Testing the Affiliation Hypothesis of Homoerotic Motivation in Humans: The Effects of Progesterone and Priming

Abstract

The frequency of homoerotic behavior among individuals who do not identify as having an exclusively homosexual sexual orientation suggests that such behavior potentially has adaptive value. Here, we define homoerotic behavior as intimate erotic contact between members of the same sex and affiliation as the motivation to make and maintain social bonds. Among both male and female nonhuman primates, affiliation is one of the main drivers of homoerotic behavior. Correspondingly, in humans, both across cultures and across historical periods, homoerotic behavior appears to play a role in promoting social bonds. However, to date, the affiliation explanation of human homoerotic behavior has not been adequately tested experimentally. We developed a measure of homoerotic motivation with a sample of 244 men and women. Next, we found that, in women (n = 92), homoerotic motivation was positively associated with progesterone, a hormone that has been shown to promote affiliative bonding. Lastly, we explored the effects of affiliative contexts on homoerotic motivation in men (n = 59), finding that men in an affiliative priming condition were more likely to endorse engaging in homoerotic behavior compared to those primed with neutral or sexual concepts, and this effect was more pronounced in men with high progesterone. These findings constitute the first experimental support for the affiliation account of the evolution of homoerotic motivation in humans.

Full text

Testing the Affiliation Hypothesis of Homoerotic Motivation in Humans: The Effects of
Progesterone and Priming
DIANA S. FLEISCHMAN, DANIEL M.T. FESSLER AND ARGINE EVELYN
CHOLAKIANS
Diana S. Fleischman: University of Portsmouth Department of Psychology, King Henry Building
King Henry I Street Portsmouth Hampshire PO12DY United Kingdom Email:
Diana.Fleischman@port.ac.uk
Daniel M.T. Fessler: Center for Behavior, Evolution, & Culture and Department of
Anthropology, 341 Haines Hall Box 9511553 University of California, Los Angeles, Los
Angeles, CA 90095-1553
Argine Evelyn Cholakians: Department of Anthropology, California State University Fullerton
800 N. State College Boulevard, Fullerton, CA 92831-3547
Acknowledgements
The authors would like to thank Lianna Burton, Thomas Butler, Brian Fredrickson, Jennifer
Gabrysh, Danielle Kaplan, Aspen Lewis and Vera Wegricht for collecting data. This research
was supported by a UCLA Council on Research grant to the second author.
Accepted for publication in Archives of Sexual Behavior

2
ABSTRACT
The frequency of homoerotic behavior among individuals who do not identify as having an
exclusively homosexual sexual orientation suggests that such behavior potentially has adaptive
value. Here we define homoerotic behavior as intimate erotic contact between members of the
same sex and affiliation as the motivation to make and maintain social bonds. Among both male
and female nonhuman primates affiliation is one of the main drivers of homoerotic behavior.
Correspondingly, in humans, both across cultures and across historical periods, homoerotic
behavior appears to play a role in promoting social bonds. However, to date, the affiliation
explanation of human homoerotic behavior has not been adequately tested experimentally. We
developed a measure of homoerotic motivation with a sample of 244 men and women. Next, we
found that, in women (n= 92), homoerotic motivation was positively associated with
progesterone, a hormone that has been shown to promote affiliative bonding. Lastly, we explored
the effects of affiliative contexts on homoerotic motivation in men (n=59), finding that men in an
affiliative priming condition were more likely to endorse engaging in homoerotic behavior
compared to those primed with neutral or sexual concepts, and this effect was more pronounced
in men with high progesterone. These findings constitute the first experimental support for the
affiliation account of the evolution of homoerotic motivation in humans.
KEY WORDS: homosexual behavior; homoerotic behavior; affiliation; progesterone;
priming.

3
INTRODUCTION
Many individuals appear capable of sexual arousal toward members of the same sex
under the right circumstances (e.g., during adolescence; in initiation rites, see Muscarella, 2006),
and the ethnographic record indicates that same-sex sexual behavior is found in a majority of
societies, spanning a wide range of social complexity, and at times occurring at fairly high
frequencies (Kirkpatrick, 2000; Muscarella, 2007; Ross & Wells, 2000; see also Rind & Yuill,
2012). In laboratory studies, self-identified heterosexual women, on average, have been shown to
have a generalized genital response to both sexes, while heterosexual men have been shown to
have a modest genital response to male sexual stimuli (e.g., Chivers, Rieger, Latty, & Bailey,
2004; Chivers, Seto, Lalumiere, Laan, & Grimbos, 2010; Rieger, Chivers, & Bailey, 2005). This
has led a number of investigators to propose that same-sex sexual behavior may have some direct
adaptive value (Kirkpatrick, 2000; Muscarella, Cevallos, Siler-Knogl, & Peterson, 2005).
Before examining the adaptationist thesis, because investigations of same-sex sexual
behavior often use inconsistent terminology (Muscarella, 1999), we must first define the question
at issue. Sexual orientation is commonly understood as patterned sexual desire that may lead to
sexual behavior with members of one or the other sex. However, consonant with the functional
explanations to be discussed below, much, if not most, same-sex sexual behavior occurs in those
who do not identify as exclusively homosexual in orientation (Diamond & Wallen, 2011;
Kinsey, Pomeroy, & Martin, 1948; Kirkpatrick, 2000). Our theoretical focus here is on a broad
category of behavior, that which is erotic and occurs with members of the same sex, and its
hormonal and motivational underpinnings (which include, but are not limited to, sexual desire);
we consider sexual orientation or identity only secondarily, as this is not isomorphic with
behavior. Muscarella (2000) defined homoerotic behavior as “same-sex sexual behavior

