In a 1926 essay, “Science and Poetry,” I.A. Richards, better known for his later book Practical Criticism and its influence on what would become the New Criticism, offers up a vivid analogy for what our bodies do in the presence of a powerful work of art:

Suppose that…we carry an arrangement of many magnetic needles, large and small, swung so that they influence one another. As we move, the perturbations in this system will be very complicated. But for every position in which we place it there will be a final position of rest for all the needles in which they will in the end settle down, a general poise for the whole system.1

When Richards goes on to claim that “[o]ur interpretation of the poem is the movement of these interests,” then if we follow his magnetic compass analogy, our “interpretation” requires above all something to measure those compass movements, and an understanding adequate to interpret the results (28). Richards himself identified the tools needed for the task, which, as he put it, “until recently could only be very incompletely carried out; the psychology of instinct and emotion was too little advanced” (21). What Richards meant by this, of course, was that with its newest developments, the “psychology of instinct and emotion” was poised to deliver a rich and complete understanding of art’s meaning, something that the discipline of criticism up to that point had been striving and failing to achieve.

No doubt the I.A. Richards who wrote those sentences would have found his own compasses spinning away at the prospects for literary criticism offered by the latest attempts to apply the sciences of instinct and emotion to art, namely what has been called most recently “neuroaesthetics.”2 For when we replace the galvanometers of the past with 21st-century functional Magnetic Resonance Imaging (fMRI) technology, or if we replace the “psychology of instinct” with the most recent developments in evolutionary psychology, we can get an even clearer picture of those “magnetic perturbations” that Richards imagined occurring in our bodies in the presence of the work of art. Evolutionary biology, for instance, tells us we’re predisposed from a long pre-human and human history of surviving through cooperation and alliance by trying to read other people’s minds to assess their trustworthiness. And fMRI, for another example, gives us a way of seeing how and when those mind-reading skills get put to use or what kinds of alterations in the brain can cause them to fail us. Interpreting works of art turns out, according to this research, to be a way of exercising all of these inclinations. But it’s not only aesthetic philosophy or art history or literature, film, theatre, music or new media studies, whose scholars and theorists have sought to benefit from the research. The present forms of what Richards termed “the psychology of instinct and emotion” are regularly deployed in the service of ethics, marketing, business management, economics, political theory, anthropology, sociology, communications, and pretty much any other disciplines and practices that live by their accounts of human motivations. With an ever clearer picture of how our minds work and of the evolutionary developments, and bodily and environmental constraints that shape those workings, the argument goes, we get a clearer picture of everything. We’re not just doing neuroaesthetics, we’re doing neurowhatever, and we’re apparently delighted with the results. This essay argues for why we should not just be delighted with the results, or rather, why we can’t be delighted with the results and still maintain a coherent account of what we’re doing when we’re doing the interpretive work of literary or art history and criticism.

A few years ago, in a book partly devoted to criticizing the theoretical claims of the Language Poetry movement, a relatively influential (as poetry movements go) self-declared avant-garde that had its heyday in the 80s and 90s, I made passing reference to some literary criticism advanced during the same period by scholars drawing on the brain research of their contemporaries, including Jerry Fodor, Gerald Edelman, and Antonio Damasio.3 And I argued then that this prototype of what we now call neuroaesthetics was making the same theoretical mistake I saw being made by the Language poets in their own self-presentation, namely a confusion of the meaning of the work of art with, on the one hand, the effects of the work on the reader/listener/beholder, and on the other hand, what we might call its causes. In other words, the Language poets embrace the same idea of the meaning of the work of art that I.A. Richards had promoted in “Science and Poetry” a half a century earlier. And the tendency, from I.A. Richards on through Language poetry through the neuroaesthetic approaches to art that I discuss below – the tendency to conflate the meaning of a work of art with the experiences that go into making the work and the experiences said to arise from it – has gone hand-in-hand with a tendency to ignore the incoherent results of that conflation.

