Friday, 22 February 2008

Making up the Mind 9: How Brains Model Minds

To recap, I'm making my way philosophically through Chris Frith's Making up the Mind. Chapter 6 starts by presenting us with an interesting discussion of our ability to discern the characteristics of moving people even on the basis of very little information. We are referred to the fascinating website BioMotionLab for some fun interactive examples of this. Frith also discusses the tiny cues (millimetre eye movements) we use to take 'the first step into someone else's mental world.' We are also treated to what is rare in the neuropsychological literature - an epistemologically-non-overblown description of the nature of mirror neurones. Our natural inclination to 'read' movements as goal-oriented actions is lucidly described. Even in a description of the experience of pain, Frith mentions with admirable honesty that he finds it hard to understand Wittgenstein's critique of traditional mentalist conceptions of the putative 'privacy' of subjective experience but, beyond some fairly unproblematic (i.e. easy to back-translate into a more phenomenologically acceptable idiom) lapses into talk of 'the privacy of experience' and the 'construction of mental models' etc., the text is conceptually clean. It is almost as though a philosophical critic has gone through the text and purged it of all the entifying, homuncular, and mereological fallacies that, as I have noted in previous blogs, bedevil previous chapters.

The difficulties soon crop up, however, when the discussion turns to the difficult problem of (pp. 151ff.) 'The Experience of Agency'. This putatively 'ubiquitous' and resolutely 'private' experience is

The experience of being in control, of deciding to do something and then doing it. Of being in control of [or, as Frith later suggests, perhaps simply having the feeling that one is in control of] our destiny. We are all agents. But there is much more to our sense of agency than performing actions to achieve goals. We make choices. We decide which goals to aim for. We decide when to perform actions. We are not just agents. We are free agents. At least for the small things in life, we all belive that we are in control and can cause things to happen. My hand is resting on the table and I am staring at my finger, waiting for it to move. Nothing happens. And yet, whenever I want to, I can lift my finger. This is the mystery of mind over matter: the way thought can make things happen in the physical world.

Now, first, this seems like a somewhat phenomenologically gauche description of agency. For what evidence is there that there is any experience characteristically associated with my agential acts? I can myself find none in my experience. Or rather, I can perhaps locate an experience of trying hard to perform some act. (Similarly, on occasion I find myself having to make decisions before I freely act; most of the time, however, my action does not require such prior deliberation.) But this feeling is notable by its occasionality: most of the time I do not find myself having to try; or to the extent that we wish with Jennifer Hornsby to make room for a broader notion of 'trying' (I myself don't so wish), the trying is not at least something which is experienced as such. (Cf. Wittgenstein, Philosophical Investigations, 622, "When I raise my arm I do not usually try to raise it".)

Take the interesting experience of staring at our finger in a state of alienation and waiting haplessly for it to move. In this peculiar mental state we can even find ourselves willing our finger to move, only finding it still stubbornly residing there. It is precisely when we are in this frame of mind that we start to imagine - as Frith himself seems to imagine - that agency is characterised by inner acts of willing, acts which then cause things (such as bodily movements) to happen. But surely what this mini-phenomenological-experiment teaches us is that everyday agency has little to do with trying or willing or inner causation. What I need to do is to move my finger, where this is not to be characterised as my doing something mental in order to cause a bodily movement, but rather as my straightforwardly doing the bodily movement itself. 'Motility is not the handmaiden of consciousness', as Merleau-Ponty notes. Or, as Schopenhauer wrote in The World as Will and Representation, vol 2, p. 36: "We do not by any means recognize the real, immediate act of will as something different from the action of the body, and the two connected by the bond of causality; both are one and indivisible. Between them is no succession; they are simultaneous."

In terms of the previously described unholy trinity (the homuncular, entifying, and mereological fallacies), the 'inner causation' story is surely homuncular. Take the following passage, cited by Frith (also in his and Sarah Jane Blakemore's paper Self-Awareness and Action) from Ian McEwan's super novel Atonement.

She raised one hand and flexed its fingers and wondered, as she had sometimes before, how this thing,this machine for gripping, this fleshy spider on the end of her arm, came to be hers, entirely at her command. Or did it have some little life of its own? She bent her finger and straightened it. The mystery was in the instance before it moved, the dividing moment between not moving and moving, when her intention took effect. It was like a wave breaking. If she could only find herself at the crest, she thought, she might find the secret of herself, that part of her that was really in charge. She brought her forefinger closer to her face and stared at it, urging it to move. It remained still because she was pretending, she was not entirely serious, and because willing it to move, or being about to move it, was not the same as actually moving it. And when she did crook it finally, the action seemed to start in the finger itself, not in some part of her mind. When did it know to move, when did she know to move it?

Now Frith's use of this literary passage is entitled 'It isn't just scientists who wonder how we can control our actions'. And that's fine. But, in keeping with the blankly empirical character of his text, no mention is made of the extraordinary state of mind, Befindlichkeit, or what Matthew Ratcliffe calls the 'feeling of being' in which, or which frame, the question being posed. And it is surely not unreasonable to suppose that not only is the characterisation of Briony's state of mind (rather than the content of her thoughts) the real purpose of McEwan's text, but also that this state is what provides the conditions of (perhaps merely apparent) intelligibility for the asking of Briony's questions. What state of self-alienation does one need to be in to view one's own hand as a machine or fleshy spider? To view it as having a life of its own? Perhaps it is to be in a state in which the homuncular view - that one is in some of kind of causal relationship with one's own body - as if one were the inner captain of its ship - as if the way to characterise intentions is as inner causes of actions which 'take effect' at certain points in time - is apt to come naturally to one. Yet Briony also considers the possibility of a non-alienated conception of agency, in which actions 'start in the finger itself, not in some part of her mind' (cf Merleau-Ponty again: "motility is not the handmaiden of consciousness"). And that, we may suppose, is actually the phenomenologically apt way of describing the situation.

