Anabela Rato

Previous research has suggested that Romance language learners of English are presented with a challenge when perceiving English stop consonant sounds in a native-like manner (e.g., Aliaga-García & Mora, 2009; Flege & Eefting, 1988). This cross-language difficulty may be related to the interplay of several factors, such as L1 attunement and/or L2 experience (Flege, Munro, & MacKay, 1995). The present study further investigated how the first language (L1) and language experience affect non-native perception of the English voiceless stops /p, t, k/. Specifically, the study aimed at investigating whether Catalan and Portuguese learners of English could identify and discriminate between aspirated and unaspirated voiceless stops, since in both Romance languages aspiration is a non-existent phonetic property (Ladefoged, 1972). Two groups of upper-intermediate learners of English (L1 Catalan (n=22) and L1 European Portuguese (n=19)) and a group of advanced Portuguese learners of English (n=22) performed two perception tasks: forced-choice identification and AX categorical discrimination. The testing stimuli consisted of word-initial voiceless aspirated stops (e.g., pot, tool, key) and word-initial consonant clusters after /s/ (e.g., spot, stool, ski) naturally produced by three British English talkers. The target (C)CVC words were embedded in the carrier phrase "This (target word)" so that the aspirated and unaspirated stop allophones would be presented in similar contexts (e.g., This pot-This spot). Findings show that the more advanced learners obtained significantly higher scores than the upper-intermediate learners, which seems to indicate an effect of language experience, in line with previous studies (e.g., Flege, Munro & Skelton, 1992). Moreover, despite the comparable VOT patterns in their L1s and the similar language proficiency in the target language, the upper-intermediate Portuguese group outperformed the upper-intermediate Catalan learners in both identification and discrimination tasks. This might be a result of greater daily exposure to the target language through outside classroom input (e.g., non-dubbing on TV), which may contribute to their FL learning success (Rubio & Lirola, 2010). All these findings taken together seem to indicate that quantity and possibly quality of L2 input in different L1 linguistic environments influence non-native speech perception, and L2 language experience promotes accurate L2 allophonic speech perception.

The present study examined whether heritage speakers (HSs) of European Portuguese (EP) who were born or moved to a German-speaking country before the age of eight years were perceived as native speakers of EP. In particular, this study intended to determine whether a change of linguistic environment, length of residence in a migrant context, length of residence in the country of origin before migration and after remigration, and age at return could predict the degree of (non)native accent in the heritage language. Thirty native Portuguese speakers assessed the global accent of 20 Portuguese-German bilinguals, five Portuguese monolinguals and five highly proficient German speakers of Portuguese as a second language (L2). The group of HSs comprised 17 speakers who returned to Portugal. The results revealed that listeners perceived a strong global foreign accent in the speech of the L2 learners, while the monolingual Portuguese speakers were clearly perceived as being native speakers of EP. The HSs' ratings were considerably closer to the monolingual average ratings, but they showed more variation, indicating that their accent may bear non-native traces. Further analyses showed that the age at which the HSs emigrated was the only significant predictor, while length of residence in the host country and in Portugal were less predictive. INTRODUCTION The study reported in this paper aims at investigating whether global accent is affected by input alterations across the lifespan, focusing on adult heritage speakers (HSs) of European Portuguese (EP) who grew up and either still live in Germany or have returned to their country of origin, Portugal. In particular, this study examined whether the accent of the HSs who return to the country of origin (henceforth, returnees) is perceived as being more native-like than the accent of those who live in the country to which they migrated.

This study investigated the short- and long-term effects of a high variability perceptual training on the production of three English vowel contrasts (/i/- /ɪ/; /ɛ/-/æ/; /u/-/ʊ/) by Portuguese learners. The perceptual training consisted of five sessions divided into two blocks that included both discrimination and identification tasks followed by immediate feedback with natural stimuli produced by multiple native talkers in different phonetic contexts. Vowel production data were collected three times, namely before training (pretest), immediately after (posttest), and two months later (delayed posttest) with a sentence-reading task. Both duration and the first two vowel formants were measured to calculate the Euclidian distance (Hz) and duration ratios (ms) between the vowels of the target contrasts. The acoustic analyses revealed that perceptual training had a significant effect on pronunciation accuracy of the target vowels, specifically in terms of vowel quality.

