Like Belgian Chocolate for the Universal Mind. Interpersonal and Media Gossip from an Evolutionary Perspective. (Charlotte De Backer)

Thinking in terms of Pleistocene adaptation suggests that we should not find gossip about, for example, how soundly so-and-so is sleeping – a fact with relatively little fitness-relevance – nearly as interesting as gossip about with whom so-and-so is sleeping.” (Barkow, 1992: 629).

“While not much has been empirically established about gossiping as a personality or as a ‘biological’ disposition, there have been some wildly fanciful speculations.” (Rosnow & Fine, 1976). In this chapter I want to outline a theoretical framework that demonstrates how gossip can be regarded as a biological disposition. Gossip is a Human Universal (Brown 1991), which indicates that it is not a purely cultural product. If all people have a tendency to exchange gossip, it makes one wonder which possible adaptive functions can be attributed to this human trait.

Focusing on the adaptive problems of survival, mating and group living that our ancestors faced in evolutionary history, I will now search for the possible adaptive function(s) of gossip. I will discuss multiple functions, because the scope of this chapter is to show that the term ‘gossip’ can better be subdivided in more refined and better-defined categories.

Gossip, Barkow (1989, 1992) says, concerns the exchange of information about other people. As I outlined in chapter 1, much of our daily talk falls under this definition, or as Barkow (1992: 628) says:

“The short answer to ‘which individuals’ [we gossip about] is relatives, rivals, mates, offspring, partners in social exchange, and the very-high ranking. The short answer to ‘what kinds of information’ [we gossip about] is relative standing and anything likely to affect it, control over resources, sexual activities, births and deaths, current alliances/friendships and political involvements, health, and reputation about reliability as a partner in social exchange.” (Barkow, 1992: 629).

This evolutionary perspective on gossip, on which Barkow focuses, embodies a broad range of topics, that each could solve specific, but very different adaptive problems. Looking for the potential adaptive value of gossip, I should classify gossip in different, smaller sub categories. As will become clear in this chapter, I suggest that gossip could solve many problems our ancestors faced. Of course, not one psychological mechanism can be able to solve many different specific problems. It seems more plausible that different specialized psychological mechanisms might have been able to solve the different specific problems I will discuss.

I will now start to outline some of the adaptive problems our ancestors faced, and consequently show how specific kinds of gossip might have been suitable solutions. The goal of this chapter is to show how we can divide gossip as a noun into different sub categories, based on the different functions of each specific kind of gossip. At this point, I do not focus on gossip as a verb. I have already discussed the uses and effects of gossip in chapter 2, and will refer to many of the results from that chapter. In the next chapter, I will also focus again on gossip as an act. I will then discuss how the different sub categories of gossip as a noun can be translated into behavioral models that explain, in evolutionary terms, when it is beneficial to receive and share such specific types of information.

I divide this chapter into four main parts. First, I outline the adaptive problems of learning and suggest how gossip can solve these problems. I then switch to problems of human mating, and the potential problem-solving of gossip in these interactions. Third, I discuss problems of group living, and the potential functions of gossip in this domain. To conclude this chapter I give an overview of how gossip can be classified based its different functions. These smaller sub categories can more easily be translated to operational definitions, which can be used for empirical research.

Buss (1999: 67-96) lists ‘food acquisition and selection’, ‘finding a place to live’ and ‘combating predators and other environmental dangers’ as human survival problems, or what Darwin called the ‘hostile forces of nature’. Our ancestors faced problems of finding food, developing good hunting skills, finding a safe place to live, and combating predator risks. For instance, the emotion of ‘fear’, which we are all confronted with at times, was (and still is) the solution to some of the adaptive problems our ancestors faced, such as avoiding bites from animals, attacks from predators, diseases, etc. We still fear some animals, diseases, heights and other things that endangered the survival of our ancestors. The reason for this is very simple -- our ancestors feared dangerous situations, and therefore avoided those situations, which gave them more chances to survive than those who did not fear these dangerous situations, risked their life, and eventually died. If the behavioral strategy of our successful ancestors was passed on (genetically) to their offspring, resulting in fear of dangerous situations again, the offspring also had more chances to survive.

Knowledge of what is dangerous, and where is a safe place to live, which animals are safe to approach and which you better avoid, was acquired by our ancestors, and still is today, through experience. When talking about ecological problems of survival, we talk about experience, and therefore about learning, since: “[a]s a matter of definition, we could say that learning occurs when behavior is modified by experience.” (Gaulin & McBurney, 2004: 172).

Learning is important for humans, and as Kaplan et al (2000, 2003) have stressed, intelligence evolved because of ecological selection, and co-evolved with a longer lifespan, unique to humans (see also previous chapter). Humans have a much longer juvenile period, in which individuals are dependent on others for food acquisition, and in which they learn skills they will need in a later stage of life (Kaplan et al, 2000, 2003). Since the growth of humans into adulthood is so slow, there is a higher cost of raising offspring. Bogin and Smith (1996) think that this problem might explain menopause in women. With what is known as the ‘grandmother hypothesis’, (see also Kaplan et al 2000), they explain that women live beyond their reproductive period, to secure extra help in raising the offspring of their own offspring.

Learning is not a uniquely human trait, other animals (especially other primates) use learning techniques as well (see Byrne & Whiten, 1988 and Whiten & Byrne, 1997), but the extended juvenile period, in which learning is central, is unique to humans. In what follows I will discuss how we learn from both real world experiences and fictitious situations. For both, our learning rests on individual experiences and others’ experiences.

“To see why, remember how selection works. Selection simply retains those genes that are good at getting into the next generation. But note that genes are only tested for their ability to cope with the environments they actually encounter; they may have unpredictable effects when they encounter new environments.” (Gaulin & McBurney, 2004: 173)

The adaptations housed within our human minds are tested for adaptive or non-adaptive outcomes in a specific environment through experience. The learning processes of our ancestors (of course, only those that had a positive outcome) resulted in adaptations that still operate today. As I outlined in the previous chapter, a fear of snakes that results in avoiding them had an adaptive outcome for our ancestors, which was carried on to successive generations. Experience teaches us to avoid snakes, but experience also teaches us that we should not fear water hoses, although they resemble snakes.

We learn some things more easily than other things. The ease of learning rests on how well acquired knowledge matches a real-world situation. In evolutionary terms, the ease of learning depends on how well knowledge matches situations that occurred in the EEA. Because these learning mechanisms are designed (by selection) to solve specific problems, it seems more plausible that our minds house multiple learning mechanisms, each designed to solve a specific problem, instead of one general learning mechanism that can solve all problems (Gaulin & Mc Burney, 2004: 171-196).

In our modern world, learning mechanisms enable us to learn some things quickly, without much effort (such as language). However, as the mismatch hypothesis predicts (see previous chapter) our modern environments differ in some aspects from the EEA. We live in rapidly changing environments, and as Henrich and McElreath (2003) have suggested, learning might be most useful under these conditions. Learning is most important when it comes to situations that mismatch the EEA. To give an example, cars are a very recent addition to our environment (in evolutionary terms). We do not fear cars (as the mismatch hypothesis predicts), and individuals learn they are dangerous through experience. If we only learn that cars are dangerous through personal experiences, chances are high we would all have been hit by a car already. First-hand experience can be very costly.

The same was true for our ancestors, as Williams (1966) and Scalise-Sugiyama (1996, 2001) explained. Learning about mobile prey, predation risks and other dangerous situations through first-hand experience might have been so very costly, that it was better to get the experience from others (second-hand).

Bandura's (1969, 1977) proximate explanation for why we learn from observing others is as follows:

"Learning would be exceedingly laborious, not to mention hazardous, if people had to rely solely on the effects of their own actions to inform them what to do. Fortunately, most human behavior is learned observationally through modelling: from observing others one forms an idea of how new behaviors are performed, and on later occasions this coded information serves as a guide for action." (Bandura, 1977: 22)

Bandura (1969, 1977) explained how learning through observation is the process of recording behavioral patterns of people without communicating with them. Watching how others behave and which strategies they use can teach us which actions to copy and which to avoid. When you see someone else acting in a successful way to solve a problem, you might copy this behavior when you are faced with the same problem. However, when you see someone failing to solve a problem, chances are high you will not copy this person’s behavior in the future. Bandura’s (1969, 1977) Social Learning Theory rests on three main principles. First, people observe others; they store and organize the information and often rehearse the strategy before actually mimicking. Second, people are more likely to copy a strategy that has an outcome they value, and, third, people are more likely to copy strategies that have admired status.

An accompanying, ultimate level of explanation, comes from evolutionary psychologists, who frame the learning capacities of humans in a cultural perspective. They use the terms ‘cultural learning’ or ‘social learning’, referring to the cultural transmission/ acquisition of information involved in these actions. In line with Bandura, Scalise-Sugiyama (1996, 2001), Henrich and McElreath (2003), and Richerson and Boyd (1992) have stressed that learning through own experiences can be too costly for an individual (see above). Observation is a cheap way to get information about successful and unsuccessful behavior strategies. We exploit the expertise of others, as Russon (1997) puts it. In effect, imitation of others can be regarded as a fast and frugal heuristic, say Gigerenzer and Todd (1999: 31-32), because it enables individuals to make decisions with limited time and energy. “Humans, unlike other animals, are heavily reliant on social learning to acquire large and important portions of their behavioral repertoire.” (Henrich & McElreath, 2003: 123).

A second reason for copying others, as Bromley (1993) and Richerson and Boyd (1992) explained, is because individual learning can result in errors. We change or corroborate our own experiences, by looking at others:

“We cannot acquire all we need or want to know through direct, first-hand experience. We have to rely on others to give us the benefit of their experience, through experience, through education and social learning. Also, because individual assessments and evaluations may be wrong, they need to be corroborated by independent observers. […] To a large extent, inter-subjective agreement defines reality.” (Bromley, 1993: 75-76)

Maybe an ancestor who first encountered a lion was lucky enough that the lion was resting after having a full meal, and was not chasing this ancestor. This experience might have (falsely) lead this ancestor to believe lions are not dangerous animals, that they should not be feared. For such reasons it can still be valuable to listen to similar experiences of other people. If this ancestor later heard that one of his band members was eaten by a lion, he might change his opinion from ‘lions are safe’ to ‘lions are dangerous’. If he later hears another member of his band was eaten by a lion, this might corroborate his thoughts that ‘lions are dangerous’.

As I argued above, individual learning might be most useful to deal with modern problems that mismatch the EEA. Boyd and Richerson (1988a, 1989a, as cited in Richerson and Boyd, 1992) similarly predicted that social learning should also be more common in changing environments. By testing models that incorporate this idea, they confirmed their hypothesis (for details see Boyd & Richerson 1988a, 1989a). In line with this, Henrich and McElreath (2003) propose that conformist learning should be prevalent in contexts where little information is present (such as changing environments). Copying what the majority does, they say, might be the best solution. However, as Henrich and McElreath (2003) comment earlier in their theory; social learning alone is not sufficient:

“While social learners do very well when they are rare, they do poorly when they are common. Without any individual learners, social learners cannot track changes in the environment, and the first individual learner entering a group of social learners always has higher fitness than the others.” (Henrich & McElreath, 2003: 125)

When environments change, individual learning precedes social learning. First, individuals have to test different strategies to learn which strategy is the best option. Once this is established, others can copy this strategy and social learning becomes the best option.

When we want to learn from others’ experience, we have two options say Henrich and Gil-White (2001). First, we can directly learn from the actions of others, through observation and info copying. Second, we can learn about the actions of others transmitted to us by a third person. This second-hand strategy, of course, requires some communication.

Learning from others by observation might not be a unique human trait, because it can be found throughout the animal world (Gigerenzer & Todd, 1999). Learning from observing others is common among other non-human primates. Primate social intelligence is regarded as the precursor of human cultural evolution. The use of observational learning is most common among apes (Boyd & Silk, 505-510) but can be found in some monkeys as well. Hauser (1988), studying vervet monkeys, concluded that these monkeys use social transmission to learn about new solutions to ecological problems. “Nonetheless, it seems clear that no other primate relies on observational learning to the same extent that humans do, and their behavior is much less variable from group to group or from region to region than the behavior of humans is.” (Boyd & Silk, 2003: 510).

Tomasello, Kruger and Ratner (1993) have argued that human imitation differs significantly from other non-human primates’ and other animals’ imitation techniques. Chimpanzees might see others using a stick to gather ants from a nest, and then imitate this behavior, but will never reach the level of imitation humans exhibit. Different from other animals, we humans do not merely copy the idea or invention (use a stick), but carefully pay attention to every action of others and try to imitate this behavior as much and as detailed as possible.

“True imitation is probably not the whole story, however, at least not in the long run.” (Henrich & McElreath, 2003: 127). In addition to getting experience from observation, humans also learn by talking about the behavior strategies of others. As I explained in the previous chapter, some researchers (e.g. Scalise-Sugiyama, 1996, 2001) have suggested that language evolved to transmit information about hunting skills. Kaplan and Hill (1992) say that information exchange in the context of foraging strategies can increase experience information:

“Aché men, for example, live in social groups consisting of about 10 hunters, and at the end of day, each man usually reports to the others in considerable detail concerning every game item that he encountered that day, and the outcome of the encounter. At minimal cost, each man gains 10 times as much information about encounter rates as he actually experiences.” (Kaplan & Hill, 1992: 196)

Mobility of hunter-gatherer communities indeed increases the opportunities to exchange information about foraging techniques, marriage and rituals. Politis (1996), analyzing the mobility and settlement patterns of the Nukak in the Colombian Amazon, attributes not solely economic reasons to the moving patterns of these people, but stresses the importance of information exchange as well. Residential mobility among the Nukak promotes contact between bands and information exchange. Mobilizing to acquire information might imply some costs to the traveler, but he reaps benefits that often go beyond pure information acquisition (Kaplan & Hill, 1992):

“Information acquisition may pay off in currencies other than calories. Steward (1938) reports that young adult males in the Great Basin frequently travelled great distances to visit and share information about food distribution with their hosts. For their hosts the value of information may have been increased food yields. The value of the information for the young men might have been to obtain welcome in groups containing marriageable women.” (Kaplan & Hill, 1992: 187)

Returning to information exchange, which is central here, Tooby and Cosmides (2001) argue that our attention will be driven to both stories about potentially damaging situations, (such as seemingly safe situations with concealed threats) and to advantageous situations that we should seek out and prolong. Social communication benefited our ancestors, giving rise to a rapid flow of experience:

“In this process of incarnation, humans, beings social and communicative organisms equipped with decoupling, are no longer limited by the slow and unreliable flow of actual experience. Instead, we can immerse ourselves in the comparatively rapid flow of vicarious, orchestrated, imagined, or fictional experience. A hunter-gatherer band might contain scores or even hundreds of lifetimes’ worth of experience whose summary can be tapped into if it can be communicated. So, vicarious experience of especially interesting events, communicated from others, should be aesthetically rewarding.” (Tooby & Cosmides, 2001: 23)

Tooby and Cosmides (2001) further comment on this by stating that today we still value stories about other people, because of these experience-related benefits. Listening to what happened to others was adaptive for our ancestors to increase their experience. Those who showed interest in stories about others had an advantage on those who did not care to hear about the behaviors of others. The interested ancestors were the winners of the evolutionary arms race. The ones who cared for experiences of others became our ancestors, and today we (as their progeny) still show interest in stories about other people, because selection favored this human trait:

“Indeed, we evolved not so long ago from organisms whose sole source of (non-innate) information was the individual’s won experience. Therefore, even now our richest systems for information extraction and learning are designed to operate on our own experience. It seems therefore inevitable, now that we can receive information through communication from others, that we should still process it more deeply when we receive it in a form that resembles individual experience, even though there is no extrinsic reason why communicated information needs to be formatted in such a way. That is, we extract more information from inputs structured in such a form. What form is this? People prefer to receive information in the form of stories.” (Tooby & Cosmides, 2001: 24)

Their argument is in line with Sperber’s (1985) explanation for why we value stories over Stock Exchange information or numbers, as I outlined in the previous chapter with the little memory experiment.

What is important to keep in mind, as Richerson and Boyd (1992) stress, is that learning mechanisms are shaped by natural selection. Individuals do not copy everyone and everything; they are selective in their copying behavior. Richerson and Boyd (1992) therefore discuss biased transmission, and the simplest form is direct bias, which means that individuals need not (re)invent behavior, but evaluate different strategies (of others) and choose from among these. And, “Again, if we suppose that selection on genes is responsible for the guiding rules behind people’s choices, direct bias will tend to cause adaptive cultural variants to spread.” (Richerson & Boyd, 1992: 65).

I agree with Henrich and Gil-White (2001) that age and sex should be important criteria in selecting whom to copy or not:

“Premodern societies have a strong, sex-biased division of labour, so children should prefer somewhat older, same-sex models. This allows children to acquire gender-relevant skills and scaffold themselves to increasingly complex skills – for copying models that are too advanced will often result in failure.” (Henrich & Gil-White, 2001: 176)

These authors also argue that we are more likely to copy healthy individuals. Clear skin, bright eyes, shiny hair, and lack of injuries are all cues of a potential state of health; we should therefore be more likely to copy those who have these features (Henrich & Gil-White, 2001: 176).

Copying older people or healthier people actually indicates a similar pattern; we copy higher status individuals. As I will argue later in this chapter, status and reputations are important, and the best option is to keep your status as high as possible (see below). It has been argued (Boyd & Richerson, 1985; Henrich, et al, 2001; Henrich and Gil-White, 2001) that mimicking the behavior of higher status people is an adaptive strategy that might lead to an increase in one’s own status. But, reputations are often complex. Being a good hunter involves different skills and perhaps even keeping to a certain diet (Henrich & Gil-White, 2001). It is often difficult to determine what exactly causes someone to be skilled and prestigious. Boyd and Richerson (1985) therefore argue that evolution might have shaped human psychology for a general copying bias rather than a specific copying bias. It is less costly to mimic the general behavioral pattern of an individual than to analyze precisely which behavioral combinations lead to success. Their General Copying Bias theory explains why we sometimes do not copy specific strategies, but the overall behavior of higher status individuals. Or as Henrich and Gil-White (2001: 167) say:

“Cultural transmission is adaptive because it saves learners the costs of individual learning. Once some cultural transmission capacities exist, natural selection favors improved learning efficiencies, such as abilities to identify and preferentially copy models who are likely to possess better-than average information. […] Such preferred models may be said to have prestige with respect to their ‘clients’ (the copiers).”

In the next chapter, I focus on how we can translate all the different kinds of gossip I lay out in this chapter into behavioral models. I will also focus in more depth on the effect of the General Copying Bias. As will become clear, this GCB distorts the way we process gossip information (see chapter 5).

In addition to learning from the real life experiences of our own and others', we also learn from fictive actions of our own and others'. Scalise Sugiyama (2001) cites Humphrey (1983, as cited in Scalise Sugiyama, 2001), who stressed the importance of experience to the survival of every individual, and referred to play and dreaming to get experience. I will not discuss the role of dreaming here, but overview the importance of play, and narratives.

“Perhaps the most commonly agreed upon definition for play is that it is a form of juvenile behavior resembling functional behavior but its form is typically exaggerated, seemingly less serious, and has individual components rearranged in unusual sequences. It has also been described as a class of behaviors in which the means are more important than the ends.” (Pellegrini & Bjorklund, 2004: 24)

The tremendous benefits play offers is that no threats, or risks endanger the actors. Pellegrini and Bjorklund (2004) distinguish for object play and fantasy play. Object play refers to manipulation of objects. They say that the function of object play is related to hunting skills. It prepares young children for the use of objects in the context of hunting in later stages of life. Similarly, Steen and Owens (2000) regard chase play, performed by 3 to 5 year olds as a cognitive adaptation to solve the evolutionary (and developmental) problem of training to become a skilled hunter. This form of pretend play is a safe mode to get skilled in chasing prey. Because no real risks of getting seriously injured are present in this pretend play, children can learn difficult skills, and train themselves without being exposed to extreme costs.

