Differential contributions of left-hemispheric language regions to basic semantic composition

Abstract

Semantic composition, i.e. the ability to combine single words to form complex meanings, is a core feature of human language. Despite growing interest in the basis of semantic composition, the neural correlates and the interaction of regions within this network remain a matter of debate.In the present fMRI study, we designed a well controlled two-word paradigm in which phrases only differed along the semantic dimension while keeping syntactic information alike. 33 healthy participants listened to meaningful phrases ("fresh apple"), anomalous phrases ("awake apple") and pseudoword phrases ("awake gufel") while performing both an implicit and an explicit semantic task. We identified neural signatures for distinct processes during basic semantic composition: The general phrasal composition process, which is independent of the plausibility of the resulting phrase, engages a wide-spread left hemispheric network comprising both executive semantic control regions as well as general conceptual representation regions. Within this network, the functional connectivity between the left anterior inferior frontal gyrus, the bilateral pre-supplementary motor area and the posterior angular gyrus specifically increases during meaningful phrasal composition. The anterior angular gyrus, on the other hand, guides more specific compositional processing evaluating phrasal plausibility. Overall, our results were stronger in the explicit task, pointing towards partly task-dependent involvement of the regions.Here we provide a separation between distinct nodes of the semantic network, whose functional contributions depend on the type of compositional process under analysis. For the first time, we show that the angular gyrus may be decomposable into two sub-regions during semantic composition.

Modelling brain representations of abstract concepts

“…Critically, the current study used a task that required participants to activate and contextualize abstract concepts. In this task, we identify the angular gyrus as a critical hub for the dynamic use of abstract knowledge, consistent with the view that this region plays a key role in combinatory linguistic processing (David & Yee, 2019;Graessner et al, 2021;Price et al, 2015;Pylkkänen, 2019Pylkkänen, , 2020. Such combinatory processing may be particularly critical for abstract concepts, which more strongly need to be contextualized in a situational way during everyday use.…”

“…Our findings yield multiple key insights into abstract concept representation: First, our findings provide novel evidence for IPC as a core area for concept coding (Binder et al, 2009). Using a task that required participants to activate and contextualize abstract concepts, we identify the angular gyrus as a critical hub for the dynamic use of abstract knowledge, consistent with the view that this region plays a key role in combinatory linguistic processing (Graessner et al, 2021;Price et al, 2015;Pylkkänen, 2019). Such combinatory processing is critical in real-life situations that require applying abstract knowledge in a situational and meaningful way.…”

Minimal phrase composition revealed by intracranial recordings

“…More broadly, pMTG is robustly associated with basic phrase comprehension deficits across a large number of studies. Data from brain stimulation implicate postero-middle temporal areas in syntactic processing (Duffau et al, 2014), and another recent red-boat study in fMRI implicated a broad portion of pMTG in meaningful phrase composition (Graessner et al, 2021). Impairments in grammaticality judgments have primarily been linked to posterior temporal lobe damage (Wilson and Saygin, 2004).…”

“…Functional neuroimaging studies using the red-boat paradigm implicate the left anterior temporal lobe (ATL), specifically the temporal pole (Antonucci et al, 2008;Lambon Ralph et al, 2012;Wilson et al, 2014;Zhang and Pylkkänen, 2018), inferior frontal regions (Graessner et al, 2021b(Graessner et al, , 2021a and posterior temporal regions (Flick and Pylkkänen, 2020;Matchin and Hickok, 2020) as crucial nodes for phrase composition with variations in the timing (180-350ms post-composition) and duration (50-100ms) of their engagement (Kochari et al, 2021).…”

Composition within and between Languages in the Bilingual Mind: MEG Evidence from Korean/English Bilinguals

“…This LATL composition effect is also observed for two-word phrases in written (Baron et al, 2010;Bemis and Pylkkänen, 2011;Westerlund et al, 2015) and auditory (Bemis and Pylkkänen, 2013;Sheng et al, 2019) comprehension cross-linguistically. Several other regions are also thought to contribute to combinatory processing: the LIPL, hypothesized to encode relational aspects of meaning (Boylan et al, 2017;Williams et al, 2017) and composition more generally (Binder and Desai, 2011;Bemis and Pylkkänen, 2013;Price et al, 2015Price et al, , 2016Graessner et al, 2021); the vmPFC, found sensitive to the semantic properties of combinatory expressions (Pylkkänen and McElree, 2007;Bemis and Pylkkänen, 2011;Pylkkänen et al, 2014;Blanco-Elorrieta et al, 2018); and the LIFG, associated with both long-distance dependencies (Stromswold et al, 1996;Leiken et al, 2015) and phrasal composition (Zaccarella and Friederici, 2015). The current study tests, for the first time, whether these regions (LATL, LIPL, vmPFC, LIFG) exhibit composition effects across code switches.…”