4
involving genital contact” (p. 52); however, as not all actions having an erotic component
involve genital contact, we adopt a broader definition, construing as homoerotic behavior all
intimate contact, be it intentionally or unintentionally erotic, involving members of the same sex,
regardless of whether it involves genital contact.
Motivation to affiliate, broadly construed, has been implicated in homoerotic behavior in
both male and female nonhuman primates (Vasey, 1995; Wallen & Parsons, 1997). Same-sex
sexual contact has been found in 67% of cultures in the Human Relations Area Files
(Kirkpatrick, 2000), and the practice of such behavior by individuals who also engage in
opposite-sex sexual contact has been prominent in many culture regions for centuries (Ross &
Wells, 2000). Working in parallel, drawing on cross-species comparisons and extensive
ethnographic and historical materials, Kirkpatrick (2000) and Muscarella (2000) presented
overlapping evolutionary models arguing that human same-sex sexual behavior serves the
adaptive function of promoting alliance formation and maintenance (Muscarella et al. 2005; see
also Rind & Yuill, 2012).
Reflecting its central role in biological fitness, a strong neurological reward system
undergirds sexual behavior, hence, it stands to reason that bonds, including between those of the
same sex (homosocial), can be strengthened via sexual behavior. The affiliation hypothesis thus
proposes that natural selection co-opted this reward system as a means of promoting same-sex
social bonds. In small-scale societies that resemble those characteristic of ancestral human
populations, social bonding and alliance formation play a number of critical roles, including
advantage in violent inter-coalition conflict (Bowles, 2009; Van Vugt, 2009), buffering against
food shortfalls (see Hill & Hurtado, 2009), alloparenting (Hrdy, 2009), and insuring against
illness and injury (Sugiyama, 2004). Moreover, while some aspects of these patterns are uniquely

5
human, the fundamental link between social connectedness and fitness has a very deep
phylogeny, as there is evidence that, among both male and female nonhuman primates, social
bonds translate into enhanced survival and reproductive success (Kuhle & Ratdke, 2013; Silk,
Alberts, & Altmann, 2003; Silk et al., 2010). Consistent with the above patterns, in the clinical
context, social connectedness has been shown to importantly influence both overall human
health and prognosis after physical trauma (see Reblin & Uchino, 2008; Seeman, 1996).
Likewise, in keeping with the adaptive value of strong social bonds, major physiological systems
regulate both aversion towards cues of rejection and motivation towards affiliating with others.
Aversion to cues of rejection has been linked to the pain system, fundamental to self-
preservation (DeWall et al., 2010; Eisenberger, Jarcho, Lieberman, & Naliboff, 2006;
Eisenberger & Lieberman, 2004; MacDonald & Leary, 2005). Conversely, affiliating with others
and engaging in cooperative exchange activates neural reward centers in the brain (Bora, Yucel,
& Allen, 2009).
Although over a decade has passed since publication of the principal papers laying out the
affiliation explanation, empirical investigations addressing it remain scarce. Recently, an
interview study found that 89% of young British men surveyed reported having kissed another
man on the lips, with the majority stating that this activity was homosocial (i.e., affiliative)
(Anderson, Adams, & Rivers, 2010). Only one study has attempted to experimentally test the
affiliation hypothesis. Muscarella et al. (2005) found that U.S. undergraduates who read
vignettes about heterosexuals engaging in homosexual behavior with a positive outcome viewed
the protagonist as having higher status and more reproductive opportunities than those in the
neutral or negative outcome conditions. However, this study examined only third-party
perceptions of the consequences of same-sex sexual contact, but did not test the core prediction

6
stemming from the affiliation hypothesis, namely that affiliative and homoerotic motivations
should be functionally linked. Here, we aim to do exactly that. We adopt a multi-pronged
strategy, examining hormonal correlates of affiliation in both sexes, and exploring the effects of
priming affiliation in men.
Progesterone
Progesterone, produced mainly in the ovaries of women and the adrenal glands of men, is
one of the principal hormones underlying parental behavior (Mehta & Josephs, 2010; Wirth,
Meier, Fredrickson, & Schultheiss, 2007). Progesterone has been implicated in the psychology of
affiliation and in response to rejection. In a repeated-measures study of women, Miller (2011)
found that attention to social stimuli both increased during the high-progesterone phase of the
menstrual cycle and was correlated with salivary progesterone. Brown et al. (2009) found that
women engaging in a task promoting social closeness showed increases in progesterone, and the
extent of this increase predicted altruistic motivation in participants towards their task partner
one week later. In both men and women, following the experience of experimental social
rejection, progesterone increased when participants were told that they would have an
opportunity to affiliate with others (Maner, Miller, Schmidt, & Eckel, 2010).
However, not all studies linking progesterone and affiliation have found consistent
results. Heterogeneous findings have resulted from studies in which affiliation was measured
through a picture story exercise (Schultheiss, Dargel, & Rohde, 2003; Schultheiss, Wirth, &
Stanton, 2004; Wirth & Schultheiss, 2006) and evoked through film clips (Schultheiss et al.,
2004; Wirth & Schultheiss, 2006), especially in male participants. This may partly reflect the use
of mating opportunities as an affiliative prime, a manipulation that constitutes a confound (see
below). Although results are not entirely uniform across methods, there are grounds for

7
concluding that progesterone is positively associated with affiliation in humans. As our goal here
was to explore the hypothesis that homoerotic behavior serves an affiliative function, we
therefore investigated whether progesterone can also predict the motivation to engage in such
behavior.
Importantly, whereas progesterone has been positively associated with affiliative
motivation in humans, it has not been positively associated with sexual motivation; indeed, there
are suggestions of a negative association. Some research has shown a decrease in sexual function
in women on hormonal contraceptives containing synthetic progestins (Adams, Gold, & Burt,
1978; Wallwiener et al., 2010; but see also Alexander, Sherwin, Bancroft, & Davidson, 1990;
Guida et al., 2005). Likewise, although the degree to which circulating progesterone affects
sexual motivation in men has not been adequately explored, synthetic progesterone injections
have been used to reduce the testosterone and sex drive of male sexual offenders (Andersen &
Tufik, 2006). Hints of a negative association between progesterone and sexual motivation thus
suggest that evidence of a positive association between progesterone and homoerotic motivation
is not explicable as a by-product of an increase in sexual motivation that would enhance
attraction to both same and opposite sex others.
If homoerotic motives function to enhance affiliation, then, given the association between
progesterone and affiliation, we can expect that progesterone should be positively correlated with
homoerotic motives. In women, position in the menstrual cycle is a principal determinant of
progesterone level; hence, we can expect female homoerotic motivation to exhibit corresponding
cyclic variation. Importantly, however, a second factor may independently influence changes in
homoerotic motivation. A fundamental characteristic of the menstrual cycle is the concentration
of fertility (i.e., the probability of conception following coitus) during the periovulatory period

