Working on a Smile: Responding to Sexual Provocation in the Workplace

Working on a Smile: Responding to Sexual Provocation in the Workplace

Scan any social situation involving both sexes and you will likely see differences in how women and men communicate both verbally and nonverbally. One nonverbal behavior concerns the fact that women tend to smile more than men. Indeed, numerous studies done over several years provide strong support for the finding that women smile more than men irrespective of whether the measure is frequency, dura­tion, or even the size or kind of smile (Hall, 1984; Hecht & LaFrance, 1998; Henley, 1977; LaFrance, Hecht, & Paluck, 2003).

A recent meta-analysis showed, however, that the size of the sex dif­ference in smiling is contingent on several factors (LaFrance, Hecht, & Paluck, 2003). In other words, women do not always smile more than men do. For example, women and men smile in comparable amounts when both sexes believe they are not being observed but women smile more than men when everyone is conscious of being evaluated. The sexes also smile in comparable amounts when they are in the same sit­uation, role, or occupation. Both these contexts suggest that there are social expectations for women to smile and for men not to (LaFrance & Hecht, 1999). In fact, smiling women are evaluated more positively than those who do not smile (Deutsch, Le Baron & Fryer, 1987). Women smile more than men when the emotional climate is tense or negative than when the emotional milieu is comfortable and positive. That women smile more when the context is strained may spring from feeling greater obligation to try and do something to set it right.

While such findings suggest when greater smiling by women is to be expected, they do not tell us much about how people perceive a smiling woman. Can people accurately differentiate smiles that indicate plea-sure or amusement from those that do not? Do observers consider the possibility that women sometimes smile because they are expected to do so? Do they entertain the hypothesis that a woman may be smiling not because she is feeling pleased but precisely because she is feeling tense and uncomfortable?

The present chapter describes a series of studies aimed at deter-mining whether observers are able to tell the difference between smiles by women that spring from honestly felt pleasurable feelings from smiles that indicate displeasure. We also explore whether women and men are equally able to detect differences in types of smiles. Might it be the case that men are more inclined to take smiles by women on face value? Some early data indicate that men may misinterpret friendly behavior by women. Specifically, Abbey (1982) found that male ob­servers were significantly more likely than female observers to inter­pret friendly behavior by women in sexual terms. We also wanted to explore the possibility that there are individual differences among men in the inclination to misinterpret a women's smile, to overlook cues in dicating that the smile reflects not pleasure but pain. In other words, are some men in some situations more inclined to misinterpret women's smiles?

In order to address this question it is necessary to show that smiles associated with positive emotion can be reliably distinguished from smiles that are not related to enjoyment or pleasure. Research has shown this to be the case: there are several kinds of smiles, only one of which actually reflects positive emotion (Ekman & Friesen, 1982). Specifically, Ekman and Friesen (1982) noted that variants of human smiles could be distinguished on the basis of a number of behavioral markers (i.e., morphology, intensity, timing, location, and laterality). Felt (or Duchenne) smiles have been distinguished from 'false' (or non-Duchenne) smiles on the basis of the presence of orbicularis oculi activation, as marked by crow's feet wrinkles at the outer eye corners (Ekman & Friesen, 1982). Several studies have obtained evidence that Duchenne smiles occur more often than other types of smiling when adult participants watched pleasant films or when they self-reported amusement during both solitary and social situations (Ekman & Friesen, 1982; Frank, Ekman, & Friesen, 1993). Non-Duchenne smile types on the other hand appear to be a kind of social lubricant, adopted when social context calls for signs of accommodation or coop­erativeness.

We began the investigation by asking how women respond when they are asked sexually provocative questions by a male interviewer during the course of a job interview. In particular, we were interested in knowing how women look when they are on the receiving end of in appropriate queries. On the one hand, by virtue of applying for a job,

female applicants need to come across in the best possible light. On the other hand, by virtue of being the recipient of sexually loaded questions, they may need to find some way to convey disinterest in that particular line of questioning. It seemed possible that they might smile but that it would not be a smile of delight. We found that to be the case.

Next, we examined whether men and women are equally able to dif­ferentiate Use or social smiles from pleasurable ones. Then, we looked more closely at individual differences among male observers. Specifically, we explored differences among men in how they interpret smiling by women in two different contexts. Before describing the spe­cific studies, we first take a look at the larger issue of why women might smile at all when the context is unpleasant.

