Duration of Performances and Lengths of Plays: A Response to Objections

It's a pleasure to participate in this colloquy with my two good friends, Joe Falocco and Steven Urkowitz. To me, they are Joe and Steve, but here I shall use their patronymics. All three of us are concerned about the time given for the performance of Elizabethan plays.They both come at this problem from a perspective different from mine. They are chiefly concerned about how Original Practices acting companies should perform Shakespeare's plays today. I understand the problem, and I don't think there's an easy answer for the OP companies, but I am focused on a different issue. What did professional London acting companies do in their theaters from roughly 1594 to 1616? In answering that question, I come at the problem from an unusual perspective. I'm not an academic. I'm a trial lawyer. In my profession, you can't form a theory first and then cherry-pick conforming facts. If you do, you will lose. You must examine the facts first, all of them. And only then form your theory, a theory that explains them all.In my 2010Shakespeare Quarterly article I sought to deal with this problem—the amount of time that Elizabethan professional acting companies themselves devoted to the performance of their plays—on a comprehensive basis. Yes, as Falocco suggests, I was rather severely constrained by SQ's word limit. But I sought to look at all the evidence, and to provide a comprehensive explanation. I began by considering the speed at which lines in Elizabethan plays were originally delivered. Then I divided the argument into three sections. I aimed first to show that performance events in Elizabethan theaters, including both the principal plays themselves and incidental entertainment, ran almost four hours. Second, that the time devoted to plays themselves varied considerably. Third, that the remainder of the four-hour period was taken up by incidental entertainment, only about a half hour of preliminary entertainment indispensable. That left a practical maximum of about three and one-quarter hours for plays themselves.1In this article I propose to address objections that have been offered since 2010, in the same topic order as the original article. Not just Falocco's objections, but those of other scholars, each of whom, as fortune has it, is also a good friend.In my article I estimated that Elizabethan plays were delivered at a speed of about twenty lines per minute. This wasn't my estimate originally. It was that of Alfred Hart, the original proponent of the idea that Elizabethan plays were delivered in two hours.2 Hart supported his estimate with good evidence, and I added more. The American Shakespeare Company in Staunton, Virginia, for example, is able to deliver plays at that speed. When they experimented with Elizabethan pronunciation, they were able to deliver the plays even faster.Falocco responds, both here and in his Shakespeare Bulletin article, that OP companies have found it difficult to deliver plays at that speed to contemporary American audiences.3 That may be so, but it doesn't tell us much about the speed at which Elizabethan acting companies delivered plays to their audiences. We have no basis to assume that Elizabethans spoke English at the same speed as twenty-first-century Americans. Even today, languages are spoken at different speeds in different geographic areas. I grew up in Mississippi and went to college in Boston. I can report that English is spoken faster in New England than in the South. I speak Spanish fluently as well, which I learned in Madrid. When I speak Spanish with Argentines, it seems as if we are talking on a bed of molasses. But when I go to Puerto Rico or the Dominican Republic—it's the same in Cuba—it takes me a day to become accustomed to the super rapid speed. Syllables are clipped and words are blended one into another without pause.We can't be certain, of course, about the speed at which Elizabethans spoke English. But several indications suggest that the pace of spoken language has slowed since that time. Back to Spanish. Puerto Rico, what is now the Dominican Republic, and Cuba were occupied by Spain early in the sixteenth century. They are all islands. It seems probable that their patterns of speech partly reflect the patterns of Spanish as it was spoken in the sixteenth century. In English, if 1930s movies are still played at original running speeds, it seems clear that Americans spoke English faster than they do now. With respect to Elizabethan English, scansion of the metered lines in plays and poetry shows that many syllables that we pronounce, they did not. Elizabethan pronunciations would also have sped spoken speech. "Love" as we pronounce it, for example, requires more breath and time than "move," which is how they pronounced "love."The best evidence for the speed at which Elizabethans spoke English is discussed in my SQ article. Thirteen plays are described by the Elizabethans themselves as taking two hours. The average length of those plays is 2,562 lines. At