Pre-velar /æ/-raising
Pre-velar /æ/-raising is a process in which some speakers raise /æ/ before /g/, but not /k/ (Bauer & Parker, 2008; Purnell, 2008; Wassink et al., 2009). This process appears to occur in Canada (Boberg, 2008), as well as some areas of the US near the Canadian border such as the Pacific Northwest and the Northern Midwest (Stanley 2018; 2019). My current research in this area focuses on the perception of this phenomenon and how it arises. I am also interested in the distribution of its production.
The production-perception link in phonologically-conditioned pre-velar/æ/-raising
My first generals paper investigated the relationship between production and perception in phonologically-conditioned. More specifically, I tested if there is a correlation between listeners’ production of /æ/ in the pre-voiced velar context and their boundary between /æ/ and /ɛ/ across the /æ/-/ɛ/ continuum, and whether or not this boundary is phonologically conditioned by the voicing of the following velar. I also investigated if there was a correlation between individuals’ productions and their self-report data.
The perception of pre-velar /æ/-raising and its relation to phonological context, production and metalinguistic awareness
For my dissertation, I investigated the perception of pre-velar raising and how variation in perception relates to phonological context, production and metalinguistic awareness. I am interested in how variation in perception arises and what this can tell us about how we form phonological categories.
Output
- Sullivan, L. (2020.) The production-perception link in phonologically-conditioned pre-velar/æ/-raising. Generals Paper 1. University of Toronto, Toronto ON. [unpublished manuscript] *Note: This manuscript incorrectly states that duration was manipulated. Duration was not manipulated. An unpublished follow-up study showed that the effects found in this study hold when duration is controlled for, though the size of the effect shrunk
- Sullivan L. (2020.) The production and perception of prevelar /æ/-raising by Canadian and American English speakers. Acoustics Virtually Everywhere: The 179th Meeting of the Acoustical Society of America. Online.
- Sullivan, L. (2021.) Beg, bag or bEYg? Reliability of self-reported /æg/-raising in North American English. Summer Phonetics/Phonology Forum 2021. University of Toronto, Online. [presentation]
- Sullivan, L. (2022.) BAG-raising in Ontario and Colorado. Montreal-Ottawa-Toronto Phonology/Phonetics Workshop 2022. University of Ottawa, Online. [presentation]
- Sullivan, L. and Rosen, N. (2023.) Vowel Patterning in Manitoba. Congrès annuel de l’Association canadienne de linguistique 2023 | 2023 annual meeting of the Canadian Linguistic Association. York University. Toronto, Ontario. [poster]
- Sullivan, L. and Rosen, N. (2023.) Bag, ban and bang: /æ/-raising in the Interlake. Change and Variation in Canada 12. University of Toronto Mississauga. Toronto, Ontario. [presentation]
- Rosen, N. and Sullivan, L. (2023.) Front vowel patterning in the Interlake region of Manitoba, Canada. In: Radek Skarnitzl & Jan Volín (Eds.), Proceedings of the 20th International Congress of Phonetic Sciences (pp. 3641–3645). Guarant International. [paper]
- Sullivan, L. (2023.) Hearing beg and bag: Variation in the perception of pre-velar /æ/-raising. In: Radek Skarnitzl & Jan Volín (Eds.), Proceedings of the 20th International Congress of Phonetic Sciences (pp. 501–505). Guarant International. [paper]
Gender-based sound symbolism in given names
I investigated the relationship between the phonology of given names and their gender. Correlations between phonological factors and gender may be suggestive of a sound-symbolic relationship. Previous research in this area has focused primarily on English and have found that these relationships exist (e.g. Cutler et al., 1990; MacAuley et al., 2018; Sidhu & Pexman, 2015; Slater & Feinman, 1985; Wright et al., 2005).
English and French
My M.A. research expanded on previous work by looking at French. A corpus analysis showed similar correlations between language and gender in both French and English, while a “name gendering” task showed that English-speaking listeners categorized nonce names as male or female based on both factors specific to both French and English and factors shared by the two languages. This suggests that further research into the possibility of cross-linguistic sound-gender correlations is warranted. More recent studies in other, less related languages, such as Japanese (Shinohara & Kawahara, 2013), Urdu (Mohsin, Sullivan & Kang, 2019), Cantonese (Wong & Kang, 2019) Armenian (Ananthathurai et al. ,2019) and Kutchi (Ananthathurai et al., 2019).
Korean
My second generals paper investigated sound patterns in Korean given names, as well as how English and Korean speakers use those patterns to assign gender to names using similar methodology to my study of English and French names. In addition to proposed cross-linguistic patterns, this study also looked at how language-specific sound symbolic patterns in Korean affected the sound patterns in given names and if and how English and Korean speakers used these patterns. The results indicate that most cross-linguistic patterns hold in Korean, however the language-specific patterns interfere with this, both in the corpus of names, and in the name gendering task, suggesting that language-specific pattern may be able to interfere with cross-linguistic ones.
Output
- Sullivan, L. (2018.) The phonology of gender in French and English given names. Summer Phonology Forum 2018. University of Toronto, Toronto ON. [presentation]
- Sullivan, L. (2018.) The phonology of gender in French and English given names. Forum Paper. University of Toronto, Toronto ON. [unpublished forum paper]
- Sullivan, L. & Kang, Y. (2019.) In Sasha Calhoun, Paola Escudero, Marija Tabain & Paul Warren (eds.) Proceedings of the International Congress of Phonetic Sciences, Melbourne Australia 2019. (pp. 2124-2128). Canberra, Australia: Australasian Speech Science and Technology Association Inc. [paper] [poster]
- Sullivan, L. (2020.) Universal and cross-linguistic gender-based sound symbolism in Korean given names. Generals Paper 2. University of Toronto, Toronto ON. [unpublished manuscript] [data]
- Kang, Y. & Sullivan, L. (2021.) Phonology of gender in Korean personal names. Korean Linguistics in Crosslinguistic Context. Cornell University, Online. [presentation]
- Sullivan, L. & Kang, Y. (2023.) Interacting sound symbolic patterns: The case of vowel sound symbolism in Korean given names. In: Radek Skarnitzl & Jan Volín (Eds.), Proceedings of the 20th International Congress of Phonetic Sciences (pp. 2144–2148). Guarant International. [paper]
Other Projects
- Web-based (Online) Data Collection – Using web-based interfaces to conduct experiments and collect production and perception data
- Sullivan, L. (2022.) Participant error in online production data collection in Gorilla. Challenges for change: A crowd-sourced brainstorming session. Labphon 18 Satellite Workshop. Online. [presentation] [write-up]
- English Demonym Allomorphy – Phonological conditioning factors of the selection of -an and -ian, including formal analysis, corpus analysis and possible future experiment
- Individual Differences in Phonetic Perception – Correlation between gender, AQ score and the perception of VOT in velar, alveolar and labial stops
- Sullivan, L. (2019.) The effect of cognitive processing style on the perceptual compensation of stop voicing for place of articulation. PsyLinCS UTM Workshop 2019. University of Toronto Mississauga, Mississauga ON. [poster]
- Pedagogy – Training TAs to provide effective feedback on student writing
- Sullivan, L., Vearncombe, E. and Sanders, N. (2021.) Grading grading: Training for consistency, efficiency, and effectiveness in marking linguistics writing. Special Session on Pedagogy, Congrès annuel de l’Association canadienne de linguistique 2021 | 2021 annual meeting of the Canadian Linguistic Association. Online. [presentation]