The conceptualist argument

Even if spirits/souls do not exist, it seems that we have an understanding of what it would be for such things to be real. Thus, we can understand what it would be for demons or angels to exist, Cartesian egos to obtain in some possible world, and God to be a spirit. In light of this, I shall advance a type of conceptualist argument for substance dualism – minimally, the view that we are spiritual substances that have bodies – based on the understandability of what it would be for something to be a spirit, e.g. what it would be for God to be a spirit. After presenting the argument formally, I shall clarify and defend its various premises with a special focus on what I take to be the most controversial one.

The conceptualist argument

Here's the argument:

1. (1) If a person S understands what it is for something to be an entity (or purported entity) e, then S has an adequate concept of e.

2. (2) If S has an adequate concept of e, then S has a distinct positive concept of e.

3. (3) Therefore, if a person S understands what it is for something to be an entity (or purported entity) e (e.g. a divine spirit), then S has a distinct positive concept of e.

4. (4) We understand what it is for God to be a divine spirit.

5. (5) Therefore, we have a distinct positive concept of God's being a divine spirit.

6. (6) If thinking (i.e. conscious) matter is metaphysically possible, it is not the case that we have a distinct positive concept of God's being a divine spirit.

7. (7) Therefore, it is not the case that thinking (i.e. conscious) matter is metaphysically possible.

8. (8) We are either thinking (i.e. conscious) matter or thinking (i.e. conscious) spirits.

9. (9) Therefore, we are thinking (i.e. conscious) spirits.

Premises (1), (2), (4), and (8)

While premises (1), (2), (4), (6), and (8) are vulnerable to attack, in my opinion, (6) is the most controversial of the bunch, so I shall give special attention to it in the next section.

Regarding (1), in ordinary language, to understand something involves comprehending, grasping, or apprehending adequately its essential characteristics. For my purposes, I think the ordinary language usage will suffice. I have a view of the proper ontological analysis of a concept and the mind's possession of a concept, but, again, I do not think that delving into such intricacies is necessary for my purposes. In affirming (1), I am making the point that understanding is a matter of having an ‘adequate’ concept. In this sense, understanding cannot be reduced to linguistic know-how. Thus, contrary to Paul Churchland, if a person had never experienced pain and had no concept of pain, but could use the word ‘pain’ to predict and explain other people's behaviours, then that person would not understand pain (Churchland (Reference Churchland1984), 53, 70–72). And in (1), I am not talking about reference. Specifically, I am not claiming that concept possession is a necessary condition for successful referring or that one can refer by way of a concept only if that concept is adequate. Instead, I am talking about having an understanding of what it is to be some entity or purported entity.

In (2), I offer a necessary condition for being an adequate concept, a condition that may fairly be taken as a cousin of Descartes' notion of an idea's distinctness. By affirming (2) I mean to say, among other things, that if one has an adequate concept of, say, a type of entity (e.g. being an animal, being a mammal, being a dog), this entails that one has a positive concept of the features unique to the entities that fall within the extension of that concept. One way to illustrate this is by way of the notion of a species-concept being a concept of a genus and differentia. Thus, I do not have a distinct positive concept of being a dog if I possess only the concept of being a mammal (or of being a mammal if I possess only the concept of being an animal). That latter concept fails to pick out features unique to dogs. To have a distinct positive concept of being a dog I would need to have a positive concept of the differentia, of that which marks out dogs uniquely and sets them apart from cats and other mammals. If I am right about (1) and (2), then while mediaeval people had a surface concept of and could refer to water, they did not understand what it is to be water.

It is possible to distinguish between a concept and the way one possesses it. To understand this distinction, note that the property of running can be possessed in different ways or modes (running quickly or slowly). Similarly, the property of being-an-appearing-of-red can be possessed by a subject in different ways (clearly, fuzzily). In the same way, there can be different modes of possession for concepts (e.g. partially, fully). Now it may well be that in acts of employing a concept to refer successfully to the members of the concept's extension, it is individually necessary and jointly sufficient to have the relevant adequate concept and an appropriate mode of possession (e.g. fully), or it may be that these are INUS conditions for such acts. For my purposes, I am merely claiming that a necessary condition for being an adequate concept is that the concept be distinct. Whatever else one wants to say about a concept's mode of possession, distinctness is a necessary condition for understanding the concept according to (2).

