Timothy Bewes’ recent, ambitious, widely admired, and award-winning Free Indirect: the Novel in a Postfictional Age is an “historico-philosophical approach” to the “great inventions of the novel”—“the laws of causality, narrative purpose, and individual motivation”—and an attempt to show what emerges when they become “unreliable” or “fall desperately short.”1 What distinguishes the “historico-philosophical approach,” Bewes says, is that it “rejects the capacity of subjective categories to explain the formal features of literary works (or anything else)” (FI, 44). What is a subjective category? One “whose unity is located within the self.” Why does the historico-philosophical approach reject these categories? Because “entities such as the writer’s beliefs or intentions, are not derivable from a reading of the work. Nor, therefore, can they be held to have any explanatory privilege regarding its meaning.”

Jensen Suther’s even more recent, equally ambitious, and at least equally admirable True Materialism: Hegelian Marxism and the Modernist Struggle is also about the novel and, although it does not call itself historico-materialist, is significantly influenced by the critic who invented the term, Georg Lukács. More to the point, questions about causation, purpose, subjectivity, and narrative are also central to Suther. But where, for example, “narrative purpose” is at the heart of the things Bewes wants to get beyond, for Suther’s “Hegelian approach,” both narrative and purpose are the sorts of things you not only cannot get beyond but cannot even coherently imagine getting beyond. “[T]here can be no novel in which the intention of an author or artist is not at work,” Suther says,2 and the “self” that Bewes rejects as irrelevant to a work’s “meaning” is actually “a non-optional condition” of its “intelligibility.”

One reason that intention is crucial in both these texts is suggested by Bewes’ remark that it is unable to explain not just the formal features of literature but also of “anything else.” That is, the link between the question of authorial intention and the question of action (i.e., doing anything) in a text like “Against Theory” (does it matter if the marks in the sand were made by waves washing against the shore rather than by someone writing with a stick?) has emerged as central. Hence, invoking Elizabeth Anscombe’s Intention, Mark Vareschi has observed that situating it as a topic in the theory of action might help “reinvigorate intention as a useful concept in literary study.”3 And although Vareschi explicitly situates Anscombe against the arguments of “Against Theory,” most of the more recent literary theoretical texts do exactly the opposite. As does the Hegelian turn to intention in texts like Suther’s and especially Nicholas Brown’s Autonomy. But, of course, the argument of “Against Theory” was that every critic who ever reads anything is, practically speaking, dealing with the author’s intention. And Bewes’ whole idea is that the “post-fictional” texts he is interested in—for example, texts that refuse to provide any “grammatical evidence of a narrator’s point of view” (like Coetzee’s Elisabeth Costello)—do not do this.

The value for Bewes of this refusal of a narrator’s point of view is its utility “for anyone interested in exploring the possibility of an end to the schema in which the meaning of the literary text has its ultimate determination in the figure of the expressive (intentional or unintentional) subject” (FI, 106). In other words, if there is no way in practice or in principle to determine the narrator’s point of view, then there is no relevant subject to appeal to. And what critics call the “difficulty” of establishing the “source of the point of view or opinion expressed” in Coetzee’s recent work amounts to “not merely an ambiguity but the dissolution of point of view as a signifying problematic” (FI, 86).

But supposing this is true about Coetzee’s narrator, how does that make it true about Coetzee? Indeed, if we are supposed to see in the narrator a refusal of the “figure of the expressive (intentional or unintentional) subject,”4 how can it possibly be true about Coetzee since, if he did not mean us to see that the narrator has no point of view, we cannot possibly be supposed to see it. Bewes cites Cora Diamond’s reading of Elizabeth Costello for its critique of all the philosopher readers who try to evaluate the arguments in the talk Elizabeth gives about her horror at the way we treat animals. In these readings, Elizabeth “is taken to be nothing other than a device for Coetzee, who remains in all such readings the expressive subject of the work” (FI, 176). But Diamond’s point is that Coetzee is not trying to present his own arguments through Costello and that instead he “gives us a view of a profound disturbance of soul, and puts that view into a complex context.”5 Whether this undoes the idea that Elizabeth is an expressive subject is unclear, but it is totally clear that it does not undo the idea that Coetzee is an expressive subject. What Coetzee is doing, according to Diamond, “expresses a mode of understanding of the kind of animal we are,” and the whole problem with his philosopher critics is that they do not see what he is doing.