SEX, POWER, AND NONVERBAL BEHAVIOR

A quarter of a century ago, several communication researchers pro-posed a set of ideas linking gender, expressiveness, and power. Nancy Henley (1977) argued that men and women display different nonver­bal behaviors. Women were described as smiling more, engaging in more eye contact, and displaying greater sensitivity cues from others than men. Robin Lakoff (1973) similarly observed that women used language differently than men, exhibiting more verbal hedges, tag questions, and super polite linguistic forms. Many studies confirmed the existence of sex differences in several verbal and nonverbal modali­ties (Baird, 1976; Burgoon, Buller, & Woodall, 1989; Frieze, Parsons, Johnson, Ruble, & Zellman 1978; Hall, 1978; Haviland, 1977; LaFrance, Hecht, & Paluck, 2003). There is debate however about how these differences are to be explained. Henley (1977) argued that they are due to the fact that women typically have less power and status than do men. Women were said to be more expressive because they have less power just as subordinates are more expressive because they have less power than their superiors. Subordinates need to demon­strate that they know their place and nonverbal cues are a particularly effective way of conveying this (Ellyson & Dovidio, 1985). For example, it pays for subordinates to signal deference as well as being especially attentive to cues coming from a superior. Lakoff (1973) too character­ized women's speech as powerless and argued that women use power-less linguistic forms such as verbal hedges to signal their place and to avoid giving offense. The shared contention is that survival may very well depend on to the ability of people with low power to respond in nonverbally sensitive and deferential ways (Bugental, Shennum, Frank, & Ekman, 2001). In short, sex and power are viewed as inextri­cably linked and communication is viewed as an important mecha­nism that keeps the link in working order.

A power explanation for sex differences in nonverbal behavior has received some, but not unanimous, support. For example, Bugental

and her colleagues (2001) noted that physically abused children ap­pear to be especially sensitive to nonverbal cues presumably because they need to be able to predict their abuser's moods and actions. With respect to smiling behavior, both Denmark (1977) and Deutsch (1990) found that having lower power was associated with showing more smiling. But other researchers have been unable to replicate a direct relationship between having less power and showing a greater ten­dency to smile. Some have found no relationship (Dovidio, Brown, Heltman, Ellyson, & Keating, 1988; Hall & Payne, 1995) while others report that high-power people sometimes actually smile more than low-power people (Halberstadt, Dovidio & Davidson, 1988).

Some recent findings provide a possible resolution to the inconsis­tent findings. Specifically, Hecht & Lance (1998) manipulated power in same-sex dyads and measured how much participants smiled. A number of potential mediators were also measured. Specifi cally, participants rated how positive they felt and how much they felt the need to please the other person during the interaction. Although we found no overall differences in how much low and high power people smiled, we did find that these two groups smiled for very different rea sons. For those with low power, how much they smiled was positively associated with how much they experienced the need to please the per-son with more power. In contrast, for those with more power, their smiling was significantly correlated with how positively they felt. In sum, when those with more power smiled it was related to their feeling good, but those with less power smiled because they felt the obligation to please. Moreover, women felt more obligated to please than did men regardless of their level of assigned power. Even when women had more power, the degree to which they smiled was positively correlated with needing to please. This association was not found among high-power men (Hecht & LaFrance, 1998). That women experience the ob­ligation to smile more because of the diffuse lower status assigned to women is compatible with expectations states theory (Ridgeway & Smith-Lon, 1999). Communicative behaviors may thus serve as one mechanism by which power inequities between the sexes are reflected and maintained.

SEX, POWER, AND THE WORK ENVIRONMENT

Employment contexts represent fertile ground for observing the man­agement of power via nonverbal behavior. To begin with, evaluation of work performance is strongly affected by how people comport them-selves nonverbally (Imada & Hakel, 1977; Spangler, 1995; Washburn & Hakel, 1973). In addition, relationships between workers and supe riors are frequently established, negotiated, and maintained via the ac­tion of nonverbal behaviors. Subordinates signal their deference and superiors signal their authority and command. However, sometimes

these relationships are disrupted when the power of the superior is used for inappropriate purposes.

One distressing misuse of power in a work context is harassment targeted at female workers. According to one estimate, between 42% and 90% of women have experienced sexual harassment (Baker, Terpstra, & Larntz, 1990) and the consensus among investigators is that approximately 50% of women have experienced sexual harass­ment of one type or another (e.g., Fitzgerald & Ormerod, 1993; Schnei­der, Swan, & Fitzgerald, 1997). Researchers have characterized sexual harassment along several dimensions including type, severity, and fre­quency. Legal scholars have taken these dimensions into account in defining two major types of harassment specifically quid pro quo and hostile environment harassment. The former is based on sexual coer­cion, namely on demands for sexual favors in return for job-related benefits or escape from retaliation and may be based on a single inci­dent (EEOC, 1980). In contrast, hostile environment harassment characterizes ongoing contexts where workers are subject to intimi­dating, hostile, or offensive work environments because of their sex (Meritor Savings Bank v. Vinson, 1986). Hostile environment harass­ment usually occurs over a period of time and incorporates behaviors such as sexist jokes and remarks, offensive language and graffiti, re-quests for sexual favors or dates, nude pinups, and sexual a-mails (Conte, 1997).

Some cases of sexual harassment are so egregious that that there is little doubt that women have been subjected to sexual coer­cion or longstanding gender-based hostility (Fitzgerald & Ormerod, 1993). In both cases, sexual harassment results in nega­tive consequences for both the individual and the organization. In­dividuals report that their emotional lives have been negatively impacted (Gutek & Koss, 1993; U.S. Merit Systems Protection Board, 1981). In fact, heightened arousal and intrusive flashbacks observed in the aftermath of severe harassment are similar to symptoms associated with post-traumatic stress disorder (Dansky & Kilpatrick, 1997; Gutek & Koss, 1993). Besides psychological repercussions, severe harassment produces decreased job satis­faction (Fitzgerald, Drasgow, Hulin, Gefland, & Magley, 1997; Gutek & Koss, 1993; Piotrkowski, 1998) and reduced job perfor­mance (Crull, 1982; Stockdale, 1998).

