Talking to Infants, Pets and Lovers
HomeFull List of Titles
1: ICSLP'98 Proceedings
Keynote SpeechesText-To-Speech Synthesis 1Spoken Language Models and Dialog 1Prosody and Emotion 1Hidden Markov Model Techniques 1Speaker and Language Recognition 1Multimodal Spoken Language Processing 1Isolated Word RecognitionRobust Speech Processing in Adverse Environments 1Spoken Language Models and Dialog 2Articulatory Modelling 1Talking to Infants, Pets and LoversRobust Speech Processing in Adverse Environments 2Spoken Language Models and Dialog 3Speech Coding 1Articulatory Modelling 2Prosody and Emotion 2Neural Networks, Fuzzy and Evolutionary Methods 1Utterance Verification and Word Spotting 1 / Speaker Adaptation 1Text-To-Speech Synthesis 2Spoken Language Models and Dialog 4Human Speech Perception 1Robust Speech Processing in Adverse Environments 3Speech and Hearing Disorders 1Prosody and Emotion 3Spoken Language Understanding Systems 1Signal Processing and Speech Analysis 1Spoken Language Generation and Translation 1Spoken Language Models and Dialog 5Segmentation, Labelling and Speech Corpora 1Multimodal Spoken Language Processing 2Prosody and Emotion 4Neural Networks, Fuzzy and Evolutionary Methods 2Large Vocabulary Continuous Speech Recognition 1Speaker and Language Recognition 2Signal Processing and Speech Analysis 2Prosody and Emotion 5Robust Speech Processing in Adverse Environments 4Segmentation, Labelling and Speech Corpora 2Speech Technology Applications and Human-Machine Interface 1Large Vocabulary Continuous Speech Recognition 2Text-To-Speech Synthesis 3Language Acquisition 1Acoustic Phonetics 1Speaker Adaptation 2Speech Coding 2Hidden Markov Model Techniques 2Multilingual Perception and Recognition 1Large Vocabulary Continuous Speech Recognition 3Articulatory Modelling 3Language Acquisition 2Speaker and Language Recognition 3Text-To-Speech Synthesis 4Spoken Language Understanding Systems 4Human Speech Perception 2Large Vocabulary Continuous Speech Recognition 4Spoken Language Understanding Systems 2Signal Processing and Speech Analysis 3Human Speech Perception 3Speaker Adaptation 3Spoken Language Understanding Systems 3Multimodal Spoken Language Processing 3Acoustic Phonetics 2Large Vocabulary Continuous Speech Recognition 5Speech Coding 3Language Acquisition 3 / Multilingual Perception and Recognition 2Segmentation, Labelling and Speech Corpora 3Text-To-Speech Synthesis 5Spoken Language Generation and Translation 2Human Speech Perception 4Robust Speech Processing in Adverse Environments 5Text-To-Speech Synthesis 6Speech Technology Applications and Human-Machine Interface 2Prosody and Emotion 6Hidden Markov Model Techniques 3Speech and Hearing Disorders 2 / Speech Processing for the Speech and Hearing Impaired 1Human Speech ProductionSegmentation, Labelling and Speech Corpora 4Speaker and Language Recognition 4Speech Technology Applications and Human-Machine Interface 3Utterance Verification and Word Spotting 2Large Vocabulary Continuous Speech Recognition 6Neural Networks, Fuzzy and Evolutionary Methods 3Speech Processing for the Speech-Impaired and Hearing-Impaired 2Prosody and Emotion 72: SST Student Day
SST Student Day - Poster Session 1SST Student Day - Poster Session 2Author Index
A B C D E F G H I
J K L M N O P Q R
S T U V W X Y Z
Multimedia Files
Acoustic And Affective Qualities Of IDS In English
Authors:
Christine Kitamura, University of New South Wales (Australia)
Denis Burnham, University of New South Wales (Australia)
Page (NA) Paper number 909
Abstract:
This study examines modifications to the acoustic and affective qualities of infant-direct speech (IDS) from birth to 12 months. The acoustic analysis of fundamental frequency shows that IDS has the highest level of mean fundamental frequency (F 0 ) in speech to infants at 6 months, and the highest level of pitch range in speech to infants at 12 months. Sex-based differences were evident with mothers using higher mean-F 0 and pitch range in speech to female than male infants. The mother's speech samples were also rated on five scales of communicative intent. Two factors, labelled 'Affective' and 'Attentional', were extracted from these scales. Analysis of variance of the derived factor scores shows that mothers have the highest scores on the Affective factor at 6 and 12 months, while the maximum on the Attentional factor occurs at 9 months. Mothers also increase their use of both these IDS components more in speech to girls than boys.
