University of Virginia Library

[Textual] Criticism and Deconstruction
by
D. C. Greetham

Deconstruction just happens:" thus Jacques Derrida at his annual City University of New York lecture in September, 1989. His point was to distinguish "deconstruction" (a philosophical disposition or a critical practice) from "deconstruction-ism" (an institutionalised or scholarly movement, and only one of the several -isms of contemporary American "Theory," which word Derrida now regards as a proper noun in its current American usage, and not as a "theory" in the conventional scientific sense). The distinction between deconstruction and deconstructionism may be very pertinent to any discussion of the nature and role of deconstruction in present-day scholarship, for the academic and political power of deconstructionism may have given a bad name to the actual critical practice of deconstruction, whether it be in literary or in textual criticism, a practice which is addressed in G. Thomas Tanselle's recent article "Textual Criticism and Deconstruction,"[1] to which this present essay is in part a response.

But if deconstruction "just happens," what exactly does it do, and how may we recognise it in action? Is it in fact really much different from many of the critical practices we are already familiar with? Is it indeed (as some of its detractors have suggested) no more than yet another, this time decadent and anti-humanist, survival of formalism, whose last incarnation was the "old" New Criticism? Such speculations cannot be fully addressed here—in part because my concerns are more textual than they are literary (in the narrow sense of that term)—but I raise them to emphasise that deconstruction is not a practice that was "invented" at the École Normale Supérieure by Derrida, then brought to this country by the "hermeneutical mafia" at Yale. As one of its major practitioners, J. Hillis Miller, insists: "The present-day procedure of 'deconstruction,' of which Nietzsche is one of the patrons, is not . . . new in our own day.

It has been repeated regularly in one form or another in all the centuries since the Greek Sophists and rhetoricians, since in fact Plato himself, who in The Sophist has enclosed his own self-deconstruction within the canon of his own writing."[2] Rather, deconstruction is an attitude towards the apparent structures embedded in works (and texts), and an attempt to interrogate those structures, initially by inverting the hierarchies which the structures represent. It thus brings to literary criticism a "suspicion" of texts similar to that long endorsed by most practising textual critics. Let us not forget that Eugene Vinaver, whose major editorial work on Malory was done before any formal "deconstructive" movement, insisted that "textual criticism implies a mistrust of texts."[3] And twenty-five years ago, Fredson Bowers was warning critics against the "sleazy editing" of the great English and American classics, and suggesting that a Vinaver-type suspicion would be salutary to critical awareness. In a justly-celebrated passage in Textual and Literary Criticism, he roundly condemned the passivity and gullibility of the literary establishment: "[I]t is still a current oddity that many a literary critic has investigated the past ownership of and mechanical condition of his second-hand automobile, or the pedigree and training of his dog, more thoroughly than he has looked into the qualifications of the text on which his critical theories rest."[4] Bowers was then calling for what has become known in critical circles as a "hermeneutics of suspicion," of which textual criticism and deconstruction are, in their different ways, exemplary. In one sense, therefore, the deconstructors have simply joined the textual party rather late.

However, it is only fair to acknowledge that the sort of suspicion or mistrust brought to literary criticism by deconstruction does not at first appear to be at all related to the critical attitude of the philological tradition. Theirs, the deconstructors', seems to be whimsical, idiosyncratic, undisciplined, destructive, and badly written; ours, the philologists', is structured, objective, disciplined, constructive, and closely argued. Theirs seeks to fragment texts and to prevent their ever becoming works by exposing their inevitable aporia (their "central knot of indeterminacy"); ours seeks to make works out of the corruptions of texts, and (at least in the eclectic tradition) to create "texts that never were," which ideally should be consistent and determinate—or such is the most frequently-endorsed editorial aim. How can textual criticism as we have

usually defined that enterprise share any ideals or practices with deconstruction, despite their similar reliance upon suspicion?

The first step in answering that large question is to show deconstruction in action, and I begin by pointing to two well-known critical acts that may be regarded as deconstructive—one textual and the other non-textual. The non-textual first: when Freud began formulating his theories of psychoanalysis, he inherited a series of structures arranged in neat hierarchical disposition—conscious/unconscious, presence/absence, spoken/unspoken, wakefulness/dreams, etc., with the first term in each pair being conceived as the primary, more valid, more reliable, or more important of the two, and with the second being defined by its distance or separation from, or negation of, the first. In each case, the effect of Freud's theories was to invert the inherited hierarchical structure, by, for example, declaring that it was the unconscious which controlled the conscious and not the other way round, and that access to the "real" nature of a psychical problem could be gained through "slips" or "lapses," and not through intended expression. Freud was "deconstructing" these hierarchies, setting them on their head, and demonstrating that the fixed, reliable relationship we had got used to was in fact unfixed and unreliable. Similarly, in the textual sphere, when Greg wrote his famous article on copy-text,[5] he had inherited a series of textual structures arranged in hierarchical disposition, the most significant of which were content/surface and early/late. The copy-text article was quite literally "revolutionary" because it turned these terms and concepts around—positing that editors interested in recovering authorial intention should be concerned more about the characteristic surface features (accidentals) of a text than the content (substantives) in selecting a copy-text, and that therefore they would probably do better to choose an early text showing these putative authorial forms rather than the latest, most "finished" (in both senses) of states as copy-text. Like Freud, Greg was "deconstructing" the received wisdom of textual criticism by inverting the received hierarchies.

If this is all deconstruction is, it may all seem rather tame, but it does not end there. The real reason for the deconstructive inversion of hierarchies is not to create other (inverted) hierarchies in their place, which will become the new received wisdom, but to question each new hierarchy as it appears, with the effect of for ever delaying or denying a fixed, reliable relationship. Deconstruction's prime philosophical agenda is thus to interrogate and repudiate any appeal to a "metaphysics of presence,"

any claim that would ground a system of difference in such a fixed and unchallengeable authority. This process of continual interrogation is usually called "différance", a play (in French) on the terms for "differ" and "defer," and, given Derrida's concern for inverting the priority of speech over writing, a play which can only be seen (in writing) and not heard (in speech), for in French there is no phonemic difference between the e and a of "différence"/"différance." This play (a characteristic Derridean jeu) is a post-structuralist response to the structuralist arrangement of pairs of concepts into positives and negatives based on their "difference" (a concept derived through structuralism from Saussure's "phonemic"—i.e., differential—analysis of language, whereby each phoneme is recognised as potentially meaning-ful by speakers of a language by the fact that is not another phoneme, but is "different"). Thus "cat" is heard as different from "mat" because [c] and [m] are phonemes in English. The most important pair in this differential calculus of language was that composite of the "sign," made up of the "signifier" (the actual form, the surface features, of the word) and the "signified" (the concept or substance to which this form referred), with the signified as the more "real," the more "transcendental" of the two. As can be readily seen, this distinction, especially in the privilege given to the signified, is very similar to the "substantives"/"accidentals" difference which Greg inherited and inverted, just as later linguists (and psychologists) were to invert Saussure. If, therefore, these differences are not fixed but may be challenged—even if they seem to embody such apparently permanent truths as substance and accidence—then it may be that we will never arrive at a definitive "system" which can support a universally relevant series of hierarchical relationships. There will never be a permanent "metaphysics of presence"—and it is this persuasion that deconstruction insists upon.