Against (2), it could be argued that when it comes to polar concepts, e.g. real/unreal, true/false, good/evil, male/female, one need not have a distinct positive concept of one member of the pair to have an adequate concept of it, that spirit/matter fits this pattern, and, thus, that we need only a distinct positive concept of matter and not of spirit.

As a response, let us first distinguish two cases. First, there are cases in which one member of the pair is ontologically and conceptually basic and the other is a privation of some sort or another (real/unreal, true/false, good/evil). In such cases, I think there is a distinct positive concept of the basic entity and not of the privation. But it seems to me that unreal, false, evil, and their ilk are not positive realities in their own right; they are mere absences, and thus we do not need a distinct positive concept of them to have an adequate concept of them. However, for any real (or purported) existent like a dog or a spirit, it cannot be a mere privation. Ontologically, it must possess positive properties, parts, and so on; conceptually, to be an adequate concept of such an entity, there must be a distinct positive concept of those positive features unique to it. A spirit is not just the absence of matter; the number two satisfies this privation condition but it is not a spirit.

Second, male/female are correlative notions in which one is not basic and the other a privation. Could ‘male’ be ‘a non-female mammalian human person’? I don't think so. Ontologically, there must be something positive to a male qua male that makes him such; being a male is not merely the absence of being a female. Conceptually, to be an adequate concept of what it is to be a male, there must be a distinct positive concept of being a male. In grasping ‘being a male’, it would not be sufficient simply to grasp ‘being a non-female mammalian human person’ even if this description could be successfully used to refer to all and only the relevant objects. Now a spirit is not merely an entity that is not-physical. It is not a mere privation. The relationship between matter/spirit is more like that between male/female than real/unreal. Thus, there needs to be a distinct positive concept of matter and spirit for there to be an adequate concept of each.

Regarding (4), while some philosophers may think that there are certain problems with the classic Western theistic concept of God, very few have identified as problematic the notion of God as a spirit (or the notion of demons or angels). Yet, prima facie, there is a problem here, and it has been identified by Colin McGinn (Reference McGinn1997, 23):

There are two reasons why ‘immaterial substance’ is not adequate to capture the notion of a spirit, and McGinn puts his finger on one of them: ‘immaterial substance’ is purely negative and leaves us needing a positive characterization of ‘spirit’. Second, while certain entities, e.g. numbers, may not be substances, it seems intelligible to take them as such. If this is so, then we need a way to distinguish a spirit from a number construed as a substance, and ‘immaterial substance’ is obviously not sufficient for this task. It seems to me that the way out of these two inadequacies, and the way to proffer a distinct concept of spirit, is to characterize a spirit as a substance that essentially exemplifies an appropriate range of (actual or potential) properties constitutive of consciousness, e.g. being a thought, belief, sensation, desire, volition.

This is what we understand when we consider a possible world with Cartesian egos, demons, or angels, or when we construe God as a spirit. Thus, Joshua Hoffman and Gary S. Rosenkrantz define a soul – a notion they take to be coherent such that souls conceivably exist – as ‘an unlocated substance which is capable of consciousness’ (Hoffman & Rosenkrantz (Reference Hoffman and Rosenkrantz1997), 5).Footnote ¹ Now, the notion of being unlocated is not an essential feature of the concept of a soul. Coherent notions of the soul have been developed according to which a soul is either unextended yet located or extended (and, therefore, divisible, yet not composed of separable parts) and located. More important for my purposes is to note that Hoffman and Rosenkrantz do not take it to be adequate to characterize the notion of a soul in strictly negative terms, and when it comes to providing a positive characterization, they rightly appeal to the features of consciousness.

It could be argued that before we understand what it is for angels, demons, or God to be a spirit, we already have to understand what it is for anything to be a spirit and we gain the latter understanding by beginning in our own case, e.g. by first-person direct awareness of our own self. We then reason by analogy or in some other way to an understanding of what it is for God to be a spirit. I am sympathetic with this point, but it would be question-begging to employ it here. So I set aside questions about the proper account of how we come to have a distinct concept of spirit in general or of God as spirit in particular. I merely assume that we do, in fact, have such a concept.

Finally, in (8) (we are either thinking matter or thinking spirits), let us grant for the sake of argument that the properties constitutive of consciousness are irreducible, uneliminable, genuinely mental properties. Premise (8) says that the two live options for most thinkers would be that the possessor of consciousness is a material object (e.g. the brain, a sub-region of the brain, the entire animal, an atomic simple), with conscious properties supervening on or in some other way attached to it, or a spiritual substance (for Thomists, an individuated substantial form with the powers of life and consciousness) of some kind or other. Should we add other disjuncts to (8)? Maybe. But if we do, this would be small comfort to physicalists. If my argument is successful, thinking matter is not possible, and the dialectic would then be among dualist views of various stripes and non-physicalist alternatives.

