Interchange I: On intention, context, and meaning, with examples

This begins with Henry Staten (HS) expanding on an example he brought in when commenting on a draft of Walter Benn Michaels’s (WM) piece.

HS had written: Consider: I say something ironically, but say it in a serious tone, and am completely convincing in my seriousness, and nothing ever happens to make anyone think I wasn’t serious. I smile and say to a student, “you’re the smartest student in the class,” when I think he’s a moron, and I clasp his shoulder affectionately and give him an A on his paper; and I privately enjoy having bamboozled this moron in this way. What’s the status of the private irony? This is a curious fact about my psychological weirdness, not anything having to do with “meaning.”

HS: Normally, if we mean an utterance ironically, we give an outward sign that this is the case—either by saying it in an ironic tone, or by a facial expression, or some other sign, perhaps not to the person I’m addressing but to someone else who is present, a sly sideways look, for example, that the addressee doesn’t see. But suppose in this case we say it, and act, in a way that is exactly how we would do it if we meant it seriously. Something has happened “in my head” here, something that we can definitely call “merely mental,” which is completely divorced from what I propose we call the scene of utterance. Have I successfully meant “you’re so smart” ironically? I meant it that way “in my head,” but did the actual utterance in its context mean that?

Now suppose immediately after I say this the addressee walks out the door, and within seconds I have a heart attack and die. And suppose further that I’m a world-famous philosopher, and my addressee had a voice recorder in his pocket because he wanted a record of having talked personally with me, and he takes it home and plays what I said to him for his friends, and they congratulate him, and he keeps the recording permanently and treasures it the rest of his life, and passes it down to his heirs, who also treasure it for generation after generation. What is the relation of my conscious sense that I meant it ironically to the worldly reception of my words? Did I successfully mean it ironically, regardless of the fact that no one ever would be able to guess this, and despite the fact that I executed the speech-act of its utterance in a way that checks off all the boxes of what Austin considered a happy speech-act?

WM: Answer to the last question—yes. A happy misinterpretation!

HS: That means you subscribe to a purely mentalistic theory of meaning.

WM: Is the difference between, say, something done accidentally and something done on purpose problematically mental in your view? Nothing is done at all in anyone’s head—but the question of the purpose of any action is not reducible to what other people think it is. And, if it were, they’d still be deploying the concept of purpose. As they do in your example where everybody takes the sentence as an expression of the speaker’s intention. That’s why the guy’s so happy! Your example doesn’t abandon intention for a millisecond. It just substitutes what’s plausible for them to believe about what the speaker meant for what he actually meant.

HS (Self-cite): I found the following in some old screed of mine, commenting on Iseminger’s article in his important collection Intention and Interpretation, and I thought it was relevant to the current debate. Iseminger writes, “what would make it the case … would be some fact about … that person’s thoughts … at the time.”1 (Car ran out of gas. But compare: my car is out of gas, and I want a ride to the gas station, but I’m a literary theorist and when someone picks me up, I concentrate on meaning “my car emerged from a cloud of argon” when I say “my cloud ran out of gas.” Now what follows? Say the other guy says, “I’ll take you down the road to the next service station.” I could then reply, “Oh, I meant …,” in which case puzzlement ensues. Or I say, yes, I do need gas also, but that wasn’t what I meant, I meant ….” Again, puzzlement ensues; I will now have to explain that I’m a literary theorist who wanted to prove …)

What this shows is that the ontological mode of positing theses about meaning collapses the complexity of the scene of meaning in a way that falsifies rather than illuminates it. I can mentally “mean” whatever I like when I say these words, but it is only to the degree that this mental act engages the scene of meaning that it is connected with the language game with the word “meaning” as this language game has evolved outside of the language game of theoretical speculation. There is, as Wittgenstein says, a before and an after of the punctual use of language, and this before and after, and not the punctual mental act accompanying the utterance, are constitutive of meaning.

