Evolutionary psychology of jealousy in romantic relationships evidence for a sexually dimorphic response mechanism in humans

Sex Differences in Jealousy
1
The Evolutionary Psychology of Jealousy in Romantic Relationships:
Evidence for a Sexually Dimorphic Response Mechanism in Humans
by
REBECCA M. ALLEN
A
thesis submitted in partial fulfillment
of the requirements for the
Degree of Bachelor of Arts with Honors
in Psychology
WILLIAMS COLLEGE
Williamstown, Massachusetts
May
11,2005
Sex Differences in Jealousy
2
Abstract
Evolutionary psychologists have proposed that men and women differ in their relative
responses to sexual vs. emotional infidelity in a romantic relationship, a proposal based
upon the presumption of evolved sex differences in mating strategies. Because sexual
infidelity signals the possibility of being cuckolded and emotional infidelity signals the
possible withdrawal of resources, men are expected to be relatively more jealous over the
former whereas women are expected to be relatively more jealous over the latter.
Although past research has demonstrated robust support for the evolutionary hypothesis,
this support has been called into question on both empirical and conceptual grounds. In
this thesis,
I
test the claims of the evolutionary hypothesis using a novel experimental
paradigm designed to address previous methodological concerns. In two studies,
participants (college students in Study

1;
married adults in Study
2)
were asked to
consider a series of hypothetical scenarios depicting cues to either sexual or emotional
infidelity and to rate their responses on a variety of continuous measures. Across both
studies, as expected, women were found to be significantly more bothered by emotional
cues than were men. Contrary to expectations, however, women and men displayed
similar responses to sexual cues, with women having a slight tendency to be more
bothered than men. The latter effect was less consistent than the former, however, giving
rise to a reliable two
-
way interaction between sex and infidelity type (sexual vs.
emotional). Although these findings were only partly supportive of my predictions, they
nonetheless demonstrate that humans do display a sexually dimorphic jealousy response,
and provide support for the contention that this is an evolved mechanism. Discussion
addresses a number of additional findings, as well as avenues for future research.
Sex Differences in Jealousy 3
Contents

I
.
Introduction 4

I1
.
Study 1 30
-
Method

34

-
Results 38

-
Discussion -44

I11
.
Study 2 48
-
Method

49

-
Results 50

-
Discussion 54

IV
.
General Discussion -59

V
.
References
66

VI
.
Acknowledgements 71

VII
.
Appendix -72

VIII . Tables and Figures 82
Sex Differences in Jealousy
4
The Evolutionary Psychology of Jealousy in Romantic Relationships:
Evidence for a Sexually Dimorphic Response Mechanism in Humans
"O!
beware, my lord, of jealousy;
It is the green
-
eyed monster which doth mock
The meat it feeds on.
"
-
William Shakespeare,
Othello
Anyone who has ever been in a romantic relationship probably has a firsthand
understanding of the emotional distress that any threat to this relationship may cause. The
very suspicion that one's mate may be romantically interested in a rival is enough to
inspire a negative reaction in anyone, such as sadness, insecurity, hurt feelings, or anger
-
in some cases, even enough anger to attack one's partner or the object of their affection

(Daly
&
Wilson, 1988). All of these negative feelings that arise in response to a romantic
partner's past or imminent infidelity can be commonly understood as manifestations of
jealousy. Jealousy over romantic relationships is a universal phenomenon that can be
seen in men and women across cultures and age groups (Buunk, Angleitner, Oubaid
&
Buss, 1996; Buss, Shackelford, Kirkpatrick, Choe, Lim, Hasegawa, Hasegawa
&
Bennett,
1999; Wiederman
&
Kendall, 1999; Harris, 2003).
Evolutionary theorists have proposed that jealousy is a response mechanism that
has been naturally selected for in humans because over the course of evolution, those
who had a negative response to signs of a mate's infidelity would be more apt to prevent
this infidelity and thus experience higher rates of reproductive success (Wiederman
&
Allgeier, 1993). This explanation appears likely when we consider the fact that many
other animals have also evolved mechanisms to reduce the possibility that their mate will
engage in extradyadic copulation. For example, among
Plecia nearcticas,
aptly laown as
"
love bugs,
"
after a male gains access to a sexually receptive female he remains attached
Sex Differences in Jealousy
5
to her in a copulatory embrace that can last for days, thereby preventing other males from

