Cognitive and Affective Responses to Expectancy Violation

All brains are, in essence, anticipation machines. -- D. C. Dennett, 1991

A fundamental aspect of human cognition is the necessity of forming expectations, based on direct or indirect experience, for how things are in the world. People form expectations about virtually everything, including how other people behave. These “expectancies” form the basis of our impressions of individual people and also the stereotypes that we hold about groups of people. Social psychologists have long been interested in how people use their expectancies to guide their social interactions, thoughts, and emotional responses concerning other people. One key interest in our lab is in studying the cognitive and affective responses elicited in people when others behave in ways that violate or contradict initial expectancies.

To study these issues, we often use a combination of measures to study cognitive and affective responses. Measures of cognitive response often include event-related brain potentials (ERPs), which are a measure of the electrical activity of the brain associated with cognitive processing of discrete stimulus events, in addition to memory (recall) measures. To measure affective responses, we have used facial electromyography, which involves placing electrodes over specific facial muscle regions to detect tiny movements in those muscles that indicate positive or negative affective responses (e.g., smiling or frowning). In one such study (Bartholow, Fabiani, Gratton, & Bettencourt, 2001), we found that expectancy-violating behaviors elicited a larger amplitude P300 component of the ERP, which is thought to reflect (among other things) online updating of working memory templates. We also found that negative (but not positive) expectancy violations elicited an immediate, negative affective reaction in the EMG (activation of the corrugator supercilii, which draws the brows together). These findings were important in a number of respects, but perhaps primarily because they showed that not all expectancy violations elicit negative affect, which had been proposed by a number of theorists (e.g., Mandler, 1990Olson, Roese, & Zanna, 1996).