HS: One last thing. I didn’t answer your question about the difference between something done on purpose and something done accidentally. I think it depends on the particular occasion. It might have everything to do with conscious intention, and it might require a more nuanced analysis, but in a standard case, it would require a reference to conscious volition somewhere along the line. For example, I might work on an assembly line and do the same operation many times per day, then screw it up one time. It might by this time have become completely automatic so that I spend my time daydreaming while I work; so nothing need have gone on in my consciousness (more accurate than ‘in my head’) when I screw it up. But I’m mortified that I did it, and this is because it’s my standing intention to do it right and lose my job. Further, when I was learning to do it, I put a great deal of conscious intention into the learning. In other cases, conscious volition might be operating the whole time. The invocation of the mental isn’t something I categorically reject; it’s just not the be-all and end-all that traditional mentalism takes it to be. What’s your analysis?

WM: The question of relative degrees of consciousness seems to me beside the theoretical point—if it’s done on purpose, you can’t explain what was done without recourse to the category of intention. So, insofar as we understand every text to be the product of an act, understanding the text involves understanding the act—not understanding some distinct mental event going on inside someone’s head, but understanding what someone was doing. When I raise my arm to reach for a dish on a high shelf, my intention is not some additional or prior act taking place inside my head (maybe I’m thinking about it, maybe I’m not—doesn’t matter). And it’s not something that can be added to (or subtracted from) the motion of my arm (Wittgenstein’s famous point). It’s built into the description of what I’m doing by moving my arm, the answer I can give to the question why. Although you say this involves question begging, it’s just as essential to your own analysis as it is to mine.

And the whole attempt to imagine the text as if no one had produced it—an attempt that’s foundational for Derrida since he regards it as built in even when of course someone did produce it—can’t help but position what the author was doing as not the explanation of the text but the cause of it—which is why he and you end up locating the intention outside the text. This is Derrida’s mistake. What he understands as radical about his account of the speech act (the necessary displacement of the “governing” intention) is in fact what’s most traditional about it—the location of the intention in a “place” from which it seeks to govern but, of course, can’t. (Obviously, it’s not the fact that it’s traditional that’s the problem.) In other words, once your options are governing/not governing, you’re screwed.

Joshua Kates (JK): I’m with you on all of this; I don’t think consciousness has anything to do with it, and I don’t think that you think consciousness has anything to do with it. But when you speak of description a little later in your second to last remark, I think things become more difficult for you. I, Anscombe, Heidegger, others—and Henry’s reference to Wittgenstein fits here—claim such descriptions arise in a context, without which 1) the doer could not hold the intention they had—in which they could not describe themselves as doing x or y; and 2) that context also includes, to varying degree, others who can also describe the doer as having such an intention without any explicit regard to the doer themselves. You seem uncomfortable with this possibility and this stipulation, unless I am mistaken.

I also don’t get what you are doing here with Derrida. The reason Derrida does not think intentions are governing is because he thinks the condition of having a (linguistic or semiotic) one entails a form of repeatability that the intention cannot fully master. Yet there also cannot be any text-y things, written or spoken or semiotic, without intention playing some role. Moreover, as an interpreter, Derrida wants both: he wants to interpret faithfully the intention, as Henry pointed out, and read what is not available as an intention but inscribed in a text (which reading also produces). Not one or the other, but BOTH. Whether this is finally viable, to some extent, is always only ascertainable in the readings—sometimes it works, in Derrida’s reading of Rousseau, sometimes not so much, in his reading of Benjamin, for example. Why Derrida wants this and what, then, is a text for him is another question … but I don’t think his project can be jammed into the causal-non-intentional box you fashion for it.

WM: Yeah, 2 (which I think is false) in no way follows from 1 (which is obviously true and which no one denies or probably has ever denied).

HS: Walter, I’m not sure what you mean when you say that 2 doesn’t follow from 1, I mean where that leaves you with what you’re agreeing and what you’re disagreeing with.

WM: So I think 1) is uncontroversial and the part of 2) that doesn’t follow from 1 is “without any explicit regard to the doer themselves.” Actually, now that I look at it again, I’m not quite sure what it means. An account of what the doer did necessarily involves regard to the doer (not sure that “explicit” matters?). And any account of what a text means just is an account of what the doer did. Which is just to say that the context the reader is in may well have an impact on what the reader understands or misunderstands the text to mean but has no impact at all on what it actually means.

This is one of the reasons I’m leery of all the context talk. Back in the day, I was a big context-is-everything guy, and in fact, when I would first teach these texts and understood myself to be in agreement with them, Derrida’s insistence on context was crucial. But as I started to think that all context could usefully mean was stuff that influenced your account of what the speaker meant and that “not saturable” could only coherently be an epistemological point, I stopped thinking that the assertion that context mattered was saying anything helpful.

