The Nature of Affect in the Structural Mere Exposure Effect

Abstract

This paper investigates the characteristics of the affective component in the structural mere exposure effect (SMEE). Two approaches are considered – fluency attribution approach (FA) and affect as predictive efficiency approach (APE) – within a predictive coding framework. Using the artificial grammar learning and affective priming paradigms, we demonstrate that a violation of implicitly learned regularities elicits an automatic negative affective response. This result suggests that SMEE can be observed without any overtly evaluative judgment. Participants’ decisions on the grammaticality of stimuli did not change this pattern. We conclude that SMEE is based on the affective response to prediction errors made by the cognitive system and may include the fluency attribution process in the later stages of processing.