University of Virginia Library

FORM AND FUNCTION IN THE ENGLISH EIGHTEENTH-CENTURY LITERARY EDITION: THE CASE OF EDWARD CAPELL
by
Marcus Walsh [*]

The forms of eighteenth-century editions of English literary texts were at least as diverse as the forms taken by literary editions in our own time, varying as they did not only with the different markets at which particular editions were aimed, but also with the naturally more fluid state of a scholarly genre in a relatively early phase of its historical development. Small-format editions for instance, aimed at general use, or theatre texts meant to be pocketed by a playhouse audience, included minimal scholarly paratexts. Early editions of Shakespeare, prepared by poets—Nicholas Rowe (1709) and Alexander Pope (1723-25)—for a general and genteel readership similarly provided sparse textual and explanatory matter. As vernacular literary editing became more and more a domain of historical scholarship, or `philology', the poetic text was increasingly accompanied by a far more elaborated apparatus, as in Zachary Grey's edition of Hudibras (1744), Thomas Newton's variorum Paradise Lost (1749), or Thomas Tyrwhitt's Chaucer (1775). The disposition and relations of that apparatus, as now, were variable. In Tyrwhitt's Chaucer, or in John Upton's 1758 Spenser, elaborated scholarly annotation was printed separately from the `clean' text. In Grey's Hudibras or Newton's Milton the text was accompanied by substantial bodies of footnotes. In Richard Bentley's 1732 edition of Paradise Lost the text page becomes a complex and interrelating pattern of various components, the text attended by substantial and typographically insistent footnotes and a battery of marginal alternative readings, and itself broken up by italics and square brackets— `hooks'—drawing attention to what Bentley thought dubious readings.

In the editing of the plays of William Shakespeare, however, an essentially `standard' format for the scholarly edition was adopted and evolved over the course of the century. Lewis Theobald, arguably the first `critical'— to use his own word, `intelligent'—editor of Shakespeare, also led the way in providing a formal model, deliberately choosing (as he told his then friend William Warburton) `to follow the form of Bentley's Amsterdam

Horace, in subjoining the notes to the place controverted.'[1] Theobald's example of providing the text with a running accompaniment of explanatory and textual footnotes was followed, with a generally rising ratio of notes to text, and more or less steadily increasing scholarly paratexts, in the editions of Pope and Warburton (1747), Samuel Johnson (1765), Johnson and Steevens (1773 and 1778), Reed (1785), Malone (1790), and Steevens (1793).

In this paper I shall be discussing the highly individual editorial work of Edward Capell, which formally is strikingly distinct from this recognizable main stream. Capell is now best remembered for his edition of Shakespeare, ten octavo volumes of the text of the plays, published in 1768, followed by a First Part of Notes and Various Readings in 1774, and the complete threevolume Notes and Various Readings of 1779 through 1783. I shall refer also to Capell's only other piece of published editing, the Prolusions . . . of Antient Poetry (1760); to Capell's holograph copy for the Shakespeare edition; and to Capell's unpublished holograph Milton, which survives with the Shakespeare holograph in the Wren Library of Trinity College, Cambridge.

Both the text and the subsequent apparatus of Capell's Shakespeare edition are innovative and distinctive. As we shall see, the Shakespearean text was published, in 1768, in a strikingly, even provocatively, `clean' form. Capell's textual practice, abandoning the textus receptus and returning to the `original copies', displayed what is arguably a radical text-critical modernity. Capell's explanatory notes, published many years after the text, and very different in book form and mise en page, are based on a daunting range and depth of contextualizing knowledge of Elizabethan and Jacobean culture and literature rivalled, amongst previous editors of Shakespeare, only by Theobald and Johnson.

Capell's hermeneutic procedures and text-editorial methodologies have been both discussed and appreciated by such scholars as Alice Walker and Brian Vickers.[2] Here I shall examine rather the `bibliographical codes' of Capell's works of English editorial scholarship: matters of volume format and makeup, the relations among texts and paratexts, styling and mise en page. In particular I shall argue that Capell's editorial publications are deliberately designed and carefully produced, reflecting not only the expertise and craftsmanship of their printer, but also the intentions of Capell himself. Both the Shakespeare and the Prolusions display Capell's thorough and thoughtful attention to form and to the effect of form on function. Both books show the clearest evidence of Capell's highly distinctive view of the editor's role, of typographic layout, of the reader's use of the text, and of the relation of the editor both to his author and to his reader.

