Psychology of Intimate Relationships Week 3

Friendship Jealousy: One Tool for Maintaining Friendships in the Face of Third-Party Threats?

Jaimie Arona Krems Oklahoma State University

Keelah E. G. Williams Hamilton College

Athena Aktipis and Douglas T. Kenrick Arizona State University

Friendships can foster happiness, health, and reproductive fitness. However, friendships end— even when we might not want them to. A primary reason for this is interference from third parties. Yet, little work has explored how people meet the challenge of maintaining friendships in the face of real or perceived threats from third parties, as when our friends inevitably make new friends or form new romantic relationships. In contrast to earlier conceptualizations from developmental research, which viewed friendship jealousy as solely maladaptive, we propose that friendship jealousy is one overlooked tool of friendship maintenance. We derive and test—via a series of 11 studies (N � 2,918) using hypothetical scenarios, recalled real-world events, and manipulation of online emotional experiences—whether friendship jealousy possesses the features of a tool well-designed to help us retain friends in the face of third-party threats. Consistent with our proposition, findings suggest that friendship jealousy is (a) uniquely evoked by third-party threats to friendships (but not the prospective loss of the friendship alone), (b) sensitive to the value of the threatened friendship, (c) strongly calibrated to cues that one is being replaced, even over more intuitive cues (e.g., the amount of time a friend and interloper spend together), and (d) ultimately motivates behavior aimed at countering third-party threats to friendship (“friend guarding”). Even as friendship jealousy may be negative to experience, it may include features designed for beneficial—and arguably prosocial— ends: to help maintain friendships.

Keywords: friendship, jealousy, friend guarding, partner choice, relationship maintenance

Supplemental materials: http://dx.doi.org/10.1037/pspi0000311.supp

Inevitably, our friends become close with other people, as when they make new friends or form new romantic relationships. How do we feel when this happens? Speaking to The New York Times, author Andrea Lavinthal confessed: “Most girls won’t admit this, but they’d rather you hit on their significant other than their best friend” (Alford, 2014). We propose Lavinthal is not unusual in experiencing a stab of “friendship jealousy” at the thought of a best friend becoming especially close with someone else. Even as

friendship jealousy might be unpleasant to experience—and per- haps even embarrassing to acknowledge—it may function to help people maintain their valued friendships.

Most descriptions liken “being jealous” to feeling some mixture of sadness, anger, and anxiety (Bringle, 1991; Hupka, 1991; Par- rott & Smith, 1993; Sharpsteen, 1991; Sharpsteen & Kirkpatrick, 1997). Research on jealousy has focused mostly on the domain of sexual and/or romantic relationships (e.g., Bhugra, 1993; Buss, 2000, 2013; Buss & Haselton, 2005; Buunk, 1981, 1982; DeSteno & Salovey, 1996; DeSteno, Valdesolo, & Bartlett, 2006; Freud, 1910). Yet, jealousy is not unique to mating relationships, as historical, anthropological, nonhuman animal, and even some modern empirical evidence can attest (e.g., Buss, 2013; Campos, Walle, & Dahl, 2010; Hruschka, 2010; Parker, Low, Walker, & Gamm, 2005). Here, we explore jealousy in a distinct social domain—friendship. We also propose that friendship jealousy might differ from romantic jealousy in several important ways.

Whereas romantic jealousy is presumed to facilitate mate reten- tion in the face of real or perceived third-party threats to romantic relationships (e.g., Buss, 1989, 2013; Buss & Shackelford, 1997; Buss, Shackelford, & McKibbin, 2008), we reason that friendship jealousy might be evoked by and guide responses to real or perceived third-party threats to valued friendships. In terms of similarities, both romantic relationships and friendships are core

This article was published Online First August 10, 2020. X Jaimie Arona Krems, The Oklahoma Center for Evolutionary Anal-

ysis (OCEAN), Department of Psychology, Oklahoma State University; Keelah E. G. Williams, Department of Psychology, Hamilton College; X Athena Aktipis and X Douglas T. Kenrick, Department of Psychology, Arizona State University.

This research was partially funded by a Philanthropic Education Orga- nization (PEO) International Scholar Award awarded to Jaimie Arona Krems. We thank Steven L. Neuberg and Thao Ha for their helpful comments on this research, and Malina Lemmons for her research assis- tance.

Correspondence concerning this article should be addressed to Jaimie Arona Krems, The Oklahoma Center for Evolutionary Analysis (OCEAN), Depart- ment of Psychology, Oklahoma State University, 116 North Murray Hall, Stillwater, OK 74078. E-mail: [email protected]

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

Journal of Personality and Social Psychology: Interpersonal Relations and Group Processes

© 2020 American Psychological Association 2021, Vol. 120, No. 4, 977–1012 ISSN: 0022-3514 http://dx.doi.org/10.1037/pspi0000311

977

components of daily life (e.g., Hruschka, 2010; Perlman, Stevens, & Carcedo, 2014). Evidence suggests that, like romantic relation- ships, friendships can facilitate health, happiness, and even repro- ductive fitness (e.g., DeScioli & Kurzban, 2009, 2011; DeScioli, Kurzban, Koch, & Liben-Nowell, 2011; Dunbar, 2018; Lewis et al., 2011; Seyfarth & Cheney, 2012; Tooby & Cosmides, 1996). Moreover, just as people compete to attract romantic partners— sometimes luring them away from existing relationships—people engage in social competition to attract friends (e.g., Barclay, 2013, 2016; Eisenbruch & Roney, 2017; Krems & Conroy-Beam, 2020). Indeed, people frequently report that third parties harm friendships (e.g., Rose, 1984). Thus, we should expect that, just as people possess responses suited to retaining their mates in the face of third-party interference, people might also possess responses suited to retaining their friends in the face of third-party interfer- ence (e.g., DeScioli & Kurzban, 2009, 2011).

Although substantial research has addressed the recurrent chal- lenges of mate retention and romantic relationship maintenance (e.g., Buss & Shackelford, 1997; Buss et al., 2008; Conroy-Beam, Goetz, & Buss, 2016; Kenrick, Griskevicius, Neuberg, & Schaller, 2010; Kenrick, Neuberg, Griskevicius, Becker, & Schaller, 2010; Pillsworth & Haselton, 2006; Welling, Puts, Roberts, Little, & Burriss, 2012), far less work has explored strategies for friend retention and friendship maintenance (Canary, Stafford, Hause, & Wallace, 1993; Fehr, 1996; Hays, 1985; Oswald & Clark, 2006; Oswald, Clark, & Kelly, 2004; Rose & Serafica, 1986; Rusbult, Olsen, Davis, & Hannon, 2004). A consideration of the recurrent challenges of friendship maintenance leads to several predictions about the architecture of friendship jealousy, which we test here.

Jealousy Can Protect Valued Social Relationships

The experience of jealousy is considered aversive, and it has been widely described as a blend of sadness, anger, and anxiety (e.g., Sharpsteen, 1991). Researchers working at various levels of analysis largely agree that jealousy is evoked when relationships are threatened by others, and can motivate behavior aimed at countering those threats (e.g., Buss, 2013; Buss, Larsen, Westen, & Semmelroth, 1992; Buunk, Angleitner, Oubaid, & Buss, 1996; Daly, Wilson, & Weghorst, 1982; DeSteno & Salovey, 1996; DeSteno et al., 2006; Panksepp, 2010; Sagarin, 2005; Seyfarth & Cheney, 2012). In the prototypical example, one becomes jealous upon perceiving that a third party is interested in one’s partner and/or that one’s partner is interested in that third party (Daly et al., 1982; Salovey, 1991).1 However, different types of jealousy may be at work in different social domains, be activated by qualitatively different threats, and motivate behavior to protect qualitatively different relationships. For example, just as romantic partners may be jealous at the attention their spouses give rivals, a child may be jealous at the attention a caregiver lavishes on that child’s siblings (e.g., Miller, Volling, & McElwain, 2000)—in- deed, behavior consistent with jealousy appears as early as the first year of life (Campos et al., 2010).2

The flourishing work on romantic jealousy provides an initial springboard for exploring the understudied phenomenon of friend- ship jealousy. Specifically, researchers taking an evolutionary ap- proach to romantic jealousy have found evidence consistent with the propositions that— because successfully maintaining romantic relationships likely increased reproductive fitness in ancestral en-

vironments, and because third parties posed recurrent threats to the maintenance of those relationships—romantic jealousy is likely one adapted tool of partner retention (e.g., Buss, 2000, 2013; Buss & Haselton, 2005; Daly et al., 1982; Scelza et al., 2019; Symons, 1979).3 Behavior consistent with romantic jealousy may be evi- denced in some nonhuman animals; among humans, it exists across cultures, and it possesses features that appear to be well- designed to solve the recurrent problem of retaining mates in the face of third-party threats: (a) romantic jealousy is evoked by cues that a romantic partner stands to be lost to a third party; (b) cues with better predictive validity of impending loss receive prioriti- zation in driving levels of romantic jealousy; and (c) romantic jealousy spurs a suite of behavioral inclinations, known as mate guarding, that are theoretically aimed at countering the threat of partner loss (Buss, 2000, 2013; Buss & Haselton, 2005; Buss et al., 1992; Buss & Shackelford, 1997; Lewis, Al-Shawaf, Conroy- Beam, Asao, & Buss, 2017; Scelza, 2014; Scelza et al., 2019; Schmitt & Pilcher, 2004). Here, we conduct a similar investigation of friendship jealousy, first making a case for the importance of friendships and the recurrent challenge of third-party threats to them.

Friendship: Beneficial Bonds

Friendships, defined as sustained medium- to long-term coop- erative alliances between genetically unrelated conspecifics (DeScioli & Kurzban, 2009; Perlman et al., 2014), are a human universal (Hruschka, 2010) and may also exist across nonhuman species (e.g., Seyfarth & Cheney, 2012; Silk, 2002, 2003). Among humans, friendships are considered central to health, happiness, and well-being by laypersons and researchers alike (Adams & Allan, 1998; Dunbar, 2018; Fehr, 1996; Hruschka, 2010; Perlman et al., 2014). For example, Americans report valuing their friends as much as they value money and employment— behind only health and family (Gallup Poll News Service, 2005), and people view friendships as a primary means to achieve a meaningful life (Benenson, 2014; Campbell, 2002; Kenrick & Krems, 2018; Krems, Kenrick, & Neel, 2017; Perlman & Peplau, 1981; Smith & Christakis, 2008). Indeed, the survivability and health benefits of maintaining friendships may be second only to those benefits accrued by quitting smoking (Dunbar, 2018; Holt-Lunstad, Smith, Baker, Harris, & Stephenson, 2015).

1 Colloquially, “jealousy” and “envy” are often used interchangeably. However, the former is evoked by the threatened loss of an existing bond, whereas the latter is evoked when one covets something someone else has (and that the envious person lacks; DelPriore, Hill, & Buss, 2012).

2 Because evolution works by descent with modification, jealousy over friendships (friendship jealousy) could have been built upon the founda- tions lain for sibling jealousy or mating jealousy— or vice versa. We focus here not on the phylogeny of friendship jealousy, but rather on the features that it might have if, like romantic jealousy, it functioned to help maintain valued but potentially threatened social bonds.

3 This is not to say that any behavior spurred by such jealousy is, today, fitness-enhancing or morally good; rather, this argument supposes that, even if some people occasionally reduced their fitness via harming their mates (e.g., in fits of jealousy-induced rage), on average, those people who experienced romantic jealousy were likely to have been more successful at maintaining their valued romantic relationships in the face of third-party threats—and, thus, would have enjoyed greater reproductive success—than those who failed to experience romantic jealousy under the same circum- stances.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

978 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

In line with this, some theory and evidence suggest that friend- ships have positive effects on fitness, defined in evolutionary terms. This makes sense, as friends have long been considered important sources of social, emotional, and material support (e.g., Campbell, 2002; Cottrell, Neuberg, & Li, 2007; Davis & Todd, 1982; Hruschka, 2010; Plickert, Cote, & Wellman, 2007). Re- search from Western cultures, studies in small-scale societies around the world, and even ethology suggest that sustained friend- ships might bolster fitness via a number of routes: by promoting an individual’s survival, providing status and resources, mitigating the negative impact of both physical and social threats, helping individuals win agonistic conflicts, and/or improving the longevity of offspring (Ackerman, Kenrick, & Schaller, 2007; Aktipis et al., 2018; Barakzai & Shaw, 2018; Barclay, 2013; Campbell, 2002; David-Barrett et al., 2015; DeScioli & Kurzban, 2009, 2011; DeScioli et al., 2011; Dunbar, 2018; Hrdy, 2011; Hruschka, 2010; Lewis et al., 2011; Roberts, 2005; Shaw, Choshen-Hillel, & Caruso, 2018; Shaw, DeScioli, Barakzai, & Kurzban, 2017; Silk, 2003; Silk, Alberts, & Altmann, 2003; Seyfarth & Cheney, 2012; Sugiyama, 2004; Tooby & Cosmides, 1996; Trivers, 1971).

Several specific theories have been proposed to account for how friendship might have bolstered fitness in ancestral environments. Whereas traditional theorizing emphasized the reciprocal ex- change of goods and support (e.g., Trivers, 1971), more recent theories have moved beyond the notion of dyadic reciprocation to explore how friendships might have helped humans survive and thrive in small, densely interconnected social groups (Barclay, 2013, 2016; DeScioli & Kurzban, 2009, 2011; DeScioli et al., 2011; Tooby & Cosmides, 1996). For example, Tooby and Cos- mides (1996) suggest that friendships helped resolve the recurrent challenge of accessing support when ill or injured. Because friends have a stake in one another’s welfare, they often provide the much-needed support to ill or injured friends that strangers would be unlikely to provide the ill or injured. Some evidence from small-scale societies supports this theorizing (e.g., Sugiyama, 2004). A similar account is the Alliance Hypothesis for Human Friendship, which holds that friendship arises, in part, from cog- nitive mechanisms designed to assemble support for future con- flicts (DeScioli & Kurzban, 2009, 2011; DeScioli et al., 2011). On this view, when social conflicts arose in small, interconnected groups, (a) people likely knew both disputants, and (b) as in most conflicts, the disputant with more supporters would win. Given the dynamics in small groups, having even just one close friend put their support behind you could make the difference between win- ning a conflict (e.g., surviving) and losing (e.g., dying).4

The Challenge of Friendship Maintenance

Despite their phenomenological and objective worth, even val- ued friendships can wane or end (Casper & Card, 2010; Rose, 1984; Rose & Serafica, 1986; Wellman, Wong, Tindall, & Nazer, 1997). For example, Wellman and colleagues (1997) found that, over the course of a decade, roughly 60% of friendships among working- and middle-class Canadians ended, with roughly half of sustained friendships becoming notably less close. Of course, friendships end for numerous reasons (de Vries & Johnson, 2002; McEwan, Babin Gallagher, & Farinelli, 2008; Rose, 1984). A primary and frequently cited reason that people give for the dis- solution of friendships they wished to maintain is interference from

third parties (e.g., Owens, Shute, & Slee, 2000; Parker, Kruse, & Aikins, 2010; Rose, 1984; Tannen, 2017). Indeed, multiple studies have found that both children and adults cite third-party interfer- ence as having caused the dissolution or diminution of a friendship (Bigelow, Tesson, & Lewko, 1996; McEwan et al., 2008; Silver- man, La Greca, & Wasserstein, 1995).

People clearly perceive third parties as possible threats to friend- ship maintenance, and such perceptions might be sufficient to evoke friendship jealousy, but do third parties truly threaten friend- ships? Laypeople and researchers alike might intuitively resist the notion that one can “lose” friends to other people— even as many readily accept that romantic relationships can be lost to others. It is true that norms of exclusivity can differ for friendly versus monogamous romantic relationships, and people generally per- ceive that friendly affection is less zero-sum than romantic affection (e.g., Davis & Todd, 1982; Sprecher & Regan, 2002). Nevertheless, several independent lines of research suggest that third-party interference may be a longstanding and genuine chal- lenge to friendship maintenance.

First, whenever individuals have some choice in picking their partners, and partners vary in desirability, individuals compete over desirable partners. Many theories of cooperation, including those dealing with human friendship thus include presumptions that individuals compete for desirable partners (e.g., Barclay, 2013, 2016; Bird, Bird, Codding, & Zeanah, 2019; Delton & Robertson, 2012; DeScioli & Kurzban, 2009, 2011; Krems & Conroy-Beam, 2020; Nesse, 2007; Noë & Hammerstein, 1994; Tooby & Cosmides, 1996). Even among the small-scale groups of nonhuman primates, individuals compete for access to desirable “friends” (e.g., Palombit, Cheney, & Seyfarth, 2001; Seyfarth & Cheney, 2012). A primary implication of theories involving part- ner choice is “the possibility of abandonment by one’s current partner in favor of a rival” (Barclay, 2013, p. 172).

Second, robust research suggests that a person can maintain only so many relationships at any one time, whether because we have finite time to invest in these relationships (Dunbar, 1993, 2008; Milardo, Johnson, & Huston, 1983; Roberts & Dunbar, 2011; Roberts, Dunbar, Pollet, & Kuppens, 2009; Zhou, Sornette, Hill, & Dunbar, 2005), and/or because we may be able to keep close track of only so many relationship partners at once (Hall, Larson, & Watts, 2011; Krems, Dunbar, & Neuberg, 2016; Krems & Wilkes, 2019; Miritello et al., 2013; Oswald et al., 2004). For example, relationships require time to build and to maintain (e.g., Miritello et al., 2013; Oswald et al., 2004), but time is a notably inelastic resource. Evidence suggests that when a friend forms a new, close relationship and/or becomes newly close with an existing relation- ship partner, this can cause one or more previous occupants of the friend’s innermost circle of affection to be relegated to an outer, less close circle (Dunbar, 2012; Roberts et al., 2009; Zhou et al., 2005). Further, this line of work might imply—likely in line with

4 Consider a group in which Steve, Tony, and Bucky are all friends, but Tony and Bucky have a conflict. Whichever friend Steve supports in the conflict will win. According to the Alliance Hypothesis, Steve is likely to throw his support behind the disputant most likely to support him in future conflicts (say, Bucky). By protecting Bucky, Steve also increases the likelihood that he will prevail in future conflicts, for he has protected a likely supporter.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

979FRIENDSHIP JEALOUSY

lay intuitions—that we feel the most threatened when potential interlopers consume our friends’ time.

Third, explicitly evolutionary views of friendship hold that third parties posed recurrent threats to friendship maintenance. For example, Tooby and Cosmides (1996) propose that people have only so many friendship slots, or niches; people attempt to become irreplaceable to friends to protect their niches from being usurped by others and, thus, to continue enjoying the benefits associated with those friendships (e.g., provisioning, social support). Somewhat similarly, the Alliance Hypothesis contends that a person’s friends are ranked hierarchically in descending order of those one would support in a conflict (e.g., between one’s own friends; DeScioli & Kurzban, 2009; DeScioli et al., 2011). All things equal, if a man’s best friend (first-ranked friend) had a conflict with his fourth-ranked friend, that man would support his best over his fourth-ranked friend, and two best friends should support one another over anyone else. However, if your best friend became closer to someone else, you would move down in your best friend’s rankings and thus be less likely to enjoy your best friend’s support in prospective conflicts with the very person who usurped your rank. Counter to lay intuitions, then, this work might imply that third parties will be perceived as threats insofar as they stand to replace us in our preferred niche or hierarchical ranking.

Known Friendship Maintenance Tactics

It is widely acknowledged that most bonds require some main- tenance to sustain (Burt, 2000; Cummings, Lee, & Kraut, 2006; Oswald & Clark, 2006; Oswald et al., 2004). Just as romantic relationship maintenance is considered a two-pronged challenge, involving maintaining a partner’s continued investment in the bond and preventing that partner’s loss or defection to someone else (Kenrick, Griskevicius, et al., 2010; Kenrick, Neuberg, et al., 2010), so too might friendship maintenance involve both of these challenges.

Again, the bulk of existing relationships research has focused on how people meet these challenges in romantic relationships; how- ever, some work has focused explicitly on friendships. This friend- ship work has been largely descriptive, taking a data-driven ap- proach to cataloging the tactics people report using (Canary et al., 1993; Fehr, 1996; Hays, 1985; Oswald & Clark, 2006; Oswald et al., 2004; Rose & Serafica, 1986; Rusbult et al., 2004). For example, Oswald and colleagues (2004) developed a typology of friendship maintenance strategies, including being positive around friends, being supportive, and spending time together. Others have cited the importance of avoiding sensitive issues and making one another laugh (e.g., Bigelow & La Gaipa, 1975; Burleson & Samter, 1994; Canary et al., 1993).

We propose that these tactics are well-suited to meet only one of the two central challenges of friendship maintenance. For example, being positive around friends and making them laugh might help maintain friends’ continued investment, but these tactics do not seem especially well suited to mitigate a friend’s loss or defection to someone else. On the other hand, friendship jealousy may be especially well designed to do just that.

Friendship Jealousy: Existing Work and New Predictions

Before we outline our predictions, we first overview existing work on friendship jealousy, which yields contrasting views that friendship jealousy is maladaptive or pathological.

Existing Research on Friendship Jealousy

In developmental psychology, there is some work on jealousy in friendships, which has built upon the theoretical foundations laid by Selman and colleagues (e.g., Selman, 1980; Selman & Schultz, 1990). Selman expected that jealousy over friends abates after adolescence, when “social-cognitive advances help older children take a more balanced view in which they recognize that no single relationship, no matter its quality, can meet all the interpersonal needs of an individual” (Parker, Ebrahimi, & Libber, 2005, p. 236). This work implies that, although children and adolescents might experience friendship jealousy, normally developing adults likely do not.

Accordingly, related empirical research has almost exclusively examined children and adolescents, and it has focused on the negative antecedents and outcomes of friendship jealousy (Kraft & Mayeux, 2016; Lavallee & Parker, 2009; Parker, Campbell, Kollat, & Lucas, 2008; Parker et al., 2005; Parker et al., 2010; Parker, Ramich, & Roth, 2009; Roth & Parker, 2001). For example, children with low self-esteem may have a high “vulnerability” to friendship jealousy (Ebrahimi, Parker, Lavallee, & Seiffke- Krenke, 2005; see also Bhugra, 1993), children who experience such jealousy may have lower friendship satisfaction (Giltenboth, 2001; Lavallee & Parker, 2009; Parker & Wargo Aikins, 2009), and children with a reputation for jealousy may be less accepted and perhaps also more frequently victimized by their peers (Parker & Gamm, 2003; Parker et al., 2010).

We do not dispute the validity of these findings. However, taking a functional perspective raises some questions about the underlying theorizing. For example, if friendship jealousy consis- tently causes solely negative outcomes, why does it remain so prevalent across eras, cultures, and perhaps even species? In the historical and ethnographic records, for example, there are numer- ous accounts of children and adults reacting with jealousy when friends seem to prefer the company of others (e.g., Hruschka, 2010; Rose, 1984; Tannen, 2017; Yalom & Brown, 2015). One possible way to reconcile empirical evidence for friendship jeal- ousy’s negative outcomes with its continued prevalence is to recognize that existing theorizing has led researchers to expect only—and, thus, to test for only—negative antecedents and out- comes of friendship jealousy. An evolutionary perspective expects that, even if friendship jealousy can be aversive to experience and occasionally harm friendships, on average, those who experienced it may have been better able to maintain their friendships (and associated benefits) than those who failed to experience friendship jealousy in the same situations.

Predictions from a Functional Perspective

If friendship jealousy arose as one tool of friendship mainte- nance, our modern psychologies might still possess several spe- cific features. As described in more detail below, in a series of 11

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

980 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

experiments, we investigate: (a) what cues do (and do not) evoke friendship jealousy; (b) whether some cues evoke greater friend- ship jealousy than others; (c) what behavior friendship jealousy might motivate; and (d) the specificity of friendship jealousy for this role as a tool of friendship maintenance—that is, whether friendship jealousy (vs. the often-concomitant emotion of sadness and/or anger) is uniquely evoked by third-party threats to friend- ships, and/or uniquely motivates threat-countering responses.

Evoked by Third-Party Threats to Friendship (But Not Friendship Loss in General)

Although we expect reports of sadness and/or anger to be concomitant with friendship jealousy—as jealousy is both theo- rized to be and reported to feel like a mix of these emotions (e.g., Bringle, 1991; Sharpsteen, 1991)— our approach expects that friendship jealousy will be evoked by the possible loss of a valued friendship to a third party. We contrast this with the threatened loss of the friendship alone (e.g., as when a friend spends more time at work). Whereas the threatened loss alone should be suffi- cient to evoke sadness and anger, only the threatened loss of a friend to a third party should evoke high levels of friendship jealousy (even as this might also evoke sadness and/or anger; Hypothesis 1). This prediction speaks to the cues that do (and do not) elicit friendship jealousy, and also begins to differentiate friendship jealousy from often-concomitant emotions.

