My research interests are Ngwi language documentation and historical-comparative analysis, tone variation and change, and factors that influence tonal system complexity.
Journal of the Southeast Asian Linguistics Society, 2024
This study investigates tone changes in Iu Mien, a Hmong-Mien language with two complex contour t... more This study investigates tone changes in Iu Mien, a Hmong-Mien language with two complex contour tones: high rising-falling (T3) and low rising-falling (T4). L-Thongkum (1988) observed younger speakers producing rising variants of T3 and T4, which she attributed to Thai contact. This study replicates L-Thongkum's study 34 years later to observe tone changes in real time and examine the potential role of peak delay, a phonetic bias resulting in a later f0 peak. 40 speakers (ages 12-84) produced T3 and T4 monosyllables in isolation and in a carrier phrase. f0 was modeled using Generalized Additive Mixed Modeling (GAMM). Results confirm incremental change toward rising variants of T3 and T4. Older speakers showed peak delay (with more of a rising contour) under conditions of shorter duration and preceding low tone. Critically, younger speakers produced peak-delayed, rising variants in all conditions, indicating generalization of the rising contour. While language contact may have initiated the changes, a phonetic bias toward peak delay generated the seeds of potential change. The change from rising-falling > rising aligns with apparent-time studies in other Asian languages, suggesting this may be a crosslinguistically unidirectional pathway.
We investigate apparent-time variation of the low falling (T21) and low level (T11) tones of the ... more We investigate apparent-time variation of the low falling (T21) and low level (T11) tones of the Black Lahu language of Yunnan, China. Linear mixed-effects modeling of spontaneous speech shows that both tones have a higher F0 trajectory among younger speakers and in certain phonetic environments. Since F0 is known to lower with increasing age, for comparison we also analyze variation in the high rising tone (T45) and find no evidence of generational difference. This suggests that the effect of age on low tones is not due to physiological change across the life span. We leave open the question of whether this result reflects a change in progress or a stable sociotonetic difference between older and younger speakers. This study contributes towards two underrepresented areas of sound change research: (1) sociotonetic approaches to tone variation in naturalistic speech styles, and (2) engagement with Indigenous scholars who are cultural insiders in small rural minority language communit...
Cite the source of the dataset as: Yang, Cathryn (2011): Lalo regional varieties: Phylogeny, dial... more Cite the source of the dataset as: Yang, Cathryn (2011): Lalo regional varieties: Phylogeny, dialectometry and sociolinguistics. Bundoora: La Trobe University.
Proc. 1st International Conference on Tone and Intonation (TAI), 2021
Studies of change across the lifespan have enabled a more nuanced understanding of segmental soun... more Studies of change across the lifespan have enabled a more nuanced understanding of segmental sound change, but this method has hardly ever been used in tone change research. This study, one of the first longitudinal studies of tone, investigates the tonal contours of King Rama IX of Thailand across a 38year period (1959-1997). During the 20 th century, Thai Tone 3 ("Falling") changed from mid-falling to high-falling. The study examines whether King Rama IX's T3 pronunciation changed during the period of investigation, and if so, whether the change was in the same direction as the community trend. The data consist of 16 recordings of the King's commencement speeches at Chulalongkorn University. F0 was extracted from 1317 T3 tokens and fitted to a cubic polynomial function. Linear mixed model results and post-hoc analysis of estimated marginal means indicate a period of change followed by stability: T3's F0 onset raised and the slope became steeper between the 1960's and the 1980's (in line with community trends), but not between the 1980's and 1990's. These findings align with previous research showing that adults have the potential to change in the post-adolescent period, but they also may resist change when that aligns with their identity.
Cite the source of the dataset as: Yang, Cathryn (2021): The phonetic tone change *high > risi... more Cite the source of the dataset as: Yang, Cathryn (2021): The phonetic tone change *high > rising: Evidence from the Ngwi dialect laboratory.
This is a comparative tone box of the Ngwi branch. Previous research has identified consonantal e... more This is a comparative tone box of the Ngwi branch. Previous research has identified consonantal effects on pitch as important precursors to the phonologization of rising tones. However, evidence from tonal evolution studies in East and Southeast Asian languages suggests that the high register (i.e., tones above mid pitch) is an additional source of rising. The current study examines the plausibility of the high > rising pathway by testing two hypotheses in a 'dialect laboratory': (1) that a rising tone in one language will correspond to a high register tone in a closely related dialect or language, and (2) that consonantal conditioning is not required for a rising tone to emerge. These hypotheses are tested on transcribed tone data from 14 Ngwi (Loloish, Tibeto-Burman) language clusters (82 language varieties in total), organized into a comparative tone box. This study examines the correspondence patterns of rising tones across a database of 82 Ngwi languages and dialects...
References: Carlson, B. (2013). The ‘new frontier’: Emergent Indigenous identities and social med... more References: Carlson, B. (2013). The ‘new frontier’: Emergent Indigenous identities and social media. In M. Harris, M. Nakata, & B. Carlson (Eds.), The Politics of Identity: Emerging Indigeneity (pp. 147–168). Sydney: University of Technology Sydney E-Press. Galla, C. K. (2016). Indigenous language revitalization, promotion, and education: function of digital technology. Computer Assisted Language Learning, 29(7), 1137–1151. https://doi.org/10.1080/09588221.2016.1166137 Jany, C. (2018). The role of new technology and social media in reversing language loss. Speech, Language and Hearing, 21(2), 73–76. https://doi.org/10.1080/2050571X.2017.1368971 Jacques, G., & Michaud, A. (2011). Approaching the historical phonology of three highly eroded Sino-Tibetan languages: Naxi, Na and Laze. Diachronica, 28(4), 468–498. Kral, I. (2014). Shifting perceptions, shifting identities: Communication technologies and the altered social, cultural and linguistic ecology in a remote indigenous context. Th...
Diachronica. International Journal for Historical Linguistics
Ground-breaking studies on how Bangkok Thai tones have changed over the past 100 years (Pittayapo... more Ground-breaking studies on how Bangkok Thai tones have changed over the past 100 years (Pittayaporn 2007, 2018; Zhu et al. 2015) reveal a pattern that Zhu et al. (2015) term the “clockwise tone shift cycle:” low > falling > high level or rising-falling > rising > falling-rising or low. The present study addresses three follow-up questions: (1) Are tone changes like those seen in Bangkok Thai also attested in other languages? (2) What other tone changes are repeated across multiple languages? (3) What phonetic biases are most likely to be the origins of the reported changes? A typological review of 52 tone change studies across 45 Sinitic, Tai-Kadai, Hmong-Mien, and Tibeto-Burman languages reveals that clockwise changes are by far the most common. The paper concludes by exploring how tonal truncation (Xu 2017) generates synchronic variation that matches the diachronic patterns; this suggests that truncation is a key mechanism in tone change.
Proceedings of the 19th International Congress of Phonetic Sciences, Melbourne, Australia 2019, 2019
Qu, Zhousuo 瞿周梭 & Cathryn Yang. 2019. Acoustic analysis of tone in Benna Hani: Tone sandhi and ne... more Qu, Zhousuo 瞿周梭 & Cathryn Yang. 2019. Acoustic analysis of tone in Benna Hani: Tone sandhi and neutralization in an atypical Tibeto-Burman language. In Sasha Calhoun, Paola Escudero, Marija Tabain & Paul Warren (eds.) Proceedings of the 19th International Congress of Phonetic Sciences, Melbourne, Australia 2019 (pp. 1967-1971). Canberra, Australia: Australasian Speech Science and Technology Association Inc.
