The Materiality of the Shakespearean Text Author(S): Margreta De Grazia and Peter Stallybrass Source: Shakespeare Quarterly, Vol

Home , Nicholas Okes, Second Folio, Shakespeare Quarterly

The Materiality of the Shakespearean Text Author(s): Margreta de Grazia and Peter Stallybrass Source: Shakespeare Quarterly, Vol. 44, No. 3 (Autumn, 1993), pp. 255-283 Published by: Oxford University Press Stable URL: https://www.jstor.org/stable/2871419 Accessed: 29-06-2019 21:12 UTC

REFERENCES Linked references are available on JSTOR for this article: https://www.jstor.org/stable/2871419?seq=1&cid=pdf-reference#references_tab_contents You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms

Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to Shakespeare Quarterly

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms The Materiality of the Shakespearean Text

MARGRETA DE GRAZIA AND PETER STALLYBRASS

FOR OVER TWO HUNDRED YEARS, KING LEAR WAS one text; in 1986, with the Oxford Shakespeare, it became two; in 1989, with The Complete King Lear 1608-1623, it became four (at least). As a result of this multiplication, Shakespeare studies will never be the same. This is not simply because we may now have more Shakespeare than before-many Lears instead of one; a mere enlargement of the canon requires no rethinking of how the works are to be prepared and interpreted. Shakespeare studies will never be the same because something long taken for granted has been cast into doubt: the self-identity of the work. We are no longer agreed on the fundamental status of the textual object before us. Is it one or more? The significance of this uncertainty cannot be overestimated. Identity and difference are, after all, the basis of perception itself, the way we tell one thing from another. The possibility of multiple texts, then, constitutes a radical change indeed: not just an enlargement of Shakespeare's works but a need to reconceptu- alize the fundamental category of a work by Shakespeare. One of the most evident results of the multiple-text issue has been mounting resentment toward the editorial tradition. Eighteenth-century editors have been blamed for substituting a composite Lear for the seventeenth-century Lears, and subsequent editors have been charged with reproducing their predecessors' conflation. A denigration of editing in general has ensued, as if editors had been passing off an artificial Shakespeare for the real. What has not been stressed, however, is the extent to which literary critics have assumed and perpetuated the terms through which Shakespeare was reproduced in the eighteenth century. While editors examined and defended their choices, critics tended simply to assume the established status of the texts they used. In recent years the two dominant modes of reading Shakespeare-formalism and historicism-have received the text at hand on faith, whether it be, for example, the Alexander edition in England or the Riverside edition in America. Both forms of criticism have

This piece began as an introduction to a collection of essays by Randall McLeod/Random Cloud/Random Clod, to whose work we remain indebted. See, in particular, Randall McLeod, "Spellbound: Typography and the Concept of Old-Spelling Editions," Renaissance and Refor- mation, n.s., 3 (1979), 50-65; "Unemending Shakespeare's Sonnet 111," Studies in English Literature, 21 (1981), 75-96; "UNEditing Shak-speare," Sub-stance, 33/34 (1982), 26-55; Ran- dom Cloud, "The Psychopathology of Everyday Art," Elizabethan Theatre, 9 (1986), 100-168; "The Marriage of the Good and Bad Quartos," Shakespeare Quarterly, 33 (1982), 421-31; "'The very names of the Persons': Editing and the Invention of Dramatick Character" in Staging the Renaissance: Reinterpretations of Elizabethan and Jacobean Drama, David Scott Kastan and Peter Stallybrass, eds. (New York and London: Routledge, 1991), pp. 88-96; and Random Clod, "Information on Information," TEXT, 5 (1991), 241-81. We also wish to thank David Bevington, Jeff Masten, Barbara Mowat, Gary Taylor, and Paul Werstine for comments that helped shape this essay.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 256 SHAKESPEARE QUARTERLY taken for granted the identity of their object without realizing how that assumption subverts their very approach. Formalists call for exacting attention to the minutiae of literary language without giving thought to the printing-house practices that have in modern editions produced them. Historicists, tracing in Shakespeare's works the discursive structures specific to the late sixteenth and early seventeenth centuries, have ignored the extent to which these structures are eighteenth-century constructs.' This inattention to the textual object catches both approaches in methodological paradox: formalists closely read printed texts as if they were authorial compositions; historicists anachronistically read Enlightenment texts as if they were Renaissance discourses. With both editing and criticism holding fast to a modern text that derives from an eighteenth-century tradition, where can we turn? The First Folio and early quartos seem our only recourse, but not because they hold out promise of getting back to "the thing itself," uncontaminated by latter-day mediations and interventions. "The thing itself," the authentic Shake- speare, is itself a problematic category, based on a metaphysics of origin and presence that poststructuralism has taught us to suspect. (Indeed it was the search for such a chimera that vexed the editorial project in the first place.) Return to the early texts provides no access to a privileged "original";2 on the contrary, for the modern reader it bars access. The features that modernization and emendation smooth away remain stubbornly in place to block the illusion of transparency-the impression that there is some ideal "original" behind the text.3 These features are precisely the focus of this essay: old typefaces and spellings, irregular line and scene divisions, title pages and other paratex- tual matter, and textual cruxes. They constitute what we term the "materiality of the text."4 Discarded or transformed beyond recognition in stan-

1 See, however, Leah Marcus's insistence on the relation between the topicality of history and the specificity of its texts in Puzzling Shakespeare: Local Reading and its Discontents (Berkeley, Los Angeles, London: Univ. of California Press, 1988), esp. pp. 43-50; and "Levelling Shake- speare: Local Customs and Local Texts," SQ, 42 (1992), 168-78. 2 This point is made by Gary Taylor in the "General Introduction" to William Shakespeare: A Textual Companion, Stanley Wells and Gary Taylor with John Jowett and William Montgomery (Oxford: Clarendon Press, 1987), pp. 3-7. 3 G. Thomas Tanselle argues influentially for this bibliographic and hermeneutic model, maintaining that a work is an abstract linguistic entity that does not "exist on paper or in sounds": "Whatever concept of authorship one subscribes to, the act of reading or listening to receive a message from the past entails the effort to discover, through the text (or texts) one is presented with, the work that lies behind" (A Rationale of Textual Criticism [Phila.: Univ. of Pennsylvania Press, 1989], p. 18, emphases added). Tanselle's search for the underlying work is like what Stephen Orgel has discussed as the desire for the "authentic" Shakespeare: "The assumption is that texts are representations or embodiments of something else, and that it is that something else which the performer or editor undertakes to reveal" ("The Authentic Shakespeare," Representations, 21 [1988], 1-25, esp. p. 24). Unlike Tanselle's idealized "work," however, Orgel's "script" has always been subject to varying historical, theatrical, and editorial contingencies. See also Paul Werstine's discussion of how one specific version of this drive- what he defines as "the desire for a certain kind of narrative"-can lead textual criticism astray, in "Narratives About Printed Shakespeare Texts: 'Foul Papers' and 'Bad' Quartos," SQ, 41 (1990), 65-86, esp. p. 82. 4 The materiality of the early Shakespearean texts was scrupulously examined by the New Bibliographers in the first half of this century, but only as a means of discovering an idealized Shakespeare. See Margreta de Grazia, "The Essential Author and the Material Book," Textual Practice, 2 (1988), 69-86; Anne-Mette Hjort, "The Interests of Critical Editorial Practice,"

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 257 dard editions, they remain obstinately on the pages of the early texts, insisting upon being looked at, not seen through. Their refusal to yield to modern norms bears witness to the specific history of the texts they make up, a history so specific that it cannot comply with modern notions of correctness and intelligibility. Though modern editions leave them behind as inert and obsolete matter, the Folio and quartos enable us to confront the historical gap between the distinguishing marks of the modern "Shake- speare" and the distinguishing marks of a particular set of early modern theatrical texts. Attention to these earlier material traces is no exercise in antiquarianism. These older forms return as active agents calling our own forms into question. When the materiality of the early texts confronts modern practices and theories, it casts those modern practices and theories into doubt, revealing that they, too, possess a specific-and equally contingent-history. It makes us face our own historical situatedness. The materiality of this essay, for instance, when it quotes from Shakespeare, at no point reproduces the materiality of the early Shakespearean texts, even when quoting from the 1623 Folio. Not only are all our Shakespearean quotations in a modern typeface, but wve have silently emended long s to s, some us to vs, some vs to us. We do this because, despite our insistence on the specificity of early modern texts, we have no desire to perpetuate the illusion that we are presenting "original" or "unedited" text, in either its worn archival or fresh simulacral form. Even if we could convince ourselves that we had an "original" or "unedited" text, we would have established not its existence but rather the persistence of the epistemological categories that make us believe in its existence. This essay is an attempt to interrogate what we take to be the four categories basic to the dominant post-Enlightenment treatment of Shake- speare. Beginning with the category of a single work, we proceed to that of the discrete word. We continue with the ostensible sources of those two signifying units: the unified character, who utters the word, and the autonomous author, who is credited with the work. This process reverses what Frederic Jameson has termed the "dynamics of the historical tribunal."5 It is not present standards that pass judgment on past forms, but rather past forms that return to try present standards.

I. WORK

As noted above, the recent debate over the texts of King Lear has raised questions about the identity of the work.6 Beginning in the eighteenth

Poetics, 15 (1986), 259-77; and Hugh Grady, The Modernist Shakespeare: Critical Texts in a Material World (Oxford: Oxford Univ. Press, 1991), pp. 57-63. Paul Werstine's discussion of the New Bibliographers' reliance on "foul papers," the purported original drafts of the author, is also relevant; see both his "Narratives" and "McKerrow's 'Suggestion' and Twentieth- Century Shakespeare Textual Criticism," Renaissance Drama, 19 (1989), 149-73. So, too, is Terence Hawkes's discussion of the sacramental resonances of this bibliographic quest, as well as of its philosophical idealism, in That Shakespeherian Rag: Essays on a Critical Process (London: Routledge, 1986), pp. 74-76. 5 On the value of this upsetting dialectic reversal of past and present, see "Marxism and Historicism" in his The Ideologies of Theory: Essays 1971-1986, 2 vols. (Minneapolis: Univ. of Minnesota Press, 1988), Vol. 2, p. 175. 6 In 1976, Michael Warren delivered the ground-breaking paper that argued for the

