My research is concerned with the stylistic use of phonetic variables to construct social personae. I am particularly interested in how the interpretation of an individual's social identity is stored in the mind and linked to linguistic (phonetic, lemma, and syntactic) information. In order to investigate this, I use a combination of qualitative and quantitative methods, including ethnography, acoustic phonetic analysis, and speech perception experiments.
At UHM, I am the convener for the Sociolinguistics Research Group, a group of made up of faculty and students who work in a variety of areas within sociolinguistics.
Both qualitative and quantitative methods were used for the study that I report on in my dissertation. I conducted a year-long ethnography at Selwyn Girls' High, an all girls' school in Christchurch, New Zealand. I became familiar with a number of distinct social groups at the school, and I observed a binary distinction among the groups based on whether they adopted or rejected the school's norms. One example of how this was expressed was whether a group chose to eat lunch in the Common Room (CR) or not (NCR).
For the linguistic analysis I conducted acoustic phonetic analysis on tokens of like from the girls' speech. I found that a girl's realisations of like depended on a combination of her social grouping (CR vs. NCR) and the function of like (quotative like as in and Mum's LIKE "turn that stupid thing off." vs. discourse particle like as in Lily was LIKE checking out my brother..) Realisations of quotative like also appears to be related to how frequently a girl used it. I argue that the phonetic realisations were used by the girls as a part of constructing their social personae. Click here to download a copy of a paper that will appear in the conference proceedings of the 5th International Gender and Language conference.
There was also a perception element to the research. Girls at the school took part in three perception experiments. For two of the experiments, participants were asked to identify the token of like. The results provide evidence that individuals use their knowledge of lemma-based phonetic variation in production in order to identify the token. I presented results from two of these experiments at LabPhon 11 in Wellington.
As a part of identity construction, I am interested in how individuals manipulate phonetic variables in order to construct different linguistic styles. In collaboration with Penny Eckert and Kyuwon Moon (Stanford), I am involved in an ongoing investigation of prosodic variables as stylistic tools in an emerging peer social order.
In my MA thesis, I examined the effect that the age attributed to a speaker could influence the perception of vowels that are involved in an ongoing sound change in New Zealand English (NZE). In NZE, the vowels in TRAP, DRESS, and KIT are shifting, such that TRAP and DRESS have raised and KIT has centralised. As a result, older speakers of NZE produce different variants than younger speakers. I conducted two experiments, and results from both provide evidence that perceivers were influenced by the age they attributed to the speaker. Results from the first experiment appear in Te Reo, and results from the second experiment are reported in a forthcoming paper.
Another recent sound change in NZE is a merger of the diphthongs in NEAR and SQUARE. As a result of the merger, many younger speakers pronounce the words really and rarely identically, and many older speakers maintain a distinction. Jen Hay, Paul Warren, and I conducted an experiment to test whether participants would perceive distinct tokens of the diphthongs differently depending on the age and social class attributed to the speaker. The results from this experiment are published in the Journal of Phonetics.
In a paper co-authored with Jen Hay and Aaron Nolan, we report on results from an experiment examining the extent to which regional labels at the top of a response sheet affects perception of the vowel in KIT. This paper can be found in The Linguistic Review. In two follow-up studies, we have found that stuffed toys associated with a region could influence perception and that perception and production could be influenced by manipulating attitudes toward a region.
In both Hay, Warren, and Drager (2006) and Hay, Nolan, and Drager (2006) we observed an apparent effect of the experimenter. For both experiments, participants were met by two different experimenters: some met with a New Zealander, and other met with a speaker of Southern California English (me!). Participants who met with me responded differently on the perception tasks. In an article in the Australian Journal of Linguistics, we present results from production: participants produced significantly different realisations of the diphthongs in NEAR and SQUARE depending on which experimenter they met with.
To test the experimenter effect explicitly, we conducted a lexical access experiment where participants were asked to identify the odd one out from word triplets, such as beer bare bear. Participants listened to auditory instructions, read by a speaker of RP (who maintains a distinction) or a speaker of NZE (who merges the diphthongs). The paper reporting on results from this experiment is forthcoming in Language and Speech.
In collaboration with Abby Walker (OSU), I am examining phonetic variation in polysemous words, depending on their reported frequency of use and their age.
Jen Hay, Andy Gibson, and I are currently working on a paper that reports on results from four experiments investigating the perception of linking and intrusive /r/ in New Zealand English. The results indicate that trends in perception are consistent with those in production, and the degree to which an individual will perceive intrusive /r/ depends on the degree to which that individual uses it.
‘Ōiwi Parker Jones (Oxford) and I are currently collecting data for a corpus-based study on phonetic variation in Hawaiian Creole English (HCE). Watch this space!
Now that I am based in Hawai‘i, I hope to work much more on Hawai‘i Creole (a term used by linguists to refer to Pidgin, which is what the language is called locally).
I have also been involved in the ONZE Project as a research assistant and an advisee. We have weekly meetings where we discuss research ideas and results (and eat Anita's famous cake when we're lucky). For more information on ONZE, please visit our website.