We present the first acoustic analysis of the tone system of Benna Hani, a poorly documented and potentially endangered non-standard variety of Hani (China: Southeastern Tibeto-Burman, Ngwi). Benna Hani is unusual for a Ngwi language in that it shows complex tone sandhi and tone neutralization. Benna Hani has only three contrastive tone categories (T1, T2, and T3), but sandhi patterns show a profusion of phonetically arbitrary variants for each category. 33 Benna Hani speakers (17 female) recorded 47 monosyllables and 36 disyllables in isolation and embedded in a carrier phrase. 10,118 tokens were extracted and normalized in semitones relative to mid level pitch. We plot mean F0 trajectories with 95% confidence intervals for monosyllables and disyllables. Acoustic analysis results cohere with native speaker intuition, namely that T1 and T3 neutralize in monosyllables and in the first syllable in a disyllabic compound, but are distinguished in the second syllable of disyllables.
Co-authors: James N. Stanford, Yang Liu (刘洋), Jinjing Jiang (蒋金晶), Liufang Tang (唐留芳)
ASK ME FOR... more Co-authors: James N. Stanford, Yang Liu (刘洋), Jinjing Jiang (蒋金晶), Liufang Tang (唐留芳)
ASK ME FOR A COPY: Endangered tone languages are not often studied within quantitative variationist approaches, but such approaches can provide valuable insights for language description and documentation in the Tibeto-Burman area. This study examines tone variation within Yangliu Lalo (Central Ngwi), a minority language community in China that is currently shifting to Southwestern Mandarin. Yangliu Lalo’s Tone 4, the rising-falling High tone, is lowering and flattening among young people, especially females, who also tend to use Lalo less frequently. Tonal range in elicited speech is shown to be decreasing as use of Lalo decreases. Concurrently, the standard deviation of the pitch of individual tones also decreases, while at the same time speakers with a narrow tonal range also show greater articulatory precision for each tone. Tonal range and standard deviation of pitch are both parameters of tonal space, the arrangement of, and relationship between, tones within the tonal system. The results from our apparent-time study suggest that tonal space provides a new avenue of sociolinguistic inquiry for tone languages.
In this paper, reduction as a key mechanism in diachronic tone change is illustrated through an a... more In this paper, reduction as a key mechanism in diachronic tone change is illustrated through an apparent-time study of tone changes-in-progress in Dali Nisu, a Tibeto-Burman minority language spoken in southwest China. 26 native speakers of Dali Nisu (11 female, 15 male) participate in a tonal production experiment. Linear mixed effects modelling results show that three reduction changes are in progress: 1) high level tone lowering, 2) low rising tone flattening, and 3) low falling tone flattening. As a result of (1) and , overall tonal range is decreasing. Voicing of syllable-initial consonants interact with the reduction changes: the magnitude of the reduction effect seen among younger, more educated speakers depends on voicing of the initial consonant. This case study provides a snapshot of tone change-in-progress, showing that tonal reduction is an important mechanism shaping tone change in syllable-tone languages.
In Picus Ding and Jamin Pelkey (eds.),
Sociohistorical Linguistics in Southeast Asia: New Horizon... more In Picus Ding and Jamin Pelkey (eds.), Sociohistorical Linguistics in Southeast Asia: New Horizons for TibetoBurman Research in Honor of David Bradley, 75–95. Leiden: Brill.
腊罗话属于彝语支的西部方言群，分布在中国云南省西部，人口少于三十万。以前，大多数腊罗话的土语都没有文献记录，并且，腊罗话曾被认为只有两种土语。该论文对腊罗话作出首次精确的分类。这个分类是基于在1... more 腊罗话属于彝语支的西部方言群，分布在中国云南省西部，人口少于三十万。以前，大多数腊罗话的土语都没有文献记录，并且，腊罗话曾被认为只有两种土语。该论文对腊罗话作出首次精确的分类。这个分类是基于在18个腊罗村的方言学研究，当中包括1,001个词语的录制、通解度测试以及社会语言学的访问。综合历时亲缘分类、通解度测试以及用来文史特编辑距离测量的语音距离的结果，该论文认为，腊罗话实际上由至少七个土语组成。语音距离与通解度有很强的相关性，而基于语音距离的NeighborNet网络分析与时间深度较浅的历时亲缘分类相一致，为方言测量学的技术提供了进一步的验证。该论文表明，这些技术在鲜为人知且濒临灭绝的本土语言的分类中用处很大，这些语言通常都迫切需要我们来保护。
Asia-Pacific Language Variation and Change, May 2015
Since Labov’s early work (e.g., 1963, 1966), sociolinguists have frequently examined change in pr... more Since Labov’s early work (e.g., 1963, 1966), sociolinguists have frequently examined change in progress on the segmental level, but much less is known about tone change in progress. The present study finds evidence of a tone split in progress in Lalo, a Tibeto-Burman language of China. While many of the world’s tone languages show historical evidence of tone splits, to our knowledge this is the first time that a tone split has been observed while it is occurring, making it possible to closely examine phonological, social, and perceptual factors. In this sociotonetic study of Lalo, 2,938 tone tokens were extracted from recordings of 38 speakers and analyzed in terms of age, sex, and educational level. Multifactorial analyses show that the temporal extent of voiced stops’ depression of Tone 1 F0 is increasing in apparent time, especially among women, while VOT of voiced stops is decreasing as educational levels improve, giving speakers more contact with Mandarin Chinese. The same 38 speakers were also given a perceptual identification task in which F0 was systematically adjusted. Mixed-effects modeling showed that listeners used multiple acoustic cues (consonant voicing, F0 onset, and F0 shape) to identify the voiced initial. These findings suggest that Lalo is undergoing a tone split that follows Beddor’s (2009) coarticulatory path to sound change.
Selected Papers from the 2009 Conference of the Australian Linguistic Society, 2009
Eastern (E) and Northwestern (NW) Lalo are dialect clusters of endangered Lalo varieties spoken i... more Eastern (E) and Northwestern (NW) Lalo are dialect clusters of endangered Lalo varieties spoken in western Yunnan, China, and belonging to the Ngwi (Loloish) branch of Tibeto-Burman. This paper, the first linguistic treatment of tone in E and NW Lalo, describes the diachronic tonal developments that distinguish E and NW Lalo from
other Lalo varieties. Based on audio recordings of cognates in one E variety and four NW varieties, I present acoustic and diachronic analyses of the varieties’ tonal systems. I compare their tonal systems with Proto-Ngwi, Proto-Lalo and Central Lalo to posit their development path from Proto-Lalo. Tone change in Lalo varieties shows interaction not only between prevocalic consonants’ laryngeal features and pitch contour, but also between phonation types and pitch height. The interactions between laryngeal features and pitch add to the understanding of how laryngeal features condition secondary tone change.