SL980909.PDF (From Author)
SL980909.PDF (Rasterized)
TOP
Acoustic Qualities Of IDS And ADS In Thai
Authors:
Chayada Thanavisuth, Department of Linguistics, Faculty of Arts, Chulalongkorn University (Thailand)
Sudaporn Luksaneeyanawin, Department of Linguistics, Faculty of Arts, Chulalongkorn University (Thailand)
Page (NA) Paper number 912
Abstract:
Infant Directed Speech (IDS) samples were collected longitudinally from six mothers speaking to their infants (three with female infants and three with male infants) when their infants were newborns, 3 months, 6 months, 9 months and 12 months of age. Adult Directed Speech (ADS) samples of these mothers were also collected to be compared with these IDS samples. This maternal speech was then examined in terms of the phonetic characteristics, specially the prosodic aspects-pitch, tempo, and loudness. Twenty utterances from each of the six mothers speaking to their infants at each of the five ages and to an adult, totaling 720 utterances (20 utterances x 6 mothers x 6 age groups) were acoustically analyzed. They are fundamental frequencies, frequency ranges, utterance duration, syllable duration, numbers of syllable per utterance, and intensity values. Results of the comparison between IDS and ADS showed that IDS used higher fundamental frequencies (263.3 Hz vs 247.99 Hz), greater frequency range in semitones (15.16 st vs 13.5 st), shorter utterance duration (1106 ms vs 1506 ms), longer syllable duration (343 ms vs 220 ms), and less numbers of syllable per utterance (3.85 vs 7.63). Concerning the age-related changes of prosodic features, there seems to be three age intervals which show unique modification to the pattern of prosodic characteristics. The first age group is the newborns as opposed to other age groups, mothers use longer utterance duration, longer syllable duration, and more numbers of syllable per utterance. In the 3- and the 6-month-old periods, the mothers' speech style changes to an increase of the mean Ff and pitch modulation. By the time the infants reach 9 and 12 months of age, there is a successive decrease of Ff of the speech directed to the 6MO to the 9MO and the 12MO. Most prosodic features of the IDS directed to the 12MO are almost similar to those of the ADS. The study of intensity values in IDS fails to reveal any significant variation due to a general phonetic problem in the instrumental study of the degree of loudness in speech. The variations of the phonetic characteristics among the 6 groups in all aspects were statistically analyzed using the Analysis of Variance (ANOVA). It was found that every variation of the prosodic features investigated across age groups was statistically significant. It seems that these phonetic characteristics in the mothers' speech are adjusted to accommodate the communicative functions between the mothers and their infants and these functions have developed according to other developments of the infants.
