PSYCHOLOGICAL SCIENCE Research Report

SEX DIFFERENCES IN : Evolution, Physiology, and Psychology David M. Buss, Randy J. Larsen, Drew Westen, and Jennifer Semmelrolh University of Michigan

Abstract—In species with internal female sources in putative offspring that are ge- mous , a woman risked having fertilization, males risk both lowered pa- netically unrelated. her mate invest in an alternative woman ternity probability and investment in ri- These multiple and severe reproduc- with whom he was having an affair (par- val gametes if their mates have sextial tive costs should have imposed strong tial loss of investment) or risked his de- contact with other males. Females of selection pressure on males to defend parture for an alternative woman (large such species do not risk lowered mater- against cuckoldry. Indeed, the literature or total loss of investment). Second, in nity probability through partner infidel- is replete with examples of evolved an- polygynous , a woman was at ity, but they do risk the diversion of their ticuckoldry mechanisms in lions (Ber- risk of having her mate invest to a larger mates' commitment and resources to ri- tram, 1975), bluebirds (Power, 1975), degree in other wives and their offspring val females. Three studies tested the hy- doves (Erickson & Zenone, 1976), nu- at the expense of his investment in her pothesis that sex differences In jealousy merous insect species (Thornhill & Al- and her offspring. Following Buss (1988) emerged in humans as solutions to the cock, 1983), afid nonhuman primates and Mellon (1981), we hypothesize that respective adaptive problems faced by (Hrdy, 1979). ^Sirice humans arguably cues to the development of a deep emo- each sex. In Study I, men and women show more paternafinvestment than any tional attachment have been reliable selected which event would upset them other of the 200 species of primates (Al- leading indicators to women of potential more—a partner's sexual or exander & Noonan, 1979), this selection reduction or loss of their mate's invest- emotional infidelity. Study 2 recorded pressure should have ope ratgde specially ment. physiological responses (heart rate, intensely on human male_sASymons ' Jealousy is defined as an emotional electrodermal response, coiTUgator su- (1979); Daly, Wilson, and Weghorst "st^te that is aroused by a perceived percilii contraction) while subjects imag- (1982); and Wilson and Daly (in press) threat to a valued relationship or position ined separately the two types of partner have hypothesized that male sexual jeal- and motivates behavior aimed at coun- infidelity. Study 3 tested the effect of be- ousy evolved as a solution to this adap- tering the threat. Jealousy is 'sexual' if ing in a committed sexual relationship tive problem (but see Hupka, 1991, for the valued relationship is sexual" (Daly on the activation of jealousy. All studies an alternative view). Men who were in- et al., 1982, p. 11; see also, Salovey, showed large sex differences, confirming different to sexual contact between their 1991; White & Mullen. 19^\It is rea- hypothesized sex linkages in jealousy ac- mates and other men presumably ex- sonable to hypothesize that jealousy in- tivation. perienced lower paternity certainty, volves physiological reactions (auto- greater investment in competitors' ga- nomic arousal) to perceived threat and metes, and lower reproductive success motivated action to reduce the threat, al- than did men who were motivated to at- though this hypothesis has not been ex- In species with internal female fertil- tend to cues of infidelity and to act on ization and gestation, features of repro- amined. Following Symons (1979) and those cues to increase paternity proba- Daly et al. (1982), our central hypothesis ductive biology characteristic of all 4,000 bility. species of mammals, including humans, is that the events that activate jealousy males face an adaptive problem not con- Although females do not risk mater- physiologically and psychologically dif- fronted by females—uncertainty in their nity uncertainty, in species with biparen- fer for men and women because of the paternity of offspring. Maternity proba- tal care they do risk the potential loss of different adaptive problems they have bility in mammals rarely or never devi- time, resources, and commitment from a faced over human evolutionary history ates from 100%. Compromises in pater- male if he deserts or channels investment in mating contexts. Both sexes are hy- nity probability come at substantial re- to alternative mates (Buss, 1988; Thorn- pothesized to be distressed over both productive cost to the male—the loss of hUl & Alcock, 1983; Trivers, 19722^The sexual and emotional infidelity, and pre- mating effort expended, including time, redirection of a mate's investment to an- vious findings bear this out (Buss. 1989). energy, risk, nuptial gifts, and mating op- other female and her offspring is repro- However, these two kinds of infidelity portunity costs. A cuckolded male also ductively costly for a female, especially should be weighted differently by men loses the female's parental effort, which in environments where offspring suffer and women. Despite the importance of becomes channeled to a competitor's ga- in survival and reproductive currencies these hypothesized sex differences, no metes. The adaptive problem of pater- without investment from both parents. systematic scientific work has been di- nity uncertainty is exacerbated in spe- rected toward verifying or falsifying their In human evolutionary history, there existence (but for suggestive data, see cies in which males engage in some were likely to have been at least two sit- postzygotic (Triv- Francis, 1977; Teismann & Mosher, uations in which a woman risked losing a 1978; White & Mullen, 1989). ers, 1972). Males risk investing re- man's investment. First, in a monoga-