Mary Thomas Crane’s Shakespeare’s Brain: Reading with Cognitive Theory arrived just at the moment of the shift from the so-called “cognitive theory” approaches to literature to the neuroaesthetic.4 In that work, Crane suggests that the field of Shakespeare studies will be forever altered once we recognize Shakespeare’s brain as the most important “material site of production” for his plays (3). It’s no longer the materiality of the theater, or the Elizabethan court or the invisible bullets of transatlantic trade that matter, in other words; it’s the materiality of the neurons and the transmitters that make them fire. While Crane turned brain research into a fairly crude instrument for criticism only a decade ago, even our most sophisticated 21st-century appeals to the neuroscience and evolutionary psychology that succeeded “cognitive theory,” share with their less persuasive counterparts of the last century the same basic logical mistake, and we don’t need any science at all to see that mistake. Indeed, we can simply return briefly to the height of the New Criticism in the last century, and more specifically, to William K. Wimsatt’s and Monroe C. Beardsley’s soon-to-be widely read polemics, “The Intentional Fallacy” (1946) and “The Affective Fallacy” (1949). As I hope to show in light of our current attraction to all things neuro, we might better call them “The Causal Fallacy” and “The Effective Fallacy.”5

Let’s start with the causal problem. For Wimsatt and Beardsley, committing the intentional fallacy means mistaking for the meaning of the work of art a whole host of things the author might have experienced or thought about in the making of the work: “revelations,” they explain, “(in journals, for example, or letters or reported conversations) about how or why the poet wrote the poem — to what lady, while sitting on what lawn, or at the death of what friend or brother” (10). Among the list of things that Wimsatt and Beardsley object to, the lawn seems especially representative of the problem, particularly when they turn to their chief example of misguided intentionalist criticism, John Livingston Lowes’s The Road to Xanadu (1927). In it, they argue, Lowes treats the books Coleridge is known to have read as the basis for “clusters of associations, like hooked atoms, which were drawn into complex relation with other clusters in the deep well of Coleridge’s memory, and which then coalesced and issued forth as poems” (11). In other words, Coleridge’s readings become just like the lawn on which the poet sat — or for that matter any place he might have been sitting or anything else that might have crossed his path, or his mind, while he was writing. These associations contribute, according to Wimsatt and Beardsley, to what they call the “gross body of life, of sensory and mental experience, which lies behind and in some sense causes every poem,” but which “can never be and need not be known in the verbal and hence intellectual composition which is the poem” (12). The problem with everything from the lawn on which the poet sat to the hooked atoms of his associative memory has to do with their status as causes of the work, and however distant or proximate they may be, they are categorically distinct from the meaning of the work. Wimsatt and Beardsley’s contemporary Cleanth Brooks put the point succinctly: we might as well inquire, “What porridge had John Keats?”6

My claim about neuroaesthetics is that whether we’re pointing to our Pleistocene ancestors‘ predilection for climbable trees in landscapes or which region of an artist’s brain is activated in the process of choosing a color array, we’re essentially working off of related versions of Brooks’s question, identifying the causes but not the meaning of the work. Neuroaesthetics is answering a set of questions about causes, while the interpretation of a work of art depends on having answers about its meaning.

Now of course Wimsatt and Beardsley in effect commit the very fallacy they mean to refute when they put intention on the same continuum with the “gross body of life, of sensory and mental experience” that they rightly identify as the causes of the work. It’s exactly that conflation that leads them to misidentify their fallacy as having to do with “intention” in the first place. The problem is evident from the beginning of their essay, in one of the most compact statements of the argument that they put forward: “[T]o insist on the designing intellect as a cause of a poem is not to grant the design or intention as a standard” (4). The problem with the causes of a poem, as we’ve already seen, as that they’re not dispositive; a large proportion of what goes on in the making of a work (porridge eaten, lawns sat on) is completely irrelevant to anything that counts as its meaning. And if it is relevant (Bartram’s Travels to Coleridge, John Day’s Parliament of Bees to T.S. Eliot), then whatever is relevant about it will be available in the poem itself. But the standard that Wimsatt and Beardsley offer as the corrective to a mistaken appeal to causes (and in their mind, to authorial intentions), namely the “internal” and simultaneously “public” evidence that consists in the fact that the poem is “embodied in language, the peculiar possession of the public,” is no more dispositive than porridge or lawns (5). We can consult our “grammars, dictionaries, and all the literature which is the source of dictionaries…all that makes a language and culture,” but we cannot, on the basis of that “public” information alone, decide what does and doesn’t apply to the poem’s meaning.7 We can only admit or rule out the evidence by appeal to the intention of the poem’s maker. In this respect, the only standard that actually could serve in the ways that Wimsatt and Beardsley have in mind is intention, which has nothing to do with what they rightly understand as the causes of the work. Better, as I have already suggested, to call their essay “The Causal Fallacy.” Indeed, if they had accepted the full entailments of the disjunction between meaning on the one hand, and causes and effects on the other, intention might never have emerged as a fallacy for them in the first place.