fertilizing her eggs. Another species of insect, Johannseniella nitida, adopts an even
more drastic strategy: Males of that species actually leave their genitalia behind, broken
off inside the female, after insemination
-
presumably to seal off the female's
reproductive opening and thus forestall takeovers from rival
males (Buss, 2003).
Considering the fact that even lowly insects have evolved specialized behaviors to
prevent the likelihood of losing their mate to a rival, it is reasonable to
surmise that
humans have also evolved an innate sense of jealousy to accomplish the same purpose.
However, a specific aspect of the evolutionary theory of jealousy that has come under
intense scrutiny is the hypothesis that humans have evolved sexually dimorphic jealousy
mechanisms that cause males and females to be differentially upset by different types of
infidelity. Specifically, this theory postulates that men will be relatively more jealous
about their mate's sexual straying, whereas women will be relatively more jealous at the
prospect of their mate's emotional involvement with a rival (Buss, Larsen,
Westen
&
Semmelroth, 1992). A great deal of research has been done to investigate this question,
and while considerable support has been found for the hypothesis, there is much debate in
the field over whether jealousy is really an evolved sexually dimorphic mechanism, or if
gender differences in jealousy can be explained by other, more social or cultural
influences (Harris, 2003).
The goal of this thesis is to offer an unbiased examination of the arguments for
and against the
evoXutionary theory of human jealousy and to build on previous studies by
developing and implementing an experimental paradigm that addresses the limitations of
these studies.
I

ultimately intend to determine a) whether there really is a sex difference
Sex Differences in Jealousy
6
in jealousy response to different types of infidelity, and b) if so, whether this difference
can be attributed to evolved mechanisms or a more socially learned explanation.
Understanding the Evolutionavy Argument
Human males and females, over the course of evolutionary history, have
undoubtedly faced differing
environmelltal pressures and circumstances affecting mating
behavior. Some of the effects of these pressures can be seen in the sex
-
differentiated
psychological tendencies and preferences shown by men and women
-
for instance,
countless studies have demonstrated that men worldwide tend to be attracted to women
with physical attributes of youth and fertility, which indicate an ability to bear healthy
children, whereas women are more attracted to signs of social dominance and material
wealth, which are cues to a man's ability to provide resources for her offspring
(e.g.
Buss, 1989; Kenrick
&
Keefe, 1992). Attraction to certain characteristics in the opposite
sex is not the only way that past evolutionary pressures might manifest themselves in
human psychology; in order to preserve their chances of reproductive success, males and
females may have also needed to be sensitive to different types of cues that could indicate
a potential threat to their mating relationship. This idea is where the evolutionary theory
of jealousy has its origin.
For any species of
animal that employs internal fertilization, the males are faced

with a reproductive dilemma: It is impossible for a male to be certain that his mate's
offspring are the result of his own fertilization or a competitor's. Different species have
evolved different methods of countering this uncertainty of paternity, from
mate-guarding
behaviors to infanticide to simply impregnating as many females as possible (Buss
8&
Shackelford, 1997; Buss, 2003). As humans are a relatively monogamous species,
Sex Differences in Jealousy
7
meaning that a man's investment of time and resources in his mate tends to be high, it
would be especially costly to him if his mate were sexually unfaithful, because all of his
energy would be devoted towards raising offspring that did not carry his own genes. Thus
it follows that over the course of evolution, human males who were sensitive to
indications of their mate's sexual straying such that they could minimize or forestall it
altogether would have been more successful at passing down their genes than males who
were not bothered when their mate showed signs of sexual interest in another man.
Females, on the other hand, can be 100% certain that their offspring carries their
own genes, so this should not have been a primary concern for them. However, since the
female was the one who bore and consequently raised her children, she would have had a
much higher probability of success if she could be certain that her mate remained
emotionally devoted to her, providing his material support and protection for the
extended time that it takes to raise a child to adulthood.
A
man's brief sexual dalliance,
although potentially troubling, would not have been
harmful to his mate's reproductive
success so long as he stayed emotionally attached and willing to share his resources with
her. Women who were vigilant to signs that their mate's affection (and hence his
resources) was being diverted to another, such that they could minimize or forestall this
occurrence, therefore should have held an evolutionary advantage over those who were

oblivious to these indications of emotional interest and attention towards a competitor.
From the logic of these scenarios comes the evolutionary theory
ofjealousy: Although
members of both sexes should be upset by
any
type of infidelity on the part of their mate
since emotional and sexual infidelity frequently co
-
occur (Harris
&
Chistenfeld, 1996;
DeSteno
&
Salovey, 1996), men are predicted to be relatively more upset
by
sexual
Sex Differences in Jealousy 8
infidelity than women, and women are expected to be relatively more upset by emotional
infidelity than men (Buss et al., 1992).
The idea of increased male sexual jealousy has been supported with statistics on
morbid jealousy, spousal abuse, and homicide in response to sexual infidelity, all of
which are more commonly found in men than women (Daly
&
Wilson, 1988).
Additionally, it has been found that women place a much higher importance on emotional
investment in a relationship than men do; a man's ability to commit is a key factor in a
woman's assessment of his potential as a mate, and women who repeatedly engaged in
sex without a high degree of emotional investment were found to have heightened levels
of anxiety and distress despite holding the
"

