publisher colophon
  • In Praise of Depth:or, How I Learned to Stop Worrying and Love the Hidden

In recent years, some prominent scholars have been making a surprising claim: examining literary texts for hidden depths is overblown, misguided, or indeed downright dangerous. Such examination, they've warned us, may lead to the loss of world Heidegger warned of (Gumbrecht), to the world-denying metaphysics Nietzsche warned of (Nehamas), or to the suspicious form of hermeneutics Ricoeur warned of (Best, Marcus, Moi). This paper seeks to suggest that, though the concerns are understandable, there's ultimately nothing to worry about. The fact that Nietzsche himself happily used metaphors of surface and depth suggests that they are not, in fact, metaphysically fraught. The fact that it's possible to appreciate surfaces at the same time as depths means that there's no real danger of losing the world. And as for depth-talk turning us into suspicious hermeneuts, that would happen only if we made two fallacious assumptions: first, that all depths are meanings; second, that all hidden features are in a text by accident. But since plenty of authors hide things deliberately, and since what's hidden often has nothing to do with propositional content, both assumptions are profoundly mistaken. Meanwhile, the surface/depth metaphor is the only thing that adequately captures the phenomenology of reading, especially when it comes to misdirection-based, hermetic, enigmatic, ironic, or satirical texts, where special activity on our part prompts a sudden leap to a radically different mode of understanding. And unlike its rivals, the surface/depth metaphor reflects a real asymmetry: depths explain surfaces, but not vice versa; surfaces are available without depths, but not the other way around. We need the metaphor, and we need to stay open to hidden depths as we read. As long as we don't come in with terrible assumptions, nothing bad will happen to us, and plenty of good things will. It's perfectly safe to go back in the water.

Can a literary text ever have hidden depths? Are bad readings ever superficial? And is it ever helpful to talk this way? I'm going to suggest that the answer to all three questions is yes. That might seem like something so obvious as not to need saying, but in recent years people have started urging us to change our vocabulary, reducing or even eliminating our reliance on metaphors of depth. And not just any people, either, but some of the most brilliant and prominent, including Toril Moi, Hans Ulrich Gumbrecht, and Alexander Nehamas. (Production of Presence, Only a Promise of Happiness, and Revolution of the Ordinary: these are three of the most interesting books I know on the subject of aesthetics.1) Moi, for example, tells us categorically that "we need to break with the picture of texts as objects with surface and depth" (RO 5). Why are such important thinkers saying this? And how can those of us who disagree defend our approach against their powerful objections?

Let's start from an example, one that I'll borrow from R. Lanier Anderson.2 Imagine you're about a quarter of the way into Pride and Prejudice. Elizabeth Bennet, who has been talking to Fitzwilliam Darcy, drops a remark about running into a certain George Wickham. Immediately Darcy's expression changes: "A deeper shade of hauteur overspread his features, but he said not a word, and Elizabeth, though blaming herself for her own weakness, could not go on."3

Superficially, everything seems pretty clear here: Darcy is an arrogant kind of person, and being reminded of Wickham—the "mere" son of a steward—has only brought his haughtiness out more strongly. The reality, of course, is very different. Yes, Darcy is arrogant; that much is true. But when Elizabeth mentions Wickham, his changed expression has nothing to do with pride. Rather, he is thinking of the fact that this Wickham recently pursued his fifteen-year-old sister, angling for her money, and came within days of persuading her to elope with him. Darcy is, perhaps, angry with Wickham; embarrassed for his own family; unsure as to whether Elizabeth knows; fearful for his sister's honor if too many people find out what happened . . . He is certainly full of emotions, but none of them is "hauteur." [End Page 145]

Austen does things like that on more than one occasion. Elsewhere, for example, Elizabeth reveals that Lydia has eloped with Wickham, and Darcy's face becomes very serious. He's actually just plotting his next (benevolent) move, but Elizabeth is convinced he's thinking what a terrible thing it is to be mixed up with such a family. The narrator doesn't say "Elizabeth surmised that Darcy was having second thoughts"; the narrator says Elizabeth knew he was. ("Elizabeth soon observed, and instantly understood it."4) And we only learn the truth some sixty pages later.5

What this means, in both cases, is that the narrator is playing a little trick on us, thanks to the magic of free indirect discourse. The narrator may seem to be simply stating the facts, but she is really showing us things as they appear to Elizabeth. It's Elizabeth who thinks that the expression on Darcy's face is one of arrogance (in the first example) or regret (in the second). And Elizabeth, as it happens, turns out to be mistaken.

Deliberate Depths

If the above interpretation is true, I hope it follows uncontroversially that there's a more and a less illuminating way to read the Austen sentence. The less illuminating way is to take it at face value, as a straightforward depiction of Darcy's emotional state; the more illuminating way is to see behind its subtle trickery and realize that Austen is setting a trap for us. Now I see no reason not to call the less illuminating reading a superficial reading: the sentence superficially presents Darcy as getting even more puffed up than he was before but turns out, at a deeper level, to be working very differently. The text has hidden depths, and a reading that does not uncover them is itself a shallow one.6

Furthermore, I see no reason to think that this is an accident. After all, Austen herself is clearly a big believer in surfaces and depths, and in the (general) superiority of depths over surfaces. (Isn't prejudice precisely a failure to get beneath appearances?) And Austen is a magisterial writer; it's hard to imagine that she simply made a mistake when having her narrator speak of hauteur. So when I say that there's a deep reading, I'm not talking about Austen's unconscious, or about diabolical social forces operating through her, or about tricks that a mischievous demon called language is playing on her. Deep readings are not necessarily readings that laugh in the face of authors, revealing things of which they are oblivious.

At least in this case, depth is something deliberately created by the author. Austen wants her readers to make a mistake, and then, ideally, [End Page 146] go on to correct it. If they are to understand what's really going on, they have work to do. Indeed, getting beneath the surface requires them to do something special. (I'm going to suggest, later on, that "having to do something special" may be the decisive criterion for determining whether it makes sense to speak of depths and surfaces.) A casual reader might very well read the sentence, misunderstand it, and move on; in order to get it right, she'd probably need to come back to it, bringing her new knowledge to bear on a sequence of words that now looks very different.

Effect, Not Meaning

Austen, interestingly, doesn't do any of the work for us. When Darcy finally tells Elizabeth what happened between Wickham and his sister, Austen does not have Darcy add "and that's why my face changed when you mentioned him to me—remember, back on page 63?"7 Nor does the narrator refer us back to the earlier moment. If we are to amend our error, we're going to have to do it for ourselves.

And this brings me to a final point about the sentence in Austen. What changes, when we read it correctly, is not just our understanding of what it means; it's also our understanding of what it does. Recognizing that Austen is setting traps for us like this—right from the famous first sentence, in fact—should nudge us in the direction of trying to figure out why. Why doesn't Austen just have her narrator tell us what Darcy was actually feeling? Or, if she wanted to maintain suspense, why not simply pass it over in silence, or tell us that Elizabeth thought she saw indignation on Darcy's features?

Anderson's explanation is an ingenious one. The reason we take the sentence for granted, he suggests, is that we already assume Darcy is an inveterately stuck-up individual. It accords perfectly with our preexisting picture of who he is. That is to say, it accords perfectly with our prejudice. It is our prejudice that causes us to get it wrong about the sentence, just as it is Elizabeth's prejudice that causes her to get it wrong about Darcy. And perhaps the experience of reading the novel is supposed to bring us up short. Perhaps it's supposed to chip away at our self-assurance, via the very process of reading. Perhaps it's supposed to help us turn ourselves into more discerning and careful readers—not just of novels but also of the world.

Not everyone will accept Anderson's theory, of course, but I hope most will agree that it is not superficial. (It is deeper, for example, than a reading that tells us Pride and Prejudice is merely a love story.) And the depth in question—crucial point—is one that is all about effect, not about meaning. [End Page 147]

Best, Marcus, Moi: Symptomatic Reading

As we saw earlier, however, there are some major theorists who would caution us against phrasing things this way. Nehamas has been on the case since the 1980s, with an essay titled "Mythology: The Theory of Plot"; Gumbrecht began in the mid-90s; Stephen Best and Sharon Marcus flew the flag in the aughts, with an essay that attracted a perhaps surprising amount of attention given the prior writings of Nehamas, Gumbrecht, and others; and in the late teens, Moi continued the fight.8 To each decade its champion. We could perhaps add Susan Sontag to this list, at least as an inspiration: though she did not attack depth, she did come out against "interpretation," in an essay of that name, back in 1966.9

Each of these theorists, I suspect, has a particular worry in mind, a particular type of bad outcome they are trying to stave off. (If you're interested in such details, read on; if not, please feel free to skip to the section titled "The Trouble with Message-Mongering.") Best and Marcus, along with Moi, appear primarily concerned about the "hermeneutics of suspicion" and its close cousin "symptomatic reading."10 For them, the notion of depth is identified with the idea that the core of every text is something that ended up there by accident, thanks to unconscious drives, political prejudices, or the sheer nature of language. To people who feel that way, finding the deep essence of the text is thus a strike against the author, that naive individual who innocently imagined she could know her own mind, or control her words, or write something that was what she wanted rather than what ideology dictated. What's shallow, according to the suspicious hermeneut, is the intended effect of the text; what's deep is its inadvertent revelation of the power of language, ideology, and/or the unconscious. If you have the misfortune to become such a hermeneut, according to Moi, an immediate consequence is that you will never again enjoy a novel, poem, or play. You can enjoy its destruction at your own hands—your brilliantly talented hands—but you will not experience surprise, wonder, or self-understanding, let alone self-transformation (RO 175–76, 220). No wonder Best, Marcus, and Moi see symptomatic reading as a danger worth averting.