Fantasy play, as Pellegrini and Bjorklund (2004) call it, refers to the player taking a stance that differs from reality. It is engaging in a fictive world, and requires higher intelligence skills and a Theory of Mind[3]. In this vein, Tooby and Cosmides (2001) say that pretend play rests on cognitive mechanisms that allow us to make representations that are decoupled from our real world (metarepresentations). The decoupling is important for the individual, to distinguish fictive situations from real situations. If this were not the case, then false information from fictive events would flow into our knowledge about the real world.

Without Theory of Mind pretend play is not possible. “Pretend play is now recognized as so fundamental an expression of the human cognitive architecture that its absence in a toddler is seen as diagnostic of a neurological impairment (autism).” ( Frith, 1989: cited in Tooby & Cosmides, 2001: 9). Taylor et al (2004) investigated this with a longitudinal study of young children, and indeed concluded that Theory of Mind is required to engage in imaginary play. Their research focused specifically on children who had imaginary friends, with whom they interacted. They noticed that this is a very common phenomenon (occurring in 65% of children), and that these imaginary interactions enable children to perform better in social contexts in later stages of their life.

In line with this thinking, Whiten and Byrne (1988b) have argued that play also occurs in other animals, but only in those who have Machiavellian intelligence skills. They refer to studies which tend to confirm that play occurs in primates (for a more detailed overview see e.g. Jolly, 1988 and Whiten and Byrne, 1988b), and say that: “One of the most plausible ‘arguments by design’ for the function of play is that it serves to gather knowledge […] to allow future flexible response to relatively novel circumstances…” (Whiten & Byrne, 1988b: 59). Recently Lewis and Barton (2004) also concluded that the occurrence of social play in non-human primates is related to an increase in the cerebellum size. The cerebellum enables cognitive and motor skills, which are both required in social play.

As an example of the ‘safety’ of pretend (or fantasy) play, Bogin & Smith (1996) explain how girls, who look adult before becoming fertile, act as if they were adult women, and get experienced, before risking to get pregnant. Boys, similarly, reach fertility earlier, but can experience adult socio-sexual feelings, such as anxiety, pleasure, pride and guilt, long before they physically appear to be adults. Their adult-like behavior is considered humorous, so that they can compete with adult men without risking their lives. For both girls and boys this adolescent phase in life offers them an opportunity to gather experience, learning how to act as adults and parents before becoming adults and parents. “From this theoretical perspective, it is profitable to view the evolution of human childhood and adolescence as adaptations for both feeding and reproduction.” (Bogin & Smith, 1996: 31.)

Play has the most value for younger children, who lack experience. Bock and Johnson (2004) analyzed the play behavior of children in Botswana, and noticed that play occurred most at a younger age and the time devoted to play decreased as age increased. This supports the idea that play prepares the younger ones for skills they will need at a later stage in their life. Also, young children seem to play with older children, to learn from their behavior, and the learning is even bi-directional. Gray and Feldman (2004), observing children playing at school noticed that young children not only play with other children from their own age, but also with adolescents. The youngsters benefit from this play because they somehow feel more secure and engage in kinds of play behavior they would not act out with same-aged children. The adolescents also learn from this playing, because this is an ideal situation to practice nurturance and leadership.

Play and fictive stories are similar, because both involve the metarepresentation of a state of affairs of other individuals (Tooby & Cosmides, 2001). Again, we need Theory of Mind to gain this form of experience. Since, processing stories involves attributing minds to fictive characters and mind-reading (Zunshine, 2003).

The major difference between the two might be that play occurs most in children and declines with age, while our interest in fictive stories remains active when growing older. Tooby and Cosmides (2001) call our interest in fiction an adult version of children’s play behavior. But children like stories too. As a matter of fact, all humans like stories about other people (Brown, 1991), and a reason for this might be because fiction reflects reality, says Carroll (1999):

“In literature the most frequent and important themes are those that concern individual identity, sexual romance, and the family. Survival is the basis of all adventure stories, and by far the largest proportion of stories that are not strictly oriented to survival are organized around the mating game, the concerns of parents for children, and family relations generally.” (Carroll, 1999: 164)

Humans value fiction (and the arts) because these stories have adaptive functions; they transmit valuable information to be used in real life (Carroll, 2002):

“Art provides an emotionally and subjectively intelligible model of reality, and it is within such models that human beings organize their complex behaviors in flexible response to contingent circumstances. The imaginative models that we construct about our experience in the world do not merely convey practical information. They direct our behavior by entering into our motivational system at its very roots--our feelings, our ideas, and our values. We use imaginative models to make sense of the world, not just to "understand" it abstractly but to feel and perceive our own place in it--to see it from the inside out. Making sense of the world in this way, through narrative and through the other arts, is both a primary psychological need and a necessary precondition for organizing our behavior in ways that satisfy all our other adaptive needs.” (Carroll, 2002: 611)

Our interest in fiction, from evolutionary perspective, is integrated into a more general framework of research on evolution and the arts (for an overview see Carroll 2002; Richardson & Steen, 2002).

The reason why fiction catches our interest is because fiction triggers our emotions (Carroll, 1999; Tooby & Cosmides, 2001):

“[…] although fiction seems to be processed as surrogate experience, some psychological subsystems reliably react to it as if it were real, while others reliably do not. In particular, fictional worlds engage emotion systems while disengaging action systems (just as dreams do). An absorbing series of fictional events will draw out of our mental mechanisms a rich array of emotional responses – the same responses that would be appropriate to those same events and persons if they were real. We care about the people involved, we identify our welfare with one or more of the characters, we may be afraid, or disgusted, or shattered, as if (in the emotional channel) those events were happening to us.” (Tooby & Cosmides, 2001: 8-9)

Fictive stories resemble situations that existed in the EEA. When hearing (or reading) about situations that resemble reality, you feel as if you are experiencing these things yourself, and that gives fiction its adaptive value. You gain experience without taking any risks and you invest very little effort (i.e. get information at very low cost).

By outlining the importance of learning to battle the hostile forces of nature, the role of communication becomes clear. Learning rests largely on the transmission of information (social or cultural learning), and this is where one of the adaptive functions of gossip plays a role. I will discuss this now in more detail.

“There is a fantastic payoff in trading hard-won knowledge with kin and friends, and language is obviously a major means of doing so.” (Pinker, 1994: 367). The core idea of my argument here is that gossip solves the adaptive problem of the costs implied with acquiring first-hand experience. Learning which strategies improve or endanger your fitness can be costly, and in the extreme can even cause your death. Acquiring information about the strategies others have tested can resolve this problem. Social learning solves the problem of individual trial and error, and it is in this view that gossip plays an important role.

Similar to what Scalise Sugiyama suggested (2001: 225): “What social or environmental factors might have triggered the emergence of storytelling? The answer to this question, I believe, is the costs involved in first-hand information acquisition.” To know what is going on around you is to your benefit, to the extreme it can sometimes save your life. “…a proper attention to gossip […] has plain survival value for a social animal.” (Derbyshire, 2001: 27). Or as Spacks (1982) reports about one of the respondents she questioned: “In the doing, in the talking, the story-telling, she encounters her own capacities and experiences her own pleasures. She also learns and relearns what she shares with the rest of humanity.” (Spacks, 1982: 32).

In the first chapter I defined gossip in the most general sense as information about the deviant or surprising traits and behaviors of third persons. The central function of gossip in resolving the problem of acquiring experience is the behavioral information transmitted through gossip. I therefore rename gossip that functions to acquire experience Strategy Learning Gossip, or abbreviated: SLG.

The subject of gossip in Strategy Learning Gossip (SLG) has a non-focused role. The identity of the gossiped subject does not really matter. He or she is a mere carrier of fitness-relevant strategies. Whether we know the gossipee or not, we will be interested in the transmitted information if we can learn from their behavior. Replacing the gossipees of SLG with any other name will not change the value of SLG. Let me explain with an example.

“Have you heard Peter was rushed to the hospital after eating at this fancy new restaurant in Ghent? He caught a severe form of food poisoning!”

Whether you know Peter or not, the SLG information remains the same. The SLG information here is “If you eat at this fancy new restaurant you risk getting seriously ill”. Whether the gossipee is Peter, Tom, Anne, or anyone else, this does not change the value of the SLG. Try to change the name in the example I gave you with another name, and it will become clear.

If you do know Peter, this form of SLG might have an extra value to you, namely that you are informed that your friend (or foe) is deadly ill. Also, if you know the gossipee of SLG, this might indicate that what happened to this person can easily happen to you as well. SLG about people from your close surrounding has relatively higher probability of applying to you as well since you live in the same environment, than SLG applying to an unknown person whose environment is unknown to you as well.

Although the gossipees of SLG are mere carriers of fitness-relevant information, certain features of the gossipee still might matter, such as their age, sex, and status. Since we do not learn from and copy just anyone, we select who we will learn from by some cues. I think the importance of certain features of the gossipee might be different for fitness-promoting and fitness-endangering SLG. To explain this, I will present you two examples, one of a fitness-promoting SLG, and one of a fitness-endangering SLG.

I start with an example of SLG about fitness-endangering situations, where personal features of the gossiped subject might not matter much. In my opinion, the recipient of this kind of information is best to avoid costly strategies that others have experienced, regardless of whether this person was older, more successful or healthier than the recipient. To illustrate, consider a five-year old boy and his father take a ride on a brand new roller coaster in a funfair. Due to a technical mistake, they are both not properly secured and fall out of their seats, getting seriously injured. When their friends spread this news (fitness-endangering SLG) around, it will not mater if they focus on the adult or the five-year old, and it will not matter if they tell this to poor people or millionaires; recipients of this information will think twice before trying out (new) roller coasters in funfairs.

When the SLG is about fitness-promoting strategies, things are different. We will not copy the beneficial strategies of others if this does not benefit ourselves, and the only way we can improve ourselves is by copying others who are healthier and wealthier. To return to my example, imagine the five-year old boy and his father visited the game-machines before going on the roller coaster ride. The boy might have run around the game-machines in this funfair to search for coins people left in these machines. When he finds plenty of these coins, this child might feel as if he is a millionaire, and other children will soon follow this successful strategy. However, you don’t see adults running around game machines and screaming for joy when they find some lost coins. Unless you are homeless, or very poor, as an adult your average status is higher than a five-year old. What is fitness-promoting for this child might not be as beneficial for you. When it comes to fitness-promoting SLG, the gossipee (carrier of fitness-relevant information) should have equal or preferable higher status than the recipient. Because older, healthier or more prestigious people are more interesting to mimic (see above) we can expect that we will exchange more fitness-promoting SLG about older, healthier and more prestigious people.

As I said in chapter 1, gossip is not about all behaviors, but rather deviant or surprising behaviors. In this context SLG translates to new strategies (which are both clearly deviant and surprising) and strategies we are familiar with, but that have a different outcome. In the latter situation, we will be surprised if the outcome of a strategy we are familiar with is different from the usual outcome. SLG is functional to learn from, and to change-or-corroborate our thoughts about our own experiences. In summary, SLG functions to teach us new strategies, and to change-or-corroborate our own experiences, with experience information from others. These two functions are similar to the functions attributed to social learning, which I described in section 2.2.2 of this chapter (see above).

What is ‘new’ to a receiver depends on the behavioral repertoire a receiver has stored in his experiences. Younger, less experienced people will benefit more by learning from SLG. Consequently, older, more experienced people might not learn as much from SLG as younger people do, but can fall back more often on the change-or-corroborate function of SLG. Extremely rare occurring situations are ideal for become topics of SLG. “Man struck by lightning, while walking in a park”, “Girl killed by flower-box that fell from window-sill”, are examples of situations that have a low frequency of occurrence. These kinds of stories spread fast, and often appear in the media as well. In chapter 7, where I focus on Media Gossip, I will come back to this. As a reminder of the boundaries of my research topic, I want to state that SLG about these events is often related to rumors and so-called Urban Legends, and is, therefore, different. As I explained in chapter 1, I restrict gossip as an overall noun (and therefore SLG) as information of which the sender has truth/false knowledge. The sender must believe the information is true, or must know the information is not true (a lie). Everything in between these certainties, which is information that we have heard about but do not know whether it is true or not, I classified as rumors, and these are not the subject of this research. Urban legends and rumors very often report about ‘strange, very uncommonly occurring events’. If the sender has truth/false knowledge, and this is the case if he or she was a witness to the event, or trusts the person he or she heard the SLG from, I will call this SLG. As soon as such SLG starts spreading around quickly, and between many people who don’t trust each other (who do not know each other very well) the SLG turns into rumors and Urban Legends.

To illustrate SLG about surprise and common situations, let me return to the example of the roller coaster incident. We learn roller coasters are dangerous, and especially people who have no own experiences with roller coasters learn a lot from this information. Still, many of us have been riding roller coasters for years without ever getting hurt. We usually consider this as a fun (and safe) activity, and do not expect different outcomes such as injuries. Hearing that people get injured distorts our general idea about roller coasters; it deviates from the ideas we have stored in our minds about funfairs and roller coasters. Hearing this kind of SLG, we might corroborate our storage of experiences with roller coasters.

Attributing this learning function to gossip, I come very close to Scalise Sugiyama’s (1996, 2001) view on why humans tell stories. Her work inspired my idea to regard gossip as a social learning heuristic. However, I do criticize that she only discusses the role of second-hand information in the context of hunting, and not in a wider social context. Learning to survive is not solely about how to secure food, but also about living with other human beings who, throughout our evolutionary past, have shaped a great deal of our behavior.

Although I have stressed the importance of learning so far in the context of natural selection, I also think that sexual and social selection shaped our learning motivations. I do not talk about sex differences (yet, see later) here, but want to stress that we not only interested in strategies that can secure our survival, but also interested in strategies that can help us gain experience in the context of mating and being a successful group-member. SLG in general should be further split into different categories, which I term as Survival SLG, Mating SLG, and Social SLG. These three forms of gossip are similar in function, but still differ enough in terms of problems they could have solved and still might solve. Let me explain each different kind in more detail.

Survival SLG concerns strategies from which we can learn how to directly improve or secure our fitness to survive. It concerns strategies that influence our health, food acquisition, living, and other basic problems of survival. Talking about how someone got ill and suffers from it is Survival SLG, from which we learn what to avoid. Talking about how someone else escaped this illness (e.g. by a healthy diet) is Survival SLG from which we learn what to mimic in the future. Talking about someone who died from eating a poisonous plant is Survival SLG from which we learn what to avoid eating in the future. Talking about someone who discovered a nutritious new food source is Survival SLG from which we learn what to eat in the future. Etc.

Survival SLG is information about life and death. It is information about the strategies of other people, that have either been successful or non-successful, and from which the receiver can learn what to avoid and what to promote to secure his or her own life.

Mating SLG is information about the successful and unsuccessful strategies others have used (tried and tested) to find and/or guard a mate, and deal with sexual rivals. I discuss the different problems of human mating more in detail in section 3 of this chapter. Gossip as an overall noun embodies different kinds of information (also called gossip) that could solve problems of human mating (see below). Central to Mating SLG is that the behavior strategies are in focus, and not as much the gossipee himself or herself, as is the case in other kinds of Mating Reputation Gossip. For instance gossiping about the successful strategy of a man to win the heart of a woman is Mating SLG from which other men can learn how to behave in the future when they want to find a mate. Talking about the unsuccessful strategy of a man to conquer the heart of a woman, is also Mating SLG, from which other men can learn how not to behave in the future.

Mating SLG is information about mating strategies. It is information about how others have behaved to find or guard a mate, in either a successful of unsuccessful way, and from which the receiver of this information can learn how to behave in the future when he or she wants to find or guard a mate. It is information to learn how to please your heart and the heart of others, and especially to learn how not go get hurt or not to hurt others.

An extra note I want to mention here concerns Mating SLG about parenting skills. Learning from others how to deal with problems concerning raising offspring has clear fitness-relevant value. It affects the inclusive fitness[4]of the parent or related family member. In the next sections, I will discuss the problems of parenting and the importance of inclusive fitness more in detail, and come back to the value of Mating SLG about parenting skills.

Group living is important because it offers co-operation possibilities and protection. I discuss the problems of group living and other functions of gossip later in this chapter. What I want to focus on here is that we learn from Social SLG how to interact with others. We learn about social norms. Recently Baumeister, Zhang and Vohs (2004) have already outlined this function of gossip. As they say:

“Anything that might ease or facilitate the process of learning these rules [guidelines for how to live in a culture] would be beneficial to the individual seeking to live in the culture. That is were gossip comes in handy. In our view, gossip is a potentially powerful and efficient means of transmitting information about the rules, and other guidelines for living in a culture. On the surface, gossip consists of stories and anecdotes about particular other people, perhaps especially ones that reflect negatively on the target. We readily concede that some of the appeal of gossip is simply learning about other people. However, we think that a second, less obvious function of gossip is to convey information about social norms and other guidelines for behavior.” (Baumeister, Zhang & Vohs, 2004: 113)

Through Social SLG we can learn how to behave in a culture. To give an example of how important this can be I refer to the anecdote John Derbyshire (2001) reported in a popular journal. He tells the story of how he used to refrain from taking part in gossip interactions at work, and how this once resulted in losing his job. The tradition at his former job, a brewery, used to be to go for a drink at lunchtime on Friday and return to work lightly intoxicated in the afternoon. The new boss, however, did not approve this behavior. Derbyshire ended up being the only one who was a little drunk after lunch, while the other colleagues had heard (through gossip) that this new boss would sanction those who drank during office hours.

Social SLG is about teaching and learning what is right and what is wrong to do in a certain social setting. It is about how to behave, and especially how not to behave. Gossip reveals the social etiquettes. Social SLG also teaches us about the social contracts we don’t know of (yet), and also reinforces the social contracts that govern a society; through Social SLG we can corroborate what we knew already.

As I will argue later in this chapter, this type of gossip is closely related to Co-operation Reputation Gossip and Calibration Reputation Gossip, which concern information about people who follow or defect social contracts or other group norms. The major difference between Social SLG and Co-operation and Calibration Reputation Gossip is that for Social SLG the focus lies on the behavioral information, while the focus of the other two are directed to the gossipee. All three different kinds of gossip often co-occur, and whether they co-occur or not depends on the relationship between the gossipers and the gossipee. If the gossipers and the gossipee do not know each other, the information will be solely Social SLG. If the gossipers know the gossipee, it will be Co-operation or Calibration Reputation Gossip as well.

Baumeister, Zhang and Vohs (2004) do not note a difference between Social SLG, Co-operation Reputation Gossip and Calibration Reputation Gossip. Still, I think it is important to do so. The example Baumeister et al (2004) give at the beginning of their article concerns a man who was betrayed in a purchase from a commercial company. This is an example of gossip information that can be classified as both Social SLG and Co-operation Reputation Gossip. We learn that it is wrong to betray others, or we can corroborate if we knew this already, and we punish the person who cheated, by giving him a bad reputation and decrease his future co-operation opportunities.

Because this classification of gossip is based purely on theoretical fundaments, and future empirical research should investigate whether extra categories are needed or not, I leave open this option by adding an ‘Other-SLG’ category. Other-SLG covers forms of SLG that cannot be classified under Survival SLG, Mating SLG, or Social SLG. Or, it can also embody gossip stories that can be classified under one or more of the three other forms of SLG, but when multiple functions are present and one of the functions is not to transmit information about survival, health or social matters.

