Can people be corrupted by made-up stories?[1] Vocabulary shifts aside, this question has been a heated one ever since the character Socrates, in Plato’s Republic, alleged that tragedy makes us soft by accustoming us to giving free rein to our feelings, and since Aristotle wrote—very likely in response—that tragedy actually purifies our emotions, in some vague way that has forever remained unspecified.[2] Succeeding centuries have witnessed all kinds of variations on the same basic disagreement, right down to contemporary claims that, on the one side, literature is the one and only mechanism for developing empathy for those not like us and, on the other, that literature is a tool of the oppressive hegemonic régime, an ideological apparatus designed to keep the populace in a catatonic trance.

It’s a little hard to find one’s way in all this, but I want, here, to say both yes and no. First the no: I want to deny to literature the inherent power to make its readers worse; such power, I’ll suggest, is granted only by institutions of reading, and it can easily be taken away by the same institutions. If individual works corrupt individual readers, this is a mark merely of the incompetence of reader, institution, or both. I will add, however, that there’s a potentially troubling corollary to this perhaps comforting thought. Second, the yes: I want to suggest that there is a particular way of being corrupted by fiction, one that we need to be especially wary of, an insidious temptation that comes along with the deep love of literature that, I assume, all literary critics once possessed, and that some are lucky enough to retain.

1. Imaginative Resistance

Some contemporary philosophers strongly imply that no such corruption is possible. There is, they say, a thing called “imaginative resistance”: when presented with a fictional world in which, say, murder is good, we find ourselves unable (or at least unwilling) to imagine it. As Tamar Gendler puts it, “I have a much easier time following an author’s invitation to imagine that the earth is flat than I do following her invitation to imagine that murder is right.”[3] And by conse-quence, I am entirely safe from any debasement at the hands of the work.

Is this correct? We might object that to use the word “murder” is already to stack the deck, since “murder,” unlike “killing,” is a moral term. What if we called it, say, “taking care of”? What, in other words, if we consider the case of mafia fiction, in which the very worst thing one can do is to report crimes to the police, and the very best thing one can do is, at times, to “take care of” an unarmed human being, someone whose only blemish is, perhaps, to have reported crimes to the police?

It is in fact a fascinating feature of certain mafia movies, and virtually all outlaw movies, that they perform an imaginative “revaluation of values” without us resisting in the slightest. Mummy movies, incidentally, have a related effect: when the ultra-rationalist—the one who insists loudly that there is no such thing as mummies—is the first to be strangled to death, we feel no sorrow, since obviously he should have known better. For the duration of the movie, we are people who would rather spend time with believers in the paranormal than with seekers of fact; we are people whose firm conviction it is that to base one’s judgments on logic and empirical evidence is to merit extermination. And perhaps this attitude, which we could term “imaginative inertia,” is the standard case. Far from resisting the different, sometimes opposite, values of the fictional world, we positively delight in trying them on for an hour or two.

2. Symmetry

So you might think: fictions can corrupt. If I watch too many mafia movies, I will start to believe that reporting crimes to the police is a dreadful thing, and killing—sorry, “taking care of”—those who do so is a glorious shining virtue. We should be wary of fiction (or I suppose excited about fiction, if you’re the kind of person who wants everyone to be corrupted; though I suppose if you want people to be corrupted, it’s not clear that this would any longer count as corruption, on your definition). Still, consider again the example of mummy movies. After I watch the 1959 version,[4] with its wonderfully caricatural mummy-denier, am I actually shamed into treating mummy-believers with greater respect in the future and exhibiting the proper dread at Tutankhamun exhibitions? I don’t think so. I became a different person during the viewing experience, but this has had no lasting impact on me—perhaps precisely because I became a different person— in my everyday life.