8
(Wilcox, Dunson, Weinberg, Trussell, & Baird, 2001). Given that all behaviors necessarily entail
opportunity costs, we can therefore expect that, if they are governed by adaptations, motivations
to engage in behaviors that are not tied directly to conception will exhibit a periovulatory
reduction. For example, in many mammals, including humans, caloric intake is reduced around
ovulation, thereby reducing foraging time in a manner that frees up time for mate-seeking and
mating (Fessler, 2003). By the same token, independent of conscious intentions as regards
reproduction, we might expect women to prioritize opposite-sex sexual contact over same-sex
sexual contact during the fertile window.
Consonant with this thesis, in a repeated-measures study over one menstrual cycle, L.M.
Diamond and Wallen (2011) found that, in a subset of non-lesbian women who felt that there
was a role for choice in their sexual desires, homosexual fantasies and desires decreased around
ovulation. Progesterone peaks during the luteal phase, approximately one week after ovulation
(Hatcher & Namnoum, 2004), when probability of conception has declined dramatically (Wilcox
et al., 2001). Hence, it is important to determine whether a mid-luteal rise in homoerotic
motivation reflects (1) the enhancing effects of progesterone as a means of achieving affiliation
or (2) declines in the costs of behaviors that compete with mating for time and energy. More
generally, the same reasoning regarding opportunity costs suggests that, in both women and men,
cues indicating the presence of favorable mating opportunities may modulate homoerotic
motivation, as evolved mechanisms can be expected to prioritize the direct fitness benefits of
potentially conceptive behavior with the opposite sex over the indirect fitness benefits of
affiliative bonding; conversely, cues of the importance of affiliation in the current context should
have the opposite effect. Accordingly, experimentally manipulating the presence of such cues
should, respectively, decrease or increase homoerotic motivation.

9
To summarize the above, we predicted that, if, in both sexes, affiliation is an important
ultimate function of homoerotic motivation, then 1) progesterone, a hormone associated with
affiliative motivation, will be positively correlated with homoerotic motivation; 2) priming
affiliative motivation will increase self-reported homoerotic motivation; 3) in women,
homoerotic motivation will decrease as a function of the probability of conception; and 4)
priming opposite-sex sexual opportunities will decrease homoerotic motivation.
Testing the above predictions requires a measure of homoerotic motivation. Existing
instruments designed to measure sexual orientation (e.g., the Kinsey Scale ) (Kinsey et al., 1948)
are unsuitable, as they examine past behavior, not current motivation. Although L.M. Diamond
and Wallen (2011) developed an instrument (modeled after Spector, Carey, & Steinberg, 1996)
intended to examine female homoerotic motivation, responses were dependent on participants’
social circumstances on the day of participation. For both studies, we develop a new measure of
motivation to engage in homoerotic behavior. In Study 1, we administer this measure to women
in conjunction with progesterone assays and assessments of menstrual cycle position. In Study 2,
we measure progesterone in men, then prime participants using word puzzles addressing
affiliation, opposite-sex sexual opportunities, or neutral concepts, and then measure homoerotic
motivation.
METHOD
Development of the measure of homoerotic motivation
In order to develop a new measure of homoerotic motivation we collected responses to an online
questionnaire and factor analyzed those questions addressing the phenomenon in question.
Participants and Recruitment

10
Participants were recruited via advertisements posted in the Volunteers section of
craigslist.org and homoads.org, the latter venue being employed in order to increase
representation of individuals who identified as other than exclusively heterosexual.
Advertisements invited volunteers to participate in an online study described as a short survey
for men and women 18+ years of age about attitudes towards same-gender behaviors. A total of
1028 individuals participated. After excluding 67 participants who did not complete the entire
questionnaire, and one participant under age 18, we were left with 960 participants (69%
women), age 18 to 79 years (M = 32.71, SD = 12.81). Initially, we did not ask participants to
report their sexual orientation, but we later added the Kinsey scale (Kinsey et al., 1948), which
was administered to 244 participants. This sample of 244 was 38.5% exclusively heterosexual,
16.8%.heterosexual with incidental homosexual contact, 6.6% heterosexual with more than
incidental homosexual contact, 9.4% bisexual, 6.6% homosexual with more than incidental
heterosexual contact, 10. 7% homosexual with incidental heterosexual contact and 11.5%
exclusively homosexual.
Materials
The questionnaire consisted of 16 statements, 5 of which directly addressed homoerotic
motivation. Four of these statements employed the present tense and addressed ongoing
experience (e.g., “The idea of kissing a [wo]man seems sexually arousing to me”). One statement
employed the past tense (“I have fantasized recently about having sexual contact with a
[wo]man”). We included this recent experience item on the basis of the logic that, whatever the
actual frequency with which such fantasizing occurred, participants would be more likely to
recall and acknowledge having had such fantasies if their current state of mind included positive

11
homoerotic motivation. For each statement, participants indicated their level of agreement using
a 5-point Likert scale (1, strongly disagree, to 5, strongly agree).
Factor Analysis
Factor analysis of the homoerotic motivation questions revealed that all five questions loaded
onto a single factor; alphas were acceptable for all participants (α = 0.87), for males alone (α =
0.90), and for females alone (α= 0.85) (see Appendix A for questionnaire items and individual
item loadings). Furthermore, the homoerotic motivation measure correlated significantly with the
Kinsey score for all participants, r(242) = .74, p < .001, for males alone, r(69) = .89, p < .001,
and for females alone, r(171) = .70, p < .001. Satisfied that our instrument successfully
addressed the topic at issue, we then proceeded to employ it in testing predictions derived from
the affiliation hypothesis.
Study 1- Progesterone’s association with homoerotic motivation in Women
Participants
A total of 194 female participants, age 17 to 22 years (M = 18.57, SD = 0.83) were
recruited from the Department of Psychology subject pool at a large university in the
southwestern U.S. Of the 90 women who reported being sexually active, two women reported
only engaging in homosexual sex, 88 reported only engaging in heterosexual sex, and none
reported sexual activity with both men and women. Of the 194 participants, 67 reported using
hormonal birth control. Among the 127 women not taking hormonal birth control, we excluded
those for whom we did not obtain a valid progesterone value, those without a valid cycle day (no
cycle day reported or a cycle day over 30), and those who did not complete the requisite
questions, leaving 92 naturally-cycling women. Women taking hormonal contraceptives have