Although milder forms of harassment may be insufficiently seri­ous to warrant legal sanctions, it is likely that they can still produce significant emotional and work repercussions. But very little is known about reactions to and effects of harassment that falls below the legal radar screen. While most research on serious sexual harass­ment has focused on its prevalence and its long-term effects (e.g., Livingston, 1982; Stockdale, 1998), researchers have largely ignored the immediate emotional and job-related effects of milder forms of sexual provocation. In the next section, we describe a study aimed at

determining how women manage this kind of untoward behavior by superiors in an employment context.

SMILING IN RESPONSE TO SEXUALLY PROVOCATIVE QUESTIONS

Women sometimes smile when their affect is not especially positive and when their need to please is high. This can happen in both work and home environments. On the job, women are more likely than men to find themselves in a predicament where they are treated as under­lings and as objects of heterosexual attention. Like male underlings they are dependent on supervisors and bosses for rewards and re-sources. Also like males in low power positions they are expected to show a modicum of respect and deference to their superiors. But un­like men they are assumed to be less competent and commanding (Ridgeway & Smith-Lon, 1999). Also unlike men, women in less powerful positions have the additional task of not giving the 'wrong message' to male bosses who may express more than workday interest in them. How do women respond to the double bind that these situa tions present (Rudman & Glick, 2001)? Specifically, we examined how female job applicants respond to a male who asks them sexually pro-vocative questions during the course of the job interview.

We decided that in order to study immediate reactions to subtle ha rassment, we would need to create an experimental situation in which research participants would be exposed to realistic harassment and contrast it with an otherwise identical but not sexually provocative sit uation. To make the situation as realistic as possible, we needed to cre ate a work context in which participants would not be aware that they were in a study. See Woodzicka and LaFrance (2001) for a discussion of the need for this since role-played or imagined reactions to harassment differ substantially from real reactions.

To create a realistic but controlled context, we devised a job inter-view in which female job applicants were assigned to either an experi mental interview or a controlled interview. In the former, participants were asked a few sexually provocative questions interspersed with more typical interview questions. In the control condition, partici­pants were asked a matched set of surprising but not sexually-provoc ative questions. Interviewees were recruited by campus posters and by advertisements placed in campus and local newspapers. Fifty women (ages 18-39) signed up to participate in the interview that was de-scribed as determining eligibility for a job as a research assistant. Par­ticipants were told that if they performed well in the interview, they would be placed on a list made available for faculty seeking to hire re-search assistants. In reality, all participants, regardless of their performance, were placed on the list.

A male interviewer asked female job applicants three experimental or control questions interspersed with other interview questions. In the experimental condition, the questions were marked with sexual in-

nuendo; in the control condition there was a set of matched questions minus the sexual allusion. The experimental and control questions were pre-tested to be comparable with respect to their 'surprise' quo­tient. All questions, whether in the 'harassment' condition or in the matched control condition, were rather surprising queries to hear in the context of a job interview. The only difference between conditions was in the amount of sexual content they contained. Specifically, dur ing the course of a sixteen question interview, participants were asked three sexually provocative questions (such as 'Do you have a boy-friend?') or'matched non-sexual questions (such as 'Do you have a best friend?'). Besides the experimental and control questions, the job interviews were handled identically. All participants were unobtru­sively videotaped during the interview.

After the interview, a female research assistant escorted partici­pants to another room where they completed a self-report affect scale and several questions designed to elicit reactions to the interview and interviewer. Participants were then fully debriefed, given an opportu nity to withdraw their self-report and video data from the study, and paid for their time. One participant in the experimental condition asked that her videotaped interview be erased. This was done immedi ately. Each participant was also given the names and phone numbers of the psychology department and human-subjects committee chair-persons in the event that she wanted to discuss the study further. Neither of the chairpersons was contacted.

From the videotapes, we coded facial expressions using the Facial Action Coding System (FACS; Ekman & Friesen, 1978). The FACS is a comprehensive system that describes all possible visible facial muscle movement. We coded frequency, duration, onset, and offset for all Ac tion Units (AUs) for a total of 6 seconds, beginning one second prior to the end of each 'experimental' target question. This period was cho­sen to capture most facial expression associated with a particular question.

We were particularly interested in determining whether interview­ees smiled during the interviews. If so, we wanted to know how much of it was Duchenne and non-Duchenne smiling. Duchenne or felt smiles involve two specific facial action units. One action entails the lip corners, which are pulled up and back; the second facial action in felt smiles is observed by the cheeks being raised causing wrinkles to form at the outer corners of the eyes. Non-Duchenne smiles, on the other hand, only involve mouth movement and show no reliable relationship to positive feelings (Ekman, Friesen, & Ancoli, 1980).

We found that participants in the sexually provocative condition smiled more than those in the non-sexualized interview. But the type of smile clearly indicated that enjoyment was not the underlying emo­tion. In response to the sexually provocative questions, nearly all (96%) of the female job interviewees displayed social, that is non-Du chenne, smiling. And in a specific comparison of the experimental and

control groups, women asked the provocative questions showed sig­nificantly more non-Duchenne than did women asked non-provoca­tive questions, t(45) = 2.02, p < .05.