But does the practice of deconstruction "just happen" (to revert to Derrida's assumption)? Consider briefly the two examples already cited—Freud and Greg. One of the hierarchies endorsed by Freud is neurotic/psychotic, with the former being adopted as the "norm" whereby the latter's "difference" is to be measured. In Deleuze and Guattari's Anti-Oedipus,[6] this hierarchy is indeed challenged—and inverted, with the neurotic replaced by the psychotic as the norm in their deconstructive system of psychoanalysis, so-called "schizoanalysis." Similarly, the male/female pair in Freud (with the female being seen as a defective "version" of the male—especially genitally) has been challenged and inverted by

some post-Freudian feminists, with the male now seen as a (genital) "version" of the female.[7] On the "textual" side, we can observe the same process of continual challenge, for while Gregian copy-text theory and the related concepts of eclecticism and intentionality have had a long and successful influence on the production of scholarly editions in the English-speaking world, some of the hierarchies of that theory have indeed been challenged and inverted. From within intentionalism, Hershel Parker has attempted to overturn the consistent/inconsistent and the whole/part pairs, declaring that Greg's rationale is "too rational" to allow "full intentionality" to be presented.[8] And from a position somewhere on the margins between intentionality and sociology, Jerome J. McGann has challenged the early/late, original/copy, author/reader, single/multiple pairs of Greg-Bowers textual criticism, preferring to regard the authorial "originary moment" of composition as only one among several in the full ontology of the work, and to deny any special privilege to this early, original, single entity.[9] No doubt other challenges will come to upset the new hierarchies created by Parker and McGann.

So deconstruction may "just happen," as Derrida suggests, and it does not need "deconstruction-ism" or deconstructors to make it happen. It is happening all around us: indeed, as I will suggest here, it has happened in the article by Tanselle which is the raison d'être of this one, and it will happen in the concluding section of my essay. It is perhaps fortunate that this is so, for it may be easier to "show and tell" about something actually occurring before our eyes as we read than it is to talk in abstractions about concepts which may appear remote from and inimical to the interests of textual critics.

And so at last I turn to "Textual Criticism and Deconstruction," to Tanselle v. Hartman et al., and to my own typographically peculiar title, "[Textual] Criticism and Deconstruction." In dealing with the function and form of such titles, I will thus have to concentrate on the "surface features" of the text—brackets, word order, and the like—and I am sure that an audience of textual critics brought up in a Gregian or post-Gregian dispensation will not have to be convinced that such concentration is quite proper in an article on textual criticism, and will appreciate that these "accidentals" are potentially very significant, for they may embody much of the "substance" of a critical argument. It is, rather

appropriately, a concern shared by textuists and deconstructors (as the play on "différance" has already shown).

Titles, and the specific form of titles particularly, are thus very important to the deconstructors, as Tanselle notes (p. 13) in his comments on the "double" title (reflecting the similarly "double" essay) in Derrida's "Living On. Border Lines," which emulates the appearance of a typical textual page and apparatus/commentary, with continual reference back and forth between the top and bottom of the page, over the "border" separating the two. This concept of border, a thin line of difference dividing one idea from its opposite or one function from another, is central to the structuralist ethic from which deconstruction derives, and is obviously important (as Derrida's parody of the textual edition shows) to our own discipline. We like to know which is which—text or apparatus. The difference that lies at the heart of the identity of the phoneme in structuralist linguistics (cat/mat etc.) is thus redolent not only in structuralist anthropology and structuralist poetics, but in the very concept of the lemma and variant so familiar in the apparatus of critical editions. The lemma (reflecting the accepted lection in the edited text) is "correct," "authoritative," "true," "sincere," "original," whereas the variants listed after the lemma are "incorrect," "unauthoritative," "false," "corrupt," or "derived." This difference is usually exemplified in the apparatus by the half square bracket]. Elsewhere, the prevalent typographical device of difference is the slash /, an increasingly common method to show this play of difference in titles,[10] and perhaps best-known in the title of Barthes' famous S/Z, a late-structuralist phonemic analysis of Balzac's novella Sarrasine. The play of difference, however, while setting up apparent bipolar oppositions of form and value, inevitably acknowledges the potential similarities between the two opposed phonemes or concepts (they are perceived as different in a particular language or cultural system precisely because they could possibly be perceived as identical in another), and the textual critic's phonemic use of] in the apparatus has a similar ambivalence. The "rejected" readings have at some time been accepted by some readers (scribes or compositors) as part of the text, and their status as variants is therefore only ever contingent, just as is the status of the accepted reading in the lemma itself. I am reminded, for example, of E. Talbot Donaldson's witty (if somewhat

sexist) analysis[11] of the editorial acceptance of readings as rather like choosing a wife, with whom one might live happily enough for some time before becoming discontented and seeking out an alternative (reading and spouse) from among those that had originally been rejected. The border line of the lemma square bracket, like all such border lines (including marriage in Donaldson's analogy) is therefore temporary and contingent, and the lemma is thus "bracketed" (quite literally), in the manner that the phenomenologist Husserl suggested that all concepts must be while they were under debate. Thus the brackets around [Textual] in my title, and thus this digression on phonemic typography, for it is precisely this problem (what does "textual" mean and what is a text?) that is at stake in Tanselle's interrogation of the deconstructors. My title thus co-opts Tanselle's but acknowledges that we will need to bracket [text] and [textual] in the discussion rather than simply assume that we already know what these terms mean before the debate begins. As I shall show later, it is precisely because Tanselle assumes one specific "phonemic" difference between "text" and "work" (and no other) that his analysis of Deconstruction and Criticism is a formal deconstruction of it, a reversal of its hierarchies.

Tanselle's title ("Textual Criticism and Deconstruction") is itself a play on the title of the work (or text?) he is examining, for he reverses the ordinal value of Hartman et al.'s Deconstruction and Criticism by making "deconstruction" the second term in the title, to be read against the grain of the first, now amplified as Textual Criticism. A small matter, but one emblematic of the primacy of the structuralist relationship in such pairings—the "violent hierarchy" which deconstruction seeks to undo. It is no accident that the Bowers volume already quoted from is called Textual and Literary Criticism (not the other way 'round), for he insists throughout that it is the literary critics' duty to be aware of, and corrected by, the findings of textual criticism before the literary enterprise can even be begun. Empowerment is thus reversed in Tanselle's title, and he has thus committed a deconstructive criticism of Deconstruction and Criticism before even beginning his analysis, for he has inverted the values of the deconstructors, turned the "violent hierarchy" on its head, deconstructed the deconstructors. It is my contention that his entire essay is precisely this—a deconstruction of deconstruction—and that this begins at the beginning, with the co-option of the title.[12]

There is one further typographic idiosyncrasy I must comment on (rather than employ) before showing Tanselle's deconstruction in action. When planning this essay, I had hoped not only to be able to "bracket" [textual] criticism, but to place it sous rature, or "under erasure," by striking through the term with a mark of erasure while still leaving it legible. This device, familiar to textual critics in the textual page diacritics and marks of erasure used in some genetic editions, and especially in transcripts of heavily-reworked manuscripts, is the usual deconstructive manner of showing that the concept under discussion (especially "philosophy") is still present in its "traces" while being absent from the terms of the current debate: the technique is, in other words, identical in signification and in form to the textual critic's marking the "present absence" of a rejected, earlier reading in the evolution of a genetic text, legible only in its orthographic "traces." It is again an example of a deconstructive method that has been predicted by perfectly orthodox textual techniques. Unfortunately, while rhetorically desirable, the sous rature mark was not possible in the normal letterpress of Studies in Bibliography, and I had to settle for the phenomenological bracket—enough to make the basic point.