When focusing on specific SLG topics, we can expect some male/female differences. Especially concerning Mating SLG, men and women will be eager to learn about different strategies. I will discuss this in more detail in the next section. Although I expect sex differences to occur in what is being gossiped about, at first glance you might not expect any sex differences in a tendency to exchange SLG. Getting experiences at low costs is equally important for men and women. Learning how to secure your life, how to attract mates and how to behave properly are all as important for men as they are for women. When asking men if they gossip, they often reply “No, not at all”, and if you then point to a specific gossip conversation you overheard from them, they will tell you that “This is not gossiping, but exchanging important information”. That is exactly what is going on when we exchange SLG; we exchange information that is so important that it can influence your fitness. Men most certainly exchange SLG, because it was so important for our male ancestors to secure their survival and reproduction, and it is still is helpful for modern men to improve their life skills.

But, there is a reason why I argue that women might show a greater tendency to engage in SLG than men. The reason lies in different relocation patterns in the past (as I described in chapter). Women were being relocated after marriage and forced to become the more socially active gender, able to integrate easily in new social environments. Our male ancestors traveled a lot as well, and visited other bands, but did not stay with these culturally different groups for the rest of their lives, as our female ancestors did when they married. Behaving according the group norms in order to be accepted as a group member must have been more important for our female ancestors. Of course, relocation in our modern societies will have an effect on both men and women, eliciting higher interest in SLG. However, on average I expect this effect to be largest for women, since they experienced this problem more often in our evolutionary past. If this is true, these sex differences should still reflect in current behavior, with women showing greater interest in SLG (especially Social SLG) than men. Fewer sex differences might be expected for Survival SLG and Mating SLG. Men traveled extensively in our evolutionary past, dealing with new environments and new problems of survival to which Survival SLG could have offered solutions. And, of course, mating is as important for men and women.

In summary, both men and women should show a tendency to engage in SLG, and this tendency will be bigger for women, especially concerning Social SLG. For all different kinds of SLG sex differences can be expected in specific topics gossiped about.

Beyond securing our own survival, reproduction is important in evolutionary terms. If our ancestors knew how to secure their own survival, but did not reproduce, we would not be here today. The same sex differences in what men and women find attractive in a potential mate are shown to occur cross-culturally (Buss, 1999: 99-160).

Our present human mating strategies are shaped by sexual selection forces throughout our evolutionary history (Buss, 1994: 5-6). Sexual selection leads to sex differences when considering what we find attractive in a potential mate. What we have to keep in mind, though, is that selection shaped our desires. What we actually do and whom we finally choose to be our partner might be a reflection of these inherited desires but can be different as well (Buss, 1994; Symons, 1979). A woman might not desire a dominant man to be her partner, but the social context she lives in might make her decide to marry a dominant man to offer her protection from environmental threats (Graziano, Jensen- Campbell, Todd & Finch, 1997: 147-148).

Because the literature on an evolutionary approach to human mating is enormous, and I just want to outline the basic problems men and women face in the context of human mating, to show how gossip could have solved some of these problems. I will not give a full, detailed overview of all research in this domain. My general outline is based mainly on the work of Buss (1994) and Buss and Schmitt (1993), complemented with some of the most important recent studies.

First of all, women desire multiple traits in men, and similarly men desire various traits in women. There is not one single mechanisms that shapes their desires, but rather multiple. Therefore, popular quotes such as “Women want rich men” do not cover the whole story (Graziano et al, 1997: 161-164). If it where that simple, mating would not be as problematic as it was and still is.

Second, although it has been suggested that humans evolved primarily for monogamous mating (Miller & Fishkin, 1997), I concur with the opposite hypothesis of Buss (1994) and Schmitt (Buss & Schmitt 1993) that humans have evolved for both long-term and short-term mating. Both men and women use different strategies, and favor different traits depending on their mating strategy (short-term or long-term). I will discuss both strategies for both sexes, in my following overview of adaptive problems in the context of sexual selection.

One of the central premises in Buss and Schmitt’s (1993) Sexual Strategy Theory (SST) focuses on the difference between long-term and short-term mating. Generally, men have a stronger desire for short-term mating, and women have a stronger desire for long-term mating, these authors say. Because of differences in gamete production and parental investment, men have more advantages of having multiple partners. From a biological perspective, men benefit more than women from short-term sexual strategies (see Trivers, 1978 for more details). Again, I want to stress that no moral statements can be made from this. Evolutionary psychologists never say men should have as many partners as possible. But in the perspective biological principles of human mating, those men who have multiple partners have more options for reproduction, which in biological perspective is more successful than having fewer partners and fewer reproduction options.

When saying that, in biological terms, men benefit more than women from short-term relations, this is not to say that women do not engage in short-term relations. As I will discuss later in this section, women have multiple sexual relations as well, and opt for short-term sexual relations. If women did not choose short-term relations, where would men find their ideal potential short-term mate?

When questioning men and women about their sexual behavior and desires, Schmitt, Shackelford and Buss (2001) noticed that men indeed have a stronger desire for short-term relations than women do. They also concluded that, on average, men more that women, have a stronger desire to have more sexual partners (see also below) and men, more than women, need less time before consenting to sex. Schmitt et al (2001) stress as well, though, that both men and women have adaptations for short-term mating, but that men are generally more oriented toward short-term mating than women (for more details see Schmitt, Shackelford & Buss, 2001).

The most convincing evidence so far that men have a stronger desire for sexual variety, and therefore have a stronger desire for short-term relations, comes again from a research of Schmitt (Schmitt et al, 2003). In co-operation with researchers from 10 major world regions (including North America, South America, Western Europe, Eastern Europe, Southern Europe, Middle East, Africa, Oceania, South/Southeast Asia, and East Asia) he collected data from 16,288 people. The results from this large-scale cross-cultural survey confirmed the hypothesis that men have a stronger desire for more sexual partners and short-term sexual relations, than women:

“This study provides the largest and most comprehensive test yet conducted on whether the sexes differ in the desire for sexual variety. The results are strong and conclusive—the sexes differ, and these differences appear to be universal. Men not only possess a greater desire than women do for a variety of sexual partners, men also require less time to elapse than women do before consenting to sexual intercourse, and men tend to more actively seek short-term mateships than women do.” (Schmitt et al, 2003: 101)

Highly related to the fact that men have a stronger desire for short-term relations, they also have a stronger desire for a larger number of sexual partners. I already mentioned the research of Schmitt, Shackelford and Buss (2001) that confirmed this. When asking men and women how many sexual partners they desired in the next thirty years, Buss and Schmitt (1993) as well noticed a significant sex difference in the mean scores of male and female respondents. On average women desired 5 sex partners and men 18. Miller en Fishkin (1997: 220-221) replicated this experiment. Their average for female respondents was 2.79, a little lower than Buss and Schmitt’s findings. But, their male respondents’ average outscored by far; Miller and Fishkin’s male respondents’ average desired number of sexual partners was 64.32. However, they had to admit that, when two outliers were removed from the sample, their average came close to the number 18 of Buss and Schmitt. Nonetheless, the overall pattern clearly indicates that men desire more partners than women. The research of Ellis and Symons (1990) regarding sex differences in sexual fantasies also indicated that men desire higher number of sex partners.

When men meet new women, they often get aroused. Buss (1994) talks about the Coolidge-effect, named after President Coolidge:

“The story is told that President Calvin Coolidge and the first lady were being given separate tours of newly formed government farms. Upon passing the chicken coops and noticing a rooster vigorously copulating with a hen, Mrs. Coolidge inquired about how often the rooster performed this duty. ‘Dozens of times each day,’ replied the guide. Mrs. Coolidge asked the guide to ‘please mention this fact to the president.’ When the president passed by later and was informed of the sexual vigor of the rooster, he asked, ‘Always with the same hen?’ ‘Oh, no,’ the guide replied, ‘a different one each time.’ ‘Please tell that to Mrs. Coolidge,’ said the president.” (Buss, 1994: 79-80)

But as Symons (1979) explained, what we want is not always what we can have. Men, he said, desire a large number of sexual partners, but reality constraints their actual behavior. In their search for multiple sexual partners men face adaptive problems.

One of the adaptive problems men face in their search of a potential female mate, is that her sexual intent might differ from his desires, and that she might even lie about this sexual intent. When men pursue short-term relations with women, they do not desire non-promiscuous women. Looseness and promiscuity (which are undesired in their long-term strategy, see below) are female qualities that increase men’s chances on reproductive success in the context of short-term mating (Buss & Schmitt, 1993; Buss, 1994). Women often deceive men in their sexual intent, which causes extra adaptive problems to men (see below).

Since sexual selection is about the potential of reproduction by securing offspring, men in their short-term (and long-term) strategies will care for the reproductive potential of women. For women there are two important factors: her reproductive value and her fertility. Buss and Schmitt (1993) give an example by comparing a 14-year-old girl to a 24-year-old girl. The first has higher reproductive value; the latter has the highest fertility. The future reproductive potentials of a 14-year-old girl exceed those of a 24-year-old girl, but the 24-year-old girl has more chances of being fertile already. When pursuing short-term goals men will favor the fertility of a woman, when he wants to obtain a long-term relation, he will be more interested in her reproductive value (favoring younger women to marry).

Because fertility and reproductive value are disguised in female humans, men face the problem of estimating both features correctly. Cues they use to do this are youth and health, say Buss and Schmitt (1993). I discuss these more in detail in the next section on men and long-term relations.

Last, when men pursue short-term relationships, they want to avoid commitment and investment (Buss & Schmitt, 1993). They resolve this problem by deceiving their potential partners (see below). And they have to, since the best strategy for men to win the heart of women for short-term relations is by displaying immediate investment. Women want to acquire resources from men, in the context of short-term relations, so that men cannot omit a form of investment. Other signals they display, to best show off their qualities as good short-term mates, are their dominance and physical attractiveness (Schmitt & Buss, 1996).

Although short-term strategies are most adaptive for men, this does not imply that they do not benefit from and engage in long-term relationships. When looking for a long-term partner the problems men face mainly concern the reproductive value of a potential mate, and her fidelity (Buss, 1994). Let me discuss both traits in a little more detail.

3.1.2.1 Women’s disguised reproductive value: the importance of youth and beauty

As I already said above, cues to detect the reproductive value of women are youth and beauty:

“Our ancestors had access to two types of observable evidence of a woman's health and youth: features of physical appearance, such as full lips, clear skin, smooth skin, clear eyes, lustrous hair, and a good muscle tone, and features of behavior, such as a bouncy, youthful gait, an animated facial expression, and a high energy level. These physical cues to youth and health, and hence to reproductive capacity, constitute the ingredients of male standards of female beauty.” (Buss, 1994: 53).

Research on what men and women display and ask for in personal advertisements, has shown that men prefer a younger partner, and explicitly ask for this in their personal ads (Butler-Smith, e.a., 1998; Greenless & McGrew, 1994; Hayes, 1995; Matthews, 1999; Rajecki, e.a., 1991; Rasmussen e.a., 1998; Sprecher, e.a., 1994; Wiederman, 1993; Willis & Carlson, 1993). The cross-cultural studies from Buss (1989) and Kenrick and Keefe (1992) gave similar results. Pawlowski and Dunbar (1999) also noticed that women less than men mentioned their age in personal advertisements. They disguised their age. After analyzing their profiles, these authors concluded that it is mostly women older than 35 who decline to mention their age.

However, men will not always prefer younger women; it depends on which stage of their life history they are in. If it is true that men have preferences for fertile women, then teenaged boys should prefer slightly older girls to younger ones. Kenrick, Keefe, Gabrielidis and Cornelius (1996) questioned teenagers about their sexual desires and found that teenage boys prefer partners to be slightly older, which proves this prediction.

What men regard as ‘attractive’ or beautiful in potential partners relies on some facial and bodily features. Related to the youth-aspect, neoteny of the face plays an important role in attractiveness. Neotenous features are, for instance, large eyes, high cheek bones, small nose and chin, full lips, etc. (Buss, 1994). To test this, Johnson and Franklin (1993) manipulated faces, using a computer program. Their results show that faces with these younger features are rated as more attractive than ‘older’ looking faces.

Next, concerning the face, the skin is very important, says Etcoff (1999), who wrote a book on beauty and evolutionary psychology. The skin, Etcoff argues, signals health, at least when she appears flawless and bright. A last facial feature that is rated attractive in women is facial symmetry. Again, it has been suggested that facial symmetry signals health (Shackelford & Larsen, 1999). Women with more symmetrical facial features are rated more attractive than women with less facial symmetry (Gangestad & Thornhill, 1997; Grammer & Thornhill, 1994).

When focusing on bodily features, two important cues are waist-to-hip ratio (WHR) and body-mass-index (BMI). Singh (1993, 1995) reported that men, in general, prefer slim women (low BMI) to overweight women (high BMI), and a WHR of 0.70. For women the WHR ranges from 0.67 to .080. Furnham, McLelland, and Omer (2003) tested Singh’s hypothesis, using drawings of either light or heavy women, with high or low WHR. British and Kenyan women both rated a WHR of 0.70 as most attractive, and preferred light women over heavier ones. Later studies using more realistic computer images (e.g. Streeter & McBurney, 2003) also confirm Singh’s hypotheses.

In a study, using front-view and profile pictures of women, Torte and Cornelissen (2001) noted that BMI and WHR are both important cues in attractiveness ratings. Both their male and female raters preferred the depicted (real) women with lower BMI, and lower WHR. In their conclusion, BMI was put forward as the most important predictor. This importance of BMI is what defines attractiveness for men and women. Low BMI seems to be more important for women than for men to be rated as attractive. This falls in line with the research of Maisey, Vale, Cornelissen, and Tovee (1999). They compared attractiveness ratings of pictures of men and women, and concluded that size (BMI) is most important for women, while shape is most important for men. For men, it is not the waist-to-hip ratio that determines their attractiveness, but rather waist-to-chest ratio (WTC). Maisey et al (1999) noted that men with higher BMI, but smaller WTC (indicating big chest) were still rated attractive, while for women BMI was the main criterion to rate attractiveness. Women with higher BMI are rated less attractive, even if they have the ideal WHR. The reason for this, these authors say:

“Men have evolved to find optimally attractive women of a particular BMI range that is a good predictor of optimal health and reproductive potential. It seems that the cues to male attractiveness serve a different function. For men, a body shape indicative of physical strength seems to be more important than simple body mass.” (Maisey, et al, 1999: 1500)

Next to beauty, for long-term sexual relationships men desire faithfulness, sexual loyalty and chastity of women. A key problem for men is paternal uncertainty. If their long-term partner cheats on them, they risk the cost of investing in offspring that is not theirs. Men therefore prefer faithful women over unfaithful or promiscuous women for long-term goals (Buss & Schmitt, 1993).

While short-term sexual strategies are more desired by men because of the greater biological benefits they can gain, long-term relationships are most desired for women, since the biological benefits following from this strategy are greater for them (see Trivers 1978 for more details). Women generally desire one single sexual partner to be romantically and sexually involved with (Ellis & Symons, 1979). They do not desire to maximize their number of sex partners, and tend to be choosy in picking out fewer partners. Because of their high parental investment, they especially face problems of finding a partner who is willing to commit and invest in them and their offspring.

According to Schmitt et al (2002) women reach their fertility peak in their early thirties. At that time they show the greatest willingness to commit to a long-term relationship, as these authors have shown by questioning 611 women about their sexual desires. Now which cues do women pay most attention to when choosing a potential mate?

Buss (1994, 1999: 105) summed up some traits women desire when choosing a long-term male partner. For instance, women prefer men who are able to invest in her and her offspring, and also are willing to do so (ability and willingness for investment). Women favor men with good parenting skills. Next, women desire protection; they favor men who can physically protect them. And last but not least, women prefer compatible men. I overview these cues briefly:

Cues that signal men’s ability to invest are, as Buss (1999: 105) says, good financial prospects (at least in our modern societies), social status, older age, ambition, size, and strength. Townsend and Levy (1990a,b) set up some nice experiments that clearly indicate that women prefer men with abilities to invest over men who lack these abilities. When presented pictures of low/medium/high attractive men, who had either a low/medium/high social occupation, women preferred men with higher occupational status, for a variety of different relationships (Townsend & Levy 1990a).

In a second experiment Townsend and Levy (1990b) again used low/medium/high attractive men. All of these men were then presented wearing different clothing styles. The clothing styles indicated high/medium/low status. The same man, wearing different clothes, had the most success when wearing clothes that signaled higher occupational status. Overall, Townsend and Levy concluded that attractiveness has some impact, but is ruled by occupational status. The research of Wiederman (1993) on personal advertisements indicates that men are aware of this; they, more frequently than women, state their (financial) access to resources in their ads.

Since occupational status and access to resources increase with age, women tend to prefer older men. Research has already confirmed this (see e.g. Butler-Smith, e.a., 1998; Greenless & McGrew, 1994; Hayes, 1995; Matthews, 1999; Rajecki, e.a., 1991; Rasmussen e.a., 1998; Sprecher, e.a., 1994; Wiederman, 1993; Willis & Carlson, 1993). And, even cross-cultural research reports the same conclusion; in his cross-cultural survey Buss (1994) noted that women one average desire their partner to be 4 years older. Similarly, Kenrick and Keefe (1992) noted that women in general tend to marry men who are 5 years older.

When looking for cues of protective abilities of a potential partner, women prefer tall men (Dunbar, 1995a; Muir, 2000), with a high shoulder-to-hip ratio (Hughes & Gallup, 2003). When placing a personal advertisement men, more than women, mention their height (Epel et al, 1996; Willis & Carlson 1993). Muscled men score higher with women, as is confirmed by the research of Faurie, Pontier and Raymond (2004) who questioned students and concluded that male students who were registered for sport’s courses had more sexual partners.

Lastly, as Buss (1994, 1999) says, women want compatible men, man who has characteristics they value, a man that can compliment them. Again, proof of this stems from research on personal advertisements. Women, more than men, ask for specific character traits they value in potential long-term partners (Deaux & Hanna, 1984; Willis & Carlson, 1993).

Though women, in general, prefer long-term over short-term sexual relations, they also engage in short-term affairs. Both single women and women in a relationship seek short-term partners, but their goals might differ.

The benefits women can get from engaging in short term sexual relations are acquiring resources (immediate investment), protection of men, and potentially assessing long-term mates (Buss & Schmitt, 1993). The cost women pay for short-term relationships is risking their reputation as a long-term mate. For long-term relationships, men do not favor promiscuous women (see above). However, if high-quality mates are rare (that is to say, if a woman is surrounded by poor men) she might be better off to engage in several short-time affairs to benefit from multiple investments, rather than relying on one source of investment. This is what Cashdan (1993) pointed out; when women are surrounded with low-quality males (e.g. prison, drug addicts, etc.) they try to form alliances with multiple males.

Single women, pursuing short-term relations, date both single men and men who are already engaged in a long-term relationship (poaching, see Schmitt & Buss, 2001), with the expected benefits of acquiring investment, protection and potential long-term relations.

Even though they are engaged in a long-term relationship, women might still seek potential extra mates. Female cuckoldry has occurrence rates of 10% in some cultures (see Baker and Bellis, 1995). Female cuckoldry seems to correlate with shifts in her menstrual cycle. Women seem to desire extra-marital partners most when they are in the most fertile period of their menstrual cycle (Gangestad et al, 2002). This desire also translates in real behavior, such as more frequent visits to single bars of women in their fertile period (Grammer et al, 1997). However, because of the implied costs of potentially loosing their long-term partner, women might also refrain from this risk-taking behavior. In similar research, Grammer et al (1997), Chavanne and Gallup (1998) noticed that women decreased risk-taking behaviors when they were in their most fertile period of their menstrual cycle (all reported researchers concern women who do not take birth control pills). However, again as Symons (1979) stresses, sexual selection shaped sex differences in desire, and not necessarily in behavior.

