Frank Kermode
Frank Kermode; drawing by David Levine

Near the end of Versions of Pygmalion J. Hillis Miller argues that reading involves two obligations. On the one hand: “to transfer to reading Henry James’s injunction to the observer of life, the novice writer: ‘Try to be one of the people on whom nothing is lost!’ ” On the other: “to reduce the inexplicable to the explicable, to find its reason, its law, its ground.” I have never respected this second duty. It seems to me to pretend to knowledge where knowledge is impossible. I would much prefer to be shown that a book or a painting is mysterious than to have someone reduce it to the explicable for my poor benefit. But Miller’s concern is a crucial one, as this passage from his book indicates:

Current criticism tends to propose one or another of the following three grounds on the basis of which the anomalies of narrative may be made lawful: (1) society, or the more or less hidden social and ideological pressures that impose themselves on storytelling and reveal themselves in oddness; (2) individual psychology, or the more or less hidden psychic pressures that impose themselves on storytelling and make it unaccountable; (3) language, or the more or less hidden rhetorical pressures, pressures from some torsion within language itself, that impose themselves on the storyteller and make it impossible for the story to maintain itself as a lucid and reasonable account.

Miller says that each of these axioms is imperialistic: “Each as a mode of explanation demands to exercise sovereign control over the others.” He remarks, too, of these axioms the “strong resistance generated in some of those to whom they are proposed.” He doesn’t say whether, in a particular case, the resistance is justified or not.

I don’t know what readers would make of Irving Howe’s Selected Writings 1950–1990 if they didn’t place a high valuation upon the axiom of society; if they didn’t regard social considerations as of the first importance. It is not to be expected that in a selection of his essays over a period of forty years Howe will always be found enforcing the same emphases, but he has never forgotten the social imperative or its bearing upon literature and criticism. In “Writing and the Holocaust” (1986) he has this paragraph:

Chaim Kaplan’s Warsaw diary, covering a bit less than a year from its opening date of September 1, 1938, is a document still recognizably within the main tradition of Western writing: a man observes crucial events and strives to grasp their significance. Kaplan’s diary shows the discipline of a trained observer; his prose is lucid and restrained; he records the effort of Warsaw Jewry to keep a fragment of its culture alive even as it stumbles into death; and he reveals a torn soul wondering what premises of faith, or delusion, sustain his “need to record.” Barely, precariously, we are still in the world of the human as we have understood it, for nothing can be more human than to keep operating with familiar categories of thought while discovering they will no longer suffice.

Howe would not compare Kaplan’s experience with his own, or claim to possess a torn soul. But many of his books and essays find him, in arduously patient tones, working with familiar categories of thought while discovering they no longer suffice. Or that they are no longer heeded. The idea of society, the sentiment amounting to a conviction that we survive, if we do, as a society: this is Howe’s axiom. During the past forty years he has had many occasions to reflect with dismay upon the fragility, in practice, of this sentiment; and upon the recurring demand to live (as Lionel Trilling phrased it) “beyond culture.” In “Anarchy and Authority in American Literature” (1967) Howe wrote:

There is no other Western culture of the past two centuries in which, to my knowledge, so many demands have been expressed for the “creation of values.” When one comes to think of it, that is really an extraordinary fact. The literatures of Europe either sustain traditional values or enlarge upon revolutionary values; but both are seen as inseparable from the social order in which the writer writes and the reader reads. In our culture we have made the unprecedented demand upon writers that they “create values” quite apart from either tradition or insurgency. What we have often meant is that they establish a realm of values at a distance from the setting of actual life, thereby becoming priests of the possible in a world of shrinking possibilities. We ask them to discover, out of their desperate clarity, a vision we can cherish, and cherish perhaps in direct proportion to our knowledge that we will not—or cannot—live by it.

There is a feasible riposte to this; that while “we” may not undertake to live by the vision we cherish, someone else may, our children may. But I don’t think I could persuade Howe to settle for that. He has lived upon his own vision for so long.


