INTERDAT is computer software which substitutes a human interviewer. INTERDAT asks questions and tries to understand the responses by attributing mental models to the interviewee. The correctness of these models is tested by forecasting the responses to new questions. INTERDAT has many possibilities to adapt its models till it reaches the desired degree of understanding. Technologically INTERDAT is an Artificial Intelligence programme which is written entirely in LISP and to our knowledge the first Al application in sociological research.

The article argues for a synthesis between analytical philosophy and social sciences as relevant and necessary. The motivation and framework of such a synthesis is outlined on the basis of a critical social science. The authors illuminate such a perspective negatively in a critique of empirical and theoretical sociology, then positively in a clarification of the critical standpoint. Four theses, two under each-aspect, are defended: 1. Concerning empirical social sciences Neither the quantitative nor the qualitative paradigm of empirical social science (...) is able to put forward adequate methods for social research. Instead, the development of reconstructive methods is proposed to combine the advantages and eliminate the disadvantages of the quantitative and qualitative paradigms. 2. Concerning theoretical sociology Macrosociological theories tend to resist empirical corrobation. Pure theoretical and philosophical justification abounds instead. In this situation the tools of analytic theory of science are proposed in order to clarify the necessary steps towards a further development of theories, which can be empirically tested. 3. Concerning the critique of society A critical social science must incorporate a theory of a just society in order to analyse social institutions in a normative way. In this context an ethical realist approach is offered which tries to fulfill two conditions for sociologically relevant normative reasonings: satisfaction of individual interests and the rational consensus of all persons concerned. 4. Concerning critique of ideology The tools of analytic philosophy can be given new application by combining them with an analysis of interests under the title of critique of ideology. (shrink)

Critical social science has to acknowledge that every fundamental critique of society implies the justification of alternative norms and institutions. Several current objections against such an explicitly normative understanding of critical social science are discussed. The following outline of a theory of a just society tries to meet two demands: the rational consensus of all individuals concerned and the satisfaction of individual interests. In societies characterized by class struggles, however, these two aims turn out to be incompatible. Therefore an ethical (...) realist approach is offered which takes into account the clarification and normative reinterpretation of interests. The tools of analytic philosophy can be given new application in combination with an analysis of interests under the title of a critique of ideology. (shrink)

Some principles of logic-oriented theories of argumentation are discussed. A sharp distinction is drawn between arguments in formal contexts and arguments in non formal contexts. It is argued, that the analysis of colloquial arguments cannot be based on the logic-oriented theories of argumentation. The outline of a more realistic theory of argumentation is given, by using defeasable inferences as a starting point. The model is applied to some everyday arguments.

Most texts on argumentation theory stress the importance of formal logic for the study of arguments. This paper raises some doubts about the usefulness of logic for the study of argumentation. In fact\ the basic analogy between logical proofs ana arguments in natural language does not seem to hold. There seems to be a basic circularity in everday arguments which cannot be reconstructed by the standard logical tools. Therefore we habe to look for some non-logical forms of representation. Some hints (...) for this new type of argument representation are given. (shrink)

Is causal analysis an appropriate method to discover causal laws? Some prerequisites for any successful application of mathematical causal models are worked out and discussed. It can be shown that these prerequisites cannot be fulfilled by present models and there are strong reasons to assume that they cannot be satisfied by any model. Therefore causal analysis will probably be of no importance for the discovery of causal laws in sociology.