The benefits women who are already engaged in a long-term relationship get form these short-term affairs are explained by the ‘sexy-son hypothesis’. This is, women who seek extramarital affairs seem to favor physically attractive men, who signal health and good genes (Buss & Schmitt, 1993). As a proof of this, Women in a long-term relationship and cheating on their husbands prefer men with facial symmetry (Gangestad & Thornhill, 1997). Facial symmetry signals health and good gene quality. Men with more symmetrical faces (less Fluctuating Asymmetry or FA), have more sexual partners (Thornhill & Gangstead, 1994), are rated as more attractive, dominant, sexy, and healthy (Grammer & Thornhill, 1994), and tend to have more masculine features (Gangestad & Thornhill, 2003).

In their cuckoldries, women seek good genes to secure the fitness of their offspring. They do not care about the willingness to invest of men with whom they cheat their long-term partner, since they already have source of investment (the long-term partner) (Buss, 1994).

3.2 Conflict between the sexes: deceiving potential partners as conflict strategy

In their search for a good potential mate, either for short-term or long-term commitment, both men and women are confronted with deceitful actions of the opposite sex. Conflicts in the context of mating not only occur within the same sex (as I will discuss more in detail below), but also between the sexes.

The so-called Error Management Theory (EMT) explains men’s overestimation of women’s sexual intents, and women’s underestimation of men’s willingness to invest. When individuals of the same sex are in conflict over the access to another-sex individual, they will sometimes use deceptive strategies in their competition. Haselton and Buss (2000; also Buss, 1994) have shown that men often show more willingness to commit than they actually desire to. They deceive women, to win their heart, by making them (falsely) believe they are willing to commit for a long-term relationship. Women do not display this commitment deception as often as men do, but outscore men, on average, with their sexual deception. They make men (falsely) believe they are interested in sexual intercourse. Reasons for this, according to Buss (1994), are that they can benefit from resource acquisition through their sexual deception, or they can increase their desirability.

Two problems result from these deceitful strategies. First, men are confronted with the problem of correctly estimating women’s willingness for sexual intercourse. Second, women are confronted with the problem of correctly estimating men’s willingness to invest and commit (Haselton, 2003). A recent study of Haselton, Buss, Oubaid, and Angleitner (in press) shows that negative emotional responses towards deceitful strategies might solve these problems. They found sex differences in emotional responses to deceitful strategies used by men and women in the mating context, which were in line with predictions derived from the literature on deceit in the context of mating. Women feel emotionally distresses if they discover men deceived them by presenting themselves as more able and willing to invest than they really were. Men, on the other hand feel more emotionally distressed if women deceived them sexually, by showing more willingness than they were actually willing to give.

But as I mentioned, conflicts in the context of mating, mainly occur within the sexes.

In general, if males compete with each other it is because of status achievement conflicts. Wilson and Daly (1985) say that age and social status are the main predictors of intra-sexual male aggression. Younger men compete with older men over status positions (young-male-syndrome) and, regardless of age, men compete with each other for status ranking positions. In general, Wilson and Daly (1985) say that poor men commit more murders than wealthier men. I will discuss more about these conflict situations in the next section about group-living problems. Here I focus on men competing for other women.

Men do indeed compete for women as well, and according to Schmitt and Buss (1996), when men compete for access to the same woman for short-term strategy, they display their immediate investment and disgrace each other on immediate investment abilities. They also disgrace competitors by signaling that their rivals are weak, lack the dominance that protects women, and is desired by women for short-term partners. Third and lastly, denigrating male competitors on their physical appearance seems to work as well. Schmitt and Buss (1996) concluded all this from their research and also noticed that a different competitive strategy is used for long-term relationship goals. Men competing for a long-term partner disgrace each other by signaling that his rival does not have future potential resources, and scores low on kindness and willingness to invest.

In general, when females are in conflict, it is mostly with other females, and the reason of the conflict is almost always access to a man (Campbell, 1999, 2001). Schmitt and Buss (1996) say that when women compete for a long-term mate, a good strategy to rule out their competitor is by signaling that she is promiscuous, or sexually unfaithful, which increases the risk of paternal uncertainty. Men don’t want promiscuous or unreliable women for long-term relations, thus this would bring competitors in disgrace. When women look for short-term relations, on the other hand, they do best by signaling that their competitors are sexual unavailable (Schmitt & Buss, 1996).

Once the perfect mate is found, either for short-term or long-term commitment, adaptive problems do not disappear. On the contrary, even when in a relationship, both men and women still face two important problems; their mate might cheat on them, and rivals from the same sex still keep a sharp look-out. When in a relationship, both men and women are confronted with the same problem of human mating; the problem of guarding your mate.

Poaching (attracting someone who is already committed to someone else) occurs both in men and women, say Schmitt and Buss (2001). Hence, it is one of the problems both men and women in relationships are confronted with, because they should be aware that others might want to attract and steal their mate (Buss & Shackelford, 1997).

A major problem men face in the context of long-term mating is paternal uncertainty. Women’s reproductive value is concealed; we cannot see when a woman reaches her fertility peak. A man risks not being the father of his children if his wife cheats on him in her fertility period (Buss, 1994; Buss & Schmitt, 1993).

A major problem women face, when it comes to guarding their partner, is the risk of loosing investment. If their long-term partner leaves them for someone else, or invests his resources in others, they loose investment that is crucial to raise their offspring (Buss, 1994; Buss & Schmitt, 1993).

For these reasons evolutionary psychologists (e.g. Buss 1994, 1999, 2000, Symons, 1979) argue that humans experience jealousy when they think their partner is cheating on them, and that sex differences in jealousy do exist. In general, they say that men show more sexual jealousy, since their biggest problem is sexual infidelity of their partner. Women, on the other hand, experience more emotional jealousy feelings, because their problem concerns the risk of their partner to invest his resources in someone else. Studies about jealousy from an evolutionary perspective seem to confirm this (e.g. Buss, 1988; Buss & Shackelford, 1997; Buunk, et al, 1996; Dijkstra & Buunk, 2001; Pietrzak, et al, 2002; Shackelford, Buss & Bennett, 2002; Schützwohl & Koch, 2004).

To conclude my overview of adaptive problems of human mating, I should actually touch on the adaptive problems of being parents. Since, human mating often results in offspring, and this opens up a new domain of adaptive problems that both our male and female ancestors were faced with, and which both men and women still face today. However, to explain these kinds of problems, it is important to understand ‘Inclusive Fitness’. I therefore discuss these problems in section 4.2.1, after I have defined inclusive fitness in section 4.1.2.1.

Now how does gossip fit into all this? Again gossip, in general, cannot solve these problems men and women face when looking for a potential mate and guarding a mate. I again suggest that specific kinds of gossip might have solved the problems of mating our ancestors faced.

In the context of mating, gossip is an ideal tool to gather information about the status of a potential mate, to get information about possible rivals and to control your partner. Miller (2000) already stressed the importance of gossip in the context of mating. However his ideas are different from what I present here. In his opinion, gossip is a tool to show off exclusive social knowledge. I agree with Miller and will discuss this in more detail in the next chapter, where I translate the different kinds of gossip that I classify into behavioral models, explaining when it is beneficial to share such information. Miller (2000) suggests that men especially benefit from sharing gossip, because they so show off their social intelligence, which attracts potential female mates. I disagree on this. Gossip in the context of mating problems (Mating Reputation Gossip, or abbreviated Mating RG) is equally important for men and women. Both our male and female ancestors faced multiple problems in the search for and the keeping of a good mate, as I just outlined above. The tendency to share and acquire Mating RG is, in my opinion, equally significant for men and women. The topics they gossip about might differ, as I will explain later. In contrast to the above section, where I discussed male / female differences separate at the end, I will discuss expected male and female differences throughout the text in this section, because of the greater relevance to this kind of gossip.

As an important overall note, I want to stress a difference between Mating RG and Strategy Learning Gossip. SLG is information where the gossiped about behavior is in focus, and gossipees are merely carriers. For Mating Gossip and also all other kinds of Social Gossip, as I will discuss in section 4 of this chapter, this is different. For Mating Gossip and Social Gossip, the gossiped about traits and behaviors cannot be detached from the gossipee. Both kinds of gossip are about reputations, the traits and behaviors of a specific person. I therefore refer to all the kinds of gossip I will discuss from here on as forms of Reputation Gossip (abbreviated RG). For all following different kinds of gossip it is important to keep in mind that trait/behavior information is attached to a specific person. Gossipees are the focus of Reputation Gossip. I will come back to this in section 5, where I summarize everything and propose my classification system with different levels.

In general I distinguish four different kinds of gossip in the context of mating. Mates Detection Reputation Gossip, Mating Structure Reputation Gossip, Mates Control Reputation Gossip and Intrasexual Conflict Reputation Gossip. Each of these different forms of communication has specific and different functions, and occurs in conversations between men and women. Each of these different kinds of gossip can still be narrowed down into smaller units, which will be useful for empirical studies on Mating Gossip. Throughout this dissertation I will only use these three different kinds of Mating Gossip by name otherwise it may become confusing for the reader. Still, I will now discuss each of these three kinds of Mating Gossip in more detail and narrow down the definitions.

3.6.1 Mates Detection Reputation Gossip and Mating Structure Reputation Gossip

3.6.1.1 Mates Detection Reputation Gossip and Mating Structure Reputation Gossip explained

Mates Detection RG solves the problem of finding good potential mates. Mates Detection RG focuses on the reputation of men and women as sexual partners. It concerns information about their ‘mating skills’; are they good or bad partners? Both men and women face problems of finding good potential mates, and are even confronted with the additional problem of the deceitful actions of the opposite sex, as I outlined above. Mates Detection RG might have solved these problems by offering valuable information about the reputation of a potential mate and his or her real intentions.

Gossipees of Mates Detection RG are people with whom the gossipers can establish encounters. Gossipers have had past encounters with the gossipee and know this person, or do not know the gossipee yet, but a future encounter with the gossipee is possible. This is to say that the gossipee is either known directly or indirectly, through some intermediary person, so that future encounters can be secured. If future encounters cannot be secured Mates Detection RG and all other forms of Detection RG have little or no value to the one receiving this information.

Men and women equally benefit from this solution to their problem of finding good potential mates. I expect no difference between men’ and women’s tendency to engage in Mates Detection RG. However, the topics they will discuss about potential mates will differ and keep in line with the above-described sex differences in mate preferences.

In the context of long-term commitment, female gossipees of Mates Detection RG will be subject of information about their age and attractiveness as cues for her reproductive value, and her reputation of being faithful or not. In this same context, male gossipees will be subject of information about their ability and willingness to invest, their protective abilities, compatibility to the interested (female) candidate, and their willingness to commit or not.

In the context of short-term relations, female gossipees will be subject to information about their age and attractiveness as cues for their reproductive value, and their reputation of being promiscuous or not. In this same context of short-term relations male gossipees will be subject to information about their ability to invest, and their physical attractiveness.

In sum, female gossipees of Mates Detection RG have more chance to be praised or criticized about their age and attractiveness than male gossipees, both in the context of short-term and long-term mating. Male gossipees of Mates Detection RG have more chance to be praised or criticized about their abilities and willingness to invest, both in the context of short-term and long-term mating.

Along with this, female gossipers of Mates Detection RG will be most interested in information about male potential partners' abilities and willingness to invest. And likewise male gossipers will be more interested in information about the age and attractiveness of female potential partners. Again, when selecting for short-term and long-term strategies, extra criteria can be added.

In our search for a potential mate, knowledge about whom is having a (sexual) relationship with whom is important. Being single increases your reputation as a good potential mate, but being in a relationship does not exclude you from the mating circuit. As mentioned earlier mate poaching does occur; both men and women potentially cheat on their partner, and sometimes leave their partner for someone else. When you are looking for a potential mate, you might therefore also look at men or women who are in a relationship. Still, it is important to know with whom a potential mate is engaged. Are they engaged with your best friend, a family member or a very jealous, aggressive partner? In these cases you might not want to interfere in this relationship. However, when your potential partner of interest is dating one of your foes, or an unknown harmless person, or even better if he or she is in a bad relationship at the moment, you might continue your strategy to win his or her heart. Mating Structure Reputation Gossip is information that informs gossipers of whom is sleeping with whom. As Barkow (1989) explained (see citation on first page of this chapter), it is far more interesting to know whom someone sleeps with, than how soundly they sleep.

Levin and Kimmel (1997) concluded from their content analysis of media gossip columns that ‘relations’ are the most gossiped about topic. Their conclusion, which I will come back to in more detail in chapter 7, indicates the importance of Mating Gossip. They did not focus on the different kinds of Mating Gossip as I define them, and so I cannot find support for the presence of Mates Detection RG in media gossip columns.

I did find some support for my ideas about Mates Detection RG in Power's (1998) thoughts on gossip. She also bases her ideas on gossip on the mating theory Buss and Schmitt (1993) proposed, and mentions how Deacon (1992, 1997) had stressed the manipulative use of gossip in the context of human mating. According to Power (1998), gossip is used both for male and female reproductive strategies. Men gossip with men about other women’s sexual behavior and reproductive value, she says. Women discuss the status of men, and warn their female friends about known philanders. Mating gossip not only occurs between male-male and female-female conversations, says Power (1998), but men gossip with women about the marital status of potential mates; they search for potential mates through gossip. And, both men and women warn their opposite sex friends about bad mates.

In general, these ideas are strongly related to what I offer. However, Power (1998) did not make divisions for different kinds of gossip, which I clearly distinguish from each other. I will mention her ideas again, in the next section on Intrasexual Conflict Gossip. However, she did not mention the valuable use of gossip as a control system to guard mates, as I do.

In their search for a potential mate, and even after they have found a mate, both men and women face threats from same-sex rivals. Other women will desire the good man you are after, and even after you have hooked up with him, you must be aware that mate poaching does happen. Being single or in a relationship, men and women must constantly be aware of these threats, and therefore constantly face the problem of same-sex rivalry. Intrasexual Conflict Reputation Gossip might have offered a solution to this problem. Detecting your sexual-rivals and lowering their reputation through gossip can make you become the winner in this battle within the sexes. Sexual Rival Detection RG functions to solve the problem of detecting whom our sexual rivals are. Sexual Rival Slander RG functions to lower the reputation of these competitors, and hence increase your own reputation as a good potential mate.

Again these specific forms of gossip require a direct or indirect connection between the gossipers and the gossipees. Receivers of Sexual Rival Detection RG benefit only if they get information about rivals whom they can encounter or, more importantly, whom their potential partner can encounter. The same rules apply for Sexual Rival Slander RG. Senders of this slanderous form of gossip only benefit if they deal directly with the gossipee or because an indirect connection to the rival exists, which makes potential future encounters a reality. Totally unknown gossipees with whom even indirect connections do not exist and with whom future encounters are not possible will not be gossipees of Mating Reputation Gossip.

Intrasexual Conflict RG is closely related to Alliance Reputation Gossip, which I will discuss later in this chapter. Both concern the use of information to deal with competitors, which Hess and Hagen (2000) refer to as ‘Informational Warfare’. Gossip, according to Hess and Hagen (2002) is equivalent to informational aggression in the sense that gossip is defined as the dissemination of information about others’ reputation with the purpose of increasing your own status relative to the status of your competitors.

Important differences between Intrasexual Conflict RG and Alliance RG are that Intrasexual Conflict RG occurs between two or more parties of the same sex, while Alliance RG can be between two or more parties of the same or different sex. The stakes of Intrasexual Conflict RG are potential mates (i.e., women battling for the same man and men battling for the same woman). The stakes of Alliance RG are not about interest in potential mates, but concerns non-sexual social relationships (as I will discuss later).

As I have mentioned above, Mating RG is equally important for men and women, and I generally do not expect sex differences in the tendency to use this kind of information. However, I do expect a sex difference in the use of Intrasexual Conflict RG, and especially for Sexual Rival Slander RG. In times of conflict, informational warfare is not the only means to battle competitors; physical aggression can be used as well. But, as Campbell (1999) suggested, women cannot afford to compete physically. She explains that physical aggression is more costly for women, since mothering is more crucial to female inclusive fitness than fathering is to male inclusive fitness. Therefore women engage more in low-level combats, such as using negative gossip. Hess and Hagen (2002) comment that informational aggression is not merely a substitute for physical aggression, but is simply more effective to use in within-group conflicts, to attack reputations.

Hess and Hagen (2002) further argue that the reputation of a female potential mate is more disguised than a male’s reputation as a potential mate. Women’s disguised reproductive value makes them an easy target to gossip about, Hess and Hagen argue, because it is more easy to lie about that then about men’s abilities and willingness to invest. In general they claim that:

“Because gossip is an excellent strategy for within-group competition, and because it is effective in attacking and defending difficult-to-assess aspects of reputation, gossip may have been a more effective weapon in female intrasexual competition than it was in male intrasexual competition, increasing selection for psychological adaptations for informational aggression in females.” (Hess & Hagen, 2002: 12)

Following their arguments, I therefore predict that women will have a stronger tendency to engage in Sexual Rival Slander RG than men.

Again, I mention that these ideas on Mating RG are very similar to Power’s (1998) vision of gossip. When focussing here on Intrasexual Conflict RG, she also suggests that gossip is best used by men and women to warn each other for potential rivals, and check the status of rivals. Women discuss the reproductive value and sexual behavior of rival women with each other, because they can manipulate the reputations of rivals, she says. Also, both men and women warn their opposite sex friends about mates' rivals, in her opinion. However, she does not mention data to confirm these thoughts. Specific research on Mating RG is scarce because researchers so far have always studies gossip in the most general sense, and focused mainly on the control and manipulation effects. That being said, I did find some research that either rejects or supports my ideas.

The most supportive findings come from McAndrew and Milenkovic (2002) research. They asked 39 male and 44 female students aged 17-22 how likely they would search for and spread around specific gossip information about specific others. The students were asked to imagine that the gossip topic concerned one of their male or female family members, friends, acquaintances, strangers or professors. Two gossip topics concerned ‘promiscuity’ and ‘sexual inability’. Their results showed that:

“Females were most interested in the behavior of other females when it came to stories about promiscuity (p < .001) and sexual infidelity (p < .05), but males had equal interest in other males and in females in these situations (p > .05). However, this was reversed when the gossip concerned an inability to have sex. In this case, males were significantly more interested in news about other males (p < .01), and females had no preference (p > .05). Females, however, were significantly more likely to pass along information about the sexual difficulties of males (p < .05) and the promiscuity of other females (p < .02). Males had no predisposition to gossip about one gender over the other in either of these situations.” (McAndrew & Milenkovic, 2002: 14)

That female students were so interested in the promiscuous and infidelity behavior of other female students indicates that they seek information they can use for Sexual Rival Slander Gossip. This information is especially useful for females when looking for a long-term mate and having to deal with competitors and, as I have outlined above, females still prefer long-term over short-term relations. Similarly, males show so much interest in the sexual inability of other males, because, in my opinion, this information is very useful when they seek a short-term mate and want to rule out competitors.

That female students were more likely to pass on this information than their male colleagues confirms the prediction of Hess and Hagen (2002) that women might have a stronger tendency to use Sexual Rival Slander RG than men.

