PART I

THEORETICAL FRAMEWORK

CHAPTER 6. How gossip develops in a lifetime (ontogeny)

Gossip isn't scandal and it's not merely malicious. It's chatter about the human race by lovers of the same. Gossip is the tool of the poet, the shop-talk of the scientist, and the consolation of the housewife, wit, tycoon and intellectual. It begins in the nursery and ends when speech is past. (Phyllis McGinley - 1905-1978, poet and writer of children's books)

1 Introduction

In the previous chapters I outlined the evolutionary background of gossip (phylogeny). I put forward a functional analysis that differentiated several adaptive functions to gossip, and translated these in behavioral models. Following Tooby and Cosmides’ (1990a) guidelines for an adaptationist analysis, I still have to give a description of the inherited adaptations (underlying mechanisms) of gossip, and how these develop in a lifetime. In this chapter I will focus on the ‘hardware’ that enables us to gossip. That is, I will give a short description about the mental mechanisms that underlie gossip, and I will also shortly outline the crucial role of emotions in this process.

After this description of the nature aspect of gossip, I turn to the nurture aspect. I will describe how the mental mechanisms get activated because of environmental triggers, and how they develop. Doing this, I also want to stress how the environment in which we develop communication and gossip today is different from the EEA, and what the consequences are for this.

To conclude this chapter I resume the debate on sex differences in the tendency to gossip, to show that gossip develops differently in a man or woman’s lifetime because of biological predispositions and differences in communication skills and social networks nowadays.

2 Gossip hardware

Just like a computer needs hardware to run programs, in order for an individual to be able to gossip, he or she must have some hardwired mental mechanisms.

2.1 Mental mechanisms underlying gossip

According to evolutionary psychologists the human brain is no blank slate at birth, but consists of mental modules. In my opinion, several mental mechanisms underlie our gossip behavior. Gossip, as a form of communication, requires language, understanding of other individuals with whom we interact, and storage of information. I will first describe two crucial mental modules that are necessary for communication in general, and applicable to all different kinds of gossip, and then outline which other modules interfere with Strategy Learning and Reputation Gossip.

If any of these underlying mechanisms is missing in an individual, he or she will not be able to gossip in a way like other people, who do not lack these underlying mechanisms, can. Gossip, like most of our behavior is the outcome of hardwired mechanisms in interaction with a given environment. Both distortions in hardware and environment cause distortions in the behavioral output.

2.1.1 Communication: language and Theory of Mind

2.1.1.1 Language mechanism

Chosmky (1972, 1980) suggested that humans possess a Language Acquisition Devise (LAD). A language module, Chomsky argued, is required for young children to learn a language spontaneously. The LAD he talks about is a module that contains a set of grammar rules. The LAD provides toddlers a Universal Grammar set that can be applied to any given language. Before a child is taught any grammar rules, it can fall back on its LAD to get a notion on how to combine different words.

As Pinker (1994) has explained, the human language module is the hardware we need, to use language in a proper way. It is a computational device with some rules that define how to use language. These simple guideline rules enable us to refer to our environment, by using symbols, which saves our time and energy to get experienced say Pinker and Bloom (1992). Still, because a language module enables us to use symbols, this does not mean that these symbols are part of our language module, says Deacon (1998: 334-340). Only a few general structures are present in our language instinct, Deacon further comments: “Instead, the best candidates for innate language adaptations turn out to be some very general structural characteristics of the primary language medium itself, speech, and the computational demands this medium imposes when it comes to symbolic analysis.” (Deacon, 1998: 339). Language circuits are the hardware that allows us to use symbols, and ontogenetic and situational contexts will define what these symbols look like.

As Pinker says: “…if there is a language instinct, it has to be embodied somewhere in the brain, and those brain circuits must have been prepared for their role by the genes that built them.” (Pinker, 1994: 299). Generally speaking brain locations regarding language use are dominantly present in the left brain hemisphere, where the regions of Broca and Wernicke are situated. Broca’s area plays a crucial part in language production and the use of grammar (Bradshaw, 1997: 93-103; Hauser, 1998: 225-238; Mithen, 1998: 121-125; Pinker, 1994: 299-313). However, Broca’s area is not the sole area in the brain for grammar use, says Pinker (1994: 309). Proof of this, he says, stems from patients whose area of Broca has been damaged. Those people are not able to speak, but can still succeed on some grammatical tests.

In line with the above, Damasio and Damasio (1992) argue (1) that three sets of structures are needed in the brain to process language, and (2) that those are not exclusively located in the left brain hemisphere. First of all, language requires a collection of neurons (located both in the right and left hemisphere) that represent concepts. Second, we need a smaller number of neurons in the left hemisphere that represent language components such as phonemes, and syntax rules. Third, and last language rests on a set of structures (also located in the left hemisphere) that combines both; that can take a concept and stimulate the production of words. To conclude this section with Pinker again: “Why has it been so hard to draw an atlas of the brain with areas for different parts of language?[…] the brain is a meatloaf.” (Pinker, 1994: 314).

2.1.1.2 Theory of Mind module (ToM)

When communicating with each other, for instance by using gossip, being able to talk is not sufficient. One needs to be able to predict his or her communication partner’s actions to come to a fully communicative process. According to Malle (2002) research more and more leads to the conclusion that language co-evolved with Theory of Mind, because they are so very related to each other. The ability of humans to attribute mental states and interpret the action of other individuals is what Simon Baron-Cohen calls ‘to mindread’ (Baron-Cohen, 1994). In order to do this, humans possess what Baron-Cohen (1994) calls a ‘Theory of Mind Mechanism’. According to Baron-Cohen (1994) the neurocognitive ToM-system consists of four components: an intentionality detector (ID), an eye direction detector (EDD), a shared attention mechanism (SAM), which represents if an individual is paying attention to the same thing as others do, and an overall theory-of-mind mechanism, which coheres the other three components.

