University of Virginia Library

III

When one understands Greg's theory and the CEAA's implementation of it, one cannot help regarding many of the recent discussions (both favorable and unfavorable) of this joint subject as naïve and parochial, and frequently as uninformed or misinformed.[43] A few,

however, do raise important issues, and it is regrettable that a survey of these discussions must begin with one of so little substance as that of Edmund Wilson. In a two-part article entitled "The Fruits of the MLA," published in the New York Review of Books on 26 September and 10 October 1968, Wilson offered what can only be called an ill-tempered and incoherent attack on the CEAA editions, making references to six volumes ostensibly under review;[44] in December of that year the article, with a postscript commenting on some of the correspondence provoked by it,[45] was published in pamphlet form as "A New York Review Book," and in 1973 it was collected into the posthumous volume The Devils and Canon Barham (pp. 154-202), edited by Leon Edel.[46] Because of Wilson's stature, this article has received a considerable amount of attention and will continue to have an audience in the future as part of his collected essays; if it had been written by a lesser figure, however, its obvious motivation and manifest confusion would have prevented its being taken seriously. Wilson makes transparent his motive for discrediting the CEAA editions by quoting, at the start, a letter he had written to Jason Epstein in 1962 setting forth the idea of "bringing out in a complete and compact form the principal American classics," based on "the example of the Editions de la Pléiade" (pp. 155-156); this undertaking he had hoped would be supported by the National Endowment for the Humanities, but the MLA, he says, "had a project of its own for reprinting the American classics and had apparently had ours suppressed" (p. 159).[47] Thus determined to find fault with the results of the MLA project, Wilson never addresses himself to the basic editorial rationale (that is, to Greg's theory) but instead is content to ridicule such matters as the laboriousness of the research involved, the extent of the apparatus, and the physical size of the volumes. The article is, uncharacteristically, full of confusions, if not inconsistencies,[48] the most egregious perhaps being his professed admiration for a "sound and full text" (p. 157) combined with his view that collation is unrewarding if it does not uncover "serious suppressions and distortions" (p. 161) or interesting variants ("the scrutinizing of variants may, in some cases, be of interest," p. 172). What Wilson is unwilling to acknowledge is that the CEAA's concern extends beyond a scholarly audience to the general public: the CEAA, he says, is "directing a republication of our classics which is not only, for the most part, ill-judged and quite sterile in itself but even obstructive to their republication in any other form" (p. 190). He fails to note that the pages of text, unencumbered in most cases by editorial intrusions, are suitable for photographic reproduction in volumes more convenient to hold and that the apparatus (which, admittedly, helps to make some of the volumes cumbersome), rather than being "sterile," may serve to generate other editions, based on differing evaluations of the evidence—or at least to encourage analysis of the editor's judgments. Wilson's piece scarcely demands any reply, but the celebrity it achieved caused the MLA to feel that some sort of official notice was appropriate, and in March of 1969 the MLA published a pamphlet entitled Professional Standards and American Editions: A Response to Edmund Wilson, containing two accounts of the history and aims of the CEAA, by William M. Gibson and John H. Fisher, along with letters from five scholars enumerating errors or confusions in Wilson's remarks.[49] Actually, all that was necessary, if a reply was to be made, was Gordon Ray's brief comment which stands as the epigraph to the pamphlet. Recognizing that "this attack derives in part from the alarm of amateurs at seeing rigorous professional standards applied to a subject in which they have a vested interest" (and thus recognizing the attraction which Wilson's position had for a number of people one might have expected to see through it),[50] Ray observes, "As the American learned world has come to full maturity since the second World War, a similar animus has shown itself and been discredited in field after field from botany to folklore. In the long run professional standards always prevail."

In contrast to Wilson's article, which makes no reference to Greg's theory, two brief essays which appeared soon after it—the work of Paul Baender and Donald Pizer—do raise questions directly about the applicability and usefulness of Greg's "Rationale." Although each of these essays is weakened by a partial misunderstanding of Greg, they at least raise issues the discussion of which may serve to clarify certain points in some people's minds. Baender, an editor associated with a CEAA edition, published in 1969 a note entitled "The Meaning of Copy-Text,"[51] which asserts that the term has become "ambiguous and misleading," principally for two reasons: first, that it is a "banner word" which "tends toward the superlative" and which thus implies "authority beyond its denotation, as though the term itself ratified an editor's choice of text"; second, that it is "not suited to the full range and complexity of editorial problems" (p. 312). The first point has nothing to do with the word "copy-text" or the concept but only with unscholarly reactions to it—unscholarly because they depend on the "prestige" (as Baender calls it) of the term rather than the arguments lying behind it. The second is of more consequence but is based on an oversimplification and distortion of Greg's position. If it were accurate to say flatly that Greg's theory is eclectic with respect to substantives but maintains "a single-text criterion" with respect to accidentals (p. 314), or if it were fair to suggest that its application to situations involving prepublication texts results in "another stage for a retrogressive pursuit of copy-text" (p. 316), then one would have grounds for claiming that it is "not suited to the full range and complexity of editorial problems." But nothing in Greg's theory, as we have seen, prohibits the emendation of accidentals in the copy-text when one has grounds for doing so; nor is it consistent with his theory to assume that a surviving manuscript must necessarily—regardless of its nature—become copy-text, since he allowed for the possibility that in some

