Maria Bernadete Marques Abaurre

Secondary stress in two varieties of Portuguese and the Sotaq optimality based computer program

Abstract This paper presents an analysis of secondary stress in Brazilian and European Portuguese (BP and EP) based on Optimality Theory (OT). We argue that a representational analysis has the following advantages: (i) generating all the facts of Brazilian and European Portuguese without postulating any cases of absolute neutralization; (ii) not forcing the usage of the notion of directionality, thus implying a simplification of the phonological theory; and (iii) being able to generate variant forms in parallel. We have developed a computer program, sotaq, that tests proposed stress systems (formulated in terms of constraint hierarchies) for both varieties of Portuguese against observed actual stresses and in large corpora, thus allowing for automatic testing of Optimality Theory predictions for secondary stress.

The interplay between spontaneous writing and underlying linguistic representations

The paper discusses the importance of incorporating naturalistic data from spontaneous written texts in language acquisition research in general, and in writing acquisition studies in particular. It is argued that spontaneous data can be revealing of the nature of local, epilinguistic procedures, which, by their turn, are seen as indices of underlying system construction. Data are analyzed, from first graders’ spontaneous writing in Portuguese, which bear on the problem of the underlying representation of phonological nasality in the language. It is finally argued that it is the two-sequence interpretation given by the orthography to this aspect of Portuguese phonology that, in the long run, forces children to restructure their «preferred» analysis pointing to a series of phonologically distinct nasal vowels in the language. The alternating nature of the considered data leads to discussion of the relationship between longitudinal and transversal language acquisition studies, and to the proposal of incorporating a micro-historical dimension to explain variability in language acquisition data. It is finally argued that the data under consideration are best accounted for by a constructivist model that attributes a constitutive status to social interaction, and to children’s history of contact with particular words and structures within the context of specific discourse events.RésuméL’article traite de l’interêt des données «naturelles» provenant de textes écrits spontanés, pour la recherche sur l’acquisition du langage en général et l’acquisition de l’écrit en particulier. On soutient l’idée que les données spontanées peuvent être révélatrices de la nature des procédures locales, épilinguistiques, qui, à leur tour, sont considérées comme des indices de la construction de systèmes sous-jacents. Les données analysées proviennent d’écrits spontanés en portugais d’élèves du Cours Préparatoire. L’analyse porte sur le problème de la représentation sous-jacente de la nasalité phonologique dans la langue. On argumente que c’est la traduction orthographique de cet aspect de la phonologie portuguaise qui, à la longue, force les enfants à restructurer leur analyse «préférentielle» et à admettre une série de voyelles nasales phonologiquement distinctes dans la langue. La nature des données considérées nous amène à discuter la relation entre les études longitudinales et transversales d’acquisition du langage, et à proposer l’incorporation d’une dimension micro-historique pour expliquer la variabilité des données d’acquisition. On soutient finalement que les données considérées sont très bien expliquées par un modèle constructiviste qui attribue un statut constitutif à l’interaction sociale et à l’histoire du contact des enfants avec des mots et des structures particulières dans le contexte d’évènements discursifs spécifiques.

DISPERSÃO E HARMONIA VOCÁLICA EM DIALETOS DO PORTUGUÊS DO BRASIL

Neste artigo, buscamos evidencia acustica para a harmonia vocalica com base na existencia de correlacao entre os valores de F1 das vogais pretonicas e tonicas. XXXX & XXXX (2009) mostram que um /a/ tonico nao desencadeia harmonia vocalica em dados de Minas Gerais, Espirito Santo e Goias. Dados de Kentowicz & XXXX (2011), no entanto, sugerem que /a/ desencadeia harmonia em dados de Pernambuco. Neste trabalho, apresentamos uma hipotese preliminar, baseada na dispersao vocalica, acerca do comportamento de vogais tonicas baixas como gatilhos de harmonia pretonica. Nossa analise baseia-se na comparacao feita entre dados de Salvador e Porto Alegre e os dados de Sao Paulo, Minas Gerais e Pernambuco de Kenstowicz & XXXX.

Apagamento vocálico e binariedade no português: uma investigação baseada em preditivas Bayesianas

As aplicacoes de modelos estatisticos em fonologia de laboratorio tem se mostrado produtivas na avaliacoes do poder de predicao de hipoteses. Neste trabalho, a forca da restricao FootBin foi testada estatisticamente, no portugues brasileiro e no portugues europeu, a partir de preditivas Bayesianas. Partimos do pressuposto de que, se a restricao FootBin for ativa, ha uma preferencia, na lingua observada, por palavras com um numero par de silabas. Com base neste pressuposto, o estudo foi realizado a partir de dados experimentais (apagamento/reducao vocalica) analisados acusticamente e tratados estatisticamente. Nossos resultados estatisticos apontam para o fato de FootBin estar ativa nas duas variedades de portugues, embora a forca para binariedade se manifeste de maneira mais forte em PB. O fato de FootBin estar ativa nas duas variedades da lingua com forcas diferentes suporta uma analise otimalista baseada em restricoes.