Many speech researchers (linguists, psycholinguists, speech therapists) who work with speech perception still struggle to prepare their data collection tests mostly due to limitations as regards programming skills and knowledge of scripting languages. Concerned with this scenario and methodological constraints, we developed an open source application software (http://www.worken.com.br/tp) that enables researchers to design not only speech perception tests, but also perceptual training tasks with immediate feedback. The objective of this paper is to show some of the main features of the software and to highlight its user-friendliness.

O objetivo deste artigo é informar pesquisadores sobre alguns procedimentos metodológicos necessários para desenvolver testes de percepção e tarefas de treinamento perceptual. Após apresentar uma breve explicação sobre testes de discriminação e identificação, o artigo mostra como elaborar um teste perceptual em dois software livres diferentes: Praat e TP. Embora o Praat seja uma ferramenta excelente para elaborar experimentos de percepção, pode não ser um software amigável para pesquisadores que não se sintam confortáveis com a linguagem de programação e/ou com o inglês técnico utilizado no seu manual. Como uma alternativa, o TP foi criado para oferecer um software para testes/treinamento de percepção (áudio e audiovisual) que fornece feedback imediato e cuja configuração intuitiva segue a estrutura utilizada em aplicativos do Microsoft Windows.

The present study investigated whether high variability perceptual training can be effective in the modification of Portuguese EFL learners’ mature perceptual patterns of three English vowel contrasts (/i/-/ɪ/, /ɛ/-/æ/, and /u/-/ʊ/). Learners’ perception was assessed three times (pretest, posttest and delayed posttest) with an identification test. The five-session perceptual training included discrimination and identification tasks with immediate feedback. The results revealed that the Portuguese learners’ performance in the identification of the target vowels improved significantly after training, and the knowledge gained during perceptual learning was retained two months after completion of the training program. Moreover, the scores of the generalization test indicate that there was robust learning of two of the target vowel contrasts. The results of this experiment support the claim that perceptual learning can occur in non-naturalistic environments within a short period of time and corroborate previous findings on the malleability of L2 adult learners’ perceptual systems.

This paper aims at presenting a new tool developed to help researchers from various areas design perception tests intuitively and without having to know programing. Furthermore, it is intended to be a tutorial of the tool in Spanish. TP (Perceptual Tests/ Perceptual Training) is an open source application software which includes perceptual tests and training tasks (sound, image and video) with immediate feedback and was created as an alternative to existing softwares (Rauber, Rato, Kluge, Santos, 2012). Firstly, the functions of the program will be described in detail. Then, the different experiments that can be designed in TP will be showed. Finally, some considerations about the program will be provided.

Este artigo apresenta os resultados de duas experiências que testaram a percepção e a produção de vogais anteriores inglesas por dez falantes nativos de mandarim. No que diz respeito a produção, foram medidos os dois primeiros formantes e a duração das vogais. Os resultados revelam que as informantes chinesas tiveram dificuldades em produzir vogais inglesas sem equivalentes na língua nativa (L1), mostrando tendência para perceber duas categorias da língua estrangeira (L2) como sendo apenas uma e para produzir as vogais da L2 com menor duração do que as informantes americanas que formavam o grupo de controlo. Quanto a percepção, os resultados de um teste de identificação mostram que a vogal da L2 com uma equivalente na L1 (/i/) foi a que teve taxas de identificação correcta mais elevadas. Os resultados demonstram que a performance perceptiva e superior a produtiva, e que a percepção e a produção estão relacionadas: as taxas de identificação mais elevadas no teste de percepção correspondiam a resultados melhores em termos de produção; e as taxas de identificação mais baixas a uma produção menos precisa.

