Looking forward does not mean forgetting about the past: ERP evidence for the interplay of predictive coding and interference during language processing

Abstract

Interference and prediction have independently been identified as crucial influencing factors during language processing. However, their interaction remains severely underinvestigated. Furthermore, the neurobiological basis of cue-based retrieval and retrieval interference during language processing remains insufficiently understood. Here, we present an ERP experiment that systematically examined the interaction of interference and prediction during language processing. We used the neurobiologically well-established predictive coding framework and insights regarding the neuronal mechanisms of memory for the theoretical framing of our study. German sentence pairs were presented word-by-word, with an article in the second sentence constituting the critical word. We analyzed mean single trial EEG activity in the N400 time window and found an interaction between interference and prediction (measured by cloze probability). Under high predictability, no interference effects were observable. Under the predictive coding account, highly predictable input is totally explained by top-down activity. Therefore the input induces no retrieval operations which could be influenced by interference. In contrast, under low predictability, conditions with high interference or with a close, low-interference distractor showed a broadly distributed negativity compared to conditions with a distant, low-interference distractor. We interpret this result as showing that when unpredicted input induces model updating, this may elicit memory retrieval including the evaluation of distractor items, thus leading to interference effects. We conclude that interference should be included in predictive coding-based accounts of language because prediction errors can trigger retrieval operations and, therefore, induce interference.