Mind: A Brief Introduction (John R. Searle) / quote 33 of 138
The argument from analogy lacks an independent check
According to the argument from analogy, I infer the existence of mental states in other people, by analogy with myself. Just as I observe a correlation of my own behavior with my mental states, so I can infer the presence of appropriate mental states in others when I observe their behavior. I have already pointed out the limitations of this form of argument. The problem is that in general with inferential knowledge there must be some independent check on the inference if the inference is to be valid. Thus for example, I might discover that a container is empty by banging on the container and inferring from the hollow sound that there is nothing in it, but this inferential form of knowledge only makes sense given the assumption that I could open up the container and look inside and thus noninferentially perceive that the container is empty. But in the case of knowledge of other minds there is no noninferential check on my inference from behavior to mental states, no way that I can look inside the container to see if there is something there.