What We Talk About When We Talk About Intention

Definitions of intention are slippery. Is it an act? A mental event? Do we detect it by logic, intuition, syntax? The question of what an author means—and how a reader knows—has a way of stalling us in the mud. In part, this is because it assumes more agreement than it earns. Answers tend to rest on unspoken commitments about how causes relate to effects. In literary studies, this tension surfaces most famously in Wimsatt and Beardsley’s The Intentional Fallacy, which treats intention as a kind of mental shadow behind the text—real, maybe, but unknowable. But why should we think of intention that way? What if, instead, intention is already in the text—not hidden behind it, but shaping what it is? This essay explores that possibility by turning to the philosophy of ordinary language—Wittgenstein, Anscombe, Cavell—and asking what kind of picture of intention emerges when we take language not as a system of signs, but as something people do.

Definitions of intention are complicated: is it an act, a mental event? How do we discern it—deduction, induction, semantics? Why does it matter—ethics, necessity, parsimony?" Also, the question, how a reader knows what an author means, has a way of committing the intentionalist to alien premises in answering. It assumes definitional agreement. This has happened at the level of academe. In the paradigm critique of intention, The Intentional Fallacy, William Wimsatt and Monroe Beardsley present intention as “the designing intellect,” distinguishing it from the text; to grant the “cause of a poem is not to grant [it] as a standard.” (Wimsatt and Beardsley 469) This idea of intention, as Kay Mitchell notes, is “a Cartesian account,” (Mitchell 18) casting it as a “mental objective” (Close qtd in Mitchell 17) hidden from public view.

Ferdinand de Suassure's semiotics emerged as a way of theorizing this model, and has been assumed in much of what counts as literary theory. In it, Saussure developed the idea of sign as signifier united to the signified, a mental concept connected to a “sound-image” (or graphic). Signifiers are material, psychological imprints on our senses. (Saussure 66) Through social tradition we inherit thoughts and language simultaneously, both emerging as two sides of the same sheet of paper. (113) Signs, then, are essential to language, material, and arbitrary with respect to what they signify. This precludes authors from the question of meaning because it locates it, not in the act of communication, but in a system of significance. Of course, positive forms of this argument have been complicated and challenged by more than a half century of theorists. But, as Moi reminds in Revolutions, it has been judged by its own criteria such that the failure of representation becomes a pretext for structuralist and post-structuralist readings. Representation is assumed in its failure, announcements basic to dominant theories of metafiction.

But this makes intentions invisible as a matter of definition. A closer look at Fallacy shows this is motivated by a deeper commitment to mental causes as disconnected, that is, indescribable, from their effects. It’s this commitment that Walter Benn Michaels takes as explaining “now longstanding debates about whether the intended meaning of a work is the…object of interpretation...” (Michaels 1) In Wimsatt and Beardsley’s picture, “the poet only moves her fingers; the rest is up to the syntactic and semantic rules of language.” (7) 

And this is how theories of action underly theories of reading. Michaels finds Donald Davidson’s account of causation as the exemplar of Wimsatt and Beardsley’s assumption. He explains, “…Davidson’s description of braking a car as ‘pressing a pedal’ and thus causing the ‘automobile to come to a stop’; his point is that we can assign ‘responsibility’ to the man braking not by ‘transfer[ring]] agency from one event to another’ (the movement of the foot would be one event, the car stopping another) or ‘by saddling the agent with a new action’ (first he pressed the pedal, then he stopped the car) but ‘by pointing out that his original action had those results’ (59). Following Anscombe, by contrast, we would say that the action the man performed was not pressing a pedal (with the consequence that the car braked) but braking.” (Davidson 59 qtd in Michaels 4)

So, if behind representation are deeper commitments about causes and effects—ideas about what we can know of causes in their effects—then a definition of intention needs a picture of causes that are somehow present in effects; the movement of the poet’s fingers and the “rules of language” must be part of a larger thing we recognize in reading. What could such a thing be? This question will guide my inquiry into intention. A rough start is to notice how Wimsatt and Beardsley’s causes (call it WB causation) make us talk. Texts become agents of communication, while writers become objects, like a que hitting a billiard ball. This manner of speaking is so common we don’t notice its special use. If you called both Moby Dick and a grandfather clock beautiful, I would know what you meant. But you’d have to explain if you called the grandfather clock Marxist. Is not Marxism a concept, a thought held by a subject? My point is not that a grandfather clock can’t be Marxist—say it was made in a socialist commune or had Raymond Williams quotations carved into its panels. But then we would say it’s Marxist in a certain respect, under certain forms of description. Thus, it becomes interesting to ask, what gives form to our descriptions? For the clock, it’s Marxist as a dedication to progressive scholarship, or as a product of the Marxist vision of work. In these examples, to define the clock as Marxist is not to find a formal definition under which clocks become Marxist in all cases. It is to pick out examples, distinct pictures of human activity containing both causes—communes, woodcarvers—and effects—clocks. 

Returning then to the strangeness of literary talk—from this, we can see how giving rules that hold for language in all cases is like giving a formal definition for Marxist clocks. Definitions, when separated from use say very little. They are like, as Ludwig Wittgenstein reminds, “a cog that has been removed from its proper place and is now turning idly in the air.” (Wittgenstein §192) Definitions of language—whether on the level of signs or forms (like metafiction)—are no different. What a text means to a reader can’t derive from its form—or from arguments about its failure to be a form—because definitions only emerge from instances of use. They can only be a picture of what we mean to do with words at a time and place. 

This is why Michaels argues in Intentionalism Again, that the question of what a “vehicle” is in an antiquated law about vehicle use “is an empirical question having nothing to do with the theory of language.” (Michaels 1989, 91) If we think of language abstractly in terms of general rules, “we are led to realize how much depends (‘everything’) on the ‘level of generality we choose.’” (93) This comes out when we imagine a definition like, “every word in the language signifies something.” Such a theory would have “so far said nothing whatever; unless we explain what distinction we wish to make.” (Wittgenstein §13) Say we make a definition for what a tool is: they all “serve to modify something.” We end up with useless classifications like, “a hammer modifies the position of a nail, a saw the shape of a board.” Wittgenstein asks, “would anything be gained by the assimilation of these expressions?” (§14) This is how, in place of definitions, intentionalist are drawn to examples of use, the descriptions of which, in the case of reading, must contain both writerly causes and their literary effects. 

But what replaces post-Saussurean reading? Answering requires qualification. Intentionalists shouldn’t forsake representation or semantic analysis. Instead, they should reclassify representation as something we do with words. But from where do we derive meaning? What does Michaels’ “empirical” process look like? The Natural Language tradition calls our attention to “forms of life.” (Wittgenstein §241) These are the modes of human activity of which language plays a part. Linguistic actions in these forms are organized by language games — “giving orders, asking questions…eating, drinking, playing.” (§25) These are as diverse as human activity itself and are structured ultimately by physiological, spatial, and social constraints, what Wittgenstein calls grammar. These govern the activities in which words are a part of acting. 

In this sketch, the intentionalist glimpses a language model for which representation is not fundamental, transforming the project of reading. Henry Cavell presents this transformation by detailing a series of commitments. These include, 

that language is shared, that the forms of life I rely upon in making sense are human forms, that they impose human limits upon me, that when I say what we ‘can’ and ‘cannot’ say I am indeed voicing necessities that others recognize, i.e. obey (consciously or not); and that our uses of language are pervasively, almost unimaginably, systematic. (qtd from Moi 52) 

Grammar limits language and forms of life limit grammar; both are known in specific language use cases. Once an outline of this use is made, the intentionalist has grounds for criteria, the beginnings of an interpretation. This method is extended by Wittgenstein’s student and interpreter, Elizabeth Anscombe to determine the use of the word “intention.” Her book Intention is helpful for two reasons. It demonstrates the Natural Language process for determining the meaning of a word and gives the intentionalist theoretical means to justify their claims. 

Anscombe identifies “intention” as a “form of description” we use to pick out certain practices and events. She identifies this use, showing how the word succeeds or fails in practice. She identifies two forms relevant to the interpretation of literature. The first is what David Schwenkler calls “forward looking,” (Schwenkler 63) intention, taking the form of doing P with a view to Q. I express an intention when I say I am crossing the street to pet a dog. But the connection between P and Q is not predictive. I could cross the street for many reasons. Anscombe wonders how to describe this connection. She does so with an appeal to the form of “outward looking” intentions. Actions under this description are explained as intentional by effects which are simultaneous with, inherent to, the act itself. Anscombe says, “if someone comes into my room, sees me lying on the floor and asks, ‘what are you doing?’ …an answer like resting or doing yoga…would be a description of what I am doing.” This would be “an expression of intention.” (Anscombe §22, 34.:3-35:1) In this way, Anscombe concludes that forms of description can be “swallowed up” by successive forms. The intention to do yoga just is an explanation for lying down, such that lying down, in this case, is doing yoga. The connection between P and Q is not predictive but transitive. 

This helps intentionalist to see texts as actions and so to reconsider the text-world dichotomy such that interpretation becomes a description of what the author meant, or, to reverse Wimsatt and Beardsley, descriptions of causes (writing) contained by their effects (a novel, et cetera). Outward looking intentions are the form of a description in which actions are described in terms of what happens. This is made clearer when Anscombe turns to complex examples. She imagines an assassin killing members of an evil government. He does so by pumping water into a cistern in which he has placed poison. The cistern feeds to the house where the government officials are staying. She notes that there are many descriptions of the assassin’s action—moving his hands up and down, pumping, feeding a cistern, killing. But she notices that “each description is introduced as dependent on the previous one,” (§ 26,45:6) such that moving one’s hands up and down just is assassinating. That which formally contains these movements is the practical knowledge of the actor, intention being the form of description containing the movements in one act. It’s the unity of something done as a means to an end. 

The intentionalist finds a home in Natural Language philosophy at precisely this point. Anscombe frames intention as a descriptive act, governed by language-game grammar and use. The forms of life in which this act of description is appropriate—think Cavell’s “necessities that others recognize”—are those in which actions are described by their effects when they are done with an end in mind. Literary interpretation is empirical to the extent we understand words as actions describable only under certain conditions. By rejecting representation and considering linguistic meaning within the acts of which they are a part, the intentionalist can fruitfully argue that texts are just the kinds of effects which “swallow up” their causes: the writer’s intention in writing.