12
lower levels of endogenous natural progesterone (Fleischman, Navarrete, & Fessler, 2010; Frye,
2006). While it is possible that the relevant variable for these women is the sum of endogenous
progesterone and exogenous synthetic progestins, the hormone assay that we employed only
measures the former. As a consequence, among our participants taking hormonal contraceptives,
the majority exhibited progesterone values below the minimum sensitivity (54%) of our assay.
We included all 67 participants taking hormonal contraceptives who completed the entire
questionnaire and did not use assay failure as an exclusion criterion. The mean ages of the
naturally-cycling and hormonal-contraceptive-using subsamples did not differ significantly from
that of the total sample.
Measures
Progesterone was assayed using enzyme-linked immunosorbent assay (ELISA)
(Salimetrics, State College, PA). The sensitivity of the assay, defined as the lower limit of
detection that can be distinguished from the zero standard, was 5 pg/mL. Mean inter- and intra-
assay coefficients of variation for the participants retained for analysis were 14% and 3.6%,
respectively.
Estimation of Conception Risk
Using published data (Wilcox et al., 2001), we calculated conception risk via either
backward counting from an actual date of next menstruation or forward counting from a previous
date of menstruation. Backward counting is more reliable for predicting date of ovulation
(Haselton & Gangestad, 2006); hence, we used it for all participants who returned a postcard
with their date of subsequent menses (44 out of 92 participants). For those participants lacking
actual date of next menstruation, we used forward counting (Gangestad & Thornhill, 1998).

13
Conception risk values ranged from 0 to .08 (M = 0.03; Mdn = 0.02). For more explanation of
this method and its validation see Navarrete, Fessler, Fleischman, and Geyer (2009).
Procedure
Upon arrival, participants salivated into vials, which were then sealed and frozen at −20
°C until analysis. Participants completed several computerized instruments in a single sitting,
and were randomly assigned to one of four orders of presentation. In addition to retrospective
information on the participant's menstrual cycle, participants were given a stamped postcard and
asked to return it upon onset of next menstrual bleeding.
We predicted that salivary progesterone would be positively correlated with homoerotic
motivation in naturally-cycling women. One previous study found that women taking hormonal
contraceptives showed higher affiliative motivation than women who were naturally cycling
(Schultheiss et al., 2003); however, another study found that women on hormonal contraceptives
did not attend more to social stimuli like naturally cycling women in the high progesterone phase
of the menstrual cycle (Miller, 2011). Accordingly, we had no predictions concerning
homoerotic motivation among participants using hormonal contraceptives.
Results
Correlations with Progesterone and Conception Risk
Log transformed salivary progesterone and homoerotic motivation were significantly
correlated, r(90) = .27, p < . 01 (see Fig. 1). Although the highest progesterone levels occur in
the luteal phase, progesterone levels first begin to rise periovulatorily, i.e., during the period of
maximal conception risk. In order to fully disentangle the effects of progesterone on homoerotic
motivation from any effects of conception risk on homoerotic motivation, we did a partial

14
correlation on progesterone levels and homoerotic motivation controlling for conception risk; the
correlation remained significant, r(90) = .27, p < .01 see Fig. 1.
Comparison of Women Taking and Not Taking Hormonal Contraceptive
Scores of homoerotic motivation did not differ significantly between women on
hormonal contraceptives (M = 2.39, SD = .99), and naturally cycling women (M = 2.23, SD =
.98), t(157) = 1.01, p = .31. Lastly, there was a non-significant correlation in the predicted
direction between conception risk and homoerotic motivation, r(90) = -.14, p = .18 (see Fig. 2).
Discussion
Consonant with the thesis that homoerotic behavior serves an evolved affiliative function,
Study 1 showed a small positive correlation in women between homoerotic motivation and
progesterone, a hormone thought to be associated with affiliation. Because nonreproductive
sexual behavior entails opportunity costs that detract from reproduction, if homoerotic behavior
is governed by evolved adaptations, it follows that homoerotic motivation should decline as a
function of conception risk. Nevertheless, Study 2 revealed only a nonsignificant trend in this
direction. However, in comparing these two results, it is important to keep in mind that the first
independent variable was measured directly, with far greater precision than the second
independent variable, as progesterone was assayed in a sample of 104 participants, while
conception risk was calculated on the basis of estimated position in the menstrual cycle in a
sample of 92 participants, within which it was possible to use the more accurate backward-
counting method in 49% of cases. Lastly, individuals vary markedly from one another in
progesterone levels (Jasienska & Jasienski 2008), and this may have been a source of substantial
noise in Study 1. If progesterone is a key part of the proximate mechanism that adjusts
homoerotic motivation so as to minimize the trade-off between alliance formation and

15
conceptive behavior, then, given our between-subjects design, inter-individual variation in
progesterone levels may have precluded detection of the effects of conception risk in Study 1.
With the above caveat in mind, the results of Study 1 provided preliminary evidence in
support of the hypothesis that human homoerotic behavior serves the ultimate function of
strengthening social bonds. Both theoretical considerations and cross-species comparisons
indicate that the hypothesized function should not be limited to females. To further explore the
affiliation hypothesis, we investigated homoerotic motivation in men
Study 2: Priming affiliation and measuring progesterone in men
Introduction
If human homoerotic behavior serves an affiliative function, then priming the concept of
affiliation and contexts in which such affiliation is valuable should increase homoerotic
motivation. Likewise, if homoerotic motivation is adaptively altered in light of the cost/benefit
ratio of homoerotic behavior, then, due to the opportunity costs of nonreproductive behavior, in
the absence of cues of the value of affiliation, cues of opportunities for opposite-sex sexual
contact should decrease homoerotic motivation. Such priming effects can be expected to be
particularly marked in men, as our species’ history of intergroup conflict appears to have
exercised strong selective pressure on human males to mark group boundaries and affiliate with
in-group members (Van Vugt & Park, 2010), while vastly lower obligate parental investment
means that men should be particularly responsive to short-term mating cues signaling the
presence of opportunities (Symons, 1979).
Conceptual priming thus offers a promising avenue for investigating the affiliation
hypothesis in men. Moreover, because men produce progesterone (albeit at substantially lower
levels than do women), combining conceptual priming with measurements of progesterone