Calibrated to friend value. At any one time, we might have multiple friends—a best friend, close friends, and perhaps some friendly acquaintances—with closer friendships providing more numerous and potentially more important benefits. Thus, the pro- spective loss of each of these friends should evoke different amounts of friendship jealousy. We expect that friendship jealousy will be sensitive to the value (i.e., closeness) of the threatened friendship (Hypothesis 2). This prediction speaks to one cue to which friendship jealousy might be calibrated.

Calibrated to cues of replacement. Evolutionary accounts of friendship imply that the most powerful cues to a friendship being threatened would be replacement threats, or cues that the possible interloper stands to usurp one’s place in the friendship. In contrast, intuition—and perhaps also friendship accounts emphasizing finite resources required for relationship maintenance (e.g., time)—im- ply that friendship jealousy would be most closely calibrated to the amount of time a best friend spends with the possible interloper. Although the amount of time a best friend spends with a person can be a cue to the stability of that person’s position in the friendship hierarchy or niche (i.e., to replacement threat), it is not determinative: one’s best friend can spend more time with others while still prioritizing one most of all. Consider, for example, that your best friend may spend 40 hr a week with “work buddies,” 20 hr with a new romantic partner, and only 3 hr a week with you, but neither the work colleagues nor the romantic partner threaten your status as best friend. By contrast, when your best friend prioritizes another friend over you (i.e., replaces you in the hierarchy or friendship niche) you stand to lose the potentially fitness- enhancing friend-mediated benefits (Burkett, 2009; DeScioli & Kurzban, 2009, 2011; DeScioli et al., 2011; Tooby & Cosmides, 1996).

Thus, whereas time spent together can be a cue of replacement threat, friendship jealousy should be strongly calibrated to more

direct cues of replacement threat (e.g., a best friend supporting the interloper over you; Hypothesis 3). We test between these predic- tions in several distinct but complementary ways. This prediction speaks to whether some cues evoke greater friendship jealousy than others.

Motivates behavior to counter threats. From an evolution- ary perspective, thoughts and feelings motivate threat-countering action (e.g., Nesse, 2019; Neuberg, Kenrick, & Schaller, 2010; Sznycer et al., 2016). Thus, friendship jealousy (but not necessar- ily sadness and/or anger alone) should motivate behavior aimed at maintaining the threatened friendship—what we term “friend guarding” (Hypothesis 4). This prediction speaks to the behavioral outputs of friendship jealousy, and also differentiates friendship jealousy from often-concomitant emotions.

Research Overview

Should these predictions find support in the 11 studies we report below, the present work would achieve several ends. Specifically, we would provide some of the first empirical evidence for a novel theory of friendship jealousy—a phenomenon that, although seem- ingly common (e.g., Alford, 2014; Rosenfeld, 2004; Hruschka, 2010), remains largely unexplored. Second, results would eluci- date the architecture of friendship jealousy in ways similar to other recent research on discrete emotions taking an evolutionary per- spective (e.g., Al-Shawaf, Conroy-Beam, Asao, & Buss, 2016; Sell, Tooby, & Cosmides, 2009; Sell et al., 2017; Sznycer, Xygala- tas, Agey, et al., 2019; Sznycer, Xygalatas, Alami, et al., 2018; Sznycer et al., 2017; Sznycer, Xygalatas, Agey, et al., 2018): showing what inputs do (and do not) evoke friendship jealousy, which cues are prioritized in driving it, and which behavioral outputs it motivates. Third, predicted findings would challenge existing conceptualizations of friendship jealousy from both early research and lay intuitions (Selman, 1980, see Alford, 2014)—that friendship jealousy is solely negative—instead aligning with evo- lutionary research on negative affect serving some beneficial ends (e.g., Sznycer, Xygalatis, Agey, et al., 2018; Sznycer, Xygalatis, Alami, et al., 2018).

Study 1

The experience of jealousy may feel like a mix of sadness and anger (e.g., Bringle, 1991; Sharpsteen, 1991). However, to the extent that jealousy is a distinct reaction well-suited for protecting friendships against third-party interference, it should be uniquely sensitive to third-party threats, whereas sadness and anger might not be. We predict that friendship jealousy will be strongly evoked by third-party threats to same-sex best friendships (relative to the threatened loss of the friendship alone), whereas sadness and anger might be evoked in both situations.

Method

Participants. We determined we would need approximately 42 participants for .80 power to detect small to medium effects (f � .20) assuming measurement correlation � .5. Fifty-six un- dergraduate students (20 females, 1 not reporting sex; Mage � 19.35; SDage � 1.51) at a university in the Southwestern United States were recruited into an in-laboratory study and received

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

981FRIENDSHIP JEALOUSY

course credit for participation in a 30-min study. All participants who completed the focal dependent variables were included in analyses.

Procedure and design. Participants were asked to give the names (first name, last initial) of same-sex others they saw in their day-to-day lives: a best friend, a close friend, and an acquaintance. Each friend type was defined for participants (e.g., “a best friend is the ONE person to whom you are closest, a person who would help you in dire times”). Best friends’ names were then piped into questions asking about that friend. We did not later ask about close friends and acquaintances; this information was gathered to ob- scure our focal aims.

Participants then gave short responses to open-ended questions about the primary activity and functionality of best friendships (“What is the primary activity that you and [name of best friend] do together?”; “What is the primary function that you fulfill for [name of best friend]? That is, what is it about your friendship that [name of best friend] could not live without, or that [name of best friend] would lose if you were no longer best friends?”). Re- sponses were piped into the four focal questions (described below) asking about reactions to the loss of the friendship (alone vs. to a third party). Questions appeared in randomized order.

To assess reactions to the loss of the best friendship alone, participants were asked (a) “If [name of best friend] stopped doing [activity] with you, would you feel . . . ?” and (b) “If [name of best friend] stopped relying on you and your friendship for [function], would you feel . . . ?”

To assess reactions to the loss of the best friendship to a third party, participants were asked (c) “If [name of best friend] started doing [activity] with another person—a same-sex stranger—in- stead of you, would you feel . . . ?” and (d) “If [name of best friend] started relying on someone else—a same-sex stranger—for [function] instead of relying on you, would you feel . . . ?”

In both conditions, reactions were assessed on a 7-point Likert- scale (1 � not at all, 7 � very much). Responses assessed included friendship jealousy (“jealousy”), emotional reactions often consid- ered concomitant with jealousy (“sadness,” “anger”), and three distractor emotions (“happiness,” “pride,” and “disgust”). Each

reaction item appeared in random order. Reactions were aggre- gated across activity and function questions.

Other measures. In this and subsequent studies, we included items not germane to focal predictions and not analyzed here, including exploratory measures (e.g., fitness interdependence; Ak- tipis et al., 2018) individual differences (e.g., self-esteem), and demographic questions (e.g., sex/gender, age, ethnicity). All mea- sures throughout were approved by the Institutional Review Board (IRB). Data for this and subsequent studies are available on Har- vard Dataverse.

Results and Discussion

We compared feelings of jealousy, sadness, and anger across conditions. The resultant 2 [Condition: loss of friendship, loss of friendship to a third party] � 3 [Reactions: jealousy, sadness, anger] within-subject analysis of variance (ANOVA) revealed a main effect of Reaction, F(2, 108) � 27.10, p � .001, �p2 � .334, qualified by the predicted Condition x Reaction interaction, F(2, 108) � 49.93, p � .001, �p2 � .480. The main effect of Condition was not significant (p � .202). (Means [SEs] for all main and distractor responses are reported in Table S1 in the online supple- mental materials). (Theorizing, testing, and interpreting sex/gender similarities and differences is beyond the scope of the current work. However, we reported related results in the online supple- mental materials. Effects largely hold across participant sex/gen- der; see online supplemental materials).

First, as predicted, reported friendship jealousy was greater in the loss to a third party condition (M � 3.86, SE � 0.28) than it was in the loss alone condition (M � 2.57, SE � 0.23, p � .001, 95% confidence interval (CI) [0.90, 1.67], �p2 � .450; see Figure 1). More important, this pattern of responding is con- sistent with the prediction that friendship jealousy is strongly evoked only by the threat of losing a friend to a third party (and not by the threat of losing a friend). In contrast, reported sadness was greater in the loss alone condition (M � 4.69, SE � 0.25) than in the loss to a third party condition (M � 4.08, SE � 0.30, p � .001, 95% CI [0.30, 0.92], �p2 � .225). Condition did

1

2

3

4

5

6

Jealousy Sadness Anger

Loss alone

Loss to a third party

tne mesrodnE

detr opeR

Figure 1. Reported levels of friendship jealousy and related emotions of sadness and anger in reaction to losing a best friendship (alone) versus losing a best friendship to a third party (e.g., to a best friend’s new friend) from Study 1. Error bars represent SEs.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

982 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

not significantly influence reported anger (Mloss alone � 3.57, SEloss alone � 0.28; Mloss third party � 3.39, SEloss third party � 0.27; p � .250).

Moreover, in the loss alone condition, both reported sadness and anger were greater than friendship jealousy (ps � .001). This was not the case in the loss to a third party condition; both jealousy and sadness were greater than anger in this condition (ps � .005), although jealousy and sadness were statistically indistinguishable from one another (p � .131).

Discussion

Consistent with our theorizing that friendship jealousy is rele- vant for dealing specifically with third-party interference in friend- ships, only friendship jealousy (and not sadness or anger) is relatively strongly evoked by third-party threats to friendships (vs. the loss of the friendship alone). The differences in reactivity to these two conditions is especially striking because both the threat- ened losses yield the same ends (loss of the friendship and related benefits). In Study 2 we further explore the possible specificity of friendship jealousy, not only as being uniquely evoked by third- party threats to friendships, but also as uniquely motivating be- havioral intentions to counter those threats.

Study 2

Study 2 facilitated several goals. First, we aimed to replicate findings from Study 1—that friendship jealousy is strongly evoked by third-party threats to best friendships, whereas sadness and anger, although often-concomitant with jealousy, are not— here, using a between-subjects design and an alternative sample. Thus, to replicate findings from Study 1, participants again reported their friendship jealousy at the prospective loss of their best friendships (alone or to third parties).

Second, we aimed to test whether friendship jealousy (vs. often- concomitant emotions) motivates friend guarding. We measure friend guarding intentions by adapting one well-established tactic of mate guarding: vigilance toward possible rivals (e.g., Buss et al., 2008; Shackelford, Goetz, & Buss, 2005). We predict that (a) third-party threats to best friendships will evoke more friendship jealousy than the threatened loss of the friendship alone, and, in turn, that (b) this friendship jealousy (but not sadness or anger) will motivate greater behavioral intentions to engage in friend guard- ing. This is not to say that third-party threats evoke neither sadness nor anger. Rather, we simply doubt that sadness and/or anger, alone, would be responsible for motivating friend guarding—a prediction consistent with existing emotions work. For example, to the best of our knowledge, no existing, parsimonious theorizing argues that sadness or anger, alone, spur partner guarding; by contrast a sizable body of work on romantic jealousy has produced cogent theorizing and supportive empirical evidence that romantic jealousy causes people to engage in mate guarding (see, e.g., Buss, 2013).

The third aim of Study 2 is to discriminate between types of friendship maintenance behavior. As noted above, friendship maintenance is a two-pronged challenge that involves both miti- gating a friend’s loss or defection to someone else (the focus here) but also maintaining a friend’s continued investment in the bond (the primary focus of existing friendship research). Whereas we

term behavior aimed at preventing the defection of a friend to a third party “friend guarding”—likely comprised of tactics such as vigilance (e.g., knowing if the best friend is with the potential interloper; Buss et al., 2008)—we term behavior aimed at securing a friend’s continued investment in the relationship “everyday friend retention”—likely comprised of tactics such as making one another laugh and being open with one another (e.g., Canary et al., 1993; Oswald et al., 2004). To explore this, we added a third condition: a no-information control condition. Participants re- ported their intent to engage in both types of friendship mainte- nance in response to the threatened loss of the friendship alone, a third-party threat, or a no-information control. (Participants in the no-information control condition did not report their affective reactions to the conditions).

We predict that (c) people will report greater intentions to engage in friend guarding-vigilance when third-party threats are salient (loss to a third-party condition) than not (no-information control condition). We do not have strong predictions about pos- sible differences in friend guarding between the loss to a third party and loss alone conditions. On one hand, the loss alone condition makes no mention of a third party, implying people might not increase their vigilance to possible interlopers. However, on the other hand, the friendship is still waning, and third-party interference may be one assumed reason why; thus, people might increase their vigilance to possible interlopers—though, still likely not to the same extent that they would when a third-party is explicitly specified.

We also explore whether (d) people report greater intentions to engage in everyday friend retention in the no-information control condition—reflecting the normal, nonthreatened state of friend- ship—than when third-party threats are salient. (We do not make strong predictions about levels of everyday friend retention in the loss alone vs. other conditions.) We explore this because, as noted above, time and energy are inelastic resources, both of which are required to maintain relationships. People might be best served by efficiently deploying these finite resources— engaging in the be- havior that seems the best tool for the job. When the job is continued friend retention (in the absence of third-party threats), engaging in vigilance over and above everyday friend retention behavior may be superfluous. Note, however, that both friend guarding and everyday friend retention may be usefully recruited when friendships are threatened by third parties. Indeed, we would not assert that these different types of friendship maintenance are completely orthogonal. To the extent that everyday friend retention tactics help people maintain friends’ continued investment, third- party threats and consequent friendship jealousy could also spur everyday friend retention.

To investigate this, participants completed a two-part experi- ment in which they were randomly assigned to one of three conditions. After filling in the names of same-sex others—includ- ing best friends, which would again be piped into later scenarios— participants reported their reactions to the threatened loss of their best friends (loss alone condition), the threatened loss of their best friends to third parties (loss to third party condition), or were shunted forward in the study without being given a scenario and reporting emotional responses (no-information control condition). The second section of the study assessed intentions to engage in friendship maintenance, with participants in all three conditions

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

983FRIENDSHIP JEALOUSY

reporting behavioral intentions to engage in friend guarding and everyday friend retention.

Method

Participants. We determined that a sample size of approxi- mately 330 participants was necessary to achieve .80 power to detect small-to-medium effects (f � .15) of differences in friend- ship maintenance types (friend guarding; everyday friend reten- tion) across the three experimental conditions, assuming a .5 correlation between friendship maintenance measures. United States-residing participants were recruited into an approximately 9-min survey on TurkPrime for small monetary compensation. Whereas over 360 participants began our survey, only 243 (127 female, Mage � 36.92, SDage � 11.05) filled out focal dependent variables and passed two data checks. One of these checks was critical to the study, and required participants to have followed directions in writing the first names of three same-sex friends (best friend, close friend, and acquaintance). People failed this check if, for example, they wrote the same name for all friends (e.g., Dan, Dan, and Dan), if they used names that were clearly not all same-sex (e.g., Eric, Dave, and Annie), or if they failed to fill out a name for the best friend. A sensitivity analysis suggests this sample size yielded .80 power to detect an effect size of f � .17.

Procedure and design. Participants were randomly assigned to one of three experimental conditions. Two of these conditions echoed those in the previous experiment (loss alone, loss to a third party); the new condition was a no-information control condition. All participants first reported some brief demographic information and filled in the first names of a same-sex best friend, close friend, and acquaintance. Participants then filled out information about the activities and functions of their best friendship and acquaintance- ship, as in Study 1. Only best friend information was germane to the study; as in Study 1, we included other questions to obscure the focal aims.

Reactions. In the first main survey section, participants in the loss alone and loss to third party conditions reported their reactions to the prompt. Responses assessed included friendship jealousy (“jealousy”), often-concomitant emotions (“sadness,” “anger”), and distractor emotions (“proud,” “afraid,” and “pity”), all of which appeared in random order and were assessed on a 7-point Likert scale (1 � not at all, 7 � very much). Because reporting these reactions in the no-information control condition would have seemed nonsensical, participants in the no-information control condition were advanced to the next survey section directly after providing their friends’ names and friendship activities and func- tions.

Friendship maintenance. In the second main survey section, participants in all three conditions reported how likely they would be to engage in friendship maintenance— both friend guarding- vigilance and everyday friend retention. For example, participants in the loss to a third party condition were prompted, “In that situation—where [name of best friend] is becoming potentially closer with a new friend than [name of best friend] is with you— how likely would you be to do each of the following?”, whereas those in the control condition were prompted, “In your best friendship with [name of best friend], how likely would you be to do each of the following?” In all, participants responded to

11 items, appearing in random order, on a 7-point Likert scale (1 � not at all, 7 � very much).

To assess behavioral intentions to engage in the friend-guarding tactic of vigilance to third party threats, we adapted two items from existing measures of mate guarding (Buss et al., 2008; Shackelford et al., 2005) by drawing from the Mate Retention Inventory— Short Form (MRI-SF; Buss et al., 2008): “Pay attention to whether [name of best friend] was spending time with a new best friend,” “Try to figure out if [name of best friend] prefers a new friend’s company to my company” (� � .81–.88).

To assess behavioral intentions to engage in everyday friend retention, we adapted nine items from existing work (Canary et al., 1993; Oswald et al., 2004), reflecting six everyday friend retention tactics: being positive (“Be cheerful and positive whenever you’re with [name of best friend]”), being open (e.g., “Have open dis- cussions with [name of best friend]” [� � .76 –.90]), assurances (“Assure each other about the importance of your friendship”), social networks (“Rely on other friends to help you through this rough patch”), avoidance (e.g., “Avoid talking about things that we disagree about” [� � .79 –.86]), and humor (“Try to make [name of best friend] laugh”). These were aggregated into one measure of everyday friend retention (� � .81).

Other items. Participants also completed other exploratory and demographic variables we did not analyze (e.g., intrasexual competitiveness, ethnicity).

Results

Is friendship jealousy (vs. sadness, anger) uniquely evoked by third party threats? Yes. To assess this, we conducted a 2 (Condition: loss alone, loss to a third party) � 3 [Reaction: jealousy, sadness, anger] mixed-factors ANOVA. This yielded significant main effects of Condition, F(1, 162) � 6.07, p � .015, �p

2 � .036, and Reaction, F(2, 324) � 77.87, p � .001, �p2 � .325, as well as a significant interaction, F(2, 324) � 76.38, p � .001, �p

2 � .320. (Means [SEs] for all main and distractor responses are reported in Table S2 in the online supplemental materials).

Exploring the interaction, first, we replicate findings from Study 1, and we support the prediction that friendship jealousy is more strongly evoked by third-party threats to friendships: People re- ported greater friendship jealousy in the third-party threat condi- tion (M � 3.57, SE � .19) than in the loss alone condition (M � 2.60, SE � .21), F(1, 162) � 12.09, p � .001, �p2 � .069, 95% CI [0.42, 1.54] (see Figure 2a). Providing discriminant evidence, people reported both greater sadness in the loss alone condition (M � 5.07, SE � .20) than in the third-party threat condition (M � 3.21, SE � .19), F(1, 162) � 47.25, p � .001, �p2 � .226, 95% CI [1.32, 2.39], and also greater anger in the loss alone condition (M � 3.17, SE � .19) than in the third-party threat condition (M � 2.30, SE � .18), F(1, 162) � 11.00, p � .001, �p2 � .064, 95% CI [0.35, 1.39]. Additionally, people in the third-party threat condi- tion reported greater friendship jealousy than other reactions (ps � .030), whereas people in the loss alone condition reported greater sadness than other reactions (ps � .001).

Assessing different types of friend retention. We conducted a 3 (Condition: control, loss alone, loss to a third party) � 2 [Friendship maintenance tactic: friend guarding, everyday friend retention] mixed-factors ANOVA to explore whether condition affected behavioral intentions to engage in these two types of

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

984 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

friendship maintenance. This yielded a significant main effect of Tactic, F(1, 235) � 43.50, p � .001, �p2 � .156, and a significant interaction, F(2, 235) � 10.28, p � .001, �p2 � .080. The main effect of condition was not significant (p � .558). (Means [SEs] for all individual tactics are reported in Table S3 in the online supplemental materials).

Exploring the interaction, we first tested the prediction that people report greater intentions to engage in friend guarding when third-party threats to friendships are salient (loss to a third party) than not (control condition). Specifically, we expected— and found— greater intent to engage in friend guarding when friendships were threatened by third parties (M � 3.97, SE � .20) compared with the control condition (M � 3.32, SE � .21), F(1, 161) � 4.73, p � .031, �p2 � .029, 95% CI [0.06, 1.24] (see Figure 2b). We additionally found a not-predicted trend to report greater intent to engage in friend guarding when friendships were threat- ened by third parties compared with the loss alone condition (M � 3.49, SE � .20), F(1, 160) � 3.08, p � .081, �p2 � .019, 95% CI [�0.06, 1.03]. There were no significant differences in reported intent to engage in friend guarding between the loss alone and control conditions (p � .593).

We also find that people reported greater intent to engage in everyday friend retention in the control condition (M � 4.63, SE � .13) than in the loss to a third party condition (M � 4.11, SE � .12), F(1, 161) � 8.78, p � .004, �p2 � .052, 95% CI [0.17, 0.87]— or in the loss alone condition (M � 4.15, SE � .14), F(1, 149) � 6.21, p � .014, �p2 � .040, 95% CI [0.10, 0.85], though that was not predicted a priori. There were no significant differences in reported everyday friend retention intent between the loss alone and loss to a third party conditions (p � .817).

Recall that we did not predict that, when faced with third-party threats, people would report greater intentions to engage in friend- guarding than everyday friend retention, as such threats should motivate behavior aimed at each of the central challenges of friend retention (both preventing defection and securing continued in-

vestment). We do not find a significant difference between these types of friend retention behavior in the loss to a third party condition (p � .435). We did find, in both the control and loss alone conditions, that people reported greater intentions to engage in everyday friend retention behavior than in friend guarding— vigilance, Fcontrol(1, 235) � 48.10, p � .001, �p

2 � .170, 95% CI [0.94, 1.68], Floss alone(1, 235) � 12.36, p � .001, �p

2 � .050, 95% CI [0.29, 1.04]. Also, as one might expect, this difference was larger in the control than the loss alone condition. This makes some sense; whereas everyday friend retention behavior should be endorsed across conditions, specifically friend guarding should be amplified only when third-party threats are salient. Taken together, these findings provide further evidence for the specificity of the behavioral outputs of friendship jealousy—that is, friend guarding.

What emotional reactions drive friend guarding? In our view, third-party threats to valued friendships evoke friendship jealousy, which, in turn, motivates people to engage in friend guarding. (Again, this is not to say that third-party threats and/or friendship jealousy do not also motivate engagement in everyday friend retention tactics; in fact, we would expect this to the extent that these tactics help maintain the friendship by securing friends’ continued investment.) To test this and, thus, underscore the pre- dicted specificity of the inputs and outputs of friendship jealousy versus sadness and anger, we performed a parallel multiple medi- ation analysis (Hayes, 2017). We tested whether the relationship between condition (loss alone, loss to a third party) and intent to engage in friend guarding was statistically mediated by friendship jealousy, sadness, and/or anger. We used 5,000 bootstrapped iter- ations to compute a bias corrected 95% CI for the indirect effects.

As shown in Figure 3, the effect of condition on friend guarding was significantly mediated only by reported friendship jealousy, b � �0.43, SE � 0.17, 95% CI [�0.82, �0.16]; the indirect effects of sadness and anger were not significant, nor was the direct effect, b � 0.31, SE � 0.30, p � .293. This supports our prediction that friendship jealousy mediates the effect of condition

1

2

3

4

5

6

Jealousy Sadness Anger

tne mesrodnE

de tro peR Re po

rt ed

En do

rs em

en t

1

2

3

4

5

6

Friend guarding - vigilance

Everyday friend reten�on

Loss alone

Loss to a third party

Control

(a) (b)

Figure 2. From Study 2: (a) Reported levels of friendship jealousy and related emotions of sadness and anger in reaction to losing a best friendship (alone) versus losing a best friendship to a third party (participants in the control condition did not complete these measures). (b) Reported behavioral intentions to engage in friend guarding—vigilance and everyday friend retention by condition (loss alone, loss to third party, and no- information control). Error bars represent SEs.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

985FRIENDSHIP JEALOUSY

on reported intentions to friend guard. Moreover, this finding also mitigates concerns that sadness and/or anger might instead be responsible for driving friend guarding. (See the online supple- mental materials for a similar analysis of everyday friend reten- tion).

