1
Note: This is a slightly revised version of the published paper.
Substance Physicalism
How should we define 'physicalism' or 'minimal physicalism'? In my view this question calls for stipulation since these are theoretical terms without a uniform use. Different views of psychophysical relations are physicalistic in different ways and to different degrees, and there is an obvious interest in clarifying and distinguishing these views and determining which are true. My aim in this paper will be to do some of the clarifying and distinguishing. Stipulation of a unique thesis as physicalism or minimal physicalism must come with a rationale, and as I have none to offer I shall not pursue this.
Some regard physicalism as the thesis that all first-order properties instantiated in the spatiotemporal world are physical properties. I shall refer to this as type physicalism or property physicalism. There are variants of this thesis generated by variants in how a physical property is characterised in terms of microphysical properties. One approach requires a physical property to be coextensive with a microphysical property. Another approach requires (as a first approximation) a physical property to logically supervene on microphysical properties. On this second approach, property physicalism can be presented in the form of a supervenience thesis--another popular way of defining physicalism. But as I think that the following discussion will apply on any reasonable view of physical property, I will mostly continue to talk of physical properties without getting more specific. I shall assume that it will be clear enough to think of a microphysical property as an assignment of fundamental microphysical parameters in some type of spatial or spatiotemporal region, where the fundamental microphysical parameters are those featuring in an ultimate microphysical theory.
A few decades ago, physicalism would more often have been characterised as the supposedly weaker thesis of token physicalism or ontological physicalism--the view that every token or particular in the spatiotemporal world is a physical particular. Many people who are discontented with physicalism would not be content with a mere denial of property physicalism, but are opposed to ontological physicalism. I do not share this opposition, but I think it is an interesting question that has not been satisfactorily answered, whether a coherent thesis of ontological physicalism can be articulated that is genuinely weaker than property physicalism. And of all views about mind-body relations, I think it is ontological physicalism whose truth or falsity has major practical consequences for how people lead their lives.
There are many kinds of particular to consider in enquiring what it is for a particular to be physical, the most important of which are events, objects, substances, and states. I take events of a certain class and states to be property exemplifications. For such particulars, ontological physicalism reduces to property physicalism. But I do not think all particulars are property exemplifications. Many have held that there is an important notion of particular, sometimes referred to as a concrete particular, for which ontological physicalism amounts to a weaker claim than property physicalism, and it is ontological physicalism for such particulars that is the topic of investigation here. I focus, for the most part, on providing a characterisation of ontological physicalism for substances (including objects), which I shall call substance physicalism. Then I suggest that an analogous argument would yield an equivalent characterisation for concrete event physicalism. Many have thought substance dualism a mysterious, perhaps incoherent, doctrine and token physicalism for events to be clear by contrast. In this paper I argue that substance physicalism and concrete event physicalism are equivalent versions of ontological physicalism and are best understood as the view that the world is governed by laws of succession with purely physical antecedents.
To this end I'm going to focus on mental substances as the obvious candidates for nonphysical substances. (I shall often simply refer to mental substances as minds, and to physical substances as bodies.) And I shall seek a characterisation of a substance dualistic world, i.e. a world in which there are nonphysical minds and nonmental bodies. The account could also be adapted to provide a characterisation of a world containing vital substances or any other putative nonphysical substance.
The term 'substance' may seem antiquated and unsuitable for modern philosophical enquiry, so let me clarify my usage here. It is often said that a substance is capable of independent existence, and I shall take this to be an essential feature of substances insofar as it rules out properties, because they need something to inhere in, as well as nonbasic, i.e. supervenient, entities such as property exemplifications. A further feature of substances I shall take to be essential is that they are continuants, i.e. particulars that occupy spatial regions and endure through time, rather than events. However, the discussion can be adapted to apply also on a four-dimensional view of substances, by reinterpreting talk of three-dimensional substances, where necessary, as talk of time-slices of four-dimensional substances. On the account I'm offering, substances include physical objects, as well as identifiable bits of liquid and gas that wouldn't readily be described as objects. Human bodies, brains and nervous systems are thus examples of substances that are relevant to this enquiry, along with minds. But as minds are sometimes thought not to be occupants of spatial regions, I'm not going to take spatial location or extendedness to be a further essential feature of substances.
When Descartes refers to res extensa and res cogitans, he seems to have in mind a spatially extended physical substance, the totality of matter, and a nonspatially extended mental substance, the totality of minds. But he sometimes refers to individual minds as substances in discussing their relations to individual bodies. So Cartesian substance dualism can be formulated as the thesis that there are physical bodies and nonphysical minds, or as the thesis that there is a physical res extensa and a nonphysical res cogitans. These appear to be equivalent, for if res cogitans is nonphysical then the individual minds it comprises must be nonphysical,[1] and if every mind is nonphysical then the totality of those minds, res cogitans, must be nonphysical. So for the sake of convenience I shall ignore these big Cartesian substances, unless otherwise specified, and focus instead on individual minds and bodies.
