The influence of lexical statistics on temporal lobe cortical dynamics during spoken word listening

Abstract

Neural representations of words are thought to have a complex spatio-temporal cortical basis. It has been suggested that spoken word recognition is not a process of feed-forward computations from phonetic to lexical forms, but rather involves the online integration of bottom-up input with stored lexical knowledge. Using direct neural recordings from the temporal lobe, we examined cortical responses to words and pseudowords. We found that neural populations were not only sensitive to lexical status (real vs. pseudo), but also to cohort size (number of words matching the phonetic input at each time point) and cohort frequency (lexical frequency of those words). These lexical variables modulated neural activity from the posterior to anterior temporal lobe, and also dynamically as the stimuli unfolded on a millisecond time scale. Our findings indicate that word recognition is not purely modular, but relies on rapid and online integration of multiple sources of lexical knowledge.

The peri-Sylvian cortical network underlying single word repetition revealed by electrocortical stimulation and direct neural recordings

“…Finally, recent work has also demonstrated that distributed cortical networks encode abstract statistical information about lexical and other linguistic features, including lexical frequency (Cibelli, Leonard, Johnson, & Chang, 2015; Leonard, Bouchard, Tang, & Chang, 2015). These studies challenge the notion of a ‘mental lexicon’ that is simply a dictionary of all known words (Gow Jr, 2012; Ullman et al, 1997), and provide evidence that many aspects of speech processing including acoustic-phonetic, phonemic, and lexical processing are deeply influenced by language statistics.…”

SyllabO+: A new tool to study sublexical phenomena in spoken Quebec French

“…Researchers in the fields of cognitive neuroscience of language, psycholinguistics, neurolinguistics, and experimental psychology can use it to create stimuli representative of the spoken language-both sublexical (e.g., syllable) and lexical (word) stimuli-controlled for normalized frequency, transition probabilities, or mutual information. Given the known effects of distributional statistics on speech perception and production, behaviorally and at the neural level (Carreiras, Mechelli, & Price, 2006;Carreiras & Perea, 2004;Cibelli, Leonard, Johnson, & Chang, 2015;Deschamps et al, 2016;Karuza et al, 2013;Leonard et al, 2015;Newport & Aslin, 2004;Pelucchi et al, 2009aPelucchi et al, , 2009bPeña et al, 2002;Saffran et al, 1996;Saffran, Johnson, Aslin, & Newport, 1999;Tremblay et al, 2012;Tremblay et al, 2016;Vitevitch, 2003;Vitevitch & Luce, 1998;Vitevitch et al, 1997Vitevitch et al, , 1999 it is important to control for these effects in order to avoid confounds that can mask other effects of interest, such as articulatory or phonological complexity. However, such important experimental control is only possible when databases providing this information exist, which was not the case for Quebec spoken French.…”

Neural correlate of the construction of sentence meaning

“…6-11) and word-level processes (e.g., refs. [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26]. However, the most distinctive feature of human language is its compositionality: the ability to create and understand complex meanings from novel combinations of words structured into phrases and sentences (27).…”