It is just this way of describing the situation, however, that Frith brings into question in the passages which follow. 'Being an agent', he says (p. 153), 'is all about cause and effect.' The discussion about agents veers confusingly straight into a discussion about the brain: First an interesting distinction is stipulatively drawn between 'physical' and 'mental' times. I am to ring a bell by pressing a button with my finger, and report the time from a clock in front of me when I do this and when the bell rings. (This is an interval in mental time. The physical time is the actual interval between button press and bell ring.) The finding is that the typical mental time is about half as long as the physical time. In an alternative condition I am not asked to ring the bell. Rather my brain is magnetically interefered with such that my finger will press the button without me myself pressing the button (i.e. we have a movement but not an action). The finding is that the (pseudo-)mental time in the second interval - i.e. the interval that I judge to occur between when my finger presses the button and when the bell rings - is about twice as long as in the first case.

This finding is described by Frith as follows:

Your brain recognizes that you are not being an agent, and so does not recognize you as causing an effect. It therefore reduces the binding of the events in time. ... The brain creates my experience of agency by binding together the causes and effects of the actions I perform.

But no justification is given for these descriptions; no explanation of what warrants it is provided. First, we have still been given no evidence for the supposition that we do normally have a 'sense of agency'. We have indeed been offered descriptions of cases in which we have a striking sense of not being an agent (for example when we find that our finger has been caused to 'press' a button via a magnetic manipulation of our brains). But nothing which would license an inference from this to the idea that we normally have a sense of being an agent has been provided.

Second, the text appears to commit a flagrant mereological fallacy: the brain is said to be 'recognising' whether or not I am an agent, and to 'recognise' whether or not it is I who cause the effect (I'm assuming here that the effect is not the moving of the finger (since then we would have a homuncular compound to the mereological fallacy) but rather the genuine effect of the tone being played). But it is people, not brains, that recognise things. Brains just consist in neurons that fire if they are caused to by external or internal stimuli.

Finally, who is supposed to be witnessing these temporally bound-together causes and effects? Frith says that this is 'my sense of agency'. But on this (homuncular) picture, it seems that we are once again placed on the receiving end of the brain's information processing: the brain produces an experience for an 'us' who has retreated to an existential point far inside the skull.


Let us now take stock of the discussion, and offer a diagnosis of what seems to have happened. Frith starts by wondering whether or not we really are agents. Perhaps, he says, we only have a sense of agency (p. 152): 'my beliefs on free will are very ambivalent. What I do know is that I have a very strong experience of free will. I feel that I am in control of my actions.' Whether or not he is 'in control of his actions' is left aside; it is rather an account of this putative experience of being in control that we will be given.

The parallels with Frith's treatment of perception are clear. Whilst we normally take it that perception puts us in touch with the world, Frith's account has it that this is merely an illusion, a sense of being in contact, created by the brain. As I argued then, the real motivation for this account was the homuncular, mereological, and entifying fallacies inscribed within it, and not the empirical data which was presented through the distorting lens of these conceptual confusions. Now we turn to action, Frith is less willing to argue definitively that our idea that we are actually the agents of our actions (or that we have 'free will' as he puts it) is an illlusion. Nevertheless the account he provides is designed to show how a putative inner sense of our own agency is a function not of a direct registering of any putative causes of our actions, but rather a sense created for us by the brain and constituted by judgements of temporal proximity of our movements and their environmental effects.

Strikingly absent is any attention given to the question of what it really means to talk of an agent who acts voluntarily and intentionally. A good place to start might be with the following: That we tend to take fellow human beings as agents unless we have reasons to doubt their agency, a 'default assumption' which we may well think constitutive of the very concept of what it means to be an agent. Or that we tend to take what someone says as the reason for their action as the reason for their action unless, again, we have reasons to suppose otherwise; another assumption which can not implausibly be taken to structure our very discourse of agency. Or that human action is voluntary to the extent that it is either not coerced from 'outside' (e.g. by a slave-master) or not marked by the presence of surprise at the occurrence of the movements one's body is making (e.g. when we spasm). Or that the meaningfulness and intentionality of action is not a function of its being caused or brought about in a particular way, but rather of its occurring in the ongoing context of a meaningfully structured human life embedded meaningfully in an environment.

Suppose instead, however, that we are, however intimately, conceived tacitly and ab initio to be lodged in our bodies like pilots trying to steer a ship from within, then it will be only too natural to suppose that the intentional or volitional character of our action is a matter of the generation of inner motor commands - or, to suppose indeed, that we may have to make do with an illusion of such commands. Disidentify oneself with one's body, consider it as a mere piece of flesh - as a 'fleshy spider on the end of [the] arm' - and it will come as no surprise that the attempt to reconnect us to it may come to seem too ambitious, and that we may instead have to put up with a mere sense of being an agent.