What is a substance dualistic world? Let us consider first a popular way of characterising physical particulars and thereby ontological physicalism proposed by Geoffrey Hellman and Frank Thompson.[2] It is that every physical particular is one that satisfies a basic positive physical predicate at a place, or is a part or sum of such particulars. Those predicates are given by a list which, according to Hellman and Thompson, might include 'is a neutrino', and 'is a four-dimensional manifold'.
One problem with this is that it is hard to make sense of a sum of particulars of different spatiotemporal kinds. Hellman and Thompson include fields along with substance-like entities such as electrons. And one might also expect a fundamental theory to contain predicates such as 'is an alpha decay', 'is a measurement' satisfied by event-like entities. For this approach to have the best chance of succeeding it seems that a restriction must be made to particulars of some spatiotemporal kind, such as events or substances, in forming sums. But in both cases it would be necessary to show how a sum is formed when there are overlaps and gaps between particulars satisfying the designated predicates. And this would clearly be revisionary in the sense that not any sum of substances or events would commonly be regarded as a substance or event.
Furthermore, it may not be clear into what spatiotemporal category to classify the particulars satisfying the various predicates proposed even by current theories of fundamental physics. Does a neutrino, photon, or quark essentially occupy a spatiotemporal region, or is it essentially capable of enduring through time? It seems better to represent fundamental physics exclusively in terms of fields rather than as dealing with microscopic events and substances.
However, the main problem with this approach is that it fails to provide an analogous positive account of what it could mean for nonphysical substances to exist in a world alongside physical substances. Some might be inclined to say that a nonphysical substance is one made from some nonphysical stuff such as ectoplasm. But it seems to me that we have no idea how to characterise nonphysical stuff, and hence no idea how any nonphysical substance could be shown to be built up from such nonphysical stuff.
What positive characterisation can be offered of a nonphysical substance? If a mind is not spatially located then it is nonphysical. But I shall argue that we can make sense of a nonphysical mind being spatially located and extended, and hence should not define a nonphysical substance as lacking spatial location or extension.
In seeking an alternative approach, let us begin with the assumption that a substance is a bearer of properties that satisfy certain conditions. A physical substance bears certain intrinsic physical properties and is spatially located and extended within a spatial region. This spatial region must satisfy certain conditions that may not be precisely specifiable, such as having a recognisable boundary given purely conventionally or in terms of different types of material. Likewise, a mental substance bears certain intrinsic basic mental properties, such as having a certain (narrow content) belief or sensation, which satisfy certain holistic coherence conditions. In each case the conditions are hard if not impossible to specify, but I do not think such specification is necessary for the purpose at hand.
Now let us consider the options for conditions under which a mental substance would also be a nonphysical substance. It might be suggested that a substance is nonphysical iff none of its intrinsic properties are physical. But such a condition does not appear to be necessary, for it would seem that a substance could be nonphysical even if it has some intrinsic physical properties.[3]
Suppose, instead, we say that a substance is nonphysical iff at least one of its intrinsic properties is nonphysical. This strikes me as necessary for a substance to be nonphysical, and I do not think it would be altogether implausible to say that it is sufficient as well. On this view, if there are mental substances and any of their intrinsic properties are nonphysical, then there will be nonphysical substances. So property dualism would entail substance dualism provided that some nonphysical properties are properties of minds. This would accommodate the traditional view that one species of substance dualism is a kind of epiphenomenalism. However, it has also been thought that substance dualism is a significantly stronger thesis than property dualism, and an interestingly different one--a difference that cannot be accounted for by the minimal further condition that some nonphysical properties are properties of minds.
The only way to capture this strong view of substance dualism, as far as I can see, is to deny that having a nonphysical intrinsic property is sufficient for a substance to be nonphysical, and to add as a further condition for a substance to be nonphysical that it play a causal role (in virtue of the causal relevance of some of its nonphysical properties). I do not think we would be prepared to say that there are, or could be, any physical substances in our world that do not have any effects. In general, a substance must play a causal role. Thus I shall seek a characterisation of substance dualism which requires that in order for there to be nonphysical minds in a world in addition to physical substances, those nonphysical minds must also play a causal role in virtue of the causal relevance of some of their nonphysical properties. I shall be aiming at a fully general characterisation, i.e. one which covers worlds that quite obviously don't match ours.