Illustrations

In 1760 Capell published his Prolusions. Perhaps no other English literary edition of the eighteenth century—certainly none before George Steevens's Twenty of the Plays of Shakespeare (1766)—is so determinedly a manifesto of editorial purpose. The full title of the work (plate 1) makes explicit Capell's claimed design: Prolusions; or, select Pieces of antient Poetry,— compil'd with great Care from their several Originals, and offer'd to the Publick as Specimens of the Integrity that should be found in the Editions of worthy Authors. The Virgilian epigraph reinforces the point:

Here Meliboeus fears that his cherished lands are in danger of passing to a godless soldiery; by analogy, Capell invites us to think of English literature as (in Dryden's translating phrase) `happy fields', which must be preserved from the hands of the barbarous, and cultivated with care and honesty.[3] The book is made up of editions of The Nutbrown Maid; Thomas Sackville's Induction; Sir Thomas Overbury's A Wife, now a Widowe; the pseudo-Shakespearean Edward the Third; and Sir John Davies's Nosce Teipsum. The Dedication reiterates Capell's intention to offer `an Example of Care and Fidelity to Persons who take upon them the Publication of our best Authors'. That `Care and Fidelity' is certainly instanced in Capell's text-editorial procedure as he explains it in his Preface: the selection, on the basis of collation, of one original authoritative text as base text or `ground-work'; its emendation either by collation with other copies or by rational conjecture; and the faithful reporting of all variant readings.[4] But the care and fidelity appropriate to the reprinting of the texts of the English literary tradition is instanced too in the make-up, format, and printing of the Prolusions. The volume was printed `for J. and R. Tonson in the Strand', by Dryden Leach, `who may be styled', as John Nichols was to put it in his Literary Anecdotes (2.453), `the Father of Fine Printing in this Country'; the colophon reads, with a proper pride and unusual specificity, `From the Press of Dryden Leach, in Crane-court, Fleetstreet. Oct. 6th. 1759'. The book is a small, carefully-printed octavo. The page measures 6.875 x 4.5 inches (175 x 114 mm.).[5] The type is a distinctive small Caslon, light-faced and beautifully cut, cleanly inked and impressed on a wove paper (to the significance of which I shall return). The text page is neat and elegant, with relatively generous margins, particularly at the foot, and minimum decoration; the opening page of

Edward III (plate 2) bears a rectangular rule frame at the head, and a plain rule under the title, but other pages have no rules or type ornaments at all. Most pages have a single line of textual notes; Capell was keen to preserve what he calls `the beauty of his page' by printing the remainder of the textual variants in separated lists rather than as footnotes. There are no explanatory notes on the page, and no line numbers. There is no catch-word. There are some curiously old-fashioned features of style, including the small black-letter used for some of the textual footnotes, but overall this is a strikingly `modern' piece of printing.

The 1768 Shakespeare is similar in format and style. It is made up of ten small octavo volumes whose pages measure 7 x 4.5 inches (178 x 114 mm.).[6] The most immediately striking feature of the title page (plate 3) of the printed work is the title itself: `Mr WILLIAM SHAKESPEARE / his / COMEDIES, HISTORIES, and TRAGEDIES.' Other eighteenth-century editions of Shakespeare named themselves in a familiar and standardised way: `The Works of Mr. William Shakespeare' (Rowe), `The Works of Shakespear' (Pope, Hanmer, Warburton), `The Works of Shakespeare' (Theobald), `The Plays of William Shakespeare' (Johnson), `The Plays of William Shakspeare' (Johnson and Steevens, Reed), `The Plays and Poems of William Shakspeare' (Malone). (The most consequential disagreement in these forms of title, notoriously, is in the spelling of Shakespeare's name.) Capell's form of title is highly unusual in its own time, but it is of course a version of the title of the first folio: `Mr. WILLIAM / SHAKESPEARES / COMEDIES, / HISTORIES, & / TRAGEDIES.' To revert to the 1623 folio title is at once to make a claim of genuineness and originality, to assert the authenticity of the text presented in these volumes. Unlike the conventional eighteenthcentury title—`The Works of Shakespeare'—Capell's chosen title gives the author primacy, putting Shakespeare's name at the head of the page. Indeed, Capell's form of words, `Mr WILLIAM SHAKESPEARE / his / COMEDIES, . . .', uses an older form of the possessive than Heminge and Condell had done a century and a half before, allowing Capell to present Shakespeare's name untrammelled by inflection. The name of Capell as editor appears nowhere on the title page, an act of self-effacement which aligns him with such another gentleman-editor as Charles Jennens. These are strategies which give Shakespeare an uncontested authority. In other eighteenth-century editions the word `Works' or `Plays' is almost always given the largest type, and the name of the editor or editors almost always appears, with lesser or greater prominence. One might contrast with Capell's, for instance, the title pages of Johnson's 1765 Shakespeare (plate 4), or the 1778 Johnson and Steevens Shakespeare. In both of these the most prominence is given to the word PLAYS, set on its own line in capitals 6 mm high. In both these title pages, Shakespeare's name is allowed rather smaller capitals, 3.5 mm high. In Johnson's 1765 Shakespeare the editor's name appears but is permitted only decently small capitals, a little over 2 mm high. In 1778, less modestly, and more closely approaching the typographical status allowed to Shakespeare

himself, the names of Johnson and Steevens are printed in capitals 3 mm high.

