Anscombe's Intention and Related Issues

Interview by Richard Marshall

'Now, one of the things that made Hare such an important target for Anscombe was that she found his moral philosophy to be totally perverse. The issue here was not over support for the bombings of Hiroshima and Nagasaki (I have no idea about Hare's position on that), nor over whether he supported giving the degree to Truman (though I'm almost certain he did: only a tiny number of Oxford faculty supported her motion to oppose it). It's rather that his position was explicitly consequentialist...'

'The pamphlet she published in 1938 with her friend Norman Daniel, which opposed Britain's entry into the Second World War, cites Aquinas directly and presents arguments right out of the Summa Theologiae. But in writing Intention Anscombe kept that influence largely beneath the surface: while she's channeling Thomistic ideas consistently throughout the text, Thomas gets referenced only in a stray footnote. I think this is another main cause of the initial obscurity of the book.'

'...perception is integral to action, and essential to our knowledge of what we intentionally do, but not because it provides evidence that justifies our beliefs about what we are doing; and that at least in some cases it is conscious visual awareness, and not merely nonconscious visual representations, that plays a role in shaping action as it unfolds.'

'... it might be impossible to resolve Molyneux's question experimentally after all, since anyone who can see well enough for the experiment to be carried out on them would already have had the opportunity to form so many cross-modal associations that the result of the experiment would be meaningless. Despite the initial appearances, it might be a question that we can answer only on philosophical grounds.'

'If there's one thread running through most of my published philosophical work, it's that introspection isn't good for shit.'

'... the only form of self-knowledge that grounds the possibility of honest assertion of our beliefs is a knowledge that isn't through evidence at all -- perhaps something like the "rationalist" self-knowledge that I described above in summarizing Moran's position.'

John Schwenkler is interested in Philosophy of mind (perception, action, self-consciousness), epistemology (self-knowledge, agent's knowledge, testimony, disagreement), philosophy and cognitive science and is the editor of the Brains blog. Here he discusses the hazards of reading Anscombe's Intention, Anscombe and Aquinas, the distinction between foresight and intention, whether his defense of the strong cognitivism thesis ‘intending is believing’ marries with Anscombe’s position,  Anscombe’s distinction between knowing our own actions and knowing of another’s, implications of agential knowledge being practical knowledge, perception and body knowledge, whether visual space awareness requires the visual awareness of space, matching feelings with seeings, perspectival seeing, self location vs minimal view, the limits of self-knowledge and introspection, mirrored self-recognition, and finally assertive honesty and non-empirical knowledge.

3:16: What made you become a philosopher?

John Schwenkler: This one was easy. In my first semester as a college freshman at the Catholic University of America I took an introductory philosophy course with Brian Shanley, a Dominican priest who went on to serve as president of Providence College. I was hooked right away by the special combination of rigor and open-endedness on display in philosophical writing. My education in those undergraduate years was overwhelmingly focused on the history of philosophy -- I mean, one semester I took a "Contemporary Philosophy" course where the core readings were from Peirce, Frege, Heidegger, and Wittgenstein! But this turned out to be a great foundation as my interests developed over the years, and also kept me mindful of how the possibilities for philosophy can be wider than is taken for granted in many of our contemporary conversations.

3:16:  Let’s start with Anscombe’s Intention which is one of those philosophy texts that strike many people as obscure as it is brilliant. Your approach has been to assess it on its own terms by making explicit who she was arguing with and against, and by making the importance of the context of when she wrote it clear too. So can you sketch for us what are the hazards of reading Anscombe without approaching it as you do, and then say what was the context – which involves Harry Truman and the nuclear bombings of Hiroshima and Nagasaki doesn’t it ?

