“This book,” Michel Foucault says at the beginning of The Order of Things, “has its origin in a text by Borges”; and indeed Borges, the author of delicate, lucid, disturbed visions of what used to be an ordered world, has become something like the patron saint of much recent French writing. A character in Borges’s story “Death and the Compass” seems even to anticipate a French connection. An amateur detective in the tradition of Holmes and Dupin, confronted with the flat-footed police inspector so familiar in such fiction, Erik Lönnrot dismisses the inspector’s simple view of a murder as “possible, but not interesting.”

You will reply (the great man continues) that reality is under no obligation to be interesting. I shall reply that reality can get along without that obligation, but a hypothesis can’t.

Later in the story, when Lönnrot has solved a complicated series of crimes more to his intellectual liking, we are told that “the mere circumstances, the reality (names, arrests, faces, judicial and penal arrangements) scarcely interested him now.”

I’m not suggesting that Foucault and Lévi-Strauss, say, are not interested in reality—Foucault is notably interested in judicial and penal arrangements—but I think it is true that they are impatient with mere circumstance undignified by theory. And the reason I mention this is that the Anglo-Saxon prejudice appears to be exactly the reverse. The first step into theory, we feel, is the first step in the rejection of reality, and we may wish to insist that Erik Lönnrot, in Borges’s story, is killed because he wouldn’t listen to the flat-footed inspector. The inspector was right about the murder, and the series of further crimes was a trap laid by a man who knew Lönnrot’s taste for intricacy and used it to lure him to his death.

So they do order these things better in France, as Sterne said, even if the order may seem to come at rather a high price. The contrast is striking in literary criticism. Roland Barthes, for example, who is in his way a practical critic, regularly falls into the waiting arms of theory, while the theories we have, like Empson’s seven types or Bloom’s six ratios, seem to have been knocked together in the critic’s back room, without benefit of tradition. Even Northrop Frye, no doubt the most distinguished literary theorist writing in English in this century, can sound more pragmatic than the toughest of pragmatists. “Whatever is of no practical use to anybody,” he says in his Anatomy of Criticism, “is expendable.” So there.

Frye is an attractive and complex case, though. Anatomy of Criticism (1957) started a vogue for the study of literature as myth, but it also offered, with considerable panache and remarkable good sense, to link up, by means of modes and genres and symbols and archetypes, a number of discrete areas of literary scholarship. “It is better to think,” Frye suggested, “not simply of a sequence of meanings, but of a sequence of contexts or relationships,” and he set out to map the major contexts and relationships of literature. Oliver Twist, for example, is a novel of a certain kind but it is also a variation on the theme of the foundling child, which goes back through Tom Jones and much popular fiction to Oedipus and Moses.

Frye was not trying to tell critics what to do, he was trying to indicate what criticism, a multiple and scattered activity at last seen as a single enterprise, might already be doing. The approach was appealing to scholars working in dark empirical corners without much sense of where anything led; and it was much opposed by people who felt that Frye was asking them to sacrifice the life of particular literary texts on the altar of a strange abstraction called “literature.” I suppose I belong to the second group myself, but I have always admired the sweep and the generosity of Frye’s understanding, and in any case an interesting puzzle remains.

A teacher of literature fully persuaded of the social and moral significance of his discipline, Frye is given to suggesting that literature is not about “life,” or at least to separating himself from those critics who think it is. Partisan of a progressive approach to criticism, he is the creator of a powerful regressive myth, which tips literature back into folklore at every other moment.

What happens, I think, is that Frye’s practice is usually more flexible than his theory, or to put that in a more complicated way, that his theory is often a displaced form of practice after all. We get a clear sense of this in A Natural Perspective (1965), Frye’s study of Shakespearean comedy and romance. It may be worth following out a piece of its argument.


Frye first distinguishes between himself and those he calls “moral” or “allegorical” critics, who clearly have designs on literature:

They feel that its essential function is to illuminate something about life, or reality, or experience, or whatever we call the immediate world outside literature.

He then suggests that popular fiction “tells us little that is credible about the life we live in”—although the crucial words here are tell and credible, since Frye is reserving the right for popular fiction, when it surfaces in the plays of Shakespeare, to show us things that are entirely incredible about the life we live in. And we arrive finally at The Winter’s Tale, which portrays in its conclusion not reality but something a good deal more real than that:

…it is the world we want; it is the world we hope our gods would want for us if they were worth worshipping.