16
allows for an independent test of the patterns found in Study 2, as we can expect an interaction
between priming and progesterone. We therefore predicted that (1) priming affiliation will
increase homoerotic motivation in men, while priming opposite-sex sexual contexts will have the
opposite effect, and (2) the effects of priming will be further mediated by progesterone level.
Participants
A total of 61 male participants, age 18 to 27 years (M = 19.56, SD = 1.92) were recruited
from the Department of Psychology subject pool at a large university in the southwestern U.S.
One participant was excluded because he did not complete all of the requisite items, leaving 60
participants.
Method
Progesterone Measure
Saliva was collected as in Study 2 before any of the questionnaires or instruments were
completed. Progesterone was assayed as in Study 2. Mean inter- and intra-assay coefficients of
variation for the participants retained for analysis were 9.4% and 6.1%, respectively. Valid
progesterone values were obtained for 59 participants.
Priming
Participants were randomly assigned to one of three priming conditions: affiliative,
opposite-sex sexual opportunity, or neutral. There were 21 participants in the Sexual condition,
21 participants in the affiliative condition and 17 participants in the neutral condition (after
exclusion of the participant for whom we did not get a valid progesterone value). Before the
measure of homoerotic motivation, participants were given a priming exercise in the form of
word fragment completion puzzles composed of 40 words in two sets of 20: affiliative (e.g., Fri_
_ds , Friends), opposite-sex sexual opportunity (e.g., Bre_ _ _s, Breasts), and neutral (e.g.,

17
Sq_ar_, Square). Within each condition, participants were randomly assigned two of three
possible lists of 20 items. In the affiliative and opposite-sex sexual priming conditions, 5 neutral
words were also included in each list to conceal the priming objective. To further avoid demand
characteristics many word fragments were selected such that more than one word could be
constructed using the fragment (e.g,. Lo_ _r, Lover; Tr_ _p , Troop). To gauge the degree of
priming achieved, completed fragments were inspected to determine the percentage of responses
that fit the intended prime. In both the affiliative and opposite-sex sexual conditions, there was a
significant increase in percentage correct from puzzle 1 (M: 64%) to puzzle 2, (M: 70%), paired
t(41) = -2.36, p = .02, suggesting that the relevant concepts were increasingly activated as the
priming process continued.
Sexual orientation. After the priming tasks and the homoerotic motivation inventory,
participants completed the (1948) Kinsey scale. Of the 60 participants, 56 indicated that they
were exclusively heterosexual and 4 indicated that they were heterosexual with incidental
homosexual contact (Kinsey 2). Scores on the Kinsey scale were randomly associated with
priming condition, F(2, 57) = 1.15, p = .32.
Results
Effect of Priming
An ANOVA was used to compare homoerotic motivation across the three priming
conditions, revealing a significant difference between conditions, F(2, 58) = 3.62, p = .03. This
analysis remains significant with the inclusion of the one participant lacking a valid progesterone
value and with the exclusion of 4 participants who report as Kinsey 2s. Post hoc t-tests reveal a
significant difference between homoerotic motivation in the affiliative (M= 1.79, SD = .61)
versus opposite-sex sexual condition (M=1.37, SD= .41) t(40) = 2.61, pone-tailed = < .01, and a

18
significant difference between the affiliative (M= 1.79, SD = .61) and neutral conditions (M=
1.47, SD = .54) t(37) = 1.64, pone-tailed = .03. Although those in the neutral priming condition
showed higher scores on the homoerotic motivation scale than those in the opposite-sex sexual
condition this was not significant (p = .52)
Interaction of Priming and Progesterone
A MANOVA model using condition as a fixed factor and log transformed progesterone
as a covariate predicting homoerotic motivation was significant overall F(5, 53) = 2.94, p < .02
revealing a significant interaction of priming condition and log transformed progesterone F(2,
53) = 3.22, p < .048, with main effects of condition F(2, 53) = 2.12, p = .13 and log transformed
progesterone F(1, 53)= 0.06, p = .82. As illustrated in Fig. 3, performing a mean split on the
basis of progesterone level revealed that men with high progesterone who were assigned to the
affiliative condition reported more homoerotic motivation than both men with low progesterone
in the affiliative condition and men in other conditions regardless of progesterone level.
Discussion
Results from Study 2 largely supported the affiliation hypothesis, as priming the concept
of affiliation appeared to positively shift men’s attitudes towards homoerotic behavior.
Congruent with the results of Study 1, such priming interacted with progesterone levels, as those
men who had the highest progesterone at baseline evinced the largest effect of affiliative
priming. Lastly, as in Study 1, there is evidence for the opportunity cost thesis, as homoerotic
motivation is lower in men primed with sexual as opposed to affiliative concepts.
GENERAL DISCUSSION
The results presented here provide the first experimental evidence for the hypothesis that
human homoerotic behavior serves the ultimate function of enhancing affiliation. In women, we

19
found a significant correlation between homoerotic motivation and progesterone, a hormone
thought to be linked to affiliation. In men, we found that priming with cues of affiliation
increased homoerotic motivation relative to men primed with other concepts, and that this effect
was particularly marked in men with high baseline progesterone. Men primed with words about
opposite-sex sexual contact showed the lowest scores on homoerotic motivation and there was
negative albeit nonsignificant correlation between conception risk and homoerotic motivation in
women. Both of these hint at adaptive design in light of opportunity costs such that homoerotic
motivation may be downregulated in the face of the possible direct benefits of reproductive sex.
While promising, our findings should nonetheless be considered preliminary, as our
investigations were subject to a number of limitations. First, while our theory predicted links
between homoerotic motivation and hormonal and situational factors associated with affiliation,
our between-subjects design necessarily limited our ability to specify the precise nature of these
linkages. It could be, for example, that those who are chronically high in progesterone
experience more homoerotic motivation overall, or that those who happened to have a higher
progesterone level (perhaps because of situational factors) at the time of participation reported
more homoerotic motivation. Our methods thus did not allow us to differentiate trait-level from
state-level facets of progesterone profile and possible corresponding patterns of homoerotic
motivation.
As we will discuss below, the affiliation hypothesis aims to explain the evolutionary
persistence of both a flexible capacity for sexual attraction to both sexes and, in a subset of
individuals, exclusive attraction to members of the same sex. However, there is a substantial
empirical gap between our measure of homoerotic motivation and sexual orientation per se,
including, among others, the possible lack of generalizability stemming from our reliance in both