JS: Right, so there are lots of things that make Intention so brutally hard to read, but surely one of the main ones is that Anscombe does absolutely nothing to set up the context for her discussion. My sense is that this, together with her near-total lack of reference to any contemporary interlocutors, is to some extent a reflection of how philosophy was done in Oxford at the time: philosophers there were writing for one another, and the audience was presumed to know already the ideas that were in the air. So one of the things I do in my book on Intention is to work on filling in some of these gaps. Thus I argue, for example, that Anscombe's lengthy discussion of "motive" early in her book is helpfully read as a response to some of Gilbert Ryle's arguments in The Concept of Mind, and that her discussion later on of the practical syllogism centers on a criticism of R. M. Hare's account in The Language of Morals.

Now, one of the things that made Hare such an important target for Anscombe was that she found his moral philosophy to be totally perverse. The issue here was not over support for the bombings of Hiroshima and Nagasaki (I have no idea about Hare's position on that), nor over whether he supported giving the degree to Truman (though I'm almost certain he did: only a tiny number of Oxford faculty supported her motion to oppose it). It's rather that his position was explicitly consequentialist in the sense Anscombe defines in "Modern Moral Philosophy": it is a position according to which, as she puts it, "'the right action' is the action which produces the best possible consequences", and so doing well is a matter of acting always "for the best". Thus, for example, we find Hare writing in The Language of Morals that in making a choice "It is the effects that determine what I should be doing; it is between the [possible] sets of effects that I am deciding"; and he gives a similar account toward the end of the book of how we are to make "decisions of principle" concerning whether or not to continue to abide by received moral views. Anscombe found this abhorrent: as a Catholic, she believed that we should be able to say from the start that certain sorts of action, such as the deliberate massacre of non-combatants in wartime, are simply off the table. However, she argued in "Modern Moral Philosophy" that the reason why contemporary philosophers couldn't recognize this possibility was not that they were bad at doing ethics, but that it was obscured by their assumptions about the nature of action and intention. And those are what she is taking on in her book.

3:16:  How are Aristotle and Aquinas important to her project – and how much did Wittgenstein influence her ?

JS: Anscombe's debts to Aristotle are pretty evident in Intention and in others of her early writings, such as her 1956 paper in Mind on the discussion of future contingents in De Interpretatione. With Aquinas, things are a bit different. We know that as an Oxford undergraduate she was significantly influenced by the Dominican priests at Blackfriars Hall, who oversaw her conversion to Catholicism. The pamphlet she published in 1938 with her friend Norman Daniel, which opposed Britain's entry into the Second World War, cites Aquinas directly and presents arguments right out of the Summa Theologiae. But in writing Intention Anscombe kept that influence largely beneath the surface: while she's channeling Thomistic ideas consistently throughout the text, Thomas gets referenced only in a stray footnote. I think this is another main cause of the initial obscurity of the book. For example, I still remember the feeling of revelation that came with realizing that her discussion of practical knowledge draws to a great extent on Aquinas' treatment of divine cognition in Part I, Question 14 of the Summa. With this background in place, so much of what she says becomes clearer.

As for Wittgenstein -- well, there is a lot that could be said here, and one difficulty for me is that I don't feel expert enough on Wittgenstein's writings to discuss their influence on her in detail. Anscombe was, of course, Wittgenstein's greatest student, and he handpicked her to edit and translate the Philosophical Investigations. Her longtime friend Anthony Kenny recently quoted Anscombe as saying, "I don't have a single idea in my head that wasn't put there by Wittgenstein" -- a characterization that's surely inaccurate (what about Aquinas?), but nevertheless points to the extent of her intellectual debts to him. I also talk in a long biographical essay on Anscombe that I wrote last year about the fondness she felt for her teacher, and the influence he had on her understanding of how philosophy should be done. (That account draws on a notebook with Anscombe's remembrances of Wittgenstein that's available in the archive of her papers held by the Collegium Institute at the University of Pennsylvania.)

3:16: What’s the distinction between foresight and intention and how does it figure in her condemnation of Truman’s action – and in framing her notions about intention in this way does that mean that she’s offering an ethical reading of intention or is it that she thinks that to get to the ethics we have to understand psychology of action or have a conceptual grip on it?