It is, Frye says a moment later, “an imaginative model of desire.”

Now Frye, in spite of these brilliant and moving abstractions, has just said that in Shakespeare there is “nothing to be abstracted from the total experience of the play.” Frye’s theory claims, rather stiffly, that literature is either about life or about itself, and Shakespeare’s plays for him are about themselves, tell their stories for the sake of the story. But his practice suggests that literature may well be about life because it is about itself, and treats the reality of desire—how could that not be a part of “life”?—with eloquent affection.

Some of the difficulty comes simply from the ordinary rigidity of conceptual language, which tends to convert assumptions into assertions. What Frye calls the “large simplifying device”—the grand dichotomy, or the vague general scheme—has the bad habit of forgetting that it is a device. But there is more than this. Frye often prefers tidiness to reality, which is another version of Erik Lönnrot’s problem, and one of the best ways of being tidy is to be complete.

Frye’s talk of “the whole of literature,” of “literature in general,” his quest for what lies “at the heart of all literature,” his taste for identifying “the two major structural principles in fiction,” or the “four primary narrative movements in literature,” of which “all stories in literature” are complications or derivations—these things are plainly as much a matter of temperament as they are a question of method. When Frye retreats, as he says in A Natural Perspective, “into a middle distance,” one has the feeling that he is relatively close to the text. There is a revealing passage in Frye’s elegant Norton lectures, now published as The Secular Scripture. “The happy endings of life,” Frye says, “as of literature, exist only for survivors,” and he adds that “after all” the Nazi regime in Germany did collapse in a few years, so that “the survivors have the more complete perspective.” He is not being unkind or thoughtless, and of course what he says is true, unless there is an even “more complete perspective” in which it all happens again. The snag is the word complete. Do we want completeness in such contexts? Aren’t there a dozen other things we would rather have?

But completeness is what Frye wants. In The Secular Scripture he explores the “imaginative model of desire” again, and this time he is after the whole story, “the epic of the creature,” the tale that fiction tells when we put it all together, the secular fable of man on his own, as distinct from the sacred myth of man in the hands of God. “A mythological universe,” Frye says, “is a vision of reality in terms of human concerns and hopes and anxieties.” It is a projection, that is, not an imitation. And the universe of fiction, the secular scripture, is best seen, Frye thinks, as a vast romance. By romance he means what we usually think of as belonging to that tradition—Spenser, Scott, William Morris, some of Chaucer, the late Shakespeare, most science fiction—but also a flickering element in all kinds of literature, in Jane Austen, for instance, or George Eliot, or Kafka, or virtually anywhere.

Romance characteristically throws its heroes and heroines into a variety of tight spots, but it gets them out again at the end. After various travails, the story Frye has pieced together offers us a portrait of a human community which goes well beyond the happy social reconciliations of comedy, indeed which seems closer to Rabelais’s noble abbey of Thélème than to any community we are likely to know:

…romance’s last vision seems to be that of fraternity, Kant’s kingdom of ends where, as in fairy tales, we are all kings and princesses. The principle of the aristocracies of the past was respect for birth; the principle of fraternity in the ideal world of romance is respect rather for those who have been born, and because they have been born.

Now the authoritative objection to what Frye calls “the total story of romance,” or for that matter the total story of any literary mode, or of literature itself, was made by W.K. Wimsatt, in a paper reprinted in his Day of the Leopards. None of this is “historical fact,” Wimsatt says sternly, and continues:


And this is so not only in the obvious sense that the stories are not true, but in another sense, which I think we tend to forget and which mythopoeic writing does much to obscure: that such a coherent, cyclic, and encyclopedic system, such a monomyth, cannot be shown ever to have evolved actually, either from or with ritual, anywhere in the world, or ever anywhere to have been entertained in whole or even in any considerable part.

I am myself rather suspicious of Frye’s seeing so much in the whole of literature and so little in the individual works he must pillage for his grander purpose, but I find I do recognize the community he evokes as the conclusion of his romance. If I may give examples of my own which seem to me securely rooted in the books they come from, the wished-for world is the liberated Italy we see at the beginning of La Chartreuse de Parme, it is the paradise of unpossessive love which Catherine Earnshaw throws away in Wuthering Heights, it is the country beyond money and dirt and decay which Dickens hints at so insistently in the fairy-tale structures of Our Mutual Friend.

