(Glen) As Skinner was fond of saying, the problem with "mental" explanations is not that they are mental, but that they are not explanations. (Bruce) Depends, doesn't it? Reasons work. "Why did you fall asleep? " "Because I wanted to rest." But a reason may not explain. "Hold on, you fell asleep while driving." To explain this behavior, we need a cause. (Gerardo) That´s exactly the problem. Reasons and causes are part of different language games. For knowing a reason, it´s enough a sincere avowal of the person, but that´s not enough for knowing a cause. For knowing a cause, we need the empirical testing of regularities between events, but that´s not necessary for knowing a reason. Wittgenstein criticized Freud for his confusion of reasons and causes: Freud may convince his patient with his new proposals of unconscious reasons of symptoms or dreams, and that might even change the patient in many ways, but even then, that´s not a causal explanation. When Freud said for example that a strong superego was the cause of self-punishment, the criteria of "a strong superego" was precisely the presence of self-punishment, so the relation is conceptual and not causal. A similar problem can be seen in more recent cognitive therapy, when it´s proposed that "negative thoughts cause depression": depression is defined by tests (e.g. the BDI) which measure the presence of negative thoughts, so the relation is conceptual and not causal. Regards, Gerardo.