liberal
"
opinion that uncommitted sex was
morally acceptable (Townsend, 1995). In another study, when 15 committed
undergraduate couples were asked to list as many jealousy
evoking situations as they
could think of, men were significantly more likely to list sexual scenarios, whereas
women tended to list situations in which their partner spent more time with or showed
emotional attachment to a potential rival (Francis, 1977, as cited in Wiederman
&
Allgeier, 1993).
Despite the fact that these and many other studies appear to support the
evolutionary hypothesis, we must be cautious in our interpretation of the data. Men do
commit more violent crimes than women in response to sexual infidelity, but this may
simply be a reflection of the fact that men commit more violent crimes
overall
than
women do. Also, as Harris points out in her 2003 review, people who are morbidly
jealous or who commit murder as a reaction to jealousy represent extreme,
possibly
mentally unstable cases, and
"
observations about population extremes are likely to offer
a
Sex Differences in Jealousy
9
very unreliable guide to species
-
typical characteristics
"

(p.
108).
In the case of the
undergraduate couples listing the predicted types of scenarios that would make them
jealous, this could just be because men tend to think about sex in general more than
women do, and women tend to think more about emotional issues, so these were the
exemplars that happened to spring to mind. Obviously, one cannot rely on homicide
statistics or general jealousy studies as proof of the evolutionary hypothesis
-
in order to
truly examine its worth, we must turn to studies that have been specifically designed to
test the question of whether the sexes differ in their sensitivity to infidelity, as predicted
by evolutionary theory.
Robust Support for the Evolutionary Theory
One groundbreaking study performed by Buss and his colleagues in 1992
provided direct support for the evolutionary theory of jealousy. In a simple experimental
design, Buss presented participants with two hypothetical scenarios. Participants were
asked to imagine that their romantic partner either formed a
"
deep emotional attachment
"
to or had
"
passionate sexual intercourse
"
with another person, and were then asked to
choose which of the two would upset or distress them more.
An
additional set of
questions asked whether their partner

"
trying different sexual positions
"
or
"
falling in
love
"
with another person would upset them more. The results fiom this simple, forced-
choice measure were as predicted by the evolutionary hypothesis: Significantly more men
than women chose the instance of sexual infidelity as more upsetting, for both sets of
scenarios.
In this same study, Buss also found preliminary physiological evidence
in
support of the hypothesis, reporting that men showed more electrodema1 activity
(EDA)
Sex Differences in Jealousy 10
and a higher pulse rate (PR) when imagining sexual infidelity than emotional infidelity,
whereas women displayed the opposite tendencies (Buss et al., 1992).
These findings in support of the evolutionary theory ofjealousy have been
replicated by many subsequent studies employing the same or similar forced
-
choice
methodology (Wiederman
&
Allgeier, 1993; Buunk et al., 1996; DeSteno
&
Salovey,
1996; Wiederman
&

Kendall, 1999; Buss et al., 1999; Cramer, Abraham, Johnson,
&
Manning
-
Ryan, 2001; DeSteno, Bartlett, Braverman,
&
Salovey, 2002; Harris, 2002;
Pietrzak, Laird, Stevens
&
Thompson, 2002; and for a meta
-
analysis, see Harris, 2003).
Different populations have been sampled, different cultures have been tested, and the
phrasing of the scenarios has been altered slightly, but across them all, one very robust
effect is that when people are asked to choose between the two different types of
infidelity, significantly more men than women indicate that sexual infidelity is worse, and
significantly more women than men choose emotional infidelity as worse. The reliability
of this finding is undeniable, even in samples from cultures as diverse as Germany, the
Netherlands, Sweden, Japan, and Korea (Buunk et al., 1996; Wiederman
&
Kendall,
1999; Buss et al., 1999). Similar results have also been found when asking men and
women which type of infidelity they would be more
likely to forgive, and which would
lead them to break up with their partner (Shackelford, Buss,
&
Bennett, 2002).
But the support for the evolutionary hypothesis isn't as clear
-
cut as it might first

