Non-accented Syllables Research Articles

Per un ascolto efficiente del portoghese europeo. Percorsi dal basso e dall’alto

Abstract – This contribution offers a discussion on the enhancement of the top-down approach in the listening of oral texts in European Portuguese, by Italian-speaking learners. Portuguese has a vowel system that is particularly rich when compared to Italian. In addition, the vowels in non-accented syllables undergo a process of reduction that makes them difficult to perceive by the ear of an Italian speaker. In order to face such a difficulty of perception, recognition and comprehension of the oral text, it is necessary on one hand to consider the specific problems at all linguistic levels, following a bottom-up perspective and, on the other hand, to focus on the contribution that inference can make to strengthen the possibilities of understanding the message.

Uncovering Tonal and Temporal Correlates of Phrasal Prominence in Medʉmba.

Characterizing prosodic prominence relations in African tone languages is notoriously difficult, as typical acoustic cues to prominence (changes in F0, increases in intensity, etc.) can be difficult to distinguish from those which mark tonal contrasts. The task of establishing prominence is further complicated by the fact that tone, an important cue to syllable prominence and prosodic boundaries cross-linguistically, plays many roles in African languages: tones often signal lexical contrasts, can themselves be morphemes, and can also interact in key ways with prosody. The present study builds on phonological generalizations about tonal patterns in Medʉmba, a Grassfields Bantu language, and uses the speech cycling paradigm to investigate relative timing of syllables varying in phrase-level prominence. Specifically, we investigate timing asymmetries between syllables hypothesized to occur at the edge of a phonological phrase, which carry a high phrase accent, and those in phrase-medial position, which do not. Results indicate significant differences in the temporal alignment of accented versus non-accented syllables, with accented syllables occurring significantly closer to positions established as prominence-attracting in previous speech cycling research. We show that these findings cannot be attributed to differences in tone alone. Findings demonstrate the importance of relative temporal alignment as a correlate of prosodic prominence. Findings also point to increased duration as a phonetic property which distinguishes between syllables bearing phrasal prominence from those which do not.

Análisis filológico de S. Ireneo «Adversvs haereses» 3,3,2b

The philological analysis of a scriptural or patristic text is the proper starting-point for its theological study and interpretation. Here are the principal conclusions of the syntactic and stylistic analysis of the famous text of S. Irenaeus: 1. The hanc enim ecclesiam is the Church of Rome, which is mentioned immediately before (3,3,2a). This is demanded a) by the deictic reference normally inherent in hic as adjective to the object last named, and b) by the explanatory significance of enim. 2. Necesse est expresses the obligation arising from the nature of things; in this case, from the constitution of the Church. Its subject in omnis Ecclesia, where omnis has both its collective and its distributive value: i. e. all the local Churches together and every one in particular. The expression necesse est convenire makes of the Church of Rome the centre of gravity and the key to unity of the whole Church. 3. The antecedent of the adjectival relative clause in qua etc. is the Church of Rome. This is demanded a) by the immediate context, b) by the text itself, c) by the function of the phrase hoc... fideles, d) by the necessity of avoiding a tautology, and e) by the stylistic devices employed. 4. The phrase ab his qui etc. is comparative in significance. 5. The whole period is built up in accordance with the rules of rhythmic prose as regards a) the beginning, b) the end of each member and phrase, and c) the number of syllables. Its stylistic structure is made up of two members: I) Ad hanc... fileles, and II) In qua... traditio. Each member is divided into four phrases. 6. Syntactically, they are both members of a single period: stylistically, they have their own substantive character and independence. This is shown a) by the beginning of each member; b) by the total parallelism in the structure of the final phrases, with the adjectival relative clause between the demonstrative adjective (eos, ea) and the antecedent (fideles, traditio); c) by the symmetrical parallelism of the two members, so that all their phrases can be superimposed; d) by the identical alternation of accented and non-accented syllables in the final clausulae, absent at the end of the six remaining phrases; and e) by the equality or quasi-equality of syllables (isocratic school) of the first member, not to be found in the second (demosthenic school). All this being so, one of the two members cannot depend upon an element present only in the other; that is to say, the relative in qua has as its antecedent, not omnis Ecclesia in 3,3,2b, but Romae Ecclesia in 3,3,2a, through the intermediary of hanc Ecclesiam in 3,3,2b. 7. Here, then, is the translation of the passage according to the syntactico-stylistic requirements and the genius of the Spanish language: For with this Church (of Rome), on account of its supreme principality, all the Church must concur, that is, the faithful from (on) every side. It has always conserved the apostolic tradition better than the (faithful, Churches) on every side. In a second article, entitled The 'Principalitas' and the 'Principatus' of the Episcopal See of Rome and its Bishop, I shall sketch the semantic field of principalitas, literally translated here by 'principality', and its theological result (the Primacy of the Pope).

