Motion events and event segmentation in american sign language
Abstract
As we may understand it, space-time is some ambient, homogeneous, and non-discrete phenomenon. However, we as human beings are able to segment space-time into chunks, or events, based on perceived start and end points (Zacks & Tversky, 2001). While event segmentation of this sort is a cognitive, species-wide ability, how events are then encoded into language is typological. Some languages focus on certain aspects of events, while other languages focus on others. Motion events have been of particular interest, specifically which elements from Talmy's catalog are encoded within a single event or sentence, or across a narrative. To that end, Bohnemeyer et al. (2007) (= B. et al.) aim to figure out how much information can be packaged within a single motion event in a given language. Their work covers a range of genetically unrelated languages, but does not take sign languages into consideration. Here I intend to extend B. et al.'s diagnostics to ASL. For want of an "Event Phrase", B et al. propose the Macro Event Property (MEP), which carves out events from a range of syntactic structures. Formally, the MEP contains any number of subevents that can be 'located' by a single time adverbial. B. et al. show that there are a number of unrelated grammatical constraints whose domain of application is the MEP. Namely, within a single macro event, a language may NOT: a. assign two of the same theta-roles to Ground DPs, (Argument Uniqueness Constraint, AUC) b. assign a particular Ground-denoting DP two theta-roles, or (Referential Uniqueness Constraint, RUC) c. encode and entail a change of direction in the path a Figure takes. (Unique Vector Constraint, UVC) Next, consider that motion events come in two flavors in ASL: there are those that are expressed by "fossilized" verbs (such as GO-TO, GO-OUT, and ZOOM) and those that are expressed by classifier constructions (e.g. w/e-CL:3-GO[drive]). While these two signing modes are differentiated by a host of syntactic, morphological and phonological characteristics, of interest here is their differing semantics: The former are time-anchored (despite being motion events), while the latter are space-anchored. As B. et al.'s constraints concern the amount of spatial information allowed within a single macro-event expression, there are grounds here to wonder whether the constraints are respected in both signing modes; in one, but not the other; or in neither. We might predict already, though, that classifiers--with their focus on space--may be allowed to express more spatial information than what B. et al.'s constraints can handle.
Degree
M.A.
Advisors
Wilbur, Purdue University.
Subject Area
Linguistics
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.