The results clearly showed that the distinction between Duchenne and non-Duchenne smiles was a meaningful one-these two smile types were negatively correlated with each other (r = -.35). In other words, as felt smiling decreased, false or social smiling increased. Moreover, there is clear evidence that smiling in response to sexual provocation is not to be taken as evidence that the women were pleased to be asked these provocative questions. We found that amount of non-Duchenne smiling was positively correlated with the degree to which participants reported feeling angry, upset, disgusted, and irrita ble and negatively correlated with feeling strong. In addition, non-Du­chenne smiling was positively correlated with perceiving the interviewer as sexist and perceiving the interview as having been sexu ally harassing. Table 7.1 shows these correlations. In short, the inter­viewees in the harassment condition noted that the interviewer was behaving inappropriately and they were not happy about it.

These findings indicate then that female interviewees recognized their predicament. They needed to appear pleasant and competent even while having to contend with a harassing interviewer. As appli­cants, the women had little power because the interviewer was the gateway to a desired job. As recipients of sexually provocative ques­tions, the job applicants had little recourse but to 'grin and bear it' which is an apt way to characterize a non-Duchenne smile. Such smil­ing indicates not equanimity at the unwelcome questions but an attempt to get through an untenable situation.

Had the non-Duchenne smiling been effective at warding off further unwelcome advances or in establishing oneself as particularly compe‑

Correlations Between Non-Duchenne Smiling and Self-Report Measures
for Interviewees Asked Provocative Questions

tent or professional then it might have been a useful strategy. It did not work that way. We found that non-Duchenne smiling in this context was actually counterproductive. From videotapes of the interviews, an independent group of 50 male and female undergraduates rated the interviewees on several dimensions including how competent and smart they appeared and how likely they would be to hire them. Judges viewed a set of brief, five-second, silent visual clips taken of the interviewees. The more interviewees displayed non-Duchenne smiling the more negatively they were evaluated on competence (r = -.47) and intelligence (r = -.53). As non-Duchenne smiling increased, ratings of her competence and intelligence decreased.

Moreover, verbal performance in the interview was negatively im­pacted. Women who were asked sexually provocative questions were less fluent in their answers than those who were asked equally surpris ing but not sexual, questions, t(46) = 2.35,p < .05. In another evalua­tion of the interviewees, judges from Yale's School of Management rated the answers by harassed women to be of poorer quality than an­swers given by non-harassed interviewees, t(46) = 1.99, p < .055. In sum, sexual harassment reduced job performance as evidenced by both nonverbal behavior and verbal facility.

Faced with sexually loaded questions, the female applicants smiled but it was not the smile of enjoyment. The interviewees clearly per­ceived that these questions were out-of-line and their non-felt smiling reflected this. Such smiles appear to have been a way to deal with the unwelcome intrusion rather than a sign of assent. But how do men in­terpret such smiling? There is the possibility that false smiles, dis­played in an attempt to appease, may be misinterpreted as indication of interest and enjoyment. Next, we describe two studies that directly address this issue.

INTERPRETING WOMEN'S SMILES

How do men interpret the smiling shown by women in response to in-appropriate sexual questions in a job interview? Is it accurately per­ceived as a social smile or is it instead misinterpreted as a sign that they find the questions pleasing and the attention welcome? We first examined whether men are less likely than women to correctly identify Duchenne and non-Duchenne smiles. Then, we looked at whether there are individual differences in how men perceive women who show non-Duchenne smiles.

Sex Differences in Identifying Duchenne and Non-Duchenne Smiles

Research strongly supports women's superior nonverbal sensitivity (Hall, 1978; Lieberman, Rigo, & Campain, 1988; Rotter & Rotter, 1988; Zuckerman, Hall, DeFrank, & Rosenthal, 1976). Across a range of emotional displays, women are more accurate decoders of facial ex-

pressions than are men (Hall, 1984). Consequently we might expect women to be more accurate in distinguishing felt from social smiles.

Thirty male and 30 female participants viewed 32 silent video clips taken of the female participants from the job interview study described above. The clips (presented in two orders) varied from one to three sec­onds in duration and consisted of eight Duchenne smiles and 24 non-Duchenne smiles. After viewing each clip, participants were asked to indicate whether the smile was genuine/real or fake/false.

Results showed that men were significantly less accurate in decod­ing non-Duchenne smiles than were women. Specifically, men were less likely than women to correctly label false smiles t(58) = 2.03, p < .05. Male participants underestimated the number of non-Duchenne smiles and overestimated the occurrence of Duchenne or felt smiles. In sum, men were less accurate decoders of different smile types shown by women and the errors were in the direction of seeing any smile by a woman as indicating genuine positive feeling. This finding corroborates previous research that found that men mistakenly inter­pret women's friendliness as indication of sexual interest (Abbey 1982; Abbey & Melby, 1986).

Likelihood to Sexually Harass and Ratings of Smiling and Non-smiling Women

In our final study, we examined whether individual differences among men might affect how they interpret a woman's smile. On the face of it, it seems likely that men who are inclined to take advantage of work sit­uations for sexual purposes wi11 be more likely to misread apparently positive nonverbal behavior from women. Indeed, research has found that certain men are more likely to engage in forms of harassment than are others (Pryor, 1987).

