Discovering African American English Speech Melodies
From left: Kristine Yu (linguistics), Meghan Armstrong-Abrami (languages, literatures, and cultures), Lisa Green (linguistics) and Brendan O’Connor (computer science)
It is well known that there are systematic differences between African American English (AAE)—a linguistic variety with set sentence structure, sound patterns, meaning, and word structure used by some, not all African Americans—and “mainstream/classroom/standard” American English. For example, speakers of AAE know that they can say she been married to mean that ‘She has been married’, but that if they put an accent on been, i.e., she BEEN married, that means that ‘She has been married for a long time and is still married’. But what does an AAE speaker do with their voice to produce an accented BEEN, and how does a listener perceive it? Such questions about the melodies of spoken AAE have been little studied. Our new collaboration and funding from the Institute for Diversity Sciences will allow us to make use of existing but unanalyzed recorded data to provide the foundations for such research.
We will share these pilot data and analyses at the international workshop on AAE speech melodies we are hosting at UMass in Fall 2018, and these pilot results and additional new fieldwork we’ll conduct based on those results will put us on track for a successful grant application to the National Science Foundation to continue this work. The discovery and understanding of patterns in spoken AAE that we will gain from the work may help address some of the widespread societal ills resulting from linguistic discrimination against speakers of AAE, as well as provide the basis for eventual improvements to speech recognition devices like Siri and Alexa for AAE speech. The research team includes UMass students Anissa Neal (linguistics), Deniz Ozyildiz (linguistics), Ayana Whitmal (linguistics), Tamira Powe (communication), and Su Lin Blodgett (computer science).