Socialism is his name for it. He refers to “the unity of socialism and democracy,” “the liberal-radical vision of the good society,” a realm of values he associates with individuality and dissent. He finds it hard to understand “why the American working class in the 1880s and 1890s engaged in very militant and even violent strikes yet did not ‘move ahead’ to any large-scale socialist beliefs.” During the years annotated by these essays, Howe has witnessed “the debacle of socialism” to the point at which his readers would be justified in asking him how this will-o’-the-wisp differs from the most outlandish dream offered by Emerson or any other American writer. In “The New York Intellectuals” (1968) Howe refers to the plight of a few intellectuals, including Nicola Chiaromonte, Paul Goodman, and Dwight Macdonald, “who wished to dissociate themselves from the postwar turn to realpolitik but could not find ways of transforming sentiments of rectitude and visions of utopia into a workable politics.” I don’t understand why he did not add his own name to that short list. As editor of Dissent, he’s entitled to that modest reward.

In the same essay Howe writes:

Once it became clear that waiting for the revolution might turn out to be steady work and that the United States would neither veer to fascism nor sink into depression, the intellectuals had little choice but to live within (which didn’t necessarily mean, become partisans of) the existing society.

That, in brief, has remained Howe’s position, and the parenthetical nuance marks his scruple.

But he finds the conditions dismal. In “Reaganism: The Spirit of the Times” (1986) he ponders “the collapse of American liberalism”:

To recall Lionel Trilling’s once-famous remark of the 1950s—that in America liberalism is the only viable political tradition—is to thrust oneself back into another world. In the face of the Reaganite victory, the organizational and ideological collapse of American liberalism has been astonishing. Hardly a politician dares acknowledge himself to be a liberal, the very word itself having come to seem a political handicap.

In such a context it may be deemed vulgar to ask: What bearing has Irving Howe’s social emphasis upon his literary criticism? But a social vision may be an incitement to criticism even when it has no chance of establishing itself in political life. A will-o’-the-wisp, if believed in and pursued over hard ground, may allow a writer to come upon his subject obliquely or desperately, which may be in that case the only way of coming upon it at all. The myth of socialism, like any other myth, enables Howe to imagine what James called “the possible other case,” the case rich and edifying where the producible reality is for the most part sordid.

As a case in point, I am thinking of Howe’s essay on “George Eliot and Radical Evil,” his introduction to Daniel Deronda. Received opinion on that novel, mainly received from Henry James and F. R. Leavis, says that the whole Jewish part of it is dead, but that the other part, the deplorable marriage of Gwendolen and Grandcourt, is George Eliot’s most powerfully imagined work. Howe is forced to agree with this, but since he is loath to see anything Jewish die, he postpones the obsequies with causes and explanations. He tries to explain why George Eliot, having set out to imagine Daniel Deronda and to summon him into particular life, found herself merely assigning to him the civilizing function that Matthew Arnold assigned to the Hellenic element in any of his readers. Howe’s essay is characteristically generous and tactful; notably in its recognition that the relation between Gwendolen and Grandcourt is finally mysterious, and that “George Eliot explores this ‘mystery’ but is too much the novelist simply to clear it up.” It is a mark of Howe’s tact that he does not offer to clear it up for her.

It is hard to say what Geoffrey Hartman’s ruling axiom is. Of Miller’s three, he would choose “Language” but not if it forced him to think reality merely an affair of words. His preoccupation over many years with the Holocaust has preserved him from that idolatry. But he is sympathetic to the language critics, and especially to Jacques Derrida and Paul de Man.

Minor Prophecies is a gathering of Hartman’s recent essays, linked by a concern for the character of the essay itself as a form of intervention. The matters upon which these essays intervene include the theory and practice of reading, methods of interpretation, the question of style, and the various styles of deconstruction. The critics he describes most fully are Leavis and de Man, but there are many vivid pages also on Derrida, T.S. Eliot, Valéry, Virginia Woolf, Habermas, I.A. Richards, Erich Auerbach, Walter Benjamin, and Kenneth Burke. Hartman’s strategy is to describe two established forms of style and to try to make peace between their exponents.