To capture this individual difference, Pryor (1987) developed an in­strument, the Likelihood to Sexually Harass Scale (LSH) to measure men's propensity to use power to elicit sex from women by threatening punishment or promising rewards. In the LSH scale, male respon­dents are given ten hypothetical scenarios that depict a male who has the power to control an important reward or punishment for a female target. They are asked to imagine themselves in the role of the male in each scenario and to indicate the likelihood that they would behave in a sexually exploitative way if there would be no negative consequences to them as a result of their actions.

In one validation study, or (1987) observed the behavior of men who scored either high or low on the LSH. Specifically, males who were pre-tested on the LSH were randomly assigned to train a female con-federate how to putt a golf ball or how to play poker. The confederate rated the frequency and nature of the participants' touching behavior toward her during the training session. Results revealed that high LSH males touched the confederate more and behaved in more sexual fash-

on towazd her than low LSH males when the training condition pro vided acceptable opportunities to do so (i.e., in the golf but not in the poker training condition).

Since the LSH scale has also been found to predict sexualized be­havior during an interview (Rudman & Borgida, 1995), it seemed likely that high scorers on the LSH would also be more likely to mis read a non-Duchenne smile, especially in a work context. We predicted that men scoring high on LSH would perceive women as happier and more flirtatious than men scoring low on the scale when she displayed non-Duchenne smiling. This was predicted to be especially the case when the smiling is described as occurring during a job interview rather than a casual conversation.

Sixty male participants viewed twelve silent videoclips from the in terview study, six of which showed a woman with a neutral facial ex pression and six showed a woman displaying a non-Duchenne smile. Because physical attractiveness might have affected participant rat­ings, the smiling and non-smiling women were matched on attractive­ness. In addition, we examined the effect of context on participants' ratings. Participants were told that the women they would observe were either 'talking to someone' or 'interviewing for a job.' Partici­pants rated each of the twelve women on a number of state and trait words (e.g., happy, flirtatious, apprehensive).

Male participants were strongly affected by whether the female tar gets showed a social smile versus not smiling at all. Regardless of like lihood to sexually harass and context, males saw women with a false smile as significantly more happy, amused, friendly, and surprised than women who did not smile. They also rated them as more flirta­tious, fake, and less apprehensive than non-smiling women.

It is interesting to note that men's ratings were affected by what they believed she was doing. When they thought she was a job appli­cant, they rated her as more afraid and uncomfortable if she was smiling than if she was not smiling. However, if they believed they were watching her in conversation, they saw her as less afraid and uncomfortable than non-smiling women. Thus, male raters as a group 'know' that smiling in a job interview does not necessarily mean that a person is happy. The critical issue was whether this would be also true of men who show a greater propensity to engage in sexual harassment.

As scores on LSH increased, so did the ratings of the smiling target as flirtatious (r = .28) and desirable (r = .45). Further, we found a sig nificant interaction between LSH and context on ratings of smiling tar-gets, F(1,49) = 4.83, p < .05. When told that the female target was interviewing for a job, men with higher scores on LSH rated her as more flirtatious and desirable than did high LSH men who were told the women were just talking to someone. In contrast, men scoring lower on LSH rated the women as more flirtatious when they believed she was just conversing. When the targets did not smile, scores on LSH

were positively correlated with ratings of them as vulnerable (r = .28), afraid (r = .31), and confused (r = .31).

In sum, men scoring high in LSH saw smiling women as more flirta tious and desirable and non-smiling women as more vulnerable and confused. Smiling as an attempt to get through an uncomfortable situ ation was seen by these high scorers on LSH as flirtatious and perhaps as a 'green light' to move forward. But it also doesn't work not to smile. Those failing to smile run the risk of being judged as afraid, vulnera ble, and confused, not exactly desirable characteristics in a potential employee.

IMPLICATIONS AND FUTURE DIRECTIONS

Social power, sex, and smiling are connected in a complex set of rela tionships. Low power people do not necessarily smile more than those with more power. Rather, in a relationship characterized by asymmet­ric power, people with little power appear to be obligated to display some level of smiling regardless of how positively they feel. The more powerful person, on the other hand, appears to have the option to smile, in the sense of only smiling when he or she feels positively in clined. In other words, it is not necessarily the case that people with low status or power smile more than those with greater power. But when they do smile it springs from the feeling of needing to please the other. Women feel this need even more acutely than men even when they are in positions of greater power.

In a job interview, the need to please an interviewer is clearly im­portant. But women in this situation have an even more complex and difficult interpersonal task if the interviewer seems interested in them as sexual objects rather than merely as job candidates. What our investigations showed was that women who are subtly ha­rassed in a work setting showed more social or fake smiling than those who are not harassed. Social smiling makes sense given the double bind they are in. Unfortunately, it doesn't work. Those who show social smiles are rated more negatively as job applicants. Fur­thermore, some men, namely those who are inclined to sexually ha­rass women, tend to see sexual potential in the smile. They see smiling, especially during a job interview, as indicating flirtatious­ness-perhaps a sign that the woman welcomes the advances. Her agenda is of another order altogether. The social smile is employed as a kind of defense against harm either to her well-being or her chances for employment.