Together our language module and ToM module enable us to communicate with other individuals. These are the minimum requirements necessary in anyone’s mind to be able to gossip. According to the differences in the sub categories of gossip, different additional mechanisms are required for Strategy Learning Gossip and Reputation Gossip. I will discuss these now.

2.1.2 Strategy Learning Gossip: storing experiences

For strategy learning gossip (SLG) an individual of course needs language and ToM to communicate with others. Especially ToM is of great importance; receivers of SLG must be able to take the attitude of the gossipee and feel ‘as if’ what happened to the gossipee happened to himself or herself. The receivers must not only be able to grasp what happened to the gossipee, but transfer this experience fictionally to his or her own potential future and estimate if the outcome of the gossipee’s behavior would be the same for himself or herself.

Additionally to language and ToM, SLG requires a storage mechanism to store the strategy information. We can only benefit from the strategies of others if we can store these in some behavioral repertoire that can guide our own future decisions, says Guth (2000). He talks about a Master Module that stores behavioral information, and is used when we encounter similar problems in our own future. Next to storing, an individual must also be able to mimic the behavior of the gossipee. That language is involved in the mimicking behavior, can be confirmed by the research of Heiser et al (2003) who noticed that Broca’s area is activated when we observe the behavior of others. But next to Broca’s area, another neurological process underlies mimicking. When we observe the behavior of others, our brain fires mirror neurons, says Ramachandran (2000). Mirror neurons enable us to mimic the behavior of others, and allow a cultural inheritance that is liberated from a purely gene based evolution.

Last, other mental modules will interact with our language- and ToM modules to enable the functions of the different specific kinds of gossip. Survival SLG will make use of other mechanisms such as ‘predator avoidance modules’, or ‘food preference modules’, while Mating SLG will be sub structured by ‘mate selection modules’, ‘mate guarding modules’ and ‘intrasexual rival modules’ (for specialized modules see e.g. Buss, 1999; Gaulin & McBurney, 2004). Group-norms SLG will be founded on ‘cheater detection modules’ (Cosmides, 1989; Cosmides & Tooby, 1992) and ‘altruist detection modules’ (Brown & Moore, 2000). Each of these separate modules function in interpersonal interactions, e.g. the ‘mate selection module’ of an individual will be activated if a boy sees an attractive girl passing by. The core idea is that gossip rests on all different kinds of modules, which I will not all discuss here in detail. Our language- and ToM module in interaction with these specialized other modules created new options for adaptive behavioral output, which resulted in the different kinds of gossip. Our mind consists of a set of tools (mental mechanisms) that in recombination with each other make each different kind of gossip an adaptive tool to solve specific problems.

2.1.3 Reputation Gossip: storing person-related information

Where SLG focuses on behavior, RG puts the gossipee more central. RG is about person-specific information, and therefore requires mechanisms that can store such information. We must be able to keep track of someone’s traits and past actions, and recall this information when necessary. Again, the outcome of different kinds of RG, such as Kin Detection RG, Co-operation RG, Mates Detection RG etc. (see chapter 4), are the product of different specialized modules interacting with our language- and ToM modules.

For instance, Kin Detection RG, requires an underlying mechanism that enables us to differentiate between kin related and non-kin related people. Co-operation RG requires a mental ‘mechanism for cheating detection’, as shown to exist by Cosmides and Tooby (1992; Cosmides, 1989), and an ‘altruist detection module’ (see Brown & Moore, 2000). Alliance RG rests on ‘alliance formation modules’, which are shown to exist by the research of Kurzban, Tooby & Cosmides (2001).

Again, I stress that these specialized mechanisms were originally selected to operate in an interpersonal context. Our cheater detection module enabled our ancestors to detect cheaters in their real life interactions. With the emergence of language, and recombination of the function of these specialized modules with language, new options arose, such as e.g. detecting cheaters by communicating about the cheating behavior of others (cooperation gossip).

2.2 Emotions as gossip regulators

2.2.1 Emotions as mechanism coordinators

Because each of the different kinds of gossip rests on the activation of different modules, some orchestration is required to secure a fast and correct outcome. This is where emotions play an important role. Emotions guide our communicative behavior (Buck, 1988; Cosmides & Tooby, 2000; Dimberg, 1997; Ekman, 1982; Evans, 2001). As Segerstrale & Molnár (1997: 14) say: “We are biologically preprogrammed for sociality and continuously monitor one another for emotional cues that serve as guides to future interactions.”

According to Cosmides and Tooby (2000) emotions can be regarded as some super ordinate program that functions to coordinate activities and interactions. Emotion programs detect evolutionary reliable cues signaling that a certain situation is present. Cosmides and Tooby (2000) give an example of how the emotional experience of ‘fear’ activate behavioral decision rules to deal with potential threats; different actions will be considered, such as hiding, flight, or self-defense. Ledoux (1998) supports the same idea and gives a detailed description of how fear guides or fight-or-flight behavior when being in a threatening situation (for more detailed report see Ledoux 1998). Another example, Cosmides and Tooby (2000) mention is how the feeling of ‘guilt’ functions to recalibrate trade-offs in welfare between self and other. Similarly Damasio (1996: 169-177) gives the example of how emotions guide our altruistic behavior. We do not rationally calculate the costs and benefits of altruistic actions, that would consume too much of our effort and most often not easily lead to an outcome. Instead, we are driven by what we feel, our emotions guide our altruistic actions in a much faster and accurate way.