It is far from clear, in other words, that fictions can, by themselves, lead a reasonably sensible reader astray. I’ll come back later to these two qualifications—the sensible reader, the fictions acting by themselves—but for now I want to point out an important corollary. The corollary, as I see it, is this: if fictions cannot corrupt, then fictions cannot edify. If I am virtuous, then I will certainly resist the promptings of Sade to rape and torture; but if I am vicious, then I will just as strongly resist the urgings of Dickens to do unto others what I would have done to myself, or to be kind to escaped convicts, or to embrace (heaven forfend) the Christmas spirit. Perhaps I will resist Dickens even if I am good, which is to say even if I share the values his texts appear to be endorsing. After all, there is something about sanctimonious fictions that makes one either burst out laughing, as Oscar Wilde famously claimed to, or respond with indignation. (I like to think of this as the Meg Ryan Syndrome: when you leave an “uplifting” film like Sleepless in Seattle, you may feel like breaking something.)

3. The Ethical Turn

I’m claiming, then, that fictions cannot, by themselves, corrupt a reasonably sensible reader, and also that fictions cannot, by themselves, edify a reasonably sensible reader. Now I recognize that this places me rather at odds with an important recent trend in literary criticism. We are living in the age of the ethical turn, and it has become commonplace to claim that literature’s defining characteristic is its ability to make people nicer (rather in the way, perhaps, that music used to be said to “soothe the savage breast”). Substantial quantities of time and journal space have been dedicated to assessing the precise ways in which literature contributes to a better society.

Some, like Mark William Roche, have suggested that literature makes us better people by apprising us of important moral truths.[5] Indeed it often sounds, in Roche, as though a literary work were nothing more than a glorified mechanism for delivering messages, like a horse-drawn UPS truck. Thus literature tells us things about life (p. 82); it communicates something to us (p. 63); it offers insights (pp. 95, 208, 211); it has an idea (p. 65); it has a message (p. 252); it has something to say (p. 63); why, it goes so far as to make arguments (p. 57), and can, as a result, mislead the unwary (p. 64).

A second group (headed by Lynn Hunt and Richard Rorty) has argued that literary texts foster empathy with an ever-widening circle of human types, gradually bringing more and more of “them” under the designation “us.”[6] And a third group, headed by Martha Nussbaum, has seen literature as fine-tuning our moral decision-making faculty, so that we are better equipped to notice and respond to subtle claims on our moral attention.[7]

In almost all cases, the salutary effect on readers is presented as automatic, inevitable, “inescapable”—as though novels were so many bricks with which to hit recalcitrant unbelievers over the head, in hopes of shaking their skepticism loose. Thus Nussbaum writes that “it is impossible to care about the characters and their fate in the way the text invites without having some very definite political and moral interests awakened in oneself.”[8]

4. Messages

I want to argue against all three approaches, before talking about what I take to be the real improvement and real corruption available at the hands of fiction. Let’s start with the idea that fiction is a mechanism for delivering helpful moral messages. Before I read such and such a story, the argument runs, I didn’t know it was good to do X and bad to do Y; but now I see the light. How convincing is this approach? Well, let’s take an example.

In the Nun’s Priest’s Tale, Chaucer tells the story of Chauntecleer, a rambunctious rooster who has what he takes to be a prophetic dream, warning him not to go out into the yard. His wife, Dame Pertelote, reassures him that nightmares are not portents but merely the result of a bad diet, and that what he needs is in fact “some laxatyf.”[9] Chauntecleer wanders out into the yard, and there, sure enough, is the fox that he dreamed about. Slyly, the fox invites Chauntecleer to sing with that beautiful voice of his and to keep his eyes closed as he does so. When the rooster’s eyes are closed, the fox snatches him up and carries him off. Chauntecleer is only saved when he tricks the fox into bragging about his victory and thus freeing him from his jaws.

What is the moral of the story? Let me tell you: the moral of the story is that dreams are prophetic. This is clearly proven by the fact that the rooster’s dream came true after all, in spite of the foolish cynicism of Pertelote. Not convinced? You think perhaps that the moral is something different—that it is good to be prudent and bad to let arrogance get the better of us? Perhaps you even believe that the story teaches us this, because, after all, Chauntecleer’s arrogance led to his downfall; we must learn from his example to be more humble and circumspect. If so, then notice that the two claims—that the story teaches us to be prudent, that the story teaches us to believe in oneiromancy—have exactly the same structure and exactly the same evidentiary basis. Both are “proven,” quote-unquote, by means of a fictional narrative. (Indeed, a rather ludicrous one: can this story really “prove” anything other than the claim that if you happen to be a talking rooster, you should beware of talking foxes?) And both have the support of the narrator, who chastises Pertelote for not believing in dreams just as much as he chastises Chauntecleer for sauntering into the yard. In fact, the oneiromancy moral has if anything more argument on its side, being supported by the large number of prophetic dreams Chauntecleer cites from the Bible, a book whose authority many, if not most, of Chaucer’s listeners could reasonably be expected to take very seriously indeed.

