Let’s turn our attention now to the most interesting case: that of the rational animal (the human being). Human beings have the capacity not only to sense and to imagine but to understand. Understanding in this sense involves the capacity to grasp something universal, a concept that can be used to formulate general propositions and questions. A relatively simple question like, Are human beings warm-blooded? for example, is something that no non-rational animal can ask. That’s not to deny that a non-rational animal could learn by induction that human beings are warm-blooded. They are capable of learning the corresponding fact but not of proposing the question to themselves. Hence, they cannot direct their actions in such a way as to investigate the question. Only a rational animal can do that.
A related capacity of rational animals is that of making judgments about what is possible or impossible. An animal can learn a general fact, but it cannot learn that the fact is necessary. It could, in principle, learn something corresponding closely to the Pythagorean theorem, but it could never learn that the theorem is necessarily true (in any Euclidean space). To put the matter more simply, the animal could learn that there is a point between any two points, but it could never learn the necessity of this fact.
Timothy Williamson has recently argued that our modal knowledge is a by-product of our knowledge of subjunctive or counterfactual conditionals. We can define possibility, necessity, and impossibility in terms of such conditionals.
Proposition p is possible iff (p might be the case if p were the case).
Proposition p is impossible iff (p would not be the case even if p were the case).
And proposition p is necessary iff (p would be the case even if p were not the case).[1]
Non-rational animals can learn the truth of subjunctive conditionals, so why can’t they gain modal knowledge?
I would respond that non-rational animals can gain true modal opinions, but they are incapable of modal knowledge. Animal learning is not sufficiently fine-grained to distinguish between robust but contingent generalization and strictly necessary generalizations. Hence, the subjunctive conditionals that they can come to know are suitably hedged. They can know something like: if p were the case and things were to continue to be generally as they have been, then q would be the case. We cannot define metaphysical possibility or necessity by the Williamsonian method in terms of such hedged conditionals.
How do rational animals gain modal knowledge? By extracting essences from the sensory data and manipulating those essences in acts of experimental cogitation (thought experiments). How do I know that, necessarily, all trilaterals have three internal angles? I can grasp trilaterality as such, and I can experience firsthand the impossibility of constructing a trilateral with more or fewer than three internal angles. This is clearly something that I do as a whole most fundamentally. Parts of my brain may well be suborned into participating in the mental process, but the brain process has the meaning and significance it does by virtue of its participation in the larger whole. And it is my rational soul that is ultimately responsible for the potentiality of my so acting.
Both Aristotle and Aquinas held that this power of abstracting universals enables my soul to be something “separate” from the body. This separation argument, unlike Descartes’ conceivability argument, has nothing to do with consciousness in general. For the hylomorphist, sensory consciousness is essentially an embodied process. It makes no sense to suppose that an immaterial entity could sense a color or a smell. Vision, smell, and the other senses are essentially material processes linking an organism to its environment. And sensory imagination is conceptually tied to sensation. To imagine a sensation is to put oneself in a state relevantly similar to that of being-appeared-to sensorily.
This is not to say that hylomorphism is compatible with physicalism. Conscious states of sensation are certainly not identical to any physical state, if by that we mean a state describable wholly in microphysical terms. Not even inorganic substances are merely physical objects. To say that something is an immaterial substance is to say something quite a bit stronger than to say that it is essentially non-physical. It’s to say that the substance has no spatially-circumscribed, prime-matter-informing parts.
Acts of pure understanding are not tied in the same way to embodied processes as sensations are. There is nothing absurd about supposing that an entirely immaterial substance might understand universal concepts. So, to the extent that we can do so, we are in a state relevantly similar to the state of such an immaterial intellect. So, it is conceivable and so prime facie possible that we could exist in a similarly immaterial state.
Suppose that we did survive the death of the body. What would be the aftermath? On the interactionist model, the soul might continue to exist and might continue to bear certain purely intellectual accidents (like understanding what a triangle is). Could it engage in extended activity—like inference or recollection? Not if this involved any form of sensory imagination. Most of us (myself included) can engage in inference only by laboriously imagining spoken or written symbols or diagrams of some kind, but some very gifted people are able to “see” logical consequences of what they know without any such imagery, and such pure intellectual activity might be possible in a disembodied state.
