The situation of universities in societies where they’re allowed to have a situation at all (where, in other words, they’re not just part of the bureaucratic apparatus) has always had a monastic coloring. In the castle and on the plain, lord and peasant go about their accustomed pleasures and assigned chores; shopkeepers along the lanes of little towns quietly pursue their little advantages—while within the cloistered halls, doctors and saints carry on their internecine feuds on topics too remote and abstract for the society outside to know or care about. Every now and then a thump, a squeak, and a wail of suppressed agony comes out of the academy; but mostly the odium theologicum expresses itself in the classroom, at the department meeting, and by pointedly ignoring in public the rogues of the opposite party. Occasionally it overflows in print, as reviews, articles, and contributions to little magazines—as well as in books assembled out of these fragments, like those currently under review.

A good deal of the recent feuding has taken place in departments of literature, and most of the banners have been planted around the question of what texts to read and how to read them. A key word with more symbolic than precise denotative meaning is “deconstruction.” All five of the books under review, though they may toy with lighter topics on the way to the major statement, come down in the end with a position on the deconstruction issue. For instance, Professor M.H. Abrams, in his strangely titled Doing Things With Texts, has assembled essays and critiques under four main heads, of which the first two are preparatory to broad assessment of the deconstructive argument, reserved for the last two. Intermingled are some explanatory discussions of the methods of Abrams’s last big book, Natural Supernaturalism, and a couple of reviews—occasional but exemplary—of books by Philip Wheelwright and Northrop Frye.

Readers of Abrams’s previous criticism will not have to be told that his prose is careful and informed, thoughtful, and therefore slow of pace, soft-spoken but firmly argued, and invariably clear. Perhaps Michael Fischer, who assembled the various essays and contributed a foreword, could have done a bit more, i.e., a bit less, to produce an economical and well-tailored book. The essay “From Addison to Kant: Modern Aesthetics and the Exemplary Art,” which directly follows one entitled “Art-as-Such: The Sociology of Modern Aesthetics,” repeats a good deal of its predecessor’s material. It has much to say, of which critics caught up in today’s catchwords need to be reminded, but once is enough.

The point Abrams makes here is that viewing the art-object—the painting on the wall, the poem on the printed page—as an object-in-itself, to be judged by purely aesthetic standards, is a relatively recent habit of mind. It derives from social circumstances of the late eighteenth and early nineteenth centuries, and carries on its face traces of specific philosophical and even theological prepossessions. Though unmistakably a historical construct, this critical premise has been so long established and so largely taken for granted that it now seems natural and self-evident. But there may be values in the work, including those of direct, didactic statement, that the artist did not intend for us to overlook, and that the work itself by its structuring requires us to recognize.

Thus the ideals of “art as such” (or “for art’s sake”) and “pure, disinterested contemplation” are revealed as partial and time-bound. Yet from this basic aesthetic position have developed a great number of modern critical notions, mostly formalist in nature, among the most recent of which a wide scattering lay claim to names like “structuralism,” “poststructuralism,” and “deconstruction.” In one of his later essays, “Construing and Deconstructing,” Abrams confronts and directly assesses the latter procedure.

This section of the book invites particular analysis because in it Abrams confronts a skilled and responsible opponent, Professor Hillis Miller, over the interpretation of a well-known and very brief text, Wordsworth’s eight-line elegy known from its first line as “A slumber did my spirit seal.”

A slumber did my spirit seal; I had no human fears:
She seemed a thing that could not feel
The touch of earthly years.

No motion has she now, no force; She neither hears nor sees;
Rolled round in earth’s diurnal course,
With rocks, and stones, and trees.

An oddity of their discussion is that both readers construe the poem in very much the same way to essentially the same effect. They agree that the poem is the utterance of a single speaker, that it concerns a female (“she”), that she is discussed from two different, successive points of view. In the first stanza she is alive and apparently likely to remain so; thus the speaker’s spirit is sealed (protected) against the basic fact of her mortality by a slumber (including the idea of a delusive dream). In the second stanza she is understood to be dead and incorporated forever in the cycles of the material universe; it is a painful but real change that the speaker is understood to regret. So far the two critics are in agreement, not only with one another, but with just about every common reader of the poem.


