Speaker: Jason Shaw
Title: Lecture #1: Modelling phonetic variation with the neural dynamics of movement preparation
Time: Friday, November 13th, 1-2:30pm
Location: 32-D461
Abstract: A Dynamic Neural Field (DNF) is a formal object developed within systems neuroscience with the intention of giving a theoretical foundation to the notions of cooperation and competition between neural populations (Amari, 1977). DNFs provide the formal foundation for a general theory of movement preparation (Erlhagen & Schöner, 2002). More recently, DNFs have been applied specifically to the preparation of speech movements and the interaction of this cognitive process with phonological representations (Gafos & Kirov, 2010; Kirkham & Strycharczuk, 2024; Roon & Gafos, 2016; Stern et al., 2022; Stern & Shaw, 2023a, 2023b; Tilsen, 2019). There are now neural-based explanations for common phonetic patterns, including contrastive hyperarticulation, trace effects in speech errors, phonetic convergence/divergence to an interlocuter, and leaky prosody. Drawing on the broader framework of Dynamic Field Theory (Schöner & Spencer, 2016), this lecture will introduce the formal definition of a DNF and applications to modelling phonetic variation.
Title: Lecture #2: Revisiting the gestural parameters of prosodic modulation with a new gesture dynamics
Time: Thursday, November 14th, 4.30-6pm
Location: 32-D461
Abstract: Fowler’s (1980) critique of extrinsic timing theories served as a launching point for the dynamical systems approach to phonological representations. Since then, there has been substantial development of theories centered on a particular dynamical system, the damped mass-spring (e.g., Browman & Goldstein, 1989; Saltzman & Munhall, 1989) and complexifications of this system (Byrd & Saltzman, 1998; Sorensen & Gafos, 2016). Early studies showed that articulatory variation across prosodic positions was difficult to capture parsimoniously with the parameters of the damped mass-spring system (e.g., Beckman et al., 1992; Cho, 2006). These shortcomings motivated theories of gesture-external modulation, which maintain the damped mass spring as the model of the gesture but conceptualize prosody as trans-gestural modulation of time (Byrd et al., 2006; Byrd & Saltzman, 2003) and/or space (Katsika et al., 2014; Saltzman et al., 2008). After presenting some new data that is problematic for these theories, I’ll revisit the gesture-intrinsic approach to prosody with an alternative proposal for gestural dynamics, developed recently by Michael Stern (Stern & Shaw, 2024).