The interaction between the person and his environment can vary between an open and a closed pole. In the behavior that might be characterised as open, emphasis is placed upon contact, mutual influencing and, in the extreme, merging between the person and his world. In the behavior that might be characterised as closed, autonomy as a separate entity and, ultimately, isolation between the person and his environment are particularly stressed. It is proposed to conceptualise this variability as a general dimension of openness versus closedness. From this concept, we derived the hypothesis that within each level of psychic adaptation (normal, neurotic and psychotic) open as well as closed forms of adaptation can be found. On the neurotic level, hysterical neurosis was assumed to be more open and obsessive-compulsive neurosis to be more closed. On the psychotic level nonparanoid schizophrenia was assumed to have more open characteristics and paranoid schizophrenia to have a more closed style. This hypothesis contains a rejection of

• the hypothesis of general openness of schizophrenia as an undifferentiated group by Van der Drift, by Federn and by Fisher and Cleveland (in their initial publications) and
• of the hypothesis of a direct relationship between pathology and openness by the same authors and between pathology and closedness by Rokeach and by Jourard.

Our hypothesis conceptualizes openness-closedness as a matter of style rather than of pathology indeed. The results show that, as expected, paranoid schizophrenics (N=30) and obsessive-compulsive neurotics (N=10) are significantly more closed than nonparanoid schizophrenics (N=30) and hysterical neurotics (N=30) on the Barrier minus Penetration score of Fisher and Cleveland (p<0.05) and on the self-disclosure questionnaire of Jourard (p<0.01), whereas differences on the Dogmatism scale (Rokeach) and on the storing systems of Zucker (Federn) and of Salome-Finkelstein (Van der Drift) were found to be statistically insignificant.