Nonverbal cues like smiling appear to be a significant component of interpersonal relationships marked by differences in power. In many situations, they act as efficient yet unspoken confirmation that every-one knows their place. But as the studies described above demon­strate, nonverbal behavior in such contexts can be conveniently misinterpreted. Superiors can read the cues at the most superficial

level and protest if anyone were to question them saying that they were merely responding to what they saw.

That women sometimes smile when they are uncomfortable and that such smiling is sometimes misinterpreted can combine to create an insidious self-fulfilling prophecy. For example, Word, Zanna, and Cooper (1974) found that self-fulfilling prophecies were mediated by nonverbal behavior in interracial interactions. We found a similar pro­cess in these job interviews. It appears that male interviewers who use their power to harass job applicants can produce interviewee behavior that reinforces sex stereotypes that portray women as less competent and more flirtatious. However, this is just one piece of a larger picture. Sexual harassment is more likely to occur in environments that rein-force gender-based power discrepancies and overlook inappropriate sexual behavior (Pryor, Giedd, & Williams, 1995).

Some men, namely those high in likelihood to sexually harass, are inclined to take advantage of their power, particularly when the work-place overlooks mild sexual harassment. Even high LSH men know that there are serious consequences of quid pro quo harassment. But they may engage in subtle sexual provocation of women that they have some power over. Subtle provocation maybe the harassment of choice because it can be denied or deemed non-serious in the unlikely event that the target reports it. Our findings suggest that women on the re­ceiving end of the subtle harassment display the non-Duchenne smile which unfortunately is not recognized as a false or appeasement smile. Instead, it is interpreted in two equally harmful ways. It may be taken as a sign of incompetence or as a signal that the target welcomes the ha­rassment. Either way the resulting behavior is attributed to undesir­able applicant attributes rather than to the untenable situation interviewers themselves put the applicants in. The attributions dimin­ish women's chances of being hired or promoted and might explain why women tend to receive lower evaluations than men during job in­terviews (Olian, Schwab, & Haberfeld, 1988). Such attributions could also lead to continued sexual harassment.

Perhaps even more pernicious is reinforced views of women as in-competent and coy. Besides individual harm, sex stereotypes maintain the gender status quo. For example, women as a group may be pro­moted and paid less because they come into an interview assumed to be less competent and serious than men. If workplaces are to become nondiscriminatory, seemingly innocuous behaviors that put women in a double bind must be discouraged.

It might be countered that women mislead others when they display false smiles whether they intend to or not. Apparently women should smile only when they feel happy; otherwise they are responsible for what follows. However, there are some problems with this simple stip­ulation. In the first place, research findings indicate that people react rather negatively to non-smiling women. Secondly, advising women to change their behavior misses the point. Instead of taking a close look

at inappropriate supervisor behavior, women are criticized for engag ing in a behavior which is not only understandable given the context, but which would be criticized if it were not displayed. The responsibil­ity falls on the perpetrator, not the victim, to alter his behavior and discontinue the power play.

When it comes to smiling, women are in a double bind, and even more so with men who have a propensity to harass. Men scoring high on LSH view a smiling woman as flirtatious and desirable and an un smiling woman as afraid and vulnerable. None of these characteristics are desirable in the workplace. Even those not likely to harass have ex­pectations regarding women's nonverbal behavior. Women are ex­pected to smile and those who do not are judged harshly (Deutsch, LeBaron, & Fryer, 1987). When a woman displays non-Duchenne smil­ing, she signals that she is non-threatening and ready to please. Though appealing in many interpersonal situations, our results sug­gest that these displays make women look incompetent in a job inter view. It falls to future work to continue to map out the interactions of varying situations, individual differences, and perceptions of women who do and don't smile.

A popular adage over the past years has been 'don't sweat the small stuff.' In the arena of sex-based harassment and employment, the op­posite is probably more advisable. Employers should be encouraged to carefully attend to instances of harassment no matter how trivial they may seem. In other words, they should sweat the small stuff, namely subtle harassment that falls under the legal radar screen. Our research demonstrates that even seemingly harmless sexual provoca­tion has pernicious effects.

If the 'sexual provocation-smiling-misinterpretation cycle' is to be interrupted, training programs must be developed and implemented that go beyond merely being in compliance with the law. First and fore most, managers should be taught that subtle harassment is not just of­fensive but may lead to performance deficits and decreased productivity. Further, severe harassment may grow out of an environ ment that condones seemingly innocuous sexual behaviors (e.g., sex ual jokes, remarks, and questions). Research suggests that women are more likely to experience sex-based harassment in workplaces where men perceive the social norms as permitting such behavior (Pryor, Giedd, & Williams, 1995). Thus, management should be urged to cre ate and support work environments that do not tolerate sexual power plays of any type. Second, training programs should help managers and employees to understand typical and gendered nonverbal re­sponses to displays of power. Specifically, employees should be taught that smiling doesn't always indicate happiness or pleasure. On the contrary, smiling, especially by low-powered individuals, frequently signals unease or discomfort. Management can be taught to use avail-able contextual cues and target characteristics to understand the meaning of and respond appropriately to women's smiling.

