The primacy of affect: Evidence and extension.

Abstract

Zajonc's (1980) primacy of affect hypothesis asserts that the emotional or affective qualities of stimuli can be processed more readily than their cognitive attributes. The present work sought to provide a direct test of this hypothesis by comparing the effects of affective and cognitive priming under extremely brief (suboptimal) and longer (optimal) exposure durations. In a series of studies, neutral stimuli (Chinese ideographs) were primed by either affective (smiling vs. angry faces) or cognitive stimuli (small vs. large shapes, symmetric vs. asymmetric shapes, and male vs. female faces) at either suboptimal (4 msec) or optimal (1000 msec) exposure durations. At suboptimal exposures only affective primes produced significant shifts in subjects' judgments of the ideographs. At optimal exposure durations this pattern of results was reversed such that only cognitive primes involving size, symmetry and gender produced significant shifts in judgments. Taken altogether, these results support Zajonc's assertion that affective reactions can occur with minimal stimulation and can therefore precede and alter subsequent cognitions. In an extension of the primacy of affect hypothesis, a final study was conducted to explore our ability to nonconsciously differentiate among discrete emotions. In a forced-choice discrimination task between a suboptimal prime and foil, only the positive emotion of happiness was differentiated at a level greater than chance from the negative emotions of anger, fear, sadness and disgust. Subjects were unable to differentiate between any of these negative emotions. Interestingly, surprise, which many researchers refuse to consider an emotion precisely because it is not affectively valenced, could not be reliably discriminated from either the positive or negative emotions. In sum, these data indicate that at the nonconscious level only gross positive and negative affective reactions, as opposed to more fine-grained emotional discriminations, are possible. It appears that while primitive positive and negative affective reactions may occur relatively early in the information-processing chain, some sort of cognitive elaboration may be required to define precisely which specific emotion will ultimately emerge. The failure to find greater emotional specificity at the nonconscious level underscores the need for a theoretical distinction between valenced affective reactions and discrete emotions.