Can prediction error explain predictability effects on the N1 during picture-word verification?

Area of Science:

• Cognitive Neuroscience
• Psycholinguistics
• Computational Neuroscience

Background:

• Predictability effects in visual word recognition are often observed in early electrophysiological responses, specifically the N1 component of event-related potentials (ERPs).
• Existing research suggests predicted words elicit lower amplitude N1s, aligning with simple predictive coding theories.
• However, findings on the magnitude and interactions of predictability effects have been inconsistent, with limitations in past study designs.

Purpose of the Study:

• To test whether early effects of predictability in visual word recognition reflect prediction error by examining the interaction between prediction magnitude and certainty.
• To investigate the validity of a simple predictive coding account in explaining the N1 component's response to predictable words.

Main Methods:

• A preregistered study using a picture-word verification paradigm.
• Manipulated the continuous predictability of target nouns (picture-congruent) based on picture-name association norms.
• Recorded event-related potentials (ERPs) from 68 participants during the task.

Main Results:

• Observed a pattern of N1 effects that was opposite to the predictions of a simple predictive coding framework.
• The interaction between prediction magnitude and certainty did not yield results consistent with basic predictive coding models.

Conclusions:

• The findings challenge the straightforward application of simple predictive coding accounts to explain early predictability effects in visual word recognition.
• Suggests that the neural mechanisms underlying word processing may involve more complex predictive processes than previously assumed.