The first style is that of the familiar essay. Historically, it issues from the French salons of the seventeenth century and from practices of conversation in France and England: it invokes “the gentle reader,” presumes upon social and moral consensus, and fears enthusiasm. Addison and Steele are among its adepts, Coleridge called it “the middle style,” and in most of its versions it looks back to Horace. It is the style of friendship, “from Dryden to Donoghue,” as Hartman agreeably describes it. The second style is that of Theory; it is “prose with a noticeable proportion of technical terms.” Difficult, experimental, speculative, rabbinic, it is a style of disenchantment, and Hartman values it in that character:

We who are bewitched by capitalism must find a means of disenchantment, of breaking the spell, and from within our historical moment. Otherwise we lapse into a transcendental form of interpretation, whereby thought realizes only itself and does not change anything outside itself.

But Hartman doesn’t claim, I hope, that this difficult style is our sole defense against the siren songs of capitalism. Or has he forgotten that Irving Howe’s resistance to them has been practiced in the familiar style?

There is a question of justice, as Hartman well knows. But the polarity he enforces between the two styles makes justice nearly impossible. He can’t avoid presenting the first as that of a genteel tradition, and then we wonder where the gentility is to be found in Leavis, Empson, Richards, G. Wilson Knight, and many other critics who supposedly embody it. In fact each of these critics puts genteel assumptions under severe scrutiny.

As the book goes on, Hartman associates the second style mainly with deconstruction. He praises that movement—if it is a movement rather than a loose fellowship of practices—for having “dismantled such essentialist values as origin (genealogy), intent (original intent), and identity (nature) by a study of the temporal aspect of human existence (how our truths remain contingent, how we are never present or transparent to ourselves…”). On the other hand (and Hartman is our most formidable two-handed critic), he acknowledges that deconstruction “has not made its situatedness clear enough or developed terms for moral questions in a way that is systematic or free of linguisticity.” When roused on a personal issue, as in the case of Paul de Man and his collaborationist writings in wartime Belgium, Hartman ignores theoretical misgivings and appeals to conventional ideas of truth, reference, meaning, context, fact, event, and self. “No, I am describing what is the case,” he replies to a hypothetical reader, as if “describing” and “case” were self-evidently stable enough to enforce the assertion. He calls a certain argument “a historical distortion,” as if the bias of “distortion” could be measured by any honest historian. Again, generalizing from his own moral capacity, he speaks of “the appalled rational self” and assumes that the phrase can bear any amount of analysis. In the same spirit he makes a concession which, if pursued, would lead him back to fellowship with the familiar essayists of the first tradition:

No judgment is purely historical. To achieve an integrated view of the life and letters of writers who interest us—by deriving a character from their written legacy and adding elements of their biography—means accepting at some level the “metaphysical” assumption that identity is personal, stable rather than intermittent, and transparent enough for scrutiny.

But this, too, is occasional. Hartman has set up an opposition between two styles of criticism and between the ideological axioms on which each of them supposedly depends, but he is too honorable to allow the invidious comparisons to stand.

He wants to retain both styles, or rather the best of both. Of the first he says, “It is crucial not to give up that public and conversational mode.” He is also greatly taken by the other style, as he finds it in Heidegger, Jabès, Blanchot, Derrida, and de Man. At one point he proposes to get the best of both styles in the form of “strange conversation,” “with the critic burrowing from within, or swerving into unexpected turns, or breaking the form altogether, as if that had been monologic not dialogic, and conversation had to be reconstituted at quite another level.” This is, as one would expect, a fair description of Hartman’s own books.

The most moving pages in Minor Prophecies come near the end. Hartman is perturbed by what he regards as an intensified demand for “a didactic, ethical, action-oriented criticism”:

Cultural criticism has become more complex and circumstantial than in the period between the World Wars, yet it still tends to see through texts rather than with them. And there are signs that new days of rage are upon us, on the part of those who are so much in touch with reality that they do not have to be in touch with language. The best or worst that might be said about critical essays is that they are, despite themselves, minor prophecies that overread the signs of the times. They are also, however, counterprophecies that light up contrary indications and the merely potential relation of the world of discourse to the world. In literary criticism the naming of what may come is not prophecy so much as the branch of a poetics described by Wallace Stevens, one that maintains us “in the difficulty of what it is to be.”

