Friday, January 28, 2011
The Reviewing of Scholarly Editions
This essay ended my string of contributions to an annual review of scholarship, for the acting editor bluntly informed me that my intemperate attacks on Fredson Bowers and G. Thomas Tanselle could not be published. Many months later, absolutely unchanged except for a bracketed note about the suppression of the essay, still perfectly temperate toward Bowers and almost benign toward Tanselle, this essay was printed in Editors’ Notes: Bulletin of the Council of Editors of Learned Journals, 10 (Spring 1991). Fewer textual scholars and literary theorists saw it there than would have seen it in the glossier home it had been destined for, but I remain grateful to the bravery of Edna L. Steeves for taking a piece she knew had been suppressed already because it dared to challenge Bowers. I post it here as the first product of my I.R.I.S. scanning, cleaned up “by hand,” italics restored, I trust, and other minor glitches corrected, such as the attribution to Jerome McGann of a book called A Critique of Modem Textual Criticism.
THE REVIEWING OF SCHOLARLY EDITIONS
My topic is the reviewing of scholarly editions of literary works after the promulgation of W. W. Greg's rationale of copy-text. Except for the Wesleyan Fielding and the Clarendon Dickens, I restrict my examples to what I know most about, editions of American prose works, philosophical writings as well as belles-¬lettres, mainly the collected editions sponsored by the Center for Editions of American Authors (CEAA) or the successor organization, the Center for Scholarly Editions (CSE), and published, most often, by American university presses. I also refer to a few single-shot editions such as the reconstruction of the original text of The Red Badge of Courage that Norton published in 1982 and one or two other texts in popular series (such as the Riverside Editions).
In adjacent terrain there are two excellent statements on the reviewing of multiple volume editions of American historical documents. In the Newsletter of the Association for Documentary Editing 2 (September 1980) Gregg L. Lint eloquently protested the American Historical Review policy of reviewing at length only the first volume in a series, and in the American Archivist 43 (Winter 1980) Fredrika J. Teute sorted out dozens of reviews according to positions taken on various much-debated issues. Closest to my present concern is Lint's declaration that the effect of the AHR policy was to relegate documentary editions "to a historical backwater, somewhere beneath monographs" (p. 1). As an editor, a former member of the CEAA advisory committee, a frequent examiner ("vettor") of CEAA editions before publication (Clemens, Cooper, Hawthorne, Irving, Thoreau), and the one who, for several years, reviewed more CEAA editions than anyone else (Brown, Clemens, Cooper, Crane, Hawthorne, Howells, and in effect Melville), I can vouch that what Lint sees as a historical backwater is bigger and darker, a Dismal Swamp where editions of the papers of John Adams and his predecessor jostle oilily against Mohamet and his Successors, and where An Amateur Laborer makes A Chance Acquaintance with George Washington's papers and George's Mother. The CEAA and CSE volumes offered reviewers dozens of chances to educate themselves and their colleagues about the theory and practice of textual editing. Driven by no Imperative Duty, most of them threw away their chances.
During the decade before the great American editing enterprise was started, the New Criticism had won all the theoretical battles and had achieved practical domination over the academy. So strong was the New Critical prejudice against anything textual, bibliographical, and biographical that most people regarded the new editions with suspicion or disdain. Still worse, the fact that the New Criticism had been entrenched for several years meant that many of the younger editors had to educate themselves as they went along, in an Operation Textual Bootstraps. More insidious was the fact that the dominant editorial theory had close and debilitating affinities with the New Criticism: W. W. Greg's rationale of copy-text ignored the relevance of information about the process of composition and revision very much the way the New Criticism disdained biographical information. The new editors often performed feats of contortion to distance their supposedly pure textual analyses from grubby biographical research.
