Centipedes aren’t social. Neither are snails or brown bears. In fact, most animals are loners that prefer solitude to company. So why don’t we?

All animals must survive and reproduce, and being social is one strategy for accomplishing these two important goals. When it comes to finding food or fending off enemies, herds and packs and flocks can often do what individuals can’t, and that’s why over millions of years, many different species have found it useful to become social. But of the thousands and thousands of social species on our planet, only we form large-scale societies of genetically unrelated individuals. The ability to live in such large groups is one of the main reasons why we developed such large brains (Sallet et al., 2011; Shultz & Dunbar, 2010; Smith et al., 2010), and it has allowed our species to thrive. Consider this: If you had rounded up all the mammals on Earth 10,000 years ago and weighed them, human beings would have accounted for about 0.01% of the total weight. Today we would account for 98%. Being social has allowed us to become the heavyweight champions of survival and reproduction, and as you are about to see, much of our social behavior revolves around these two basic goals.

To survive, an animal must find resources such as food, water, and shelter. These resources are always scarce, because if they weren’t, then the animal population would just keep increasing until they were. Animals solve the problem of scarce resources in two ways: by hurting each other and helping each other. Hurting and helping are antonyms, so you might expect them to have little in common. But as you will see, these seemingly antithetical behaviors are actually two solutions to the same problem (Hawley, 2002).

Aggression

The simplest way to solve the problem of scarce resources is simply to take what you want and kick the stuffing out of anyone who tries to stop you. Aggression is behavior whose purpose is to harm another (Anderson & Bushman, 2002; Bushman & Huesmann, 2010), and it is a strategy used by just about every animal on the planet. Aggression is not something that animals do for its own sake, but as a way of getting the resources they desire. The frustration-aggression hypothesis suggests that animals aggress when their desires are frustrated (Dollard et al., 1939), and it is pretty easy to see aggression through this lens. The chimp wants the banana (desire), but the pelican is about to take it (frustration), so the chimp threatens the pelican with its fist (aggression). The robber wants the money (desire), but the teller has it all locked up (frustration), so the robber threatens the teller with a gun (aggression).

381

The frustration–aggression hypothesis is right as far as it goes, but many scientists think it doesn’t go far enough. They argue that the actual cause of aggressive behavior is negative affect (more commonly known as feeling bad) and that a frustrated desire is just one of many things that can induce it (Berkowitz, 1990). If it is true that animals aggress when they feel bad, then anything that makes them feel bad should increase aggression, and some evidence suggests it does. For example, laboratory rats that are given painful electric shocks will attack anything in their cage, including other animals, stuffed dolls, or even tennis balls (Kruk et al., 2004). People who are made to put their hands in ice water or to sit in a very hot room are more likely to blast others with noise weapons or punish others by making them eat hot chilis (Anderson, 1989; Anderson, Bushman, & Groom, 1997). The idea that aggression is a response to negative affect may even explain why so many acts of human aggression—from violent crime to athletic brawls—are more likely to occur on hot days when people are feeling irritated and uncomfortable (see FIGURE 12.1).

Of course, not everyone aggresses every time they feel bad. So who does and why? Research suggests that both biology and culture play important roles in determining if and when people who feel bad will aggress. Let’s start by examining biology.

Biology and Aggression. If you wanted to know whether someone was likely to aggress and you could ask them only one question, it should be this: “Are you male or female?” (Wrangham & Peterson, 1997). Violent crimes such as assault, battery, and murder are almost exclusively perpetrated by men—and especially by young men (Strueber, Lueck, & Roth, 2006). Although most societies encourage males to be more aggressive than females (more on that shortly), male aggressiveness is not merely the product of socialization. Studies show that aggression is strongly correlated with the presence of a hormone called testosterone, which is typically much higher in men than in women, in younger men than in older men, and in violent criminals than in nonviolent criminals (Dabbs et al., 1995).

One of the most reliable ways to elicit aggression in males is to challenge their status or dominance. Indeed, three quarters of all murders can be classified as “status competitions” or “contests to save face” (Daly & Wilson, 1988). Contrary to popular wisdom, it isn’t men with low self-esteem but men with unrealistically high self-esteem who are most prone to aggression, because such men are especially likely to perceive others’ actions as a challenge to their inflated sense of their own status (Baumeister, Smart, & Boden, 1996). Men seem especially sensitive to these challenges when they are competing for the attention of women (Ainsworth & Maner, 2012).

