Special speech registers: Talking to infants, pets, the hearing impaired, foreigners and computers

Abstract: Humans speak in specific ways depending on the nature of the audience. For example, compared to adult-directed speech, Infant-Directed Speech has heightened pitch characteristics, heightened positive affect, and hyperarticulated vowels. Here the results of studies of Infant-Directed Speech (InfantDS), PetDS, ForeignerDS, HearingImpairedDS and ComputerDS are presented. The results will lead to the discussion of three issues. First, there are separable independent components of conversational speech: attentional, expressed via pitch characteristics of the voice; affective, expressed via rated emotion in the voice; and linguistic, expressed via didactic devices such as vowel hyperarticulation. Second, humans are particularly good at perceiving and acting upon, even unconsciously, these separable voice components. Third, humans are particularly good at perceiving the abilities and needs of their audience and responding to these by producing speech with differential weighting, even unconsciously, of the attentional, affective and linguistic components of their speech. The nature of special speech registers and the ways that humans use these will be discussed in terms of what makes a human conversation a meeting of minds; the relationship between language and thought; the nature of ComputerDS and what makes human conversations difficult to be replicated by embodied conversational agents.

Bio: Following an honours degree (University of New England, Australia, 1974), and Ph.D. (Monash University, 1980), Burnham was 18 years in the School of Psychology at the University of NSW, before taking up the position of Inaugural Director of MARCS Auditory Laboratories, Sydney, Australia http://marcs.uws.edu.au/. MARCS Labs has over 60 members and specialises in Speech and Music research. Burnham’s own research concerns five areas of speech perception and production: Ontogenetic studies of infants’ and children’s speech perception development; phonetic, attentional and emotional aspects of Special Speech Registers to infants, foreigners, pets, lovers, and computers; Cross-Language studies on the relationship between speech perception and vocabulary, reading , and second language learning; Auditory-Visual speech perception studies with infants, children and adults within and across languages; and Lexical Tone, an understudied but prevalent speech feature affording investigations of speech-music, and segmental-suprasegmental relationships. Burnham is President, Australasian Speech Science and Technology Association (ASSTA); Executive member, ARC Research Network on Human Communication Science (HCSNet); member of the International Advisory Council (IAC) and Interspeech Steering Committee (ISC) of the International Speech Communication Association (ISCA); and Co-founder/Vice-Chair, Auditory-Visual Speech Perception Association (AVISA). Burnham has been funded continuously by the Australian Research Council (ARC) and other grant bodies for the past 25 years, and has directed various large inter-disciplinary, international research projects. Currently he is lead CI on an ARC Discovery grant on tone perception (2009-2012); lead CI of an ARC and National Health and Medical Research Council Special Initiative, the $3.4M Thinking Head project (2005-2011) encompassing Embodied Conversational Agents and Human-Robot Interaction; and lead CI of the $1M ARC infrastructure project., the Big Australian Speech Corpus (Big ASC, 2010), in which 3 hours of speech data from 1,000 people all around Australia are being collected.