Discussion

Study 2 supported several key predictions about friendship jealousy and also addressed the concern that emotions other than friendship jealousy might motivate friend guarding. First, we rep- licated the pattern of responding from Study 1, further suggesting that friendship jealousy (but not sadness and anger) is strongly evoked by third-party threats. Second, friendship jealousy (but not sadness or anger), in turn, motivated intentions to engage in a tactic of friend guarding (vigilance) thought well-designed to help defend against third-party threats to relationships (e.g., Buss et al., 2008). Findings mitigate concerns that emotions often concomitant with jealousy—sadness and/or anger—might instead motivate friend guarding. Third, we also find that intention to engage in friend guarding was greatest in the loss to a third party condition, again consistent with predictions and the overall putative model. (In contrast, intention to engage in everyday friend retention was greatest in the control condition).

Studies 3a and 3b

Studies 3a and 3b facilitated three main goals. First, we tested the prediction that (a) friendship jealousy is sensitive to friend value (i.e., closeness), such that the prospective loss of closer friends evokes greater friendship jealousy.

Second, we test between intuitive and functionally derived pre- dictions as to which cues friendship jealousy is more strongly calibrated—time versus replacement threat. We do so in multiple complementary ways. One way that we do this—in both Studies 3a and 3b—is by varying the type of relationship between one’s best friend and the potential interloper. Consider two situations wherein a person might be expected to react with jealousy: In one situation their best friend forms a new romantic relationship; in a parallel situation, their best friend forms a new same-sex friendship. A new

romantic relationship often makes steep demands on one’s time (Dunbar, 2012; Johnson & Leslie, 1982; Milardo et al., 1983; Roberts et al., 2009); presumably, time a best friend spends with a romantic partner cannot also be spent with a friend. Thus, one might intuitively expect a person to feel strong friendship jealousy at a best friend’s forming a new romantic relationship—particu- larly if such jealousy were driven by the prospective loss of time spent with the best friend (“time threat”).

However, a person’s best friend and romantic partner likely fulfill some distinct functions. Because a new romantic partner is less likely than a new friend to usurp one’s place in their best friendship, one will feel less friendship jealousy when their best friend forms a new romantic relationship than a new same-sex friendship—particularly if such jealousy were driven by the pro- spective usurpation of one’s best friendship (“replacement threat”). Note also that this prediction makes a clear, though rather obvious, distinction between jealousy over friends versus romantic partners; that is, one might expect romantic jealousy to be evoked more strongly when interlopers are prospective mates, and friendship jealousy to be evoked more strongly when interlopers are prospec- tive friends.

In summary, we predict that (b) interlopers who are romantic partners evoke less friendship jealousy than those who are same- sex friends. We find support for this prediction is Study 3a, and examine a potential confound linked to those findings in Study 3b: One might wonder whether people’s implicit conceptions of a best friend’s new romantic relationship—specifically as being short- or long-term in nature—might play a role in lowered friendship jealousy when interlopers are romantic partners (vs. same-sex friends). Thus, in Study 3b, we make explicit the type of romantic relationship in which the best friend is engaged. Because long-term partners clearly pose a greater replacement threat than short-term partners (e.g., one-night stands), we predict that (c) long-term partners will evoke greater friendship jealousy than short-term partners. However, we still expect that interlopers who are friends will evoke greater friendship jealousy than inter- lopers described as (short- or long-term) romantic partners. Addi- tionally, in Study 3b we extend this by testing the prediction that

Condition: Loss to third party, Loss

alone

Jealousy b=-0.43, SE=0.17

95%CI=[-0.82, - 0.16]

Sadness b=0.13, SE=0.15

95%CI=[-0.17, 0.44]

Anger b=0.14, SE=0.10

95%CI=[-0.02, 0.40]

Friend guarding - vigilance

b = -0.94, p = .001

b = 1.82, p < .001

b = 0.87, p = .001

b = 0.46, p < .001

b = 0.07, p = .412

b = 0.16, p = .097

Figure 3. Parallel multiple mediation model depicting the effect of condition (loss to a third party, loss alone) on friend guarding-vigilance as mediated by endorsement of reactions (jealousy, sadness, and anger) from Study 2.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

986 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

(d) close friends’ new friendships also evoke greater friendship jealousy than do their new romantic relationships.

In Study 3b, we also provide an additional, complementary exploration of whether friendship jealousy is more strongly cali- brated to replacement than time threat. We do this by keeping constant the type of relationship but varying the features of the interloper; interlopers are either same- or other-sex friends. Be- cause a person’s same- and other-sex friends may serve some distinct functions (e.g., Bleske & Buss, 2000; Buss & Schmitt, 1993; Lewis et al., 2011), same-sex friends might be perceived to pose greater replacement threats (but similar time threats) than other-sex friends. Thus, we predict that (e) a best friend’s new same-sex friend will evoke greater friendship jealousy than will a best friend’s new other-sex friend. This provides an additional, though again obvious, distinction from romantic jealousy; for the majority of people (i.e., heterosexuals), romantic jealousy is evoked by the threat of an other-sex (vs. same-sex) rival, whereas friendship jealousy should be more strongly evoked by a same-sex (vs. other-sex) rival. Thus, in both Studies 3a and 3b, we test the prediction that interlopers presumably posing greater replacement threats (same-sex friends vs. other-sex romantic partners or other- sex friends) evoke greater friendship jealousy.

Popular reporting has highlighted the particular concern that, were one to introduce their best friend to their close friend, those two might hit it off and become closer to one another (e.g., Rosenfeld, 2004). Thus, for ecological validity and for exploratory purposes, we include two types of same-sex interlopers here: a same-sex stranger and a same-sex other who is the participant’s own close friend (and who does not have an existing relationship with the focal best friend). Whereas we would still predict that either type of same-sex friend will evoke greater friendship jeal- ousy than an other-sex romantic partner, we do not have predic- tions about the relative levels of friendship jealousy evoked when the interloper is a same-sex stranger versus one’s own close friend. On one hand, the anecdotal reporting mentioned above implies that this is a very perturbing situation, and one might speculate that, should the best friend and one’s own close friend hit it off and become potentially closer to one another, one might be doubly replaced and feel greater friendship jealousy. However, on the other hand, research suggests that such network consolidation—as when our previously unacquainted friends become friends with one another— can increase the likely longevity of the focal friend- ship (Benenson, 2014; Benenson, Nicholson, Waite, Roy, & Simp- son, 2001). Thus, the friendship between one’s best and one’s close friend might decrease concern about losing either friendship and, ultimately, attenuate friendship jealousy.

Third, we again tested the prediction that (f) friendship jealousy positively predicts behavioral intentions to friend guard (in Study 3a). Here, we expand our measurement of friend guarding beyond vigilance by creating a novel friend guarding scale, adapting from mate guarding work (MRI-SF; Buss et al., 2008; Shackelford et al., 2005). The MRI-SF assesses engagement in 19 mate guarding tactics (e.g., monopolization, threaten rivals, and derogate mate to rivals). Our friend guarding scale assesses intent to engage in 12 tactics (e.g., vigilance, separation, monopolization) that represent reconceptualizations and/or combinations of mate-guarding tac- tics. Whereas we do not have strong predictions about the relative employment of different friend guarding tactics, we do predict that (g) people will report greater intentions to guard their best friends

against interlopers presumably posing greater replacement threats (i.e., those who are friends vs. new romantic partners).

Method

Participants. Study 3a. Pilot data from a U.S. online community sample

suggested that a sample size of 150 participants was suitable to detect significant differences in reported friendship jealousy be- tween focal scenarios (at .80 power for effects f .20). Because we also planned to explore our novel friend guarding measure in this study, and would need participants to spend a great deal more time on the survey, we recruited into an hour-long survey as many undergraduate participants as possible during a single term at a university in the U.S., aiming to at least double the sample size used in the pilot. We ultimately recruited 483 individuals who filled out sex information, with 466 individuals (263 female; Mage � 19.06, SDage � 2.04) completing focal measures of a three-part task for course credit. Post hoc power analysis suggests this gave us power ( .99) to detect small effects assuming a .5 correlation among measures.

Study 3b. Participants (N � 303; 125 female; Mage � 19.44, SDage � 1.32) were recruited from an undergraduate pool and participated after first completing an unrelated study on stigma (involving no priming) and a distractor task, with 289 participants (122 female) completing focal dependent variables. The focal task took approximately 20 min. Sample size was dictated by power analyses for the unrelated study. Post hoc analysis reveals this yielded .99 power to detect small effects, assuming .5 correlation between measures.

Procedure. In the first part of the task, participants reported basic demographics (age, sex) and a range of information about same-sex friends and friendships. Specifically, participants re- ported the first name and last initial of same-sex others in their day-to-day lives: a best friend, a close friend who was not already friends with the best friend (close friend #1), a second close friend who was not already friends with the best friend (close friend #2), and an acquaintance. Each relationship type was described (e.g., “A best friend is the ONE person to whom you are closest, a person who would help you in dire times”). Participants were asked, “What is the first name and last initial of your SAME-SEX best friend? (If you have a best friend from growing up, but don’t see this person on a daily basis, please tell us about your best friend from school or work instead. If you have a same-sex romantic partner you consider your ‘best friend’ please INSTEAD choose a same-sex best friend in whom you are not sexually interested.).”

Friendship scenarios. In the experimental section of the stud- ies, participants were asked to imagine different scenarios regard- ing friends forming a new, potentially closer relationship with another person. Each scenario used the names of the friends reported in the first part of the task. To give a specific example, participants were instructed: “Imagine that [Your Acquaintance] and another same-sex person met up and started to really enjoy one another’s company. You haven’t previously met this other person, but [Your Acquaintance] is spending a lot of time with them. They’re becoming fast and close friends—maybe even closer with one another than you are with [Your Acquaintance].”

Study 3a. In Study 3a, participants imagined and reacted to five scenarios, presented in random order. There were three sce-

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

987FRIENDSHIP JEALOUSY

narios with friends of varying closeness (acquaintance, close friend #1, best friend) forming a new, potentially closer friendship with a same-sex stranger. These scenarios allowed us to test predictions about friendship jealousy being calibrated to friend value (i.e., closeness).

Two additional scenarios asked participants to imagine the best friend forming new, potentially closer relationships with the par- ticipant’s own close friend (close friend #2), and with a new romantic partner. Along with the scenario in which the best friend forms a new, potentially closer friendship with a same-sex stranger, these scenarios allowed us to test predictions about friendship jealousy being calibrated to replacement threat, as in- stantiated via interloper type.

Study 3b. Study 3b largely replicated the design of Study 3a with three modifications. The first modification further ex- plored replacement threat by comparing interlopers who were same-sex versus other-sex new friends (i.e., keeping interloper type constant and instead varying replacement threat via fea- tures of the interloper). The other two modifications explored the role of replacement threat by further assessing the predicted difference in friendship jealousy toward interlopers who are same-sex friends versus romantic partners. First, participants were also asked to imagine close friends forming new, poten- tially closer romantic relationships (length unspecified). Sec- ond, participants were asked to imagine two related, specific scenarios in which the best friend formed a new romantic bond: one in which the best friend formed a new, potentially closer short-term romantic relationship (e.g., a one- or few-night stand), and one in which the best friend formed a new, poten- tially closer long-term romantic relationship.

In all, participants imagined seven scenarios. Four involved target individuals forming new, potentially closer friendships: an acquaintance with a same-sex stranger; a close friend with a same-sex stranger; a best friend with a same-sex stranger; a best friend with an other-sex stranger. Again, this allows us to test whether friendship jealousy is calibrated to friend closeness and replacement predictions. Three involved target individuals forming new, potentially closer romantic relationships—a close friend forming a new, potentially closer romantic relationship; a best friend forming a new, potentially closer romantic rela- tionship with a short-term partner; and a best friend forming a new, potentially closer romantic relationship with a long-term partner. In comparison with other scenarios, these help us test whether friendship jealousy is calibrated to replacement threat versus time threat (as does comparing best friends with same- versus other-sex friends). Not all possible scenarios in a fully “balanced” design are included (e.g., close friends forming a new, potentially closer friendship with an other-sex stranger); we included only those that replicated or critically extended Study 3a.

Reported friendship jealousy. After each scenario, partici- pants reported emotional responses. In Study 3a, responses in- cluded friendship jealousy, commonly associated emotions (anger, sadness), and also distractor reactions (happy, proud, relieved, dismayed, guilty, resentful, and disgusted). In Study 3b, responses assessed included friendship jealousy, anger, sadness and also happiness, fear, and pride. Reactions were assessed on 7-point Likert-scales (1 � not at all, 7 � very much).

Friend-guarding intentions (Study 3a). In Study 3a, we as- sessed behavioral intentions to engage in friend guarding. After completing parts one and two of the experiment, participants were asked to recall three scenarios (the best friend becoming poten- tially closer with a same-sex stranger, with close friend #2, and with a new romantic partner), in randomized order. After recalling and reimmersing themselves in each scenario, participants filled out a 44-item friend guarding scale (e.g., “How likely would you be to . . . monopolize your best friend’s time?”; � � .97) that we created by adapting the MRI-SF (Buss et al., 2008; Shackelford et al., 2005). Friend-guarding intentions were assessed on a 7-point Likert-scale (1 � not at all likely, 7 � very likely). See Appendix A for all items and tactics.

Results and Discussion

Study 3a. An omnibus test exploring reported friendship jeal- ousy across all scenarios revealed a significant effect of scenario, F(4, 1860) � 288.93, p � .001, �p2 � .383.

5

Study 3b. We first ran an omnibus test assessing differences in reported friendship jealousy across the seven scenarios, F(6, 1728) � 126.11, p � .001, �p2 � .305.

Does friendship jealousy vary as a function of friend value (i.e., closeness)? Yes. The highest levels of friendship jealousy were evoked in scenarios where new, potentially closer friend- ships were formed between best friends and strangers, followed by close friends and strangers, followed by acquaintances and strangers.

Study 3a. Consistent with predictions, participants reported less friendship jealousy at the prospective loss of acquaintances (M � 1.33, SE � .04) than close friends (M � 2.27, SE � .08, p � .001, 95% CI [�1.09, �.79], �p2 � .238) or best friends (M � 3.47, SE � .09; p � .001, 95% CI [�2.34, �1.96], �p2 � .512), and additionally reported less friendship jealousy at the prospective loss of close friends than best friends (p � .001, 95% CI [�1.39, �1.03], �p2 � .271; see Figure 4a).

Study 3b. Replicating Study 3a, the prospective loss of best friends to same-sex strangers evoked greater reported friendship jealousy (M � 3.95, SE � .10) than did the prospective loss of close friends to same-sex strangers (M � 2.40, SE � .09, p � .001, 95% CI [2.161, 2.628], �p2 � .414), and the prospective loss of close friends to same-sex strangers evoked greater reported friend- ship jealousy than did the prospective loss of acquaintances to same-sex strangers (M � 1.57, SE � .07, p � .001, 95% CI [.651, 1.030], �p2 � .208; see Figure 4b).

Does friendship jealousy vary as a function of “replacement threat”? Yes. As predicted, interlopers likely posing a greater replacement threat (but perhaps a lesser time threat; same-sex strangers, participants’ own close friends) evoked greater friend- ship jealousy than interlopers likely posing a lesser replacement threat (but perhaps a greater time threat; a romantic partner).

5 Given the focus of the current work, and in light of the results of Studies 1 and 2, we discuss at length only findings for friendship jealousy in Studies 3a– 8b. Means (SEs) for other responses are reported in the online supplemental materials. We assess various emotional responses across studies. In general, most threats to friendship prompt (nuanced patterns of) friendship jealousy and often-concomitant negative emotions, whereas friends forming new romantic relationships additionally prompt some positive responses (e.g., pride, happiness).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

988 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

Study 3a. People also reported significantly lower levels of friendship jealousy at the best friend forming a new romantic relationship (M � 2.24, SE � .08) versus a new friendship with either a stranger (M � 3.47, SE � .09, p � .001, 95% CI [�.1.41, �1.07], �p2 � .306) or with one’s own close friend (M � 4.00, SE � .09, p � .001, 95% CI [�1.94, �1.58], �p2 � .439). This again implies that friendship jealousy may be more strongly calibrated to replacement than time threat. Although we did not predict this a priori, we also find that people reported greater friendship jealousy when the interloper was their own close friend than a same-sex stranger (p � .001, 95% CI [0.35, 0.65], �p2 � .130).

Study 3b. Extending results from Study 3a, we replicate this pattern for the prospective loss of close friends: people reported significantly greater friendship jealousy at the close friend forming a new, potentially closer friendship (M � 2.40, SE � .09) than a new, potentially closer romantic relationship (M � 1.91, SE � .08), F(1, 295) � 32.17, p � .001, 95% CI [0.31, 0.64], �p2 � .098.

People also reported greater friendship jealousy at the best friend’s new friendship with a same-sex stranger (M � 3.95, SE � .10) than an other-sex stranger (M � 2.79, SE � .10), F(1, 289) � 135.22 p � .001, 95% CI [0.96, 1.35], (�p2 � .312). This further suggests that friendship jealousy is strongly calibrated to replacement threat (that is presumably greater for new same- than other-sex friends) and not necessarily time threat (that is presumably equal for new same- and other-sex friends; see Figure 5a).

Again replicating and extending results from Study 3a, people reported greater friendship jealousy at best friends’ new friend- ships with same-sex strangers (M � 3.95, SE � .10) than either best friends’ new short-term romantic relationships (M � 2.04, SE � .08), F(1, 297) � 240.19 p � .001, 95% CI [1.67, 2.16], �p

2 � .447, or long-term romantic relationships (M � 2.27, SE �

.09), F(1, 297) � 260.40 p � .001, 95% CI [1.50, 1.92], �p2 � .467. People reported greater friendship jealousy at the best friend’s forming a new long- versus short-term romantic relationship, F(1, 298) � 6.62 p � .011, 95% CI [0.05, 0.37], (�p2 � .022).

Friend-guarding intentions (Study 3a). Does friendship jealousy positively predict behavioral inten-

tions to friend guard? We first created a best friend jealousy composite, consisting of participants’ reported jealousy in reaction to each scenario wherein best friends formed new, close bonds (with same-sex strangers, own close friends, romantic partners; � � .77), and a best friend friend-guarding intentions composite, consisting of reported behavioral intentions to friend guard in reaction to each of these scenarios (� � .99). Consistent with predictions, this aggregated measure of best friendship jealousy was significantly, positively, and strongly associated with reported intent to guard best friends, r(468) � .527, p � .001. We note, however, as we discuss below, that intentions to friend guard are low (but see Studies 2, 5b, 6, 8a, and 8b).

We also conducted other tests supporting the notion that friend- ship jealousy predicts friend guarding, even over and above sad- ness and anger (see online supplemental materials).

Do friend-guarding intentions vary in accordance with predictions? Yes. We conducted a 3 [Interloper] � 12 [Tactic] repeated measures ANOVA, which yielded main effects of Inter- loper, F(2, 924) � 91.34, p � .001, �p2 � .165, and Tactic, F(11, 5082) � 141.56, p � .001, �p2 � .235, as well as an Interloper � Tactic interaction, F(22, 10,164) � 16.15, p � .001, �p2 � .034.

First, aggregating over the 12 tactics, as predicted, people re- ported significantly greater intentions to guard against interlopers who were same-sex strangers (M � 1.88, SE � .05) than romantic partners (M � 1.50, SE � .03), F(1, 462) � 139.57, p � .001, �p

2 � .232, 95% CI [0.32, 0.44]. Similarly, people reported signif- icantly greater intentions to guard against interlopers who were their own close friends than romantic partners, F(1, 462) � 72.77

1

2

3

4

5

Re po

rt ed

Fr ie

nd sh

ip Je

al ou

sy

1

2

3

4

5

(a) (b)

Figure 4. In both (a) Study 3a and (b) Study 3b, the lowest levels of friendship jealousy were reported when one’s acquaintance became potentially closer with a same-sex stranger, followed by when one’s close friend became potentially closer with a same-sex stranger, and the highest levels were reported when one’s best friend became potentially closer with a same-sex stranger. Error bars reflect SEs.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

989FRIENDSHIP JEALOUSY

p � .001, �p2 � .136, 95% CI [0.18, 0.28]. Additionally, people reported greater intention to friend guard when interlopers were same-sex strangers than their own close friends (M � 1.73, SE � .04), F(1, 465) � 36.10, p � .001, �p2 � .072, 95% CI [0.10, 0.20].

Second, exploring the interaction reveals that this overall pattern largely holds across tactics of friend guarding (see Table 1). In general, different rivals cause people to report different levels of friend-guarding intent; and these intentions are greatest in response

to rivals best positioned to usurp one’s friendship (i.e., same-sex rivals as compared with romantic partners).

Does friendship jealousy statistically mediate the relationship between presumed replacement threat (instantiated via rival type) and friend-guarding intentions? Yes. Because we reason that new friends pose greater replacement threats than do new romantic partners, we predict that best friends forming new friendships versus romantic relationships would evoke greater friendship jeal-

Table 1 Intentions to Engage in 12 Friend-Guarding Tactics (M, SE) Across Types of Rivals

Tactic

Rivals

Same-sex stranger Own close friend Romantic partner

Vigilance 2.18 (.07)a 2.07 (.07)b 1.83 (.06)c Separation 2.06 (.06)a 1.87 (.06)b 1.45 (.04)c Monopolization 1.99 (.07)a 1.80 (.06)b 1.40 (.04)c Induce jealousy 2.00 (.06)a 1.91 (.07)b 1.52 (.05)c Punish/threaten friend 1.58 (.04)ab 1.53 (.04)ab 1.34 (.04)c Emotional manipulation 1.77 (.05)a 1.68 (.05)b 1.44 (.04)c Derogate rival 1.65 (.05)a 1.38 (.04)bc 1.32 (.03)bc Self/commitment enhancement 2.48 (.06)a 2.35 (.06)b 2.00 (.06)c Possession signals 2.41 (.08)a 2.22 (.07)b 1.99 (.07)c Derogation of own friend 1.35 (.05)a 1.29 (.03)b 1.21 (.04)c Direct aggression toward rival 1.32 (.04)a 1.27 (.03)b 1.19 (.03)c Indirect aggression toward rival 1.77 (.05)ab 1.39 (.04)ab 1.32 (.04)c

Note. Means sharing subscript do not differ significantly (p � .05).

Figure 5. Reported friendship jealousy in reaction to multiple scenarios in which same-sex best friends form new, potentially closer relationships with third parties from (a) Study 3b and (b) Study 4, and (c) retrospectively, recalled friendship jealousy in reaction to multiple lived experiences when best friends formed close same-sex friendships, close romantic relationships, or same-sex acquaintanceships (from Study 5a). Error bars represent SEs.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

990 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

ousy, and, in turn, increased intentions to friend guard. Specifi- cally, then, a same-sex stranger—and perhaps also the partici- pant’s own close friend—should evoke greater replacement threat than romantic partners and, thus, we expect friend-interlopers to evoke greater friendship jealousy and, in turn, friend guarding, than the romantic-interlopers.

We conduct tests of statistical mediation two ways. First, we aggregate across both same-sex friend-interlopers—same-sex strangers and participants’ own close friends—in computing friendship jealousy and friend guarding, comparing these aggre- gates against scores in response to romantic-interlopers. Second, we compare only the same-sex friend interloper against the romantic- interloper. Both tests yield the same pattern of results, supporting predictions (see the online supplemental materials for the latter test). To explore this—and, thus, our model for friendship jealousy (see Figure 6)—we used MEMORE, an SPSS macro for conducting mediation analyses on within-subjects variables (Montoya & Hayes, 2017).

To compare friendship jealousy and friend guarding in reaction to friendships versus romantic relationships, we aggregated re- ported friendship jealousy when best friends formed new, poten- tially closer friendships (i.e., with same-sex strangers and with participants’ own close friends; � � .79) and we similarly aggre- gated reported friend guarding for friendships (i.e., when interlop- ers were same-sex strangers and participants’ own close friends; � � .98). We then entered the two measures of reported friendship jealousy—when best friends form new friendships (aggregated) versus new romantic relationships—as mediators of the relation- ship between rival type and friend-guarding intentions using 5,000 bootstrapped iterations to compute a bias corrected 95% CI for the indirect effects. This yielded a significant indirect effect (b � .13, SE � 0.03; 95% CI [0.09, 0.19]), suggesting that greater friendship jealousy is evoked when interlopers are friends versus romantic partners, and, in turn, this friendship jealousy motivates greater behavioral intentions to guard the best friendship. The direct effect

remained significant, b � .17, SE � 0.03, p � .001, 95% CI [0.11, 0.23] (see Figure 6). Parallel multiple mediation analysis further suggests that friendship jealousy— but not sadness and anger— predicts friend-guarding intentions (see online supplemental ma- terials).