VOL. 3. NO. 4. JULY 1992 Copyright © 1992 American Psychological Society 251 PSYCHOLOGICAL SCIENCE

So\ UinViences in Jealousy

STUDY 1: SUBJECTIVE 70 DISTRESS OVER A PARTNER S EXTERNAL INVOLVEMENT 60- This study was designed to test the hypothesis that men and women differ in which form of infidelity—sexual versus 50- emotional—triggers more upset and sub- jective distress, following the adaptive Pcrcentage 40- logic just described. Reporting More Distress to Sexual Infidelity 3Q Method After reporting age and sex, subjects 20- (N = 202 undergraduate students) were presented with the following dilemma: 10 -

Please think of a serious committed romantic relationship that you have had in the past, that Sexual Infidelity versus Sexual Infidelity versus you currently have, or that you would like to Deep Emotional Inildellty Love Infidelity have. Imagine that you discover that the per- son with whom you've been seriously in- volved became interested in someone else. What would distress or upset you more (please circle only one): 70 (A) Imagining your partner forming a deep emotional attachment to that person. 60- (B) Imagining your partner enjoying pas- sionate sexual intercourse with that other person. 50-

Subjects completed additional ques- tions, and then encountered the next di- Percentage 40- Reporting lemma, with the same instructional set, More Distress to but followed by a different, but parallel, Sexual Infidelity 30 choice:

(A) Imagining your partner trying different sexual positions with that other person. (B) Imagining your partner falling in love 10 - with that other person.

Results Have Been in Committed Have Not Been In Committed Sexual RdaUonship Sexual Relationship Shown in Figure 1 (upper panel) are the percentages of men and women re- Fig. 1. Reported comparisons of distress in response to imagining a partner's sexual porting more distress in response to sex- or emotional infidelity. The upper panel shows results of Study 1—the percentage of ual infidelity than emotional infidelity. subjects reporting more distress to the sexual infidelity scenario than to the emotional The first empirical probe, contrasting infidelity (left) and the love infidelity (right) scenarios. The lower panel shows the distress over a partner's sexual involve- results of Study 3—the percentage of subjects reporting more distress to the sexual ment with distress over a partner's deep infidelity scenario than to the emotional infidelity scenario, presented separately for emotional attachment, yielded a large those who have experienced a committed sexual relationship (left) and those who have not experienced a committed sexual relationship (right). and highly significant sex difference (x^ = 41.56, df^ 3,/7< .001). Fuily60%of the male sample reported greater distress greater distress over a partner's emo- with 32% more men than women report- over their partner's potential sexual infi- tional attachment to a rival.7 ing greater distress over a partner's sex- delity; in contrast, only 17% of the fe- This pattern was replicated with the ual involvement with someone else, and male sample chose that option, with 83% contrast between sex and love. The mag- the majority of women reporting greater reporting that they would experience nitude of the sex difference was large. distress over a partner's falling in love

252 VOL. 3, NO. 4, JULY 1992 PSYCHOLOGICAL SCIENCE

David M. Buss et al.