HS: Good stuff, thanks for the clarification.

JK: On “without any explicit regard to the doer themselves,” how about this? The three of us, Josh, Walter, and Henry, are continuing our discussion in the backroom of a dive bar where it is hot. In the middle of making a point, Henry gets up and walks to a window that only he is positioned to see and opens it. I say to Henry, yes, I was hot too. Henry, however, was thinking about his point concerning Wittgenstein the whole time and never thought anything about the temperature of the room. Nevertheless, following Anscombe we can say Henry intended to cool down the room; he opened the window to make the room cooler. Of course, he might say, “oh, yeah” in response to me, or he might just keep talking about Wittgenstein—either way, however, we have his action under a certain description, which gives us his intention.

Reading literature—or again, really anything—Hobbes, Plato, the constitution—is a little like that, by my lights. This part of the text appears there for a purpose, which means we view it as part of the author’s intention—it’s an intentional object in this sense. Sometimes we find other parts of a text that make us think we got the first description wrong or that we need to tweak it somehow. Though in some cases, possibly Plato or Spinoza—we might think the contradiction or conflict is on purpose. The difference between most texts and other long discourses and Henry’s opening the window is that no one can account for all of the former and so different descriptions, from different readers, at different times may proliferate, though not without bounds or without overlap, and all are descriptions of the author’s intention.

WM: What was Henry doing? He was opening the window. Why was he doing it? Because he was hot and intended to make the room cooler. Does it matter what he was thinking of while he was doing it? No. So maybe we all agree on this (although I can’t for the life of me see how this is without any regard to the doer). But if we do agree on this, and we do see it as the same kind of thing involved in understanding literary texts, you guys should just start calling yourselves intentionalists like me. And if your idea of literary interpretation is that it’s descriptions of the author’s intention—some of which are compatible because we’re describing different aspects of that intention (so there’s no conflict), and some of which (insofar as we’re differing about what she was trying to do) are incompatible (so they can’t both be true—no pluralism allowed), I don’t see how a single syllable of that isn’t just straight Against Theory.

HS: I agree with Josh’s 2, in his reply to this.

Additionally, I remark that the action he describes is validly analyzed in terms of cause and effect. My body felt uncomfortably hot, and this caused me unthinkingly to get up and open the window. We say I opened it intentionally because it was no accident, but if one respects the nuances of the grammar of use, it’s clear that this use of “intention” is different from those in which there isn’t this kind of strong physical element and no conscious thought. (Although it seems to me that we do have to introduce the notion of degrees of consciousness here because I had to be in some sense conscious of what I was doing, “subliminally,” one might say, although not in my focal consciousness. I really do think that Husserl would be relevant here, especially the Logical Investigations, but also Ideas I, where he does such a profound analysis of the structure of intentionality. And that reminds me: Josh, when you use “intentional object” to mean “my purpose, the aim of my speech or action,” you’re talking in a very non-Husserlian way). You, Walter, want to reduce all intentional action to one simple structure, but this is just to sweep a whole boatload of necessary distinctions under the rug.

As far as what I call myself, I don’t actually call myself an intentionalist because I think that’s a term that can only create confusion, as I explained in my Syndicate response to Walter (also why I avoid talking about meaning).2 I think these terms automatically evoke traditional mentalism, and there’s no point in pretending we can just re-define them and throw them out there. I do speak of intention, but I’m careful to call it “techne-intention.” So, in my account, intention is mediated by the know-how that has been wired into the brain-body by culture; whereas Walter wants intentionality to function in a way that needs no mediation. Derrida, of course, sees language as the differantial medium of thought itself, “in the head.” The first pages of his discussion of Rousseau in Grammatology are the best place to see how he thinks about this. (There’s important new work on Peirce that is adding new dimensions to how I think about this.) One of these technai is of course language, which is considerably more than a mere code. I just re-read Culler’s pp. 18–19 in Structuralist Poetics in which he talks about Saussure and Mounin on the sign, which nicely states why language is much harder to pin down than code.3 The basic thing that I find relevant here is that the sign is defined by Saussure as “the combination of a concept and an acoustic image.” The dicey bit is the concept. Walter’s notion, as near as I can figure out, is that when I use a chain of signifiers, they automatically come married to their accompanying concepts. But 1) a concept, in Saussure’s conception, is a mental something, so I need Walter to define the structure of the sign in his non-Saussurean use of it, such that concepts/ideas play no role in the definition and 2) since Saussure’s definition (which he expands by comparing the relation of sign and signified to the recto and verso of a sheet of paper) implies that, as Culler says, “for every signifier there’s a particular positive concept hidden behind it.” Mounin then runs with this definition by talking about “decoding language.” As Culler then comments, “this is a version of Derrida’s ‘metaphysics of presence’ which longs for a truth behind every sign: a moment of original plenitude when form and meaning were simultaneously present to consciousness and not to be distinguished …. It is assumed that we should still try to pass through the signifier to the meaning that is the truth and origin of the sign and of which the signifier is but the visible mark, the outer shell.”

