Presentation

Brendan Costello¹, Chiara L. Rivolta^1,2, Núria Sánchez^1,2, Francisco Vera^1,3, Marcel Giezen¹; ¹BCBL (Basque Centre on Cognition, Brain and Language), ²University of the Basque Country (UPV/EHU), ³Universidad Rey Juan Carlos

Research into the processing of sign languages is limited by the unavailability of lexical characteristics (or large corpora) for most sign languages. Lexical databases for British Sign Language (BSL) and American Sign Language (ASL) include measures such as familiarity and iconicity ratings or phonological neighbourhood density (Vinson et al., 2008; Sehyr et al., 2021). As part of an ongoing project to expand the Spanish Sign Language LSE-Sign database, a lexical database with 2400 LSE signs (Gutiérrez-Sigut et al., 2016), we collected familiarity and iconicity ratings from 90 deaf signers (half of whom were native signers) for a subset of 300 lexical signs in the database. These signs were chosen to be representative of the full database (in terms of phonological form) and to include a broad range across the frequency and iconicity dimensions; additionally, 200 of the signs had meanings represented by an image in the Multipic database (Duñabeitia et al., 2018). The familiarity ratings show a broadly normal distribution whereas the iconicity ratings tend toward a binomial distribution, with signs being rated as either highly iconic or not iconic. The data reveal a negative correlation between familiarity and iconicity, confirming a pattern found for other sign languages (Sehyr et al., 2021). [We are currently in the process of collecting iconicity ratings for these 300 signs from sign-naïve hearing individuals to examine how experience with a sign language influences the perception of iconicity. Additionally, we are developing a measure of phonological neighbourhood density and hope to be able to report the results of this analysis also.] To examine the impact of these lexical properties on sign processing, we carried out a lexical decision task with 200 of the rated signs plus 200 pseudosigns (created by changing the handshape, location or movement of a real sign so that it no longer had a meaning). Results from forty-two deaf signers (half of whom were native signers) showed a clear lexicality effect: responses to real signs were faster and more accurate compared to pseudosigns. Analysis of the responses to real signs revealed a facilitatory effect of familiarity: signs with higher familiarity ratings had more accurate and faster responses (with no difference between native and non-native signers). Iconicity was also associated with greater accuracy, but the effect on reaction times depended on sign language background: native signers – but not non-native signers – responded faster to iconic signs. [We also collected data on a picture naming task for these 200 signs from the same participants and will be able to report the results of this study. We also intend to incorporate the phonological neighbourhood density index into the analysis of the lexical decision and the picture naming responses.] These results reveal that lexical access in a signed language bears similarities to spoken language access – we see typical lexicality and familiarity effects – but there are also modality-specific effects: iconicity impacts how signs are processed, and native signers appear to be more sensitive to this visual property than non-native signers.

Topic Areas: Signed Language and Gesture, Meaning: Lexical Semantics