EVOLUTIONARY ORIGINS OF SEX DIFFERENCES IN JEALOUSY ? Questioning the " Fitness " of the Model

Evolutionary psychology has become a popular frameM'ork for studying jealousy. Much of this popularity can be attributed to work by Buss and his colleagues showing an apparent relation between an individual's sex and jealousy for certain types of infidelity (i.e., sexual vs. emotional) that is consistent with evolutionary theory (Buss, Larsen, Westen, & Semmelroth, 1992). In two .studies, we take issue with these findings and argue that the relation between sex and jealousy reported by Buss and his colleagues is more properly explained by considering individuals' beliefs concerning the covariation between sexual and emotional infidelity. Evolutionary psychology has become a popular perspective from which to study jealousy. In accord with this perspective, the origins of jealousy are ascribed to the evolutionary history of humans (Buss. 1991. 1995). and the psychological mechanisms thought to be responsible for the evocation of jealousy are evaluated with respect to their present or past adaptive benefits (e.g., Wiederman & Allgeier, 1993). The popularity of the evolutionary perspective among investigators studying jealousy can be attributed in part to an influential article by Buss and his colleagues showing a sex difference in the intensity of jealousy in response to different types of infidelity (Buss. Larsen, Westen, & Semmelroth. 1992). Men reported more jealousy in sittiations involving sexual rather than emotional infidelity, but women reported more jealousy in situations involving emotional, as opposed to se.xual. infidelity. These sex differences in the elicitors of jealousy arise, according to the evolutionary model, as a consequence of their fitness-enhancing capabilities (Buss et a!., 1992). Fitness refers to the ability to pass on genetic material by raising offspring to the age of sexual maturity (Daly & Wilson, 1983). Briefly stated, evolutionary theory predicts that males in species employing internal fertilization are vigilant of possible sexual contact by their mates with other males; this behavior is designed to prevent cuckoldry. Females of biparental species with internal fertilization have no doubt concerning their genetic link to offspring and are therefore predicted to be vigilant of threats concerning the absconding of the male, not of the sexual act itself; the male's continued presence aids in the successful rearing of the offspring (Buss et al., 1992; Daly & Wilson, 1983). In order to evaluate these predictions for human jealousy. Buss et al. (!992) conducted three studies. In all the studies, the dependent variable of interest was which of two types of infidelity (sexual or emotional) would evoke more intense jealousy. Address correspondence to Peter Salovey, Department of Psychology, Yale University, P.O. Box 208205. New Haven, CT 06520-8205. As defined by Buss et al. (1992). sexual infidelity involves actual sexual contact between individuals; emotional infidelity involves the expression of a deep affection for and attachment to another individual. Participants in these studies were asked to imagine situations representing each type of infidelity. It was expected that men would be more distressed by sexual than by emotional infidelity and women would be more distressed by emotional than by sexual infidelity. Two of the three studies presented subjects with a forcedchoice question asking them simply to indicate which of the two types of infidelity would cause them more distress. In accordance with the evolutionary perspective. Buss and his colleagues found a significant sex difference in the choice of infidelity type; women were much more likely than men to indicate that the emotional infidelity event resulted in more distress. Also, physiological data were collected as a measure of emotional arousal in response to imagining each type of infidelity. Men showed significant elevations in electrodermal activity in response to imagining the sextial as compared with the emotional infidelity situation; the reverse pattern was found in women. Data from other physiological indices (pulse rate and electromyographic activity of the corrugator supercUii] were not as clear (Buss et al.. 1992). Based on these findings. Buss et al. (1992) concluded that their predictions were supported, and that only the "evolutionary psychological frameworks . . . generated the sexdifferentiated predictions in advance and on the basis of sound evolutionary reasoning" (p. 255). We challenge this interpretation of these findings. A major threat to the credibility of results based on a paradigm that does not use random assignment of participants to conditions is the influence of unmeasured variables (Abelson, 1995; Bollen. 1989). Such misspecification can lead to the acceptance of spurious results. When individuals are not assigned randomly to conditions, there is less assurance that the other dimensions upon which the individuals vary are balanced within the ensuing analyses. Consequently, causal claims with reference to the measured independent variable (or variables) may be compromised. In the case at hand, individuals, of course, entered the study as men or women; sex cannot be assigned randomly. Therefore, the specter that variables correlated with sex were not balanced in subsequent analyses must be a concem. This situation is not ustially problematic when examining sex differences in a descriptive manner; if men behave a certain way because of a third variable that is correlated with sex, it may still make sense to speak in terms of a sex difference. However, when claims are attributed to sex based solely on genetically influenced predispositions, as opposed to other socially derived influences, lack of random assignment signals possible problems; there can be VOL. 7, NO. 6, NOVEMBER 1996 Copyright © 1996 American Psychoiogicai Society 367 PSYCHOLOGICAL SCIENCE Sex Differences in Jealousy? little confidence that observed sex differences are not due to other nongenetic variables correlated with sex. Extra care must be taken in such situations to examine the influence of alternative explanatory variables. We suspect that the findings presented by Buss et al. (1992) can be explained by what we term the double-shot hypothesis. Simply stated, some individuals believe that emotional and sexual infidelity are not independent events. Consequently, they will select the type of infidelity that more implies the occurrence of the other when asked to indicate which one would make them more jealous. For instance, emotional infidelity, for certain individuals, may imply that sexual infidelity has occurred or soon will occur. These perceptions of nonindependence, moreover, may be correlated with sex in some samples, with women more likely than men to expect that emotional infidelity by their partners implies associated sexual infidelity. If the double-shot hypothesis is correct, it might explain the results obtained in the forced-choice paradigm used by Buss et al. (1992). According to this hypothesis, women select emotional infidelity as more distressing in the forced-choice paradigm because emotional infidehty really represents two types of infidelity as opposed to one. Certain types of infidelity bother individuals more than others because they represent a double shot of infidelity; the occurrence of both these types of infidelity is no doubt more troubling than either individually and also signals a greater threat to the relationship. Such beliefs concerning the nonindependence of these two types of infidelity cannot be traced to genetic causes according to the usual evolutionary arguments in this domain; females should always be concerned with loss of attention and resources, whether or not males engage in extradyadic sexual activity. Instead, it seems more likely that perceptions of the nonindependence of these types of infidelity are derived through socialization; men and women, because of past experience, may hold differing beliefs concerning the implications of the each of the two types of infidelity. Moreover, perceptions of the nonindependence of sexual and emotional infidelity seem likely to vary not only across, but also within, sex. To the extent that this is the case, and to the extent that these perceptions account for variance in jealousy reported by individuals, perceptions of nonindependence would seem to create a more powerful and parsimonious accotint of jealousy than would an account based on evolution. The following two studies were designed to evaluate this alternative explanation for the findings reported by Buss et al. (1992). In each case, we expected to show that the reported relation between biological sex and jealousy for certain types of infidelity was due to the association of sex with differing perceptions of the nonindependence of the two types of infidelity. We predicted that sex would provide no unique explanatory ability beyond that associated with these perceptions of nonindependence, and because sex is not the causal agent of such perceptions, but may simply covary with them, confirmation of these predictions would identify the relation between sex and infidelity choice as misspecified.