Discussion of 'Computer Simulation of a Model of Neurotic Defence Processes'

Empirical determination of the parameter values. The first question is whether there is any sense in designing a model for which as yet there exist no methods of measurement for assessing values of parameters and variables (anxiety, pleasure gain, cathexis, defence potentials, etc.). By this same argument, however, one can also question the whole metapsychology of psycho-analysis, which is verbally formulated, for there is no doubt that at the present time it can be tested only by the analyst's use of estimations of variables in the setting of the cure. Directly from the observations made in the analytical situation, we have deduced a system of variables which are normalized in relation to each other and thus admit of measurement. The advantage of this procedure is that all variables and relations occurring in the system can be formulated precisely and the system can be tested for its inner consistency. The simulation is not a 'real life' simulation. Rather it is concerned with a 'meta-model' of a complex kind in which, by means of simulation, a great number of different defence processes produced by variation of the parameter values may be represented.

The input values are first taken in approximated form exactly in the same way as the analyst uses his observation of the patient as information for his own thought process. Afterwards they are improved iteratively by continued simulation.

Scales from 0 to 1 were introduced as an ad hoc metric. We intend to obtain the values for the parameter by means of scaling techniques in the estimation of the values of certain variables during psychoanalytical interviews. Anyone applying metric procedures in this field is very well aware that the objectification of such assessments presents a highly complex problem, as yet not fully resolved. First, a model of measurement strategy for the assessments must be found (cf. Beckmann, 1968). Once satisfactory procedure