382

Women can be aggressive too, of course, but their aggression tends to be focused on challenges to their resources rather than to their status. Women are much less likely than men to aggress without provocation or to aggress in ways that cause physical injury, but they are only slightly less likely than men to aggress when provoked or to aggress in ways that cause psychological injury (Bettencourt & Miller, 1996; Eagly & Steffen, 1986). Indeed, women may even be more likely than men to aggress by causing social harm—for example, by ostracizing others (Benenson et al., 2011) or by spreading malicious rumors about them (Crick & Grotpeter, 1995).

Culture and Aggression. Aggression has a biological basis, but it is also strongly influenced by culture (see FIGURE 12.2). For example, violent crime in the United States is more prevalent in the South, where traditional notions of honor require men to react aggressively when their status is challenged (Brown, Osterman, & Barnes, 2009; Nisbett & Cohen, 1996). In one set of experiments, researchers either did or did not insult American students from northern and southern states. When insulted, Southerners were more likely to experience a surge of testosterone and to feel that their status had been diminished by the insult (Cohen et al., 1996). And when a large man walked directly toward them as they were leaving the experiment, the insulted Southerners got “right up in his face” before giving way, whereas Northerners just stepped aside. On the other hand, when participants were not insulted, polite Southerners stepped aside before Northerners did. Clearly, culture plays an important role in determining whether our innate capacity for aggression will actually lead to aggressive behavior (Leung & Cohen, 2011).

383

Cooperation

Aggression is one way to solve the problem of scarce resources, but it is not the best way, because when individuals work together, they can often each get more resources than either could get alone. Cooperation is behavior by two or more individuals that leads to mutual benefit (Deutsch, 1949; Pruitt, 1998), and it is one of our species’ greatest achievements—right up there with language, fire, and opposable thumbs (Axelrod, 1984; Axelrod & Hamilton, 1981; Nowak, 2006). Every roadway and supermarket, every robot and smartphone, every ballet and surgery is the result of cooperation, and it is difficult to think of an important human achievement that could have occurred without it.

Risk and Trust. So why don’t we all cooperate all the time? Cooperation can be beneficial, but it can also be risky, and a simple game known as the prisoner’s dilemma illustrates why. Imagine that you and your friend have been arrested for hacking into a bank’s computer system and stealing a few million dollars. You are now being interrogated separately. The detectives tell you that if you and your friend both confess, you’ll each get 10 years in prison for felony theft, and if you both refuse to confess, you’ll each get 1 year in prison for disturbing the peace. However, if one of you confesses and the other doesn’t, then the one who confesses will go free and the other will be put away for 30 years. What should you do? If you study FIGURE 12.3, you’ll see that you and your friend would be wise to cooperate. If you both trust each other and both refuse to confess, then you will both get light sentences. But if you trust your friend and then your friend double-crosses you, your friend will go free and you will spend the next 30 years in a cell.

The prisoner’s dilemma game illustrates a basic fact: Cooperation benefits everyone, but only if everyone cooperates. If someone doesn’t cooperate, then everyone else pays a price. This fact makes it hard for us to decide whether we should or should not cooperate. We know that if everyone pays their taxes, then the tax rate stays low, and everyone enjoys the benefits of sturdy bridges and first-rate museums. But we also know that those who cheat and don’t pay their taxes will get all the same benefits— and that we will have to pay for them! If we could be sure that everyone would pay their taxes we’d be happy to pay ours, but we don’t want to pay ours if they aren’t going to pay theirs. What to do? Is there any way to know whom we can trust?

384

Groups and Favoritism. In fact, there is. A group is a collection of people who have something in common that distinguishes them from others, and every one of us is a member of many groups—from families and teams to religions and nations. Although groups differ in many ways, they all have one thing in common, which is that group members generally trust each other to be honest, fair, and nice. Prejudice is a positive or negative evaluation of another person based on the person’s group membership, and discrimination is positive or negative behavior toward another person based on the person’s group membership (Dovidio & Gaertner, 2010). One of the defining characteristics of groups is that members are positively prejudiced toward fellow members and tend to discriminate in their favor (DiDonato, Ullrich, & Krueger, 2011). Even when people are randomly assigned to be members of meaningless groups such as “Group 1” or “Group 2,” they still give preferential treatment to members of their own group (Hodson & Sorrentino, 2001; Locksley, Ortiz, & Hepburn, 1980). Simply knowing that a person is one of us and not one of them, is sufficient to produce prejudice and discrimination (Tajfel et al., 1971). Because group members can be relied on to favor each other, group membership makes cooperation less risky.