Relações entre aspectos ortográficos e fonético-fonológicos de fonemas oclusivos

OBJETIVOS: descrever o desempenho ortográfico de crianças - quanto ao registro de consoantes oclusivas do Português Brasileiro; verificar se a acento influencia a ocorrência de erros no interior das palavras; e classificar o padrão da distribuição dos erros. MÉTODOS: foram analisadas 210 produções textuais referentes a 14 propostas temáticas, coletadas durante o ano de 2001, quando os sujeitos cursavam a 1ª série do Ensino Fundamental. Para descrever o desempenho das crianças, para o primeiro objetivo, os dados foram divididos em Acertos e Erros; para segundo objetivo, os erros foram classificados em sílabas acentuadas e não-acentuadas, e, por fim, para o terceiro objetivo, os erros foram subclassificados em omissões, erros ortográficos e erros fonológicos. Os erros fonológicos foram subdivididos, ainda, conforme ocorressem dentro e fora da classe. RESULTADOS: para o primeiro objetivo, encontraram-se 5.746 possibilidades de ocorrência de consoantes oclusivas na posição de ataque silábico simples, com maior quantidade de acertos; para o segundo objetivo, observou-se maior ocorrência de erros em sílabas não-acentuadas; e por fim, para o terceiro objetivo, foi possível observar predomínio de erros fonológicos, seguidos pelas omissões e pelos erros ortográficos, tanto no interior das sílabas não-acentuadas quanto das acentuadas. Ainda conforme o terceiro objetivo, nos erros fonológicos, observou-se sua maior ocorrência dentro da classe; e novamente, não-influência do acento. CONCLUSÃO: conclui-se, então, a importância da análise dos aspectos fonéticos-fonológicos na aquisição da ortografia infantil, bem como a relevância da distribuição dos acertos tanto como dos erros.

Characteristics of the acquisition of sonorant consonants orthography in Brazilian children from a São Paulo municipality.

To describe the spelling performance of children with regard to the record of sonorant consonants in Brazilian Portuguese language, to verify if the errors in their records were influenced by the accent in the word, and to categorize the kinds of errors found. For this current survey, 801 text productions were selected as a result of the development of 14 different thematic proposals, prepared by 76 children from the first grade of primary school, in 2001, coming from two schools of a city from São Paulo, Brazil. Of these productions, all words with sonorant consonants in a syllabic position of simple onset were selected. They were then organized as they appeared as pre-tonic, tonic, and post-tonic syllables, unstressed and tonic monosyllables. The following was observed: the number of hits was extremely higher than that of errors; higher occurrence of errors in non-accented syllables; higher occurrence of phonological substitutions followed by omissions and, at last, orthographic substitutions; and higher number of substitutions that involved graphemes referring to the sonorant class. Considering the distribution of orthographic data between hits and errors, as well as their relationship with phonetic-phonological aspects, may contribute to the comprehension of school difficulties, which are usually found in the first years of literacy instruction.

Glottal marking of vowel-initial words in German

Glottal marking of vowel-initial German words by glottalization and glottal stop insertion were investigated in dependence on speech rate, word type (content vs. function words), word accent, phrasal position and the following vowel. The analysed material consisted of speeches of Konrad Adenauer, Thomas Mann and Richard von Weizsäcker. The investigation shows that not only the left boundary of accented syllables (including phrasal stress boundary) and lexical words favour glottal stops/glottalization, but also that the segmental level appears to have a strong impact on these insertion processes. Specifically, the results show that low vowels in contrast to non-low ones favour glottal stops/glottalization even before non-accented syllables and functional words.

