[Implicit causality in language: criteria for selection of stimulus material in studies of verb causality].

Studies dealing with the implicit causality in verbs have shown that even minimal descriptions of interpersonal events (e.g. "Michael apologizes to Peter" or "Vera admires Karen") systematically elicit attributions toward the sentence subject or sentence object. However, in the majority of existing studies, the stimulus materials (i.e., interpersonal verbs) have not been selected randomly: Verbs were selected either because they had often been used in previous studies, or they were counterbalanced with regard to a number of additional criteria (valence, derivational form, etc.), and therefore, a truly random sampling of stimulus verbs were impossible. In the present study, the criteria for selecting interpersonal verbs are varied in order to compare two groups of verbs, namely, verbs which have been used very often in previous studies versus a random sample of interpersonal verbs. It is shown that the classical findings concerning the perceived causes of interpersonal verbs are less pronounced for the random sample than for the non-random sample of interpersonal verbs. However, even for the random sample of verbs, an impressive amount of variance in causal attributions is explained by different verb types.