2018 Friday Session B 1645

Friday, November 2, 2018 | Session B, Conference Auditorium | 4:45pm

The Development of Inter-Lexical Inhibition from 1st Grade to Adulthood
C. Blomquist, B. McMurray

Spoken word recognition is described as a competition process: As acoustic information unfolds over time, similar-sounding candidates (e.g., cap, cat) compete until one word can be disambiguated and “wins”. The dynamics of lexical competition develop from infancy through adolescence (Fernald et al., 1998; Rigler et al., 2015; Sekerina & Brooks, 2007), with competition resolution increasing in efficiency with age. However, it is not clear what components of competition are changing over development. One potential mechanism is inhibition between words, by which activated candidates suppress competitors (Dahan et al., 2001). The present study asked whether this inhibition changes with development.

We isolated inter-lexical inhibition (c.f. Dahan et al., 2001) with stimuli in which the coda of a target word (e.g., the /p/ from cap) was spliced onto the onset of a competitor (e.g., the /kæ/ from cat to make ca(t)p). This word-splice condition boosts activity for the competitor word. If inhibition is operative, this will inhibit activation of the target word, slowing its recognition. This was compared with a nonword-splice condition where the onset came from a nonword (e.g., ca(ck)p) that would not inhibit the target. A difference in target activation across the word-splice and nonword-splice conditions indicates an effect of inter-lexical inhibition above and beyond the effect of coarticulatory mismatch.

Lexical activation was measured in a visual world task in which participants saw four pictures on a screen, heard a word, and selected the referent while eye movements were monitored. We tested typically developing monolingual listeners from two age groups: 7- to 8- year-olds (N=40), 12- to 13-year-olds (N=42) and adults (N=40).

All age groups showed significantly more looks to the target in the nonword-splice condition relative to the word-splice condition (p < 0.0001), suggesting similar patterns of inhibition. However, given recent evidence that inhibition can be tuned with experience, we also examined the first half of the experiment as a purer measure of inhibition, since there was less opportunity for retuning of competition dynamics. The expected effect appeared for adults and older children (Figure 1), with a significant difference between the word-splice and nonword- splice from 732 to 2000 msec for adults, and from 644-1118 msec in 12- to 13-year olds.

However, for 7- to 8-year-olds there was no effect of inhibition during the first half of trials, suggesting evidence for inhibition across all trials may have been a result of learning over the course of the experiment.

Differences in inter-lexical inhibition across age suggest continued development of even basic skills like word recognition far later than is typically expected. From an applied perspective, this result is of interest because children with developmental language disorder show atypical patterns of real time word recognition (McMurray et al., 2010) that could be related to the late development of these inhibitory processes. Further, inter-lexical inhibition can be tuned in short-term training paradigms (Kapnoula & McMurray, 2016); this suggests that lexical inhibition may be amenable to interventions designed to address atypical patterns of processing in individuals with language disorders.