The ability to trust and cooperate is one of the main benefits of being in a group. But groups also have costs. For example, when groups try to make decisions, they rarely do better than the best member would have done alone—and they quite often do worse (Minson & Mueller, 2012). One reason is that groups don’t fully capitalize on the expertise of their members (Hackman & Katz, 2010). For instance, groups (such as a school board) often give too little weight to the opinions of members who are experts (the professor) and too much weight to the opinions of members who happen to be high in status (the mayor) or especially talkative (the mayor). Groups are also susceptible to the common knowledge effect which is the tendency for group discussions to focus on information that all members share (Gigone & Hastie, 1993). The problem with this is that the information everyone shares (the size of the gymnasium) is often relatively unimportant, whereas the truly important information (how a school in a different district solved its budget crisis) is known to just a few. In addition, group discussion often acts as an “amplifier” of initial opinions. Group polarization is the tendency for groups to make decisions that are more extreme than any member would have made alone (Myers & Lamm, 1975). A group whose members come to the table with moderate opinions (“We should probably just renovate the auditorium”) can end up making an extreme decision (“We’re going to build a new high school!”) simply because, in the course of discussion, each member was exposed to many different arguments in favor of a single position (Isenberg, 1986). Finally, members of groups care about how other members feel and are sometimes reluctant to “rock the boat” even when it needs a good rocking. Groupthink is the tendency for groups to reach consensus in order to facilitate interpersonal harmony (Janis, 1982). Harmony is important (especially if the group is a choir), but studies show that groups often make poor decisions in order to achieve it (Turner & Pratkanis, 1998). For all of these reasons, groups underperform individuals in a wide variety of tasks.

385

The costs of groups go beyond bad decisions. People in groups sometimes do terrible things that they would never do alone, such as rioting, lynching, and gang-raping (Yzerbyt & Demoulin, 2010). Why do we sometimes behave badly when we assemble in groups?

One reason is deindividuation, which occurs when immersion in a group causes people to become less aware of their individual values. We may wish we could grab the Rolex from the jeweler’s window or plant a kiss on the attractive stranger in the library, but we don’t do these things because they conflict with our personal values. Research shows that being assembled in groups draws our attention to others and away from ourselves, and as a result, we are less likely to consider our own personal values and instead adopt the group’s values (Postmes & Spears, 1998).

A second reason why groups behave badly is diffusion of responsibility, which refers to the tendency for individuals to feel diminished responsibility for their actions when they are surrounded by others who are acting the same way. For example, studies of bystander intervention— which is the act of helping strangers in an emergency situation—reveal that people are less likely to help an innocent person in distress when there are many other bystanders present, because they assume that one of the other bystanders is more responsible than they are (Darley & Latané, 1968; Latané & Nida, 1981). If you saw a fellow student cheat on an exam, you’d probably feel more responsible for reporting the incident if you were taking the test in a group of 3 than in a group of 3,000 (see FIGURE 12.4).

If groups make bad decisions and foster bad behavior, then might we be better off without them? It seems unlikely. Not only do groups enable cooperation, which has extraordinary benefits, but one of the best predictors of a person’s general well-being is the quality and extent of his or her group memberships (Myers & Diener, 1995). People who are excluded from groups are typically anxious, lonely, depressed, and at increased risk for illness and premature death (Cacioppo & Patrick, 2008; Cohen, 1988; Leary, 1990). Groups may cause us to misbehave, but they are also the key to our happiness and well-being.

386

Altruism

So far, the picture we’ve painted of human beings isn’t all that rosy: People aggress against each other in order to get resources, and they cooperate when doing so provides them with greater benefits than aggression does. Okay, sure, we all like benefits. But aren’t we ever just nice to each other?

Altruism is behavior that benefits another without benefiting oneself, and for centuries, scientists and philosophers have argued about whether people are ever truly altruistic. That might seem like an odd argument to have. After all, people give their blood to the injured, their food to the homeless, and their time to the elderly. We volunteer, we tithe, we donate. Isn’t that evidence of altruism?