Gumbrecht: Loss of World

As for Gumbrecht, his particular worry is that an excessive focus on depth will place us at a distance from our own lives. (Gumbrecht, it should be noted, does not oppose the metaphor in general, but merely wishes to decrease its prominence, in order to expand the space for [End Page 148] other possibilities.) In the first instance, the problem is the way in which we engage with the objects around us: "To interpret the world," says Gumbrecht, "means to go beyond its material surface or to penetrate that surface in order to identify a meaning . . . that is supposed to lie behind or beneath it" (PP 25). (Notice the twin assumptions (1) that interpretation is always a matter of depth—going behind or beneath—and (2) that depth is always a matter of meaning.) Now this obsession with depth, Gumbrecht adds, cuts us off from our instinctive relationship to our surroundings, which is one of connectedness. Instead of feeling part of the world, a body among bodies, we detach ourselves from it in order to set ourselves apart as observers. Something related holds for the events in our lives: we will never actually experience them if we are always mining them for meanings.11

And something related holds, too, for literary texts. The humanities, writes Gumbrecht, have "separated us from everything that could not be described as or transformed into a configuration of meaning" (PP 92); this, he says, is what explains "the positive value that our languages quite automatically attach to the dimension of 'depth.' If we call an observation 'deep,' we intend to praise it for having given a new, more complex, particularly adequate meaning to a phenomenon" (PP 21). In other words, the humanities—literary criticism included—have been relentlessly focused on extracting meanings from texts, wrongly thinking that doing so will yield a better account of them, and this process has destroyed the immediate, authentic, connected experience we could be having. In Martin Heidegger's terms, it has produced a "loss of world" (PP 92). Some readers will recognize here a powerful kinship not just with Heidegger but also with Sontag, who famously said that "to interpret is to impoverish, to deplete the world—in order to set up a shadow world of 'meanings.'"12 That, again, sounds like a danger worth avoiding.

Nehamas: Bad Metaphysics

Gumbrecht's overarching worry that talk of surfaces and depths will lead to regrettable consequences is shared, albeit with different specifics, by Nehamas. Discussing Sontag's "Against Interpretation," Nehamas writes that "her argument depends on a very sharp distinction between the superficial and the deep, the apparent and the real, the here and the beyond, which in turn invokes a controversial metaphysical picture."13 In other words, Sontag's targets—those eager delvers who love to mine texts for Marxist or Freudian meanings—are committing themselves not just to a dubious way of reading but to a dangerous metaphysics, [End Page 149] the two-world model Nietzsche warned us about. (That, surely, is the force of the phrase "the beyond," whose presence in the sentence is otherwise hard to explain.) Thinking of artworks in terms of surfaces and depths nudges us in the direction of a bad metaphysics, the kind that may ultimately lead us to downgrade the real.

As to why it's a dubious way of reading, the key point for Nehamas is that the use of surface/depth metaphors commits us to imagining that there are two entirely separate levels of analysis, with clear demarcation between them and no interplay among their elements. "What something is," Nehamas claims, "is not nearly as independent of what it means as it may seem to be" (OPH 121); an element we might want to call "surface" cannot be fully understood without what we might want to call "depth." Thus it is not that Darcy's apparent hauteur is surface and his real feelings, along with Austen's trickery and project, are depth. Rather, we are talking about a single complex textual entity, something like Austen's overall project worked via misdirection.

Whereas the surface/depth metaphor suggests a level of meaning that is fully grasped and that is later replaced by the truth, says Nehamas, the reality is a partial reading that is later completed by a fuller account.14 When we refine our assessment of the text, we are merely filling in details, not ditching a shallow understanding in favor of a deep one. "Even in those cases where we say that an action or a text means something other than what it appears to mean," writes Nehamas, "we do not have two meanings, one real and one apparent. All we have, even in the case of psychoanalytic or Marxist interpretation, is a series of progressively more complicated, detailed, and sophisticated hypotheses."15 Or again, "In asking 'What is the point of this story?' we are not moving on to a new level, but we are trying to give a more detailed interpretation of that story."16

So the "depth" elements cannot be separated from the "surface" elements, since the latter are what they are, in part, by virtue of the former; and we do not replace surface with depth, but merely fill in a partial understanding with more information. Third and finally, the demarcation between "surface" and "depth" is not stable. "What counts as appearance," Nehamas writes, "is simply what we take to be indisputable at some particular moment" (OPH 124)17—and what we take to be indisputable can easily change, causing a "surface" element to flip into a "depth" element, or vice versa. "The distinction between what something merely seems to mean and what it really means, between surface and depth, cannot be drawn systematically," argues Nehamas; "there is no clear line on one side of which everything is appearance, separated forever from the reality that lies on the other."18 To talk in terms of surfaces and depths is thus to indulge in a profoundly misleading theory [End Page 150] of interpretation. And to indulge in that theory is to flirt with some dangerous metaphysics.

The Trouble with Message-Mongering

I don't think you can fault people like Gumbrecht for taking on the assumption that artworks are all about sending messages, and that this message obsession deprives many artworks of their true power.19 There are copious significant ambitions available to aesthetic objects—including emotion-generation, question-raising, defamiliarization, transfiguration, the transmission of a way of seeing, the production of formal models, and the training of mental capacities—and to neglect all of these in favor of a universalized didacticism is to subject our interactions with artworks to serious depletion.20

Most of the options above have been known about for some time. Emotion is already in Aristotle; transfiguration is in Nietzsche; question-raising is in Roland Barthes and Toni Morrison; vision is in Marcel Proust and Simone de Beauvoir. Yet for some reason they've been largely forgotten about, and in many quarters it would seem that the only kind of depth people can imagine is a depth of meaning. Somewhat demoralizingly, Best and Marcus seem themselves to buy into the idea: that, I assume, is why they call for critics to "indicate what the text says about itself" (SR 11), "register what the text itself is saying" (SR 8), relay "an individual text's presented meaning" (SR 11), and "understan[d] its verbal meaning" (SR 10); after all, "texts can reveal their own truths" (SR 11), and "the moments that arrest us in texts . . . can themselves indicate important and overlooked truths" (SR 18).21 That's an awful lot of saying, meaning-delivering, and truth-telling for one short article.22 This assumption is also, I suspect, why their other main alternative is unadorned surface description. Best and Marcus feel "drawn," they say, "to modes of reading that attend to the surfaces of texts rather than plumb their depths" (SR 1–2), and so they propose to focus on "what is neither hidden nor hiding" (SR 9), so that "simply paraphrasing a text" becomes a highly valuable enterprise (SR 10).23

To my ears, these are rather surprising things to say, and they can perhaps only be explained by the fact that Best and Marcus explicitly equate interpretation with the act of "assigning a meaning to a text" (SR 1). Gumbrecht, as we saw, defines interpretation in much the same way, calling it "the identification and/or attribution of meaning" (PP 1). But as Nehamas and Moi point out, there is really no need to understand interpretation so narrowly: the term also covers the attribution of significance, and significance is a beast of an entirely different stripe.24 [End Page 151]

The Trouble with Symptomatic Reading

In short: the assumption that artworks are nothing but glorified vehicles for the transmission of ideas, whether deliberately or inadvertently, deprives many of them of the power to enhance or transform our lives. And so, to say it again, you can't fault people like Gumbrecht for attacking it. You also can't fault people like Best, Marcus, and Moi—or for that matter Paul Ricoeur and Susan Sontag and Eve Sedgwick and Rita Felski—for taking on the hermeneutics of suspicion. Best, Marcus, and Moi are right to invite us to read books on their own terms, allowing each to show us how to read it, rather than assuming in advance that we know what to do with it (SR 11; RO 216). (Moi agrees with Beauvoir, and I agree with her, that literary reading should be an adventure; finding the same thing in every text—something, indeed, that you already knew going in—is surely the opposite of that.)25 It is entirely reasonable to worry whether globalized suspicion could end up eliminating the pleasure or self-knowledge or emotion or transformation many texts have to offer us. It is entirely reasonable to worry whether symptomatic reading tends to satisfy itself with too little by way of evidence, imperiously sets itself above ostensibly benighted authors, and frequently distorts artworks, those curious square pegs that it relentlessly forces into its neat round holes.26 Suspicion, of course, is sometimes warranted; but it isn't always warranted, and when it shows up uninvited, it can be a book-crushing impediment to aesthetic experience.

None of this, however, implies that we need to get rid of depth. Depth didn't make us believe in ubiquitous didacticism; depth didn't sign us up to a two-world model of metaphysics; depth didn't tempt us to read suspiciously; depth didn't cause us to drive square pegs into round holes. Other, more local errors did that. And I think we can tackle those local errors on their own, rather than taking the sledgehammer of depth-denial to the nut of suspicion.

Depth is Not the Enemy

My sense, then, is that the scholars I've mentioned are mistaken in thinking that depth is the problem. Gumbrecht is mistaken in thinking that someone who finds hidden depths can only be a meaning-monger; Nehamas is mistaken in thinking that the use of depth metaphors in literary analysis has untoward metaphysical implications;27 Moi is mistaken in thinking that moving beyond symptomatic reading requires setting aside the surface/depth distinction; Best and Marcus are mistaken in [End Page 152] thinking that suspicion is the only possible kind of depth. They can all have everything they want without sacrificing a very serviceable metaphor.

Consider, again, the example we began with. As we've seen, there's a shallow reading of the hauteur passage and a deeper reading of the hauteur passage and, more broadly, a shallow understanding of what's going on in the novel and a deeper understanding of what's going on in the novel, one that takes into account the pervasive use of free indirect discourse. In neither case does the better understanding have anything to do with forces beyond Austen's control; on the contrary, we are dealing with elements Austen deliberately introduced into the novel. And in neither case does the better understanding have anything to do with propositional content; on the contrary, Austen probably thinks that a reader who comes away thinking they have learned the lessons they need to learn about life and are now wise individuals—I won't name names—are readers as foolish as Mary, a character whose main reason for being in the novel is surely to serve as an implicit warning to us. What's going on in Pride and Prejudice, then, is not the transmission of a message that Austen has to impart to us, and it's not the thwarting of Austen's ambitions by language, ideology, or the unconscious. Austen is achieving exactly what she wants to achieve—namely, to give her readers the opportunity to become a little better at suspending judgment. That capacity, which is a form of know-how rather than a principle, is what is really on offer here, for those who get beneath the surface.28

So depth does not have to consist in inadvertent revelations, and indeed it does not have to consist in revelations at all. What a text is hiding beneath its surface can, as Proust knew, be a vision of the world;29 it can be a set of tensions—deliberate tensions, designed to spark reflection;30 it can be an ironic attitude, again deliberate, toward its own contents;31 or it can be an intended effect, such as the fine-tuning we found in Austen. Depth is multifarious; it is often deliberate; it is frequently salutary; and there is nothing wrong, it seems to me, with calling it by its name.

