Jazz composers and scholars have praised American pianist Bill (William John) Evans as a superior improviser. But oddly, considering the renown gained by Evans, his music is only recently being analyzed in any theoretical detail. An essential aspect of Evans' style, in my opinion, is his ability to satisfy the harmonic obligations of the theme and, at the same time, establish a maximum degree of rhythmic contrast with the metrical pattern of the original harmonic succession. This paper examines the rhythmic relationship of Evans' melodies to the chord changes over various timespans of the theme and its variations. A method of accent analysis is developed to describe this relation and is applied to excerpts firm two of his solos with a distinct bebop character. According to the provided criteria, this analytical method defines the strongest melodic accents ("points of emphasis") in the melodic accent pattern in relation to the different chordal accents generated by the theme's harmonic progressions and to a lesser extent by the interpolated substitutes that together constitute the hypermetrical accent pattern. In comparing these accent patterns and the accent processes of "coincidence", "synchronization", "cross-accent", and "convergence" they articulate we determine some of the techniques Evans uses to create rhythmic contrast between the theme and its variations. On the grounds of this study, one may conclude that the rhythmic structures of melody shown to be in "Peri's Scope" and "Beautiful Love" are characteristic of Evans' improvisations, especially in the early stages of his career. Hence, one may use this type of analysis to investigate the improvisational styles of Evans' predecessors: innovators such as Charlie Parker and Bud Powell, and those of some of his contemporaries, such as the pianist Red Garland. Finally, the original concepts, terms, and processes identified by this method of accent analysis may help the jazz student to recognize the possible relations of melodic pitches to underlying chords during the course of performance. It seems likely, then, that this type of analysis can form the basis for a pedagogical method that has a distinctly rhythmic-harmonic aspect, an aspect that deserves more attention in jazz education. / Arts, Faculty of / Music, School of / Graduate

Fluency of processing – the ease with which one extracts information from stimuli – affects a variety of cognitive processes over and above the influence of declarative content. Although this influence has been extensively demonstrated in a variety of different domains (Alter & Oppenheimer, 2009), there are virtually no studies exploring this effect with auditory material. Moreover, although research on modality differences suggests that people process auditory information differently than they process visual or written information (Conway & Gathercole, 1987; Markman, Taylor & Gentner, 2007), there are no studies that directly compare the effects of processing fluency on judgments across different modalities. The current dissertation reports two sets of studies, one investigating the effects of processing fluency on cognitive judgments in the auditory modality, and a second exploring cross-modal differences in processing fluency. The first set of studies showed that although foreign-accented speech is more difficult to process, this disfluency does not affect cognitive judgments. In the second set of studies, two experiments show that disfluency in processing affects judgments of truth (Experiment 1) and the intention to purchase a product (Experiment 2) only with written – non-verbal – material. Experiment 3 investigates one possible explanation for the limited influence of processing fluency in speech: because people tend to focus on conceptual information over low-level acoustic information when processing language (Lahiri & Marslen-Wilson, 1991; Gow & Gordon, 1995; Mattys, White & Melhorn, 2005; Norris, McQueen & Cutler, 1995), distortions to the superficial features of the speech signal is likely to have limited impact on how people process the conceptual content. In Experiment 3 participants are primed to attend to the superficial features of foreign-accented speech. The results showed that when people are primed to attend to features that make foreign-accented speech difficult, non-native speech has an impact on subsequent judgments of truth. Overall, the studies presented here show that listeners can extract content from speech, even when it is distorted. They also show that when attention is directed to low-level acoustic features of speech, processing fluency effects becomes apparent. / text

Automatic speech recognition technology is becoming increasingly widespread in many applications. For dictation tasks, where a single talker is to use the system for long periods of time, the high recognition accuracies obtained are in part due to the user performing a lengthy enrolment procedure to ‘tune’ the parameters of the recogniser to their particular voice characteristics and speaking style. Interactive speech systems, where the speaker is using the system for only a short period of time (for example to obtain information) do not have the luxury of long enrolments and have to adapt rapidly to new speakers and speaking styles. This thesis discusses the variations between speakers and speaking styles which result in decreased recognition performance when there is a mismatch between the talker and the systems models. An unsupervised method to rapidly identify and normalise differences in vocal tract length is presented and shown to give improvements in recognition accuracy for little computational overhead. Two unsupervised methods of identifying speakers with similar speaking styles are also presented. The first, a data-driven technique, is shown to accurately classify British and American accented speech, and is also used to improve recognition accuracy by clustering groups of similar talkers. The second uses the phonotactic information available within pronunciation dictionaries to model British and American accented speech. This model is then used to rapidly and accurately classify speakers.

