CBSU bibliography search
To request a reprint of a CBSU publication, please
click here to send us an email (reprints may not be available for all publications)
Characterizing lexical complexity computations in the fronto-temporal language network
Authors:
BOZIC, M., SU. L., WINGFIELD, C. and MARSLEN-WILSON, W.
Reference:
Seventeenth Annual Meeting of the Cognitive Neuroscience Society, A86
Year of publication:
2010
CBU number:
7163
Abstract:
Bilateral fronto-temporal systems play a key role in speech comprehension. Evidence suggests that their activation is modulated by variations in the lexical complexity of the incoming input. General processing demands for linguistically simple words (e.g. dark) activate a bilateral subsystem, while combinatorial processes reflecting the presence of linguistic complexity (e.g. regular inflectional morphemes, play+ed) engage a left hemisphere perisylvian subsystem (Marslen-Wilson & Tyler, 2007). We applied Representational Similarity Analysis (RSA; Kriegeskorte et al, 2008) to establish information content and test the type of computation performed in fronto-temporal areas during processing of these different input types. Consistent with findings from univariate analyses, results showed that combinatorial processes triggered by the presence of inflectional suffixes (-ed) produce inter-correlated patterns of representation in left pars triangularis (BA45). Competition between multiple candidates for lexical access (e.g., in words with embedded stems, claim-clay), which increase general processing demands, correlated with the representation pattern in bilateral pars orbitalis (BA47). In a follow-up study we varied combinatorial and general processing demands in both inflected and derived words (e.g. played vs bravely), to test further how different complex inputs modulate the computations performed in fronto-temporal areas. The results show that inflected and derived words present different challenges for the speech processing system, and suggest a dynamic modulation of the type of computation performed in different linguistic contexts. This novel approach to studying language comprehension allows more precise characterization of the information carried and processed in the fronto-temporal language network.
MRC Cognition and Brain Sciences Unit