The book ends there, but I assume Hartman will meditate further on the issues he has described. I would like to take part in a strange conversation with him on the passage just quoted; on the meaning and force of “reality” and “language” in that invidiously comparative second sentence; on the character of “signs” that “days of rage” are impending; on the word “world” in “the merely potential relation of the world of discourse to the world.” The meanings of too many central words are left dangling.

Hartman would, I suppose, put Frank Kermode among the familiar essayists and think of him as representing the best of that tradition. Kermode would be content with that arrangement. He thinks it good that professors write for the newspapers and magazines, and he quotes with approval one of the less offensive precepts of Lord Chesterfield: “Speak the language of the company you are in; and speak it purely, and unlarded with any other.” His own style assumes a social consensus, it addresses men and women who are deemed to share common sense, literacy, a certain span of public allusion, and willingness to be interested in ideas upon which the survival of our species does not necessarily depend. Never gruff, he becomes irritable when he finds colleagues refusing to mind their own business. In a review-essay on Marshall McLuhan he says:

He was a pioneer in the movement, now seemingly irresistible, which has carried English literature professors out of literature into larger and, one must suppose, more exciting studies—philosophy, law, psychoanalysis, history, “culture,” “theory” and prophecy.

That was written three years ago, and it marks the occasion, I take it, when Kermode gave up his longstanding interest in Theory. An essay on Paul de Man, written in 1989, has this passage suggesting why he did so.

Furthermore, even if one breathes the air of pure theory, it must sometimes seem strained to argue that it is always impossible to say what one means, even if the statement you wish to make is that it is always impossible to make such a statement; or to combine this belief with the belief that one can and should intend to say, and say, what will make sense of de Man’s life as a whole.

That sentence indicates, and is evidently intended to indicate, how far Kermode is willing to go to placate the opposition, and how wearisome he finds the necessity of doing so. In another essay of 1989 he permits himself to cry “Enough!”

Anyone presuming to review works of modern literary theory must expect to be depressed by an encounter with large quantities of deformed prose. The great ones began it, and aspiring theorists usually carry their heads grotesquely to one side in emulation of these models. What begins as servile mimicry soon becomes a pathological condition.

Wishing to put the case of Paul de Man behind him, Kermode urges others to do the same:

Trilling’s students, when he introduced them to the abyss of the Modern, gazed into it politely, said “how interesting!” and passed by. Others may do the same to de Man’s abyss, and carry on thematizing and totalizing because it is their pleasure to do so, even if it is shamefully human to do so; and they have a long history of resistance to puritanical imperatives. As a rule they will do so without reference to the youthful errors of Paul de Man, and the insiders should now be happy to stop worrying about them and get on with their necessary and impossible projects.

So Kermode gets on with his own necessary and possible projects.

The Uses of Error is an ample selection of the essays and reviews he has published over the past several years. Some of them are tributes to the scholarship of others. Kermode has always been a scholar in his own right, and he has admired the great ones who spend their lives taking possession of a subject. When he mentions Frances Yates, Erwin Panofsky, Ernst Kantorowicz, Aby Warburg, among others, it is with a degree of reverence never elicited from him by the work of a mere critic. His spiritual home is the Warburg Institute, where the primacy of scholarship is taken for granted. His own contributions to learning have referred mainly to Shakespeare, Spenser, and Donne; and, among the modern authors, to the literature of the 1890s culminating in Yeats. In recent years he has given his talent to the study of the Bible. His main interest has always been in the history of ideas and images, and in what happens to these when they are brought to their destiny as form.

The new collection includes reviews and essays on Augustine, Joachim of Fiore, the Bible, Shakespeare, Renaissance politics; a Victorian section (Tennyson, Gissing, Wilde); and a large showing of moderns and contemporaries (Wells, Forster, Orwell, Cyril Connolly, Betjeman, Roy Fuller, Umberto Eco, Julian Barnes, and many more). There is also an autobiographical section, so interesting that I urge Kermode to write a full memoir, not merely snatches of one.

