In Chapters 1, 6, and 7, we discussed reasons for thinking that a very standard model of mental representations should be separated frommuch of what is going on today in building a science of the mind. To be sure, there may be heuristic reasons to talk about what a neural signal is saying or to regard a burst of neural activity as about some episode in the environment. Nonetheless, such talk has a price to pay to be legitimate in a science of the mind. The price is high, and we have seen good reasons to think that current theories of content cannot pay it.