Premise (6)

As I mentioned above, (6) seems to me to be the most controversial premise of my argument, so I want to give it some focused attention in what follows. According to (6), whenever someone allows for the possibility that there is such a thing as thinking matter and uses mental properties to characterize matter, he will have no way to give distinct, positive content to the notion of immaterial spirit. By claiming that thinking matter is possible, it follows that the various mental properties of consciousness – sensation, other forms of awareness, thought, belief, desire, volitional choices done for the sake of ends – could characterize material substances. If this is so, then granting the reality of immaterial spirits for the sake of argument, one cannot use mental properties to characterize the nature of a spirit, since those properties are consistent with both a spirit and a material substance. To get at the nature of spirit, one must look elsewhere, but here is where the difficulty lies.

To probe the problem more deeply, let us begin with an analysis of the notion of a substance.Footnote ² In general, a substance is a primitive, underived unity of actual and potential properties or parts at a time, it sustains absolute sameness through accidental change, and it has an essence that answers the most fundamental question ‘What kind of thing is this?’ and that grounds its membership in its infima species.

Note that the characterization just given is a formal one, that is, it merely states necessary conditions for any thing to count as a substance without giving material content necessary to distinguish one kind of substance from another. When it comes to characterizing specific kinds of substances, it seems that there are only two things one can do: state the formal criteria and list a set of positively characterized, distinct dispositions or potentialities as the material content for a specific kind of substance. For example, a characterization of an atom of gold as a substance might go something like this: it is a primitive, underived unity of actual and potential properties or parts at a time (having the power to melt at such and such degrees, dissolving in aqua regia, having atomic number 79 which, in turn, is the potential to attract a certain number of electrons, to resist inertial changes to degree thus and so), it sustains absolute sameness through accidental change, and it has an essence (being gold) that answers the most fundamental question ‘What kind of thing is this?’ and that grounds its membership in its infima species (the natural kind gold).

Note that the formal aspects of the definition of gold are the same as those for a substance in general and the material content is listed in parentheses. Note also that the various properties listed in parentheses that give material content to ‘being gold’ so as to distinguish it from ‘being iron’ are dispositions or potentialities, properties that are actualized by a gold atom if certain conditions obtain (e.g. gold will dissolve in aqua regia if placed in it). Substances are more than their potentialities; indeed, while I don't have a knock-down argument for this, it seems reasonable to say that something can have potential properties only if it has actual properties: no potentiality without actuality! Unfortunately, it is very difficult to say what the essence of a substance is without making reference to its potentialities. Thus, we say that gold is that which has the potential to do x, y, and z; iron is that which has the potential to do f, g, and h; and so on.

By way of application, it is hard if not impossible to characterize adequately a specific kind of substance without reference to the various potentialities it has by its very nature and which distinguish its essence from the essence of other kinds of substances. But now a difficulty arises for those who think that thinking matter is possible: By taking the various potentialities for thought, feeling, sensation, and so forth to be consistent with being a material and a spiritual substance, the very notion of being a spirit is rendered vacuous, and surely this is too strong. Even if there are no spiritual substances, surely the claim that there are is understandable and filled with distinct content, yet the thinking matter thesis renders empty any attempt to give content to the notion of a spirit.

During the days of John Locke (who defended the thesis that thinking matter is possible), those who accepted the possibility of thinking matter had a terrible time giving any distinct content to the notion of a spirit. About all they could say was that it has bare being. But this is hardly informative and, in any case, it could be used to characterize a bare particular, the number two or any other ‘immaterial’ entity. Even if there happen to be no spirits, it is surely wrong to say the very notion of a human spirit or of God as a spirit is vacuous and only capable of formal characterization.

Most dualists would characterize a spirit as an immaterial (e.g. unextended, without solidity, and not composed of separable parts) substance with the formal characteristics of a substance in general and with the ultimate potentialities of sensation, emotion, thought, volition, belief, and so forth. Notice that the material content to being a spirit has a negative component (being without solidity, unextended). But that is not a sufficient condition to have an intelligible concept of spirit, especially one adequate to distinguish it from other entities that satisfy the negative component. And it is precisely the capacities for consciousness that provides that material content. But if those capacities are consistent with a material substance, then one cannot appeal to them to characterize distinctly a spirit any more than one could appeal to the features of being a mammal to provide a distinct concept of dogs versus cats.