Substance dualist theories are typically divided into interactionist, epiphenomenalist, and parallelist theories, though I shall be pointing out later that there is at least one further kind that fits into none of these categories. Interactive substance dualism is the most natural and popular version of substance dualism. It is the view that there is interaction between substances in both directions. Minds and bodies certainly appear to interact. There would be physical to mental interaction when events such as skin piercings cause sensations such as pains. And there would be mental to physical interaction when mental events of deciding to do something cause voluntary movements of one's body (in action), or when thoughts or perceptions give rise to involuntary bodily movements and secretions such as trembling and sweating (in expression of emotion).[4]
Epiphenomenalism is the view that there is physical to mental interaction, but that minds have no causal powers. As should now be clear, I think that it is better classified as a species of substance physicalism. The parallelist view is that minds and bodies do not interact in either direction, but unfold fully according to laws or principles of their own special sort. I shall give an argument later that this view is probably incoherent--an argument that tells equally against the view that the world contains substances which aren't involved in any causal relations at all.
However I do want to argue that interactive substance dualism is coherent, and to this end I want to respond to the well-known objection raised in the seventeenth century literature, most famously by Princess Elisabeth in her correspondence with Descartes, that interaction between different substances, and in particular mental to physical interaction, is inconceivable.[5] It is an objection that to this day finds adherents.
Now the question how substance to substance causation is conceivable can be raised about all four types of it: physical to physical, mental to mental, physical to mental, and mental to physical. A good answer for Descartes to have given Elisabeth would be to say that no further explanation of the intelligibility of substance to substance causation is needed than that there are fundamental laws or principles of succession in accordance with which such causation occurs. How is it conceivable that the earth keeps the moon in orbit? Answer: it accords with a fundamental physical law, in this case gravity. How is it conceivable that deliberation can lead to one's choosing a course of action? In response, the substance dualist may invoke mental to mental laws, or state that deliberation brings before the agent's mind some available options from which a selection is made by a libertarian free choice. It is only this libertarian case that I need to describe as governed by a mental to mental principle (for lack of a better term) to distinguish it from those governed by laws. (This explains my frequent reference to laws and principles.) How is it conceivable that bodies cause sensations? To answer this Descartes need do no more than point to the physical to mental laws that he invokes in his sixth Meditation, where he says that 'whenever [the brain] is disposed in a given way it gives the same indication to the mind.'[6] And finally to the case of special interest to Elisabeth -- how is it conceivable that minds cause body movements? The substance dualist should offer a similar account of the intelligibility of action and expression of emotion, this time invoking mental to physical laws. Indeed Descartes suggests that such an account can be offered in the case of emotion.[7] In sum, laws and principles of succession of these sorts would suffice to account for the intelligibility of any case of substance to substance causation, unless there were some further reason for thinking that they conflicted with other laws or logical truths, or were incoherent.[8]
Let us look in more detail at these four kinds of laws and principles. The physical to physical laws would state that when the fundamental physical parameters are distributed a certain way in someone's body, or more generally in a spatial region, then they will be distributed a certain other way after a specified time interval in that region. They are not fundamental physical laws, such as the inverse square law of gravitational attraction (or some more accurate law of gravitation). Rather, they are derived from such laws by taking as antecedent a physical description of a body or region of space and allowing the consequent to give a physical description of the body or region of space after a specified time interval in accordance with all the fundamental physical laws.[9] Such laws are derivable for any choice of specified time interval. Let us assume for the moment that they are deterministic. I will examine the consequences of relaxing this requirement later.
The physical to mental laws governing sensation would state that when a body or brain moves into a state of a certain kind, then after a specified time interval an associated mind experiences a sensation of a certain kind. To give these laws the best chance of matching our world, the specified time interval would presumably have to be less than one second. Again let us initially assume that such laws are deterministic. Descartes effectively offers such a law-based view of mental to physical interaction in the passage quoted above from his sixth Meditation. He refers to such laws to explain thirst and pain and would presumably do so for all types of sensation.
In the case of mental to physical interaction, the governing law would state, e.g., that whenever a mind wills[10] a certain kind of limb movement, then after a specified time interval an associated body moves into a state of a certain kind. To give the best chance of matching our world, the specified time interval would again have to be less than one second, and the kind of body state would not be the willed kind of limb movement, but rather an electro-chemical configuration of a certain kind in the brain which typically causes that limb movement through the workings of fundamental physical laws.[11] This is effectively just reversing Descartes' account of physical to mental interaction, as he suggests in the letter quoted above for the case of emotion. Again let us assume for the time being that such laws are deterministic. Having the body respond deterministically to the mental act of willing should not be held to compromise the freedom of the will, for the locus of freedom on any account must be in the mental process up to and including the mental act of willing.