Insofar as he accepts Diamond’s reading,6 we could say that Bewes does see what Coetzee is doing but does not see that he is doing it. In fact, his whole picture of intention makes it in principle impossible for him to see it. That is what it means to proclaim that “entities such as the writer’s beliefs or intentions … are not derivable from a reading of the work.” What is wrong with this is not that the opposite is true—it is not that intentions actually are derivable from a reading of the work. The problem is rather the picture of the author’s intention as being in one place (in his head? in any event, somewhere outside the work) while the work is someplace else, so you might be able to “attach” one to the other (FI, 109). This is what Stanley Cavell, under the influence of Anscombe, called the “bad picture” of intention, and once you have got that picture, it makes complete sense to refuse what Bewes calls “a pre-Barthesian” notion of “critical-authorial sovereignty” and to insist that whatever the author was thinking does not have “any explanatory privilege regarding” the text’s “meaning.” If the text, once it has been produced, stands on its own, why, for example, should what was going on inside the author’s head matter more than what is going on inside the reader’s head? Or why should it matter what is going on inside anyone’s head?

But the basic point of Anscombe’s good picture—a point that Suther understands when he describes the author’s intentions as “constitutive of and internal to the thing itself” (TM, 97)—is that the intention is not some “entity” existing prior to or independent of the act. And the entire point of the wave-poem in “Against Theory” was that to see the marks on the beach as a poem—even to see them as letters in English words arranged in lines—was to see them as written by someone who, just for starters, knew English, knew how to write, and had the concept of the poetic line. We do not go outside the poem to see that it is written in English; to see it as a poem is already to see it as written in English. Seeing the marks as words in a poem is seeing them as made to do something, and reading the poem is seeing (or, if it’s hard, figuring out) what they were made to do. So, questions about what the text means are intrinsically (right from the start, not optionally) questions about what the writer means. And our interest in what the writer meant is not a function of granting her “sovereignty” over the text’s meaning; it is a function of our identifying it as a text and understanding it to mean anything at all.

Suther makes a version of this point when he says that Barthes’ refusal of “authorial intention” involved not only the death of the author but the death of the reader and even of the text since without

the idea of readerly responsibility that the idea of authorial intention presupposes, what a work means is whatever it is taken to mean. …  If a work means whatever one takes it to, then there is nothing to which one bears any responsibility. If the author dies and there is … no determinate meaning about which we can be right or wrong, then the reader dies—because there is no one to occupy the recognitive position defined by responsibility to such a determinate meaning about which we can be right or wrong. (TM, 280)

Suther here usefully runs through one of the criticisms of the idea that the author’s intention can be dispensed with by an interpreter: if the author’s intention is not the object of interpretation, then there can be no right or wrong interpretations because there is nothing to be right or wrong about,7 and if there is nothing to be right or wrong about—if the work just means what a reader takes it to mean—then the reader is not actually reading a text, just reporting his reaction to it. Hence there can be no such thing as agreement (responding in the same way is not agreeing) or disagreement (responding differently is not disagreeing). So, the problem with the death of the author is not that it threatens our ability to get true interpretations or that it elevates the reader but that it makes reading disappear and makes all the activities that go with understanding what someone has written (or said)—really all the activities that go with any use of language, which is to say, language itself—literally incomprehensible: we have no account of what understanding or misunderstanding a text would be. The proper form of the old war cry “all readings are misreadings” would actually be “there are no such things as readings.” In the terms that Suther picks up from the post-Sellars tradition in American philosophy, the space of reasons is replaced by the space of causes.