JS: As I understand her, Anscombe believes that the distinction between foresight and intention boils down to a difference in what is part of a person's practical knowledge of their action: she says that intentional action is known without observation, and the knowledge of it is "the cause of what it understands", while this isn't true of what is done with mere foresight. Another way to put this, which is supposed to come to the same thing, is that only what's done intentionally falls within the means--end order that's elicited by the sense of the question "Why?" that asks for a person's reasons for acting. This latter account of the distinction can be hard to sustain, though, since where doing one thing, such as polluting a river, is a foreseen side effect incurred in the course of doing another, such as increasing profits for one's company, it can be appropriate to answer the question "Why are you polluting the river?" with a statement like, "In order to increase profits". (This is one place where the discussion in my book gets things flatly wrong, but I'm doing some work now that tries to improve on it.)

Now, according to Anscombe this distinction is very important to sound ethical thinking: she believes, for example, that the intentional killing of innocents is what's wrong in any circumstances, while actions that will lead foreseeably to innocent deaths can be justified. (Can be, but not necessarily are: it's no part of double-effect reasoning that foreseen side-effects are ethically irrelevant.) But this doesn't mean that the distinction is itself an ethical one. Rather, it's a distinction that we use all the time just in describing what people do and why. That's why it is, in the first place, supposed to be a topic for what Anscombe calls "the philosophy of psychology" -- and so one that we are, according to the programme laid out in "Modern Moral Philosophy", supposed to approach only after "banishing ethics totally from our minds". It is, however, a significant question whether such a thing is so much as possible, given the evidence compiled by Joshua Knobe and others for the extensive influence of ethical considerations on the way we describe what human beings do. Those of us who wish to follow Anscombe have not, to my mind, done enough so far to address this last question directly.

3:16: So what is intention according to Anscombe and does your defense of the strong cognitivism thesis ‘intending is believing’ marry with Anscombe’s position?

JS: This is such a hard question. In her book, the closest Anscombe gets to a direct answer is when she writes in section 47 that "the term 'intention' has reference to a form of description of events" -- a position she contrasts with one according to which intention is "an extra property" of intentional actions "which a philosopher must try to describe". The latter position is the one we tend to take for granted in analytic philosophy today: we assume that whatever intention is, it must be a mental state, and that intentional actions are the actions that are caused by a person's mental states in the right way. But then what is Anscombe's alternative to this? I want to put it this way: she is saying that we bring the concept of intention to bear whenever we describe what a person (or animal) does in a way that gives her "Why?"-question application to it -- for example, when we describe what further someone is doing in doing something more immediate (say, making an omelet by cracking some eggs), or describe someone as doing something so that something else will come about (say, attending college to earn a degree), and so on.

But that's still not a direct answer to the question you asked. When philosophers ask, "What is intention?", they usually mean something like: What is it to intend to do something? And my paper with Beri Marušić, "Intending Is Believing", answers the latter question in a way that I think has a decent claim to fit with Anscombe's own ideas. We argue there that to intend to do something is nothing more than to believe, on the basis of practical reasoning, that one is going to do it. For me, the deep attraction of this position is that it helps to capture the way that, from the first-person perspective, the only concern of practical deliberation is to answer the question of what I will actually do. And the paradoxical-sounding slogan in the title of our paper is supposed to be a way to capture this phenomenon: our claim is that as an agent, the answer I give to the question "What will I do?", where this answer is given through practical reasoning rather than reflection on evidence for the judgment that I'll do this or that, just is my intention to do the thing I say I will.

3:16: Is Anscombe’s distinction between knowing our own actions and knowing of another’s secure in the terms she sets out, and have you an alternative way of securing it, if indeed it can be? Are there special sources for knowing our own intentions or is there a special domain or even a special character that enables this?

