Reciprocal expressions and the maximal typicality hypothesis

In two experiments, we study the effects of verb concepts on the interpretation of reciprocal expressions in Dutch and Hebrew. One experiment studies Hebrew to test a previous account, the Strongest Meaning Hypothesis, which suggests that listeners resolve ambiguity in reciprocal sentences using the logically strongest meaning that is consistent with the context. The results challenge this proposal, as participants often adopt a weaker meaning than what the Strongest Meaning Hypothesis expects. We propose that these results reflect the sensitivity of reciprocal quantifiers to verb concepts, which is modelled by a new principle, the Maximal Typicality Hypothesis (MTH). For any given reciprocal sentence, the MTH specifies a core situation: the maximal situation that is also maximally typical for the verb concept. The MTH predicts reciprocal sentences to be maximally acceptable in the core situation and, under certain conditions, in situations that contain it, but substantially less acceptable in other situations. To test this prediction, we conducted a two-part experiment among Dutch speakers: (a) a membership test that ranks typicality preferences with different verbs; (b) a truth-value judgement test with reciprocal sentences containing these verbs. The results show that the typical number of patients per agent varies between verbs, with a significant effect of these preferences on reciprocal quantification: the stronger the verb concept's bias is for one-patient situations, the weaker is the interpretation of reciprocal sentences containing it. These results support the MTH as a basis for a general theory of reciprocal quantification.