Seven Key Points

At this stage we can start to draw things together. What I've been suggesting is that we should continue to make room for notions of superficiality and depth in our assessment of artworks, and there are seven main considerations to bear in mind in this context. First, literary texts frequently contain or imply features that are hidden. Second, these features are in no way limited to "meanings"; they can also be things like covert complexities and non-obvious intended effects. Third, [End Page 153] some of those hidden features are there by accident, just as the suspicious hermeneuts told us all along. That doesn't mean, however, that all hidden features in a work got in there by mistake. On the contrary, and this is my fourth point, many have been deliberately included by the author. Hence, fifth, reading for depth does not have to be suspicious or symptomatic reading; when we hunt for treasure the author has generously hidden for us, we are not going against her intentions, outsmarting her, or exposing her failures, but simply doing just what she hoped we would do.32

Now when elements are hidden deliberately, they typically tend to be important. That's point six. We generally cannot do without them if we want to have the full experience offered by the artwork, and—seventh point—they often cast the rest of the work in a radically new light. Understanding the hauteur passage in Pride and Prejudice doesn't just clear up a local mystery about Darcy's frame of mind; it encourages us to think completely differently about the way in which the novel works, the nature of our reading experience, and what the entire enterprise is for. We are not, in fact, reading a morality tale in which Austen warns us of the perils of arrogance and blinkeredness (that would be a Mary kind of novel, and nobody wants to be Mary, least of all Austen). Instead, we are undergoing a process in which we may, if we are alert and willing, give ourselves practice in withholding judgment.

To be sure, hidden features do not always effect such a radical transformation of our understanding. Not all texts have buried features, let alone buried features that make a big difference, and it would be as much of a mistake to impose a deep interpretation on a straightforward text as to flatten out a complicated one.33 Still, Pride and Prejudice is by no means an anomaly. And with novels like that, isn't it reasonable to say that there are hidden depths?34 Isn't it reasonable to prefer a reading that takes them into account over one that doesn't? Is it really always a prejudice, as Michel Tournier appears to think, to believe that depth is better than breadth?35

The Text Becomes a Different Object

Let me insist on the last main point: when it comes to ambitious artworks, hidden depths frequently have to do with the nature and/ or function of the work as a whole. I might go into Jonathan Swift's A Modest Proposal thinking that it's a serious and horrifying tract written by the eighteenth-century equivalent of Ayn Rand; but at a certain point, unless I'm really obtuse, I'm going to realize the whole thing is [End Page 154] a scathing satire.36 This isn't like misreading a word, or thinking that a character has red hair rather than brown. Here, my new understanding transforms my entire sense of what the piece is. It completely upends my hypotheses about what it's for, what I'm supposed to do with it, and why people should bother reading it in the first place.

The same is true, if Anderson is right, for Pride and Prejudice: we go in thinking it's just a comedy of marriage and come out understanding it's also an opportunity to practice our detachment. It's true too, if some of us are right, for Plato's Symposium, where we go in thinking it's just a collection of philosophical ideas and come out understanding it's also a workout for our reasoning muscles. The Symposium turns out to be a dramatically different object than what we had suspected (a set of exercises, not a treatise); it requires a dramatically different kind of engagement (refutation, not reverence); and it serves dramatically different ends (training, not teaching).37

The examples could be multiplied endlessly. Think of a "comedy" that turns out, on closer inspection, to be tragic. Think of the parables of Jesus in the gospel of Mark.38 Or think of the story of Abraham nearly killing Isaac. Moi tells us that "nothing is hidden" in the Abraham story (RO 189; NIH 45), yet surely it hides plenty, not least the most important thing of all: namely, what we are supposed to do with it. Perhaps we are supposed to figure out what Abraham is thinking, so we can be more like him—but perhaps we are supposed to be less like him. (Some biblical commentators believe that God was testing Abraham's morality, not his faith, and that Abraham was supposed to say, in the words of Bob Dylan, "Man, you must be puttin' me on.") Or perhaps we are simply supposed, as Kierkegaard's narrator thinks, to be mystified.39 These are very different approaches, and not one of them is there to be seen on the surface of Genesis.40

Our encounters with texts like those above—enigmatic texts, ironic texts, and the like—are encounters in which our initial understanding is not just mistaken but radically mistaken. The text before us becomes a new kind of object, requiring a new set of operations, offering a new set of experiences, designed to have a whole new function. The point of the exercise is not even close to what we had initially suspected.

And it's precisely because the text becomes a drastically different entity—and because this happens only on some occasions, not on others—that the replacement metaphors offered by people like Nehamas do not seem adequate, from a phenomenological standpoint. No metaphor, of course, is ever going to be a perfect fit, but some, I think, do a better job of capturing the phenomenology of reading. When a text becomes a whole new object before our eyes thanks to our new understanding of [End Page 155] what it's up to, that feels like a shift in level, not just a filling in of details. "Darcy's face expressed his haughtiness" is not incomplete; it is mistaken.

Reading a novel with hidden depths is not like scanning a landscape, or turning an object around in our hands, or opening the curtains wider to get a better view. Rather, it's like seeing a magician pull a rabbit out of a hat, or feeling the ground falling out from beneath our feet, or staring as the good guy rips off a mask to reveal a villainous face underneath. It's not just that we didn't see the whole picture; it's that we got the key points totally wrong. It's almost as if we hadn't been seeing them at all.

There's a Difference in Kind

Why should we say, though, that our hypothesis about intended effect, in cases like these, is "deep"? Well, partly because it's crucially important, shedding light on the operations and value of the work as a whole; partly because it's not available to immediate inspection; partly, in Austen-like cases, because something initially blocks us from seeing it, there being a deliberately placed temptation to read badly. But I'd also add another reason, which will be my eighth main point: our hypothesis about what's really going on in a work, about what it's designed to do, about what contribution we're supposed to make to let that happen—that hypothesis is a thing of a completely different nature from whatever else we may know or guess about it. Once I suspect what's really going on in Pride and Prejudice, I don't just have a thousand and one pieces of information rather than a thousand. The difference is not just quantitative; it is a difference in kind.

This goes against something Nehamas tells us about reading. As we saw earlier, Nehamas takes interpretation to be a smooth, continuous line from maximal ignorance to maximal knowledge: we never make a leap to a "new level" but merely form "progressively more complicated, detailed, and sophisticated hypotheses" about the text in our hands.41 That's one reason why, according to him, it makes no sense to speak of depth.

But consider, by way of analogy, the following scenario. (I'm borrowing it from a real work of fiction, but since I hate spoilers with a passion, I'm not going to name it. The first rule of essay writing is never to ruin surprises.) You make a new friend. He strikes you as tough, confident, and free-thinking. You go on to learn that he's resolutely free of materialism (and, perhaps the flipside of this, rather slovenly and unhygienic). Then you notice that he's charismatic, intelligent, and gregarious, with astonishing, almost pied-piper-like leadership skills. Pretty soon you [End Page 156] start to suspect that his moral standards are questionable. Oh, and then you discover that he isn't actually real, but is in fact just a figment of your imagination. Is the last revelation on a par with everything else? No. As Immanuel Kant famously said in response to the ontological argument, existence is not a property.42 So it's misleading to say that we progressively developed a clearer and clearer picture of our friend, our understanding rising steadily as time went by; rather, the realization that he never existed was a game-changer. It sent our gradually rising graph line crashing through the floor.

So too with our overall comprehension of texts. As long as I'm just noticing additional details, such as the fact that Darcy has a sister, that his first name is Fitzwilliam, or that his mansion is impressive, the difference in my attitude is just a difference in degree. But when I figure out Austen's secret design, my understanding has made a qualitative leap. Darcy's first name and Austen's artistic project are items that belong to massively distinct mental categories, as heterogeneous as baby goats and thermonuclear war.43 What it means for something to have "depth" is that its explanation requires bringing in something like that, something of a different nature, something that does a kind of work no additional detail ever could.44

There is, in fact, a crucial asymmetry between whatever we think of as a text's surface and whatever we think of as its depth. (For those keeping score, this will be my ninth point.) Depths very often explain surfaces; surfaces, by contrast, rarely explain depths. Thus Austen's project accounts for the existence of the hauteur sentence, and not vice versa.45 In addition, while surfaces are available without depths, depths are not available without surfaces. It is easily possible, for example, to enjoy Pride and Prejudice for its characters and situations, entirely ignoring or overlooking the free indirect discourse and thus failing to understand Austen's ultimate project. But there is no way to notice the free indirect discourse without imagining the characters and situations. Similarly, it is easily possible to enjoy the adventures of Aslan the lion without knowing that Aslan is a Jesus figure—when I read C. S. Lewis's The Chronicles of Narnia books as a nine-year-old, I myself was blissfully unaware—but it is impossible for a reader to see Aslan as a Jesus figure without seeing Aslan as (also) a lion.46 That double asymmetry, of dependency and explanation, once again justifies us in speaking of separate levels. You can iron them all you like, but texts will not stay flat. [End Page 157]

"Nothing is Hidden"—Or is It?

So to say it again, many texts have hidden features; these hidden features stand in an asymmetrical relationship to their obvious counterparts; while some have crept in there by accident, others have been deliberately planted by their creators; they are often vital; and when we discover them, we make a qualitative leap into a whole new way of reading, the text being transformed—just like that friend of yours who turned out not to exist—into a whole new kind of object.

Not everyone, of course, would agree with the account I have just given. Moi, for example, denies the very first premise. No, she says, literary texts do not include hidden features;47 as Ludwig Wittgenstein claims about language generally, "nothing is hidden."48 We may occasionally be confused about something in an artwork, Moi acknowledges, and then later clear up the confusion, but this does not mean that anything was hidden from us. "When the fog in our head lifts," she explains, "we often feel that we should have seen what the problem was all along. For then the solution often seems excruciatingly obvious. How could we have missed it? It was never hidden. We just failed to see it" (RO 184–85). "Note the difference," she adds elsewhere, "between assuming that a text is hiding something from us, and assuming that the problem is in me" (RO 181).