cases a later, rather than an earlier, text is the appropriate choice. One of Baender's illustrations[52] rests on a basic confusion (of which Baender is not alone guilty) between "copy-text" and "printer's copy." Baender cites a situation in which the number of authorial alterations in a later printing makes it more convenient for the editor to use a reproduction of that later printing as the basis for his text, entering onto it the readings of the first printing wherever the later readings are not judged to be authorial. Such a procedure, of course, does not violate Greg's theory (however risky it may be in practical terms, since one is increasing the probability that nonauthorial readings may inadvertently be allowed to remain in the text); but Baender's feeling that one follows the procedure "despite this convention of copy-text" makes clear that he is not focusing on the distinction between "text," meaning a particular arrangement and formal presentation of a group of words, and "printer's copy," meaning a specific physical copy of a text furnished to the printer. Greg's "copy-text" is a "text"—which can exist in more than one physical embodiment (for example, the individual copies of an edition)—and Greg did not comment on the manner in which that text should be reproduced for the use of the compositor who is setting type for the editor's new edition. The CEAA Statement does go on to recommend, for obvious practical reasons, the use of a photographic reproduction of the copy-text as printer's copy; but not to follow this course, whether for convincing or questionable reasons, does not in itself contradict Greg's theory, since no theoretical matter is at issue.[53]

Two years later Donald Pizer raised again,[54] but in broader terms, the question of the applicability of Greg's theory to recent literature

by enumerating five ways "in which copy-text theory is unresponsive to the distinctive qualities of [that is, the historical circumstances lying behind] modern American texts" (p. 148).[55] Although Pizer calls attention to some issues that deserve careful consideration, his article is ineffective as an argument against the general usefulness of Greg's rationale because it fails to distinguish between theoretical and practical concerns and to recognize fully the lack of dogmatism in Greg's approach. The last three of his points are irrelevant to an analysis of Greg's theory—what they are relevant to is a consideration of the particular kind of edition (in the sense of text plus apparatus or other commentary) appropriate for modern (nineteenth- and twentieth-century) American literature. While this subject is of course a legitimate matter for debate, the issue is only confused by the implication that the adoption of Greg's theory determines the nature of the apparatus (or whatever accompanies the text) as well as of the text itself. Thus his third point—that the multiplicity of manuscripts, typescripts, and proofs which survive for some modern works makes the task of recording all variant readings excessively onerous[56] —presupposes that something in that theory of copy-text necessitates a complete record of variants, for he concludes: "the theory of copy-text either hinders the preparation of critical editions or encourages the production, at immense expense, of unusable editions" (p. 151). But whether or not one wishes to follow the practice of CEAA editions in recording variants (and the CEAA does not require as an absolute rule that all pre-copy-text variants be noted in print) has nothing to do with whether or not one edits a text in accordance with Greg's theory; and naturally the job of editing a reliable text is complicated by the survival of numerous documents, for the variants in them must be examined carefully regardless of whether a listing is to be published. Pizer's fifth point is a related one, dealing also with apparatus: he objects to clear text in a "critical edition" because turning to the back of a book to consult the apparatus is more difficult than looking at the foot of a page, and he disapproves specifically of the sections of apparatus which the reader of a CEAA edition must "juggle" (p. 152). The possibility that a more efficient apparatus can be devised is always open; but the plan of the apparatus does not alter the editorial procedure, and a dislike of "the tendency toward clear-text publication" cannot through any argument become an "objection to copy-text theory."[57] The fourth of Pizer's observations amounts to nothing more than the recognition that some editors may choose to edit works which some readers deem unworthy of the effort expended. He speaks of "the absolutism of copy-text procedures"—meaning the uniform treatment of major and minor works—without acknowledging that the decision to edit is a critical evaluation in itself. Not all the CEAA editions are "complete" editions, and those that are reflect—rather than any requirement of Greg's theory—the critical belief that the stature of the authors involved demands full-scale investigation of even their lesser pieces.[58] Very few people (and certainly not the CEAA) would dissent from the view that—since time and money are not unlimited—"practical editions" must suffice for many literary works; but there will never be complete agreement on exactly what works those are.

Pizer's first two objections, in contrast, do raise questions about theory, but not, as he implies, solely about Greg's theory; they are serious questions which any editor must face, whether in the context of Greg's rationale or not. It is Pizer's contention that Greg's theory, by leading an editor normally to adopt the accidentals of a manuscript in preference to those of a first printing, ignores the fact that modern authors sometimes "rely on the taste" of particular publishing-house

editors, who thus "have increasingly participated in the creative process of their authors." He argues, in other words, that an author who expects or encourages certain kinds of alterations to be made in the publisher's offices must be said to prefer or "intend" the resulting text. "If an author," as Pizer concisely puts it, "within such a relationship and for whatever motives, accepts an editorial change or suggestion, his acceptance is the equivalent of a creative act, even though the act is the initial responsibility of an editor" (p. 148). The aim of Greg's theory, with which no scholarly editor would quarrel, is to establish the text which the author intended; and by concentrating on unmodernized texts it aims to establish the author's intended text in respect to accidentals as well as substantives. What constitutes the author's "intention" is of course the crucial question, and in answering it the editor must always depend, to a greater or lesser degree, on his critical insight. It is axiomatic that an author's own statements of his intention, when they exist, do not, for a variety of reasons, necessarily coincide with his actual intention—the only guide to which is the work itself. An author may acquiesce in his publisher's decisions and then rationalize his behavior; or he may genuinely be grateful for changes which make his work, in one way or another, more acceptable (and salable) to the public; or he may approve of alterations in many other kinds of situations—without truly believing that the result quite represents his own style or approach. What appears in a prepublication form of a text is normally a better representation of the author's habits than what appears in a first printing, and the text of a fair-copy manuscript or typescript reflects the author's intention, whether or not it turns out to be his final intention in every respect. It is true, as Pizer says, that choosing "an early copy-text encourages a frame of mind which requires later variants to 'prove themselves' as authorial rather than as editorial or printer's variants" (p. 149); but such would seem to be the safest course in most instances, since the author's responsibility for a later reading—especially in accidentals—is normally less certain than his responsibility for an early one.[59] Of course, such editorial caution may occasionally produce a text reflecting "an author's discarded rather than final intentions," but at least it reflects his, rather than someone else's, intentions. The editor's critical judgment—his literary taste exercised in the light of his intimate knowledge of the author and all known relevant external evidence—must finally determine the case; and there is nothing in Greg's theory to prevent him, on this basis, from deciding that the later variants have indeed "proved themselves." If, however, he starts from the assumption that the author and the publisher's editor are creative collaborators, he will, to be sure, produce an unmodernized text—in the sense that it reflects the author's period—but it may be far from the text which the author wished (finally, or at any other time).[60]

