Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Poster Slams
An fMRI investigation of the interplay between representational content and syntactic function during the comprehension of language describing state change
Poster D48 in Poster Session D with Social Hour, Friday, October 7, 5:30 - 7:15 pm EDT, Millennium Hall
Yanina Prystauka1, Gerry Altmann2; 1UiT The Arctic University of Norway, 2University of Connecticut
Comprehending language describing change, e.g. “The chef will chop the massive onion”, requires comprehenders to represent the changed object in, minimally, its initial and end states (Altmann & Ekves, 2019). Previous work (Hindy et al., 2012; Solomon et al., 2015) shows that once instantiated, both the initial and end states are maintained in memory, allowing the unfolding language to refer to either state, e.g. “The chef will chop the massive onion, and then/but first he will smell the onion”. But what brain mechanism might control which representations become (more) activated, and when? What is its role in mediating between the syntax used to convey event structure and the intended meaning of that structure? Behavioral evidence (Prystauka et al., submitted) shows that representations of both the chopped and unchopped onion are active after “…will chop the massive onion”, but the unchopped state is considerably less active (if at all) after the participial form “… will choose the chopped onion”. Hindy et al. (2012) and Solomon et al. (2015) had shown, for “…will chop…”, that this simultaneous activation of the distinct states of the onion results in increased activation, relative to “…will weigh…”, in left posterior ventrolateral prefrontal cortex (pVLPFC); they equated this effect with increased competition (for selection) between the distinct object-state representations of the onion. Here, we ask about “…the chopped onion”; whereas after “…will chop…”, representations of both the unchopped and chopped onion are available for subsequent reference, they are not after “…the chopped onion” – what are the consequences of this for pVLPFC activation and its role in the construction of event representations? fMRI was acquired while participants (N=22) performed a sentence reading task in which we manipulated the degree of change described by the verb (chop/weigh) and the modifier type (massive/chopped). Following Hindy et al., (2012), we used a Stroop task to first identify conflict-sensitive voxels in pVLPFC per individual and used these as our region of interest (ROI); increased activation in these voxels is an index of competition (January et al., 2009). We anticipated increased activation in the chop-massive condition relative to the weigh-massive condition. Of critical interest was how the weigh-chopped condition would pattern – more like weigh-massive (suggesting no or little competition between object states), or more like chop-massive (suggesting competition and dual activation)? Activation in the ROI was higher in the chop-massive condition than in the weigh-massive condition, suggesting dual state activation following state-change verbs (replicating our prior studies). Critically, activation was higher for sentences with participles (weigh-chopped) than sentences with adjectives (the weigh-massive and chop-massive conditions). We propose that whereas greater activation of stroop-sensitive voxels in pVLPFC after chop-massive than after weigh-massive reflects competition (i.e. representational conflict that interferes with selection), the additional increase after weigh-chopped reflects the additional control required to actively suppress (c.f. Grindrod et al., 2008) the representation of the onion in its unchopped state, reflecting the function of the participial form. pVLPFC is thus central to controlling the accessibility of key components of event representation during event comprehension.
Topic Areas: Meaning: Combinatorial Semantics, Control, Selection, and Executive Processes