Abstract

Sentences with more than one logic operator are in principle ambiguous between different interpretations, depending on the assignment of logic scope. This aspect, largely overlooked in previous studies, could provide the right distinction that we need to capture children's early misinterpretation of epistemic negative sentences. Grounded on the notion of semantic strength (Horn 1989), the opposition between weak (e.g. might not) and strong (e.g. cannot) meanings will be introduced, in order to motivate the empirical prediction that an asymmetry between different kinds of negative sentences should be observable in language development. This idea will be explored in two new experiments on Standard Italian, an ideal testing ground given its extremely simplified modal paradigm. The results confirm the prediction that, at age five, children do not have problems in comprehending negative epistemic sentences in general. Experiments 1 and 2 show instead that difficulties are largely confined to weak negative sentences. An additional result, found in Experiment 1, is the difference between children and adults in positive sentences. This result is accounted for in terms of focus and a related assertion of exhaustivity. In Experiment 2, a contextual manipulation is introduced to test this hypothesis.