Publication was not a decision to be taken lightly. Folios were big books and expensive to produce, so the First Folio was very ambitiously priced at £1 (for an edition bound in calfskin; unbound copies were a little cheaper). A copy of the sonnets, by comparison, cost just 5 pence on publication-or one forty-eighth the price of a folio. Even so the First Folio did well and was followed by second, third, and fourth editions in 1632, 1663-1664, and 1685.

The idea of the First Folio was not just to publish plays that had not before been seen in print but to correct and restore those that had appeared in corrupt or careless versions. Heminges and Condell had the great advantage that they had worked with Shakespeare throughout his career and could hardly have been more intimately acquainted with his work. To aid recollection they had much valuable material to work with-promptbooks, foul papers (as rough drafts or original copies were known) in Shakespeare’s own hand, and the company’s own fair copies-all now lost.

Before the First Folio all that existed of Shakespeare’s plays were cheap quarto editions of exceedingly variable quality-twelve of them traditionally deemed to be “good” and nine deemed “bad.” Good quartos are clearly based on at least reasonably faithful copies of plays; bad ones are generally presumed to be “memorial reconstructions”-that is, versions set down from memory (often very bad memory, it seems) by fellow actors or scribes employed to attend a play and create as good a transcription as they could manage. Bad quartos could be jarring indeed. Here is a sample of Hamlet’s soliloquy as rendered by a bad quarto:

To be, or not to be, I there’s the point,

To Die, to sleepe, is that all? I all.:

No, to sleepe, to dreame, I mary there it goes,

For in that dreame of death, when wee awake,

And borne before an everlasting Judge,

From whence no passenger ever returned…

Heminges and Condell proudly consigned to the scrap heap all these bad versions-the “diverse stolne, and surreptitious copies, maimed, and deformed by the frauds and stealthes of injurious impostors,” as they put it in their introduction to the volume-and diligently restored Shakespeare’s plays to their “True Originall” condition. The plays were now, in their curious phrase, “cur’d, and perfect of their limbes”-or so they boasted. In fact, however, the First Folio was a decidedly erratic piece of work.

Even to an inexpert eye its typographical curiosities are striking. Stray words appear in odd places-a large and eminently superfluous “THE” stands near the bottom of page 38, for instance-page numbering is wildly inconsistent, and there are many notable misprints. In one section, pages 81 and 82 appear twice, but pages 77-78, 101-108, and 157-256 don’t appear at all. In Much Ado About Nothing the lines of Dogberry and Verges abruptly cease being prefixed by the characters’ names and instead become prefixed by “Will” and “Richard,” the names of the actors who took the parts in the original production-an understandable lapse at the time of performance but hardly an indication of tight editorial control when the play was reprinted years later.

The plays are sometimes divided into acts and scenes but sometimes not; in Hamlet the practice of scene division is abandoned halfway through. Character lists are sometimes at the front of plays, sometimes at the back, and sometimes missing altogether. Stage directions are sometimes comprehensive and at other times almost entirely absent. A crucial line of dialogue in King Lear is preceded by the abbreviated character name “Cor.,” but it is impossible to know whether “Cor.” refers to Cornwall or Cordelia. Either one works, but each gives a different shading to the play. The issue has troubled directors ever since.

But these are, it must be said, the most trifling of bleats when we consider where we would otherwise be. “Without the Folio,” Anthony James West has written, “Shakespeare’s history plays would have lacked their beginning and their end, his only Roman play would have been Titus Andronicus, and there would have been three, not four, ‘great tragedies.’ Shorn of these eighteen plays, Shakespeare would not have been the pre-eminent dramatist that he is now.”

Heminges and Condell are unquestionably the greatest literary heroes of all time. It really does bear repeating: only about 230 plays survive from the period of Shakespeare’s life, of which the First Folio represents some 15 percent, so Heminges and Condell saved for the world not only half the plays of William Shakespeare, but an appreciable portion of all Elizabethan and Jacobean drama.

The plays are categorized as Comedies, Histories, and Tragedies. The Tempest, one of Shakespeare’s last works, is presented first, probably because of its relative newness. Timon of Athens is an unfinished draft (or a finished play that suffers from “extraordinary incoherencies,” in the words of Stanley Wells). Pericles doesn’t appear at all-and wouldn’t be included in a folio edition for another forty years, possibly because it was a collaboration. For the same reason, probably, Heminges and Condell excluded The Two Noble Kinsmen and The True History of Cardenio; this is more than a little unfortunate because the latter is now lost.

They nearly left out Troilus and Cressida, but then at the last minute stuck it in. No one knows what exactly provoked the dithering. They unsentimentally tidied up the titles of the history plays, burdening them with dully descriptive labels that robbed them of their romance. In Shakespeare’s day there was no Henry VI, Part 2, but rather The First Part of the Contention Betwixt the Two Famous Houses of York and Lancaster, while Henry VI, Part 3 was The True Tragedy of Richard Duke of York and the Good King Henry the Sixth-“more interesting, more informative, more grandiloquent,” in the words of Gary Taylor.

Despite the various quirks and inconsistencies, and to their eternal credit, Heminges and Condell really did take the trouble, at least much of the time, to produce the most complete and accurate versions they could. Richard II, for instance, was printed mostly from a reliable quarto, but with an additional 151 good lines carefully salvaged from other, poorer quarto editions and a promptbook, and much the same kind of care was taken with others in the volume.

“On some texts they went to huge trouble,” says Stanley Wells. “Troilus and Cressida averages eighteen changes per page-an enormous number. On other texts they were much less discriminating.”

Why they were so inconsistent-fastidious here, casual there-is yet another question no one can answer. Why Shakespeare didn’t have the plays published in his lifetime is a question not easily answered either. It is often pointed out that in his time a playwright’s work belonged to the company, not to the playwright, and therefore was not the latter’s to exploit. That is indubitably so, but Shakespeare’s close relationships with his fellows surely would have ensured that his wishes would be met had he desired to leave a faithful record of his work, particularly when so much of it existed only in spurious editions. Yet nothing we possess indicates that Shakespeare took any particular interest in his work once it was performed.

This is puzzling because there is reason to believe (or at least to suspect) that some of his plays may have been written to be read as well as performed. Four in particular-Hamlet, Troilus and Cressida, Richard III, and Coriolanus- were unnaturally long at 3,200 lines or more, and were probably seldom if ever performed at those lengths. The suspicion is that the extra text was left as a kind of bonus for those with greater leisure to take it in at home. Shakespeare’s contemporary John Webster, in a preface to his The Duchess of Malfi, noted that he had left in much original, unperformed material for the benefit of his reading public. Perhaps Shakespeare was doing likewise.