Because Sexual Rival Slander RG is about attacking the reputation of others, this is clearly a form of so-called ‘bad gossip’. Slandering rivals is information with a negative connotation. If girls and women, as Hess and Hagen (2002) predicted, and McAndrew and Milenkovic (2002) confirmed, engage more in this kind of gossip, it might be expected that the overall gossip conversations of girls and women contain more derogatory gossip. Levin and Arluke (1985) tested this. Their hypothesis was not based on Hess and Hagen’s (2004) ideas, but was instead derived from the common sense definition of gossip as bad talk, which is popularly associated with women. They analyzed the difference between girls’ and boys’ gossip conversations among college students but could not confirm their hypothesis that girls might be more derogative in their gossip. Even though this might indicate that the suggestions may be wrong, I did find other studies that can also support the above-described ideas.

As I explained in chapter 1, I did not exclude lies from my general definition of gossip. In the context of Intrasexual Conflict RG, lying about rivals’ reputation can be a good strategy to compete with them. Kuttler et al (2002) analysed the gossip strategies of 384 preadolescents. They noticed that when preadolescent girls want to attack the reputation of other girls, they often spread lies about them. Self-serving motivations and jealousy are the drives for adolescents to share false knowledge, say Kuttler et al (2002), and girls more than boys believe jealousy to be the major drive to spread lies through gossip.

Although the search for a good potential mate and the battle within the sexes can be expected to occur most among preadolescents and adolescents, these things occur among older people as well. For instance, Percival (2000) reports that female elderly people in sheltered housings use gossip as a tool to manipulate each other’s reputation. Women dominated the sheltered houses he observed. Men, as a minority group, were often the source of jealousy among the women, and gossip was used to attack the reputation of rivals, he reports.

A last form of support stems from the research domain on gender and discourse, (e.g. the research of Coates, 1989, 1993, 1996). As Cameron (1997) commented, a lot of these analyses focus on the conversation style and ways of interactions, and less on the content of what is being discussed. However, they also focus on the competitiveness over status in the talk between women, which is strongly related to what I call female Intrasexual Conflict RG. Guendouzi’s (2001) findings are also of interest. She analyzed the content of gossip conversations in only-female groups, and concluded that women do a lot of, what she calls, ‘bitching’. Guendouzi concludes that, besides co-operation, much of women's talk involves competition:

“Bitching, I would argue, is a form of talk that is power-driven, enabling its participants to compete for social capital through contrasting their own behavior with that of others in their peer groups. The women’s talk analysed here shows bitching to be a highly powerful ideological form of talk that relies on cooperative conversational strategies to function as a means of competing for ‘peer-validated’ versions of femininity in the social marketplace.” (Guendouzi, 2001: 47)

Though, not exactly focused on Intrasexual Conflict RG, I think this last statement indicates that what I outlined above about the functioning of Intrasexual Conflict RG can be confirmed. However, again, researchers so far have not focused on the different sub definitions of gossip as I propose here. I think future researchers should make use of these more specific definitions, and further elaborate on Mating Gossip in particular.

The last form of Mating RG that I define is Mates Control RG. I suggest that this form of gossip might have solved (and still solves) the problems of mate guarding that our ancestors faced. When involved in a relationship, both men and women still have to deal with same-sex rivals, and will still use Intrasexual Conflict RG to deal with these specific problems. However, beyond dealing with same-sex rivals, men and women face the problem that their partner might cheat on them. Gossip about the cheating behavior of our spouses, or the spouses of our friends, might have been an ideal tool for our ancestors to stay informed. I mention the partners of our friends as well, since as we do not want to be cheated by our own partner, we will not want this to happen to our family and friends either.

As I explained above, it is important to understand a difference between ‘emotional cheating’ and ‘sexual cheating’. For women, the biggest threat, in biological terms, is that their partner invests his time and resources in another woman (emotional cheating). For men, their biggest threat, in biological terms, concerns the sexual infidelity of their partner; they risk investing their resources in a child that is not theirs. These differences in costs for men and women, translate in differences in jealousy (see above), and in my opinion in different interest in gossip topics.

More specifically, I predict that women have a higher likelihood than men to become a gossipee of Mates Control RG that focuses on sexual infidelities of women in a relationship. And vice versa, I expect men to have a higher likelihood to become a gossipee of Mates Control RG that focuses on emotional infidelities of men in a relationship.

Here again, I do not expect male / female differences in the tendency to exchange this kind of gossip. From the gossipers’ perspective, both men and women will want to control the actions of a specific person, who is committed in a relationship. Highest interest can be expected if the gossipee is their own partner. Still, interest can also be present when the infidel gossipee is someone they know or the partner of someone they know.

So far, I have not yet found data to support these ideas. As I mentioned above, many researchers have already tested and confirmed the evolutionary predictions for sex differences in jealousy. But so far, no research has investigated the specific role of gossip as a mates control strategy.

What I haven’t mentioned so far, but is important to discuss in the context of Mating RG, is gossip about the parental skills of men and women. A lot of our daily gossip conversations concern the good or bad parenting skills of other people, especially among people who have children of their own. I have already discussed that learning how to be a good parent can be transferred through Mating SLG. Still, we not only want to learn about parenting skills through gossip, but label specific people with a good or bad reputation as a parent. As mentioned above, I first must elaborate on ‘Inclusive Fitness’ before I can explain the problems of parenting.

Parental skills can be ideal stakes for certain forms of Social Gossip, therefore, I will discuss this more in detail in section 4 of this chapter.

“Humans are an obligately interdependent species, unable to survive and reproduce outside a group context.” (Caporael & Baron, 1997: 328)

In what follows, I will sketch a last compilation of adaptive problems our ancestors faced. I here focus on adaptive problems of group living. The presence of other individuals has shaped our mind, as much as ecological and sexual (which already implies the presence of others) selection pressures did. I discuss two main areas of adaptive problems here: problems of co-operation and problems of conflict.

Co-operation is so common in everyday life all over the world that it does not need to be proven that it exists. Co-operation is not a recent human behavior, but has a long history, and is as present among hunter-gatherer societies, as it is today. Milton (1988) for instance, comments about the importance of co-operation in food acquisition:

“Cheaters (clever manipulators) in [such] a foraging system might survive over the short run, but would be unlikely to persist over the long run as eventually there would be no one left to carry out the work of hunting and gathering or no one wishing to hunt or gather with the selfish manipulator. Therefore, in my view, one of the strongest selective pressures on early humans would have been to develop the ability for co-operative behaviour, delayed gratification, and the sharing of highly desirable and essential goods.” (Milton, 1988: 304)

Behavioral ecologists have tried to explain why hunter-gatherers share food, and cooperate in the acquisition of food as well (e.g. see Alvard & Nolin, 2002; Gurven, in press; Hill, 2002; Hawkes et al, 1997; Winterhalder, 1997). Since, as obvious as co-operation might seem, it is a puzzling phenomenon in evolutionary terms.

Co-operation involves altruistic acts of individuals. And, as Trivers (1974: 250) says: “An altruistic act can be defined as one that harms the organism performing the act while benefiting some other individual, harm and benefit being defined in terms of reproductive success.” Therefore, altruism (and co-operation) needs an explanation in evolutionary terms. Why would individuals help others to increase others’ benefits, while implying costs to themselves? “If natural selection followed the classical models exclusively, species would not show any behavior more positively social than the coming together of the sexes and parental care.” (Hamilton, 1964: 1).

Cosmides and Tooby (1992: 167-170) discuss the problem of altruism that has been solved by the explanation of two main theories: kin selection, as explained by Hamilton, and reciprocal altruism, put forward by Trivers. I discuss both in more detail.

Hamilton's (1964) first work was redefining Darwin’s term ‘fitness’. The classical, Darwinian definition of fitness refers to the effect solely on the number of children an individual can have (Tooby & Cosmides, 1992: 168). Using a mathematical model, based on relations between related persons in one generation, Hamilton concluded there’s more to fitness than the classical definition embodies. Fitness is more than just the possibilities of reproduction and survival of an individual. What Hamilton (1964) called inclusive fitness, can be described as follows:

“Inclusive fitness may be imagined as the personal fitness which an individual actually expresses in its production of adult offspring as it becomes after it has been first stripped and then augmented in a certain way. It is stripped of all components which can be considered as due to the individual's social environment, leaving the fitness which he would express if not exposed to any of the harms or benefits of that environment. This quantity is then augmented by certain fractions of the quantities of harm and benefit which the individual himself causes to the fitnesses of his neighbours. The fractions in question are simply the coefficients of relationship appropriate to the neighbours whom he affects…” (Hamilton, 1964: 8)

On the one hand, inclusive fitness is dependent on environmental factors, but on the other hand inclusive fitness contains more than classical fitness, since it is augmented by contributions the individual makes to the fitness of kin related people. How much you can augment your own fitness, by helping family members, depends on how related you are. Relatedness between two individuals can be expressed by Wright’s (1922, cited in Hamilton, 1964) coefficient of relatedness (r). This coefficient, r, varies from 0 (unrelated) to 1 (identical twins). For offspring and siblings r=0.5, for aunts, uncles, nieces and nephews r=0.25, and so forth.

Basing his ideas on the notion of the inclusive fitness of an individual, Hamilton (1964: 14-16) could explain altruism between related individuals. He described four classical situations between two interacting related individuals. In the first scenario, both can interact in such a way that both benefit from the act. Second, in some interactions both members lose. Third, an individual can be altruistic, which means he sacrifices himself in benefit of the other. Fourth, and lastly, an individual can act egoistic, which means he or she wins and his or her partners loses. Selection, Hamilton (1964) argued, will favor the first option and rule out the second one.

But how will selection operate in the two last actions? Hamilton (1964) searched for some continuity, a general rule that can tell us when altruism will occur and could explain when it is useful for an individual to help a related family member. He came up with the following mathematical rule: k > 1/r. In which ‘k’ refers to the ratio between the benefit and the costs (losses) of the individual acting altruistic. And: “This means that for a hereditary tendency to perform an action of this kind to evolve the benefit to a sib must average at least twice the loss to the individual, the benefit to a half-sib must be at least four times the loss, to a cousin eight times and so on.” (Hamilton, 1964: 16).

A practical example of his simple rule is the alarm call of birds. Imagine a bird nest being threatened; mother bird will signal the danger in order to keep the intruder away. Doing so she will endanger her own life to protect the life of her offspring. She is related to each of her offspring with r=0.5, and when she can save 4 of them the k ratio is: profit 4/loss 1. Given Hamilton’s rule, this brings us to a final result of 4>1/0.5. So according to Hamilton, the mother bird’s act was very useful.

We now have explanations for three of the four most common interactions between family members, but what about the last one; how can we explain selfish behavior? Simple: by turning around the formula: k < 1/r. If an altruistic act towards a related other asks too much effort of ourselves, this is, when the costs are bigger than the benefits of the action, the act will not be favored by natural selection (according to Hamilton’s rule).

Nepotism, which refers to Hamilton’s kin selection theory, occurs universally (Daly, Salmon & Wilson, 1997). People are more willing to help their kin-related family members than other people. People also show less violence towards kin-related individuals. Daly and Wilson (1988) described this in detail in their book ‘Homicide’, in which they report that violence occurs more between non-related individuals. One special situation they focus on is adopted or stepchildren, who more often are the victim of abuse by their stepparents or adopting parents (see Daly & Wilson, 1988 for more details).

Nancy Segal, known for her research on twins, also tested Hamilton’s rule. Her prediction was that monozygotic (MG) twins should show more cooperative behavior than dizigotic twins (DZ). Using a prisoner’s dilemma game (see also below), she could confirm this prediction: MZ twins displayed more co-operation than DZ twins (Segal & Herschberger, 1999).

Hamilton’s rule, of course, only explains altruistic acts, and co-operation between related individuals, and does not offer an explanation for co-operation between unrelated members. I will come back to this in the next section. First, I want to mention the crucial issue for Hamilton’s idea; that some kin recognition is required. Of course we cannot always know who is related to us and who is not. Still, kin-recognition systems have been hypothesized to exist in humans. Lieberman, Tooby and Cosmides (2003) have shown that people regard sexual relations between two parties that have been raised together (co-residence) as incest. Co-residence, they argue, predicts genetic relatedness, and can therefore be used as a cue for kin-recognition.

Another cue, we might use to detect kin related people is facial similarity. DeBruine (2002) tested people on their willingness to cooperate with strangers. She either showed them pictures of faces that were significantly different from the respondent’s face, or pictures of faces that were quite similar to the face of the respondent. Her results indicate that people are more trusting of strangers that look fairly similar, and are more willing to cooperate with them. DeBruine (2002) assumes that this might indicate that people feel more related to others with whom they share facial similarities.

A last cue people can rely on for kin-recognition is body odor. Porter, Cernoch and Balogh (1985) experimented with the recognition of kin related members by using odor as a cue. They presented respondents with odor samples of people, who were unknown to the respondents. The respondents were asked to match odor samples of people they thought were related to each other. They concluded that the odors of mothers and their offspring were correctly matched. Matching these same mothers to their husbands by using their body odor as a cue did not result in correct matches. Mothers and their offspring are kin related, and mothers and their husband do not share genetic material. From their experiments, these authors concluded that kin related people have similar body odors, and therefore odor can be used as a cue for kin recognition.

Looking at human co-operation form an evolutionary perspective, Hamilton only explained co-operation among related individuals. Co-operation makes some sense, when we help kin-related people, but when talking about helping ‘strangers’ we really come to the ‘evolutionary puzzle of co-operation’ (Fehr & Gächter, 2002).

Although Hamilton’s Kin Selection theory seems to be only applicable to related individuals, Trivers (1985) argued it could also be used to explain co-operation among non-related individuals. Individuals will cooperate with each other, he argues, as long as the outcome of the interaction is beneficial; if the benefits outscore the costs. What Trivers labelled as Reciprocal Altruism refers to the simple principle “if you help me, I will help you in return”.

Trivers (1985) already argued that co-operational transactions between individuals do not need to occur bilaterally at the same time. An individual need not to reciprocate immediately, but can do this at a later point in time. Transactions where reciprocity is delayed are referred to as Indirect Reciprocity (see also below).

However, As Trivers (1985) stated, this simple principle can only work in populations where no cheaters occur. Cheaters or free-riders are individuals who benefit from the altruistic acts of others without returning a favor.

Contributions to explain human co-operation that focused on the free-rider problem came from game-theory and computational theory. I first explain the computational theory, and will focus on game theoretical approaches to human communication in the next section.

Cosmides and Tooby (1987, 1992) put forward a computational theory for social exchange. They defined the information processing problems that each individual is faced with when he or she has to solve a problem of social exchange. According to these authors, problems associated with social exchange are, first of all, that an individual must be able to recognize other individuals. Second, individuals must be able to remember some aspects of one’s past interactions with other individuals. And last, Cosmides and Tooby (1987, 1992) suggest that individuals should also look at the general reputation of a potential cooperator, as well as the chance to meet this person in the future. They organized their ideas in a ‘grammar of social contracts’:

“Just as grammar of the English language is a set of rules for distinguishing well-formed sentences from ill-formed sentences, a grammar of social contracts is a set of rules for distinguishing well-formed social contracts from ill-formed social contracts. It includes the set of assumptions about the rules governing social exchange that must somehow be incarnated in the psychological mechanisms of both participants. Without these assumptions, much of what people say, mean, and intend to do in exchange situations could not be understood or anticipated.” (Cosmides & Tooby, 1987: 72)

Cosmides and Tooby (1987) outline their grammar of social contracts in terms of costs and benefits involved in the exchange action. Social contracts concern algorithms of the form “If P then Q” in a social exchange context. For instance “If you do my dishes, I will pay you $20”. Translated terms of costs and benefits mean that the one who proposes the contract (x) first asks for a benefit B from the acceptor (y), for which he or she will return a favor, which means implying a cost to himself or herself. If P means a cost for the acceptor C(y) and a benefit for who proposes B(x). Not P means zero costs for the acceptor and zero benefits for who proposes. Q implies a benefit for the acceptor B(y) and a cost for who proposes C(x). Not Q implies zero costs for who proposes and zero benefits for who accepts the social contract.

Cosmides and Tooby (1987, 1992) presented respondents violations on such “If P then Q” contracts, making use of the Wason-selection-task (see Cosmides 1989), and they noticed that on average people are better in solving such problems if they are presented in a social context. When the violation of the simple rule “If P then Q” was presented in an abstract sense, respondents showed significant lower scores, meaning they had more troubles solving such problems in abstract contexts.

To rule out the possible effect of ‘availability’, which points to the effect of familiarity with the presented social situation that might cause the better results of solving the Wason-selection-task in social settings, Cosmides (1989) played with different social settings. She presented respondents non-familiar social settings, such as “If a man eats cassava root, then he must have a tattoo on his face”, and “If you eat duiker meat, then you have found an ostrich eggshell.” (Cosmides, 1989: 211), and again asked her respondents to play detective and search the violation of the social contract (the cheater). Cosimides’ (1989) results ruled out the effect of familiarity:

“No matter how wildly unfamiliar the rule’s terms, social contract problems elicited social contract responses. Furthermore, if associations between specific terms were responsible for the pattern of results on social contract rules, then descriptive rules using the same unfamiliar terms should have elicited the same pattern; they did not. No theory whose predictive and explanatory power rests on associations between specific terms used in a social contract rule can explain the result of these experiments.” (Cosmides, 1989: 228)

What Cosmides and Tooby (1992: 220) have shown with their social contract theory is that humans possess a mental mechanism to detect cheaters in social settings. “In other words, the empirical record is most parsimoniously explained by the hypothesis that the evolved architecture of the human mind contains functionally specialized, content-dependent cognitive adaptations for social exchange.” (Cosmides and Tooby, 1992: 220).

In addition to Cosmides and Tooby’s (1992) cheater detection mechanism, Brown and Moore (2000) suggested we also have a mechanism to detect altruists. They adapted Cosmides’ (1989) Wason-selection-task, so that respondents should solve the problem of detecting altruists instead of cheaters. And:

“This study suggests that participants can solve altruist-detection Wason problems significantly better than control Wason problems (abstract and school versions). This finding is analogous to the findings on cheater-detection by Cosmides (1989). Altruist-detection tasks elicit the switching response (i.e., “not-P” and “Q”) for reversed rules. The switching response supports the notion that subjects are detecting altruists rather than giving logically correct responses. Furthermore, participants did not solve cheater-detection tasks better than altruist-detection tasks. These results are consistent with the evolutionary hypothesis that humans are sensitive to information regarding the genuineness of the altruistic behaviour performed.” (Brown & Moore, 2000: 33)

In the game-theory domain, the problem or free-riding is classically known as the Prisoner’s Dilemma. The Prisoner’s Dilemma presents the situation where two prisoners are questioned about a crime they committed together. Both of them do well by denying, because this will give both of them an equally moderate benefit of not being considered as guilty. However, when they are told that they will be minimally punished if they confess, then admitting they committed the crime becomes tempting. This is because confession results in an even bigger benefit than when both deny. If they both confess they equally benefit. However, if one of them confesses and the other one denies then benefits are unequal for both prisoners. In this situation one gets a maximum benefit (because he denies the facts), and the other will suffer a maximum punishment (he has confessed being the criminal) (Axelrod & Hamilton, 1988).

Axelrod & Hamilton (1981) designed a computer game, based on the prisoner’s dilemma. In their game, two players who both co-operate both get a moderate benefit. However, they can get a bigger benefit if they defect on an altruistic player (who co-operates). Axelrod (1984) set up a competition to solve this problem. Different contributors came up with strategies, and the winning strategy was the Tit-for-Tat. This simple strategy starts with co-operation and then copies what his opponent did in the previous round. If the opponent defects, Tit-for-Tat defects in the next round. If the opponent co-operates, Tit-for-Tat co-operates in the next round. Tit-for-Tat won the competition because it was an Evolutionary Stable Strategy (ESS), which means that no other strategy that played against Tit-for-Tat could win from it. Tit-for-Tat could therefore not be ruled out over generations by competing strategies.

