Free will does not mean actions are uncaused. It means they are caused by persons. The popular notion that persons can be reduced to mechanistic neurological processes is self-refuting—if true, its advocates are uttering not words but meaningless sounds. The “causes” of actions, Thomas Szasz reminds us, are called reasons. But those cannot be reduced to brain activity, however much they may depend on it.
I don't want to get into get questions of reductionism. I just want to get into Richman's unmotivated question-begging assumption:
There is simply no reason whatsoever to accept this proposition unless one has already accepted that determinism (or the thesis of mechanistic causation) is false.
We have just about as much reason to believe this:
But what do we know to be true? Words mean things. Yes, they do! I'm meaning things right now, if you know what I'm saying. And you do! So, words, not just sounds (or pixels, or whatever.) Great.
And this is even relevant to questions of causation HOW?
What don't we know? Whether or not determinism is true.
But we know that it is possible to meaningfully communicate. So, if determinism is true, we know that determinism is compatible with meaningful communication. If indetermism is true, then we know that indeterminism is compatible with meaningful communication. We cannot, however, know a priori that either is incompatible with meaningful communication.
Richman is right. Persons cause actions. But he is wrong to assert that persons are not some deterministically governed physical fact about the world under a different description. Because he just doesn't know that. And he is wrong to think that it matters. Questions about personhood, agency and responsibility have nothing much to do with questions about reductionism or the ultimate nature of causation.
We know that our words mean things, and we know that under the right conditions we cause our own behavior and are responsible for what we do. However, though we cause our own behavior when a gun is to our head, we are not fully responsible for it. And if a drug is messing with the neurological conditions necessary for normal deliberation and choice, then responsibility is mitigated, whether or not behavior is fully reducible to neurological events. Similarly, if certain kinds of brain disorders undercut the neurological underpinnings of normal deliberation and choice, then responsibility may be mitigated.
The question of whether I am responsible when coerced does not turn on deeper metaphysical questions. And neither does the question of whether I am responsible if I have a brain lesion that short-circuits that conditions for normal agency.
I agree with Richman that free-will is self-evident. I deny that the self-evidence or experience of excercising free-will carries with it any information about the truth of reductionism or determinism, or that it tells us very much about the conditions for ascribing responsibility.