that length a play would be delivered at 21.3 lines per minute, considerably more rapidly than my own estimate. That rapid speed is corroborated by a statistical cross-check. The lengths of the thirteen plays are more or less normally distributed along a "bell curve." But the shortest would, at 21.3 lines per minute, have been delivered in one hour and 32 minutes, and the longest in two hours and 24 minutes. Just about the range one would expect for plays described as taking two hours.4The number of those thirteen plays and the average of their line lengths are important for this purpose. Elizabethans just weren't into precision. You can't take any one example and assert that it means something. In his accompanying article, Falocco points out that Dekker describes his play If This Be Not Good as providing "three howres of mirth," an elapsed time that would require a delivery speed considerably slower than 21.3 lines per minute. But If This Be Not Good is also described by Christopher Beeston, in a preface to Heywood's Apology for Actors, as taking two hours, which would require a delivery even more rapid than 21.3.5I next collected in my SQ article evidence concerning the length of the Elizabethan performance event. The evidence shows that performance events lasted almost four hours. That was true before the mid-1590s, when performance events began between 3:30 and 4:00 p.m., after evening prayer, and ended between 7:30 and 8:00 p.m. And it was true after the mid-1590s, when performance events began at 2:00 and ended near 6:00 p.m. Scholars who still want to believe that plays could last no longer than two hours have never come to grips with that evidence. Let me deal with their scattershot objections.Richard Dutton suggests that I have misinterpreted events before 1594. He is mistaken. Yes, both the Queen's Company and the City of London complained about the inconveniences caused by performances in the open theaters in Shoreditch after dark. But that only shows that such performances did in fact occur. Why would they complain about something that didn't happen? Dutton also consistently confuses the lengths of the performance events in this period with the lengths of time devoted to the plays themselves. Both before 1594 and after, as I hope I showed, incidental entertainment occupied the residual time not taken by the plays themselves.6Dutton goes on to suggest that the most reliable evidence regarding plays' ending times after 1594 is a promise made in 1594 by the Lord Chamberlain's actors.7 The City had long objected to performances that began before the end of evening prayer, around 3:30 p.m. Somewhat inconsistently, the City also complained about suburban performances that ended, in winter, in dark. The Lord Chamberlain's actors wanted to begin at 2 p.m. In exchange, they promised, they would end their performances "betwene fower and fiue."8 The promise thus proves that the actors were willing to say what their prospective benefactors wanted to hear. But it proves nothing more. The evidence, much of it cited in the SQ article, is overwhelming that after 1594 performance events ended near 6 p.m. That evidence concerns not promises but actual events.Falocco, in both his accompanying and SB articles, argues that plays could not have continued past sunset. He isn't the first person to suggest that Hamlet could not have held up Yorick's skull in the dark. But this is, again, a cultural assumption by folks who assume that our culture also applies to peoples of the past. We are accustomed to artificial light. Elizabethans were accustomed to dealing with darkness and light lent only by the moon and the stars.Let's put our cultural assumptions aside and look at the evidence I adduced in my SQ article. Phillip Stubbs reports in 1583 that people attend the outdoor theaters in Shoreditch both "night and daye." Gervase Babington reports in the same year that plays in "open courtes" are especially dangerous "on the night." The City of London sarcastically responds to the Queen's Company petition in 1584 that "if in winter the dark do cary inconvenience," then let "no playeing be in the dark." Philip Henslowe loaned money in 1598 to the Lord Admiral's Company "when they fyrst played dido at nyght." John Webster complains in 1612 that his play, The White Devil, was acted in winter "in so open and blacke a Theater, that it wanted . . . a full and understanding" audience.9Even if we ignored all that evidence, there would be a further problem. I laid out substantial evidence that before the mid-1590s, performance events began at the two outdoor theaters in Shoreditch, the Theater and the Curtain, after evening prayer ended, between 3:30 and 4 o'clock. Plays could not have begun at those times and ended in winter before dark. Let's assume, against the evidence, that plays themselves began at about 3:45, without preliminary entertainment. A two-hour play beginning then would, from mid-October through mid-February, have ended in the dark. A play beginning at 3:45 in mid-December began at