You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster E31, Friday, November 10, 10:00 – 11:15 am, Harborview and Loch Raven Ballrooms

Establishing a Bio-Marker of Object-State Competition

Yanina Prystauka1,2, Zachary Ekves1,2, Gerry Altmann1,2;1University of Connecticut, 2The Connecticut Institute for the Brain and Cognitive Sciences

Altmann and colleagues have postulated the need to represent not simply types (e.g. onions as a kind of vegetable) and tokens (a specific onion grounded in space and time) but also token-states (the onion peeled or chopped). fMRI studies have demonstrated during comprehension of action sentences such as “The chef will chop the onion. Then, she will weigh the onion” that the different states of the onion (intact/chopped) are simultaneously active. Selecting between the situationally-relevant state-representations of the same token (the onion) at the end of the 2nd sentence results in increased activation in brain areas also recruited during Stroop interference (VLPFC). This suggests that such selection entails conflict resolution. Conflict is not found in sentences describing minimal or no change, such as “The chef will smell the onion. Then, she will weigh the onion” [1]. Moreover, introducing a new token instead of referring back to the same token, as in “The chef will chop the onion. Then, she will weigh another onion” eliminates the need to resolve the competition and doesn’t cause increased activation in VLPFC [2]. The present study builds on the above, and seeks to relate research on sentence processing, cognitive control, episodic memory and meaning representation using brain oscillations and time-frequency analysis. Specifically, we ask (i) how the effects found with fMRI manifest in EEG and (ii) whether these effects differ from known effects due to lexical and referential ambiguities. EEG was acquired using a 256-channel HydroCel Geodesic Sensor Net while participants (N=26) were reading sentences (N=320) presented to them one word at a time. Half of the sentences were designed to elicit the object-state change effect and were similar to those introduced above. Another half were fillers that either had a lexically ambiguous noun, or a referentially ambiguous pronoun. A time-frequency analysis of EEG power, synchronized from the onset of the final noun in the second part of the sentence or the referentially ambiguous pronoun, revealed a significant increase in alpha and lower beta frequencies only in sentences describing significant change and referring back to the same token – an effect we believe might be an EEG marker for conflict due to multiple object-states. This finding is consistent with literature relating alpha oscillations to cortical inhibitory processing which fits the hypothesis advanced by [1]: selecting the relevant token state required inhibition of the irrelevant one. Moreover, effects in beta frequency might be related to the maintenance and prediction hypothesis [3]: maintaining the current cognitive set might require more resource in the case of the substantial change condition than in the case of the minimal change condition. 1. Hindy, N. C., Altmann, G. T., Kalenik, E., Thompson-Schill, S. L. (2012). The Journal of Neuroscience, 32(17), 5795-5803. 2. Solomon, S. H., Hindy, N. C., Altmann, G. T., Thompson-Schill, S. L. (2015). Journal of Cognitive Neuroscience. 27(12), 2324-2338. 3. Lewis, A. G., & Bastiaansen, M. (2015). Cortex, 68, 155-168.

Topic Area: Meaning: Discourse and Pragmatics

Back to Poster Schedule