The discrepancy between the six new plays of 1623–1624 and the estimated seventeen of 1594–1603 is not a mark of conflict in the evidence. Times had changed. The King’s men needed only a third of the new plays that they had produced in earlier years. The use of a private theater largely accounts for this change, for the seats of Blackfriars could be filled four or five times over by the audience from a single performance at the Globe. What is really significant is that the King’s men presented the same number of new plays as the Prince’s men, and that the practices of Shakespeare’s fellows were in harmony with those of other companies.
Only an idolatrous love of Shakespeare can lead us to conclude that from 1599 to 1609 the Lord Chamberlain’s men produced appreciably fewer plays than the other companies did. All were in lively competition, in which, as Platter noted, “those which play best obtain most spectators.” To maintain that the Globe company produced only four or five new plays a year, we must prove that Shakespeare’s plays were of such popularity that they could be repeated again and again while other companies had to change their bills daily. However, we have no evidence to show that this was the case. Certainly, Falstaff was a perennial favorite, but so was Barabas the Jew. A play such as Richard II was old by 1601. Twelfth Night, or Malvolio, held the stage, it seems, but so did The Spanish Tragedy, or Jeronimo. Yet Henslowe’s schedule reveals that the old war horses such as Jeronimo, The Jew of Malta, Faustus, and Tamberlaine, altogether, provided no more than 11 per cent of the performances of the Lord Admiral’s company throughout the entire recorded period and no more than 6 per cent in any one year (see Appendix A, chart ii). We should like to think that Shakespeare’s work had more commercial appeal than Marlowe’s or Kyd’s. But can we suppose that it had a popularity, let us say, five or six times greater? A sobering thought on the enigma of popularity must strike us when we realize that Pericles was, if its succession of quartos offers any evidence, more popular than Antony and Cleopatra, and that The Winter’s Tale, if Court performances are any measure, appealed to royalty more than King Lear. Furthermore, once we eliminate the plays which in all likelihood were given few performances, such as Troilus and Cressida, All’s Well and Measure for Measure, we are left with too few Shakespearean plays to sustain a theatrical company in the London of 1600. A reference to the list of Court performances between 1603 and 1642 verifies the pattern reflected in Henslowe’s records. Aside from their first appearances before James, when they presented many old stand-bys, the King’s men usually offered the latest plays to Their Majesties, and when Shakespeare died, the works of other writers rapidly superseded his at Court.[13] Like the commonalty, royalty expected to see the current “hit.”
The plays we now regard as great literary works were struck off in the harassing atmosphere of a commercial enterprise. Most of the plays were failures or temporary successes. Most of those produced by the Admiral’s men played their few, in many cases very few, performances and passed away without any further trace but the notation by a shrewd businessman. Of the one hundred and thirteen plays listed by Henslowe between 1592 and 1597, sixty-seven would certainly be unknown without the Diary and another twelve would probably be unknown (see Appendix A, chart i). However, among the thirty-four plays that would be otherwise known, only twenty-seven are extant, or about 24 per cent of the plays listed by Henslowe. By assuming that the twenty-nine extant Globe texts represent a similar percentage of the Globe repertory, we arrive at a conclusion that 116 plays were actually produced by the company between 1599 and 1609. But during these years the theater suffered closings of extraordinary duration because of the plague.[14] In addition, the Globe period is calculated from September, 1599, to March, 1609. Actual playing time, therefore, amounted to about seven and a half years. This estimate divided into the 116 new plays gives us a result of 15.6 plays as the average number of new works offered by the Globe company each year. Actually, in estimating these figures, some allowance must be made for Shakespeare’s superiority. How much, however, is virtually impossible to say. Nor is an actual figure necessary as long as we realize that the repertory systems of all three companies were fundamentally the same. In effect, the figures that we have for the Lord Admiral’s and Worcester’s men are a far safer guide to actual Globe practice than any other evidence.