But, in addition, Miller undertakes a second-level reading of the poem, which he doesn’t hesitate to call “allegorical.” For this reading he draws on a number of other Wordsworth texts written at different periods of the poet’s life, on details of his biography, and on cosmogonies with which one can only conjecture whether he was sympathetic or even acquainted. Among other things, this second-level reading converts the girl who is the chief focus of the poem into a potential sexual partner, into a surrogate for Wordsworth’s own mother, who died when he was eight, and into a symbol of the rising and setting sun. (Her name, we have come to believe from other poems, was “Lucy,” though Wordsworth does not use that name in this poem. The name “Lucy,” though it could just as well equate with “Lucifer,” is taken to refer to the sun as the source of light, and perhaps through the old sun-Son pun, to the logos.) Thus her loss to the universal fate of mortality can be equated with the poet’s loss of that ground outside the system of language which is always needed but never available. On this reading the poem not only asserts allegorically but enacts (in the death of the logos) its own essential meaninglessness, and leaves us (if I understand the finest of Hillis Miller’s fine points) in the authentic position of shivering in awe over the abyss of the unknowable and unrepresentable.

Abrams is very gentle and permissive in his attitude toward this second-level reading, which even his considerate nature is tested to take seriously. The principle of reading one text in the light of other texts by the same author is an old and established one but its aid has traditionally been to clarify and unify the basic text one is studying. To explode that basic text into a bundle of unrelated and unresolvable contradictions at the expense of its human meaning has generally been recognized as a possibility, and with ingenuity and determination, the operation can be performed on this or any other text. No law of God or man forbids this procedure. But the effect of the two-level reading of Wordsworth’s eight poignant lines is to create two quite different and ultimately unreconcilable texts.

The assured conviction of the pure (or uncanny) deconstructionists that all texts not only can but must be reduced to total indeterminacy (a.k.a meaninglessness) underlies most forms of the deconstruction enterprise; and the root of this compulsion is a conviction about the nature of language. Having construed the text, and then deconstructed that construal, the critic is understood to be freed to exercise his ready wit in full liberty from the constricting and tyrannical intentions of the original author or text; the resulting misreadings, if strong enough, are legitimized as a way of bringing out meanings of an original text to which the language—seen as an organism with a life of its own—was alert, though the original author was not. The text is of interest, in fact, chiefly as an occasion for liberating a more inspiring and important text, that produced by the critic. Indeed, the less the text amounts to (as in analyses of pop culture, soap operas, beer commercials, and television spots), the greater the opportunity for the critic to shine.

It is to this new freedom of the critic from service to the text as given that Abrams most vigorously objects. The alternative, as he repeatedly asserts, is not to suppose that all texts are simple and univocal—speaking with a single voice to a single effect. On the contrary, complex texts are subject to many different interpretations, and different premises will lead to different judgments. But the ideal of reconciling the problems of a text is very different from the ideal of exploding them; and it isn’t always clear which ideal the different people calling themselves deconstructionists, propose to serve. They often find themselves being misunderstood—so often that, given their principles, one can’t help suspecting that some of them welcome the misunderstanding as an occasion for shifting the focus of discussion from the old text to the new. The intrusive critic is easy to resent—which shouldn’t blind one to the fact that widening the scope of critical discussion often serves permanently valuable ends.


Frank Kermode, like Abrams, is an established exponent of the “old” criticism; his new book, An Appetite for Poetry, gathers together ten essays almost equally divided between a critique of deconstructionists, an appreciation of three poets and a critic, and discussions of critical and textual problems, particularly as exemplified in readings of the Bible. Kermode’s interest in Biblical studies goes back a couple of books, and in this field, where enthusiastic amateurs find a happy hunting ground, he dwells, to his credit, among the strict scholars. In general, Kermode is a crisp and forceful writer—the noun is an honorific. Like Abrams, he distinguishes the first level of deconstructionists—Jacques Derrida and Paul de Man specifically—from their less literary and often more schematic epigones.

Without getting overbearingly administrative about it, he deplores the impoverishment of the literary curriculum by making Shakespeare, for example, a set of preachments on the doctrines of modern feminism, or by teaching texts from the subliterature in preference to the “classics,” however they are defined. In only one instance does he scold or hold contemporary practice up to scorn; but in concluding an essay on the common reader, he speaks very generally and very severely of those who, pursuing novelty or popularity, shortchange their students and their readers. It is the universities, he says, and the daily reviewers who come so largely from universities, whose function it is to foster the next generation of common readers: “And every narcissistic, venal, or impudent review, every clever academic stunt, is a dereliction of this duty of continuance and creation.”