From this standpoint, seeking to dispense with authorial intention is a problem not because we should not do it but because we cannot possibly do it. And Suther sometimes makes this point. But what he says here (and other parts of his argument take up this rhetoric) is not just that we cannot do it but that it would be irresponsible to try—that the reader “bears” a “responsibility” to the text. But how can there be responsibility without irresponsibility? And if reading just is trying to figure out what some author meant, how can anyone read irresponsibly? Or, more precisely (since we could just skip bits, not look up words we do not know, et cetera), how can responding to a text without interesting oneself in what the author meant count as irresponsible? It would perhaps be irresponsible to claim that one’s response to a text was really an interpretation of it, but the irresponsibility would not be in the response itself (it makes me think of my grandma) but in the description of that response as a reading (it is about my grandma), and even then the description would be not so much irresponsible as mistaken. The strength of Suther’s account here is that he describes what reading is; the peculiarity is that he then converts his description of what reading is into an account of what, if we behave responsibly, it ought to be.

This sense that readers are failing if they do not attend to the author’s intentions is sometimes matched in Suther’s text by the sense that authors (or speakers) can fail if they do not properly attend to their readers. The structure of this worry is explicitly Hegelian, and once again, Suther convincingly puts it in non-optional terms: there is a “‘demand for recognition’ built into intentional action itself,” and this demand is not just to our empirical audience but to the “norms” that make action possible: “we are responsible not only to one another but also to a practice or institution itself” (TM, 39). And, of course, here too, we see the question of responsibility appear, the hint of the optional. But, here again, just as the reader cannot help but attend to the writer’s intention (whether or not he correctly understands it), the writer cannot help but seek recognition, whether or not she succeeds in getting it.

In Suther, it looks like what provides the slippage from what you cannot help but do to what you ought to do is the distinction between “subjective purpose” and “objective intention.” “Sentences,” he says, “mean what they do as bearers of intentions,” they are the actions of “individual speakers.” But they cannot really count as their actions unless the speakers meet the “demand for recognition” (TM, 34). And what gets recognized need not be what was originally intended: “Because an agent’s intention must be embodied in either utterances or acts (in perceptible signs or perceptible bodily movements), an agent’s subjective purpose can come into conflict with her objective intention” (TM, 308). Why isn’t the agent’s subjective purpose dispositive? Because “an intention is not a private psychological content that precedes a speech act,” because it is not “simply what the agent says it is,” because an agent can “get” herself wrong or “deceive” herself. The point, then, of the category of “objective” intention is that it acknowledges but transcends “subjective purpose”; it is “what can be ascribed to the agent as the embodied speaker or actor that she is.” Her speech act means what she intended it to mean, but what she intended it to mean is not necessarily identical to her “subjective purpose”; it is what objectively turned out to be her purpose.

But what does objective add to subjective? How can the objective intention differ from the subjective one? Clearly what you say or write can fail to accomplish what you meant it to accomplish. But the fact that you do not laugh at my joke (or are not convinced by my argument) does not alter its meaning. Equally clearly, my account of my meaning may be at least partially mistaken. I thought of myself as just trying to amuse, but even I can be brought to see that I was trying also (and maybe even instead) to hurt. But this redescription does not alter my subjective intention; it just gives a more accurate account of it. Actually, the only way you can make the objective intention different from the subjective one is by adopting the same picture of intention that Bewes has, the one that identifies it with a “private psychological content” to which something physical (you raise your arm, you touch the keyboard) needs to be added. Or (the same thing in a different direction) you raise your arm, to which your intention to raise it needs to be added. But the whole point of the Anscombian/Hegelian account of intention to which Suther is committed is that it refuses this description of the intentional act.

So, when Suther insists that “to grasp an agent’s intention” is “to attend to her actual deeds and not (just) to her explicit statements of intent” (TM, 37), this is right (the author’s statements of her intention are not necessarily dispositive) but not in a way that suggests her intention depends on “whether she is recognized as living up to” it. We can see what Suther is worried about (that is, why he feels nervous about “subjective purpose” and why he worries about living up to your intention) when we note that the argument I have just summarized begins with the question of the “intentionality of the language-user” but ends with the question of someone who betrays his friend but “has convinced himself” that it was not really betrayal, that he acted “in his friend’s interest” (TM, 308). But however we understand the moral judgment here, the question of what someone means by what she says does not involve her living up to her intention and is not in the slightest dependent on it being recognized by anyone else.