2.2.2 Emotional regulation of gossip

“The fact that many who disapprove of gossip find themselves engaging in it testifies to its emotional power.” (Spacks, 1985: 10-11). Different emotions indeed play an important role in our daily gossip behavior. Ekman (1982, 1994) distinguishes six basic universal emotions: happiness, sadness, angriness, fear, surprise and disgust. Each of these basic emotions drive us to gossip. And, besides these basic emotions, higher cognitive emotions play a crucial role as well. Higher cognitive emotions are, as Evans (2001) says, social emotions like love, guilt, shame, embarrassment, pride, envy, and jealousy. These emotions helped our ancestors cope with increasingly complex social environments. Unlike the basic emotions, these higher cognitive emotions can be different according to cultural differences. It also takes a longer time to build up and get rid of these higher cognitive emotions, says Evans (2001). Jealousy for instance can last for a long time and will sometimes motivate us to negatively manipulate the reputations of rivals through gossip over an extended period of time. Nevertheless these higher cognitive emotions are of great importance in the context of gossip. What I want to put forward here is how our basic emotions trigger our gossip behavior. We need short-time emotional triggers to get the gossip started, and be continued.

2.2.2.1 Surprise: gossip and the unknown

I already stressed the importance of surprise in chapter 1, when I put forward a general definition of gossip. Traits and behaviors become gossip if they elicit feelings of surprise in the receiver, as I explained with the example of ‘brushing your teeth’ in different contexts (see chapter 1). Surprise is the emotion that drives our initial interest in gossip. It focuses our attention on unexpected events, changes in someone’s behavior pattern, fitness-relevant strategies that were so far unknown or unclear to us. Surprise is the emotion that shifts gossip from other forms of communication about third parties.

2.2.2.2 Happiness: why gossip is like chocolate

Words are powerful antidepressants (Evans, 2001). Feelings of joy emerge when a receivers gets gossip information that can increase fitness opportunities; fitness-promoting SLG, Kin Detection RG, Altruist Detection RG, Ally Detection RG and Mates Detection RG for instance give information to a receiver about future potential opportunities to use strategies that will increase his or her fitness, future potential opportunities to cooperate or mate, etc. This gives joy and excitement for the receiver, and to the sender as well if the benefits can rebound to him or her.

Gossip sometimes elicits feelings of catharsis, of joy (see chapter 2). Dunbar (1998a,b) compared human gossip to grooming among other primates. When primates groom each other, endorphins, natural opiates, are released in their brains. If Dunbar’s argument is true, that gossip and grooming are comparable, each functioning to bond individuals, then it might be the case that endorphins are released when we engage in a gossip conversation, which could explain our feelings of joy and catharsis. Dunbar (1998a) reports that laughter is associated with the stimulation of opiate production, so that telling jokes literally make us feel good. Telling jokes stimulates opiate production, chemically making us feel bonded with our coalition partners, and since communication requires less time investment than physical grooming, this joking is a cheap substitute that allows us to invest our precious time in other activities:

“Telling jokes allows us to stimulate opiate production in our grooming partners even when we don’t have the time to sit there doing it physically. We can get on with the other important activities of ecological survival – traveling, hunting gathering, preparing and eating food.” (Dunbar, 1998a: 191)

He further agues that the same might be true for gossip. Gossip contains a lot of humoristic aspects, and has been compared to humor (Morreall, 1994; see also chapter 2). However, gossip is not always a stimulator of happiness. We might feel happy at the very moment of the act of gossip, feeling bonded with our partners in crime when for instance telling negative reputation gossip about non-coalition members. Still, as time passes we might start to regret our acts, fearing recursions. Gossip is an instant stimulator of endorphins, but can potentially turn out bad. In this sense, gossip is comparable to eating chocolate. Evans (2001) reports that it is widely known for humans that chocolate stimulates the release of endorphins, and therefore eating chocolate stimulates feelings of happiness. But:

“However, research has shown that, while most people feel more positive and energetic immediately after eating a chocolate bar, this effect soon wanes, and an hour afterwards they tend to feel even worse than they did before eating the chocolate in the first place.” (Evans, 2001: 96-97)

Since most often the same is true for gossip, namely that gossip is tempting and first boosts our happiness but can turn out bad on the long run, I consider gossip being the “chocolate for our universal minds”.

2.2.2.3 Sadness, disgust, and angriness: when gossip is bad

Feelings of sadness, disgust, and angriness will be present when the content of gossip turns our bad. We might feel sad and/or angry that we do not succeed to reach a certain goal. These negative emotions will then motivate us to search for new options, and trigger our interest for Strategy Learning Gossip. Complaining to your friend that you have tried to conquer the heart of that one girl in so many ways, but have not succeeded, might trigger your friend to give you some good advice ‘Do this and then you will succeed’ is a form of problem-solving Strategy Learning Gossip. So feelings of sadness, and angriness will urge an individual to express his failures, to which another person will react by giving solutions to solve the problem.

Applying these emotional drives to Reputation Gossip results in somewhat similar thoughts. Sadness, angriness, disgust, just like other negative higher cognitive emotions will appear if a specific person has done something wrong –defected on someone, violated rules- and we spread around these events. These negative emotions are lighters of reputation gossip, they motivate senders of bad RG. An endurance of this sending is secured by our negative higher cognitive emotions, such as jealousy. Jealousy among rivals, for instance, in the context of allies or mating will start gossip to be spread around. As Hess and Hagen (2002), Power (1998) and Kuttler et al (2002) have shown, young girls gossip when they feel jealous about other girls with whom they compete for access to boys and allies.

2.2.2.4 Fear: when gossip is frightening

Fear arousals are related to precautions, and will play a role in Strategy Learning Gossip that warns for fitness-endangering situations. When we fear that bad things that happened to others could happen to ourselves or people we care for we will warn our friends by sharing important SLG about dangerous situations. Fear triggers our tendency to share SLG to protect ourselves and the people we care for.