As in his books, so now in these pieces, Kermode likes to show that we live in continuity as well as in change. He doesn’t report that there is nothing new under the sun, but that the degree of novelty, in a particular case, has been exaggerated. Responsive to good work in any form, he is also cordial to those people whose personalities exceed their works. Cyril Connolly, for instance. Kermode recalls with notable gratitude how, during the War, he waited for each issue of Connolly’s Horizon, a magazine all the more to be treasured because it took so little notice of the war. Of The Unquiet Grave, which Connolly published in 1944 under the pseudonym “Palinurus,” Kermode writes:

Palinurus was Aeneas’s pilot. One night he fell asleep and disappeared overboard; reaching the shore, he was murdered by savages. Aeneas met him on his visit to the underworld, when Palinurus asked to be properly buried, worried that he had only an unquiet grave. This ancient hero, a skilful failure, fascinated Connolly, who published under the name of Palinurus an account of opportunities refused, of women pursued, caught, and lost—a series of poignant, silvery, self-pitying meditations on his life and culture, on his beautiful failures that made success seem vulgar.

Much of Kermode is in that word “silvery,” which alludes without fuss to the Latin writers of the Silver Age whom Connolly loved, and to the strict inferiority of silver to the gold that Connolly never in any sense achieved.

My favorite essay in The Uses of Error is the one that gives the book its title, a weaving together of several related works in literature and art; a passage from Job, St. Jerome’s Latin translation of it, the version of it in the English Bible, Georges De La Tour’s painting “Job Visited by His Wife,” Muriel Spark’s novel The Only Problem. Kermode is concerned with the transformation of one text by another, an act of interpretation in which a new sense of the original is produced, “antithetical but acceptable.” His example is the passage about Nicodemus in St. John’s Gospel, in which Nicodemus, “a ruler of the Jews” came to Jesus by night, saying that he recognized him as a teacher sent by God. The anecdote is transformed in Henry Vaughan’s great poem “The Night”:

Wise Nicodemus saw such light
As made him know his God by night.
Most blest believer he!

The point of this is to say that “the history of interpretation, the skills by which we keep alive in our minds the light and the dark of past literature and past humanity, is to an incalculable extent a history of error.” But Kermode wouldn’t have it otherwise, or try to correct the error. Never one to drive the reader into a corner or to insist upon an impasse, he immediately revises “error” to read a history “of ambiguity, of antithetical sense.” To recall a distinction from an earlier book of Kermode’s, there are carnal readings of a text, which are all the same, and spiritual readings, where the differences come. There is no reason to beat our breasts or lie awake sweating about indeterminacy. In effect, Kermode’s advice is Emerson’s: “Do your own thing.” All he asks is that the thing be also interesting to someone else.

Hillis Miller’s axiom in Versions of Pygmalion is Language; but he is concerned to show that the consequences of this choice, for reading, are morally strenuous. His book is an extension of The Ethics of Reading (1987), in which he indicated what such a reading would entail. In the third chapter of the new book he writes:

One of the major functions of literary theory is as a critique of ideology, that is, a critique of the taking of a linguistic reality for a material one. The ideology in this case involves the hidden assumptions of our procedures of teaching and of the general institutionalization of literary study. So theory has become more and more a subject of study for its own sake. But literary theory is of little or no use unless it is “applied,” used. It must be active, productive, performative. What theory produces ought to be new readings, in the broadest sense of that word.

Hence, in Versions of Pygmalion, his readings of Henry James’s What Maisie Knew and “The Last of the Valerii,” Melville’s “Bartleby the Scrivener,” Kleist’s “Der Findling,” and Blanchot’s L’arrêt de mort.

