Location: Room 357 Levin Building
Department of Linguistics
University of Pennsylvania
Speakers converge to linguistic variants they don’t hear: The case of the Southern /ay/ vowel
It has been well established that language users shift the way they speak to become more similar to their interlocutors, known as convergence. Studies on convergence have shown that speakers imitate the same linguistic forms they observe an interlocutor using. What is not known is whether speakers also converge by using linguistic forms they expect an interlocutor to use, even with these forms are absent in the input. I refer to this as Expectation-Driven Convergence, which entails converging toward a previously heard and then recalled linguistic target that is not observed locally. I find evidence for expectation-driven convergence in a controlled laboratory study. Participants converge toward a model talker’s Southern dialect by producing more monophthongal (i.e., more Southern) tokens of /ay/ (as in ride) after listening to a Southern-accented talker who never produces the /ay/ vowel. I use a Word Naming Game designed to elicit tokens of /ay/ and other Southern-accented vowels, which allows participants’ vowel productions to be compared before and after exposure to a Southern or Midland (control) talker. Findings add to a growing body of evidence showing that social knowledge influences linguistic behavior and shed light on the mental relationships between social and linguistic information.