If “The Intentional Fallacy” misses its own point, failing to understand that the meaning we’re after cannot be explained except by an appeal to authorial intention, Wimsatt and Beardsley are obviously not all confused about the fact that the meaning of a work cannot be explained by appeal to its causes. And by the same token, they also rightly understand that the meaning cannot be explained by appeal to the effects of the work. In “The Affective Fallacy,” the “affective” critic in question is, in the simplest formulation, someone who mistakes the feelings inspired by a work of art — the work’s emotional and physiological effects on the reader/beholder — for the meaning of the work. Coleridge again is the case in point for Wimsatt and Beardsley’s argument: “The tourist who said a waterfall was pretty provoked the silent disgust of Coleridge, while the other who said it was sublime won his approval. This…was not the same as if the tourist had said, ‘I feel sick,’ and Coleridge had thought, ‘No, I feel quite well’” (27). When we make claims about the value of a work of art, we take our claims to hold not just for ourselves, but for others as well — if I say the waterfall is sublime I think it should count as sublime for anyone. If we didn’t have this normative expectation of our judgments, we’d have no basis for arguing that one work of art is better than another, much less disagreeing about their meaning, much less appealing to evidence to support our arguments for their meaning or their value.

For the most part, “The Affective Fallacy” is concerned with readers whom the authors identify as “affective” insofar as they, like the tourist viewing the waterfall, “testify[…] to what poetry does to themselves” (31). But another kind of “affective critic” comes to light when Wimsatt and Beardsley make brief mention of just the sort of laboratory research where the galvanometer – or fMRI — might offer insight into human responses to art. And what those who “testify to what poetry does to themselves” share in common with those who “coolly investigate what it does to others,” is once again the impulse to treat the effects of the work of art as indistinguishable from its meaning (31). Wimsatt and Beardsley cite the researchers who “inquire what kinds of colors are suggested by a line of Keats, or […measure] the motor discharges attendant upon reading it…. The affective critic,” they write, “is actually able, if he wishes, to measure the ‘psychogalvanic reflex’ of persons subjected to a given moving picture” (31). Thus, in one experiment, “‘Students have sincerely reported an ‘emotion’ at the mention of the word ‘mother,’ although a galvanometer indicated no bodily change whatever. They have also reported no emotion at the mention of ‘prostitute,’ although the galvanometer gave a definite kick’” (31). Of course, Wimsatt and Beardsley are invoking these experiments in order to shore up their argument against affective criticism, claiming that these researchers are committing, as I have already begun to suggest, the same affective fallacy as readers who think a poem means the sadness or joy or whatever other “kick” it happens to give them.

Now it’s easy to see how certain neuroscientific approaches to art could easily commit either of the fallacies — that is, how they mistake the causes of the work or the effects of the work for its meaning. Pointing to the fact that, as Mark Johnson explains in The Meaning of the Body, our bodies have a front and back, and much of our literature represents backwards and forwards movement, is certainly a way of talking about the material contingencies that in some very literal sense “cause” the work, but it by no means guarantees us an account of the work’s meaning.8 Take for example Michael Fried’s argument in Courbet’s Realism that the painter ever more relentlessly sought to make paintings in which through various manipulations of perspective and the picture plane it was as if the painter’s own body could merge with the depicted world occupying the canvas.9 Or Charles Palermo’s reading of Joan Miró’s The Policeman (1925) in Fixed Ecstasy: Joan Miro in the 1920s, where the hand of the policeman, with its ambiguously arrayed thumb and digits, allows us to imagine it both facing palmward, gesturing out of the painting toward the beholder, and simultaneously facing back into the painting, as if facing the canvas from the viewer’s side of the picture plane, an “allegorization,” as Palermo puts it, of the painter’s “own actions.”10 Or, for that matter, take the famous comment Miró himself made about Courbet’s painting, The Stormy Sea (1869): “One feels physically drawn to it as by an undertow. It is fatal. Even if this painting had been behind our backs, we would have felt it.”11

