Abstract
After a brief discussion of levels of explanation in psychopharmacology, evidence is reviewed which suggests that the lithium ion may produce behavioural effects by an action on information processing mechanisms. The clinical implications of such a view are noted. It is suggested that chlorpromazine may also affect stimulus analysis. A model of affective disorders, based upon the idea of stimulus analysis dysfunction, is then elaborated and suggestions are put forward concerning the relationships between different types of psychoactive drugs.