The problem is in me: if there's a fog between me and the text, that fog is on my glasses, not on the surface of the work. All I need to do is give my glasses a good wipe. Or, to use Moi's own metaphor, all I need is "self-therapy" (RO 185). Once I've turned myself into the kind of person who is able to appreciate this particular artwork, everything will seem dazzlingly obvious to me; nothing will be opaque any more; I won't even be able to believe that I was ever confused. ("How could we have missed it?")

Is that really how most cases are, though, phenomenologically speaking? The first time you read Pride and Prejudice, you take the "hauteur" passage at face value and think Darcy is becoming even more puffed up. You go back and read it again and realize that you fell victim to narrative misdirection, pulled off by means of some brilliantly handled free indirect discourse. But surely you don't berate yourself for not having noticed it the first time. You don't wonder how you could possibly have missed it. On the contrary, you retain a clear sense that practically every reader is going to fall at this same hurdle. Your friends will get it wrong; your students will get it wrong; pretty much everyone you can imagine will get it wrong. You yourself may get it wrong next time, at least temporarily. That's because this trick is in the book, not in us. The fog is on the words, having been blown in there by Austen; it's not (or [End Page 158] at least not just) on our glasses, having emanated from our incompetent hearts. You silently compliment Austen on her skill at constructing such an ingenious device—a device, precisely, of concealment. Of course there's something hidden in there. Austen deliberately hid it.

To put it another way, when you read the hauteur sentence a second time, its irony becomes something you notice—but that you notice as something hidden. Think, here, of a situation where you hide a present in a room and watch your favorite nephew gleefully look for it. You know exactly where the present is; maybe you can even see it, from where you're standing. But you see it as something that is not seen (by everyone), something that is not immediately visible.49 The feeling of hiddenness is a very real, and very important, part of the phenomenology of reading. The thrill of discovery is a vital engine of narrative pleasure; and without concealment, no discovery is possible.

For Those Who Need More Evidence

The Austen example, of course, is not entirely clear-cut, since, as I mentioned earlier, a certain degree of "self-therapy" is indeed appropriate (even if it doesn't lead to us thinking we were fools for not seeing the truth all along). So let's turn to a less equivocal case: the last two lines of Stéphane Mallarmé's famous sonnet "Ses purs ongles." Those lines read, in part, "dans l'oubli fermé par le cadre, se fixe / De scintillations sitôt le septuor." In other words, "in the forgetfulness enclosed by the window, the septet of scintillations immediately fixes itself"—or, more prosaically, a constellation of seven stars appears in an otherwise empty window.50 So far so fine. But when you set aside the meaning of the words and listen to the sounds, something else emerges. As Robert Greer Cohn famously observed, the final line ("De scintillations sitôt le septuor") is counting out the numbers cinq, six, sept.51 And when you hear that, you start to wonder whether the "cadre" of the previous line didn't start the whole thing off. Cadre, scintillations, sitôt, septuor: the poem itself is tallying the stars.

That's a truly beautiful effect. And it's a hidden effect. I've recited that poem to myself hundreds of times, and I've never considered it something obvious, something I should have noticed immediately. Mallarmé concealed it from me. Yes, he hoped some people would hear it—eventually. But not everyone, and not the first time. This is hermetic writing, and it belongs to an extensive and distinguished tradition. It would be a strange result if we ended up denying that hermetic writing ever existed. [End Page 159]

Glorious elements like these resemble ingenious crossword puzzle clues. When we solve them, we don't think we should have seen through them right away. We don't think our friends will have no trouble with them. On the contrary, we silently thank the setter for having made it difficult for us. Even if there were some kind of work we could do on ourselves—some kind of "self-therapy"—that would cause us to see through all crossword clues, or all literary devices, we wouldn't want to do it, and we wouldn't want our friends to do it either.

A Moderate Moi

At this point it may be tempting to attribute a more moderate view to Moi. (After all, her book—to say it again—was one of the great theoretical highlights of the last decade.) Perhaps her claim is only that some texts lack hidden elements, not that all texts lack hidden elements. If that were the case, we could take her to be saying the following: the surface/ depth metaphor implies that all artworks have hidden elements, and that an artwork automatically hides something, merely by virtue of being an artwork. The surface/depth distinction is thus inaccurate (inasmuch as many works of literature contain nothing that's hidden in any meaningful way) and dangerous (inasmuch as it fosters the hermeneutics of suspicion). As long as we retain the surface/depth distinction, we will be in danger of becoming symptomatic readers.52

This more moderate position has a great deal to recommend it. I'm not convinced, however, that we really need to oust the surface/depth distinction in order to overcome the universalist variant of symptomatic reading, there being other perfectly good arguments against it. Nor am I convinced that someone who believes in hidden depths will automatically start finding them in every text, let alone become a hermeneut of suspicion. (I myself am a pretty good case in point, I think, having rarely if ever engaged in suspicious hermeneutics, and having done my fair share of arguing against its excesses.53)

And I'm not convinced, finally, that the surface/depth distinction automatically implies that every artwork contains hidden elements. After all, we sometimes say of texts that they are "all surface," or "twodimensional," or "monolithic," or "empty," or "transparent." (Even if we imagined a literary text as an object, we could imagine that object as being made of Perspex, as being flat, as being made of uniform material, as having windows, or as being hollow, with nothing inside it.) Thus we can employ the very same metaphor to deny that a given text should be read suspiciously. The metaphor itself does not commit us to anything [End Page 160] much, it seems to me, when it comes to concealment.

The main problem, however, is the fact that Moi doesn't always stick to the modest version of her argument. Rather than saying that some texts lack hidden elements, she prefers the unconditional statement "nothing is hidden." She writes, in a wholesale way, that "claims about hiddenness . . . in literary criticism are empty" (RO 179). She says that we can reject the hidden/shown distinction—in general—"without losing anything at all" (RO 179). She claims, without qualification, that "there simply is no need to think of texts . . . as hiding something" (RO 178). She proposes, as a totalizing demand, that "we need to break with the picture of texts as objects with surface and depth" (RO 5). And she writes, equally categorically, that "metaphors of . . . hidden and shown don't actually describe what readers are doing" (RO 185).54

Thus Moi is not, in fact, saying (in these passages at least) that some texts lack elements that warrant being called hidden; she's saying that all texts lack elements that warrant being called hidden. Any time we refer to a hidden feature, therefore, we are making an empty statement. If we think that readers ever look for elements that authors have concealed from immediate view, we are categorically wrong, because metaphors of hiddenness do not describe what readers are doing.

But as I've been trying to show throughout this paper, to speak in terms of a quest for hidden treasure is, in many cases, to describe very well what readers are doing. The notion of depth can be given reasonably determinate content, and it does a better job, in my view, of capturing the phenomenology of surprising texts than does the account of wiped glasses. There is, at the end of the day, nothing empty about it. Nor is there anything dangerous about the hidden/shown distinction—it does not automatically turn us into suspicious hermeneuts—and it is far from clear, in addition, that it can be jettisoned without cost.

Special Work is Required

But, someone might say, the words are all directly in front of us, right there on the page before our eyes. How could Austen or Mallarmé conceal something, even if they wanted to? Behind what? Surely everything is in plain sight!55

That's a compelling objection, but it's not one I take to be decisive. The reason is something I briefly mentioned earlier, and that I'll now reintroduce as my tenth main point: some objects demand a special kind of interpretive work. [End Page 161]

Fig. 1. Hans Holbein the Younger, The Ambassadors (1533) National Portrait Gallery, London
Click for larger view
View full resolution
Fig. 1.

Hans Holbein the Younger, The Ambassadors (1533) National Portrait Gallery, London

Think, here, about Hans Holbein's Ambassadors (Fig. 1). By now, of course, most of us are very familiar with this painting, but imagine what it must have been like for its earliest viewers. How many were anything other than mystified by that rugged, almost baguette-shaped blob in the foreground, poking into the folds of the bishop's robes? How many of them gave up trying to figure it out? How many, by contrast, moved their bodies so that they were looking down at the blob from a very narrow angle to the right of the canvas, or so that they were looking up at the blob from an equally odd angle to the left of the canvas? Moving to these weird positions is not something we normally do with paintings.

Or think, if you prefer, of Vivian Darkbloom, the character in Lolita.56 "Vivian Darkbloom," it turns out, is an anagram of "Vladimir Nabokov." But how many readers notice that without having it pointed out to them? Rearranging letters is not something we ordinarily do with a [End Page 162] character's name.

Or think, again, of Mallarmé's sonnet. To hear the numbers quatre, cinq, six, and sept, you have to engage in a strikingly singular operation: you have to distract yourself from the meaning of the whole words, while at the same time allowing yourself to notice other words within the resulting pattern of sounds. Listening for shadow-words is not something we usually do with sentences.

In strictly literal terms, perhaps, all of these features are perceivable. But that doesn't mean they are not hidden. On the contrary, we can tell they are hidden by the fact that most people miss them. And if most of us miss these hidden treasures (at first), that's not because there's something wrong with us—it's because uncovering hidden treasure requires us to dig in an unusual direction. We won't hear those numbers in Mallarmé unless we distract ourselves from the meaning; we won't see that skull in Holbein unless we move our body to a bizarre viewing position; we won't pick up the author's name in Lolita unless we start rearranging letters. We don't get to see it, or hear it, unless we change the way we play the game. To find the hidden treasures, we have to perform special mental operations.

Nor is it significant that we can eventually figure out such puzzles. Your nephew will eventually find his present, too, but that doesn't mean it's not hidden. (To hide something is not to make it impossible to discover; it is merely to make it difficult to discover.) When it comes to things we are looking for, in fact, three scenarios are possible. Sometimes they are right in front of our nose; sometimes they are stashed away for good, definitively beyond detection; sometimes they are concealed temporarily, available to us but not right away, not without us doing something to find them. Mallarmé's numbers, Holbein's skull, and Austen's hauteur all fall into the third category.