Discussion

Studies 3a and 3b yield support for several key predictions. Regarding the inputs of friendship jealousy, we find that it is calibrated to cues of friend value—such that threats to more valuable (i.e., closer) friendships evoke greater friendship jealo- usy—and also seems to prioritize cues of replacement threat over time threat, a proposition we test more fully in subsequent exper- iments. Regarding the outputs of friendship jealousy, we find support for the prediction that it may spur friend guarding; we also test this further in subsequent experiments.

Additionally, we find that people report greater friendship jeal- ousy when interlopers are the participants’ own same-sex friends than when interlopers are strangers, but we also find that people seem more intent to friend guard against same-sex strangers than their own close friends. Why might this seeming inconsistency have occurred here?

We speculate that this strong friendship jealousy evoked by best friends becoming potentially closer with our existing friends might fail to drive similarly strong intentions to friend guard because, relative to internal feelings, this external behavioral output (friend guarding) can potentially harm the interloper, the existing best friendship, and the existing close friendship. Consider a person who engages in friend guarding by disparaging one of their own close friends to their own best friend (or vice versa); this tactic of friend guarding could harm the person’s reputation for being a good friend in the eyes of the friend being guarded, thereby threatening the person’s existing best friendship, and if the inter- loper were to find out about this disparagement—as when the best

Replacement Threat (instantiated via

rival type)

Friendship Jealousy

Friend-guarding Behavioral Intentions

Figure 6. Proposed statistical mediation model showing that reported friendship jealousy statistically mediates the relationship between replacement threat, instantiated via interloper type (same-sex friends, other-sex romantic partner), and intentions to engage in friend guarding from Study 3a.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

991FRIENDSHIP JEALOUSY

friend being guarded and the interloper become close—this could further threaten that person’s existing close friendship. By con- trast, such friend guarding would likely be far less costly when engaged in against a same-sex stranger. For example, disparaging a stranger (vs. one’s own close friend) to one’s best friend is less likely to make one look like a poor friend.

Study 4

Study 4 facilitated two main goals. First, we again test the prediction that friendship jealousy is more sensitive to replacement threat than time threat (by varying interloper type). Second, ex- tending Studies 3a and 3b, we test these predictions in a between- subjects manner. Participants reported their reactions to one of four hypothetical scenarios: same-sex best friends becoming potentially closer with a same-sex stranger, same-sex best friends becoming potentially closer with the participant’s existing close friend (who was not previously friends with the focal best friend), same-sex best friends becoming potentially closer with a romantic partner, or participants’ own romantic partners becoming potentially closer to an other-sex stranger (i.e., a mate poacher). (Although our theory does not make predictions about romantic jealousy, one might understandably wonder how levels of reported friendship jealousy and better-studied romantic jealousy compare to one another; thus, we also explore people’s reported romantic jealousy in response to the threatened loss of their romantic partners.) Conducting this study between subjects mitigates concerns that the within-subjects designs of Studies 3a and 3b elicited implicit and/or explicit comparisons, and were thus problematic for testing our predic- tions. One might also find the between-subjects design to be more ecologically valid.

Method

Participants. We determined a sample size of approximately 350 participants was necessary for .80 power to detect small-to- medium effects (f � .15) specifically in friendship jealousy be- tween conditions. We recruited 384 U.S. participants from TurkPrime who began an approximately 7-min survey for small monetary compensation; of those, 268 participants (139 female) passed two attention checks, reported sex information, and com- pleted focal measures. This yields .80 power to detect small to medium effects (f � .17).

Procedure and design. Similar to previous studies, partici- pants first filled out brief demographic information and filled in names of same-sex others, including best friends. Participants were then randomly assigned to read and respond to one of four hypothetical scenarios about: best friends becoming potentially closer to a same-sex stranger, best friends becoming potentially closer to the participant’s own same-sex close friend (whom the best friend did not previously know), best friends becoming po- tentially closer to an other-sex romantic partner, romantic partners becoming potentially closer to a person of the same sex as the participant. Unlike previous studies, participants in the condition wherein best friends form new romantic relationships were first instructed to imagine that their best friends were not currently in a romantic relationship. Participants reported their reactions—jeal- ousy, often-concomitant emotions of sadness and anger, and dis- tractor emotions (pity, fear, and pride)— on a 7-point Likert scale (1 � not at all, 7 � very much).

Results and Discussion

Does jealousy vary as a function of replacement threat? Yes. We replicate results from Studies 3a and 3b, which had within-subjects designs, here using a between-subjects design (and an adult community sample).

An omnibus test assessing reported jealousy across the four scenarios yielded a significant effect of scenario, F(3, 209) � 20.88, p � .001, �p2 � .231. Consistent with predictions, both instances when interlopers were same-sex friends evoked greater reported friendship jealousy (same-sex stranger: M � 3.91, SE � 0.24; same-sex own close friend M � 3.43, SE � 0.25) than when the interloper was an other-sex romantic partner (M � 2.64, SE � 0.24), Fstranger(1, 109) � 14.39, p � .001, �p

2 � .117, 95% CI [0.62, 1.93] and Fclosefriend(1, 104) � 5.59, p � .020, �p

2 � .051, 95% CI [0.12, 1.47]. There was no significant difference in re- ported friendship jealousy when interlopers were same-sex friends (p � .116; see Figure 5b). This suggests that the within-subjects aspect of previous studies was not solely responsible for the predicted pattern of findings and, thus, mitigates the possible concern that our pattern of findings was because of a within- subjects design.

We also find that people reported significantly greater jealousy in scenarios where their romantic relationships were threatened (M � 5.28, SE � 0.25) than in those where their best friendships were threatened, Fstranger(1, 105) � 15.67, p � .001, �p

2 � .130, 95% CI [2.05, 0.68], Fclosefriend(1, 100) � 28.62, p � .001, �p

2 � .223, 95% CI [1.16, 2.53], Fromanticpartner(1, 104) � 68.19, p � .001, �p2 � .396, 95% CI [2.01, 3.27]. This is consistent with notions that, even as both romantic and friendly affections might be genuinely finite, exclusivity norms about these relationship types—and specifically relatively greater exclusivity norms for romantic relationships—affect responding (e.g., Davis & Todd, 1982).

Studies 5a and 5b

The majority of jealousy research examines reactions to hypo- thetical scenarios (e.g., Buss, 2013; Daly et al., 1982; DeSteno & Salovey, 1996; Parker et al., 2005, 2010), as did Studies 1– 4. One might wonder, however, whether people are able to accurately assess and report how they would feel in such situations; after all, people are thought to be poor affective forecasters (e.g., Wilson & Gilbert, 2005). To complement the findings above, and to address the concern that findings from Studies 1– 4 might be driven solely by errors in affective forecasting, Studies 5a and 5b asked partic- ipants to recall various times when same-sex best friends became close with new same-sex friends, with romantic partners, and/or formed new same-sex acquaintanceships. To do this, we asked participants to recall real-world experiences. Although less con- trolled than hypothetical scenarios, this provides a complementary and perhaps more ecologically valid means for exploring friend- ship jealousy.

Additionally, Studies 5a and 5b provide an empirical test of our previous assumption that friends are perceived to pose greater replacement threats (but lesser time threats) than romantic part- ners, and vice versa. Here, participants gave retrospective reports of how much each interloper threatened to “replace” them in their friendships, as well as how much time the best friend spent with this new person (when the participant was not also included).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

992 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

Studies 5a and 5b also allow us to empirically test whether re- placement and/or time threat statistically mediates the relationship between interloper type and friendship jealousy.

Study 5b also further explores both friend-guarding behavior and subsequent friend retention. Whereas theory and evidence thus far provide some support for the prediction that friendship jealousy (but not sadness or anger) motivates behavior aimed at countering threats that evoked it, here, we are able to begin to assess the related– but tangential, and non-critical– question of whether friend guarding is actually efficacious in countering those threats.

In all, participants reported the likelihood that each bond—the best friend’s same-sex friend, the best friend’s romantic partner, the best friend’s same-sex acquaintance—posed a replacement threat and a time threat, and reported the amount of friendship jealousy they felt at the time the event occurred. In Study 5b, participants then also reported the extent to which they engaged in several tactics of friend guarding at the time and whether they successfully retained their best friendships. To overview our spe- cific predictions, we expect that (a) people will recall having experienced greater friendship jealousy when rivals were friends than romantic partners (or acquaintances). We predict that (b) same-sex friends were perceived as having posed a greater replace- ment threat than romantic partners (or acquaintances), but that (c) romantic partners were perceived as having posed a greater time threat than same-sex friends (or acquaintances). This perceived replacement threat (but not time threat) should mediate the rela- tionship between interloper type and friendship jealousy. We also expect that (d) people will report more friend guarding when rivals were friends than romantic partners (or acquaintances), and that (e) friendship jealousy mediates the relationship between interloper type and friend guarding. Finally, we explore friend retention rates across rival types, and we test whether friend guarding is perhaps positively associated with friend retention today.

Method

Participants. Study 5a. We determined a sample size of approximately 160

participants was necessary to detect small-to-medium effects (f � .15) in friendship jealousy between rival types (with .99 power, and measurement correlation � .05). We recruited 212 U.S. par- ticipants from TurkPrime who participated in an approximately 12-min study in return for small monetary compensation. Of these, 184 participants (99 women, 3 no sex reported; Mage � 38.12, SDage � 11.93) were determined to be genuine (i.e., not bots) based on answers to open-ended questions and reported experi- encing at least one of the friendship events. All participants re- porting their sexes and filling out focal dependent variables were included in analyses.

Study 5b. Following analyses from Study 5a, we attempted to recruit the same number of usable participants ( 180). Partici- pants were recruited from TurkPrime into an approximately 15- min study in return for small monetary compensation. Of the 278 U.S. participants (156 women, 1 no sex reported) who began the survey, 249 participants (144 women, 1 no sex reported Mage � 37.50, SDage � 11.88) were determined to be genuine (i.e., not bots) based on answers to open-ended questions and also re- sponded as having experienced at least one of the friendship events.

Procedure. Participants were asked questions about their same-sex best friendships across their lifetimes, and were given the same description of a best friend as in previous experiments. They were asked to recall if one or more of three unique events oc- curred: (a) best friends becoming newly close with a same-sex person (i.e., a friend), (b) best friends forming new close romantic relationships, and (c) best friends making new same-sex acquain- tanceships.

All text in prompts was the same between Studies 5a and 5b, with one exception—in Study 5b, we noted that the same-sex friendships could have been newly close; that is, the new friend did not have to be a new person and/or same-sex stranger per se (e.g., that rival could have been an existing friend with whom the best friend simply became newly close). To ensure that best friends’ new friendships and romantic relationships were at least similarly close in quality— echoing the controlled hypothetical scenarios from previous studies—we added text in the two prompts asking participants to recall events when, for example, a best friend formed a new [romantic relationship] during the time that they were friends—“in particular a new [romantic relationship] that may have been just as close as your friendship with your best friend was.”

We included acquaintanceships only to assess whether the pat- tern of responding seen toward hypothetical scenarios from previ- ous studies replicated via this complementary paradigm; we did not anticipate this instance to be associated with much friendship jealousy or friend guarding.

Participants were also instructed that, should any event have occurred more than once, they should think about the event that was the most meaningful for them at the time. Participants reported whether each event happened, maybe happened, or did not happen. For each event that participants recalled as having happened or as maybe having happened, participants answered questions about that event. The order in which questions about each of the three events appeared was randomized.

Retrospective replacement threat. Participants responded to two items assessing the extent to which rivals threatened to replace them (“How much did your best friend’s [new romantic partner] threaten to ‘take your place’ in your best friend’s affections?”; How much did your best friend’s [new romantic partner] threaten to fulfill the same functions for your best friend that you were currently fulfilling?”;� � .89 –.92) on 8-point bipolar scales (1 � not at all, 8 � very much). In Study 5a, participants responded to only the first of these two items.

Retrospective time threat. Participants reported the extent to which rivals took up the best friend’s time (“How much time did your best friend spend with this [new romantic partner] (without you there)?”) on an 8-point bipolar scale (1 � not at all, 8 � very much).

Retrospective friendship jealousy. Participants were asked “At the time that it happened to you—that your best friend formed a new [romantic relationship]—to what extent did you feel . . . ?” Participants reported their reactions, including friendship jealousy, anger, sadness, and also two distractor emotional reactions (hap- piness, pride) on a 7-point Likert-scale (1 � not at all, 7 � very much).

Retrospective friend-guarding behavior (Study 5b). In Study 5b, participants were also asked, “At the time that it happened to you—that your best friend formed a new [romantic relation-

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

993FRIENDSHIP JEALOUSY

ship]—to what extent did you engage in each of the following behaviors?” We assessed five friend-guarding tactics that received high endorsement in Study 3a (Vigilance [� � .94], Separation [� � .92], Induce Jealousy [� � .92], Self/Commitment Enhance- ment [� � .92], Possession Signals [� � .92]) via the same items as used in Study 3a (overall � � .97).

Successful retention (Study 5b)? In Study 5b, participants were also asked “At the time that it happened to you—that your best friend formed a new [romantic relationship]— did you suc- cessfully maintain a relationship with your best friend that was the same or highly similar in quality?”, and responded either “No” or “Yes.”

Results

Study 5a. Of 181 participants, 142 (82 women) reported that, during the focal best friendships, their best friends formed (or maybe formed) new, close friendships with same-sex strangers, 152 (86 women) reported that their best friends formed (or maybe formed) new, close romantic relationships, and 161 (92 women) reported that their best friends formed (or maybe formed) new same-sex acquaintanceships.

Study 5b. Of 249 participants, 197 (114 women) reported that their best friends formed (or maybe formed) close friendships with same-sex friends, 208 (125 women) reported that their best friends formed (or maybe formed) close romantic relationships, and 211 (124 women) reported that their best friends formed (or maybe formed) new same-sex acquaintanceships.

Retrospective replacement threat. Study 5a. An ANOVA exploring differences in replacement

threat recalled by participants across rival types yielded a signif- icant effect, F(2, 232) � 31.44, p � .001, �p2 � .213. Participants recalled greater replacement threat when rivals were same-sex strangers (M � 4.68, SE � .21) than romantic partners (M � 4.23, SE � .21; p � .023, 95% CI [0.06, 0.84], �p2 � .038) or same-sex acquaintances (M � 3.08, SE � .20; p � .001, 95% CI [1.18, 2.03], �p2 � .328). Participants also recalled greater replacement threat when rivals were romantic partners than same-sex acquain- tances (p � .001, 95% CI [0.73, 1.58], �p2 � .195).

Study 5b. An ANOVA exploring possible differences in re- placement threat yielded a significant effect of rival type, F(2, 338) � 50.33, p � .001, �p2 � .229. Replicating findings from Study 5a, when rivals were same-sex friends, participants recalled greater replacement threat (M � 4.90, SE � .16) than when rivals were romantic partners (M � 4.45, SE � .18; p � .010, 95% CI [.10, .76], �p2 � .030) or same-sex acquaintances (M � 3.30, SE � .16; p � .001, 95% CI [1.26, 1.89], �p2 � .329). Participants also recalled greater replacement threat when rivals were romantic partners than same-sex acquaintances (p � .001, 95% CI [.82, 1.47], �p2 � .230). Findings suggest that interlopers are perceived as posing greater replacement threats when they are new close friends as compared to romantic partners (or acquaintances).

Retrospective time threat. Study 5a. An ANOVA exploring possible differences in time

threat recalled by participants across rival types yielded a signif- icant effect, F(2, 228) � 37.39, p � .001, �p2 � .247. As expected, participants recalled greater time threat when rivals were romantic partners (M � 5.36, SE � .21) than same-sex strangers (M � 4.01, SE � .15; p � .001, 95% CI [0.84, 1.86], �p2 � .177) or same-sex

acquaintances (M � 3.16, SE � .19; p � .001, 95% CI [1.65, 2.76], �p2 � .397). Participants also recalled greater time threat when rivals were same-sex strangers than same-sex acquaintances (p � .001, 95% CI [0.40, 1.31], �p2 � .111).

Study 5b. An ANOVA exploring possible differences in time threat also yielded a significant effect of rival type, F(2, 338) � 103.14, p � .001, �p2 � .379. Replicating findings from Study 5a, participants recalled greater time threat when rivals were romantic partners (M � 6.28, SE � .13) than same-sex friends (M � 5.35, SE � .12; p � .001, 95% CI [.66, 1. 21], �p2 � .210) or acquain- tances (M � 4.07, SE � .14; p � .001, 95% CI [1.87, 2.57], �p2 � .414). Participants also recalled greater replacement threat when rivals were same-sex friends than same-sex acquaintances (p � .001, 95% CI [.99, 1.58], �p2 � .231). Findings support the previ- ous assumption that interlopers are perceived as posing greater time threats when they are new romantic partners as compared with close friends (or acquaintances).

Retrospective friendship jealousy. Study 5a. Echoing the pattern above for replacement threat

(but not time threat), an ANOVA exploring participants’ recalled friendship jealousy yielded a significant effect of rival type, F(2, 230) � 50.03, p � .001, �p2 � .303; participants recalled experi- encing greater friendship jealousy when rivals were same-sex strangers (M � 3.72, SE � .18) than romantic partners (M � 3.09, SE � .18; p � .001, 95% CI [0.30, .96], �p2 � .111) or same-sex acquaintances (M � 2.03, SE � .15; p � .001, 95% CI [1.35, 2.11], �p2 � .474). Participants also recalled greater friendship jealousy when rivals were romantic partners than same-sex ac- quaintances (p � .001, 95% CI [0.70, 1.40], �p2 � .179; see Figure 5c, above).

Study 5b. Replicating findings from Study 5a, and also echo- ing the pattern above for replacement threat (but not time threat), an ANOVA exploring possible differences in recalled friendship jealousy yielded a significant effect of rival type, F(2, 332) � 45.11, p � .001, �p2 � .214. When rivals were same-sex friends, participants recalled greater friendship jealousy (M � 3.73, SE � .15) than when rivals were romantic partners (M � 3.30, SE � .15; p � .001, 95% CI [.14, .73], �p2 � .048) or same-sex acquaintances (M � 2.40, SE � .13; p � .001, 95% CI [1.04, 1.63], �p2 � .294). Participants also recalled greater friendship jealousy when rivals were romantic partners than same-sex acquaintances (p � .001, 95% CI [.65, 1.16], �p2 � .177).

Does replacement threat (or time threat) drive friendship jealousy? We test this using MEMORE, an SPSS macro for conducting mediation analyses on within-subjects variables (Mon- toya & Hayes, 2017). We conducted parallel multiple mediation of the influence of replacement and time threats on friendship jeal- ousy, using 5,000 bootstrapped iterations to compute a bias cor- rected 95% CI for the indirect effects.

Study 5a. Consistent with predictions, the indirect effect of interloper type on friendship jealousy was significantly mediated by replacement threat, b � 0.18, SE � 0.09, 95% CI [0.03, 0.39], but not by time threat, b � �0.02, SE � 0.09, 95% CI [�0.22, 0.14]. The direct effect was significant, b � 0.48, SE � 0.16, p � .004, 95% CI [0.15, 0.81].

Study 5b. As shown in Figure 7—and replicating Study 5a— the indirect effect of interloper type on friendship jealousy was significantly mediated by replacement threat, b � 0.18, SE � 0.08, 95% CI [0.05, 0.36], but not by time threat, b � �0.06, SE � 0.07,

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

994 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

95% CI [�0.20, 0.07]. The direct effect was significant, b � 0.31, SE � 0.15, p � .034, 95% CI [0.02, 0.60].

Retrospective friend-guarding behavior. In Study 5b, we ran a 3 [Rival type] � 5 [Friend-guarding tactic] mixed-factors ANOVA to explore recalled friend guarding. This yielded signif- icant effects of rival type, F(2, 290) � 163.55, p � .001, �p2 � .530, and tactic, F(4, 580) � 35.81, p � .001, �p2 � .198, and a significant interaction, F(8, 1160) � 32.20, p � .001, �p2 � .182.

First, aggregating over tactics, we find that people reported greater friend-guarding behavior toward rivals who were same- sex friends (M � 3.01, SE � .11) than romantic partners (M � 2.53, SE � .11; p � .001, 95% CI [0.38, 0.59], �p2 � .359) or same-sex acquaintances (M � 2.09, SE � .11; p � .001, 95% CI [0.82, 1.03], �p2 � .684). Participants also reported greater friend-guarding behavior when rivals were romantic partners than same-sex acquaintances (p � .001, 95% CI [0.35, 0.54], �p

2 � .356). Second, people recalled using four out the five friend-guarding tactics more when rivals were same-sex friends than when rivals were romantic partners or same-sex acquain- tances (ps � .005) and also when rivals were romantic partners than when rivals were acquaintances (ps � .005). The exception was Inducing Jealousy, which people recalled having deployed more when rivals were romantic partners than same-sex friends (ps � .001). See Table 2 for all means (SEs).

How often did people retain best friendships across inter- loper type? Of participants responding to each prompt, 90.6% (valid percent) reported successfully retaining best friends against same-sex acquaintances, 41.6% (valid percent) reported success- fully retaining best friends against same-sex rivals, and 47.4% (valid percent) reported successfully retaining best friends against romantic partners. To our knowledge, this provides some of the first evidence on actual friend retention in the face of third-party threats to friendships.

Using a McNemar test, we find that people reported retaining best friendships more often when rivals were acquaintances than friends (p � .001 for 156 cases) or romantic partners (p � .001 for 161 cases), as one might expect if acquaintances pose less replace- ment threat than other potential interlopers. There were no differ- ences in reported retention when rivals were same-sex friends versus romantic partners (p � .282 for 156 cases).

Does friendship jealousy drive friend guarding? A primary prediction was that best friends’ friendships evoke greater friendship jealousy than their romantic relationships, and that these differences in friendship jealousy, in turn, motivate dif- ferences in friend-guarding behavior. We used 5,000 boot- strapped iterations to compute a bias corrected 95% CI for the indirect effects to test this. The analysis yielded a significant indirect effect of friendship jealousy, b � 0.04, SE � 0.02, 95% CI [0.01, 0.10], supporting the proposed model. The direct effect was significant, b � 0.43, SE � 0.05, p � .001, 95% CI [0.33, 0.53].

Might friendship jealousy-motivated friend guarding lead to greater friend retention? We first explored whether friendship jealousy might be associated with greater downstream friend re- tention. We then explored whether there was a positive association between friend guarding (aggregated) and friend retention. Finally, we additionally assessed the exploratory question of whether friendship jealousy-motivated friend guarding might predict friend retention.

Friend guarding significantly predicted retention when rivals were friends, Wald(1, N � 173) � 4.70, p � .030, Exp(B) � 1.28, 95% CI [1.02, 1.59]. Although the trend was in the same direction, this relationship was not significant when rivals were romantic partners, Wald(1, N � 194) � 2.05, p � .153, Exp(B) � 1.18, 95% CI [0.94, 1.47]. For consistency, we also explored acquaintances. There was not a significant relationship between guarding and retention, Wald(1, N � 162) � 0.94, p � .759, Exp(B) � 0.95, 95% CI [0.67, 1.34].6

Next, we used 5,000 bootstrapped iterations to compute a bias corrected 90% CI for the indirect effects to test whether friendship jealousy over best friends becoming close with same-sex friends

6 We speculate as to why we find this null result here. Friendship jealousy and related friend guarding are quite low when interlopers are acquaintances, consistent with the notion that acquaintances are not typi- cally considered replacement threats. Thus, whereas some acquaintances might progress to friends and perhaps replacement threats, many might not; many acquaintanceships might dissolve or fail to progress. Thus, people might simply not friend guard against their best friends’ acquaintances and yet nevertheless achieve high best friend retention.

Interloper type (friend,

romantic partner)

Replacement threat

(Mfriend – Mrom) b=0.18, SE=0.08

95%CI=[0.05, 0.36]

Friendship jealousy (Mfriend –

Mrom)

b = 0.44, p = .006

b = -0.93, p < .001

b = 0.40, p < .001

b = 0.07, p = .353Time threat (Mfriend – Mrom) b=-0.06, SE=0.07

95%CI=[-0.20, 0.07]

b = 0.31, SE=0.14, p = .034, 95%CI = [0.02, 0.60]

Figure 7. Parallel multiple mediation models depicting the effect of interloper type on friendship jealousy as mediated by replacement and time threats from Study 5b.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

995FRIENDSHIP JEALOUSY

led to more friend guarding and, in turn, more friend retention (a binary outcome variable). The analysis yielded a significant indi- rect effect of friend guarding, b � 0.06, SE � 0.04, 90% CI [0.01, 0.13], supporting that exploratory proposition. The direct effect was not significant, b � 0.04, SE � 0.09, p � .646.