with a rival = 59,20, 4f' = 3, p < Ptilse rate intercourse with this other person. Try to .001). A photoplethysmograph was attached feel the feelings you would have if this to the subject's right thumb to monitor happened to you." the pulse wave. The signal from this The instructions for emotional infidel- STUDY 2: PHYSIOLOGICAL pulse transducer was fed into a Grass ity imagery were identical to the above, RESPONSES TO A PARTNER S Model 7P4 cardiotachometer to detect except the italicized sentence was re- EXTERNAL INVOLVEMENT the rising slope of each pulse wave, with placed with ''Imagine that your partner the internal circuitry ofthe Schmitt trig- is falling in love and forming an emo- Given the strong confirmation of jeal- ger individually adjusted for each subject tional attachment to that person.^'' Phys- ousy sex linkage from Study I, we to output PR in beats per minute. iological data were collected for 20 s sought next to test the hypotheses using following the subject's button press indi- physiological measures. Our central cating that he or she had achieved the measures of autonomic arousal were Electromyographic activity Bipolar EMG recordings were ob- image. Subjects were told to "stop and electrodermai activity (EDA), assessed relax" for 30 s between imagery trials. via skin conductance, and pulse rate tained over the corrugator superciiii (PR). Electrodermai activity and pulse muscle. The EMG signal was relayed to a wide-band AC-preampIifier (Grass rate are indicators of autonomic nervous Results system activation (Levenson, 1988). Be- Model 7P3), where it was band-pass fil- cause distress is an unpleasant subjec- tered, full-wave rectified, and integrated Physiological scores tive state, we also included a measure of with a time constant of 0.2 s. The following scores were obtained: muscle activity in the brow region of the (a) the amplitude of the largest EDA re- face—electromyographic (EMG) activity sponse occurring during each 20-s trial; ofthe corrugator superciiii muscle. This Procedure (b) PR in beats per minute averaged over muscle is responsible for the furrowing After electrode attachment, the sub- each 20-s trial; and (c) amplitude of EMG ofthe brow often seen in displays ject was made comfortable in a reclining activity over the corrugator superciiii of unpleasant emotion or affect (Frid- chair and asked to relax. After a 5-min averaged over each 20-s trial. Difference lund, Ekman, & Oster, 1987). Subjects waiting period, the experiment began. scores were computed between the neu- were asked to image two scenarios in The subject was alone in the room during tral imagery trial and the jealousy induc- which a partner became involved with the imagery session, with an intercom on tion trials. Within-sex t tests revealed no someone else—one sexual intercourse for verbal communication. The instruc- effects for order of presentation of the scenario and one emotional attachment tions for the imagery task were written sexual jealousy image, so data were col- scenario. Physiological responses were on a form which the subject was re- lapsed over this factor. recorded during the imagery trials. quested to read and follow. Each subject was instructed to engage Jealousy induction effects in three separate images. The first image Table 1 shows the mean scores for the Subjects was designed to be emotionally neutral: physiological measures for men and "Imagine a time when you were walking women in each of the two imagery con- Subjects were 55 undergraduate stu- to class, feeling neither good nor bad, ditions. Differences in physiological re- dents, 32 males and 23 females, each just neutral." The subject was instructed sponses to the two jealousy images were completing a 2-hr laboratory session. to press a button when he or she had the examined using paired-comparison t image clearly in mind, and to sustain the tests for each sex separately for EDA, image until the experimenter said to PR, and EMG.'sJhe men showed signifi- Physiological Measures stop. The button triggered the computer cant increases in EDA during the sexual to begin collecting physiological data for imagery compared with the emotional Physiological activity was monitored 20 s, after which the experimenter in- on the running strip chart of a Grass imager2(r = 2.00, ^/ = 29, p < .05). structed the subject to "stop and relax." Women showed significantly greater Model 7D polygraph and digitized on a The next two images were infidelity laboratory computer at a 10-Hz rate, fol- EDA to the emotional infldehty image images, one sexual and one emotional. than to the sexual infidelity image\(/ = lowing principles recommended in Ca- The order of presentation of these two cioppo and Tassinary (1990), 2.42, df - 19, p < .05). A similar-pattern images was counterbalanced. The in- was observed with PR.^^en showed a structions for sexual jealousy imagery substantial increase in PR to both im- Electrodermai activity were as follows: "Please think of a seri- ages, but significantly more so in re- Standard Beckman Ag/AgCl surface ous romantic relationship that you have sponse to the sexual infidelity image\(' = electrodes, filled with a .05 molar NaCl had in the past, that you currently have, 2.29, df=3l,p< ,05).^ Women showed solution in a Unibase paste, were placed or that you would like to have. Now elevated PR to both images, but not dif- over the middle segments of the first and Imagine that the person with whom ferentially so^The results of the corru- you're seriously involved becomes inter- third fingers of the right hand. A Wheat- gator EMG were similar, although less stone bridge apphed a 0.5-V voltage to ested in someone else. Imagine you fmd out that your partner is having sexual strong. Klen showed greater brow con- one electrode. traction to the sexual infidelity image,