JS: This is one of the places where my thinking about these matters has changed a lot since I first started working in this area. Right from the beginning, I was convinced that readers of Anscombe weren't paying sufficient attention to her claim that the knowledge of one's actions is distinctive in being practical knowledge, and not just in being knowledge that we have "without observation". And before long I came to connect this idea to what I think is the fundamental insight of Richard Moran's great book Authority and Estrangement: that what distinguishes the first-person perspective isn't the special (say, introspective) mode of access that we have to what we think, want, and do, but rather the kind of rational authority that we exercise over these matters.

However, in many of my early papers I argued for a position that I no longer endorse, namely that the knowledge of one's actions could be both practical and perceptual -- that is, justified by the perception of what one is doing. I now think this is mistaken, and that the mistake arises from a failure to apprehend the distinctive form of practical reasoning. For example, suppose I reason that, since I need to get to the airport by noon and the bus that I need to get there is going to depart from the stop nearby at 11:30, therefore I'll leave for the stop no later than 11:25. In this case, the reasoning that takes me to this conclusion, which I identify with the belief that this is when I'll leave, has as a premise something that I know through observation, viz. that the bus will depart at 11:30. But is my conclusion a piece of observational knowledge? Of course not: I use this information to decide, rather than to predict, that this is what I'm going to do. And I want to argue that the same thing holds of the role that perception plays in the knowledge of what I'm intentionally doing right now. I try to work this position out in the penultimate chapter of my book, but I'll admit that the details of what I say there are not totally convincing even to me.

3:16:  Does the thesis that agential knowledge is practical knowledge entail that an agent always has non-observational knowledge of what she is intentionally doing?

JS: I'm going to stick with "No", which is the answer I've given consistently throughout all the things I have written on this topic. My most recent version of that answer is that in the normal case a person's practical thinking about what she's doing is always practical knowledge of what she's doing, but that it's possible for there to be exceptional cases in which a person acts intentionally but without knowledge of what she does.

There is, however, the following major barrier to upholding that answer, which I'm embarrassed to admit I haven't addressed sufficiently in anything I have written so far. For a follower of Anscombe, an important characteristic of intentional action is that when a person is doing something, call it "A", intentionally, the statement "Because she is doing A" can be an answer to the question of why she is doing something else. (For example: "Why is Jack peeling bananas?"---"Because he is making banana bread.") But there are powerful reasons for thinking that this kind of explanation presupposes a person's knowledge of the fact that is their reason for acting: so if Jack doesn't know he is making banana bread, then that can't be the reason why he is slicing bananas. Set against the arguments supporting the claim that we don't always know everything we do intentionally, which are powerful in their own right, this argument leaves us in a familiar position of philosophical aporia. I don't really know what is the right way out.

3:16: . Are we partly zombies when we use perception to guide our body actions, and does this mean that we might misidentify the object of an episode of bodily awareness, or would that be down to different reasons ?

JS: These are separate questions, as I see them. The first has to do with the role of conscious awareness in shaping action as it unfolds. In a paper that I published with Robert Briscoe we criticized the view, developed at length in David Milner and Melvyn Goodale's book The Visual Brain in Action, that the visual guidance of action is entirely unconscious or "zombie"-like. (This view has a lot of currency in visual neuroscience and has been endorsed by many philosophers, too.) Robert and my argument was that this position is overstated, and that the empirical evidence supports a more moderate, and less counterintuitive, position on which action is guided nonconsciously only when it's habitual, routine, or otherwise easy to execute. I'm not sure how successfully we made our case -- some philosophers have taken it seriously and there are neuroscientists who are sympathetic to it too, but I know that Goodale has not been convinced!