Discussion

First, people reported greater friendship jealousy when rivals were same-sex friends than romantic partners (or same-sex ac- quaintances), replicating findings from Studies 3a, 3b, and 4 (using hypothetical scenarios)— here with recalled real-world events. This suggests that errors in affective forecasting, alone, are un- likely to have caused the results found in previous studies.

Second, people reported greater replacement threat when rivals were same-sex friends versus romantic partners (or acquaintances), and reported greater time threat when rivals were romantic part- ners than friends (or acquaintances). In combination with analyses showing that replacement threat (but not time threat) mediates the relationship between rival type and friendship jealousy, this pro- vides further evidence that friendship jealousy is more strongly calibrated to replacement than time threat. Moreover, it bolsters confidence in the assumptions that different interlopers are per- ceived to pose different levels of replacement and time threats. In Study 5b, as predicted, people also reported greater friend- guarding behavior when rivals were friends than romantic partners (or acquaintances). Related mediation analyses suggest that friend- ship jealousy significantly statistically mediated the relationship between rival type and friend-guarding behavior.

To our knowledge, Study 5b also provides some of the first evidence of friend retention rates in the face of possible third-party threats, thus, speaking to the modern-day efficacy of jealousy- motivated friend guarding. Reports suggest that people retained their best friends 90% of the time when rivals were acquain- tances, but only 45% of the time when rivals were friends or romantic partners. Note, however, that we asked participants to recall salient friendship-threatening events; these low retention numbers may reflect that people are more likely to recall events that had effects on their friendships (i.e., friendship dissolution). Additionally, we emphasize that there is quite the gulf between friendship-jealousy spurred friend guarding and downstream friend retention. Moreover, evolutionary accounts of emotions would not depend on modern friendship jealousy leading to more friend retention (rather, such a perspective would make testable predictions about the inputs and outputs of such an emotion, as we do here). We return to this important point in the General Discus- sion.

Study 6

Preregistered Study 6 (osf.io/53vbr) builds on Studies 5a and 5b and manipulates online friendship jealousy (vs. a relatively neutral friend-related control condition) to further explore whether friend- ship jealousy causes friend-guarding intentions.7 To do this, Study 6 adapts a paradigm from existing emotions work (e.g., Ashton- James & Tracy, 2012; Ekman, Levenson, & Friesen, 1983) that evokes emotional experience in the moment. After reading a prompt instructing participants to experience either friendship jeal- ousy at a friend becoming potentially closer with another same-sex person (friendship jealousy) or a typical day spent with their best friend (relatively neutral friend-related control), participants wrote in detail about their current affect, physiology, and musings. Next, participants reported their currently felt affect (a manipulation check) and the extent to which they would engage in friendship maintenance behavior— both friend guarding (our focal dependent variable) and also everyday friend retention—akin to Study 2.

We predict that participants in the friendship jealousy condition will (a) experience greater friendship jealousy and (b) report greater friend-guarding intentions compared with those partici- pants in the relatively neutral friend-related control condition. In the latter condition, we expect participants might feel neutrality (no strong emotionality) and/or happiness. Finally, we additionally explore whether currently felt friendship jealousy (vs. other emo- tions) might statistically mediate the relationship between condi- tion and reported intentions to engage in friend guarding.

Method

A power analysis indicated that 351 participants would be required for .80 power to detect small- to medium-sized effects (f � .15) in friend guarding between the two conditions. We opened a study to 400 U.S. participants on TurkPrime and included bot and attention checks. Of 413 participants who filled out the two critical open-ended items at the start of the survey (same-sex others’ names and responses to manipulation), which also served as bot checks, 340 participants (191 women, 1 other/missing; Mage � 39.96, SDage � 13.09) passed both checks, and also

7 To ensure that our manipulations were successful in an online com- munity sample, we ran a preregistered pilot (osf.io/86bza). Data from the pilot support preregistered predictions and are further replicated in Study 6. The primary difference between the pilot and the present study is our use of friend-guarding scale items; whereas the pilot assessed only friend guarding-vigilance (as in Study 2), here, we assess multiple friend- guarding tactics (akin to Study 5b). Results are reported in detail in the online supplemental materials.

Table 2 Means (SEs) for Friend-Guarding Behavioral Tactics Used Toward Different Rivals

Tactic Same-sex friend Romantic partner Same-sex acquaintance

Vigilance (� � .89) 3.35 (.14) 2.37 (.13) 2.08 (.13) Separation (� � .94) 2.57 (.13) 2.23 (.12) 1.93 (.11) Inducing jealousy (� � 91) 2.25 (.12) 2.17 (.12) 1.98 (.13) Self/commitment enhancement (� � .92) 3.56 (.12) 2.88 (.12) 2.24 (.12) Possession displays (� � .92) 3.34 (.13) 2.64 (.12) 2.20 (.12)

Note. All comparisons between rival types were significant (p � .005).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

996 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

correctly answered a late-in-study attention check question (an- swering a multiple choice question correctly by writing “ethnicity” in the related text box), and were included in analyses.

Procedure and design. As in previous studies, participants first reported some brief demographic information (sex, age) and were instructed to fill out the names of same-sex others— here, a best friend and an acquaintance. The best friend’s name would be piped into the manipulation prompts, whereas the acquaintance name was asked only as part of our bot check, similar to previous studies. Participants then underwent the manipulation, which in- volved putting oneself in a friendship jealousy-evoking scenario or in a typical day spent with the best friend, and then writing a detailed open-ended response about what they are currently expe- riencing. What participants wrote here served as the second critical bot check.

We adapted the established Relived Emotion Task (Ekman et al., 1983) to manipulate participants’ online affect—friendship jealousy versus a relatively neutral friend-related control. Here, we asked participants to imagine rather than to relive an emotion because reliving a past, salient friendship jealousy event may color participants’ responses on the focal measure (intent to friend guard the current best friend).8 For example, past instances of friendship jealousy may be especially salient because they led to the disso- lution of the best friendship; subsequently asking participants about their intentions to guard a current best could be confusing. We avoided such potential issues here by modifying the traditional relived emotion task, eliciting participants’ emotions in response to a hypothetical friendship jealousy (or control) event related to their current best friends (rather than a potential former friend), render- ing the present paradigm a combination between a hypothetical scenario approach and a traditional relived emotion approach.

In the friendship jealousy condition, participants were instruct- ed: “Imagine that you are feeling very jealous because [name of best friend] has become quite close with another same-sex friend, maybe even closer to that new friend than [name of best friend] is to you. Everyone has, at some time in their lives, felt jealousy when their friends seemed to like another friend better. It’s OK to feel this. Give yourself a moment to close your eyes and really experience this jealousy. Describe in as much detail as possible the jealousy that you are feeling. For example, what’s going on in your mind, and what are you thinking while you are feeling this jeal- ousy? What’s going on in your body—your gut, your arms and legs, your heartbeat—while you are feeling this jealousy?” We included information about friendship jealousy being normative to encourage participants to honestly experience and report their feelings, in line with previous work (Ashton-James & Tracy, 2012; Zammuner, 1996).

Our relatively neutral friend-related control condition was adapted from Ashton-James and Tracy (2012); therein, participants in the neutral condition detailed their activities that day, whereas here participants were instructed: “Imagine that you are spending time with [name of best friend]. There’s no special occasion, rather, it’s just a typical day when you’re spending time together— akin to those you might have had not too long ago. Take some time to really visualize each experience that you are doing on this regular day. Give yourself a moment to close your eyes and really experience each of your activities during this regular, typical day. Describe in as much detail as possible everything that you and [name of best friend] would be doing. For example, what’s going

on in your mind and what are you thinking while you are going about your everyday activities? What’s going on in your body— your gut, your arms and legs, your heartbeat—while you are going about your everyday activities with [name of best friend]?”

Manipulation check. Next, participants were asked to report the extent to which they were currently feeling five affective states—friendship jealousy (“jealousy”), neutral (“no strong emo- tions [neutral]”), “happiness,” “sadness,” and “anger”— on 7-point Likert-type scales (1 � not at all, 7 � very much).

Friendship maintenance. We assessed participants’ behav- ioral intentions to engage in two types of friendship maintenance behavior: friend guarding and everyday friend retention. Partici- pants read, “Take a moment to re-experience what you were feeling in your mind and body. Given your feelings right now, how likely would you be to . . .” and responded to 17 items on a 7-point Likert-scale (1 � not at all, 7 � very much).

We assessed friend guarding via 12 items measuring five tactics that we explored in previous studies: We used two items for Vigilance (e.g., “Check up on [best friend] to see if they’re spending time with other people”; � � .91); three items for Inducing Jealousy (e.g., “Develop and show off other new friend- ships you made”; � � .82); and five items for Possession Signals (e.g., “Show off how close you and [best friend] are on social media [e.g., posting pictures of you two together])”; (� � .84). Although people scored highly on the Self/Commitment Enhance- ment friend-guarding tactic in Study 3a, leading us to use it in Study 5b, some of these items had potential theoretical overlap with everyday friend retention items (e.g., related items assessed being positive in the former and avoiding potentially negative subjects in the latter), so we did not use this tactic. Similarly, although people scored highly on the Separation friend-guarding tactic in Study 3a, leading us to use it in Study 5b, we did not assess Separation because the items required the presence of a likely interloper (e.g., “Try to get your best friend to avoid places where this other person will be”), which would have been confus- ing and nonsensical for participants in the control condition (where no interlopers were mentioned or expected). Instead, we added Monopolization, the items for which did not require the presence of an interloper (e.g., “Monopolize [best friend]’s time”; � � .85). We aggregated scores on these five friend-guarding tactics into one overall measure of friend guarding (� � .94).

To assess everyday friend retention, we used five items from Study 2 (e.g., “Be cheerful and positive whenever you’re with [best friend]”; � � .57).9 Unlike Study 2, we did not include the item assessing the social networks tactic (“Rely on other friends to help you through this rough patch”), because it would have been confusing and nonsensical for participants in the control condition (where no rough patch is mentioned or expected).

8 We thank Jessica Tracy for helpful feedback on developing this ma- nipulation.

9 Because of this reliability score, we created a second composite of friend retention that excluded the item “Avoid talking about things that we disagree about.” This second composite had an improved score (� � .65). Predicted results do not differ when using either composite (see online supplemental materials).

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

997FRIENDSHIP JEALOUSY

Results and Discussion

Manipulation check. We ran a 2 (Condition) � 5 [Affective reactions] mixed-factors ANOVA, which yielded main effects of condition, F(1, 339) � 57.52, p � .001, �p2 � .145, affective reactions, F(4, 1356) � 67.74, p � .001, �p2 � .167, and a significant interaction, F(4, 1356) � 92.88, p � .001, �p2 � .215.

In line with predictions, people reported currently feeling greater friendship jealousy in the friendship jealousy condition (M � 4.13, SE � .13) than in the relatively neutral friend-related control condi- tion (M � 1.45, SE � .14), F(1, 339) � 196.50, p � .001, �p2 � .367, 95% CI [2.30, 3.06]. As one might expect, given the links between jealousy and basic affective states of both sadness and anger, people also reported currently feeling more sadness in the friendship jealousy condition (M � 3.55, SE � .13) versus the control condition (M � 2.05, SE � .14), F(1, 339) � 64.87, p � .001, �p2 � .161, 95% CI [1.13, 1.87], and currently feeling more anger in the friendship jeal- ousy condition (M � 2.65, SE � .11) versus the control condition (M � 1.45, SE � .11), F(1, 339) � 60.03, p � .001, �p2 � .150, 95% CI [0.90, 1.51]. People in the friendship jealousy condition also reported currently feeling significantly greater friendship jealousy than any other reaction (ps � .001).

People reported currently feeling greater neutrality in the control condition (M � 4.01, SE � .14) versus the friendship jealousy condition (M � 3.19, SE � .13), F(1, 339) � 17.21, p � .001, �p

2 � .048, 95% CI [0.43, 1.20], and people also reported currently feeling greater happiness in the control condition (M � 4.84, SE � .13) versus the friendship jealousy condition (M � 3.15, SE � .12), F(1, 339) � 89.00, p � .001, �p2 � .208, 95% CI [1.34, 2.05]. People reported currently feeling significantly greater happiness than any other reaction (ps � .001) in the relatively neutral friend-related control condition; excepting happiness, people re- ported currently feeling significantly greater neutrality than any other reaction (ps � .001) in the control condition. This seems sensible, given the prompt was to immerse oneself in a typical day with one’s best friend.

Does friendship jealousy cause people to engage in greater friend guarding? We first conducted a 2 (Condition) � 2 [Friend- ship maintenance type] mixed-factors ANOVA. (We also report the similar results of a 2 (Condition) � 5 [Tactic] ANOVA in the online supplemental materials.) This yielded main effects of condition, F(1, 339) � 29.38, p � .001, �p2 � .220, type, F(1, 339) � 407.98, p � .001, �p2 � .546, and a significant interaction, F(1, 339) � 95.66, p � .001, �p2 � .220.

Exploring the significant interaction, we find, as predicted, that people in the friendship jealousy condition reported greater behavioral intent to engage in friend guarding (M � 3.68, SE � .10) than those in the control condition (M � 2.30, SE � .11), F(1, 339) � 84.45, p � .001, �p2 � .199, 95% CI [1.08, 1.67]; see Figure 8. This suggests that online experiences of friendship jealousy cause increased intent to engage in friend guarding.

We also find that people reported greater behavioral intent to engage in everyday friend retention in the relatively neutral friend- related control condition (M � 4.85, SE � .08) than in the friendship jealousy condition (M � 4.57, SE � .08), F(1, 339) � 6.24, p � .012, �p

2 � .019, 95% CI [0.06, 0.50], though this effect was comparatively smaller. This echoes similar findings from Study 2. People in both conditions reported greater intent to engage in everyday friend reten-

tion than friend guarding (ps � .001, �p2control � .554, �p 2 jealousy �

.147). Reanalysis in two additional, preregistered subsamples:

Does friendship jealousy cause people to engage in greater friend guarding? As per our preregistration, we planned to conduct two ancillary analyses of this same key prediction sup- ported above—that people in the friendship jealousy condition report greater intent to friend guard than those in the control condition. We did so because, even in light of the manipulation check, one might be concerned, for example, that not all partici- pants were truly and/or strongly feeling friendship jealousy in the friendship jealousy condition; perhaps, then, noise is causing us to find support for our prediction. Perhaps among only those report- ing strong currently felt friendship jealousy, our key prediction would no longer hold. Perhaps self-report of currently felt emo- tions was problematic because people were uneasy reporting their jealousy or anger (even as we attempted to mitigate this). Thus, to be circumspect, we conducted ancillary tests to explore the robust- ness of our results among those participants most likely to have been affected by the manipulations as intended—in line with a modern multiverse approach (e.g., Steegen, Tuerlinckx, Gelman, & Vanpaemel, 2016).

We conducted the exact same ANOVA as reported above in each of two subsamples of data. We did so in a first subsample (n � 196) that included only those participants in the friendship jealousy condition whose manipulation check scores indicated strong levels of self-reported currently felt friendship jealousy (scores of 4 or higher on the 7-point scale) and those partici- pants in the control condition whose manipulation check scores indicated low levels of currently felt negative affect—friend- ship jealousy, sadness, anger (scores below 4 on those negative emotions). We also did this in a second subsample (n � 266) that included only those participants whose reflected affect scores indicated strong levels of currently felt friendship jeal- ousy in the friendship jealousy condition (scores of 4 or higher on a 7-point scale) and low levels of currently felt negative affect in the control condition (scores below 4 on negative emotions). Here, scores were coded by a research assistant (RA) who was blind to hypotheses and manipulations; the RA read participants’ written responses to the manipulations and then

1

2

3

4

5

6

7

Friend guarding Everyday friend reten�on

Friendship jealousy

Rela�vely neutral friend- related control

tnetnIlaroivaheB detropeR

Type of Friendship Maintenance Behavior

Condi�on

Figure 8. Reported behavioral intent to engage in friend guarding and everyday friend retention as a function of condition. Error bars represent SEs.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

998 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

coded them for the extent to which they reflected the same currently felt emotions that we measured via participant self- report (also on a 7-point scale). (For each emotion, correlations between participant self-report and RA-coded emotions was greater than .300 [ps � .001].) In both subsamples, we again find support for our prediction—people in the friendship jeal- ousy condition reported greater intent to friend guard than people in the control condition. Our results seem robust, and findings imply that noise is unlikely to be driving findings. See online supplemental materials for more information about these subsamples, coding, and detailed results.

Does currently felt friendship jealousy (or other affective reactions) mediate the relationship between condition and friend guarding? Returning to the full sample of participants passing bot and attention checks, we explored this mediational pathway using PROCESS Model 4, allowing multiple mediators (friendship jealousy, neutrality, happiness, sadness, and anger), using 5,000 bootstrapped samples and 95% bias corrected CIs. As illustrated in Figure 9, and consistent with expectations derived from our model, the indirect effect was significant for friendship jealousy, b � 1.39, SE � 0.14, 95% CI [1.14, 1.68], and not for any other affective reaction. The direct effect was no longer significant, b � 0.07, SE � 0.15, p � .656. This further suggests that online experiences of friendship jealousy drive friend guard- ing.

Study 7

One of the most intriguing findings from several of the preced- ing studies (Studies 3a–5b) is that people reported greater friend- ship jealousy when friends formed new same-sex friendships (likely posing higher replacement threat, but lesser time threat) than new romantic relationships (likely posing higher time threat, but lesser replacement threat). In combination with mediation

analyses from Studies 5a and 5b, those findings support the func- tional prediction that friendship jealousy is more strongly cali- brated to cues of replacement threat, even over other cues that our intuitions might expect—specifically, the amount of time best friends spend with other people. Indeed, we expect that cues with better predictive validity of impending loss (i.e., replacement threat) should receive prioritization in driving levels of friendship jealousy. To further explore this, Study 7 orthogonally manipulates replacement and time threats, manipulating the proposed drivers of friendship jealousy.

Recall that an Alliance Hypothesis perspective on friendship contends that a person’s friends are ranked hierarchically in de- scending order of those one would support in a conflict between those friends (DeScioli & Kurzban, 2009; DeScioli et al., 2011). Relative to lower-ranked friends, higher-ranked friends presum- ably receive support in conflict and preferential access to other friend-mediated benefits (e.g., resources, information; DeScioli et al., 2011; see also de Waal & Waal, 2007). An especially strong illustration of replacement threat might be a best friend siding with and/or conferring a contested resource on someone else over oneself.

To vary replacement threat, then, participants imagined that the best friend was attending a desirable social event at which space is limited; the best friend could take only one friend as their guest to this event, and both the participant and the new friend voiced desires to attend. In the no-information control condition, the participant is informed only that the best friend can take one of them to the event. In the low replacement threat condition, the best friend decides to take the participant (not the new friend). In the high replacement threat condition, the best friend decides to take the new friend (not the participant). To vary time threat, partici- pants were asked to imagine that a best friend has formed a new, close same-sex friendship, and that the best friend is spending

Condition: Control,

Friendship jealousy

Jealousy b=0.77, SE=0.19

95%CI=[0.80, 1.50]

Sadness b=0.13, SE=0.13

95%CI=[-0.10, 0.41]

Anger b=0.00, SE=0.11

95%CI=[-0.11, 0.33]

Friend guarding- vigilance

b = 2.21, p < .001

b = 1.38, p < .001

b = 0.35, p < .001

b = 0.07, p = .200

Neutral b=0.01, SE=0.03

95%CI=[-0.05, 0.09]

Happiness b=0.13, SE=0.13 95%CI=[-0.10,

0.40]

b = -0.64, p < .001

b = -1.96, p < .001

b = 1.60, p < .001

b = -0.02, p = .729

b = -0.07, p = .237

b = 0.08, p = .280

Figure 9. Parallel multiple mediation model depicting the effect of condition on friend guarding, as mediated by self-reported currently felt affect from Study 6.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

999FRIENDSHIP JEALOUSY

either a lot (high time threat) or a little (low time threat) time with this person.

Whereas we expect that friendship jealousy is more strongly calibrated to replacement than time threat, we still expect time threat to affect friendship jealousy in some situations. That is, insofar as the amount of time a best friend spends with a new friend cues the extent to which that new friend might pose a current or eventual replacement threat, we would expect time threat to influence friendship jealousy. Thus, we predict that (a) when there is no information about replacement threat (i.e., in the no-information control condition for replacement threat), time threat will enhance friendship jealousy: The more time that the best friend spends with the new friend, the more friendship jeal- ousy should be evoked. However, (b) time threat will have less effect on friendship jealousy than will replacement threat: When available, information about replacement threat should largely override the effects of time threat, such that high replacement threat evokes high levels of friendship jealousy, and low replace- ment threat evokes low levels of friendship jealousy.

Method

Participants. A power analysis suggested we would need 432 participants for .80 power to detect small-to-medium effects (f � .15) in reported friendship jealousy given our six-cell between- subjects design. U.S. participants were recruited from Amazon Mechanical Turk (MTurk) into an approximately 12-min study in return for small monetary compensation. Given expected attrition, we attempted to recruit 450 participants. There were 503 partici- pants who started the survey and reported their sex; of these, 428 participants (218 women; Mage � 38.14; SDage � 11.94) filled out focal dependent variables and were included in analyses.

Procedure. As in previous studies, participants first answered basic demographic questions about themselves (sex, age) and questions about same-sex friends before imagining and responding to one of six possible scenarios, given the 2 (Time threat: high, low) � 3 (Replacement threat: low, no-information control, high) between-subjects design.

Time threat was operationalized as the amount of time one’s best friend was spending with a new same-sex friend. Replacement threat was operationalized via a party scenario. All participants read that their same-sex best friends had been invited to the exclusive birthday party of a same-sex acquaintance they want to know better. Participants are ultimately informed that their best friends can take a guest— but both they (the participants) and their best friends’ new friend voiced desires to be that plus one. In the no-information control condition, the scenario stops there. In the low replacement threat condition, participants are informed that their best friends have decided to take them to the event (instead of the new friend); in the high replacement threat condition, partici- pants are informed that their best friends have decided to take the new friend to the event (instead of the participant; see Appendix B for scenarios).

Reported friendship jealousy. After reading the scenario, participants reported their reactions, including friendship jealousy, sadness and anger, and distractor reactions (disgust, guilt, pride, happiness, relief, enthusiasm, nothing) on a 7-point Likert-scale (1 � not at all, 7 � very much).

Results and Discussion

Running the 2 (Time threat) � 3 (Replacement threat) ANOVA yielded a marginally significant main effect of time threat, F(1, 422) � 3.29, p � .071, �p2 � .008, such that scenarios presenting high time threat evoked marginally greater reported friendship jealousy (M � 2.91, SE � .11) than those presenting low time threat (M � 2.62, SE � .11). We also found a significant main effect of replacement threat, F(2, 422) � 112.93, p � .001, �p2 � .349, such that high replacement threat scenarios evoked signifi- cantly greater reported friendship jealousy (M � 4.38, SE � .14) than control scenarios (M � 2.40, SE � .14); p � .001, 95% CI [1.60, 2.36], �p2 � .229, and control scenarios evoked significantly greater reported friendship jealousy than low replacement threat scenarios (M � 1.52, SE � .14; p � .001, 95% CI [.55, 1.22], �p2 � .088; see Figure 10). We also found the predicted interaction, F(2, 422) � 5.13, p � .006, �p2 � .024.

In line with predictions, only in the no-information control condition for replacement threat did time threat drive results; more time spent with the best friend evoked greater reported friendship jealousy (M � 2.90, SE � .19) than did less time spent with the best friend (M � 1.90, SE � .20), F(1, 422) � 13.38, p � .001, 95% CI [.46, 1.54], �p2 � .031.

Time threat did not affect reported friendship jealousy in the low replacement threat condition (when the best friend chose the participant over the new friend), which evoked relatively low levels of friendship jealousy overall (Mhigh-time � 1.45, SEhigh-time � .19; Mlow-time � 1.58, SElow-time � .20; p � .630). Time threat also did not affect reported friendship jealousy in the high replacement threat condition (when the best friend chose the new friend over the participant), which evoked relatively high levels of friendship jealousy overall (Mhigh-time � 4.39, SEhigh-time � .20; Mlow-time � 4.38, SElow-time � .20; p � .984).

At both high and low time threat, high replacement threat evoked greater reported friendship jealousy than did no- information control or low replacement threat (ps � .001). At low time threat, friendship jealousy did not differ as a function of low versus control replacement threat (p � .248); at high time threat, low replacement threat evoked less friendship jealousy than did control replacement threat (p � .001, 95% CI [�1.99, �0.92], �p

2 � .116).