This question leads to a consideration of eclecticism, and Pizer's second point is that an eclectic text, incorporating later substantive readings into an earlier copy-text, violates the integrity (or "imaginative 'feel,'" as he calls it) of individual stages of an author's work. The result, which "may incorporate changes made by the author over many years," is, he says, "a text which never existed and which has little or no critical interest" (p. 150). Certainly it never existed, for a critical text by definition differs from any single extant documentary form of the text; but whether it is of critical interest depends on how well the editor has performed his task, for his aim is to produce a text which accords with the author's intention more fully than that of any given extant document or printing. The fact that an author may make alterations in a work over a long period of years does not necessarily mean that they reflect different conceptions of that work; when they do, then of course each version should be edited separately as a work in its own right (following the theory of copy-text with regard to each). But surely it blurs a critical distinction to insist that every revision "constitutes a distinctive work with its own aesthetic individuality and character" (p. 149).[61] What this argument leads toward, obviously, is the abandonment of the editor's critical function and the restriction of editing to the production of accurate facsimiles. It is somewhat puzzling that Pizer is reluctant to allow the scholarly editor to attempt

a historical reconstruction of the author's intended text, when he is quite ready to believe that contributions of the original publisher's editor were accepted by the author as furthering his intentions. And it is paradoxical that a person who objects to the uniform editing of major and minor works for its failure to make "critical distinctions" ("which is what the study of literature is all about for most scholars and students") should disapprove of texts that involve an editor's critical judgment and should hesitate to offer to the public clear-text editions without apparatus, since they constitute "only the editor's beliefs about the author's final intentions" (p. 152). If, as he recognizes, editing is "in varying degrees an aesthetic enterprise," the "editor's beliefs" command respect to the extent that the editor is at once a careful historian and a sensitive critic; and the existence of insensitive editors casts no more doubt on the undertaking as a whole than the existence of obtuse literary critics does on the activity of literary analysis. When Pizer calls Greg's theory of copy-text "'scientific' in its central impulse" because it "establishes a principle (albeit a flexible one) that is supposed to work in every instance" (p. 153), he disregards the fact that the principle is "flexible" for the very reason that it places no restriction on the operation of informed judgment.

In the months following the appearance of Pizer's article, several communications stimulated by it were published in the pages of the same journal. Norman Grabo, in April 1971, and Hershel Parker, at greater length in October, criticized Pizer's position.[62] Then in November John Freehafer, applauding Pizer, set forth what he considered to be three additional "major deficiencies of the CEAA editions."[63] It is significant that the deficiencies are said to be "of the CEAA editions" and not of Greg's theory, for what Freehafer objects to is not Greg's approach but the way it has been put into practice in CEAA editions, along with the decisions reflected in those editions about the kinds of material to be presented. His first two points are patently argumentative: the CEAA editions, he believes, exhibit "a failure to learn from the best editorial practice of the past," because the history of Shakespearean scholarship has shown that the "empty boasts" of an editor like Theobald prove in the long run to be of little substance

(whereas critical discussions, like Johnson's, are often of lasting value); and they demonstrate "a failure to present literary works as such" by not providing critical analyses[64] and explanatory notes, by being "almost totally concerned with bibliographical questions." The first point springs from the CEAA use of the word "definitive." As I have said before, this word was an ill-advised choice and has been too freely used; nevertheless, it should be clear to any reader of a CEAA apparatus, from its discussion of various problematical points, that CEAA editors are not claiming (nor did Greg expect editors following his rationale to be able to claim) that they have made all the right decisions and thus produced a "definitive" text; all they can aim for as a goal is to provide a definitive apparatus, recognizing that it is at least possible sometimes to establish facts. The decision to emphasize the history of the text (including the history of critical reaction to it) in CEAA introductions and afterwords is obviously related to this point, for those essays constitute another part of the apparatus, directed toward laying out what historical facts can be established.[65] That these editions are historically oriented, however, does not mean that they fail "to present literary works as such" but simply that they do not present literary works accompanied by any one critical interpretation.[66]

