Dualism and the Problem of Interaction

Daniel Dennett raises a well-known and oft rehearsed difficulty for any dualist metaphysics of the human person in the following remarks:

“The standard objection to dualism was all too familiar to Descartes himself in the seventeenth century, and it is fair to say that neither he nor any subsequent dualist has ever overcome it convincingly. If mind and body are distinct things or substances, they nevertheless must interact; the bodily sense organs, via the brain, must inform the mind, must send to it or present it with perceptions or ideas or data of some sort, and then the mind, having thought things over must direct the body in appropriate action (including speech)…but anything that can move a physical thing is itself a physical thing.” (1991, pp. 33-5)

Call the objection raised here the problem of interaction (POI). One might simplify the argument by reducing it to a set of incompatible theses, as follows:

1) Minds and bodies are distinct (mental and physical) substances. (dualism)
2) Minds and bodies must interact.
3) Anything that can move a physical thing is itself a physical thing. (physicalism)

Since denying (2) nowadays seems to strain credulity (despite the fact that many early Cartesians did just that), the best hope for resolving the problem would seem to be a denial of either (1) or (3). If one believes that (1) and (2) are true, then (3) must be false; and likewise, if one believes that (2) and (3) are true, then (1) must be false. Dennett denies (1). I believe we are better off to deny (3).

In the first place, POI, at least as stated by Dennett (one of its chief modern proponents) comes very close to begging the question. As Trenton Merricks (2007) points out, POI is close in form to the following (nearly) question begging argument: If God exists, then something non-physical (God), causes physical events; but nothing non-physical can cause a physical event; therefore God does not exist. Both this argument and POI rest on the thesis that causation can occur only between physical entities. But what grounds can there be for this believing in this thesis other than a commitment to physicalism? Surely there is no way that physicalism could be empirically verified. Objecting to (1) on the grounds that it conflicts with (3), then, (nearly) begs the question against the dualist.

Unfortunately, the dualist runs the risk of making the same mistake, and begging the question against the physicalist, if she rejects (3) simply because it conflicts with her dualism. So we might ask, are there any paradigmatically neutral reasons to reject (3), reasons which don’t simply assume the truth of (1) and falsehood of (3)? I think there are.

Reason #1: phenomenal awareness. We are directly aware of what it is like for our mind to move our body. We desire to type a letter and our hand moves, we decided to go to the coffee machine and our legs move, we want to whistle a tune, and our lungs and lips move. Generally speaking our mind moves our body in a rather simple manner. Commands in the form of barely perceptible intentional thoughts towards the body are issued and action (usually) follows immediately. It seems like there is no physical apparatus by which the mind directs the body. The bodies response to our mental commands “feels” instantaneous and unobstructed, seldom requiring any exertion (except perhaps in cases where the body is impaired). Of course, such awareness is not in itself enough to establish that the mind is non-physical. It merely establishes that, whatever the mind is, it is causally powerful, and its relation to the body “feels” simple and free of physical complication.

Reason #2: Non-physical causation. Jonathan Schaffer, David Lewis, and Charlie Martin have argued, fairly convincingly, that the connection view of causation (a thesis which (3) rests upon) is simply false. As Schaffer says “there are many ways to wire a causal mechanism,” and, one might add, not all of them require connection. Consider a mechanism like the following: a radioactive substance is placed in a chamber with sensors designed to record each instance of an isotope peeling off the substance and passing through the sensor (and out of the chamber). When the supply of isotopes runs out (when the sensor ceases recording hits at the usual rate), an explosive device is triggered, and an explosion results. In this case it seems perfectly correct to say that the bomb’s detonation is caused by the absence of radioisotopes in the chamber. But that absence is not physical. [1] if this is true, then, contra POI (3) it is possible to have causes that are non-physical.

Reason #3: Theism. Those who wish to hold that theism is even possibly true, should not be inclined to except POI (3). As Merricks’ argument showed us, to except POI (3) gives one reason to deny the possibility of God’s existence. But surely no theist would deny this possibility. So clearly no theist (whether dualist or materialist) should accept three. Furthermore, no agnostic should accept three. Most agnostics believe that there is, at the very least, some possibility that God exists. But if there are no non-physical causes, then God must not exist, therefore there is no possibility that God exists. If there is a possibility that God exists then there must be a correlating possibility that there are non-physical causes. Thus POI (3) is simply too strong even for the agnostic.

So these are three reasons why I think POI (3) should be rejected. None of these reasons begs the question against physicalism, nor does any assume that dualism is true.

[1] Most physicalists will deny that are such things as absences, preferring instead to explain absences away by appeal to second order states of affairs or facts about the world. This is a long and labyrinthine literature, suffice it to say that there are plenty who believe in non-physical causation (or causation by absence), and they do not do so based on a desire to champion dualism.