By contrast, the possibility of misidentifying an object of bodily awareness is one that comes up in the philosophical literature on immunity to error through misidentification. Roughly, a judgment of the form "x is F" is immune to error through misidentification when there's no question of its manifesting knowledge that something is F but a mistake about which thing that is (i.e., whether it's x rather than something else). A lot of philosophers have argued that many judgments about the state of one's own body have this characteristic: if I judge, for example, on the basis of first-person awareness that I am warm, or in pain, then while I might be wrong about that (for I could mistake for a pain what is really a tickle), I couldn't be wrong about whether it's myself that I am aware of. And this is sometimes taken to show that bodily awareness isn't a form of perception, since it's in the nature of a perceptual capacity that it concerns a range of different objects. But my paper on this topic challenges that reasoning, arguing that it's possible to misidentify which part of one's body a given episode of bodily awareness concerns, and proposing that we might understand bodily awareness as perception "from within" of the state of one's bodily parts.

3:16: You ask this so what’s the answer: Does visual space awareness require the visual awareness of space? How should we think about the role of visual spatial awareness in perception and perceptual knowledge?

JS: The first question you ask is from the title of one of my earliest published papers. (It began as a chapter of my Ph.D. dissertation.) Spelled out a bit, what I mean it to be asking is whether the visual awareness of spatial objects, properties, and relations is necessarily of them as positioned -- that is, located and oriented -- within a wider space. And there's a noble tradition of answering this question "Yes": for example, Kant argues that answer in the part of the Transcendental Aesthetic where he is critiquing Leibniz's view of how the concept of space is formed, and Husserl and Wittgenstein give what I think are similar arguments in some of their writings. 

But in my paper I discuss a case from the neurological literature of the patient RM, whose condition was studied at length by the UC Berkeley psychologist Lynne Robertson -- someone I worked with a bit during graduate school. And RM seems to have been in the sort of state that these philosophers say is impossible: he saw things, and could identify their shapes up to isomorphism, but did not seem to experience them as positioned in any wider space. For me, part of what makes this case so interesting is not just its bearing on the first-order philosophical question in the title of my paper, but also the further questions it raises about how philosophical arguments intersect with experimental findings. Lately, as my work in action theory has begun to draw on research in linguistics and experimental philosophy, I find myself wrestling with these questions once more, and without any satisfying way to answer them.

Now, the second question is about how our perception of the world, through vision in particular, relates to our knowledge of what we are intentionally doing. And some of what I've said above indicates the sort of answer that I want to give here: that perception is integral to action, and essential to our knowledge of what we intentionally do, but not because it provides evidence that justifies our beliefs about what we are doing; and that at least in some cases it is conscious visual awareness, and not merely nonconscious visual representations, that plays a role in shaping action as it unfolds.

3:16: How should we understand the ability of a baby, say, to match what they feel with what they see? Is this about associationist belief formation or hard-wired sensory modalities that make things look like they feel?

JS: The version of this question that's discussed in Locke's Essay is not about a baby, but about a "man born blind" who eventually is "made to see". Philosopher's call it "Molyneux's question" because it was posed to Locke by a man named William Molyneux. And Molyneux and Locke both thought that such a person could come to recognize seen shapes as the same as the shapes he knew through touch only on the basis of associative learning. This position has some intuitive backing, and it also seems to be supported by experimental evidence: most famously, a study that was published in Nature Neuroscience found that newly sighted individuals could match seen shapes with seen ones and felt with felt, but could not match a seen shape with one that they had perceived only by touching it with their hands.

Now, here's a funny story. I first learned of that study when my mother sent me a (real, made from paper) clipping from a New York Times article about it. The article left me skeptical at first, but then I read the study in detail and came to believe that it really did show what the authors had claimed. And this result was a huge surprise to me, since under the influence of Gareth Evans's arguments in his famous paper on the Molyneux question, I had long held that sight and touch shared a common form that was given, not by learned associations, but by a perceiver's capacity for spatially directed action. So as a good empiricist (though not in the Lockean sense) I set about revising those views in the face of this new experimental evidence. But that further thinking led me to the conclusion that I defend in my paper, "Do Things Look the Way They Feel?": that the study in question didn't demonstrate the Lockean conclusion after all, since (to put things very simply) the participants in it probably couldn't see well enough for their failure in the touch-to-vision matching task to have the significance attributed to it.