Linguistics of the Tibeto-Burman Area, 35.2.113-137., 2012
Lalo is a Central Ngwi (Loloish) language cluster spoken in western Yunnan, China by fewer than 3... more Lalo is a Central Ngwi (Loloish) language cluster spoken in western Yunnan, China by fewer than 300,000 speakers. Previously, most Lalo varieties were undocumented, and Lalo was thought to have only two dialects. This paper presents the first rigorous classification of Lalo languages, based on a dialectological study conducted in eighteen Lalo villages, which included the recording of 1,001-item vocabularies, comprehension tests, and sociolinguistic interviews. Synthesising results from diachronic subgrouping, intelligibility, and phonetic distance measured by Levenshtein distance, this paper argues that the Lalo cluster actually comprises at least seven closely related languages. Phonetic distance correlates strongly with comprehension, and the NeighborNet network based on phonetic distance concurs with the diachronic subgrouping at shallow time depths, providing further validation of the dialectometric techniques. This paper suggests that such techniques are useful in the classification of lesser-known, endangered, indigenous languages, which often have an urgent need for language maintenance efforts.
International Journal of Humanities and Arts Computing, 2008
Levenshtein distance, also known as string edit distance, has been shown to correlate strongly wi... more Levenshtein distance, also known as string edit distance, has been shown to correlate strongly with both perceived distance and intelligibility in various Indo-European languages (Gooskens and Heeringa, 2004; Gooskens, 2006). We apply Levenshtein distance to dialect data from Bai (Allen, 2004), a Sino-Tibetan language, and Hongshuihe (HSH) Zhuang (Castro and Hansen, accepted), a Tai language. In applying Levenshtein distance to languages with contour tone systems, we ask the following questions: 1) How much variation in intelligibility can tone alone explain? and 2) Which representation of tone results in the Levenshtein distance that shows the strongest correlation with intelligibility test results? This research evaluates six representations of tone: onset, contour and offset; onset and contour only; contour and offset only; target approximation (Xu & Wang, 2001), autosegments of H and L, and Chao's (1930) pitch numbers. For both languages, the more fully explicit onset-contour-offset and onset-contour representations showed significantly stronger inverse correlations with intelligibility. This suggests that, for cross-dialectal listeners, the optimal representation of tone in Levenshtein distance should be at a phonetically explicit level and include information on both onset and contour.
Lawu is a severely endangered, undocumented Ngwi (Loloish) language spoken in Yunnan, China. This... more Lawu is a severely endangered, undocumented Ngwi (Loloish) language spoken in Yunnan, China. This paper presents a preliminary sketch of Lawu phonology based on lexico-phonetic data recorded from two speakers in 2008, with special attention to the tone splits and mergers that distinguish Lawu from other Ngwi languages. All tone categories except Proto-Ngwi Tone *3, a mid level pitch, have split, conditioned by the voicing of the initial segment. In the conditioning and effect of these tone splits, Lawu shows affinity with other Central Ngwi languages such as Lisu and Lahu and is provisionally classified as a Central Ngwi language.
Journal of the Southeast Asian Linguistics Society, 2024
This study investigates tone changes in Iu Mien, a Hmong-Mien language with two complex contour t... more This study investigates tone changes in Iu Mien, a Hmong-Mien language with two complex contour tones: high rising-falling (T3) and low rising-falling (T4). L-Thongkum (1988) observed younger speakers producing rising variants of T3 and T4, which she attributed to Thai contact. This study replicates L-Thongkum's study 34 years later to observe tone changes in real time and examine the potential role of peak delay, a phonetic bias resulting in a later f0 peak. 40 speakers (ages 12-84) produced T3 and T4 monosyllables in isolation and in a carrier phrase. f0 was modeled using Generalized Additive Mixed Modeling (GAMM). Results confirm incremental change toward rising variants of T3 and T4. Older speakers showed peak delay (with more of a rising contour) under conditions of shorter duration and preceding low tone. Critically, younger speakers produced peak-delayed, rising variants in all conditions, indicating generalization of the rising contour. While language contact may have initiated the changes, a phonetic bias toward peak delay generated the seeds of potential change. The change from rising-falling > rising aligns with apparent-time studies in other Asian languages, suggesting this may be a crosslinguistically unidirectional pathway.
We investigate apparent-time variation of the low falling (T21) and low level (T11) tones of the ... more We investigate apparent-time variation of the low falling (T21) and low level (T11) tones of the Black Lahu language of Yunnan, China. Linear mixed-effects modeling of spontaneous speech shows that both tones have a higher F0 trajectory among younger speakers and in certain phonetic environments. Since F0 is known to lower with increasing age, for comparison we also analyze variation in the high rising tone (T45) and find no evidence of generational difference. This suggests that the effect of age on low tones is not due to physiological change across the life span. We leave open the question of whether this result reflects a change in progress or a stable sociotonetic difference between older and younger speakers. This study contributes towards two underrepresented areas of sound change research: (1) sociotonetic approaches to tone variation in naturalistic speech styles, and (2) engagement with Indigenous scholars who are cultural insiders in small rural minority language communit...
Cite the source of the dataset as: Yang, Cathryn (2011): Lalo regional varieties: Phylogeny, dial... more Cite the source of the dataset as: Yang, Cathryn (2011): Lalo regional varieties: Phylogeny, dialectometry and sociolinguistics. Bundoora: La Trobe University.
Proc. 1st International Conference on Tone and Intonation (TAI), 2021
Studies of change across the lifespan have enabled a more nuanced understanding of segmental soun... more Studies of change across the lifespan have enabled a more nuanced understanding of segmental sound change, but this method has hardly ever been used in tone change research. This study, one of the first longitudinal studies of tone, investigates the tonal contours of King Rama IX of Thailand across a 38year period (1959-1997). During the 20 th century, Thai Tone 3 ("Falling") changed from mid-falling to high-falling. The study examines whether King Rama IX's T3 pronunciation changed during the period of investigation, and if so, whether the change was in the same direction as the community trend. The data consist of 16 recordings of the King's commencement speeches at Chulalongkorn University. F0 was extracted from 1317 T3 tokens and fitted to a cubic polynomial function. Linear mixed model results and post-hoc analysis of estimated marginal means indicate a period of change followed by stability: T3's F0 onset raised and the slope became steeper between the 1960's and the 1980's (in line with community trends), but not between the 1980's and 1990's. These findings align with previous research showing that adults have the potential to change in the post-adolescent period, but they also may resist change when that aligns with their identity.
Cite the source of the dataset as: Yang, Cathryn (2021): The phonetic tone change *high > risi... more Cite the source of the dataset as: Yang, Cathryn (2021): The phonetic tone change *high > rising: Evidence from the Ngwi dialect laboratory.
This is a comparative tone box of the Ngwi branch. Previous research has identified consonantal e... more This is a comparative tone box of the Ngwi branch. Previous research has identified consonantal effects on pitch as important precursors to the phonologization of rising tones. However, evidence from tonal evolution studies in East and Southeast Asian languages suggests that the high register (i.e., tones above mid pitch) is an additional source of rising. The current study examines the plausibility of the high > rising pathway by testing two hypotheses in a 'dialect laboratory': (1) that a rising tone in one language will correspond to a high register tone in a closely related dialect or language, and (2) that consonantal conditioning is not required for a rising tone to emerge. These hypotheses are tested on transcribed tone data from 14 Ngwi (Loloish, Tibeto-Burman) language clusters (82 language varieties in total), organized into a comparative tone box. This study examines the correspondence patterns of rising tones across a database of 82 Ngwi languages and dialects...
