Denis Burnham - Academia.edu (original) (raw)
Papers by Denis Burnham
8th Annual Conference of the International Speech Communication Association, Aug 27, 2007
A number of studies have shown that mothers hyperarticulate vowels in their Infant Directed Speec... more A number of studies have shown that mothers hyperarticulate vowels in their Infant Directed Speech to their 6-month-old infants. Here we investigate the possibility that such hyperarticulation might also occur for lexical tone for mother-infant dyads in tone language environments, and possible changes in mothers’ vowel and such tone hyperarticulation in IDS across the infant’s first year. IDS from a total of 22 native Cantonese speaking mothers was recorded, 11 when their infants were 3-, 6-, and 9-month-old, and another 11 when their infants were 6-, 9-, and 12-monthold, and mothers asked to use nine target words in their speech; one for each of the three corner vowels /i/, /a/ and /u/), and another six for each of the Cantonese tones on the vowel /i/. Vowel hyperarticulation was investigated using first and second formant values, and tones using fundamental frequency onset and offset [1]. Preliminary results for 5 mothers presented here indicate that both vowel and tone hyperarticulation occur, but that while vowel hyperarticulation emerges around 6 months and increases from 6 to 9 to 12 month, tone hyperarticulation occurs only at 6 and 9 months. The results suggest that, as for vowels, tone space is hyperarticulated in IDS, but returns to Adult Directed Speech levels earlier for tones than vowels. Possible reasons for this are discussed, as are future studies with other tone languages with smaller tonal inventories.
Number of vowel tokens elicited from each participant in the infant-directed speech (IDS), adult-... more Number of vowel tokens elicited from each participant in the infant-directed speech (IDS), adult-directed speech (ADS), and exaggerated speech (ES) conditions.
This paper presents a study of audio-visual production of the four Mandarin lexical tones on word... more This paper presents a study of audio-visual production of the four Mandarin lexical tones on words in citation form and in sentences. OPTOTRAK motion capture data of the head and face of a Mandarin speaker were modelled using both PCA and guided-PCA. For each tone, correlations between F0 values and the different face and head components were calculated. Results show that there are visual parameters related to the different F0 patterns of each tone. Moreover differences were found in both duration and correlational patterns between words produced in citation and in sentential forms. The results show that there are identifiable visual correlates of lexical tone but the difference between citation and sentential forms has implications for materials used in production and perception studies of Mandarin lexical tones, and possibly those in other languages. Index Terms: audiovisual speech production, tone languages, Mandarin, OPTOTRAK, motion capture. 1.
When talking to babies, adults invariably use a special speech register characterized by elevated... more When talking to babies, adults invariably use a special speech register characterized by elevated fundamental frequency (pitch), exaggerated into-nation contours, and high affect (1, 2). It has also been found that mothers hyperarticulate vowels when addressing their infants but not when speaking to other adults (3). This phenomenon is ubiquitous, occurring across various languages— English, Russian, Swedish, and Japanese (3, 4)— and is thought to facilitate infants ’ linguistic de-velopment by amplifying the phonetic character-istics of native language vowels (3). However, the very ubiquity of this speech style means that it is practically impossible to obtain direct evi-dence of its function as a language-teaching de-
This paper presents results concerning the exploitation of visual cues in the perception of Manda... more This paper presents results concerning the exploitation of visual cues in the perception of Mandarin tones. The lower part of a female speaker's face was recorded on digital video as she uttered 25 sets of syllabic tokens covering the four different tones of Mandarin. Then in a perception study the audio sound track alone, as well an audio plus video condition were presented to native Mandarin speakers who were required to decide which tone they perceived. Audio was presented in various conditions: clear, babble-noise masked at different SNR levels, as well as devoiced and amplitudemodulated noise conditions using LPC resynthesis. In the devoiced and the clear audio conditions, there is little augmentation of audio alone due to the addition of video. However, the addition of visual information did significantly improve perception in the babble-noise masked condition, and this effect increased with decreasing SNR. This outcome suggests that the improvement in noise-masked condit...
Hearing Eye II, 2013
... age by infants (Kuhl et al., 1992), and that this ability may be constrained and guided by pr... more ... age by infants (Kuhl et al., 1992), and that this ability may be constrained and guided by pre-setlanguage-general phonetic values (Polka, 1995 ... For consonants this modulation is first evident at 7-11 months, when sensitivity to non-native consonant contrasts is gradually ...
