Tools for the transcription of Portuguese prosody
Conventions
A P-ToBI record of an utterance includes three main tiers time-aligned with the audio file, and more precisely with the sound wave, the spectrogram and the F0 contour, as illustrated below.
The audio file, with its wave form, spectrogram and F0 contour, constitute continuous phonetic records of the utterance.
The three main tiers are the obligatory symbol strings: the Tone tier, the Orthographic tier and the Break Indices tier. The first and third are point tiers, and the second is an interval tier.
The three obligatory symbol strings can only contain phonological events. However, depending on the purposes of the transcription or on the stage of the transcription (e.g., initial stage, interim analysis), phonetic annotations can be included. In some cases, these can be added to the main tiers. Examples are phonetic events such as a delayed peak (<), an early peak (>) or a raised peak (¡), which may be annotated in the Tone tier. Another examples are the deletion of segments (marked with curved parentheses – ‘ve(r)’) or the insertion of segments (marked with squared parentheses – ‘mar[ɨ]’), which can be annotated in the orthographic tier.
Alternatively, the use of a phonetic tier (a forth tier) where temporary annotations or hypotheses yet to be evaluated are made, basically resorting to the same set of symbols, may be the best option especially in cases where the variety under analysis is still under-studied and/or a full-fledged analysis of the aspects of the target intonation system under observation is yet unavailable.
Below, you may find an example:
Task: Reading Task
Semantic interpretation: Information-seeking yes/no question
Speaker: EN
Variety: Bragança
Utterance: Ela foi ver o mar?
Translation: She went to see the sea?
A P-ToBI record of an utterance includes three main tiers time-aligned with the audio file, and more precisely with the sound wave, the spectrogram and the F0 contour, as illustrated below.
The audio file, with its wave form, spectrogram and F0 contour, constitute continuous phonetic records of the utterance.
The three main tiers are the obligatory symbol strings: the Tone tier, the Orthographic tier and the Break Indices tier. The first and third are point tiers, and the second is an interval tier.
- Tone tier: an autosegmental-metrical transcription of the intonation contour, including the tonal categories (pitch accents and boundary tones) and reflecting their phonological organization: for example, prenuclear, nuclear and post-nuclear pitch accents, tone association to the relevant syllable in pitch accents (marked by an '*') and to the relevant syllable and prosodic level in boundary events (marked by the diacritic '%' for IP and 'p' for the PhP).
- Orthographic tier: an orthographic transcription of each (dictionary) word in the utterance.
- Break Indices tier: a prosodic transcription of the phrasing obtained according to the morphosyntax-to-phonology mapping, the phonological constraints that play a role in the language/variety, as well as realization constraints. As in the ToBI tradition, a numeric index indicates the degree of juncture:
0 = Clitic, 1= Prosodic Word, 2 = Prosodic Word Group, 3 = Phonological Phrase, 4 = Intonational Phrase.
Importantly, prosodic breaks are annotated independently of tonal boundary marks (as tonal marking of some of the prosodic levels may vary across varieties of the language, and/or be optional).
The three obligatory symbol strings can only contain phonological events. However, depending on the purposes of the transcription or on the stage of the transcription (e.g., initial stage, interim analysis), phonetic annotations can be included. In some cases, these can be added to the main tiers. Examples are phonetic events such as a delayed peak (<), an early peak (>) or a raised peak (¡), which may be annotated in the Tone tier. Another examples are the deletion of segments (marked with curved parentheses – ‘ve(r)’) or the insertion of segments (marked with squared parentheses – ‘mar[ɨ]’), which can be annotated in the orthographic tier.
Alternatively, the use of a phonetic tier (a forth tier) where temporary annotations or hypotheses yet to be evaluated are made, basically resorting to the same set of symbols, may be the best option especially in cases where the variety under analysis is still under-studied and/or a full-fledged analysis of the aspects of the target intonation system under observation is yet unavailable.
Below, you may find an example:
Task: Reading Task
Semantic interpretation: Information-seeking yes/no question
Speaker: EN
Variety: Bragança
Utterance: Ela foi ver o mar?
Translation: She went to see the sea?
© Copyright © 2015. All rights reserved. Design by Pedro Oliveira.
 |
 (PTDC/CLE-LIN/119787/2010) |
 |
 |
 |
 |
|---|