Whilst this view is appealing, there are a number of crucial objections – some of which we looked at earlier.

First of all, we might argue that our own knowledge of our mental states is not based upon observing our own behaviour. I don’t say, “Oh look! I am bleeding! I must be in pain. I’d better cry.”

Secondly, it is difficult to translate all beliefs into sentences that do not rely on mental terminology. For instance, to say that someone is behaving in a particular way is not the same as saying that they have a certain belief. To say that “Jim is hitting a ball” is not to tell us much about his intentions or beliefs. But to say that “Jim contemptuously smashed the ball for 6 runs”, implies something going on mentally (Jim had a certain attitude to hitting the ball – i.e. contempt).

Another problem is related to the idea of Zombies or robots. In such imaginary cases, the behaviourist view would not give us any criteria for distinguishing them from “normal” humans. This is because there would be no qualia for humans, zombies or robots to possess. If you’re a behaviourist, this may be fine, but you are then left with the choice of admitting that zombies and robots have minds or that behaviourism does not solve the problem of other minds.