Handshape Identification Across Contexts

Handshape is a core phonological parameter of sign languages, serving as a primary cue for lexical distinction (Battison, 1978; Brentari, 1998). Prior research in American Sign Language (ASL) demonstrates that handshapes that share phonological features are more perceptually confusable (Lane et al., 1976; Stungis, 1981). However, these studies used a limited set of 20 handshapes, often collapsed across other phonological manual parameters (movement, location, and palm orientation), and tested only small samples of participants (signers n=3, non-signers n=3). Consequently, it remains unclear how handshape identification is influenced by broader phonetic contexts. This study builds on prior work by examining perception of a larger, more phonologically diverse set of 36 ASL handshapes across systematically varied contexts (three locations × three movements × three palm orientations). With a larger participant sample, we can assess context-specific effects and interactions across location, movement, and orientation. Stimuli will be presented with controlled visual degradation (Stungis, 1981). Participants—both Deaf signers exposed to ASL before age 5 and hearing non-signers—will complete a 36-alternative forced-choice identification task. Trial-by-trial responses will generate detailed confusion matrices to quantify perceptual similarity and systematic misidentifications. This study aims to address several open questions. First, which handshapes are robustly identified under perceptual degradation? Second, how does the phonetic context—location, movement, and palm orientation—modulate perception of handshapes? Third, how does language experience shape sign perception, both overall identification accuracy and the pattern of errors? Finally, how well do current theoretical models of handshape representation (e.g., Brentari’s Prosodic Model, 1998; the feature set used in Stungis’s revised LBB2 confusion model, 1981; or the continuous model proposed by Stungis, 1981) predict handshape confusability across signers and non-signers?