References: Carlson, B. (2013). The ‘new frontier’: Emergent Indigenous identities and social med... more References: Carlson, B. (2013). The ‘new frontier’: Emergent Indigenous identities and social media. In M. Harris, M. Nakata, & B. Carlson (Eds.), The Politics of Identity: Emerging Indigeneity (pp. 147–168). Sydney: University of Technology Sydney E-Press. Galla, C. K. (2016). Indigenous language revitalization, promotion, and education: function of digital technology. Computer Assisted Language Learning, 29(7), 1137–1151. https://doi.org/10.1080/09588221.2016.1166137 Jany, C. (2018). The role of new technology and social media in reversing language loss. Speech, Language and Hearing, 21(2), 73–76. https://doi.org/10.1080/2050571X.2017.1368971 Jacques, G., & Michaud, A. (2011). Approaching the historical phonology of three highly eroded Sino-Tibetan languages: Naxi, Na and Laze. Diachronica, 28(4), 468–498. Kral, I. (2014). Shifting perceptions, shifting identities: Communication technologies and the altered social, cultural and linguistic ecology in a remote indigenous context. Th...
Diachronica. International Journal for Historical Linguistics
Ground-breaking studies on how Bangkok Thai tones have changed over the past 100 years (Pittayapo... more Ground-breaking studies on how Bangkok Thai tones have changed over the past 100 years (Pittayaporn 2007, 2018; Zhu et al. 2015) reveal a pattern that Zhu et al. (2015) term the “clockwise tone shift cycle:” low > falling > high level or rising-falling > rising > falling-rising or low. The present study addresses three follow-up questions: (1) Are tone changes like those seen in Bangkok Thai also attested in other languages? (2) What other tone changes are repeated across multiple languages? (3) What phonetic biases are most likely to be the origins of the reported changes? A typological review of 52 tone change studies across 45 Sinitic, Tai-Kadai, Hmong-Mien, and Tibeto-Burman languages reveals that clockwise changes are by far the most common. The paper concludes by exploring how tonal truncation (Xu 2017) generates synchronic variation that matches the diachronic patterns; this suggests that truncation is a key mechanism in tone change.
Proceedings of the 19th International Congress of Phonetic Sciences, Melbourne, Australia 2019, 2019
Qu, Zhousuo 瞿周梭 & Cathryn Yang. 2019. Acoustic analysis of tone in Benna Hani: Tone sandhi and ne... more Qu, Zhousuo 瞿周梭 & Cathryn Yang. 2019. Acoustic analysis of tone in Benna Hani: Tone sandhi and neutralization in an atypical Tibeto-Burman language. In Sasha Calhoun, Paola Escudero, Marija Tabain & Paul Warren (eds.) Proceedings of the 19th International Congress of Phonetic Sciences, Melbourne, Australia 2019 (pp. 1967-1971). Canberra, Australia: Australasian Speech Science and Technology Association Inc.
We present the first acoustic analysis of the tone system of Benna Hani, a poorly documented and potentially endangered non-standard variety of Hani (China: Southeastern Tibeto-Burman, Ngwi). Benna Hani is unusual for a Ngwi language in that it shows complex tone sandhi and tone neutralization. Benna Hani has only three contrastive tone categories (T1, T2, and T3), but sandhi patterns show a profusion of phonetically arbitrary variants for each category. 33 Benna Hani speakers (17 female) recorded 47 monosyllables and 36 disyllables in isolation and embedded in a carrier phrase. 10,118 tokens were extracted and normalized in semitones relative to mid level pitch. We plot mean F0 trajectories with 95% confidence intervals for monosyllables and disyllables. Acoustic analysis results cohere with native speaker intuition, namely that T1 and T3 neutralize in monosyllables and in the first syllable in a disyllabic compound, but are distinguished in the second syllable of disyllables.
Co-authors: James N. Stanford, Yang Liu (刘洋), Jinjing Jiang (蒋金晶), Liufang Tang (唐留芳)
ASK ME FOR... more Co-authors: James N. Stanford, Yang Liu (刘洋), Jinjing Jiang (蒋金晶), Liufang Tang (唐留芳)
ASK ME FOR A COPY: Endangered tone languages are not often studied within quantitative variationist approaches, but such approaches can provide valuable insights for language description and documentation in the Tibeto-Burman area. This study examines tone variation within Yangliu Lalo (Central Ngwi), a minority language community in China that is currently shifting to Southwestern Mandarin. Yangliu Lalo’s Tone 4, the rising-falling High tone, is lowering and flattening among young people, especially females, who also tend to use Lalo less frequently. Tonal range in elicited speech is shown to be decreasing as use of Lalo decreases. Concurrently, the standard deviation of the pitch of individual tones also decreases, while at the same time speakers with a narrow tonal range also show greater articulatory precision for each tone. Tonal range and standard deviation of pitch are both parameters of tonal space, the arrangement of, and relationship between, tones within the tonal system. The results from our apparent-time study suggest that tonal space provides a new avenue of sociolinguistic inquiry for tone languages.
In this paper, reduction as a key mechanism in diachronic tone change is illustrated through an a... more In this paper, reduction as a key mechanism in diachronic tone change is illustrated through an apparent-time study of tone changes-in-progress in Dali Nisu, a Tibeto-Burman minority language spoken in southwest China. 26 native speakers of Dali Nisu (11 female, 15 male) participate in a tonal production experiment. Linear mixed effects modelling results show that three reduction changes are in progress: 1) high level tone lowering, 2) low rising tone flattening, and 3) low falling tone flattening. As a result of (1) and , overall tonal range is decreasing. Voicing of syllable-initial consonants interact with the reduction changes: the magnitude of the reduction effect seen among younger, more educated speakers depends on voicing of the initial consonant. This case study provides a snapshot of tone change-in-progress, showing that tonal reduction is an important mechanism shaping tone change in syllable-tone languages.
In Picus Ding and Jamin Pelkey (eds.),
Sociohistorical Linguistics in Southeast Asia: New Horizon... more In Picus Ding and Jamin Pelkey (eds.), Sociohistorical Linguistics in Southeast Asia: New Horizons for TibetoBurman Research in Honor of David Bradley, 75–95. Leiden: Brill.
腊罗话属于彝语支的西部方言群，分布在中国云南省西部，人口少于三十万。以前，大多数腊罗话的土语都没有文献记录，并且，腊罗话曾被认为只有两种土语。该论文对腊罗话作出首次精确的分类。这个分类是基于在1... more 腊罗话属于彝语支的西部方言群，分布在中国云南省西部，人口少于三十万。以前，大多数腊罗话的土语都没有文献记录，并且，腊罗话曾被认为只有两种土语。该论文对腊罗话作出首次精确的分类。这个分类是基于在18个腊罗村的方言学研究，当中包括1,001个词语的录制、通解度测试以及社会语言学的访问。综合历时亲缘分类、通解度测试以及用来文史特编辑距离测量的语音距离的结果，该论文认为，腊罗话实际上由至少七个土语组成。语音距离与通解度有很强的相关性，而基于语音距离的NeighborNet网络分析与时间深度较浅的历时亲缘分类相一致，为方言测量学的技术提供了进一步的验证。该论文表明，这些技术在鲜为人知且濒临灭绝的本土语言的分类中用处很大，这些语言通常都迫切需要我们来保护。
Asia-Pacific Language Variation and Change, May 2015
Since Labov’s early work (e.g., 1963, 1966), sociolinguists have frequently examined change in pr... more Since Labov’s early work (e.g., 1963, 1966), sociolinguists have frequently examined change in progress on the segmental level, but much less is known about tone change in progress. The present study finds evidence of a tone split in progress in Lalo, a Tibeto-Burman language of China. While many of the world’s tone languages show historical evidence of tone splits, to our knowledge this is the first time that a tone split has been observed while it is occurring, making it possible to closely examine phonological, social, and perceptual factors. In this sociotonetic study of Lalo, 2,938 tone tokens were extracted from recordings of 38 speakers and analyzed in terms of age, sex, and educational level. Multifactorial analyses show that the temporal extent of voiced stops’ depression of Tone 1 F0 is increasing in apparent time, especially among women, while VOT of voiced stops is decreasing as educational levels improve, giving speakers more contact with Mandarin Chinese. The same 38 speakers were also given a perceptual identification task in which F0 was systematically adjusted. Mixed-effects modeling showed that listeners used multiple acoustic cues (consonant voicing, F0 onset, and F0 shape) to identify the voiced initial. These findings suggest that Lalo is undergoing a tone split that follows Beddor’s (2009) coarticulatory path to sound change.
