[Structure and intensity of surprise]

Two experiments investigated the multicomponent theory of emotion intensity proposed by Frijda, Ortony, Sonnemans, and Clore (1992) in the case of surprise and compared it with the more traditional "feeling element" approach to emotion intensity. In both experiments the participants performed a choice reaction time task for a certain number of trials. In the last trial they were surprised by an unexpected change of appearance of the stimuli. In Experiment 1 the change concerned either an action-relevant or an action-irrelevant, affectively neutral stimulus. In Experiment 2 the change concerned either an affectively positive or an affectively negative stimulus. Response delay in the surprise trial served as the index of the interruption of ongoing processes caused by surprise. Immediately after the surprise trial the participants completed a surprise intensity questionnaire that was an adaptation of the general emotion intensity questionnaire used by Sonnemans and Frijda (1994). In contrast to the findings of Sonnemans and Frijda (1994), the present findings were consistently better in line with the traditional feeling element approach to emotion intensity than with the multicomponent theory. Possible reasons for the failure to support the multicomponent theory of emotion intensity for the case of surprise are considered.