I will now briefly indicate how Tanselle's commentary on Deconstruction and Criticism is itself a deconstruction of that book—in the various ways I have already alluded to. Tanselle's frequent quarrel with the deconstructors appears to be that they have appropriated, or perhaps misappropriated, his very discourse, and specifically the distinction he wishes to make between two terms, "text" and "work," and the different conceptions those two terms may represent.[13] And like Tanselle, I believe

that there is a useful distinction to be made here, but a distinction which is useful to textual criticism, not necessarily to criticism or deconstruction, and not really permanent or unchallengeable. Tanselle's book A Rationale of Textual Criticism (1989) is in large part concerned with offering a wide-ranging and highly sophisticated interdisciplinary discussion of the various textual ontologies of different disciplines—literature, music, painting, sculpture, film, and so on. To this end, the distinction he posits between "text" and "work" in this book is very apt: the former is the concrete, specific embodiment (document or other physical state) of the conceptual entity "work." As Tanselle quite properly notes throughout his Rationale, in some disciplines (e.g., painting and sculpture) the two may compete for the same space, for there is no separate "work" beyond the singularity of its concrete manifestation or "text"; but in other disciplines (e.g., literature), the "work" may be finally unknowable, precisely because the document/text is not necessarily identifiable (according to his division of the terms) with this conceptual "work" lying behind it and is certainly not to be confused with it.

However, this view of the nature and limitation of "text" and "work" is useful (and accurate) only if one shares the assumption that the physical manifestation (text) is not identical with the conceptual entity (work) in disciplines like literature. As Jerome McGann insisted in response[14] to

a recent essay in which I had posited this difference,[15] the apparent clarity of Tanselle's orderly distinction does not hold if (like McGann) one holds that the concrete is not only the way in which we may know the work but is the work itself. Within the body of the textual discipline, McGann would therefore offer another system of differences for these terms, and not necessarily that suggested by Tanselle as a defence and exemplification of what McGann sees as an untenable "idealist" position. But this disagreement over terms among the textual critics is not my concern here, but rather the effect of the terminological distinction in Tanselle's "reading" of Deconstruction and Criticism, for throughout his "Textual Criticism and Deconstruction" essay he reads Hartman et al. "against the grain" of his own book the Rationale, continually upbraiding the deconstructors for having failed to make the same distinction that motivated the Rationale. Tanselle is very open about this technique, directing the reader early in his essay (note 2) to the Rationale for a discussion of the issues "at greater length." In this way, he reverses the priorities of the two works (i.e., deconstructs the first through the second), moving the "text"/"work" distinction of the Rationale into the centre of the Deconstruction collection, and marginalising the other concerns of that volume. This is a "classic" deconstructive mode, and Tanselle does it with considerable facility. For example, he chides Hartman for using "the word 'text' to mean 'work,'" (p. 2) asks "are we to conclude that he is using 'text' as a synonym for 'poem' and 'work'?" (p. 3) and, most tellingly, insists that his "text"/"work" distinction is "nevertheless a central one" (p. 8) despite the deconstructors' misprision of the terms. Perhaps equally telling for the displacement of one book by another is his comment that "A place does occur . . . where a distinction between 'text' and 'poem' is made, but it is not the one I am making" (p. 6; emphasis added).

The same dissatisfaction, and the same determination to move "text"/"work" to the centre of Deconstruction and Criticism, is repeated throughout Tanselle's essay and is much too frequent a charge to be chronicled fully here. Just a few examples will show the technique: "how we define 'texts' is the crux of the matter" (p. 19; "his equating of versions of works with the texts that survive in documents" (p. 10); "It is remarkable that a group of critics so preoccupied with theoretical matters should have failed to include in their discussions some recognition of the difference between the stationary arts . . . and the sequential arts" (p. 29). In these and many other such comments throughout the "Textual Criticism and Deconstruction" essay, Tanselle places his perfectly

valid "phonemic" difference of "text" and "work" at the centre of Deconstruction and Criticism and judges the values of that book by its failure to codify its arguments around the same phonemic difference. Deconstruction and Criticism is deconstructed (figuratively disassembled into a series of inconsistent utterances, inadequate because of their cavalier ignoring of the "text"/"work" distinctions of the Rationale) in the construction of "Textual Criticism and Deconstruction."

These recognised terminological inadequacies in Deconstruction and Criticism lead Tanselle to a second type of deconstruction—the exposure of the aporia or "fundamental knot of indeterminacy" in the collection. This can be done by a reader-response demonstration of a reader's problematic encounter with the text, as witness the following: "the reader will have assumed that these matters are to be pursued as the essay develops. But . . ." (p. 3); "One thinks, at first, that Bloom is going to make this point. . . . Immediately, however . . ." (p. 5); "Hartman's vagueness here will be noted only by those who have thought about the difference between texts of documents and texts of works" (p. 20). In such cases, Tanselle acts as a Fishian innocent reader, continually surprised and jolted by the twists and turns of the writer's utterance, and shows how such a reader's expectations (expectations founded upon an acceptance of a specific definition of "text" and "work") are constantly disappointed by the actual expression of the authors of Deconstruction and Criticism.

The exposure of inconsistency can also be done by a direct analysis of the aporia itself, rather than by acting as innocent reader, as witness the following: "But the confusion of the opening sentence undercuts everything else" (p. 8); "he is using a metaphor that undercuts what he wishes to say" (p. 17); "Yet he seems to see no awkwardness in letting 'texts' also mean the 'scriptures' or 'poems'" (p. 22); "This oversight causes him to confuse . . . and, as a result, to write an unpersuasive brief for deconstruction" (p. 28). This method (particularly the last swipe, implying that deconstruction has been ill-served by the deconstructors) is again a perfectly acceptable deconstructive mode, for it does not claim to deconstruct, but merely to demonstrate how deconstruction has already happened in the body of the text. As Hillis Miller has remarked, "Deconstruction is not a dismantling of the structure of a text but a demonstration that it has already dismantled itself."[16] And that is what Tanselle achieves—a demonstration that because of the logical and terminological inadequacies of Deconstruction and Criticism, that book has already "dismantled itself."

A third (and related) way in which Tanselle deconstructs is to use

the dual dissatisfaction of failed expression and reader unease as grounds for an actual re-writing of the text of Deconstruction and Criticism where it is perceived to be lacking or defective. This editorial desire is quite proper, of course, but it is also an instance of Bloom's doctrine of "strong misprision" (the desire of the belated poet/critic to rewrite the precursor in order to subdue him). Tanselle's rewriting and strong misprision ranges from the optative—a desire that x should have done so-and-so—to an actual reconstruction of, and "correction" of the precursor. Witness the following: "If he had pursued this question, he might have . . ." (p. 9); "But what he might also have said is . . ." (p. 4); "How much richer the passage would be if it resonated with . . ." (p. 6); "The point that I believe Derrida wishes to make could be more coherently expressed as follows . . ." (p. 15); "Hartman's account . . . cannot/but seem thin compared to what it would have been if . . . had been a part of it" (pp. 22-23); "This question would perhaps be more interesting than the one he has chosen to discuss" (p. 23); "the sentence . . . is not well phrased (it could begin . . .)"; and here, Tanselle does actually rewrite de Man's sentence [n14], as he does Derrida's on p. 15. Criticisms of omission are, of course, the common stock of reviewing, but Tanselle's consistent rewriting of the text of Deconstruction and Criticism is more than this—it is a co-option of that text, and (again) a deconstruction of it.

And finally, Tanselle deconstructs the text of Deconstruction and Criticism by "playing" with the text, by indulging in a Derridean jeu, a fanciful and amusing commentary on it according to the principles of his own discipline—textual criticism and editing. For example, he includes a humorous note on a possible textual "error" in de Man's essay: "I have placed 'like' in brackets because the word in the printed text is 'line.' It seems unlikely that de Man meant to say 'Is the status of a text line the status of a statue?', for a 'text line' would seem to mean a unit or building block of a text and would therefore not be parallel with 'statue,' a whole work. The matter must remain uncertain, however—as, indeed, the constitution of all texts of works is uncertain. This typographical error, if it is that, illustrates the necessity for deciding on the makeup of the text as a part of the act of reading. It also shows how documents can be unreliable witnesses to past intentions, even though we can never know with certainty the precise extent of that unreliability" (n16).

