Is it really possible that, after so many centuries of tinkering and bickering, the editors still have not produced a text of King Lear as Shakespeare would have wanted it? That the world has persuaded itself that this play is the master’s masterpiece—without noticing that the editors have jumbled together words, lines, and episodes (from Q and F, the quarto text of 1608 and the folio text of 1623) that were actually meant to replace each other, and that sometimes make little sense as now printed? If so, we must salute the text of King Lear as the most successful hoax in the long history of Shakespeare fabrications.
Several books and articles have recently argued that Q and F represent different versions of the play which, like the 1805 and 1850 manuscripts of Words-worth’s The Prelude, should be treated as separate entities and not conflated, as they have been for almost three hundred years, in a single text. Any modern edition of Shakespeare one may consult—the Arden, the Penguin, the Riverside—presents a composite text, usually one that follows F but accepts many readings from Q. Challenging an editorial tradition that has survived for so long, the authors must expect some resistance, and Messrs. Taylor and Warren should perhaps have helped the reader of the most ambitious criticism so far of the “single version” theory by placing it in a wider context. Stanley Wells’s introduction, “The Once and Future King Lear,” was a splendid conference paper given in Stratford on Avon in 1982; it made everyone sit up, read out as it was with conviction and in ringing tones.
On that occasion there was no time to look far beyond King Lear; what one now wants to know is whether the editors of King Lear have been so incompetent (if indeed they have been) because a “single version” theory was fashionable in Shakespeare editing generally—or whether “the two versions of Lear” is a theory that has itself become fashionable only because the discussion has been largely restricted to a single play. “Stanley has really swallowed it,” I heard one delegate mutter, as the applause died down; new though it seems, the “two versions” theory has already had a bandwagon effect, and that will stiffen the resistance.
Yet the idea that Shakespeare revised his plays has been around for a long time. It was once held that the first printed texts of Romeo, Henry V, Merry Wives, and Hamlet were “first drafts,” later improved; according to this view Shakespeare first wrote:
To be or not to be, ay, there’s the point.
To die, to sleep—is that all? Ay, all.
No: to sleep, to dream. Ay, marry, there it goes…
and then thought he could do better. In 1909, however, A.W. Pollard showed that these early texts were not first drafts but unauthorized reconstructions from memory—or “bad quartos”—of the plays as we know them. The first quartos of Richard III and Lear resembled these bad quartos in some respects, though not in all, and remained a puzzle. From the 1930s to the 1960s there appeared many ingenious studies of the text of Lear, one of which, by Madeleine Doran, anticipated Taylor and Warren (in 1931) by arguing at length that Q (the first quarto of 1608) looks like a first draft rather than a bad quarto. But at this very time more and more bad quartos were identified (Leo Kirschbaum published a “Census of Bad Quartos” in 1938), and the tide was running the other way.
Miss Doran’s opponents included E.K. Chambers, W.W. Greg, A.W. Pollard, and Quincy Adams (Oxford, Cambridge, the British Museum, and the Folger Shakespeare Library)—an unequal contest, even if Miss Doran, then a recent graduate, was later to become one of the most admired Renaissance scholars in the United States. She partly retracted her theory in 1941, though she still held that Q “must represent an earlier form of the play than the Folio.” Others shared her belief, but for almost fifty years the weight of opinion tipped the balance decisively the other way.
If Miss Doran was in the right, as is now suggested, it is worth asking why her views did not prevail. The answer is that by the 1920s revision theories had got out of hand. F.G. Fleay, J.M. Robertson, and J. Dover Wilson in the New (Cambridge) Shakespeare had had so much fun detecting “revision” in every play of Shakespeare that E.K. Chambers who, so to speak, had taken charge of the Elizabethan stage in four encyclopedic volumes1 decided that enough was enough. Chambers slapped down the “disintegrators” (i.e., those who contended that Shakespeare or others revised his plays) in an acidulous British Academy lecture (1924) that has become a classic. The disintegrators, he observed, “approach the point where scholarship merges itself in romance”—a remark that may well be echoed by some reviewers of The Division of the Kingdoms. Miss Doran therefore arrived at the wrong time, just when the good ship Shakespeare had swung around and sailed off majestically in another direction.
Apart from the fact that “disintegrators” were out of fashion, the theory that Lear was revised had to contend with another kind of semi-invisible resistance. From the 1920s to the 1960s so much exciting pioneer work was concerned with textual corruption that an assumption silently took root that all textual variants are best explained as corruptions. (Many illnesses are caused by viruses, so perhaps all illnesses are caused by viruses.) From the 1920s to the 1940s major advances were made in the understanding of bad quartos, led by Peter Alexander; from the 1950s the emphasis switched to studies of the distinctive habits of the various compositors in the print shop, thanks to brilliant work by Charlton Hinman and Alice Walker—and, quite naturally, the editors of King Lear busied themselves with bad quarto theories and then threw themselves with equal enthusiasm into compositor analysis (viruses again?).
