Tuesday, September 26, 2006

Mailbag: Dualism and Causal Exclusion

[István Aranyosi writes in with the following remarks...]

Karen Bennett has recently been working on showing that the exclusion problem cannot be escaped by the dualist. I think this is a quite widely held view nowadays. I have recently been working on showing the contrary: that it does not even arise for dualism, correctly understood. As Karen rightly points out [PDF], the exclusion problem is the fact that the following five propositions seem to be inconsistent, so one of them has to be denied:

1. DISTINCTNESS: mental events are distinct from physical events.
2. COMPLETENESS: every physical effect has a sufficient physical cause.
3. EFFICACY: mental events can cause physical events, and can do so in virtue of their mental properties.
4. NON-OVERDETERMINATION: effects of mental causes are not systematically overdetermined in a bad way (unlike the death of firing squad victims).
5. EXCLUSION: no effect has more than one sufficient cause unless it is overdetermined.

I argue in my paper [PDF] that one of the theses that a dualist should uphold is that behaviour itself is not purely physical, but it is ‘mentally enriched’, as I like to put it. The idea is not ad hoc: think of ‘my arm’s rising’ versus ‘my raising my arm’; the first is a purely physical effect, the second, according to me as a dualist, is both physical and mental. This yields a premise to be added to the five above:

6. Some effects of some mental causes are not purely physical.

I then try to show that adding 6 to 1-5 yields a consistent set of propositions. Take an effect e of a mental cause c, say ‘my raising my arm’ as caused by ‘my wanting to raise my arm’. The cause is both physical and mental: it’s both a wanting and a neural property. The effect is both physical and mental: it’s both a mere upward motion and an intentional raising. All this is compatible with:

1. Distinctness: wanting is a distinct property from the neural one.
2. Completeness: it holds for purely physical causes and a qualified version holds for not purely physical events: ‘the instantiation of the physical aspect of every effect has a sufficient physical cause’.
3. Efficacy: again it is interpreted as: events with mental aspects can cause events with physical aspects (and can do so partly in virtue of their mental aspect)
4. Non-overdetermination: the mental aspect of the cause is responsible for the mental aspect of the effect, mutatis mutandis for the physical aspect, so there’s no overdetermination
5. Exclusion: the effects of mental causes do not have either a physical or a mental sufficient cause; both aspects of the cause event jointly cause them.

Finally, let me clarify why I speak of some effects of some mental causes in (6). Paradigmatic mental causes are those like believing, wanting, desiring, etc. These are the ones for which (6) is true. Non-paradigmatic or derivatively mental causes are the ones that are effects of paradigmatic ones, for example ‘raising an arm’. Raising an arm can cause ‘destroying the spider’s web’, and this effect is purely physical even if caused by a mentally enriched cause.


  1. Here's a possible worry: intuitively, it seems like my raising my arm is just the event of my mind causing my arm's raising. If your mind has nothing to do with the physical aspect of your arm's movement, then it's an awfully limited sort of "efficacy" you're granting the mental here.

    (But I haven't had time to read the paper yet; perhaps this objection is addressed there?)

  2. Yes, this is indeed an objection I consider in the paper. I’m not sure how credible my reply is to the physicalist, but assuming you are a naturalistic property dualist, which is my assumption in the paper, you will think that the world or many regions of it are characterised by causal interconnections among constitutively dual events –mental/physical- so the notion of efficacy that you work with is one according to which these mental/physical events bring about other mental/physical events. So for a dualist the question “does the mental aspect of your cause event have any impact on the physical aspect of your effect event?” does not arise, because the world doesn’t work like that: it is events in their full constitution that do the causal work. However, in the paper I talk about a contrastive approach to mental causation, according to which sentences about mental causation are to be understood as:

    C qua mental rather than qua physical causes E qua mental rather than qua physical.

    But the ‘rather than…’ phrase is not to be understood as eliminating what follows in the place of the dots, unlike in Jonathan Schaffer’s contrastive approach to causation where the phrase is used to eliminate some irrelevant conditions from the status of being causes.

    So there are two ways of talking about causation: either among causes in their full constitution or contrastively as above. The first yields a notion of efficacy that suits the dualist views, though the physicalist may not like it (so much worse for the physicalist). The second is such that it would even be strange for the mental aspect of a cause event to directly impact on the physical aspect of an effect event (it impacts it indirectly via impacting its mental aspect; recall that it’s naturalistic dualism, so there are psychophysical laws of nature).

  3. This is a picky logical point, but it seems funny to me to say that adding 6 to 1-5 makes 1-5 consistent. If adding 6 leaves you with a consistent set, then 1-5 are already consistent without 6. What 6 does is show you how 1-5 are consistent by helping show that the supposed inconsistency is not present.

    I didn't read the Bennett paper, but it seems to me that the dualist can escape the exclusion problem in other ways. Leibniz, of course, would deny efficacy, as would property dualists. I think George Bealer (who does face this issue as an event dualist) denies 5, but I'm not very sure of that.

  4. Yeah, sure, you're right about the consistency. What I offer by 6 is a way to interpret 1-5 so that they are not in fact inconsistent. (I say in the post that '1-5 seem to be inconsistent')

    Regarding the dualist escape, of course, there are at least as many ways for the dualist to "escape" the problem as there are ways to deny some proposition from 1-5, but not all of them are considered escapes. I have tried to offer a real escape by accommodating all of 1-5.

  5. Oh sorry, there is one proposition that the dualist shouldn't deny, namely, the distinctness of the mental from the physical...

  6. Distinctness is a proposition about property dualism, now that I think of it. Substance dualists and (mere) property dualists accept that, but (mere) event dualists like Bealer need not.


Visitors: check my comments policy first.
Non-Blogger users: If the comment form isn't working for you, email me your comment and I can post it on your behalf. (If your comment is too long, first try breaking it into two parts.)