1

2

3

4

5

Low replacement threat

Control (no- informa�on)

High replacement threat

Low �me threat

High �me threat

ysuolaeJ pihsdneirF

detropeR

Replacement threat

Time threat

Figure 10. Reported levels of friendship jealousy as a function of time threat and replacement threat from Study 7. Error bars represent SEs.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1000 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

That time threat did not affect reported friendship jealousy when replacement threat information was available suggests that people use time threat as a cue to replacement threat. That is, when more direct information about replacement threat is available, such information more strongly drives friendship jealousy. Thus, the overall pattern of results suggests that friendship jealousy is cali- brated more closely to replacement threat than to time threat.

Studies 8a and 8b

Study 8a and preregistered Study 8b (osf.io/esv4n) explore the proposed model of friendship jealousy. Specifically, we test the prediction that interlopers posing a presumably higher replacement threat (the best friend becoming close with a new friend vs. a new romantic partner; 8a) or interlopers specifically described as pos- ing a higher replacement threat (a friend whom your best friend chose over you vs. a friend over whom your best friend chose you; 8b) evoke greater friendship jealousy and, in turn, evoke more intense friend-guarding behavior.

To measure friend-guarding behavior, we adapt a behavioral- choice paradigm from established work on stereotyping and prej- udice, wherein chosen seating distance away from a target is an index of antipathy toward that target (e.g., Amodio & Devine, 2006). Here, we task participants with making a similar choice— but for seating their best friends and the interlopers. Seating distance (how many seats apart) is an index of the separation that participants create between their best friends and potential inter- lopers.

Method

Participants. U.S. participants were recruited from TurkPrime into an approximately 8-min study on social emotions in return for small monetary compensation. For each study, we aimed to attain a sample size of 225. We used two strong attention/ bot checks: assessing whether participants (a) followed critical directions in naming three same-sex friends and (b) correctly responded to an attention-check prompt. Taking a multiverse ap- proach (Steegen et al., 2016), we first conducted primary analyses on the data having used both exclusions; because results do not appreciably differ depending on whether we analyze those passing the one or both checks, we present data from those passing only the critical first check because it allows for a larger sample size. (We replicate findings reported below using data from passing both checks in the online supplemental materials).

Study 8a. Of 270 participants responding to “What is your sex?”, 202 (86 women) passed check 1 (Mage � 36.54, SDage � 10.33). Sensitivity analysis suggests this yields sufficient (.80) power to detect small- to medium-sized effects (f � .20).

Study 8b. Of 243 participants responding to “What is your sex?”, 193 (84 female) passed check 1 (Mage � 36.34, SDage � 11.24). Sensitivity analysis suggests this yields sufficient (.80) power to detect small- to medium-sized effects (f � .20).

Procedure. As in previous studies, participants first reported basic demographic information and answered questions about their same-sex friends before reading and responding to one of two possible scenarios.

In Study 8a, participants were randomly assigned to read about a best friend becoming potentially closer with either a same-sex

friend or an other-sex romantic partner (Sarah and Mike). In Study 8b, participants were randomly assigned to read about a best friend choosing to take the participant to a desirable event (over the best friend’s new same-sex friend who also wanted to attend) or about that best friend choosing to take the new friend to a desirable event (over the participant who also wanted to attend). For both studies, participants then reported their affective reactions to the scenario and completed the seating distance question.

Reported friendship jealousy. Following the scenario, par- ticipants reported their emotional reactions, including friendship jealousy and distractor emotions (Study 8a: happy, proud, pity; Study 8b: guilt, pity, afraid, comfortable)— on 7-point Likert- scales (1 � not at all, 7 � very much). See online supplemental materials for all Means [SEs] of distractor reactions.

Chosen seating distance (friend guarding). After reporting their reactions, participants read a second prompt: “You’re about to take a big trip for work, AND you’ve been working on helping your group of friends plan a surprise birthday party for one of your mutual friends. Everything is set, but you have one more task to do before you leave for your trip (meaning that, unfortunately, you won’t be at the party). You have to make the seating arrangements. Most of it is locked in; that is, there are some people at the side tables that you can’t move around. But there are other people that you have to assign seats for—including your best friend [name of best friend] and [interloper], who’s also coming. You’re helping out with the seating at the last minute, so no one will know that you made these seating arrangements.”

Participants viewed a seating arrangement chart (see Figure 11) and chose where to seat the best friend and also the interloper (seats A through I). These decisions formed the basis of our second dependent variable, chosen seating distance, which we created by taking the absolute value of the distance between seat choices for the best friend and interloper. Distance scores of 0 —meaning best friends and interlopers were sat in the same chair—were not included in analyses.

Stage

A B C D E F H IG

Figure 11. Image viewed by participants in Studies 8a and 8b.

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1001FRIENDSHIP JEALOUSY

Results and Discussion

Reported friendship jealousy. Study 8a. We explored reported friendship jealousy caused by

interloper types via a one-way ANOVA, finding the predicted effect of interloper type, F(1, 197) � 7.50, p � .007, �p2 � .037, 95% CI [0.22, 1.33]; people reported greater friendship jealousy when interlopers were friends (M � 4.10, SE � .20) than romantic partners (M � 3.33, SE � .20).

Study 8b. We explored reported friendship jealousy caused by interloper types via a one-way ANOVA, finding the predicted effect of replacement threat, F(1, 227) � 57.41, p � .001, �p2 � .233, 95% CI [1.53, 2.61]; people reported greater friendship jealousy when interlopers posed a higher replacement threat (M � 4.76, SE � .19) than a lower replacement threat (M � 2.69, SE � .20).

Chosen seating distance (separation). Study 8a. We conducted a one-way ANOVA, finding that

people chose to seat best friends about one seat farther away from friend-interlopers (M � 2.79, SE � .21) than romantic- interlopers (M � 1.69, SE � .20), F(1, 186) � 14.53, p � .001, 95% CI [0.53, 1.68], �p2 � .072.

Study 8b. We conducted a one-way ANOVA, finding that people chose to seat best friends about one seat farther away from rivals posing high replacement threat (M � 3.63, SE � .24) than low replacement threat (M � 2.49, SE � .26), F(1, 171) � 10.33, p � .002, 95% CI [0.44, 1.83], �p2 � .057.

Does reported friendship jealousy mediate the relationship between interloper type and intensity of friend-guarding behavior? Yes. To explore the overarching model—that inter- lopers posing higher replacements threats evoke more friendship jealousy, and, in turn, more intense friend guarding, we used PROCESS Model 4 (Hayes, 2017) for 5,000 bootstrapped samples to compute a bias corrected 95% CI. Whether replacement threat was operationalized via interloper type (Study 8a) or explicitly manipulated (Study 8b), greater replacement threat evoked more friendship jealousy and more intense friend-guarding behavior.

Study 8a. The analysis indicated a significant indirect effect of replacement threat on friend guarding behavior (b � 0.25, SE � 0.11, 95% CI [0.07, 0.52]), with the direct effect also significant (b � 0.8, SE � 0.28, p � .003, 95% CI [0.30, 1.41]).

Study 8b. The analysis indicated a significant indirect effect of replacement threat on friend guarding behavior (b � 0.74, SE � 0.24, 95% CI [0.30, 1.25]), which rendered the direct effect non- significant (b � 0.39, SE � 0.40, p � .331, 95% CI [�0.40, 1.18]).

General Discussion

We derived predictions about the features that friendship jeal- ousy might possess if it were well-designed to help us solve the likely recurrent challenge of retaining friends in the face of third- party threats. In 11 studies— exploring friendship jealousy via hypothetical scenario, recalled experience, and manipulated online experience—we supported predictions regarding the architecture of this emotion: elucidating which inputs do (and do not) evoke friendship jealousy, which cues are prioritized in driving levels of friendship jealousy, and which behavioral outputs friendship jeal- ousy seems to propel (i.e., friend guarding).

First, we predicted and found that friendship jealousy was evoked at the prospective loss of best friends to third parties (vs. at the prospective loss of friends alone [e.g., as when a friend moves away]). Results are consistent with the expectation that friendship jealousy, unlike often concomitantly experienced emo- tions of sadness and/or anger, is uniquely evoked by third-party threats to friendships. Second, friendship jealousy is sensitive to the value of the threatened friendship (or friendship closeness), such that the prospective loss of best friends evokes greater friend- ship jealousy than does the prospective loss of close friends or acquaintances. Third, friendship jealousy is also strongly cali- brated to cues that one is about to be replaced. That is, drawing on evolutionary accounts of friendship (e.g., DeScioli & Kurzban, 2009, 2011; DeScioli et al., 2011; Geary, 1998; Silk, 2003; Tooby & Cosmides, 1996; Trivers, 1971), we predicted and found that friendship jealousy is more strongly calibrated to cues that another person is going to usurp one’s place in the focal friendship (re- placement threat) than it is to best friends simply spending more time with other people (time threat). Fourth, we consistently find that friendship jealousy drives friend guarding, a suite of behavior seemingly well-designed to mitigate third-party threats to friend- ship.

Across studies, these effects are seemingly specific to friendship jealousy relative to often-concomitant emotions of sadness or anger. It is unlikely that these effects are solely attributable to issues of study design or methodological paradigm (e.g., using within- vs. between-subjects design, using hypothetical scenarios). Rather, we support predictions across studies that use hypothetical scenarios, recalled reactions, and manipulate online emotion.

The proposition that friendship jealousy might have arisen to serve some beneficial ends challenges existing concepts of friend- ship jealousy in scientific research (e.g., Selman, 1980) and also lay intuition (Alford, 2014), which frame jealousy as maladaptive and/or pathological (for a similar review regarding romantic jeal- ousy, see Buss, 2013). Early accounts of friendship jealousy as- serted that such “immature” feelings abate after adolescence in normally developing individuals. A strong version of this argu- ment might expect friendship jealousy would not be present in the majority of adults (Selman, 1980; Selman & Schultz, 1990). How- ever, the present work reveals robust and contextually dependent levels of friendship jealousy among adults, presumably the major- ity of whom are normally developing. Rather, findings reported here are in line with an evolutionary view of friendship jealousy.

Is Friendship Jealousy Really Adapted?

We took a functional approach to derive the present predictions, beginning with the premise that jealousy may be a complex reac- tion tailored by natural selection to coordinate cognition, motiva- tion, physiology, and behavioral outputs in response to recurrent adaptive problems (Nesse, 1990; Sznycer et al., 2016; Tooby & Cosmides, 1996). Yet, some of our predictions might be similarly derived from other theoretical approaches or, sometimes, even intuition. How can one make a case that friendship jealousy is an adapted response to an ancestrally recurrent challenge? After all, unlike bones, cognition and behavior do not fossilize.

Making the case for a recurrent problem linked to fitness. To make such a case, one might first explore whether sustained friendships likely provided recurrent fitness benefits, and whether

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1002 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

third parties likely posed recurrent threats to the maintenance of those friendships. That is, might there be a recurrent problem tributary to reproductive success for friendship jealousy to “solve”?

As noted above, theory and evidence suggest that sustained friendships may have long provided benefits— both across human cultures and also among nonhuman primates (Ackerman et al., 2007; Aktipis et al., 2018; Barclay, 2013; Benenson, 2014; Boesch, 2009; Campbell, 2002; David-Barrett et al., 2015; De Waal & Waal, 2007; DeScioli & Kurzban, 2009, 2011; DeScioli et al., 2011; Dunbar, 2018; Hrdy, 2011; Hruschka, 2010; Lewis et al., 2011; Roberts, 2005; Seyfarth & Cheney, 2012; Schino, 2001, 2007; Shaw et al., 2017; Silk, 2002, 2003; Sugiyama, 2004; Tooby & Cosmides, 1996; Trivers, 1971). Among some nonhuman pri- mates, for example, the support of merely one ally in agonistic interactions within small, interconnected groups can have huge consequences on whether a disputant leads the group and gleans the related benefits, or suffers expulsion (e.g., Bergman & Shee- han, 2013; de Waal & Waal, 2007; Seyfarth & Cheney, 2012). Particularly among female nonhuman primates, the support of merely one ally can mitigate the negative effects of social stres- sors—with major impacts on fitness, such as extending one’s life or the life of one’s offspring (e.g., Palombit et al., 2000; Palombit, Seyfarth, & Cheney, 1997; Silk et al., 2009, 2010; Weingrill, Lycett, & Henzi, 2000). Fitness benefits of sustained friendships have also been demonstrated in other social species, including dogs, horses, cows, and dolphins (e.g., Cameron, Setsaas, & Lin- klater, 2009; Connor, 2007; Cook, Prichard, Spivak, & Berns, 2018).

There is also theory and evidence to suggest that third parties could have been recurrent threats to friendships. Both modern and ancestral humans—as well as our contemporary nonhuman pri- mate cousins—interacted in densely interconnected social groups, competed for desirable friends, and invested finite resources in some of those friends over others (DeScioli & Kurzban, 2011; Feld, 1981; Krems & Conroy-Beam, 2020; Seyfarth & Cheney, 2012). Third-party threats do not require a constant stream of new possible friends in and out of groups; rather, it takes as few as three people for a friendship to be threatened: two existing friends and a third party. Thus, even as ancestral humans did not encounter nearly as many new potential friends (e.g., same-sex, similarly aged strangers) as we might today, evidence suggests that they navigated a complex web of social challenges that could be prompted by allies becoming closer with other existing allies and/or newcomers entering one’s social group (e.g., Bird et al., 2019).

If sustaining friendships in the face of third-party threats could have been a recurrent and fitness-linked challenge, then following Williams (1966; see also Alcock, 1993; Barkow, Cosmides, & Tooby, 1992; Lewis et al., 2017; Schmitt & Pilcher, 2004), we can ask whether friendship jealousy shows evidence of “special de- sign” to solve such a challenge. The present work begins to provide some evidence for this, as friendship jealousy appears to be evoked by some threats (but not others), and to spur behavior aimed at mitigating such threats.

Building a nomological network of evidence that friendship jealousy could have helped meet that recurrent challenge. An argument for functional significance can also be addressed by a nomological network incorporating diverse forms of evidence

(Schmitt & Pilcher, 2004). Such an approach integrates not only (a) theoretical and (b) empirical evidence (e.g., the experimental evidence we present here; see also Burkett, 2009), but also (c) cross-cultural evidence suggesting a psychological phenomenon is reliably developing (i.e., “universal”; Brown, 1991), and (d) phy- logenetic evidence of homologous behavior in nonhuman animals, especially modern primates, suggesting a phenomenon may be highly conserved (e.g., Tinbergen, 1963), as well as other forms of evidence (e.g., genetic, physiological).

With regard to (c) cross-cultural evidence, accounts of friend- ship jealousy are found across cultures in the ethnographic record (e.g., Tarascan farmers in Mexico [Friedrich, 1965], Pashtun herd- ers [Lindholm, 1982], Copper Inuit [Condon, 1987], and Guate- malan villagers [Reina, 1959]; see Hruschka, 2010). Although we cannot assess nonhuman animals’ internal states as we can with humans’, (d) there is behavioral evidence consistent with friend- ship jealousy in nonhuman animals (e.g., de Waal & Waal, 2007; Harcourt, 1992; Schino, 2001; Seyfarth, 1977; Seyfarth & Cheney, 2012). For example, DeScioli and Kurzban (2011) refer to work on chimpanzees “jealously prevent[ing] the formation of rival rela- tionships” (p. 211), referring to alliances, not romantic bonds.

In line with Schmitt and Pilcher’s (2004) suggestions, it could also be interesting to consider other modes of evidence. One could imagine future work exploring physiological responses to distinct friendship threats (e.g., electromyographic activity, skin conduc- tance), testing whether reactivity patterns match those of friend- ship jealousy reported here (e.g., Pietrzak, Laird, Stevens, & Thompson, 2002; Takahashi et al., 2006).

Ultimately, we do not advance the claim that our findings prove friendship jealousy is an adaptation, nor do we suggest that our findings cannot be explained post hoc by alternative frameworks. Rather, we note that our predictions were derived from a functional perspective, that our findings are consistent with the notion of friendship jealousy as a psychological mechanism well-designed to respond to likely recurrent threats of third-party interference, and that no existing framework has generated or tested the predic- tions we describe here.

Potential Limitations and Future Directions

We argue that friendship jealousy is potentially beneficial inso- far as it might spur behavior that helps individuals maintain friendships. However, it is worth asking whom friendship jealousy might benefit: actors, friends, and/or third parties? In particular, we speculate that friendship jealousy is beneficial for actors more so than actors’ friends or third parties. Indeed, we have not speculated on nor explored the impact of friendship jealousy on focal friends or third parties, as future work should do. Related though, Owens and colleagues (2000) describe harmful possessiveness over friends among teenage girls in Australia. Such possessiveness, which might be a means of friend guarding, is most likely to be harmful to the focal friend and/or the third party, insofar as it prevents the focal friend and third party from enjoying the benefits that might have been associated with the precluded friendship.

Additionally, although we argue that friendship jealousy can facilitate some positive ends— even as it may be negative to experience—we would not expect that actors feel only aversive reactions when friends form new bonds. Some scenarios might evoke mixed emotions (e.g., both jealousy and pride), for example,

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1003FRIENDSHIP JEALOUSY

as when a best friend forms a romantic relationship. Exploring the full range of these emotions was beyond the scope of the present work.

One might be concerned that, because jealousy is a negative feeling, and/or because friendship jealousy might be considered shameful to experience, there are problems in assessing jealousy via self-report. On this view, one might expect a response bias toward consistently low levels of reported friendship jealousy. However, reported levels of friendship jealousy not only system- atically varied in line with nuanced predictions, but they also went above the midpoint of the scale, suggesting that such a response bias is unlikely to be problematic here.

Participants here come from Western, Educated, Industrialized, Rich, and Democratic samples (WEIRD; Henrich, Heine, & Norenzayan, 2010). We would expect friendship jealousy to exist across nations, cultures, subsistence patterns and so on—and evi- dence suggests that it does (Hruschka, 2010). Future work should include large-scale studies, studies in non-Western samples, and studies in small-scale societies to assess both the possible univer- sality of friendship jealousy as well as possible cultural differences therein (see, e.g., Scelza, 2014; Sznycer, Xygalatas, Alami, et al., 2018a, Sznycer, Xygalatas, Agey, et al., 2018). Understanding differing cultural norms about friendship exclusivity, for example, and/or features of different local ecologies (e.g., number of pro- spective friends we might encounter; Schug, Yuki, Horikawa, & Takemura, 2009) could lead to empirically testable predictions about systematic differences in friendship jealousy (see, e.g., Scelza et al., 2019).

Friend guarding. We explored friend guarding to provide support for the notion that friendship jealousy motivates expected behavioral inclinations. There are likely to be multiple outputs of friendship jealousy, with friend guarding being just one. Our preliminary investigation suggests further exploration of friend guarding is warranted (see Schützwohl, Joshi, & Abdur-Razak, 2019). Indeed, research on romantic jealousy spurred myriad pa- pers on mate guarding, and we would hope to see future work explore the nuances of friend guarding.

A first question for such future work has to do with a limitation here: at least in Study 3a, self-reported friend-guarding intentions were relatively low. Why? It is possible that our full scale included mating-relevant tactics that people eschew when guarding friends and/or failed to include important tactics unique to friend guarding. Another possibility is that friend guarding is simply not efficacious in modern environments and, thus, adults do not engage in it. Because of the observations of friend guarding in nonhuman animals (e.g., Seyfarth & Cheney, 2012), children and adolescents (e.g., Owens et al., 2000), and in adults across cultures (e.g., Hruschka, 2010), we doubt that this is apt an account. Rather, perhaps adults’ friend guarding is quite subtle. After all, people rarely rate “possessiveness” highly on their list of friendship preferences. Indeed, we would not dispute that some jealousy- evoked reactions could harm friendships, as has been found among children (e.g., Parker et al., 2010). Perhaps compared with chil- dren, adults are more adept at engaging in tactics of friend guard- ing that mitigate the undesirable appearance and/or effects of friendship jealousy, while nevertheless acting on that experience in ways that successfully facilitate friendship maintenance.

A second question has to do with the efficacy of friend guarding. More important, an evolutionary argument does not depend on

modern friend guarding being efficacious. Rather, the approach we take presumes that friendship jealousy arose in response to a recurrent, fitness-relevant threat (third parties taking our friends), and that our ancestors who experienced friendship jealousy in germane situations were better able to keep those friends (and enjoy the benefits of doings so). We would not expect, however, that all friend guarding would necessarily be adaptive today (though it may still be adapted); for example, friend guarding in modern contexts could be maladaptive in the clinical sense, such that it can cause some negative outcomes (e.g., among children; Parker et al., 2010). The present work provides initial evidence hinting that friend guarding might be efficacious; participants who recalled having engaged in more friend guarding seemed to have more success in retaining their best friends. However, much more research is needed to explore the costs, benefits, and efficacy of friend guarding (for actors, friends, and third parties).

Other new avenues for friendship research. Recently, Kelci Harris and Simine Vazire wrote, “given how important friendships are for health and well-being, it is surprising that friendship has not received more attention from social and personality psychologists” (Harris & Vazire, 2016, p. 647). We would similarly sound the call to reinvigorate friendship research, and we believe that the present work serves as one illustration for how researchers might begin to address this issue. That is, even as friendship and other social relationships (e.g., with kin, romantic partners) may represent distinct social domains, we used the wealth of research on roman- tic relationships to provide a model for identifying both important challenges (e.g., relationship maintenance) and also possible tools for meeting those challenges (e.g., friendship jealousy) that were understudied in friendships.

In this fashion, we suggest just a few areas of future investiga- tion, beginning with explorations of friend value (in mating: e.g., Buss, 1989; Conroy-Beam & Buss, 2016a, 2016b, 2017; Edlund & Sagarin, 2014; Lewis et al., 2017; Li, Bailey, Kenrick, & Linsen- meier, 2002; Schmitt, 2010). For example, people high in friend value are in greater demand on the friend market (Krems & Conroy-Beam, 2020); they might be the targets of more friendship poaching attempts and perhaps also be more jealously guarded by their existing friends. Similarly, just as some romantic interlopers are considered more threatening to same-sex rivals than others (e.g., especially attractive women), so too might certain friend interlopers be considered especially threatening and likely to evoke high levels of friendship jealousy.

Related, Pat Barclay (2013) actually suggests a polygyny thresh- old, “whereby individuals must choose between befriending high- value partners who offer less time to each of many friends versus befriending low-value partners who offer much more devotion to fewer friends” (p. 170). This raises the question of whether people might use differing friendship strategies—akin to mating strategies (e.g., Buss & Schmitt, 1993; Simpson & Gangestad, 1992)— whereby some invest less in any one friend, and others invest heavily in one friend at a time. If so, who pursues these different strategies, and what are their consequences for strategists’ health and well-being?

Slightly further afield, one might also use existing work on romantic relationships as a playbook to explore understudied fea- tures, such as friendship (dis)satisfaction, sustainment, and/or dis- solution—and the features that predict them (in mating: e.g., Amato, 2000; Meltzer & McNulty, 2010; Meltzer, McNulty, Jack-

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1004 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

son, & Karney, 2014). For example, Conroy-Beam et al. (2016) found that the interplay between the mate values of an actor, their romantic partner, and the actor’s possible alternative mates better predicted actor relationship satisfaction than did having a romantic partner who fulfilled all the actor’s mating preferences. No similar work explores this in friendships. Indeed, with increasing attention being paid to loneliness—that affects roughly a third of people in industrialized countries—and its negative consequences (e.g., in- creased mortality risk) as a public health problem (Cacioppo et al., 2002; Cacioppo & Cacioppo, 2018; Murthy, 2017), unlocking the secrets to successfully sustaining friendships would seem to be a pressing problem with important practical implications.

Conclusion

Friendship jealousy possesses features one would expect of an apt tool for friend retention—specifically, a tool that is well designed to help prevent the loss of valued friends to other people. Friendship jealousy seems uniquely evoked by threats that others might pose to our friendships, is calibrated to how much we value our friends and to how likely it is we might lose those friends (and, thus, any friendship-derived benefits), and friendship jealousy also seems to motivate us to engage in behavior aimed at preventing such loss (friend guarding). In summary, friendship jealousy might be unpleasant to experience, but it may nevertheless help us maintain our friendships, making it one important but previously overlooked tool of friendship maintenance.

References

Ackerman, J. M., Kenrick, D. T., & Schaller, M. (2007). Is friendship akin to kinship? Evolution and Human Behavior, 28, 365–374. http://dx.doi .org/10.1016/j.evolhumbehav.2007.04.004

Adams, R. G., & Allan, G. (Eds.). (1998). Contextualizing friendship. Placing friendship in context (pp. 1–17). Cambridge, UK: Cambridge University Press.