In the tenth Book of Ovid’s Metamorphoses, Pygmalion creates a statue of a girl. Gazing at it in wonder and love, he transforms it into Galatea, brought to life by his desire: she is, as Miller says, “the mirror image of his desire.” Miller associates this creation with one of the cardinal rhetorical figures, prosopopoeia, the trope which ascribes a face, a voice, a name to the absent, the inanimate, or the dead. As in Hardy’s poems about his wife, prosopopoeia is a device to compensate, however inadequately, for death, the ultimate loss. It is the figure of conjuring or summoning.

Miller associates it further with the “performative” capacity of language, that is, the utterance that allows the speaker to do something by means of the words spoken. I now pronounce you man and wife. Performatives, as in J.L. Austin’s How to Do Things with Words, are distinguished from constatives, which present a true or false account of the facts of the case. Miller wants to emphasize the performative character of language, its capacity to do things, because he wants to hold speakers, writers, and readers responsible for their actions. As in The Ethics of Reading, his motto is the sentence in the Preface to The Golden Bowl in which James says that “the whole conduct of life consists of things done, which do other things in their turn.”

The first consequence of Miller’s terms is that his reading of What Maisie Knew and “Bartleby the Scrivener” is surprisingly conventional. It is a slow, thick reading, dense with interrogation and implication, but it has the effect of recovering the ideology of character, identity, and personality which many theorists forbid us to invoke. Galatea has a name; so have Maisie, Mrs. Wix, Bartleby, and the other characters. Having a name, they apparently can’t be denied other attributes, qualities of moral bearing and insistence. After a while, in Miller’s readings, it wouldn’t seem bizarre to ask how many children Lady Macbeth had, and was she a caring mother; and how did Bartleby come to be so obstinate? Wanting conclusions that bear on the performance of moral acts, he asks questions that seem to me intrusive:

Just how is the reader to evaluate Maisie’s act? Is it ethically admirable? Should we try to emulate it in our own life? What would it mean to do that?

I think it sufficient to imagine acting as Maisie acted, and excessive to interrogate her.

Miller is very acute in showing how Maisie conjures into a semblance of existence certain ideal persons by comparison with whom she judges each of the characters surrounding her. Maisie’s capacity to project these ideal persons is what James calls her power of “wonder.” The novel has to end when Maisie has outgrown this capacity; after that, as James says, there would have to be “another scale, another perspective, another horizon.” Miller deals splendidly with these matters, but he needed no very elaborate theory to achieve this result. His has always been a scruple that reads well by reading slowly.

Again, in his reading of “Bartleby the Scrivener,” Miller is patient, perceptive—as we know him always to be—but his purposes are not really different from those of other readers. He tries to figure out what makes Bartleby tick and to make sense of what Elizabeth Hardwick called Bartleby’s “resistance to amelioration.” Indeed, Miller’s procedure here is so like any other careful reader’s that he disappoints me by proposing yet another interpretation of Melville’s story: Bartleby is “a ghost, a walking dead man, a living statue, a kind of zombie possessed by who knows what malicious spirit.” This seems to me to let the reader off the hook; it domesticates the mystery by removing the provocation of it.

At a certain point in the book, Miller seems to realize that Pygmalion, prosopopoeia, and performatives have committed him to an old-fashioned way of reading; complete with recognition of characters, plots, motives, and so forth. At that point, if I read him aright, he turns to Blanchot’s strange fiction, L’arrêt de mort (Death Sentence) as another other critic might for similar reason turn to Beckett, Borges, or Calvino, to force him out of his vocabulary. Blanchot is a novelist and critic. As a critic of fiction he is best known for those essays in The Gaze of Orpheus which demonstrate the force of the neuter in narrative. What Kafka teaches us, according to Blanchot, is that “story-telling brings the neuter into play.” Kafka’s use of the neuter helps us see that in the realistic novel the novelist refuses to say “I” but delegates that power to other people. As Blanchot says in “The Narrative Voice”:

The novel is filled with little “egos”—tormented, ambitious, unhappy, though always satisfied in their unhappiness; the individual asserts himself in his subjective richness, his inner freedom, his psychology; the novelistic narration, that of individuality—not taking into consideration the content itself—is already marked by an ideology to the extent that it assumes that the individual, with all his particular characteristics and his limits, is enough to express the world, that is to say, it assumes that the course of the world remains that of the individual.