“Romance’s last vision,” wherever it is found, thus performs the task that Frye sets for all fiction, that of countering the absurdity of death with the absurdity of coherence. Scheherazade, Frye reminds us, was telling stories for her life. Her stories were her life.

Perhaps even the storytellers in Boccaccio’s Decameron are not simply running away from the plague that rages outside their walls. “Once upon a time”: the formula invokes, out of a world where nothing remains, something older than history, younger than the present moment, always willing and able to descend again once more.

I don’t like older than history much, and I prefer Denis Donoghue’s formulation for a similar enterprise, which helps to “preserve as fundamental truths the values which are otherwise homeless.” For this is what Frye’s view of romance does. It reclaims an ideal in the face of its constant betrayal, and it is in history that we need such visions, if we can hang on to them. The true absurdity is not death but the appalling lives so many people lead, and the world is there to be changed rather than transcended. But if we didn’t have pictures of transcendence, we should probably not remember why we want the changes, and the main point here is the urgency with which Frye’s large schemes, however wide the loops they make, finally speak to the “life we live in”; the vigor with which Frye pursues the traces of desire in literature.

Spiritus Mundi—the title is taken from Yeats’s “Second Coming”—is a collection of twelve essays which investigate and summarize, with much charm and some petulance, Frye’s continuing concerns: the liberal uses of education, the life of books, his sense of “all literature” as caught up in a “mythological universe.” There is a remarkable essay on Yeats’s A Vision, and another on charms and riddles, where Frye displays prodigious learning as well as his usual intellectual agility. There are personal notes too, indications of the relevance—to use what Frye calls a “neo-Nazi slogan”—of Spengler and Frazer, and passages like this one, which perfectly illustrates Frye’s critical habits and tone:

I have had some influence, I know, but I neither want nor trust disciples, at least as that term is generally understood. I should be horrified to hear of anyone proposing to make his own work revolve around mine, unless I were sure that that meant a genuine freedom for him. And if I have no disciples I have no school. I think I have found a trail, and all I can do is to keep sniffing along it until either scent or nose fails me.

If Frye’s elegance, here as elsewhere, sometimes gets the better of him, this serves to show him, once again, on his personal perch between theory and practice. “Life has no shape; literature has,” he writes, sounding like Oscar Wilde on an evangelical day. He claims the highest practical validity for literary theory, since it enables us to see how literature supplies the many deficiencies of mere “life.” Important as this claim is in the respects I’ve suggested above, it does of course bring us perilously close to the world of Erik Lönnrot. It is true, no doubt, that life has no shape, but “like most truths,” as John Bayley says on another subject in his Characters of Love, “it is not as true as all that.” It is also true that life is full of shapes we don’t like or can’t recognize.

Bayley himself, in The Uses of Division, suggests a working notion rather than “anything much in the way of critical ‘theory’ “—the quotations marks around the suspect word amount almost to a sneer. His notion is division, the sense of a writer at odds with himself, which is probably a more common critical principle than Bayley seems to think it is. It is apt to produce a certain quantity of melodrama about what goes on “deep down” in a writer, at a “deep dark level,” to borrow phrases Bayley uses of Jane Austen and Kipling, not the deepest and darkest of authors perhaps, but the notion clearly makes sense. The difficulty is to avoid suggesting that the divisions are really healed in the “equilibrium” or the “artistic whole” of the work—again the words are Bayley’s—since then you sacrifice a vivid disunity to a rather dull old sense of “form.” In the best essays in this book—on Kipling and Keats—Bayley keeps all the necessary contradictions wonderfully alive, finds several Kiplings marching side by side, and explores the ways in which the “badness” of Keats’s poetry, the energy of its more than occasional vulgarity, is inseparable from its greatness. (“Hyperion is, so to speak, not bad enough, too full of hardwon decorum.”)

Bayley is not always considerate toward his readers, and at times The Uses of Division made me feel I had stumbled into the middle of a tutorial, or picked up a book review halfway through, but generally, here as in his earlier books, Bayley is as intelligent and thoughtful and decent an instance as one could wish of how much a critic can do without much in the way of critical theory, and he understands as few theorists do what he calls “the real and unforgiving dimension of history.” Admittedly, such a critic has only to relax his vigilance, or not to have reread his texts lately, and he will slip into gossip. This is what happens when Bayley compounds silliness with non sequitur by telling us that “Proust is nothing if not a social scientist” and then informing us that “his father and brother were doctors.”

