Sentence disambiguation in the phonology-syntax-semantic interface

• Autor(a) principal: Othero, Gabriel de Ávila
• Data de Publicação: 2018
• Outros Autores: Teixeira, Mariana Terra
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Letras de Hoje (Online)
• Texto Completo: https://revistaseletronicas.pucrs.br/ojs/index.php/fale/article/view/28698
• topic: Syntactic ambiguity; Syntactic parsing; Phonology-syntax-semantics interface; Ambiguidade sintática; Processamento sintático; Interface fonologia-sintaxe-semântica; Psicolinguística

We start from the results presented by Angelo & Santos (2017) on the disambiguation of sentences based on prosodic clues to try to clarify an issue that was unsolved in their research. The authors applied a comprehension task regarding sentences with NP1-V-NP2-Attribute structure (e.g. O pai visitou o filho feliz, ‘The father visited the son happy’ – ambiguous structure in Portuguese). One of their conclusions was that syllable stretching in the NP2 formed distinct prosodic domains between NP2 and the Attribute ([visitou o filhoϕ] [felizϕ], ‘[visited the sonφ] [happyφ]’), and that favors a non-local interpretation of the Attribute (one in which the Attribute is interpreted as modifying the subject, NP1, and not object, NP2). However, the authors noticeda variation between interpretations with local and nonlocal apposition in the comprehension of sentences without prosodic elongation. That is, according to the authors, the absence of prosodic marking would not favor one or the other interpretation, whereas elongated prosodic marking would favor a non-local interpretation. In our work, we argue that the disambiguation of this type of sentence depends basically on the resolution of a conflict between three differentgrammatical principles that are acting together in the disambiguation of sentences with this structure: a prosodic principle, a syntactic-computational principle and a semantic-pragmatic principle. In order to prove our point, we developed an off-line sentence comprehension test and applied it to 270 informants. Our results point to an interaction between these three principlesin the resolution of the ambiguity of these sentences.