Capell's title page is not on the face of it less cluttered than those of other Shakespearean editions. Following the title itself, Capell provides a description of his book, which insists once more on its textual authenticity—`faithfully republish'd' from the original quartos and folios—and sets out what the reader may expect to find, including the Notes and Various Readings, promised for `some other Volumes'. (Such a description was of course necessary in Capell's time, when the title page doubled as a blurb, set up as advertisement in the bookseller's shop or in the street.) There is, as so often in the eighteenth-century vernacular literary edition, a classical motto, asserting in rather familiar fashion the superhuman genius of Shakespeare (`Qui genus humanum ingenio superavit' [`Who surpassed the race of man in understanding'], Lucretius, De rerum natura, 3. 1056). And there is a simple imprint. Nevertheless, this is an unpretentious and well-proportioned page. There are neither rules nor (unlike the pompous page of Sir Thomas Hanmer's Oxford Shakespeare) ornaments. If Capell's does not have the simplicity of early Foulis title pages—the 1743-45 edition of Shaftesbury's Characteristicks, for instance, or the 1756 Horace Opera—or the monumental severity of Baskerville's 1757 Virgil Opera, it nonetheless appears strikingly light and well-balanced. The balance is aided by the clear separation and spacing of the elements of the title—author's name, the different genres of the works, the account of the editorial enterprise in this volume and in the promised Notes and Various Readings, the motto, the imprint. The lightness is substantially effected by the very extensive use of italics, in contrast with the predominant roman upper case of Johnson's 1765 edition, or the 1778 Johnson / Steevens edition, or, more starkly, with the lapidary mass and precision of Baskerville's large rounded roman capitals.

The effect achieved in this title page—within the space and conventions available clean, elegant, classicising—is substantially as intended by Edward Capell as editor of these volumes, as we can see from comparing it with the title page as it appears in Capell's holograph manuscript, now in the Wren Library at Trinity College (plate 5). Dryden Leach as printer of course must have had a substantial input, in terms of choice of fount and other matters, and Capell and Leach presumably had opportunities to consult before and after their previous work together on the Prolusions. Nevertheless, the typography and layout of the printed page, its line-breaks, punctuation, spacing, capitalisation, and italicisation, are exactly as set out in the fair copy provided by Capell. The only detail that differs is the imprint, where the absence of the name of printer or publisher—the manuscript supplies a mere `&c.'—might suggest that Capell had deliberately designed the page before he knew for certain who his printer and publisher would be. At least as probably, however, those choices were already made, and Capell was simply leaving Leach and Tonson free to specify their preferred exact wording.

So close a match between the author's manuscript and the printed title page is remarkable. It was not of course unusual for authors to set out their

own title pages, as James McLaverty has demonstrated in his seminal article on the subject, `Questions of Entitlement'.[7] After 1728, according to McLaverty, Alexander Pope probably designed his own title pages. He was capable of calligraphic imitation of type-headings, as in the manuscript of An Essay on Criticism. [8] McLaverty shows, however, using the theoretical example of a comment by Moxon in Mechanick Exercises, and the practical instance of William Wake's Exposition of the Doctrine of the Church of England, that while it was common for authors to draft carefully-designed title pages, it was normal for compositors to make alterations so as to `order [the author's] Work the better'. Capell seems to have needed, or tolerated, no such ordering at all.

Capell's prescription of the codicography and typography of his book, however, goes far beyond the title page. The text-page of the printed Shakespeare is distinctive, and in many respects similar in styling to that of the Prolusions (plate 6). Again, the format is a small octavo. There are particularly generous margins, almost an inch at the top and right, and almost an inch at the foot; the page has the proportions, and aspires to the elegance, of a larger format work. The typeface is a small Caslon, slightly more heavily cut than that used in the Prolusions (there is, incidentally, a Leach/Caslon connection; Dryden Leach printed, in 1764, A specimen of printing types by W. Caslon and Son). There are no catch-words, no line numbers, and no ornaments apart from straight rules, and the plain rectangular rule frame used in the headings of poems in the Prolusions recurs, as on this first page of Hamlet. As in the Prolusions there is a single line of textual notes at the foot of the page (plate 7), but no explanatory notes on the page, and as in the Prolusions this is announced as deliberate policy. In the Introduction to the 1768 Shakespeare Capell excuses himself, with a politely contemptuous irony, for not including the `accustom'd and laudable garniture of Notes, Glossaries,' etc. He explains this decision partly on the grounds of the difficulty of distributing notes amongst the volumes; partly on the grounds of his intention to provide notes quite separately; and partly, and significantly, because `a very great part of the world, amongst whom is the editor himself, profess much dislike to this paginary intermixture of text and comment' (p. 30).

