–Skinner boxes showed more promise to explain all human behavior than the most promising areas of AI research (neural nets, say) and yet the whole program vanished into Chomsky’s critique, many parts of which seem obvious in retrospect (e.g. learning language requires knowing meanings while stimulus-response mechanisms don’t). Sure, maybe it’ll be different this time.
-The information one gets out of a Skinner box is not so much about rats as about a way in which rats will distort their behavior in certain highly contrived and alien circumstances. We do not get insight into an isolated variable but an aberrant and confused response to aberrant and confused environments, even if the aberration is statistically predictable. The same can be said of training human beings to reach for their wallets when you tell them that something is “for a limited time” or “limit 10 per customer”.
-Treating something in the brain will not suffice by itself to cure all mental illness and neither will talk therapy even assuming that they could permanently eliminate all symptoms. If someone is depressed because they have no friends or because they are evil or mistaken about the value of life, you don’t cure them either by neutralizing the somatic element involved in the depression or by talking them out of it in a way that doesn’t address the actual cause of their depression – even though in both cases they will cease being depressed. Unlike other sorts of medicine, mental health cannot separate what is moral from what is a cure, and so the moral enters into the intrinsic practice of psychology while it is always relatively separate from other sorts of medicine. A guy will still be cured if you kill small children to eliminate his arthritis but he won’t be cured of a depression arising from having no friends by giving him a drug or convincing him that he doesn’t need friends.