Sixty Australian English speaking toddlers were tested in a longitudinal study at 30, 33, and 36 ... more Sixty Australian English speaking toddlers were tested in a longitudinal study at 30, 33, and 36 months on vocabulary size, phoneme sensitivity, language-specific speech perception, and articulation accuracy. Vocabulary size was measured with the Australian English adaptation of the MacArthur Communicative Development Inventory and the Peabody Picture Vocabulary Test III. Phoneme Sensitivity (PS) comprised scores from mispronunciation detection, rhyme detection, and nonword repetition tasks. Language Specific Speech Perception (LSSP) was calculated by subtracting the score for nonnative speech perception from the native score, indicating the degree of specialisation in the native language. Articulation Accuracy (AA) was measured with an adaptation of the Queensland Articulation Test. Results showed (i) linear improvements in all new measures, appropriately depicting the developmental trend; (ii) significant correlations between AA and vocabulary size; (iii) predictability of vocabul...
This study investigated the effects of aging, auditory and auditory-visual perception of lexical ... more This study investigated the effects of aging, auditory and auditory-visual perception of lexical tone of native Thai listeners. Elderly and younger Adults’ discrimination of the 5 Thai tones was investigated in audio-visual (AV), audio-only (AO), and visual-only (VO) conditions at two inter-stimulus intervals (ISIs) [500 and 1500 ms] in clear and noisy conditions. Generally, the Elderly performed more poorly than the Adults, but in both groups there was similar ranking of the relative discriminability of tone contrasts. Notably, in noise there was better tone discrimination in AV than in AO, and this was equally the case for young Adults and the Elderly. This shows that the elderly can and do use visual speech information to augment auditory perception of tone. The Elderly also benefitted more from the 1500 ms ISI suggesting that their tone perception is better when a more phonological (rather than acoustic) mode of processing is available.
Research on tone languages is relatively scarce compared with that on non-tonal languages. Nowher... more Research on tone languages is relatively scarce compared with that on non-tonal languages. Nowhere is this more apparent than in first language (L1) acquisition, especially the acquisition of lexical tone production. Here we present the first comprehensive lexical tone production data from 120 children (from 18 months to 6 years of age) and 12 adults learning Thai as their L1. Participants’ pronunciations of each of 20 words (4 with each of the 3 level and 2 contour Thai tones) were recorded and analysed for fundamental frequency (F0) at 13 points along the course of the production. Analyses of normalized F0 over time suggest that tone plots become more differentiated over age. To investigate this more closely, tone spaces were computed via plots of F0 offset against F0 onset. Tone ellipses based on these were used to derive Tone Differentiation Scores. Analyses of these show linear improvement over age which was more pronounced for level-level than level-contour tones. Over and abo...
In a previous study we have found that non-tone language speakers are able to form lexical tone c... more In a previous study we have found that non-tone language speakers are able to form lexical tone categories through extracting frequency distribution in training, but only when attention is directed towards the distribution [12]. This study extends the distributional learning literature by investigating how tone language speakers’ linguistic experience with tones affects their distributional learning of non-native lexical tones. Native Mandarin listeners were presented with a Thai lexical tone minimal pair distributed either unimodally (promoting formation of a single category), or bimodally (promoting two category formation). Assessment of performance in a discrimination task before and after exposure showed that the Bimodal Distribution group improved significantly from Pretest to Posttest whereas the Unimodal group did not. These results suggest that tone language speakers capitalise on their experience in using pitch phonemically to form the appropriate number of lexical tone cat...
Due to limitations of cochlear implant (CI) technology in transmitting fundamental frequency (f0)... more Due to limitations of cochlear implant (CI) technology in transmitting fundamental frequency (f0), there is poor perception of lexical tone by CI users. Here simulated cochlear implant audio is used to investigate the role of visual speech information in Mandarin tone discrimination. Results showed that visual information improved tone perception in both Mandarin and Australian English listeners, particularly in CI simulation conditions, and particularly for tone pairs where tonal distinctions are cued by non-f0 components, suggesting that tone discrimination in CI users can be trained.
Newborn infants come equipped with a sophisticated auditory system, allowing them to perceive fin... more Newborn infants come equipped with a sophisticated auditory system, allowing them to perceive fine grained phonetic distinctions. This phonetic detail is not audible through the lowpass filter of the mother’s womb, whereas prosodic information is. Thus when born, infants have speech preferences based on prosody not on phonetic segments. The experiments here concern the role of prosodic and phonetic information in infants’ preference for their native language. English language environment infants of 6, 71⁄2, and 9 months were tested for their preference of two lists, one of Thai and one of English words. The availability of lexical tone and stress information were manipulated by using 1and 2-syllable words, and of prosodic and phonetic information by low-pass filtering. The results show that infants cannot use the tonal information on 1-syllable words to recognise their native language, and that stress and phonetic information work in tandem for native language recognition.