By the 1950s Sir Walter Greg had become the central figure in “Elizabethan” bibliography and textual criticism, and Greg did not lose sight of the possibility that Shakespeare could have revised some of his plays. But, as the present writer observed in The Stability of Shakespeare’s Text (1965), Greg never fully grappled with the awkward problem that once you concede that there has been some revision in a two-text play, you are no longer entitled to assume that all variants, apart from self-evident examples of revision, must be corruptions. Several of the contributors to The Division of the Kingdoms have returned to this problem, and propose solutions that differ radically from Greg’s.
For nearly sixty years those who believed that Shakespeare revised some of his plays have therefore had to struggle against elusive historical forces, a resistance largely conditioned by temporary scholarly fashions. Whether the arguments for revision were good or bad (and some were bad), a senior figure always appeared with a big stick and clubbed down the “disintegrators.”
A very recent example is particularly instructive: Harold Jenkins, no friend to revision theories, found himself driven by his own argument (and against his will, it seems) to conclude that Shakespeare interpolated forty lines in act 2, scene 2, of Hamlet. Fredson Bowers of the University of Virginia, in a long and careful review, remarked, “It is one of Professor Jenkins’s most particular virtues that he is uncompromising in his opposition to the view that Shakespeare tinkered with his texts in search of literary as against theatrical improvement” (The Library, September 1983). Jenkins at the same time complained that “there has been too much irresponsible conjecture about Shakespeare’s supposed revisions” (New Arden Hamlet, 1982, p. 5); and Bowers once wrote of the “astonishing amount of variation” in plays such as Hamlet and Othello that “some of these differences must no doubt be imputed to Shakespeare’s revising his own work” (On Editing Shakespeare, 1966, p. 134). We all wield the big stick—and believe in it, when applied to others. Even those who plead for a particular instance of revision, or who have accepted the possibility in the past, still turn in horror from the very notion of a self-revising Shakespeare.
For nearly sixty years the “disintegrators” had a bad press, and now, quite suddenly, things have changed. George Walton Williams, writing in Shakespeare Survey (Cambridge), agrees that Shakespeare revised King Lear, and that the familiar composite text confuses the plot line, “blurs the delicately indicated expectations,” misrepresents both the first and second versions, and, quite simply, “is not Lear“; and the Clarendon Press (Oxford) has now produced the expensive volume under review, which grew out of the same conviction. How did it happen? The credit belongs, in the first instance, to Michael Warren—who startled the World Shakespeare Congress in Washington, DC, with a short paper, “Quarto and Folio King Lear and the Interpretation of Albany and Edgar,” in 1976. Warren examined the additions, omissions, and variants in the speeches assigned to Albany and Edgar and concluded that the differences between Q and F cannot be ascribed to chance, but were dictated by different conceptions of the two men when the Q and F texts came into being; in short, Warren identified a strategy of revision. Supporting books and articles soon followed on both sides of the Atlantic. One, by Gary Taylor (1980), neatly argued that “the apparent motive for many of the major F changes from the Q text is to strengthen the structure of act IV.”
F has done this by cutting superfluities…and by strengthening the narrative line, largely by accelerating and clarifying the movement toward war.2
Taylor, in other words, identified a second strategy of revision.
Taylor and Warren are the editors of the collection of essays under review, several of which report even more strategies of revision. At first such a multiplication of strategies may seem surprising—but let us leave that until later, and first find out what the reviser is supposed to have done. Michael Warren, in “The Diminution of Kent,” explains that “the part of Kent undergoes such substantial change between Quarto and Folio in the last two and a half acts” that F reduces him to “a marginal figure until his entrance just prior to Lear’s final scene.” The argument hitches on to Warren’s earlier paper, showing that F’s “diminution” of Kent is connected with the same text’s magnification of Edgar: “during the third and fourth acts Kent and Edgar perform similar functions,” so Kent’s loss is Edgar’s gain.
Warren also notes that F reduces “the number of passages of static speech or dialogue about moral issues in the play,” and consequently cuts some of Kent’s other speeches (one of Kent’s functions in Q was to serve as a “choric” commentator, and that is how readers of the modern synthetic or conflated text remember him). But Kent’s “diminution” already began in Q, where all of his “major plot functions are taken from him” during and after act 3, scene 6, and “a previously dynamic figure, Kent-as-Caius becomes static.” The F reviser therefore continued a rethinking of Kent’s role that probably started during the composition of the Q version. In one short paper, it will be observed, Warren asks us to accept a number of distinct but converging “strategies of revision.”
Warren’s argument has important implications for the literary critic, as also has John Kerrigan’s analysis of the changes in the Fool’s part.
