Jerry Fodor’s causal theory of content is a well-known naturalistic attempt purporting to show that Brentano was wrong in supposing that physical states cannot possess meaning and reference. Fodor’s theory contains two crucial elements: one is a notion of “asymmetric dependence between nomic relations,” and the other is an assumption about the nature of the “causally operative properties” involved in the causation of mental tokens. Having dealt elsewhere with the problems Fodor’s notion of asymmetric dependence poses, we show in this paper a difficulty with the other element of his theory concerning what kinds of properties are the causally operative ones in the tokenings of a semantic symbol in the brain of a perceiver. After presenting this difficulty, we examine three possible responses a Fodorian might make to our criticism.