appear. One finding that throws a snag into the evolutionary argument is the fact that,
though the percentage of men who are more upset by sexual infidelity is significantly
higher than the percentage of women who are, that number is still
frequently lower than
50%. This means that the majority of men in many samples chose emotional infidelity as
Sex Differences in Jealousy 1 1
more upsetting than sexual, even though the evolutionary theory seems to predict that this
should not be the case. Also, the variability of the percentages of men and women who
are more upset by one type of infidelity vs. the other across samples is disconcertingly
high. For instance, in one cross
-
cultural study,
76%
of the American male sample were
more upset by the sexual than the emotional scenario, whereas only 38% of the Japanese
male sample were (Buss et al., 1999). Obviously, culture and other factors do have a
substantial effect on which scenario a person is more likely to pick, and evolutionary
theorists are quick to point out that evolution does not give rise to
"
hard
-
wired,
"
inflexible tendencies, but rather predispositions that are subject to variation according to
cultural and situational variables (Wiederrnan
&
Allgeier, 1993). Moreover, the crucial
fact remains that in both of those samples, as well as in all of the other studies, the
percentage of the female equivalent that chose sexual infidelity as more distressing was
still significantly lower than the percentage of males who did. The key argument put forth

by the proponents of evolutionary theory is for a
sex difference
in jealousy, and this sex
difference is indisputably robust across virtually all studies replicating Buss's original
forced
-
choice method.
The Double
-
Shot Hypothesis: An Alternative Explanation
Though most researchers, using Buss's methodology, have had no trouble
replicating his findings of a significant sex difference in what type of infidelity
maltes
people more jealous, there has been significant controversy over how those findings
should be interpreted. A few psychologists, such as David
DeSteno and Christine Harris,
have proposed alternative, non
-
evolutionary explanations for these results, explanations
that make sense of the findings not by appealing to domain
-
specific evolved tendencies
Sex Differences in Jealousy 12
but to more domain
-
general human reasoning ability. The most well
-
known name for this
cognitive interpretation of Buss's data is the
"

double
-
shot hypothesis.
"
This hypothesis
asserts that men and women will be more upset about the type of infidelity that they
believe is most likely to imply that the other type of infidelity is occurring as well
(DeSteno
&
Salovey, 1996). In other words, women will be especially jealous about
emotional infidelity because they believe that if a man is in love with someone he is
likely to be having sex with her
(i.e. emotional and sexual infidelity), whereas it is
probable that a man will willingly have sex with someone without being in love with her,
(i.e. sexual without emotional infidelity). Conversely, according to this hypothesis, men
believe that it is possible for a woman to be emotionally attached to someone without
having sex, but if she is having sex with someone then she is probably also in love with
him (see Harris
&
Christenfeld, 1996, for a similar argument). Thus, regarding Buss's
results, it appears as though gender is the variable that affects jealousy, when in reality it
is individuals' rational beliefs about the opposite sex that cause them to show a
sex-
differentiated response.
In order to test this theory, a differential infidelity implication (DII) scale was
created by subtracting a participant's likelihood judgment that sexual infidelity would
also imply emotional infidelity
from their judgment that emotional infidelity would also
imply sexual infidelity (DeSteno
&

Salovey, 1996). Thus, a positive DII would indicate
that the participant believed that emotional infidelity implies sexual infidelity
more than
the other way around (expected for women), a negative DII would indicate the opposite
(expected for men), and a DII around zero would mean that the participant believed that
the two types of infidelity implied each other equally. The first study using this
scale
Sex Differences in Jealousy 13
found that, although sex initially did predict jealousy reactions, this effect was rendered
nonsignificant when DII was calculated and controlled for statistically. A participant's
DII could significantly predict which type of infidelity he or she would choose as worse
-
women tended to have a positive DII, with a majority choosing emotional infidelity as
worse, and men's
ID11 hovered around zero, with 49% choosing sexual as worse (DeSteno
&
Salovey, 1996). Another more recent study with a homosexual sample also found
support for the double
-
shot hypothesis using the DII scale
-
overall, jealousy seemed to
be more influenced by the sex of the unfaithful partner than by the sex of the subject
(Dijkstra, Groothof, Poel, Laverman, Schrier,
&
Buunk, 2001).
Evolutionary Response to the Double
-
Shot Argument
Proponents of the evolutionary theory of jealousy have put forth two main

arguments in response to the challenge of the double
-
shot hypothesis. First, they argue
that the double
-
shot hypothesis is incomplete in that it doesn't provide an account as to
why
men and women across cultures would have these sex differences in beliefs about
infidelity (Buss, Larsen,
&
Weston, 1996; Buss et al., 1999; Shackelford et al., 2002). It
is possible that these beliefs could be a result of differential socialization rather than
evolutionary pressures, but then why would all boys across cultures be taught to believe
that women were more likely to have sex only when they were already in love and vice
versa for girls? The evolutionary account provides an explanation for this, which is that
women must consider a man's willingness to commit to her and her children before she
will mate with him, whereas men are under no such evolutionary obligation and thus find
sex without emotional attachment much more palatable. Social
-
cognitive accounts such
Sex Differences in Jealousy 14
as the double
-
shot hypothesis cannot offer a parsimonious explanation for why people's
rational beliefs about infidelity would be so similar around the world.
Second, the DII scale used by DeSteno and his colleagues as evidence against the
evolutionary hypothesis may be misleading. Taking a factor that correlates highly with
sex, such as beliefs about infidelity, and examining its relationship to different jealousy
reactions does not prove a causal relationship. As Buss points out, one might perform the
same statistical procedure with such sex