8th Annual Conference of the International Speech Communication Association, Aug 27, 2007
A number of studies have shown that mothers hyperarticulate vowels in their Infant Directed Speec... more A number of studies have shown that mothers hyperarticulate vowels in their Infant Directed Speech to their 6-month-old infants. Here we investigate the possibility that such hyperarticulation might also occur for lexical tone for mother-infant dyads in tone language environments, and possible changes in mothers’ vowel and such tone hyperarticulation in IDS across the infant’s first year. IDS from a total of 22 native Cantonese speaking mothers was recorded, 11 when their infants were 3-, 6-, and 9-month-old, and another 11 when their infants were 6-, 9-, and 12-monthold, and mothers asked to use nine target words in their speech; one for each of the three corner vowels /i/, /a/ and /u/), and another six for each of the Cantonese tones on the vowel /i/. Vowel hyperarticulation was investigated using first and second formant values, and tones using fundamental frequency onset and offset [1]. Preliminary results for 5 mothers presented here indicate that both vowel and tone hyperarticulation occur, but that while vowel hyperarticulation emerges around 6 months and increases from 6 to 9 to 12 month, tone hyperarticulation occurs only at 6 and 9 months. The results suggest that, as for vowels, tone space is hyperarticulated in IDS, but returns to Adult Directed Speech levels earlier for tones than vowels. Possible reasons for this are discussed, as are future studies with other tone languages with smaller tonal inventories.
Number of vowel tokens elicited from each participant in the infant-directed speech (IDS), adult-... more Number of vowel tokens elicited from each participant in the infant-directed speech (IDS), adult-directed speech (ADS), and exaggerated speech (ES) conditions.
This paper presents a study of audio-visual production of the four Mandarin lexical tones on word... more This paper presents a study of audio-visual production of the four Mandarin lexical tones on words in citation form and in sentences. OPTOTRAK motion capture data of the head and face of a Mandarin speaker were modelled using both PCA and guided-PCA. For each tone, correlations between F0 values and the different face and head components were calculated. Results show that there are visual parameters related to the different F0 patterns of each tone. Moreover differences were found in both duration and correlational patterns between words produced in citation and in sentential forms. The results show that there are identifiable visual correlates of lexical tone but the difference between citation and sentential forms has implications for materials used in production and perception studies of Mandarin lexical tones, and possibly those in other languages. Index Terms: audiovisual speech production, tone languages, Mandarin, OPTOTRAK, motion capture. 1.
When talking to babies, adults invariably use a special speech register characterized by elevated... more When talking to babies, adults invariably use a special speech register characterized by elevated fundamental frequency (pitch), exaggerated into-nation contours, and high affect (1, 2). It has also been found that mothers hyperarticulate vowels when addressing their infants but not when speaking to other adults (3). This phenomenon is ubiquitous, occurring across various languages— English, Russian, Swedish, and Japanese (3, 4)— and is thought to facilitate infants ’ linguistic de-velopment by amplifying the phonetic character-istics of native language vowels (3). However, the very ubiquity of this speech style means that it is practically impossible to obtain direct evi-dence of its function as a language-teaching de-
This paper presents results concerning the exploitation of visual cues in the perception of Manda... more This paper presents results concerning the exploitation of visual cues in the perception of Mandarin tones. The lower part of a female speaker's face was recorded on digital video as she uttered 25 sets of syllabic tokens covering the four different tones of Mandarin. Then in a perception study the audio sound track alone, as well an audio plus video condition were presented to native Mandarin speakers who were required to decide which tone they perceived. Audio was presented in various conditions: clear, babble-noise masked at different SNR levels, as well as devoiced and amplitudemodulated noise conditions using LPC resynthesis. In the devoiced and the clear audio conditions, there is little augmentation of audio alone due to the addition of video. However, the addition of visual information did significantly improve perception in the babble-noise masked condition, and this effect increased with decreasing SNR. This outcome suggests that the improvement in noise-masked condit...
Hearing Eye II, 2013
... age by infants (Kuhl et al., 1992), and that this ability may be constrained and guided by pr... more ... age by infants (Kuhl et al., 1992), and that this ability may be constrained and guided by pre-setlanguage-general phonetic values (Polka, 1995 ... For consonants this modulation is first evident at 7-11 months, when sensitivity to non-native consonant contrasts is gradually ...