There is striking further evidence of the extent to which Capell's textpage, like the title page, embodies his intentions. Most editors at this time gave their printers a more-or-less heavily marked up and interlineated copy of a previous text (Theobald, for instance, sent to the printer a heavilymarked copy of the 1728 second edition of Pope's Shakespeare).[9] Capell however provided a remarkably neat holograph fair copy, embodying his own rather than a traditionary text, with virtually every detail of layout and

typography made explicit, and hardly a correction in sight (plate 8). Comparing the published text with Capell's surviving manuscript demonstrates that Dryden Leach followed his lead closely in matters of layout, rules, spacing, capitalisation, and italicisation. There are three differences between Capell's manuscript and the printed book. Two are slight: the abbreviated names of characters, marked for caps and small caps in the manuscript, are italicised in the printed book, and the catch-words of Capell's manuscript disappear in print. The third change is more substantial and significant, purifying the page and certainly made according to Capell's intention. Capell's holograph transcript includes marginal initials identifying the provenance of readings derived from other moderns: `R.' for Rowe, `P.' for Pope, `T.' for Theobald, and so on. They are dropped in the printed text, he explains in his Introduction, because they are `unsightly', as well as because `his only object has been to do service to his great Author' (1.24n.). This is a privileging of text over gloss, the restored scripture of the author over the intrusions of the editor's secondary responsibility.[10] The correspondence between Capell's transcript and the printed book is otherwise extremely exact. Remarkably, it extends as far as the number of lines per page, and page divisions, which, in printed text and manuscript coincide exactly up to page 32. In both the text occupies 136 pages. Capell's manuscript is written on a very good quality paper with the line divisions pricked out exactly in the right hand margin, ready for use. Each page of the manuscript, and of the book, affords 32 lines to the poetic text. Interestingly enough the manuscript of Capell's neverpublished edition of Paradise Lost also has 32 lines per page.[11]

Two further characteristics of the Capellian page style contribute to the production of a clean and uncluttered, a classically simple, textual effect. Firstly, capitalisation is light; Capell's manuscripts and published work are in the vanguard of the process by which capitalisation was purged from the literary text. In the manuscript and printed texts of Capell's Shakespeare only proper names, and the first word of each line of poetry, have initial capitals. In the Trinity College manuscript of Capell's edition of Paradise Lost there is no initial capital for the first word of each line, except for the first word of each verse paragraph. The same policy is followed in all of the texts (except Edward III) printed in Capell's Prolusions (plate 9); it seems reasonable to assume that this was a feature of the manuscript for the Prolusions (which is not known to be extant). Capell's use of italicisation is more problematic. Decades before Capell's Prolusions or Shakespeare, such authors as Alexander Pope (David Foxon has argued)[12] were already dropping the use of italics for proper names, partly in the search for a more classical textual simplicity, partly because they felt such typographic emphases assumed an unsophisticated audience. Capell's texts, however, both printed and manuscript, regularly

italicise names. In an equally old-fashioned manner, Capell uses italics, not quotation marks, to identify quotation and direct speech. In the manuscript of Paradise Lost direct speech is always italicised; the lengthy speeches of Raphael, for instance, appear on the page as solid blocks of italic. The same styling is used in the Prolusions, notably in The notbrowne Mayde (plate 10). This may be an old-fashioned usage, but the effect is purifying and classical; instead of solid roman type disfigured by heavy double quotation marks at the start of every verse line, we have whole pages of italic print, elegant of line and light in weight, recalling and possibly meant to recall the humanist italic pages of Aldus Manutius.

A final and it seems to me significant feature of the Capellian text is the paper on which it is printed. Both the 1760 Prolusions and the 1768 text of Shakespeare were printed on wove paper—papier vélin. Unlike laid paper, with its characteristic chain and laid lines produced by the mould, wove is a smooth-surfaced paper made on fine mesh screens woven from high-quality brass wire; a process originated by James Whatman the Elder in the 1750s. Wove paper was probably first used in Baskerville's 1757 Virgil, whose volumes are made up of a mixture of sheets of laid and wove paper, all subjected to Baskerville's glazing process. Wove paper production was new and experimental in 1757, and according to John Balston, in the major study of James Whatman, it was not until 1759, and in particular for the quarto Paradise Regained of that year, that Whatman developed the process to a level which satisfied his own criteria.[13] Wove paper did not enter general use for another three decades. Though the Foulises in Glasgow used unusually fine papers, and indeed used Whatman papers from 1747, they did not turn to wove papers until this became the industry norm in the 1790s.[14] Wove paper was used only in a very small minority of the publications of Baskerville, or of Dryden Leach. Leach for instance resorts to wove paper for such aristocratic special printing as An Elegy on the Death of Lady Boynton, a 5page quarto pamphlet published in 1768, but not for more common books.

The stipulation of wove paper for both the Prolusions and the Shakespeare therefore seems entirely remarkable in 1760 and 1768, and seems in all likelihood to have been proposed by Capell. No doubt Capell, and Leach, valued the quality of Whatman's new wove paper itself, and hoped by its use to enhance the perceived quality of their printed text. Perhaps also, for the same more technical reason that Baskerville had Whatman develop new papers, including a wove paper, to use with his new lighter type, Leach and Capell thought it would help in the even impression of the relatively fine-cut Caslon type used for the Prolusions and Shakespeare. If so, the choice of wove paper may be thought of as a condition of the typography of these volumes.

