Rethinking verb directionality in emerging and established sign languages

A Language and Cognition Seminar

Speaker: Dr Adam Schembri, theDepartment of English Language and Linguistics, University of Birmingham.

Abstract

There has been a growing interest in the role emerging sign languages can play in our understanding of language evolution. In particular, work on home sign systems (e.g., Goldin-Meadow & Brentari, 2017), Nicaraguan Sign Language (e.g., Kocab et al., 2015) and Al Sayyid Bedouin Sign Language (e.g., Sandler, 2017) has provided a range of evidence about how grammar develops in new sign languages. Often this data is compared to current understanding of established sign languages, such as American Sign Language (e.g., Montemurro et al., 2019). However, existing assumptions about the grammar of established sign languages, such as British Sign Language (BSL) and its related variety Auslan (the majority sign language of Australia), are beginning to be challenged by new corpus data (e.g., Johnston, 2018; Fenlon et al., 2018).

This paper will discuss concern that claims about language emergence need to be solidly grounded in a detailed understanding of the structure and use of both emerging and established sign languages. In particular, this paper will focus on the nature of verb directionality in sign languages. Indicating verbs in BSL, and other sign languages, can be directed towards locations in space associated with their arguments. The development of verb directionality has been studied in emerging sign languages, including Nicaraguan Sign Language, (Senghas & Coppola, 2001), as well as in artificial sign language studies (Motamedi et al., 2018). The data from both domains suggests that directionality develops over time from an initial stage in which arguments are represented by the signer’s own body to one in which arguments are associated with abstract locations in space around the signer’s body. Work drawing on spontaneous data has revealed, however, that the properties of indicating verbs in established sign languages are not as well understood as the literature might suggest. Fenlon et al. (2018) investigated a range of linguistic and social factors in 1,436 indicating verb tokens collected from a corpus of BSL, widely believed to be one of the oldest existing sign languages. Unlike previous claims based on elicited data (e.g., Morgan et al., 2006), the analysis shows that directionality in BSL is not obligatory (as had also been reported for Auslan, see de Beuzeville et al., 2009). Directionality is, in fact, variable, and conditioned by several linguistic factors, such as constructed action (a type of enactment used to represent referents), animacy, and co-reference. Moreover, the corpus data show that the use of abstract spatial locations away from the signer’s body is not the preferred strategy in BSL (Cormier et al., 2015).

This paper reports additional factors that may influence when a BSL sign is modified directionally for argument reference. Using the BSL Corpus indicating verb dataset, the influence of definiteness and variable argument noun phrase presence was investigated. This was based on the suggestion that definite referents might be established with more specific locations in the signing space compared to indefinite referents, as suggested by work on Catalan Sign Language (Barberà, 2016), and thus be more likely to trigger indicating verbs being directed towards these locations.

The study also explores whether the presence of an explicitly expressed argument in the clause may be significant, something not previously investigated. Results suggest that definiteness is indeed a significant factor, with definite arguments more likely to trigger directionality in indicating verbs than indefinite arguments. Variable argument presence was also important with modification more likely in clauses with null argument expression. Thus, there are a complex range of factors that influence the use of verb directionality in a ‘mature’ sign language. This has important implications for an understanding of the emergence and development of morphological ‘complexification’ in sign languages, as the use of abstract space is clearly not obligatory, and – unlike what is suggested in the emerging and artificial sign language literature – actually appears to be disfavored. We need to be mindful of this when making claims about emerging versus established sign languages in discussions of sign language emergence, and its contribution to language evolution more generally.