Psychology appears to be captivated by a framework that forces one to think in contrary positions. Hence we attribute the causes of behavior to either individual or society; and, when we search causes in the individual, then either his mental structure would be involved or his bodily constitution. The basis of this framework is the tendency to conceive causal explanation as the reduction of phenomena to the laws or mechanisms of one particular level of organization. Such a reductionism unavoidingly fails when one strives for the explanation of complex behavior. The last dccade there is growing consensus that reductionism is unwise. However, the recognition alone that complex behavior is produced by interacting social, mental and physiological factors does not suffice. We need a feasible interactionistic approach, that provides directives for the construction of theories in which mechanisms of different levels gear into one another. Here some proposals will be discussed that might meet this need, illustrated by an attempt to explain anorexia nervosa.