That fear triggers not only the tendency to share warning SLG about negative events, but also motivates us to share more joyful gossip stems from the research of Anthony (1973). Anthony (1973), analyzing the spread or rumors about the closing of a school, predicted that anxious people should be more tended to spread rumors around. Although her research focuses on rumors and not on gossip, her explanation is still useful. Anthony says that anxious people spread more rumors than non-anxious people. Anxious people like positive, pleasant rumors because these act as anxiety-reducing, while an unpleasant rumor is desired by anxious people because of precautionary reasons: you can better be prepared of the scary things that can happen to you, she says. To test her hypothesis, Anthony (1973) presented the rumors to 88 high school students. Using the Taylor Manifest Anxiety Scale, she could control for the anxiety level of her respondents, and split up her group in high anxiety and low anxiety subgroups. Her results show that “Since the two high anxiety groups had almost complete transmission, while the low anxiety group had very little, tentative support is given to the anxiety hypothesis of rumor transmission.” (Anthony, 1973: 97).

This role of fear is present in the spread of gossip as well say Rosnow and Fine (1976):

“Like rumormongering, gossiping appears to be mediated by subtle anxieties and thus seems directed at reducing or avoiding possible strains or diffusing a discomfort. This is not to say that gossiping cannot arouse anxiety; but one person’s dissatisfaction may be a source of comfort to another.” (Rosnow & Fine, 1976: 87)

To finish this part on the role of emotions in our gossip behavior, let me conclude with a statement of Rosnow and Fine (1976) that nicely outlines how different emotions interact when they drive our gossip actions:

“Chronic and acute anxieties behave like the transmitter chemicals which aid in the transfer of an excitation from one neuron to another. When anxieties are concentrated, they should increase (up to a point) the efficiency of rumor and gossip transmissions. Hence, with rumor, ambiguity is the proponent energy used to excite, anxiety is the transmitter chemical which speeds information on its way.” (Rosnow & Fine, 1976)

We first feel surprised when we observe some remarkable event, or hear a gossip. Our attention is caught. In a next step, joy, fear, anger, sadness or disgust will make us spread around the message to others, who again will first feel surprised, and then caught by one or more of the other emotions that trigger our tendency to share this fitness-relevant information with others again.

3 Activating and running the gossip hardware

Mechanisms are necessary for every individual in order to develop normal behavior. If people lack language mechanisms they will not develop language. If people lack emotions, they are sometimes worse off than people who do have emotions. As Evans (2001) says, if someone would lack fear, he or she might wonder to approach a tiger or not. If we would lack feelings of disgust, we would eat rotten food and get sick, etc. Our mental modules are a necessity to stay alive and what makes the activation of such mechanisms unique is that we acquire them without putting in too much effort. As Cosmides and Tooby (1994a: 330) say: “Psychological mechanism should develop in all normal human beings, without any conscious effort, and without any formal instruction.”

3.1 Development of communication skills

In what follows I will give an overview of how gossip, as communication, emerges and develops almost spontaneously in almost any human being.

3.1.1 First emotions

A newborn baby cannot talk, but does show emotions already. Damasio (1996: 131-139) supposes that basic universal emotions are hardwired and present at birth. That we do not copy our emotional expressions from seeing others laugh or cry, stems from the fact that newborn blind babies show emotions as well (Evans, 2001). However, as Damasio (1996) comments, even though babies show emotional reactions on certain stimuli, such as sounds, movements, shapes, these are not yet associated with objects or subjects.

3.1.2 Perception of objects and subjects

Next to showing emotions at birth, newborn infants show a preference for faces, and recognize faces. As early as a few days after birth babies show more interest in social actors than any other object. Humans, or pictures of faces or objects that resemble faces elicit interest in babies (Smith, 1988). This supports the idea that humans have an innate preference for faces.  (for an overview see Johnson & Morton, 1991).

3.1.3 Ontogeny of language

After emotions and recognition of interaction partners is established, infants start to develop language, they develop speech. That this happens very rapid and automatically, considered the complexity of a language system, again supports the idea that innate mechanisms for language are present (Pinker, 1994).

Language comprehension precedes language production, says Pinker (1994). Before his or her first birthday a baby understands words and some time after this first birthday he or she starts to produce them. The reason for this production delay is due to physical constraints. Until the age of one, babies’ larynx is located at a higher level, to prevent choking when drinking. Around the age of one the larynx descends, which enables the infant to produce new sounds, which are necessary for speech production (Hauser, 1998; Savage-Rumbaugh, Shanker & Taylor, 1998: 73).

Still, the idea that language comprehension comes before language production in children, has been refuted by the research data from Reznick and Goldfield (1992), who suggest that both occur at the same time. A spurt in the development of both language comprehension and production occurs in the second year, and follows the insight of the child that names refer to things, which they realize at the end of the first year. In any case, by the age of two a toddler is able to speak and hold a (basic) conversation.

A mental mechanism does not get activated if the proper environmental stimulus is not present. If we never encounter snakes in our entire life, our fear mechanisms for snakes will not get activated. The same is true for language. If an individual would never encounter another human being, he or she would not develop language, because a minimum requirement for our language module to get activated is speech of other human beings, says Pinker (1994). Of course it seems hard to test this; we cannot lock up an newborn infant, depriving it from any social contact and investigate if he or she would develop language. However, sadly enough, these cases do exist. Feral children, who are deprived from social contacts with other human beings, are an interesting case to investigate whether certain mental mechanisms, such as our language module, get activated when the social trigger of other human beings is lacking. Data on language use of wild- and feral children all shows the same result; these individuals are mute:

“The outcome is always the same: the children are mute, and often remain so. Whatever innate grammatical abilities there are, they are too schematic to generate speech, words, and grammatical constructions on their own.” (Pinker, 1994: 277)

Researchers, and others, have tried to teach language to feral children, after they were liberated from their isolation. A well known case is Genie, who was locked in a room, isolated from any social contact till the age of 13 (see Newton, 2003). She managed to speak an immature language, but never succeeded to master the full use of our human language. The reason for this is because the activation of our language module should occur in the first stage of our life. There is a critical period for language acquisition, says Pinker (1994). If children are not surrounded by a speech-rich environment before the age of six, they will not develop a full use of language:

“In sum, acquisition of a normal language is guaranteed for children up to the age of six, is steadily compromised from then until shortly after puberty, and is rare thereafter. Maturational changes in the brain, such as the decline in metabolic rate and numbers of neurons during the early school-age years, and the bottoming out of the number of synapses and metabolic rate around puberty, are plausible causes.” (Pinker, 1994: 293)

The ontogeny of language nicely illustrates that human behavior is not about nature or nurture, but about nature and nurture. If individuals lack a biological basis, they will not develop full language use. The same is true for individuals who are deprived of the proper nurture aspects, if they miss an environment of speech, they will not develop language. The critical period even more proofs that nature and nurture go hand in hand. If the nurture aspects are not present at the time wherein the nature aspects are ready to be activated, then language does not develop completely. If language would be completely learned, people such as Genie, would be able to speak like other adults after sufficient training. Unfortunately for them, this is not the case in reality.

3.1.4 Ontogeny of social intelligence

The last step for the infant to take in the process of becoming a skilled communicator is developing Theory of Mind. Learning a language is a social skill. Every time a child learns the meaning of a word, says Bloom (2001), he or she learns something about the thoughts of other people. The basis of developing ToM concerns eye contact and recognition of other human beings, which occurs early after birth. However, a notion that the mind of others differs from one’s own mind only develops at the age of four to five. To test this, researchers use false-belief tests, such as the classical Sally-Ann tests, as first developed by Wimmer and Perner (1983). Baron-Cohen et al (1985) used an adapted version of this test to explore the development of ToM with young children and concluded that until the age of four, children cannot solve these tasks properly.

The Sally Ann test presents children with pictures of two dolls. In this story Sally and Ann are in a room and Sally hides a marble in a basket, covers it and departs. While Sally is gone, Ann moves the marble from the basket and puts it in a box. Then Sally enters the room again, and then the question for the respondents follows: “Where do you think will Sally look for the marble?”. Children up until the age of four, lacking a fully developed ToM answer “In the box”, whereas children age 4-5 and up answer correctly “In the basket.” Children who have not yet developed ToM cannot understand that the mental states of others differ from their own mental state. Children with autism do not develop a well functioning ToM. The performance of autistic children on the Sally-Ann test is below average (Baron-Cohen et al, 1985). Children and adults with autism lack ToM, they show; Baron-Cohen (1995) calls it mindblindness.

Most autistic children have severed problems using language, says Bloom (2001). Exceptions are people with Williams Syndrome (WS), a special form of autism. These people do not have severe damage in their social abilities, and can use language just like normal adults do. From this it follows that the degree of language deficits goes hand in hand with the degree of distortions in an individual’s ToM:

“In general, then, the extent of the language deficit found in autistic children may be a direct function of the severity of the theory-of-mind deficit. A severe theory-of-mind deficit might leave children without the ability to orient preferentiality to speech, share attention, or follow eye gaze, and they might never be able to grasp the notion of an arbitrary sign, leading to no word learning at all. A less severe impairment might make word learning possible but limited and idiosyncratic. And in some cases, the theory-of-mind impairment might be sufficiently mild so as to leave word learning fairly unimpaired, although such individuals might still have problems with aspects of language such as irony and metaphor.” (Bloom, 2001: 80)

In sum, it takes a child up to the age of 4-5 to fully develop a communication system, taken into account that this child does not have any biological defects (aphasia, deafness, muteness, etc.) and grows up in a normal social environment. At the age of five, under these right conditions, a child must already be able to communicate with others, and since the most important mechanisms for gossip (language and ToM) are activated, children of that age must be able to grasp gossip as well. Whether they will be ready to use all the different kinds of gossip, I distinguish, depends on the activation of other specialized mechanisms that interact with language and ToM. For instance, Mating SLG, Mates Detection RG,  Sexual Rival Slander RG, and Mates Control RG will only occur at a later age, if an individual’s mating mechanisms get activated.

3.2 How our social network develops

Next to discussing how communication develops in a lifetime, I will now give a brief overview of the development of individuals’ social network. Gossip requires gossip partners, and who these are depends on the structure of our social network.

3.2.1 From parents to friends

As a baby, we explore most of our world in the company of an adult (mostly the mother). Babies of 7-9 months old become attached to their parents and feel safer to explore the environment in their presence. In the period from one to three years of age other kids become increasingly important for our development. Young children start to engage in play, and especially social play with other children, which is of great importance to learn how to interact with others and be able to predict the behavior and intentions of others (Smith, 1988). Play is an important aspect of the human life span (see also chapter 4), and outlines the first social context wherein children and young adults learn to interact with others. Family and friends are the social partners with whom an individual first uses gossip.

Social environments of children are not as complex as those of adolescents and adults. Children grow up in their family, and besides going to school and playing with friends, they are not part of many different social groups. In my opinion, the modern childhood might not be very different from our ancestral children’s social environment. Even the fact that nowadays babies are often raised in children’s day-care, might not be much different from hunter gatherer societies where collective childcare was not uncommon either (Buss, 1999).

3.2.2 Complexity of social relations in industrialized societies

For adults, the industrialized world frames individuals in much more complex social settings. Nowadays we do not live in small cooperative groups of kin related people, cooperating and being in conflict with other small groups of kin related people. As adolescents we live with our family, attend school and engage in after-school activities. As an adult we get study-companions if we go to the university, we get colleagues, and every time we switch jobs we get new ones, our sexual partners might open up new social networks, leisure time activities, such as going to the gym, taking art classes, etc.

Emler (1994) distinguishes three clusters of social relations in our current industrialized societies. First of all he talks about our ‘personal relations’ with acquaintances, friends, family, and so on. These contacts are characterized by quality, loyalty, and personal ties. Second, he mentions our ‘structural relations’, which are formal relations like in an office. Last, we are all part of ‘categorical clusters’, relations with others based on categories like gender, race, class, nationality, etc.  And as Emler (1994: 128) comments “Young adults interact mainly with people they know personally.” But as we grow older personal and structural relations become equally important and interfere with each other.