According to Trivers (1985: 389-392) the three essential features of the Tit-for-Tat strategy that contribute to its success are; first of all, that this strategy never starts with defecting others. Second, it is successful to never defect on an individual who has never defected on you (which Tit-for-Tat respects). Third and last, it is important to be forgiving towards individuals who only defected on you once.

The weakness of Tit-for-Tat, says Wright (1994: 198) is that is might not reflect what is going on in real life human co-operation: "In real life, co-operation isn't a matter of black and white. […] So the human rules for reciprocal altruism are likely to be a bit less binary than TIT FOR TAT's". Nevertheless, Tit-for-Tat could solve the free-rider problem and therefore explain the evolution of co-operation.

Another solution to solve the free-riding problem that has been proposed within the field of game theory is called Strong Reciprocity (see e.g. Fehr & Fischbacher, 2003; Fehr, Fischbacher & Gachter, 2002; Fehr & Fischbacher, 2004; Fehr & Gachter, 2002; Gintis, Bowles, Boyd & Fehr, 2003). Strong reciprocity suggests that we implement punishment in our social interactions. If free-riders are being punished for their behavior, the problem of cheating can be controlled. However, in order to punish free-riders, those who punish have to pay a cost in doing so, therefore Fehr and Gächter (2002) talk about ‘altruistic punishment’. Using a public goods game, controlling for punishment to be present or not, they concluded that in the condition where free-riders were punished, co-operation maintained a higher level over time, whereas in the condition that lacked punishment the level of co-operation decreased over time. The proximate mechanism that triggers punishment is a strong negative emotion (Fehr & Gächter, 2002). Fehr and Fischbacher (2004) even noticed that people punish free-riders, when they were merely witness of the violation of a co-operation norm, and are not the victim themselves; they refer to his as third party punishment.

The last field of research I want to discuss here investigates the role of reputations to explain human co-operation.

Reputations are of great impact, and importance. This is not only the case in our current, modern societies, but also in current indigenous societies that most closely resemble the EEA; the hunter-gatherer societies that still exist today. It has been shown that among hunter-gatherers individuals can gain access to important resources, such as food and mates, by having a good reputation. If they loose their good reputation, this often implies that they loose their access to resources, such as food and mates, as well (for more details on this see e.g. Chagnon 1988; Gurven, Allen-Arave, Hill, and Hurtado 2000; Hawkes 1991; Hawkes 1993; Hawkes, O’Connell, and Blurton Jones 2001; Marlowe 1999; Patton 2000; Smith & Bliege Bird 2000; Sosis 2000; Sugiyama and Chacon 2000).

Other proof of the importance of reputations in resource acquisition once again stems from game-theories. I will not discuss this here, because it would lead me too far off course, but I want to stress that the importance of reputations has been proven by others (for details see e.g. Milinski, Semmann, Bakker, and Krambeck 2001; Milinski, Semmann, and Krambeck 2002; Wedekind and Milinski 2000).

The importance and benefits of reputations has been applied to theories of co-operation, such as indirect reciprocity and costly signaling. ‘Costly-signaling’ or ‘show-off’ theories simply refer to the fact that individuals act altruistically (which is costly) to signal their qualities as mates or social partners. They invest costs in their altruistic deeds, to get return benefits of having a ‘good reputation’, which in the end increases their chances of acquiring resources, such as mates, or social partners who are willing to cooperate with them due to this good reputation (see e.g. Alvard & Nolin, 2002; Gintis et al, 2001; Gurven, in press; Hawkes 1991; Smith & Bliege Bird 2000).

In the context of reciprocal actions, reputations are useful to learn whether future social partners have defected on others or not (e.g., Enquist & Leimar 1993; Pollock & Dugatkin 1992). Indirect reciprocity was already mentioned by Trivers (1985, see also above), and refers to delayed returned benefits. After Trivers, who focused on bilateral transactions, others have applied this theory to transactions between more than two players. This is to say, if A helps B, while B helps C, B can ask C to return a favor to A. In this situation, all players will benefit, but the transactions are no longer bilateral, but can be expanded. This makes co-operation in larger groups possible, but requires good reputations. If, for instance C does not have a good reputation, A might not accept the offer form B (that C will return a favor), and the whole co-operation between these three players would fall apart. Reputations are crucial, because they keep a record of someone’s past interaction with other group members (see e.g. Alexander 1987; Leimar and Hammerstein 2001; Nowak and Sigmund 1998; Panchanathan & Boyd 2003, 2004).

As some (e.g., Enquist and Leimar 1993; Leimar and Hammerstein 2001; Panchanathan & Boyd 2003) have stressed, reputations involve information exchange, and therefore communication. To end this section on human co-operation, let me overview what has been said explicitly about the role of communication in the context of co-operation.

Ulbaek (1998: 41) says that: “Language evolved in the Homo lineage not because of superior hominid intelligence, but because of special conditions: the development of reciprocal altruism as a way of gaining fitness by sharing and helping.” Ulbaek claims that language is helpful in the context of reciprocal altruism. The reason for this is that language enables individuals to communicate (shared) values, and creates reputations.

First of all, Cosmides and Tooby (1987, 1992) address the importance of language in their social contract theory, by pointing to the fact that language is useful to communicate one’s values to others (although this is also possible on nonverbal terms). And, researchers have noticed indeed that communication leads to an increase of trust and cooperativeness in the context of social dilemmas, such as the prisoner’s dilemma, or other economic games. If participants are allowed to communicate with each other prior to playing a social dilemma game, the level of co-operation increases (e.g. Orbell & Dawes, 1993; Orbell, Van de Kragt & Dawes, 1988).

In larger social settings where interactions between each individual with each other individual become scarcer, distortions in predicting the behavior of others will occur more frequently. This is an argument in Dunbar’s (1998) theory on the evolution of social language use (see also previous chapter). Cheaters can occur more easily in larger groups, he argues, and the exchange of social information about others can resolve this. By talking about others (i.e. gossip) cheaters can be detected.

Using computer simulation to test whether co-operation can evolve in large social groups Cox et al (1999) found that: “Co-operation can evolve as a dominant and stable strategy in very large social groups provided that certain conditions are met.” (Cox, Sluckin & Steele, 1999: 374). These constraints are that there need to be enough interactions between the group members and each individual should have memory capacity to store information about other players. These constraints are in line with Dunbar’s social brain hypothesis, which claims that our brain expanded because of changes in social group sizes, and the need to maintain productive social relationships.

The reason why language enables co-operation in larger social settings is because it constructs and manipulates reputations. Since, as Panchanatan and Boyd (2003) say:

“If group size is small, perhaps individuals can monitor the goings on of all others and thus properly attribute standing to a partner observed defecting on another. As group size increases, however, this assumption seems implausible. Language seems to offer individuals access to information about others that they were not able to observe directly. Integrating this hearsay with personally observed information, individuals may be able to accurately track the standing of other group members. This argument is parsimonious with the observation that something like indirect reciprocity seems to be extremely rare in nature with the notable exception of humans. Without effective communication, reputations may only exist in the context of a stable dyad. Once communication is possible, an individual’s reputation takes on more global characteristics.” (Panchanathan & Boyd, 2003: 123)

Further support that language plays a role in the detection of cheaters can be found in the research of Etcoff et al (2000), who analyzed the lie detection skills of aphasics (people who are unable to understand words). They noticed that aphasics significantly outscore people with no language impairment when detecting lies about emotion. They are better equipped to recognizing non-verbal cues. “Perhaps damage to the circuitry underlying language comprehension results in the growth of compensatory skills in recognizing non-verbal behaviour.” (Etcoff et al, 2000: 139).

Along side problems of co-operation, individuals living in social groups are also confronted with problems of conflict. I will start here with an explanation of how conflict arises between kin-related members. More specifically, I will discuss the problems of raising offspring, to which I already referred in section 3.5 of this chapter. As I indicated in section 3.6.4 of this chapter, a specific kind of gossip might be able to resolve adaptive problems of this kind. I will then turn to problems of conflict between non-related group members.

Trivers (1974) studied parent-offspring conflicts. His theory rests on the principle that parents are related with their offspring with r = 0.5. The core idea of Trivers’ (1974) theory is that offspring always want more investment from their parents, than their parents are willing to give. These differences in interest result in conflicts between parents and their offspring.

Interesting to mention here is Buss’ (1999) suggestion that parents tend to invest more in their offspring if their investment can result in returned benefits. Buss (1999) therefore points to the fact that child abuse is not common among parents and their unhealthy children or stepchildren. Still, child abuse appears even less among healthy children and genetically related offspring. Daly and Wilson (1988) have also shown that age is an important factor here. Although child abuse occurs seldom, it is even less present among children who have reached their reproductive age.

Also keep in mind that the relatedness between parents and children is not always r=0.5. As Buss (1999) comments, fathers are never a 100% sure their children are theirs, they are confronted with the problem of paternal uncertainty, to which I come back in the next section. But even more important to mention here are stepchildren and adopted children, with whom the stepparents or adopting parents have no genetic relatedness. In line with Trivers’ prediction, Daly and Wilson (1988) reported that stepchildren and adopted children are on average more often subject to child abuse than offspring with whom the parents are related.

Beyond this direct parent-offspring conflict, Trivers (1974) also reports that conflict arises within the offspring. Parents, being equally related to all of their offspring will want their kids to act altruistic towards each other as soon as the benefits from their acts outscore the costs. From the offspring’s point of view, however, helping their sister or brother only benefits themselves if the benefits are twice as large as the costs of investment. Therefore children act less altruistic with each other than their parents want them to. Or as Trivers (1974: 260) says: “Parents are expected to socialize their offspring to act more altruistically and less egoistically than the offspring would naturally act, and the offspring are expected to resist such socialization.” (Trivers, 1974: 260).

Hunter-gatherer societies or often regarded as peaceful communities. Still, this is not in line with reality. Embler (1978) analyzed the data of 50 modern hunter-gatherer societies and on average in 64% of all cases warfare occurred at least once every two years. Only ten percent (including the !Kung in Africa) were rated as having no or rare warfare. Conflict is common among hunter-gatherer societies and occurs both between and within bands (Embler, 1978).

Although the mobility of hunter-gatherer societies hinders the accumulation of property or material goods, hunter-gatherer communities are not always egalitarian communities. Inequalities exist and struggles for status therefore occur (Cashdan, 1980). In ancestral environments, small cooperative groups were in conflict with other small cooperative groups (Krebs & Denton, 1997; Lewin, 1993; Tooby & Devore, 1987).

“In these environments, discriminating between friend and foe would have been valuable, as it is today. Making a mistake about whether another person was a friend or an enemy could have been lethal, so it is plausible to assume that it would have been adaptive to categorize others quickly and decisively and to be wary of strangers.” (Krebs & Denton, 1997: 27)

As I outlined above, reputations give access to resources. Reputations probably emerged in the context of cooperative behavior, but once they existed they became the target of conflicts over hierarchies and over achieving social status, because of these reputations. Having a good reputation leads to a higher (social) status, and aggressive behavior can increase your power over others; it leads to social status (Buss, 1999), and: “Status, reputation, and honor are far from trivial, however. Because humans evolved in the context of small groups (e.g. Alexander, 1987; Tooby & DeVore, 1987), a loss of status could have been catastrophic in the currency of survival and reproduction.” (Buss, 1999: 294).

The reason why humans are so ‘aggressive’ is due to of the benefits an individual can reap from this behavior. Buss (1999: 281-285) sums up some benefits, of which some are related to human mating (inflict costs on intrasexual rivals, deter infidelity), but most concern benefits in the context of group living. Aggression and conflict are defense mechanisms against attacks; they can deter future aggression, and be used to acquire the resources of others.

Being aggressive, or forceful can increase your social status, and allow you uphold a leader position. Henrich and Gil-White (2001) distinguish two kinds of leadership-traits. A man can be a leader because of his dominance power; forceful leaders use aggression to strive for status. But leadership can also rest on prestige, when the leader is good in persuading others. The latter achieve leadership because they excel in something: “Prestige rests on merit in the eyes of others (rather than force deployed against them), and promotes the admiration of inferiors (not their fear), a desire for proximity (not distance), and periods of sustained observation (not furtive glances).” (Henrich & Gil-White, 2001: 170).

Prestige rests on merit in the eyes of others. In times of conflict, co-operation again plays an important role. Through co-operation, an individual can surround himself or herself with allies, whom can contribute to his or her prestige, and can offer protection for hostility threats of other individuals and bands.

In their conflict with others, individuals also co-operate; they form alliances to compete with other individuals or other alliances. In our struggles with others we need support, and we not only fall back on family members, but we also have friends, or other allies to fall back on. As Tooby and Devore (1987) have suggested, conflict in the EEA most probably occurred between small bands of people. And these small bands rely on some special form of co-operation, they can be considered as what we nowadays call ‘friends’. Friendships are a form of long-term co-operations.

Coalition formation is not strictly limited to friendships, but friends are so important to have that people who lack them not only feel lonely, but also become more depressed. As Hartup and Stevens (1997) have pointed out, most studies about the importance of friendships focus on children or adolescents. However, as argued by these authors, friendship plays an important role throughout our entire lifespan. Friendship correlates with well-being; they foster our self-esteem and can therefore be considered as important cognitive and affective resources. Friendship relations have developmental advantages and lead to good outcomes.

The reason friendship is so important, say Tooby and Cosmides (1996), can be explained with the Banker’s paradox: “The harsh irony of the Banker’s paradox is this: just when individuals need money most desperately, they are also the poorest credit risks and, therefore the least likely to be selected to receive a loan” (Tooby & Cosmides, 1996: 131). Translated to human social life, this shows that those who needs help the most, are the most costly to others, and therefore risk being left alone.

Friendships, as Tooby and Cosmides (1996) argue, can resolve this problem since friends do not always count costs and benefits of their transactions (Silk, 2003). Friends support one another, tolerate a lot more from one another than other non-related individuals, make resources available for one another, and protect one another (Cords, 1997). Friendship is something special; the puzzle of friendship, as Silk (2003) calls it, is that it differs on two important aspects from reciprocal altruism that explains co-operation (see above):

“Friendship in contemporary Western societies seems to be based on two fundamentally incompatible rules. The first rule is that it is inappropriate to keep careful and accurate track of benefits given and received from friends, or to help friends with the explicit expectation of being repaid. This is not just rhetoric; in the laboratory, people obscure their own contributions to joint tasks with friends and avoid keeping track other of their friends’ contributions. The second rule is that costs and benefits should be balanced in relationships with friends. Friendships are expected to be based on equality, and people seem to be dissatisfied with relationships in which the benefit-cost balance is tipped in favor of themselves or in favor of their partners.” (Silk, 2003: 50)

The existence of these two rules implies that people value reciprocity in relationships with friends and strangers, and rely on the mechanisms of Tit-for-Tat reciprocal altruism to regulate their behavior toward strangers, but not toward friends. We have no models of the evolution of reciprocity that can accommodate both these rules. Theoretical work on reciprocity generally suggests that natural selection will favor strategies that are highly sensitive to recent interactions and require contingent (but not necessarily) equal distribution of benefits. The psychology of friendship contradicts the logic of these models.” (Silk, 2003: 50)

Friendships are different from other cooperative actions between non-related people. It has been suggested, says Silk (2003), that friendship is similar to the relationships between kin-related individuals, because it evolved in an environment (EEA) where our ancestors lived together with close relatives, and where individuals needed not to distinguish between kin and non-kin. But “[E]ven in those small foraging societies, people interact regularly with both relatives and non-relatives […]” (Silk, 2003: 47). Cords (1997) believes friendship was selected for because it returns so many benefits to individuals. She lists five in total: joint access to food patches; support, acquisition and maintenance of dominance; make food sources available to each other (point out were they are); tolerance and protection; and finally potential mates, since friends sometimes become mates. This last benefit is something that appears in our closest relatives as well; male-female relationships in baboons are considered a form of mating opportunities (Silk, 2003). What the exact underlying mechanism of friendship might be is not clear at this moment, but it is clear that friendship returns benefits, and individuals do best to secure a few or more good friends.

The strategy to secure friends, Tooby and Cosmides (1996) say, is to make yourself irreplaceable. One must be able to provide what others value, but cannot obtain (easily) elsewhere. The ability to offer special skills, combined with maintaining social relations with people who value and need these skills make a person irreplaceable, and worthwhile for a friendship relationship. Individuals can expect to be jealous of others who display similar unique skills, since this might threaten the irreplaceable position they hold.

Some friends are not like others. As people commonly say “only in times of need or suffering, does one gets to know his or her real/ true friends”. Besides these true friends, we all have our fair weather friends, who are more volatile. It is important for individuals to recognize the difference between both and as Tooby and Cosmides (1996: 135) say: “[t]he adaptive problem of discriminating true friends from fair weather friends would have been a formidable signal detection problem for our ancestors.”

Since we should divide our time as optimally as possible, and friendship requires spending time together (Cords, 1997), individuals probably possess a specialized computational mechanism to decide whom they want to spent time with. According to Tooby and Cosmides (1996: 136-137), the following rules can guide humans in the decisions they have to make when they decide with whom they should spend their time:

1. How many slots are already filled? If you have few friends, you will more easily decide to spent time with someone.

2. Who emits positive externalities? We not only seek friends with good reputations as altruists, but also desire side-effects (positive externalities). We e.g. desire good tool-makers, game locators, etc. We search for people with skills that can be helpful to us.

3. Who is good at reading your mind? Dyads who understand each other can cooperate more easily.

4. Who considers you irreplaceable? If you can be easily replaced, you better look for other opportunities to invest your time in.

5. Who wants the same things you want? If someone values the same things as you do, and has similar goals, this person will also make your local world beneficial for you by his or her actions.

4.3.2.3 Maintaining friends: keeping up and reconciliation when the friendship is broken

“Losing a valued friend, being able to spend less time with the friend, becoming less valued by that friend, or at the extreme, social isolation, may be more costly than being cheated.” (Tooby & Cosmides, 1996: 138).

An importance feature in this coalition forming with friends is that we should not only seek who are true and best friends, but we have to maintain these bonds. As Cords (1997) says, we often need reconciliation if we have harmed a friend. Friends have to protect their bond from disruptive aggressive conflict attempts of external forces, such as other (non-ally) individuals. Humans often do this by clearly indicating who is their friend and who is not. We classify others as ‘the good’ or ‘the bad’.

In all this coalition formation, humans seem to be biased in their perception of other individuals; in-group members, allies and friends are seen more positively and out-group members, or non-allies, more negatively. We like our friends, we dislike our enemies, and we seem to be very categorical in our thinking about others. Krebs and Denton (1997) believe this has a clear adaptive value:

“Inasmuch as other people perceive us in term of our social identities and treat us as members of the group to which we belong, it is in our interest to enhance the status of our groups. This does not involve group selection because we are, in effect, using the groups with which we are associated to foster our own fitness. When we enhance the value of our in-group, we, as a member of the group, may gain more than we have sacrificed on behalf of the group.” (Krebs & Denton, 1997: 35)

Krebs and Denton (1997) further suggest that, regardless of the fact that we make judgments rather quickly and often stick to them for quite a while (especially if the first judgment is a negative judgment), our cognitive programs to make judgments should be flexible. We often switch alliances when it is in our interest to do so, and therefore we should be able to adapt our judgments according to these switches.

Now, how can gossip help to solve some of these problems of co-operation and conflict in the context of group living? Again, the problems of group living, which I outlined above, are numerous, and differ from each other. One single mechanism cannot solve all of these different kinds of problems. Again, I distinguish different kinds of gossip that each could have solved one of these specific problems. Social selection pressures, in combination with natural and sexual selection pressures have shaped our interest in different kinds of gossip, because this information was as important to our ancestors as food. Without this kind of information they were not able to solve certain problems, which decreased their survival and reproduction options.