But the case for theory and Erik Lönnrot is not simply that hypotheses have to be more interesting than reality. It is that we can’t make any sustained sense of reality without them—reality for the critic being the bundles of words we call literature. A person who says he has no theory is usually the prisoner of several conflicting and unacknowledged theories. Even so, I do think good books are livelier and more shifting than the theories we chase them with, and I would want to distinguish between theoretical curiosity in a critic, which seems to me nearly essential, and the possession of a particular theory, which is probably the last thing he needs. It is true, as the structuralist theoretician Tzvetan Todorov has said in a recent interview,* that poetics and criticism are different things. But Todorov adds that there’s no reason why the same person can’t pursue both. What we need, I think, is not a separation of critical and theoretical powers and not some fraudulent dream of a happy medium between the two, but a ground where theory and practice can talk to each other.

This is precisely what Denis Donoghue’s new book offers. The Sovereign Ghost contains seven essays, which mediate on the various uses and activities of the imagination. Donoghue takes for granted an interest in literary theory, but he also takes for granted close attention to particular literary texts, and the result is very alert and engaging work, attuned to theory but not lost in it. Donoghue argues for the imagination where Lionel Trilling, for example, argued for the “mind,” because the imagination seems to Donoghue a larger and more liberal faculty. “The imagination is the name we give to the mind when it is prepared, if feeling requires it, to see everything change except itself.” “Imagination would then mean the act of a mind determined to keep open every possibility of creation, freedom, play, and pleasure.”

Sometimes Donoghue’s “imagination” seems synonymous with the capacity to write or read literature, which would make it something like the author of Frye’s secular scripture: “Imagination is the intellectual power which we call upon to unify the whole range of experience, however heterogeneous that range may be.” But Donoghue is not after large configurations for their own sake. He wants arrangements of ideas which will allow him to look at a series of relations between the imagination and the world—in Shakespeare, in American literature, in Pope, in Wordsworth, in T.S. Eliot, in Allen Tate—and his arrangements never displace the texts themselves. Donoghue pauses to note internal rhymes in Dante, for example, or a literary rat in The Waste Land:

The rat creeping through the vegetation has only as much to do with animal life as is required to incite a certain feeling in the speaker. The rat has crept into the words and lost itself there….

He explains why “Where did you die?” in one of King Lear’s last speeches is preferable to the sensible alternative “When did you die?”: “because the entire scene turns upon place and the dislocation of the King’s mind…. The next time Lear speaks he asks, ‘Where have I been? Where am I?’ ”

In the title essay, Donoghue defends his own version of the liberal imagination against the structuralists’ more depleted and democratic interpretation, where the speaking subject (and hence the imagination) is not a given authority, but an identity to be constructed out of the language it uses. Donoghue grants the truth of this picture—it is a way of describing reading—but objects to the “implication that the subject is merely the sum of its occasions.” And when Jacques Derrida launches his attack on the hegemony of the voice and the rule of consciousness, because voice and consciousness imply an illusory and reactionary notion of human freedom, Donoghue sees the point but calmly differs:

I cannot regret either the privileged status of consciousness or the idiom of its privilege; it seems to me to be sustained by our sense of the bodily condition of our lives, and of our relation to time and earth.

What Donoghue is saying here, and earlier in the same essay, is that structuralism and its cousins represent an option, not an enemy or a promise of salvation. “If I were required to make a leap of faith,” he says, “only one leap being allowed, I would aim for the Romantic assumption which features imagination and subjectivity as primary terms…. But…I am not persuaded that we have come to the stage of a single choice and battle stations.” I’m not sure I would make exactly the same leap, but Donoghue’s general position here seems to me exemplary. If there is a fence we should choose sides, not sit on it; but we shouldn’t hurry to see fences where there may be only bridges we haven’t managed to meet on. Donoghue’s practice, even more than Frye’s or Bayley’s, shows us how to tackle theory without succumbing to Erik Lönnrot’s scorn for mere circumstance. Reality is under no obligation to be interesting, but a hypothesis is not obliged to lead us into labyrinths from which there is no escaping.

This Issue

April 14, 1977