Cortical representation of the constituent structure of sentences

Abstract

Linguistic analyses suggest that sentences are not mere strings of words but possess a hierarchical structure with constituents nested inside each other. We used functional magnetic resonance imaging (fMRI) to search for the cerebral mechanisms of this theoretical construct. We hypothesized that the neural assembly that encodes a constituent grows with its size, which can be approximately indexed by the number of words it encompasses. We therefore searched for brain regions where activation increased parametrically with the size of linguistic constituents, in response to a visual stream always comprising 12 written words or pseudowords. The results isolated a network of left-hemispheric regions that could be dissociated into two major subsets. Inferior frontal and posterior temporal regions showed constituent size effects regardless of whether actual content words were present or were replaced by pseudowords (jabberwocky stimuli). This observation suggests that these areas operate autonomously of other language areas and can extract abstract syntactic frames based on function words and morphological information alone. On the other hand, regions in the temporal pole, anterior superior temporal sulcus and temporo-parietal junction showed constituent size effect only in the presence of lexico-semantic information, suggesting that they may encode semantic constituents. In several inferior frontal and superior temporal regions, activation was delayed in response to the largest constituent structures, suggesting that nested linguistic structures take increasingly longer time to be computed and that these delays can be measured with fMRI.

Disentangling meaning in the brain: Left temporal involvement in agreement processing

“…the "who", "what", "where" and "when"), irrespective of the part of speech that conveys them. According to the authors, this fits nicely with fMRI data showing ATL involvement in semantic combinatorial mechanisms (Pallier et al 2011), and connects to a larger literature on semantic memory and conceptual combination that assumes the ATL to be an amodal semantic hub (Baron et al 2010;Patterson et al 2007).…”

“…Within this network, the finding of differential responses for Person and Number Violations (Person Violation > Number Violation) would provide evidence for featurespecific mechanisms in the building and interpretation of a sentential relation, in line with theoretical analyses (Bianchi, 2006;Mancini et al 2013;Sigurdsson, 2004Sigurdsson, , 2008. Based on this, one could expect to find a dissociation i) in areas that are thought to support the extraction of morphosyntactic information to build syntactic structure (Hagoort, 2005;Pallier et al 2011); ii) in areas supporting the mapping between 10 morphosyntactic and semantic-discourse information, to derive the overall interpretation of a sentence. In the former case, we expect a difference between person and number to emerge in quantitative terms, because of the higher position in syntactic structure that person occupies compared to number.…”

“…Specifically, access to information encoded in a higher syntactic node could generate greater cognitive cost and so greater activation in the posterior portion of the MTG/STS (Hagoort, 2005;Lau et al 2008;Pallier et. al.…”

“…However, pMTG involvement was not limited to correct-agreement sentences: when contrasting the two incorrect conditions (Person > Number Violation contrast), Person Violations were found to generate a greater response in this area than Number violations, suggesting differential sensitivity to the type of information manipulated in verbal morphology. As reviewed above, functional interpretations about the involvement of the pMTG in language processing range from proposals advancing its role in the extraction of morpholological information for building syntactic structure (Pallier et al 2011) or in the retrieval of syntactic frames from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 20 semantic memory (Hagoort, 2005), to views emphasizing the centrality of this region in mechanisms at the syntax-semantics interface, such as the analysis of verb-argument relations (Bornkessel-Schleswesky and Schleswesky, 2013;Bornkessel et al 2005;Friederici, 2011). The current set of data allows us to distinguish between the two proposals.…”

“…A better framework for the discussion of these data is therefore provided by studies and models that assume pMTG involvement in the extraction of information from the input to build syntactic structure (Hagoort, 2005;Hickock and Poeppel, 2007;Lau et al 2008;Molinaro et al 2015;Pallier et al 2011). In this respect, the quantitative dissociation found between person and number would result from their different underlying structural representations (Bianchi, 2006;S.…”