This is both salutary, and in its irony, very funny, a play with the text reminiscent of the elaborate textuality of Nabokov's Pale Fire. And entering further into the spirit of this editorial play on the text in the deconstructive essay, one could perhaps suggest that, according to the classical doctrine of lectio difficilior probior est ("the more difficult reading is the more moral"), the "text line" reading, because of its seeming

opacity (but not complete implausibility) is more likely to be authorial than compositorial.

Another amusing example of such a Derridean jeu is Tanselle's foot-note 29, on a Miller misquotation. Commenting on the source of the word "uncanniest," Tanselle conducts a (mock?)-bibliographical analysis of the possible (though unlikely) discrepancies in the various editions of translations of Heidegger, concluding with the speculation that "it seems unlikely that the copy Miller used had a different reading. Perhaps, instead, the change from 'strangest' to 'uncanniest' in Miller's quotation somehow springs from his knowing the association of 'uncanny' and 'unheimlich' in translations of and commentaries on Freud . . ." (n29). It is a beautifully articulated jeu, especially in its invocation of a Freudian "slip" on Miller's part, dependent on Miller's putative reading of a Freudian text. It is a textual game well played, and is another deconstructive turn against the text of Deconstruction and Criticism.

Thus, in these several ways, Tanselle's reading of Deconstruction and Criticism becomes an act of deconstruction itself. And most importantly for this present discussion, in his placing of his interpretation of "text" and "work" at the centre of his reading of Deconstruction and Criticism, Tanselle has changed and inverted the priorities of that book by reading it against the grain of his own work A Rationale of Textual Criticism. The relationship between Deconstruction and Criticism and "Textual Criticism and Deconstruction" via A Rationale of Textual Criticism is thus simply another example of Derrida's dictum that "deconstruction just happens" and is, moreover, a demonstration of Derrida's insistence that "One text reads another. . . . Each 'text' is a machine with multiple reading heads for other texts" (p 107).

One might argue that these methods used by Tanselle in his long review of Deconstruction and Criticism are really no different from the traditional criticism practised in such reviews. The reviewer's responsibility is to point out the omissions and obfuscations, the inadequacies and the failures of the book under review. It may even be within the reviewer's brief to suggest ways in which the book could be improved by rewriting, even perhaps to do some of that rewriting oneself. And, of course, we must expect that the reviewer's own work and persuasions will act as a de facto criticism of the infelicities discovered and described. All of this happens in Tanselle's essay, and the easy answer to the demurral that such criticism is not specifically (or only) deconstructive might simply be to accept Miller's claim that deconstruction is "analytic criticism" as such, (p. 252) and to assume therefore that all such analysis is deconstructive. However, "Textual Criticism and Deconstruction" is all this and more: it is not only analytical, but it is querulous of the

oppositional hierarchies (e.g., between literature and criticism, or between critic and text) set up in Deconstruction and Criticism, and insistent upon the importation of a rival set of hierarchies as the means of criticism; further, in its dismantling of the logic and coherence (if such they be) of the book and in its shift in empowerment (judging the effectiveness of a book of criticism by its conformity to another source of critical power, and ironically playing with and emasculating the text by employing the techniques of another discipline), Tanselle's essay moves beyond the generally-accepted limits of "analytic criticism" and becomes a witty, careful, and very effective piece of deconstructive criticism. The deconstructors hoist with their own petard.

I hope it is therefore clear that I am not supposing that there is anything malicious, untoward, or unheimlich in Tanselle's having [mis]read Deconstruction and Criticism. On the contrary, as I have already suggested, Bloom's theory of "strong misprision" would doubtless commend such a [mis]reading, and Miller's insistence that deconstructive readings merely expose the aporia that is already there, together with Derrida's assurance that "deconstruction just happens" would both accept Tanselle's using his book to read theirs as a quite proper interpretative stance. However, the next stage of the deconstructive enterprise is to show that such a [mis]reading does not arrest the [re]interpretation, that another [mis]reading will produce another set of reversed priorities—in other words, that there are no "metaphysics of presence" or "definitive editions" of works, editions that will never have to be revised, editions that contain in text and notes a permanent, fixed, hierarchical arrangement of the values inscribed therein. For just as Greg could deconstruct earlier belletristic textual criticism, and just as McGann et al. can then deconstruct Greg, so Tanselle's re-editing (for such it is, in form as well as intention) of Deconstruction and Criticism does not close the process but invites further re-reading.

This is my final task in the showing of deconstruction in action. Can I offer another re-reading of Deconstruction and Criticism that will deconstruct Tanselle's deconstruction, just as one edition re-edits another? If I were to reverse Tanselle's hierarchies, and look for congruence rather than difference, common cause rather than dissension, between the deconstructors and the textual critics, what did Tanselle's reading miss that might be of value to a textual critic looking to Deconstruction and Criticism for evidence of such similarities between the two disciplines rather than the differences? This is a part of the project to "domesticise" deconstruction, and to demonstrate that, with a shift in perspective, a quite different valuation may be put on the aims and practices of the deconstructors. I now offer such a consensual reading—in the order

used by Tanselle, and by the authors of Deconstruction and Criticism.

Harold Bloom's essay "The Breaking of Form" is concerned with the function of form as an interpretative device, in charting "our awareness, however precarious, that the sequence of parts is only another trope for form. Form, in poetry, ceases to be trope only when it becomes topos, only when it is revealed as a place of intervention" (p. 2). These places of intervention are perceived by Bloom as critical or creative, but in both of these senses they are inevitably a part of the textual (and particularly the editorial) act and prerogative. For just as textual commentary has been seen as a series of interventions, even "guerrilla raids" upon a text,[17] so the breaking of the form of a text and the reconstitution of it in another form is the strongest imaginable intervention and a perfect exemplification of Bloom's "misprision." All editors, best-text or eclectic, are healthily guilty of this misprision of documents, a suspicion that their evidence is somehow suspect and must be remade according to a different ethic (authorial rather than scribal, shall we say?). Form is thus the means whereby the editor asserts that there has indeed been an intervention, and that form is indeed no longer simply a trope but has become the very means and subject of the editorial desire—its topos, repeated, as such topoi inevitably are (if they are to become topoi, recurrent motifs), over and over again. As editors, we therefore must agree, with Bloom, that "innocence of reading is a pretty myth" (p. 6) and that "critical reading [i.e. critical editing, which is the concrete manifestation of critical reading] aspiring toward strength must be as transgressive as it is aggressive" (p 7). It is transgressive in the strict sense, that it crosses borders, the borders between reading and [re]writing, between text and commentary (for the new constructed text is a commentary on the old one), and between literature and criticism (for, similarly, new texts of literature are both new literature and criticism of that literature at the same time, in an osmosis of "host" and "parasite" much more cohesive than that posited later in this collection by J. Hillis Miller).

Furthermore, Bloom's insistence that "I only know a text, any text, because I know a reading of it, someone else's reading, my own reading, a composite reading" (p. 8) will come as no surprise to those textual critics (most of us, I would guess) who have spent the bulk of their professional lives trying to overcome the naivety of students (and faculty) who do not realise that interpretations do indeed depend upon texts, and that it really does matter which text one has read. As Bloom rather obviously notes "The Milton, the Stevens, the Shelley, do not exist."

