Real-Time Processing of ASL Signs: Effects of Linguistic Experience and Proficiency

Word recognition in spoken language is known to be a dynamic, incremental, and continuous process. As words unfold in time, listeners activate a range of potential lexical candidates, including those that share semantic (Yee & Sedivy, 2006) and phonological (Allopenna, Magnuson, & Tanenhaus, 1998) features with the target word. Among the methodological techniques used to probe real-time recognition, the use of eye-tracking technology has been particularly informative, in that eye movements provide a rapid and detailed metric for determining the locus of the listener’s visual attention. In contrast, much less is known about how individual lexical items in a sign language are recognized as they unfold in real time. Signed languages such as American Sign Language (ASL), which are produced manually and perceived via the visual channel, have been shown to be acquired in a similar manner to spoken language when exposure begins at birth (Mayberry & Squires, 2006; Newport, 1985). It has been a challenge to probe the on-line recognition of signs, however, due to the fact that the visual channel is required both to recognize signs and to demonstrate comprehension via eye gaze. Although sign languages share many linguistic properties with spoken language, the manual modality and visual perception of signs has led to questions regarding the degree of sub-lexical representation that exists for single signs. Commonly, signs are thought to be composed of at least three distinct parameters, namely handshape, location, and movement. These parameters have frequently been considered as contributing partial lexical information with varying degrees of salience. For example, in several gating studies, the order of parameter recognition as signs were presented in increasing segments led to the finding that location information was identified early in signs, followed by handshape, and finally movement information (Emmorey & Corina,1990; Clark