Aktipis, A., Cronk, L., Alcock, J., Ayers, J. D., Baciu, C., Balliet, D., . . . Winfrey, P. (2018). Understanding cooperation through fitness interde- pendence. Nature Human Behaviour, 2, 429 – 431. http://dx.doi.org/10 .1038/s41562-018-0378-4

Alcock, J. (1993). Animal behavior: An evolutionary approach (5th ed.). Sunderland, MA: Sinauer.

Alford, H. (2014, October, 24). Striking a friendship match: The etiquette for having your friends befriend each other. The New York Times. Retrieved from https://www.nytimes.com/2014/10/26/fashion/the- etiquette-for-having-your-friends-befriend-each-other.html

Al-Shawaf, L., Conroy-Beam, D., Asao, K., & Buss, D. M. (2016). Human emotions: An evolutionary psychological perspective. Emotion Review, 8, 173–186. http://dx.doi.org/10.1177/1754073914565518

Amato, P. R. (2000). The consequences of divorce for adults and children. Journal of Marriage and the Family, 62, 1269 –1287. http://dx.doi.org/ 10.1111/j.1741-3737.2000.01269.x

Amodio, D. M., & Devine, P. G. (2006). Stereotyping and evaluation in implicit race bias: Evidence for independent constructs and unique effects on behavior. Journal of Personality and Social Psychology, 91, 652– 661. http://dx.doi.org/10.1037/0022-3514.91.4.652

Ashton-James, C. E., & Tracy, J. L. (2012). Pride and prejudice: How feelings about the self influence judgments of others. Personality and Social Psychology Bulletin, 38, 466 – 476.

Barakzai, A., & Shaw, A. (2018). Friends without benefits: When we react negatively to helpful and generous friends. Evolution and Human Be- havior, 39, 529 –537.

Barclay, P. (2013). Strategies for cooperation in biological markets, espe- cially for humans. Evolution and Human Behavior, 34, 164 –175. http:// dx.doi.org/10.1016/j.evolhumbehav.2013.02.002

Barclay, P. (2016). Biological markets and the effects of partner choice on cooperation and friendship. Current Opinion in Psychology, 7, 33–38. http://dx.doi.org/10.1016/j.copsyc.2015.07.012

Barkow, J. H., Cosmides, L., & Tooby, J. (Eds.). (1992). The adapted mind: Evolutionary psychology and the generation of culture. New York, NY: Oxford University Press.

Benenson, J. F. (2014). Warriors and worriers: The survival of the sexes. New York, NY: Oxford University Press.

Benenson, J. F., Nicholson, C., Waite, A., Roy, R., & Simpson, A. (2001). The influence of group size on children’s competitive behavior. Child Development, 72, 921–928.

Bergman, T. J., & Sheehan, M. J. (2013). Social knowledge and signals in primates. American Journal of Primatology, 75, 683– 694. http://dx.doi .org/10.1002/ajp.22103

Bhugra, D. (1993). Cross-cultural aspects of jealousy. International Review of Psychiatry, 5, 271–280. http://dx.doi.org/10.3109/0954026930 9028317

Bigelow, B. J., & La Gaipa, J. J. (1975). Children’s written descriptions of friendship: A multidimensional analysis. Developmental Psychology, 11, 857– 858. http://dx.doi.org/10.1037/0012-1649.11.6.857

Bigelow, B. J., Tesson, G., & Lewko, J. H. (1996). Learning the rules: The anatomy of children’s relationships. New York, NY: Guilford Press.

Bird, D. W., Bird, R. B., Codding, B. F., & Zeanah, D. W. (2019). Variability in the organization and size of hunter-gatherer groups: For- agers do not live in small-scale societies. Journal of Human Evolution, 131, 96 –108. http://dx.doi.org/10.1016/j.jhevol.2019.03.005

Bleske, A. L., & Buss, D. M. (2000). Can men and women be just friends? Personal Relationships, 7, 131–151.

Boesch, C. (2009). The real chimpanzee: Sex strategies in the forest. Cambridge, UK: Cambridge University Press. http://dx.doi.org/10.1017/ CBO9780511627132

Bringle, R. G. (1991). Psychosocial aspects of jealousy: A transactional model. In P. Salovey (Ed.), The psychology of jealousy and envy (pp. 103–131). New York, NY: Guilford Press.

Brown, D. E. (1991). Human universals. New York, NY: McGraw-Hill. Burkett, B. N. (2009). Friendship, jealousy, and the Banker’s Paradox

(Unpublished doctoral dissertation). University of California, Santa Bar- bara, Santa Barbara, CA.

Burleson, B. R., & Samter, W. (1994). The social skills approach to relationship maintenance: How individual differences in communication skills affect the achievement of relationship functions. In D. J. Canary & L. Stafford (Eds.), Communication and relational maintenance (pp. 61–90). San Diego, CA: Academic Press.

Burt, R. (2000). Decay functions. Social Networks, 22, 1–28. http://dx.doi .org/10.1016/S0378-8733(99)00015-5

Buss, D. M. (1989). Sex differences in human mate preferences: Evolu- tionary hypotheses tested in 37 cultures. Behavioral and Brain Sciences, 12, 1–14. http://dx.doi.org/10.1017/S0140525X00023992

Buss, D. M. (2000). The dangerous passion: Why jealousy is as necessary as love and sex. New York, NY: Simon & Schuster.

Buss, D. M. (2013). Sexual jealousy. Psihologijske Teme, 22, 155–182. Buss, D. M., & Haselton, M. (2005). The evolution of jealousy. Trends in

Cognitive Sciences, 9, 506 –507. http://dx.doi.org/10.1016/j.tics.2005.09 .006

Buss, D. M., Larsen, R., Westen, D., & Semmelroth, J. (1992). Sex differences in jealousy: Evolution, physiology, and psychology. Psycho- logical Science, 3, 251–256. http://dx.doi.org/10.1111/j.1467-9280.1992 .tb00038.x

Buss, D. M., & Schmitt, D. P. (1993). Sexual strategies theory: An evolutionary perspective on human mating. Psychological Review, 100, 204 –232. http://dx.doi.org/10.1037/0033-295X.100.2.204

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1005FRIENDSHIP JEALOUSY

Buss, D. M., & Shackelford, T. K. (1997). From vigilance to violence: Mate retention tactics in married couples. Journal of Personality and Social Psychology, 72, 346 –361. http://dx.doi.org/10.1037/0022-3514 .72.2.346

Buss, D. M., Shackelford, T. K., & McKibbin, W. F. (2008). The mate retention inventory-short form (MRI-SF). Personality and Individual Differences, 44, 322–334. http://dx.doi.org/10.1016/j.paid.2007.08.013

Buunk, B. (1981). Jealousy in sexually open marriages. Alternative Life- styles, 4, 357–372. http://dx.doi.org/10.1007/BF01257944

Buunk, B. (1982). Anticipated sexual jealousy: Its relationship to self- esteem, dependency, and reciprocity. Personality and Social Psychology Bulletin, 8, 310 –316. http://dx.doi.org/10.1177/0146167282082019

Buunk, B. P., Angleitner, A., Oubaid, V., & Buss, D. M. (1996). Sex differences in jealousy in evolutionary and cultural perspective: Tests from the Netherlands, Germany, and the United States. Psychological Science, 7, 359 –363. http://dx.doi.org/10.1111/j.1467-9280.1996 .tb00389.x

Cacioppo, J. T., & Cacioppo, S. (2018). The growing problem of loneli- ness. The Lancet, 391, 426. http://dx.doi.org/10.1016/S0140- 6736(18)30142-9

Cacioppo, J. T., Hawkley, L. C., Crawford, L. E., Ernst, J. M., Burleson, M. H., Kowalewski, R. B., . . . Berntson, G. G. (2002). Loneliness and health: Potential mechanisms. Psychosomatic Medicine, 64, 407– 417. http://dx.doi.org/10.1097/00006842-200205000-00005

Cameron, E. Z., Setsaas, T. H., & Linklater, W. L. (2009). Social bonds between unrelated females increase reproductive success in feral horses. Proceedings of the National Academy of Sciences of the United States of America, 106, 13850 –13853. http://dx.doi.org/10.1073/pnas.0900 639106

Campbell, A. C. (2002). A mind of her own: The evolutionary psychology of women. Oxford, UK: Oxford University Press. http://dx.doi.org/10 .1093/acprof:oso/9780198504986.001.0001

Campos, J. J., Walle, E. A., & Dahl, A. (2010). What is missing in the study of the development of jealousy. In S. L. Hart & M. Legerstee (Eds.), Handbook of jealousy: Theory, research, and multidisciplinary approaches (pp. 312–328). Malden, MA: Wiley-Blackwell.

Canary, D. J., Stafford, L., Hause, K. S., & Wallace, L. A. (1993). An inductive analysis of relational maintenance strategies: Comparisons among lovers, relatives, friends, and others. Communication Research Reports, 10, 3–14. http://dx.doi.org/10.1080/08824099309359913

Casper, D. M., & Card, N. A. (2010). “We were best friends, but . . .”: Two studies of antipathetic relationships emerging from broken friendships. Journal of Adolescent Research, 25, 499 –526. http://dx.doi.org/10.1177/ 0743558410366596

Condon, R. G. (1987). Inuit youth. New Brunswick, NJ: Rutgers University Press.

Connor, R. C. (2007). Dolphin social intelligence: Complex alliance rela- tionships in bottlenose dolphins and a consideration of selective envi- ronments for extreme brain size evolution in mammals. Philosophical Transactions of the Royal Society of London Series B, Biological Sci- ences, 362, 587– 602. http://dx.doi.org/10.1098/rstb.2006.1997

Conroy-Beam, D., & Buss, D. M. (2016a). Do mate preferences influence actual mating decisions? Evidence from computer simulations and three studies of mated couples. Journal of Personality and Social Psychology, 111, 53– 66. http://dx.doi.org/10.1037/pspi0000054

Conroy-Beam, D., & Buss, D. M. (2016b). How are mate preferences linked with actual mate selection? Tests of mate preference integration algorithms using computer simulations and actual mating couples. PLoS ONE, 11, e0156078. http://dx.doi.org/10.1371/journal.pone.0156078

Conroy-Beam, D., & Buss, D. M. (2017). Euclidean distances discrimina- tively predict short-term and long-term attraction to potential mates. Evolution and Human Behavior, 38, 442– 450. http://dx.doi.org/10.1016/ j.evolhumbehav.2017.04.004

Conroy-Beam, D., Goetz, C. D., & Buss, D. M. (2016). What predicts romantic relationship satisfaction and mate retention intensity: Mate preference fulfillment or mate value discrepancies? Evolution and Hu- man Behavior, 37, 440 – 448. http://dx.doi.org/10.1016/j.evolhumbehav .2016.04.003

Cook, P., Prichard, A., Spivak, M., & Berns, G. S. (2018). Jealousy in dogs? Evidence from brain imaging. Animal Sentience, 3, 1.

Cottrell, C. A., Neuberg, S. L., & Li, N. P. (2007). What do people desire in others? A sociofunctional perspective on the importance of different valued characteristics. Journal of Personality and Social Psychology, 92, 208 –231. http://dx.doi.org/10.1037/0022-3514.92.2.208

Cummings, J. N., Lee, J. B., & Kraut, R. (2006). Communication technol- ogy and friendship during the transition from high school to college. In K. Kraut, M. Brynin, & S. Kiesler (Eds.), Computers, phones and the internet: Domesticating information technologies (pp. 265–278). New York, NY: Oxford University Press.

Daly, M., Wilson, M., & Weghorst, S. J. (1982). Male sexual jealousy. Ethology and Sociobiology, 3, 11–27. http://dx.doi.org/10.1016/0162- 3095(82)90027-9

David-Barrett, T., Rotkirch, A., Carney, J., Behncke Izquierdo, I., Krems, J. A., Townley, D., . . . Dunbar, R. I. (2015). Women favour dyadic relationships, but men prefer clubs: Cross-cultural evidence from social networking. PLoS ONE, 10, e0118329. http://dx.doi.org/10.1371/journal .pone.0118329

Davis, K. E., & Todd, M. J. (1982). Friendship and love relationships. Advances in Descriptive Psychology, 2, 79 –122.

DelPriore, D. J., Hill, S. E., & Buss, D. M. (2012). Envy: Functional specificity and sex-differentiated design features. Personality and Indi- vidual Differences, 53, 317–322. http://dx.doi.org/10.1016/j.paid.2012 .03.029

Delton, A. W., & Robertson, T. E. (2012). The social cognition of social foraging: Partner selection by underlying valuation. Evolution and Hu- man Behavior, 33, 715–725. http://dx.doi.org/10.1016/j.evolhumbehav .2012.05.007

DeScioli, P., & Kurzban, R. (2009). The alliance hypothesis for human friendship. PLoS ONE, 4, e5802. http://dx.doi.org/10.1371/journal.pone .0005802

DeScioli, P., & Kurzban, R. (2011). The company you keep: Friendship decisions from a functional perspective. In J. I. Krueger (Ed.), Social judgment and decision making (pp. 209 –225). New York, NY: Psychol- ogy Press.

DeScioli, P., Kurzban, R., Koch, E. N., & Liben-Nowell, D. (2011). Best friends: Alliances, friend ranking, and the MySpace social network. Perspectives on Psychological Science, 6, 6 – 8. http://dx.doi.org/10 .1177/1745691610393979

DeSteno, D. A., & Salovey, P. (1996). Jealousy and the characteristics of one’s rival: A self-evaluation maintenance perspective. Personality and Social Psychology Bulletin, 22, 920 –932. http://dx.doi.org/10.1177/ 0146167296229006

DeSteno, D., Valdesolo, P., & Bartlett, M. Y. (2006). Jealousy and the threatened self: Getting to the heart of the green-eyed monster. Journal of Personality and Social Psychology, 91, 626 – 641. http://dx.doi.org/ 10.1037/0022-3514.91.4.626

de Vries, B., & Johnson, C. (2002). The death of friends in later life. Advances in Life Course Research, 7, 299 –324. http://dx.doi.org/10 .1016/S1040-2608(02)80038-7

De Waal, F., & Waal, F. B. (2007). Chimpanzee politics: Power and sex among apes. Baltimore, MD: JHU Press.

Dunbar, R. I. M. (1993). Coevolution of neocortical size, group size and language in humans. Behavioral and Brain Sciences, 16, 681– 694. http://dx.doi.org/10.1017/S0140525X00032325

Dunbar, R. I. M. (2008). Cognitive constrains on the structure and dynam- ics of social networks. Group Dynamics, 12, 7–16. http://dx.doi.org/10 .1037/1089-2699.12.1.7

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1006 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

Dunbar, R. I. M. (2012, March, 21). Can the Internet buy you more friends? Tedx Observer Talk. Retrieved from https://www.youtube.com/ watch?v�07IpED729k8

Dunbar, R. I. M. (2018). The anatomy of friendship. Trends in Cognitive Sciences, 22, 32–51. http://dx.doi.org/10.1016/j.tics.2017.10.004

Ebrahimi, N., Parker, J. G., Lavallee, K., & Seiffke-Krenke, I. (2005, April). Adolescent self-esteem and jealousy’s reciprocal relations over time. Paper presented at the Biennial Meeting of the Society for Re- search in Child Development, Atlanta, GA.

Edlund, J. E., & Sagarin, B. J. (2014). The mate value scale. Personality and Individual Differences, 64, 72–77. http://dx.doi.org/10.1016/j.paid .2014.02.005

Eisenbruch, A. B., & Roney, J. R. (2017). The Skillful and the stingy: Partner choice decisions and fairness intuitions suggest human adapta- tion for a biological market of cooperators. Evolutionary Psychological Science, 3, 364 –378. http://dx.doi.org/10.1007/s40806-017-0107-7

Ekman, P., Levenson, R. W., & Friesen, W. V. (1983). Autonomic nervous system activity distinguishes among emotions. Science, 221, 1208 – 1210. http://dx.doi.org/10.1126/science.6612338

Fehr, B. (1996). Friendship processes. Newbury Park, CA: Sage. http://dx .doi.org/10.4135/9781483327440

Feld, S. L. (1981). The focused organization of social ties. American Journal of Sociology, 86, 1015–1035. http://dx.doi.org/10.1086/227352

Freud, S. (1910). Contributions to the psychology of love. Papers XI, XII, XIII in Collected Papers, 4, 192–235.

Friedrich, P. (1965). A Mexican cacicazgo. Ethnology, 4, 190 –209. http:// dx.doi.org/10.2307/3772729

Gallup Poll News Service. (2005, December). Gallup Poll social series: Lifestyle. Retrieved from http://brain.gallup.com/documents/ questionnaire.aspx?STUDY�P0512059&p�3

Geary, D. C. (1998). Male, female: The evolution of human sex differences. Washington, DC: American Psychological Association. http://dx.doi .org/10.1037/10370-000

Giltenboth, K. J. (2001). Expressions of jealousy in friendships and ro- mantic relationships: Influence of gender and type of rival. Unpublished honors thesis. Pennsylvania State University, PA.

Hall, J. A., Larson, K. A., & Watts, A. (2011). Satisfying friendship maintenance expectations: The role of friendship standards and biolog- ical sex. Human Communication Research, 37, 529 –552. http://dx.doi .org/10.1111/j.1468-2958.2011.01411.x

Harcourt, A. H. (1992). Coalitions and alliances: Are primates more complex than non-primates? In A. H. Harcourt & F. B. M. de Waal (Eds.), Coalitions and alliances in humans and other animals (pp. 445– 471). Oxford, UK: Oxford University Press.

Harris, K., & Vazire, S. (2016). On friendship development and the Big Five personality traits. Social and Personality Psychology Compass, 10, 647– 667. http://dx.doi.org/10.1111/spc3.12287

Hayes, A. F. (2017). Introduction to mediation, moderation, and condi- tional process analysis: A regression-based approach. New York, NY: Guilford Press publications.

Hays, R. B. (1985). A longitudinal study of friendship development. Journal of Personality and Social Psychology, 48, 909 –924. http://dx .doi.org/10.1037/0022-3514.48.4.909

Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33, 61– 83. http://dx.doi.org/ 10.1017/S0140525X0999152X

Holt-Lunstad, J., Smith, T. B., Baker, M., Harris, T., & Stephenson, D. (2015). Loneliness and social isolation as risk factors for mortality: A meta-analytic review. Perspectives on Psychological Science, 10, 227– 237. http://dx.doi.org/10.1177/1745691614568352

Hrdy, S. B. (2011). Mothers and others. Cambridge, MA: Harvard Uni- versity Press.

Hruschka, D. J. (2010). Friendship: Development, ecology and evolution of a relationship. Berkeley, CA: University of California Press. http://dx .doi.org/10.1525/california/9780520265462.001.0001

Hupka, R. B. (1991). The motive for arousal of romantic jealousy: Its cultural origin. In P. Salovey (Ed.), The psychology of jealousy and envy (pp. 252–270). New York, NY: Guilford Press.

Johnson, M. P., & Leslie, L. (1982). Couple involvement and network structure—A test of the dyadic withdrawal hypothesis. Social Psychol- ogy Quarterly, 45, 34 – 43. http://dx.doi.org/10.2307/3033672

Kenrick, D. T., Griskevicius, V., Neuberg, S. L., & Schaller, M. (2010). Renovating the pyramid of needs: Contemporary extensions built upon ancient foundations. Perspectives on Psychological Science, 5, 292–314. http://dx.doi.org/10.1177/1745691610369469

Kenrick, D. T., & Krems, J. A. (2018). Well-being, self-actualization, and fundamental motives: An evolutionary perspective. In E. Diener, S. Oishi, & L. Tay (Eds.), Handbook of well-being. Salt Lake City, UT: DEF Publishers. Retrieved from http://nobascholar.com

Kenrick, D. T., Neuberg, S. L., Griskevicius, V., Becker, D. V., & Schaller, M. (2010). Goal-driven cognition and functional behavior: The fundamental-motives framework. Current Directions in Psychological Science, 19, 63– 67. http://dx.doi.org/10.1177/0963721409359281

Kraft, C., & Mayeux, L. (2016). Associations among friendship jealousy, peer status, and relational aggression in early adolescence. The Journal of Early Adolescence, 38, 385– 407. http://dx.doi.org/10.1177/ 0272431616670992

Krems, J. A., & Conroy-Beam, D. (2020). First tests of Euclidean prefer- ence integration in friendship: Euclidean friend value and power of choice on the friend market. Evolution and Human Behavior, 41, 188 – 198. http://dx.doi.org/10.1016/j.evolhumbehav.2020.02.003

Krems, J. A., Dunbar, R. I., & Neuberg, S. L. (2016). Something to talk about: Are conversation sizes constrained by mental modeling abilities? Evolution and Human Behavior, 37, 423– 428. http://dx.doi.org/10.1016/ j.evolhumbehav.2016.05.005

Krems, J. A., Kenrick, D. T., & Neel, R. (2017). Individual perceptions of self-actualization: What functional motives are linked to fulfilling one’s full potential? Personality and Social Psychology Bulletin, 43, 1337– 1352. http://dx.doi.org/10.1177/0146167217713191

Krems, J. A., & Wilkes, J. (2019). Why are conversations limited to about four people? A theoretical exploration of the conversation size con- straint. Evolution and Human Behavior, 40, 140 –147. http://dx.doi.org/ 10.1016/j.evolhumbehav.2018.09.004

Lavallee, K. L., & Parker, J. G. (2009). The role of inflexible friendship beliefs, rumination, and low self-worth in early adolescents’ friendship jealousy and adjustment. Journal of Abnormal Child Psychology, 37, 873– 885. http://dx.doi.org/10.1007/s10802-009-9317-1

Lewis, D. M. G., Al-Shawaf, L., Conroy-Beam, D., Asao, K., & Buss, D. M. (2017). Evolutionary psychology: A how-to guide. American Psychologist, 72, 353–373. http://dx.doi.org/10.1037/a0040409

Lewis, D. M. G., Conroy-Beam, D., Al-Shawaf, L., Raja, A., DeKay, T., & Buss, D. M. (2011). Friends with benefits: The evolved psychology of same- and opposite-sex friendship. Evolutionary Psychology, 9, 543– 563. http://dx.doi.org/10.1177/147470491100900407

Li, N. P., Bailey, J. M., Kenrick, D. T., & Linsenmeier, J. A. (2002). The necessities and luxuries of mate preferences: Testing the tradeoffs. Journal of Personality and Social Psychology, 82, 947–955. http://dx .doi.org/10.1037/0022-3514.82.6.947

Lindholm, C. (1982). Generosity and jealousy: The Swat Pukhtun of northern Pakistan. New York, NY: Columbia University Press.

McEwan, B., Babin Gallagher, B., & Farinelli, L. (2008). The end of a friendship: Friendship dissolution reasons and methods. Research pre- sented at the National Communication Association Annual Conference, San Diego, CA.

Meltzer, A. L., & McNulty, J. K. (2010). Body image and marital satis- faction: Evidence for the mediating role of sexual frequency and sexual

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1007FRIENDSHIP JEALOUSY

satisfaction. Journal of Family Psychology, 24, 156 –164. http://dx.doi .org/10.1037/a0019063

Meltzer, A. L., McNulty, J. K., Jackson, G. L., & Karney, B. R. (2014). Sex differences in the implications of partner physical attractiveness for the trajectory of marital satisfaction. Journal of Personality and Social Psychology, 106, 418 – 428. http://dx.doi.org/10.1037/a0034424

Milardo, R. M., Johnson, M. P., & Huston, T. L. (1983). Developing close relationships: Changing patterns of interaction between pair members and social networks. Journal of Personality and Social Psychology, 44, 964 –976. http://dx.doi.org/10.1037/0022-3514.44.5.964

Miller, A. L., Volling, B. L., & McElwain, N. L. (2000). Sibling jealousy in a triadic context with mothers and fathers. Social Development, 9, 433– 457. http://dx.doi.org/10.1111/1467-9507.00137

Miritello, G., Moro, E., Lara, R., Martínez-López, R., Belchamber, J., Roberts, S. G., & Dunbar, R. I. (2013). Time as a limited resource: Communication strategy in mobile phone networks. Social Networks, 35, 89 –95. http://dx.doi.org/10.1016/j.socnet.2013.01.003

Montoya, A. K., & Hayes, A. F. (2017). Two-condition within-participant statistical mediation analysis: A path-analytic framework. Psychological Methods, 22, 6 –27. http://dx.doi.org/10.1037/met0000086

Murthy, V. (2017). Work and the loneliness epidemic. Harvard Business Review (Brighton), Sept 28, 2017. Retrieved from https://hbr.org/ coverstory/2017/09/work-and-the-lonelinessepidemic

Nesse, R. M. (1990). Evolutionary explanations of emotions. Human Nature, 1, 261–289. http://dx.doi.org/10.1007/BF02733986

Nesse, R. M. (2007). Runaway social selection for displays of partner value and altruism. Biological Theory, 2, 143–155. http://dx.doi.org/10.1162/ biot.2007.2.2.143

Nesse, R. M. (2019). Good reasons for bad feelings: Insights from the frontier of evolutionary psychiatry. New York, NY: Dutton.