Moi tells us that the way to figure texts out is just to keep looking; Nehamas tells us that the way to do it is to go sideways, rather than down. ("To understand something better is not to isolate it," Nehamas writes, "[and] to delve into its depths: it is to see how it is like and unlike every thing that surrounds it" (OPH 121).57) But I'm not going to see the anagram in Nabokov merely by continuing to look. I'm not going to hear the numbers in Mallarmé by comparing that poem to other poems. (What I need, on the contrary, is to stay within the sonnet and start doing something special.) And I'm not going to understand Austen's project (merely) by comparing her work to Charles Dickens and Anthony Trollope. Someone might say that I should compare Pride and Prejudice to novels that don't use free indirect discourse, but how would I know to make such a comparison if I hadn't already noticed [End Page 163] the device? Why, in addition, would I bother, if I hadn't already guessed at its importance? And if all I end up with is a set of comparisons and contrasts, rather than a hypothesis about Austen's hidden project, can I really be said to understand her novel better?58

Toril Moi's Objection

Fascinatingly, and instructively, Moi flat-out denies the point I just made. No, says Moi, we never need to do anything special with literary texts. "Critics who think they are uncovering hidden truths don't read any differently from critics who don't," she writes. "In fact, even the most suspicious critics . . . —Sherlock Holmes and Sigmund Freud—don't do anything special. They simply look and think" (NIH 34; RO 178). Or again: "Freud isn't digging under the surface. He looks at and listens to his analysand's expressions, and thinks" (NIH 42; compare RO 186).

But, one wants to say, what is covered by the "thinking" part? What is Freud doing, exactly, when he's thinking, as opposed to just looking and listening? I can only imagine one possibility: he's trying to find an explanation for what he sees and hears. And since it's Freud, surely that explanation has to do with hidden parts of the psyche; surely it involves setting aside how things appear at first glance in favor of secret and insidious homunculi such as the Id. In other words, surely Freud is digging under the surface; that's precisely what's going on while he's thinking.59

The same is true for Sherlock Holmes. Here's what Moi has to say about him: "It's not that the others look at the surface, whereas Sherlock looks beneath it. It is that he pays attention to details they didn't think to look at" (RO 186, emphasis removed). Now it is indeed true that Holmes pays better attention than anyone else. But is that really the only thing he is doing? Isn't he, like Freud, also doing some thinking? And doesn't that thinking include the search for an explanation?

Consider all those times when Holmes points a detail out to Watson and Watson has no idea what to do with it. In "A Case of Identity," for example, there's a typewritten signature on a letter, and Holmes says it's "conclusive." "Of what?" asks Watson, completely at a loss.60 You can notice all the details you like, but if you don't have the ability to put them together into an overarching account—to explain the reasons for them being the way they are—then you are not a detective. Even when Holmes and Watson are noticing exactly the same things, then, Holmes is doing something Watson isn't. Holmes isn't just paying better attention; he is also looking beyond the appearances to their hidden causes. (The connection to art is direct: full appreciation involves figuring out, [End Page 164] or at least making a good guess at, the cause of what we see before us, which is to say the intention of a postulated author.61)

Indeed, the very reason that Holmes does a better job of the noticing part is that he is not just noticing. Holmes is noticing selectively, and his principle of selectivity is driven by his system of interpretation. Since, in any scenario, there's a quasi-infinite quantity of details we could attend to, it turns out that it can never be enough to look, in a general sense; we must also know where to look. Holmes doesn't test his clients to see if they can fly unaided. He doesn't check to see that they're breathing air. Instead he looks at their fingers, their boots, their handwriting. Why? Because he has more than a good pair of eyes: he has a theory about the kind of thing that's worth attending to. But to have that kind of theory—to know which details can potentially count as relevant, which observations can potentially count as usable data—we need to get beneath the surface. We need to go deep.

That, indeed, is why analysts get trained. Indeed, of the three categories of expert that Moi mentions here, all receive instruction in their respective fields. Detectives are trained to detect. Psychoanalysts are trained to analyze. Literary critics are trained to close-read. (In a different context, Moi acknowledges the last point.62) Since we're in the realm of symptoms, we could add medical doctors too. Doctors have to know what causes a given combination of headaches and chills. They don't just look; they explain. And if they are also better observers of bodies, it is precisely because they possess knowledge about medical causes and effects.

To put it another way, each of these domains involves a special way of looking: there are particular things it makes sense to look for, and particular ways to test them. If my copy of Pride and Prejudice has a bit of fluff on it, I know I'm not supposed to include that in my interpretation. A detective, by contrast, might pick it up with gloves and bag it. Detectives attend to fibers and fingerprints; film theorists attend to editing transitions and camera movements; art historians attend to brushstrokes and signature placements; literary critics attend to enjambment and free indirect discourse. I for one would never have noticed enjambment or free indirect discourse if I had not been told to look for them. If we're not on the alert for things like that, we can gaze for a hundred years without getting closer to seeing them.

So people who have been trained to read literature are doing something special. When we rearrange the letters of "Vivian Darkbloom," that's not just looking harder; that's looking differently. When we hear Mallarmé's sonnet counting out the stars, we are not just listening longer; we are listening differently. We are performing a radically distinct mental operation. [End Page 165]

Non-Sequential Cases: Allegory and Parody

Before I conclude, let me quickly mention a type of depth that differs from the ones we've been considering so far. Austen's hauteur, Mallarmé's numbers, Nabokov's letters, and Holbein's skull: these fall into a category of depth effect that we might call sequential. In sequential cases, there's something in an artwork that we can't see at first, but that we can come to notice eventually. We have to do something in order to notice it; unless a friend or teacher or web page short-circuits the whole thing by revealing the secret to us, we're going to have to rearrange letters, listen for phonemes, crane our neck, or do something else that's inhabitual. There's an order of noticing, and the order of noticing matters. It's built in to the artwork, and it's an essential part of the reading or viewing experience.

There are, however, instances of depth that do not involve sequentiality. Consider allegory: the pigs in Animal Farm are (also) Stalin and Trotsky, just as Aslan in the Narnia books is (also) Jesus. Or consider parody. Most readers, I suspect, find Samuel Richardson's Pamela (1740) either devastatingly powerful or unbearably preachy. (Let's face it, the full title is Pamela; or, Virtue Rewarded.) If you're in the latter camp—as I may possibly be—you are quite likely to have a great time with Shamela, a book Henry Fielding published one year later (1741). But I defy anyone to enjoy Shamela, at least the way it's meant to be enjoyed, without knowing about Pamela. A reader picking Shamela up cold, so to speak, simply won't get the joke.63

Similarly, anyone who hasn't read or heard of a poem comparing eyes to stars, lips to coral, breasts to snow, and cheeks to roses will have a hard time appreciating William Shakespeare's Sonnet 130 ("My mistress' eyes are nothing like the sun") to the fullest. Young Frankenstein, This is Spinal Tap, Blazing Saddles, Scary Movie . . . In all these cases, there's something as it were behind the artwork, namely the text or tradition that is being made fun of. We need to know about that text or tradition in order to get the joke, and it's entirely possible for us not to get it.64

What this means is that something is hidden here too; it's just that it's hidden from some and not from others, as opposed to being hidden now and revealed later. (To people in the know, everything is crystal clear, and understanding is immediate. We know what's going on; we know we know what's going on; we know that others don't know what's going on; and all of that contributes to our enjoyment.) That's what separates this kind of scenario from the ones we were talking about earlier, where there's a delay between an initial misapprehension and a subsequent self-correction. We still need to do something, when processing these [End Page 166] texts, that we don't do with every work of fiction: namely, measure the distance between the book in our hands and books we read earlier. But this kind of processing comes easily to seasoned appreciators of romance, horror, westerns, or whatever the parodied genre may be.

To the extent that something is hidden at least from some readers, it is reasonable to speak of "levels": at one level, Sonnet 130 is a rather back-handed compliment to a love partner; at another level, it is (also) a critique of poetic exaggeration. Since intertexts and contexts are generally not visible on the surface of the works in question, it will not be enough to follow Moi's advice and just keep looking. No amount of inspection of Animal Farm will reveal that Snowball is Trotsky or that Napoleon is Stalin.65 No amount of inspection of This is Spinal Tap will reveal the existence of an entire tradition of "behind the music" documentaries, with their relentlessly repeated topoi. If you know it, you know it; if you don't, you'd better find out, by going beyond the book or film you're currently engaged with.

In many such cases, of course, it is possible for readers or viewers to enjoy a work, to some extent, without knowing what it is alluding to. Recognizing the layers—getting the full joke—will almost certainly enhance appreciation, but it isn't a necessary ticket to entry. Consider those fantastic poems in Lewis Carroll's Alice books, such as "Speak Roughly to Your Little Boy" and "You are Old, Father William." Thanks to the sterling work of Martin Gardner, we now know that these poems are parodies of po-faced, didactic Victorian poetry; there was a popular nineteenth-century poem that sternly admonished folks to speak gently to their little boys, and the original Father William was a saintly, preachy type, not a cantankerous old man who is likely to kick you downstairs.66 But "Speak Roughly to Your Little Boy" is a hilarious poem with or without its Victorian target, and we can appreciate it on its own terms even as we also enjoy the satire. Similarly, it's fun to continue imagining Napoleon and Snowball as pigs, even as we also imagine them as rivals for power in the fledgling Soviet Union. And this should alert us to an important side point, which is that perceiving hidden depths does not necessarily mean we jettison the surface above them. Contrary to what we heard earlier from Gumbrecht, the discovery of a deeper level does not always cause us to forget or dismiss or disdain what's immediately apparent. Eleventh and final point, then: deep reading and superficial reading can happily coexist. [End Page 167]


To sum it all up, then: artworks frequently contain features that are hidden. These features are in no way limited to "meanings," but include things like secret complexities and non-obvious intended effects. Many such features are there intentionally, having been deliberately introduced by the author. Accordingly, one can read for depth without being a suspicious or symptomatic hermeneut.

Further, there are two different categories of secret feature. There's the kind we see immediately, where what it means for it to be hidden is simply that other readers do not see it, either because they don't know the intertext that's being alluded to or because they've never been introduced to the allegorical code. And then there's the kind where it takes a while, because we have work to do—special, unusual work—to make the thing appear, like warming up those invisible ink messages some of us used to write as kids.

In either case, the hidden elements may well be categorically distinct from the non-hidden elements; we are talking about a difference in kind, not just a difference in degree. And the hidden elements tend, as well, to be highly consequential. They often throw an entirely new light on the nature, function, and significance of the text as a whole, revealing us, or our fellow readers, to have been radically mistaken about it. They do not always require us to jettison the surface, but they do make things look markedly different. And even when their contribution does not upend our entire modus operandi, we cannot do without them if we wish to have the full experience offered by the artwork.