(p. 8) Precisely—they do not exist as authors independent of the texts in which their authorship is enshrined, and these texts are variable, not singular or monodic. As even the CSE emblem admits, there is no the to an "approved text" only the indefinite an.

The singularity of a text is as much a product of editorial suppression (of the "inauthentic" or "insincere"?) as it is of augmentation or construction of texts. In all editing, it is thus a concept similar to Hill's coinage "addomission"[18] (the joint principles of inclusion and exclusion in variation), which is used in producing editorial singularity; as Hill's ambiguous term suggests, an editor may not always be able to determine which is which in comparing one text to another. But by such suppression of evidence, by confirming one reading out of many as acceptable, most editors (or at least those editors trying to construct single-state, eclectic texts) are always trying to restrict meaning, to fence it in and circumscribe it. Thus, Bloom's talk of the "authentic poem" having achieved its "dearth of meaning by strategies of exclusion" (p. 15) can be seen as an expression of one aspect of the ideology of eclecticism, as opposed to, say, the ideology of geneticism, which in general seeks to expand meaning without such suppression. But even the genetic editor will be concerned at mapping the changes in meaning, in the exclusions made from one stage to another, so that Bloom's definition of the "strong authentic allusion" as observable in "what the later poem does not say, by what it represses" (p. 15) would be recognised by the geneticist. Now, I acknowledge that Bloom is speaking of different works rather than different texts of the same work (to adopt Tanselle's terminology for the moment), but as any student of Yeats or Auden (or Chaucer or Shakespeare or Wordsworth or Shelley or Emerson or Dickinson) can testify, the genetic growth from one text or state or version to another text or state or version, or the growth of versions even within a single text or document, can display a similar variability, and therefore a similar series of acts of repression as those observable in the sort of progression from work to work that Bloom is concerned about. The important point is that the principle of exclusion and repression, authorial or editorial, is already familiar to the textual critic: Bloom's deconstruction has already been domesticised.

There is another textual issue raised in the Bloom essay that needs some comment here. Bloom invokes the authority of the Stoic school of Pergamanian philology in defending his own interest in "the revisionary ratios that take place between texts" (p. 14). He notes: "Ratios, as a critical idea, go back to Hellenistic criticism, and to a crucial clash

between two schools of criticism, the Aristotelian-influenced school of Alexandria and the Stoic-influenced school of Pergamon. The school of Alexandria championed the mode of analogy, while the rival school of Pergamon espoused the mode of anomaly. The Greek analogy means 'equality of ratios,' while anomaly means a 'disproportion of ratios.' Whereas the analogists of Alexandria held that the literary text was a unity and had a fixed meaning, the anomalists of Pergamon in effect asserted that the literary text was an interplay of differences and had meanings that rose out of these differences. Our latest wars of criticism thus repeat battles fought in the second century B.C. between the followers of Crates of Mallos, Librarian of Pergamon, and the disciples of Aristarchus of Samothrace, Librarian of Alexandria" (p. 14). And I would contend that the repetition of battles observable among the literary critics can be seen in similar ideological or methodological battles still being fought among textuists. Thus, Bloom's reference to the presumed "unity" of a text to the Alexandrian librarians reinforces the supposition that the extensive collection of multiple copies of the same work by the Alexandrians was to assist in the formation of linguistic and authorial rules of analogy whereby norms of utterance could be formulated and then used to identify spurious usage in texts. This "collational" system of the Alexandrians, based on the assumption of an ideal authorial text of which the surviving documents were only corrupt remaniements, is, I believe, similar in its ideology and methods to that of modern eclectic editing, which also regards the extant texts as vehicles for the restoration of a putative authorial form. I would go further, and claim that, while the Alexandrians embraced the empirical methodology of Aristotle, the desire for an appeal to an absolute, an ideal beyond the capacity of an individual text to preserve, owes more to the Neoplatonism of Greek Alexandria than it does to a technical reliance on Aristotle. One should note, for example, that some of the Roman apologists for analogy went as far as demanding that actual contemporary grammatical expression be amplified by such "ideal" forms where these did not exist in the preserved language.[19] On the other, Pergamanian, side of the debate, I would hold that the linguistic doctrine of anomaly is similarly no philosophical accident in a city dominated by the Stoic assumption that all material remaniements are inevitably flawed and that it is therefore both impractical and impious to attempt an Alexandrian resuscitation of a grammatical or authorial ideal usage. Instead, the Pergamanians and their followers insisted on a description of language (and therefore authorial usage) based entirely on the aberrations of preserved documentary forms. And it is this same reliance that motivates best-text editing, an assumption that a putative ideal authorial form is not recoverable (or not testable given the inevitably corrupt evidence of surviving documents) and that the best recourse is therefore a fidelity to the system of individual utterance—aberration or anomaly—that one finds in a particular document. Bloom uses the analogy/anomaly opposition to reinforce his support of "uncanny," "anomalous" criticism of the deconstructive mode, criticism looking for the play of difference rather than the play of similarity. I see the same opposition as being useful in charting the philosophical oppositions between two very different approaches to editing. We have both been taken to task by Donald H. Reiman, who, in an unpublished lecture, "Anyone for Pergamon?," declares that there is little historical evidence to support the opposition in the terms suggested by Bloom. Reiman's attack on Bloom's critical history does not, however, impinge upon the relation I am now positing between Bloom's evaluation of the current terms of the opposition in criticism between "canny" and "uncanny" critics, analogists seeking unity and anomalists seeking difference, and the similar unity/difference schools of editorial work shown in eclecticism and best-text theory. The arguments for (and against) deconstruction are already paralleled in the dialectics of textual criticism.

And so with de Man. This essay appears to make some bibliographical/textual concessions by echoing one of the traditional aims of textual criticism, "the establishment of texts whose unreliability is at least controlled by more reliable means" (p. 40). The various scientific or technical pretensions of various editorial schools, or schools founded upon technical disciplines (say, new bibliography on analytical and descriptive bibliography, or textual analysis upon statistics and probability theory) might make it seem to an outsider like de Man that such relative or desired or even measurable reliability might be an inevitable rationale for textual work, and outsiders like the deconstructors therefore often translate this desire into certainty or confidence, without realising the contingency that is always acknowledged in all reconstituted texts.[20] Thus,

de Man's opposition of the philological and bibliographical is based upon ignorance or blindness, and therefore sets up a false opposition. Indeed, much of the body of his essay repeats in different language Bloom's concern with the changes brought about by exclusion—this time with a vocabulary of "erasure," "effacement," and "disfiguration" (p. 46). He observes such erasures in a number of figures—of forgetting and remembering (pp. 50-51), of veiling and unveiling (p. 53), burial and archaeology (p. 67), but all such figures emblematise the same opposition of lost and found, known and unknown, original and changed, and most are (interestingly enough) a part of the traditional vocabulary of textual scholarship (e.g., memorial reconstruction, archetypal unveiling etc.) If de Man's language is similar to that of textual criticism, what of the concepts?

Consider this: "How can a positional act, which relates to nothing that comes before or after, become inscribed in a sequential narrative? How does a speech act become a trope, a catachresis which then engenders in its turn the narrative sequence of an allegory? It can only be because we impose, in our turn, on the senseless power of positional language the authority of sense and meaning" (p. 64). The expression is indeed somewhat arch and opaque, but the insistence on the forced imposition of order through invoking the supposed power of positional language is no more than a recognition of the sort of changes wrought, and the differing empowerment constructed, as a result of the various plays on titles I discussed earlier in this essay. The "disfiguration" that is de Man's subject is the conscious, unconscious, or completely accidental series of reinscriptions that mark all acts of copying, of reading, and, of course, of editing. Editing is a particularly powerful "disfiguration" of the documentary tradition, even when it seems to endorse it by the selection of a "best" text, for the selection of the "best" is itself a disfiguration of pure history, which is usually more arbitrary, less personal, and less monodic than best-text editing.

