Four experiments found benevolent sexism to be worse than hostile sexism for women's cognitive performance. Experiments 1-2 showed effects of paternalist benevolent sexism and ruled out explanations of perceived sexism, context pleasantness, and performance motivation. Experiment 3 showed effects of both paternalist and complementary gender differentiation components of benevolent sexism. Benevolent sexism per se (rather than the provision of unsolicited help involved in paternalism) worsened performance. Experiment 4 showed that impaired performance due to benevolent sexism was fully mediated by the mental intrusions women experienced about their sense of competence. Additionally, Experiment 4 showed that gender identification protected against hostile but not benevolent sexism. Despite the apparently positive and inoffensive tone of benevolent sexism, our research emphasizes its insidious dangers.Keywords: benevolent and hostile sexism, discrimination, performance and working memory, ingroup identification, sense of competence Sexism is expressed in a variety of ways, some more subtle than others and some more hostile than others. Examples of this attitude are gender-related humor, sexist name-calling, sexual harassment, and employment discrimination. Gender discrimination, like any other type of discrimination, is less and less accepted in our modern societies, and legislation has been designed to promote gender equality. In Europe, for instance, Article 141 of the European Commission's treaty states that any job classification system that is used for determining pay must be based on the same criteria for both men and women and be so written as to exclude any sex discrimination. Nonetheless, even if obvious forms of sexism are often socially condemned, more subtle forms of sexism are not. Worse, they sometimes are promoted, as they may sound positive in tone or even seem gallant or chivalrous. Consider, for instance, a man helping a woman to carry luggage, paying for her meal at a restaurant, or complimenting her caring abilities. One might ask what is wrong with that. The problem is that people do not notice that these behaviors may be threatening. These kinds of behavior might, however, be expressions of paternalism or else might ambivalently flatter women while simultaneously implicitly suggesting their inferiority. Paternalism suggests that men should take responsibility for the welfare of women, who might not be able to carry luggage by themselves, might not earn enough money to pay at the restaurant, and should be remembered for their qualities (warmth) rather than their weak points (competence). In this article, we propose that paternalism as well as the other aspects of benevolent sexism, by suggesting women's lack of ability, are devastating to women's performance. Sexism Versus PatronizingHostile and benevolent sexism often go hand in hand and result in ambivalent attitudes toward women (Glick & Fiske, 1996. Hostile sexism is an obviously antagonistic attitude toward women, who are often viewed as t...

Building upon the social emotion model (Smith, 1999), we examined the combined impact of categorization context and social identification on emotional reactions and behavioural tendencies of people confronted with the victims of harmful behaviour. Depending on conditions, participants were led to categorize the victims and themselves in the same common group or in two distinct subgroups of the larger common group. We also measured participants' level of identification with the group that was made contextually salient. As predicted, emotional reactions of anger and their associated offensive action tendencies were more prevalent when participants were induced to see the victims and themselves as part of the same group and when they were highly identified with this common group. In line with appraisal theories of emotion, we also found that the emotional reaction fully mediated the impact of categorization context and identification on action tendencies. We discuss the data with respect to their implications for the role of emotion in improving intergroup relations.

In this paper, we examined reactions to situations in which, although one is not personally involved, one could see oneself connected to either the perpetrators or the victims of unfair behavior. We manipulated participants' similarity and measured their identification to either one of two groups which participants later learned was the victim or the perpetrator of harmful behavior. As predicted, making salient similarities to the victims lead participants to: 1) appraise the perpetrator's behavior as more unfair; 2) experience more anger; and 3) be more likely to take action against it and less prone to show support for it as a function of their level of identification with their salient ingroup. In sharp contrast, focusing participants' attention on their similarities to the perpetrators reversed this pattern of findings: Compared to high identifiers, low identifiers appraised the behavior as more unfair than high identifiers, which made them feel angry (and guilty) and less likely to show support for the perpetrator's behavior. The data also provide strong support for a mediational model in which appraisal of the situation colors the emotional reaction which in turn orients action tendencies. We discuss the implications of our findings for the issue of group-based emotions.

Two experiments were run in The Netherlands and Belgium 1 week after the terrorist attacks against the World Trade Center in New York on September 11, 2001. The aim was to investigate whether social categorization affected emotional reactions, behavioral tendencies, and actual behaviors. Results showed that focusing participants' attention on an identity that included American victims into a common ingroup led them to report more fear and stronger fear-related behavioral tendencies and to engage more often in fear-related behaviors than when victims were categorized as outgroup members. Results are discussed with respect to appraisal theories of emotion and E. R. Smith's model of group-based emotions.