Part of the problem, also, was the antagonism engendered by the politics of organizing the CEAA and lobbying for federal funding. To the victors went the spoils, and the losers were widely regarded as spoil-sports. Lewis Mumford's and Edmund Wilson's abuse of the CEAA in the New York Review of Books (1969) is infamous history now, and looking coolly back at the scene of carnage G. Thomas Tanselle and I have said enough about the dubious motives, misreadings, and inaccuracies of these Grand Old Men. (See my summary in Flawed Texts and Verbal Icons [Evanston: Northwestern University Press, 1984], pp. 64-65.) Now one can admit that the victors sometimes behaved more like rambunctious late nineteenth-century American poli¬ticians than decorous late twentieth-century academicians, for having gained posses¬sion of the feeding trough some of them publicly scorned those they had pushed away. When Fredson Bowers lacerated conscienceless hacks of textbook-makers (as he did at the 1965 MLA) he not only set up a false distinction (I'd put the Riverside The House of the Seven Gables or the Norton Critical Edition of Maggie against the Ohio State House and the Virginia Maggie), he alienated people with residual good will toward textual editing and polarized the academy to the point that to be critical of Bowers was to be labeled an enemy of responsible editing. Over the years there was even some none-too-subtle intimidation, as when thorough, and thoroughly damning, reviews were suppressed. [Months after the preceding sentence was written, this essay was itself suppressed.] People were treated badly, and you had to identify the hidden agenda of grievances whenever you tried to interpret a new review.
The biggest part of the problem was the heritage of respect for authority. Bowers's stature as the chief disciple of Greg and the most eminent American bibliographer was such that his authority as an editor prevailed even when he was editing someone on whom he had not previously been an authority. There was, also, an engrained sense that if a literary critic like Sylvere Monod knew Dickens intimately (as we all knew he did) and was devoted to his service (as we all knew he was), then he would be just the editor to give fullest respect to the authority of his author. This belief in the mystical authority of authorities revealed itself unexpectedly, as when Jack Stillinger in reviewing Little Dorrit for Nineteenth-Century Fiction 35 (March 1981) declared the question of copy-text was settled merely because Philip Gaskell and Sylvere Monod were "firmly on the side of using printed versions as copy-texts" (p. 545). (To challenge the Dickens editions, as Bowers and Tanselle did, required only moderate skepticism and routine examination of the editorial policies.)
When CEAA volumes and other modem textual editions began limping from the presses, editors of journals felt no duty to review them extensively or seriously. It is easy to accuse the boards of many journals and annuals of choosing a pack of incompetent literary critics, choosing old-time editors (which meant writers of explanatory notes), or choosing aggrieved friends of authorities who had been shoved aside by New Bibliographers. I complained in Flawed Texts and Verbal Icons (p. 63) about two sets of reviewers, the persnickety ones who gleefully pointed out typographical errors and the sycophants who abased themselves before the master-editors. The snits may have been less offensive than the idolators, one of whom said: "The blend of astute critical judgment and technical expertise needed to effect this text" is "awe inspiring." Bowers's methodology inspired "a firmer faith that here are stories as close as human ingenuity can bring them to what Stephen Crane meant to create." The language was quasi-religious: awe inspiring editing led to firmer faith; yet a less devout motive may be discerned in some professions of faith, such as that in Luther S. Mansfield's fast move past the apparatus in the textually simple Redburn: "For the readers willing to take the scholarly accuracy of the text on faith, when this apparatus is present, the major interest of this volume may well be the thirty-eight page ‘Historical Note’" (American Literature 42 [November 1970], p. 406). When you're bored, say you are taking it on faith.
By its association with the American Literature section of MLA, American Literature ought to have set the standard for reviewing the new CEAA and CSE volumes lavishly and rigorously, at least ought to have set a policy of identifying and devoting more space to volumes that raised complex textual issues than to volumes that involved comparatively routine textual problems. Instead, American Literature sent out the first volumes for review, then disposed of later volumes with brief notices. No reviewer bothered to determine what was complex enough or new enough or challenging enough to warrant special attention. To make matters worse, American Literature printed unusually long and hostile reviews of scholarly editions by scholars with vested academic interest (and on a petty scale even vested commercial interest) in opposing the texts under review. Was I the only one who protested that this policy invited unpleasant misconstrual?
Some journals did better. Nineteenth-Century Fiction, perhaps because of the early presence at UCLA of Leon Howard and other people interested in editing, frequently allowed reviewers four or five pages, or even more, and encouraged them to review two or more volumes of an edition. David V. Erdman welcomed discussions of editorial issues in the Bulletin of the New York Public Library and its short-lived successor, the Bulletin of Research in the Humanities. Proof printed essay reviews on different CEAA editions, and Review has carried on where Proof left off, treating British editions as well as American. Still other incisive reviews of the new editions came, less frequently, in JEGP, Resources for American Literary Study, Modern Philology, American Literary Scholarship, and particularly in three special issues of Studies in the Novel (on Hawthorne, on Textual Studies in the Novel, on Stephen Crane).