Why, then, do we want to believe the prudence moral but not the oneiromancy moral? Because the former scenario is realistic and the latter is not. In real life, people (and not just farm animals) are punished for their hubris, but in real life, dreams do not point to the future. Very well. But the problem is that we have assumed the very thing we set out to prove. We say that the story fails to convince us that dreams are prophetic because it offers, as its only evidence, a series of tales that do not seem likely; but the reason such stories do not seem likely is that dreams are not prophetic. In circular fashion, the Nun’s Priest’s Tale convinces us only of what we already believed before we began to read it. Which means, strictly speaking, that it convinces us of nothing at all.

None of this has deterred Chaucer scholars from finding all kinds of edifying messages in the Nun’s Priest’s Tale. As one such scholar puts it, “The [Nun’s Priest’s] Tale could only have been written for a medieval audience which looked at life seriously.... If we turn to the poetry, we can see that it is of a kind which could only proceed from a fine moral concern.”[10] We’re supposed to learn from it not to be like the rooster, or not to be like the Nun’s Priest, or to be like the widow who opens the tale, or to read it as an allegory of the Fall or (finally) of the Church. I actually see this as a mark of desperation. Critics know they are supposed to draw an edifying lesson, and so they make one up; in a sense it doesn’t matter which, as long as the purported piety matches our own.

Now this, elegantly enough, is exactly the procedure of the Nun’s Priest himself. Just like the Chaucer scholars, he assures us that the story has a moral—indeed, he makes explicit what they leave implicit, viz. the belief that every story has a moral. And just like them, he lavishes upon us a whole heap of morals: it’s dangerous to be arrogant; “murder will out”; “destinee ... ma[y] nat been eschewed”; “ever the latter ende of joye is wo”; and even, charmingly enough, “mulier est hominis confusio” (“woman is the ruin of man”).[11]

All of these morals surely cannot be true at once. If it is really true that the world is ruled by divine Providence, such that arrogance is always punished and murder always revealed, then it is not true that joy is always followed by sadness: if God rewards the good, then some joys are surely allowed to endure. Further, if destiny rules the show, then there is not much sense in castigating Chauntecleer for his lack of prudence.

The Nun’s Priest’s Tale is, in fact, a parody of didacticism, a story that reminds us of how extraordinarily easy it is to draw apparently edifying lessons from any narrative. As long as our listeners already subscribe to a particular piety, they will happily take a story to be illustrating it; they will indeed consider it to emerge automatically from the story, as the only possible inference. They will go so far, under certain circumstances, as to consider the story all the evidence it needs.

It is not even the case that we learn this from the Nun’s Priest’s Tale—not even the case, that is, that we learn how ineffectual fictions are as a tool for conversion. The Nun’s Priest’s Tale is a parody of didacticism, but true to its own implicit principles, it fails to teach even the impotence of literary instruction. Had it done so, then there would surely not exist today that voluminous and intensely earnest bibliography of devout interpretations. The Nun’s Priest’s Tale is a story that fails to prove even its own futility—and that, in so doing, vindicates itself.

One final note about the Nun’s Priest’s Tale. At one point, the Nun’s Priest tells us that we should pay attention to his story because it is true.[12] This is a vital moment, a crucial admission: the admission is that fictions lack something that histories possess, namely the ability to prove something possible. That, incidentally, is what explains the James Frey controversy and its recent fellows. Fictions cannot offer any substantiation for their implicit claims, and accordingly it is always open to us to dismiss them as fantastical. This may well be why a group of fifteen conservative scholars and public policy leaders, when asked by the right-wing journal Human Events to name what they considered the “ten most harmful books of the 19th and 20th centuries,”[13] didn’t mention D. H. Lawrence, or Emile Zola, or Bertolt Brecht; they listed Freud, Marx, Nietzsche, Lenin, Hitler, Mao, Comte, Kinsey, Skinner, and Mead. Not a single work of fiction ended up on the list.