To solve the above-described problems, I suggest the following different kinds of gossip might have offered a solution to these problems: Kin Structure Reputation Gossip, Co-operation Reputation Gossip, Alliance Reputation Gossip and lastly, Calibration Reputation Gossip. I will explain these more in detail, and also show that further classifications must be made for some of these divisions. Overall I classify these different kinds of gossip as Social Reputation Gossip (Social RG), because they all function to solve the problems of being social and, therefore, problems of group living. I end this fourth section with a special note on the importance of gossip regarding parental skills, which will be present in many of the different kinds of gossip, as I have defined them, throughout this chapter.

A first important type of Social RG is Kin Structure RG. We all talk about who is related to whom, because this kind of information is of value to us. As Hamilton (1964) outlined with his inclusive fitness theory, we benefit from the successes of kin related people and likewise, unsuccessful actions of these relatives rebound to own losses in inclusive fitness. From the perspectives of genes (through which selection operates) knowing who is related to who is valuable, because the actions of others influence our own and others’ inclusive fitness. That we are aware of this is nicely referred to by Merry (1984: 281) who said that: “Because gossip and scandal tarnish the name of an entire family, kinsmen are anxious to prevent family members from behaving in a dishonorable or immodest fashion.”

Daly, Salmon and Wilson (1997: 283) said that: “In all societies, people are motivated to inquire how strangers and new acquaintances might be genealogically linked to people they already know, and feel that they have acquired useful social information when such links are uncovered.” Kin Structure RG is a tool to detect kin relatedness. It is a valuable tool, since kin relatedness is not outwardly visible, (even though we might rely on cues such as co-residence, odor similarity and facial resemblance) we are often unaware or uncertain about who is related to whom.

The division I label Co-operation Reputation Gossip is gossip that functions to label specific persons with a reputation of being a cheater or an altruist. It concerns information about who obeys and who violates social contracts. Social contracts are important to maintain the bonds between individuals; free-riders and cheaters threaten the bonds between individuals. As I explained earlier, co-operation in larger social settings requires reputation knowledge, and I already stressed the importance of communication in the construct and manipulation of these reputations. Co-operation RG is an ideal tool to construct reputations about co-operators and enables co-operation in larger social settings.

Again gossipees of Co-operation RG are individuals with whom the gossipers can have future potential encounters. Gossipees of Co-operation RG are directly or indirectly known to the gossipers. Totally unknown gossipees with whom future encounters are not possible are not subjects of Co-operation RG.

Because it is both important to label specific people with the reputation of ‘cheater’ (with whom you will not want to co-operate in the future) and ‘altruist’ (with whom you will want to co-operate in the future), I further narrow down Co-operation Reputation Gossip to Cheater Detection RG and Altruist Detection RG.

Cheater Detection RG is information about the deceitful behavior of a specific third person, with the goal to label this person with the negative reputation of being a cheater. Cheaters are people who violate social contracts, and need to be punished for their behavior to keep co-operation in a social group possible. Spreading around such information will make future cheating actions of this person less easy to perform.

Examples of Cheater Detection RG are: “Lisa is so unreliable; she always cancels dates at the very last moment”, “Robin promised to help me paint my kitchen last week and never showed up”, “Have you heard that everyone except Lauren’s brother came to help when she moved into her new apartment?” etc.

Altruist Detection RG is information about the altruistic behavior of a specific third person with the goal to label this person with the positive reputation as an altruist. Spreading around such information will make future co-operational actions easier to establish for this gossipee.

Examples of Altruist Detection RG are: “Marie is such a good host, she always spoils her guests”, “Tom is a very helpful person; whenever you need help you can count on him”, “If Brad had not shown up to help me clean the house after the party last week, I would have been in trouble!” etc.

“Have you heard, Robin and Nancy waited for Tom to pick them up and bring them to the airport, but Tom never showed up?! They finally took a taxi, but arrived too late, missed their flight and were too late to attend his brother’s wedding!”

Both examples concern the violation of a social contract. Spreading around this information about these gossipees affects their reputation. However, the first violation is more important than the second one. Both Tom and Sophie violated a social contract, but Tom’s behavior clearly harmed other people, while Sophie’s behavior seems harmless. What Sophie did was a violation of a social contract that is embedded in many official law systems; drinking at a young age. Her ‘cheating’ behavior did not harm other people, but potentially could have harmed herself or others.

It is important to keep such differences in mind. Cheating is variable and can occur in many different degrees. The worst degrees of cheating are actions were other people have suffered from the cheater’s behavior (as in Tom’s case). We can label this ‘actual cheating’. A lesser degree of cheating concerns potentially harmful actions, where the cheater violated a social contract, but no others were harmed (as in Sophie’s case). These forms of cheating are violation of official laws without effectively harming others (e.g. drinking at the age of 14, or taking drugs without any consequences due to your actions). I label these as ‘potential cheating’, since there is potential to harm others. The importance of differentiating between these two degrees of cheating will become clearer in the next chapter, where I focus on the costs and benefits of sharing and acquiring Cheater Detection RG.

Similar arguments can be made for Altruist Detection RG. People can clearly benefit other people (actual altruists), and people can potentially benefit other people (potential altruists). To illustrate both, imagine that Lucy sees a little boy falling in the lake and the boy cannot swim. She jumps in the water and saves his life. This will be spread around, and I label this as Altruist Detection RG about an ‘actual altruist’. Her actions clearly benefited a third person (the boy). Now consider that she jumps in the water, but unfortunately too late, the boy has already drowned. Lucy’s behavior here did not result in clear benefits to third parties, but the fact that her behavior potentially could have saved the boy’s is the relevant information. She will still be thanked for her effort, and still get a reputation increase for acting altruistic. This is Altruist Detection RG about a ‘potential altruist’.

Another comment I would like to make here is that being labeled as a ‘cheater’ or an ‘altruist’ for one for your past actions does not make you a cheater or an altruist in general. Josh, being a cheater on the road because he drives too fast, might be a really helpful person towards his friends and family when they ask for his help, which makes him altruistic. Lucy, although she jumped into the water to save the life of a young boy, might have a bad reputation for breaking promises to lend a helping hand when needed.

Being labeled as a cheater or an altruist in a specific situation will make future co-operation in a similar situation more or less difficult to establish. However, this does not mean that all of your future co-operative actions will be similarly influenced.

As I mentioned above in section 2.4.3.3, Co-operation RG is closely related to Social SLG, but I want to stress again that both have clearly different functions. Social SLG functions to exchange fitness-relevant behavior strategies; senders can teach receivers how to behave, and receivers learn what is right and what is inappropriate behavior through the actions of others. Subjects of Social SLG are mere carriers of this information.

For Co-operation RG, gossipees are the focus. They get labeled with a reputation of being a cheater or being an altruist. The function of Co-operation RG is to construct and manipulate the reputations of third parties as co-operators, and therefore influencing the future co-operational opportunities of the gossipees. Co-operation RG also teaches us fitness-relevant information; we learn who is a cheater and who is an altruist. However, notice that Social SLG teaches us about behavior strategies, while Co-operation RG teaches us about specific others. The difference between both will become clearer in the next chapter, where I focus on the costs and benefits of exchanging both forms of gossip.

Both Social SLG and Co-operation RG can be present in the same piece of information and the presence of the latter depends on the relationship between gossipee and gossipers. Only if they know each other, either directly or indirectly (through mutual acquaintances) can information be labeled as Co-operation RG. Labeling unknown persons, with whom we do not interact and with whom we do not co-operate, as cheaters or altruists has no value, and is a waste of effort.

Where as Co-operation RG concerns information about one-shot single events of co-operational transactions of individuals, Alliance RG is information about the strong, long-lasting bonds between individuals. As I explained above, alliances are very important in the context of group living and conflict. In our evolutionary past our ancestors faced the problems of finding good allies and maintaining bonds with them. In my opinion, four kinds of gossip information might have functioned to solve these problems: Ally Detection RG about the skills of others, Ally Structure RG to learn who is allied to whom, Ally Maintenance RG to maintain the social structure, and Calibration Reputation Gossip to track changes in the traits/behaviours of our social network members.

4.4.3.1 Individual seeks allies: Ally Detection Reputation Gossip and Ally Structure Reputation Gossip

First of all, we need to look for coalition partners; we search for friends, both for true friends and fair-weather friends. As Tooby and Cosmides (1996) have said, we look at the skills of people to decide whether to bond with these people as friends or not. This is not the only criterion, but it is an important one.

I label Ally-detection Gossip as information about the skills of specific third persons. This form of information can be a useful tool in our search for allies. As mentioned above, we seek friends and allies to fill in niches of our social network. Although this often happens unconsciously, we become friends and form coalitions with people that in some way can be of help to us. For instance, when thinking about building a house, you might be more tuned to find someone who is a skilled architect or a skilled interior designer, and become allied with these people. When you are raising children, your attention will be focused on people with good parenting skills, who might be able to give you a hand when needed. When you start studying, you will be interested to find senior students with good grades who might be able to give you some useful advice, and so on.

Your needs will define with whom you will spend most of your time at specific moments. Once your house is built, your architect and designer friends might become less important, and allegiance with them may fade away again. Ally-detection Gossip serves to inform us about the skills of people we do not know yet, and who might be worthwhile allies to us. Ally-detection gossip also reminds and updates us about the skills of others we know already, and with this information we choose with whom to (temporarily) tighten or loosen our bonds.

In our search for allies we often rely on ethnic markers. Examples of ethnic markers are language, culturally transmitted beliefs and behavior traits of an individual. When we select individuals with whom we want to co-operate, say McElreath, Boyd and Richerson (2003) we select them on the basis of markers. We choose partners with somewhat similar traits and attitudes. Ally Detection Reputation Gossip therefore not only focuses on the skills of others that can be of use to us, but also on the markers of others, and more specifically on markers that are similar to the markers of the gossipers or markers that are different from what the gossipers value. In the latter case Ally Detection RG functions to detect non-allies.

In our search and maintenance of allies, it is important to know who is allied to whom. Krebs and Denton (1997) have suggested that we should be able to make quick and correct judgments about others. As they said, we tend to view in-group members more positively and out-group members more negatively. Gossip about the categorization of other individuals might have been beneficial to our ancestors to enable quick judgments. Information about who can be considered a friend or an enemy could have been life-saving information. Knowing that an unknown person is allied to one of our friends can help us make quick estimations to decide if this person will become our friend or foe.

Ally-structure Gossip reveals the structure of our social network, gives information about who is allied to whom, and can help us to easily classify new people as friends or foes, based on their bonds with people we do know.

Once we have established social contacts, we need to maintain our bonds. Gossip again can be an ideal tool in this context. By spreading good reputation gossip about our allies, and bad reputation gossip about our non-allies, we clearly outline the boundaries of our social network structure. I label this good/bad-reputation manipulative form of gossip Ally Maintenance RG.

Gossipees of Ally Maintenance RG are individuals who are part of our social network. We know them directly, because of past encounters, or we know them indirectly through an intermediary person, so that future encounters can be secured. There is an additional restriction for gossipees of Ally Maintenance RG. We spread positive gossip about friends and negative gossip about foes, but what about neutral others? Neutral others are not gossipees of Ally Maintenance RG.

Ally Maintenance RG has another beneficial effect; praising allies increases the status of the individuals that we are associated with, and in the end, returns as individual benefits. Similarly, if we lower the social status of non-allies, our relative status position increases as well, which again returns as individual benefits. Ben-Ze’ev (1994) illustrated this functioning of gossip, when he said:

“In addition to the respect gained by conveying interesting behind-the-scenes information, gossipmongers, who often have a low self-image, may gain some respect by conveying information which is slightly damaging to others. Lowering the evaluations of others may seem to them to somehow increase their own.” (Ben-Ze’ev, 1994: 19)

This evolutionary view on Ally Maintenance RG is not new; McAndrew and Milenkovic (2002) already argued that from an evolutionary perspective, individuals should be eager to spread good gossip about allies and bad gossip about non-allies. Negative information, that can decrease the reputation of another individual, is interesting to use when concerning higher status people or rivals, they say. Lowering the status of people who are already lower status than us makes no sense. On the other hand, when elevating the status of allies one can relatively increase his or her own status, they argued.

To test their assumptions, McAndrew and Milenkovic (2002) surveyed 83 respondents (39 males, 44 females, ranging in age from 17 to 22) about how likely they would spread around good/bad information about friends, relatives, acquaintances, professors, or strangers. The results of their survey shows that individuals are, indeed, more likely to pass along negative gossip about potential adversaries (strangers and powerful others) and to not pass along negative information about allies (friends and relatives). Positive gossip was likely to be spread around when concerning allies and not spread around when the subjects were non-allies.

“People actively seek information about others that will be most useful in social competition. We seek exploitable, damaging information about high-status people and non-allies; we actively disperse status-enhancing information about our allies; and we keep a very watchful eye on our friends.” (McAndrew & Milenkovic, 2002: 17)

My comment on their research is that I do not go along with the idea that we are simply more eager to spread bad gossip about higher status people. Whether we will spread negative information about others depends on whether these people are allies or non-allies, in my opinion. A higher status person who is allied to us will not become the target of our slandering gossip conversations. Since we can bask in the glory of being connected to this person, it is far more beneficial for an individual to praise such higher status allies.

Further, the results from McAndrew and Milenkovic (2002) only support part of the Ally Maintenance RG predictions. They proofed that we are indeed more prone to spread good news about allies, but their results about ‘non-allies’ should be taken with some skepticism. First of all, professors are not always non-allies; some students might look up to their professors, and regard them as allies. Secondly, strangers are not necessarily non-allies. If we do not know them yet, we might not have decided yet if these people are friends or foes, so Ally Maintenance RG might not come into play at this point. I also suggest that the ‘good’ and ‘bad’ information they used should be tested. For instance, drug abuse was used as ‘bad’ gossip, but among students (their research population) using drugs might not necessarily have such a negative connotation as it has among adults.

We gossip about members of our social networks in order to better understand them (Goldsmith & Baxter, 1996). Unexpected changes in someone’s behavior pattern are important for our long-term co-operational interactions with others. When we interact with others, we try to predict the behavior of other people. As I will argue in chapter 6, we ‘mind-read’ other people to predict their actions, and act on these predictions. Knowing that a friend of ours gets easily hurt, when we comment on his or her doings, we might refrain from giving him or her comments, because we have learned from past experiences that this ends badly. We uphold ideas about all the people we live with; we form images, based on our past interactions. How others see us, and how we see others defines our and their reputations (Bromley, 1993).

What I label ‘Calibration Reputation Gossip’ functions to calibrate our image of the other people we interact with. Calibration of someone’s image is necessary when he or she deviates from the average others, or when he or she deviates from himself or herself. I will explain this a little more detailed.

What I label ‘Other Deviance Calibration RG’ concerns information about the traits and behaviors of a specific other person that deviates from what an average other person would be like or do. If all men wear pants and Luke wears skirts, he deviates from the average, norm in his group. Gossip about his strange behavior is Other Deviance Calibration RG. Other Deviance Calibration RG concerns violations of social etiquettes; behaving nicely, shaking hands, eating with knife and fork (in some cultures) and so on.

You could argue that Luke ‘cheats’ on this norm and should be labelled a ‘cheater’. Still, Luke’s behavior does not effectively nor potentially harm other people. Other Deviance Calibration RG is different form Cheater Detection RG because the latter functions to label gossipees as cheaters to decrease their co-operation opportunities, while the first merely labels a gossipee as ‘different’. Being merely different from the norm without causing harm does not require that your co-operation opportunities should be restricted. Other Deviance Calibration RG does not affect our co-operation opportunities, which makes it different from Cheater Detection RG. The other difference between these two is that Cheater Detection RG labels gossipees with a bad reputation of ‘cheater’, while Other Deviance Calibration RG labels a gossipee as ‘different’. Being different can decrease your status, but in the eyes of some people can also increase your status.

While Other Deviance Calibration RG focuses on individuals who stray from the norm, Self Deviance Calibration RG concerns information about individuals who suddenly deviate from their own regular pattern.

When someone changes some of his or her characteristics we need to change our view of this person, and recalibrate the beliefs we uphold about this person. It is important to do this, in order to predict the actions of this person in our interactions with them. Because we cannot constantly keep track of every member or our social network, gossip can be an ideal tool to adjust or calibrate the images of our social network members. If you know Bob as a person who hates traveling, then hear that he has traveled twice in the last month and now loves to travel, is useful to adjust your image of him. And next time you look for a travel companion, you might even consider asking him to join you. Self Deviance Calibration RG is information about the changes in the regular characteristics and behavior patterns of our social network members. It functions to adjust our images, so that we can correctly predict the actions of our social network members. Self Deviance Calibration RG informs us about changes in the appearance of social network members, changes in their regular habits and so on. Gossipees of Self Deviance Calibration RG do not violate social norms, but they violate their own habits. They suddenly do something you could not have predicted.

For instance, consider this bit of gossip: “Nancy went to the hairdresser”. If Nancy goes to the hairdresser every week, chances are low that her social network members will exchange this information about her. However, if Nancy hardly ever goes to the hairdresser, chances rise that this piece of information will be shared. This piece of information then becomes Self Deviance Calibration RG. Remember that in chapter 1 I explained that gossip, in the most general sense, must be restricted to information that surprises us or that is deviant from what an average person does. For Self Deviance Calibration RG the surprise factor is important; the behavior of Nancy is not deviant from what other average people might do, but it surprises us because she suddenly behaves differently from her standard behavior pattern.

People who very often change their image, therefore, are more gossiped about. If your actions cannot be predicted, people will be more eager to exchange information about what you do, to somehow get an understanding of you and increase their ability to predict your actions.

To end my section on Alliance RG, let me point out again that this form of gossip is very similar to Intrasexual Conflict RG, which I discussed in section 3.6.2 of this chapter. Both forms of gossip concern the detection and manipulation of others. The difference between both is that we seek sexual rivals when using Intrasexual Conflict RG and both allies and non-allies by using Alliance RG. Alliance RG has a broader spectrum; it concerns friends and foes, while Intrasexual Conflict RG only concerns our rivals. Similarly with Ally Maintenance RG we both praise our friends and slander our foes, while Sexual Rival Slander RG, as the name itself says, only focuses on slandering rivals.

Before turning to sex differences in Social RG, I want to make one final remark on all different kinds of Alliance RG. Although most gossip is about one specific gossipee, we do gossip about multiple gossipees as well. Especially Alliance RG will not always be one-person-directed, but will very often take the form of judging groups of people. “They are the enemy”, “We are good vs. they are bad” forms of utterances can be expected under Alliance RG.

I have not yet discussed whether or not we can expect sex differences to occur in the context of Social RG. Generally speaking, I clearly expect to find a difference in the tendency to engage in Social RG. More specifically, I predict that women have a stronger tendency to exchange Social RG. I do not focus on sex differences in the specific content of different kinds of Social RG. I discussed sex differences in gossip content quite extensively in section 3.6 of this chapter, because in the context of Mating RG clear predictions can be made about sex differences in gossip content. For Social RG, just like for Strategy Learning Gossip, sex differences in content have much less at stake.

In section 3.6.2.2 of this chapter, I already explained that women find gossip to be a more effective tool than men, because it is more costly for to battle conflicts with physical aggression. In chapter 3, I explained that patrilocation occurred throughout our evolutionary past. Women migrated to a new social group, which means that they lived with more non-kin related others than male ancestors did. Therefore, our female ancestors have faced more problems of dealing with new social network members. Learning who is related to whom, who is a cheater and who is an altruist, who has skills we value, and who is allied to whom were problems our female ancestors have faced more than our male ancestors.