Even the best-text editors perform their tasks because of the acknowledged absence of the author to do the job for them (or even, perhaps, because of a distrust of authors as well as documents), and thus when de Man observes that "[i]n Shelley's absence, the task of thus reinscribing the disfiguration now devolves entirely on the reader" (p. 67), he is not only confirming Bloom's assertion that we can never know the Shelley,

but also acknowledging the editorial prerogative, as the most empowered reader of a text, to commit the disfiguration on behalf of this absent originator. There is a paradox in this relationship, for while most textual critics have (until recently, anyway) invoked the "metaphysics of presence" as the rationale for editing (that is, they have appealed to the author's presence as the ground for their decisions), they have usually been dependent upon that very author's absence (through death or the commitment of inner speech to the public sphere of writing) for the reconstruction of the presence. Thus even intentionalist editors have all along anticipated a familiar piece of Derridean deconstruction, for one of Derrida's earliest and most typical acts of deconstruction was to deny the apparent primacy of speech (presence) over writing (absence), a primacy asserted by Saussure's privileging of a phonecentric system of phonemes, and to assert that the full value of the Saussurean sign was conceptually impossible without writing, which thus had primacy over speech. Thus, the intentionalists are employing the Derridean inversion of absence over presence in seeking to discover the present author who is manifest only through the marks or "traces" of absence in the documentary remains—in "writing," or écriture.

This paradox would, I am sure, have been recognised by de Man had it been pointed out to him, and he comes near to articulating another paradox of editing when he notes of The Triumph of Life that a reading of the text of this work "establishes that this mutilated textual model exposes the wound of a fracture that lies hidden in all texts. If anything, this text is more rather than less typical than texts that have not been thus truncated" (p. 67). The truncaton he is referring to is Shelley's accidental death, which left the poem as a "fragment," but while de Man thus percipiently sees this truncation as symptomatic of the "state" of texts, the implication is still that most critics and readers resist the acknowledgement of such a "fracture" in the construction of works and prefer to see them "whole" and "completed." He is probably right in this characterisation of the typical non-textual critic, but his assertion would be woefully inadequate in describing the versioning, fragmentalist, revisionary type of textual criticism represented by such contemporary textuists as Reiman, Gabler, Shillingsburg, McGann, Urkowitz, Taylor, Warren, and Foley, for whom the "fracture" is indeed the norm. Again, textual criticism has anticipated and domesticated the agenda of the deconstructors. Did de Man realise, I wonder, when he penned the following passage that he was writing a (somewhat rhetorical) description of the aims of textual criticism? "And to read is to understand, to question, to know, to forget, to erase, to deface, to repeat—that is to say, the endless prosopopoeia by which the dead are made to have a face and a voice

which tells the allegory of their demise and allows us to apostrophise them in our turn" (p. 69).

As already observed, Derrida's double essay "Living On. Border Lines" imitates (or parodies) in its very textual appearance the dual form of the page of a scholarly edition, with "text" above and "commentary" or "apparatus" below, in reduced type. Tanselle briefly comments on this device (pp. 13-14), but neither he nor (I would hazard) Derrida fully appreciates the inherent possibilities of the parody. Thus, the "lower" ("Border Lines") text is indeed quite properly concerned at first with discussing, allusively and figuratively, the relationship between the commentary/apparatus and "the text itself" above. Noting (in the text of "Living On") that the commentary is usually thought of as "only a textual supplement," an "in other words" for the text proper, Derrida nowhere takes up this frequent theme of the "other words" and applies it to the formal mechanism of an apparatus, which is constructed precisely to find a home for these "other words" of the text. The apparatus is nothing but the text in other words (rejected words), and Derrida demonstrates this by his continuous "and so on and so forth" (back and forth) from text to commentary, without ever drawing on the formal, textual link that would have confirmed this hypothesis. As is to be expected, Derrida attempts to subvert (or invert) the traditional primacy of text to supplement (a familiar thesis of his deconstructive criticism) by embedding authorial instructions in the supplement that should, if carried out, authorise the translation of the text appearing in the nominally superior, but actually dependent, position. But, through chance or ill-will on the part of the translators, these instructions from the author are reproduced but then ignored, thus simultaneously seeming to endorse Derrida's inversion while practically failing to carry out the necessary measures that would ensure it. Thus Derrida may characterise the commentary in "Border Lines" by such instructions (e.g., "My desire to take charge of the Translator's Note myself. Let them [the translators] also read this band as a telegram or a film for developing (a film "to be processed," in English?)" pp. 77-78), but when it comes to specific wishes or commands (e.g., "This would be a good place for a translator's note" [p. 79] or "To be quoted in its entirety" [p. 135]), the authorial will is countermanded, for the extract is not quoted in full as Derrida had demanded of the translators, who similarly do not insert a "translator's note" at the "good place" suggested by the author. So by what is the authorial will countermanded? Paradoxically, by the translator's desire for fidelity to the text and therefore fidelity to the authorial will. And so here is the double bind: the translator is the virtuous copyist or compositor who writes or sets only litteratim what the author's text had already

inscribed, but in so doing, in being faithful to the letter not the spirit, fails to carry out authorial intention in its wider implications. Derrida is at one level (quite literally) making a statement about the inherent subservience and inferior status of apparatus, commentary, and supplement in our culture, but might there not also be an equally contentious statement about intention as well?

The problem with Derrida's commentary is its uniformity (as Tanselle briefly notes [p. 14]). A compositorial (or maybe an authorial) desire for equally-balanced openings has resulted in right and left, recto and verso, of each opening, having equal lines of commentary (seven lines of commentary on each page). This effectively breaks any substantive relation between text and apparatus by the imposition of a formal or aesthetic requirement, which is perhaps Derrida's idea—that substantive relations of this sort are illusory. But in employing a standard, all-too-uniform textual-page division, the essay misses an important point, and an important opportunity, about the power and reflexivity of text and commentary. We have all seen scholarly editions, particularly of works with multiple witnesses, in which the readings adopted by the editor in the text proper are sustained, (literally) supported, and empowered by the rejected readings listed and discussed in the apparatus and commentary below. Scholarly editors have indeed been occasionally reprimanded for the weight and freight of such voluminous charting of variance and of arguments for textual substantiation (the Kane/Donaldson edition of the B text of Piers Plowman is often cited in this regard in my own period, with a couple of lines of text supported by fifty-odd lines of apparatus and commentary). The paradox in this intertextual relationship—and the one to which Derrida might be allusively referring—is that the authentication of the "primary" text above is wholly dependent on the description and evaluation of the "rejected" readings below. The editors must successfully demonstrate the inadequacies of the lower text in order to convince the reader that the upper text is authentic. Thus textual empowerment passes from the lower to the upper, which is therefore (in Derridean terms) "supplementary" or "secondary" and not "original" or "primary." It is a deconstructive inversion with which all multiple-text editors are familiar (and presumably comfortable), but to the deconstructor it must look like a perfect exemplification of the paradoxically primary role of the supplement. It is my suspicion, however, that the formal exigencies of Derrida's text imply that this particular deconstructor has probably not fully appreciated, or fully articulated, the paradox.[21]