On the one hand, to see the artists’ commitments to rethinking the function of the picture plane in relation to both artist and beholder would be, following artists like Courbet or Miró, or critics like Fried or Palermo, to see those commitments as arising out of a set of problems about painting as such, problems which the work itself theorizes and displays. On the other hand, if we’re thinking like Mark Johnson, we can simply add these examples to our bucket of evidence that the human mind is structured by our bodily orientation in space, and hence so is our art. Put that way, the difference between having an account of the meaning of the work and having an account of its causes is not only easy to see, but, I would argue, an easy strike against the kinds of neuro approaches I’ve been describing thus far. And it’s an easy strike if only because the bucket of evidence for, say, our spatial orientation and its constraints on our imagination produces a general lack of differentiation between individual works of art, since presumably they’re all going to index that spatial orientation in one way or another. In short, we can as easily talk about one work or another if all we care about is understanding the shared human conditions that constrain them; whereas the spatial orientation that most humans share by virtue of our bodies (barring unusual discrepancies in ability or bodily function) will not help us understand why any given artist attempts to revise or ignore existing paradigms such as the picture plane or single-point perspective. But the question of whether something in the art functions as an allegorization, to recall Palermo’s reading, necessarily goes beyond anything we can explain by appeal to our shared bodily conditions. Indeed, two different paintings looking exactly the same might produce the same effects, but if each allegorized something different, they would necessarily have two entirely different meanings.

So far I’ve been arguing that the causes and effects (or, say, inputs/outputs or trigger mechanisms — however one might choose to name them) that neuroscience and evolutionary psychology identify and explain for us are incommensurate with the fundamental interest we have in intention, one that is basic to even our most ordinary acts of interpretation. But of course one might make the argument that, even if we grant that intentions operate in a categorically distinct register from causes and effects, nevertheless our special interest in intention is itself best explained by appeal to its neurological and evolutionary functions. This I take to be one of the central claims Blakey Vermeule’s recent book, Why Do We Care about Literary Characters?12 Through a provocative and elegant array of evidence ranging from Chaucer’s Canterbury Tales to Ian McEwan’s Atonement, Vermeule argues quite forcefully for the myriad ways in which we like to pay attention to and practice interpreting the actions and motives of others. And it doesn’t matter whether we’re concerned about real people or fictional ones. Narratives involving particularly complex epistemological problems with respect to characters’ motives entertain us because we need to practice detecting whether someone is deceiving us; it’s essential to our survival. Given our fascination with discerning complex motives, and through them complex alliances, no wonder then that God and gossip are two of Vermeule’s privileged examples. I won’t address the fascinating work that Vermeule does with gossip, mainly because I want to devote the remaining space of this essay to what she calls “God novels” — novels usually involving an omniscient narrator and a certain degree of self-reflective commentary on the capacity of the fiction either to determine (or seemingly fail to determine) the lives of its characters. God-novels are key in what Vermeule calls the “high mind-reading tradition”: “At the center of the high mind-reading novel is a narrator who adopts the standpoint of an agent with full access to strategic social information and who parcels out the information at markedly different rates, placing some characters and even the reader in a temporarily blinded position…A high mind-reading novel only makes sense in the presence of God, who stands as the final guarantor of full access to social information” (129). What we commonly imagine that God to be like is of particular interest to Vermeule.