I have attempted to argue that these two major editorial enterprises by Edward Capell are distinctive in many aspects of their form, and that their distinctiveness can be shown in important ways to stem from Capell himself. It seems highly unlikely that the publishing house of Tonson had any part in the page or volume design of Capell's 1768 Shakespeare (equally, the Tonsons could scarcely have contemplated publishing an edition of Shakespeare which competed directly, in form or market, with the edition by Samuel Johnson in whose publication they had shared fewer than three years earlier). It is certainly true that the printer Dryden Leach was a willing and expert co-adjutor, responsible of course for the practical execution, and the choices involved in the execution, of Capell's wishes. No doubt many aspects of the two physical books I have been considering are the result of consultation and cooperation between Capell and Leach. More than that, it has been convincingly argued that Leach was in the forefront of the move towards a simpler, neater typography in the 1750s and 1760s. In his extraordinarily suggestive article on the `Revolution in the Layout of Books in the Eighteenth Century', Nicolas Barker contrasts a page of the 1754 Act establishing the British Museum, with its large and blocky type, heavy capitalisation and cluttered marginal notes, with the Statutes and Rules Relating to the Inspection and Use of the British Museum, in a small neat type, without obtrusive capitals or marginalia, printed by Dryden Leach in 1768—the same year as he printed Capell's Shakespeare.[15] Leach had both the aesthetic sympathies and the craft skills to print the books Capell required. Nonetheless, the typographical features of Capell's two books, while they conform very exactly to Capell's holograph, do not coincide with a regular Leachian house style. For example, Christopher Smart's subscription-published volume of 1765, containing his Song to David, Hymns and Spiritual Songs, and Translation of the Psalms, was printed by Leach. It is not especially `modern' in style, with its heavily roman title page, and double-quotation marks used throughout at the beginning of every quoted line within the verse.

Whatever the nature of the collaboration of Leach and Capell, and whatever their relative input, it seems clear that the style of the Prolusions and Shakespeare are part of that revolution of typographical layout in the mideighteenth century which is usually taken to originate with the work of Andrew and Robert Foulis in Glasgow from 1742 onwards. As Nicolas Barker has described it, the main characteristic of the new style led by the brothers Foulis was neatness, and its main vehicle the printing of the classics. The Foulis style certainly influenced Baskerville and Leach, and seems also to have influenced Capell. It is easy to rehearse a list of features that Capell's printed editions have in common with at least the smaller format reprints of the classics that emerged from the Foulis press in the 1740s and 1750s: a

general neatness and plainness; lightness of type; small type sizes; a virtual absence of type-figures; absence of catch-words; minimal capitalization; the use of fine papers; the use of small paper sizes.[16]

The difference between the page of Foulis and Capell, on the one hand, and the larger, busier and closer-printed variorum pages of Samuel Johnson's 1765 Shakespeare or the Johnson-Steevens editions of 1773 and 1778, is obvious. Why was Edward Capell at so much pains to produce, in 1768, an edition in page format and physical properties so much at odds with what had by then become a significant norm in Shakespearean editing? Why was he so concerned to separate from his critical text of Shakespeare the detailed and copious textual, explanatory and illustrative materials he was already preparing, and in which indeed he had an enormous intellectual and personal investment? Capell's `clean', unmixed, unannotated text displays many of the characteristics, and some of the impulses, of humanist printing. Early printed editions of classical authors, of Horace for example, mimicked the densely populated page of the medieval manuscript, in which accreting commentaries surrounded the text. As many scholars have pointed out, this structure was challenged by the humanist scholar printers—Aldus Manutius, notably— who sought to strip away the accumulations of diverse commentary, and restore the intended authorial text, as they thought it, to its central and privileged place, uncluttered and uncontested by rival voices. This humanist tendency in printing was revived by Foulis and by Baskerville as a vehicle for (predominantly) the poetic works of classical antiquity (plate 11). E. J. Kenney suggests, of the Baskerville editions of classical authors, that they

Capell's `clean', unmixed text, courteously refraining from imposing explicatory annotation upon the reader, insisting on the primacy of the author and the authenticity of the text, belongs to a similar humanist tradition. Fine in physical quality, beautifully printed, clear on the page, it was no doubt aimed at least in part at a broadly similar market. It is designed not for the shelf or the library desk, but for the pocket—the moderately moneyed pocket —and familiar reading. It is the only text-critical edition of Shakespeare which, in its first-published format, might have been carried conveniently to, and used by a spectator at, the playhouse. Capell was a scholar, and substantial parts of his Shakespearean scholarship, in the form of his Notes and Various Readings, would follow years later in, significantly, a much larger format, and with an altogether different page style. In 1768, however, his priority was to produce a text of Shakespeare's plays which would visibly and functionally represent the dramatist as a classic of vernacular literature, a classic which might be read in the same way, and in recognisably the same

formats, as the classics of Greek and Roman literature had come to be read.