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 258 SHAKESPEARE QUARTERLY century, editors commonly assumed that both the 1608 quarto Lear and the 1623 Folio Lear imperfectly represented the Lear Shakespeare had written; they therefore attempted to recover the authorial Lear by conflating the two texts. Since 1976, however, an impressive band of textual scholars has insisted upon the anachronism of the eighteenth-century conflated play and has argued that both the quarto Lear and the Folio Lear should be recognized as discrete plays. What formerly appeared as one play in all collections of the Works has now been printed as two in the 1986 Oxford Collected Works. In Michael Warren's 1989 The Complete King Lear 1608-1623, the play has been photoduplicated as four discrete units: the 1608 quarto True Chronicle Historie, the 1619 quarto True Chronicle History, the 1623 Folio Tragedie, and Warren's own two-in-one Parallel Texts of the first and last of these three facsimiles.7 Indeed, four is a conservative tally; since all but the last unit consist of loose rather than bound pages, the materials lend themselves to the constructing of additional textual units, conceivably an exponential number of them. Nor is the problem of a work's nonidentity with itself limited to King Lear: Hamlet has recently been published as three texts, and multiple-text editions have been considered for Othello and Troilus and Cressida.8 As the several titles for Lear demonstrate, what editors have taken as a single work possesses multiple names as well as multiple texts. Titles for a given play vary not only from title page to title page but also from title page to Stationers' Register entry, so that the 1608 quarto is registered as "Mr Wllm Shakespeare his historye of Kinge Lear.... "In this instance even the

recognition of the quarto and Folio Lears as separate texts, noting that three other scholars were simultaneously and independently arguing for the same distinction. Two books brought general attention to the issue, Steven Urkowitz's Shakespeare's Revision of King Lear (Princeton, N.J.: Princeton Univ. Press, 1980) and Gary Taylor and Michael Warren's collection of essays, The Division of the Kingdoms: Shakespeare's Two Versions of King Lear (Oxford: Clarendon Press, 1983). Two texts of Lear are printed in William Shakespeare: The Complete Works, Stanley Wells and Gary Taylor, gen. eds. (Oxford: Clarendon Press, 1986). For a summary of the Lear controversy, see Grace loppolo, Revising Shakespeare (Cambridge, Mass.: Harvard Univ. Press, 1991), pp. 1-5 and 163-67. 7 Warren notes that "the book demonstrates the problem of the term King Lear. King Lear here is not a single play presented in sanitized form, domesticated, processed for easy consumption" (The Complete King Lear 1608-1623 [Berkeley: Univ. of California Press, 1989], p. vii).. 8 The Three-Text Hamlet: Parallel Texts of the First and Second Quartos and First Folio, AMS Studies in the Renaissance 30, ed. Paul Bertram and Bernice W. Kliman (New York: AMS Press, 1991). Multiple-text modern editions for Hamlet, Othello, and Troilus, as well as Lear, were prepared by the editors of the Oxford 1986 edition cited above. See also Stanley Wells and Gary Taylor, "The Oxford Shakespeare Re-viewed by the General Editors," Analytical and Enumerative Bibliography, n.s., 4 (1990), 6-20. On textual instability generally, see E.AJ. Honigmann, The Stability of Shakespeare's Texts (London: Edward Arnold, 1965), and Stephen Orgel, "What is a Text?" in Kastan and Stallybrass, eds., pp. 83-87; on the problem of the Lear texts, see Taylor and Warren, eds. On the instability of individual plays, see: for Hamlet, Barbara Mowat, "The Form of Hamlet's Fortunes," RenD, 19 (1988), 97-126, Joseph F. Loewenstein, "Plays Agonistic and Competitive: The Textual Approach to Elsinore," RenD, 19 (1988), 63-96, Paul Werstine, "The Textual Mystery of Hamlet," SQ, 39 (1988), 1-26; for Troilus, Taylor, "Troilus and Cressida: Bibliography, Performance, and Interpretation," Shake- speare Studies, 15 (1982), 99-136; for Henry V, Taylor, Three Studies in the Text of 'Henry V' (Oxford: Oxford Univ. Press, 1979). For extensive accounts of revisionist activity, see Wells and Taylor, Textual Companion, pp. 16-18; Taylor, "Revising Shakespeare," TEXT, 3 (1987), 285-304; and loppolo, passim.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 259

genre of the play is mutable, identified in the 1608 quarto as "Chronicle Historie," in the entry as "historye," and in the Folio as "Tragedie."9 In certain cases titles do not correspond closely enough to warrant confident identification of a text with a Stationers' Register entry. For example, we cannot be sure what Blount's entry of 20 May 1608 for "The booke of Pericles prynce of Tyre" was intended to protect. Was it his right to copy the 1609 The Late And much admired Play, Called Pericles, Prince of Tyre. With the true Relation of the whole Historie, adventures, andfartunes of the said Prince . . ., or George Wilkins's 1608 novella The Painfull Adventures of Pericles Prince of Tyre. Being the true History of the Play of Pericles, or the 1607 reprint of Laurence Twine's prose romance The Pattern of Painful Adventures ... That Befell unto Prince Apollonius? Or did it protect Blount's right to some text that no longer exists; or indeed did it protect his right to print all of the texts referred to here? For purposes of assigning ownership, one title could comprehend different texts with similar titles, even when their titles distinguished them by genre, as in the Lear and Pericles examples above. The first person to register that title might become the de facto owner of all similarly titled texts to appear in the future. As Peter W. M. Blayney has noted, Blount and Jaggard seem to have faced this problem when attempting to secure the rights to the plays of the First Folio.'0 Records suggest, for instance, that they had to negotiate for the rights to the anonymous The Taming of a Shrew (1594) and The Troublesome Reign of John (1591) in order to establish their title to Shakespeare's The Taming of the Shrew and King John. One suspects that, analogously, the anonymous King Leir and Shakespeare's Lear were not clearly distinct textual properties. Just as a single name could refer to a variety of texts, so, too, a single text could be referred to by a variety of names. Thus the play published in 1607 as The Famous History of Sir Thomas Wyatt, assigned on the title page to Thomas Dekker and John Webster, appears in Henslowe's diary as both "A playe called Ladey Jane" and "the playe of the overthrowe of Rebelles. "" The confusion of titles and texts in multiple Lears and Hamlets seems, then, to have been typical in the early modern printing houses,-an example of what D. F. McKenzie has termed "the normality of non-uniformity."'2

9 On the fluidity of generic divisions in the Folio, see Orgel, "Shakespeare and the Kinds of Drama," Critical Inquiry, 6 (1979), 107-23; on their fluidity in eighteenth-century editions, see de Grazia, Shakespeare Verbatim (Oxford: Clarendon Press, 1991), p. 149, n. 42. 10 The First Folio of Shakespeare (Washington, D.C.: The Folger Shakespeare Library, 1991), p. 2. 11 Gerald Eades Bentley, The Profession of Dramatist in Shakespeare's Time, 1590-1642 (Prince- ton, N.J.: Princeton Univ. Press, 1971), pp. 203-4; Henslowe's Diary, ed. R. A. Foakes and R. T. Rickert (Cambridge: Cambridge Univ. Press, 1961), pp. 218-19. 12 D. F. McKenzie coins this phrase in his discussion of the variables involved in determining printing-house averages in "Printers of the Mind: Some Notes on Bibliographical Theories and Printing-House Practices," Studies in Bibliography, 22 (1969), 1-75, esp. pp. 8-13. Peter Blayney also notes the regularity of irregularities in his account of the proofing and printing order of the 1608 quarto of Lear in The Texts of King Lear and their Origins, Volume 1: Nicholas Okes and the First Quarto (Cambridge: Cambridge Univ. Press, 1982), pp. 89-218. Marion Trousdale discusses the recurrence of disuniformity in Renaissance theatrical and printing-house practices, as well as the New Bibliography's inadequacies in dealing with it, in "Diachronic and Synchronic: Critical Bibliography and the Acting of Plays" in Shakespeare: Text, Language, Criticism: Essays in Honour of Marvin Spevack, Bernhard Fabian and Kurt Tetzeli von Rosador,

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 260 SHAKESPEARE QUARTERLY

This nonuniformity is hard to capture in modern texts of Shakespeare, even when the editor wishes to take into account the textual and titular diversity of these early publications. How is the very name of a text to appear? In the case of multiple-text plays, how many variants between texts of a given play warrant the reproduction of the play in multiple forms? In short, how can reproduction be true to either a title or a text when that title or text is not-to our eyes, at least-true even to itself? Mechanical copying in Shakespeare's day, not unlike scribal copying, varied its object in duplicating it, affording a particularly concrete instance of the poststructuralist insight that every repetition introduces difference. Paradoxically, while modern photographic technology enables us to reproduce, both rapidly and relatively cheaply, the labor-intensive produc- tions of the Renaissance printshop so that we can manufacture facsimiles without resetting type, this new technology is radically incompatible with the Renaissance "original" it would duplicate. There is no monadic "original" and hence no authenticity either, for the latter depends on the former. As Walter Benjamin perceives in discussing mechanical reproduction: "The presence of the original is the prerequisite to the concept of authenticity."'13 Stephen Orgel in both his critical and editorial work argues that the concept of authenticity is not applicable to what must be imagined as an "unstable, infinitely revisable script."'14 As Orgel insists, the performance has always been regarded as malleable and permeable; the script "is essentially unstable and changes as the performers decide to change it."''5 Even when removed from theatrical contingency and consigned to what historians of the book have termed (perhaps prematurely with respect to the Renaissance) "the fixity of print," the text remains provisional. Because it has commanded more editorial attention than any other early modern publication, Shakespeare's First Folio best illustrates the unsettled nature of a printed text.'6 Charlton Hinman's 1968 Norton facsimile of the 1623 First Folio displays the variability that was built into the very practices of printing-house production.'7 Because of the printing-house practice of correcting proof during the course of printing and then indiscriminately assembling corrected and uncorrected sheets, it is highly probable that no two copies of the Folio are identical. Peter Blayney has recently proposed another explanation for the variations among Folio copies: probably be-

eds. (Hildesheim, Zurich, New York: Olms-Weidmann, 1987), pp. 304-14, and "A Second Look at Critical Bibliography and the Acting of Plays," SQ, 41 (1990), 87-96. 13 Illuminations, trans. Harry Zohn (New York: Shocken Books, 1969), p. 220. 14 "The Authentic Shakespeare" (cited in n. 3, above), p. 24; see also his "What is a Text?" in Kastan and Stallybrass, eds. 15 "Shakespeare Imagines a Theater," Poetics Today, 5 (1984), 549-61, esp. p. 558. For illustrations of various transformations to which the text has yielded historically and presently, see "The Authentic Shakespeare"; Orgel's unpublished manuscript "Acting Scripts, Perform- ing Texts" also addresses this topic. 16 According to Peter Blayney, "virtually all English books printed before the eighteenth century vary to some extent from copy to copy" (First Folio, p. 15). In a lecture delivered at the University of Pennsylvania in April 1990, McLeod demonstrated a particularly radical disuniformity among copies of Holinshed's Chronicles; copies were bound in different orders and with different interpolations with the result that-in the most literal sense-there is no single narrative in the Chronicles. 17 The Norton Facsimile: The First Folio of Shakespeare, ed. Charlton Hinman (New York: W. W. Norton, 1968).

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 261

cause of a delay in obtaining the rights to Troilus and Cressida, "three distinct issues" of the Folio were successively sold in 1623: one without Troilus, one with Troilus, and one with both Troilus and the Prologue to Troilus.'8 As Hinman readily admitted, the Norton facsimile reproduces "an ideal representation," which "almost certainly never has been realized in any actual copy of the edition."'9 A modern simulation compiled from thirty of the Folger's copies, the facsimile is designed to represent "what the printers of the original edition would themselves have considered an ideal copy."20 This "ideal representation" bears witness to the unattainable goal of achiev- ing an exact reproduction of a text that never possessed a single or fixed form. Moreover, as Gary Taylor remarks, "photography itself, like any other medium, communicates its own message: that what you see is real, accurate, genuine."'2' The process that appears best suited for duplicating the material text ends up reproducing only one of its multiple forms. Facsimiles confer a sanctity upon the particulars of the duplicated text, hypostatizing forms that were quite fluidly variable at their publication. Photo-reproduction arrests the textual drift characteristic of both the mechanics and semantics of the early quartos and First Folio. Michael Warren's The Complete King Lear designedly resists this arrest by opening up the textual proliferation that was endemic to early modern printing practices. Ironically, modern editions of Shakespeare achieve something of the same destabilization: while aiming to reproduce the "ideal text," their very number and variety through the centuries approximate the instability of the early playtexts. The First Folio's status as "the chief authority" in determining the Shake- speare canon is also problematic.22 This corpus was anticipated by a very different one in 1619 when Thomas Pavier made the first attempt to publish the "collected works" of Shakespeare.23 If not for the lord chamberlain's letter forbidding the publication of plays belonging to the King's Men without their permission, it might have been Pavier's 1619 quartos rather than Blount and Jaggard's 1623 Folio that defined Shakespeare's canon. Shakespeare's Collected Works would then have consisted of Pericles, A Yorkshire Tragedy, and The first part of the true and honorable history, of the life of Sir Iohn Old-castle, plus seven of the Folio's thirty-six plays. In 1664 another seventeenth-century printer attempted to define the Shakespeare corpus: Philip Chetwinde in the second edition of the Third Folio added to the thirty-six plays of the First and Second Folios not only Pericles but seven

