The role of verb type, noun phrase position and case marking in sentence comprehension in Spanish-speaking children

ANALÍ R. TABOH; CAROLINA A. GATTEI; DIEGO E. SHALOM

Bucarest (modalidad híbrida)

Congreso; 8th Bucharest Colloquium of Language Acquisition; 2023

Universidad de Bucarest

Understanding who does what to whom entails weighing both syntactic and semantic cues in order to assign thematic roles that suit the lexico-semantic structure of the verb. Such cues could be morphological case marking, noun phrase (NP) position, animacy, among others. There is evidence that children’s sentence comprehension is primarily based on a strategy that prioritizes NP order (Biran & Ruigendijk, 2015; Dittmar et al., 2008; Schipke et al., 2012) and that they are able to correctly use case marking cues over NP order rather late (e.g., around age seven in Dittmar et al., 2008). The aim of this work was to assess sentence comprehension and, in particular, the development of strategies for the integration of different types of information (i.e., morphological case marking, NP order, and verb type). Spanish object-experiencer psychological verbs offer an interesting opportunity for this because, while the canonical order of the language is subject-verb-object (SVO), their Experiencer surfaces as an object and they are more frequently used in object-verb-subject (OVS) order (Gutiérrez-Bravo, 2007).We designed a sentence comprehension task following the truth-value judgment paradigm (Crain & Thornton, 1998). We constructed semantically reversible sentences with activity verbs (e.g., gritar, to yell) and object-experiencer psychological verbs (e.g., gustar, to like) in SVO and OVS order. While the canonical order is SVO for activity verbs and OVS for object-experiencer psychological verbs, both types of verbs can be used in either order (see the Examples section). Two case marking cues were present in all sentences: the particle a preceding the object NP and a dative clitic, correferential with the object, preceding the verb. Participants were 162 typically-developing children from Buenos Aires, Argentina, aged three to nine years. In the task the participants heard a short story and were shown a picture, after which a puppet said a sentence to describe the situation (e.g., “¡A Burro le gusta Cerdo!”, Donkey likes Pig). Children had to indicate whether the sentence was correct or incorrect. The data were fitted with logistic mixed effects models to analyze the effects of age, verb type and NP order (canonical or non-canonical for that verb) on accuracy, controlling for maternal education and verb frequency.Comprehension was significantly better when the NP order matched the canonical order of the verb type (p