It is, I believe, this relationship which is the significant editorial theme of Derrida's essay, but he does touch upon other matters that have textual import. For example, he alludes to the "functioning of the title, the transformation of its relationship to the context and of its referentiality" (p. 117) in a manner similar to my discussion earlier in this essay. He speaks of the way in which "[o]ne text reads another, of how "e[ach] "text" is a machine with multiple reading heads for other texts" (p. 107), in an acknowledgement of the intertextuality of multiple-text works and, of course, of the sort of "reading" of one book by another that I have suggested occurs in Tanselle's reading of Deconstruction and Criticism through A Rationale of Textual Criticism. Furthermore, it is to my mind ironic that Derrida is forced to conduct his discussion of the deconstructive concept of "invagination" (and "double invagination") entirely at the figurative level, since the "border line" between his two essays does not, in fact, involve any of the folding or double folding that the figure represents, whereas our own discipline, particularly in the work of the analytical bibliographer, is directly concerned with the physical problems of folding and double folding in the imposition and format of printed books, especially as that can impinge upon the total "meaning" of the book. As Derrida notes, "Invagination is the inward refolding of la gaine [sheath, girdle], the inverted reapplication of the outer edge to the inside of a form where the outside then opens a pocket. . . . Like the meaning "genre" or "mode," or that of "corpus" or the unity of a "work," the meaning of version, and of the unity of a version, is overrun, exceeded, by this structure of invagination" (pp. 97, 102). Derrida's figurative language of folding as it acts upon meaning and genre or version is limited to the conceptual and metaphorical in his essay—and in deconstructive discourse generally—but the figure can find concrete exemplification in the conscious bibliographical control over meaning and genre exerted by authors such as Jonson and Pope in preparing the differing public and bibliographical appearance of their works in differing formats with different (en)foldings. Thus the publication in 1616 of Jonson's Works (i.e. plays) in folio rather than in quarto (the expected format for popular, ephemeral, vernacular entertainments

like the drama) was a deliberate play on the genre-defined limits of bibliographical format, and Jonson was thus trying to change the cultural "meaning" (and genre) of his oeuvre by the move from double to single folding ("invagination"). A similar point was being made in Pope's issuing of his Homer translations in varying formats dependent on the cultural level and expectations of the audience, for Pope astutely sensed that genre, and thus meaning, was in part created by the various audiences' expectations of bibliographical format. Another well-known example of how the effects of bibliographical folding can affect meaning and genre (this time without authorial involvement) is the problem of casting-off of copy in the imposition for the Shakespeare First Folio, where miscalculations in the amount of manuscript text that would fit into a particular gathering caused prose to be set as verse (to waste space) and verse to be set as prose (to save space) in the outer leaves of the gathering: in this case, the meaning dictated by the necessity of the text's fitting into the space created by the bibliographical folding was actually observable in a technical change of genre. I would doubt that Derrida had Jonson or Pope or Shakespeare in mind in his consideration of "invagination" and its effects, but the figure is nonetheless striking, and is again an example of how textual critics are already predisposed to a consideration of some of the complex issues of meaning and form that are only now beginning to animate other critics.

Geoffrey Hartman's essay "Words, Wish, Worth: Wordsworth" specifically takes up the previously-mentioned deconstructive theme of speech and writing, presence and absence, in his discussion of the figures of voice and sight in Wordsworth. "[T]he greatest deceit voice has practiced is to represent itself as repressed by the written word. Derrida argues that it is writing that really suffered the repression, by being considered a mere reduction or redaction of the spoken word" (p. 207). Hartman's vocabulary ("redaction" in particularly) confirms that the deconstructive relationship between speech and writing, and the editorial complicity in overcoming the limitations of apparent absence in writing, do have textual significance of the sort I have already discussed. A redaction is indeed perceived in the ideology of textual criticism as a "reduction" of the power of an original, even though there may not always be a net reduction in the actual text that undergoes redaction. Through the genealogical assumption of "divergent variation," a downward multiplication of corruption and a consequent loss of authority as one charts the transmission of a text in a stemma, each redaction buries the original voice of the speaker-author in a cacophony of rival voices, each of which questions the authority of this original in its act of reinscription (and transmission). Hartman speaks for the aims of the

traditional Lachmannian textual critic using the genealogical method to arrive at an archetypal reading when he suggests that such an "interpreter zealously redeems the buried voice of the text" (p. 207). Classical Lachmannian textual criticism is indeed an attempt at a "redemption" of something lost, and the image of burial (along with that of archaeology, cleaning, and revivifying) is a familiar one for such textual work—from Erasmus through to the higher criticism of Altertumswissenschaft.

However, Hartman goes on to note that such interpreters engage in this type of archaeology "instead of understanding how texts eclipse voice and speak silence" (p. 207). This understanding is linked by Hartman to the similar awareness of "the deceptive relation between speech acts and being-in-time" (p. 206), for the very" utterance of human wishes . . . reveals, through such phenomena as texts, an 'untimely,' that is, residual and deferred element" (p. 206). Again, such thoughts, which are clearly imagined to be disturbing or unsettling to the literary critic, who may be used to making an immediate correlation between the speaking voice of a text and the speaking voice of an author, should be quite familiar to textual critics, who have traditionally based their discipline upon an open acknowledgement of the "residual" and the "deferred" in a text's transmission. In fact, it might be argued that if there were no such deferral, and no residue of an original, in the subsequent stages of textual dissemination, then there would no reason for textual critics to practice at all. If each text, each redaction, of a work contained "timely utterance" rather than "untimely," then each text could presumably have potentially equal authority and there would be no need for the adjudication of the textual critic. Now, it might be suggested that it is an assumption such as this that has caused the textual critics to despair at the naivety of literary critics imagining that all texts are "timely" and all texts authoritative. How refreshing, then, to find a deconstructive critic who is eagerly pointing out to his colleagues in the literary establishment that the textual critics have been right all along, and that literary critics must now recognise the "deferral" and the "residue" which characterises the buried voice of written texts! As I suggested earlier in this essay, the hermeneutics of suspicion is a predisposition shared by the deconstructors and the textuists.

However, this particular sort of suspicion (of residue and deferral) is acknowledged, indeed traded upon, only by those textual critics of what one might call an "idealist" persuasion, critics who regard the "untimely" physical contingencies of a text's survival as a liability and at the same time as a challenge. Such critics are well represented in some of the beautifully articulated cadences of Tanselle's Rationale, where he wrestles with the problems of deferral and residue. Consider this passage, for

example: "What every artifact displays is the residue of an unequal contest: the effort of a human being to transcend the human, an effort continually thwarted by physical realities. Even a document with a text of the sort not generally regarded as art—a simple message to a friend, for example—illustrates the immutable condition of written statements: in writing down a message, one brings down an abstraction to the concrete, where it is alien, damaged here and there through the intractability of the physical" (pp. 64-65).

Yes, indeed, says Hartman, alienation and deferral, damage and untimeliness, are the norms of the written mode, which is continually striving against its residual status. Hartman goes even further, in suggesting that "[t]here is no authentically temporal discourse, no timely utterance, except by resolute acts of writing (p. 207, emphasis added). What may make an act of writing "resolute" is unclear in Hartman's essay (except, perhaps, in its success in replicating the "light" and "insight" which is the other subject of the essay). I would therefore suggest, given the similar language and similar concerns of Hartman and Tanselle, that the most (perhaps the only) resolute act of writing is the reinscription which is the very process and product of textual editing. For only such scholarly editing is both aware of the fallen conditions of the deferred text and has the skill and insight (if practised with imagination and authority) to unbury the text and the authentic discourse which the redaction conceals. Such anyway is the most common aim of scholarly editing in these past two millenia or so.