Not necessarily, because behaviors that appear to be altruistic often have hidden benefits for those who do them. For example, ground squirrels make a loud noise when they see a predator, which attracts the predator’s attention and puts them at increased risk of being eaten, but which allows their fellow squirrels to escape. Although this behavior appears to be altruistic, it actually isn’t because the helpers are genetically related to the helpees. Any animal that promotes the survival of its relatives is actually promoting the survival of its own genes (Hamilton, 1964). Kin selection is the process by which evolution selects for individuals who cooperate with their relatives, and cooperating with related individuals is not truly altruistic. Cooperating with unrelated individuals isn’t necessarily altruistic either. Male baboons will risk injury to help an unrelated male baboon win a fight, and monkeys will spend time grooming unrelated monkeys when they could be doing something more interesting (which is just about anything). But as it turns out, the animals that give favors tend to get favors in return. Reciprocal altruism is behavior that benefits another with the expectation that those benefits will be returned in the future, and despite the second word in this term, it isn’t truly altruistic (Trivers, 1972). Indeed, reciprocal altruism is merely cooperation extended over time.

The behavior of nonhuman animals provides little evidence of genuine altruism (cf. Bartal, Decety, & Mason, 2011). But what about us? Are we any different? Like other animals, we tend to help our kin more than we help strangers (Burnstein, Crandall, & Kitayama, 1994; Komter, 2010), and we tend to expect those we help to help us in return (Burger et al., 2009). But unlike other animals, we do sometimes provide benefits to complete strangers who have no chance of repaying us (Batson, 2002; Warneken & Tomasello, 2009). We hold the door for people who share precisely none of our genes and tip waiters in restaurants to which we will never return. And we do more than that. As the World Trade Center burned on the morning of September 11, 2001, civilians in sailboats headed toward the destruction rather than away from it, initiating the largest waterborne evacuation in U.S. history. As one observer remarked, “If you’re out on the water in a pleasure craft and you see those buildings on fire, in a strictly rational sense you should head to New Jersey. Instead, people went into potential danger and rescued strangers” (Dreifus, 2003). Human beings can be truly altruistic, and some studies suggest that they are even more altruistic than they realize (Miller & Ratner, 1998).

387

All animals must survive and reproduce, and social behavior is useful for survival. But it is an absolute prerequisite for reproduction, which doesn’t happen until people get very, very social. The first step on the road to reproduction is finding someone who wants to travel that road with us. How do we do that?

Selectivity

With the exception of a few well-known celebrities, people don’t mate randomly. Rather, they select their sexual partners, and as anyone who has lived on Earth for more than 7 full minutes knows, women tend to be more selective than men (Feingold, 1992; Fiore et al., 2010). When researchers asked an attractive person to approach strangers on a college campus and ask, “Would you go out with me?” they found that roughly half of the men and half of the women agreed to the request for a date. On the other hand, when the attractive person said to strangers, “Would you go to bed with me?” the researchers found that none of the women and three quarters of the men agreed to the request (Clark & Hatfield, 1989). There are many reasons why a woman in this particular situation might say no (Conley, 2011), but research suggests that women tend to be choosier than men in most other situations as well (Buss & Schmitt, 1993; Schmitt et al., 2012).

One reason for this is that our basic biology makes sex a riskier proposition for women than for men. Men produce billions of sperm in their lifetimes, their ability to conceive a child tomorrow is not inhibited by having conceived one today, and conception has no significant physical costs. On the other hand, women produce a small number of eggs in their lifetimes, conception eliminates their ability to conceive for at least 9 more months, and pregnancy produces physical changes that increase women’s nutritional requirements and put them at risk of illness and death. Therefore, if a man mates with a woman who does not produce healthy offspring or who won’t do her part to raise the children, he’s lost nothing but 10 minutes and a teaspoon of bodily fluid. But if a woman makes the same mistake, she has lost a precious egg, borne the costs of pregnancy, risked her life in childbirth, and missed at least 9 months of other reproductive opportunities.

So basic biology pushes women to be choosier then men. But culture and experience can push equally hard, and in a different direction (Petersen & Hyde, 2010; Zentner, & Mitura, 2012). For example, women may be choosier than men simply because they are approached more often (Conley et al., 2011) or because the reputational costs of promiscuity are higher (Eagly & Wood, 1999; Kasser & Sharma, 1999). Indeed, when sex becomes expensive for men (e.g., when they are choosing a long-term mate rather than a short-term date), they become every bit as choosey as women (Kenrick et al., 1990), and relatively minor changes in the courtship ritual can actually cause men to be choosier than women (see the Real World box). The point is that biology makes sex a riskier proposition for women than for men, but cultures can exaggerate, equalize, or even reverse those risks. The higher the risk, the more selective people of both genders tend to be.