20
studies on samples composed almost exclusively of young U.S. undergraduate students, the vast
majority of whom were heterosexual.
Our hormonal measures were also subject to some limitations. First, we relied on salivary
assays of progesterone, a less direct measure than serum assays. Second, our measure of
conception risk in women was subject to uncertainty as regards menstrual cycle position, as we
were able to employ the more reliable backward-counting method in less than half of the sample.
Future investigations should therefore employ ovulation tests or estradiol assays in order to
measure conception risk.
Homoerotic motivation as we measured it is, of course, subject to the limitations of self-
report. Although we do not have evidence that our measure of homoerotic motivation actually
predicts the extent to which individuals engage in same-sex sexual behavior, nonetheless, with
the caveat that cultural norms undoubtedly shape the extent to which motivation translates into
action, this is plausible given that those who are bisexual or homosexual (as indicated by the
Kinsey scale) display greater homoerotic motivation on our scale. Homoerotic motivation is a
necessary but not sufficient precursor to engaging in homoerotic behavior; however, it may well
be the case that this motivation is not always experienced consciously (e.g., Adams, Wright, &
Lohr, 1996).
Our studies examined only one potential neurochemical promoter of affiliation,
progesterone. Testosterone has been shown to promote prosocial behavior in some contexts (e.g.,
van Honk, Montoya, Bos, van Vugt, & Terburg, 2012). More broadly, oxytocin has been
implicated as an affiliative hormone in many studies (for reviews, see Bartz, Zaki, Bolger, &
Ochsner, 2011; Campbell, 2008). However, concerns about such findings have recently been
raised in light of problems potentially plaguing the measurement of oxytocin in humans. This is

21
particularly important given that salivary measures of oxytocin may be especially questionable
(McCullough, Churchland, & Mendez, 2013), as the invasive and intimate nature of blood draws
is such that care would need to be taken not to create contexts of affiliation that could, in turn,
interact with the phenomenon of interest here. In the event that such issues can be resolved, it
will be important to explore whether, as the affiliation hypothesis predicts, oxytocin too
undergirds homoerotic motivation.
Our findings constitute preliminary evidence that homoerotic motivation serves the
adaptive function of promoting alliance formation and maintenance. Our studies primarily
explored such motivation in individuals who are not exclusively homosexual, yet our findings
potentially shed light on the latter as well. Exclusively homosexual orientation has posed an
evolutionary puzzle, as the percentage of those with this orientation remains fairly stable despite
its self-evident reproductive costs.
There is evidence for increased fertility in the maternal kin of exclusively homosexual
men, suggesting a history of sexually antagonistic selection (Ciani, Cermelli, & Zanzotto, 2008;
Iemmola & Camperio, 2009; VanderLaan, Forrester, Petterson, & Vasey, 2012). However, at
most, this can explain only some of the variance, and only in one sex; indeed, numerous
investigations have failed to find evidence of a direct adaptive benefit of exclusive
homosexuality (for a review see Rahman & Wilson, 2003). Importantly, the genetics underlying
exclusive homosexuality have been elusive and not easily replicated (Gavrilets & Rice, 2006),
suggesting that many loci may be involved, each of which makes only a small contribution to
sexual orientation.
If homoerotic motivation is governed by many genes, then, in combination with gene-
environment interactions, we would expect a continuous distribution of sexual attraction (see

22
Poiani & Dixson, 2010), with those in the middle of the spectrum generally having the highest
fitness under ancestral conditions and constituting the bulk of the distribution. Indeed, the
affiliation hypothesis predicts that, in societies where homoerotic behavior is not highly
proscribed, exclusively heterosexual individuals can be expected to possess fewer and/or
shallower fitness-enhancing alliances. Although the observable distribution of female sexual
orientation is, in part, consistent with this model, the bimodal distribution of male sexual
orientation in Western samples seems to defy it (M. Diamond, 1993). However, the heightened
social stigma attending male homoerotic (or even affectionate homosocial) behavior in these
populations may well explain the sex difference in these distributions, i.e., the sex difference
observed in the degree of continuity of the distributions of sexual orientation likely stems from
the greater proscriptions applied by Western cultures to male homoerotic behavior relative to
female homoerotic behavior. Importantly, recent research reveals not only greater continuity in
the distributions of sexual orientation in each sex than previously recognized, but also cohort
effects explicable in terms of changing cultural norms that shape both action and identity (Savin-
Williams & Vrangalova, 2013); relatedly, there is considerable evidence that both the targets of
attraction and self-construed identity often vary across time, particularly in women (L.M.
Diamond, 2012, Kuhle & Radtke, 2013). Future research will thus clearly need to address the
multiple biological, social, and contextual factors shaping human homoerotic motivation.
Although much of the literature on the evolution of homosexuality portrays sexually
antagonistic selection, overdominance, and homozygosity at multiple loci as mutually exclusive
accounts, such exclusivity is not intrinsic to the theories themselves. Accordingly, in light of our
findings, and in light of the apparent frequency of the capacity for same-sex arousal in humans,
we believe that it would be a mistake for exciting developments in one area to squelch

23
investigations in the other. Indeed, as evolutionary studies of sexuality come of age, the time has
arrived for a fuller exploration of the thesis that, consonant with patterns evident in nonhuman
primates, the capacity for homoerotic behavior serves an adaptive affiliative function in humans.

24
REFERENCES:
Adams, D. B., Gold, A. R., & Burt, A. D. (1978). Rise in female-initiated sexual activity at
ovulation and its suppression by oral contraceptives. New England Journal of Medicine,
299, 1145–1150.
Adams, H. E., Wright, L. W., & Lohr, B. A. (1996). Is homophobia associated with homosexual
arousal? Journal of Abnormal Psychology, 105, 440–445.
Alexander, G. M., Sherwin, B. B., Bancroft, J., & Davidson, D. W. (1990). Testosterone and
sexual behavior in oral contraceptive users and nonusers: A prospective study. Hormones
and Behavior, 24, 388–402.
Andersen, M. L., & Tufik, S. (2006). Does male sexual behavior require progesterone? Brain
Research Reviews, 51, 136–143.
Bartz, J. A., Zaki, J., Bolger, N., & Ochsner, K. N. (2011). Social effects of oxytocin in humans:
Context and person matter. Trends in Cognitive Sciences, 15, 301–309.
Bora, E., Yucel, M., & Allen, N. B. (2009). Neurobiology of human affiliative behaviour:
implications for psychiatric disorders. Current Opinion in Psychiatry, 22, 320-325.
Bowles, S. (2009). Did warfare among ancestral hunter-gatherers affect the evolution of human
social behaviors? Science, 324, 1293-1298.
Brown, S. L., Fredrickson, B. L., Wirth, M. M., Poulin, M. J., Meier, E. A., Heaphy, E. D.,…
Schutheiss, O. C. (2009). Social closeness increases salivary progesterone in humans.
Hormones and Behavior, 56, 108–111.
Campbell, A. (2008). Attachment, aggression and affiliation: The role of oxytocin in female
social behavior. Biological Psychology, 77, 1–10.

