Dreßing points out that my claim that intentions have a causal role to play in the online control of action confronts me with the problem of mental causation. She also suggests that this problem is more pressing for me than it is for the accounts of the functions of intentions proposed by Bratman and Anscombe. While I agree that the problem of mental causation is an issue for me, I disagree with her assessment that it isn’t as serious a worry for these accounts.
First, let me clarify that when I talk about conscious intentions and their causal role, I am concerned with what Ned Block (1995) calls access consciousness rather than with phenomenal consciousness. In other words, my claims are about intentions qua conscious states exploiting and conveying information globally available in the cognitive system for the purposes of reasoning, speech, and high-level action control. My account thus faces the “easy” problems of consciousness rather than the “hard” problem (Chalmers 1995). I share Chalmers sanguine assessment about phenomena pertaining to access consciousness:
There is no real issue about whether these phenomena can be explained scientifically. All of them are straightforwardly vulnerable to explanation in terms of computational or neural mechanisms. (1995, p. 201)
This is not to say, however, that in confining oneself to phenomena of access consciousness one can eschew all metaphysical conundrums. In particular, as pointed out by Dreßing, the mere fact that Cartesian dualism has fallen out of favour and that the vast majority of philosophers and cognitive scientists are nowadays willing to embrace some form of materialist monism doesn’t insure the dissolution of philosophical worries about mental causation. The version of the problem of mental causation that non-reductive physicalists, whatever their exact persuasion, are confronted with is the Causal Exclusion Problem: how could mental properties play a causal role given that they appear to be screened off by their physical realizers?
Dreßing argues that this problem is more pressing for my view than for the pragmatic (Bratman) or the epistemic (Anscombe) creation myths, the reason being that these latter two teleological myths are about prior intentions and that neither “require any assumption about causality, as they do not involve a mind-world directed causality, but rather an intra-mental mental causality” (Dreßing this collection, p. 6). Dreßing also claims that Velleman’s view does not imply any explicit claim about causality either, since on this view intentions are a spandrel or a by-product.
I disagree with this assessment for three reasons. First, as I explained in section 2, while the speculative character of these stories qua answers to the why-question may justify labelling them as creation myths, the stories also offer answers to the what-question. In that regard their claims about the epistemic or pragmatic roles of intentions should be taken as empirical claims. Thus, even if we go along with Velleman’s claim that a capacity for intentions is a spandrel and that the epistemic and pragmatic functions of intentions are not teleofunctions, they are nevertheless functions in the ordinary functionalist sense and we still need an explanation of how intentions can play these epistemic and pragmatic roles.
Second, the Causal Exclusion Problem is a problem for anyone espousing a non-reductive form of materialist monism,[1] whether their primary concern is with intra-mental causation or with mind–world causation. Suppose that a state S has the mental property M (e.g., the property of being an intention to go to London on Monday) and a physical basis P, suppose that S' has the mental property M' (e.g., the property of being an intention to buy a train ticket to London) and a physical basis P', and suppose that S'' has the mental property M'' (e.g., the property of being a being a belief that one will go to London on Monday) and a physical basis P''. On a Bratmanian pragmatic account of intentions, I would want to be able to say that my intention to go to London on Monday causes, via further means-end reasoning, my intention to buy a ticket to London. But how can the mental property M of S play a causal role in bringing about a state S' with mental property M', given that they appear to be screened off by the physical properties P and P'? Similarly, with regard to the epistemic function of intentions, how could I say that my intention to go to London on Monday causes my belief that I will go to London on Monday, given that mental properties M and M'' appear to be screened off by the physical properties P and P''?
Third, while it is true that on Bratman’s account future-directed intentions may only cause behaviour through the mediation of present-directed intentions, still Bratman insists that the whole point of having a capacity for intentions is to produce behaviour that contributes in the long run to our securing greater desire-satisfaction. Similarly, on a reliabilist reading of Anscombe’s epistemic claim that intentions embody knowledge of our actions, they do so because intentions reliably cause what they represent. As Velleman puts it, “[u]nless an intention with the content ‘I'm going to move my toe’ reliably causes my toe to move, it won't amount to practical knowledge” (Velleman 2007, p. 201). Thus, Bratman, Anscombe and Velleman cannot be exonerated from the task of explaining how mental states can cause behaviour.
With respect to the problem of mental causation, we are all in the same boat. The metaphysical standing of my account is no less or more precarious than the standing of these other accounts. Are we then all metaphysically doomed? Readers should not hold their breath; I have no new, unassailable solution to the problem of mental causation to offer. Yet, it would certainly be premature to claim that the problem of mental causation is insoluble. Many lines of response have been proposed and are currently being explored (for a review, see Robb & Heil 2014). I cannot discuss all these accounts here. Let me just say that the approach I find most congenial stems from Fred Dretske’s work (Dretske 1988, 2004) on psychological explanations of behaviour. Dretske distinguishes between triggering causes and structuring causes, where a triggering cause is an event that initiates or triggers a causal chain of events, and a structuring cause the cause of the process or setup that makes a given triggering cause produce the effect it does. To take an example from Dretske (2004), moving a computer mouse is the triggering cause of cursor movement, but hardware and programming are the structuring causes of cursor movement. Dretske’s central claim is that mental states and events are best analysed as structuring rather than triggering causes of behaviour. On this view there is no competition between physical and psychological or mental explanations, since they have different explananda. While the triggering physical properties explain bodily motion, i.e., explain why bodily motions occur at a certain point in time, the structuring mental properties explain behavior, i.e., they explain why in circumstances of a certain sort, bodily motions of this kind rather than that kind are produced.
Much work remains to be done in order for us to understand more precisely how structural causes operate and in particular how they can do so in the dynamic way needed to account for the plasticity and flexibility of human behaviour. In this respect, Dretske’s account remains largely under-developed (for recent work on this issue, see e.g., Slors 2015; Wu 2011). Dretske’s approach in terms of structuring causes has the great merit, however, of offering a potential solution to the Causal Exclusion Problem and to let us see how explanations in terms of physical properties and explanations in terms of mental properties may not compete but rather complement each other. As we will now see, thinking of intentions as structural causes of action rather than triggering causes can also help us assuage certain empirical worries.