Several studies have demonstrated that second/foreign language (L2/FL) speech learning is a challenge to late learners (i.e., adolescents or adults) in terms of perception and production of certain non-native phonemic and phonetic contrasts (Moyer, 2013). The interaction of different factors might explain learners’ difficulties, namely age of onset of learning (AOL), amount of native (L1) and non-native language (NNL) use over time, quantity and quality of NNL input, and the interference from the L1 phonological system (Piske, 2007). The Speech Learning Model (SLM), proposed by Flege (1995), hypothesizes that difficulties in perceiving and, consequently, in producing non-native contrasts are due to the (dis)similarities between the L1 and the NNL phonological systems. The L1 sound system is likely to hinder the formation of new non-native (L2/FL) phonological categories. However, a considerable number of cross-language studies has revealed that phonological learning is attainable for late learners, and their abilities in perceiving and producing segmental and suprasegmental non-native contrasts can improve, since the mechanisms used in the acquisition of the L1 sound system remain intact over the lifespan and can be applied to L2/FL learning (Flege, 1995). Experimental studies that investigated the effects of perceptual training on non-native speech sound perception and production reported its success not only in the modification of adult learners’ perceptual patterns, but also in the improvement of their pronunciation accuracy, confirming, thereby, the plasticity of L2/FL learners’ mature perceptual system (e.g., Aliaga-Garcia, 2013; Pereira & Hazan, 2013; Wang, 2008; Wang et al., 2003). Difficulties in the perception of non-native vowel contrasts have been widely described as a significant part of the problems learners have in L2/FL phonological acquisition/learning (Strange, 2007). Therefore, the present study investigated the effects of perceptual training on the learning of three English contrasts (/i/-/ɪ/; /ɛ/-/æ/; /u/-/ʊ/) by a group of EFL (English as a Foreign Language) learners. This set of vowel contrasts was selected due to reported difficulties European Portuguese native speakers have in perceiving and producing them (Flege, 1994, as cited in Flege, 1995; Rato et al., 2013). The English phonological categories /ɪ/, /æ/ and /ʊ/ tend to be assimilated to the Portuguese vowel sounds /i/, /ɛ/ and /u/, respectively, and no distinction between the two vowels of each pair is made, due to their acoustic and articulatory proximity. Specifically, this study investigated (i) whether a high variability perceptual training, which included stimuli with different phonemic contexts produced by multiple native talkers, had a positive effect on the perception of the English target segments; (ii) if transfer of improvement to oral production was observed; (iii) whether perceptual learning generalized to identification of new words produced by novel talkers; and (iv) if long-term training effects remained. The participants’ perception was assessed three times with an identification test designed with natural stimuli: (1) before the auditory training – pretest; (2) immediately after the training was over – posttest; and (3) two months later – delayed posttest.  The perceptual training program consisted of five sessions divided into two blocks, which included discrimination tasks and identification sequences followed by immediate feedback. Production was tested simultaneously in the three phases by means of a sentence-reading task with the target vowel segments. The results show that the Portuguese learners’ performance in the identification of the English vowels improved significantly, and perceptual gains were retained two months after completion of the training sessions. Moreover, the results of the generalization test indicate that there was robust learning of the two front vowel pairs. Acoustic analyses of spoken data revealed that phonological learning transferred to production. In sum, these results support the claim that perceptual learning can occur in a formal non-naturalistic environment within a short period of time and corroborate previous findings on the malleability of L2/FL adult learners’ perceptual systems.