Of the four essays central to Kermode’s book—those on Milton, Empson, Eliot, and Wallace Stevens—I think very likely the one on Stevens, relating him (surprisingly) to Heidegger, may well be the most far-reaching and speculative. But his appreciation of Empson is both timely and likely to be influential. William Empson was a cantankerous and fiercely independent man; he was also a critic of many interests, and though his thinking was erratic it was frequently brilliant. His book on The Structure of Complex Words, after lying dormant for almost fifty years, is starting to attract interest again, because its emphasis on the equations to be found in compound words shows that everyday language may be just as indeterminate as literary language. “Honest,” for example, can be used to soften while reaffirming social distinctions, as in “my honest friend Bob here,” or to describe a good fellow who is not too uppity to be a friendly companion in a bar, or with very different connotations of “an honest woman,” or with assertive egalitarianism, as in “an honest man’s the noblest work of God.” This variety of meanings is grist to the deconstructionist mill, and in the simplest of senses, it’s absolutely right. All language, if you examine it scrupulously and pick its components apart deliberately, turns out to be made of the same loose texture. As Seneca said some time ago, “Confusum est quidque in pulverem sectum est,” that is to say, “Chop it up fine enough and nobody will know what it is.”

Readers have been more apt to look narrowly at literary language because the nuances and undertones may be more artful there; everyday speech is just as easy, or perhaps easier, to dissect into indeterminacy, but most people don’t try to do it—for the crude purposes of the marketplace, crude diction is perfectly adequate. So what Empson exposed in The Structure of Complex Words, with great zest and intelligence as well as some perversity, is congruous only in a limited way with the deconstructionist program. Complex Words was an extension and intensification of the earlier and better-known Seven Types of Ambiguity; both books were devoted to widening, complicating, and enriching the reading process. Empson indulged in a lot of marginal forays—his almost algebraic poetry, his angry assault on Milton’s God, his battles for Basic English, among others—so that it’s hard to get a coherent picture of his truculent, intense presence. Kermode does well to remind us of the entirety of Empson’s literary career, its scattered structure, as well as its strokes of what he can only call genius.

About most of Kermode’s concluding essays on Biblical criticism I find myself unfit to pass informed judgment. Over the centuries hardly a subject has been more hotly or variously debated. A lot of the disputes are antiquated now (at least in the minds of thoughtful readers), but new ones spring forth so that if one starts from very near scratch, catching up is by now just about impossible. But I’ll make an exception for the essay on “Divination,” which has to do with conjectural emendation of a corrupt text, i.e., inspired or informed guess-work. Shakespearean examples are easiest to present: for example, the old crux in which Dame Quickly, describing Falstaff on his deathbed (Henry V, II, iii), is made by the authoritative Folio to say, “his nose was as sharp as a pen and a table of green fields.” Theobald in the eighteenth century emended to “and ‘a babbl’d of green fields,” but a stubborn rear-guard action is still being fought over other possibilities. The green fields may be a green baize tablecloth, or alternatively a cemetery, of which Faltaff’s face is a table, i.e., a prophetic picture. This is a hard case to decide because there are several semi-workable choices; other cruxes present us with only two choices, both desperately uncomfortable.

There are plenty of occasions when following the best surviving version of a text condemns one to absolute nonsense; and there the editor worth his salt has to open his inner eye, gather his forces, and take a flying leap at what will constitute a reading. Of course one can go fearfully, ludicrously wrong; but without audacity carried sometimes to the point of folly, one can’t hope to go right. It is an exercise for the Evil Knievels among textual scholars, and Kermode clearly relishes the perils and possibilities of the exercise.

Protocols of Reading by Robert Scholes borrows its title from Jacques Derrida; it means simply “rules of reading.” Derrida says he doesn’t know any good ones, and Scholes says he isn’t going to try to supply any, so that leaves open the question what the book is about, and I must say the question still lingers when one has finished reading it. Scholes declares himself a semiotician, which is a reader and interpreter of signs, a formula that covers a lot of ground, from looking at the clouds to responding to a Stop sign. As for what special protocols we need for reading a text, my first thoughts were humble, not to say vulgar, ones. I thought a good reader ought to know the language in which his text is written, the various meanings of the various words, and the conventions of punctuation. Then, depending on the subject matter discussed in the text, he should have some preliminary information about it; no one will make much of a text on advanced mathematics who hasn’t mastered some elementary and intermediate math. And if a book is going to use technical terms without explaining them, refer glancingly to out-of-the-way philosophers, embark on topics without a rationale, and drop them without reaching a conclusion, protocols of reading may not be enough; some protocols of writing may be called for.

