The Effect of Syntactic and Semantic Cues on Lexical Access in Broca’s Aphasia

• ¹ University of California, San Diego, SDSU/UCSD Joint Doctoral Program in Language and Communicative Disorders, United States
• ² San Diego State University, School of Speech, Language, and Hearing Sciences, United States

BACKGROUND Current research with listeners with Broca’s aphasia (LWBA) supports the presence of a delay in lexical processing (Ferrill, et al., 2012). However, it is not currently clear whether it is access to the meaning of the individual words that is delayed, or whether there is delayed building of phrasal categories (e.g., merging the Determiner, the, with the Noun, girl to form the NP, the girl) which then may result in a failure to integrate the NP into the ongoing sentence. To address this question we take advantage of the unique properties of adjectives, which can provide a salient syntactic cue, indicating that a noun is forthcoming, and a semantic cue, describing unique features of the noun they modify. The following study examines whether on-time lexical access is evinced in LWBA (compared to control groups) when a syntactic or semantic cue is present. We hypothesize that if the deficit is due to delayed structure building, a semantically neutral adjective will remediate the deficit, whereas if it is due to delayed access to the meaning of individual words, an additional semantic cue may help evince on time lexical access. METHOD Data from 21 unimpaired listeners (YNC) and 5 LWBA were included. All LWBA survived a single unilateral left hemisphere stroke, and were premorbidly right handed, monolingual English speakers. We used an on-line cross-modal picture priming (CMPP) task, with sentences presented at a normal rate of speech. Listeners were presented with active sentences containing prenominal adjectives (in bold, below) in one of two conditions: BIAS condition (1) containing a semantically biased adjective, and NON-BIAS condition (2) containing a semantically neutral adjective matched for length and frequency. 1. The duck led the dirty*1 pig*2 along the winding country road [BIAS] 2. The duck led the dizzy*1 pig*2 along the winding country road [NON-BIAS]. In each sentence a visual probe was presented either at the offset of the adjective (*1) or the object noun (*2) during ongoing auditory presentation. Priming was measured by comparing response times to related and control probes – faster response times to related probes indicate a priming effect, thus activation of the lexical item. RESULTS Preliminary statistical analyses revealed that, for the YNC group, significant priming at the offset of the adjective in the BIAS condition (p=.03), but not at noun offset. For the NON-BIAS condition, no priming was found at the offset of the adjective (p>.05), but was observed at noun offset (p=.05). Results for LWBA demonstrated no priming at adjective offset for either condition; however, priming was observed at noun offset for the BIAS condition only (p=.01). DISCUSSION These results suggest that the delay in lexical processing for LWBA is due to delayed access to the meaning of individual words rather than due to delayed structure building. Although lexical access was facilitated for LWBA when a semantic cue was provided, this pattern was still protracted when compared to neurologically unimpaired listeners who were able to anticipate an upcoming noun in that condition.