sunset, leaving only about a half-hour until the end of twilight. Would a half-hour have been sufficient? Kyd's Spanish Tragedy usually is dated around 1587. Its original version contains 2,898 lines. At 20 lines per minute, it required two hours and twenty-four minutes. It could not have ended until 6:10, from October through February, well past dark.10Falocco assumes that plays could not have continued past dark without artificial light. The evidence says that they did continue past dark. It doesn't matter whether this was done solely by the light of the stars and the moon or otherwise. But artificial light probably was used, chiefly by cressets, suspended iron pots holding burning oily substances. I laid out the evidence in the SQ article. It isn't just from Randle Cotgrave's 1611 French/English dictionary, but also from court testimony in 1603 that "cressett lights" were used at the Boar's Head, an outdoor theater, "in Wynter." And R. B. Graves shows, extensively, that cressets were used in this period to light outdoor Continental European stages.11The fact that cressets were used at the Boar's Head only in winter is significant. Cotgrave particularizes the use of cressets to "Play-houses." Elizabethan theaters did have a short passageway, the hall between the entrance and the yard and galleries. But that short hall did not distinguish playhouses from many other buildings. So it seems unlikely that the cressets were used to light the halls. Far more likely, they were used, as on Continental outdoor stages, to illuminate the stage. Even if they were used to light the short hall, they were used for that purpose only "in Wynter," so we still know that the theater was in use after dark.Graves does tentatively conclude, as Falocco notes, that artificial lighting was not used to illuminate the Elizabethan stage. But Falocco fundamentally misunderstands Graves's argument, which is that as darkness fell upon the stage, the eyes of Elizabethan audiences gradually became accustomed to the lessening light. A sudden illumination of the stage by artificial light would momentarily have blinded them. But Graves notes that if cresset lights were lit from the beginning of the play, their lighting effects would have been incorporated by the human eyeball as daylight diminished into darkness. In any event, Graves isn't saying that plays didn't continue past dark; he's saying that they did. He just doesn't make the same cultural assumption as Falocco, that artificial lighting was necessary.Graves points out instead that the many stage directions calling for candles and torches would have been useless in broad daylight. No light from those objects would have been apparent to the audience.12 As Urkowitz points out in his accompanying article, moreover, the ending of Romeo and Juliet would have been quite effective in the dark, under the light of torches. And what about Hamlet, by the way? In the play's timeline it ends at night. Wouldn't its catastrophe have been especially effective under torchlight?I argued next in SQ that within Elizabethan performance events the lengths of time devoted to performing the plays themselves varied substantially. From about an hour and a half to about three hours and a quarter. I based this argument on three categories of evidence, that is, the wide variety in the lengths of printed plays, the absence of any apparent effort in editing playhouse manuscripts to reduce the playscripts to any standard length, and the numerous references to plays requiring not just two, but three, hours to perform.Falocco and Lukas Erne aim almost exclusively at the first category of evidence, the wide variety in the lengths of printed plays. When I consider their objections, I am once again puzzled by the epistemological methods of my academic colleagues.13 If you want to know the truth, you must consider all the evidence. You can't say that this fact or that fact supports my theory, so therefore my theory is right. My point was that play performance times varied substantially. That explanation accounts for all the evidence.The wide variety in the lengths of printed plays is significant. No doubt, a play as printed was not necessarily the same play text that was performed. It would depend upon the manuscript the printer used. If he used a playhouse manuscript, the printed text probably was performed. If he used an authorial manuscript, the performed version probably contained variations. But there is one constant in both cases. The underlying scripts were written for for-profit acting companies by professional writers. This was a commercial market. In such a market, sellers offer goods intended to meet buyers' expectations.The sellers in this case, the authors, knew what their buyers expected. They sold their proposed plays to the acting companies, as Henslowe's records show, in a regular procedure. First, they prepared a synopsis, a "plot," and/or sample scenes, which they showed to the acting companies and discussed with them. In those discussions, the actors almost certainly were