388

Attraction

For most of us, there is a very small number of people with whom we are willing to have sex, an even smaller number of people with whom we are willing to have children, and a staggeringly large number of people with whom we are unwilling to have either. So when we meet someone new, how do we decide which of these categories the person belongs in? Many things go into choosing a date, a lover, or a partner for life, but perhaps none is more important than the simple feeling we call attraction (Berscheid & Reis, 1998). Research suggests that this feeling is caused by situational, physical, and psychological factors. Let’s examine them in turn.

Situational Factors. We tend to think that we select our romantic partners on the basis of their personalities, appearances, and so on—and as you’ll see in a moment, we do—but we only get to select from the pool of people we’ve met, which is why physical proximity is one of the best predictors of whether two people will end up in any kind of relationship (Festinger, Schachter, & Back, 1950; Nahemow & Lawton, 1975). Most of us end up marrying someone who lived or worked or went to school in the same places we did. Proximity not only provides the opportunity for attraction, but it also provides the motivation: People work especially hard to like those with whom they expect to have interaction (Darley & Berscheid, 1967). When you are assigned a roommate or an office mate, you know that your day-to-day existence will be a whole lot easier if you like the person than if you don’t, and so you go out of your way to notice the person’s good qualities and ignore the bad ones.

389

Proximity provides something else as well. Every time we encounter a person, that person becomes a bit more familiar to us, and people generally prefer familiar to novel stimuli. The mere exposure effect is the tendency for liking to increase with the frequency of exposure (Bornstein, 1989; Zajonc, 1968). For instance, in some experiments, geometric shapes, faces, or alphabetical characters were flashed on a computer screen so quickly that participants were unaware of having seen them. Participants were then shown some of the “old” stimuli that had been flashed on the screen as well as some “new” stimuli that had not. Although they could not reliably say which stimuli were old and which were new, the participants did tend to like the old stimuli better than the new ones (Monahan, Murphy, & Zajonc, 2000). Although there are some circumstances under which “familiarity breeds contempt” (Norton, Frost, & Ariely, 2007), for the most part, it tends to breed liking (Reis et al., 2011).

Physical Factors. One of the strongest determinants of attraction is a person’s physical appearance. One study found that a man’s height and a woman’s weight were among the best predictors of how many responses their personal ads received (Lynn & Shurgot, 1984), and another study found that physical attractiveness was the only factor that predicted the online dating choices of both women and men (Green, Buchanan, & Heuer, 1984). Good-looking people have more sex, more friends, and more fun than the rest of us do (Curran & Lippold, 1975), and they even earn about 10% more money over the course of their lives (Hamermesh & Biddle, 1994; see FIGURE 12.5). Appearance is so powerful that it even influences non-romantic relationships: For example, mothers are more affectionate and playful when their children are more attractive (Langlois et al., 1995).

So yes, it pays to be beautiful. But what exactly constitutes beauty? The answer to that question varies across cultures. In the United States, for example, most women want to be slender, but in Mauritania, young girls are forced to drink up to 5 gallons of high-fat milk every day so that they will someday be obese enough to attract a husband. As one Mauritanian woman noted, “Men want women to be fat, and so they are fat. Women want men to be skinny, and so they are skinny” (LaFraniere, 2007). In the United States, most men want to be tall, but in Ghana, most men are short and consider height a curse. “To be a tall person can be quite embarrassing,” said one particularly altitudinous Ghanaian man. “When you are standing in a crowd, the short people start to jeer at you,” said another (French, 1997).

390

But while different cultures have somewhat different standards of beauty, those standards also have a lot in common (Cunningham et al., 1995). For example, in all cultures, faces are generally considered more attractive when they are bilaterally symmetrical—that is, when the left half is a mirror image of the right (Perrett et al., 1999). Symmetry is a sign of good genetic health (Jones et al., 2001; Thornhill & Gangestad, 1993), and so nature seems to have designed us to be attracted by it. Of course, attraction is one thing and action is another. Studies show that while everyone may desire the most beautiful person in the room, most people tend to approach, date, and marry someone who is about as attractive as they are (Berscheid et al., 1971; Lee et al., 2008).

Psychological Factors. A person’s physical appearance is often the first thing we know about them, so it isn’t surprising that it determines our initial attraction (Lenton & Francesconi, 2010). But once people begin interacting, they quickly move beyond appearances (Cramer, Schaefer, & Reid, 1996; Regan, 1998). People’s inner qualities—their personalities, points of view, attitudes, beliefs, values, ambitions, and abilities—play an important role in determining their sustained interest in each other, and there isn’t much mystery about the kinds of inner qualities that most people find attractive. For example, intelligence, sense of humor, sensitivity, and ambition seem to be high on just about everybody’s list, whereas “sadistic serial killer” typically ranks rather low (Daniel et al., 1985).