So, my question, Walter, is how you define the sign in a way that isn’t subject to this response.

Also, I’m a bit puzzled by your refusal to allow the mental dimension any place in your analysis. To me, it looks like you’re so avid for immediacy that you won’t even brook such a mediation as thought itself—maybe because you’ve taken to heart Derrida’s argument that thought itself is already riven by differance.

JK: a) Obviously, I am not Walter, but I think he and I don’t invoke the mental precisely to avoid the issues pertaining to consciousness that you yourself raise. That is not to say that different degrees of awareness (consciousness, if you want) can play a role in different situations, but the main point, as you say, is that meaning is not the effect of some ghostly operation, some mental event, again, neither for me nor, I think, for Walter (though in his essay on Anscombe I think he inflects her work a bit in this direction). I also don’t think Walter is interested in signs, except as they appear in use by a speaker or author, where they do not play a regulative role—sort of like Grice. (It would take too long now, but maybe I’ll come back to it, but I read Saussure very differently from you and Culler, if I am following. The concept/signified is not the difficult part—as you say, it’s always linked to a signifier, and is also located in the collective repository of a language synchronically and collectively understood—so there is no mentalism; it’s with the signifier and its status that the difficulty for Derrida and in a different way Walter lies—and yet in a different way still for me.)

b) You write: “‘I’m careful to call it ‘techne-intention.’ So in my account, intention is mediated by the know-how that has been wired into the brain-body by culture.” This is also a long discussion, but I would point out that Anscombe distances herself from techne when discussing Aristotle’s practical syllogism (62–63 of her Intention). She does this because technai, she argues, only get us general rules, hence nothing that can be active in a context, in a situation, where intention alone emerges. We might have conditionals: if we want a low-acid wine, pick the grapes early. Yet for Aristotle and her, you also need someone wanting something on an occasion (here a low-acid wine) to have praxis imbued with intention. This is important, I think, and I alluded to this when writing to Walter since it is difficult to identify the actions and expressions in use with what they are when not being used (see Charles Travis’s and others’ occasionalism). I take this to be Wittgenstein’s point in paragraph 48ff of Philosophical Investigations (with the colored squares) when discussing names. Names only name something when used to talk about something (in some language game); without use, they do not actually name. (Are there in RRBGGGRWW, where each names a colored square in a 9×9 grid, 9 names or 4? Witt says it doesn’t matter as long as there is no confusion when playing the game, but this is a different case from the name in use.)

P.S. My last point is probably better and certainly more succinctly put by noting the following. Rugs were once made by hand; you can get a machine to make rugs, transform the techne; but you can’t get a machine to want to make one, nor then can you get it to intend to make one.

HS: If I get into your remarks, which raise a whole BOATLOAD of new problems, I would start to sink into an endless debate, so I’m going to try to ignore most of it, except to say that Walter does have recourse early in his piece to the notion of the signified’s being tied to the signifier, which means concept to sound, and all I’m asking for from him is that he gives me his non-mentalist definition of “concept.” As far as Anscombe and intention, if you guys spent as much time studying Wittgenstein as you’ve spent following Anscombe, you’d be much better off. But in any case, I’ve developed the notion of a techne far beyond anything that she imagines, so what she thinks about techne is irrelevant.

HS: Walt, a simple yes or no question, to which I will only accept a simple yes or no answer. Can I say “blah blah blah” and mean “the wind sprang up at 4 o’clock?”

