The Grammar of Sign Language Movement

Sign languages are naturally, fully-fledged languages equally complex to spoken languages. Sign languages are formed by manual (hands and arms) and non-manual components (e.g. position of head, chin, eye brows or upper body) in the three-dimensional signing space in front of the signer. Similar to spoken words that are composed of individual sounds, also signs are formed by discrete parts. Signs are formed by the hand configuration (shape and orientation), place of articulation and the movement component. In this project the movement component, i.e. the dynamics within the signed signal is in focus. Pre- vious work suggest that kinematic measures of hand movement (how far hands travel, how fast hands travel, how fast hands come to a stop, etc.) reflect underlying meaningful distinctions in the language. For instance, it has been shown that stress is marked in American Sign Language by a faster movement, that is signs that are stressed are signed faster than unstressed signs. Further, the movement dynamics differentiates two diverging groups of verbs: a) telic verbs that describe an event with an endpoint (e.g. arrive) and b) atelic verbs that lack such and endpoint (e.g. analyze). It has been shown that verbs describing an event with an endpoint (but not verbs lacking this end- point) have a sharper ending movement to a stop. However, to date few is known about which of the various kinematic variables are used within sign language production and in which contexts they are used to convey linguistic distinctions within meaning and grammar. In this project the motion capture methodology and the EMG (Elektromyografie) method are used to examine the linguistic role of the different kinematic variables in Austrian Sign Language (ÖGS). Two departments of the University of Salzburg work in close collaboration and combine their exper- tise regarding the analysis of sign languages (Linguistics Department) with the expertise to analyze complex motion sequences (Sports Department). It is predicted that there is a general mapping from body movement dynamics to the grammatical structure of sign languages. This mapping from body dynamics to grammatical structure is hypothe- sized to make the perception of distinctions as easy as possible for the sign language perceiver. We expect to develop novel analytical techniques, grounded in motion capture expertise, that will contribute critical answers to linguistic puzzles that require empirical techniques to resolve. Knowledge of the characteristics of the dynamics within signing can be used to improve automatic recognitionranslation of signed material, construction of animated avatar models, sign language teaching, interpreter training, and development of educational materials. This project is the first instrumental analysis for ÖGS and will provide important information on its underlying grammar.