-
related factors as height or testosterone levels
and come to the erroneous conclusion that those were in fact what caused the
demonstrated sex difference in jealousy (Buss et al, 1996). A simpler and more effective
way to test the validity of the double
-
shot hypothesis would be to repeat the forced
-
choice self
-
report methodology, but to use scenarios that either specified that each type of
infidelity was
exclusive
of the other, or scenarios in which
both
types of infidelity
occurred and participants were asked which aspect of the infidelity upset them the most.
Both of these suggestions were tested in subsequent studies, and the results in
general did not bode well for the double
-
shot hypothesis. When participants were asked
to choose whether sexual
without
emotional infidelity or emotional
without
sexual
infidelity was worse, and when participants were presented with a scenario of
both
sexual
and emotional infidelity and asked which

aspect
was worse,
in
all situations the
predictions of the evolutionary hypothesis were borne out. Men still chose sexual
infidelity as the most upsetting more often than women did, and vice versa (Buss et al.,
1999; Cramer et
al., 2001). When DII was calculated in these studies, using the same
methodology that DeSteno used, mixed results were found: in one sample,
DII alone was
a significant predictor of jealousy but when it was combined with sex it lost its
Sex Differences in Jealousy 15
significance (Buss et al.,
1999), and in another sample, DII was not statistically reliable
for men and marginally significant for women, but in the direction opposite from what
the double
-
shot would predict (Cramer et al., 2001). Clearly, results using the DII scale
are mixed and inconclusive at best, and findings from subsequent studies designed to
separate emotional from sexual infidelity more distinctly have found a sex difference that
cannot be explained by the double
-
shot hypothesis.
Alternative Methods of Measurement
The data that have been discussed up to this point can for the most part be
characterized as generally supportive of the evolutionary theory of jealousy, with a few
questions left unanswered (such as why, in some samples, the majority of men were more
upset by emotional infidelity, and to what extent variables like culture and beliefs can
affect evolved jealousy tendencies). However, there is one crucial limitation that applies
to all of the studies that have been discussed thus far: they have all employed the same,

simple, forced
-
choice, self
-
report methodology to measure jealousy. Though this method
does have its strengths, it also has significant limitations, one of which being that
forced-
choice scenarios are extremely artificial. In everyday life, no one could ever actually
choose
whether they would prefer their mate to have an exclusively sexual affair or an
exclusively emotional one. This may mean that the forced
-
choice methodology is a poor
measure of actual human jealousy, or that it measures a conscious, rational response
rather than an evolved
"
gut
"
reaction.
One
study designed to test this possibility imposed a cognitive load on
participants by asking them to remember a string of numbers while they considered the
two infidelity scenarios
(DeSteno et al., 2002). Since evolved mechanisms are supposed
Sex Differences in Jealousy 16
to be unconscious and automatic, if the evolutionary theory ofjealousy is correct then
undermining a person's conscious processing ability with a cognitive load should
increase or at least have no effect on the sex difference in jealousy responses. In actuality,
although the cognitive load manipulation had no effect on men's ratings of which type of
infidelity upset them more, it caused women to be more likely to list sexual infidelity as

more upsetting (65% of those under cognitive load vs. 36% of those in control condition;
DeSteno et al., 2002). These results call the forced
-
choice method of measuring jealousy,
as well as the evolutionary hypothesis itself, sharply into question. In order to redeem and
strengthen the evolutionary hypothesis, confirmatory data is needed from other methods
of measurement. If Buss's findings are limited to the forced
-
choice methodology, and are
observed only under conditions of minimal cognitive load, then they may reflect some
"
methodological artifact resulting from a specific and effortful decision strategy invoked
by the format of the question,
"
rather than an evolved tendency (DeSteno et al., 2002).
Results from Continuous Measures.
One potential alternative to the forced
-
choice
measure is asking participants to rate their emotional reactions to the infidelity scenarios
on a continuous, Likert
-
type scale. A number of studies have attempted to use this
methodology, but thus far these attempts have yielded mixed results. Most researchers
have found that when they ask participants to rate their level of upset or distress from
Buss's scenarios on a continuous scale, there is no significant interaction between gender
and infidelity type, and hence no sex difference in jealousy reactions to the two types of
infidelity
(Wiederman
&