What about babies, though? While fascinating work has been done in this vein by the psychologist Andrew Meltzoff, as I see things there are at least two big difficulties with using experiments involving infants, instead of a grown child or adult who's recently been given sight, in an attempt to address the Molyneux question. One is that we can't ask infants what they perceive: instead we have to rely on indirect evidence about how infants direct their attention in response to novel stimuli, and things like that, in order to get at what their experience is like. But it's possible that these observed effects are due to processes that operate below the level of conscious awareness -- which is what, as I understand it, the Molyneux question is supposed to be about. 

And the second difficulty is that infants, no less than those who are "made to see" only later in life, develop their visual capacity only gradually -- a process that is accompanied by a lot of tactual exploration, as any parent can tell you. (Trust me: I am an expert in young children.) All of this, together with my reservations about past experimental work that I explained just above, has made me start to think that it might be impossible to resolve Molyneux's question experimentally after all, since anyone who can see well enough for the experiment to be carried out on them would already have had the opportunity to form so many cross-modal associations that the result of the experiment would be meaningless. Despite the initial appearances, it might be a question that we can answer only on philosophical grounds.

3:16: And what happens when I see a coin from a perspective that makes it look slanted – am I really seeing the perspectival shape or just imagining it?

JS: This is another fundamental question in the philosophy of perception that can be traced back to early modern empiricism: for example, Hume writes in the Treatise that "It is commonly allowed by philosophers, that all bodies, which discover themselves to the eye, appear as if painted on a plain surface". Now, most philosophers and psychologists today tend to reject this position, holding instead that the human visual system yields a representation of the world as extending in three dimensions: so, for example, a coin that's tilted away from your eyes will look to you to be circular, not elliptical. But even so, a lot of philosophers think that what Hume says here is at least partly right, since they believe that visual perception includes viewer-relative or "perspectival" shapes and sizes in addition to "objective" or view-invariant ones. And so there's a huge debate among philosophers of perception concerning how to understand this perspectival character.

How can we tell whether these philosophers are right in thinking that there are perspectival shapes in visual experience? It might seem like this should be easy: after all, when we introspect our own experience we can just observe whether or not these perspectival shapes are there. But Robert Briscoe has written what I think is a really terrific paper that serves to challenge that assumption. He argues that what philosophers take to be the presence of perspectival shapes in visual experience might be due to the imaginative construction of those shapes in a process that he calls "make-perceive". And I don't think we should trust our ability to tell on the basis of introspection whether we are perceiving perspectival shapes or merely make-perceiving them.

Okay, so what alternatives are there to the introspective approach? In a paper that I wrote with Assaf Weksler, we imagine a couple of ways that the question could be addressed through the tools of experimental psychology -- though without actually running any of these experiments ourselves! However, a more recent paper (here's a nice summary written for a popular audience) out of Chaz Firestone's lab at Johns Hopkins presents a series of studies that try to do just this, albeit not quite in the way that Assaf and I envisioned. The authors of that paper take their results to support the popular position I described above: they conclude that visual experience includes perspectival shapes in addition to objective ones, since (for example) the presence of objectively circular but perspectivally elliptical shapes in a display interferes with visual search for shapes that are objectively elliptical. For what it's worth I'm not totally convinced by these results, though I've not yet gotten around to writing up why -- in any case I think it's a terrific study that's a paradigm of how experimental philosophy of perception should be done.

3:16:  There’s a dispute between a ‘self-location thesis’ on the one hand and a ‘minimal view’ on the other – why and how do you argue that phenomenology makes the minimal view inadequate?

JS: Right now I'm sitting at my laptop, which is positioned in front of me. I see where it is. But in seeing where it is, do I merely see that it is: to the front? (This is how the "minimal view" would have it.) Or do I also see that it is: in front of me? If the latter description is right, then visual experience is "self-locating" in a rich way: in addition to presenting things as they are in relation to the position where I am, it also presents my position -- though not, of course, as a position that lies within the field of view.

