Ability

In the last post I mentioned the need for an analysis of ability in order to tell whether my account of the freedom required for moral responsibility is compatible with determinism. Here, I am going to take a stab at providing such an analysis.

First off, it seems obvious to me that, if I have the ability to do something, then it is possible for me to do it. Some will likely disagree with me about this, and many will likely claim that there is more to ability than the possibility of acting in a certain manner, but it is at least a reasonable starting point for our investigation. I am not committed to this being a complete analysis of ability, but it seems unlikely that any other part of an analysis would conflict with determinism if this part does not.

The puzzle which arises with an analysis of ability in terms of possibility is as to the kind of possibility. A compatibilist, of course, must deny that it is possibility in the sense of being compatible with a proposition stating how things were in the distant past together with the laws of nature (or stating whatever might be determinative of whatever happens). Jack Spencer at MIT has argued that ability needs to be understand in a manner compatible with our having the ability to do things which are logically impossible, and I find his argument compelling (Able to Do the Impossible, Mind (2017) 126: 465-96). So the possibility in question cannot be one which entails even such a weak kind of possibility as logical possibility.

Abilities are possessed by individuals. Further, such abilities can be carried around from situation to situation. I thus propose that we understand ability to phi as holding with respect to an agent when the agent, taken as she is at the time she is said to have the ability to phi, possesses the properties in virtue of which some individuals who actually phi, do so. Or, better, to decouple abilities completely from their exercise, we could better say that an agent possesses the ability to phi when that agent possesses the properties which are required for an agent to phi.

This need not be the only case where an agent can be said to have an ability. For instance, some abilities may be built up from lower-level abilities. If A has the ability to teach something in general, and B has the ability to learn it in general, then A, ceteris paribus, has the ability to teach it to B, even if no one has ever taught it to B. The ceteris paribus clause is required in this case to deal with cases where the learning style of B conflicts with the teaching style of A, such that A can teach a topic to students in general, but not to B, because A lacks the ability to teach the topic to students of the sort which B is--but this would be clear in a more fine-grained account of the abilities of A and B, which would still not require A's ability to include the ability to teach a topic to B in particular.

Jack Spencer argues that there are cases where there are necessarily unexercised abilities. That is, that it is possible for someone to have an ability in a world where no one exercises the ability. His example involves a lonely genius in a deterministic world who could, but never does, discover what things were like in the distant past and the complete description of the laws of nature. Again, however, by analyzing ability at the level of the properties required of an agent to phi, we can simply note that the lonely genius has all the required properties himself, and simply is never caused to utilize those properties.

The difficulty for my view is to say what keeps an agent from trivially having the ability to go 10 feet into the air simply because he has the properties required to do so in the event that a giant gust of wind blew him up into the air. There are a couple ways we could solve this problem. In the given case, we might require that the mode of ascent be specified. The agent lacks the power to propel himself 10 feet into the air under his own power, even if he contributes to the height he ascends with the gust of wind. Alternatively, we might specify that abilities only count under normal conditions. Thus, the agent in this example only has the ability to ascend 10 feet if gusts of wind of this kind are normal. I suspect that the latter better tracks the way we talk about abilities, while the former is more precise for metaphysics. To put it another way: when we talk about abilities, we mean abilities under normal conditions, and so it is the context of what counts as normal which specifies which ability in the more precise sense we are talking about.