Allgeier, 1993; DeSteno
&
Salovey, 1996; DeStens et
aP.,
2002;
Sabini
&
Green, 2004). In response to these findings, Buss argues that continuous
measures are not reflective of evolutionary sex differences because sf the
"
ceiling
Sex Differences in Jealousy
17
effect.
"
This means that men and women respond to both infidelity scenarios at a very
high level ofjealousy
(i.e., they rate themselves as
"
extremely upset
"
for both), because it
is evolutionarily advantageous for men and women to have a jealous reaction to both
types of infidelity since they both signal threats to one's relationship and they tend to
overlap so much in real life (Buss et al, 1999). Thus, because ratings of both types of
infidelity cluster near the upper range of the response scale, it is difficult if not impossible
to determine which of the two is relatively worse for men vs. women unless participants
are forced to choose between them.
A
few recent studies have gathered data from continuous measures that appear to

support the evolutionary hypothesis, but these studies can be called into question on
methodological grounds. For instance, one study found a significant sex difference in the
predicted direction on a continuous measure, but the continuous ratings were taken
after
the forced
-
choice question was asked, so subjects' responses could easily have been
influenced by what they had previously seen (Sagarin, Becker, Guadagno, Nicastle,
&
Millevoi, 2003). Another study found a significant interaction between gender and
infidelity type, but the phrasing of the question was different
-
instead of rating how
"
upset or distressed
"
they would feel, participants were asked to indicate how "jealous"
they would be, which is a different question than Buss originally asked (Becker, Sagarin,
Guadagno, Millavoi,
&
Nicastle, 2004). These findings are too inconsistent to conclude
that the evolutionary hypothesis is supported by data from continuous measures of
jealousy. This method of measurement does hold promise, but nothing conclusive can be
demonstrated until the problem of the
"
ceiling effect
"
is dealt with.
Sex Differences in Jealousy
18

Results from Physiological Measures.
One type of reaction that is not subject to
self
-
presentation bias or altered by rational beliefs is the human body's unconscious
physiological response to certain stimuli. Many jealousy studies, including the original
one by Buss and his colleagues in 1992, have measured participants' physiological
responses to imagined infidelity situations, in an attempt to determine whether the
jealousy manifests itself physically in the sex
-
differentiated manner hypothesized by
evolutionary psychologists. Unfortunately, a brief review of these studies will show that
data from physiological measures of jealousy are even more mixed and inconclusive than
those from continuous measure studies, and it is questionable whether this method is even
an accurate measure of jealousy at all.
Buss and his colleagues chose to measure participants' electroderrnal activity
(EDA), pulse rate (PR), and electromyographic activity (EMG) as methods of assessing
their physiological arousal in response to situations of imagined infidelity (1992). They
found that men showed greater EDA for sexual infidelity than for emotional, women
showed greater EDA for emotional infidelity than for sexual, men had a higher PR for
sexual infidelity than emotional, women displayed no difference in PR for either
infidelity scenario, and EMG data showed no significant effects for men or women (Buss
et al., 1992). These data were interpreted as generally supportive of the evolutionary
hypothesis, but Buss overlooked a number of problems. First of all, the EDA readings
showed extreme variance, which sheds some suspicion on their value as a method of
measurement (for example, the mean EDA reading for men imagining the sexual
infidelity was 1.30 with a standard deviation of 3.64). Secondly, rather than comparing
the measures between sexes, Buss considered them separately for each sex
(e.g. men's
Sex Differences in Jealousy

19
EDA from
sexual
infidelity vs. men's EDA from
emotional
infidelity), which is
inconsistent with the method he used to analyze the forced
-
choice responses (e.g.
percentage of
men
who picked sexual as worse vs. percentage of
women
who picked
sexual as worse). Lastly, and perhaps most importantly, there is no guarantee that these
physiological measures have anything to do with jealousy
-
they are only assessing the
body's level of arousal, which could be due to jealousy or perhaps to a different sort of
sensation, such as the physical arousal one might experience when fantasizing about any
sort of sexual, act, not specific to infidelity (Grice
&
Seely, 2000).
More physiological studies followed the first one, attempting to address some of
these limitations and increase power by testing more subjects. The results of these studies
varied wildly, from supporting the evolutionary hypothesis on all measures (Piertzak et
al.,
2002), to finding no sex difference in arousal (Harris, 2000), to finding some results
that were the exact opposite of what the evolutionary theory would predict (Grice
&