Factors affecting phonetic variation of American English /k/

Word-final consonant reduction is commonly observed in American English, but variation in word-initial consonants is less thoroughly studied. Several considerations suggest that initial consonants might resist reduction; a more nuanced hypothesis is that initial consonants resist reduction in prosodically strong positions (i.e., at phrase edges and pitch accents) but may be reduced in weaker positions (e.g., phrase medially, in nonaccented syllables), particularly in un-self-conscious speech. A corpus of 16 spontaneous dialogues (8 speakers), prosodically labeled using the ToBI transcription system, provides a testbed for this view: the speakers were pairs of close friends engaged in an absorbing task which distracted them from paying close attention to their speech. Preliminary inspection of initial /k/’s from 3 speakers revealed a range of pronunciations (as described in earlier work), from normal prevocalic /k/ (with a silent closure period followed by release noise), to ‘‘leaky’’ /k/’s with some noise during the closure period and higher-amplitude release noise, to fricated /k/’s with constant-amplitude noise throughout, and offered some support for reduction in weak prosodic contexts. Further analysis will compare prosodic context with other factors governing phonetic variation (e.g., word frequency); implications for speech recognition and production will be discussed. [Work supported by NIH DC00075 and DC02978.]

The Implicated Ballad

We used to think of the classic oral ballad of the British Isles and English-speaking North America as ancient, timeless, eternal. Indeed, we used to think of most folklore genres as equally ancient, timeless, and eternal. But the evidence is mounting that as a genre in English and Scots oral tradition the classic ballad may be very time-bound. At its start, obviously, the genre could not have emerged before the beginning of vernacular stanzaic song, nor indeed before the English and Scots languages emerged in the thirteenth century. The evidence of the written record (and the parallel written record of Scandinavian balladry) suggests, moreover, that the form originated as a literate entertainment of late medieval elite culture, and did not settle comfortably into oral tradition until the midsixteenth century, taking on at that time the oral characteristics that have been described by Andersen (1985), Buchan (1972), and others. Ironically, this first appearance of an oral ballad tradition in the British Isles is almost exactly contemporary with the rise of popular culture, and especially a popular press, in those islands. As a result, from the very beginning the tradition has been “contaminated” by popular broadside texts. But the element of popular culture that may have had the strongest impact on this oral tradition is the cheap guitar that began to be available late in the nineteenth century. The guitar affected both tune and text. Singers who accompanied themselves by strumming chords on a guitar naturally adapted the tonality and contour of their tunes to fit classic harmonic structures. The tonality shifted from monophonic modal airs to major and minor tunes with implied harmony. (Some Dorian tunes survived because airs in this scale—like the minor, but with a raised sixth—are usually adaptable to an accompaniment consisting of two successive minor chords, such as Dm and Em). This loss of melodic freedom was accompanied by loss of rhythmic freedom. An unaccompanied singer is not forced to be rhythmically regular, but once the singer starts strumming, regularity follows. Such rhythmic regularity in melody in turn forced increased regularity in prosody. The old four-beat or three-beat line with variable numbers of non-accented syllables still appeared in some unaccompanied performances, but alongside these appeared ballads in a more even

Exploring underlying pitch targets in English statements

This study investigates F0 contours and their potential association with underlying pitch targets in English statements with and without narrow focus. Eight native speakers of American English read aloud 21 sentences at 2 speaking rates with 7 repetitions. The initial, middle, and final words in these sentences vary in focus status (focus or no focus) and word length (monosyllabic, disyllabic, or trisyllabic). F0 curves are extracted using a method that combines automatic vocal cycle detection and manual rectification. Preliminary results show that F0 peaks occur earlier when the vowel of the accented syllable is phonologically long than when it is phonologically short. This seems to suggest that a nonfinal pitch accent in an English statement is probably associated with a high static target. Preliminary analyses also show that F0 contours of nonaccented syllables are influenced more by the preceding accent than by the following accent. Furthermore, the influence of the accented syllable on the F0 of the following (but not the preceding) nonaccented syllable is found to reduce over time. These patterns seem to suggest that inter-accent F0 contours in English probably also result from certain underlying pitch targets rather than from linear or nonlinear interpolation between adjacent pitch accents.

The automatic generation of French declarative prosodic structures

The fully automatic generation of a prosodic structure for a text-to-speech system has remained an elusive task. Most systems still depend on extensive user-specified timing and F0 markers to improve the approximation of a natural prosody. A three-stage processing system that provides fully automatic prosodic control for a wide variety of French declarative sentences has been developed. Perceptual testing has demonstrated a high degree of acceptability, as well as an excellent resistance to degradation induced by telephone transmission. The initial phonological processing phase, comprising phrase identification, liaison, chaining, and syllabification rules, produces a phonemic chain marked for prosodic, lexical, and syllable boundaries. The second processing phase calculates the temporal structure on the basis of a statistical model incorporating some 25 segmental and suprasegmental parameters. The final phase specifies accented and nonaccented syllables and applies Fujisaki modeling for the generation of the F0 contour. The system has been integrated into a TTS system that functions in real time on high-end personal computers. Examples are available on the laboratory’s ftp site (ftp://eliot.unil.ch/pub/LAIP/speech/LAIPTTS). This report will concentrate on the derivation of the relevant control parameters. [Work supported by FNRS, OFES.]

