Polonius thought that the Player’s speech was too long. Dr. Johnson said that nobody would wish Paradise Lost any longer. One might say the same of Meistersinger or Götterdämmerung. On the other hand, Robert Schumann defended Schubert’s ‘heavenly lengths’ against detractors. Size is a pretty primitive criterion, but it enters powerfully into our tacit or explicit judgments.
Can one say anything useful about the length of plays in the EMD corpus? If you exclude the shorter genres of interludes and masques, there remain 300 plays that are classified as tragedy, comedy, tragicomedy, or history. Leaving aside genre distinctions and charting the number of plays per decade against the average length of plays you can see that the length of plays increases into the 1590’s and then moves up and down within a fairly narrow range.The interquartile figures for the entire period range from 16,443 for the top of the lowest quartile to 21,034 for the bottom of the highest quartile with a median of 18,879. You won’t go wrong much if you remember that the typical Early Modern play has a length of 19,000 +/3,000 words.
The truth value of the chart depends of course on the representativeness of the corpus. Consider the various ways in which the 300 plays in it might not be representative:
- I picked most but not all the plays from the TCP collection.
- The TCP collection is a (growing) subset of the digitized page images of microfilmed print books and reflects the various priorities of the people who picked this rather than that text for transcription.
- The microfilmed texts include many but not all of the printed books that have come down to us.
- What has come down to us reflects choices and accidents in the history of textual transmission. It is not necessarily a representative sample of what was performed, published, or read at the time.
On the other hand, 300 plays are 300 plays, and the various biases would have to reinforce each other at every step for a sample of that size to be seriously misleading. Thus one may with some caution take the current EMD corpus as representative and conclude from the chart that
- the production of plays rises sharply in the (late) eighties
- stays at a high plateau between 1590 and 1620
- declines in the 1620’s and 30’s and precipitously during the Civil War period
There is nothing striking new about these conclusions, but for a newcomer the chart tells a simple story very clearly. Note, however, that the figures for length before 1580 and after 1640 may be distorted by low sample size.
Let us now turn to the box plot about differences in length by genre of the 146 comedies, 89 tragedies, 41 tragicomedies and 23 histories. The box plot lets you see more immediately the differences between the descriptive statistics outlined in the table below.
Length of plays by genre
From these figures you can infer confidently that playwrights and their audiences had firm expectations about the dimensions of what the prologue to Romeo and Juliet calls “the two hour traffic of our stage”. Tragedies, comedies, and tragicomedies do not differ much in length. It is a little surprising that comedies are longer than tragedies. But the list of tragedies includes translations of ancient plays, which typically are between half and two-thirds the length of modern plays.
Why are history plays on average longer by at least 10%? The history play is the smallest of the dramatic genres, and it is dominated by Shakespeare, the author of ten of the two dozen works in the EMD corpus. Shakespeare’s history plays are longer than those of his contemporaries, as is shown by these summary statistics:
Length of history plays
But Shakespeare’s plays are generally longer than those of his contemporaries. This is clear from a box plot that compares the length of 155 plays written between 1589 and 1612 by Shakespeare and others. The implications of that fact emerge from another look at the earlier boxplot of length by genre. What plays are represented by the asterisks that mark high or low outliers? There are six high outliers for comedy:
- Killigrew, The parson’s wedding (37,946)
- Jonson, Every man out of his humour (37,149)
- Killigrew, Thomaso, part II (36,852)
- Jonson, Bartholomew Fair(35,081)
- Chapman, Two wise men and all the rest fools (33,880)
- Killigrew, Thomas, part I (32,643)
Killigrew was largely a writer of closet dramas. The two tragic outliers are Hamlet and Gascoigne’s Glass of Government (a ‘tragicall comedie’ that is hard to classify. Shakespeare wrote 12% of the plays in the EMD corpus, but a third of his tragedies are in the top quartile by length, and only two (Timon and Macbeth) are below the median.
Cymbeline is the outlier among the tragicomedies, and the four longest plays in that genre include Killigrew’s Princess and Bellamira Her Dream, pt 2, as well as The Winter’s Tale.
It appears from this survey that length has some value as an indicator of literary ambition. Remember in this context the eyebrows that were raised when Jonson published his plays as ‘works’ — a move that elevated the plays from the humble status of scripts to a higher plateau of textuality:
To Mr Ben Johnson demanding the reason
why he called his plays works.
Pray tell me Ben, where does the mystery lurk,
What others call a play you call a work.
Thus answered by a friend in Mr Johnsons defense.
The authors friend thus for the author says,
Ben’s plays are works, when others works are plays.
(Cited from King’s College Information Services)
A play that runs to more than 20,000 words needs to be cut in performance. Why were those extra words written in the first place? Did Shakespeare find himself in the position of Kant, who apologized to a correspondent for writing a long letter because he did not have the time to write a short one? Or was Shakespeare like Jonson in thinking of his plays as ‘works’?