25
Ciani, A. C., Cermelli, P., & Zanzotto, G. (2008). Sexually antagonistic selection in human male
homosexuality. PloS ONE, 3, e2282.
Chivers, M. L., Rieger, G., Latty, E., & Bailey, J. M. (2004). A sex difference in the specificity
of sexual arousal. Psychological Science, 15, 736–744.
Chivers, M. L., Seto, M. C., Lalumiere, M. L., Laan, E., & Grimbos, T. (2010). Agreement of
self-reported and genital measures of sexual arousal in men and women: A meta-analysis.
Archives of Sexual Behavior, 39, 5–56.
DeWall, C. N., MacDonald, G., Webster, G. D., Masten, C. L., Baumeister, R. F., Powell, C.,…
Eisenberger, N. I. (2010). Acetaminophen reduces social pain. Psychological Science, 21,
931.
Diamond, L. M. (2012). The desire disorder in research on sexual orientation in women:
Contributions of dynamical systems theory. Archives of Sexual Behavior, 41, 73-83.
Diamond, L. M., & Wallen, K. (2011). Sexual minority women’s sexual motivation around the
time of ovulation. Archives of Sexual Behavior, 40, 237–246.
Diamond, M. (1993). Homosexuality and bisexuality in different populations. Archives of Sexual
Behavior, 22, 291–310.
Eisenberger, N. I., Jarcho, J. M., Lieberman, M. D., & Naliboff, B. D. (2006). An experimental
study of shared sensitivity to physical pain and social rejection. Pain, 126, 132–138.
Eisenberger, N. I., & Lieberman, M. D. (2004). Why rejection hurts: A common neural alarm
system for physical and social pain. Trends in Cognitive Sciences, 8, 294–300.
Fessler, D. M. T. (2003). No time to eat: An adaptationist account of periovulatory behavioral
changes. Quarterly Review of Biology, 78, 3–21.

26
Fleischman, D. S., Navarrete, C. D., & Fessler, D. M. T. (2010). Oral contraceptives suppress
ovarian hormone production. Psychological Science, 21, 750-752.
Frye, C. A. (2006). An overview of oral contraceptives: mechanism of action and clinical use.
Neurology, 66, S29-S36.
Gangestad, S. W., & Thornhill, R. (1998). Menstrual cycle variation in women’s preferences for
the scent of symmetrical men. Proceedings of the Royal Society B: Biological Sciences,
265, 927–933.
Gavrilets, S., & Rice, W. R. (2006). Genetic models of homosexuality: generating testable
predictions. Proceedings of the Royal Society B: Biological Sciences, 273, 3031-3038.
Guida, M., Di Spiezio, S. A., Bramante, S., Sparice, S., Acunzo, G., Tomaselli, G. A., …Nappi,
C. (2005). Effects of two types of hormonal contraception—oral versus intravaginal—on
the sexual life of women and their partners. Human Reproduction, 20, 1100 –1106.
Haselton, M. G., & Gangestad, S. W. (2006). Conditional expression of women’s desires and
men’s mate guarding across the ovulatory cycle. Hormones and Behavior, 49, 509-518.
Hatcher, R. A., & Namnoum, A. B. (2007). The menstrual cycle. In R. A. Hatcher, J. Trussel, &
A. L. Nelson (Eds.), Contraceptive technology (18th ed. pp. 7-17). New York: Ardent
Media.
Hill, K., & Hurtado, A. M. (2009). Cooperative breeding in South American hunter–gatherers.
Proceedings of the Royal Society B: Biological Sciences, 276, 3863.
Hrdy, S. (2009). Meet the alloparents: Shared child care may be the secret of human evolutionary
success. Natural History, 118, 24–29.

27
Iemmola, F., & Camperio, A. C. (2009). New evidence of genetic factors influencing sexual
orientation in men: Female fecundity increase in the maternal line. Archives of Sexual
Behavior, 38, 393–399.
Jasienska, G., & Jasienski, M. (2008). Interpopulation, interindividual, intercycle, and intracycle
natural variation in progesterone levels: A quantitative assessment and implications for
population studies. American Journal of Human Biology, 20, 35-42.
Kinsey, A. C., Pomeroy, W. B., & Martin, C. E. (1948). Sexual behavior in the human male.
Bloomington, IN: Indiana University Press.
Kirkpatrick, R. C. (2000). The evolution of human homosexual behavior. Current Anthropology,
41, 385–413.
Kuhle, B., & Radtke, S. (2013). Born both ways: The alloparenting hypothesis for sexual fluidity
in women. Evolutionary Psychology, 11, 304-323.
MacDonald, G., & Leary, M. R. (2005). Why does social exclusion hurt? The relationship
between social and physical pain. Psychological Bulletin, 131, 202-223.
Maner, J. K., Miller, S. L., Schmidt, N. B., & Eckel, L. A. (2010). The endocrinology of
exclusion. Psychological Science, 21, 581-588.
McCullough, M. E., Churchland, P. S., & Mendez, A. J. (2013). Problems with measuring
peripheral oxytocin: can the data on oxytocin and human behavior be
trusted?. Neuroscience & Biobehavioral Reviews, 37(8), 1485-1492.
Mehta, P. H., & Josephs, R. A. (2010). Social endocrinology hormones and social motivation. In
D. Dunning (Ed.), Social motivation. (pp. 171-189). New York: Psychology Press.

