How closely should a theory of thought and behaviour reflect the perceptions of those who are its objects? We would have no qualms if a theory of canine psychology wandered very far from how dogs perceived themselves and the world. We suppose that dogs have no theoretical conception of themselves or of the world at all, so the question would barely arise. But if a psychologist theorizes about our minds, in psychological terms rather than in neurological terms, it seems that the theory ought to keep reasonably close to our own perception of ourselves. We think that we are formed and driven by experiences, thoughts and emotions. We expect the psychologist to draw on the vocabulary that we use in our folk psychology, and to use it in the same way, so that the conclusions make sense to the rest of us. New connections may be made. Trends of previously unnoticed significance may be highlighted, as when Stephen Pinker, in The Better Angels of our Nature, assembles a range of economic and social factors that help to explain the diminution in our level of violence. But there still seems to be an expectation that the theory will stay close to everyday understanding. Furthermore, this expectation is often met. Articles in Psychological Review, for all their talk of cognitive architecture and their reconceptualizations of phenomena, are still written in a language that is adjacent to the language of folk psychology.
What support does this expectation have? We should not impose any such constraint on the sciences in general. If we imposed it on physics and chemistry, expecting them to use languages that were adjacent to the language in which we describe the everyday objects around us, most of the progress that has been made in the past century would not have been made. It is hard to see how we could justify imposing it on psychology by saying that the discipline is about our minds, and that we know our own minds from the inside. The data we get from the inside may be distorted in all sorts of ways. Moreover, we can imagine aliens, without a sense of our own minds from the inside, devising a human psychology that would be very effective in predicting or explaining thought and behaviour, but to which we could not relate in the way that we can relate to folk psychology, because the aliens had not drawn on that folk psychology. One option seems to be left. Our discipline of psychology is still too immature to construct a full enough set of robust, contentful, concepts, using its internal resources. It still needs to piggy-back on the content that is supplied by folk psychology. So long as it needs to do that, a requirement to stay close to the language of folk psychology imposes a useful control, ensuring that theorizing does not go off on an undisciplined frolic. But any such dependence on folk psychology is not guaranteed to continue.