Interpretations of the Fool based on the received text have swung wildly between the extremes represented by Empson’s blathering natural and Orwell’s canny rationalist. In neither version of the play is the character uncomplicated; but, broadly speaking, the Empsonian view of him rests on those elements in the conflated text which derive from Q, while the Orwellian interpretation depends on ingredients drawn from F. Between Q and F comes a substantial rewrite—involving some 54 lines of a part only about 225 lines long when conflated—which significantly alters the Fool’s personality.
Students of Shakespeare’s imagery will be equally interested in Thomas Clayton’s “Revision in the Role of the King.” Clayton examines F alterations that “are part of an evident, emphatic, and functional pattern that combines the progressive peeling away of layers of attire, appearance, affectation, and at last anatomy itself,” and comments on F additions such as “Since now we will divest us both of rule…” (1.1.49-50), “Come, unbutton here” (3.4.108), “Plate sin with gold, / And the strong lance of justice hurtless breaks; / Arm it in rags, a pigmy’s straw does pierce it” (4.6.165ff).
On a different tack, Roger Warren (not to be confused with his namesake) explains F’s two biggest omissions from act 3, scene 6, the mock trial and Edgar’s closing soliloquy. Both are cut because they are
given more effective (and, in the Folio, expanded) treatment later in the play…. Just as the omission of the mock trial abridges the repetition of effects from the previous scenes, it also avoids anticipating important elements of 4.6.
Here then—to pause for a moment—we have an extraordinary number of revision strategies, replacing the older theory that Q is simply a bad quarto. Too many, perhaps? Shakespeare, we are told in Taylor and Warren’s book, changed his conception of several characters, tightened his plot (cutting out duplications, choric speeches, and other “superfluities”), and reinforced the play’s central ideas and images. Can this be the marvelously fluent dramatist of whom his colleagues said that “we have scarce received from him a blot in his papers”? Initially one reacts with skepticism. After all, the hypothesized “first version” must have been, already, a masterpiece—was it really necessary to revise it so drastically? On reflection, though, one has to admit that Shakespeare, as a dramatist, always did much more than was strictly necessary, and that for a writer so fluent the postulated revisions were not drastic at all. If he ever did set out to improve a play, the myriad-minded awareness suggested by these essayists, each of whom deals with only a tiny part of the whole, is exactly what we should have expected.
When did he do it? Gary Taylor thinks in 1609–1610, perhaps for the opening of the Blackfriars theater. He argues that the rewriting began on a copy of the printed Q, i.e., in 1608 or later (Shakespeare would not take the prompt-book with him if he revised at Stratford); and Taylor produces stylistic and vocabulary tests (not easy to follow) that would date F’s additions with the “last plays,” i.e., after 1608.
A group of essays examines the editorial tradition, and asks why things went so badly wrong. Steven Urkowitz traces how, in the eighteenth century, editors came to take for granted “the undemonstrated assertion” that Q and F go back to a single arch-text. Paul Werstine challenges “the long-standing assumption, inherited from eighteenth-century editors, that printing-house agents were responsible for massive corruption”; he defends F’s editors and compositors, and concludes that “only a handful of the variants between the Quarto and Folio versions of Lear can be attributed to the two compositors.” (I find this hard to believe. Is there only a handful of manifest errors in F?) Because editorial commentary “typically contrasts single words from Q and F,” writes Randall McLeod, in another rousing indictment, it usually presupposes “a choice of one word and rejection of the other.” McLeod shows that when we are not brainwashed by editorial guidance, both words frequently make good sense. MacD.P. Jackson analyzes the distribution of variants in the two texts, and finds that it “can…only be plausibly explained as the result of authorial revision.” The “most densely variant act” is the first, and here some of the variants are “interrelated and…produced systematic differences.”
It is not, of course, a mere matter of chance that all the contributors speak with one voice from the ends of opposed winds (from the United States, Canada, New Zealand, and England, to be precise). Several of them discussed the text of Lear at the annual meeting of the Shakespeare Association of America in 1980, and then teamed up: this volume is part of a campaign—some will say a conspiracy—that has now continued for some years. I happen to think that they have a very strong case, but not everyone agrees. There will be opposition from many places. There will be thunder from Charlottesville, Virginia.
Yet when the dust has settled, this volume will be seen to mark an important turning point for Shakespeare’s editors and critics. For more than thirty years the editors have been overconfident that the new practices of bibliography could detect textual corruption, and have pursued a policy of “corrective editing” where scholarship too readily merges itself in romance; as Francis Bacon said, “rash diligence hath done great prejudice” to the “true correction” of authors—for, “as it hath been wisely noted, the most corrected copies are commonly the least correct.” King Lear is not the only play with a “conflated text” that will have to be unscrambled by future generations. (This will be less straightforward than some of the essayists assume, and will take time.) As for the critics, if they take the trouble to follow Shakespeare’s revising pen, they will be surprised to learn how very consciously he must have pondered the minutest implications of what he wrote. This, also, has been said before, but was never so fully demonstrated. We owe an immense debt to Taylor and Warren, for their own subtle essays and for masterminding this important campaign.
February 2, 1984