The girl ate the apple, didn’t she? Assessing event culmination in Italian adult and child native speakers.

The present dissertation investigates the phenomenon of event (non-)culmination in Italian, focusing on how adult and child native speakers interpret telic-perfective sentences. Through a combination of theoretical discussion, experimental evidence, and acquisitional perspective, this work aims to clarify whether Italian speakers accept the non-culmination reading for telic-perfective sentences and whether culmination is part of their truth-conditional meaning (i.e., an entailment) or rather the result of a pragmatic inference. According to traditional analyses (Dowty, 1979; Krifka,1989, 1998; Tenny, 1994; inter alia), telic predicates such as ‘mangiare il panino’ (lit. to eat the sandwich) denote events with an inherent endpoint. When these predicates are combined with the perfective aspect, sentences are expected to entail that the event expressed by the predicate reached the culmination point, namely that the sandwich was entirely eaten. Nevertheless, cross-linguistic research has shown that this entailment does not always hold. Indeed, in many languages, telic-perfective sentences can be used to describe incomplete events, a phenomenon known as non-culmination. Whether Italian allows such a reading remains an open empirical question. Moreover, the absence of explicit morphological marking, for both lexical and grammatical aspects, makes Italian an opaque language with respect to event culmination and, therefore, an interesting empirical ground to test. A series of experimental studies was conducted, administering Truth-Value Judgment Tasks (TVJT) to both Italian adult and child native-speakers. First, adults’ interpretation of telic-perfective sentences was tested. Results revealed that Italian adults frequently accepted telic-perfective sentences as descriptions of partially completed events, suggesting that culmination is not always entailed, but can be cancelled, challenging the predictions of traditional analyses (Krifka, 1989; 1998). In addition, the verbs’ classification based on the Scalar Approach (Rappaport Hovav, 2008; 2014 - i.e, Punctual vs. Durative Change of State vs. Incremental Theme verbs) did not fully predict verbs’ behavior. Therefore, a cluster analysis was conducted based on participants’ acceptance rates of the Partial Result scenario. Four verb clusters emerged, capturing graded differences in how strongly predicates require culmination. A further experiment validated this data-driven classification, showing that culmination is not uniformly encoded across all telic predicates and revealing its gradient nature. The acquisitional perspective added a further dimension to our understanding of event culmination by investigating how Italian children, aged 4-to-7, interpret telic-perfective sentences. Children showed early sensitivity to the same distinctions as adults, even though more broadly. Indeed, the fine-grained distinction among verbs found in adult participants is not yet fully developed, even in first-year primary school children. Finally, in the last experiment, negation was used as a diagnostic tool to disentangle whether culmination constitutes a semantic entailment or a pragmatic implicature. Both adults and children were tested on positive (i.e., ‘ha mangiato il panino’) and negative (i.e., ‘non ha mangiato il panino’) telic-perfective sentences, used to describe Partial Result situations. Results showed an asymmetrical pattern: participants tended to accept positive telic-perfective sentences while rejecting their negative counterparts in the same context. This asymmetry supports the view that culmination is not semantically entailed, but rather a defeasible implied. In conclusion, the findings demonstrate that in Italian, the non-culminating reading is acceptable and that event culmination is not uniformly encoded across all telic predicates. Instead, accomplishment verbs are distributed along a continuum, with verbs differing in how strongly they require culmination.