VOI d IlIIV tOQT 253 PSYCHOLOGICAL SCIENCE

Sex DitTcrences in Jealousy

lo Ihe emolional infidelity image than and not significant: Women reported that Table 1. Means and standard women who had not been in a committed they would experience more distress dt'viittitms on physiological sexual relationship. The rationale was about a partner's emotional infidelity measures during tWD that direct experience of the relevant than a partner's sexual infidelity, regard- imagery conditions context during development may be nec- less of whether or not they had experi- essary for the activation of the sex- enced a committed sexual relationship Imagery linked weighting of jealousy activation. (Y^= 0.80, #= l,ns). Measure type Mean SD ' For men, the difference between Males those who had been in a sexual relation- EDA Sexual 1,30 3.64 Subjects ship and those who had not is large and ErriLiIional -0.11 0.76 highly significant. Whereas 55% of the Pulse Sexual 4.76 7.80 Subjects for Study 3 were 309 under- rate Emotional 3.00 5.24 graduate students, 133 men and 176 men who had experienced committed Brow Sexual 6.75 32.96 women. sexual relationships reported that they EMG Emotional 1.16 6.60 would be more distressed by a partner's sexual than emotional infidelity, this fig- Females EDA Sexual -0.07 0.49 Procedure ure drops to 29% for men who had never Emotional 0.21 0.78 experienced a committed sexual rela- Pulse Sexual 2.25 4.68 Subjects read the following instruc- tionshlgjx' = 12.29, df = \,p< .001). rate Emotional 2.57 4.37 tions: Sexual jealousy in men apparently be- Brow Sexual 3.03 8.38 comes increasingly activated upon expe- Emotional 8.12 25.60 EMG Please think of a serious or committed roman- rience of the relevant relationship. tic relationship that you have had in the past, Note. Measures are expressed as that you currently have, or that you would changes from the neutral image like to have. Imagine that you discover that condition. EDA is in microsiemen the person with whom you've been seriously DISCUSSION units, pulse rate is in beats per minute, and EMG is in microvolt involved became interested in someone else. units. What would distress or upset you more The results of the three empirical {please circle only one): studies support the hypothesized sex (A) Imagining your partner falling in love linkages in the activators of jealousy. and women showed the opposite pattern, and forming a deep emotional attachment to Study I found large sex differences in although results with this nonautonomic that person. reports of the subjective distress individ- measure did not reach significance (/ = (B) Imagining your partner having sexual uals would experience upon exposure to 1.12. df = 30, p < .14, for males: r - intercourse with tbat other person. a partner's sexual infidelity versus emo- -1.24, df = 22, p < .12, for females). tional infidelity. Study 2 found a sex link- The elevated EMG contractions for both Alternatives were presented in stan- age in autonomic arousal to imagined dard forced-choice format, with the or- jealousy induction trials in both sexes sexual infidelity versus emotional infi- der counterbalanced across subjects. delity; the results were particularly support the hypothesis that the affect ex- Following their responses, subjects were perienced is negative, i asked: "Have you ever been in a serious strong for the EDA and PR. Study 3 rep- or committed romantic relationship? licated the large sex differences in re- (yes or no)" and "If yes, was this a sex- ported distress to sexual versus emo- STUDY 3: CONTEXTS THAT ual relationship? (yes or no)." tional infidelity, and found a strong ACTIVATE THE effect for men of actually having experi- JEALOUSY MECHANISM enced a committed sexual relationship. Results These studies are limited in ways that The goal of Study 3 was to replicate call for additional research. First, they and extend the results of Studies 1 and 2 The results for the total sample repli- pertain to a single age group and culture. using a larger sample. Specifically, we cate closely the results of Study 1. A Future studies could explore the degree sought to examine the effects of having much larger proportion of men (49%) to which these sex differences transcend been in a committed sexual relationship than womeiv-'(19%) reported that they different cultures and age groups. Two versus not having been in such a rela- would be more distressed by their part- clear evolutionary psychological predic- tionship on the activation of jealousy. ner's sexual involvement with someone tions are (a) that male sexual jealousy We hypothesized that men who had ac- else than by their partner's emotional at- and female commitment jealousy will be tually experienced a committed sexual tachment to, or love for, someone else greater in cultures where males invest relationship would report greater subjec- (^^ = 38.48, #= 3,p< .001). heavily in children, and (b) that male tive distress in response to the sexual in- The two pairs of columns in the bot- sexual jealousy will diminish as the age fidelity imagery than would men who tom panel of Figure I show the results of the male's mate increases because her had not experienced a high-investing separately for those subjects who had reproductive value decreases. Second, sexual relationship, and that women who experienced a committed sexual rela- future studies could test the alternative had experienced a committed sexual re- tionship in the past and those who had hypotheses that the current findings re- lationship would report greater distress not. (for women, the difference is small flect (a) domain-specific psychological

