ABSTRACT
ABSTRACT Purpose to investigate prosodic boundary effects on the comprehension of attachment ambiguities in Brazilian Portuguese and to test two hypotheses relying on the notion of boundary strength: the absolute boundary hypothesis (ABH) and the relative boundary hypothesis (RBH). Manipulations of prosodic structure influence how listeners interpret syntactically ambiguous sentences. However, the role of prosody in spoken language comprehension of sentences has received limited attention in languages other than English, particularly from a developmental perspective. Methods Twenty-three adults and 15 children participated in a computerized sentence comprehension task involving syntactically ambiguous sentences. Each sentence was recorded in eight different prosodic forms with acoustic manipulations of F0, duration and pause varying the boundary size to reflect predictions of the ABH and RBH. Results Children and adults differed in how prosody influenced their syntactic processing and children were significantly slower than adults. Results indicated that interpretation of sentences varied according to their prosodic forms. Conclusion Neither the ABH or the RBH explained how children and adults who speak Brazilian Portuguese use prosodic boundaries to disambiguate sentences. There is evidence that the way prosodic boundaries influence disambiguation varies cross-linguistically.
RESUMO Objetivo investigar os efeitos de fronteiras prosódicas na compreensão de ambiguidades sintáticas no português brasileiro além de testar duas hipóteses baseadas na noção de intensidade de fronteira: a hipótese de fronteira absoluta (ABH) e a hipótese de fronteira relativa (RBH). Manipulações da estrutura prosódica influenciam como os ouvintes interpretam frases sintaticamente ambíguas. No entanto, o papel da prosódia na compreensão da linguagem oral tem recebido atenção limitada em línguas além do inglês, particularmente do ponto de vista do desenvolvimento. Método Vinte e três adultos e 15 crianças participaram de uma tarefa computadorizada de compreensão de frases envolvendo frases sintaticamente ambíguas. Cada frase foi gravada em oito formas prosódicas diferentes com manipulações acústicas de F0, duração, e pausa, variando o tamanho da fronteira prosódica de modo a transparecer as previsões da ABH e RBH. Resultados Crianças e adultos diferiram em como a prosódia influenciou o processamento sintático; as crianças foram significativamente mais lentas que os adultos. Os resultados indicaram que a interpretação das frases variou de acordo com suas formas prosódicas. Conclusão Nenhuma das hipóteses (ABH ou RBH) explica como crianças e adultos falantes do Português brasileiro utilizam as fronteiras prosódicas para desambiguar frases. Há evidências de que a maneira com a qual os limites prosódicos influenciam a desambiguação de frases varia entre os idiomas.
ABSTRACT
Sentence-comprehension deficits have been described in patients with primary progressive aphasia (PPA). However, most instruments to address this domain in more detail and in a clinical context have not been adapted and translated into several languages, posing limitations to clinical practice and cross-language research. Objectives: The study aimed to (1) test the applicability of the Brazilian version of the Test for Reception of Grammar (TROG2-Br) to detect morphosyntactic deficits in patients with PPA; (2) investigate the association between performance in the test and sociodemographic and clinical variables (age, years of formal education, and disease duration); (3) characterize the performance of individuals presenting with the three more common variants of PPA (non-fluent, semantic, and logopenic) and mixed PPA (PPA-Mx) and analyze whether TROG-2 may assist in the distinction of these clinical profiles. Methods: A total of 74 cognitively healthy participants and 34 individuals diagnosed with PPA were assessed with TROG2-Br. Overall scores (correct items, passed blocks), types, and categories of errors were analyzed. Results: In controls, block scores were significantly correlated with years of formal education (Spearman's r = 0.33, p = 004) but not with age. In PPA, age, education, and disease duration were not significantly associated with performance in the test. Controls presented a significantly higher performance on TROG2-Br compared to PPA individuals and their errors pattern pointed to mild general cognitive processing difficulties (attention, working memory). PPA error types pointed to processing and morphosyntactic deficits in nonfluent or agrammatic PPA, (PPA-NF/A), logopenic PPA (PPA-L), and PPA-Mx. The semantic PPA (PPA-S) subgroup was qualitatively more similar to controls (processing difficulties and lower percentage of morphosyntactic errors). TROG2-Br presented good internal consistency and concurrent validity. Discussion: Our results corroborate findings with TROG-2 in other populations. The performance of typical older adults with heterogeneous levels of education is discussed along with recommendations for clinical use of the test and future directions of research.