The people who became competent reviewers were learning how to edit by editing. Any list of notable reviewers would include the editors 0 M Brack, Jr., Robert H. Hirst, Joseph McElrath, James B. Meriwether, and David J. Nordloh. Several fine reviewers were veterans of Meriwether' s South Carolina classes: Thomas L. McHaney, Noel Polk, James L. W. West III. Of the competent reviewers from outside the editorial establishment, very few pursued their arguments through two or more reviews or other textual essays and still fewer educated themselves through dealing with a series of different textual questions. John Freehafer, who deplored what a critic might conclude about Hawthorne upon encountering in the Centenary Marble Faun "sham Briticisms, false personifications, pretentious capitals, and miscorrections" (Studies in the Novel 2 [Winter 1970], p.499), soon gave up on the CEAA rather than continuing to try to improve it. For many years the best reviewing of the CEAA and CSE volumes went unpublished-some of the "vetting reports" by Meriwether, Michael Millgate, Nordloh, for instance. The education of everyone concerned with the CEAA was retarded by the decision that CEAA "vetting" reports were to be kept secret, sent only to the executive (later advisory) committee and the editor of the volume being "vetted." For a good decade at our very cleverest we reinvented the wheel every few months.
The sum total of stringent and elaborate essays on single volumes or several volumes of an edition was very small. John Freehafer's essay on The Marble Faun in Studies in the Novel 2 (Winter 1970) was perhaps the most notable review of one work. Reviews of the in-progress Hawthorne, Howells, Melville, and Thoreau editions appeared in the first four volumes of Proof. Surely the best review of a completed edition was David Nordloh' s "On Crane Now Edited" in the special Crane issue of Studies in the Novel 10 (Spring 1978). There were very few comprehensive essays on the CEAA. PBSA commissioned me to write a general article, "The CEAA: An Interim Assessment" (First Quarter 1974) that proved useful to two subsequent surveyors, Peter Shaw in American Scholar 45 (Winter 1975-76) and Tom Davis in Library 32 (March 1977). For the special textual issue of Studies in the Novel 7 (Fall 1975) Tanselle wrote a major article which amounted to a multiple review of both British and American editions, "Problems and Accomplishments in the Editing of the Novel." Tanselle's (unsigned) essay in PMLA 92 (September 1977) on the founding of the CSE included a review of major reviews of the CEAA. But a look at the density and scale of Tanselle' s analyses and summaries reminds one anew just how few were the attempts to grapple strenuously with individual CEAA volumes or the whole enterprise.
As Thomas L. McHaney said in the "Forum" of the "Textual Studies in the Novel" special issue of Studies in the Novel 7 (Fall 1975), p. 399, reviewers had seldom raised "the important questions." Conspicuously, the utility and possible hazards of creating an eclectic text were seldom discussed in these reviews of eclectic editions. Donald Pizer insisted in the Bulletin of the New York Public Library 75 (March 1971) that each authorially revised form "constitutes a distinctive work with its own aesthetic individuality and character" (p. 149). As I pointed out in BNYPL 75 (October 1971), the "aesthetic individuality and character" of most editions after the first is often specious, the result of a host of new compositorial changes mixed we any new authorial revisions. The issues are extremely delicate, but the fullest examination, Bowers's "Remarks on Eclectic Texts" (Proof 4, 1975) emphasized practical ways of creating eclectic texts rather than principles that might make you want to create one. No reviewer of the CEAA and CSE volumes has made more than a feint toward identifying--much less exploring--the basic historical and aesthetic arguments for and against eclectic editions.