254 VOL. 3, NO. 4, JULY 1992 PSYCHOLOGICAL SCIENCE

David M. Bu.ss et al.

adaptations to cuckoldry versus poten- molded post hoc to fit the findings— Principles of psychophysiology: Phyxical, so- tial investment loss or (b) a more do- cial, and inferential element.-^. Cambridge. En- something perhaps true of any set of gland: Cambridge University Press, main-general mechanism such that any findings. None but the Symons (1979) Daly, M.,& Wilson, M. (1988a). Evolutionary social thoughts of sex are more interesting, and Daly et al. (1982) evolutionary psy- psychology and family . Science. 242, arousing, and perhaps disturbing to men 519-524. chological frameworks, however, gener- Daly, M., & Wilson, M. (1988b). Homicide. Haw- whereas any thoughts of love are more ated the sex-differentiated predictions in thorne, NY: Aldine. interesting, arousing, and perhaps dis- advance and on the basis of sound evo- Daly, M., Wilson. M., & Weghorst, S.J. (1982). turbing to women, and hence that such Male sexual jealousy. Ethology and Sociobiol- lutionary reasoning. The recent finding ogy. 3. 11-27. responses are not specific to jealousy or that male sexual jealousy is the leading Erickson, C.J., & Zenone, P.G. (1976). Courtship infidelity. Third, emotional and sexual cause of spouse battering and homicide differences in male ring doves: Avoidance of infidelity are clearly correlated, albeit cuckoldry? Science. 192. 1353-1354. across cultures worldwide (Daly & Wil- Francis, J.L. (1977). Toward the management of het- imperfectly, and a sizable percentage of son, 1988a, 1988b) offers suggestive evi- erosexual jealousy. Journal of Marriage and men in Studies 1 and 3 reported greater dence that these sex differences have Family Counseling. 10. 61-69. distress to a partner's emotional infidel- Fridlund, A., Ekman, P-, & Oster, J. (1987). Facial large social import and may be species- expressions of emotion. In A. Siegman & S. ity. Emotional infidelity may signal sex- wide. Feldstein (Eds.), Nonverbal behavior and ual infidelity and vice versa, and hence communication (pp. 143-224). Hillsdale, NJ: both sexes should become distressed at Eribaum. Hrdy, S.B.G. (1979(. Infanticide among animals: A both forms (see Buss, 1989). Future Acknowledgments—This research was review, classification, and examination of ihe research could profitably explore in supported in part by National Institute of implications for the reproductive strategies of greater detail the correlation of these Mental Health (NIMH) Grant MH-44206- {sfna\cs. Ethology and Sociobiology. I. 14—40. 02 to David M. Buss; Research Scientist Hupka, R.B. (1991). The motive for the arousal of forms of infidelity as well as the sources Development Award KO1-MH00704 and romantic jealousy: Its cultural origin. In P. Sa- of within-sex variation. Finally, the in- NIMH Grant MH 42057 to Randy Larsen; lovey (Ed.), The psychology of jealousy and triguing finding that men who have expe- and Biomedical Research Support Grant en\y (pp. 252-270). New York: Guilford Press. SO7 RR07050-25 from the National Insti- Levenson, R.W. (1988). Emotion and the autonomic rienced a committed sexual relationship tutes of Health to Randy Larsen and nervous system: A prospectus for