Seely, 2000). In one all
-
male sample, participants were aslted to imagine a baseline
sexual scenario and a baseline emotional scenario as well as the sexual and emotional
infidelity scenarios, and there was no difference between physiological response to the
baseline sexual and the sexual infidelity
-
in other words, imagining a sexual situation
was more arousing to men than imagining an emotional situation, regardless of whether
that situation involved infidelity or not (Harris, 2000). A reasonable interpretation of
these mixed results may thus be that this method of measurement is simply not useful for
assessing sex differences in jealousy.
Reactions to Actual vs. Hypothetical Infidelity.
One common theme thoughout
all of these different methods of measuring jealousy
is
that they all rely on the
Sex Differences in Jealousy
20
participants' ability to imagine a hypothetical situation of infidelity and react to it as if
they were actually experiencing it. Clearly, this is not an optimal way of measuring
someone's actual jealousy reaction because there is no way to know whether they would
react in a similar fashion if the infidelity actually happened to them. Ethical concerns
prohibit studies that would put participants in a situation where they honestly believe that
their partner is cheating on them, so the only other possibility for examining people's
responses to actual infidelity is to find a sample of participants who have been cheated on
in the past, and ask them to indicate retrospectively which aspect of the infidelity
bothered them more.
A
review of self

-
report studies with participants who had been victims of
infidelity in the past shows that overall there is no evidence for a sex difference in which
aspect of the infidelity causes the most distress (Harris,
2003).
Males and females,
including homosexual as well as heterosexual participants, both tend to focus on the
emotional aspects of their partner's unfaithfulness when recollecting it. However, it can
be argued that this method of assessing jealousy is even more dubious than asking for
reactions to hypothetical infidelity. The correlational nature of these data
(i.e.,
participants are not assigned to conditions randomly and there is every possibility that
certain types of people are the victims of infidelity more than others) renders causal
conclusions impossible. Moreover, retrospective reports such as these are notoriously
unreliable and subject to any number of biasing influences. Finally, predictions derived
from evolutionary theory are most relevant to individuals' tendency to become jealous
over potential threats to their relationship, not to their emotional responses
after
the
infidelity situation has already come and gone. This being the case, until someone can
Sex Differences in Jealousy
21
design a study that examines immediate reactions to actual infidelity without arousing
ethical concerns, hypothetical situations of infidelity are the best method available for
assessing jealous responses.
Potential Moderating Factors
As we have seen, there are many factors other than gender that can have a
significant effect on a person's feelings ofjealousy. It is important not to disregard these
factors when considering the data that have been collected on the subject thus far. If the
sex difference found by Buss is not universal, that is, if one of these factors is strong

enough to erase it entirely, then this throws the validity of the evolutionary hypothesis
into question.
An
evolved tendency should be present in all samples, regardless of other
influences such as age or culture.
Age/Marital Status as a Moderating Factor.
The majority of evolutionary
jealousy studies have been performed with American undergraduate students as subjects.
The sex difference in response to forced
-
choice questions is robust within this population,
but as the age of participants increases, the data tend to become somewhat less reliable.
Considerably fewer studies have been done with older participants, so it is difficult at this
point to ascertain whether age is a significant moderator of jealousy.
A
meta
-
analysis of
the few studies that have included participants older than
25
years shows that age does
indeed have a significant effect on which type of infidelity people indicate is worse
-
the
sex difference is stronger for college students than for older participants (Harris,
2003).
Importantly, though, the sex difference is still significant for these older samples.
One notable factor
associated with age is the marital status of the participants.
Only one study has been done to examine the effect of marital status on forced

-
choice
Sex Differences in Jealousy 22
jealousy responses, and this has provided some intriguing results (Voracek, 2001). A
community sample of Austrian adults, with an average age of 32 and an age range of 18
-
58, was provided with Buss's original forced
-
choice infidelity items (
"
passionate sex
"
vs.
"
deep emotional attachment
"
) and asked to choose which would upset or distress them
more. Of those who were unmarried but in a committed relationship, 58.5% of men and
28.3% of women indicated that the sexual infidelity would be worse, a significant
difference. The married sample, on the other hand, displayed almost
no sex difference at
all: 27.1% of men and 27.7% of women indicated that sexual infidelity would be worse.
This striking result raises some interesting questions. Can marriage eliminate the
robust sex difference that has been found across countless samples? If so, does this cast
doubt on the notion that the sex difference is an evolved tendency? Clearly, further study
of married samples is warranted, but even if this result were to be replicated and marriage
were found to eliminate the sex difference in jealousy, caution should be taken before
completely ruling out the evolutionary hypothesis. Marriage is a social contract that binds
two people together for life and is very difficult to break. Consequently, married
individuals have a much greater assurance that their relationship will be safe against

rivals. This assurance is purely a result of the contract and all of the societal regulations
that accompany it
-
the institution of marriage would not have existed early on in human
evolutionary history, so our ancestors would not have had this type of guarantee. It could
be that, once an individual is secure in the knowledge that it would be very difficult for a
rival to threaten his or her relationship, evolved gut reactions of jealousy are subsumed
by this assurance and replaced by more rational or cognitively
-
influenced answers.
Unmarried couples are therefore likely to be a better source for testing evolutionary
Sex Differences in Jealousy 23
theories, because their relationships are not protected by the cultural shield of wedding
vows. However, further study of married samples could still provide valuable insight into
whether there truly are evolved sex differences in jealousy.
Cultural Differences as a Moderating Factor.
One of the key results needed to
support a theory of evolved psychological differences is that the findings must replicate
in cross
-
cultural studies. For the most part, the evolutionary theory of jealousy has
performed well on this test. The forced
-
choice methodology has been employed with
undergraduate samples in the U.S., Germany, the Netherlands, Sweden, Japan, and
Korea, and every single experiment has found a significant sex difference overall in
which type of infidelity is more upsetting (Buunk et al., 1996;
Wiederrnan
&
Kendall,