Would I myself still exist after death? In life, my soul is a proper part of me (on the interactionist model) and so not identical to me. After death, only my soul survives. Is it then identical to me? There is a Dion/Theon problem here. Consider Tibbles the cat, who loses its tail at time T. Before T, there seems to have been a proper part of Tibbles consisting of all the parts except the tail. Call this Tibbles-Minus. Before T, Tibbles ≠ Tibbles-Minus. But after T, it seems that Tibbles = Tibbles-Minus. This result is inconsistent with the eternity of identity. The best solution is van Inwagen’s—there was no such thing at Tibbles-Minus before T. The parts other than Tibbles’ tail did not compose anything. But clearly van Inwagen’s solution is of no help here. My soul clearly exists before my death, and it is not now identical to me now. So, it can’t be identical to me after death. So, it seems that I cannot survive my death, even if my soul can. (This conclusion is usually labeled corruptionism. The contrary position is survivalism.)
One possible alternative to corruptionism would be to argue that the substance has as parts not only the soul and the body (and its parts) but also all of its accidents. Now, we can apply van Inwagen’s solution, denying that my soul and my accidents do not compose anything before my death. Since we are counting the substantial form as part of the substance, it is probably reasonable also to count the accidents as parts of the substance. This alternative would entail that the person can survive only so long as he has at least one (mental) accident.
What about the emanationist model? On this model, neither the form (soul) nor the accidents are parts of the substance, either before or after death. At death, the distinction between the soul and the substance should persist. Now the soul should emanate a partless, immaterial substance. This immaterial substance would now be the bearer of all of my (mental) accidents.
But isn’t my essence that of a rational animal? If so, how can I exist without being an animal? I agree—I can’t. But perhaps rational animals need not have bodies, in the sense of proper parts, and need not be related to any bits of prime matter. An immaterial animal is a strange concept, but so is the survival of death. (This is also a worry for the interactionist version of survivalism.)
Let’s turn finally to free will, Suppose that I have deliberated extensively and am now ready to make a decision, a decision that is undetermined by all causally prior facts. How is this supposed to work on an emanationist model? Presumably, I have two causal powers: one to produce change A in myself (resulting in behavior b(A)), and a second to produce change B in myself (resulting in behavior b(B)). It is causally undetermined which power I shall in fact exercise. Suppose I exercise the power to produce change A. Change A will (given the WPSP) necessarily be accompanied by appropriate changes in some of my material parts, presumably parts of my cerebral cortex. Empirically speaking, what is happening will be very similar to the story that would be given by the interactionist model. The difference lies only in the identification of the fundamental agent: the whole organism in one case (emanationist), and the soul in the other (interactionist).
Quantum mechanics can be of some help here. Let’s us a collapse model first. One can think of the free choice as being accompanied or realized by some collapse of one or more quantum wavefunctions associated with parts of the brain. The collapse is diachronically undetermined but not synchronically undirected. The whole person or the soul simultaneously determines how the collapse shall happen (not, of course, in those terms, but in terms of the intended result of the collapse).
The whole organism and its integral parts are constantly precipitating such collapses throughout the body and, for the most part, these collapses are not directed to any intended effect.
On a no-collapse but one-world interpretation, like Traveling Forms, we can suppose that the whole organism actualizes the branch that corresponds to the chosen alternative.
How does Born’s rule interact with free choice? Presumably, when I made my A/B choice, each choice was associated with some vector in the Hilbert space (or some superselection sector in the space of spaces), and Born’s rule should apply, assigning an objective probability to each choice as a function of the wave amplitude associated with each. Are such objective probabilities compatible with a truly free or autonomous will? Both answers (Yes and No) have some plausibility.
We could suppose that Born’s rule and the wave amplitudes correspond to some kind of internal inclination of the will. So long as the will is merely inclined and not necessitated to choose as it does, we can evade van Inwagen’s famous consequence argument for incompatibilism. In other words, it’s not clear that there is any incompatibility between free will and probabilistic pre-determination—much less clear than is the incompatibility between free will and pre-necessitation.
Suppose, however, that probabilistic causation is incompatible with free will. Then we would expect free-will collapses of the wavefunction to be an exception to Born’s rule. Well, for all we know, they are. We can infer inductively that Born’s rule applies in every case relevantly similar to those observed in the laboratory, but free-will-induced collapses are not relevantly similar.
There is one more loophole we might exploit. Suppose that every exercise of free will involves the nearly simultaneous collapse of a large number of distinct local wavefunctions (I’m assuming here a collapse model, but the same point could be made by means of a large number of near-simultaneous branching events.) Some of these combinations result in behavior b(A) and other in b(B). The will could affect the joint probabilities of the collapses in such a way as to make b(A) overwhelmingly likely while respecting the Born-rule probability of each individual collapse.
[1] Williamson’s definitions only work if we set aside per impossibile conditionals. I don’t think we should, but I’ll ignore this point arguendo.
The Analytic Thomist 