No description available.

This paper seeks to address the issue by examining two factors that potentially affect a listener’s perception of foreign accented speech: degree of familiarity (as acquired through a work or personal environment) and amount of ESL or linguistic training. Speech samples were recorded from 18 international students from Hispanic, Asian, and Middle-eastern backgrounds and across all proficiency levels as designated by their academic English program. Six native English speakers were also recorded to serve as a basis for comparison. Listeners were drawn from two pools: people with ESL and/or linguistic training (n=42) and laypersons with no such specialist training (n=36). After completing a background questionnaire to assess familiarity with foreign accented speech, each listener rated all 24 speech samples on the dimensions of comprehensibility, degree of accent, and communicative ability. Results indicate that participants with ESL/linguistic training rate foreign accented speech more positively on all three dimensions than laypersons with no such training. Additionally, degree of familiarity with foreign accented speech is positively correlated with how participants rated the accented speech samples. a number of highly significant interactions between these and other factors including sex of the speaker, proficiency level of the speaker, and L1 family of the speaker were found as well.

Le but du travail était la comparaison de l’accentuation des deux langues : français et lituanien. J’ai essayé de trouver les ressemblances et les différences de l’accentuation. L’accentuation est une science qui s’intéresse à la nature des accents dans de différentes langues et les fonctions que l’accent accomplit. C’est un phénomène provenant de l'augmentation de la durée syllabique, de l'intensité sonore et de la hauteur mélodique sur certaines syllabes de l'énoncé, elle fait une partie de la prosodie. A l'écoute de locuteurs parlant français ou lituanien, nous remarquons que certaines syllabes seront mieux perçues que d'autres. Ces syllabes « remarquables » sont porteuses d'un accent ; on les appelle des syllabes accentuées. L'accent est la mise en valeur d’une syllabe. Nous avons vu que les systèmes d’accentuation des deux langues sont très différents. Tout d’abord nous avons aperçu que le lituanien a un accent de mot qui est libre, tandis que le français possède un accent de groupe de mots qui est fixe. L’accent lituanien n’est pas attaché à une syllabe déterminée dans tous les mots tout en ayant une place fixée dans chacun d’eux. Prenons un mot de trois syllabes, par exemple. Nous pourrons dire que, dans une langue donnée, l'accent est libre quand il peut tomber aussi bien sur la première, la deuxième ou la troisième syllabe des mots de cette langue. En lituanien l'accent se déplace de cette façon dans « áudeklas », « saldaĩnis » et « lygumà », mais la liberté ne... [to full text]

Made available in DSpace on 2015-05-14T12:42:48Z (GMT). No. of bitstreams: 1 Arquivototal.pdf: 1180446 bytes, checksum: ab14305390450b5a3863652943e78075 (MD5) Previous issue date: 2012-04-19 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / The phenomenon that we propose to analyze in this work is the accent of the Brazilian Portuguese (based on PB), according to Câmara Júnior (1970, p. 63), it is characterized as a greater inspiration power, or the emission intensity, of the vowel of a syllable in contrast with other vowels . The general objective of this work is to discuss the criteria which define the Brazilian Portuguese accent pattern. In order to achieve that goal, a work was carried out based on data previously taken in several studies there are in Brazil. We understand that it is not necessary to take more data, as long as there are variation studies with true results which are worth to support our thesis and according to it, the primary accent of the Brazilian Portuguese is free, variable (on the edges), unpredictable, therefore, it can not be predicted by any rule or restriction unless stipulations are considered, what ends up bringing much stress to the Metrical phonology. We decline the Metrical Phonology as it is constituted nowadays for that theory consider the primary accent as secondary and the secondary as primary, what brings inaccurate results, not trustable enough because the criteria used in that theory sensitivity meaningful to a sort of support what characterizes the Brazilian Portuguese accent seem not to be appropriate to an analysis of that system. We postulate that the primary accent is considered as primary, as states the First Primary Accent Theory PAF proposed by van der Hulst (1997, 2006), that is the guiding of this work. According to the PAF, there are two necessary criteria for placing the primary accent: the special syllable and the dominium. Based on that theory the special syllable is the one which gets a lexical mark motivated for getting that feature. Concerning the dominium, this involves the two last syllables (right to left), being, in this case, an optional extra metrical fact. In that point, we disagree with the approach proposed by van de Hulst for we postulate the enlargement from two to three syllables next to the right end the group of three syllables is perfectly predictable by the Metrical Phonology, with a difference from that theory, so that the third part of that group is closed. / O fenômeno que nos propomos a analisar neste trabalho é o acento do português brasileiro (doravante PB) que, segundo Câmara Júnior (1970, p. 63), se caracteriza como uma maior força expiratória, ou intensidade de emissão, da vogal de uma sílaba em contraste com as demais vogais silábicas . O objetivo geral deste trabalho é discutir os critérios que definem o padrão acentual do português brasileiro. Para alcançar esse objetivo, desenvolvemos um trabalho baseado em dados previamente levantados pelos diversos estudos já existentes no Brasil. Entendemos não ser necessário levantar mais dados, uma vez que já há estudos de variação com resultados confiáveis que servem para apoiarmos nossa tese, segundo a qual o acento primário do português brasileiro é livre, variável (na subjacência), imprevisível, portanto, não pode ser previsto por qualquer regra ou restrição salvo se se lançar mão de estipulações, o que termina por gerar grande sobrecarga à Fonologia Métrica. Rejeitamos a Fonologia Métrica tal como constituída atualmente por essa teoria tratar o acento primário como secundário e o acento secundário como primário, o que gera resultados imprecisos, pouco confiáveis porque os critérios adotados por essa teoria sensibilidade ao peso e adoção de um tipo de pé como o caracterizador do acento do português brasileiro - parecem não ser adequados para a análise desse sistema. Postulamos que o acento primário seja tratado como primário, tal como defende a Teoria do Acento Primário Primeiro PAF , proposta por van der Hulst (1997, 2006), que é a teoria norteadora deste trabalho. Segundo a PAF, dois são os critérios necessários para a localização do acento primário: a sílaba especial e o domínio. Essa teoria entende por sílaba especial aquela que recebe uma marca lexical motivada diacronicamente ou por ser um empréstimo. Quanto ao domínio, este compreende as duas últimas sílabas da borda (direita ou esquerda), sendo, nesse caso, a extrametricidade opcional. Nesse ponto, discordamos da abordagem proposta por van der Hulst por postularmos a ampliação de duas para três sílabas próximas à borda direita a janela de três sílabas é perfeitamente previsível pela Fonologia Métrica, com a diferença que, para essa teoria, a terceira banda dessa janela é sempre fechada.

Until very recently language development research classified the language learner as belonging to one of two discrete groups – monolingual or bilingual. This thesis explores the hypothesis that this is an insufficient description of language input and that there are sub-groups within the monolingual category based on the phonological variability of their exposure that could be considered akin to that of bilingual toddlers. For some monolingual toddlers, classified as monodialectal, their language exposure is generally consistent, because both of their parents speak the dialect of the local area. Yet for other toddlers, classified as multidialectal, the language environment is more variable, because at least one of their parents speaks with a dialect that differs from the local area. It is considered that by testing this group of multidialectal toddlers it will be possible to explore the effect of variability on language development and how increased variability in the bilingual linguistic environment might be influencing aspects of language development. This thesis approaches the influence of variability from three areas of interest: phonetic specificity of familiar words using a mispronunciation paradigm (Experiments 1 and 2), target recognition of naturally occurring pronunciation alternatives (Experiments 3 and 4) and use of the Mutual Exclusivity strategy in novel word learning (Experiment 5). Results show that there are differences between the two dialect groups (monodialectal and multidialectal) in a mispronunciation detection task but that toddlers perform similarly with naturally occurring pronunciation alternatives and in their application of the Mutual Exclusivity strategy. This programme of work highlights that there is an influence of linguistic variability on aspects of language development, justifying the parallel between bilingualism and multidialectalism.

No description available.

Recent work has shown that preschoolers track informants’ past reliability concerning familiar information and labels, and they use this information to judge the correctness of novel information and labels they provide. But linguistic factors also sway children’s choices for social interaction, for which native-accented speakers are preferred. The present study uses the selective trust paradigm to consider how accentedness interacts with speaker reliability with native- and foreign-accented informants. The results show that speaker reliability and accentedness affect four-year-olds’ choices, but the impact of these factors differed by response type. Preschoolers preferred to ask the native-accented speaker for information, regardless of his reliability. However, in choosing which label to learn, preschoolers selected the reliable speaker’s label, regardless of accent, and correctly identified the unreliable speaker. This study provides evidence suggesting that young children separate their social biases from their objective assessment of novel information. / text