Selected Papers from the 2009 Conference of the Australian Linguistic Society, 2009
Eastern (E) and Northwestern (NW) Lalo are dialect clusters of endangered Lalo varieties spoken i... more Eastern (E) and Northwestern (NW) Lalo are dialect clusters of endangered Lalo varieties spoken in western Yunnan, China, and belonging to the Ngwi (Loloish) branch of Tibeto-Burman. This paper, the first linguistic treatment of tone in E and NW Lalo, describes the diachronic tonal developments that distinguish E and NW Lalo from
other Lalo varieties. Based on audio recordings of cognates in one E variety and four NW varieties, I present acoustic and diachronic analyses of the varieties’ tonal systems. I compare their tonal systems with Proto-Ngwi, Proto-Lalo and Central Lalo to posit their development path from Proto-Lalo. Tone change in Lalo varieties shows interaction not only between prevocalic consonants’ laryngeal features and pitch contour, but also between phonation types and pitch height. The interactions between laryngeal features and pitch add to the understanding of how laryngeal features condition secondary tone change.
Linguistics of the Tibeto-Burman Area, 35.2.113-137., 2012
Lalo is a Central Ngwi (Loloish) language cluster spoken in western Yunnan, China by fewer than 3... more Lalo is a Central Ngwi (Loloish) language cluster spoken in western Yunnan, China by fewer than 300,000 speakers. Previously, most Lalo varieties were undocumented, and Lalo was thought to have only two dialects. This paper presents the first rigorous classification of Lalo languages, based on a dialectological study conducted in eighteen Lalo villages, which included the recording of 1,001-item vocabularies, comprehension tests, and sociolinguistic interviews. Synthesising results from diachronic subgrouping, intelligibility, and phonetic distance measured by Levenshtein distance, this paper argues that the Lalo cluster actually comprises at least seven closely related languages. Phonetic distance correlates strongly with comprehension, and the NeighborNet network based on phonetic distance concurs with the diachronic subgrouping at shallow time depths, providing further validation of the dialectometric techniques. This paper suggests that such techniques are useful in the classification of lesser-known, endangered, indigenous languages, which often have an urgent need for language maintenance efforts.
International Journal of Humanities and Arts Computing, 2008
Levenshtein distance, also known as string edit distance, has been shown to correlate strongly wi... more Levenshtein distance, also known as string edit distance, has been shown to correlate strongly with both perceived distance and intelligibility in various Indo-European languages (Gooskens and Heeringa, 2004; Gooskens, 2006). We apply Levenshtein distance to dialect data from Bai (Allen, 2004), a Sino-Tibetan language, and Hongshuihe (HSH) Zhuang (Castro and Hansen, accepted), a Tai language. In applying Levenshtein distance to languages with contour tone systems, we ask the following questions: 1) How much variation in intelligibility can tone alone explain? and 2) Which representation of tone results in the Levenshtein distance that shows the strongest correlation with intelligibility test results? This research evaluates six representations of tone: onset, contour and offset; onset and contour only; contour and offset only; target approximation (Xu & Wang, 2001), autosegments of H and L, and Chao's (1930) pitch numbers. For both languages, the more fully explicit onset-contour-offset and onset-contour representations showed significantly stronger inverse correlations with intelligibility. This suggests that, for cross-dialectal listeners, the optimal representation of tone in Levenshtein distance should be at a phonetically explicit level and include information on both onset and contour.
Lawu is a severely endangered, undocumented Ngwi (Loloish) language spoken in Yunnan, China. This... more Lawu is a severely endangered, undocumented Ngwi (Loloish) language spoken in Yunnan, China. This paper presents a preliminary sketch of Lawu phonology based on lexico-phonetic data recorded from two speakers in 2008, with special attention to the tone splits and mergers that distinguish Lawu from other Ngwi languages. All tone categories except Proto-Ngwi Tone *3, a mid level pitch, have split, conditioned by the voicing of the initial segment. In the conditioning and effect of these tone splits, Lawu shows affinity with other Central Ngwi languages such as Lisu and Lahu and is provisionally classified as a Central Ngwi language.
This volume is the first work to systematically investigate the diachronic and synchronic relatio... more This volume is the first work to systematically investigate the diachronic and synchronic relationships between regional varieties of Lalo, a Ngwi (Loloish) language cluster spoken in western Yunnan, China. The empirical basis for the research is linguistic data, as well as intelligibility tests and sociolinguistic interviews on contact, dialect perceptions, and ethnolinguistic vitality from nineteen Lalo-speaking villages. The volume uses these data to present a phonological and lexical reconstruction of Proto-Lalo, as well as a phylogenetic subgrouping of the different Lalo varieties. As a complement to this, a synchronic classification of Lalo varieties according to phonetic distance, intelligibility, and speaker perceptions is also given. This combination of methodologies enable an integrated synchronic and diachronic depiction of Lalo dialect diversity.
Previous research has proposed a direct path from consonantal effects on F0 to the development of... more Previous research has proposed a direct path from consonantal effects on F0 to the development of a rising tone value. However, findings from tone change studies in Asian languages suggest an additional pathway: a high tone (i.e., with a tonal target in the upper pitch range) may evolve into a rising tone, without consonantal conditioning. This study examines tone change pathways to rising in the Ngwi (Loloish) languages. Among the 11 Ngwi language clusters examined, a new rising tone value has unambiguously developed from an historically high tone in 9 of them. In several clusters, prevocalic consonants conditioned a tone split, but in other clusters, consonantal conditioning does not appear to have played a role. The findings suggest that a high tonal target is itself a condition favorable to the development of a rising tone value, as reaching a high tonal target in connected speech frequently entails a rising contour.
In this paper, reduction as a key mechanism in diachronic tone change is illustrated through an a... more In this paper, reduction as a key mechanism in diachronic tone change is illustrated through an apparent-time study of tone changes-in-progress in Dali Nisu, a Tibeto-Burman minority language spoken in southwest China. 26 native speakers of Dali Nisu (11 female, 15 male) participate in a tonal production experiment. Linear mixed effects modelling results show that three reduction changes are in progress: 1) high level tone lowering, 2) low rising tone flattening, and 3) low falling tone flattening. As a result of (1) and (3), overall tonal range is decreasing. Voicing of syllable-initial consonants interact with the reduction changes: the magnitude of the reduction effect seen among younger, more educated speakers depends on voicing of the initial consonant. This case study provides a snapshot of tone change-in-progress, showing that tonal reduction is an important mechanism shaping tone change in syllable-tone languages.
