Academic journal article Psychonomic Bulletin & Review

Can You Experience 'Top-Down' Effects on Perception?: The Case of Race Categories and Perceived Lightness

Academic journal article Psychonomic Bulletin & Review

Can You Experience 'Top-Down' Effects on Perception?: The Case of Race Categories and Perceived Lightness

Article excerpt

Published online: 18 December 2014

© Psychonomic Society, Inc. 2014

Abstract A recent surge of research has revived the notion that higher-level cognitive states such as beliefs, desires, and categorical knowledge can directly change what we see. The force of such claims, however, has been undercut by an absence of visually apparent demonstrations of the form so often appealed to in vision science: such effects may be revealed by statistical analyses of observers' responses, but you cannot literally experience the alleged top-down effects yourself. A singular exception is an influential report that racial categorization alters the perceived lightness of faces, a claim that was bolstered by a striking visual demonstration that Black faces appear darker than White faces, even when matched for mean luminance. Here, we show that this visually compelling difference is explicable in terms of purely low-level factors. Observers who viewed heavily blurred versions of the original Black and White faces still judged the Black face to be darker and the White face to be lighter even when these observers could not perceive the races of the faces, and even when they explicitly judged the faces to be of the same race. We conclude that the best subjectively appreciable evidence for top-down influences on perception does not reflect a genuinely top-down effect after all: instead, such effects arise from more familiar (if subtle) bottom-up factors within visual processing.

Keywords Lightness perception · Race perception · Modularity · Top-down effects · Cognitive penetrability


In contrast to the traditional "modular" view of perception, according to which the unconscious inferences that determine what we see are driven largely or only by the patterns of light striking the eyes (Fodor, 1983;Pylyshyn,1999), a surge of recent research has suggested that higher-level states such as intentions, desires, social attitudes, and categorical knowledge can directly affect what we see. For example, it has been reported that desirable objects such as money or chocolate appear closer than do neutral or undesirable objects (Balcetis &Dunning,2010); that wearing a heavy backpack makes hills look steeper (Bhalla & Proffitt, 1999); and that reflecting on an unethical deed makes the world look darker (Banerjee, Chatterjee, & Sinha, 2012). Such empirical reports, among hundreds of others, have revived claims (previously popular during the "New Look" movement from the middle of the last century) that perception is a 'constructive' process that consults the rest of the mind for input in delivering a visual percept of the environment (for recent reviews, see Collins & Olson, 2014; Dunning & Balcetis, 2013;Lupyan,2012; Proffitt, 2006; Vetter & Newen, 2014; for philosophical commentary, see Raftopoulos & Zeimbekis, 2014).

Top-down demonstrations?

These studies advertise their results as effects on perception per se; the relevant manipulations are said to literally alter our visual experiences. However, an awkward fact about these claims is that nearly all such examples fail to be subjectively noticeable-a fact that one can appreciate for oneself in everyday life, given the general nature of such claims. For example, if you put a $1 bill next to a $100 bill, the latter does not appear closer; if you stand at the base of a hill and put on a heavy backpack, the hill does not look noticeably steeper; and if you think now of a past transgression, the room does not seem noticeably dimmer. Of course, such observations do not by themselves rule out genuinely perceptual interpretations of such phenomena, given how much visual processing occurs unconsciously. But the absence of subjectively appreciable evidence is, at the very least, unusual within the broader context of vision science-where data and experiments about what we see are routinely accompanied by "demonstrations" in which interested observers can experience the relevant phenomena for themselves in often-dramatic fashion. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.