Turning to page 56 of Scholes’s book, one finds him dealing basically with “vulgar nihilism,” defined as the belief that “since there is no Truth, there is no error either, and all beliefs are equal.” Just who subscribes to this simple yet capacious belief one is not told, but apparently Jacques Derrida has put forward an alternative to it. However, Scholes continues, we will approach Derrida and his alternative through a very different thinker, Stanley Rosen, whose thought derives from Plato, and who opposes Richard Rorty as well as Derrida because he or they derive(s) from “Heidegger’s interpretation of Anaximander, with the crucial difference [contributed in an unspecified text by an unspecified author] that Sein is replaced by Nichts.” To employ this sort of prose in a discussion of protocols of reading seems to me not far from insulting. Let the writer express his thought in clear and reasonable English, and we’ll read him perfectly well.

Though there probably aren’t any formal rules of reading (and if there were, they would be obsolete as soon as formulated) there are many ways in which better (more informed, more imaginative, more alert) readers can help the rest of us. As a rule, this process takes place most successfully in the format of conversation. Any individual assertion made in the course of a conversation can be in error, but all statements are subject to limitless correction and modification. And this thought highlights a limitation of printed books that becomes particularly apparent when they are discussing matters which have been defined by a long process of give-and-take.

Scholes tells us that a major element of Derrida’s theory of communication was developed in controversy with John Searle; but he doesn’t adequately describe what Searle said or which parts of his argument impinged on Derrida. All through these several books of meta-and meta-meta-criticism, one encounters decisions being triumphantly recorded over opponents who have scarcely been allowed to utter a word in their own defense, let alone to share in the last word. In this somewhat extended sense, the deconstructionists are often no less univocal than their traditional antagonists.

Not wholly alien to the rules of reading and writing, though no more easy of formulation, are what used to be called the rules of evidence. Moderate attention to them enables the reader to follow the writer’s train of thought. For example, Scholes undertakes to read the character of the late nineteenth century from thumbnail biographical sketches of three men—Joyce, Mussolini, Lukacs—without noting that a wholly different late nineteenth century could be constructed from virtually any other trio of arbitrarily selected individuals. So where did the idea come from that controlled the selection? There’s no knowing, no way of knowing. The book concludes with an extended comparison between advertising and poetry, and an argument, based mainly on Middle-march, that the novel as a form is interestingly intertwined with the Protestant ethic. Semioticians, who can choose from the infinite world of signs those they want to discuss should beware of the signals they send in their act of choice.

Framing the Sign by Jonathan Culler is yet another collection of mainly occasional critical pieces—essays, articles, lectures, introductions, and contributions to collections. It is almost all recent material, well organized, crisply written (where the subject permits), and unusually strenuous in its thinking. A very positive feature is Culler’s careful avoidance of the vertical pronoun and severe concentration on the business at hand. The result is an impressive variety of essays on a range of topics from trash and tourism to cultural literacy and a comprehensive linguistics of writing, to all of which he applies, with persistence, sharp intelligence, and a tendentiousness that some people may find narrow, the same deconstructionist critical approach.

Whatever his personal character, his book shows no traces at all of a sense of humor. His claim on page 40 that while formerly the history of criticism was part of the history of literature, “now the history of literature is part of the history of criticism,” puts one is mind of the frog in Aesop who tried (with explosive consequences) to puff himself up to the dimensions of an ox.

More striking is the desuetude of literature as such in the deconstructionist’s economy. Culler largely avoids mentioning it, least of all to imply that pleasure can be found in it. He has nothing to say of Blake or Mallarmé, nothing of Donne or Dante or Hopkins or Beckett. The names are mentioned here not as talismans, but as first entries in a far, far longer list of authors in whom Culler’s preoccupation with critical theory has allowed him to take no visible interest.

The book is divided into four parts: one on institutions; one on three modern critics; one on culture; and one titled “Framing Language.” The section on institutions concerns mainly departments of English and the premises of offering instruction in literature; it is supplemented by a youthfully harsh assault on religion, primarily Christian religion. (Wisely, Culler did not try to deconstruct the Koran, or the ban placed on Salman Rushdie might have been invoked.) The three critics are Empson, Bachelard, and Paul de Man, and the latter is given an extended and detailed treatment at the expense of the former two. The application of deconstruction to modern culture takes place under the heads—as disparate as one could want—of law, tourism, and rubbish. In the final unit Culler edges close, in several considerations of the language of fiction, to discussing literary matters, but linguistic arguments (including a couple of bodacious ones) take over, leaving behind any reader whose central interest is not critical methodology.