However different the form of Capell's 1768 and Samuel Johnson's 1765 Shakespeares, it is nonetheless possible to compare Capell's principle here with Samuel Johnson's stated belief that `Notes are often necessary, but they are necessary evils'. In his Preface Johnson famously urges the newcomer to Shakespeare to read straight forward, to

Only after this initial, unmediated and uninterrupted experience is the reader advised to `attempt exactness, and read the commentators'. Johnson here recommends an initial aesthetic response to the work in its entirety, very much comparable not only with Pope's generally anti-philological position, but also, and more specifically, with Pope's humanist distinction, in the Aristarchus passage of the Fourth Book of the Dunciad, between fragments and the glorious whole; or with his insistence that `men of a right Understanding generally see at once all that an Author can reasonably mean'.[19] Johnson allows the necessity of notes, and even the value of the formidable apparatus of variorum commentary. And clearly the annotations in his own 1765 edition are one of the major contributions to the philological side of the argument, the scholarly investigation of Shakespearean readings and meanings. Yet notes are still in his view an intrusion, explaining the crux but failing to interpret the whole work: `particular passages are cleared by notes, but the general effect of the work is weakened'. From the hermeneutic point of view, Johnson's emphasis on the local and limited interpretative possibilities of the note might put us in mind of some post-modern attacks on annotation as traditionally conceived: for instance, Anne Middleton's characterization of explanatory notes as `scattered raids upon local meaning', which cannot finally explain their whole text, and which are `insistently anti-narrative (or counter-narrative)', `tending to occlude the horizontal coherence of the text for the vertical plenitude of superimposed or parallel forms of information'.[20] Certainly the radical simplicity of Capell's text, far more than Johnson's proto-variorum format, is hospitable to `general effect', to narrative, to horizontal coherence.

Nevertheless the extreme original purity of Capell's 1768 edition was at odds with a rather widespread contemporary presumption amongst textual

scholars that notes should immediately accompany the text. The classical bibliographer Edward Harwood, in his 1775 View of the Various Editions of the Greek and Roman Classics, felt able to use this criterion without the need for self-justification, preferring the Amsterdam edition of Bentley's Horace (1713) on the ground that `the text and notes are exhibited in the same page' (p. 172). There is a telling echo in Harwood's words of Lewis Theobald's stated preference for Bentley's Amsterdam Horace as formal model. And a number of Capell's contemporary readers and reviewers found Capell's plan unusual, aberrant, and baffling. `What shall we do with an edition of Shakespeare without notes?' asked the Monthly Reviewer, writing immediately after the publication of Capell's bare text in 1768. The Critical Reviewer, in the same year, looks in vain for explanations of Capell's textual `ipse dixits', and acidly remarks that all the commentary which Capell was already promising in 1768 `might have been inserted in the edition before us, without hazarding its character of being a well printed book.' And Charles Jennens, in his elaborately collated edition of Lear published in 1770, judged `that man . . . greatly mistaken in his ideas of beauty, who prefers the handsome appearance of a page in black and white, to the quick and easy information of his readers in matters necessary to be known for their becoming proper judges of the sense of the author, and the goodness of the edition'.[21]

Other writers found more value in the principle of linear and unhindered reading that the form of Capell's edition so strongly favoured. A writer in the English Review in 1784, with not only the 1768 bare text but also Capell's subsequent extensive apparatus in front of him, congratulated Capell on having printed the notes `by themselves, not forcing them upon the student of Shakespeare, but leaving the perusal to his choice'. The question of where to place the notes is represented as one of `the great questions that divide the learned world', and is settled in favour of a mode of linear reading of poetry which foregrounds passion, inspiration and sensibility:

Eighteenth-century editors and reviewers had, clearly, their own often sophisticated theoretics of editorial function and presentation, and conducted their own debates. It is also possible however to understand the intentions of Capell, and the response of his contemporaries to his work, in the light of a much more recent theoretics of the book. Jerome McGann (with others, and especially Donald McKenzie) has taught us to think of the text `as a laced

network of linguistic and bibliographical codes'.[23] McGann herds under the term `bibliographical codes' such features as binding, paper quality, book pricing, page layout, typography, and the general spatial relations and interactions among the `primary' text and the different components of editorial apparatus. Many of these features have implications for the sale, marketing and audience of books. All may in principle enact the intentions of the makers of books, and affect the understanding and response of readers.