What’s especially striking about our imagination – and what is equally striking about the analogy Vermuele ends up drawing between the most common impressions of divine omniscience and our fantasies about what perfect access to social information would look like – is that the perfect and the imperfect, the divine and the human, the all-knowing and the ignorant, cleave precisely along the faultline of causality. In a psychological experiment cited by Pascal Boyer, Vermuele explains, researchers

asked people to tell them what God is like. People gave a wide range of descriptions with certain features in common. For example, people said that God has the power to do many different things at one time, unlike humans, who are bound to do things sequentially. People were then given stories in which God is shown to be doing several things at once. However, when the people repeated the story again several hours later, they invariably described God as doing those things in sequence — first one, and then the other. People import specific narrative inferences from their own experiences to the story, even though they know, abstractly, that God is capable of doing all those things at once. Putting this finding more generally, people seem to have one set of logic centers for abstract ideas and another for what happens as their experience unfolds. I might want to write a novel. But the sheer difficulty of sitting down every day for five hours, feeling miserable, and struggling with my unruly tangle of sentences might prove too much for me. (145-6)

Another way of putting the claim that our experience unfolding belongs to one “logic center” in our brains while our abstract ideas belong to another would simply be to say that there is a fundamental logical difference between them. And that fundamental distinction plays out in the very next two sentences Vermeule writes, for it is precisely the breach between our intention (“I might want to write a novel”) and our experience (“difficulty of sitting down,” “feeling miserable,” and “struggling” that will “prove too much”).

Vermeule depicts this breach even more vividly in a brilliant reading of Ian McEwan’s Atonement, in which a scene involving a young college-aged girl and boy who love each other but don’t know it yet takes place in a garden where a third character, a 13-year-old girl, observes them and misinterprets their actions, initiating a chain of painful consequences. As Vermeule explains, the 13-year-old character is also a budding novelist, so that the scene she observes becomes the basis for a scene she imagines one day writing. And Vermeule is especially interested in a passage in which the mature character, having had a successful career as a novelist, reflects on the act of authorship and its relation to other kinds of acts:

How can a novelist achieve atonement when, with her absolute power of deciding outcomes, she is also God? There is no one, no entity or higher form that she can appeal to, or be reconciled with, or that can forgive her. There is nothing outside her. In her imagination she has set the limits and the terms. No atonement for God, or novelists, even if they are atheists. It was always an impossible task, and that was precisely the point. (McEwan, cited on 135)

For Vermeule part of the fascination that McEwan’s novel offers is its ability to deploy in ingenious ways “human social signaling” that can “‘trigger inferences’ about education, leisure time, communities of taste, and ultimately class background” (134). But we might say that another part of its fascination has to do with the ways in which the privileged vantage on that social information – not just in the sense of having privileged access to the signaler’s intentions but in the sense of having intended them in the first place – itself displays the difference between intending and whatever actual effects might follow from the achieved intent. Like the difference between the God who achieves all at once and mortals who do things sequentially, the difference between the temporal unfolding of outcomes and the atemporality of intention is for McEwan’s Godlike novelist precisely the difference between causes and effects on the one hand, and “the attempt that was all” on the other. And it’s easy to thread this back through the question of atonement that McEwan’s character raises: outcomes and the causes that bring them about are the sorts of things for which one might atone, but the logic of atonement does not apply for intentions, assimilable to neither cause nor effect.

The great puritan theologian, Jonathan Edwards, saw the difference between divine and human agency precisely in terms of the logical distinction between intentionality and causality.13 In his 1758 treatise, The Great Doctrine of Original Sin Defended, Edwards’s defense hinges on defining divine creation logically rather than phenomenologically and on distinguishing intended effects from actual ones.14 The larger argument of Edwards’s treatise, as we shall see, is an effort to justify holding human beings accountable for their sin while at the same time insisting on God’s absolute sovereignty and omniscience as their creator. The problem, as Edwards imagines it, is that Adam’s posterity is held responsible for something Adam himself actually did. Or as Edwards puts it, some might think that “imput[ing] Adam’s sin to his posterity…is unjust and unreasonable, inasmuch as Adam and his posterity are not one and the same” (220). Because we are not Adam, and because Adam is the one who sinned, the argument goes, conferring Adam’s guilt on us is “unjust and unreasonable,” and thus inconsistent with the intentions of a just God. Edwards refutes the claim by contradicting the very idea that “Adam and his posterity are not one and the same,” arguing that they are only as different as the “root of the tree” is from its “branches.” In short, the mistake is to focus on the parts without taking account of the whole. “God,” by contrast, according to Edwards, “looked on [Adam’s] posterity as being one with him. And though he dealt more immediately with Adam, it yet was as the head of the whole body, and the root of the whole tree; and in his proceedings with him, he dealt with all the branches, as if they had been then existing in their root” (220).