As lovers of literature, we need be grieved little by the disappearance of 75 per cent of the plays, at least judging from contemporary response. Generally the plays that have come down to us were the more popular pieces. Either they were printed, or discussed, or alluded to. At the same time they were played more frequently. The seventy-nine plays which we know only through Henslowe provided 496 performances in five years. The other thirty-four played 403 performances in the same period. On an average we find the plays otherwise known to us played nearly twice as many performances as those mentioned by Henslowe.
Those pieces that attained popularity and whose stage life extended over a period of years run like strong threads through the repertory of an Elizabethan company. But between the strands there was much filler, plays which spoke their brief piece upon the platform and departed within a few months. Seven to eight performances were the average number for a play. Many did not attain even this many representations. Three out of every ten plays had no more than one or two performances. Less than one out of ten went beyond twenty performances. An extensive and actually wonderful process of winnowing out the chaff was at work. This process was the repertory system. As a result of it, the plays that could bring back an audience year after year survived to speak for the age (see Appendix A, chart ii).
The process of winnowing out the ineffectual pieces was supplemented by the custom of revivals. Periodically, plays of the recent past would be brought back to the stage for another run. Usually the pattern of performances for a revival would follow that of a new play: close-packed performances at first and a tapering off until representation ceased. The Spanish Tragedy, or as Henslowe entitles it, Jeronimo, offers a clear example of the process at work. In March, 1592, it was presented for three performances, in April, again for three performances, in May it reached its peak with five performances, and in June played twice. The hiatus in the summer and fall of 1592 interrupted the normal cycle. On resumption of playing in December, Jeronimo appeared again, was repeated twice in January for the last times. These performances were by Strange’s men. Four years later, on January 7, 1597, the Lord Admiral’s men revived it as a “new” play, indicating that it had been substantially revised. Subsequent performances followed with diminishing frequency with intervals of 4, 6, 5, 9, 10, 28,[15] 14, 21, 26, 29 days. The play was further revived in September or October of 1601, this time with additions by Jonson.
Twenty plays in Henslowe’s list show definite evidence of revival, either during the 1593–1597 period or the 1599–1601 period. Only Doctor Faustus shows continual performance from 1594 to 1597. Originally revived on October 2, 1594, it was performed from time to time by the Lord Admiral’s company which did not allow more than four months to elapse between performances. There was a later revival toward the end of 1602.
Among the nineteen remaining plays the manner of revival varied somewhat. Nine of them seem to have been altered or enlarged considerably for the revival. Usually these plays had been off the stage for several years. Fortunatus was reworked by Dekker in November, 1599, after it had lain idle for three and a half years. Jeronimo, as we saw already, had not been offered for four years when it was revived as “new” in January, 1597. Tambercame, Part II, was three and a half years old when presented as “new” on June 11, 1596. Two of the plays, Friar Bacon and Friar Bungay and Phaethon, show evidence of alteration as do the rest, but specifically for Court performances. Though there is no certainty that revivals in the public playhouse occurred at the same time, it is not unlikely, as we shall discover.
One advantage of the Elizabethan method of revivals—abetted by the absence of copyright laws—was that it enabled a writer to rework his own or someone else’s work. Through how many versions, for example, did the narrative of Hamlet pass to reach its final stage? We know of three at least: the one played by the “Lord Admeralle men & Lorde Chamberlen” at Newington Butts on June 11, 1594; the one contained in the 1603 Quarto; the one announced as “newly imprinted and enlarged to almost as much againe as it was, according to the true and perfect Coppie.” The constant sifting of the repertory not only screened out hack pieces, it also provided time for the refinement of masterworks.
In instances where no proof of literary revision exists, there is evidence sometimes of theatrical revision. Four plays from four to six and a half years old were revived after 1597. The purchase of properties for them indicates that they received new productions. Of the last six of the twenty plays revived, only the cessation of playing and, after an extended lapse of time, the resumption of performances tell us that they were revived.