How much wit and wisdom do we want our mate to have? Research suggests that we are most attracted to those who are similar to us (Byrne, Ervin, & Lamberth, 1970; Byrne & Nelson, 1965; Hatfield & Rapson, 1992; Neimeyer & Mitchell, 1988). Indeed, one of the best predictors of whether two people will marry is their similarity in terms of education, religious background, ethnicity, socioeconomic status, and personality (Botwin, Buss, & Shackelford, 1997; Buss, 1985; Caspi & Herbener, 1990).

391

Why is similarity so attractive? First, it’s easy to interact with people who are similar to us because we can instantly agree on a wide range of issues, such as what to eat, where to live, how to raise children, and how to spend our money. Second, when someone shares our attitudes and beliefs, we feel more confident that those attitudes and beliefs are correct, and that’s always a comfort (Byrne & Clore, 1970). Third, if we like people who share our attitudes and beliefs, then we can reasonably expect them to like us for the same reason, and being liked is a powerful source of attraction (Aronson & Worchel, 1966; Backman & Secord, 1959; Condon & Crano, 1988). Although we tend to like people who like us, it is worth noting that we especially like people who like us and who don’t like anyone else (Eastwick et al., 2007).

Relationships

Once we have attracted a mate, we are ready to reproduce. (Note: It is perfectly fine to pause for dinner.) Human reproduction ordinarily happens in the context of committed, long-term relationships (Clark & Lemay, 2010). Only a few animals have such relationships, so why are we among them?

One answer is that we’re born half-baked. Because human beings have large heads to house their large brains, a fully developed human infant could not pass through its mother’s birth canal. So human infants are born before they are fully developed. That means they need a lot more care than one parent can provide, and that’s one reason why human adults tend to do their reproducing in the context of committed, long-term relationships.

In most cultures, committed, long-term relationships are signified by marriage, and ours is no exception. The probability of marrying by age 40 is about 81% for American men and 86% for American women (Goodwin, McGill, & Chandra, 2009). But when asked, people generally don’t say that they got married in order to solve the big-headed baby problem; they say that they got married because they were in love. The fact that people marry for love is obvious, but it only became obvious in the past century or so (Brehm, 1992; Fisher, 1993; Hunt, 1959). Ancient Greeks and Romans got married, but they considered love a form of madness (Heine, 2010). Twelfth-century Europeans got married but thought of love as a game to be played by knights and ladies of the court (who happened to be married and not to the knights). Throughout history, marriage has traditionally served a variety of decidedly unromantic functions—ranging from cementing agreements between clans to paying back debts—and in many cultures, that’s how it is still regarded. In fact, it wasn’t until the 17th century that Westerners started to think that love might be a reason to get married.

But what exactly is love? Psychologists distinguish between two basic kinds: passionate love, which is an experience involving feelings of euphoria, intimacy, and intense sexual attraction, and companionate love, which is an experience involving affection, trust, and concern for a partner’s well-being (Acevedo & Aron, 2009; Hatfield, 1988; Rubin, 1973; Sternberg, 1986). The ideal romantic relationship gives rise to both types of love, but the speeds, trajectories, and durations of the two experiences are markedly different (see FIGURE 12.6). Passionate love is what brings people together: It has a rapid onset, reaches its peak quickly, and begins to diminish within just a few months (Aron et al., 2005). Companionate love is what keeps people together: It takes some time to get started, grows slowly, and need never stop growing (Gonzaga et al., 2001).

392

The most recent U.S. government census statistics indicate that for every two couples that get married, roughly one couple gets divorced. But why? Marital satisfaction is only weakly correlated with marital stability (Karney & Bradbury, 1995), suggesting that relationships break up or remain intact for reasons other than the satisfaction of those involved (Drigotas & Rusbult, 1992; Rusbult & Van Lange, 2003). Social exchange is the hypothesis that people remain in relationships only as long as they perceive a favorable ratio of costs to benefits (Homans, 1961; Thibaut & Kelley, 1959). Relationships offer both benefits (love, sex, and financial security) and costs (responsibility, conflict, loss of freedom), and people maintain them as long as the ratio of the two is acceptable. Three things determine whether a person will find a particular cost-benefit ratio to be acceptable: