Conclusion

“Although I am fully convinced of the truth of the views given in this volume I by no means expect to convince experienced naturalists whose minds are stocked with a multitude of facts all viewed, during a long course of years, from a point of view directly opposite to mine. But I look with confidence to the future to young and rising naturalists, who will be able to view both sides of the question with impartiality.”(Charles Darwin)

1 Filling the gap of knowledge about gossip

Gossip is omnipresent in everyone’s daily life. Indulging in gossip is addictive. Even though we might want to stop engaging in the exchange of gossip, we can’t. We can’t turn it of. Studies looking at the content of daily conversations have shown that gossip takes up a big part of our communicative behavior (Goldsmith & Baxter, 1996; Dunbar, 1998b). Gossip occurs in every society. Brown (1991) listed gossip as a uHuman Universal, as a human trait that is present in every culture of the world. Still, despite the universal and dominant presence of gossip, this does not reflect in the scientific field. Research on gossip is rather scarce.

In the 1960’s first real interest in gossip as a scientific research topic arose when Max Gluckman (1963, 1968) and Robert Paine (1967) argued about the function of gossip in our everyday life. After them research on gossip flourished for a while. Mainly anthropologists conducted studies to investigate the role of gossip in our everyday lives. Next to these participant observation studies, interest also grew from psychologists, sociologists and communication studies, who not only used participant observation to investigate gossip, but also made use of questionnaires and experimental designs.

In my opinion, the research on gossip that has been done so far misses out on some important aspects. First of all, as Foster (2004) has recently argued, there lacks a clear definition of what gossip is about and what it is not. Various researchers use various definitions of gossip. Some see it only as negative talk behind someone else’s back, while others generalize gossip to any talk about a given individual, present or not. Because the definitions of gossip are sometimes so very different, it is hard to compare studies on gossip. In my view ‘gossip’ is a too complex phenomenon to put forward ‘the’ definition of it. Gilmore (1978) shared my opinion in this, and already suggested to classify gossip in smaller sub definitions. He classified gossip according to the different uses of gossip in a Spanish rural community. His classification is applicable in this cultural setting, but cannot be applied across cultural borders, because it was funded on specific cultural habits common to the community where Gilmore conducted his research. So far no one has ever tried to put forward a classification system of gossip that can be applied to any culture, and that could solve the problem of defining gossip for scientific scopes.

Second, the research on gossip has until now almost exclusively focused on proximate explanations to explain this human trait. In explaining behavior, two levels of explanation can be distinguished. As Tinbergen (1976) has pointed out, proximate explanations deal with direct causal mechanisms, while ultimate explanations look at evolutionary developed causes of traits in a population. Basically proximate explanations offer answers to “How” questions, while ultimate explanations complement these with answers to “Why” questions Standard social scientists in general focus on proximate explanations for human behavior, while evolutionary psychologists complement these with ultimate explanations. It is important to keep in mind that both scientific paradigms are not contradictory to each other, but complement each other.

Most research on gossip so far, stems from within the standard social sciences. The last decade a few evolutionary psychologists have looked at gossip as well, but none of them has come up with a complete theory of this human trait. What is lacking in the research on gossip is a synthetic theory, combining the knowledge of the standard social sciences and the knowledge of evolutionary psychology to explain how gossip operates and why it emerged and still functions in our current daily lives. Only such an overall theory can lead to a complete understanding of the many aspects of a complex human trait such as gossip.

Third, and last, the research on gossip has so far almost completely focused on Interpersonal Gossip, occurring in face-to-face settings. However, we cannot deny that in our Western societies gossip is transmitted through other channels than face-to-face communication. The mass media have enabled new transmission systems to spread gossip about individuals on a large scale. Although standard social scientists have focused on the relation between audiences and celebrities, and have looked at phenomena such as stardom and idolatry, they have not yet put forward answers to why almost anyone craves gossip about these celebrities. Media Gossip is a highly influential source of information that reaches enormous audiences in all Western societies. Still, again this does not reflect in the scientific field. Despite the dominant presence of Media Gossip, research about this phenomenon is scarce.

For these three reasons I thought it was necessary to set up a research study on gossip that can fill these three gaps of knowledge about gossip. To reach my goals I have opted to first of all conduct a theoretical literature review of the research on gossip so far. I presented this in part I of this dissertation (chapter 1-7). I looked at all critical points of discussion among researchers who have tried to define gossip. From this critical overview I came up with a very general definition of gossip (chapter 1). Second, I looked at the most important and relevant studies that have tried to explain the effects gossip has on our daily lives (chapter 2), and added to this a new approach to investigate the function of gossip, using an evolutionary framework (chapter 3-6).

Evolutionary psychologists (e.g., Buss, 1999; Cartwright, 2000; Barkow, Cosmides & Tooby, 1992; Barrett, Dunbar & Lycett, 2002; Gaulin & McBurney, 2004; Symons 1979; Tooby & Cosmides 1990) say that our human psychology was shaped by the process of natural selection. In their view our brain houses adaptations, or problem-solving mechanisms. The outcome of these problem-solving mechanisms results in higher chances of survival and reproduction for an individual. Our psychological mechanisms evolved to solve specific, recurring problems faced by our hunter-gatherer ancestors.

Every behavioral outcome is the result of a biological basis (nature) in interaction with the environment (nurture). The historical context takes a central position within the framework of evolutionary psychology. It is the past environment wherein selection shaped human behavior. Important to understand is that when evolutionary psychologists refer to the Environment of Evolutionary Adaptedness (EEA), they do not refer to a place or time, but to a statistical composite of selection pressures that shaped an adaptation (Tooby & Cosmides, 1990a). When they do refer to a time and place in our evolutionary past, that is relevant in adaptationist terms, they refer to the Pleistocene. They do so because for approximately 99% of our evolutionary past our ancestors lived as hunter-gatherers (Hawkes, O’Connell & Rogers, 1997).

The combination of both looking at the proximate and ultimate explanations of human gossip offer a complete theory to explain gossip, as I thought was lacking in the research on gossip. In a last theoretical chapter (7) I have applied this theory to the recent phenomenon of Media Gossip.

Next to this theoretical literature review I also set up some empirical studies to focus on certain aspects of gossip, which I presented in part II of this dissertation (paper 1-8). I opted to use both qualitative and quantitative studies. For the qualitative studies I made use of focus group interviews to investigate the presence of Interpersonal Gossip in our daily lives (paper 1) and the use of Media Gossip (paper 5). For the quantitative studies about Interpersonal Gossip I made use of a paper-and-pencil test, I distributed among 168 students (paper 2), a recall test experiment conducted with 85 students (paper 3). To investigate Media Gossip I additionally used a quantitative content analysis on 852 gossip articles, selected from three Flemish gossip magazines (paper 4). Last I distributed a large-scale survey about various aspects of the use and interest in Media Gossip among 104 American students and 838 Flemish respondents. The results of this survey are presented in paper 5 to paper 8.

2 Frequently asked questions about gossip

To resume my general conclusions I distilled from both my theoretical literature review and my empirical studies, I will now answer some general frequently asked questions concerning gossip. I start with one of the most crucial questions: “What is gossip?” I then turn to an explanation for our addiction to the exchange of gossip, and put forward an answer to “Why we all consume Media Gossip”. Next, I will explain that women indeed gossip more than men. And, I will end this concluding section with resuming the link between ‘gossip’ and ‘chocolate’.

2.1 What is gossip

‘Gossip’ embodies so many aspects, that this vague term almost says nothing. Because of the complexity of this human trait, it is best to classify gossip in smaller parts that can more easily be defined. A first way to classify gossip is as Post (1994) has said, to distinguish for gossip as a noun and gossip as a verb. Gossip is both information spread to others (noun, message) and a habit we all engage in (verb, action).

On this very general level, gossip as a noun can be defined as:

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

This general definition of gossip as a noun embodies most of the definitions other researchers have used so far. This general definition is close to how anthropologists define gossip. It is not the most general definition of gossip that has been given so far. The most general definition of gossip is the one Dunbar (1998b) uses in his approach to gossip. He classifies every conversation about an individual as gossip. Even talk about yourself is included in his definition. I have opted not to do this, because self-disclosure (talk about yourself) might be very related to gossip, but in my opinion the very essence that defines gossip is that it concerns information about other individuals.

On this very general level gossip as an act can be defined as the exchange of how I just defined gossip as a noun. Of course these definitions are so very general that again they do not fully answer the question “What is gossip?” These definitions are comparable to how a horse breeder would explain to laypeople what horses are. If the horse breeder wants to fully explain what horses are, he or she must mention that there exist different breeds of horses that are all similar, but still are different. Gossip in this very general sense is comparable to ‘furniture’. It is an overall noun that compromises many different sub aspects. ‘Furniture’ refers to ‘chairs’, ‘tables’, ‘beds’, and so on. But chairs have different functions than tables, and beds, and if we fully want to understand how people use furniture and why they use furniture, we focus on the sub categories. Imagine you have a visitor who has never seen any furniture you own before, and he or she asks you why you have all these different strange things in your house. To explain the function of furniture, you will focus on explaining what chairs are used for, what tables are used for and what beds are used for, but you will not explain the use of furniture in general.

If we want to get a complete picture of how gossip influences our daily lives, I also believe we should not try to explain gossip as a whole, but look at the different sub aspects gossip embodies. I have therefore further narrowed down both gossip as an act and gossip as a noun to more refined categories. Let me start with resuming my classification system of gossip as a noun.

As a noun I have classified gossip according to the different functions that can be attributed to gossip in the most general sense. In total I have distinguished for 16 different definitions of gossip, based on the different functions gossip can have. For a description of these 16 sub categories of gossip as a noun I refer to table 1.

Table 1. 16 definitions of gossip according to their functional design

As a noun, the focus of gossip can either be directed to the gossiped about behavior itself, or the focus can go towards the gossiped about traits/behaviors attached to a specific person. In the first case I talk about Strategy Learning Gossip (SLG). Here gossipees are mere carriers of fitness-relevant information about the strategies they have tested. Whether they are known directly or indirectly (through one or more intermediary others) or not, to the gossipers does not matter. What matters is the behavioral information gossiped about. I further narrowed down SLG to Survival SLG, Mating SLG and Social SLG. For a description of these I refer to table 1.

When the gossiped about traits/behaviors attached to a specific gossipee have a connotation that disappears when the traits/behaviors are detached from the gossipee, I talk about Reputation Gossip. Bromley (1993) has explained that reputations are build up by what we are (traits) and what we do (behaviors) and how others perceive this. When gossip as a noun is about traits/behaviors that get (extra) value because of their link to a specific gossipee, reputation information is present in this gossip message. I have further narrowed down Reputation Gossip to Mating RG and Social RG. Mating RG concerns information about the reputations of others as good (potential) mates or rivals in the search and maintenance of a mate. Social RG is about the reputations of individuals as social agents in non-sexual interactions with others. Both are further narrowed down again. For a detailed overview I again refer to table 1.

Important to note about the classification system I here put forward is that one piece of information can be classified under different sub categories. To understand gossip as a noun, gossip as an act cannot be omitted. If the gossipers do not know a gossipee, a piece of information might only be Strategy Learning Gossip to them. If they do know the gossipee, this same piece of information can be both Strategy Learning Gossip and Reputation Gossip to the gossipers. “Have you heard Jeremy filed for divorce? This means one of the most eligible men is back on the market!” can be Mates Detection RG to a female receiver and Sexual Rival Detection RG to a male receiver (in the context of heterosexual relationships). This illustrates how the connotation of gossip heavily depends on the context wherein this piece of information is exchanged.

This classification system is funded on an adaptationist analysis where I focused on functions that could solve problems occurring in the EEA. Because of their adaptive value, these functions are assumed to be universally present today. Therefore I believe that this classification system is applicable in any given society.

An indication that the 16 different kinds of gossip information I distinguish for are present in the current daily conversations of Flemish people follows from my focus group interviews. Using a semi-structured survey I asked a total of 103 Flemish respondents some questions about their daily gossip conversations. I did not use my classification system to structure my semi-structured survey, to avoid I would direct my interviewees’ answers too much. Still, from their reports I found indications of presence for all 16 sub categories of gossip. Only Kin Detection RG and Ally Detection RG were forms of gossip as a noun my respondents hardly mentioned to talk about. All others were present in their daily conversations. Mating SLG and Mating RG was dominantly present among the younger respondents and the presence of this specific form of gossip decreased as the age of the respondents increased. Survival SLG, Social SLG and Social RG were present among all age groups.

Next to classifying gossip as a noun, I also have distinguished different forms of gossip as an act. Looking at gossip as a behavior strategy I make clear distinctions between Interpersonal Gossip and Media Gossip. The first concerns the exchange of all different kinds of gossip as a noun in an interpersonal face-to-face setting. This is the form wherein gossip emerged and evolved throughout our evolutionary past. It is this classical form of gossip that could solve problems our ancestors faced (see also below). I do comment however, that nowadays Interpersonal Gossip not only is exchanged through face-to-face communication, but can be transmitted through other channels, such as written letters, e-mail, (mobile) phone, messenger programs and so on. From my focus group interviews follows that especially young respondents reported to use other than face-to-face interfaces to exchange gossip messages.

Media Gossip is gossip spread to a large, unknown audience. With Media Gossip I refer to the messages that are present in (gossip) magazines, newspapers, soaps, movies, biographies, and so on. Gossip messages are dismissed through media that reach large numbers of receivers. The sender of Media Gossip is (almost) unknown to his or her audience. Media Gossip tales about the traits and/or behaviors of gossipees that are ‘unknown’ to the gossipers. Unknown in the sense that real-life encounters with the gossipees of Media Gossip are non-existing. Media Gossipees are either celebrities, known for appearing in the media on a regular basis, or ‘public unknowns’. These last are individuals who appear in the media on a short-term basis because they have experienced something extraordinary (see also below).

Looking at Media Gossip and Interpersonal Gossip both as an act and as a noun, these forms of gossip are different on both levels. Even when Interpersonal Gossip is exchanged through new media such as e-mail or messenger programs, the features that distinguish Interpersonal Gossip from Media Gossip remain. The first main difference between both forms of gossip exchange (as an act) concerns the audience they reach. Interpersonal Gossip is exchanged between gossipers who know each other more or less. This is, Interpersonal Gossip is exchanged by individuals who have met each other personally or through an artificial medium. The audience, a sender of Interpersonal Gossip shares his information with, is relatively small compared to the large audiences Media Gossip reaches. Second, Media Gossip is different from Interpersonal Gossip in terms of the gossipees of the message (gossip as a noun). When we exchange gossip interpersonally, we talk about gossipees we either know directly, or whom we know indirectly through one or more intermediary others. Gossipees of Interpersonal Gossip are therefore (in)directly known to the gossipers. This is different for Media Gossip. Gossipees of Media Gossip are celebrities or public unknowns who are not known by the gossipers. The gossipers might feel ‘as if’ they know celebrities (see below) but there is no direct or indirect connection to the gossipees of Media Gossip.

Of course we interpersonally gossip about celebrities and public unknowns as well. Therefore I distinguished a third form of gossip funded on specific criteria as an act and as a noun: Interpersonal Media Gossip. Interpersonal Media Gossip resembles Interpersonal Gossip as an act and Media Gossip as a noun. Interpersonal Media Gossip is about the traits and/or behaviors of celebrities or public unknowns, and is shared in an interpersonal setting to a small audience of (in)direct known receivers. For an overview of how I distinguish Interpersonal Gossip, Media Gossip and Interpersonal Media Gossip, I refer to table 2.

Table 2. Different as nouns and verbs: Interpersonal Gossip, Media Gossip and Interpersonal Media Gossip

* Interpersonal Media Gossip resembles Interpersonal Gossip as an act and Media Gossip as a noun.

2.2 Why do we gossip

Knowing what gossip is in the most general sense, how we can classify gossip as a noun into 16 different sub categories, and how we can distinguish for Interpersonal Gossip, Media Gossip and Interpersonal Media Gossip, I will now answer the next frequently asked question: “Why do we gossip?”. I here focus on explaining why we share Interpersonal Gossip. An explanation for why we gossip about celebrities using Media Gossip and Interpersonal Media Gossip follows in the next paragraphs. To explain why we gossip in interpersonal settings I fall back on proximate explanations from the standard social sciences and ultimate explanations from evolutionary psychologists and my own adaptationist analysis.

Interpersonal Gossip is the oldest form of gossip. It is a form of social communication, making use of language. In my adaptationist analysis I have given an overview of the historical roots of language (see chapter 3). Although discussion exists about when and how language first emerged, I consider language as a product of evolution that is hundreds of thousands years old. Language most probably evolved out of older forms of communication and cognition that today are still present in other animals as well.

Regarding language, and therefore Interpersonal Gossip, as a product of evolution, I assume that different selection pressures have shaped our gossip behavior that exists today. I do not talk about one selection pressure, but multiple selection pressures. More specific I talked about natural, sexual and social selection pressures. All three selection pressures act on the level of individuals or, more precisely on the level of genes, as Williams (1966) has explained.

Natural selection pressures refer to the hostile forces of the environment our ancestors lived in, and the problems they faced to secure their survival. I have argued that learning is extremely important. In our problem solving decision making, we often rely on past experiences. Learning is getting experienced and having knowledge about which strategies are successful and which are not. Because investing own time and energy in testing strategies can be so very costly to an individual, he or she can benefit from others’ trial and errors. Learning that some food is poisonous or that you can die from certain snake bites for instance, are experiences that are so costly they can take away your life. In these situations individuals benefit from gaining this knowledge by getting the information not from private experiences, but from others’ experiences. Individuals vicariously learn from the unsuccessful strategies of others. The same is true for successful strategies of others. We not only learn what not to do, but by talking about the strategies of others, we also learn what we best do in the future when confronted with a similar situation. I then talk about Strategy Learning Gossip. We talk about the behaviors of gossipees to learn how to behave and how not to behave. Of course not solely natural selection pressures have shaped our minds to indulge in the exchange of SLG. Also sexual and social selection pressures have shaped our tendency to acquire and share SLG. I therefore have distinguished for Survival SLG (natural selection), Mating SLG (sexual selection), and Social SLG (social selection; see table 1).

Within the standard social sciences, this learning use of gossip has been discussed as well. They talk about the informative effect gossip has. Gossip informs group-members about what is going on in the environment. Individuals benefit from the actions of others. They mimic the strategies of other individuals, and sometimes adjust their own behavior to become similar to other group members. This last phenomenon has been explained thoroughly by Festinger (1954) in his Social Comparison Theory.

Turning to the sexual selection pressures that shaped our gossip behavior, I here resume the problems of human mating our ancestors have faced. As Symons (1979) says, men and women have different desires in what they find attractive in potential mates. These desires do not necessarily translate in actual behavior, which is important to keep in mind. These desires have been shaped by sexual selection processes, because these desires were once (and some still are) beneficial in the arms race of reproduction. What I do want to stress again, is that I here put forward what is biologically beneficial for men and women to do in their search for potential mates. I do not at all claim that these strategies are what men and women should do when searching for a potential mate. No moral consequences can be drawn from nature, as David Hume already pointed out in the 18^th century.

In biological terms men benefit most from female potential partners with high fertility and reproductive value. Of course these features are disguised in women, which results in the fact that men face(d) problems of estimating the fertility and reproductive value of potential female mates. Cues they use to estimate these criteria are youthfulness and physical attractiveness. It has been shown that young, physical attractive women score higher in terms of fertility and reproductive value then their older and/or less attractive colleagues. Further men biologically benefit from either aiming for promiscuous or faithful potential partners. When their goal is to have a short-term sexual affair, men biologically seen benefit most from preferring promiscuous women over faithful women. When men want to aim for a long-term commitment, their biological best strategy is to prefer faithful women over promiscuous women. Again, in biological terms, women benefit most from long-term relationships with potential partners that are able and willing to invest in their offspring. Women rely on the wealth status and willingness to commit of male potential partners, because these signal ability and willingness to invest.

In their search for potential partners, both men and women are confronted with similar problems. In their search for a good potential mate, either for short-term or long-term commitment, both men and women have to face the deceitful actions of the opposite sex. In addition to this, men and women encounter problems of same-sex competitors that aim for the same potential partner. Third, once they are involved in a sexual relationship, men and women stumble upon problems of potential infidelity of their partner. Biologically seen sexual infidelity of a partner opposes the biggest threat to men, while women more than men are confronted with costs when their partner invests in other women (emotional infidelity).

I have suggested in chapter 4 how different kinds of gossip can solve these mating problems. Mates Detection Reputation Gossip and Mating Structure Reputation Gossip function to inform individuals about the reputation of others as good or bad potential mates, and the availability of potential mates. Sexual Rival Detection Reputation Gossip functions to inform individuals about the potential threat of same-sex competitors. Sexual Rival Slander Reputation Gossip functions to ruin the reputation of same-sex competitors. Mates Detection Reputation Gossip functions to signal the cheating behavior of other individuals (for an overview see table 1). All these different kinds of gossip are all forms of Reputation Gossip. This is because the focus is centered on the gossipees; individuals benefit from learning who their potential mates and competitors are. Mating Strategy Learning Gossip as well can solve problems in the domain of human mating. By hearing about the successful and unsuccessful strategies in the context of human mating problems, receivers can vicariously learn how to act in the future to increase their successes. Mating SLG does not focus on the gossipees, but on the gossiped about behavior.

The function of gossip in the context of human mating problems has been acknowledged by other evolutionary psychologists (e.g. Buss & Schmitt, 1993; Hess & Hagen, 2002, 2004a, 2004b). However, within the standard social sciences this use of gossip has not been stressed before. It has been argued that love topics are popular gossip topics (Levin & Kimmel, 1977), but so far standard social scientists had not focused on the use of gossip in the context of human mating. They have put forward some effects of gossip that are clearly present in these specific functions. More specific, standard social scientists have stressed the use of gossip as a learning device, as I already mentioned above. However, they have not distinguished for learning about specific others and learning about behaviors of others, as I have distinguished for with Reputation Gossip and Strategy Learning Gossip. Standard social scientists have also focused a lot on the manipulative use of gossip. Inspired by Paine’s (1967) view on gossip they have stressed how individuals use gossip to slander foes and praise friends to increase private relative reputations. This effect reflects in the functioning of Sexual Rival Slander RG.

In general, since the contributions of evolutionary psychologists to a better understanding of the role of gossip in the context of human mating have been scarce as well, I think this part of my research can be considered as one of the main contributions towards a better understanding of this aspect of gossip in general. I also focused on gossip in the context of human mating in one of my empirical studies. I used a recall test, which I presented to 85 students. I wanted to investigate if male and female respondents recalled different aspects of the Mating RG stories I presented them. I had predicted that single female respondents would remember more about the single male gossipees and that single male respondents would recall more about the single female gossipees than vice versa (Mates Detection RG). Still, my results did not confirm this. Also the prediction that more would be remembered about troublesome relationships (Mating Structure RG), because this signals poaching opportunities, could not be confirmed with this study. What I did confirm was my prediction that cues about physical attractiveness would be recalled more for a female gossipee than for a male gossipee. Although not significant and not as strong as for the gossip about attractiveness, I also found some support for the fact that cues about wealth status are recalled more for a male than for a female gossipee (Mates Detection RG).

As a last spectrum of adaptive problems our ancestors have faced, I now turn to how social selection pressures have shaped our gossip behavior. Humans are social beings, and have always been so. Our ancestors lived together in small groups of co-operative individuals that were often in conflict with each other as well (Tooby & Devore, 1987). In biological terms co-operation asks for some explanation. Hamilton (1964) and Trivers (1971, 1985) have put forward explanations for why we act altruistic both towards kin-related and non kin-related others. Co-operation rests on reciprocity. When we favor others, we expect return benefits. Free-riders -individuals who benefit from the altruistic actions of others without returning benefits- threaten co-operation. It has been shown however that humans are good in detecting cheaters in social settings (Cosmides & Tooby, 1992).

For long-term co-operative bonds, such as friendships, we best (in biological terms) seek allies that have skills valuable to us. Friends fill up niches, and becoming someone’s friend rests on the principle of making yourself irreplaceable (Tooby & Cosmides, 1996). Both in this context of long-term co-operative interactions and short-term interactions with others we must be able to predict the actions of others. What is labeled as Theory of Mind (Baron-Cohen, 1995) refers to the ability to mind-read other people, and to predict their actions. Humans seem to be good in this. Cues relevant to predict the actions of others are looking at what any other average person usually does in this situation (rests on experience) and looking at the regularities in your interaction agent’s behavior pattern. If individuals deviate from what other individuals usually do, or suddenly deviate from their regular behavior pattern, we must keep track of this and calibrate our attitude towards this individual.

In chapter 4 I have shown how specific forms of Social Reputation Gossip can solve some of these social problems our ancestors faced in their interactions with others. Cheater Detection RG and Altruist Detection RG function to inform individuals about the threat of cheaters and the benefit of altruists operating in their environment. Furthermore these forms of gossip also function to punish cheaters and praise altruists by decreasing or increasing their reputation as a co-operator, and therefore increasing or decreasing their future co-operative opportunities. Ally Detection RG functions to inform individuals about the skills of others in their social environment. This information is helpful in the search for allies/friends. Ally Structure RG and Kin Structure RG function to inform individuals about the non kin-related and kin-related bonds of other individuals in their social environment. Ally Maintenance RG functions to increase the relative status of an individual by increasing the reputation of allies and decreasing the reputation of non-allies. Last, Other Deviance Calibration RG and Self Deviance Calibration RG function to inform individuals about changes in the behavior pattern of other individuals with whom they interact (for an overview see table 1).

Overall I have distinguished between 16 different functions that can be attributed to gossip in general. Each of these functions explains why we use a specific form of gossip, which I have all put forward in my classification system of gossip as a noun (see above). These functions reflect in the uses of gossip standard social scientists have focused on so far. Summarizing the effects gossip has in our daily interactions with others, according to the standard social scientists, I distinguished for 4 main effects. Gossip is used to (1) learn from others, (2) to control and (3) manipulate others, and because (4) gossip is entertaining, it relieves and it has a cathartic effect on the gossipers.

The learning effect of gossip goes hand in hand with the learning function I have attributed to gossip in several of the specific functions. The standard social scientists explained on a proximate level that gossip is used as a learning device. Evolutionary psychology does not contradict this, but rather adds to this, ultimate functions to explain why we learn by using gossip. Learning is either focused on behavior strategies (Strategy Learning Gossip) or on specific other individuals (Reputation Gossip). Learning from the behavior of others is so important, that it can even save your life. Because self-experiences can be too costly, we vicariously learn from the actions of others. Learning is also adaptive when it concerns other individuals. Learning who is a good potential mate can increase our reproductive opportunities. Learning about cheaters and altruists among us can increase our co-operative successes. Learning about others is necessary to calibrate our attitudes when someone suddenly deviates from what we would have predicted he or she would have done. In biological terms individuals benefit from the fruitful co-operative interactions with others. We learn about other individuals to secure these co-operative interactions run smoothly and result in benefits rather than in costs to ourselves.

The control effect of gossip, as stressed by Gluckman (1963, 1968) explains how gossip is used to control other individuals. Again evolutionary psychology adds to this an underlying ultimate explanation for why we feel the urge to control others by using gossip. As mentioned in the previous paragraph, co-operation is highly beneficial to every human being. The fact that we humans are social rests on co-operation. Although co-operation is common in everyday life that some might not wonder why we all act altruistic, this is an evolutionary puzzle. In biological terms co-operation is not that obvious. Trivers (1985) and Hamilton (1964) have explained how co-operation rests on reciprocity. Cheaters are threatening forces to reciprocal actions. Cheaters threaten the existence of groups based on reciprocal co-operative interactions. Because individuals benefit from co-operation, and cheaters threaten to destroy these benefits, we humans feel the urge to control others, to detect and punish cheaters. That is why we use gossip as a control device.

In paper 2 I have presented the results of my replication of a study Wilson et al (2000) had done. I presented students some fictive stories about the cheating and non-cheating actions of students. Other students started gossiping about the deceitful actions of these cheaters, but some also spread lies, accusing innocent people as cheaters. My results, in line with Wilson et al’s (2000) results show that Cheater Detection RG about the deceitful actions of others is not disapproved, while Cheater Detection RG about innocent people (abuse of Cheater Detection RG) gets clearly disapproved.

While the learning effect and the control effect are both highly beneficial for receivers of gossip, the manipulative effect of gossip is most beneficial for senders of gossip, as explained in chapter 5. The manipulative use of gossip, as stressed by Paine (1967), translates in several of the specific functions I have attributed to gossip in general. Some specific forms of Reputation Gossip, such as Sexual Rival Slander RG and Ally Maintenance RG are clearly manipulative. They function to increase or decrease the reputations of the gossipees. These manipulative forms of gossip are in line with Paine’s (1967) view on gossip that centered on the manipulation of reputations. Most standard social scientists and even some evolutionary psychologists (e.g. Barkow, 1989, 1992; Hess & Hagen, 2002; McAndrew & Milenkovic, 2002) focus on the manipulative use of gossip only in the context of the manipulation of reputations of the gossipees. However, as I have shown in chapter 5, where I presented optimization models and decision trees that explain why we share both Strategy Learning Gossip and Reputation Gossip with others, manipulation also accounts for the sharing of non-reputation loaded Strategy Learning Gossip. Senders of SLG do not manipulate reputations of gossipees, but knowledge of receivers. Senders can gain private benefits from increasing the knowledge of their allies with true SLG and depriving non-allies from this fitness-relevant time- and life-saving information. Senders also benefit from sharing false SLG with non-allies, and therefore potentially threatening the fitness of their non-allies. Such false SLG is optimally not shared with allies.

The reason why (ultimate explanations) we manipulate the reputations of gossipees with Reputation Gossip and the knowledge of receivers with Strategy Learning Gossip is because these manipulations rebound to private benefits. Manipulating knowledge of receivers, senders can increase their relative fitness-status. Manipulating the reputations of gossipees, senders can increase their relative social status, which in the end affects their fitness as well.

The only effect I have not discussed so far is the entertainment effect of gossip. Standard social scientists say we gossip because we love it, because it relieves and makes us feel good. What could be the underlying ultimate explanation for this effect? Why does gossip make us feel so good? Brown (1991) has suggested that one of the fundamental assumptions of evolutionary psychology is that matters closely related to our survival and reproduction, have a likelihood to make us feel good. We enjoy, we like certain habits because these strategies were optimal for our Pleistocene foraging ancestors. Those ancestors who enjoyed gossip have won the arms race. Our gossipy ancestors were the winners. They are our ancestors, and because they liked gossip, and because gossip was good for them, we like gossip. This of course leads me to the next question that will now pop up in evolutionary minds: “Is gossip then an adaptation?”

2.3 Is gossip an adaptation

Here I focus on Interpersonal Gossip and not on Media Gossip and Interpersonal Media Gossip, to which I turn in the next section. The short answer to this question is: “No, Interpersonal Gossip is not an adaptation”. Let me explain.

In summary, Williams (1966) has put forward three criteria that specify adaptations. Adaptations are (1) reliable, (2) efficient, and (3) economical. First, that adaptations are reliable refers to the fact that adaptations occur universally. All members of a species must develop the trait under ‘normal’ conditions before the trait can be called an adaptation. Interpersonal Gossip is listed as a human universal (Brown, 1991), so I could take for granted that Interpersonal Gossip passes this first criterion. However, I immediately comment on this that Interpersonal Gossip defined as ‘talk about other people’ might be a human universal. I have classified Interpersonal Gossip into 16 sub categories. Whether these sub definitions of Interpersonal Gossip occur universal, and can all be considered reliable has not been tested so far.

Next, adaptations are specialized in the sense that they solve specific, narrow problems (efficiency). Interpersonal Gossip in the general sense cannot pass this criterion. I have shown that Interpersonal Gossip does not have one single specific function to solve one specific problem, but houses different functions that each can solve specific problems. Therefore the short answer to “Is Interpersonal Gossip an adaptation” is a definite “No”. The sub categories of Interpersonal Gossip I distinguished can each solve specific narrow problems. The 16 different kinds of Interpersonal Gossip I have listed can each pass this second criterion Williams (1966) has put forward. Therefore, I can say that, for instance “Cheater Detection Reputation Gossip is an adaptation”.

Third, and last, adaptations are economical. With this Williams (1966) points to the fact that adaptations are costly. They have developmental and maintenance costs. But these costs have to be kept minimal. If mechanisms would be too costly, their functioning cannot result in benefits to the individual who has to bear the high costs of developing and maintaining the mechanism. Is Interpersonal Gossip in general, or as an overall noun of 16 specific kinds of Interpersonal Gossip costly? Not at all. As I have explained in chapter 6, in my opinion our Interpersonal Gossip behavior is the outcome of the joint action of several mechanisms. For all 16 sub categories of Interpersonal Gossip I distinguish for we rely on our Language Mechanism that has been suggested to exist (Pinker, 1995), our Theory of Mind Module, as put forward by Baron-Cohen (1995) and several emotions. I have further commented that each of the 16 sub categories of Interpersonal Gossip additionally rests on the functioning of other mechanisms as well. For instance, Cheater Detection RG is the outcome of the mechanisms, mentioned above, in joint action with our Cheater Detection Module, as shown to exist by Cosmides and Tooby (1992). All 16 sub classifications of Interpersonal Gossip are not cheap to house our minds; they are extremely cheap to house our minds. The adaptive outcome of all these different kinds of Interpersonal Gossip rests on the joint action of mental mechanisms that already exist to solve specific other functions.

The trouble with the problem-solving of each of the 16 different kinds of Interpersonal Gossip is that they are not optimal. True, learning from the behaviors of others saves our time and energy, and can save our lives, but Strategy Learning Gossip is not as reliable as our own experiences. If we want to be sure that our problem-solving strategy is fully reliable, private experiences are still better than sharing Strategy Learning Gossip. Cheater Detection RG can save our time and energy in detecting cheaters, it can reduce the risks of being defected by cheaters. Still, again Cheater Detection RG is not as reliable as our own experiences with cheaters in the past. Even though I have defined gossip as ‘reliable’ information, in the sense that senders of gossip has true/false knowledge about their gossip message, the exchange of all gossip still faces reliability problems. As stressed in my decision trees in chapter 5, where I outlined fast and optimal decision rules to acquire and share gossip in a beneficial sense, reliability is the first cue receivers of both SLG and RG must rely on.

To conclude, I compare Interpersonal Gossip with a toaster-oven. A toaster-oven can grill and bake food and toast bread. But an oven can bake better than a toaster-oven. A grill can grill better than a toaster-oven. And a toaster makes better toasts than a small toaster-oven. The great benefit of a toaster-oven is that it is cheap and takes little space. Even though the functions of a small toaster-oven might not be performed at the best, people buy them because they are cheap and do not cover up their whole kitchen. Interpersonal Gossip is comparable. The functions the 16 different kinds of gossip I have distinguished between might not be performed at the best either. Other problem-solvers might have better performances to solve problems of learning, of finding and guarding mates, of detecting cheaters, or skilled allies. But Interpersonal Gossip is cheap. Cheap, because the hardwired tools necessary to enable Interpersonal Gossip are already housed in our minds. The necessary tools are there. The only thing Interpersonal Gossip behavior requires is the joint action of these different mental mechanisms. So, if it only took that little effort, then why not benefit from the fact that Interpersonal Gossip can solve problems at a minimum. If the overall benefits outscore the very low cost of enabling humans to gossip, then why not accept the low cost and gain from the cheap benefits?

2.4 Why do we gossip about celebrities

While I assume that the 16 different sub categories of Interpersonal Gossip might be seen as adaptations, if it could be shown that they occur cross-culturally, this does not count for Media Gossip and Interpersonal Media Gossip. Media Gossip and Interpersonal Media Gossip are no adaptations. Both forms of gossip are merely an eye blink in evolutionary terms of existence. They are too recent products to be the result of a selection process. Selection acts slowly and takes many generations passing on an adaptive problem-solving trait.

Still, despite the fact that both phenomena are so very recent, I have put forward an evolutionary explanation for why we value (Interpersonal) Media Gossip. Within the standard social sciences, interest in Media Gossip has been explained by the fact that audiences can learn from media messages and by the fact that some viewers engage in parasocial relationships with celebrities (Caughey, 1984) and therefore show interest in their private lives. Let me explain on both a proximate and ultimate level.

I start with the learning effect of Media Gossip, which on an ultimate level, I have incorporated in my Learning Hypothesis to explain interest in Media Gossip. Media Gossip teaches audiences about the misfortunes of public unknowns and the victories of celebrities. Audiences can learn that they are better of than the public unknowns, and can escape in the world of glitter and glamour celebrities live in. So say standard social scientists. But why do we want to learn from the misfortunes of public unknowns, and why do we escape into the glitter and glamour? Well, acquiring information about the strategies of public unknowns is nothing less then acquiring Strategy Learning Gossip. As I have explained, Strategy Leaning Gossip focuses on the gossiped about fitness-relevant strategies, not on the gossipees.

Media Gossip triggers our interest in Strategy Learning Gossip. The media environment mismatches with the Pleistocene, wherein Interpersonal Gossip was selected for. Interpersonal Gossip solved problems our ancestors faced. Media Gossip might not be a product of evolution, but still triggers our evolved interest in gaining knowledge about the successes and failures of others, because this benefits us.

We vicariously learn from the behavior of others, even of those we do not know. In this perspective of the importance of learning, it is not that strange that we all crave gossip about the failures and successes of others. Anyone can benefit from such information to decide how he or she best acts in the future when confronted with similar problems to solve.

The Learning Hypothesis explains why we all show interest in Media Gossip about celebrities and public unknowns. We not only care for the bad information, but for the successes as well, because we can learn from both. We can learn what to mimic and what to avoid at all costs. In my content analysis of three Flemish gossip magazines (paper 4) I have shown that indeed both celebrities and public unknowns are gossipees of gossip about their failures and successes.

However, celebrities have an extra benefit to become gossipees of Strategy Learning Gossip of Media Gossip. They are prestigious people. It has been shown (Henrich & Gil-White, 2001) that mimicking higher status individuals can increase your own status. Tricky in this process, however, is estimating what exactly contributes to the prestige of others. What exactly leads to the fact that some individuals become celebrities? Their talent as great musicians or actors? Maybe, but not always. Some are just known because they are famous. Because it is so very difficult to estimate what exactly contributes to the prestige of high status others, Boyd and Richerson (1985) assume that evolution did not favor individuals who first carefully analyzed every strategy that could be the reason for success before mimicking this strategy, but rather suppose that selection favored a General Copying Bias. By simply mimicking the overall traits and behaviors of higher status others in general, individuals can be sure they also mimic the trait or behavior that contributes to success. They do not first have to waste their time in estimating what exactly is the key to the success of the higher status others.

This General Copying Bias explains why we sometimes mimic the overall image of celebrities. In chapter 5 I have explained that this General Copying Bias can influence the way we process Strategy Learning Gossip and the way we decide whether to mimic the gossiped about strategy or not. In a full elaborated decision process, receivers first look at the outcome of the gossiped about strategy. If this outcome is beneficial, receivers optimally decide to mimic. If this outcome is costly, they best decide not to mimic. If the outcome of the strategy is unclear, they best rely on the General Copying Bias. If the gossipee has a higher status than themselves, they optimally decide to mimic the unclear-outcome strategy, while they best decide not to mimic unclear-outcome strategies of lower status gossipees. A less elaborated decision process that receivers of Strategy Learning Gossip can use, is first looking at the status of the gossipee before estimating the outcome of the gossiped about strategy. This might be an easier way to decide when to mimic and when not to mimic gossiped about strategies. The dangerous aspect of this less elaborated decision process, however, is that not only neutral-, but also positive- and negative-outcome strategies will be mimicked if the gossipee has a higher status than the receiver. This explains why some start smoking because celebrities do, why some starve themselves almost to death because celebrities do, and so on.

Next to the learning effect of mass media gossip messages, the standard social scientists have also put forward a theory on parasocial interactions. What is special about our interactions with celebrities is that celebrities reveal their private lives to us, and we show emotions towards them; but we don’t share our private lives with them, and they don’t show emotions towards us (personally). Interactions in which reciprocity is lacking are called parasocial interactions (Horton and Wohl, 1956; Rubin, Perse, and Powell, 1985, Turner, 2004). Some of us establish parasocial bonds with celebrities and crave gossip about their parasocial network members for the same reasons why they crave gossip about real life social network members. I label this explanation for our interest in Media Gossip the Parasocial Hypothesis.

On an ultimate level, Barkow (1989, 1992) has explained why this happens. He believes that celebrity gossip is a recent by-product of Interpersonal Gossip. Our modern minds are not adapted to recent emerged environments, such as the media environment. The modern mass mediated environment mismatches the Environment of Evolutionary Adaptedness, wherein our minds were shaped by natural selection. Especially photographical, audiovisual stimuli are a too recent phenomenon for our stone aged minds to be adapted to. When we see an image of a media character (a celebrity), our brain processes this information as an encounter with this person: ‘as if’ we meet them in real. If this happens regularly, as is the case with celebrities who are secured of long-term media coverage, our brains start accumulating these encounters and make us (falsely) believe these people are part of our social networks.

Important differences in predictions of the Learning Hypothesis and the Parasocial Hypothesis concern (1) who is interested in Media Gossip, and (2) which kinds of gossip as a noun appeal to the Media Gossip audience. The Learning Hypothesis explains an interest in Media Gossip from all receivers. The degree of interest might vary, with a higher interest coming from individuals who are less experienced. Still, learning from the behaviors of others can be of value to anyone. In contradiction to this, the Parasocial Hypothesis only explains the interest of some receivers: those who establish parasocial bonds with celebrities. These are individuals who are exposed to media products a lot, so that frequent encounters with celebrities are ensured.

Second, the Learning Hypothesis explains our interest in Strategy Learning Gossip of both celebrities and public unknowns, while the Parasocial Hypothesis explains some people’s interest in Reputation Gossip about celebrities. Still, overall both hypotheses do not completely contradict, but rather complement each other. Most Media Gossip messages are not Strategy Learning Gossip or Reputation Gossip, but Strategy Learning Gossip and Reputation Gossip. “Womanizing ways led to the shock split between Charlie Sheen and his pregnant wife Denise Richards” is Strategy Learning Gossip from which all (male) receivers learn that womanizing can threaten your relationship (Mating SLG). Receivers who ‘know’ Charlie Sheen get Reputation Gossip out of this as well. They learn Charlie Sheen is a bad potential mate (Mates Detection RG). Whether Reputation Gossip is present depends on the relationship between the gossipee and the gossipers. If the gossipers know the gossipee, or here in the context of Media Gossip ‘as if’ they know the gossipee, Reputation Gossip is present. The presence of Strategy Learning Gossip depends on the fact that fitness-relevant behavior information is present or not.

From the results of my papers 5-8, where I investigated the appeal of Media Gossip, I generally conclude that the Learning Hypothesis has more explaining power than the Parasocial Hypothesis. Especially at a young age receivers of Media Gossip are interested because they can learn from what celebrities do, to achieve prestige. Adolescents love gossip about highly recognized international celebrities, to learn about their success. As age increases, interest in Media Gossip decreases. Reasons for this might be that older individuals have more experiences stored in their memory, and have a lower urge to acquire experiences through Strategy Learning Gossip. Second, older individuals are relatively less lower-status than celebrities are, and might not care as much about their doings than the younger individuals who relatively are much more lower-status than celebrities. However, interest in Media Gossip does not completely disappear. From the result of my large-scale survey stems that elderly people consume Media Gossip quite often. Their motivation however is different from the motivation of younger receivers of Media Gossip. From my focus group interviews I noticed that elderly people are attracted to Media Gossip because they lack real life social contacts. For them the Parasocial Hypothesis explains better why they consume Media Gossip, and also use Interpersonal Media Gossip to uphold their few real life social contacts. Another result I got from my research on Media Gossip is that women show a bigger interest than men, especially among younger receivers. This leads me to the next question: “Do women really gossip more than men?”

2.5 Who gossips most: men or women

Gossip is a social activity. It has been shown that women in general are more socially active than men. Already at a young age girls are more intimate, they exchange more information, whereas boys are more aggressive and solitary in their play behavior (Lansford and Parker, 1999). Some (e.g. Ben-Ze’ev, 1994) have argued that according to traditional cultural expectations women’s activities are more confined to family and friendship networks, and form strong ties, whereas men’s activities involve more distant relationships with colleagues.

I agree that women might be the socially most active sex, but think this is rather due to different selection pressures in our evolutionary past as well, than solely due to the pressures of nowadays cultural expectations. In chapter 3 I have argued that women and men faced different adaptive problems due to a sexual division of labor, and a difference in relocation patterns. Especially this last phenomenon contributes to the fact that women today are the socially most active gender. Our female ancestors were faced with more adaptive problems of socializing with new social network members because they moved to the band of their husband after marriage. Patrilocality, which means the husband lives with his wife and his relatives, occurs in 70% of our human societies (Burton et al, 1996). Studies tend to confirm that in our evolutionary past women migrated more often than men to his residence after marriage (Oota et al, 2001, Rodseth et al, 1991). Our female ancestors were more confronted with adaptive problems due to relocation, such as learning about their new social environment and establishing new social contacts. If this is true, we could suppose that those behavioral adaptations still exist today, and we could expect females to be the socially more active sex.

And this of course translates into a more active gossip behavior of women. Studies that have investigated this are scarce. But, Levin and Aluke (1985) and Nevo and Nevo (1993, 1994), have focused on sex differences in amount of time dedicated to gossip, and the content of these conversations. Levin and Arluke (1985) their results show that women spent more time to gossip about other people than men (71% of women’s conversations and 64% of men’s). Very similar sex differences in gossip resulted from the research of Nevo and Nevo (1993, 1994). Using a tendency-to-gossip-questionnaire, Nevo and Nevo (1993, 1994) noticed that female respondents got higher scores than their male colleagues.

Of course this does not mean that men do not gossip at all. Both men and women gossip, but women devote some more time to it. Interesting differences also arise when focusing on the specific topics discussed in gossip. Levin and Arluke (1985) have found that both men and women talk equally about dating, sex and the physical appearance of others. Men talked more about sports figures and women more about the evaluation of the behavior of others (control gossip).

So far, the few studies on sex differences in men and women’s gossip have all focused on gossip in general. When looking at the 16 different kinds of gossip I have distinguished, we might get a better idea if men and women’s gossip are really different. In chapter 5 I have argued that women might have a stronger tendency to exchange Strategy Learning Gossip about what is acceptable and non-acceptable behavior (Social SLG), because of patrilocality in our evolutionary past. I expect men and women to be equally interested in Survival SLG, and have an equal tendency to exchange Mating SLG. However, for this latter form of gossip sex differences can be expected in specific topics gossiped about. I have outlined that men and women have different desires in what they want from potential partners (see above). These sex differences in desires will reflect in interest in gossip. For instance, because physical attractiveness is on average more important for women than for men, I expect women to be more interested in Mating SLG about successful and unsuccessful strategies other women have used to improve their physical attractiveness.

For the 13 different kinds of Reputation Gossip I expect that both men and women can benefit from exchanging these kinds of gossip information. Again I expect sex differences in the interest in specific topics, and especially for Mating RG. For instance for Mates Detection RG, male gossipees can expected to be more subject to gossip information about their wealth status, while female gossipees will be more subject to Mates Detection RG about their physical attractiveness.

Further, not focusing on sex differences in content topics, but on sex differences in tendency to exchange Reputation Gossip, for some forms of Reputation Gossip it can be argued that women have greater benefits than men. As Campbell (1999) and Hess and Hagen (2002) have argued, when being in conflict women cannot as easily as men afford to compete physically. Women engage more in low-level combats, such as using negative gossip. Hess and Hagen (2002) comment that this ‘informational aggression’ is not merely a substitute for physical aggression, but is simply more effective to use in within-group conflicts, to attack reputations. Agreeing with these authors, I therefore suppose that women are more eager to engage in manipulative forms of Reputation Gossip, such as Sexual Rival Slander RG and Ally Maintenance RG. The studies from Hess and Hagen (2002, 2004a, 2004b) also confirm this.

Sexual Rival Slander RG and Ally Maintenance RG are of the most negative forms of gossip, which are more used by women than by men. These are the forms of gossip that attract most attention in our daily use, because of their negativity impact. Therefore, it is not so very surprising to me that in the popular use of the word ‘gossip’, gossip is linked with negative aspects, and with women’s talk. The direct association of gossip with women, in my opinion stems from the negative view on gossip most people uphold. This also follows from my focus group interviews on the use of Interpersonal Gossip (paper 1). I found indications that younger adolescents do not associate gossip only with girls. Their view on gossip is not as negative as older respondents. Older respondents’ answers support my idea that the associations of gossip with women’s talk stem from a negative view on gossip.

In short, both men and women gossip. However, what they like to gossip about is different for some of the 16 kinds of gossip information I distinguish for. And, women are expected to have a bigger tendency to use manipulative gossip, than men.

2.6 Why is gossip comparable to chocolate

A last question I here mention concerns my dissertation title. Why do I compare gossip with chocolate? There are several reasons. First of all, as I said in my introduction, both gossip and chocolate are addictive. Even though we sometimes decide to never eat chocolate again, and to never indulge in gossip again, we cannot resist. We can’t turn it off.

Second, chocolate makes us feel good. Eating chocolate stimulates the serotonin production in the brain. This chemical reaction brings us in a happy mood. It has been suggested by Dunbar (1998b) that this chemical reaction might be the same for exchanging gossip. However, these happy moods are not long-lasting. Hours after eating chocolate our mood swings from positive to negative, because chocolate makes your blood sugar and serotonin levels rise and then swiftly drop back down. A similar effect occurs when we gossip. At the very moment we exchange gossip, we feel good. We feel connected with our partners in crime. Gossip creates feelings of connection, because you talk about mutual acquaintances (at least when exchanging Reputation Gossip), and this explicitly makes you feel bonded to your gossip-partner(s). However, by the time we go home we might start regretting what we said, if our gossip messages were negative, because we fear retaliations of the gossipees. So, both the consumption of chocolate and negative loaded Interpersonal Gossip start off happily but imply some costs. Notice that positive loaded Interpersonal Gossip does not have this retaliation threat, and therefore can be considered to be even better than chocolate.

I here want to comment again, that the overall consumption of Media Gossip and all indulging in Interpersonal Media Gossip do not have this negative threat of retaliations. Gossiping about public unknowns and celebrities is using a ‘safe mode’ of gossip. Especially Interpersonal Media Gossip has extreme benefits. By talking about celebrities as parasocial mutual acquaintances with real life social network members, we feel (as if) we are connected through a mutual known gossipee, we feel bonded and therefore feel happy. And minutes, or hours after our gossip conversation we do not start feeling bad, even when we said something mean, nor do we feel threats of retaliations, because these media gossipees do not know us, and will not take revenge. So, if we want to feel good, a safe and easy to access tool to use is Interpersonal Media Gossip.

As a last note on why I associate gossip with chocolate, I have to refer to a quote of Steven Pinker. In his book How the mind works Pinker (1997) refers to music and fiction as being cheesecake for the mind: "I suspect that music is auditory cheesecake, an exquisite confection crafted to tickle the sensitive spots of at least six of our mental faculties” (Pinker, 1997: 534). Music and other forms of arts and fiction push pleasure buttons that trigger adaptations. Pinker explains how music triggers several mental faculties. As I have said earlier, I think this is true for certain types gossip (e.g. Media Gossip) as well. Gossip triggers several mental modules that in joint action produce the outcome of gossip behavior. But, while Media Gossip and Interpersonal Gossip are like art, triggering adaptations, but not resulting in adaptive-outcome behavior, in my opinion the 16 different kinds of Interpersonal Gossip I distinguished for do result in adaptive-outcome behavior.

3 Suggestions for future research on gossip

Although I have filled some important gaps of knowledge about gossip, research about gossip does not end with this doctoral dissertation. Gossip is tremendously complex, and fascinating to investigate. I therefore suggest future research to continue the search for a complete understanding of this dominantly present human trait. To guide their scientific adventures, I here outline some suggestions for future research. These suggestions are drawn from the fact that my own research did not cover the whole gossip-story, and some faults have inevitably slipped into my research project.

First of all, the classification system for gossip I put forward needs to be tested. Only if it can be shown that each of these specific forms of gossip exist cross-culturally, they can each pass Williams’ (1966) first criterion of ‘reliability’ to test whether they are adaptations or not. I funded this classification system on an adaptationist analysis, searching for functions that could have solved problems occurring in the EEA. If these 16 different kinds of gossip might have solved problems our ancestors have faced, we can expect these 16 different kinds of gossip to be present cross-culturally. If it can be shown that some or all of these 16 different kinds of gossip do not occur in every culture, my theory on gossip needs to be revised.

I have used a qualitative study (focus group interviews) to get an idea of the presence of the 16 kinds of gossip I distinguish for. However, no generalizations can be drawn from this study to a broader population. I therefore suggest future research to make use of quantitative research methods to investigate whether all 16 kinds of gossip can be traced back in our current daily conversations. I also advice future research to not only focus on pieces of information that can be classified in this system, but to also focus on pieces of information that can be classified as gossip in the most general sense, but still cannot be classified in any of the 16 sub categories I have distinguished for. This classification system is a first step towards a better understanding of gossip as a noun, but falsification and proof are necessary to turn this classification system into a valuable measurement tool.

Next, for future research on Interpersonal Gossip, I suggest to look at the discrepancies that occur when gossip is exchanged through different media. From my focus group interviews follows that young respondents indicate they use a lot of new technologies to exchange gossip messages. E-mail and especially messenger programs are favorite channels to transmit gossip. The older respondents I interviewed reported that they are very skeptical towards these new media to exchange gossip, because in their opinion gossip becomes less reliable and more negative when not transmitted face-to-face. I have not further investigated this, and think this is an interesting field of future research of gossip to look at.

Subsequently, research on Interpersonal Media Gossip must also focus more on sex differences in both the tendency to engage in Interpersonal Gossip, and the differences in topics gossiped about by men and women. Research on sex differences in Interpersonal Gossip is scarce, and I acknowledge that I have not contributed very much to fill this gap of knowledge. What I did do was presenting a more detailed theoretical framework that explains and predicts sex differences. This framework is funded on my classification system of Interpersonal Gossip as a noun that is footed on an adaptationist analysis.

Further, especially gossip about relations seems to appeal. Levin and Kimmel (1977) concluded from their analyses that gossip columns report a lot about the love affairs of gossipees. My content analysis of three Flemish gossip magazines confirmed this. In the context of Interpersonal Gossip, information about the love affairs of others as well can be very valuable to gossipers. In my theoretical framework I have stressed how gossip could function as a problem solver in the context of human mating. However, as I already mentioned above, this domain is still highly unexplored by both standard social scientists and evolutionary psychologists. In my empirical work I focused a little on Mating Gossip in paper 3, 4, and 8. Still, I suggest future research to explore this domain more thoroughly in the future, and this both in the context of Interpersonal Gossip and Media Gossip.

Then, Media Gossip in general is a fairly unexplored field of research. Content analyses of gossip magazines are scarce. David and McLeod (2003) have set up a nice cross-cultural study to investigate the presence of gossip on front-pages of newspapers in the last 200 years. A similar cross-cultural study about the content of Media Gossip in gossip magazines would be very interesting as well. Gossip magazines sell well in most countries. The codebook I used in my study of the Flemish gossip magazines, enlists traits/behavior categories derived from evolutionary psychology theories. These categories represent traits and behaviors that are expected to be present in every culture. Therefore this research instrument would be ideal to use in such a cross-cultural study.

Besides research on the content of Media Gossip, I also suggest future research to focus more on why people are so attracted to these messages. I have put forward to hypothesis to explain our interest in Media Gossip, and have tested these in my empirical studies. However, I here again comment that more, and especially cross-cultural research is needed to proof or falsify my predictions. In general, I conclude that interest in Media Gossip rests more on the fact that audiences want to learn, than on the fact that they want to keep up with their parasocial friends and foes. This contradicts some studies that have shown that the parasocial bonds between celebrities and their audience causes media consumption, and therefore the consumption of Media Gossip. Future research can give a better perspective on which of the two hypotheses I distinguished has more explanation power.

And last, Dunbar (1998b) has suggested that gossip stimulates the serotonin production in the brain. I think it would be very interesting to test this. I have explained above that negative loaded Interpersonal Gossip might stimulate an increase of serotonin production at the very moment the gossipers exchange their messages, but might later result in more negative moods because of fears of retaliation. This is not the case for positive loaded Interpersonal Gossip or for Media Gossip and Interpersonal Media Gossip in general. I suggest it would be very interesting to investigate (1) if these behaviors influence serotonin production, and (2) if this is different for the different kinds of gossip I here mention.

I know that still many mysteries about gossip need to be unraveled, and I can imagine as well that many more mysteries will rise along the future road.

Imagination is more important than knowledge.

Knowledge is limited. Imagination encircles the world.
Albert Einstein