1999; Buss et al., 1999). A meta
-
analysis of the data from these studies grouped by
region
(US,
Asia, and Europe) found that this variable was not a significant moderator of
gender differences in jealousy (Harris, 2003). Culture certainly does influence
participants' overall pattern of responses
-
for instance, a higher percentage of American
men than Japanese men chose sexual infidelity as the most upsetting (76% vs.
38%),
but
this trend was the same for American and Japanese women as well, so across cultures, the
sex difference remains robust (Buss et al., 1999).
Sexual Orientation as a Moderating Factor.
Homosexuality has always posed an
interesting conundrum for evolutionary psychologists, because as of yet there is no
convincing Darwinian explanation for why some people would find themselves attracted
to members of their own gender with whom they could not reproduce
(Muscarella,
11
999).
Due to this fact, most evolutionary studies of sex differences tend to exclude data from
homosexual individuals because they complicate the hypothesis. It is difficult to know
Sex Differences in Jealousy 24
what to conclude from these data, but there have been a few evolutionary studies of
jealousy using homosexual samples, and it is important not to disregard or overlook this
unique subset of the population.
A meta

-
analysis of data from studies with homosexual and heterosexual samples
reveals that, predictably, there is a stronger sex difference in jealousy for heterosexual
participants than for gays or lesbians (Harris, 2003). One study found a main effect of
sexual orientation such that lesbians and gays were more likely than straight men and
women to choose emotional infidelity as worse, which is consistent with previous
findings that sexual exclusivity in a relationship is in general more important to
heterosexual men and women than homosexuals (Harris, 2002). Disregarding this main
effect, however, there was still
a
sex difference such that more gay men than lesbians
chose sexual infidelity as the worse of the two (though the data was not shown in any
form other than a bar chart, so it was impossible to tell if this difference was statistically
significant). Another study found the exact opposite result, that
5
1
%
of lesbians and 32%
of gay men rated sexual infidelity as the most upsetting, but findings from this study were
suspect for two reasons: the sample was recruited entirely from gay bars in several Dutch
towns, so 37% of participants were under the influence of alcohol when they participated
in the study, and there was no heterosexual control group with which to compare their
data (Dijkstra et al., 2001).
From the insufficient amount of data that has been accrued using homosexual
samples, the most solid conclusion one can come to is that the evolutionary theory
of
jealousy overall lacks support among this population. This is not surprising, given the
fact that homosexuality itself is a phenomenon that still defies evolutionary explanation.
Sex Differences in Jealousy
25

Thus,
I
must do as other evolutionary theorists have done and set this challenge aside for
the moment, but this is certainly an area that would benefit greatly
from further research
in the future.
Jealousy as a Speczfic vs. Broad Mechanism
Evolutionary psychology is a notoriously controversial area of study (Gould,
1991
;
Eagly
&
Wood, 1999) and this controversy has been played out in the literature on
jealousy. Many opponents of the evolutionary theory of jealousy argue against it based on
a particular interpretation of what the theory maintains. Some critics have adopted a very
specific, narrow, rather simplistic definition that can be referred to as
"
JSIM,
"
the
"
jealousy as a specific innate module
"
definition. According to
JSIM,
"
because sexual
jealousy arose as a solution to cuckoldry, and emotional jealousy arose as a solution to
resource loss, each problem would have been unique to each sex,
"

thus leading to the
formation of jealousy as a distinct module in the brain that is different for men and
women (Harris,
2003).
This model leaves very little room for conscious thought or
variation in behavior; all humans are expected to respond to infidelity in a certain
predetermined manner based on the evolved tendencies of their gender. Clearly, if this
were what was meant by the
"
evolutionary theory ofjealousy,
"
then it would have
already been disproven many times over by the fact that jealousy varies by culture, that
men are frequently more distressed by emotional than sexual infidelity, that married
subjects show less of a sex difference than unmarried subjects, etc.
The goal of evolutionary psychology is not
to distill human behavior into a
catalog of instincts and mechanical responses bred into us by natural
selection, but
to
examine how humans interact with thelr environment and determine what role,
if
any,