envisioning performance of the plays, and thus communicated to the writers what they were expecting in the completed scripts. Add a comic scene here. Expand a role there. And so forth. The discussion surely included an envisioned running time. The actors were considering how long they thought the play could sustain their audiences' interest, and what would be the running times for the incidental entertainments they hoped to pair with the play. The writers often were paid a small advance after these conferences, but they were paid the bulk of their fees after their manuscripts were accepted.14What would have happened if the writers submitted manuscripts shorter or substantially longer than the ones the actors wanted? We're not talking here about cultural assumptions, but about how commercial markets function. That hasn't likely changed since the beginning of markets. In our own times, I can report from personal experience in the market for print newspaper op-eds. Newspapers have set word limits. Submissions slightly exceed those limits with easily cuttable penultimate paragraphs. Other submissions are almost always rejected. Similar results are described by the many manuals on writing scripts for TV shows and movies. TV shows have fixed time limits, of course. Movie producers, if they accept a script pitch, tell the writers the expected running time. The manuals all say that if you substantially exceed those limits your script will be rejected. A little extra length is again acceptable, in case something must be cut for reasons unrelated to length.15Would the result have been different in Elizabethan times? If the submitted manuscript was substantially longer than the one the actors wanted, who was to do the necessary cutting? If it was shorter, who would write the additional lines? As they rehearsed the actors could have deleted lines they didn't think were working and made minor emendations. And some actors, such as Shakespeare himself, gradually transformed themselves into playwrights. But it seems improbable that the actors generally set about redrafting their playscripts. Their time necessarily was occupied principally with memorizing, rehearsing and acting their ever rotating and evolving list of plays. As Henslowe's records show, moreover, they were also occupied with numerous tasks related to running their business—hiring employees, going to court, acquiring costumes and properties, and of course acquiring playscripts.Writing those playscripts was what the actors hired the writers to do. They were running a business; they had to make rational business decisions. If the writers delivered scripts to them that varied substantially from what they had requested, they rationally would either have rejected the scripts altogether or demanded revisions before acceptance and payment. They would not have paid for the deficient scripts and then undertaken to rewrite them. Knowing that this would be the result, the writers themselves presumably delivered the scripts that were asked of them.Printed playscripts vary widely in length: at the lowest end around 1,600 lines, about an hour and twenty minutes running time. If those plays actually ran two hours, why weren't the added lines published? At maximum, printed play texts ran to about 3,400 lines, a running time of almost three hours. Most plays fell in the middle of these extremes. They are, as noted in my SQ article, normally distributed along the bell curve. But they all varied in length too. Seven plays are longer than 3,400 lines.16 At the outer extreme are Shakespeare's Q2 Hamlet, 3668 lines; Folio Richard III, 3,570 lines; Jonson's Every Man Out of his Humor, 4,163 lines; and Bartholomew Fair, 4,179 lines.17 The Q2 Hamlet and the F RIII would have required a few minutes more than three hours. EMO and Bartholomew Fair would have required about three hours and a half. I conjectured in SQ that the maximum time for plays themselves was about three and one-quarter hours. If so, both the Q2 Hamlet and the F RIII could have been performed in full. Some lines from EMO and Bartholomew Fair necessarily were cut. But it wouldn't have been difficult to cut about 263 and 279 lines from them, respectively. Both Erne and Falocco, in his accompanying article, point out that Jonson's Induction to Bartholomew Fair makes a claim as to the play's running time. What shall we make of Jonson's candor here? Should three and one quarter hours have been described as "two houres and an halfe, and somewhat more"? Literally speaking, Jonson was not incorrect.Most of the longest plays, yes, were written by Shakespeare and Jonson. I discussed why that probably was so in my SQ article. I don't have much to add. Shakespeare and Jonson apparently were the prestige playwrights. Jonson insisted upon £10 per play when lesser playwrights were paid only £6. Presumably, audiences wanted to hear more of their plays and less of incidental entertainment. I did not, as Erne suggests, "imagine" that Jonson was attentive to the needs of the customers for his writing. That