This is because the objectivity that recognition is supposed to provide is already built into the subjectivity it is supposed to modify. Of course, if you think of the author’s intention as “private psychological content,” you may well demand that her actions express that content and either worry that they will not or imagine that something else might take that content’s place. But if (like Suther and me and almost everyone else who has been defending intentionalism in literary theory) you think of her intention as part of the description of her act, you should see her as having already expressed it. Why? Because in order to be able to produce any speech act at all you need a language. Wittgenstein memorably makes this point—“It is only in a language that I can mean something by something”8—with his challenge: “Can I say ‘bububu’ and mean ‘If it doesn’t rain I shall go for a walk’?” And although the point of the question is that the answer is “no,” that point is made even more vividly when the answer can be reimagined as yes.

Why is the answer no? According to Elmar Unnsteinson, it is because “[u]nless the context is a very odd one, the speaker will not believe that it is possible to utter ‘bububu’ with that communicative intention.”9 But this is vulnerable to two sorts of objections. The first is that it is not at all clear that Wittgenstein (as opposed, say, to Grice) would require the speaker to have a communicative intention. The speakers he imagines walking around all day “talking to themselves” do not,10 and he does not claim that they do not have a language. And the second is that the context does not have to be all that odd. If, for example, in answer to the question, “What are your plans for this afternoon?” I said to someone in my Wittgenstein reading group, “bububu,” I could easily mean and be understood as meaning, “If it doesn’t rain I shall go for a walk.” More fundamentally, not believing that or caring whether anyone will understand me, I might still, looking pensively at the clouds, say it to myself. And, of course, people who have never read the Philosophical Investigations but know people who have might also come to use this new idiom (although some non-readers might occasionally use it incorrectly—“bababu”—and need to be taught how to do it right).

The point that matters is just that if it can make sense to the speaker to say “bububu” and mean “if it doesn’t rain I shall go for a walk,” the making sense to the speaker (the explanation the speaker could give of what she is doing) is all that is needed for the speaker to have meant something. And my use of the idiom “bububu” has no more relation to private psychological content than does my use of any other idiom, say, “raining cats and dogs.” Which is to say, what it means for it to make sense for the speaker is not for her to be able to say what was in her head when she said it but for her to recognize the relevance of the question which, as Anscombe says, is relevant to any intentional act: not what is going on inside you when you say “bububu” (or what was going in inside you before you said “bububu”) but what do you mean by saying “bububu?” Why are you saying that?

The worry that underlies the sense that both writers and readers can be irresponsible—the writer by failing to have the right relation to her intention, the reader by failing to attend to the writer’s attention—is incoherent. Everyone who produces a speech act produces a text that means what she means by it; everyone who reads one is understanding (or misunderstanding) what she meant by it. This is the force of the non-optional—the reason why intentionalism cannot be a choice—the reason, really, why there is no such thing as intentionalism.

Bewes, by contrast, has basically Wimsatt and Beardsley’s view of the relevance of intention: their “A poem comes out of a head not a hat” is his “entities such as the writer’s beliefs or intentions” are “no doubt operative in the work’s composition” (FI, 45). The mistake here is not in denying that the intention as cause determines the meaning; it is in thinking of the intention as a cause, as if it were the impetus to writing a sentence rather than a description of what I mean by writing the sentence. Hence it is absolutely true that the intention is not derivable from the text and that it has no privilege in determining its meaning but only because it is already built into the text and is its meaning. And hence also Bewes cannot succeed for a second in establishing an alternative “mode of reading,” one that will disregard the intended meaning. When, for example, he praises Sebald’s work for refusing a “primary, nonobjectifying gaze that precedes the writing,” a gaze that would perforce be “located in the brain of Sebald himself“ (FI, 60), he apparently understands himself as identifying in Sebald the escape from intentionality. But the question of what is in Sebald’s brain is irrelevant, while the question of what he is doing is inescapable—on Bewes’ account, he is denying the relevance of his own intentions.

