This paper argues that the position on free will which is defended in ‘Freedom: An Impossible Reality’ is not, as Tallis claims, a compatibilist view, but actually a version of libertarianism. While endorsing many aspects of that libertarian view itself, the paper raises questions about how one of the central arguments for Tallis’s view is supposed to work, and queries whether it really follows from the fact that we need to stand apart from nature in a certain sense, in order (...) to develop the kind of abstract knowledge that is constituted by the body of scientific law, that our own actions are not mere manifestations of what Tallis calls the ‘habits of nature’. It is also suggested that while a strong case can be made for many varieties of human exceptionalism, Tallis’s view of animal behaviour may be too simple and that there are examples of animal agency which cannot be explained merely by the associative learning which appears to be the highest grade of animal cognition that Tallis countenances. (shrink)