JK: Let’s say I have been playing a game, with you, or perhaps better, a child. Each of us has said “the wind sprang up at 4 o’clock” ten times in turn. If at that point I say “blah blah blah,” it could well mean “the wind sprang up at 4 o’clock,” and under the right circumstances, it might again mean that in the future.

HS: That’s not my question. All you’ve done is posit a case in which “blah blah” has become coded for a particular group, whereas what I’m asking is whether my intention alone can make it mean that.

JK: Well, I don’t see it that way. Everything we say and write comes from exposure to what other individuals—and to this degree, groups—have said and done. In the case of “blah blah blah,” we have ways of using this phrase, though I gave you a way it could be used. (Sort of like Derrida’s citationality all the way down.) Many other unprecedented uses can be envisioned—”a full blast” when we have a full-suited deck of cards [i.e., 4] and so on. No one is saying, certainly not me, but I don’t think Walter, that there are intentions without expressions. The question is the status of the latter: are they codes, linguistic systems and so on, all of which to me seem to lead back to ideality of some stripe. (Previously I tried to frame a version of this question for you in terms of a discussion of Wittgenstein’s in PI: I don’t know if you missed that, did not want to respond, or had none.)

WM: Trying to stay out of this for now but quick observation—the example is, as Henry presents it and in his terms, a mentalist fantasy of a speech act. Anscombe talks about someone trying to move some object without touching it by concentrating on it real hard and willing it to move and then points out that you also can’t move your arm by concentrating on it real hard and willing it to move. The speaker in Henry’s example isn’t meaning anything because she isn’t even trying to mean something.

HS: What would constitute actually trying to mean it? Here’s another example. When I get migraines, it completely scrambles my verbal faculty, so that I’m not only unable to utter actual words (it just comes out a jumble of sounds) or even to THINK the words; yet I know just what I meant by the inarticulate jumble. On your account, I take it, that would mean that that’s what the jumble meant. Is that right?

JK: I am not sure at this point what the larger argument is (especially if you, Henry, are now suggesting that you can know what you mean without any words? Or am I mistaking you?)

To me, it would be helpful to take a step back:

a) You, Walter, want to say a text is an intention and to explain how that can be you draw on Anscombe’s account of intention as action under a description, here presumably the description being what the author intended. (As an action it requires words, terms or something, I believe that are expressed/used, no?)

b) You, Henry, take a text to be some sort of physical expressions or symbols, meaning what they do owing to the employment of a techne, thus allowing others, not just the author’s understanding to constitute meaning.

c) For what it’s worth, I myself take a text to be an event, wherein receivers establish the descriptions that allow for identifying what an author intended; these descriptions depend upon context, which includes worldly things and other things that have been said or written (which explains the role played by tokens or words without any appeal to types or codes).

In the case of Henry, I don’t see how what is found in techne relates to what is said or written without abandoning reference to things in the world. Such reference demands use, action. You will say techne allows this because it’s a learned practice, but I don’t see how the symbols or tokens in use and those in the techne can be mapped onto one another. The endeavors I know to map a speech act back onto a code or even a practice generally conceived all fail. (I thought that was part of being a Wittgensteinian—disallowing such a translation from use to anything else, to anything like what Peirce calls thirdness.)

In the case of Walter, almost the reverse: you stress intention and action but embrace some kind of (linguistic) conventions, no? Though for you these only are what they are in use, you give them some footing of their own, since you talk about them in their own right. I don’t see how you do that unless you accept intentionless codes, languages, signifiers, and so on.

HS: Josh, you move too fast into too deep waters for me, especially at this point in our discussion. I’m trying to posit simply instances in which precise problems arise and trying to get answers from Walter—and you, if it were possible—that stick to the immediate example without deploying a whole theory around it. So, in this case, the answer is yes, I know perfectly well what I want to say, but it won’t come out my mouth—or in the form of words in my head either. This isn’t a theory, it’s an experience. Same with my physical movements. I want to zip up my jacket, but my hands bumble around and can’t do it. I can’t command my actions any better than my words. I suppose this might give Walter grist for his mill since it will suggest to him support for his notion that meaning is just an action like any other, but that’s my experience.