Freehafer's third point, however, raises an issue which deserves to be commented upon, even though what must be said is implicit in Greg's theory and will therefore seem redundant to some readers. He

complains that the CEAA editions have failed "to use Greg's theory of copy-text with sufficient boldness and imagination to reconstruct ideal authorial texts of many of the works being edited" (p. 419). In support of this proposition he cites the differences in the texture of accidentals between The Scarlet Letter and The House of the Seven Gables in the Ohio State edition, resulting from the fact that copy-text for the former is a first printing (the manuscript not having survived) and for the latter is a manuscript; these differences, he says, can be regarded as "valid reconstructions of the author's intentions . . . only on the incredible supposition that within a year Hawthorne turned from a passionate devotion to house-styling to a passionate rejection of it" (p. 422). What this argument fails to notice is that Greg's theory, as a scholarly procedure, must operate on the basis of the available materials for a given text and aims at reconstructing the author's intention insofar as surviving evidence permits. One can well believe, with Freehafer, that Hawthorne's preferences did not shift so drastically within a year. But can one therefore say that the features of one known manuscript would also have been those of another, now missing, manuscript from approximately the same time, and that an editor would on that basis know how to set about inserting those features into the text for which no manuscript survives? Answering No could perhaps be called unimaginative, but one should then add that to be more "imaginative" would be inconsistent with the scholarly goal of exercising critical judgment within the bounds set by ascertainable fact and documentary evidence. The belief that the accidentals in one CEAA text should be identical to those in another contemporary text by the same author stems from an assumption that the CEAA goal is to reconstruct the author's "intention" in an absolute sense, rather than in the more realistic sense of that intention for which there is documentary evidence for a particular work. Naturally the editor's knowledge of the author's practice in other works, for which a different range of documents exists, ought to play a role in any decision he makes; but it would be a rare instance indeed in which such knowledge was so certain and comprehensive that the editor could feel confident in his ability to repunctuate or respell for the author without introducing far more readings that never existed than those that did. Anyone who wishes to take a more "imaginative" approach and to interpolate the habits of one manuscript or a group of manuscripts into the texts of other works would of course be able to examine and utilize the evidence present in the texts and apparatuses of the relevant CEAA volumes. A second illustration of Freehafer's is again indicative of a fundamental misunderstanding of what kind of text the CEAA is attempting to provide. Turning to a different period, he cites two recent editions of Dryden's The Indian Emperour (one in the University of California Press Works, 1966, and the other in the University of Chicago Four Tragedies, 1967) and observes that, by selecting two different copy-texts, these editions present, even after editing, two very different texts. Since both attempt to reconstruct the author's intention, both should theoretically, he says, "have arrived at identical texts" which "agree word for word, letter for letter, comma for comma" (p. 422); that they do not so agree he attributes to an unimaginative use of Greg's theory, to "tyranny of the copy-text." But Greg, precisely because he recognized the role of imagination and judgment, would never have expected two editors to make all the same choices and emerge with identical texts. What the scholarly editor is striving to do is to put his critical judgment at the service of recognizing what the author intended, and no one, including the CEAA editors, would claim that any one attempt at this is the final or "definitive" one. Freehafer's urging of a more imaginative use of Greg's rationale to produce an "author's ideal text" seems rather at odds with his criticism of the Ohio State Hawthorne, both here and in an earlier detailed discussion of The Marble Faun,[67] for making too many emendations; the existence of arguable emendations and variants suggests the impossibility of universal agreement on critical issues, and a more imaginative approach would not be likely to lessen the range of disagreement. Several times Freehafer speaks of "definitive texts"—not "definitive editions"—and in that earlier essay says that how definitive the Hawthorne edition is "largely depends upon how the editors have used their collations, concordances, and other data in establishing Hawthorne's texts" (p. 487); however, the distinction between a "critical" text and a "definitive" edition (which embodies such a text along with other information) cannot be overlooked if debates about these matters are to get anywhere. Pizer, too, in his response in December 1971 to Grabo and Parker,[68] reiterated the need for "flexibility" and for resistance to "the tidy and neat," apparently without recognizing that CEAA critical texts and their apparatuses reflect those qualities.

The same month saw the appearance of the first volume of Proof, which contained a long essay by Morse Peckham, "Reflections on the Foundations of Modern Textual Editing."[69] Peckham is the only critic of Greg's theory thus far to explain his criticisms in the context of a thoughtfully developed analysis of the nature of human communication. Most of the previous comments, as we have noticed, either arose from a misunderstanding of Greg or dealt with largely superficial matters; Peckham, on the other hand, attacks Greg's central assumptions by setting forth a view of human behavior incompatible with them. Although I shall try to show why his argument does not seem

to me to invalidate Greg's rationale, I hope it will be clear, at the same time, that Peckham is raising the kind of fundamental questions that have been too little discussed. His essay—aside from its examination of whether analytical bibliography can be regarded as "scientific"[70] —attacks Greg's theory in two respects: (1) it denies that substantives and accidentals can be meaningfully segregated; (2) it denies that the reconstruction of a text representing the author's intention is a meaningful (or attainable) goal. Although the argument supporting the first can be seen as consistent with and deriving from the larger propositions underlying the second, the first point can be taken up separately and is discussed first by Peckham.

The distinction between substantives and accidentals, Peckham says, was necessary to Greg because of the nature of the material he was dealing with: "the sparse and inconsistent punctuation in [Renaissance] dramatic manuscripts that have survived" (p. 124).[71] But, he adds, the distinction "is useless outside of his very special class of texts" (p. 125), because most later authors (and some Elizabethans as well) were aware that punctuation affects meaning and were not helpless victims of a house-style imposed by their publishers. Punctuation, he argues, does more than affect meaning, for, without punctuation, "it is frequently impossible to decide on that meaning":

