I am currently focusing on how high- and low-frequency neural oscillations may provide signatures of bottom-up and top-down information streams during sentence comprehension. This work is grounded in predictive coding ideas about how sentence-level meaning is constructed, and I am particularly interested in linguistic prediction in sentences and discourse. Linguistically I have a wide range of interests, and have worked on or am currently working on everything from picture naming in Dutch to grammatical violations in Zulu, and many other topics in between. Most of this work is from the perspective of what can be learned about cognitive processing by investigating neural oscillations and/or event-related potentials. The current project uses sentence constraint as a proxy for the predictability of a target word in a sentence and semantic anomaly as a proxy for prediction error. A key component of the project is to manipulate the degree to which participants are required to focus on top-down rather than bottom-up information while reading, by visually degrading the target words in the sentence. We hypothesise that this shift will lead to enhanced oscillatory signatures reflecting top-down information, and possibly larger prediction errors for anomalous target words.