The “Meddling-In” of Affective Information: A General Model of Automatic Evaluation Effects

This is a formatted version of the final manuscript. Although it is almost identical to the printed version, it is not a copy of it! In a recent text on visual cognition, Houghton, Tipper, Weaver, and Shore (1996) illustrated the basic problem of selective attention by inviting their readers to imagine a fictitious peanut-eater, whose only aim is to eat peanuts. Equipped with a sensory apparatus to locate peanuts in his visual field and an effector to grasp one peanut at a time, the problem arises how this organism will behave in the case of two peanuts placed in front of its sensory system. What mechanisms will prevent the peanut-eater from endlessly grasping into the emptiness between the two peanuts, thereby starving to death? An adequate cognitive apparatus must be tuned to process objects in its visual field in an asymmetrical manner, that is, to respond efficiently to target objects while ignoring distractor objects (see also Neumann, 1987). But what is the target and what is the distractor? Because one peanut is just as good as the other, higher order goals (like the wish to eat peanuts) do not sufficiently determine which of the objects is to be considered as the target of the next action, thereby defining other objects as distractors. In this regard, the peanut-eater simile reminds one of the old allegory of Buridan's ass who starved to death in the midst of two bundles of hay because of its inability to come to a decision about which bundle it should eat. In contrast to the donkey, we as humans are equipped with the ability to decide; it seems as if we are not only able to pursue top-down higher order goals that determine target objects (like bundles of hay) but are also outfitted with a bottom-up sensitivity for differences in valence between potential targets.