The few important editorial essays of these years were usually by Tanselle. For Bowers's Studies in Bibliography Tanselle wrote a series of magisterial general essays--several of which are collected in Selected Studies in Bibliography (Charlottesville: University Press of Virginia, 1979)--on such topics as principles for editorial apparatuses, Greg's theory of copy-text, the editorial problem of final authorial intention, and the central questions of editing. These were most often general surveys where an attempt to describe developments on a wide front led Tanselle to quote from the editors' own descriptions of their activities rather than to evaluate what they had in fact done. In his lucid and comprehensive article on "Some Principles for Editorial Apparatus" (Studies in Bibliography 25 ), Tanselle said that the typical Bowers apparatus had come in recent years to represent "one possible standard" for apparatuses for scholarly editions (p. 45), but despite the caution of his phrasing he helped to standardize Bowers because he was so lucid and because he did not go on to explore--or urge others to explore--textual situations where that apparatus would not tell the serious student the most important things he might want to know. The purpose of explaining a large enterprise must partly account for Tanselle's habit of discussing Bowers's editorial principles but not scrutinizing his actual editorial performances. Tanselle's 1970s essays on editing are classic general statements, but they are not, except indirectly, part of the reviewing of the CEAA and CSE', and they tended, on the whole, to deflect criticism from Bowers's editing at a time when the whole textual enterprise would have benefited from open discussion. While Tanselle's essays dealt with "the central questions of editing" in a narrow sense, they did not deal with some of the central questions of textual scholarship (such as the bearing on textual study of cognitive psychology and research into creativity) or textual criticism (such as study of the aesthetic implications of textual evidence). Tanselle chose another road (a higher one, I can now testify) and none of the rest of us turned into ideal reviewers.
In realistic terms, how could we have? The ideal reviewer would have to have possessed or to have acquired, pronto, a high level of expertise in the life of a writer being edited as well as in textual and editorial theory and practice, and occasionally he would have had to acquire some knowledge of analytical bibliography. Then he would have had to devote to a volume of any complexity not the three or four days that he might put into reviewing a critical book but weeks--weeks in which he would have redone some of the editorial work from scratch in order to verify it and supplement it.
Starting with the most accessible part of the editorial apparatus, the ideal reviewer would have examined the "historical" essay that purported to layout the circumstances of composition, publication, reception, and (sometimes) subsequent critical reputation. He would have found that the biographical experts were fallible. William Charvat repeated the "germ theory" of the composition of The Scarlet Letter when the available evidence should have made it clear that James T. Fields had nothing to do with the length the novel reached even though he had everything to do with its being published as anovel rather than the longest piece in a collection of tales. Leon Howard and I overlooked a document crucial for the dating of Melville's enlargement of Pierre. Fredson Bowers overlooked a well-reported pre-publication excerpt from The Red Badge of Courage which ought to have been used as copy-text since it was set directly from now missing manuscript pages (Henry Binder laid out the facts, in an essay, not a review, in PBSA [First Quarter, 1978]); Bowers also overlooked a previously-reported letter to Crane in which Ripley Hitchcock accepted The Third Violet for Appleton but made it clear that he had reservations about it (Henry Binder printed the letter in Studies in the Novel 10 [Spring 1978]). Almost never did a reviewer challenge the editors for biographical lapses. (Needless to say, reviewers seldom recognized new research when there was new research. An exception was the frequent recognition that Martin Battestin had some news to tell in Tom Jones.)
Inspired by Tanselle' s argument that what makes an edition definitive is the apparatus, not the presentation of a text, the ideal reviewer would have dirtied himself in the textual lists, but only a handful of reviewers devoted any thought to what a list is for, how well a list fulfills its purpose, and whether or not the purpose is worthwhile. At times by blindly imitating Bowers's practices or by overzealously inventing new practices, editors created apparatuses which were not only clumsy and error-ridden but so misconceived that no user could find from them the information he would most naturally want to have. Being lucid of mind himself, and perfectly capable of following complicated rules (such as those by which everything not specified as so-and-so is considered to be such-and-such), even Tanselle has demanded too little of apparatuses. The academic establishment as a whole-biographers, literary historians, literary critics (not to mention theorists)--has demanded too little of textual apparatuses. Yet as Melville said, there is an aesthetics in all things--even in editorial apparatuses. You wouldn't know it from reading the reviewers who lost sight of the use people might make of a list of variants for a particular work--real people who love literature and are concerned with the process of literary creation more than they are with the vagaries of compositors.