18 As Blayney notes, the second version also contained two copies of the last page of Romeo and Juliet, the second crossed out by hand (First Folio, pp. 21-24). 19 p. xii. See also G. Thomas Tanselle, "Reproductions and Scholarship," SB, 42 (1989), 25-54, esp. p. 37. 20 Hinman, ed., pp. xxii-xxiii. In preparing the facsimile texts for The Complete King Lear 1608-1623, Michael Warren (cited in n. 7, above) followed Hinman's precedent. See Marion Trousdale's discussion of Folio reproduction in "A Trip Through the Divided Kingdoms," SQ, 37 (1986), 218-23; see also "A Second Look," pp. 91-92 and 94. 21 Textual Companion (cited in n. 2, above), p. 4. 22 E. K. Chambers, William Shakespeare: A Study of Facts and Problems, 2 vols. (Oxford: Clarendon Press, 1930), Vol. 2, p. 207. 23 For a description of Pavier's attempt, see Chambers, Vol. 1, pp. 133-37; and W. W. Greg, The Shakespeare First Folio: Its Bibliographic and Textual History (Oxford: Clarendon Press, 1955), pp. 9-16.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 262 SHAKESPEARE QUARTERLY others. Four of these plays (Pericles, A Yorkshire Tragedy, The London Prodigall, and I SirJohn Oldcastle) had been ascribed to Shakespeare on the title pages of earlier quartos, and three (The Lamentable Tragedie of Locrine, The True Chronicle Histories of the whole life and death of Thomas Lord Cromwell, and The Puritane or Widow) had been previously ascribed to "W.S."24 For over fifty years this "new" Shakespeare canon replaced the "old": it was reproduced by the 1685 Fourth Folio and Rowe's 1709 edition; Pope removed the "new" plays in 1725 but reintroduced them into his 1728 edition (in a separate volume). The Third Folio is instructive in exemplifying how authorship is constructed on the basis of previous title-page attributions: plays we regard as apocryphal were considered Shakespeare's because his name or initials appeared on the title page. Seventeenth-century publishers like Chetwinde seem to have been less concerned with who wrote a given play than with its previous ascription in print. They also had to secure the right to copy before a play could be incorporated into the canon, as is demonstrated by Blount's and Jaggard's efforts to register with the Stationers' Company plays previously unpublished and to negotiate with other publishers for those in print. Pericles, it has been suggested, was excluded from the Folio because there were questions regarding not Shakespeare's authorship but rather Blount and Jaggard's ownership.25 Since the eighteenth century, attempts to de- termine whether a work belongs in the Shakespeare canon emphasize evidence of Shakespearean authorship (style, rhyme, word use, etc.). Yet the principle of selection that led to publications like the Folio may have had largely to do with the status of a group of playtexts in the acting company and in the Stationers' guild.

II. WORD

The problem of reproduction exists not only in regard to a work (its title, its text, the textual corpus to which it belongs) but also in regard to that minuscule unit that in multiples comprises it: the single word.26 This category has always commanded attention in the production of texts, early

24 Chambers, Vol. 1, pp. 532-37. The acceptance of this new, more "complete" canon as also the more authoritative is suggested by the Bodleian Library's disposal of the First Folio upon acquiring the Third (Blayney, First Folio, p. 34). Seventeenth-century playlists also suggest the provisionality of the Shakespeare canon: in the list attached to The Careless Shepherd (published by Richard Rogers and William Ley in 1656), not only Edward II but also Edward III and Edward IV are attributed to Shakespeare; in the list attached to The Old Law (printed by Edward Archer in 1656), The Arraignment of Paris, The Chances, Mucedorus, Hoffman, The Merry Devil of Edmonton, and both parts of Hieronimo are all attributed to Shakespeare, while Loves labor lost is attributed to William Sampson; in Francis Kirkham's list, appended to Tom Tyler and his Wife (1661), The Two Noble Kinsmen is attributed to Fletcher alone; see W. W. Greg, "Attribution in the Early Play-Lists 1656-1671," Edinburgh Bibliographical Society Transactions, 2 (1938-45), 303-29. 25 As noted above, Troilus and Cressida was left out of some early copies of the 1623 Folio, doubtless because Blount and Jaggard succeeded only at the last minute in securing the rights to it; there is also evidence that Richard II, I and 2 Henry IV, and Richard III came close to being excluded for the same reason (Blayney, First Folio, pp. 2-4 and 17-21). 26 The word one itself multiplies promiscuously in the sixteenth and seventeenth centuries, appearing as "an, "ain," "yane," "oon," "own," "oowne," "won," "wone," "woon," "wan," "o," "oo," "on," as well as "one." (It is no wonder that Spenser turned to a more orthographically stable Latin for the name of his steadfast Una.)

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 263 modern as well as modern, for errors must be caught and corrected. An incorrect Folio or quarto word is generally assumed to be the result of accident: the author nodded; a compositor slipped ("sleeepe," "woman- dood," "inongh"); a letter broke in the printing process.27 Yet identifying an accident can be difficult when dealing with materials produced prior to the establishment of standards of correctness. Before a field of regularity has been determined-before lexical and grammatical standardization-by what norms can an irregularity or "crux" be identified? This is not to say that the category of error did not exist in Renaissance England (errata entries alone suffice to prove otherwise); rather, what we identify as anomalies might to a Renaissance reader have been quite literally typical. Let us take a striking and familiar example from The Tragedie of Macbeth. In most editions of the play, the witches refer to themselves as the "weird sisters," and editors provide a footnote associating the word with Old English wyrd or "fate." But we look in vain to the Folio for such creatures. Instead we encounter the "weyward Sisters" and the "weyard Sisters"; it is of "these weyward Sisters" that Lady Macbeth reads in the letter from Macbeth, and it is of "the three weyward sisters" that Banquo dreams.28 The OED supports the editorial modernization of "weyward" to "weird" by recording "weyward" as a possible Renaissance spelling of "weird." It would seem, then, that editors have correctly modernized. As it turns out, however, Macbeth itself is the OED's only example of this spelling, and Theobald's 1733 printing of "weyward" as "weird" is the OED's sole authority for identifying the two forms. In other words, it is an eighteenth- century editor who here constructs not only the editorial tradition but also the research tools (i.e., the OED entries) with which that tradition might be interrogated. Another emendation is possible: the Folio's "weyward" could be modernized as wayward (the very adjective Hecate applies to Macbeth29), a simple vowel shift that effects a striking semantic one, transposing the sisters from the world of witchcraft and prophecy (conjured up by Theobald's "weird") to one of perversion and vagrancy. In this instance the OED does offer non-Shakespearean precedents for spelling "wayward" as "weyward" (with weywarde occurring in the early sixteenth century). This cannot, however, eliminate the possibility of the undocumented "weird." Even if the OED's orthographic history were exhaustive in citing Macbeth as the only instance of "weyward" for "weird", the reading would still be possible for Renais- sance readers who in both manuscript and print would frequently encounter rare or nonce orthographies.30 Without such lexical tools as the OED, on what basis could they rule it out?

27 Jeanne Addison Roberts draws attention to a particularly interesting example of such a break in" 'Wife' or 'Wise'-The Tempest 1. 1786," SB, 31 (1978), 203-8. For the significance of the discovery, loss, and rediscovery of this reading, see Stephen Orgel, "Prospero's Wife," Representations, 8 (1984), 1-13. 28 TLN 130 (1.3.32), 355 (1.5.8), 596 (2.1.20). Except where otherwise specified, quotations from and through-line-number citations of Shakespeare's plays follow Hinman. They are followed by parenthetical act, scene, and line numbers from The Riverside Shakespeare, ed. G. Blakemore Evans (Boston: Houghton Mifflin, 1974). 29 TLN 1441 (3.5.11). 30 Random Clod notes how the modern dictionary functions as "the editor's army." He also observes that there were no Shakespeare Editors in Shakespeare's time ("Information upon Information" [cited in the unnumbered acknowledgment note, above], p. 248).

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 264 SHAKESPEARE QUARTERLY

The word weyard occurs three times in Folio Macbeth: "the weyard Women," "the weyard Sisters," and "the Weyard Sisters."3' The OED again cites Macbeth as the first use. This form's approximation of the OED's other cognates for "weird"-"wyrde," "wyerde," "weard"-might support Theobald's "weird", especially in light of the play's source, the account of the reigns of Duncan and Macbeth in Holinshed's Chronicles. In Holinshed the witches are described as "the weird sisters, that is (as ye would say) the goddesses of destiny, or else some nymphs or feiries" (who are surprisingly well dressed in the woodcut illustration).32 But by forgoing "weyward" and "weyard" in favor of "weird", modern editors presuppose a notion of the signified which retroactively governs the field of the signifier. For the comparison of Macbeth's epithets with Holinshed's suggests that the field of the signifier is a field of difference-a field where "true etymologies" are displaced by the orthographic play that allows for (indeed cannot prevent) the intermingling of signifiers and the coining of ever-new etymologies. The signifiers that name the three sisters in the Folio also enact the very linguistic errancy that, as several recent readings have noted, is associated with them.33 There is an uncanny irony to the weirdlwyrd emendation, as if editors would prefer the overdeterminations of the three Fates to the indeterminacy of their less predictable cousins/cognates. But the Folio's "weyward" and "weyard" sisters and the Chronicles' "weird" sisters belong not within the straitened confines of a later monologic orthography but in a semantic field that, not yet ruled by lexical statute, accommodates verbal vagrancy. Within that same linguistic field is air-"ayre" in the Folio-the element into which the three sisters and their apparitions vanish, a word repeated more often in this play than in any other of Shakespeare's works. Appearing from and disappearing into air, the sisters hail both Macbeth and Banquo with news of heirs: Macbeth will be king but Banquo will father a line of kings, a disturbing genealogical nonsequitur that compels Macbeth to order the deaths of both the father and his heir, though the escape of the latter drives Macbeth back to the witches to know what he has already been told-"Shall Banquo's issue ever / Reigne in this Kingdome?"34 The witches repeat their prophecy by staging a proleptic enactment: the show of Ban- quo's heirs stretching out till doomsday. This prophecy, too, fails to follow logically from what precedes it (i.e., ostensible signs of Macbeth's invincibility), and once again Macbeth lashes out at heirs, not Banquo's but Macduff's "Babes, and all unfortunate Soules / That trace him in his

31 TLN 983 (3.1.2), 1416 (3.4.132), 1686 (4.1.136). 32 The Firste volume of the Chronicles of England, Scotlande, and Irelande (London: Lucas Harrison, 1577), "The Historie of Scotlande," p. 243. 33 For recent studies of Macbeth that associate the three sisters with the linguistic excess and indeterminacy that characterizes their signifier, see: Harry Berger, Jr., "The Early Scenes of Macbeth: Preface to a New Interpretation," and Stephen Mullaney, "Lying Like Truth: Riddle, Representation and Treason in Renaissance England," both in English Literary History, 47 (1980), 1-31 and 32-47; Terry Eagleton, William Shakespeare (Oxford and New York: Basil Blackwell, 1986), pp. 1-8; and Janet Adelman, Suffocating Mothers: Fantasies of Maternal Origin in Shakespeare's Plays, Hamlet to The Tempest (New York and London: Routledge, 1992), pp. 130-46. 34 TLN 1646-47 (4.1.102-3).

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 265

Line."35 Unlike Macduff and Banquo, Macbeth possesses no perpetuating line ("No Sonne of mine succeeding"36) and therefore no claim on the future. Indeed, he substitutes the witches' illusory assurances for the end- less succession of heirs that keeps Banquo alive, a substitution evident in his fatal boast to Macduff, which ascribes to his own paltering charm the invincibility of air that successive heirs confer upon Banquo:

As easie may'st thou the intrenchant Ayre With thy keene Sword impresse, as make me bleed: Let fall thy blade on vulnerable Crests, I beare a charmed Life....37

Macbeth's imperfect manhood is registered also in states of fear, like that brought on when he imagines murdering his king: in modern editions that "horrid image doth unfix my hair"; in the Folio, however, it unfixes "my Heire," allowing for a double unsettling-both of his manhood and of Scotland's dynasty.38 Hair and heir also combine, almost surrealistically, in the final scene of the play when old Siward, learning that his heir, Young Siward, died "like a man," shows himself able to take the news of his son's death "like a man" by declaring: "Had I as many Sonnes, as I have haires, / I would not wish them to a fairer death."39 Finally, Macbeth's response to the show of kings provides a remarkable instance of a phonetic and semantic convergence of what are for us three distinct words-air, heir, and hair:

Thou art too like the Spirit of Banquo: Down: Thy Crowne do's seare mine Eye-bals. And thy haire Thou other Gold-bound-brow, is like the first: A third, is like the former.40

Modern editions, of course, must opt for one of the three modern choices, usually hair.4' But all three possibilities (and their semantic connections) reside in the Folio's "haire." Indeed, we have to let hair slide into air-not the infected and filthy air on which the witches ride and hover, but the noble air of majesty, as in the archaic debonair-to avoid the absurdity of Macbeth's fixating on a succession of identical heads of royal hair. Nor can heir be excluded, for clearly it is Banquo's train of heirs that resemble him and one another; in addition, heir works best in apposition to the metonymic "Gold-bound-brow." How- ever phantasmagoric this kind of semantic slipping and sliding may seem to a modern sensibility, Renaissance textuality encourages it. A text that allows hair to appear as "hair," "heir," "heire," "heere," and "here" and heir as

35 TLN 1705-6 (4.1.152-53). 36 TLN 1054 (3.1.63). 37 TLN 2448-51 (5.8.9-12), emphasis added. 38 TLN 246 (1.3.135). 39 TLN 2488 (5.9.9), 2496-97 (5.9.14-15); M. M. Mahood also comments on these lines in Shakespeare's Wordplay (London: Methuen, 1957), p. 141. 40 TLN 1659-62 (4.1.112-15). 41 Johnson proposed air and Steevens endorsed it with a citation from The Winter's Tale, though without remarking on the supporting presence of heir: "Your father's image is so hit in you, / His very air, that I should call you brother" (5.1.127-28); see Macbeth: A New Variorum Edition of Shakespeare, ed. H. H. Furness (Philadelphia: J. B. Lippincott, 1873). We are indebted to Jamie Saegar's unpublished work on the hair/heir/air homonym.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 266 SHAKESPEARE QUARTERLY

"heir," "aire," "are," "haire," and "here" is going to be open to runnings of all kinds: run-ins, run-ons, runovers, runoffs. Modernization requires that this slippage be contained. A costly requirement, as both Fredson Bowers and Stanley Wells have acknowledged. But even the preservation of Folio or quarto spellings cannot solve the problem, for modern readers have in effect internalized the modernizing process that the editor performs on the text. The modern orthographic system that presides over their reading will fix the Folio's "Heire" so as to preclude the possibility of hair. Thus facsimiles and old-spelling editions are also problematic, for modern linguistic categories are mobilized there too, erasing an earlier system that was defined by phonetic, orthographic, and semantic plasticity.42 But how much hangs upon such small verbal points? Not much, it might be argued-although a text is literally composed of just such small points. When confronted with a crux, the editor looks for the right single word; in the case of weird, editors conjecture that the word came from Shakespeare's source. But the boundaries that separate weywardlweyardlweirdlwayward into four discrete and mutually exclusive lexical units had not yet been drawn and systematically reproduced. Until dictionaries fixed these boundaries, cognates blurred, phonetically and orthographically, without regard to the post-lexical determinations that subsequently divided them.43 Whether Holinshed or Shakespeare or a given scribe or compositor of either author's work determined a given form is less significant than the capacity of a word in the language's preregulative or generative phase to take multiple forms.44 This is precisely what baffles the project of retrieving the correct word, for it is a semantic field and not a single word that needs to be retrieved, and our linguistic categories efface that field in the very act of attempting to elucidate it. In its place a regularized text appears which, often on the basis of what the editor thinks was intended, selects for us a single word from the larger field. In modernized editions the mutable Renaissance signifier disappears: weird ceases to be linguistically (and epis- temologically) vagrant; heir ceases to be philologically (and genealogically) unsettling.

III. CHARACTER

In a recent essay Harry Berger, Jr., has reversed the logic of traditional character criticism. A long tradition beginning with Pope, peaking in A. C. Bradley, and continuing into this century particularly through Freudian

42 See Fredson Bowers, On Editing Shakespeare (Charlottesville: Univ. Press of Virginia, 1966), pp. 155-65, and Essays in Bibliography, Text, and Editing (Charlottesville: Univ. Press of Virginia, 1975), pp. 289-95; see also Stanley Wells, Modernizing Shakespeare's Spelling (Oxford: Clarendon Press, 1979). 43 On the indefinition of boundaries between words, see de Grazia, "Homonyms Before and After Lexical Standardization," ShakespeareJahrbuch, 127 (1990), 143-56; and Peter Stallybrass, "Shakespeare, the Individual, and the Text" in Cultural Studies, Lawrence Grossberg, Cary Nelson, and Paula Treichler, eds. (New York and London: Routledge, 1992), pp. 593-610. 44 On the generative and restrictive phases of vernaculars, see John Earl Joseph, Eloquence and Power: The Rise of Language Standards and Standard Languages (London: Frances Pinter, 1987); for an account (perhaps exaggerated) of Shakespeare's contribution to the generative phase in English, see Bryan A. Garner, "Shakespeare's Latinate Neologisms," ShStud, 15 (1982), 149-70.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 267 psychoanalytic criticism has understood character to precede language. Characters are imagined as having developed prior to and independent of the plays in which they appear and as speaking a language that reflects this experiential and psychological history. Berger reverses this relation, arguing that dramatic character is subsequent to and dependent upon language: "speakers as characters are the effects rather than the causes of their language.... " For the critic this reversal has important methodological implications: semiotic analysis must precede psychoanalysis, or, more specifically, "Speakers don't have childhoods unless and until they mention them."45 Berger's critique derives from Lacan's emphasis on the construction of subjectivity in discourse, but it might have been grounded in less theoretical matters. In a modern edition of a play, the list of dramatis personae precedes the play, suggesting that characters preexist their speeches. Shakespeare's first readers, however, received no such suggestion, for none of the quartos published in his lifetime feature lists of characters; the Folio includes lists for only seven out of thirty-six plays, and in every case the list appears after rather than before the play.46 Readers had to arbitrate for themselves the boundaries of identity, constructing (or failing to construct, or refusing to construct) "individual" characters in the process of reading. Quite literally without a program and therefore not programmed to encounter a group of unified characters, they instead had to negotiate an array of positionalities relating to rank, family, gender, age, and even the specific personnel of the theatrical company. Although many Restoration single-play editions of Shakespeare's plays included dramatis personae, the first edition to affix a list to every play was Nicholas Rowe's 1709 The Plays of Shakspeare.Y Rowe's immediate successor, Alexander Pope, was therefore the first editor to read Shakespeare with a list of dramatis personae preceding every play-which might explain his elaboration of the individuality of Shakespeare's characters. In the preface to his 1725 edition, Pope insists that other authors copy copies, but the very notion of mimesis belittles the originality of the Bard: "If ever any Author deserved the name of an Original, it was Shakespear"; "His Characters are so much Nature her self, that 'tis a sort of injury to call them by so distant a name as Copies of her." Pope asserts that "[e]very single character in Shakespear is as much an Individual, as those in Life itself," and every speech so characteristic that its utterer, even in the absence of speech prefixes, could be identified.48

45 "What Did the King Know and When Did He Know It? Shakespearean Discourses and Psychoanalysis," South Atlantic Quarterly, 88 (1989), 811-62, esp. pp. 813 and 823. 46 As Barbara Mowat has pointed out, four of these dramatis personae are based on the transcripts of Ralph Crane, who might have been following the model of the Jonson folio, which did feature lists before each play ("Nicholas Rowe and the Twentieth-Century Shake- speare Text," forthcoming in the Proceedings of the Fifth World Shakespeare Congress, Tokyo, August 1991). Jonson's folio generally followed the precedent set by translations of classical drama, thereby conferring upon itself the status of classic. See Orgel, "Shakespeare Imagines a Theater" (cited in n. 15, above), and de Grazia, Shakespeare Verbatim (cited in n. 9, above), pp. 33-37. 47 See Mowat, "Nicholas Rowe." 48 "Preface to Edition of Shakespeare (1725)," quoted here from Eighteenth Century Essays on Shakespeare, ed. D. Nichol Smith (Glasgow: James MacLehose and Sons, 1903), pp. 47-62, esp.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 268 SHAKESPEARE QUARTERLY

Not only did early texts lack dramatis personae lists; they also lacked another modern textual marker of stable identity: uniform nomenclature. In the early texts a character's designation can vary in stage directions, in speech prefixes, and in the text itself. A reader would therefore have had to negotiate the various names for the Roman people (or peoples?) in the Folio's stage directions for The Tragedy of Coriolanus: "Mutinous Citizens," "the Plebeians," "the People," "the Rabble," "a Troope of Citizens."49 If the denomination of the multi-headed "Hidra" changes, so, too, does that of the "too absolute" Coriolanus, whose designations include "Mar.," "Mart.," "Martius.," and "Cor.," "Cori.," "Corio.," "Coriol.," as well as "Com." and "Mene."50 In Macbeth we find that vagrant nomenclature is hardly restricted to the Three Sisters. While the stage directions clearly distinguish "Seyton" in 5.3 and 5.5 from "Seyward" in 5.4 and 5.6, the speech prefixes undo the clarity of that distinction: Seyton is once "Seyt." but otherwise "Sey."; in 5.4 Seyward moves from "Syew." to "Syw." to "Sey.," and in 5.6 he appears only as "Sey." Instability of naming occurs not only within single texts but also between different "versions" of the "same" play. A stage direction in Romeo andJuliet reads in Q1, "Enter Servingman"; in Q2, "Enter Will Kemp"; and in F 1, "Enter Peter."5' And within the single text of Q2 Romeo andJuliet, the "Lady Capulet" of modern editions is fractured into "Ca. W.," "Capu. Wi.," "La.," "M.," "Mo.," "Old La.," "Wi.," and "Wife." In the stage directions there is a further proliferation of her names: "Madame" and "Lady of the House. "52 Ronald B. McKerrow, in his Prolegomena for the Oxford Shakespeare, pur- sues the double strategy of first limiting this proliferation ("in Romeo and Juliet, Capulet alternates with Father, and Lady Capulet with Mother") and then deleting it:

To follow the original texts in this irregularity would, however, be unnecessar- ily confusing to a reader, and as, after all, these speech-prefixes are merely labels intended to show to whom the various speeches are to be attributed, it seems to me an editor's clear duty to treat them as labels and to make the labels uniform.53

The rhetoric of "after all" and "merely" leads to the "clear duty" of the editor to impose uniformity and clarity and, in the process, to ensure the easy readability of character. Curiously, some Renaissance play manuscripts suggest the literal sense in which character is posterior to speech. William B. Long, for instance, notes that "probably all of the speakers' names were

p. 48. See also Cloud, " 'The very names of the Persons'" (cited in the unnumbered acknowledgment note, above). 49 TLN 2 (1.1.1), 1552 (2.3.154), 1950 (3.1.228), 1993 (3.1.262), 3053 (4.6.128). 50At TLN 1086 (2.1.179), modern editors emend "Com." to "Cor.," and at TLN 1962 (3.1.237), they emend "Mene." to "Cor." 51 An excellent conceited Tragedie of Romeo and Iuliet (Qi); The Most Excellent and lamentable Tragedie, of Romeo and Iuliet (Q2); THE TRAGEDIE OF ROMEO and IULIET (Fl, TLN 2680 [4.5.101]). 52 Paradoxically, in the earlier, supposedly "bad" Qi, the fracturing of Lady Capulet is far less radical. In 1.3 we find the speech prefixes "W:," and "Wife:," and thereafter "M:," "Mo:," and "Moth:." See Cloud, "'The very names of the Persons.'" 53 Prolegomena for the Oxford Shakespeare: A Study in Editorial Method (Oxford: Oxford Univ. Press, 1939), pp. 56-57.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 269 added to Woodstock after the text was written (a not unusual theatrical practice), most by the writer of the text (Hand S) but many by two others (Hands A and B)."54 Pope would have been hard put to apply his theory of the distinctness of Shakespeare's characters to the one sample we have of a dramatic dialogue generally accepted as in Shakespeare's hand, a few pages from The Booke of Sir Thomas More, to which we will return. The generic speech prefixes "other," "oth," and "o" are ascribed to Shakespeare; their deletion and replacement by proper names ("Geo bett," "betts low," "william," etc.) is credited to a subsequent hand.55 When dramatic nomenclature is made uniform, the variability of Eliza- bethan and Jacobean theatrical books and manuscripts is phased out. Furthermore, without venturing into the text-performance debate, it may at least be suggested here that this character instability is not limited to the written or printed text. Onstage the doubling of actors might also have destabilized character. As the title of A lamentable tragedy mixed ful of pleasant mirth,. conteyning the life of Cambises confuses generic identity, so, too, its title-page "division of the partes" confuses personal identity: thirty-eight parts are divided among eight actors. The parts for "one man" are "Lord," "Ruf," "Commons Cry," "Commons complaint," "Lord smirdis," and "Ve- nus." A single actor thus plays aristocrat and commoner, proper noun and common noun, male and female.56 If we widen our scope beyond theatrical texts, we find not only the proliferation of names to designate a "single" person but also the opposite: the repetition of a single sign to designate different persons. Ruth Luborsky and Elizabeth Ingram's English Books with Woodcuts, 1536-1603 catalogues the recycling of the same woodcut in illustrations of different historical and fictional figures.57 In the 1577 edition of Holinshed's Chronicles, for instance, 212 cuts are used to make 1,026 images. The same image of a ruler is used to depict Japhet, Cuneday or Margan, Vectius Velanus, and Pertinax. It was reused again in another publication altogether, Caradoc of Llancarfan's The Historie of Cambria (1582), to depict "Trahaern the sonne of Caradoc." Even more puzzling to later bourgeois conceptions of unique identity is the reuse in the 1577 Holinshed of a woodcut familiar as an image of Edward VI to represent James VI of Scotland (the future James I of England).58 Identities that modern critics would distinguish converge in a single mechanically reproduced image; identities they would make uniform split into multiple names. If there is no fixity of character in the shifting prefixes of the quartos and Folio, neither is there fixity of gender, despite the typographic separation of male from female in the seven Folio character lists. But is this separation as absolute as the eighteenth-century dramatis personae suggest when the

54 "'A bed / for woodstock': A Warning for the Unwary," Medieval & Renaissance Drama in England, 2 (1985), 91-118, esp. p. 96. 55 See in Wells and Taylor, Textual Companion, a transcript from the Booke of Sir Thomas More manuscript of the 147 lines thought to be in Shakespeare's hand (pp. 463-67). 56 Thomas Preston, A lamentable tragedy (London, 1585), title page. 57 Scholar Press, forthcoming. 58 Ruth Samson Luborsky, "Connections and Disconnections between Images and Texts: The Case of Secular Tudor Book Illustration," Word and Image, 3 (1987), 74-85. See Raphael Holinshed, The Firste volume of the Chronicles of England, Scotlande, and Irelande, "The Historie of Englande," pp. 1, 20, 66, and 77, and "The Historie of Scotlande," p. 505; and Caradoc of Llancarfan, The Historie of Cambria (London, 1584), p. 112.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 270 SHAKESPEARE QUARTERLY headings "Men" and "Women" precede the lists and a space separates them? If the binaries were that clear cut, would the transvestism of a Ganymede or Cesario be imaginable? Numerous recent studies have described the controvertibility of gender not only in a theater that assigned female roles to males but also in medical and anatomical treatises that attributed homologous genitals to both sexes (and allowed for spontaneous sex changes) as well as in a widespread fascination ranging from the folkloric to the legal with the category (or noncategory) of the hermaphrodite.59 Randall McLeod has recently located a striking instance of this controvertibility in two states of an engraving of James I and his family by Willem van de Passe.60 The second state of the engraving shows alterations made to reflect changes in the royal family, but not as we might expect. Instead of engraving each child anew to denote growth and maturation, the younger children have been moved into the bodies formerly occupied by their elder siblings (thereby themselves making room for new arrivals). The individuated bodies represent not individuals but rather relational places within the family. Even gender subscribes to place, so that a pantalooned figure is now a young princess (Louisa) when it had previously been her elder brother (Mauritius). This same gliding between master and mistress is present in Shake- speare's notoriously shifty "Master Mistris" sonnet but is hardly limited to this sonnet, number 20. In the 1609 quarto the gender of the addressee is unspecified in over four fifths of the sonnets.6' As a result, seventeenth- century readers who transcribed them into manuscript collections or commonplace books were free to represent them as addressed to either a man or a woman.62 In his 1640 collection, Poems: Written by Wil. Shake-speare.

59 On Galenic medicine, see Thomas Laqueur, Making Sex: Body and Genderfrom the Greeks to Freud (Cambridge, Mass.: Harvard Univ. Press, 1990), esp. pp. 63-148; and on theories of gender and the English Renaissance theater, see Stephen Greenblatt, Shakespearean Negotia- tions: The Circulation of Social Energy in Renaissance England (Berkeley: Univ. of California Press, 1988), "Fiction and Friction," pp. 66-93. On male to female transformation, see Patricia Parker, "Gender Ideology, Gender Change: The Case of Marie Germain," Critl, 19 (1993), 337-64. For the implications of that transformation on the stage, see LisaJardine, Still Harping on Daughters: Women and Drama in the Age of Shakespeare (Brighton, U.K.: Harvester, 1983), pp. 9-36; Laura Levine, "Men in Women's Clothing: Anti-theatricality and Effeminization from 1579 to 1642," Criticism, 28 (1986), 121-43; Phyllis Rackin, "Androgyny, Mimesis, and the Marriage of the Boy Heroine on the English Renaissance Stage," PMLA, 102 (1987), 29-41; Stephen Orgel, "Nobody's Perfect, or Why Did the English Stage Take Boys for Women?" SAQ, 88 (1989), 7-29; Jean Howard, "Crossdressing, the Theatre, and Gender Struggle in Early Modern England," SQ, 39 (1988), 418-40; and Marjorie Garber, Vested Interests: Cross- Dressing and Cultural Anxiety (New York and London: Routledge, 1992), pp.21-40 and 118-27. On the Renaissance hermaphrodite, see Ann RosalindJones and Peter Stallybrass, "Fetishizing Gender: Constructing the Hermaphrodite in Renaissance Europe" in Bodyguards: The Cultural Contexts of Gender Ambiguity, Julia Epstein and Kristina Straub, eds. (London and New York: Routledge, 1991), pp. 80-111. 60 Paper delivered at the International Emblem Conference, Glasgow, August 1990. The engraving, entitled Triumphus Iacobi Regis Augustaeque Ipsius Prolis, was made circa 1622-24 and re-engraved after 1625. 61 In his unpublished essay "Imagination," Randall McLeod points out that the sexual identity of the addressee is not specified as male until Sonnet 19 and is unsettled in the very next sonnet. 62 Sonnet 2, for example, is entitled "To one that would die a maid" in five seventeenth- century manuscripts (Gary Taylor, "Some Manuscripts of Shakespeare's Sonnets," Bulletin of theJohn Rylands University Library of Manchester, 68 (1985), 210-46, esp. p. 217). For evidence

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 271

Gent, which contains most of the sonnets, John Benson took this liberty. In one sonnet he converted the masculine pronouns to feminine; in two others, he changed masculine nouns to neutral ones.63 In addition, having arranged the sonnets into groups, he assigned four of the groups titles specifying a female beloved, though they contain only sonnets in which gender is indeterminate in the quarto. Benson's converted collection established a tradition that continued until 1780, when Edmond Malone's an- notated edition established another and even longer-lasting tradition.64 Malone returned to the 1609 quarto but definitively foreclosed the option held out by its unspecified pronouns. Like the eighteenth-century dramatis personae, Malone's note (which first appears in the Preface and is repeated in the text after Sonnet 126) fixed the number, gender, and identity of the text's "characters." It decreed: "To this person [Mr. W.H.] whoever he was, one hundred and twenty-six of the following poems are addressed. The remaining twenty-eight are addressed to a lady."65 In response to the "scandal" of the male beloved specified by Malone, scholars hastened to "fix" the gender of the beloved as continuously female. The notorious 1796 forgeries of William Henry Ireland included a letter from Elizabeth I to Shakespeare "proving" that the sonnets were addressed to her.66 George Chalmers, while conceding that Ireland's documents were forgeries, defended in two lengthy tomes Ireland's identification of Shakespeare's beloved with Elizabeth I. Despite the homophobic hysteria of his own response, Chalmers, in defending Elizabeth I as the beloved, is able to show that the "certainties" of Malone's genderings are unfounded. In support of his eccentric argument, Chalmers rightly notes that "the greatest philologists, and philosophers, of her reign, addressed [Elizabeth] both as a male, and female"; more perti- nently, he argues that the pronoun "his" in the Renaissance was widely used not only for the neuter (as modern scholars have long known) but also for the feminine:

His, her, and him, were frequently confounded: and the personal pronoun, his, was often used in a neutral sense, and in the same manner, him, in those days, often referred to it.... Our grammarians have not, I think, observed, that the pronoun his was, in those days, not only used in a neutral sense, but in afeminine sense.67

that before Malone's 1780 edition most of the sonnets were assumed to be addressed to a mistress, see de Grazia, Shakespeare Verbatim (cited in n. 9, above), p. 155, n. 57. 63 For an explanation of how and why Benson's scant changes generated the commonplace that Benson bowdlerized or "white-washed" the sonnets to protect Shakespeare's reputation, see de Grazia, "The Scandal of Shakespeare's Sonnets," SS, 45 (1993), 35-49. 64 On the reprintings of Benson's 1640 edition up through the eighteenth century (and beyond) and the supplanting of the 1609 quarto, see Hyder Edward Rollins, A New Variorum Edition of Shakespeare: The Sonnets (Philadelphia: J. B. Lippincott, 1944), Vol. 2, pp. 29-36. Malone's edition first appeared in Supplement to the Edition of Shakespeare's Plays Published in 1778 by SamuelJohnson and George Steevens in Two Volumes (London, 1780), Vol. 2. 65 The Plays and Poems of William Shakspeare, 10 vols. (London: Baldwin, 1790), Vol. 10, p. 191. 66Miscellaneous Papers and Legal Instruments under the hand and seal of William Shakspeare (London: Egerton, 1796). 67 An Apologyfor the Believers in the Shakspeare-Papers (London: Egerton, 1797), pp. 53-54, and A Supplemental Apology for the Believers in the Shakspeare-Papers (London: Egerton, 1799), pp. 68-69. On Ireland's and Chalmers's gender-bending, see Stallybrass, "Editing as Cultural

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 272 SHAKESPEARE QUARTERLY

The best evidence for Chalmers's tenuous claim that his and her were interchangeable can be found, ironically, in Malone's own editions of The Winter's Tale, Love's Labor's Lost, and As You Like It, where he carefully notes what he takes to be the use of his for her, sometimes attributing it to the printer's error and sometimes to Shakespeare's habitual practice ("our author frequently takes such licenses").68 Though we might suspect Chalm- ers's motives and evidence, he does demonstrate the degree to which male and female are differentiated by editorial fiat in the "definitive" 1780 dispensation. It is by editorial fiat, too, that the speaker in the sonnets is identified with "William Shakespeare" as opposed to, for example, the generic male lover that Benson assumes in 1640. If Malone seems less peremptory than Benson in his arbitrations, it may be because we are still reading the sonnets under his editorial sway. As recent work on the construction of gender has demonstrated, any strict dispensation may be historically precipitous. Neither heterosexuality nor homosexuality existed as categories in early modern Europe, for the simple reason that neither the standard nor its deviation(s) had yet been prescribed.69 The latter-day incontrovertible male/female binary-re- flected in Malone's division of the sonnets-was not yet in place. In addition to a two-sex model, Thomas Laqueur has documented the Galenic one-sex model that obtained in the Renaissance. In this model women were viewed as incomplete men, though, as Stephen Orgel has recently proposed, at the same time that this model privileges man over woman, it also threatens male identity with the possibility that "we are all, in essence, really women."70 Further, one need only turn to the official grammar book of the period for the suggestion that there were neither one nor two but rather many categories of gender; William Lily's Latin Grammar, the only grammar prescribed for use in schools in Shakespeare's time, classifies nouns into seven genders, including male, female, neuter (neither male nor female), doubtful gender (either male or female), and epicene (both male and female). Simon Forman's use of the word it in his patients' records to refer to both boys and girls might illustrate any of these categories.7' If gender is to function as a precondition for character in the Renaissance, it needs to be

Formation: The Sexing of Shakespeare's Sonnets," Modern Language Quarterly, 54 (1993), 91-103. 68 Vol. 3, p. 230. Gary Taylor has located the source of this confusion in Elizabethan paleography and orthography: "in an Elizabethan secretary hand, terminal s was often almost impossible to distinguish from r, and in contemporary orthography her could be spelled with a medial i; in such circumstances, a 'hir' and a 'his' are materially identical, and can only be differentiated by cultural context" ("Textual and Sexual Criticism: A Crux in The Comedy of Errors," RenD, 19 [1989], 195-225, esp. p. 217 and n. 17). 69 See Michel Foucault, The History of Sexuality, trans. Robert Hurley, 2 vols. (New York: Pantheon, 1978), Vol. 1; and Alan Bray, Homosexuality in Renaissance England (London: Gay Men's Press, 1982). 70 "Call Me Ganymede: Shakespeare's Apprentices and the Representation of Women," unpublished manuscript. See also Levine (cited in n. 59, above). 71 A Shorte Introduction of Grammar (London, 1567), sig. A6. See Elizabeth Pittenger's tren- chant comments on how Renaissance pedagogy resisted the homology between grammatical and "natural" gender in "Dispatch Quickly: The Mechanical Reproduction of Pages," SQ, 42 (1991), 389-408, esp. pp. 400-401. For the observation on Forman's undifferentiated gender pronoun, we are indebted to Barbara Traister; see her Window on Elizabethan London: The Papers of Simon Forman, forthcoming.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 273 plastic enough to accommodate structures that may be both more uniform and more multiform than the male/female binary. From theater to anatomy to grammar: we have certainly ranged beyond the physical confines of the text in order to demonstrate that modern critics have constructed character more in the image of standard dramatis personae lists than of variable speech prefixes and shifting ungendered pronouns. It is a ranging that is itself illustrative of the effects of attending to the material text. Our modern habit of probing the text for an imagined complexity of character may have drawn us away from the more sensible yet no less intricate markings of discursive personal (non)identity. If Jonathan Goldberg is right in arguing that in the Renaissance "there is no notion of human character save as a locus of inscription,"72 we need to make sure that it is Renaissance "charactering" we are analyzing and not modern.

IV. AUTHOR

Our post-Enlightenment critical tradition has imagined the author stand- ing above or beyond the categories thus far considered, generating words, constructing characters, and creating texts that form his collected works. But all the above illustrations lend support to the simple but profound insight that "whatever they may do, authors do not write books."73 Stationers constructed the Folio canon (as well as rival collections); later editors added dramatis personae lists at the beginning of each playtext; compositors composed the Folio's "weyward sisters." And these agents also played a part in producing "Shakespeare". Starting with his name: how did it happen that the eighty-three variant spellings recorded by E. K. Chambers eventually took the single form Shakespeare?74 While the six supposed autographs that have been preserved are disuniform, none separates the two syllables of his last name with an e: "Willm Shaksp," "william Shakspe," "Wm Shakspe," "william Shakspere," "Willm Shakspere," "William Shakspeare."75 Why does the scripted "Shak" become the printed "Shake"? Randall McLeod has shown that when the compositor set k and long s against each other in italic type, the letters would each begin to bend or break, because they both usually kerned (i.e., the typeface of each letter projected beyond the small typebody behind it).76 To

72 "Hamlet's Hand," SQ, 39 (1988), 307-27, esp. p. 316. In "Textual Properties," Goldberg notes: "That in the two Lears different characters may speak the same lines, that the same characters (characters with the same proper names) speak different lines, suggests the radical instability of character as a locus of meaning in the Shakespearean text" (SQ, 37 [1986], 213-17, esp. p. 215). See also Goldberg's Writing Matter: From the Hands of the English Renaissance (Stanford: Stanford Univ. Press, 1990), where, by tracing character to the routine impresses of material graphology, he has done nothing less than overthrow almost three centuries of "character study," including the concept of interiority which "might be no more than the depth of the incision of characters in stone or wax" (p. 314). 73 Roger E. Stoddard, "Morphology and the Book from an American Perspective," Printing History, 17 (1987), 2-14; quoted here from Roger Chartier, "Texts, Printing, Readings" in The New Cultural History, Lynn Hunt, ed. (Berkeley: Univ. of California Press, 1989), pp. 154-75, es. p. 161. Chambers, Vol. 2, p. 371. 75 Chambers, Vol. 2, pp. 504-6. 76 "Spellbound" (cited in the unnumbered acknowledgment note, above), p. 60. For the interchanging of the first and last es in the setting of the two imprints on the title page of the

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 274 SHAKESPEARE QUARTERLY avoid breakage (and the ensuing fine), a compositor would set a neutral typebody between k and long s. The problematic "Shakspeare" would then be set in italics as "Shak-speare," "Shakespeare," or even "Shake-speare." The printer's habit of separating k and long s was frequently retained even when the typeface was roman rather than italic, so that the roman-type title page of the 1608 Lear prints "M. William Shak-speare" and that of the 1609 sonnets "SHAKE-SPEARES".77 The standard spelling of the author's name is not that of the author's hand but that of the printer's press and reflects not a personal investment in the question of identity but rather an economic one in the preservation of typeface.78 Authorship is hardly an authorial construct, for the very form of the author's name is here a printing-house production.79 It is what Jerome McGann has termed our "hypnotic fascination with the isolated author" that has led us to ignore the degree to which the production of a literary work "is a social and institutional event" rather than an individual creation.80 So compelled, we have been strangely blind to the varieties of entitlement registered on the title pages of playtexts: first and invariably, the bookseller (who owned the text); second, the printer (if and when different from the bookseller); third, frequently but not invariably, the acting company who had first performed the play, along with a report of the success with which the performance had met; and finally, the writer.8' The relative insignifi- cance of the author is particularly striking in the case of Shakespeare. Of his plays published before 1600, seven of the eight were first printed anony-

sonnets, see McLeod's "A Technique of Headline Analysis, with Application to Shakespeares Sonnets, 1609," SB, 32 (1979), 197-210. 77 Malone attempts to supplant this "inauthentic" spelling with the supposedly authentic "Shakspeare," but the 1864 Globe edition (following the Cambridge edition) reinstates the Folio spelling and makes it standard, indeed global. 78 On the immanent quality of signatures and their incompatibility with technology, see Goldberg's opening remarks in "Hamlet's Hand." For a Derridean critique of metaphysical graphic investment and an historical alternative, see the first and last chapters of Goldberg, Writing Matter. 79 On the emergence of authorship from printing-house practices, see Joseph F. Loewen- stein's account of the relation between italic typeface and authority/authorship: "One might say that the author, the modern proprietary author, descends from a typeface" ("Idem: italics and the genetics of authorship,"Journal of Medieval and Renaissance Studies, 20 [1990], 205-24, esp. p. 224). De Grazia makes a related claim about the relation between authorship and quotation marks (often interchangeable with italics) specifically in regard to Shakespeare; see "Shake- speare in Quotation Marks" in The Appropriation of Shakespeare: Post-Renaissance Reconstructions of the Works and the Myth, Jean I. Marsden, ed. (New York and London: Harvester Wheatsheaf, 1991), pp. 57-71. For Shakespeare's name as a typographical ornament, see Jeff Masten, "Textual Reproduction: Collaboration, Gender, and Authorship in Renaissance Drama" (Univ. of Pennsylvania Ph.D. diss., 1991), chap. 3. For other kinds of printing-house factors that figured in the construction of authorship, see Leah Marcus's discussion of the First Folio in Puzzling Shakespeare (cited in n. 1, above), pp. 1-32; and de Grazia's in Shakespeare Verbatim, pp. 14-48. See also the following essays by Joseph Loewenstein: "The Script in the Market- place," Representations, 12 (1985), 101-14; "For a History of Intellectual Property: John Wolfe's Reformation," English Literary Renaissance, 18 (1988), 389-412; "Plays Agonistic and Compet- itive" (cited in n. 9, above); "The Archaeology of Miltonic Genius: Bacon andJonson on Pursuit of Property," paper delivered at the English Institute, Harvard University, Boston, August 1989; and his forthcoming The Authorial Impression: Intellectual Property in the English Renaissance. 80 A Critique of Modern Textual Criticism (Chicago: Univ. of Chicago Press, 1983), p. 100. 81 As Marion Trousdale notes, "Dramatic texts in particular ... cannot be reduced to a single author or a single form because they are in origin the collective endeavors described on their title pages" ("A Trip" [cited in n. 17, above], p. 223).

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 275 mously.82 Even after 1600, when the name regularly appears, it is as an advertising device deployed by the stationer; as the epistle to the 1622 quarto of Othello attests, "The Authors name is sufficient to vent his worke." It apparently continued to function as such, judging by the number of non-Folio plays printers continued to assign to Shakespeare up through the Restoration.83 Like the names of characters, the name of the playwright is itself a variable material sign inscribed in books, not a fixed essence that lies imperceptibly behind the text. Not only the orthographic form of the author's name evolves from printing-house factors but also the name's semantic reference. Over twenty years ago G. E. Bentley demonstrated how inattention to stationers' practices leads to misreadings of the author's name on title pages. Comparing title pages and the Stationers' Register entries with Henslowe's diary, Bent- ley noted a discrepancy between attribution in print and actual payment.84 Although a play might be attributed in print to one author, several authors, as we have already noted, were paid for their contributions. The name on the title page does not, therefore, give us full (or even accurate) information about who wrote what, even in the case of well-known plays.85 Its use is determined by the guild founded to regulate and survey all aspects of the book trade. It may be helpful to consider title-page attributions and Statio- ners' Register entries in terms of provenance rather than authorship, in terms of the King's Men rather than Shakespeare.86 Here the acting company's role in prescribing the use of Shakespeare's name also must be acknowledged, for it was, we must assume, the company that attached his name to the plays when they sold manuscripts to the press and ordered playbills advertising performances. The choice of his name may reflect not his authorship, in any traditional sense, but rather his centrality to the company in multiple capacities (as playwright, actor, shareholder), not to

82 Of the first editions of these eight plays, only Love's Labor's Lost was attributed to Shakespeare on the title page. Richard II and Richard III were printed without attribution in 1597, but in 1598 second editions of both plays were published which claimed Shakespeare as author. In the case of Richard II, though, the authorship claim was dropped in the third edition, issued in the same year. 83 See Chambers, Vol. 1, pp. 537-38. 84 pp. 200-206 (cited in n. 11, above). 85 According to Bentley, "even a cautious reading of the documents suggests that as many as half of the plays printed between 1590 and 1642 were written collaboratively, and in Shakespeare's period, if Henslowe's diary can serve as index, nearly two-thirds" (pp. 204-5). While Bentley's conclusions remain uncontested, their implications have hardly begun to register in Shakespeare studies, despite Stephen Orgel's repeated denial of the self-evident assumption that the "authority of a text derives from the author" and his numerous demon- strations of nonauthorial determinations in the theater and printing house, from Shakespeare's time to our own. The benefit of taking Bentley and Orgel seriously can be seen in Scott McMillin's The Elizabethan Theatre and The Book of Sir Thomas More (Ithaca, N.Y., and London: Cornell Univ. Press, 1987), and Jeff Masten's "Beaumont and/(n)or Fletcher: Col- laboration and The Interpretation of Renaissance Drama" (ELH, 59 [1992], 337-56). Masten's forthcoming Textual Intercourse: Collaboration and Authorship in a Homosocial Context stages the contest between single and multiple authorship in the sixteenth and seventeenth centuries and spells out its practical and theoretical implications. 86 It is not insignificant that the injunction against printing the King's Men's plays has always been understood as one against printing Shakespeare's plays, i.e., acting company and author are taken to be synonymous; for a critique of this conflation, see Roslyn Lander Knutson, The Repertory of Shakespeare's Company, 1594-1613 (Fayetteville: Univ. of Arkansas Press, 1991), pp. 1-14.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 276 SHAKESPEARE QUARTERLY mention his distinctive loyalty to that company for which he wrote exclu- sively.87 The authorial name ties the work not to a sole agent or "onlie begetter" but to a productive and reproductive network.88 Nor is this function atypical, as recent discussions of earlier and overlapping modes of ascription indicate. As E. P. Goldschmidt has shown of medieval books, Peter Beal of Renaissance manuscripts, and J. W. Saunders and Arthur Marotti of poetic miscellanies, the names to which texts are ascribed designate a number of productive functions-not only authorship (itself a pro- foundly imitative activity at this period) but also the functions of copyist, reviser, compiler, respondent, setter-to-music, and coterie circle, among others.89 Yet it is the category of solitary and unitary authorship that Shakespear- eans have been most loathe to forgo, as the recent New Textualism demonstrates. Its recognition of multiple texts for single plays has inaugurated what has been repeatedly hailed as a revolution in Shakespeare studies. With the demonstration that the standard Lear text was constructed in the eighteenth century, and its corollary, that other multiple-text plays have been subjected to the same kind of conflation, the venerable tradition of composite texts has been disrupted. After Gary Taylor's series of articles and Grace ioppolo's 1992 Revising Shakespeare, Shakespeareans can no longer take the unity and integrity of their standard Shakespeare for granted. The effect has been positively (in both senses) unsettling, for readers as well as editors. In his multiple-text edition of Lear, Michael Warren urges a "differential reading" of textual details that would trouble precisely the categories this essay has been querying: the fixed signifier, the single text, unified character. Surprisingly, though, the category of author has remained essentially in place, even though it now is imagined less in terms of inspirational flashes than of repeated acts of adjustment to theatrical circumstances. This is the case despite the New Bibliographers' attention in the early decades of this century to the complexity of the material texts and the difficulty of discovering an uncontaminated origin. Yet, as several scholars have recently commented, the New Bibliographers' emphasis on materiality was betrayed from the start by a higher devotion to the authorial holograph, what they referred to as foul papers, a term whose nitty-grittiness belied its purely

87 On Shakespeare's fidelity to a single acting company, see Bentley, p. 279. With one exception, the plays assigned to Beaumont and Fletcher were also written for the King's Men (Bentley, p. 35). 88 For a discussion of the functions of authorship, see Michel Foucault, "What is an Author?" in Language, Counter-Memory, Practice, ed. and trans. Donald F. Bouchard (Ithaca, N.Y.: Cornell Univ. Press, 1977), p. 123. 89 See E. P. Goldschmidt, Medieval Texts and Their First Appearance in Print (London: Printed for the Bibliographical Society, 1943 [for 1940]); J. W. Saunders, "From Manuscript to Print: A Note on the Circulation of Poetic MSS. in the Sixteenth Century," Proceedings of the Leeds Philosophical and Literary Society, 6 (1951), 507-28; Arthur F. Marotti, "Malleable and Fixed Texts: Manuscript and Printed Miscellanies and the Transmission of Lyric Poetry in the English Renaissance" and "Manuscript, Print, and the English Renaissance Lyric," both in New Ways of Looking at Old Texts: Papers of the Renaissance English Text Society 1985-1991, W. Speed-Hill, ed. (Binghamton, N.Y.: Medieval & Renaissance Texts & Studies, 1993), pp. 159-73 and 209-21, and his Manuscript, Print, and the English Renaissance Lyric (Ithaca, N.Y.: Cornell Univ. Press, forthcoming); Peter Beal, Times Literary Supplement, 3 January 1986, p. 13.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 277

imaginary and idealized status.90 With this transcendental signifier, New Bibliography warded off the ravages of two generations of Disintegration- ists, who had, on the basis of scientific metrical analysis, attributed large stretches of the Works to non-Shakespearean hands (earlier writers, collab- orators, revisers, etc.).9' While the Disintegrationists' detection of numerous nonauthorial hands ended up depleting the Folio canon, the New Bibliographers' projection of a single authorial hand preserved it in its entirety from the start. Scott McMillin analyzes a striking example of this sacral devotion to Shakespeare's hand in his fine study of The Booke of Sir Thomas More, a manuscript ascribed to six different hands.92 As the unique sample of the imagined authorial holograph behind the entire canon, the manuscript was dear to New Bibliographers. It was also, however, something of an embar- rassment, for the only putative record of Shakespeare's playwrighting bore witness not to his authorship but to his collaboration. The project then became the identification of Shakespeare's hand and its isolation from the five others, a particular challenge when one of the hands, Hand C, resem- bled Shakespeare's, Hand D, especially at those points when Hand C was "probably copying the work of Hand D."93 As McMillin points out, despite their kinship, Hand D was identified by the New Bibliographers as that of a "dramatist," while Hand C was that of a mere "playhouse functionary," a distinction that upholds a series of other hierarchies-genius/scribe, mind/hand, master/slave.94 This politics, as McMillin discerns, runs through the manuscript's analysis, perceptible in the bias that "gives the privilege of genius to authors and holds those responsible for the material conditions of literature-actors, for example, but also scribes, stationers, and paper manufacturers . . . -as more or less contemptible."95 The Oxford edition deliberately reproduces this bias when it extricates the Hand D pages from the work of other hands and both transcribes and edits them "almost as though they were fragments of a lost work."96 The claim to have found a Shakespearean manuscript in The Booke of Sir Thomas More itself rests upon shaky ground: in particular, the assumption that one can trace in the play the hand of Shakespeare's signature (and the "by me" that precedes one of those signatures) in the 147 lines attributed to Shakespeare. The "authorship" of Shakespeare's signature has indeed

90 See Werstine, "Narratives," (cited in n. 3, above), p. 72, n. 24, and p. 81. 91 Hugh Grady, "Disintegration and Its Reverberations" in The Appropriation of Shakespeare (cited in n. 79, above), pp. 111-27. 92 McMillin's study is cited in n. 85, above. 93 Textual Companion (cited in n. 2, above), p. 461. 94 Compare Michael Bristol's comments on New Bibliography's distinction between the ideal authorial text and the venal workplace of players and printers: "The crucial boundary line is the one that separates the life of the mind from the sphere of manual labor" (Shakespeare's America, America's Shakespeare [London and New York: Routledge, 1990], p. 106). 95 McMillin, p. 154. Similarly, Werstine notes that the desire to find a unified, individual agent behind the text has been "produced by the desire for a certain kind of narrative, one which calls into being certain individuals-solitary author or lone actor-for the purpose of holding them solely responsible for the production of the most diverse phenomena. As critics, we in the twentieth century have long constructed plays as unities in our readings of them; these unities have been secured in textual criticism through the metonymy of constructing unified agents as the origins of the printed texts" ("Narratives," p. 82). 96 Textual Companion, p. 461.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 278 SHAKESPEARE QUARTERLY multiplied under recent scholarly scrutiny. In Shakespeare in the Public Records, Jane Cox writes of the six "authenticated" signatures that survive:

It is obvious at a glance that these signatures [excepting the two that appear on deeds connected with the purchase of the Blackfriars house] are not the signatures of the same man. Almost every letter is formed in a different way in each.... Which of the signatures reproduced here is the genuine article is anybody's guess.

Cox's "at a glance" is itself suspect, since it presumes a standard of uniformity among "authentic" signatures which was by no means the norm in the Renaissance.97 But her argument at least suggests the problem with taking Shakespeare's "authentic" signatures as a secure starting point for the identification of his autograph writing. In addition, her extensive analysis of sixteenth-century wills further complicates the identification, for she observes that "there are hardly any originals" among the signatures. Of the fifty-five wills proved in Prerogative Court in the same month as Shakespeare's will, for example, there are "numerous examples of 'forgeries' of witnesses' signatures." In the case of Shakespeare's will, it is quite possible that

. . . the clerk who wrote the will "forged" the signatures. Until the Statute of Frauds of 1667 there was no necessity for a will to bear the testator's signature at all. Manuals of the period indicate the form preferred by the doctors of civil law, namely that a will should be signed on every page and witnessed, but virtually any form was acceptable so long as it seemed to be a true representation of the dying man's wishes.... The legal sanctity of the signature was not firmly established.... Wills were proved by the executor's oath, nothing more, unless objections were raised by some interested party, in which case witnesses would be examined. It was not until later in the seventeenth century that handwriting experts began to be used by the court.98

The signature of Shakespeare may thus itself be a collaborative field, not the private property of a single individual. The construction of a single autograph from six disparate signatures is not unlike the construction of a single play from multiple texts: in both cases many are subsumed into one. New Textualism has boldly resisted this tradition of collapsing difference. The variant texts-even the "bad quIar- tos"-can now each be recognized as autonomous and deserving of textual and critical attention. So, too, repetitions and anomalies within a text formerly discarded as corruptions can now be preserved as authorial re- writings. But to account for textual disuniformity by appealing to authorial revision may be like resorting to pathology for an explanation of auto- graphic disuniformity: a revising hand or a palsied hand can, it must be

97 "Shakespeare's Will and Signature" in Shakespeare in the Public Record Office, David Thomas, ed. (London: Public Record Office, 1985), p. 33. For a critique of the normative assumption of uniformity, see Goldberg, Writing Matter, pp. 241-43. Peter Blayney has pointed out to us the radical difference between the secretary and italic signatures of Nicholas Okes; see his The Texts of King Lear (cited in n. 12, above), pp. 225 and 229. 98 CoX, p. 34. Cox's research would seem to supply the evidence S. Schoenbaum calls for to support the claim that in the Elizabethan period (and beyond), "a testator's name may represent the signature of another in his stead" (William Shakespeare: Records and Images [New York: Oxford Univ. Press, 1981], p. 98).

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 279

admitted, contain multitudes.99 The recognition of "multiple texts and variant passages is compromised by a theory of revision that ends up unifying and regulating what it had dispersed and loosened: all intertextual and intratextual variants are claimed in the name of a revising Shakespeare. The notion that Shakespeare revised does indeed introduce a major change in how his authorship is perceived: once hailed as a spontaneous writer never stopping to blot his papers, he now is a more deliberate one who revises both passages and texts according to second thoughts, after- thoughts, and new theatrical circumstances. The conversion is not, however, as radical as it appears. One model of Romantic genius has replaced another: the documented practices of a revising Wordsworth or Keats have replaced the fictive fantasy of an impulsive and inspired poet. In order to efface nonauthorial work, Shakespeare is imagined collaborating with himself over time, himself a committee of one.'00 Paul Werstine reminds us of all the other agents that drop out of this picture: "[Shakespeare's] texts were open to penetration and alteration not only by Shakespeare himself and by his fellow actors but also by multiple theatrical and extratheatrical scriven- ers, by theatrical annotators, adapters, and revisers (who might cut or add), by censors, and by compositors and proofreaders."'0' The notion of "Shakespeare the reviser" may thus be used to undercut the very promise it holds out, allowing critics to fall back into post-Enlight- enment categories even as they deny the editorial tradition that instated and sustained those categories. Indeed, it readily lends itself to a Man-and- Works criticism, for each multiple text constitutes a canon in miniature in which the author's personal and artistic development can be charted from revision to revision; the hermeneutic substratum is thereby also extended, each reworking providing additional ground for excavational in-depth reading in "terms always more or less psychological," as Foucault anticipated. 102 Though now possessing the documents that could dispel it, we are in danger of remaining hypnotically fascinated by the isolated author. Nevertheless, the New Textualism has provided us with the means by which we can interrogate that fascination. Gary Taylor, for instance, justi- fies the Oxford Shakespeare's choice of texts as a deliberate opting for those most distant from "the allegedly private text," those which most fully capture the collaborative aspects of drama ("the most socialized of all literary forms").'03 Even more explicitly, Taylor argues in his innovative recent work on Thomas Middleton not only that Middleton "provides us with an inescapably collaborative model of textual production" but also that we now need to start editing Shakespeare in light of Middleton rather than vice versa. 104 We need, in other words, to rethink Shakespeare in relation to our new knowledge of collaborative writing, collaborative printing, and the historical contingencies of textual production.

99 For the various diagnoses of Shakespeare's failing health and unsteady hand at the signing of his will, see Schoenbaum, p. 99. 100 On the cultural resistance to "authorship by committee," see Bristol, pp. 118-19. 101 "Narratives," p. 86. 102 Foucault, "What is an Author?" (cited in n. 88, above), p. 127. 103 Textual Companion, p. 15. See also Jerome McGann, The Textual Condition (Princeton, N.J.: Princeton Univ. Press, 1991). 104 "The Renaissance and the End of Editing" in Palimpsest: Editorial Theory Across the Humanities, George Bornstein, ed. (Ann Arbor: Univ. of Michigan Press, forthcoming in 1993).

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 280 SHAKESPEARE QUARTERLY

V. PAPER

If there is any single obstacle between us and such a project, it is the sense that the value of Shakespeare lies elsewhere, in the inner regions of the text rather than in the practices recorded on its surfaces, in what Pierre Mach- erey has termed the "postulate of depth."'05 And this indeed is the hermeneutic that the standard modern edition encourages, its legibility producing the illusion that Shakespeare can be seen through the text, an illusion Roger Chartier has recently referred to as the "spontaneous and misleading image which readers have of their relation to the text as being a transparent and purely intellectual one."''06 The clean and familiar textual surface allows reading to proceed unencumbered past matter and into the heart of the matter-into Shakespeare's "meaning." The standard edition thereby pro- motes a binarism between surface and depth in which the former leads to the latter. Sustaining this binarism is the division of labor Jerome McGann has referred to as "the schism that characterizes current literary studies.'9107 Editors treat the exterior surface of the text while literary critics probe its interior significance. But if we reject depth as the object of analysis, we will at the same time have to transform our notion of surface. No less than depth, surface is locked into the dichotomy of outer/inner, form/content, appearance/reality. Perhaps a more helpful way of conceptualizing the text is to be found outside metaphysics, in the materials of the physical book itself: in paper. Indeed the crucial quality of paper-its absorbency-eludes the dichotomy. Only because of its absorbency is paper permeable by the black spots of ink. In addition, paper retains the traces of a wide range of labor practices and metamorphoses. In Shakespeare's time paper owed its existence to the rag-pickers who collected the cloth (itself the residue of sheets and clothes) from which it was made. In the sheets of a book, bedsheets began a new life, after the rags had been turned into "stuff" and then into paper by vatmen, couchers, and layers. And it was commonplace for Elizabethan authors to anticipate another use when the pages of the book were returned to serviceable paper-paper to wrap groceries or to light tobacco.'08 The Shakespearean text is thus, like any Renaissance book, a provisional state in the circulation of matter, a circulation that involved an extraordi- nary diversity of labors. As an anonymous eighteenth-century poem puts it:

105 A Theory of Literary Production, trans. Geoffrey Wall (London and New York: Routledge & Kegan Paul, 1985), p. 81. 106 "Meaningful Forms," Liber, 1 (1989), 8-9, esp. p. 8. 107 "The Monks and the Giants: Textual and Bibliographical Studies and the Interpretation of Literary Works" in Textual Criticism and Literary Interpretation, Jerome McGann, ed. (Chicago and London: Univ. of Chicago Press, 1985), p. 187. 108 See Ben Jonson, Discoveries, no. 732, in The Complete Poems, ed. George Parfitt (London: Penguin Books, 1975), p. 392. See also his Epigram III, "To My Bookseller." For the multiple uses of pages, see Thomas Nashe, "The Induction to the Dapper Mounsier Pages of the Court," The Unfortunate Traveller in The Works of Thomas Nashe, ed. Ronald B. McKerrow, 5 vols. (London: Sedgwick and Jackson, 1910), Vol. 3, p. 207. In his reading of this narrative as "an informal phenomenology of the page," Jonathan Crewe notes that Nashe is the first (according to the OED) to use "page" as printed sheet (Unredeemed Rhetoric: Thomas Nashe and the Scandal of Authorship [Baltimore: Johns Hopkins Univ. Press, 1982], p. 69).

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 281

RAGS make paper, PAPER makes money, MONEY makes banks, BANKS make loans, LOANS make beggars, BEGGARS make RAGS.109

And the rags which quite literally composed the works of the National Bard were themselves the heterogeneous products of an international capitalist industry. The rags were vagrant, collected from all over Europe to be processed in the great papermaking industries of France and Italy. The lack of paper in England, due at first, no doubt, to the absence of any significant linen industry, meant that the Shakespearean text (like the vast majority of other English Renaissance texts) was a "foreign" body."10 It was also, in William Prynne's eye, a luxurious body. "Shackspeers Plaies," he com- plained of the Second Folio, "are printed in the best Crowne paper, far better than most Bibles.""' Other, less luxurious "foreign" bodies were used in the printshop: the words of what was to become a classic text were printed in an ink that mingled not only ingredients like juniper gum, linseed oil, and lampblack but also the residual traces of the urine of the

109 Quoted as the epigraph to Dard Hunter's Papermaking: The History and Technique of an Ancient Craft, 2d ed., rev. and enl. (New York: Knopf, 1947). Sixteenth- and seventeenth- century protests against the exporting of rags bear witness to the centrality of rags in the papermaking process. See the petitions by Richard Tottly (1585) and Francis Windebank (1640) quoted in W. W. Greg, A Companion to Arber (Oxford: Clarendon Press, 1967), pp. 38 and 354. 110 On the making and importation of paper, see D. C. Coleman, The British Paper Industry 1495-1860: A Study in Industrial Growth (Oxford: Oxford Univ. Press, 1958). Coleman notes that the protection of the woolen industry, as well as climatic problems, hindered the development of a linen industry in England. The great majority of paper used in English books in the late sixteenth and early seventeenth centuries came from France (pp. 17-18). On the relation between the production of linen and paper in Germantown, Richard Frame wrote in his A Short Description of Pennsylvania (Philadelphia, 1692):

One Trade brings in imployment for another, So that we may suppose each Trade a Brother; From Linnin Rags good Paper doth derive, The first Trade keeps the second Trade alive ... So that the Flax, which first springs from the Land, First Flax, then Yarn, and then they must begin, To weave the same, which they took pains to spin. Also, when on our backs it is well worn, Some of the same remains Ragged and Torn; Then of these Rags our Paper it is made....

The common eighteenth- and early nineteenth-century American watermark "SAVE RAGS" graphically recalls the connection between the rag-picker and reading. See Dard Hunter, Papermaking in Pioneer America (Philadelphia: Univ. of Pennsylvania Press, 1952), pp. 24 and 18. 1 l l William Prynne, Histrio-Mastix (London, 1633), unpaginated preface, quoted as referring to the First Folio by William Jaggard, Shakespeare Bibliography (New York: Frederick Ungar, 1959), p. 495; and by Marcus (cited in n. 1, above), p. 2. Peter Blayney has pointed out to us in conversation, however, that it was the Second Folio that was printed on expensive crown paper.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms 282 SHAKESPEARE QUARTERLY printshop workers, who each night used urine to soak the leather casing of the balls that inked the press." 2 It is these material practices that, even when noted, are ignored in favor of a transcendent "text" imagined as the product of the author's mind. There is a further consequence of identifying the text's abstraction with the author's mind: the erasure of the collaborative process of editing itself. Malone consolidated a tradition in which editor hides behind author. The presence of the editor is everywhere; but everywhere it is relegated to the status of the merely secondary. As Stephen Orgel has demonstrated, the very notion of producing the "authentic" Shakespeare has been a major stimulus to the proliferation of differentiated Shakespeares. Gary Taylor's Reinventing Shakespeare provides a cultural history of just this proliferation."l3 And Barbara Mowat's cogent discussion of Hamlet illustrates the paradoxical relation between the pursuit of authenticity and textual proliferation:

Driven by a desire to recover the Hamlet that Shakespeare wrote-to replace passages lost through actors' excisions or printers' errors, to restore words garbled by interfering players or incompetent printers, to take Hamlet out of the hands of previous editors and print Shakespeare's own Hamlet (either his original text or his final text, but in any event his text), editors from Rowe to Hibbard have searched for Shakespeare's words, and the result has been almost as many Hamlets as there have been editors. 114

In presenting the "true" Shakespeare, editors have effaced editing itself as a form of (collaborative) production. New formats, inclusions and exclu- sions, orderings, scholarly apparatuses, glosses, and orthographies construct new Shakespeares. And why not? If, as we have argued, there is no "original," then later editions cannot be accused of a falling off and away, for there is no fixed point from which such falling could be measured. Or rather, they can be accused of such a decline only if, instead of accepting the pressures of time and the inevitable social and political stakes of editing, they claim to be the true Shakespearean text while laboring in the production of something else. There is no intrinsic reason not to have a modernized, translated, rewritten "Shakespeare." In an important sense, that is all we can have, because the material signs of early modern quartos and folios will themselves necessarily mean differently when read within new systems of textual production. One of the strengths of Stephen Booth's edition of the sonnets is its visual display of the space between two different modes of textual inscription: the 1609 quarto and the modernized "transcription."''15 The gutter separating

112 See George Walton Williams, The Craft of Printing and the Publication of Shakespeare's Works (Washington, D.C.: Folger Books, 1985), p. 48; see also Colin H. Bloy, A History of Printing Ink, Balls, and Rollers 1440-1850 (London: Adams and Mackay, 1967), p. 53. 113 Orgel, "The Authentic Shakespeare" (cited in n. 3, above); and Taylor, Reinventing Shakespeare: A Cultural History, From the Restoration to the Present (New York: Weidenfeld and Nicolson, 1989). 114 "The Form of Hamlet's Fortunes" (cited in n. 8, above), p. 117. 115 Shakespeare's Sonnets (New Haven: Yale Univ. Press, 1977). For a critique of Booth's implicit assumption that his "transcription" reproduces the quarto, see John Barrell, Poetry, Language and Politics (Manchester: Manchester Univ. Press, 1988), "Editing Out: The Dis- course of Patronage in Shakespeare's Twenty-Ninth Sonnet," pp. 18-43.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms THE MATERIALITY OF THE SHAKESPEAREAN TEXT 283 the two texts introduces a division in printer's alphabets, in the use of the apostrophe (and, consequently, in the material marking of possession), in punctuation, in orthography, in grammar, in the very mechanics of type- setting. The collected facsimiles of Warren's The Complete King Lear open up a vastly wider range of textual possibility within the seventeenth century itself, both among different printings and among different formes of the same printing. Its three unbound Lears (Q 1, Q2, and F 1), each succeeded by either uncorrected or corrected pages, allows, indeed coaxes, the reader to assemble any number and combination of pages. Perhaps we should imagine ourselves critically positioned in this great bibliographic divide-whether stitched like Booth's gutter or unstitched like Warren's loose pages-in the space of historical difference. It might take our minds off the solitary genius immanent in the text and removed from the means of mechanical and theatrical reproduction. This genius is, after all, an impoverished, ghostly thing compared to the complex social practices that shaped, and still shape, the absorbent surface of the Shakespearean text. Perhaps it is these practices that should be the objects not only of our labors but also of our desires.

This content downloaded from 128.227.131.182 on Sat, 29 Jun 2019 21:12:54 UTC All use subject to https://about.jstor.org/terms