Sixty Australian English speaking toddlers were tested in a longitudinal study at 30, 33, and 36 ... more Sixty Australian English speaking toddlers were tested in a longitudinal study at 30, 33, and 36 months on vocabulary size, phoneme sensitivity, language-specific speech perception, and articulation accuracy. Vocabulary size was measured with the Australian English adaptation of the MacArthur Communicative Development Inventory and the Peabody Picture Vocabulary Test III. Phoneme Sensitivity (PS) comprised scores from mispronunciation detection, rhyme detection, and nonword repetition tasks. Language Specific Speech Perception (LSSP) was calculated by subtracting the score for nonnative speech perception from the native score, indicating the degree of specialisation in the native language. Articulation Accuracy (AA) was measured with an adaptation of the Queensland Articulation Test. Results showed (i) linear improvements in all new measures, appropriately depicting the developmental trend; (ii) significant correlations between AA and vocabulary size; (iii) predictability of vocabul...
This study investigated the effects of aging, auditory and auditory-visual perception of lexical ... more This study investigated the effects of aging, auditory and auditory-visual perception of lexical tone of native Thai listeners. Elderly and younger Adults’ discrimination of the 5 Thai tones was investigated in audio-visual (AV), audio-only (AO), and visual-only (VO) conditions at two inter-stimulus intervals (ISIs) [500 and 1500 ms] in clear and noisy conditions. Generally, the Elderly performed more poorly than the Adults, but in both groups there was similar ranking of the relative discriminability of tone contrasts. Notably, in noise there was better tone discrimination in AV than in AO, and this was equally the case for young Adults and the Elderly. This shows that the elderly can and do use visual speech information to augment auditory perception of tone. The Elderly also benefitted more from the 1500 ms ISI suggesting that their tone perception is better when a more phonological (rather than acoustic) mode of processing is available.
Research on tone languages is relatively scarce compared with that on non-tonal languages. Nowher... more Research on tone languages is relatively scarce compared with that on non-tonal languages. Nowhere is this more apparent than in first language (L1) acquisition, especially the acquisition of lexical tone production. Here we present the first comprehensive lexical tone production data from 120 children (from 18 months to 6 years of age) and 12 adults learning Thai as their L1. Participants’ pronunciations of each of 20 words (4 with each of the 3 level and 2 contour Thai tones) were recorded and analysed for fundamental frequency (F0) at 13 points along the course of the production. Analyses of normalized F0 over time suggest that tone plots become more differentiated over age. To investigate this more closely, tone spaces were computed via plots of F0 offset against F0 onset. Tone ellipses based on these were used to derive Tone Differentiation Scores. Analyses of these show linear improvement over age which was more pronounced for level-level than level-contour tones. Over and abo...
In a previous study we have found that non-tone language speakers are able to form lexical tone c... more In a previous study we have found that non-tone language speakers are able to form lexical tone categories through extracting frequency distribution in training, but only when attention is directed towards the distribution [12]. This study extends the distributional learning literature by investigating how tone language speakers’ linguistic experience with tones affects their distributional learning of non-native lexical tones. Native Mandarin listeners were presented with a Thai lexical tone minimal pair distributed either unimodally (promoting formation of a single category), or bimodally (promoting two category formation). Assessment of performance in a discrimination task before and after exposure showed that the Bimodal Distribution group improved significantly from Pretest to Posttest whereas the Unimodal group did not. These results suggest that tone language speakers capitalise on their experience in using pitch phonemically to form the appropriate number of lexical tone cat...
Due to limitations of cochlear implant (CI) technology in transmitting fundamental frequency (f0)... more Due to limitations of cochlear implant (CI) technology in transmitting fundamental frequency (f0), there is poor perception of lexical tone by CI users. Here simulated cochlear implant audio is used to investigate the role of visual speech information in Mandarin tone discrimination. Results showed that visual information improved tone perception in both Mandarin and Australian English listeners, particularly in CI simulation conditions, and particularly for tone pairs where tonal distinctions are cued by non-f0 components, suggesting that tone discrimination in CI users can be trained.
Newborn infants come equipped with a sophisticated auditory system, allowing them to perceive fin... more Newborn infants come equipped with a sophisticated auditory system, allowing them to perceive fine grained phonetic distinctions. This phonetic detail is not audible through the lowpass filter of the mother’s womb, whereas prosodic information is. Thus when born, infants have speech preferences based on prosody not on phonetic segments. The experiments here concern the role of prosodic and phonetic information in infants’ preference for their native language. English language environment infants of 6, 71⁄2, and 9 months were tested for their preference of two lists, one of Thai and one of English words. The availability of lexical tone and stress information were manipulated by using 1and 2-syllable words, and of prosodic and phonetic information by low-pass filtering. The results show that infants cannot use the tonal information on 1-syllable words to recognise their native language, and that stress and phonetic information work in tandem for native language recognition.