As far as your formulation of what I think a text is, I think I can accept your brief statement of it, but not your statement that I allow both sides to “constitute meaning,” because I deny that there’s any such thing as meaning (and no, I don’t mean a physical thing, I mean whatever self-identical whatever it is that mentalists and Walter in his own, purportedly non-mentalist way, think meaning is). As I keep saying, meaning is a word-concept that we deploy in a variety of ways for a variety of uses. Very simply, I adhere to Wittgenstein’s “meaning is use.”

WM: Side note (exasperated)—meaning is whatever, when you say I misunderstand SEC, you think I’m misunderstanding. Which is how we use meaning all the time, and when recourse to “meaning as use” means we can’t use it for whatever’s being misunderstood, it’s lost its point. Follow up (not exasperated at all)—the jacket thing is of course grist for my mill. You can fail to mean just like you can fail to do anything else. And you can’t button your jacket either in your head or without intending to. And the fact that you can always fail to mean does not mean that you must always fail to mean (Derrida) or that when you do fail to mean, the text you produced somehow gets to mean something else.

HS: Whoa! “You can fail to mean just like you can fail to do anything else.” I distinctly remember your saying earlier in our discussion that you cannot fail to mean. Did I dream it? You raise for me all sorts of interesting “grammatical” questions. You think the grammar of “mean” is the same as that of “do,” but don’t you also treat “intend” as having the same grammar? “I intended x” is analogous to “I meant x,” or even synonymous with it, no? But is intending also a form of doing? Isn’t there all the difference in the world between intending x and doing x? So then, either the grammar of “mean” is not analogous to that of “do,” or it isn’t analogous to that of “intend.”

Here’s more: if I mean x, do I have to know that I mean x? Can I mean x without knowing that I do? Without combing back through SEC, I’m pretty sure Derrida does not exactly say that “you always fail to mean.” What he says is directed against the metaphysics of presence (something that you continually elide, an elision that makes all discussion of SEC extremely limited and misleading), and his point is that meaning is never a “plenum,” an “absolute fullness.” I understand that you’re not interested in the metaphysical questions, but in all fairness, you should grant that that is, after all, his real target and that your arguments are tangential to his concerns. I think there’s a real value in what you do, and I think this value would come out much more effectively if you didn’t waste your energy convicting Derrida of saying things that he doesn’t actually say in the sense that you take them. I think The Purveyor of Truth says it in the most concise and quotable way: he’s writing against Lacan’s “a letter ALWAYS reaches its destination,” and he doesn’t say in reply that it NEVER reaches its destination—he says even if it arrives it continues to be tormented by the internal possibility of its non-arrival. I see this as a bit of darkly French over-subtlety, but very cool, nevertheless, as another way of saying that the sign as long as it persists can disseminate further, like a seed that lies dormant for a thousand years before sprouting. What I don’t think you give him credit for is his central concern with the productive nature of the sign. On the other hand, it’s true that his work had a huge negative influence in the way that it was received, as though he were actually saying that everything is language and that therefore everything is interpretation and that we can interpret things in whatever way we or our “interpretive community” decide to interpret them, with no regard for what they were intended to mean. That’s why I’m basically in accord with what you do, Walter, because I loathe that stuff. And although Jacques did try to some degree to correct this misreading, he didn’t try nearly hard enough. The simple minds of humanity can only encompass two polar antitheses: either meaning is a metaphysical presence/fullness, or there’s no holds barred.

As far as your first sentence, I completely agree. Meaning is what we call what we try to understand in a text like SEC, a discursive text that is positing theses. But I find it misleading to say that meaning is what we’re looking for when we read a poem. Brooks says reading a poem is an experience, like experiencing dance or music, and I find that a much richer notion.

JK: If this is supposed to be me, Walter, it’s not my position: “when recourse to ‘meaning as use’ means we can’t use it for whatever’s being misunderstood, it’s lost its point.”

(In practice, I don’t think it’s any of our positions, which is why we keep going round and around.)

HS: What Josh said.


1. Gary Iseminger, “An Interpretational Demonstration?,” in Intention and Interpretation, ed. Gary Iseminger (Philadelphia: Temple University Press, 1992), 76–96.

2. See Staten’s reply to Michaels here: Henry Staten, “Techne Theory,” Syndicate, November 22, 2021,

3. Jonathan Culler, Structuralist Poetics: Structuralism, Linguistics and the Study of Literature (Ithaca, NY: Cornell University Press, 1975).