The options for adult social contacts are plenty in our current societies, and contrast from the simple structured environments our ancestors lived in. For our ancestors it might have been fairly easy to decide whether another individual was an ally or not, but for modern human beings this has become extremely complex. Especially since the social groups we are part of often overlap and our social network becomes a spider web of interactions between different members. Nowadays we no longer have allies and non-allies, but we have school-allies, university-allies, colleague-allies, gym-allies, art school-allies, etc. and for every social activity we have our non-allies as well. Not all colleagues are allies, not all people from art school are allies etc. If this were not complex enough for an individual to handle, he or she also has to take into account that some of our school-allies might form themselves allies with people from the gym we know as well, e.g. with our art school-non-allies. Your one colleague whom you consider as a friend (ally) might be befriended with your worst enemy at work (colleague non-ally).

This complexity in social structure affects our gossip behavior of course. As I outlined in the chapter 5 distinguishing between allies and non allies is crucial to decide with whom an individual will share Strategy Learning Gossip and about whom an individual will spread good or bad Reputation Gossip. Since it is hard to make a difference between allies and non-allies in our modern societies, gossip might be more difficult to use in a functional way, than it was for our ancestors.

4 Ethnographic studies about gossip across the lifespan

To support the above described ideas how gossip develops over the human lifespan, I will now present some ethnographic studies that focus on gossip and age. I will start to sum up some studies about children’s use of gossip and discuss how the use of gossip changes across the lifespan.

4.1 Children’s gossip

Studies about gossip among children are rare. Mettetal (1983) analyzed the role of gossip among young girls. Her conversation analyses shows that girls at the age of 6 to 7 do not gossip very much, but as age increases gossip becomes more and more important and by the time they reach the age of 11-12, a third of girls’ conversations are gossip. In choosing their gossip partners these young girls are not very selective, it seems. Mettetal had predicted that gossip would occur among friends, and less among acquaintances, but her results show that young girls gossip with both equally.

Another study on gossip among children shows data on both girls and boys. Fine (1977) reported about gossip among very young children (from the age of three). He says that by the age of three, toddlers start talking about the traits and actions of other children (Fine, 1977). According to Fine (1977) gossip plays an important role in the education of children. Through gossip they learn about the social environment they are living in, they learn what is allowed to do and what not, and by telling tales, their memory is being trained.  With this he refers to the use of what I call strategy learning gossip (SLG), which indeed is most valuable for the least experienced among us, which children often are.

Next to the presence of SLG, also reputation gossip (RG) prevails among children, according to Fine’s (1977) reports on children’s gossip. For instance, he describes that when a child performs poor in class, this will be spread around by other kids. Small children might not be as good in this manipulative strategy, but “By the time a child reaches preadolescence, he or she is sufficiently aware of social interaction use of gossip as a means of impression management as effectively as adults do.” (Fine, 1977: 184). The reason why small children perform less on using reputation gossip in a manipulative sense towards other children, might be because they have not yet a fully developed Theory of Mind.

Kuttler et al (2002) for instance, conducted experiments about the understanding of manipulative gossip with preadolescents, with an average age of 9 years old. They concluded first that children of that age were capable of using manipulative gossip. And, to their surprise they noticed that younger children were more skeptical towards gossip than older children. I think that younger children might have difficulties with understanding that what others believe about someone might differ from their own beliefs about that person. Around the average age of 4-5 children understand that others’ beliefs might differ from their own. Children who are of that age might not belief that another child’s belief about a gossipee differs from their belief about that gossipee. And even though they have developed ToM, it might take some time to apply this correctly on the beliefs about other people, and higher-order beliefs such as ‘he said that she said..’ utterances.

Different from adult gossip is that children do not feel embarrassed to admit they gossip. Gossip is not regarded as something bad by kids, and as a result they even gossip in front of the target. This is something that will stop occurring frequently when they turn in to adolescents (Fine, 1977).

4.2 Does gossip decrease with age

There are several reasons to assume that younger people benefit more from gossip than older people. Emler (1994), for instance said that younger people (adolescents) have more personal social contacts than adults do. Adults spend a lot of their time in structural relations with colleagues. It is arguable that this offers more potential situations for adolescents to gossip, because they have more social contacts, and interact most often with personal related others with whom they have shared allies and non-allies to gossip about. Sitting in an office all day, where you might communicate with some colleagues about other colleagues you mutually know, is different from hanging out with a large number of school companions and friends who all know most of your own friends and acquaintances. From his observations of a Black American ghetto in Washington D.C., Hannerz  (1967) concludes that:

“Younger people are usually more mobile; they therefore develop and maintain contact, although not necessarily frequent contact, with a larger number of people. Men continue to do so into later life to a greater extent; women become more tied down by their families and develop somewhat more intensive contacts with a smaller group of other women in the more immediate environments, usually including close female relatives such as mothers, sisters and (later) daughters.” (Hannerz, 1967: 51)

Having more social contacts to gossip about and gossip with, opens more options for adolescents to use Reputation Gossip.

The same can be said about Strategy Learning Gossip; having more social agents to gossip with offers more opportunities to exchange experiences. Additionally, I have stressed before that SLG is most beneficial for the least experienced individuals, and across the lifespan younger people can benefit more from gaining experience information than older, more experienced people.

However, even though interest in gossip might be higher at a younger age, I do not say that this interest disappears as people grow older. Saunders (1999) collected data on gossip among elderly women, aged between 75 and 102, who were all part of a SOWN group (Supportive Older Women’s Network) in Philadelphia. She reports that these women gossip quite a lot. They gossip about group members, non-group members and celebrities. The topics most gossiped about are fashion, dating and problems at the Center. One could argue that Saunders only collected data on women’s gossip, and that gossip about elderly men might not flourish that much in elderly homes. And this brings us to the final debate on interpersonal gossip I wish to discuss.

5 Sex differences in gossip

In chapter 3 I outlined how sexual selection and sexual division in labor imposed different adaptive problems to our male and female ancestors. Biological sex differences will cause men and women to behave differently. What I here overview is how gossip nowadays still develops differently for men and women.

5.1 Sex differences in communication skills

Due to sex differences in the division of labor in our ancestral past, women have better communication skills argues Joseph (2000). Women’s activities such as child rearing, gathering, and domestic tool construction and manipulation contributed more to the functioning of their Broca’s area than men’s hunting activities, which do not require speech, he says. Women are therefore superior in grammar and vocabulary use. This difference in evolutionary pressure resulted in sex differences in the functional organization of the brain when it come to language. When women speak, both their left and right brain regions are activated, whereas for men only the left brain regions light up (Shaywitz et al, 1995). Although statements like this have been contested, most evidence is in support of this view (see e.g. Kansaku, Yamaura & Kitazawa, 2000).

Starting at an early stage in life, sex differences in verbal skills still occur today. Even at the age of two, when children start talking, sex differences are present, with girls performing better than boys (Galsworthy, et al, 2000). When looking at the difference in performance of boys and girls aged between 5-16 on verbal tasks, Kramer et al (1997) concluded that young girls overall perform better than young boys. These difference are maintained at later stages in life. Looking at studies of episodic memory, for instance, women seem to score better than men. But not on all episodic tests, say Lewin, Wolgers and Herlitz (2001). They compared men’s and women’s performances on verbal, nonverbal and visuospatial episodic memory tasks. Results show that women get better scores on the first two, but men outscore women on visuospatial episodic memory tasks. This is, because men overall perform better on visuospatial tasks, and they can use their expertise in these verbal tasks as well.

Miller (2000) has contested the idea that women are the verbally most active sex. According to his opinion sexual selection boasted men’s verbal skills. Proof of this he gives by referring to the fact that the creators of literary products such as books, music texts, and poetry are dominated by men. He further argues that men use their verbal skills to impress women. As soon as they have conquered the heart of a woman, they let go of these verbal seducements, which results in the complaint of married women that their husband does not talk enough. I think his theory has to be looked at with some skepticism. I agree that literary products, as forms of arts, are used by men to impress women, and that women might prefer creative over non-creative men. However, I suggest that we should not mix up verbal, language skills with literary products. Written texts are recent products in evolutionary terms, and not comparable to oral skills, which go back a lot longer in our ancestral past.  It might be that men outscore women on cultural productivity, but this is different than to claim that men would perform better on verbal skills than women do.

As a last note I mention that, besides being better skilled, women’s language also contrasts from men’s speeches in content. Women’s language today is often depicted as being more emotional than men’s language (Tannen, 1991).  Women talk more about relationships and feelings, they talk more about other people, and their stories focus on community. Men talk most about themselves; they discuss contest situations, and talk less about feelings (Romaine, 1999).

5.2 Sex differences in sociality

“Boys tend to have larger social networks than girls, who usually have one or two girlfriends with whom they play regularly. To some extent the size of these groups may be determined by the different types of activities they engage in. It takes only three girls to skip rope or two to play house, while more boys are needed for team sports such as football.” (Romaine, 1999: 123)

Romaine’s (1999) arguments could easily be countered with similar examples as ‘it only takes one boy to build blocks, two boys to race cars, but it takes a team of girls for a ballet performance’. But still her argument has been supported. Girls nowadays remain the socially most active gender it seems. This difference starts already at a very young age. Analyzing the social behavior of 3rd-, 4th-, and 5th-grade children Lansford and Parker (1999) noticed that girls are more intimate; they exchange more information, whereas boys are more aggressive and solitary in their play behavior.

And even at an adult age these differences remain says Putnam (2000):

“Informal social connections are much more frequent among women, regardless of their job and marital status. Married or single, employed or not, women make 10-20 percent more long-distance calls to family and friends than men, are responsible for nearly three times as many greeting cards and gifts, and write two to four times as many personal letters as men. Women spend more time visiting friends, though full-time work blurs this gender difference, by trimming friends for both sexes. Keeping up with friends and relatives continues to be socially defined women’s work. Even in adolescence (an not only in the United States), women are more likely to express a sense of concern and responsibility for the welfare of others – for example, by doing volunteer work more frequently.” (Putnam, 2000: 94-95)

Ben-Ze’ev (1994) has 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 with Ben-Ze’ev that indeed women have stronger allies with family and friends, and men engage more in looser contacts with peers and co-workers. However, I do not think these differences are due to cultural expectations as he says, but rather are the product of different selection pressures of men and women. Uchida (1992) suggests to not solely looking for causes in sex differences in communication within cultural influences. She discusses how both the “Difference/cultural” and “dominance/power-based” approach attribute gender differences in communication style, due to cultural differences between the sexes. The first approach explains how the communication of men and women are different, but is equal. The second approach claims that gender differences in communication are due to the general dominance of men in societies. But as Akida (1992) comments, cultural difference alone cannot explain our language difference and miscommunications. Men and women are different at birth, and gender as a social construct only explains part of the differences in language use of men and women. Akida (1992) urges researchers to use a framework that allows to explain how through language use, gender is created. To my opinion evolutionary psychology offers this alternative approach. For more details I refer to chapter 3 where I outlined the sex differences in sociality from evolutionary perspective.

5.3 Sex differences in gossip

The last and central question for this chapter to be solved is maybe the most discussed one, when it comes to gossip: “Why do men care so much about gossip if it were solely a women’s business?”

