About the Project
Dr Joost Rommers (University of Aberdeen)
Dr Agnieszka Konopka (University of Aberdeen)
This project uses advanced analyses of electrical brain activity (EEG) to investigate the neural mechanisms of error-based learning during predictive language comprehension in native and non-native speakers of a language.
The brain is often said to be a ‘prediction machine’, which would have at least two benefits. First, when predictions are confirmed, a brain that predicts can run ahead of the input and deal with stimuli efficiently and at high speed. Second, when predictions are disconfirmed, a brain that predicts can use the resulting prediction error signals to learn and generate updated predictions in future encounters with similar input.
Human language, the most advanced communication system that evolution has produced, unfolds on a particularly rapid time scale (around 250 written words or 200 spoken syllables per minute). A wealth of evidence suggests that readers and listeners predict upcoming information, which likely helps them deal effectively with the speed at which language arrives, as long as predictions are confirmed.
However, it is much less clear what happens when predictions are disconfirmed. Recent work has highlighted several possible EEG correlates of prediction error, but their link to learning has not been systematically investigated. Doing so is expected to help clarify how the brain weighs the input (recent experience) against the prediction (reflecting cumulative experience with language and world knowledge) when learning from prediction error and thereby adapting its predictions for future processing. To what extent are predictions malleable, and does this depend on the strength of the existing prediction? Does having a stronger existing prediction enhance learning from unexpected input (because the input results in stronger prediction error and updating), or does it impair learning (because the existing prediction carries a stronger weight than the input)?
To address such questions, this project will examine putative neural signatures of prediction error as a function of factors affecting prediction strength. Prediction strength can be manipulated by varying recent input during the experiment (e.g., Rommers & Federmeier, 2018), it can be measured as sentence constraint capturing long-term predictions acquired outside of the lab (e.g., Rommers et al., 2017), and it can be measured as between-reader differences in linguistic knowledge that affords predictions (vocabulary knowledge in the case of native speakers and overall proficiency in the case of non-native speakers; Konopka et al., 2018).
This project will capitalize on the multidimensionality and high temporal resolution of EEG. The PhD candidate will use advanced time-domain and spectral analyses of electrophysiological signals to study error-based learning, which will involve developing a high level of expertise in digital signal processing and programming (Matlab, R), both skills highly useful in industry and academia. The project would be suitable for candidates with a background in psychology, neuroscience, or linguistics who have interests in language and electrophysiology.
Please send your completed EASTBIO application form, along with academic transcripts to Alison McLeod at [Email Address Removed]. Two references should be provided by the deadline using the EASTBIO reference form. Please advise your referees to return the reference form to [Email Address Removed].
Candidates should have (or expect to achieve) a minimum of a 2:1 UK Honours degree, or the equivalent qualifications gained outside the UK, in a relevant subject.
Rommers, J., & Federmeier, K. D. (2018). Predictability’s aftermath: Downstream consequences of word predictability as revealed by repetition effects. Cortex, 101, 16-30.
Rommers, J., Dickson, D. S., Norton, J. J. S., Wlotko, E. W., & Federmeier, K. D. (2017). Alpha and theta band dynamics related to sentential constraint and word expectancy. Language, Cognition and Neuroscience, 32(5), 576-589.
Based on your current searches we recommend the following search filters.
Based on your current search criteria we thought you might be interested in these.