Theorievorming als formulering van ervaring

Samenvatting

(Uit nr. 6, 1960, p.299: In het artikel „Theorievorming als Formulering van Ervaring” , Soc. Gids no. 3/4 1960 van de hand van F. Bergsma zijn nog enkele storende zetfouten in de formules geslopen. De auteur zal geïnteresseerden op verzoek gaarne correcties verschaffen. Adres Joh. Vermeerlaan 34, Baarn, tel. 4458.) Apart from designing distribution methods and measuring techniques for statistical and mathematical theory construction, the methodology of sociology should attempt a logistical theory construction based on wholly qualitative data. Since qualitative case studies in sociology can be considered as descriptions of co-occurrences of class attributes or of positions in rank orders, they can be formulated as existential conjunctions of n one-place, or of n many-place predicates in synthetical propositions. For n class attributes which are m-chotomous there are mn combination-possibilities. For n>l rank-orders without tied rankings there are 2n-1 possibilities of combination, and for n>1 rank-orders with tied rankings there are 3n+ possibilities. A systematic search for those possibilities of combination which persistently fail in the data may result in negative existential conjunctions. These are always logically equivalent to positive general implications, valid for the data, but hyp'ly generalizable beyond them. The known logical paradoxes of material implication can be abolished by introducing a narrower conditional, namely the general implication of notalways-valid predicates by not-always-invalid ones. These conditionals are more in agreement with empirical conditionals than are material implications. This means, however, that they rule out all counterfactual ideal-types of the explanatory theories. The calculus of predicates can then be applied to the formularized empirical results of the case comparisons. This procedure gives a ready formulation of F. Znaniecki's 'analytical induction.' In social anthropology and historical sociology in particular, the comparative method is often limited to the analysis of qualitative variables of nonrandom macro-cases, selected for their adequacy. Therefore examples have been taken here from the folk Ur field, where qualitative multivariable conditionings embracing a change in time could easily be translated into the formulae of the logic of predicates, and were calculable by the laws of this discipline. Modified AA.