Poster C2, Friday, August 17, 10:30 am – 12:15 pm, Room 2000AB

Pre-output language monitoring in sign production

Stephanie Ries1, Soren Mickelsen1, Linda Nadalet1, Megan Mott1, Katherine J. Midgley1, Phillip J. Holcomb1, Karen Emmorey1;1San Diego State University

A domain-general monitoring mechanism is proposed to be involved in overt speech monitoring. This mechanism is reflected in a medial frontal component, the error negativity (Ne), present in both errors and correct trials (Ne-like wave) but larger in errors than correct trials. In overt speech production, this negativity starts to rise before speech onset and is therefore associated with inner speech monitoring. Here, we investigate whether the same monitoring mechanism is involved in sign language production. Twenty Deaf signers (ASL dominant) and sixteen hearing signers (English dominant) participated in a picture-word interference paradigm in ASL. Distractor words could be semantically-related, identical, or unrelated to the picture name. As in previous studies, ASL naming latencies were measured using the keyboard release time (any manual hesitations after keyboard release were removed from the data). The analysis comparing the Ne to the Ne-like wave included only the participants who produced more than 5 errors overall (12 Deaf signers, 11 hearing signers). EEG results revealed a medial frontal negativity peaking within 15 ms after keyboard release in the Deaf group. This negativity was larger in errors than correct trials, as previously observed in spoken language production. No clear negativity was present in the hearing group. We then analyzed the effect of the distractor type on behavior and on the amplitude of the Ne-like wave over all Deaf participants (N=20). Semantically-related distractor-word pairs and identical distractor-word pairs yielded faster reaction times than unrelated distractor-word pairs (p=.020 and p<.001 respectively). However, there was no effect of distractor type on the amplitude of the Ne-like wave. We therefore also conducted a post-hoc analysis in which we divided the semantically-related stimuli into categorically-related (e.g., truck and car, n=32) and associatively-related (e.g., king and crown, n=16) picture-word pairs. The associatively-related picture-word pairs yielded marginally shorter RTs (p=.062) and a marginally smaller Ne-like wave in comparison with the unrelated picture-word pairs (p=.082). There were no differences between the other conditions for Ne-like wave amplitudes, though RTs were also marginally shorter in the categorically-related versus unrelated conditions. Our results indicate that a similar medial frontal mechanism is engaged in pre-output language monitoring in sign and spoken language production. These results suggest that the monitoring mechanism reflected by the Ne/Ne-like wave is independent of output modality (i.e. spoken or signed) and likely monitors pre-articulatory representations of language. In particular, this medial frontal monitoring mechanism may be sensitive to lexical access difficulty, as indexed by the marginal facilitation effect for associatively-related compared to unrelated picture-word pairs. Differences between groups may be linked to several factors including differences in language proficiency, with more variable lexical access to motor programming latencies for hearing signers, and lower error awareness in hearing signers compared to Deaf signers.

Topic Area: Control, Selection, and Executive Processes

Back