One of my earliest papers offers an argument that visual experience is, in fact, self-locating in this rich way. Briefly, the argument is that this self-location thesis is required to understand how there can be a visual experience of change in one's bodily position -- as when, for example, you're sitting in a train car and the train next to you pulls forward, so it looks to you like you're moving back. (The idea is that the minimal view can't account for the difference between such an experience and a corresponding experience where things in the world are seen to move in the opposite way.) Does that argument succeed? Well, today my skepticism about the value of introspection makes me worry that some of the key premises in it are insufficiently supported, and there's also a critical discussion of it in Louise Richardson's paper "Sight and the Body" that raises some important objections that I don't yet have an answer to. I suppose the ball is in my court!

3:16:  What are the limits of self-knowledge and how important is introspection in this?

JS: If there's one thread running through most of my published philosophical work, it's that introspection isn't good for shit. It's unreliable, first of all, even as a means of discovering what conscious experience is like. Second, it's close to useless as a way of knowing things like what you believe or intend, since none of this is such as to show up as an introspectible aspect of your conscious life. Finally, the idea that self-knowledge comes through introspection encourages a bizarrely passive picture of self-understanding that leaves out a main part of what's distinctive about the first-person standpoint. To use an image from Moran's Authority and Estrangement that's stuck with me ever since I first encountered it, that picture suggests that I could have perfect self-knowledge if only I had a mindreading device that scanned my brain and revealed to me all my thoughts and perceptions -- and yet if I had such a device and used it to read your mind instead of my own, nevertheless I wouldn't have anything like self-knowledge of your mind!

If not introspection as a source of self-knowledge, then what else? A lot of what I've said here is meant to be in the vein of Moran's position, which is sometimes called a form of "rationalism". On this view, we know our minds largely through the exercise of the self-conscious rational capacities through which we make our minds up. One thing that's left aside here is the question of how we know our conscious experiences -- this capacity has to be accounted for in another way. But anyone with a view like this also needs to have some things to say about the cases where we don't know our minds after all, as when we deceive ourselves or have attitudes that are somehow repressed -- if this isn't due to a lapse in introspection, then what explains it instead? This question is sometimes thought to be devastating for a position like Moran's, but I've argued in something I wrote that discusses a couple of recent books on self-knowledge (they are Quassim Cassam's Self-Knowledge for Humans and John Doris's Talking to Our Selves) that in fact these are phenomena that a psychologically realistic "rationalism" should have no trouble at all in accounting for.

3:16:  What’s the best way of understanding the cognitive processes of ‘mirrored self-recognition’ and does this have implications for cognition beyond recognizing ourselves in mirrors?

JS: The question of how we recognize ourselves in mirrors is tied to the questions about self-consciousness and immunity to error through misidentification that I've touched on a few times above. Start with the premise that a judgment "I am F" is an instance of self-consciousness if my knowledge that someone is Fis inseparable from the knowledge that I am: so I could be wrong about whether anyone is F but not wrong about whether, if someone is F, it is me. This seems to hold for the self-ascription of perceptual states and of sensations like pains and tickles, as well as of states of one's own body where these are grounded in first-person awareness of bodily position and movement.

This last fact -- that we sometimes have self-conscious knowledge of bodily states -- seems to count against the Cartesian view that each of us is really an immaterial ego. But most philosophers who accept this argument will draw the line at the knowledge of our bodies that we have "from within": they think that any knowledge of ourselves that comes from looking at our bodies, or seeing our bodies reflected in mirrors, isn't self-conscious knowledge in the strict sense. And I've tried, though not yet in any great detail, to develop an account of mirror recognition on which it doesn't, after all, involve a judgment identifying oneself with the person one sees in the mirror, and so is a candidate for a radically anti-Cartesian form of self-consciousness. Unfortunately I am not at all sure that this account holds up, partly because of worries about whether the concept of immunity to error through misidentification can do the work that my argument requires of it.

