," which Gurr reads as "meaning 5 p.m.,"75 but which Hirrel glosses as "between five and six, but approaching six o'clock."76 If Gurr is correct, and the 2,015 lines of The Tempest took three hours to perform, then the 3,264 lines of Cymbeline would take four hours and fifty minutes. If Hirrel is right, and the Tempest took four hours, then Cymbeline would last approximately six hours and a half.77 There is little evidence that performances at the Globe or Blackfriars ever lasted as long as these estimates for Cymbeline suggest.Wags have long opined that academic battles are so fierce because the stakes are so low. The debate surrounding the nature of early modern performance texts, however, has important contemporary ramifications. Theatre practitioners who believe that Shakespeare's longer plays were originally staged in their entirety tend to be, as Gurr puts it, "reluctant to cut the sacred text."78 They therefore often stage performance scripts that are too long for the needs of their audiences. The work of Urkowitz and Hirrel, I fear, encourages this conservative pattern. Meanwhile, companies who aggressively abridge these plays to accommodate the exigencies of their particular theatrical moment should know that they are not betraying tradition. Instead, they emulate a practice pursued by Shakespeare's original company, at least in outdoor performance. My response to Urkowitz and Hirrel will hopefully provide such practitioners with scholarly support for their endeavors.