5. Empathy

So much for the message view. What, however, about the empathy view? Is it true that empathy for fictional entities guarantees more charitable treatment of one’s neighbors, those at least whose representatives feature in our imaginative reading? Well for one thing, it’s not a given that understanding will lead to compassion. For another thing, it’s not a given that imaginative identification with those who are not like us is an absolute good, such that its presence automatically enhances, and its absence automatically lowers, the moral status of any given situation. Moral behaviour is not just a matter of making benevolent, broad-minded judgments (for example, “people from a different country may be just as nice as people from my country”). It is also a matter of making the appropriate negative judgments (for example, “murderers are not nice people”). This means that if fictions really do train me to be compassionate toward all others, then I may actually start becoming less moral.

If you don’t believe me, just imagine someone who reads Vladimir Nabokov’s Lolita and concludes that abduction and sexual abuse are really not so terrible after all. This person would be a bad reader, made morally worse by excessive identification. For Lolita, in fact, the ideal reader is one who continually stands back from his or her empathy; indeed, the peculiar power of the work derives from the perpetual feeling of unease generated by the oscillation between disgust, connivance, and disgust at our connivance.[14] Moralists will of course tell us that this is quite right, since that is how Lolita teaches us that paedophilia is to be avoided and condemned, and that, in general, we should pay attention to the desires of other people before selfishly using those people as mere means. But how many readers do not already think this before picking up the book? We don’t learn this from Lolita; on the contrary, a healthy moral compass is the price of admission, the price of entry into that vertiginous affective space.[15] And if we did learn that, then this would rather give the lie to the idea that we automatically empathize with novelistic heroes, gradually expanding the category “we” until it encompasses all of “them.” We don’t, and we shouldn’t. It is dangerous to empathize indiscriminately. It is almost as though Nabokov wrote Lolita with no other purpose than to provide a ready example with which to defeat the simplistic view of realist fiction as caritas-based saviour of the world.

6. Fine-tuning

The Nussbaum view—the claim, that is, that literary texts give us practice with morally complicated situations—manages to survive both of these critiques, which is one reason why it is by far the most interesting of the three approaches. Even here, however, there is a problem. Or rather, a pair of problems: the amorality objection and the immorality objection.

Nussbaum assumes that when we take an intensive, vigorous, sustained, detailed, painstaking interest in the moral entanglements of other lives, we do so in a moral way and emerge improved as a result. But there is such a thing as an intensive, vigorous, sustained, detailed, painstaking yet detached interest: it is called gossip. And as irony would have it, the very work Nussbaum considers the archetype of morally improving fiction—Henry James’s The Golden Bowl—features a gossip of world-class caliber. A clear reductio ad absurdum of Nussbaum’s position, the Fanny Assingham case shows, if it shows anything, that a fascination with (or even a fine awareness of) interpersonal niceties need not fuel any concern for our fellow human being. Fine awareness is, to phrase it in Jamesian terms, no guarantee of rich responsibility.

Worse yet, it is possible to take an intensive, vigorous, sustained, detailed, painstaking interest in the moral entanglements of other lives for vicious reasons. It is quite true that if we are to acquit ourselves of all our moral obligations toward other people, we must know them very well. But the implication does not work the other way around. On the contrary, intimate knowledge of someone may be the key to effectively torturing him or her. If you want evidence of this, just have a sibling. (I myself have two.) Or consider Room 101 in Orwell’s 1984 or, for that matter, the ingeniously sadistic Mlle Vinteuil, whom Proust’s narrator describes as an “artist of evil.”[16] For every Mrs. Dalloway, there is an O’Brien; for every Amélie Poulain, there is an Iago. Nussbaum may well be right that literature is excellent training for the picking up of nuances of personality and behaviour; there is just no guarantee, unfortunately, that this skill will be put to altruistic use. Indeed, if Lisa Zunshine is to be believed, literature presents an evolutionary advantage precisely because detailed knowledge of other minds makes them that much easier to control.[17]

