The effects of sentence structure on consonant and vowel articulations. Whilst machine-generated speech is generally of good quality at the level of the single word, it is noticeably less natural-sounding at the level of the sentence. This project examines an important aspect of the naturalness of human speech, namely, the effect that sentence structure has on individual consonants and vowels. A break-down of this naturalness is seen in some speakers who have suffered traumatic brain injury: suc ....The effects of sentence structure on consonant and vowel articulations. Whilst machine-generated speech is generally of good quality at the level of the single word, it is noticeably less natural-sounding at the level of the sentence. This project examines an important aspect of the naturalness of human speech, namely, the effect that sentence structure has on individual consonants and vowels. A break-down of this naturalness is seen in some speakers who have suffered traumatic brain injury: such speakers perform well when asked to utter a short word, but struggle when asked to produce a longer string of sounds. A better understanding of the interaction between speech sounds and sentence structure will lead to improvements in the treatment of speech disorders, and in the quality of human-machine communication.Read moreRead less
How Strict is the Mother Tongue? Using Dialects to Probe Early Speech Perception and Word Recognition. This project will 1)advance knowledge of toddler word representations and their developmental precursors; 2) contribute to theories of phonological vs phonetic properties of spoken language; 3) explain how experience with the ambient language shapes children's phonological and lexical development. Moreover, the findings will 4) offer crucial new insights into sources of developmental disorders ....How Strict is the Mother Tongue? Using Dialects to Probe Early Speech Perception and Word Recognition. This project will 1)advance knowledge of toddler word representations and their developmental precursors; 2) contribute to theories of phonological vs phonetic properties of spoken language; 3) explain how experience with the ambient language shapes children's phonological and lexical development. Moreover, the findings will 4) offer crucial new insights into sources of developmental disorders (language delay, dyslexia) leading to improved early diagnosis and treatment; 5) bear on issues of second language learning; and by understanding the process by which young learners handle dialect variability, 6) provide insights into how automatic speech recognition systems can be made more robust to dialectal and foreign accent differences.
Read moreRead less
Development of second language phonetic and phonological categories. The majority of the world's population speaks two or more languages, yet we know little about how multiple languages are accommodated within a single speaker. Why do children appear to learn a second language 'like a native' but adults invariably develop a clearly perceptible foreign accent? This project investigates a little-known fact - adult second language speakers also 'hear' with a foreign accent. As Australia becomes inc ....Development of second language phonetic and phonological categories. The majority of the world's population speaks two or more languages, yet we know little about how multiple languages are accommodated within a single speaker. Why do children appear to learn a second language 'like a native' but adults invariably develop a clearly perceptible foreign accent? This project investigates a little-known fact - adult second language speakers also 'hear' with a foreign accent. As Australia becomes increasingly multilingual increasing our understanding of the human capacity for language learning would strengthen Australia's social and economic fabric by leading to improved educational practices, work prospects for migrants and, most importantly, understanding between cultures.Read moreRead less
An instrumental investigation of consonant sequences in a northern Australian language. Australian indigenous languages are of great interest, due in part to their unique phonetic structure relative to many other languages of the world. Most advances in speech science and phonetic theory are based on studies of English, or other European languages, yet an important goal of phonetic science is to account for speaking and listening processes that are deemed to be universal. Our proposal seeks ....An instrumental investigation of consonant sequences in a northern Australian language. Australian indigenous languages are of great interest, due in part to their unique phonetic structure relative to many other languages of the world. Most advances in speech science and phonetic theory are based on studies of English, or other European languages, yet an important goal of phonetic science is to account for speaking and listening processes that are deemed to be universal. Our proposal seeks to address key aspects of current phonetic theory and models of speech sound production, by providing data from an indigenous Australian language.Read moreRead less
Unveiling the mystery of tone perception: How does native language prosody affect adults' perception of foreign tones? This project provides new knowledge about how humans perceive non-native lexical tone categories. The results of this large-scale cross-language study will indicate how native languages constrain human perception of non-native speech (consonants, vowels, intonation). These data will greatly facilitate the research in second language acquisition and teaching, speech perception mo ....Unveiling the mystery of tone perception: How does native language prosody affect adults' perception of foreign tones? This project provides new knowledge about how humans perceive non-native lexical tone categories. The results of this large-scale cross-language study will indicate how native languages constrain human perception of non-native speech (consonants, vowels, intonation). These data will greatly facilitate the research in second language acquisition and teaching, speech perception modeling including automatic speech recognition with tone languages, human speech processing. The findings will also be useful to clinical speech and hearing research for speech and hearing impaired persons, and commercial applications in foreign language teaching, and computer assisted language learning for language learners.Read moreRead less