Anyone who set out to use the textual lists in Maggie to find out how the 1896 edition differs from the 1893 and how the Virginia edition differs from 1893 could spend hundreds of hours without ever determining this information--partly because the emendations list is extremely long, partly because it is filled with errors (as is the Historical Collation), partly because the pattern of the Virginia Edition is to adopt the 1896 readings so that the putative copy-text is hardly that at all. My guess is that the Virginia Edition was set from a marked copy of 1896--for it contains the inevitable errors that result when actual printer's copy is not a photographic copy of the copy-text if the copy-text is a print. An apparatus becomes simply unusable when many errors occur in the lists; when the lists are misleadingly and inadequately defined; and when many unnecessary emendations are made, especially when many are taken from a later edition (you can't keep chronology straight if the variants you make note of in the margins are the early readings).
There are subtle relationships between the form of the lists and the editor's sense of what stages of textual history are most significant. The lists reveal that the editors were preoccupied with printed forms of the text and especially with book forms, when magazine publication preceded book form. In the Cooper Edition the list of "Rejected Readings" consists mainly of readings in "authorial" editions (editions Cooper supervised or at least authorized) which the editors have judged to be non-authoritative. That is, in the case of volumes edited from manuscript they mainly consist of misreadings made by the first compositors, misreadings never corrected by Cooper in later editions. In the case of volumes for which the first edition is copy-text, the list consists mainly of words in later authorized editions which the editors think are not changes made by Cooper but by others, primarily compositors. There is nothing inherently wrong about printing a list of words you do not adopt because you are pretty sure they are non-authorial, but sometimes the lists are long--eighteen pages in The Pathfinder--a lot of space to devote to words you think are non-¬authorial. Whatever the economic justifications, the effect of the policy is to valorize the non-authorial printed variants over the variants which survive from the author's active engagement in the creative process.
Even Tanselle has by implication endorsed this valorizing of printed variants: in his overview in Studies in the Novel 7 (Fall 1975), p. 346, he criticized the editor of Oliver Twist for listing only almost all of the differences in wording between manuscript and print but went on to assume that she, as a matter of course, must have listed all verbal variants in printed texts. I have seen NEH applications where would ¬be editors proposed to ignore a great English writer's revisions between manuscript and serial and between serial and book yet promised to record all the compositorial variants between late English and late American editions. Almost all editing inspired by Greg and Bowers has tended to valorize printed variants over manuscript variants, so that in a subtle, silent way evidence of the creative process was suppressed. The printing of these elaborate lists of rejected variants seems to me a case of doing meticulously something that is not the most desirable thing to do. As apparatuses proliferated and became standard, they came to be monuments to their makers, not lists that contained valuable information. Some imitative Bowersean apparatuses are as bad as or even worse than the worst of Bowers's owe--witness the useless nineteen-page Kent State University Arthur Mervyn "Historical List of Substantive Variants," which does not include variants in authorial editions.
The ideal reviewer would have slogged back and forth from list to text to the essay which lays out the textual policies of the edition. These essays are often formidably difficult to understand, yet in them basic editorial theory is promulgated. Sometimes the essays contain minefields where explosive departures from Greg's rationale are hidden under innocent-looking discussion of something so innocuous as an emendation policy. In Maggie Bowers violates a crucial part of W. W. Greg's "The Rationale of Copy-Text" (Studies in Bibliography 3 [1950-51])--indeed, the part where Greg made his momentous half-step beyond Ronald B. McKerrow's argument in Prolegomena for the Oxford Shakespeare (1939) that once having decided that an edition, "taken as a whole,” contains authorial variants, an editor "must accept all the alterations of that edition, saving any which seem obvious blunders or misprints" (p. 18). Looking at the so-called "indifferent" readings in Maggie, Bowers found the variants so "neutral, perhaps, as to require adjudication on a rather impressionistic basis" (p.lxxiii). This subjectivity flies in the face of Greg's sensible comment that if there is no reason for altering the copy-text reading, "the obvious thing seems to be to let it stand" (p. 31). Bowers's emending hardens into textual dogma in the insistence that once an author's revising hand is established in a reprint, every substantive variant should be presumed innocent (or authorial) unless proven guilty. Here Greg's best-known disciple, without saying so, is repudiating Greg's rationale of copy-text. By ignoring the evidence that the publisher had intervened in the text of Maggie and also ignoring the well-documented tendency of compositors to make casual alterations which are not always recognizable as errors, Bowers retrogresses behind the mature wisdom of Greg's rationale, which advocates discriminating among late variants and adopting only those the editor is convinced are authorial, back toward McKerrow's outdated argument. A conscientious reviewer would have seized on Bowers's retrogression, for Greg's rationale rests on his rejecting the all-or-nothing policy, and if you are going to print an eclectic text, Greg's theory is still the only game in town.