very precisely is Dekker's caricature of Jonson as Horace in Satiromastix. Jonson, like any other professional writer, needed to sell playscripts to acting companies willing to buy them, and, in his case, willing to pay £10 for them.Falocco and Erne tax me for failing specifically to acknowledge title pages to published plays that say the accompanying text contains passages not performed. There are exactly four such title pages, among about 600 plays published during the period. The claim is, for example: "Containing more than hath been Publickely Spoken or Acted."18 The fact that these statements are made only on four title pages suggests that inclusion in print of passages cut in performance was unusual. If it were routine, why point it out these four times?Now, were cuts, plus additions, made to original authorial texts in almost all cases? Very probably. Henslowe's records show that after their playscripts were completed, the authors read the scripts to the acting companies, typically at tavern.19 As they listened, the actors undoubtedly were thinking about how the play would be performed. Some things they heard, no doubt, they wanted cut. At other points, they wanted material added. Correspondingly, Eric Rasmussen shows that play texts were revised extensively before they were performed, often by addition.20 I wrote a play once, a comedy, Meet Ben Jonson! The house had already accepted my submission, after requested revisions. I was also required to attend early rehearsals. Each time the director insisted that I cut passages that she or the actors considered too long and boring, and instead add more jokes. Each night I obeyed her instructions. The goal was to make a better play, not to reduce the play's running time, which had already been determined by the house. Still, it would have been possible, should anyone have been foolish enough, to publish the playscript with portions that were not actually performed.As I point out in SQ , surviving professional playhouse manuscripts do indeed reflect cuts, but those cuts are made for editorial reasons, not for the purpose of achieving any standardized running time. Falocco in his accompanying article points to a set of "prompt books" apparently prepared for performances sometime between 1625 and 1635 in Padua, Italy, by an amateur acting group led by Sir Edward Dering. The "prompt books" consist of markings on First Folio prints of several Shakespeare plays that are overwhelmingly deletions. But the practices of an amateur acting group performing in Italy around 1630 tell us little about what was done in England between 1594 and 1616 by professional actors. As I note in SQ , amateur groups also performed plays for six and even seven hours.21The script settled upon for a play's first performance run was not necessarily the same as the scripts used in subsequent revivals. It would be in the actors' interests, after all, to draw patrons by saying there were new things in the revived play. Henslowe often records additions to previously performed plays.22 Five title pages tell us that the accompanying print is "enlarged," "augmented," or "amplified," with "new additions." Several entries in Master of the Revels Sir Henry Herbert's records show that he reviewed new scenes added to "ould" plays.23It will be obvious, once again, that if additions were made for revivals of previously performed plays, there was no standard running time for play performances. And, as I point out in SQ , the additions could cause a play to run well past 3,000 lines. With its additions, the 1602 quarto of The Spanish Tragedy ran to 3,143 lines. Erne responds: "it has long been known" that original material was "erased" when the new lines were added.24 And how have we long known that? Because twentieth-century critical judgments say that the added material would not have been performed side-by-side with the original material, albeit "the 'joins' are quite neat."25 Twentieth-century critical judgments are important, but not with respect to seventeenth-century facts. We know nothing of the kind.Despite our critics' opinions, the fact is clear. The 1602 quarto title page, the title page upon which Erne elsewhere places so much stock, plainly states that the script as there printed represents the play "as it hath of late been divers times acted." The Spanish Tragedy was a staple of the acting repertoire. Would the actors really have cut passages their audiences expected to see and hear, in deference to the opinions of critics three centuries later?The printed texts of Tempest and Cymbeline, Falocco argues in his accompanying article, yield inconsistent results as to running time, if both plays began after two and ended near six. Dutton also argues about supposed disjunctions in Tempest's schedule.26 There's an obvious answer to these objections. Falocco and Dutton fail to account for the intermediary function of incidental entertainment. Both Tempest and Cymbeline, if scholars are correct, were written to be performed originally at the Blackfriars, where interact