research on differ dramatically from those who have David M. Buss through the University of autonomic specificity. In H. Wagner (Ed.), So- cial psychophysiology: Theory and clinical ap- Michigan Office of the Vice President for not, whereas for women such experi- plications (pp. 17-42). London: Wiley. Research. The authors thank Michael ences appear to be irrelevant to their se- Mellon, L.W. (1981). The evolution of love. San lection of emotional infidelity as the Chen, Martin Daly, Todd DeKay, Bruce Francisco: W.H. Freeman. Ellis, Arlette Greer, Kurt Hoop, Tim Power, H.W. (1975). Mountain bluebirds: Experi- more distressing event, should be exam- Ketelaar, Neil Malamuth, David Schmitt, mental evidence against altruism. Science. ined. Why do such ontogenetie experi- and Don Symons for helpful suggestions 189. 142-143. ences matter for men, and why do they on earlier versions of this article. Salovey, P. (Ed.). (1991). The psychology of jeal- ousy and envy. New York: Guilford Press. appear to be irrelevant for women? Symons, D. (1979). The evolulion of human sexual- ity. New York: Oxford University Press. Within the constraints of the current Teismann, M.W,, & Mosher, D,L. (1978). Jealous studies, we can conclude that the sex dif- conflict in dating couples. Psychological Re- REFERENCES ports. 42. 1211-1216. ferences found here generalize across Thomhill, R,, & Alcock, J. (19831. The evolution of insect mating systems. Cambridge, MA: Har- both psychological and physiological Alexander, R.D., & Noonan, K.M. (1979), Conceal- vard University Press, methods—demonstrating an empirical ment of ovulation, parental care, and human Trivers, R, (1972). Parental investment and sexual social evolution. In N. Chagnon & W. Irons robustness in the observed effect. Th£_ selection. In B. Campbell (Ed.), Sexual selec- (Eds.), Evolutionary biology and human social degree to _whiGh--these. sexrlinked^Uciv, tion and [he descent of man. 1871-1971 (pp. behavior (pp. 436-453). North Scituate, MA: 136-179). Chicago: Aldine. tors correspond tp thejiygmhesized sex- Duxbury. White, G.L,, & Mullen, P,E. (1989). Jealousy: The- Bertram, B.CR. (1975). Social factors influencing linked adaplLve^^roblerns lends sup^ft' ory, research, and clinical strategies. New reproduction in wild lions. Journal of Zoology. York: Guilford Press, to th^ evoltitionary psychologi^ljrame- }77. 463^82. Wilson, M., & Daly. M. (in press). The man who work from which they were derived. Al- Buss, D.M. (1988). From vigilance to violence: Tac- mistook his wife for a chattel. inJ, Barkow, L. ternative theoretical frameworks, includ- tics of mate retention. Ethology and Sociobi- Cosmides, & J. Tooby (Eds.), The adapted ing those that invoke culture, social con- ology. 9. 291-317. mind: and the gener- Buss, D.M. f 1989). Conflict between the sexes: Stra- ation of culture. New York: Oxford University struction, deconstruction, arbitrary tegic interference and the evocation of anger Press, parental socialization, and structural and upset. Journal of Personality and Social powerlessness, undoubtedly could be Psychology. 56, 735-747. (RECEIVED 7/16/91; REVISION ACCEPTED Cacioppo, J.T., & Tassinary, L.G. (Eds). (1990). 12/13/91)

VOL. 3. NO. 4 rin Y 255