7. The Corruption of Moralism

Literature cannot corrupt, then, any more than it can morally improve. Or rather, literature cannot corrupt on its own. But it can corrupt, ironically enough, when coupled with the belief that it is morally beneficial. For if we believe that fiction is the place to go to receive moral instruction, we may become deeply confused about what is right and what is wrong. Anyone who can be converted into a view by a fiction can be converted out of it by a fiction. If book XI of the Odyssey is enough to turn me into a pacifist on Monday, then book IX of the Iliad, which I read the following day, will be enough to make me believe in the glory of violence; if the Nun’s Priest’s Tale makes me prudent in the morning, then The Open Road will make me reckless by nightfall. Our culture is full of competing values and of stories to “prove” any one of them; conversions through fiction are simply not reliable.

In our zeal to learn moral lessons from fictions, we may leap to some strange conclusions. We may start to believe, for example, that race relations in Los Angeles really are as they are depicted in the absurdly overblown movie Crash.[18] Or we may dismiss Ralph Ellison’s masterpiece Invisible Man as a flop—this, believe it or not, is the stated view of Houston Baker—because it “tells” us that the proper response to racism is sitting in a basement draining the city’s power and thus “fails” to predict the civil rights movement (as though that were its obvious duty).[19]

And finally, if we think that literature is morally improving, we may end up feeling that it is enough to read books in order to be virtuous. This, in fact, is the complaint of Rousseau in the Lettre à d’Alembert: “When a man has admired fine actions in fables, and wept over imaginary sufferings, what more can be demanded of him? Is he not pleased with himself? Does he not congratulate himself on his noble soul? Has he not acquitted himself of all he owes virtue by the homage he has just rendered it?” “The most useful effect of the best tragedies,” concludes Rousseau, “is to reduce all the duties of man to some passing, sterile, and fruitless emotions, to make us applaud our courage in praising that of others, our humanity in pitying the ills we could have healed.”[20] Thus too Diderot, in the Paradoxe sur le Comédien, notes that the typical citizen leaves his vices at the door only to “take them up again on the way out.... I have often seen wicked men next to me taking deep umbrage at actions which they would not have failed to commit if they had found themselves in the same circumstances.”[21]

The ironic truth is that the best way to make sure that literature corrupts is to pretend that it improves. Only if it is taken to be morally beneficial will literature do us moral harm. And this is where I return to the qualification that literary texts don’t corrupt or improve us on their own. Institutions of reading have some say in how artworks are received, and it is to some extent up to us to protect future readers from the corruption of moralism. Future generations need to be rescued from our well-meaning attempts to save their souls.

8. Our Corruption: Aestheticized Cognition

But there is another type of corruption that literature inflicts on us. And this is the corruption of our faculty for judgment.

There is, in academia, a phenomenon that I would call the contagion of the object of study. You’ll have noticed that people who study Proust tend to end up writing long, extremely convo-luted sentences; I know I do. People who study Hemingway, by contrast, keep their prose terse and rugged. Beckettians are massively opaque and anaphoric, while Heideggerians often strike an oracular tone. All of this is perfectly harmless; the harm comes when we become accustomed, through long acquaintance with artworks, to judging everything by its power and not by its cogency. This is the deep contagion of the object of study to which literary critics are exposed, a corruption that results not from contact with a particular text but from sustained engagement with literary texts in general. If we are not careful, we start to judge everything—even our own ideas—on the basis of its beauty. It no longer matters, even to us, whether a view is true or false; it only matters if it is interesting, powerful, elegant. Once we are fully infected, we apply to scholarship the evaluative criteria we derive from literature, seek from criticism the same effects as we seek from fiction. One often hears the claim that criticism should take the form of literature, which, to my mind, is rather like saying the instructions for an oven ought to be in the form of a cake.