ACKNOWLEDGMENTS

We gratefully acknowledge Michelle Moore, Betsy Paluck, Hal Ersner-Hershfield, and Kelly Ganley for their help with these studies.

Correspondence concerning this article should be addressed to Julie A. Woodzicka, Department of Psychology, Washington and Lee Univer­sity, Lexington, VA, 24450, e-mail: woodzickaj@wlu.edu or to Marianne Lance, Department of Psychology, We University, PO. Box 208205 New Haven, CT 06520-8205, e-mail: marianne.lafrance@yale.edu.

REFERENCES

Abbey, A. (1982). Sex differences in attributions for friendly behavior: Do males misperceive female's friendliness? Journal of Personality and So­cial Psychology, 42, 830-838.

Abbey, A., & Melby, C. (1986). The effects of nonverbal cues on gender differ­ences in perceptions of sexual intent. Sex Roles, 15, 283-298.

Baird, J. E., Jr. (1976). Sex differences in group communication: A review of relevant research. Quarterly Journal of Speech, 62, 179-192.

Baker, D. D., Terpstra, D. E., & Larntz, K. (1990). The influence of individual characteristics and severity of harassing behavior on reactions to sexual ha­rassment. Sex Roles, 22, 305-325.

Bugental, D. B., Shennum, W., Frank, M., & Ekman, P (2001). 'True lies': Chi1-dren's abuse history and power attributions as influences in deception de­tection. In V. Manusov & J. H. Harvey (Eds.), Attribution, communication behavior, and close relationships. Advances in personal relations (pp. 248-265). Cambridge, U.K.: Cambridge University Press.

Burgoon, J. K., Buller, D. B., & Woodall, W. G. (1989). Nonverbal communica­tion: The unspoken dialogue. New York: Harper-Collins.

Conte, A. (1997). Legal theories of sexual harassment. In W. O'Donohue (Ed.), Sexual harassment: Theory, research, and treatment (pp. 50-83). Boston: Allyn & Bacon.

Crull, P. (1982). Stress effects of sexual harassment on the job: Implications for counseling. American Journal of Orthopsychiatry, 52, 539-544.

Dansky, B. S., & Kilpatrick, D. G. (1997). Effects of sexual harassment. In W. O'Donohue (Ed.), Sexual harassment: Theory, research, and treatment (pp. 152-174). Boston: Allyn & Bacon.

Denmark, F. L. (1977). Styles of leadership. Psychology of Women Quarterly, 2, 99-113.

Deutsch, F. M. (1990). Status, sex, and smiling: The effect of role on smiling in

men and women. Personality and Social Psychology Bulletin, 16, 531-540. Deutsch, F. M., LeBaron, D., & Fryer, M. M. (1987). What is in a smile? Psychol‑

ogy of Women Quarterly, 11, 341-351.

Dovidio, J. F., Brown, C. E., Heitman, K., Ellyson, S. L., & Keating, C. F. (1988). Power displays between women and men in discussions of gender-linked tasks: A multichannel study. Journal of Personality and Social Psychol­ogy, 55, 580-587.

EEOC. (1980). 'Title 29-Labor, Chapter XIV-Part 1604-Guidelines on Dis­crimination Because of Sex under Title VII of the Civil Rights Act, as

Amended Adoption of Interim Interpretive Guideline,' Washington, DC: U.S. Government Printing Office.

Ekman, P., & Friesen, W. V. (1978). Facial Action Coding System Manual. Palo Alto, CA: Consulting Psychologists Press.

Ekman, F, & Friesen, W. V. (1982). Felt, false, and miserable smiles. Journal of Nonverbal Behavior, 6, 238-252.

Ekman, F, Friesen, W. V., & Ancoli, (1980). Facial signs of emotional experi­ence. Journal of Personality and Social Psychology, 39, 1125-1134.

Ellyson, S. L., & Dovidio, J. F. (1985). Power, dominance, and nonverbal behav ior: Basic concepts and issues. In S. L. Ellyson & J. F. Dovidio (Eds.), Power, dominance, and nonverbal behavior (pp. 1-27). New York: Springer.

Fitzgerald, L. F., Drasgow, F., Hulin, C. L., Gefland, M. J., & Magley, V. J. (1997). Antecedents and consequences of sexual harassment in organizations: A test of an integrated model. Journal of Applied Psychology, 82(4), 578-589.

Fitzgerald, L. F., & Ormerod, A. J. (1993). Breaking silence: The sexual harass­ment of women in academia and the workplace. In F. Denmark & M. Paludi (Eds.), Psychology of women: A handbook of issues and theories (pp. 553-582). Westport, CT: Greenwood Press.

Frank, M. G., Ekman, P., & Friesen, W. V. (1993). Behavioral markers and recognizability of the smile of enjoyment. Journal of Personality and Social Psychology, 64, 83-93

Frieze, I. H., Parsons, J. E., Johnson, P.B., Ruble, D. N., & Zellman, G. L. (1978). Women and sex roles: A social psychological perspective. New York: W. W. Norton.

Gutek, B. A., & Koss, M. P. (1993). Changed women and changed organiza­tions: Consequences of and coping with sexual harassment. Journal of Vo­cational Behavior, 42, 28-48.

