An ERP index of real-time error correction within a noisy-channel framework of human communication

Abstract

Recent evidence suggests that language processing is well-adapted to noise in the input (e.g., speech errors, mishearing) and readily corrects the input via rational inference over possible intended sentences and probable noise corruptions. However, it remains unclear whether this inference takes the form of an offline re-analysis or a rapid, real-time correction to the representations of the input. We hypothesize that noise inferences happen online during processing and that well-studied ERP components may serve as a useful index of this process. In particular, a reduced N400 effect and increased P600 effect appear to accompany sentences where the probability that the message was corrupted by noise exceeds the probability that it was produced intentionally and perceived accurately. Indeed, semantic violations that are attributable to noise--for example, in "The storyteller could turn any incident into an amusing antidote", where the implausible word "antidote" is orthographically and phonologically close to the intended "anecdote"--elicit a reduced N400 effect and larger P600 effect. Further, the magnitude of this P600 effect is shown to relate to the probability that the comprehender will retrieve a plausible alternative. This work thus adds to the growing body of literature that suggests that many aspects of language processing are well-adapted to noise in the input and opens the door to electrophysiologic investigations of these processes