Thus "an educated author produces his punctuation as he produces his words; together they make up an unbroken semantic continuum." Clearly Peckham is correct in believing that no fixed line separates punctuation (or other "accidentals") from wording in the expression of meaning; and I am not aware of any editor who accepts Greg that would take issue with this point. But it does not therefore follow that no practical distinction can be made between them. Greg, of course, insisted that he was concerned with a "practical," not a "philosophic," distinction; but Peckham finds illogical (because it does seem to claim a "philosophic" basis for the distinction) Greg's footnote in the "Rationale" which asserts that punctuation "remains properly a matter of presentation," despite the fact that it can affect meaning. Now that footnote, it must be admitted, is not written with Greg's characteristic clarity, but the point he was getting at (as the drift of his whole essay suggests) is not, in my opinion, illogical. A paraphrase might go something like this: "Although punctuation and spelling are, from a theoretical (or 'philosophic') point of view, inseparable from words in the written expression of meaning, in practice people (i.e., scribes, compositors, and even authors at times) do react to them as if they were somehow less significant." What Greg meant by a "practical" distinction is one which, however mistaken it may be, has in fact operated to govern human behavior; and, since the editor is concerned with analyzing the behavior of certain individuals, such a distinction may be useful to him. It is certainly true, as Peckham later points out (p. 145), that Elizabethan compositors felt freer to depart from the punctuation and spelling of their copy than later compositors. But does not a modern publisher's editor generally feel less compunction about inserting a comma than altering a word? Does not the author who acquiesces to a suggested change of punctuation more readily than to one of wording, or who believes that his punctuation but not his diction actually demands revision, feel that there is some sort of distinction? So long as one can say, "I think my quotation is accurate, though it may differ in a mark of punctuation here and there," and not be regarded by most people as uttering nonsense, one can believe that a "practical" distinction between the two does widely exist in people's minds. To the extent that punctuation and spelling are popularly regarded as distinct from what is being said—and it scarcely requires demonstration that they are, and have been, so regarded—the transmission of texts is correspondingly affected. However much an editor may deplore the confusion behind this attitude (analogous to the popular oversimplification of the relation between form and content), it is his business to take into account, as realistically as he can, the factors that influence textual transmission. (Of course, some accidentals do have less effect upon meaning than others: a comma marking a phrase-ending that would be recognized even without the comma serves less purpose than one which marks the beginning of a nonrestrictive clause. But no definite line separates this second type of accidental, sometimes called "semi-substantives," from the first, which also, though more subtly, may affect the sense.) What I take Greg to be saying, then, is that the editor distinguishes substantives and accidentals not because he believes that he is making a valid conceptual distinction between two elements in written language but because the distinction is one which is likely to have been made by the persons who have been involved in the transmission of any given text (and which therefore may be useful in segregating different features of that text which may have been accorded different treatment).

Obviously Greg does not expect an editor to be bound by this distinction in his own thinking, for he makes no requirement that the editor always accept the accidentals of a first edition or that he always accept all the accidentals of whatever text he selects as copy-text. He merely observes that, given the popular tendency to be less careful with accidentals than with substantives, more of the author's accidentals are likely to be present in a first edition than in later editions. And, of course, the whole point of attempting to recover the author's accidentals is that they do indeed constitute an important part of his expression. The distinction between substantives and accidentals has no influence on what an editor decides to do when he believes that he has convincing reasons for doing a particular thing,[72] but when he does not have such reasons, the distinction enables him to make a decision in accord with what common experience shows to be a widespread attitude (one which is thus likely to have been operative in any given instance).[73] Although English spelling has become more fixed over the centuries and styles of punctuation have altered, I see no evidence that the popular conception of spelling and punctuation as the accouterments of words has shifted[74] —or any reason, therefore, not to find

Greg's approach applicable to later writings. Greg's choice of the terms "substantive" and "accidental" was, as I have said before, unfortunate, and the fun which Peckham has with them, calling them "strangely medieval," is deserved; there would have been fewer misunderstandings (and certainly fewer complaints about unnecessary jargon) if, as Peckham suggests, plain terms like "words," "punctuation," and "spelling" had been employed. But surely the point Greg was getting at is not completely hidden behind the terms he chose.[75]

Peckham's principal argument, however, deals not with accidentals but with the concepts of "text" and "author." He believes that many literary scholars—including Greg and his followers, who attempt to establish the author's intended text—are guilty of literary hagiolatry, exalting the ideas of "author" and "work of art" in ways not consistent with the nature of human communication. An author, he says, is simply an organism which produces utterances, not as a result of any special inspiration but as a result of being human:

But the author is different from other utterers in that he assembles a series of utterances into what "he judges to be a discourse" and makes this series available to others, proposing "that they too judge it to be a discourse" (p. 138). The development of the discourse up to that point has involved a combination of producing utterances and changing (or revising) them;[76] thus the author, even before his work becomes public, has already been in the position of looking back over something previously written, reacting to it as a reader, since he is not at that moment the producer. This process, Peckham argues, continues indefinitely: sometimes other human beings (such as publishers' editors) react to and change the discourse, and sometimes the author continues to change it. Each is responding to a particular version, and each "can make a change acceptable to the author or to anyone else involved" (p. 141). The "textual editor" is but one more human organism in this sequence, producing one more version of "a postulated work, that is, of a construct" (p. 128). Whether a valid distinction can be made between changes by the author and by others, therefore, turns on As a result, one cannot speak meaningfully of a single ideal "text" of a work; if the development of the concept of individuality ("self-mediated divergence from a cultural norm") had not caused the editor to confer "sainthood" on the supposed "author" and exalt certain works as canonical (p. 149), he would realize that he is "simply continuing an activity initiated by the author" (p. 144).

