The interaction between mouth actions and signs in Swedish Sign Language as an L2

In this study, we observed several patterns related to interaction and the synchronization of mouth actions and hands among L2 learners of Swedish Sign Language (SSL) compared to native signers. Previous research on signed languages has examined the synchronization of mouthings and mouth gestures (e.g. the edited volume by Boyes Braem & Sutton-Spence 2001; Crasborn et al. 2008; Johnston et al. in press). Another line of sign language research has investigated phonological errors made by L2 learners of sign languages (adult learners of signing as a second language) across a limited number of languages, primarily in the use of manual parts (e.g. Rosen 2004) as well as in the use of non-manual parts (e.g. McIntire & Reilly 1988), not including mouth actions. The current study draws from both of these research areas in an effort to answer two questions: (i) Do L2 learners use mouthings borrowed from spoken language to a greater extent than L1 (native) signers? And (ii) how do borrowed mouthings and mouth gestures interact with manual signs? In other words, what are the distribution and the scope of mouthings with respect to prosodic constituents of SSL? We based this study on an analysis of an L2 Swedish Sign Language corpus (Mesch & Schonstrom 2014), which consists of 9:06 hours of data from 17 different L2 signers, and a control group of 3 deaf native L1 signers who provided 0:34 hours of video. For the analysis, we sampled data consisting of various materials (interviews, picture and video retellings) from six L2 learners and compared it to parallel data from the control group. With respect to question (i), our analysis revealed a greater use of mouthings borrowed from spoken Swedish among the L2 group, and for (ii), we found a lack of prosodic features in spreading/interaction between mouthings and signs in SSL as an L2. Compared to the L1 control group, L2 learners either overused or avoided mouthing. Among L2 speakers, our analysis also revealed that Swedish function words (e.g. som ‘as’) often appeared as mouthings without corresponding manual signs, thus being articulated simultaneously with a “mismatched” sign (as in Example 1). Furthermore, the interaction of signs and mouthing was often dependent on Swedish mouthing: whereas L1 signers produced the pattern in Example 2, in which mouthing belonging to the first unit spread to the second unit, the L2 learners’ mouthings often followed a strict 1-to-1 pattern, in which mouthings accompanied single manual signs and rarely spread across sign boundaries. As shown in this study, linguistic factors impacting SSL as an L2 include bilingualism and different modalities, i.e. how mouthing and signs interact. This has implications for L2 teaching, in how L2 learners should be taught to use “unvoiced” articulations of spoken words with manual signs. For future research, it would be useful to compare these results with those of deaf people who are late learners of SSL, since they rarely have a spoken language as an L1 (and thus lack that type of interference).