But not quite. The "idealist" position articulated by Tanselle and endorsed by Hartman, whereby the deferred and the physical are contingent and unauthoritative, would be challenged by the sociological approach recently favoured by such textual critics as McGann and McKenzie. For such critics, Hartman's redaction and Tanselle's alienation are neither reduced nor alienated versions of something else (a "timely utterance" or an "original intention"); rather, the redactive and the alienated are the things themselves, not mere representations: it is not only the unfortunate condition of the text to be deferred, it is its necessity and its ontology as well. There is no other way of knowing the text, and no other way that its timeliness can ever be manifest. Paradoxically, therefore, it is the "traditional" textual criticism of the Greg-Bowers school and its apologist the idealist Tanselle that is closer to the deconstructive ethic of Hartman, and the more recent current social textual criticism of McGann and McKenzie that would disavow such a textuality.

The final essay in the collection, J. Hillis Miller's "The Critic as Host," offers an etymological, philosophical, and critical evaluation of the sort of relationship I have already described in the discussion of

text and commentary/apparatus—that of guest and host. As is well-known to linguists, the two terms, "guest" and "host," are ultimately drawn from the same Indo-European root, and this etymological congruence emphasises and confirms the duality of the roles within the osmotic relationship, as I suggested in the evaluation of the dual source of empowerment between text and apparatus. As Miller puts it, "A host is a guest, and a guest is a host. . . . This subverts or nullifies the apparently unequivocal relation of polarity which seems the conceptual scheme appropriate for thinking through the system" (p. 221). As a deconstructive critic, an adherent to a discipline and approach frequently criticised for being parasitical on the body of "real" literature, Miller is inevitably concerned to valorise criticism and deconstruction by demonstrating that the rigid distinction which supposes that literature has primacy over and is always different from criticism is imperfect and itself open to criticism—and deconstruction.

What is interesting for the textual critic is that Miller must therefore confront a series of charges (that criticism is ancillary, potentially destructive like a virus, and practised by those who cannot write literature) that are uncannily similar to the charges often levelled against the work of textual critics. For example, Wayne Booth's attack on deconstructive criticism as denying the "obvious or univocal reading" (p. 217) is of a spirit with Edmund Wilson's and Lewis Mumford's attacks[22] on textual critics for having encumbered the texts they edit with the unnecessary recording of variants and multiplicities of meaning, for having placed their own purposes and scholarly desiderata above those of the simple reader of simple texts, who desires a monodic, single-state, authorised, "univocal" reading instead of the complex texts and apparatus of the eclectic or genetic editor. For Wilson and Mumford, the prevailing figure was barbed wire; for Booth, it is the parasite—for both, it is a conviction that criticism of this type keeps the reader away from the text rather than helping him to understand its "obvious or univocal reading."

So yet again, there might be common cause, or at least a common enemy, between the deconstructors and the textual critics, but there is more to it than a similar political [dis]advantage. And it is in Miller's essay that this commonalty appears most strikingly. This is largely because Miller chooses to concentrate on the mediating effects of deconstruction, on its calling into question the metaphysical assumptions of criticism, on its attempt "to resist the totalizing and totalitarian tendencies of criticism" (p. 252). The hermeneutics of suspicion practised by deconstructors and textuists (particularly the textual criticism of recent

years) is inherently antipathetical to such totalisation, for both are primarily concerned with a continuous evaluation of evidence in order to "reveal hitherto unidentified meanings and ways of having meaning in major literary texts" (p. 252). Both deconstruction and the scholarly editing of multiple texts proceed from the assumption that history has failed to disclose or appropriate all the potential forms of meaning in a work as it is manifest in those several texts; and even though genetic editing may appear to offer a more overt acceptance of heterodoxy than, say, eclectic or best-text editing, the recognition of the necessity to overcome the widespread acceptance of the unmediated univocal meaning in texts is common to all editorial disciplines, otherwise there would no need for editorial interposition and mediation in history. As Miller notes, "The hypothesis of a possible heterogeneity in literary texts is more flexible, more open to a given work, than the assumption that a good work of literature is necessarily going to be 'organically unified'" (p. 225). Now, it could be argued, and I have myself argued,[23] that the need for organic unity in a work has been present in both the New Criticism and its coterminous textual equivalent, new bibliography: both found the figure of the "well-wrought urn" beguilingly attractive, and both sought to resuscitate it out of the apparent ironies and complexities in the texts of works. But the characteristic method of the eclectic editing arising out of the new bibliography was nonetheless to display the multiplicities of meaning for the intelligent reader, and to assume that the multiplicity and heterogeneity were not ancillary to, or preparatory to, an understanding of the work, but were a necessary part of the work's very texture. The ironies and variables of the text(s) might appear to be reconciled in the eclectic edition, but such editions did not deny, but rather gloried in, the potentially endless permutations that the apparatus/commentary charted. This was the "barbed wire" that Wilson and Mumford objected to. In Miller's language, what scholarly editing does, even eclectic and best-text editing, is first of all to refuse to acknowledge the unquestioned authority of the name of the author "printed on the cover of a book entitled Poetical Works" (p. 243); second, to acknowledge that all such books and such texts contained therein are chains "of parasitical presences—echoes, allusions, guests, ghosts of previous texts" (p. 225); and third, to demonstrate in the presentation of evidence, that "the critic's attempt to untwist the elements in the texts he interprets only twists them up again in another place and leaves them always a remnant of opacity, or an added opacity, as yet unravelled" (p. 247). It is true that the fragmentalist and revisionist schools of textual criticism already alluded to would seem to offer the closest parallel to Miller's elucidation of the problems of the unravelling of the elements of a text, but let us not forget that the very word text (and textual) is a cognate of not only the Latin textus—the sense of "authority"—but of textile as well, the woven fabric whose threads are forever threatening to come unravelled, and whose identity as a total or coherent entity is entirely dependent on the skills of the weaver in seeming to hide the construction of the fabric. The successful textual critic is, like the successful deconstructor, one who can perceive both the ravelling and the unravelling, the singularity and the heterogeneity, the construction and the dissolution. In analysing the term deconstruction, Miller insists that "tying up is at the same time a loosening" (p. 251), and that [d]econstructive criticism moves back and forth between the poles of these pairs, proving in its own activity . . . that there is no deconstruction which is not at the same time constructive, affirmative. The word says this in juxtaposing "de" and "con" (p. 251).

It is my contention that textual criticism and scholarly editing have this same de/con/structive balance or bipolar opposition. We talk of reconstruction and resuscitation, of perceiving the whole out of the corruptions of the fragment, but our work is a continuous acknowledgement of the power of the corrupt and the fragmentary, and even our editions may lay bare the otherwise unnoticed and separate threads of the textual garment as much as they present a completed artifact for the reader.

In thus rereading Deconstruction and Criticism according to principles different from Tanselle's, and emphasising the points of similarity rather than difference, I have presented what is in effect a deconstruction of Tanselle's deconstruction of Deconstruction and Criticism. The very difference between my essay and Tanselle's is thus an example of deconstructive différance (the pun on differ and defer mentioned earlier), since it offers a "phonemic" difference which also promotes and represents a deferral of final meaning, just as Tanselle's essay, by questioning the values and terminology of the deconstructors, not only differs from them but also defers a final meaning to "text" and "work." Such a chain of related readings, one book being read against another, is only to be expected, and is in any case no different from the re-editing of works by differing editorial persuasions, each subsequent one deconstructing the assumptions of the former. Just as Johnson re-edits Shakespeare in part to undo Pope and Theobald, and just as the CSE/CEAA editions were undertaken in part to undo the belletristically-derived texts of the

nineteenth and early twentieth century, so one rereading both acknowledges and undoes the earlier one. If deconstruction is only "analytic criticism as such" (as Miller insists), then it should come as no surprise to discover that we are all doing it, willy-nilly, or that deconstruction may not be as inimical to criticism, textual or otherwise, as we had supposed.

Notes