28
Miller, S. L. (2011). Hormones and social affiliation: Menstrual cycle shifts in progesterone
underlie women’s attention to signs of social support. Unpublished doctoral dissertation,
Florida State University, Tallahassee.
Muscarella, F. (1999). The homoerotic behavior that never evolved. Journal of Homosexuality
37, 1-18.
Muscarella, F. (2000). The evolution of homoerotic behavior in humans. Journal of
Homosexuality, 40, 51–77.
Muscarella F. (2006) The evolution of male-male sexual behavior in humans: The alliance
theory. Journal of Psychology & Human Sexuality 18, 275-311.
Muscarella, F., Cevallos, A. M., Siler-Knogl, A., & Peterson, L. M. (2005). The alliance theory
of homosexual behavior and the perception of social status and reproductive
opportunities. Neuroendocrinology Letters, 26, 771–774.
Navarrete, C. D., Fessler, D. M. T., Fleischman, D. S., & Geyer, J. (2009). Race bias tracks
conception risk across the menstrual cycle. Psychological Science, 20, 661-665.
Poiani, A., (2010) Poiani, A., & Dixson, A. F. (2010). Animal homosexuality: a biosocial
perspective. Cambridge: Cambridge University Press.
Rahman, Q., & Wilson, G. D. (2003). Born gay? The psychobiology of human sexual
orientation. Personality and Individual Differences, 34(8), 1337-1382.
Reblin, M., & Uchino, B. N. (2008). Social and emotional support and its implication for health.
Current Opinion in Psychiatry, 21, 201-205.
Rieger, G., Chivers, M. L., & Bailey, J. M. (2005). Sexual arousal patterns of bisexual men.
Psychological Science, 16, 579–584.

29
Rind, B., & Yuill, R. (2012). Hebephilia as mental disorder? A historical, cross-cultural,
sociological, cross-species, non-clinical empirical, and evolutionary review. Archives of
Sexual Behavior, 41, 797-829.
Ross, M. W., & Wells, A. L. (2000). The modernist fallacy in homosexual selection theories:
Homosexual and homosocial exaptation in South Asian society. Psychology, Evolution &
Gender, 2, 253–262.
Savin-Williams, R. C., & Vrangalova, Z. (2013). Mostly heterosexual as a distinct sexual
orientation group: A systematic review of the empirical evidence. Developmental Review,
33, 58-88.
Schultheiss, O. C., Dargel, A., & Rohde, W. (2003). Implicit motives and gonadal steroid
hormones: Effects of menstrual cycle phase, oral contraceptive use, and relationship
status. Hormones and Behavior, 43, 293–301.
Schultheiss, O. C., Wirth, M. M., & Stanton, S. J. (2004). Effects of affiliation and power
motivation arousal on salivary progesterone and testosterone. Hormones and Behavior,
46, 592–599.
Seeman, T. E. (1996). Social ties and health: The benefits of social integration. Annals of
Epidemiology, 6, 442–451.
Silk, J. B., Alberts, S. C., & Altmann, J. (2003). Social bonds of female baboons enhance infant
survival. Science, 302, 1231-1234.
Silk, J. B., Beehner, J. C., Bergman, T. J., Crockford, C., Engh, A. L., Moscovice, L. R., Wittig,
R. M., et al. (2010). Strong and consistent social bonds enhance the longevity of female
baboons. Current Biology, 20, 1359–1361.

30
Spector, I. P., Carey, M. P., & Steinberg, L. (1996). The Sexual Desire Inventory: Development,
factor structure, and evidence of reliability. Journal of Sex & Marital Therapy, 22, 175-
190.
Sugiyama, L. S. (2004). Illness, injury, and disability among Shiwiar forager-horticulturalists:
Implications of health-risk buffering for the evolution of human life history. American
Journal of Physical Anthropology, 123, 371–389.
Symons, D. The Evolution of Human Sexuality. New York: Oxford University Press, 1979.
Van Honk, J., Montoya, E. R., Bos, P. A., van Vugt, M., & Terburg, D. (2012). New evidence on
testosterone and cooperation. Nature, 485, E4–E5.
Van Vugt, M. (2009). Despotism, democracy, and the evolutionary dynamics of leadership and
followership. American Psychologist, 64, 54-56.
Van Vugt, M., & Park, J. H. (2010). The tribal instinct hypothesis: Evolution and the social
psychology of intergroup relations. Visions in Conflict: International Perspectives on
Values and Enmity, 1, 20.
VanderLaan, D. P., Forrester, D. L., Petterson, L. J., & Vasey, P. L. (2012). Offspring production
among the extended relatives of Samoan men and fa’afafine. PLoS ONE, 7, e36088.
Vasey, P. L. (1995). Homosexual behavior in primates: A review of evidence and theory.
International Journal of Primatology, 16, 173–204.
Wallen, K., & Parsons, W. A. (1997). Sexual behavior in same-sexed nonhuman primates: Is it
relevant to understanding human homosexuality? Annual Review of Sex Research, 8, 195-
223.
Wallwiener, M., Wallwiener, L. M., Seeger, H., Mueck, A. O., Zipfel, S., Bitzer, J., &
Wallwiener, C. W. (2010). Effects of sex hormones in oral contraceptives on the female

31
sexual function score: A study in German female medical students. Contraception, 82,
155–159.
Wilcox, A. J., Dunson, D. B., Weinberg, C. R., Trussell, J., & Baird, D. D. (2001). Likelihood of
conception with a single act of intercourse: Providing benchmark rates for assessment of
post-coital contraceptives. Contraception, 63, 211–215.
Wirth, M. M., Meier, E. A., Fredrickson, B. L., & Schultheiss, O. C. (2007). Relationship
between salivary cortisol and progesterone levels in humans. Biological Psychology, 74,
104–107.
Wirth, M. M., & Schultheiss, O. C. (2006). Effects of affiliation arousal (hope of closeness) and
affiliation stress (fear of rejection) on progesterone and cortisol. Hormones and Behavior,
50, 786–795.

32
Figure Captions
Figure 1- Scatterplot of the relationship between log transformed progesterone and homoerotic
motivation in women (n=92)
Figure 2- Average homoerotic motivation by cycle day for 92 naturally-cycling women. Days
11-15 of the menstrual cycle are those on which ovulation is most likely to occur.
Figure 3- The interaction of progesterone (mean split) and experimental condition on homoerotic
motivation in men (n=59)

33

34

35

36
Appendix A: Factor analysis loadings for each question and total alpha for all participants, males
and females
All participants
Male
participants
Female
participants
n
960
298
662
The idea of kissing a (wo)man seems sexually
arousing to me.
0.90
0.90
0.89
(Wo)men’s bodies are erotic.
0.80
0.90
0.70
If a (wo)man made a pass at me I would feel
disgusted. –reverse scored
0.75
0.74
0.74
I have fantasized recently about having sexual
contact with a (wo)man.
0.84
0.88
0.84
I could not imagine the touch of a (wo)man
being sexually arousing.
0.81
0.82
0.80
Alphas for total scale
0.87
0.90
0.85