3:16: Why do you say honesty in assertion requires non-empirical knowledge? What does it require then?

JS: This is the thesis of a paper I wrote with Eric Marcus that considers what's required to assert something honestly. Our starting point is the observation that it's not sufficient for an assertion to be honest, that you believe what you assert: for you might have no idea that you believe what you said, perhaps because the belief was repressed. (For example, if you have a subconscious belief that you had a terrible childhood, you might tell a friend, "I had a terrible childhood", just to make the friend feel better -- and this won't be an honest statement even though it's a statement of something you do, in fact, believe.) We then argue that any knowledge of your belief that you might acquire through evidence, whether "introspective" or otherwise, still won't be enough for you to assert that belief honestly. (Continuing the example above, suppose you engaged in a process of self-examination that revealed to you this belief about your childhood. Could you then say, honestly, to a friend that you had a terrible childhood? Of course not: the belief would be known to you only "at a distance", in what Richard Moran calls an alienated way.) And our conclusion is that the only form of self-knowledge that grounds the possibility of honest assertion of our beliefs is a knowledge that isn't through evidence at all -- perhaps something like the "rationalist" self-knowledge that I described above in summarizing Moran's position. So this is a good example of my slogan that introspection isn't good for shit, or at least for much: even when introspection does give us knowledge of what we believe, this knowledge isn't of the kind that enables us to speak our minds.

3:16: And finally, for the readers here at 3:16, can you recommend five books that would take us further into your philosophical world?


G. E. M. Anscombe, Intention -- This goes first for obvious reasons, but not without an honest description of how very, very hard it was for me to find my way into Anscombe's arguments, and how much of the book I found totally impenetrable at first. I believe that Intention more than repays the effort that's required to understand it, but I sympathize with those who'd rather spend their time on something else.

Richard Moran, Authority and Estrangement -- Perhaps no book has had such an immediately transformative impact on me as this one. It's been about 10 years now and I can still recall where I was when I first started reading Authority and Estrangement, and remember the feeling of having it reshape my thinking almost a sentence at a time. I often point to it as a book that shows how the tools of analytic philosophy can be given a real-life bearing: it's the rare piece of philosophy that's helped me to understand myself better than I had before, as opposed to just thinking differently about some abstract questions.


John McDowell, Mind and World -- This book had about the effect on me in my early years of graduate school that Moran's book had on me afterwards. I reread Mind and World recently to prepare for something I wrote to mark the 25th anniversary of its publication, and I was consistently impressed at how much of it remains fresh and exciting. McDowell's core thesis -- that perceptual experience involves the operation of conceptual capacities -- generated a massive literature that I think got a bit tired, but his account of the nature of human subjectivity and agency remains tremendously important, as is his vision of how philosophy should relate to science and to its own history.


Michael Thompson, Life and Action -- This is another difficult book whose treasures have taken me time to discover. In addition to Thompson's compelling articulation of neo-Aristotelian accounts of life, action, and practice, the book is also exciting in its philosophical method, which focuses on uncovering what Thompson calls the different forms that thought can take when we think about these different things. I don't yet have a way of explaining this "form"-talk that I myself find wholly satisfactory, but I'm doing my best to get there.


Beth Levin, English Verb Classes and Alternations -- Everyone you interview has got a surprise book, right? Anyway, this is mine. Under the influence of Anscombe I have been trying for a while to explain why we should work to investigate fundamental concepts through the investigation of what Wittgenstein would call their "grammar". Levin's book, which is basically a work of descriptive linguistics, provides a window into the grammar of a huge range of ordinary English verbs, showing how this grammar is tied to the conceptual structures that those verbs realize. (There's a more accessible treatment of this body of research in Steven Pinker's book The Stuff of Thought.)


Richard Marshall is biding his time.

Buy his second book here or his first book here to keep him biding!

End Times Series: the index of interviewees

End Time seriesthe themes