In her (2004) Varieties of Meaning, Ruth Millikan makes the claim that “no interesting theoretical line can be drawn” between biological purposes and intentional purposes. I argue that, contrary to her view, there are some interesting lines to be drawn. It is plausible that both intentions and the neural mechanisms that lie behind them have proper functions, but this does not license the inference that intentions are purposeful only because of their proper biological function. I use the proximate/ultimate distinction to argue that agents’ intentions are proximately purposeful, while their neural substrates are ultimately purposeful, and therefore that the former are not reducible to the latter, even if one adopts Millikan’s account of derived proper functions.
