Interactive Atlas of the Prosody of Portuguese

Funder

Organizational Unit

Publications

Onde (ainda [j])há o fenómeno? Contributo para o estudo da inserção de glide entre vogais centrais

Publication . Oliveira, Pedro; Paulino, Nuno; Cruz, Marisa; Vigário, Marina

We investigate glide insertion to break a hiatus between central vowels (V1_V2) in four regions of the Northern variety of European Portuguese (two urban and two rural), three discourse modalities and two age groups. We show that the process is optional and is bound by the intonational phrase domain in all regions. Both prosodic and sociolinguistic factors affect the frequency of occurrence of the phenomenon in variable ways among regions. Globally, insertion is more frequent when V2 bears higher levels of prominence and when V1 belongs to a clitic. Regions further north tend to insert more and (semi-)spontaneous tasks favor insertion. Age does not correlate with the frequency of insertion. We discuss the implications of these findings for the understanding of the synchronic status of the phenomenon.

2014Book part

Open access

Para a prosódia do foco em variedades do Português Europeu

Publication . Cruz, Marisa; Frota, Sónia

The present paper examines the prosodic realization of narrow focus in southern varieties of European Portuguese - Algarve (Alg) and Alentejo (Ale) -, in two speech styles (reading and role-play interview). Data were analysed for nuclear pitch accent (NPA) configuration, NPA realization (pitch range), and post-focal behaviour in early focus utterances. The results show that the type of NPA in focused declaratives is more homogeneous across varieties and speech styles than the one used in interrogatives: H*+L in the former; different strategies to express focus in the latter. Early focus statements reveal post-focal subordination of the H+L* pitch accent, realized with a compressed range across varieties.

2012Book part

Open access

Interface sintaxe-fonologia: desambiguação pela estrutura prosódica no português brasileiro

Publication . Gravina, Aline; Fernandes-Svartman, Flaviane

O objetivo deste artigo é a análise das estratégias prosódicas de desambiguação de sentenças em português brasileiro (PB), em contexto de ambiguidade sintática e choque acentual. Nossa hipótese é a de que a desambiguação dessas sentenças se dê através de diferentes estratégias prosódicas de desfazimento do choque acentual. Para testar essa hipótese, foram realizados experimentos de produção e de percepção com falantes do PB. Os resultados obtidos foram analisados à luz da Fonologia Prosódica (NESPOR; VOGEL, 1986) e, em termos entoacionais, à luz da Fonologia Entoacional (PIERREHUMBERT, 1980; LADD, 1996, 2008). Nossos resultados revelam que: (i) quando o desfazimento do choque de acentos se dá por retração acentual, há pistas da formação de um único sintagma fonológico e a interpretação é a de que a segunda palavra envolvida no choque se refere à imediatamente precedente; (ii) quando o desfazimento do choque se dá pela inserção de pausas e/ou pela atribuição de um acento tonal a cada palavra envolvida no choque, há pistas da formação de dois sintagmas fonológicos e a interpretação obtida é a de que a segunda palavra envolvida no choque se refere não à imediatamente precedente, mas à outra da sentença.

2013Journal article

Open access

Tune or Text? Tune-text accommodation strategies in Portuguese

Publication . Frota, Sónia; Cruz, Marisa; Castelo, Joelma; Barros, Nádia; Crespo-Sendra, Verònica; Vigário, Marina

In Portuguese, different strategies for dealing with tune-text accommodation have been reported. However, no systematic research has been conducted exploring crucial cases of complex nuclear melodies realized in nuclear words with final stress, as in yes-no questions. Based on reading and semi-spontaneous data from ten regions in Brazil and eleven regions in Portugal, this study reveals that Brazilian and European Portuguese globally differ with respect to the strategies implemented: in Brazilian Portuguese, the text is preserved and the melody is changed (mostly by means of tonal truncation); in European Portuguese, the melody is preserved and the text is changed through various strategies, including schwa epenthesis. Faithfulness to the text or to the tune is thus a relevant dimension of variation both across and within languages, and text changes (through vowel lengthening, vowel split, vowel epenthesis, or blocking of vowel deletion) are crucial means to support tune realization that have recently been found in unrelated languages.

2016Conference object

Open access

Distinguishing emphatic and Prosodic Word initial stresses: evidence from Brazilian Portuguese

Publication . Toneli, Priscila; Vigário, Marina; Abaurre, Maria Bernadete M.

In European Portuguese (EP), emphatic stress and initial stress have been reported to be optionally assigned to the first (or in some cases the second) syllable of a Prosodic Word (PW) ([1]). In Brazilian Portuguese (BP), initial stress (and/or H-tone) has been claimed to be assigned with reference to the primary stress position and be dependent on the number of pretonic syllables within a PW ([2]). [3], [4] and [5] suggest that in BP secondary stress assignment essentially signals the beginning of the PW in emphatic contexts, however. Although [5] reports that in emphatic contexts the initial stress and the ‘H-tone’ can coincide with a secondary stress, the nature of this type of stress and the difference between emphatic stress and PW initial stress in BP is in general not discussed. In this paper we argue that, although the two types of stresses in BP are tonally signaled, they are distinct, both in function and in distribution. Empirical data from two varieties of Portuguese spoken in Brazil (Paraná and Minas Gerais states) are presented, showing that the emphatic stress has a wider distribution than the initial stress, in neutral contexts. The emphatic stress may occur in any syllable from the stressed syllable leftwards, within the PW, including the syllable immediately adjacent to word-stress. The initial stress, by contrast, is found on the first or second pretonic syllable of PW, and there is a minimal distance of two syllables between initial stress and word-stress (e.g. governaDOres ‘governors’) ([2]). We argue that the initial stress is an edge phenomenon, marking PW initial positions, unlike the emphatic stress. In both cases, the tonal association is evidence for the PW domain in BP, because neither type of stress exceeds the limits of this domain (i.e. none of them can appear in post-tonic syllables of non-final PW).

2014Book part

Open access