January 22, 2020 | 2:00PM
Green 410
Justine Mertz, who is visiting from Université Paris Diderot - Paris 7, will be presenting on her work with Diane Brentari, detailing measures of phonological complexity in sign language. See below for more details.
On how to measure phonological complexity in sign languages
A measure of phonological complexity of signs is an important challenge in developing assessment tests for evaluating language competencies and language impairments in Deaf populations. In our study, we explore a new data-driven measure of complexity based on perceptual/articulatory criteria. We designed a repetition task to assess sign complexity in non-signers by evaluating fluency and accuracy in handshape, orientation, location and movement. We then compare this data-driven measure of complexity with a theory-driven measure based on sign language (SL) feature geometry. We adapted a phonological model for SLs (the Prosodic Model, see Brentari 1998) into a complexity scale for the three phonemic classes, i.e. handshape, location and movement. Currently, our data include 50 signs in French SL and 15 signs in Italian SL. I will discuss several interesting (preliminary) results: 1) the two measures indeed converge; 2) handshape is the phonemic class that better correlates with overall accuracy; 3) some signs still show considerable divergence.