Do Facial Movements Express Emotions or Communicate Motives?Brian ParkinsonDepartment of Experimental PsychologyOxford University, United KingdomThis article addresses the debate between emotion-expression and motive-communica-tion approaches to facial movements, focusing on Ekman’s (1972) and Fridlund’s(1994) contrasting models and their historical antecedents. Available evidence sug-geststhatthepresence of others either reducesor increases facial responses,dependingon the quality and strength of the emotional manipulation and on the nature of the rela-tionship between interactants. Although both display rules and social motives provideviable explanations of audience “inhibition” effects, some audience facilitation effectsare less easily accommodated within an emotion-expression perspective. In particular,emotion is not a sufficient condition for a corresponding “expression,” even discount-ing explicit regulation, and, apparently, “spontaneous” facial movements may be facil-itated by the presence of others. Further, there is no direct evidence that any particularfacial movement provides an unambiguous expression of a specific emotion. However,information communicated by facial movements is not necessarily extrinsicto emotion.Facial movements not only transmit emotion-relevant information but also contributeto ongoing processes of emotional action in accordance with pragmatic theories.What’s in a smile, a scowl, a grimace? People attacha variety of meanings to faces. Some of their attribu-tions accurately reflect what another person is actuallyfeeling, thinking, or doing, whereas other inferencesare of more dubious validity. For example, we are ca-pable of detecting quite precisely the direction ofsomeone’s visual attention (e.g., Kobayashi &Kohshima, 1997; Rutter, 1987) and register at an im-plicit level how closely attuned their facial movementsare to those of others, including ourselves (e.g.,Bernieri, Davis, Rosenthal, & Knee, 1994; Bernieri,Reznick, & Rosenthal, 1988). However, many of ourconclusions about a person’s character, attitude, or in-tentions based on first impressions of their facial fea-tures later turn out to be false. Further, although we be-lieve that seeing someone’s face helps us determinewhether they are telling the truth, in fact most peopleare relatively poor at detecting deception from nonver-bal cues, including facial movements (e.g.,Zuckerman, DePaulo, & Rosenthal, 1981). How much,then, do we read into faces and how much do we readout from them (Russell, 1997)? In particular, does ourdeeply held conviction that faces express emotions fallinto the former or the latter category? How well can wetell what people are feeling simply by looking at theirfaces?Since the 1990s, the familiar view that faces directlyexpress emotions has come under increasing scrutiny(e.g., Russell, Bachorowski, & Fernández-Dols, 2003;Russell & Fernández-Dols, 1997). Perhaps the most in-fluential alternative is that faces are not surfaces onwhich private affective meanings are somehow madevisible but rather tools for communicating social mo-tives to specificaddressees (Fridlund, 1994). Accordingto this view, the idea of expression as an outpouring ofsomething that was first inside is misleading (see alsoEkman, Friesen, & Ellsworth, 1972, p. 3). Emotion doesnot leak out into the interpersonal world. Instead, inten-tions are shared, transmitted, or coordinated betweenfaces. Thus, my “angry” display serves as a warning toyou to back off in case I attack (Fridlund, 1994), regard-less of what emotion I happen to be experiencing. Fur-ther, experiencing anger does not automatically producean impulse to make this expression, unless I also have anaggressive social motive and unless there is someoneelse around to whom the aggressive message might ap-propriately be transmitted.In this article, I review arguments and evidence re-garding the issue of whether faces express emotionsor communicate motives and intentions. I draw threebasic conclusions. The first is that many supposedlyemotional facial movements cannot be explained sim-ply in terms of emotions and display rules. The sec-Personality and Social Psychology Review2005, Vol. 9, No. 4, 278–311Copyright © 2005 byLawrence Erlbaum Associates, Inc.278A few passages of this article are revised and extended from ma-terial included in “Emotion in social relations: Cultural, group, andinterpersonal processes” (2005, Psychology Press). I am grateful toAlan Fridlund, Tony Manstead, and an anonymous reviewer for con-structive comments on a previous version.Requests for reprints should be sent to Brian Parkinson, Depart-ment of Experimental Psychology, Oxford University, South ParksRoad, Oxford OX1 3UD, UK. E-mail: [email protected] Not Copyond is that some apparent advantages of the alterna-tive motive-communication account are partlyundermined by its underspecification of the centralconcepts of social motive and audience attunement.Third and finally, a fuller articulation of dynamicemotion processes in their interpersonal context mayrender many (but not all) of the distinctions betweenemotions and social motives redundant. My centralaim is to point the way toward a more comprehensiveaccount of facial movement in social interaction thatproperly situates emotional communication in its ev-eryday functional context.This article is structured in five main sections. In thefirst historical section, I discuss Darwin’s (1872/1998)theory as a precursor to the contemporary emotion-ex-pression view and Dewey (1894) and Mead’s (1934)pragmatist accounts as forerunners of the motive-com-municative approach. Second, I present Ekman’s(1972) and Fridlund’s (1994) competing contemporarytheories in more detail. Third, I provide a brief reviewof the various sources of evidence that emotions andexpressions are directly linked and conclude that theyprovide clear support only for a probabilistic connec-tion. In the fourth and central section, I evaluate empir-ical literature addressing how the presence of otherpeople influences facial movements, focusing on evi-dence for audience inhibition and facilitation. This evi-dence is more readily accommodated within a mo-tive-communication paradigm, but this is partlybecause of the flexibility of some of its concepts. Fifth,I reevaluate the theories on conceptual as well as em-pirical grounds, drawing on the ideas and findings pre-sented earlier.Early TheoriesCharles