Although this is a greatly simplified summary of Peckham's analysis, I think that it does not distort the main outlines of his position. But one does not have to disagree with this general position in order to

believe that such editors as those of the CEAA volumes are pursuing a sensible, meaningful, and useful goal. The "textual editor" whom Peckham describes—he defines the term as subsuming "both analytical bibliographer and textual critic" (p. 141) —is naturally, in Peckham's general terms, just another person making changes in a text; but it would seem to be more illuminating to go on and note how he is to be distinguished from others who do that. For persons who make changes in pieces of writing—and are admittedly engaging in basically similar actions—fall into two groups, those performing scholarly editing and those performing what may be called "creative editing."[77] There is no reason why one cannot regard a piece of writing as the common product of all beings who have come in contact with it and reacted to it; when it is viewed in this way, any change, made at any time, whether by the original publisher's editor, by the author, or by a later "editor," has the same status and may be judged to have improved the work, harmed it, or left it the same. From this point of view a critic is not performing his function conscientiously if he does not alter the work to make it, according to his standards, more satisfying than it has ever been before. There is, as I say, no objection to this procedure—so long as one's goal is critically rather than historically oriented. But the scholar sets a goal of historical reconstruction.[78] That the "author" has some individuality is suggested, even in Peckham's approach, by the recognition that he initiated the discourse, which is then operated upon by himself and others. If that discourse is of sufficient interest, a historical interest may also attach to the initiator; and if the same being initiates a number of such discourses, the interest may be correspondingly greater. What the scholarly editor attempts—recognizing the difficulty of the task and even the impossibility of its absolute achievement—is to remove from the discourse those features for which the initiator was not responsible.[79] The result is not necessarily what the editor himself prefers but what he believes to be the author's contribution to a given discourse. The scholarly editor is thus a different kind of responder from the others in the chain Peckham is talking about; it may be that the editor, if he lives in the mid-twentieth century, cannot avoid reacting in part in mid-twentieth-century terms, but his aim is to use his critical faculties[80] to place himself in the frame of reference of the author and the author's environment. That such an aim is impossible of full attainment does not invalidate it as a guideline for a direction in which to move, despite Peckham's labeling of this attitude as "pure hagiolatry" (p. 138).[81]

The difficulty with accepting Peckham's statement of the case is evident when he remarks that the concepts of "text" and "author" require the "textual editor" to "produce a definitive edition, which he cannot do, instead of producing a new version more satisfactory for some specific purpose than any existing version, which he can do" (p. 151). What Peckham says the editor can do is in fact what CEAA editors do (and realize they are doing): they produce a critical (not definitive) text which they believe to be more satisfactory for the purpose of the historical study of literature than any previous text,

and they regard the edition embodying that text as "definitive" only in its recording of certain classes of data. Part of the problem, throughout the essay, is Peckham's interchangeable use of "text" and "edition" and his belief that CEAA editors really think they are producing definitive texts.[82] Perhaps, indeed, this is the fundamental problem, for his concluding section (pp. 153-155), recognizing that the "textual editor" may decide to produce a text representing any given stage in the history of a work, goes on to assert, "No misplaced confidence in inadequately based theory can justify his evasion of the problems of an empirical situation." But when one observes that Greg's approach is an attempt to confront the empirical realities involved in the reconstruction of a particular stage in the history of a work and that it does not proclaim the result to be the only useful text of the work (even for historical study), the issue Peckham raises is no longer an issue. It seems to me that Peckham's final description of "the task of the textual editor" is—after one has penetrated the vocabulary—accurate: But when he proceeds to say that there is no definitive version to be arrived at and "no one set of instructions" to follow, he is responding to a nonexistent argument. Much of Peckham's essay helpfully focuses on the nature of written language, and his suggestion that editors ought to be aware of the nonliterary uses to which their apparatuses can be put (as in a study of human behavior) is worth serious consideration; but as a critique of Greg's "Rationale" and the CEAA editions, it misses the mark.

During the following year (1972) there appeared two books with general-sounding, but somewhat misleading, titles, James Thorpe's Principles of Textual Criticism and Philip Gaskell's A New Introduction to Bibliography. Each raises some questions, either explicitly or implicitly, about the validity of CEAA procedures—questions which, by this time, seem very familar. Thorpe's most direct comment on the CEAA—a brief discussion of its Statement—is related to his underlying belief that textual criticism has become too bibliographical in approach and that bibliographers are trying to make textual criticism

a "science." Despite his assemblage of quotations intended to serve as background, these issues are in fact illusory: for the leading bibliographers over the years have recognized that textual criticism can never be mechanical and that bibliography is simply one tool among several useful in dealing with textual problems.[83] Thorpe paints a picture of bibliographers greedy to annex the whole "province of textual criticism," as he calls it; but whether the present emphasis of textual criticism is excessively bibliographical is a question that cannot be approached in general or theoretical terms but only in relation to the details of specific situations. After all, if bibliography offers one kind of evidence to the textual critic, he cannot sensibly say that he desires only so much, and no more, of that kind of evidence; but if his attention to those details causes him in a particular case to neglect his search for letters or documents or other kinds of external evidence, then obviously he can be criticized in that instance for undue concentration on one type of evidence. What Thorpe tries to argue, however, is that the CEAA Statement, by requiring attention to bibliographical details, implies such attention to be "the efficient cause of an ideal edition" (p. 72).[84] The Statement, he believes, reflects "the view of a text as a system of infinitely perfectible details, by which scrupulous attention to all details will ultimately yield ideal results" (p. 57). Although he does not wish "to suggest that meticulous care is pedantry" (p. 76), he does suggest that close analysis of what seem to be unimportant variants is a waste of time (e.g., p. 74). He does not acknowledge the fact that laborious collation of texts[85] and analysis of variants accomplish just as much when they demonstrate the absence of significant variants—or the presence of variants only in "the relatively trivial matters of spelling, punctuation, and capitalization" (p. 51)[86] —as when they show the existence of dramatically different readings. And no editor that I have heard of ever claimed that "scrupulous attention" to details and "meticulous care" are "a complete substitute for intelligence and common sense" (p. 78).[87] One must agree with Thorpe's later insistence (pp. 179-183) on the importance of a thorough knowledge of the author's works and period and of a diligent search for external evidence. But there is nothing inherent in the attention to bibliographical detail which prevents an editor from giving attention to other essential matters. The CEAA Statement does set forth the importance of accuracy in collating and proofreading, but it also points out the necessity for knowing the author's works and for searching out all relevant documents bearing on the history of a text[88] —and the CEAA editions have repeatedly been responsible for the uncovering of new documents and the assembling of comprehensive collections of reference material. An editor who neglects any part of his duty is open to criticism, and Thorpe's conclusion that editors should exploit "every kind of relevant evidence" (p. 79) is unexceptionable; but his belief that the "strongly bibliographical cast" (p. 103) of the CEAA Statement leads to a "glorification of method" (p. 79) rests on the fallacious assumption that attention to one kind of detail necessarily involves the neglect of other kinds. Some editors may of course be guilty of neglecting evidence, but it seems perverse to search for the cause of their incompetence in their careful attention to one kind of relevant detail.