Halberstadt, A. G., Dovidio, J. F., & Davidson, L. A. (1988, October). Power, gender, and smiling. Paper presented at the annual meeting of the Society for Experimental Social Psychology, Madison, M

Hall, J. A. (1978). Gender effects in decoding nonverbal cues. Psychological Bulletin, 85, 845-857.

Hall, J. A. (1984). Nonverbal sex dierences. Baltimore: John Hopkins Uni­versity Press.

Hall, J. A., & Payne, R. M. (1995). Smiling in relation to trait and manipu­lated dominance/status. Unpublished manuscript, Northeastern Univer­sity, Boston.

Haviland, J. M. (1977). Sex-related pragmatics in infants' nonverbal commu­nication. Journal of Communication, 27, 80-84.

Hecht, M. A., & Lance, M. (1998). License or obligation to smile: The effect of power and sex on amount and type of smiling. Personality and Social Psychology Bulletin, 24, 1223-1342.

Henley, N. M. (1977). Body politics: Power, sex, and nonverbal communica­tion. Englewood Cliffs, NJ: Prentice Hall.

Imada, A. S., & Hakel, M. D. (1977). Influence of nonverbal communication and rater proximity on impressions and decisions in simulated employ­ment interviews. Journal of Applied Psychology, 62, 295-300,

Lance, M., & Hecht, M. A. (1999). Obliged to smile: The effect of power and gender n facial expression. In P Philippot, R. S. Feldman, & E. J. Coats (Eds.), The social context of nonverbal behavior (pp. 45-70). Cambridge, U.K.: Cambridge University Press.

Lance, M., Hecht, M. A., & Paluck, E. L. (2003). The contingent smile: A meta‑

analysis of sex differences in smiling. Psychological Bulletin, 129, 305-334. Lakoff, R. (1973). Language and woman's place. Language in Society, I(2),

45-80.

Lieberman, D. A., Rigo, T. G., & Campain, R. F. (1988). Age-related differences in

nonverbal decoding ability. Communications Quarterly, 36(4), 290-297.
Livingston, J. A. (1982). Responses to sexual harassment on the job: Legal, or‑

ganizational, and individual actions. Journal of Social Issues, 38(4), 5-22. Meritor Savings Bank v. Vinson. (1986). 477 U.S. 57, 40 FEP Case 1822. Olian, J. D., Schwab, D. P., & Haberfeld, Y. (1988). The impact of applicant gen‑

der compared to qualifications on hiring recommendations: A meta-analy‑

sis of experimental studies, Organizational Behavior and Human

Decision Process, 41(2), 180-195.

Piotrkowski, C. S. (1998). Gender harassment, job satisfaction, and distress among employed white and minority women. Journal of Occupational Health Psychology, 3, 33-43.

or, J. B. (1987). Sexual harassment proclivities in men. Sex Roles, 17, 269-290.

Pryor, J. B., Giedd, J. L., & Williams, K. B. (1995). A social psychological model

for predicting sexual harassment. Journal of Social Issues, 51, 69-84.
Ridgeway, C. L., & Smith-Lon, L. (1999). The gender system and interaction.

Annual Review of Sociology, 25, 191-216.

Rotter, N. G., & Rotter, G. S. (1988). Sex differences in the encoding and decoding

of negative facial emotions. Journal of Nonverbal Behavior, 12(2), 139-148.
Rudman, L. A., & Borgida, E. (1995). The afterglow of construct accessibility:

The behavioral consequences of priming men to view women as sexual ob‑

jects. Journal of Experimental Social Psychology, 31, 493-517. Rudman, L. A., & Glick, P. (2001). Prescriptive gender stereotypes and back‑

lash toward agentic women. Journal of Social Issues, 57, 743-762. Schneider, K. T., Swan, S., & Fitzgerald, L. F. (1997). Job-related and psycho‑

logical effects of sexual harassment in the workplace: Empirical evidence

from two organizations. Journal of Applied Psychology, 82, 401-415.
Spangler, L. (1995). Gender-specific nonverbal communication: Impact for

speaker effectiveness. Human Resource Development Quarterly, 6,

409-419.

Stockdale, M. S. (1998). The direct and moderating influences of sexual ha­rassment pervasiveness, coping strategies, and gender on work-related out-comes. Psychology of Women Quarterly, 22, 521-535.

U.S. Merit Systems Protection Board. (1981). Sexual harassment of federal

workers Is it a problem? Washington, DC: U.S. Government Printing Office. Washburn, P. V., & Hakel, M. D. (1973). Visual cues and verbal content as influ‑

ence on impressions formed after simulated employment interviews. Jour‑

nal of Applied Psychology, 5(8), 137-141.

Woodzicka, J. A., & Lance, M. (2001). Real versus imagined gender harass­ment. Journal of Social Issues, 57, 15-30.

Word, C. O., Zanna, M. P., & Cooper, J. (1974). The nonverbal mediation of self-fulfilling prophecies in interracial interaction. Journal of Experimen­tal Social Psychology, 10(2), 109-120.

Zuckerman, M., Hall, J. A., DeFrank, R. S., & Rosenthal, R. (1976). Encoding and decoding of spontaneous and posed facial expression. Journal of Per­sonality and Social Psychology, 34, 966-977.