Marjorie Herbert, in collaboration with Professor Acrisio Pires, presented her QRP work, "Code-blending or Distinct Grammar? Contact Signing in the American Deaf Community," at the LSA 2017 Annual Meeting in Austin Texas. Find out more about her project below:

The outcomes of contact between American Sign Language (ASL) and spoken English within the d/Deaf bimodal bilingual individual, and the community, remain contested and in need of further research attention. Furthermore, they are of particular theoretical consequence, since unlike in situations of unimodal contact, there is no overlap of phonetic material between the two languages, introducing a novel and innovative way to study how language systems interact (Emmorey et al. 2005, 2008, 2015). The current study addresses one dimension of this contact, ‘contact signing’ (Lucas & Valli 1989, 1992), as employed by d/Deaf signers in the U.S.

In previous work, Lucas & Valli define contact signing as a broad category of bimodal productions that are not entirely predicted by either the grammars of natural ASL or spoken English. This brings into question whether contact signing is governed by a grammar distinct from those of these two languages, or whether it is a type of bilingual behavior more akin to code-switching, or rather code-blending, as linguistic material can be produced simultaneously on the hands and the mouth. These authors conclude the former; however, other studies of this phenomenon (most notably Kuntze (2000) and Hoffmeister & Moores (1987)) have found signers with early exposure to ASL to be more adept at producing contact signing, as compared to those who acquired it after age 6. These results indicate the latter conclusion.

The present study contributes to the debate by using sophisticated statistical modelling to track how certain features identified in previous research co-vary with one another under multiple conditions and various controls. 38 d/Deaf participants were recruited in the Chicago, IL area. In preliminary analyses, we focused on four features: ASL classifier constructions, ASL pronoun usage, English mouthing, and fingerspelling. The former two are hypothesized to be derived from natural ASL, whereas the latter represent contact effects with spoken English (Battision 1978, Davis 1989, Brentari & Padden 2001, Lucas & Valli 1989, 1992).

Preliminary analysis, using a generalized linear mixed model and taking the linguistic data only, i.e. the four features mentioned above, into account, yielded two interesting results. English mouthing was found to be a predictor of fingerspelling (positive correlation) and of classifier use (negative correlation). The correlation between English mouthing and fingerspelling is less surprising, since both are thought to be English-derived. The second result is interesting, since in principle, mouthing and classifiers are not mutually exclusive, as they are transmitted primarily via separate articulatory channels, the mouth and the hands. This finding suggests some interference at the grammar level, which may follow from a code-blending analysis, rather than the separate grammar account. Furthermore, varying competencies in the two languages were found to predict variation in English mouthing: when reported competence in ASL and English were taken as predictors of English mouthing, both were found to be statistically significant, with p-values of less than 0.02. This result is also potentially indicative of code-blending on the part of these bilinguals.