“If gossip were used by women to talk only abut makeup and clothes, it would be of no concern to men; it is because we know very well that gossip spills over into topics of choice and equality that the patriarchy has good reason to worry about women’s gossip.” (Ayim, 1994:  98-99)

and

“We are all familiar with the stereotype that women ‘gossip’ and ‘chatter’ while men ‘talk shop’, but actual research reveals that men talk much more than women across a wide range of contexts, e.g. in husband-wife interaction, TV discussions, meetings. Women are expected to remain silent, so when they talk, it is noticed and commented upon negatively.” (Romaine, 2000: 122)

Women might be the socially most active gender, having larger and more active social networks, being offered more opportunities to gossip. But by no means can anyone claim that men refrain from gossip. An observation where Gelles (1989), as a researcher on women and gender, put focus on was the fact that in the case study about a romantic affair in 18th century France and Great-Britain she analyzed, a man was the main subject. The fiancé is both the subject of gossip, and an active participant. “While a man is central, clearly both men and women gossip in this tale, which is far closer to the reality that we know to be the truth of the issue of “Who gossips?”.” (Gelles, 1989: 676).

Topics might differ, but when it comes to devotion of time to gossip, some researchers even claim that no sex differences exist. But let me start by the beginning, sex differences among young children.

5.3.1 Sex differences in gossip at a young age

Already at the young age when children start to gossip, boys as much as girls engage in this activity. Evaldsson (2002) observed preadolescent boys, aged between seven and ten at a Swedish after school program for one and a half year between 1989 and 1990. The boys came from low-income suburbs of Sweden. Evaldsson analyzed their gossip conversations, and noticed that boy’s gossip did not differ much from girls’ gossip. She compared her results with those from Goodwin (1980) who observed the gossip among preadolescent girls in an afro-American society. Both girls and boys gossip about the deviant and inappropriate behavior of other children. Girls, more than boys ask teachers to interfere when conflicts occur, and boys tend to talk more behind the back of others.

Still the major difference between girls and boys according to Evaldsson (2002) is the fact that boys use gossip to gain power in a group. Two-against-one situations are established through gossip.

“Records from long-term fieldwork demonstrate that gossip telling does important work for some of these boys with low-income backgrounds to position themselves as powerful and normal while others become deviant and powerless.” (Evaldsson, 2002: 221)

The boys’ gossip language is typically masculine, dominating and used to establish power. Evaldsson (2002) also comments, that even though girls’ and boys’ gossip can be used very similar to report about the deviant behavior of others and to solicit support and strengthen solidarity, this does not mean that sociality is as important for boys as it is for girls:

“These similarities, I would argue, do not indicate that the boys’ social organization has similarities to the girls’ social concerns, but that boys and girls construct overlapping social identities in social interaction.” (Evaldsson, 2002: 221)

5.3.2 Sex differences in adult gossip

“To begin with the obvious: men as well as women gossip. Both sexes acknowledge this fact, men with an air of large concession (“even we sometimes indulge in this foolish female activity”), women with defensive or aggressive energy (“men do it too, it can’t be so bad”).” (Spacks, 1982: 21)

When Fine and Rosnow (1978) wrote an article on what gossip really is about in 1978, they did not know of any sex differences in adult gossip “Yet there has been no research addressed to the correlation between sex and gossiping.” (Fine & Rosnow, 1978: 165). They did report however about data that said that women reveal more information about themselves than men do. And, since Fine and Rosnow consider self talk as a form of gossip, this might in their context mean that women gossip more than men. However, later Dunbar, Marriott and Duncan (1997) –who also consider self-talk to be gossip- showed that men talk more about themselves and use the argument to claim no sex differences exist when it comes to the amount of time we spent to gossip.

More profound studies of sex differences in gossip (with the exclusion from self talk) stem from Levin and Aluke (1985) and Nevo and Nevo (1993, 1994). Both researchers focused on sex differences in amount of time dedicated to gossip, and the content of these conversations.

5.3.2.1 Do women gossip more than men

Eavesdropping the conversations of 76 male and 120 female student college students in the US, Levin and Arluke (1985) focused on the amount of time spent to gossip and what is gossiped about. 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 their tendency-to-gossip-questionnaire, Nevo and Nevo (1993, 1994) noticed that female respondents got higher scores than their male colleagues.

5.3.2.2 What do men and women gossip about

However, looking more into detail, Nevo and Nevo concluded that only one topic is gossiped about more by women than by men, and this is the physical appearance of others. This is different from the result Levin and Arluke (1985) found in their eavesdropping study. Focussing on the topics discussed in gossip, 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) (Levin & Arluke, 1985). Maybe men do not like to admit tot talk about the physical appearance of others. Using questionnaires, we must always take into account that people sometimes answer socially desirable answers, which might explain the difference in results.

Levin and Arluke (1985) further expected women’s gossip to be more derogatory than men’s, but they could not find proof for this. Both men’s and women’s gossip consisted of both positive (27%) and negative (25%) references to others. But, women and men do clearly differ in who they talk about. Women talk more about friends and family, whereas men gossip about colleagues and celebrities. “In sex-role terms, the male students may have been “afraid to get close.” (Levin & Arluke, 1985: 284).

In sum, men do gossip, but talk more about colleagues, their profession and this in a rather non-detailed manner, which clearly differs from the content of women’s gossip, which focuses on friends’ and family’s private lives. This simply reflects the sex differences in communication skills and sociality. Women are verbally stronger than men, and talk with more emotion, which might explain their more detailed manner of gossiping. Women also have the closest ties with social contacts, who most often are friends and family, which reflects in their gossip content.

In conclusion of this debate on sex differences in gossip, I confirm the common sense idea that women indeed are more eager to gossip than men. But rather than seeing this as a negative fact, this simply stems from the fact that women are verbally better skilled in general, and that they have larger social networks with stronger ties, offering more potential individuals to gossip with and about. I also refer to chapter 4 again, where I discussed sex differences for the specific kinds of gossip I distinguished.