Rising Intonation in Australian English. The project will model the intonation of Australian English through an analysis of spoken dialogues from over 200 speakers. The study of the different types of rising intonation across different population groups will advance our knowledge of how Australian English is different from other accents, how it has changed in the last 40 years, and how accent is related to social class in Australia. The outcomes include publications on the theory of intonation, ....Rising Intonation in Australian English. The project will model the intonation of Australian English through an analysis of spoken dialogues from over 200 speakers. The study of the different types of rising intonation across different population groups will advance our knowledge of how Australian English is different from other accents, how it has changed in the last 40 years, and how accent is related to social class in Australia. The outcomes include publications on the theory of intonation, accent change, and its relationship to social class as well as a working model of intonation that can be implemented in a system for synthesising Australian English speech.Read moreRead less
The relationship between speech production and perception in Australian language speakers: implications for speech development and learning in Aboriginal children. Chronic ear infection blights the life of at least 50% of Aboriginal Australians. In a vicious cycle that extends from generation to generation, it leads to hearing loss, educational disadvantage, socio-economic disadvantage and environmental depredation, which once again leads to ear (and many other) infections. This is a unique atte ....The relationship between speech production and perception in Australian language speakers: implications for speech development and learning in Aboriginal children. Chronic ear infection blights the life of at least 50% of Aboriginal Australians. In a vicious cycle that extends from generation to generation, it leads to hearing loss, educational disadvantage, socio-economic disadvantage and environmental depredation, which once again leads to ear (and many other) infections. This is a unique attempt by researchers across academic disciplines to study the role of language in educational disadvantage and whether this disadvantage might be made worse for Aboriginal children by the early use of English at school. We ask whether, on purely acoustic or linguistic grounds, communicating in an Aboriginal language might offer improved educational and health outcomes for Aboriginal children in the early years.Read moreRead less
Uptalk in Australian English Intonation. Australian English speakers often use rising instead of falling intonation at the end of sentences which are not questions. This phenomenon of uptalk is one of the unique traits of the Australian English accent. The project will model uptalk by analyzing spoken dialogues from 150 speakers from three major populations, Sydney, Melbourne, and regional Victoria. The intonation patterns of regional and non-Sydney populations have been poorly investigated ....Uptalk in Australian English Intonation. Australian English speakers often use rising instead of falling intonation at the end of sentences which are not questions. This phenomenon of uptalk is one of the unique traits of the Australian English accent. The project will model uptalk by analyzing spoken dialogues from 150 speakers from three major populations, Sydney, Melbourne, and regional Victoria. The intonation patterns of regional and non-Sydney populations have been poorly investigated, so this kind of study is needed to get a more complete understanding of Australian English intonation and the Australian accent. Outcomes will include publications on intonation and laboratory phonology, sociophonetics, and a working model of intonation that can be implemented in speech output systems for Australian English.Read moreRead less
English as a spoken lingua franca in ASEAN: a study of its linguistic and socio-cultural features. The project will collect the first corpus of Lingua Franca English in Asia. It will thus complement the two European LFE projects currently underway and will put Australia at the forefront of this new research. Given that the findings of the research have the potential to revolutionise the ways English is taught and learned throughout the ASEAN region, as it takes the focus away from the impositio ....English as a spoken lingua franca in ASEAN: a study of its linguistic and socio-cultural features. The project will collect the first corpus of Lingua Franca English in Asia. It will thus complement the two European LFE projects currently underway and will put Australia at the forefront of this new research. Given that the findings of the research have the potential to revolutionise the ways English is taught and learned throughout the ASEAN region, as it takes the focus away from the imposition of an external native speaker norm of English. The research thus places Australia in the forefront of culturally and contextually sensitive work into English language teaching and research. The project will also benefit all communities within Australia where English plays a role as a lingua franca.Read moreRead less
Using written language to probe speech recognition models. Speech recognition models fall into two principal classes, with fundamentally different processing architectures. Feedback models (e.g. TRACE, McClelland & Elman, 1986) allow lexical knowledge to exert top-down control over phonemic analysis. Feedforward models (e.g. Merge, Norris, McQueen & Cutler, 2000) assume that information flow is entirely bottom-up. Our project adopts an innovative approach to testing between these model classe ....Using written language to probe speech recognition models. Speech recognition models fall into two principal classes, with fundamentally different processing architectures. Feedback models (e.g. TRACE, McClelland & Elman, 1986) allow lexical knowledge to exert top-down control over phonemic analysis. Feedforward models (e.g. Merge, Norris, McQueen & Cutler, 2000) assume that information flow is entirely bottom-up. Our project adopts an innovative approach to testing between these model classes, by examining the influence of written-word knowledge on speech perception. To distinguish the models, contrasts must test different processing levels and examine strategy effects. TRACE favors broad effects with limited strategic influence; Merge favors lexical effects that are necessarily sensitive to strategic factorsRead moreRead less