Behaviorism: Methodological approach in or philosophy of psychology, that insists one should use only evidence about the observable behavior of humans and animals and not about their mental states, except in so far as these can be distinguished by behavior.
As a methodological approach, this makes sense to the extent that it makes sense to do with fewer assumptions than with more.
As a philosophy of psychology it makes no sense, since everyone knows at least for oneself that one has experiences; that one's externally observable behaviour often holds clues to this; but that there is much that one experiences that others aren't aware of. By parity, this is true of all other human beings.
Besides, behaviorism seems to reduce human beings and animals to black boxes about which nothing can be said except such as is observable in their behavior, as if their internal states do not matter or must be fundamentally incomprehensible. Neither seems to me to be the case, though I agree it is easy to attribute internal states to an organism which it does not have. But this does not always happen, as anyone knows who loves any other person who loves one.
For a version of behaviorism in philosophy see Gilbert Ryle's "The Concept of Mind", and Russell's review of it in "My Philosophical Development".