So Cloze yet so Far: N400 Amplitude is Better Predicted by Distributional Information than Human Predictability Judgements

  title={So Cloze yet so Far: N400 Amplitude is Better Predicted by Distributional Information than Human Predictability Judgements},
  author={James A. Michaelov and Seana Coulson and Benjamin K. Bergen},
—More predictable words are easier to process—they are read faster and elicit smaller neural signals associated with processing difficulty, most notably, the N400 component of the event-related brain potential. Thus, it has been argued that prediction of upcoming words is a key component of language comprehension, and that studying the amplitude of the N400 is a valuable way to investigate the predictions we make. In this study, we investigate whether the linguistic predictions of computational… 
1 Citations

Figures and Tables from this paper

Distributional Semantics Still Can’t Account for Affordances
Can we know a word by the company it keeps? Aspects of meaning that concern physical interactions might be partic-ularly difficult to learn from language alone. Glenberg and Robertson (2000) found


Prediction during language comprehension: benefits, costs, and ERP components.
  • C. Van Petten, B. Luka
  • Psychology
    International journal of psychophysiology : official journal of the International Organization of Psychophysiology
  • 2012
Dissociable effects of prediction and integration during language comprehension: Evidence from a large-scale study using brain potentials
The results challenge the view that semantic facilitation of predictable words reflects the effects of either prediction or integration, and suggest that facilitation arises from a cascade of processes that activate and integrate word meaning with context into a sentence-level meaning.
Great expectations: Specific lexical anticipation influences the processing of spoken language
When listening to a constraining discourse people rapidly make very specific predictions about the remainder of the story, as the story unfolds, which are not simply based on word-based automatic activation, but take into account the actual message of the discourse.
Word predictability effects are linear, not logarithmic: Implications for probabilistic models of sentence comprehension.
Thirty years and counting: finding meaning in the N400 component of the event-related brain potential (ERP).
The effectiveness of the N400 as a dependent variable for examining almost every aspect of language processing is emphasized and its expanding use to probe semantic memory is highlighted to determine how the neurocognitive system dynamically and flexibly uses bottom-up and top-down information to make sense of the world.
CHAPTER 15 A Look around at What Lies Ahead: Prediction and Predictability in Language Processing
It is argued for the importance of investigating such linguistic prediction as yet another example of a neural system in which probability estimation is inherent, with a proposal to move beyond the debate of whether there is linguistic prediction, toward focusing research on how pre-activation may occur and what is pre-activated.
Brain potentials during reading reflect word expectancy and semantic association
The amplitude of the N400 component of the was found to be an inverse function of the subject's expectancy for the terminal word as measured by its ‘Cloze probability’, which suggests N400 may reflect processes of semantic priming or activation.
Limits on lexical prediction during reading
An exploratory data analysis of word form prediction during word-by-word reading
To better understand the empirical justification for positing an association between prenominal article predictability and scalp potentials, a wide-ranging exploratory data analysis (EDA) approach is conducted, pooling the original data with extant data from two follow-up studies.