Pity the ideal reviewer who confronts Bowers's arguments for superseding the manuscript as copy-text in the Harvard University Press Pragmatism (1975). Bowers calls the manuscripts "something close to drafts," in so "rough" a state of inscription that they cannot "be thought of as normal authorial printer's copy" (p. 199). Suspending skepticism, the ideal reviewer would have to pursue Bowers outside an edition to find an explicit elaboration of his divergence from Greg. I quote from Tanselle's summary in Studies in Bibliography 34 (1981), p. 36: "Whereas Greg criticized 'the tyranny of the copy-text' . . . by arguing in favor of the use of editorial judgment to determine authorial revisions or corrections that should be incorporated into the copy-text, Bowers points out a different kind of tyranny: Greg's concept of divided authority, he says, ‘has been so welcome to recent textual critics that they have had a tendency to overreact against any other rationale' and thus 'have been loath to accept any suggestion of a return to unified authority even when the special situation warrants it." ("Unified authority" means a later print when manuscript or some earlier print survives: here Tanselle is quoting Bowers's "Scholarship and Editing," PBSA 70 [Second Quarter 1976], p. 180.) As usual, Tanselle is more the historian of Bowers's altering notions than critic of them, but Bowers's recent quarreling with Greg deserved scrutiny, particularly his questioning "whether Greg's advice is a good editorial principle to adopt under changed conditions from those of Renaissance compositorial and scribal free-wheeling" ("Greg's 'Rationale of Copy-Text' Revisited," Studies in Bibliography 31 , p. 155.)
When I asked Carolyn Jakeman for xeroxes of some chapters of the Houghton's manuscript of Pragmatism, I found that the manuscripts are not rough drafts at all, but perfectly legible fair copy, from which the first printers set very nicely. The printers made a few errors, of course, and since Bowers had relegated the manuscript to an inferior status he did not collate it carefully and did not notice misreadings which have persisted through every printed version. For instance, at one point in "Pragmatism and Humanism" (Lecture 7), the manuscript reads "logics, geometries, or arithmetics"--altered from the earlier "logic, geometry, arithmetic"; the first printer mistook the perfectly legible "geometries" for "geometrics," and every subsequent text, including Bowers's, prints "geometrics." Bowers's arguments are elaborate and intimidating, but they are also specious. Where James's manuscripts survived they ought to have been copy-text. The situation is a vindication of Greg's insights, and a condemnation of Bowers's rationalizations, which should be seen for what they are, not as valuable "extensions" of Greg.
For years the genuine textual issues, small and great, went unexamined by reviewers or only carelessly examined. Editors mired themselves in minutiae, fussing over perfectly acceptable spelling variants and old-fashioned punctuational practices, often regularizing spelling and punctuation in violation of their claims to be printing unmodernized texts. All the while, they were blindly following a rationale of copy-text that (although perfect in some very simple situations) was incompatible with what is known of the creative process, since, denying that the process is a process, it assumes that an author's aesthetic control over anything he writes lasts as long as he lives. Editors, in short, did not rethink editorial theory and editorial apparatus or, more basically, rethink Greg.
And while textualists and reviewers alike have been confused or flippant, or merely ponderous, on most of the textual-editorial issues, both of theory and procedure, they have seemed almost unaware of a much more important area, that ambiguous terrain where textual and biographical evidence create aesthetic consequences. Biography has been kept away from editorial theory; creativity theory has been kept away from editorial theory-and is banished in Jerome McGann's modish A Critique of Modern Textual Criticism (Chicago 1983), an influential pastiche of the views of James Thorpe and Donald Pizer; cognitive psychology (with its rich new discoveries about human memory) has been kept out of editorial theory. Until editors care passionately about literature again, until editors approach the creative process with wonder and awe, we are not going to have great triumphs of textual editing and we are not going to have great new reviews of scholarly editions.