Of course, if that is what Sebald’s doing, he is making the same mistake Bewes is. Which is certainly possible. The problem for Bewes is not that he cannot be right about what Sebald is doing; it is that he cannot be right about what he himself is doing. He can be right about what Sebald is doing because there is no reason that Sebald cannot (albeit mistakenly) understand himself to be producing texts that achieve a certain “non-intentionality.”11 But he cannot be right about what he himself is doing because literally (necessarily) every account he gives of every text he discusses is an account of what the author was doing.

A striking fact about Free Indirect is that Bewes himself is aware of but is unimpressed by this criticism. He devotes several chapters to Deleuze’s idea of the “sensorimotor break”—the possibility of moments in movies when the camera records what it sees without any reference to what the characters in the film might be doing or to how the viewer of the film might understand the significance of what he is seeing. Deleuze calls them “‘pure optical and sound situations,’” “when a pragmatic function is replaced by a pure ‘seeing function’” (FI, 200). It is a better example of unintended text than any text can be because (for all the reasons people have invoked in relation to photography) the camera can bypass intentionality in ways that writing cannot. But, Bewes points out, Jacques Rancière has argued that when we recognize the “collapse” of meaning in these moments we recognize it thematically, as “allegorical images or diegetical depictions of characters (or authors) experiencing the collapse” (FI, 240). So, although the appeal to the camera gives us something a little closer to the nonintentional than any writing machine has, the actual deployment of pure seeing to some effect in a movie—precisely because it is being deployed to some effect—puts us right back where we were with Sebald: it is one thing to be about the refusal of intentionality; it is a different thing to be unintended.

How does Bewes respond to this criticism? It is not only “a contention of” his book but an “article of faith,” he says, that there is something more than “the kind of thought” that is “content” with the world where texts get interpreted,12 that there is a thought “that is not content” just to mean or understand (FI, 248). And in a series of nine (!) consecutive rhetorical questions (most of them beginning with “what if”), he wonders whether Rancière does not actually share his own “commitment to a dimension of experience or thought that passes all understanding.” Perhaps, when he criticizes Deleuze, he is not “speaking in his own voice,” “What if we read him for what, in the formal and discursive context of his writing, he is not able to reflect on?” (FI, 249). “How else to register the unspeakable,” Bewes asks, “than by not speaking it?” (FI, 253).

Setting aside his recourse here to almost the worst possible picture of intention (if you are really not speaking, even to yourself, how can we tell that what you are not speaking of is the unspeakable?), the “what if” strategy is actually a characteristic one for Bewes, whose faith several times takes the form of hoping that critics who appear to be saying one thing might really be saying the opposite. Sometimes it is sort of a joke, as when he invites us to imagine that Mark McGurl, dismissing “the absurd declarations of the ‘death of the author’ that were heard in the 1960s,” is really presenting himself as “something like a literary character” and that therefore “the buttonholing tone” of a phrase like “‘absurd declarations’” can be (reassuringly) taken more to distance the author from the dismissal of Barthes than to commit himself to it (FI, 109). But there are no jokes in Rancière, and both the hope and the faith are real. And if the hope is a little implausible, the need for faith is defensible. After all, almost all arguments require at least a little faith. But, as we have already seen, the problem is not so much that faith is required, it is that the thing we are supposed to believe in—a “pure” “speech act” (FI, 216), which is to say a speech act that is neither speech nor an act—is internally incoherent. “What if my fantasy were coherent” is a heavier lift than “what if it were real.” Bewes imagines the novel escaping the world in in which we speak “interpretatively—using conjunctions such as because, although, since, if or even if” (FI, 241). The difficulty here is not only imagining such a world but imagining that there would be any novels, much less writers and readers, in it. Swapping out earnest Mark McGurl for one that is more like Thackeray gives you a different world but with all the same conjunctions. Getting rid of the conjunctions too just gives you silences supposed to (but, without the normativity of “because” et al., unable to) mean their meaninglessness.

