Mapping the sounds of New Orleans: Acoustic measurements of vowel mergers

New Orleans is a city with a complex history, tumultuous social landscape, and distinctive linguistic features. Nevertheless, it shares some phonetic trends with other urban centers in the US South, such as the de-merger of vowel mergers (Tillery & Bailey, 2004; Labov et al., 2006; and Austen, 2020). This talk aims to discuss methods for acoustically measuring vowel mergers and to interpret the sociolinguistic implications of these measurements. I will examine the PIN∼PEN and the FEEL∼FILL mergers acoustically, drawing on a socially balanced and representative dataset of 115 sociolinguistic interviews. Using Pillai scores and Bhattacharyya’s affinity to measure overlap, I will first demonstrate why the latter is preferred for statistical analysis. Additionally, for the FEEL∼FILL merger, generalized additive mixed models (GAMMs) are employed to model formant movement directly, and I will highlight the flexibility and nuanced nature of this methodology. Through acoustic analysis, linear mixed models, and GAMMs, I will show that there are overlapping trends influenced by factors such as age, gender, ethnoracial affiliation, vowel duration, and phonetic environment. Geographic variation in formant movement will be modeled directly and visualized on geospatial maps. This study advances our understanding of vowel mergers in the urban US South and provides a model for investigating merging in other socially diverse environments.