For Edwards, the way that humans make mistakes about the relationship between God’s creative agency and Adam’s sin, and between Adam’s sin and the sins of his posterity is by treating contingencies as necessities (and vice versa). The decisive mistake, according to Edwards, is not so much thinking that because our condition is a contingent effect of Adam’s sin (and his sin, in turn, a contingent effect of God’s creation) it would be unjust to impute Adam’s guilt to us. Rather, the problem has to do with treating the relevant relation between Adam’s sin (or further back, God’s creation) and our condition as causal in the first place. When Edwards writes that “it does not at all necessarily follow, that because there was sound, or light, or colour, or resistance, or gravity, or thought, or consciousness, or any other dependent thing the last moment, that therefore there shall be the like at the next,” he is emphasizing the degree to which nothing is guaranteed to follow from any given moment (224). And there’s an equally important corollary to that claim, as Edwards explains: “the present existence, either of this, or any other created substance, cannot be an effect of its past existence” (223). It is here that Edwards reveals the fundamental condition of discontinuity between cause and effect that characterizes his descriptions of nature and that for him, defines temporal existence: “The existences (so to speak) of an effect, or thing dependent, in different parts of space or duration, though ever so near one to another, do not at all coexist one with the other; and therefore are as truly different effects, as if those parts of space and duration were ever so far asunder” (223). The connection to the argument for original sin is now easier to see, though it only emerges in reverse. According to Edwards, to see Adam’s sin as a cause of our condition (and to see our guilt as unjustified), we also have to believe that Adam’s sin and our condition do not “coexist”: “The force of the reasons brought against imputing Adam’s sin to his posterity (if there be any force in them) lies in this, That Adam and his posterity are not one” (226).

For Edwards, however, Adam and his posterity do “coexist,” ontologically united in an expression of divine will: “the derivation or the evil disposition to Adam’s posterity, or rather, the co-existence of the evil disposition, implied in Adam’s first rebellion, in the root and branches, is a consequence of the union that the wise Author of the world has established between Adam and his posterity” (221). The “world” of Edwards’s divine “Author” is complete from the start. Rather than unfolding in a chain of causes and effects, as Edwards explains, “its existence in each successive moment, is altogether equivalent to an immediate production out of nothing” (224).

I’ve taken this long theological detour not to make a point about God, but rather about the logic of intentionality that common ideas about God and Vermeule’s deployment of them lay bare. Moreover, it’s a logic that’s explicit in Vermeule’s own efforts to bring our aesthetic fascination with social information into line with scientific research. My objection to that line of inquiry isn’t an objection to the research as such, although Ruth Leys’s work on the uses of neuroscience in affect theory make clear that there are serious problems in the research methodology of a number of recent and highly influential experiments, and that those problems extend precisely from a set of mistakes about intention.15 Rather it’s that if we are interested in giving a good account of the meaning of any work of art, then focusing strictly on what kinds of emotional or instinctual or bodily triggers move its maker, or on what kinds of responses the work in turn triggers in its receivers focuses us on inputs and outputs in a way that simply cannot compel our interest in any given work over any other given work – or for that matter, in art as opposed to anything else that might yield similar results. If what I care about is what my brain does when certain social information makes me anxious or filled with love or with loathing, it’s not clear why a novel of manners (or a poem or a painting) would be any more privileged object of study than a middle-school cohort. What does compel our interest in the work of art as opposed to the middle-school cohort is the set of concerns that make us care about the form it takes, the decisions the author made to make it work one way rather than another – in short, we care about the intention. To have an account of what our brains do in making the work or in responding to it is to have an account of the causes and effects of the work. To have an account of its author’s intentions – without which, we have no aesthetic interest in the work – is to have an account of something that cannot be assimilated to causes or effects.