Since Suther does not insist on the importance of doing something that cannot be done but suggests instead the importance of doing something that cannot not be done, his version of the problem is pretty anodyne, and my reservations here are meant more as a friendly amendment—meant, that is, not to challenge what he does but to qualify his description of what he does. For example, in his description of his own practice—“‘absolute’ readings of exemplary individual novels, in order to grasp them in terms of what they themselves intend to do, what they claim therein that literature ought to be” (TM, 121)—the invocation of the Hegelian “absolute” may represent a kind of ambition (it is not non-optional), but the grasping the texts in terms of what they themselves intend and what they think literature ought to be is not. Even Bewes, who would prefer not to, is doing that. In other words, although Bewes and Suther have fundamental theoretical disagreements, they do not really have any serious methodological disagreements. Indeed, it is not even clear what serious methodological disagreement would look like.

The question of method or even the weaker “approach” (favored by Suther) in literary theory has often been understood as involving choices not only about how to figure out what the text means but about what constitutes its meaning—this is the intentionalism that insists we ought to look for the author’s intention, and it is the anti-intentionalism that says we ought to focus instead on the text itself or refuse the original historical meaning of the text and focus on readers’ responses to the text. But both these scenarios depend on the picture of the intention as somewhere outside the text (like in Sebald’s brain). Once we get over that picture, we see that looking for the author’s intention and focusing on the text are just two different descriptions of the same thing—reading. And while reading the text is definitely important if you want to understand it, no one has ever said, “My approach to understanding Elizabeth Costello was to read it.”

So, trying to understand what the author was doing is not an approach; maybe instead there can be different approaches to trying to understand what the author was doing. But even here it is useful to deflate the idea of methodological differences. What makes someone a Marxist critic, for example, is not a particular way of doing criticism but a set of beliefs about the way the world is structured and about what counts as an explanation of people’s behavior. Thinking that the exploitation of labor by capital is fundamental is different from thinking that the Oedipus complex is, but that is a difference about what is in the text, not about ways of discovering what is in the text.

Perhaps what is left for method are more specific local differences like that between the formalist and whatever we think the alternative to formalism might be. Several critics have convincingly argued that the “method wars” that supposedly took place in literary criticism during the last decade (between, say “critique” and “reparative reading”) were not really about method. Others, worried about the fate of literary criticism as a discipline and hoping to burnish its credentials, have added that there is in fact wide agreement among otherwise quite different critics that close reading is the default method of literary studies today. But what does close add to reading? As a way of understanding what a text means, close reading is just paying close attention to what the writer does and does not do. But “My approach to understanding Elizabeth Costello was to read it” does not become the methodological foundation for an academic discipline by being turned into “My approach to understanding Elizabeth Costello was to read it carefully.”13

In the last couple of years, there has been a flood of books and essays mainly in defense of close reading, and insofar as close reading is understood as a practice of writing (a set of procedures you can follow for writing an essay rather than a set of procedures you should follow for reading a poem), there is no theoretical reason to object to them. Whether any of those defenses is convincing is another question. On the face of it, it is hard to see why, for example, writing essays that begin by noticing something that looks relatively minor about a text and then go on to show how the text is in fact structured around whatever the minor thing was you first noticed can help save English departments, much less democracy. But that does not mean those essays are not worth writing or that they do not count as at least a genre of literary criticism. It is just that there is no particular link between the genre of writing that we identify as close reading and the activity of playing close attention.

Indeed, we do not even think there is a special connection between literature and paying close attention. No one, for example, thinks close reading is the “proprietary method” of people who have just bought a new outdoor grill (“some assembly required”) or are learning to make Marcella Hazan’s Bolognese. What makes literary close reading literary is just that it pays attention to the kinds of things poets and novelists are interested in. In the instructions for assembling grills or on the uses of a new metal, line endings do not matter; in some poems, even the spaces between the letters matter. Describing the difference between them is just describing the differences between what the guy who wrote the instruction manual and, say, the guy who wrote “The Instruction Manual” were doing.

This is the reason that there is no such thing as intentionalism, either in literary criticism or in manual reading. Intentionalism is not an approach because there are no other approaches. But, as a description, it nonetheless has a certain value precisely because it makes clear the irrelevance of method. If we want to save literary criticism, it is literature we need to defend.