I personally don't mind if people use a different metaphor in place of "depth."67 I would gladly hear them talk of levels, or layers, or anything they like. But whatever term people come up with, it had better include all these features. Surface reading isn't going to do the trick, I'm afraid, and neither is continuing to look, and neither is ever-expanding breadth. They don't capture the phenomenology, and they seem to lead to some rather baffling claims.

Whoever thought that depth was the enemy? There's nothing wrong with depth. The water is lovely. Dive on in. [End Page 168]

Joshua Landy
Stanford University
Joshua Landy

Joshua Landy is the Andrew B. Hammond Professor of French and Professor of Comparative Literature at Stanford University, where he codirects the Initiative in Philosophy and Literature. His books include Philosophy as Fiction: Self, Deception, and Knowledge in Proust (2004), How to Do Things With Fictions (2012), and, as coeditor, The Re-Enchantment of the World: Secular Magic in a Rational Age (2009).


. Hans Ulrich Gumbrecht, Toril Moi, and Alexander Nehamas have been extremely generous with their time in discussing the content of this paper with me. I've tried as hard as I could to make sure my characterization of their positions is accurate; if it isn't, that is by no means their fault. My deep ("deep"!) gratitude also goes out to everyone at the Chicago Philosophy and Literature Workshop, including Jonathan Lear, Eliza Little, Ruth Martin, Thomas Pavel, and Robert Pippin; everyone at the London Aesthetics Forum, including Stacie Friend, Ken Gemes, Alex Grazankowski, Andrew Huddleston, and Oda Ottosen; everyone who was kind enough to listen to me at Yale University, including Dudley Andrew, Morgane Cadieu, Marta Figlerowicz, Martin Hägglund, Jonathan Kramnick, Paul North, David Quint, Ayesha Ramachandran, Pierre Saint-Amand, and Katie Trumpener; everyone who was kind enough to listen to me at Cornell University, including Jonathan Davenport, Hannah Karmin, David LaRocca, Klas Erik Molde, and Jan Steyn; and, closer to home, Lanier Anderson, Halle Edwards, Katy Meadows, Manuel Vargas, Nirvana Tanoukhi, Blakey Vermeule, and Karen Zumhagen-Yekplé. Immense thanks to all of you.

1. Toril Moi, Revolution of the Ordinary: Literary Studies after Wittgenstein, Austin, and Cavell (Chicago: Univ. of Chicago Press, 2017) (hereafter cited as RO); Hans Ulrich Gumbrecht, Production of Presence: What Meaning Cannot Convey (Stanford, CA: Stanford Univ. Press, 2004) (hereafter cited as PP); and Alexander Nehamas, Only a Promise of Happiness: The Place of Beauty in a World of Art (Princeton, NJ: Princeton Univ. Press, 2007) (hereafter cited as OPH). I will use the abbreviation "NIH" to designate Moi's essay "'Nothing Is Hidden': From Confusion to Clarity; or, Wittgenstein on Critique," Critique and Postcritique, ed. Elizabeth S. Anker and Rita Felski (Durham, NC: Duke Univ. Press, 2017), 31–49. I will refer throughout to Nehamas's "Writer, Text, Work, Author," in Literature and the Question of Philosophy, ed. Anthony J. Cascardi (Baltimore: Johns Hopkins Univ. Press, 1987), 265–91; Nehamas's "Mythology: The Theory of Plot," in Essays on Aesthetics: Perspectives on the Work of Monroe C. Beardsley, ed. John Fisher (Philadelphia: Temple Univ. Press, 1983), 236–52; and Nehamas's "Art, Interpretation, and the Rest of Life," Proceedings and Addresses of the American Philosophical Association 78, no. 2 (2004): 25–42.

2. R. Lanier Anderson has not published this reading of Pride and Prejudice; I am drawing on occasions when we taught the book together.

3. Jane Austen, Pride and Prejudice (New York: W. W. Norton, 2000), 63.

4. Austen, Pride and Prejudice, 180 (emphasis mine).

5. The revelation in question takes place on pp. 241–42.

6. It is perhaps worth spelling out that the surface/depth metaphor can be used both of texts and of interpretations. (My thanks to Nirvana Tanoukhi for urging me to make this clear.)Just as a snorkeler can make a shallow dive into a deep ocean, so too a critic can perform a superficial reading of a text with hidden depths. (Conversely, it's not a great idea to attempt a deep dive into the shallow end of the pool.)

7. Austen, Pride and Prejudice, 133.

8. Sharon Marcus and Stephen Best, "Surface Reading: An Introduction," Representations 108, no. 1 (2009): 1–21 (hereafter cited as SR).

9. Susan Sontag, "Against Interpretation," in Against Interpretation and Other Essays (New York: Picador, 2001), 3–14.

10. Moi states that it is necessary to abolish the notion of depth in order to weaken the hold of symptomatic reading: "to get away from the belief that suspicion is the only possible attitude for a serious literary critic, we need to break with the picture of texts as objects with surface and depth" (RO 5).

11. Natural phenomena are one example: "it is extremely difficult," writes Gumbrecht, "for us not to 'read,' not to try and attribute meaning to that lightning or to that glaring California sunlight" (PP 106).

12. Sontag, "Against Interpretation," 9.

13. Nehamas, "Art, Interpretation, and the Rest of Life," 33, emphasis mine. As Nehamas put it to me in an email (January 5, 2017), "The notion of depth encourages the metaphysical approach, because if you accept it, you need to find bottom somewhere, and only some foundational metaphysics can give you that."

14. Again I am drawing to some extent on personal communication. In a written comment on a version of this paper (January 4, 2017), Nehamas explained that his target is "the notion that what you call 'superficial' is a level of meaning, fully grasped, and to be replaced by the deeper truth. Whereas I think that what you call superficial is only partial[,] to be completed by the full interpretation." I am grateful to Nehamas for the extremely clarifying exchanges.

15. Nehamas, "Writer, Text, Work, Author," 277.

16. Nehamas, "Mythology," 193 (emphasis mine).

17. Cf. "the surface of a work (or its appearance) includes those of its aspects that are not, on some particular occasion, in question" (Nehamas, "Art, Interpretation, and the Rest of Life," 34).

18. Nehamas, "Art, Interpretation, and the Rest of Life," 34. This idea becomes clearer in Nehamas's discussion of plot, which one might perhaps think of as a synecdoche for surface more generally. It is not, says Nehamas, that we "once and for all settle the plot of a novel and then go on to its meaning" (Nehamas, "Mythology," 193). Rather, "the plot is in each case taken to be constituted by those features of the work that are . . . not assumed to be controversial" (Nehamas, "Mythology," 192). "Answered questions tend to be considered as part of the plot," Nehamas notes; "unanswered ones, or those answers about which agreement is yet to be reached, tend to remain parts of its meaning" (Nehamas, "Mythology," 193).

19. "Whenever conveying or exemplifying an ethical message is supposed to be the main function of a work of art," Gumbrecht notes quite rightly, "we need to ask—and indeed the question cannot be eliminated—whether it would not be more efficient to articulate that same ethical message in rather straightforward and explicit concepts and forms" (PP 103). On this point, cf. Richard Shusterman: "propositional truth, we philosophers need reminding, is not the only worthy reason for reading and interpreting" (Shusterman, Surface and Depth: Dialectics of Criticism and Culture [Ithaca, NY: Cornell Univ. Press, 2002], 69.)

20. For these and other non-message-based effects of fictions, see Joshua Landy, How to Do Things with Fictions (New York: Oxford Univ. Press, 2012), 4–7.

21. Things are admittedly a little murky here, because the Best and Marcus piece is the introduction to a special issue of a journal. Accordingly, part of it is devoted to summaries of the essays in the issue. And since the essays are (as is common even for special issues) rather heterogeneous, one could easily come away with the impression that all kinds of reading strategy fall under the "surface reading" designation. But in those cases where something is at stake other than the transmission of "meanings" and "truths," it is hard to see how we are dealing with a refusal of depth. If the aim of Best and Marcus is to reject symptomatic reading in favor of a plurality of approaches, that is entirely welcome; I see no reason, however, for them to lump the alternative options under the designation "surface reading." I am grateful to Ruth Martin for helping me think this through.

22. The same may perhaps be true of Heather Love, who embraces "a method that. . . . considers what texts do say, rather than what they don't or can't" (Love, "Close but not Deep: Literary Ethics and the Descriptive Turn," New Literary History 41, no. 2 [2010]: 383; emphasis mine). Love seems to recommend replacing a practice of looking to literary texts for messages with a practice of looking to literary texts for descriptions. The text thus becomes a purveyor of fact rather than of truth, but its role as transmitter of propositional knowledge remains, as far as I can tell, unquestioned.

23. I can't help finding such extreme conclusions a signal that something may have gone wrong with the earlier argumentation. A similarly curious endpoint is reached in "Mythology," where Nehamas presents artworks as offering nothing more than a sophisticated brand of entertainment: "great literary works do not give their readers something in addition to entertainment," he writes, "but entertainment of the highest sort" (Nehamas, "Mythology," 195). This contrasts sharply with Nehamas's own later view (OPH 129–30 and elsewhere) according to which each artwork we love sends us off on a quest, one that can enrich or impair our lives—and sometimes alter the very nature of who we are—in dramatic and unpredictable ways.

24. Nehamas: "Interpretation need not be seen as the revelation of a text's hidden meaning" (Nehamas, "Writer, Text, Work, Author," 276). Moi: "It is possible to write about interpretation without thinking in terms of metaphors of surface and depth." ("NIH" 44)

25. See RO 217–21. Rita Felski, too, thinks that readers should be "touched, troubled, perhaps even transformed," and that the hermeneutics of suspicion is deadeningly dull: "there is no moment of revelation, no startling of consciousness, no transformation of thought. . . . what a text ultimately portends is foretold by a prior theoretical-analytical scheme" (Felski, The Limits of Critique [Chicago: Univ. of Chicago Press, 2015], 65, 64).

26. One example is Paul de Man's reading of Proust, which I discuss in Philosophy as Fiction: Self, Deception, and Knowledge in Proust (New York: Oxford Univ. Press, 2004),195–96nn47–48.

27. Tellingly, Nietzsche himself seems perfectly happy to use metaphors of surface and depth, which he generally deploys in a fairly conventional way. Depths, for example, are things that are hidden by surfaces (Beyond Good and Evil, trans. Walter Kaufmann [New York: Random House, 1966], §32; Twilight of the Idols, trans. R. J. Hollingdale [London: Penguin, 1990]; "The Four Great Errors," §3). Surfaces are also often deceptive (Human, All Too Human: A Book for Free Spirits, trans. Marion Faber and Stephen Lehmann [Lincoln: Univ. of Nebraska Press, 1994], §4; Beyond Good and Evil §24). And they are where the really interesting material is to be found (The Gay Science, trans. Kaufmann [New York: Random House, 197]), §354; Nietzsche's Werke [Leipzig: C. G. Naumann, 1903], 13: 52–53; The Birth of Tragedy, trans. Kaufmann [New York: Vintage, 1967]; The Will to Power, trans. Walter Kaufmann and R. J. Hollingdale [New York: Random House, 1967], §853:3; Twilight of the Idols 9:20; The Genealogy of Morals, trans. Walter Kaufmann [New York: Random House, 1969], 2:22 and 3:23). Accordingly, philosophers who can plumb such depths are themselves "deep," while those who cannot are "shallow" (Beyond Good and Evil, §191; Twilight of the Idols, "Expeditions of an Untimely Man," §45; Human, All Too Human, §262; Gay Scimce, §173; Gay Science poems, §3; The Birth of Tragedy, §7). The thinkers Nietzsche approves of tend to be deep in a second way, too, in as much as they hide their genuine thoughts from others (Beyond Good and Evil, §40, §289, §290). Nietzsche does not appear to have worried, then, that his own (conventional) use of such metaphors was likely to lead to bad metaphysics. Neither, I feel, should we.

28. On the distinction between knowledge and know-how, see Gilbert Ryle, "Knowing How and Knowing That," Collected Essays 1929–1968 (London: Hutchinson, 1971), 2:212–25.

29. For Proust, see Landy, Philosophy as Fiction, 59–65 and 114–16, and compare Landy, "Still Life in a Narrative Age: Charlie Kaufman's Adaptation," Critical Inquiry 37, no. 3 (2011): 513–14.

30. Cf. Milan Kundera, The Art of the Novel, trans. Linda Asher (New York: Harper, 1988), 6, 7, 18, 134; Kundera, The Curtain: An Essay in Seven Parts, trans. Asher (New York: Harper, 2008), 70–71; Mikhail M. Bakhtin, The Dialogic Imagination, trans. Caryl Emerson and Michael Holquist (Austin: Univ. of Texas Press, 1981), 311, 341; and Toni Morrison: "I don't want to give my readers something to swallow. I want to give them something to feel and think about," in Nellie McKay, "An Interview with Toni Morrison," Contemporary Literature 24, no. 4 (1983): 420.

31. I believe this to be the case in some of Plato's dialogues. See Landy, "Philosophical Training Grounds: Socratic Sophistry and Platonic Perfection in Symposium and Gorgias," Arion: A Journal of Humanities and the Classics 15, no. 1 (2007): 63–122.

32. On this point, compare Felski: "Why not think of a text as gradually yielding up its interpretive riches rather than being probed for its unconscious contradictions? . . . There is no need to resort to repression, in other words, to account for contradiction, nuance, or implicit meaning." And accordingly, "depth interpretation does not have to be antagonistic . . . . interpretation can be respectful, even reverential, in tone" (Felski, The Limits of Critique, 66, 55, 57).

33. I am grateful to Klas Erik Molde for pressing me on this issue. I am certainly not saying that all texts are layered in the way I've described. Other things I'm not saying: (1) that any hidden element counts as a source of depth; (2) that hidden depths always improve a literary work; (3) that the absence of hidden depths, conversely, automatically impairs a literary work; (4) that reading for depth is ipso facto preferable. (After all, a deep reading of a shallow text is often just an overreading.) And while we're on the subject of caveats, I'm also not saying (5) that matters of depth and superficiality are obvious, or objective, or decidable in advance. Judgments as to whether a given text has hidden depths, as to whether a given reading is superficial, and as to what kind of reading is most apt, are like most other judgments in literary criticism: they are local to a particular object or group of objects; they have to be argued for separately; and they are subject to endorsement or disendorsement by other readers. None of this, however, has stopped us in the past, and neither should it stop us in the future, from making provisional claims of various kinds about poems, plays, and novels.

34. Moi might well agree that readers have work to do in order to figure out Austen's intention; she would deny, however, that Austen's intention is hidden. All features of a work, Moi writes, are visible ("to ask 'Why this?'. . . . is to ask about what's there, on full display, in the poem, or painting, or film" [RO 203, emphasis mine]), and all intentions can be inferred directly from those features, since "intention" is nothing other than the work being done by them. ("To ask 'why this?' . . . . is to ask what work this feature does in the text." (RO 203). Hence, Moi declares uncompromisingly, "[the question] 'Why this?' is unrelated to the metaphysics of the hidden" [RO 181]). But what if an author's project involves the intention to hide something? Austen wants us to discover the truth about Darcy's feelings, but she wants us to get it wrong first. Austen wants us to detect the free indirect discourse in the hauteur sentence, but she wants us to miss it first. Austen wants us to deduce what Pride and Prejudice is really for, but she wants us to make mistakes first. Austen, in short, is deliberately withholding both the what and the why; her intention is to hide (temporarily) some of the mental states of her characters, and her intention is even to hide (temporarily) part of her own intention. To be sure, we do eventually infer from Austen's use of free indirect discourse to a postulated intention, just as Moi suggests, but we do not do so right away. (As we'll see below, things do not have to be hidden forever in order to count as hidden.) And the intention we reconstruct is in part the intention to conceal something. So the "why this?" question, in cases like these, is not at all unrelated to the metaphysics of the hidden.

35. Quoting Michel Tournier, Nehamas warns us of the "strange prejudice which sets a higher value on depth than on breadth" (OPH 124). It's true, certainly, that depth isn't always more valuable than breadth, but surely it sometimes is?

36. Intriguingly, Nehamas himself uses metaphors of depth when discussing Socrates's use of irony. "Irony often insinuates that something is taking place in you that your audience is not allowed to see," he writes; "it suggests depth" (Nehamas, The Art of Living: Socratic Reflections from Plato to Foucault [Berkeley: Univ. of California Press, 1998], 67; compare also 68, 88). I think Nehamas is right that irony suggests depth—and I think this also applies to cases of authorial irony in literature. The cases are not identical, but one might perhaps think that Nehamas has at least prima facie reasons for attributing depth to certain literary texts, in the way he attributes depth to certain historical figures.

37. For all of this, see Landy, "Philosophical Training Grounds," esp. 91–97.

38. Landy, How to Do Things with Fictions, 43–65.

39. "Abraham I cannot understand; in a way, all I can learn from him is to be amazed" (Søren Kierkegaard, Fear and Trembling, trans. Alastair Hannay [London: Penguin, 1986], 66). "This was the point to which the whole preceding discussion was intended to lead. Not to make Abraham more intelligible thereby, but in order that his unintelligibility might be seen . . . for, as I have said, I cannot understand Abraham, I can only admire him" (Kierkegaard, Fear and Trembling, 136).

40. Moi admits that the Abraham story is mysterious, but she denies "that the text hides anything" (RO 190); "nothing is hidden," she says, in Genesis 22:1–19 (RO 189). Thus for example "God's intervention isn't the revelation of a secret, it's just what God does" (RO 189). But should we really say that nothing is hidden in God's behavior toward Abraham? God tells Abraham to kill his son, but he does not say why; he does not say how this command squares with his earlier promise to make Abraham "the father of many nations" (Gen. 17:4, New International Version [Grand Rapids: Zondervan Bible Publishers, 1978]); he does not say whether or not it is a test, or if so, a test of what. God, of course, has answers to these questions (as Kierkegaard sees it), but he is just not giving them. He is deliberately withholding them from Abraham—as, indeed, is only natural if the ordeal is some kind of test. The same thing operates, for Kierkegaard, at the level of the telling. God, the author of the Bible, knows what his purpose is in including the Abraham story. He knows what we readers are supposed to do with it, how we are to let it change us, and what it is for. But God keeps all that to himself too—as, indeed, is only natural if we are to be tested. From the point of view of someone like Kierkegaard, then, the story is full of studious reticence: Abraham is deliberately and successfully withholding from Isaac what his plans are (Genesis 22:8); God is deliberately and successfully withholding from Abraham what his plans are; God is deliberately and successfully withholding narrative information from us, including Abraham's mental states; and God is deliberately and successfully withholding from us the point of our engagement with the tale. It seems to me that concealment by design is precisely what we mean by the verb "to hide." So I'm not sure why it makes sense to say that "nothing is hidden" in the Abraham story.

41. Nehamas, "Mythology," 189; Nehamas, "Writer, Text, Work, Author," 277. As we saw above, Nehamas believes that "even in those cases where we say that an action or a text means something other than what it appears to mean, we do not have two meanings, one real and one apparent. All we have, even in the case of psychoanalytic or Marxist interpretation, is a series of progressively more complicated, detailed, and sophisticated hypotheses" (Nehamas, "Writer, Text, Work, Author," 277). But think of a standard bit of psychoanalytic interpretation: the patient thought his/her dream was about cigars, say, but it was really about sex. Here it feels a bit strained to say that the sex interpretation is merely a more complicated, detailed, and sophisticated version of the cigar interpretation; I think most of us would see it as something radically different—a change of subject, almost—where the initial hypothesis has not been refined but has simply been discarded. Similarly, our initial hypothesis about Jonathan Swift's A Modest Proposal (horrifying propaganda!) is simply discarded once we form our new hypothesis (brilliant satire!). Note, again, that "it's satire" is not more complicated or more detailed than "it's propaganda"; the new hypothesis needs only to be better, in the sense of explaining the data more effectively.

42. "'Being' is obviously not a real predicate; that is, it is not a concept of something which could be added to the concept of a thing. . . . By whatever and by however many predicates we may think a thing ... we do not make the least addition to the thing when we further declare that this thing is" (Immanuel Kant, Critique of Pure Reason, trans. Norman Kemp Smith [Edinburgh: Macmillan, 1929], sec. B 626–27).

43. When we engage with artworks, we frequently ask ourselves why things are the way they are; as Moi puts it, "why this?" is a very natural question to ask ourselves (RO 180–82). It is equally natural for us to formulate these questions in purposive terms: Why does Austen choose to use free indirect discourse? Why does Plato choose to put bad arguments in Socrates' mouth? What we end up with is a hypothesis about intention, albeit one attaching to a fictive "postulated author" rather than to the real-life empirical writer. (For all of this, see Nehamas, "The Postulated Author: Critical Monism as a Regulative Ideal," Critical Inquiry 8, no. 1 [1981]: 133–49.) But notice that a hypothesis about the overall project behind an artwork is not just another feature of that artwork, to add to its length, its plot, its dramatis personae, etc. It's an emergent property of an artwork, and it lives, so to speak, in a separate ontological space. You cannot explain blobs of paint merely by referring to other blobs of paint; you need a hypothesized intention, and that is a creature of an entirely different species.

44. That is why Nehamas's account of plot feels unsatisfactory to me. Nehamas, you'll recall, says that "answered questions tend to be considered as part of the plot" and that "the plot is in each case taken to be constituted by those features of the work that are . . . not assumed to be controversial" (Nehamas, "Mythology" 193, 192). Quite a few plot points, however, are controversial or even indeterminate. It is never stated, for example, exactly how many legs the arthropod Gregor Samsa has; yet whatever number a given reader guesses, those legs generally remain part of the plot. (Even a reader who thinks "Gregor has some unspecified number of legs" may well be thinking of this as simply a plot point.) Conversely, there are plenty of answered nonplot questions: everyone agrees that Voltaire's Candide is about theodicy and that Jean Racine's Phèdre is in rhyming alexandrines, yet no one considers the topic or verse form to be part of the plot.

45. "Explanation," here, should be understood in the causal sense. The existence of gravitational forces of attraction explains—in the sense of giving a reason for—the fall of an apple; the fall of an apple, by contrast, does not explain (in the sense of giving a reason for) the existence of gravitational forces of attraction. Likewise, Austen's narrative tricks are not the reason she formed her project; on the contrary, her project is the reason why the narrative is full of tricks. And again, George Orwell's decision to have us think of Napoleon as a farmyard version of Joseph Stalin is the reason that Orwell has Napoleon exile Snowball; Napoleon exiling Snowball, by contrast, is not the reason for Orwell's decision to present Napoleon as a farmyard Stalin. I am grateful to Andrew Huddleston for helping me to think this through.

46. Some people read Animal Farm and just see pigs and horses; others also see Revolutionary Russia. Some people read Pride and Prejudice and just see men and women trying to get married; others also see free indirect discourse. But there's no such thing as a reader of Pride and Prejudice who notices the free indirect discourse and not the marriageable youth, just as there's no such thing as a reader of Animal Farm who notices Leon Trotsky and not Snowball the pig. The set of readers who "get it" is a proper subset of the total readership; no one understands the deep without understanding the shallow. And this asymmetry is exactly what you'd expect if there were a relationship of surface to depth: everyone starts on the surface, and only some break through to what's beneath.

47. See RO 180. And compare RO 178: "there simply is no need to think of texts ... as hiding something." It might of course be thought that Moi is only referring to language in general—not to literary language—when she quotes Wittgenstein's "nothing is hidden," and that she only means nothing is hidden behind words, as opposed to behind texts. But Moi herself seems to want to carry the thought over to literature. She uses "nothing is hidden" as part of the section heading for a discussion of literary reading (RO 180), and she claims that we should "think of a poem, a play, a novel as a particularly complex action," just as Wittgenstein "thinks of utterances as actions" (RO 180). If nothing is hidden, for Wittgenstein, behind the actions that are utterances, nothing is hidden, for Moi, behind the actions that are literary texts.

48. Ludwig Wittgenstein, Philosophical Investigations, trans. G. E. M. Anscombe, P. M. S. Hacker, and Joachim Schulte (Hoboken, NJ: Wiley-Blackwell, 2010), §435.

49. As my wonderful former student Jake Kohn put it to me in the course of conversation, situations like these make us feel smart, not foolish: rather than thinking "silly me, how could I have missed that?" we think instead "look what I found!" Again, it seems to me that we really want our critical vocabulary to capture accurately the phenomenology of reading. And the metaphor of hidden depths—depths that are often difficult to discern, even for those who keep looking—does a pretty good job of that.

50. Stéphane Mallarmé, Oeuvres Complètes (Paris: Gallimard, 1998), 1:37–38. My (clumsy) translation.

51. Robert Greer Cohn, Toward the Poems of Mallarmé (Berkeley: Univ. of California Press, 1979), 244. Of course, the first syllable of "septuor" does not sound exactly like "sept," but the effect is close enough for most listeners to hear the numbers being counted out.

52. I am grateful to Moi for some very patient exchanges on this topic.

53. See Landy, Philosophy as Fiction, 72–73, 163n13, 171n47, 195–96nn47–48; Landy, How to Do Things with Fictions, 8, 91–2, 126, 129, 136–7, 139–40, 175n84, 181n52, 186n97, 209n7–8, 214n43, 219n88; Landy, "Conditional Goods and Self-Fulfilling Prophecies: How Literature (as a Whole) could Matter Again," SubStance 42, no. 2 (2013): 53–55; Landy, "Philosophy to the Rescue," Philosophy and Literature 31, no. 2 (2007): 405–12; and Landy, "Deceit, Desire, and the Literature Professor: Why Girardians Exist," Republics of Letters 3, no. 1 (2012),

54. For these quotations, compare, respectively, NIH 35, NIH 35, NIH 34, and NIH 41.

55. Moi writes, we recall, that "to ask 'Why this?'. . . is to ask about what's there, on full display, in the poem, or painting, or film. Nothing is hidden" (RO 203). As we are about to see, however, it is not sufficient for an element to be present in a text in order for it to count as being "on full display." Plenty is hidden.

56. Vladimir Nabokov, Lolita (New York: Knopf Doubleday, 2010), 4, 31, 221.

57. Compare "Writer, Text, Work, Author," where Nehamas claims that "coherence, unity, and meaning are generated through the proliferation of surfaces, not through the discovery of a single principle that underlies them" (Nehamas, "Writer, Text, Work, Author," 280–81). This is somewhat surprising given what Nehamas says elsewhere: "to interpret a text is. . . . to construe it as someone's production, directed at certain purposes. . . . since texts are products of expressive actions, understanding them is inseparably tied to understanding their agents . . . meaning therefore depends on an author's intentions" (Nehamas, "The Postulated Author," 144–45). Here, the proliferation of surfaces is not enough, by itself, to allow us to generate meaning; we also require an author-postulate. (My guess is that the author-postulate is also what unifies the various features of the text, which can now be seen as working together toward the hypothesized intentions.) The author-postulate, I'd like to say, is precisely the single principle that underlies the surfaces, without which no "meaning" can (reasonably) be attributed to them.

58. Another way to put the same point: going sideways may help us answer some of the what questions, but it is not going to solve the why questions. As we read on in Pride and Prejudice, for example, we learn that we were wrong about Darcy, and thus (if we did not already know) that Austen is deploying some brilliant narratorial trickery. We do not, however, learn what the point of the trickery is. That requires speculation as to the motives that lie behind what we see in front of us. It is hard, therefore, to accept Nehamas's claim that "in asking 'What is the point of this story?' we are not moving on to a new level, but we are trying to give a more detailed interpretation of that story" (Nehamas, "Mythology," 193). If I have a hunch as to Austen's project, I don't just have additional details; I have something of an entirely different nature. I have an account not just of what Austen did but of why she did it.

59. Moi does at one point admit the existence of hiddenness in Freud's thinking: "What is hidden, according to Freud, is the meaning of it all . . . . What is hidden, then, is the unconscious motivation for the dream" (RO 187). But she also says, as we saw earlier, that "Freud isn't digging under the surface" (RO 186). I'm not sure how to reconcile these two claims. When people use metaphors of depth and digging with regard to Freud, what they mean, precisely, is that he registers a set of symptoms ("surface") and infers from these to a hidden unconscious motivation ("depth"). To say, then, that Freud infers from symptoms to motivations and that he does not dig under the surface is to say something that's very hard to understand.

60. Sir Arthur Conan Doyle, Sherlock Holmes: The Complete Novels and Stories: Volumes 1 and 2 (New York: Random House 2003), 299.

61. Again, see Nehamas's "The Postulated Author," passim.

62. "All critics, whether literary or not, take for granted that new practitioners need long and serious training" (RO 218).

63. Shamela does contain an exchange of letters between two imaginary parsons in which they refer to Pamela, but an uninstructed reader might well take that book to be as imaginary as the parsons. The parsons themselves consider it not a novel but a piece of reportage.

64. We could also mention nonparodic forms of allusion, such as the traces of Homer in James Joyce's Ulysses. Here too there is plenty to love without the Odyssey, but even more to love once we recognize its presence. And here too there is something you just have to know: if you've never heard of Homer, you aren't about to find out that Leopold Bloom is a latter-day Odysseus no matter how hard you look at the text.

65. A fortiori, no amount of staring at a Renaissance painting will tell you that lilies stand for purity, orchids for deceit, and irises for regeneration, let alone that a woman with wings and a helmet stands for victory. As is well known, allegorical systems are often highly conventional; if you don't know the relevant set of conventions, you are out of luck.

66. Lewis Carroll and Martin Gardner, The Annotated Alice (New York: W. W. Norton), 1999.

67. My guess is that most of us (myself included!) don't use the terms "deep" or "depth" very much. That said, I suspect that almost all of us use the term "superficial," and use it a lot. Part of that asymmetry may have to do with the general aversion of literary folk to enthusiasm. (We're much better at being negative than at being positive; being positive makes us uncomfortable.) But part of it may also be that the word "deep" has more than one usage. Given that there are those who refer to ideas as deep—meaning something like a combination of important, nonobvious, and true—we may fear being misunderstood in applying the term to a text or a reading. That is still no reason to jettison the metaphor, especially when options like "layers" and "levels" are available.

Additional Information

Print ISSN
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.