Blanchot is thinking of the novel so far as it fulfills the requirements of one of Miller’s three grounds of narrative: individual psychology. But in the novel as Kafka leaves it, according to Blanchot,

narrative governed by the neuter is kept in the custody of the “he,” the third person that is neither a third person nor the simple cloak of impersonality. The “he” of narration in which the neuter speaks is not content to take the place usually occupied by the subject, whether the latter is a stated or implied “I” or whether it is the event as it takes place in its impersonal signification. The narrative “he” dismisses all subjects, just as it removes every transitive action or every objective possibility. It does this in two forms: (1) the speech of the tale always lets us feel that what is being told is not being told by anyone: it speaks in the neuter; (2) in the neuter space of the tale, the bearers of speech, the subjects of the action—who used to take the place of characters—fall into a relationship of nonidentification with themselves: something happens to them, something they cannot recapture except by relinquishing their power to say “I”…

Miller needs Blanchot’s commentary not only for its bearing upon the N. and J. in L’arrêt de mort but to rescue him from the ideology of individual psychology into which, in his account of What Maisie. Knew, he has—not stumbled but—marched. He needs to be able to say that in L’arrêt de mort (and if there, elsewhere too) fiction can reach “beyond personification to the voice that is not so much impersonal as neuter, ‘ne-uter,’ neither the one nor the other.”

But he still wants to keep Blanchot’s neuter within the performative mode:

Blanchot’s implicit contribution to the theory of speech acts is a demonstration of the existence of this third possibility, neither the red language of contracts, promises, excuses, christenings, and marryings in the “real” world, nor the white language of frivolous or “poetic” performatives that make nothing happen. These speech acts are expressed in the neuter of a fictitious language that nevertheless is productive. Such performatives are detached, like a translation or like a formula I utter in a foreign tongue, from all the subjective, intentional, and social contexts that tie the Austinian performative to the presupposition of a speaking and intending “I.” Still they function to call forth something—it, the thing, the thought, the truth, a truth beyond any “I” or any social context.

I’m not convinced by the last part of that passage, with its claim that Blanchot’s neuter fictions are detached from all contexts. It is impossible to speak or write a word without having it draw some of its historical and social experience into the occasion of the performance. Even the words of Mallarmé’s “Un Coup de dés” are old warriors by the time they come to his poem, and they retain their scars between the white spaces on the page. But Miller’s commentary is immensely helpful in one of the most difficult problems of narrative theory; how to deal with passages, as in omniscient narration or “free indirect style,” which can’t be assigned to a speaker. These passages seem to involve acts, but not agents.

I note that Miller speaks of Blanchot’s neuter as a third possibility; so it does not veto the other two. A writer remains free to choose and then to take responsibility for the choice. Nothing in Miller’s book undermines this freedom. So it seems to me a work of edification, not of deconstruction; it is edifying as an example of attentive reading, practical criticism. Miller chooses his companions, and they include de Man for his work on Kleist in Allegories of Reading, Blanchot for the fiction and theory of narrative, Derrida for Psyché: Inventions de l’autre. But he does not claim that these are the only companions a critic may decently choose.

So where are we now? We have known since Northrop Frye’s Anatomy of Criticism that there are at least four forms of fiction, five if we count Finnegans Wake. The four are: novel, romance, confession, and anatomy. Each implies a different sense of life, a different ideology of understanding. I assume that criticism is willing to be just as diverse as fiction. There are many forms of criticism, far more than my reading of Howe, Hartman, Kermode, and Miller has suggested. None of the axioms on which a work of literature or a work of criticism proceeds is free of ideology: and in that respect none of them is shamed in the presence of the others. There is a question of competence, of knowing what one is doing and being able to do it. T.S. Eliot once remarked that the best method in criticism is to be very intelligent. The four critics I have been reading are masters. There is also a question of freedom, the right to do one’s own thing. None of the books under consideration reduces one’s freedom, or brings us to a new day of rage.

This Issue

August 15, 1991