Culler’s concerns are what they are, and he pursues them with spirit and tenacity. In an omnibus review it is impossible, even if one were qualified, to follow out lines of argument so rigorous and so subtly pursued. Instead, I’d like to say a few words about Paul de Man, both as he appears in Culler’s account and as he is represented in a posthumous collection of Critical Writings, 1953–1978, edited by Lindsay Waters. I should begin by saying that, though I never shared Paul’s literary point of view, and never felt myself an intimate, we were good friends during our Cornell years (1960–1968), and I experienced his death as a profound personal loss. He was a unique person in my life, not just a colleague, but a mind so sinuous and subtle that it never ceased to challenge me. Many people seem to have felt that way.

From reading Culler’s account, I was confirmed in my sense of the intense interior dialectic that went into his thought, as well as the dark negativity that kept him from bringing that dialectic to unequivocal conclusions. There were other contrasts, as between the playful wit of his personal manner and the severe authority of his writing, between the wide range of his humanist reading and his sharp dismissal of humanist philosophy. I never felt I could assemble the Paul de Mans I knew into a unified figure—there were too many layers. Generally, I knew where he was at, but couldn’t always tell where he was coming from.

I don’t feel that Culler helps me much here. He follows, more closely than I could do, the manifold interweavings of the dialectic de Man carried on with himself, but he doesn’t give me any sense of the sharp turns and complex processes through which de Man passed to get where he was going. It’s not altogether fair, but it expresses a truth to say that de Man as Culler presents him comes off as a deconstructionist sausage machine turning out linked formulas predictably, indefinitely.

De Man didn’t want to bridge the gap between language and reality, or history, or action; he didn’t want to establish a “ground” for language, and thus didn’t want to unify—falsely, as he thought was inevitable—experience with the contrivances of art. “He denied himself not only the consolation of philosophy but the pleasure of fiction.” Denis Donoghue, from whose forceful essay I adapt this phrase,* attributes de Man’s fierce prepossession on this point to an undeclared war he was waging in secret over his own untidy past. Perhaps so, though the argument seems to me simplistic. We all have a skeleton or two in the closet; as Donoghue says himself, all de Man had to do to dispose of his past, any time after 1960 or so, was to mention it. But I think the turn in his thought that took place about 1967–1968 can’t be persuasively connected with the events of 1941–1942.

It is interesting that on the purely intellectual level de Man was concerned, even in the early 1950s, with Montaigne, whose division between reality and appearances is a key theme of the “Apologie de Raymond Sebond.” Montaigne’s standards of “the truth” in the matter of reality are not just high, they are absolute; that, of course, justifies the other Essais in being inconsistent, imperfect, approximate. And that, setting aside Montaigne’s apparatus of theological terminology, was very much the crux, defined as the search for an absolute “ground” on which to rest the logocircular world of language, on which de Man got himself so bafflingly and painfully hung up. He was long on “the truth” at a certain level of ultimate authenticity; down here, with us, it was evidently another story. But that the path leading to his concluding position—dead-end though it seems—was developed by processes of controlled and critical thought, I cannot doubt. One dealt with de Man on the technical level of ideas, analyses, interpretations; psychological shortcuts won’t bring one out in the same neck of the woods, not at all.

Mr. Waters, in his compressed but brilliant introduction to the Critical Writings, doesn’t dilute de Man’s thinking, but he emphasizes its personal contours and moments of change to make it appear a particularly human process. There were, clearly, moments of hesitation and impasse, transitions marked by leaps both conceptual and linguistic, as from Rousseau and Nietzsche to Hölderlin, Heidegger, and Walter Benjamin. The fulcrum of the great turn he made in 1967–1968 seems to have been an attempted emancipation from preoccupations with the idea of self and the problems of subjectivity, and this involved revaluation of an entire experience with and commitment to Hegel. More de Man texts will be published, particularly a volume called Aesthetic Ideology, which will doubtless make some things clearer and other, perhaps, less clear. But at least some of the outlines of a careful, even painful, development are now discernible.

If one is interested in de Man’s thought as a whole, the miscellaneous critical writings collected here are clearly secondary. But with Waters’s introduction to give them order and context, they display some of the mingled toughness and sensitivity that made him a mage for two generations of very good students. His ascendancy wasn’t always to their advantage; some tried to become little de Mans overnight, without going through the chastening process of his development. But even to one who was never an acolyte, the asceticism of his thinking was always an inspiration.

This Issue

March 1, 1990