The doctrine of eliminative materialism holds that belief-desire psychology is massively referentially disconnected. We claim, however, that it is not at all obvious what it means to be referentially (dis)connected. The two major accounts of reference both lead to serious difficulties for eliminativism: it seems that elimination is either impossible or omnipresent. We explore the idea that reference fixation is a much more local, partial, and context-dependent process than was supposed by the classical accounts. This pragmatic view suggests that elimination (...) is not the prime model for understanding the complex relations between the mind and brain sciences, and that we have little ground for concluding that in general psychological kinds do not exist. We suggest that reference changes are better seen as continuous rather than completely eliminative. (shrink)

The paper examines Davidson′s discussion of Frege on the problem of predication. Simple declarative sentences are unities that are true or false; how do predicates contribute to this kind of semantic unity? According to Davidson, the problem cannot be solved by assigning referents to predicates, since this leads to an infinite regress. Frege famously contributes the idea that predicates are “incomplete” or “unsaturated” functional expressions, mapping objects to truth-values. However, he takes predicates to refer to concepts and thus is exposed (...) to Davidson′s argument. I will show that Davidson is only right with regards to “Frege′s semantics”. Yet Frege′s logical doctrine of the decomposition of a sentence – I shall argue – allows for a resolution of this problem. (shrink)