Abstract submitted to 9th International Conference on Evolutionary Linguistics, Kunming August 25... more Abstract submitted to 9th International Conference on Evolutionary Linguistics, Kunming August 25-27, 2017
Transcriptions based on recordings from two anonymous Limi speakers (one male and one female), re... more Transcriptions based on recordings from two anonymous Limi speakers (one male and one female), recorded in 2008 in Wumulong Town, Yongde County, Lincang Municipality, Yunnan Province, China.
Northwestern Yunnan is home to many small, endangered Central Ngwi groups, several of which have ... more Northwestern Yunnan is home to many small, endangered Central Ngwi groups, several of which have never before been documented: Taliu, Tagu, Tazhi, Lang’e, Naruo, Ghomo, Popei, Kuansi, Kuamasi, Laizisi, Zibusi, and Sonaga. This historical dialectology study is the first to examine the historic and connections between these groups and their Central Ngwi neighbors. Synthesizing evidence from linguistics, geography, history, and anthropology, we argue that 1) these groups form a unique cluster, the Taloid group, and 2) ancestors of the Taloid cluster included soldiers sent by the Nanzhao Kingdom (8th-9th centuries) to guard against invasion from the north.
Word lists from published sources (Chen 2010, Castro et al 2010, Duan 1998, Zhou 2004) and from recent fieldwork in Yongsheng and Eryuan counties are compared with Proto-Ngwi (Bradley 1979) to track each variety’s phonological development. The proposed cluster shows a set of shared phonological and lexical innovations; these innovations are then compared to innovations found in the Laloid cluster (Yang 2015) and in the Loloid (Lolo-Lipo-Limi) cluster (Yang 2017). Many of the innovations are shared with Laloid; others are unique to Taloid; only a few are shared with Loloid. This pattern suggests that the Taloid languages share a close genealogical relationship with each other, while further back in history they share a common ancestry with the Laloid cluster, a group closely associated with the Nanzhao Kingdom. Shared customs with the Lalo, not seen among the neighboring Bai ethnicity, also support this conclusion. Within the Taloid cluster, the varieties in Dali Prefecture show several changes induced through contact with the Bai language, while those in Yongsheng County are more likely to show influence from Chinese rather than Bai.
The relationships between Taloid languages and other Central Ngwi languages are further explored using the web-based Gabmap program (Nerbonne et al 2011) to conduct cluster analysis and multi-dimensional scaling based on phonetic distance. Findings from these quantitative methods provide further support for subgrouping the Taloid languages together, while also revealing a surprising amount of diversity across Taloid’s comparatively narrow geographical distribution.
The Taloid groups are located along an east-west axis that cuts across northern Dali and southern Lijiang through to southern Sichuan. The distribution pattern is suggestive of a military front facing north, an interpretation coherent with evidence from oral and written history, suggesting that the Taloid ancestral group were sent as soldiers to guard against threats from the Tibetan and Tang kingdoms. Over a thousand years in the same basic region, without much later expansion to new areas, helps explain the diversity seen among the Taloid’s relatively small populations. By combining evidence from various fields, the history of these under-documented, endangered language varieties can finally begin to be understood.
滇西北彝语他留土语群：忠诚的南诏战士 滇西北地区分布着彝语西部方言许多小的濒危土语，包括他留、它谷、它之、拉务、纳若（水田）、厄毛柔、颇佩、夸恩斯、夸玛斯、莱兹斯、子逋斯以及锁内嘎等。其中的几种土语从来没有被深入研究过。本文首次从历史方言学的角度对这些土语之间以及这些土语与相邻的彝语方言之间的历史和语言关系进行研究。基于语言学、地理学、历史学和人类学的研究证据，本文提出以下两个观点：（1）这些土语组成了一个独特的土语群——他留土语群；（2）他留土语群应该是南诏国派出抵御北部入侵的士兵的后代。 通过将已出版的以及最近在永胜县、洱源县田野调查的词汇表与原始彝语构拟进行比较，可以追溯每一种土语的语音历史发展脉络。研究表明，上述土语群显示出一系列语音和词汇方面的共同创新。这些共同创新有的是他留土语群独有的，有的是与腊罗土语群共同拥有的，而与倮倮土语群共同拥有的创新却不多。这意味着他留土语群各土语之间存在着紧密的系谱关系，而且从历史上看，他留土语群和与南诏国蒙舍诏关系密切的腊罗土语群的祖先是同源的。他留土语群和腊罗土语群拥有一些共同的文化习俗，而这些习俗在白族文化中没有出现，这一人类学的证据也支持了以上结论。 运用网络软件Gabmap (Nerbonne et al 2011)测量语音距离并进行聚类分析和多维标度， 更进一步证实了土语之间的关系。以上定量方法的结果支持把这些土语划分为他留土语群，并且发现它们虽然分布在窄小的地理范围内，但内部之间的差异性却比较大。 他留土语群横跨大理北部、丽江南部直至四川南部，沿一条东西轴分布。该分布模式可能反映南诏国时期面对北方战争前线的形势。这一解释与来自民间相传的和文字记载的历史是一致的，即他留人的祖先曾被派驻前线以抵御吐蕃和唐朝的威胁。超过一千年的时间一直生活在一个比较窄的地理范围，几乎没有迁徙，这一历史有利于说明该土语群人口较少却有如此大的多样性的原因。综合来自不同领域的证据，这些未被记载的濒危土语得以被初步了解。
Presentation at NWAV-AP 4 in Chiayi, Taiwan.
This variationist fieldwork-based study examines t... more Presentation at NWAV-AP 4 in Chiayi, Taiwan.
This variationist fieldwork-based study examines the connection between social factors and tone change from convex contour to flat. We conducted acoustic sociophonetic analyses on Tone 4 of Yangliu Lalo, an endangered Tibeto-Burman language whose community is currently shifting to Southwestern Mandarin. Hypothesis: As speakers’ use of Lalo decreases, the likelihood of a flat realization of Tone 4 and reduced F0 span increases, while speakers who use Lalo more have a dynamic Tone 4 contour and a wider F0 span. Methods: We recorded wordlists and phrases with 34 Lalo speakers in Yangliu Village (both genders, ages 20-72, 71 words/speaker). We extracted 8,856 pitch tokens, all of which were stressed monosyllables between mid-tones in connected speech or as citation forms, normalized in semitones relative to the mid-tone, and normalized for duration as 200 relative-time units per syllable. The same speakers were interviewed on their language use and attitudes. Results: Using Rbrul (Johnson 2009), we ran linear mixed-effects models on Age, Gender, Education, Lalo Use (a numerical index combining 1st language learned, reported frequency of Lalo use, and current best language), Language used in Family/Affinal Domains, Attitude Toward Lalo, and Voicing of Onsets, with Speaker and Word as random effects. The statistical model finds (R2=0.239) that the pitch peak occurs later in syllables with a voiced onset (p<0.0001), spoken in carrier phrase form (p<0.0001), for younger speakers (p=0.001), and for for those with stronger Lalo use (p = 0.003). The other factors and interactions were not significant. We also ran multiple regression on F0 span, defined as the MeanT4 (high) minus Mean T2 (low), between Relative Time 50 and 150. Using input fixed factors of Age, Gender, Lalo Use, Domain, Attitude, Education, Child Language, Frequency, Importance, the step-up/step-down regression model selected one significant factor: Lalo Use (R²=0.31, p=0.000391; est. = 0.18 semitones for every 1 point on the 10-point Lalo Use index). Again, no significant interactions between fixed factors were found. Analysis: The voicing effects reflect the classic tone perturbation of onsets (Hombert 1978), but the effects of language use are novel. While the flat variant minimizes articulatory effort and reduces redundant perceptual cues (a simpler level tone system), we find that usage effects play a role in this change and in the reduction of F0 span. It is not likely that contact with Baoshan or Standard Mandarin would induce a reduction of F0 span, since both have more complex tonal systems and wider F0 spans than Yangliu Lalo. Instead, decreasing F0 span represents a new balance between reduction of articulatory effort and maximization of perceptual distinctiveness (Flemming 2004).