All of these situations concern problems of ‘finding about who surrounds you’. Some of the forms of Social RG focus on teaching us information about specific others with whom we –or allies of ours- interact, or might interact with in the future. Kin Structure RG, Co-operation RG, Ally Detection RG and Ally Structure RG all teach us about the reputation of specific others. I therefore argue that due to sex differences in our evolutionary past, women today still show a stronger interest in these forms of detective Social RG.

Support for the idea that women should have a stronger interest in Ally Maintenance RG comes from Hess and Hagen (2002). Arguing that kin should compete less than non-kin, Hess and Hagen (2002) suppose that women, more than men, were exposed to within-group competition throughout our evolutionary past (also due to the patrilocality effect). Hess and Hagen (2002) put forward a theory on how gossip can be used as an aggression strategy to battle within-group conflicts. Since women, more than men, were exposed to such conflict situations, they argue that women are more specialized than men to use gossip in a manipulative sense.

Hess and Hagen (2002) experimentally tested some of their assumptions on how gossip is used as a manipulative strategy in within-group competitions. Their results can be summarized as follows: First of all, negative gossip is more likely to be distributed when the gossiped subject is competitive with the gossiper. This conclusion is in line with McAndrew & Milenkovic’s research (2002) I mentioned earlier. Next, they found that having friends protects women from negative gossip, and being a member of a coalition decreases the possibility that someone will spread negative information about you.

Hess and Hagen (2002) summarized their arguments why it is beneficial for women to uphold strong ties (allies) with non-kin. First, coalitions provide more eyes and ears to collect information; second, coalitions can more thoroughly analyze this information; third, Coalitions offer more vectors to disseminate information; fourth, gossip, reported by more than one person, might be more believable; fifth and lastly, coalitions can protect individuals by providing alibis and by providing evidence against accusations.

Because these strong ties with non-kin are so important to maintain, women might benefit more from using Ally Maintenance RG. Since our female ancestors faced problems of upholding ties with non-kin more than our male ancestors, we can expect women today to show a stronger tendency to exchange Ally Maintenance RG as well.

The final form of Social RG is Calibration RG. I do not expect sex differences to occur in the tendency to exchange such information. Both our male and female ancestors benefited equally from getting updated information about the changes in behavior patterns of their band members. Also, in regards to content, I do not expect sex differences to occur.

I mentioned above that being parents involves specific problems of willingness to invest between parent-and-child and siblings. Because parental skills are so important to secure the survival of offspring, I expect this to be a much gossiped about topic. It is important to learn how to be a good parent. In the EEA, raising offspring occurred in co-operation with other women (Caporael & Baron, 1997), therefore it was important for our ancestors to know about the parenting skills of specific members of their band. Since inclusive fitness explains that the fitness of related members rebounds to benefits to our own inclusive fitness, we will all be concerned about the parental skills of relatives.

For instance, I expect Ally Detection RG to discuss parental skills. If you have children of your own, or a family member or a friend of yours has children, you might be particularly interested in the parental skills of others, to become allied with these people.

Social RG will deal with this kind of information, but, as I also mentioned, parental skills will be the subject of Mating RG, since raising offspring is the result from mating strategies (see above 3.6.4). Gossip about parental skills can also instruct the gossipers how to be better caregivers, which I discussed in section two and classified under Mating SLG. However, Mating Gossip and Social Gossip about parental skills have a different goal; in this context, gossip about parental skills is used to label specific persons (gossipees) with the reputation of being a good or a bad parent.

Mates Detection RG will include parental skills because information about who is a good or a bad parent can be of value in the search for potential mates. Parental skills can also be discussed in Intrasexual Conflict RG; gossiping about the bad parental skills of rivals can lower their reputation as a good mate.

To end this overview of all the different types of gossip, and before presenting a concluding overview, I want to focus on the fact that using gossip nowadays might be more complex than gossip in the EEA.

As Tooby and Devore (1987) have suggested, our ancestors most probably consisted of small bands of people co-operating with each other and being in conflict with other small bands of people. Bands were fairly stable, social structures and were clear and fairly easy to understand. Our ancestors knew the people they met daily; these were their band members and most of these were even kin related others. Bands moved and also met other bands, but this in a context of clear and simple social structures. Therefore it probably was fairly easy for our ancestors to discriminate between friend and foe. Ally-maintenance Gossip was easy to use, since they most probably knew who to gossip about in a nice way and who to disgrace.

I think that the context we live in today is far more complex than the hunter-gatherer societies our ancestors lived in, which constitute the context wherein cognitive mechanisms for judgment-making evolved. As Krebs and Denton (1997) suggested, our ability to make judgments about others might be somewhat flexible. However, I think that this flexibility might not be sufficient enough to operate beneficially in our current complex societies. Nowadays we do not live in one band of social network members; we now have our family, our friends, our colleagues, the people from the gym, the people from hobby clubs, the pals from soccer, and so on. If this is not complex enough, we also have to keep in mind that some of our friends are befriended with people we dislike; our colleagues might be our worst enemies, or might be befriended with our worst enemies and so on.

What I want to stress is that Western Societies nowadays are more complex in social structure than the societies our ancestors lived in. The context wherein gossip emerged and evolved is different from our complex social networks of today and, in my opinion, this makes gossip more difficult to use nowadays, at least with respect to some specific kinds of gossip I distinguished for. I think this is especially true for Ally Maintenance RG. We might know for ourselves who is our friend and who is our foe, but nowadays it is less clear how these friends and foes are regarded by other social network members. Slandering those we dislike increases our status if we share this Ally Maintenance RG with others who also dislike these people. However, talking bad about others’ friends can turn against us, instead of being beneficial.

Other forms of gossip I made a distinction for might have increased in value because of our more complex and scattered social networks. For instance, Kin Structure RG and Ally Structure RG might have more value nowadays to learn who is related to whom and who is allied to whom.

Still, these are mere speculations. It might be useful to test in the future how many different forms of gossip, as I have defined them, are present in modern indigenous societies, who most closely resemble the EEA, and compare these results with similar studies in our Western societies.

Because I outlined the different types of gossip spread over different theoretical parts of this chapter, I feel it would be useful to summarize them again. Here, I review the classifications, functions and effects of gossip, and frame them in a classification tree for the use of future researches of gossip. Before outlining the different steps of classification, I will link the above-described functions of gossip to the effects of gossip I outlined in chapter 2.

All the functions I have now outlined result in effects when they are used. More specifically, they result in effects that have been described by standard social scientists, as I have explained in chapter 2. To show you how the functions I have outlined give a better understanding of gossip than the effect-study of chapter 2, let me explain how effects and functions are related to each other. Effects offer proximate explanations of how gossip operates, while the functions give a more detailed and why-focused explanation. In chapter 2, I summarized the effect of gossip on ‘learning’, ‘control’, ‘manipulation’, and ‘entertainment’. I here explain how different functions underlie these effects.

In chapter 2 I outlined how gossip is used to gain information about the behavior of others (chapter 2 section 3.2.1) and to compare our behavior to that of others (chapter 2 section 4.2.2): the learning effect of gossip.

The underlying functions of the learning effect are mainly explained with Strategy Learning Gossip. As I have suggested, we benefit from learning about the actions of others because own trial-and-error experiences can be too costly. SLG offers cheap, fitness-relevant information, and teaches us how to behave and how not to behave.

However, learning is not restricted to SLG, it occurs through Reputation Gossip (RG) as well. What I have stressed is that SLG teaches us about behavior strategies, while RG teaches us about specific persons. Some forms of Reputation Gossip explain why learning about specific persons is valuable; to get information about the reputation of potential mates (Mates Detection RG), to learn who is related to whom by kinship or allies (Kin Structure RG and Ally Structure RG) and to learn about changes in someone’s behavior pattern (Calibration RG). A special form of learning concerns Co-operation RG, which I mainly link to the control effect that social scientists often attribute to gossip.

Numerous researchers have stressed the control effect of gossip (for an overview see chapter 2 section 3.2.3). I mentioned two important reasons why gossip functions to control others; we benefit from controlling our mates and the mates of allies (Mates Control RG) and we benefit from controlling others’ fulfilment of social contracts. Co-operation RG explains how receivers can learn who is a cheater and who is an altruist and how senders can punish cheaters and praise altruists with this kind of information. The reason why gossip functions to control others is because individuals benefit from co-operation and group living (as explained in section 4 of this chapter) and cheaters threaten social bonds established through social contracts while altruists should be awarded for their co-operative actions.

In chapter 2 I outlined how the control effect of gossip is linked to a bonding effect (chapter 2 section 3.2.2). Co-operation RG is only a partial explanation of why this happens. Further functional explanations come from Alliance RG, and more specificly from Ally Maintenance RG. Praising our allies and attacking non-allies maintains the structure of our social network and results in return individual benefits of (relative) increase of social status. This happens because having a high social status leads to numerous benefits, as I have outlined in section 4 of this chapter.

The use of Co-operation RG and Ally Maintenance RG both result in a manipulative effect of gossip, which has also been recognized by other social scientists (see chapter 2 section 4.2.1). Besides these two forms of Reputation Gossip, Intrasexual Conflict RG as well is used to manipulate others and, in this context of rivalry mating, preferably in a negative sense.

However, we not only manipulate other people by directly acting on their reputation, as Reputation Gossip explains. When we share SLG with others, we also manipulate the knowledge of other people. In the next chapter, where I focus on the behavioral models for SLG and RG, this will become clearer.

A last effect, which social scientists have mentioned and I have discussed in section 4.2.3 of chapter 2, is the entertainment effect of gossip. I have not specified one specific kind of gossip that has an entertaining function for the simple reason that all kinds of gossip can be entertaining. We love to learn from the behaviors of others; we love to praise our friends and slander our foes; we love to hear who is a cheater and who acts altruistic, and so on. All gossip is possible entertainment, and framing this in an evolutionary perspective indicates that the use of gossip might have adaptive outcomes. As Barkow (1989, 1992) has stressed, we love to gossip for the same reason we love food; because it was beneficial for our ancestors and helped them in their struggle for existence and reproduction.

An important overall rule that should be kept in mind is that the classifications I propose can co-occur in one piece of gossip. What I want to stress is that one gossip story can have several functions. One piece of gossip can even have different functions for different senders and receivers. Let me explain.

In my classification of gossip, the first step divides Strategy Learning Gossip from Reputation gossip, based on the focus on behavioral information and/or gossipee. This first level of classification is the most important one to understand. SLG is information where the gossipee can be replaced with any other subject, and the value of SLG will not change when you do so. RG is information about a very specific person, and replacing the gossipee with another subject takes away the value of RG (it can change into RG about another gossipee or lose complete value).

Whether information can be classified as RG or not depends on the relationship between gossipers and gossipee. If gossipers and gossipees know each other directly or indirectly (which means someone they know knows the gossipee), information can be Reputation Gossip. If gossipers and gossipees do not know each other (in)directly gossip information cannot be RG. The relationship between gossipers and gossipees is a key point for all different kinds of RG. How the gossipers and gossipees are related to each other will define how gossip information can be further labelled into different RG divisions. If gossiper and gossipee are rivals, a gossip story can be Sexual-rival-slander Gossip, and if they are friends the same piece of information can be Ally-maintenance Gossip and so on.

To further classify SLG and RG, I looked at the different adaptive problems our ancestors faced and to which gossip might have offered a solution. For an overview of these different adaptive problems I refer to the graphic outline of my classification tree, and for details about these problems to the previous sections of this chapter.

I have divided SLG into three sub-categories: Survival SLG, which concerns information about life and death; Mating SLG, from which we can learn how to become successful lovers; and Social SLG, which informs us about rights and wrongs according to the norms that govern the group where we are acting out strategies.

Because the further narrowed-down structure of Reputation Gossip is more complex, I will explain in more detail in the next section.

Throughout this chapter I have framed different kinds of Reputation Gossip in the context of Mating and general Group Living. To give a clear overview of how the different kinds of Reputation Gossip can be similar to each other, I structure them here in a different framework.

To begin, Reputation Gossip is gossip about specific gossipees. It concerns the traits and behaviors attached to a certain person, what we can call reputations (Bromley, 1993). It is important to keep in mind that Reputation Gossip is only valuable for gossipers who know the gossipee directly or indirectly (through others). By living among other individuals we know, we can divide our social network in ‘mates’ (including potential mates), ‘kin-related others’ (family), ‘allies’ and ‘non-allies’. For all these different ‘other individuals’ information about ‘who they are’ and ‘how they are related to each other’ is beneficial for keeping up ties with these people.

In section 3 and 4 of this chapter, I defined three different kinds Reputation Gossip: Detection RG, Structure RG and Maintenance RG. If I multiply these with the three different kinds of social network members (mates, family, allies/non-allies), I should come to 9 different kinds of Reputation Gossip. But, if you refer to table IV.1 below, you will see that three of them are missing; Kin Detection RG, Mates Maintenance RG and Family Maintenance RG.

Table IV.1. An overview of the different kinds of Detection, Structure and Maintenance Reputation Gossip forms for mates, family members and others as gossipees

Let me explain why some are ‘missing’. Starting with Detection RG, this kind of gossip concerns information that can help us solve the problem “Who is qualified to be my ‘x’”. And ‘x’ could then be ‘mate’ ‘family’, ‘ally’ or ‘non-ally’. We look for skills and reputations of people when we search for mates and friends; we want good mates and valuable friends. But we do not search for family members. We cannot choose kin-related others; they are either kin-related or they are not, and there is nothing we can do about this. Kin Detection RG does not make any sense.

You might comment ‘yes, but we search for relatives, also by using gossip’. Indeed, we gossip a lot about who is related to whom, searching for people that might be related. But this is embedded within Kin Structure RG. Structure RG concern connections between people, and you need at least two to establish a connection. Detection RG can be about one gossipee, whereas Structure RG is applicable to mates, family and others; connections can take the form of sexual relations, kin relations or social relations.

Lastly, we need to maintain our social structure. Why did I not label three kinds of Maintenance RG? Because for all three categories of people (mates, family and allies/non-allies) the same gossip-rule applies in order to maintain bonds: spread good gossip about your loved ones and bad gossip about your foes. Family members can be our loved ones, but can also be our enemies at times. Mates are our loved ones, but can become foes or non-allies in the future. Ally Maintenance RG covers all information that praises our loved ones (allies) and slanders ours foes (non-allies that we dislike). This term covers all forms of Maintenance RG and embodies all different categories of social network members. As I will explain in the next chapter, when using this form of gossip it is important to pay attention to whether an ally or non-ally is kin-related to us or not, before we exchange Ally Maintenance RG about them, because this can result in extra benefits or potential costs (see chapter 5).

Besides the six types of Reputation Gossip that I have discussed so far, I defined the following other types: Sexual Rival Detection RG, Sexual Rival Slander RG, Cheater Detection RG, Altruist Detection RG, Mates Control RG and Calibration RG. I will discuss these now.

Sexual Rival Detection RG and Sexual Rival Slander RG are both forms of Mating RG and special forms of Detection RG and Maintenance RG. The first concerns information about who is our competitor in our search for potential mates. Sexual rivals are non-allies, but I decided to label this form of gossip separately, since it has a very specific function: to detect sexual rivals, who are still significantly different from other foes. Further, a sexual rival can actually be one of our friends (i.e. allies) as well. Similarly, Sexual Rival Slander RG is related to Ally Maintenance RG but differs in a key way. As the name indicates, the first only concerns bad information meant to slander competitors. The latter is both good and bad gossip, and is a form of Social RG meant to deal with problems of group living in a non-mating context.

Cheater Detection RG, Altruist Detection RG and Calibration RG are also mentioned separately, because they are all applicable to all members of our social network: mates, family, allies and non-allies. For all individuals we interact with, or people we learn about through a third party (indirect interactions), it is important to know whether the gossipee is a cheater or altruist and if he/she has changed his/her regular behavior pattern. We must be able to predict the behaviors or everyone we interact with, therefore Calibration RG is valuable for every gossipee with whom we interact, or with whom others we know interact. The same is true for Cheater Detection RG and Altruist Detection RG, whether someone is a mate, friend, family member, or other person, it is important to know the reputation of these people as co-operators, in order to help our decision-making (co-operate or not) in future interactions with these people.

To end my overview, I mention Mates Control RG as a special form of Cheater Detection RG. It concerns information about the cheating behavior of our mates, or the mates of people we know. This form of gossip functions to detect cheating in the context of mating, which is different from cheating in a social non-mating context.

To give an overview of all different kinds of Reputation Gossip, I refer to the graphic classification tree, attached at the end of this chapter.

“Information about the deviant or surprising (which both depend on the context) traits and behaviors of a (or more) third person(s) (most often non-present, but potentially present in the conversation), where the goal of the conversation is non-professional, and the sender has true/false knowledge of the gossip content.” (see chapter 1)

This definition is too general to operationalize for research. A classification is required.

Making use of an effect-study and an evolutionary functional analysis, I propose following classification system:

Person focus Gossip

If you replace the subject of the gossip story with any other subject, the value of the message changes, because the value is totally subject-dependent.

Reputation Gossip (RG)

Behavior and traits here are attached to a specific person. Therefore this is Reputation Gossip (RG), since reputations are the behaviors and traits of a specific person. RG functions to learn about specific persons and manipulate your own and others’ reputation.

To understand RG, knowledge about the relation between gossiper and gossipee is necessary.

Mating Reputation Gossip

Social Reputation Gossip

Problem:

Finding mates

Problem:

Intrasexual rivals

Problem:

Keeping mates

Problems:

Social contracts

Problem:

Who is kin-related

Problems:

Long-term co-operation and conflict within and between groups

Mates-detection RG

Mating-structure RG

Intrasexual Conflict RG

Mates-control RG

Co-operation RG

Kin Structure RG

Alliance RG

Calibration RG

Sexual-rival-detection RG

Sexual-rival-slander RG

Cheater-detection RG

Altruist-detection RG

Ally

Detect-ion RG

(skills)

Ally

Structure RG

Ally maintenance RG

Other

Deviance

Calibration RG

Self

Deviance Calibration

[3] The ability of humans to attribute mental states to other people and interpret their action is what Simon Baron-Cohen calls ‘to mindread’ (Baron-Cohen, 1994). Autistic people lack ToM. I discuss this more in detail in chapter 6, paragraph 2.1.1.2 and 3.1.4.

[4]I explain the term ‘Inclusive Fitness’ in section 4.1.2.1.

[5] Here I only discuss mating strategies and the impact of Mating Gossip in the context of heterosexual relations. Looking at homosexual relations and the impact of Mating Gossip in this context is an interesting body of research as well, but due to time constraints, I have not (yet) incorporate this here. I encourage future research on this subject.

	Mates	Family	Others (allies & Non-allies)
Detection RG	Mates-detection Gossip	x	Ally-detection Gossip
Structure RG	Mating-structure Gossip	Kin-structure Gossip	Ally-structure Gossip
Maintenance RG	x	x	Ally-maintenance Gossip

Gossip

Focus on Behavior The focus is on the transmitted behavioral information			Focus on Subject The focus is on the gossipee

Subject Independent gossip Replacing the gossipee with any other subject does not affect the value of the message			Subject Dependent Gossip Replacing the gossipee with any other subject does affect the value of the message

Behavior Focus Gossip If you replace the subject of the gossip story with any other subject, the value of the message is not decreased, because the value is totally behavior-dependent and almost subject-independent.

Strategy Learning Gossip (SLG) Instead of investing cost, time and energy yourself to increase your experiences, you profit from the experiences of others

Problems of life and death strategies	Problems of how to attract, keep and guard mates	Problems of behaving appropriately in interactions with others

Survival Strategy Learning Gossip	Mating Strategy Learning Gossip	Social Strategy Learning Gossip