“Successful theory must indeed explain the facts, but it would be madness to make it a constraint upon acceptable theory that it explain the ‘facts’ as they may currently be conceived by us… The ‘facts’, as currently conceived and observed by us, form the starting point for theoretical inquiry, but its successful pursuit may well reveal that we should vacate that starting place as hastily as possible. Large-scale intellectual progress will involve the wholesale rejection of old explananda as frequently as it involves the wholesale introduction of new explanantia.” ― Paul Churchland
I’ve come to think that disagreements surrounding whether or not free will is incompatible with determinism often hinge on more fundamental disagreements concerning the nature of conceptual revision. Take, for instance, the following common argument in favor of incompatibilism:
P1) If the existence of actions that actualize metaphysical possibilities is a necessary condition for free will, then incompatibilism is true.
P2) The existence of actions that actualize metaphysical possibilities is a necessary condition for free will.
C) Therefore, incompatibilism is true.
Premise one is uncontroversial. If determinism were true, then there would be no actions that satisfy the proposed necessary condition, and hence incompatibilism would follow. The truth of the second premise, however, is questionable.
For comparison, consider the now superseded biological theory of vitalism. Vitalism was an attempt to explain the difference between organic and inorganic matter. The theory claimed that while living things possess a vital spirit, non-living things do not. Put another way, the vitalist held that a necessary condition for life was the possession of a vital spirit, resulting in a strict dichotomy between living and non-living things. As the science of biology matured, it was eventually discovered that the vitalist concept of life was mistaken. It was realized that biological entities are composed out of the same kinds of elements as non-biological entities. The difference was not in the kinds of things that living and non-living things are, but rather the ways they are arranged. Unlike non-living things, the components within biological entities are suitable organized so as to perform certain functions (e.g., metabolism, maintaining homeostasis, developing, responding to stimuli, reproducing, etc.). This alternative functional conception of life, however, meant that the vitalist concept of life needed to be revised, replacing the necessary condition with a prototypical structure capable of accounting for the diversity of the functions involved, as well as the fact that each function can be exemplified to a greater or lesser extent within individual entities.
What I take the vitalist example to show is that obtaining an increased understanding of a phenomenon often requires revising concepts that are defined in terms of necessary and sufficient conditions into more nuanced concepts capable of accounting for a plurality of factors, each of which can be exemplified to a greater or lesser degree. Furthermore, such a realization would appear to have direct implications concerning the above argument. In the absence of a principled reason for thinking that the revision of the incompatibilist notion of free will is inherently illegitimate, I fail to see how premise two is defensible.
One possible incompatibilist response would be to assert that the comparison of incompatibilism with vitalism rests on a false analogy. While it may be true that the vitalist conception of life can be revised, the incompatibilist conception of free will cannot. This is because in the case of vitalism, once one comes to learn that the phenomenon of life can be conceived of functionally, one achieves a state of katanoesis, wherein which one intuitively apprehends that the proposed explanation has sufficiently explained the phenomenon. In the case of the incompatibilist concept of free will, however, no possible revisions exist which can enable katanoesis, and hence there are grounds for maintaining that no possible revisions are possible.
I think this response is misguided for several reasons. First, I fail to see why we should accept the existence of katanoesis as a criterion for legitimate explanations. Second, I think there is both historical and contemporary evidence that providing vitalists with evidence in favor of the functionalist conception of life regularly fails to generate katanoesis, and hence the suggestion that the comparison is disanalogous, for that reason, seems incorrect. And third, even if katanoesis were taken as a criterion for legitimate explanations, given that our empirical understanding of the mind is still in its relative infancy, it remains an open possibility that as more is learned about the structure and dynamics of volitional action that katanoesis will someday be achieved.