| Linguistic Short-Term Memory (STM) refers to the capacity-limited ability to actively maintain relevant linguistic information over short periods of time. Most studies of STM capacity limits are based on repetition of sequences of items in the same order as presented (serial recall). While for auditory linguistic (or spoken) material, a capacity limit in serial recall (i.e. span) of about 7+/−2 words is generally observed, the STM span for American Sign Language material is of 4.9+/−1 items. The present work shows that, contrary to previous interpretations, the shorter signed STM span cannot be attributed to phonological factors or to longer sign duration as compared with spoken word duration. The present studies also show that similar signed STM spans are present in ASL for both Deaf native ASL signers and native ASL/English native bilinguals, indicating that the lower signed STM span cannot be attributed to lower memory abilities in deaf individuals. Furthermore, the present studies show that the advantage of serial recall for spoken material over signed material is also present in backward spans, where participants are required to recall items in the reverse order. These results on serial recall tasks suggest that the ‘magical’ span of 7+/−2 in STM might rely, at least in part, on the striking ability of the auditory system to store temporal order information, and also on the coincidental reliance on ordered recall in traditional STM assessment. Importantly, the present work shows that despite their shorter STM span, comparable working memory (WM) resources are present in Deaf ASL signers and hearing English speakers during language use, indicating similar abilities to maintain and manipulate linguistic information online. The finding of a STM span of 4.9+/−1 signs in Deaf and Hearing native ASL signers confirms the view that the STM span of 7+/−2 items in speakers is an exception, probably owing to the reliance of STM measures on temporal order storage. The present studies call for adjustments in the norms used to measure mnemonic abilities in users of signed languages. |