At this point, suppose one were to raise the following objection: mammal is generic and that is why it will not do. But why cannot there be shared specific properties? Of course, they cannot both be characterized by all and only the same collection of positive constitutive properties, but that is not what is being suggested, for material substance possesses other positive constitutive properties. Granting that there are such shared specific properties and not shared properties at the level of genus, this does not rule out the possibility of a distinct positive conception of the various species concepts.Footnote ³

To illustrate the objection, consider a slice of the history of the development of our concept of the electron from the Thomson to the Bohr to the Quantum electron (electron_T, electron_B, and electron_Q, respectively). Let electrons_T have constitutive species-properties A–F, electrons_B have C–H, and electrons_Q have E–J. Here is a case in which we have a distinct positive concept of each type of electron where those concepts overlap at the level of species and not genus.

Does this objection work against my position? For three reasons, I don't think so. First, it is consistent with the counterargument's core claim, as illustrated by the electron example, that the shared properties, in this case E and F, are generic and not specific properties. In this case, being an electron is a genus concept and there are different species of electrons that obtain in different possible worlds. So the argument does not count as a counterexample to my use of the generic notion of being a mammal listed above.

Second, if one is not persuaded by my first argument on the grounds that there are cases where the shared conception is at the species level, then this point does not provide a good analogy with the spirit/thinking matter case. Why? Because this latter case is, indeed, a situation in which we have a common genus (consciousness) and two different species-concepts under it – thinking matter (i.e. material consciousness, consciousness modified in a physical sort of way so as to have a physical bearer) and spirit consciousness. The objector under consideration acknowledges that if there are two species under a common genus, they each must have unique, positive, constitutive properties that need to be grasped to have an adequate concept of the species. In this case, my argument goes through. And it is important to say that my argument is consistent with but does not require a strong doctrine of infimae species as a general metaphysical commitment. It only requires that this notion apply to the spirit/thinking matter case as I have argued above. If one rejects a strong doctrine of infimae species, then I could adjust the premises of my argument so that it applies only to those cases where this doctrine is applicable, and the spirit/thinking matter case is one of them.

Third, there is another disanalogy between my spirit/thinking matter case and the envisioned scenario of my interlocutor: in the case he/she envisions, there are still distinctive, positive attributes to which appeal can be made to provide a distinct, positive concept of each species. For the Thomson, Bohr, and Quantum electrons, those attributes would be A–B, G–H (relative to the Bohr electron) or C–D (relative to the Quantum electron), and I–J, respectively. But in the case of spirit, there are no such distinctive, positive attributes available to provide such a concept. ‘Spirit’ has no non-overlapping, positive, species-specific properties of its own, once we grant the possibility of thinking matter.Footnote ⁴

Returning to the main line of argument and by way of application, suppose one were to say that a mind is anything with the powers of consciousness, including, say, a brain with emergent conscious properties, and a spirit is an unextended mind. We are now in a position to see that such a notion of spirit is vacuous. Why? Because the distinctive component of the definition is ‘being unextended’. This is because the powers of consciousness are consistent with not being a spirit – e.g. being a conscious atomic simple – and, thus, they cannot be what constitutes a spirit and makes it unique. But when one says that God is a spirit, one does not mean merely that God is unextended. The distinct content of ‘spirit’ in such a case is precisely the fact that God possesses and is essentially characterized qua spirit by the powers of consciousness.

In summary, the admission of thinking matter renders the notion of a spirit materially vacuous and, therefore, inadequate to serve as a distinct concept of a spirit. If thinking matter is metaphysically possible, it is not the case that we have a distinct concept of God's being a spirit. If the properties of consciousness are consistent with a physical and ‘spiritual’ possessor, those properties cannot be employed in forming a distinct concept of spirit any more than the properties of being a mammal can be used to form a distinct concept of being a dog as opposed to being a cat, and one is left with attempts to characterize ‘spirit’ in purely negative or vacuous terms; these attempts do not provide what is necessary for a distinct concept of spirit, and this entails that we do not understand what it is for God to be a spirit. But we do understand what it is for God to be a spirit. So ‘thinking matter’ is metaphysically impossible. And, given (8), we turn out to be spirits.Footnote ⁵