Neuberg, S. L., Kenrick, D. T., & Schaller, M. (2010). Evolutionary social psychology. In S. T. Fiske, D. T. Gilbert, & G. Lindzey (Eds.), Hand- book of social psychology (5th ed., Vol. 2, pp. 761–796). New York, NY: Wiley. http://dx.doi.org/10.1002/9780470561119.socpsy002021

Noë, R., & Hammerstein, P. (1994). Biological markets: Supply and demand determine the effect of partner choice in cooperation, mutualism and mating. Behavioral Ecology and Sociobiology, 35, 1–11. http://dx .doi.org/10.1007/BF00167053

Oswald, D. L., & Clark, E. M. (2006). How do friendship maintenance behaviors and problem-solving styles function at the individual and dyadic levels? Personal Relationships, 13, 333–348. http://dx.doi.org/ 10.1111/j.1475-6811.2006.00121.x

Oswald, D. L., Clark, E. M., & Kelly, C. M. (2004). Friendship mainte- nance: An analysis of individual and dyad behaviors. Journal of Social and Clinical Psychology, 23, 413– 441. http://dx.doi.org/10.1521/jscp.23 .3.413.35460

Owens, L., Shute, R., & Slee, P. (2000). ‘I’m in and you’re out . . .’. Psychology Evolution & Gender, 2, 19 – 46. http://dx.doi.org/10.1080/ 14616660050082906

Palombit, R. A., Cheney, D. L., & Seyfarth, R. M. (2001). Female–female competition for male ‘friends’ in wild chacma baboons (Papio cyno- cephalus ursinus). Animal Behaviour, 61, 1159 –1171. http://dx.doi.org/ 10.1006/anbe.2000.1690

Palombit, R., Cheney, D., Fischer, J., Johnson, S., Rendall, D., Seyfarth, R., & Silk, J. (2000). Male infanticide and defense of infants in chacma baboons. In C. van Schaik & C. Janson (Eds.), Infanticide by males and its implications (pp. 123–152). Cambridge, NY: Cambridge University Press. http://dx.doi.org/10.1017/CBO9780511542312.008

Palombit, R. A., Seyfarth, R. M., & Cheney, D. L. (1997). The adaptive value of ‘friendships’ to female baboons: experimental and observa- tional evidence. Animal Behaviour, 54, 599 – 614.

Panksepp, J. (2010). The evolutionary sources of jealousy: Cross-species approaches to fundamental issues. In S. L. Hart & M. Lagerstee (Eds.), Handbook of jealousy: Theories, principles, and multidisciplinary ap-

proaches (pp. 101–120). New York, NY: Wiley-Blackwell. http://dx.doi .org/10.1002/9781444323542.ch6

Parker, J. G., Campbell, C., Kollat, S., & Lucas, A. (2008). The role of contingent self-esteem in adolescents’ jealous and aggressive responses to friendship interference. Paper presented at the annual meetings of the International Association of Relationship Researchers, Providence, RI.

Parker, J. G., Ebrahimi, N., & Libber, K. (2005, April). Body image insecurity and rival’s physical attractiveness in adolescents’ jealousy surrounding friends and romantic partners. Paper presented at the biennial meetings of the Society for Research in Child Development, Atlanta, GA.

Parker, J. G., & Gamm, B. K. (2003). Describing the dark side of pread- olescents’ peer experiences: Four questions (and data) on preadoles- cents’ enemies. New Directions for Child and Adolescent Development, 203, 55–72. http://dx.doi.org/10.1002/cd.89

Parker, J. G., Kruse, S. A., & Aikins, J. W. (2010). When friends have other friends: Friendship jealousy in childhood and early adolescence. In S. L. Hart & M. Legerstee (Eds.), Handbook of jealousy: Theory, research and multidisciplinary approaches (pp. 516 –546). New York, NY: Wiley-Blackwell. http://dx.doi.org/10.1002/9781444323542.ch22

Parker, J. G., Low, C. M., Walker, A. R., & Gamm, B. K. (2005). Friendship jealousy in young adolescents: Individual differences and links to sex, self-esteem, aggression, and social adjustment. Develop- mental Psychology, 41, 235–250. http://dx.doi.org/10.1037/0012-1649 .41.1.235

Parker, J. G., Ramich, C., & Roth, M. (2009). Third party peers as interlopers: Role of outside peer interference in preadolescents’ feelings and thoughts surrounding friendship disappointments. Unpublished manuscript.

Parker, J. G., & Wargo Aikins, J. (2009). Understanding the antecedents, relationship implications, and perpetuation of the disposition to friend- ship jealousy at the transition to junior high school. Unpublished man- uscript.

Parrott, W. G., & Smith, R. H. (1993). Distinguishing the experiences of envy and jealousy. Journal of Personality and Social Psychology, 64, 906 –920. http://dx.doi.org/10.1037/0022-3514.64.6.906

Perlman, D., & Peplau, L. A. (1981). Toward a social psychology of loneliness. Personal Relationships, 3, 31–56.

Perlman, D., Stevens, N., & Carcedo, R. J. (2014). Friendship. In J. Simpson & J. Dovidio (Vol. Eds.), P. Shaver & M. Mikulincer (Series Ed.), APA Handbook of personality and social psychology: Vol. 2. Interpersonal relations and group processes (pp. 463– 493). Washing- ton, DC: American Psychological Association.

Pietrzak, R. H., Laird, J. D., Stevens, D. A., & Thompson, N. S. (2002). Sex differences in human jealousy: A coordinated study of forced- choice, continuous rating-scale, and physiological responses on the same subjects. Evolution and Human Behavior, 23, 83–94. http://dx.doi.org/ 10.1016/S1090-5138(01)00078-2

Pillsworth, E. G., & Haselton, M. G. (2006). Women’s sexual strategies: The evolution of long-term bonds and extrapair sex. Annual Review of Sex Research, 17, 59 –100.

Plickert, G., Cote, R. R., & Wellman, B. (2007). It’s not who you know, it’s how you know them: Who exchanges what with whom? Social Net- works, 29, 405– 429. http://dx.doi.org/10.1016/j.socnet.2007.01.007

Reina, R. E. (1959). Two patterns of friendship in a Guatemalan commu- nity. American Anthropologist, 61, 44 –50. http://dx.doi.org/10.1525/aa .1959.61.1.02a00070

Roberts, G. (2005). Cooperation through interdependence. Animal Behav- iour, 70, 901–908. http://dx.doi.org/10.1016/j.anbehav.2005.02.006

Roberts, S. G., & Dunbar, R. I. (2011). Communication in social networks: Effects of kinship, network size, and emotional closeness. Personal Relationships, 18, 439 – 452. http://dx.doi.org/10.1111/j.1475-6811 .2010.01310.x

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1008 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

Roberts, S. G., Dunbar, R. I., Pollet, T. V., & Kuppens, T. (2009). Exploring variation in active network size: Constraints and ego charac- teristics. Social Networks, 31, 138 –146. http://dx.doi.org/10.1016/j .socnet.2008.12.002

Rose, S. M. (1984). How friendships end: Patterns among young adults. Journal of Social and Personal Relationships, 1, 267–277. http://dx.doi .org/10.1177/0265407584013001

Rose, S. M., & Serafica, F. C. (1986). Keeping and ending casual, close and best friendships. Journal of Social and Personal Relationships, 3, 275–288. http://dx.doi.org/10.1177/0265407586033002

Rosenfeld, L. (2004, September 1). Our mutual friend: How to steal friends and influence people. New York Magazine. https://nymag.com/nymetro/ shopping/features/9748/

Roth, M. A., & Parker, J. G. (2001). Affective and behavioral responses to friends who neglect their friends for dating partners: Influences of gender, jealousy and perspective. Journal of Adolescence, 24, 281–296. http://dx.doi.org/10.1006/jado.2001.0408

Rusbult, C. E., Olsen, N., Davis, J. L., & Hannon, P. A. (2004). Commit- ment and relationship maintenance mechanisms. In H. T. Reis & C. E. Rusbult (Eds.), Close relationships: Key readings (pp. 287–303). Phil- adelphia, PA: Taylor & Francis.

Sagarin, B. J. (2005). Reconsidering evolved sex differences in jealousy: Comment on Harris (2003). Personality and Social Psychology Review, 9, 62–75. http://dx.doi.org/10.1207/s15327957pspr0901_5

Salovey, P. (1991). The psychology of jealousy and envy. New York, NY: Guilford Press.

Scelza, B. A. (2014). Jealousy in a small-scale, natural fertility population: The roles of paternity, investment and love in jealous response. Evolu- tion and Human Behavior, 35, 103–108. http://dx.doi.org/10.1016/j .evolhumbehav.2013.11.003

Scelza, B. A., Prall, S. P., Blumenfield, T., Crittenden, A. N., Gurven, M., Kline, M., . . . Shenk, M. K. (2019). Patterns of paternal investment predict cross-cultural variation in jealous response. Nature Human Be- haviour, 4, 20 –26.

Schino, G. (2001). Grooming, competition and social rank among female primates: A meta-analysis. Animal Behaviour, 62, 265–271. http://dx.doi .org/10.1006/anbe.2001.1750

Schino, G. (2007). Grooming and agonistic support: A meta-analysis of primate reciprocal altruism. Behavioral Ecology, 18, 115–120. http://dx .doi.org/10.1093/beheco/arl045

Schmitt, D. P. (2010). Human mate choice. In M. P. Muehlenbein (Ed.), Human evolutionary biology (pp. 295–308). New York, NY: Cambridge University Press. http://dx.doi.org/10.1017/CBO9780511781193.022

Schmitt, D. P., & Pilcher, J. J. (2004). Evaluating evidence of psycholog- ical adaptation: How do we know one when we see one? Psychological Science, 15, 643– 649. http://dx.doi.org/10.1111/j.0956-7976.2004 .00734.x

Schug, J., Yuki, M., Horikawa, H., & Takemura, K. (2009). Similarity attraction and actually selecting similar others: How cross-societal dif- ferences in relational mobility affect interpersonal similarity in Japan and the USA. Asian Journal of Social Psychology, 12, 95–103. http:// dx.doi.org/10.1111/j.1467-839X.2009.01277.x

Schützwohl, A., Joshi, N., & Abdur-Razak, F. (2019). Competitor dero- gation in romantic jealousy and friendship rivalry. Evolutionary Behav- ioral Sciences. Advance online publication. http://dx.doi.org/10.1037/ ebs0000190

Selman, R. L. (1980). The growth of interpersonal understanding: Devel- opmental and clinical analyses. New York, NY: Academic Press.

Selman, R. L., & Schultz, L. H. (1990). Making a friend in youth: Developmental theory and pair therapy. Chicago, IL: University of Chicago Press.

Sell, A., Sznycer, D., Al-Shawaf, L., Lim, J., Krauss, A., Feldman, A., . . . Tooby, J. (2017). The grammar of anger: Mapping the computational architecture of a recalibrational emotion. Cognition, 168, 110 –128.

Sell, A., Tooby, J., & Cosmides, L. (2009). Formidability and the logic of human anger. Proceedings of the National Academy of Sciences, 106, 15073–15078.

Seyfarth, R. M. (1977). A model of social grooming among adult female monkeys. Journal of Theoretical Biology, 65, 671– 698. http://dx.doi .org/10.1016/0022-5193(77)90015-7

Seyfarth, R. M., & Cheney, D. L. (2012). The evolutionary origins of friendship. Annual Review of Psychology, 63, 153–177. http://dx.doi .org/10.1146/annurev-psych-120710-100337

Shackelford, T. K., Goetz, A. T., & Buss, D. M. (2005). Mate retention in marriage: Further evidence of the reliability of the Mate Retention Inventory. Personality and Individual Differences, 39, 415– 425. http:// dx.doi.org/10.1016/j.paid.2005.01.018

Sharpsteen, D. J. (1991). The organization of jealousy knowledge: Roman- tic jealousy as a blended emotion. In P. Salovey (Ed.), The psychology of jealousy and envy (pp. 31–51). New York, NY: Guilford Press.

Sharpsteen, D. J., & Kirkpatrick, L. A. (1997). Romantic jealousy and adult romantic attachment. Journal of Personality and Social Psychology, 72, 627– 640. http://dx.doi.org/10.1037/0022-3514.72.3.627

Shaw, A., Choshen-Hillel, S., & Caruso, E. M. (2018). Being biased against friends to appear unbiased. Journal of Experimental Social Psychology, 78, 104 –115. http://dx.doi.org/10.1016/j.jesp.2018.05.009

Shaw, A., DeScioli, P., Barakzai, A., & Kurzban, R. (2017). Whoever is not with me is against me: The costs of neutrality among friends. Journal of Experimental Social Psychology, 71, 96 –104. http://dx.doi.org/10 .1016/j.jesp.2017.03.002

Silk, J. (2002). Using the ‘F’-word in primatology. Behaviour, 139, 421– 446. http://dx.doi.org/10.1163/156853902760102735

Silk, J. (2003). Cooperation without counting: The puzzle of friendship. In P. Hammerstein (Ed.), Genetic and cultural evolution of cooperation (pp. 37–54). Cambridge, MA: MIT Press.

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., . . . Cheney, D. L. (2009). The benefits of social capital: Close social bonds among female baboons enhance offspring survival. Proceedings of the Royal Society of London B: Biological Sciences, 276, 3099 –3104.

Silk, J. B., Beehner, J. C., Bergman, T. J., Crockford, C., Engh, A. L., Moscovice, L. R., . . . Cheney, D. L. (2010). Strong and consistent social bonds enhance the longevity of female baboons. Current Biology, 20, 1359 –1361. http://dx.doi.org/10.1016/j.cub.2010.05.067

Silverman, W. K., La Greca, A. M., & Wasserstein, S. (1995). What do children worry about? Worries and their relation to anxiety. Child Development, 66, 671– 686. http://dx.doi.org/10.2307/1131942

Simpson, J. A., & Gangestad, S. W. (1992). Sociosexuality and romantic partner choice. Journal of Personality, 60, 31–51. http://dx.doi.org/10 .1111/j.1467-6494.1992.tb00264.x

Smith, K. P., & Christakis, N. A. (2008). Social networks and health. Annual Review of Sociology, 34, 405– 429. http://dx.doi.org/10.1146/ annurev.soc.34.040507.134601

Sprecher, S., & Regan, P. C. (2002). Liking some things (in some people) more than others: Partner preferences in romantic relationships and friendships. Journal of Social and Personal Relationships, 19, 463– 481. http://dx.doi.org/10.1177/0265407502019004048

Steegen, S., Tuerlinckx, F., Gelman, A., & Vanpaemel, W. (2016). In- creasing transparency through a multiverse analysis. Perspectives on Psychological Science, 11, 702–712. http://dx.doi.org/10.1177/ 1745691616658637

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 Anthro- pology, 123, 371–389. http://dx.doi.org/10.1002/ajpa.10325

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1009FRIENDSHIP JEALOUSY

Symons, D. (1979). The evolution of human sexuality. New York, NY: Oxford University Press.

Sznycer, D., Al-Shawaf, L., Bereby-Meyer, Y., Curry, O. S., De Smet, D., Ermer, E., . . . McClung, J. (2017). Cross-cultural regularities in the cognitive architecture of pride. Proceedings of the National Academy of Sciences, 114, 1874 –1879.

Sznycer, D., Tooby, J., Cosmides, L., Porat, R., Shalvi, S., & Halperin, E. (2016). Shame closely tracks the threat of devaluation by others, even across cultures. Proceedings of the National Academy of Sciences of the United States of America, 113, 2625–2630. http://dx.doi.org/10.1073/ pnas.1514699113

Sznycer, D., Xygalatas, D., Agey, E., Alami, S., An, X. F., Ananyeva, K. I., . . . Tooby, J. (2018). Cross-cultural invariances in the architecture of shame. Proceedings of the National Academy of Sciences of the United States of America, 115, 9702–9707. http://dx.doi.org/10.1073/pnas .1805016115

Sznycer, D., Xygalatas, D., Alami, S., An, X. F., Ananyeva, K. I., Fuku- shima, S., . . . Tooby, J. (2018). Invariances in the architecture of pride across small-scale societies. Proceedings of the National Academy of Sciences of the United States of America, 115, 8322– 8327. http://dx.doi .org/10.1073/pnas.1808418115

Takahashi, H., Matsuura, M., Yahata, N., Koeda, M., Suhara, T., & Okubo, Y. (2006). Men and women show distinct brain activations during imagery of sexual and emotional infidelity. NeuroImage, 32, 1299 – 1307. http://dx.doi.org/10.1016/j.neuroimage.2006.05.049

Tannen, D. (2017). You’re the only one I can tell: Inside the language of women’s friendships. New York, NY: Ballantine Books.

Tinbergen, N. (1963). On the aims and methods of ethology. Zeitschrift für Tierpsychologie, 20, 410 – 433. http://dx.doi.org/10.1111/j.1439-0310 .1963.tb01161.x

Tooby, J., & Cosmides, L. (1996). Friendship and the banker’s paradox: Other pathways to the evolution of adaptations for altruism. Proceedings of the British Academy, 88, 119 –144.

Trivers, R. L. (1971). The evolution of reciprocal altruism. The Quarterly Review of Biology, 46, 35–57. http://dx.doi.org/10.1086/406755

Weingrill, T., Lycett, J. E., & Henzi, S. P. (2000). Consortship and mating success in chacma baboons (Papio cynocephalus ursinus). Ethology, 106, 1033–1044. http://dx.doi.org/10.1046/j.1439-0310.2000.00 616.x

Welling, L. L., Puts, D. A., Roberts, S. C., Little, A. C., & Burriss, R. P. (2012). Hormonal contraceptive use and mate retention behavior in women and their male partners. Hormones and Behavior, 61, 114 – 120.

Wellman, B., Wong, R. Y. L., Tindall, D., & Nazer, N. (1997). A decade of network change: Turnover, persistence and stability in personal com- munities. Social Networks, 19, 27–50. http://dx.doi.org/10.1016/S0378- 8733(96)00289-4

Williams, G. C. (1966). Adaptation and natural selection. Princeton, NJ: Princeton University Press.

Wilson, T. D., & Gilbert, D. T. (2005). Affective forecasting: Knowing what to want. Current Directions in Psychological Science, 14, 131– 134. http://dx.doi.org/10.1111/j.0963-7214.2005.00355.x

Yalom, M., & Brown, T. D. (2015). The social sex: A history of female friendship. New York, NY: HarperCollins.

Zammuner, V. L. (1996). Felt emotions, and verbally communicated emo- tions: The case of pride. European Journal of Social Psychology, 26, 233–245.

Zhou, W. X., Sornette, D., Hill, R. A., & Dunbar, R. I. (2005). Discrete hierarchical organization of social group sizes. Proceedings of the Royal Society of London, Series B: Biological Sciences, 272, 439 – 444. http:// dx.doi.org/10.1098/rspb.2004.2970

Appendix A

Friend Guarding Measures Used (Data Shown for Best Friend � New Same-Sex Friend) in Study 3a (Overall � � .97)

Vigilance � � .77 Check up on your best friend, to see if they’re with this person or not Snoop through your best friend’s social media to see if they’re hanging out with this person

Separation � � .86 Stay close to your best friend when you’re at parties or in places with this other person Try to get your best friend to avoid places where this other person will be Try to keep your best friend from chatting with this other person Try not to let your best friend hang out with this person when you’re not there

Monopolization � � .85 Spend all your free time with your best friend so they can’t become closer with this other person Monopolize your best friend’s time

Induce jealousy � � .87 Develop and show off other new friendships you made Show interest in becoming best friends with other people to make your best friend jealous Try to make your best friend jealous of your friendships with other people

(Appendices continue)

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1010 KREMS, WILLIAMS, AKTIPIS, AND KENRICK

Appendix A (continued)

Punish/threaten friend � � .81 Become angry when your best friend hangs out too much with this other person Threaten to end your best friendship Yell at your best friend for becoming close with this other person Act overtly jealous when your best friend hangs out with this person

Emotional manipulation � � .75 Tell your best friend how dependent you are on them Make your best friend feel guilty for becoming close with this person

Threaten to harm yourself Act sad when your best friend talks too much about or hangs out too much with this other person

Derogate rival � � .91 Say negative things about this other person’s appearance to your best friend Start or repeat negative gossip about this other person to your best friend Point out this other person’s flaws to your best friend Talk about this other person’s promiscuousness to your best friend Talk to your best friend about how manipulative and untrustworthy this other person is Talk to your best friend about a time when this other person was mean to you or to a mutual friend

Self/commitment enhancement � � .83 Show your best friend how committed you are to them Enhance your own physical appearance Go out of your way to be nice to your best friend Emphasize your love and caring toward your best friend Give in to your best friend’s every whim

Possession signals � � .93 Introduce your best friend as “your best friend” to other people Tell everyone how close you and your best friend are Mention to this other person how close you and your best friend are Show off how close you are with your best friend when other people are around (e.g., by making inside jokes) Show off your commitment to your best friend (e.g., getting best friend jewelry, matching tattoos, and matching jerseys) Show off how close you and your best friend are on social media (e.g., posting pictures of you two together)

Derogation of best friend Tell this other person negative things about your best friend so that they don’t pursue a friendship with them Indirect aggression toward rival � � .84

Stare coldly at this other person Exclude this other person from social gatherings Get your other friends to be mean to this other person

Direct aggression toward rival � � .77 Be mean to this other person Confront this other person for trying to steal your best friend Hit this other person Vandalize the property of this other person

Appendix B

Scenarios from Study 7

Imagine that [name of best friend] and another man (woman), [name of best friend]’s new close friend, have started to really enjoy one another’s company. You didn’t know this new man very well before he (she) became friends with [name of best friend] just recently, but now [name of best friend] and he (she) have clearly become close. [Name of best friend] has introduced this new person into your friend group, as well.

Low time threat: You notice that [name of best friend] and this new man (woman) don’t hang out together a lot; they aren’t spending lots of time together—they don’t have lunch together or hang out together much at all.

High time threat: You notice that [name of best friend] and this new man (woman) are hanging out together a lot; they are really spending lots of time together— having lunch together and hang- ing out together a lot.

Low replacement threat: A few days ago, you found out that your best friend has received a much-wanted invitation to an upcoming birthday party–it’s a party for another guy (girl) you really like, but haven’t had time to get to know very well. This party is going to be a swanky, exclusive dinner party. Although you’re not that close with the person having the party, your best friend is. And a few of your other, mutual friends have also been invited.

(Appendices continue)

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1011FRIENDSHIP JEALOUSY

It’s a dinner party at a small new restaurant, so space is limited. But your best friend is allowed to take one friend to the party with him (her). You really want to go! You also know that [name of best friend] ’s new friend already asked [name of best friend] if he (she) would take him (her).

Both you and [name of best friend] ’s new close friend really want to go, so your best friend is going to have to choose which one of you to take to the party.

Today, [name of best friend] tells you that he has decided to take you to the party instead of his (her) new friend.

High replacement threat: A few days ago, you found out that your best friend has received a much-wanted invitation to an upcoming birthday party–it’s a party for another guy (girl) you really like, but haven’t had time to get to know very well. This party is going to be a swanky, exclusive dinner party. Although you’re not that close with the person having the party, your best

friend is. And a few of your other, mutual friends have also been invited.

It’s a dinner party at a small new restaurant, so space is limited. But your best friend is allowed to take one friend to the party with him (her). You really want to go! You also know that [name of best friend] ’s new friend already asked [name of best friend] if he (she) would take him (her).

Both you and [name of best friend] ’s new close friend really want to go, so your best friend is going to have to choose which one of you to take to the party.

Today, [name of best friend] tells you that he has decided to take his (her) new friend to the party instead of you.

Received August 8, 2019 Revision received May 18, 2020

Accepted May 20, 2020 �

T hi

s do

cu m

en t

is co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

ti on

or on

e of

it s

al li

ed pu

bl is

he rs

. T

hi s

ar ti

cl e

is in

te nd

ed so

le ly

fo r

th e

pe rs

on al

us e

of th

e in

di vi

du al

us er

an d

is no

t to

be di

ss em

in at

ed br

oa dl

y.

1012 KREMS, WILLIAMS, AKTIPIS, AND KENRICK