SL980912.PDF (From Author)
SL980912.PDF (Rasterized)
TOP
Pragmatic Characteristics of Infant Directed Speech
Authors:
Sudaporn Luksaneeyanawin, Chulalongkorn University (Thailand)
Chayada Thanavisuth, Chulalongkorn University (Thailand)
Suthasinee Sittigasorn, Chulalongkorn University (Thailand)
Onwadee Rukkarangsarit, Chulalongkorn University (Thailand)
Page (NA) Paper number 913
Abstract:
A longitudinal study of the pragmatic characteristics of the Infant Directed Speech (IDS) of middle class Thai mothers when talking to their babies at birth (Newborn), and when the infants were 3-, 6-, 9-, and 12-months-old was conducted. Three aspects were investigated: (1) the verbal acts which the mothers use with their infants using the framework of Thai linguistic action verbs (LAVs), and speech act verbs (SAVs) found in IDS; (2) the use of interrogative sentences in IDS both in quantitative and qualitative terms, and (3) the use of final particles incorporating both the status and mood particles which the mothers use in IDS. Regarding SAVs, overall IDS was made up of 38.6% expressives, 23.2% assertives, 20% directives, 15.3% questions, and 2.4% interaction-management. Comparison of SAVs in IDS directed to the five age groups shows a similarity in the pragmatic characteristics of the 3-and 6-month-olds, as opposed to the 9- and 12-month-olds. IDS to Newborns is unique compared with these two groups. It was found that 27.8 % of the IDS utterances were interrogatives, with a peak incidence of interrogatives at 3 months. The incidence of Yes-No interrogatives decreased as the age of the infants increased, whereas the incidence of WH- interrogatives increased as the age of the infants increased. Regarding the type of information mothers were seeking in WH- interrogatives, VP information-seeking is prominent for the Newborn and 3-month speech, but decreased as the infants got older. NP information-seeking was low for speech to the Newborns and 3-month-olds, and increased in speech to the 6-, 9-, and 12-month-olds. The information-seeking in adjuncts (location, time, manner etc.) was high in the 9- and 12-month-olds, when the infants were able to move about and started to explore their physical world. Interrogatives are used mostly for didactic functions; mothers asked and then answered the questions themselves. A large number of repetitive question forms are found, and these are used to draw the attention of infants in the mother-child interaction. Analysis of the use of the final particles reflects very clear didactic and emotive functions in IDS. The pragmatic characteristics of IDS during the first twelve months of the infants' lives varied significantly across ages. Mothers adapted their communication strategies in talking to their babies. These variations were found to help accommodate the mother-child interaction from the time when the infants were not able to communicate at all (Newborns), to the time when they started to communicate with nonverbal acts such as gazing, smiling and crying with eye contact (3 to 6 months), and in turn to the time when they started to communicate with hand touching and rudimentary verbal communication (9 to 12 months). IDS input is not just input data for infants to learn linguistic forms, but plays an important role in the development of the linguistic behaviour with the ultimate goal of communicating with other human beings successfully.
SL980913.PDF (From Author)
SL980913.PDF (Rasterized)
TOP
Are You My Little Pussy-Cat? Acoustic, Phonetic And Affective Qualities Of Infant- And Pet-Directed Speech
Authors:
Denis Burnham, University of NSW / University of Western Sydney Macarthur (Australia)
Elizabeth Francis, University of NSW (Australia)
Ute Vollmer-Conna, University of NSW (Australia)
Christine Kitamura, University of NSW (Australia)
Vicky Averkiou, University of NSW (Australia)
Amanda Olley, University of NSW (Australia)
Mary Nguyen, University of NSW (Australia)
Cal Paterson, University of NSW (Australia)
Page (NA) Paper number 916
Abstract:
It has been suggested that infant-directed speech (IDS) is a special speech register conducive to gaining infants' attention, eliciting and expressing affect, and teaching infants about the phonology of the ambient language. Certain similarities have been observed between IDS and pet-directed speech (PDS), but until now these have not been systematically studied. In this study IDS, PDS, and adult directed speech (ADS) are compared on acoustic, phonetic, and affective measures. It has been suggested that IDS should be more tuned to a linguistic didactic function than PDS, but that the two should be similar acoustically and affectively. The results show that IDS and PDS are generally different from ADS, but that IDS and PDS do not differ substantially, either acoustically, or phonetically. As the expected difference between IDS and PDS was not obtained, it could be suggested that dependent variables thought to measure the functions of special speech registers may not necessarily do so unequivocally.
SL980916.PDF (From Author)
SL980916.PDF (Rasterized)
TOP
Special Speech Registers: Talking To Australian And Thai Infants, And To Pets
Authors:
Denis Burnham, University of NSW / University of Western Sydney Macarthur (Australia)
Page (NA) Paper number 915
Abstract:
Infant directed speech (IDS) is the special speech register used to talk to human infants. This paper provides an historical and theoretical background to the study of IDS and special speech registers, as an introduction to four papers which follow. These are concerned with Australian IDS [Kitamura & Burnham], Thai IDS [Thanavisuth & Luksaneeyanawin; Luksaneeyanawin, Thanavisuth, Sittigasorn, & Rukkarangsarit], and Australian pet directed speech [Burnham, Francis, Vollmer-Conna, Kitamura, Averkiou, Olley, Nguyen, & Paterson].