Vários estudos têm demonstrado que, na aprendizagem de uma língua não materna (LNM), a perceção e a produção de determinados contrastes fonológicos e fonéticos não nativos são um desafio para aprendentes tardios (adolescentes ou adultos) (Moyer, 2013). Diversos fatores podem explicar essas dificuldades, tais como a idade do início da aprendizagem, a frequência de uso, a quantidade e a qualidade de exposição à LNM e a influência do sistema fonológico da língua materna (L1) (Piske, 2007). O Speech Learning Model (SLM), desenvolvido por Flege (1995), explica que as dificuldades percetivas e, consequentemente, produtivas se devem ao facto de o sistema de sons linguísticos da L1 impedir a formação de novas categorias fonológicas para a segunda língua ou língua estrangeira (L2/LE). No entanto, um número considerável de estudos tem revelado que aprendentes tardios podem aprender a perceber e a produzir contrastes segmentais e suprassegmentais não nativos, uma vez que os mecanismos usados para aprender o sistema de sons da L1 são ativados na aprendizagem de uma LNM e permanecem intactos durante toda a vida (Flege, 1995). Estes estudos empíricos, que investigaram os efeitos do treino percetivo na perceção e produção de sons não nativos, reportaram a sua eficácia não somente na modificação de padrões percetivos, mas também na melhoria da capacidade produtiva dos mesmos, confirmando assim a plasticidade do sistema percetivo dos aprendentes de LNMs (por exemplo, Aliaga-Garcia, 2013; Pereira & Hazan, 2013; Wang, 2008; Wang et al., 2003). Os resultados de extensa investigação indicam que as dificuldades na perceção de contrastes vocálicos não nativos são uma parte significativa dos problemas que os aprendentes revelam na aquisição/aprendizagem fonológica de uma L2/LE (Strange, 2007). Portanto, no presente estudo, investigaram-se os efeitos do treino percetivo na aprendizagem de três contrastes vocálicos da língua inglesa (/i/-/ɪ/; /ɛ/-/æ/; /u/-/ʊ/) por um grupo de aprendentes de inglês como LE. Estes três contrastes foram escolhidos devido às dificuldades percetivas e produtivas que falantes nativos de português europeu revelam na sua aprendizagem (Flege, 1994, citado em Flege, 1995; Rato et al., 2013). As categorias fonológicas /ɪ/, /æ/ e /ʊ/ da L2/LE tendem a ser assimiladas como sons da L1, /i/, /ɛ/ e /u/, respetivamente, não se verificando qualquer distinção entre as vogais dos três contrastes, devido à sua proximidade acústica e articulatória. Especificamente, pretendeu-se (i) observar o efeito de um treino percetivo de alta variabilidade, que incluiu estímulos produzidos por vários locutores nativos em diferentes contextos fonológicos, na melhoria da capacidade percetiva dos segmentos-alvo; (ii) averiguar a transferência da melhoria para a produção oral; (iii) verificar a generalização para novos contextos e novos falantes; e (iv) analisar os efeitos do treino a longo prazo. A perceção dos participantes foi testada três vezes com uma tarefa auditiva de identificação com estímulos naturais: (1) antes do treino – pré-teste; (2) imediatamente depois do treino – pós-teste; e (3) dois meses mais tarde – teste de retenção. O programa de treino consistiu em cinco sessões, divididas em dois blocos, que incluíram tarefas de identificação e de discriminação auditivas seguidas de correção imediata. A produção foi testada, igualmente, em três fases, através da leitura de frases veículo, contendo palavras com os segmentos vocálicos. Os resultados demonstram que os aprendentes portugueses melhoraram significativamente na identificação das vogais-alvo e essa melhoria da sua competência percetiva manteve-se dois meses após o término do treino. Para além disso, os resultados do teste de generalização indicam que houve uma aprendizagem robusta dos dois contrastes vocálicos anteriores. As análises acústicas das produções dos informantes revelaram também uma transferência da aprendizagem para a produção oral. Estes resultados suportam a afirmação de que a aprendizagem ao nível da perceção de fala pode ocorrer em contextos formais, num curto período de tempo, e corroboram resultados anteriores sobre a maleabilidade dos sistemas percetuais fonológicos de aprendentes adultos de uma LNM.

This thesis, developed within the Conversational Analysis and Sociolinguistics frameworks, aims to investigate the discursive and linguistic strategies involved in the organization of the conversational interaction underlying the TV talk show interview. By focusing on a corpus of analysis which includes twenty-four interviews taken from two late night talk shows, The Tonight Show with Jay Leno and Late Night with Conan O’Brien, broadcast by the North-American channel NBC between September 2008 and January 2009, this study is intended to capture the defining features of this hybrid conversational interaction by comparing two variables. The first one is the institutional vs. non-institutional character of the conversational interaction. The second variable is the gender-based differences latent in the discourse and interactional behaviour of the interview participants. It is expected that the semi-institutional and gender-oriented nature of the talk of these interviews is displayed in the heterogeneity of the distinguishing features pertaining to both casual and monitored conversation in terms of sequence organization of the interview, turn-taking system, topic control, narrative construction, lexical choice and participants footing.

O presente trabalho, que se inscreve nas áreas de investigação da Análise Conversacional e da Sociolinguística, pretende investigar as estratégias discursivas e linguísticas que regem a organização da interacção conversacional das entrevistas dos talk shows televisivos. Partindo de um corpus de análise constituído por transcrições de vinte e quatro entrevistas dos talk shows nocturnos Late Night with Conan O’Brien e The Tonight Show with Jay Leno, transmitidos pelo canal norte-americano NBC entre Setembro de 2008 e Janeiro de 2009, pretende-se apresentar as especificidades desta interacção conversacional de natureza híbrida, através de uma análise comparativa de duas variáveis. Em primeiro lugar, as manifestações de institucionalidade em contraponto com as características particulares de uma interacção conversacional não-institucional. Em segundo lugar, as representações de género subjacentes ao discurso e ao comportamento interaccional dos intervenientes nas entrevistas em foco. Propomo-nos avaliar até que ponto estas duas dimensões de análise se reflectem numa heterogeneidade de aspectos específicos, quer das conversas espontâneas, quer das interacções conversacionais preparadas, nomeadamente nos padrões sequenciais da entrevista, nos mecanismos do sistema de alternância de vez, na gestão de tópico, na construção de narrativas, na escolha lexical e no posicionamento dos participantes na interacção comunicativa.