Prosodic constraints on glottalization of vowel-initial syllables in American English

Earlier work on the glottalization of vowel-initial syllables sought an account in terms of the morphosyntactic hierarchy and isolated intonational facts, without accounting for the possible role of prosodic constituent structure. In a corpus of digitized, prosodically labeled speech (FM radio news style, 371 vowel-onset syllables produced by a single female speaker), (a) a syllable which begins in intonational constituent and is pitch accented is likely to show a glottal onset (91%), (b) constituent-initial position and pitch accent each separately increase the likelihood of glottal onset (66% and 47%, respectively). Noninitial nonaccented syllables are much less likely to be glottalized (16%; some of these syllables may have a pitch accent that is small in comparison with the strongly realized accents often produced in FM news speech). Constituent-initial glottalization occurs even for reduced-vowel syllables, which earlier work suggested was rare: 83% of reduced-vowel syllables in constituent-initial position glottalized, but only 24% in noninitial position. Preliminary results for a second speaker show a similar pattern, illustrating both the value of revisiting earlier analyses with prosodic structure in mind, and the usefulness of a large prosodically labeled database of continuous speech.

Análisis filológico de S. Ireneo «Adversvs haereses» 3,3,2b

The philological analysis of a scriptural or patristic text is the proper starting-point for its theological study and interpretation. Here are the principal conclusions of the syntactic and stylistic analysis of the famous text of S. Irenaeus: 1. The hanc enim ecclesiam is the Church of Rome, which is mentioned immediately before (3,3,2a). This is demanded a) by the deictic reference normally inherent in hic as adjective to the object last named, and b) by the explanatory significance of enim . 2. Necesse est expresses the obligation arising from the nature of things; in this case, from the constitution of the Church. Its subject in omnis Ecclesia , where omnis has both its collective and its distributive value: i. e. all the local Churches together and every one in particular. The expression necesse est convenire makes of the Church of Rome the centre of gravity and the key to unity of the whole Church. 3. The antecedent of the adjectival relative clause in qua etc. is the Church of Rome. This is demanded a) by the immediate context, b) by the text itself, c) by the function of the phrase hoc... fideles , d) by the necessity of avoiding a tautology, and e) by the stylistic devices employed. 4. The phrase ab his qui etc. is comparative in significance. 5. The whole period is built up in accordance with the rules of rhythmic prose as regards a) the beginning, b) the end of each member and phrase, and c) the number of syllables. Its stylistic structure is made up of two members: I) Ad hanc... fileles , and II) In qua... traditio . Each member is divided into four phrases. 6. Syntactically, they are both members of a single period: stylistically, they have their own substantive character and independence. This is shown a) by the beginning of each member; b) by the total parallelism in the structure of the final phrases, with the adjectival relative clause between the demonstrative adjective ( eos, ea ) and the antecedent ( fideles, traditio ); c) by the symmetrical parallelism of the two members, so that all their phrases can be superimposed; d) by the identical alternation of accented and non-accented syllables in the final clausulae, absent at the end of the six remaining phrases; and e) by the equality or quasi-equality of syllables (isocratic school) of the first member, not to be found in the second (demosthenic school). All this being so, one of the two members cannot depend upon an element present only in the other; that is to say, the relative in qua has as its antecedent, not omnis Ecclesia in 3,3,2b, but Romae Ecclesia in 3,3,2a, through the intermediary of hanc Ecclesiam in 3,3,2b. 7. Here, then, is the translation of the passage according to the syntactico-stylistic requirements and the genius of the Spanish language: For with this Church (of Rome), on account of its supreme principality, all the Church must concur, that is, the faithful from (on) every side. It has always conserved the apostolic tradition better than the (faithful, Churches) on every side. In a second article, entitled The 'Principalitas' and the 'Principatus' of the Episcopal See of Rome and its Bishop , I shall sketch the semantic field of principalitas, literally translated here by 'principality', and its theological result (the Primacy of the Pope).