References Hombert, Jean-Marie. 1978. Consonant Types, Vowel Quality, and Tone. In V. A. Fromkin (ed) Tone: A Linguistic Survey. New York: Academic Press. 77-111. Johnson, Daniel Ezra. 2009. Getting Off the GoldVarb Standard: Rbrul for Mixed Effects Variable Rule Analysis. Language and Linguistics Compass 3(1):359-83. Flemming, E. (2004). Contrast and perceptual distinctiveness. In B. Hayes, R. Kirchner, & D. Steriade , Phonetically based phonology (pp. 232-276). Cambridge: Cambridge University Press.
Uploads
Papers by Cathryn Yang
Canberra, Australia: Australasian Speech Science and Technology Association Inc.
We present the first acoustic analysis of the tone system of Benna Hani, a poorly documented and potentially endangered non-standard variety of Hani (China: Southeastern Tibeto-Burman, Ngwi). Benna Hani is unusual for a Ngwi language in that it shows complex tone sandhi and tone neutralization. Benna Hani has only three contrastive tone categories (T1, T2, and T3), but sandhi patterns show a profusion of phonetically arbitrary variants for each category. 33 Benna Hani speakers (17 female) recorded 47 monosyllables and 36 disyllables in isolation and embedded in a carrier phrase. 10,118 tokens were extracted and normalized in semitones relative to mid level pitch. We plot mean F0 trajectories with 95% confidence intervals for monosyllables and disyllables. Acoustic analysis results cohere with native speaker intuition, namely that T1 and T3 neutralize in monosyllables and in the first syllable in a disyllabic compound, but are distinguished in the second syllable of disyllables.
ASK ME FOR A COPY: Endangered tone languages are not often studied within quantitative variationist approaches, but such approaches can provide valuable insights for language description and documentation in the Tibeto-Burman area. This study examines tone variation within Yangliu Lalo (Central Ngwi), a minority language community in China that is currently shifting to Southwestern Mandarin. Yangliu Lalo’s Tone 4, the rising-falling High tone, is lowering and flattening among young people, especially females, who also tend to use Lalo less frequently. Tonal range in elicited speech is shown to be decreasing as use of Lalo decreases. Concurrently, the standard deviation of the pitch of individual tones also decreases, while at the same time speakers with a narrow tonal range also show greater articulatory precision for each tone. Tonal range and standard deviation of pitch are both parameters of tonal space, the arrangement of, and relationship between, tones within the tonal system. The results from our apparent-time study suggest that tonal space provides a new avenue of sociolinguistic inquiry for tone languages.
Sociohistorical Linguistics in Southeast Asia: New Horizons for TibetoBurman Research in Honor of David Bradley, 75–95. Leiden: Brill.
other Lalo varieties. Based on audio recordings of cognates in one E variety and four NW varieties, I present acoustic and diachronic analyses of the varieties’ tonal systems. I compare their tonal systems with Proto-Ngwi, Proto-Lalo and Central Lalo to posit their development path from Proto-Lalo. Tone change in Lalo varieties shows interaction not only between prevocalic consonants’ laryngeal features and pitch contour, but also between phonation types and pitch height. The interactions between laryngeal features and pitch add to the understanding of how laryngeal features condition secondary tone change.
Canberra, Australia: Australasian Speech Science and Technology Association Inc.
We present the first acoustic analysis of the tone system of Benna Hani, a poorly documented and potentially endangered non-standard variety of Hani (China: Southeastern Tibeto-Burman, Ngwi). Benna Hani is unusual for a Ngwi language in that it shows complex tone sandhi and tone neutralization. Benna Hani has only three contrastive tone categories (T1, T2, and T3), but sandhi patterns show a profusion of phonetically arbitrary variants for each category. 33 Benna Hani speakers (17 female) recorded 47 monosyllables and 36 disyllables in isolation and embedded in a carrier phrase. 10,118 tokens were extracted and normalized in semitones relative to mid level pitch. We plot mean F0 trajectories with 95% confidence intervals for monosyllables and disyllables. Acoustic analysis results cohere with native speaker intuition, namely that T1 and T3 neutralize in monosyllables and in the first syllable in a disyllabic compound, but are distinguished in the second syllable of disyllables.
ASK ME FOR A COPY: Endangered tone languages are not often studied within quantitative variationist approaches, but such approaches can provide valuable insights for language description and documentation in the Tibeto-Burman area. This study examines tone variation within Yangliu Lalo (Central Ngwi), a minority language community in China that is currently shifting to Southwestern Mandarin. Yangliu Lalo’s Tone 4, the rising-falling High tone, is lowering and flattening among young people, especially females, who also tend to use Lalo less frequently. Tonal range in elicited speech is shown to be decreasing as use of Lalo decreases. Concurrently, the standard deviation of the pitch of individual tones also decreases, while at the same time speakers with a narrow tonal range also show greater articulatory precision for each tone. Tonal range and standard deviation of pitch are both parameters of tonal space, the arrangement of, and relationship between, tones within the tonal system. The results from our apparent-time study suggest that tonal space provides a new avenue of sociolinguistic inquiry for tone languages.
Sociohistorical Linguistics in Southeast Asia: New Horizons for TibetoBurman Research in Honor of David Bradley, 75–95. Leiden: Brill.
other Lalo varieties. Based on audio recordings of cognates in one E variety and four NW varieties, I present acoustic and diachronic analyses of the varieties’ tonal systems. I compare their tonal systems with Proto-Ngwi, Proto-Lalo and Central Lalo to posit their development path from Proto-Lalo. Tone change in Lalo varieties shows interaction not only between prevocalic consonants’ laryngeal features and pitch contour, but also between phonation types and pitch height. The interactions between laryngeal features and pitch add to the understanding of how laryngeal features condition secondary tone change.
western Yunnan, China. The empirical basis for the research is linguistic data, as well as intelligibility tests and sociolinguistic interviews on contact, dialect perceptions, and
ethnolinguistic vitality from nineteen Lalo-speaking villages. The volume uses these data to present a phonological and lexical reconstruction of Proto-Lalo, as well as a phylogenetic
subgrouping of the different Lalo varieties. As a complement to this, a synchronic classification of Lalo varieties according to phonetic distance, intelligibility, and speaker
perceptions is also given. This combination of methodologies enable an integrated synchronic and diachronic depiction of Lalo dialect diversity.
