What semantic priming experiments tell us about aspectual features: cross-linguistic comparison

Abstract

This abstract reports and contrasts two psycholinguistic studies of durativity and resultativity in Italian and Russian. Unlike most experimental research, focused on ontological properties of aspectual classes of predicates (Todorova et al. 2000, Gennari & Poeppel 2003, Pylkkänen & McElree 2006, Bott 2008), we adopt feature-based approach to verbal semantics. Semantic priming as “improvement in speed or accuracy to respond to a stimulus, when it is preceded by a semantically related stimulus” (McNamara 2005) is used to show that a given feature is present in semantic representation of lexical meaning. We followed the general design in Bonnotte (2008), who tested French achievement and activity verbs. The author concluded that only positively-valued features benefit from priming: activities and achievements facilitated the processing of activities in the durativity task (where targets denoting durable situations had to be detected), whereas in the resultativity task (where situations with a clear outcome had to be detected) priming was displayed on achievements with achievement primes. Thus, durative verbs yield priming more easily than resultatives, presumably because durativity is a continuous feature and durativity a binary one. In our first study, Italian achievement and activity infinitives were used. The following pattern emerged: achievements were primed in both tasks and activities in the durativity task, always with activity primes. The only feature tested in Russian was resultativity; the data included simple imperfectives, and resultative and delimitative perfectives. Some interesting differences with respect to Italian were observed. For instance, facilitating priming was detected on activities only, confirming Bonnotte’s conclusions on their contextual adaptability. Furthermore, although the decision latencies are comparable, accuracy was higher in Russian: 0.93 vs. 0.86. The grammaticalized nature of Russian aspect and morphological cues (prefixes) certainly facilitated the task. Addition of tense morphology had similar effects in Italian: imperfect speeded up the processing of activities, and perfect the processing of achievements. Russian delimitatives (e.g. po-rabotat’ ‘work for a while’) yielded revealing findings, since they are bounded but atelic and thus represent an in-between category. The proportion of yes/no answers (0.43/0.57) in resultativity task suggests that the speakers place them between both domains, closer to activities than to resultatives. This supports the differentiation between boundedness and telicity from theoretical and behavioural perspective. Pairwise comparison of these results (and others still in progress) can help providing meaningful generalizations on the nature of durativity and resultativity in Romance and Slavic, and also verifying their empirical status, at the core of much theoretical work on aspect (Dölling et al. 2008).

Alessandra Zarcone

Professor of Language Technologies and Cognitive Assistants

Computational linguist with a background in NLP and in psycholinguistics, working on AI, NLP and human-machine interaction.