entertainment was necessary while the candles were cut and relit. Such entertainment needn't have taken longer than the time for the candles. But it could have taken longer. Thus, the texts of Tempest and Cymbeline tell us only how much time was devoted to performing the playscripts themselves. They do not tell us how much time elapsed during performance of both the playscripts and the interact entertainment, from the beginning of the plays to their ends.We can, however, deduce that total elapsed time in Tempest, because Tempest itself lays out the schedule for the play's performance. The time devoted to entertainment between acts of Tempest apparently was substantial. The play begins, as I explain in SQ , sometime after 2 p.m., "at least two glasses." And it will end by 6, as it will occupy "the time twixt six and now." It does indeed end near six, "on the sixth hour," six being the time "our work should cease." Falocco cites Andrew Gurr to say that "the sixth hour" means "5 pm." That's just not correct. "The sixth hour" in English means the time between 5 and 6 p.m. Dutton points out that in the fifth act the boatswain reports the ship had split "but three glasses since." But Dutton miscounts. The ship splits after the play begins, after 2. As three glasses have since passed, that means, once again, that the play is now in the hour between 5 and 6 p.m.27Falocco suggests that Shakespeare could have written his longer plays with the idea that different versions of them could have been performed on different occasions. He cites several conjectures concerning such possibilities, but there is no evidence for any of them. And I agree with Urkowitz, in his accompanying article. The memories of Elizabethan actors were by our standards truly remarkable. But is it realistic to think that in addition to keeping multiple plays in memory at one time, they also kept multiple versions of individual plays? Wouldn't that feat have been even more difficult than the first? Mnemonics work when the routine is consistent.Falocco taxes me for excluding the "Bad Quartos" from my analysis. But we don't actually know what these Bad Quartos represent, or even whether any single explanation accounts for them all. I'm inclined to agree in part with the traditional theory that they are pirated versions, although I think that they were pirated for illicit performance, not print. If that is so, they have little to tell us about professional company practices.Macbeth, as printed in the Folio and only there, is not a Bad Quarto. But it was cut, in my estimation severely so. Where, for example, are the comic scenes? No one knows why these cuts were made. Let me hazard a guess. When the play was first performed, James was king of England, as well as king of Scotland. Large passages were cut because they may have offended the new king from Scotland, who was also the acting company's new patron. Cuts for such an unusual reason would not have represented the norm. A more normal case would be the Folio Hamlet. It very probably represents an acting version of the play. Nevertheless, it cuts only 230 lines from the Q2 version of the play, probably based on an authorial manuscript, and adds 70.I considered lastly in SQ the residual question. If the performance events regularly lasted almost four hours and the plays themselves occupied varying amounts of time, what accounted for the difference? I argued that the remaining time was taken up by incidental entertainment, pre-show performances, interact music, and post-show jigs or dances. Preliminary entertainment lasted at least a half hour. Other segments could occupy varying amounts of time, from no time at all, if the segment was skipped, to as long as an hour. As far as I'm aware, no one has disagreed with the argument I made. Richard Preiss agrees that incidental entertainment occupied the residual time, but suggests that the clowns and the audiences negotiated control of the time devoted to clowning, and thus that, if necessary, plays were cut on the fly.28 Preiss's most convincing evidence comes from the 1580s, Richard Tarlton and the Queen's Company. We may reasonably doubt whether those practices were continued by the major companies after 1594.As I noted at the outset, my concern is about what happened in the Elizabethan theater itself. I don't think that my research provides an answer for contemporary Original Practices companies, a principal concern of Falocco and Urkowitz. I agree that modern audiences may have difficulty following a play delivered at twenty lines per minute. But that is not the only respect in which OP companies cannot replicate original practices. If they were to use Elizabethan pronunciation, the language would be more difficult to understand than current Glaswegian. If they were to have hucksters wandering through the audiences selling hazelnuts and ale, there would be chaos. If they were to have prostitutes plying the galleries, the authorities would shut them down. "OP," I think, just has to be a euphemism for coming as close as we can.