Word lists from published sources (Chen 2010, Castro et al 2010, Duan 1998, Zhou 2004) and from recent fieldwork in Yongsheng and Eryuan counties are compared with Proto-Ngwi (Bradley 1979) to track each variety’s phonological development. The proposed cluster shows a set of shared phonological and lexical innovations; these innovations are then compared to innovations found in the Laloid cluster (Yang 2015) and in the Loloid (Lolo-Lipo-Limi) cluster (Yang 2017). Many of the innovations are shared with Laloid; others are unique to Taloid; only a few are shared with Loloid. This pattern suggests that the Taloid languages share a close genealogical relationship with each other, while further back in history they share a common ancestry with the Laloid cluster, a group closely associated with the Nanzhao Kingdom. Shared customs with the Lalo, not seen among the neighboring Bai ethnicity, also support this conclusion. Within the Taloid cluster, the varieties in Dali Prefecture show several changes induced through contact with the Bai language, while those in Yongsheng County are more likely to show influence from Chinese rather than Bai.
The relationships between Taloid languages and other Central Ngwi languages are further explored using the web-based Gabmap program (Nerbonne et al 2011) to conduct cluster analysis and multi-dimensional scaling based on phonetic distance. Findings from these quantitative methods provide further support for subgrouping the Taloid languages together, while also revealing a surprising amount of diversity across Taloid’s comparatively narrow geographical distribution.
The Taloid groups are located along an east-west axis that cuts across northern Dali and southern Lijiang through to southern Sichuan. The distribution pattern is suggestive of a military front facing north, an interpretation coherent with evidence from oral and written history, suggesting that the Taloid ancestral group were sent as soldiers to guard against threats from the Tibetan and Tang kingdoms. Over a thousand years in the same basic region, without much later expansion to new areas, helps explain the diversity seen among the Taloid’s relatively small populations. By combining evidence from various fields, the history of these under-documented, endangered language varieties can finally begin to be understood.
滇西北彝语他留土语群：忠诚的南诏战士
滇西北地区分布着彝语西部方言许多小的濒危土语，包括他留、它谷、它之、拉务、纳若（水田）、厄毛柔、颇佩、夸恩斯、夸玛斯、莱兹斯、子逋斯以及锁内嘎等。其中的几种土语从来没有被深入研究过。本文首次从历史方言学的角度对这些土语之间以及这些土语与相邻的彝语方言之间的历史和语言关系进行研究。基于语言学、地理学、历史学和人类学的研究证据，本文提出以下两个观点：（1）这些土语组成了一个独特的土语群——他留土语群；（2）他留土语群应该是南诏国派出抵御北部入侵的士兵的后代。
通过将已出版的以及最近在永胜县、洱源县田野调查的词汇表与原始彝语构拟进行比较，可以追溯每一种土语的语音历史发展脉络。研究表明，上述土语群显示出一系列语音和词汇方面的共同创新。这些共同创新有的是他留土语群独有的，有的是与腊罗土语群共同拥有的，而与倮倮土语群共同拥有的创新却不多。这意味着他留土语群各土语之间存在着紧密的系谱关系，而且从历史上看，他留土语群和与南诏国蒙舍诏关系密切的腊罗土语群的祖先是同源的。他留土语群和腊罗土语群拥有一些共同的文化习俗，而这些习俗在白族文化中没有出现，这一人类学的证据也支持了以上结论。
运用网络软件Gabmap (Nerbonne et al 2011)测量语音距离并进行聚类分析和多维标度， 更进一步证实了土语之间的关系。以上定量方法的结果支持把这些土语划分为他留土语群，并且发现它们虽然分布在窄小的地理范围内，但内部之间的差异性却比较大。
他留土语群横跨大理北部、丽江南部直至四川南部，沿一条东西轴分布。该分布模式可能反映南诏国时期面对北方战争前线的形势。这一解释与来自民间相传的和文字记载的历史是一致的，即他留人的祖先曾被派驻前线以抵御吐蕃和唐朝的威胁。超过一千年的时间一直生活在一个比较窄的地理范围，几乎没有迁徙，这一历史有利于说明该土语群人口较少却有如此大的多样性的原因。综合来自不同领域的证据，这些未被记载的濒危土语得以被初步了解。
This variationist fieldwork-based study examines the connection between social factors and tone change from convex contour to flat. We conducted acoustic sociophonetic analyses on Tone 4 of Yangliu Lalo, an endangered Tibeto-Burman language whose community is currently shifting to Southwestern Mandarin. Hypothesis: As speakers’ use of Lalo decreases, the likelihood of a flat realization of Tone 4 and reduced F0 span increases, while speakers who use Lalo more have a dynamic Tone 4 contour and a wider F0 span.
Methods: We recorded wordlists and phrases with 34 Lalo speakers in Yangliu Village (both genders, ages 20-72, 71 words/speaker). We extracted 8,856 pitch tokens, all of which were stressed monosyllables between mid-tones in connected speech or as citation forms, normalized in semitones relative to the mid-tone, and normalized for duration as 200 relative-time units per syllable. The same speakers were interviewed on their language use and attitudes.
Results: Using Rbrul (Johnson 2009), we ran linear mixed-effects models on Age, Gender, Education, Lalo Use (a numerical index combining 1st language learned, reported frequency of Lalo use, and current best language), Language used in Family/Affinal Domains, Attitude Toward Lalo, and Voicing of Onsets, with Speaker and Word as random effects. The statistical model finds (R2=0.239) that the pitch peak occurs later in syllables with a voiced onset (p<0.0001), spoken in carrier phrase form (p<0.0001), for younger speakers (p=0.001), and for for those with stronger Lalo use (p = 0.003). The other factors and interactions were not significant.
We also ran multiple regression on F0 span, defined as the MeanT4 (high) minus Mean T2 (low), between Relative Time 50 and 150. Using input fixed factors of Age, Gender, Lalo Use, Domain, Attitude, Education, Child Language, Frequency, Importance, the step-up/step-down regression model selected one significant factor: Lalo Use (R²=0.31, p=0.000391; est. = 0.18 semitones for every 1 point on the 10-point Lalo Use index). Again, no significant interactions between fixed factors were found.
Analysis: The voicing effects reflect the classic tone perturbation of onsets (Hombert 1978), but the effects of language use are novel. While the flat variant minimizes articulatory effort and reduces redundant perceptual cues (a simpler level tone system), we find that usage effects play a role in this change and in the reduction of F0 span. It is not likely that contact with Baoshan or Standard Mandarin would induce a reduction of F0 span, since both have more complex tonal systems and wider F0 spans than Yangliu Lalo. Instead, decreasing F0 span represents a new balance between reduction of articulatory effort and maximization of perceptual distinctiveness (Flemming 2004).
References
Hombert, Jean-Marie. 1978. Consonant Types, Vowel Quality, and Tone. In V. A. Fromkin (ed) Tone: A Linguistic Survey. New York: Academic Press. 77-111.
Johnson, Daniel Ezra. 2009. Getting Off the GoldVarb Standard: Rbrul for Mixed Effects Variable Rule Analysis. Language and Linguistics Compass 3(1):359-83.
Flemming, E. (2004). Contrast and perceptual distinctiveness. In B. Hayes, R. Kirchner, & D. Steriade , Phonetically based phonology (pp. 232-276). Cambridge: Cambridge University Press.