Categorical versus gradient properties of handling handshapes in British Sign Language (BSL) : evidence from handling handshape perception and production by deaf BSL signers and hearing speakers

Sign languages include partially lexicalised signs (known as depicting constructions, DCs) that have been argued to blend linguistic and non-linguistic components, although it is unclear what these components are. To describe object handling, signers produce handshapes that represent how the hands shape for handling, but it has not yet been fully established whether the continuous object size is described by discrete handshapes in British Sign Language (BSL). The thesis examines whether experience with sign language influences perception and comprehension of BSL handling handshapes. In the first study, categorical perception (CP), using the identification and ABX discrimination tasks, is examined for handling handshapes (HHs) in BSL. The experiments reveal that adult deaf BSL signers and hearing non-signers perceive continuous HHs categorically while remaining perceptive to gradient aperture changes. Deaf BSL signers were more accurate than hearing non-signers when discriminating between handshape stimuli; this is likely due to visual language experience. However, reaction times showed no processing advantage suggesting that categorisation of BSL HHs has a general, visual-perceptual rather than linguistic basis. The second study examines whether deaf BSL signers compared with hearing non-signers express and interpret gradient sizes of manipulated objects categorically in discourse. Handling of objects gradiently increasing in size was recorded in BSL narratives, in English narratives via co-speech gesture and pantomime; recordings were shown to another group of judges who matched handling productions with the objects. All participants reliably associated smaller objects with smaller apertures and larger objects with larger apertures; however, in BSL and co-speech gesture, handshapes were not completely interpreted as gradient variations in comparison with pantomime. When gestures become more strategic or unusual, e.g. pantomime, speakers introduce finer-grained encoding of object sizes. The discontinuous patterns suggest that HHs have underlying representations outside of the linguistic realm; their categorisation arises from visual-perceptual experience that is embodied through interaction with real life entities. In discourse, handling constructions are partly conventionalised and may become decomposable in BSL overtime but it is suggested here that general cognitive and perceptual factors contribute to the conventionalisation, rather than purely linguistic. Further, the findings from both experiments lend support to the argument that HH category structure is graded. This thesis contributes to debates about the relationship between visual perception and language processing and the complex interface between language and gesture and highlights the nature of language as a multimodal phenomenon.