A more consequential matter which Thorpe takes up is the treatment of accidentals (pp. 131-170).[89] After providing a sampling of statements from authors of various periods, stressing their indifference to accidentals, and a historical survey of printers' manuals, suggesting that printers over the years have felt an obligation to "correct" accidentals, Thorpe concludes that "probably in most cases" the author "expected the printer to perfect his accidentals" and that therefore "the changes introduced by the printer can be properly thought of as fulfilling the writer's intentions" (p. 165). It seems to me that there are two basic difficulties with Thorpe's position. The first is that quotations from authors and from printers' manuals are not comparable, because the former are statements of personal opinion (often prompted by specific situations), while the latter are public announcements of recommended general practice. Thus Thorpe's evidence from the printers' manuals[90] is sufficient to show that printers have widely regarded the alteration of accidentals in copy as part of their function; but his evidence from individual writers by no means can be generalized upon to suggest that in any given instance the chances favor an author's having been indifferent to the handling of accidentals. The conclusion would seem to follow—contrary to Thorpe—that, without convincing evidence on the other side, an author's manuscript stands a better chance of reflecting his wishes in accidentals than does a printed text. Here the second difficulty arises—in Thorpe's conception of an author's "intention." In his opening chapter—his well-known essay on "The Aesthetics of Textual Criticism"[91] —he asserts, "While the author cannot dictate the meaning of the text, he certainly has final authority over which words constitute the text of his literary work" (p. 10). As a result of this distinction between "meaning" and "words," Thorpe tends to accept at face value an author's statement about wording, without focusing on the fact that the motivations influencing such a statement may be just as complex as those lying behind a statement of intended "meaning." Although he recognizes that, in the absence of an authorial statement, the intended wording must be arrived at through a critical analysis of all available evidence (p. 193),

at various points he implies that the existence of a statement settles the matter—as when he says that "the personal testimony by the author as to his intentions is plainly the most primary textual evidence that there can be" (p. 109).[92] This point of view leads to an uncritical acceptance of an author's remarks about his indifference to accidentals (or his preference for those in the printed text). The upshot of Thorpe's discussion is his astonishing recommendation that "the editor will do best to spend only a modest amount of his time on accidentals—mainly a losing cause—and devote himself to matters of substance" (p. 168). It is difficult to reconcile Thorpe's readiness to believe that an author preferred the printer's accidentals with his strict view of "the integrity of the work of art" (pp. 14-32); and it is hard to see how an editor whose aim is to establish the author's intended text, in accidentals as well as in substantives,[93] can justify the decision in advance to spend a "modest" amount of time on the accidentals. Like the earlier discussion of bibliographical detail, this chapter on accidentals reflects a peculiar view of scholarly endeavor: it suggests, in effect, that a scholar's sense of perspective is shown less by his ability to evaluate and integrate data than by his prior decision to limit his consideration of certain clearly relevant areas.[94]

If Thorpe's book, weakened by such contradictions, does not manage to serve the useful function of fairly surveying "the basic principles which underlie the practice of textual criticism" (p. vii), neither does Philip Gaskell's chapter on "Textual Bibliography" (pp. 336-360) in A New Introduction to Bibliography provide the kind of basic summary of current thinking which one might expect of an "introduction." Although his exposition of "Copy-Text" (pp. 338-343) does not specifically mention Greg's rationale, he does provide an accurate statement of its general application, with one important exception.[95] He is unwilling to push that rationale to its logical conclusion and recognize that a fair-copy manuscript, when it survives, becomes the copy-text, except when there is convincing evidence pointing toward the first (or some later) edition as the proper choice.[96] His argument rests on the same assumption as Thorpe's:

He concludes that it "would normally be wrong, therefore, rigidly[97] to follow the accidentals of the manuscript, which the author would himself have been prepared—or might have preferred—to discard" and asserts that "in most cases the editor will choose as copy-text an early printed edition, not the manuscript" (p. 340). Later, he reiterates that "the manuscript if it survives, will be consulted but will not be followed in accidentals unless the compositor appears to have misrepresented the author's intentions" (p. 358). Although he allows for situations in which the manuscript is the proper choice, he places the presumption of authority with the first printed text. I have already commented on the difficulties of maintaining such a position, but I should perhaps call attention to the way in which Gaskell's wording itself reveals some of them. To say that an author is "implicitly endorsing" the accidentals of the proofs when he lets them stand is not at all the same as to say that he prefers them, and it ignores the economic (and other) factors which may have influenced his decision; similarly, to believe that an author "would himself have been prepared . . . to discard" the accidentals of his manuscript is not the same as to believe that he wished to discard them, and it surely does not give an editor license to carry out that discarding. Gaskell asserts that the accidentals of a first edition, despite "the process of normalization carried out in the printing house," will "still be closer both to the text that the author wanted, and to the reading of his manuscript, than the altered accidentals of the second and third editions" (p. 340). This statement is true, but the reference to manuscript readings undermines the general argument: if there is any desirability in having the accidentals resemble those of the manuscript, then the manuscript ought to be chosen for copy-text in the first place; on the other hand, if the author's preference is for the first-edition readings, then the manuscript is irrelevant in this context. Gaskell raises further doubts in the reader's mind by citing the example of Thomas Hardy, who, "in revising his printed texts for new editions, appears to have changed the normalized accidentals back to the forms of the original manuscript" (p. 342). Even though Hardy may not be a typical case, his revision illustrates the point that a writer may acquiesce in printing- or publishing-house styling without preferring it. Is not the more reasonable approach, then, to presume, until contrary evidence is adduced, that a manuscript reflects the author's intentions in accidentals, rather than to begin with the presumption that it does not?

A further confusion is introduced by the argument that an author's accidentals may stand in need of correction. Gaskell doubts "whether it is worth preserving thoroughly bad punctuation just because it is the author's" (p. 358) and later advises, "Let us carry out the author's intentions wherever we can, but not to the extent of taking pride in reproducing the manifest inadequacies of his accidentals" (p. 359). But punctuation which seems "bad" to the editor may have seemed appropriate to the author; and if the editor's aim is to preserve what the author wrote, rather than his own "improvements" upon it, he cannot very well say that he will pursue the author's intention only up to a point, and no farther. Gaskell's belief that "an editor may reasonably aim at consistency in his final version" (p. 358) suggests that he is thinking of a modernized text,[98] although most of his comments seem to be concerned with editions that aim to recover the author's intentions. At any rate, it is true that his discussion never focuses on the differences in purpose between modernized and unmodernized texts:

The illogic of this passage results from the fact that two kinds of editions are being talked about simultaneously. Since "unsatisfactory" accidentals may not be "in contravention of the author's usual practice," the editor is being instructed here to emend only those "unsatisfactory" accidentals which are not characteristic of the author, thus producing a partially regularized, but not modernized, text. But when he has done that, he does not still have open to him the option of emending "according to the conventions . . . of his own" period—that is, of modernizing. After all, there would be no point in selecting the readings in need of emendation on the basis of whether they are uncharacteristic of the author and then to emend them on the basis of present-day practice. Either the editor decides to establish, as accurately as he can, the author's own accidentals; or he decides to make all the accidentals conform to the practice of his own time. The fact that the former approach is necessary for scholarly (that is, historical) study does not, of course, mean that there may not be occasions on which the latter is more appropriate. But the failure to distinguish carefully between the two cannot lead to clear thinking about editorial problems. It is unfortunate that Gaskell's discussion gives the impression of describing (as one would expect an "introduction" to describe) current generally accepted practice; beginners who turn to it for guidance will be puzzled and misled.[99]

This account of the CEAA's application of Greg's rationale to American literature and of the critical reaction to it suggests several observations. To begin with, one must recognize that, when Wilson expressed surprise at the "violence and venom" of the correspondence

aroused by his article, he was calling attention to characteristics which have unfortunately been manifested by a number of persons in this debate, on both sides. It may be gratifying to some editors to find that people care enough about editing to become emotionally involved in theoretical discussions, but scholarship is not advanced by arguments which rest on preconceptions or vested interests or clashes of personality. There can be no doubt that some of this debate has sunk to that level, and the opponents of CEAA policy are not the only ones at fault. What is particularly unfortunate is that so much time and energy has been poured into arguments about superficial or nonexistent issues, when there are so many issues of importance that remain to be considered. The belief that bibliographical and textual work is not humanistic simply because it tries to establish facts or utilizes mechanical aids—and that those engaged in it therefore do not really care about literature—is obviously an emotional rallying cry, not a proposition to be seriously entertained. Similarly, the view that editors who follow Greg are engaged in a mysterious, complex procedure with an elaborate, arcane terminology can only be regarded as an invention of those who are temperamentally disinclined to perform editorial work, for it would be uncharitable to believe that they actually find these concepts and terms a strain on the intelligence. As emotional reactions, these attitudes are understandable, and proponents of Greg's theory have sometimes done their part to provoke them; but as intellectual arguments, there is simply nothing to them.

I am not suggesting that the entire controversy has been frivolous; but even the more serious arguments have so often resulted from a misunderstanding of what is really an uncomplicated approach that one is puzzled to account for them in any but emotional terms. Neither am I saying that Greg's theory and the CEAA application of it ought not to be criticized and analyzed, for any serious intellectual position can only benefit from thoughtful constructive criticism. The point, indeed, is that there has been too little—scarcely any—of this kind of criticism. Yet much of fundamental importance remains to be thought about. The question of what is meant by authorial intention, of how that intention affects the treatment of punctuation, of what differences may be required in working with a typescript rather than a holograph manuscript—such matters as these, when disentangled from self-serving attacks on or defenses of particular editions, need more discussion. Now that a considerable interest in editorial matters has been aroused, a great deal can be accomplished if the collective effort of those interested is expended constructively. No one pretends to have solved all

the problems, but solutions can best be approached by a positive effort to understand what accomplishments have so far been achieved and to build on them. Presumably all readers are interested in seeing reliable texts of American literary works made widely available; it would be unfortunate if those who share a common goal allowed themselves to be diverted by controversy from keeping that goal at the center of their attention and working together to attain it.[100]