Jerome McGann's discussion of bibliographical codes has been mainly concerned with primary writings, but bibliographical codes have consequences, as McGann is aware, not only for the original published documentary form of literary writings, but also for the edited forms in which they are subsequently mediated. Observing that `the very physique of a book will embody a code of meaning which the reader will decipher, more or less deeply, more or less self-consciously', McGann argues that

Especially pertinent in McGann's discussion of the form of the critical edition, and its effect upon reading, is his distinction between linear reading and radial or spatial reading. Linear reading—what Tristram Shandy would call `reading straight forwards, in search of the adventures' (1.20.65)—is what the unexamining reader does in the case of more or less straightforwardly non-self-reflexive texts. By contrast, literary scholarship provides numerous examples of texts which encourage radial reading, in search rather (to borrow Tristram's words once more) of `the deep erudition and knowledge which a book of this cast, if read over as it should be, would infallibly impart'. Of such books amongst the most striking are critical editions, the Cornell Wordsworth, for example, or the Oxford English Texts editions:

McGann has provided some of the most valuable conceptual yardsticks for assessing the forms of the scholarly edition, but in recent years he has been only one, if one of the most prominent, of a number of theorists and scholars addressing issues arising out of the forms in which literary texts are presented and annotated. That discussion has tended to focus on the ideological, on issues in particular of power, ownership, appropriation, and authority. Such

issues have been traced with particular perseverance by one of the most prominent advocates of a post-modern approach to editing, D. C. Greetham, who has argued that

Greetham identifies just such a set of ideological and procedural `instructions' in `traditional' clear-text eclectic editing, which, he alleges, `deliberately tries to prevent radial reading by concentrating the reader's eye on the "text itself" of the text page proper'. Greetham suggests, with other textual theorists, that the New Bibliography and the Greg/Bowers school of eclectic editing was naturally linked with the New Criticism, whose `critical concentration' on `the text itself'

Such texts conceal, according to the arguments employed by Greetham, their own motives, functions, and contradictions. They disguise the inevitable contingency of the text. They implement hierarchical structures in which the authorial text is privileged and apparatus occupies an inferior place. Indeed, Greetham remarks that some editions in his own field of Middle English

Greetham's strictures apply no doubt with a particular precision to Capell's Shakespeare: to its offering of `an apparently seamless and perfected text', its privileging of the author and the author's writing, its ejection from the text page of most (though not all) evidences of textual variance and transmissional indirection, its fragmented and drawn-out serial publication.

The adoption of a clear-text format however, despite the crimes of disguise and coercion which have been imputed to it by Greetham and others, does not by itself rule out an awareness and pursuit of textual variance, a recognition of the complexities of meaning and of the need to explicate meaning, or an understanding of the connection between the text-critical and the text-hermeneutic enterprise. Concerned as he is with textual appearances, Capell is nevertheless deeply involved in the reasoned and supported text choices and exegetical acts of interpretative editing. Though Capell's text itself is humanist in its regard for the beauty of the page and integrity of the text, his project taken as a whole is humanist in another, philological sense: in its detailed historical literary scholarship, its text-critical responsibility, and its emphasis throughout on knowledge and interpretation. The text of

his 1768 edition was always intended to be followed by a commentary and a comprehensive textual apparatus. The first volume of Capell's apparatus and commentary, his Notes and Various Readings to Shakespeare, did not appear until 1774, and the work was first published in its complete three-volume form after Capell's death, more than a decade after the appearance of the text, from 1779-1783. These three substantial quarto volumes present an extraordinary complexity of editorial material. There is a Glossary of Shakespearean words; lists of Various Readings to each individual play; a set of exegetical Notes on each play; a collection of illustrative passages called The School of Shakespeare; or, authentic Extracts from divers English Books, that were in Print in that Author's Time; and a verbal Index to the School.

These disparate editorial materials presented in the three volumes of the Notes and Various Readings were designed to work together in justifying textual choices and in explicating the text. Capell makes this combined purpose explicit in his Advertisement to the 1774 Notes and Various Readings: `the sole intent of the "Notes," is—to establish the Author's text, and to explain it'.[26] I provide reproductions of pages from Capell's Glossary, Notes, and School which all relate to a celebrated Hamlet interpretative crux (plates 12, 13, 14), the line which reads, in Folio and Quarto, `Unhouseled, disappointed, unaneled', and which reads, in the text of Capell's 1768 edition, `Unhousel'd, unanointed, unanneal'd'. Capell's Note on this passage refers to his Glossary definition, which reads:

In the Notes Capell explains that he prefers to the reading of all the old editions the `modern correction' unanointed (the `correction' is to be found in Pope, Warburton, and Theobald—in his 1733 edition, though not in Shakespeare Restored). The modern reading, Capell argues, provides the un- prefix for all three terms, and in the three original terms extreme unction is lacking. He rejects `disappointed' and `unappointed' on the grounds that `appointing is a general word, . . . whereas the passage requires a specific one'. Capell here provides an emendation which modern editors consider quite unnecessary; but the editorial materials of the Notes and Various Readings provide what amounts to a carefully reasoned explanation, though the emendation in the `clean text' of 1768 taken alone would appear tacit, arbitrary, unjustified. The other two terms of the triplet are explained by passages in the School, which can be hunted down by the curious reader who is prepared to make careful use of the School's Verbal Index. Aneyled is exemplified, and hence explained, by three passages from Thomas More's Works (1557). `Houseled' is exemplified and explained by two passages from the second

volume of Holinshed's Chronicle, and a passage from Henry Chettle's Hoffman (1631).[27] The method of the School, in this instance as elsewhere, is a particularly pure form of the annotating editor's characteristic tool, interpretation by adduction. Exegesis is not thought necessary at this point; it is enough to select the pertinent parallel, and to highlight with quotation marks the word explained. The reader is left to bring this adduction together with the frankly interpretative comment on `houseled' in the separate note.

In fact, if readers are to gather together Capell's account of this line in Hamlet in its entirety, they must consult no fewer than six differentiated and separated components: the `clean' text of 1768; the Variant Readings for Hamlet; the Notes; the Glossary; the Verbal Index to the School of Shakespeare; and the School itself. It is true that readers are offered a good many direction signs: each note is referred to a page and line number of the text; the Notes frequently refer to the Glossary; the Index to the School enables the location of passages appropriate to the particular crux; within the passage, the word which Capell is concerned to illustrate is flagged by quotation marks; and, very occasionally, a passage in the School is there referred to a particular passage in the text of one of the plays. But these directions re-affirm what a complex maze of pathways Capell has created.

The radial reading of Shakespeare in Capell's edition is made an especially disorientating experience by the difference in size between Capell's small volumes of Shakespearean text and his vastly larger volumes of apparatus, presented in a normal format of eighteenth-century prestige scholarship. The Monthly Review of 1776, addressing itself to the single volume of Notes which had been published in 1774, complained both of the tardiness and of the bulk of these first fruits of Capell's commentary, of `the awkwardness of huge quarto volumes to a text given in small octavo, . . . coming like heavy Falstaff so long after the battle.'[28] Normally the volumes of an edition— Pope's or Theobald's Shakespeare, Thomas Newton's Milton— `range with' one another in bibliographical format. Such volumes are designed to be bound consistently and to stand together on a gentleman's or gentlewoman's library shelf. Capell's 1768 Shakespeare `ranges' only with his earlier Prolusions; it is a volume not for the shelf or the library desk, but for the pocket and familiar reading, like the Aldine Horace, and unlike Bentley's quarto Milton, or even Johnson's relatively bulky octavo Shakespeare. In format Capell's 1768 text is abstractable, suited to the purposes of linear reading, of reading as poetry. It is mis-sized in its function as a component in Capell's completed edition, which may be thought of as a complex machine for a particularly elaborated and ergonomically problematic version of the kind of radial reading that McGann describes, in which: `one moves around the edition, jumping from the reading text to the apparatus, . . . from one of these to the notes or to an appendix, . . . and so forth.' In Capell's separation of the text from notes and apparatus in time as well as in space, however, there are resemblances with

those modern editions of medieval texts which, D. C. Greetham alleges, precisely prevent radial reading (as Greetham himself rather differently conceives it) `by publishing the textual notes and full apparatus in a separate volume, sometimes years after the text proper has been available'. Greetham accuses modern intentionalist eclectic editors of service to what he calls `the hobgoblin of readability', and no doubt Capell, whose textual methods are an almost textbook example of early eclecticism, is guilty of the same charge; though for Capell `elegant readability' is a good fairy, and no hobgoblin.

The paginary forms of Capell's edition articulate a mode of scholarly communication very different from that of Bentley's Milton or Johnson's Shakespeare. Capell and Bentley offer a radical formal contrast: on the one hand, Bentley's almost medieval or post-modern page, in which the editorial workings are not only exposed, but infiltrate and destabilise the Miltonic text; on the other, Capell's complex editorial structure in which Shakespeare's text is at the same time determined and liberated, and the reader effectively allowed the choice of linear or radial reading. There is an almost equally dramatic contrast between the clean unmarked text of Capell, followed at so great a temporal and formal distance by the numerous components of an apparatus generated by one solitary scholar, and the Johnson or Johnson/ Steevens Shakespeares, which lead the reader towards what Johnson calls `exactness' of interpretation through their cooler on-page conversations of plural expert commentators, the variorum edition being above all the place where the disparate kinds of explanatory knowledge possessed by many different scholars may be brought together, and where obscurities left by one may be resolved by the `happier industry, or future information' of another. These differences between editors, these contests on the page, are one measure of the tension in the eighteenth-century world of letters between the different varieties of philological and gentleman humanisms. I have suggested that in Johnson's statement on the necessary evil of annotation we see an evidence of that stress. No doubt Capell's edition might be seen as an attempt to address this cultural and literary problem; to invent an editorial form which allows for rapprochement between the demands of these two very different, but in Johnson's wish and perhaps in Capell's deed, consonant positions.

Notes