Now this corruption takes two forms, an optimistic and a pessimistic. The optimistic has recently been epitomized by Mark William Roche, who claims any number of virtues for literature. We’ve already seen him alleging that literature improves us by delivering salutary moral messages. But he also thinks it improves us by presenting positive models (p. 246); and negative models (p. 225); and utopian scenarios (p. 46); and dystopian scenarios (p. 240); and via empathy (p. 60); and by “mak[ing] visible for us the absolute” (p. 18); and by encouraging a contemplative attitude, à la Kant (p. 37). In case this is not enough, Roche adds that “by appealing to our senses, art reawakens in us an awareness of our own being as partly nature, which in turn renders more transparent our dependence on the realm of nature beyond us” (p. 216). Further, since beauty is symmetrical and balanced, our reception of it will cause us to strive for justice.[22] And since both literary texts and human beings are inexhaustible, if we come to appreciate a great work, we will automatically come to appreciate our neighbor too (p. 230). Finally, since each literary text is unique and each phenomenon of nature is unique, reading fictions will turn us into environmentalists (p. 232). Somehow he forgot to mention that literature also cures pleurisy, malaria, and the common cold.

So much for the optimistic variant of aestheticized cognition, a bout of reverse engineering in which we look around for benefits we might want literature to deliver and attempt to find a mechanism by which that might conceivably happen, rather than inferring from actual properties of art objects to their likely effects on the world. If anything, however, the pessimistic variant is even more common. We like to say, for example, that the author, who was invented in the eighteenth century, is a tyrannical imposition of power structures; that everything is language, which is to say nothing but an endless chain of signifiers; that all binaries are hierarchical and therefore oppressive; that reading serves only to lend us social prestige; and so on. We operate on what I’d call an inverse pragmatism: if it’s convenient, it must be false.

Why do we do this? Partly, to be sure, it’s because we think that this skepticism will bring about social change. Unfortunately, however, skepticism can be marshalled in the service of all kinds of positions. Thus, as Bruno Latour recently pointed out, global warming deniers have begun seizing on the purported fallibility of science to advance their own agenda of inaction. “Entire PhD programs,” writes Latour rather contritely, “are still running to make sure that good American kids are learning the hard way that facts are made up, that there is no such thing as natural, unmediated, unbiased access to truth, that we are always the prisoner of language, that we always speak from one standpoint, and so on, while dangerous extremists are using the very same argument of social construction to destroy hard-won evidence that could save our lives. Was I wrong to participate in the invention of this field known as science studies? Is it enough to say that we did not really mean what we said?”[23]

Political wishful thinking is one reason, then. Another is that it is thrilling to dance near abysses, to feel on one’s face the bracing wind of deep, metaphysical, all-encompassing doubt. A third is that it protects us from ever being fooled. (Though as Nietzsche says, doubting everything can be the best way of deceiving ourselves.[24]) A fourth is that asceticism, paradoxically, carries with it a yield of pleasure, crowning its practitioners with an imaginary halo of virtue.[25] The world-weary, hard-bitten stance never goes out of fashion, whereas optimism will always seem tacky, and judiciousness unsexy. But a fifth is surely that it would be beautiful if things were that way. How intoxicating if it turned out that everything we see is a lie—not just some things, but all things! How thrilling if we were the powerless pawns of language and social forces! A belief in malevolent forces ruling the world is just as enchanting, ultimately, as a belief in benevolent forces ruling the world.

This is how literature has corrupted us. We’ve made exaggerated claims in both directions: the wild optimism of Roche and company, the staggering pessimism of de Man and company. And in the process we have lost allies, lost readers, lost students. We need, today, a new kind of daring. We need to dare to be truthful, measured, painstaking, even empirical. We need to dare to be positive without being starry-eyed. We need to take ourselves seriously so that others will take us seriously. We need to understand the value of literature not as moral improver but as agent of clarification, as agent of reenchantment, as formal model, and as training-ground for mental capacities. We need to overcome our corruption, to stop being (as our beloved fictional mafiosi would put it) “on the take.” In short, we need to believe for once what we ourselves are saying.