Paradoxically, the extreme view that dominance isn’t real but exists only in the minds of researchers was also held by Robert Hinde, a prominent British ethologist and now an emeritus professor at the University of Cambridge, who in the 1970s convinced many other primatologists that monkeys and apes do have social relationships. While he was writing on various aspects of primate social relationships, Hinde also published a series of articles in which he argued that dominance “does not exist in a concrete empirical sense but it may have usefulness as an explanatory concept.” In his view, dominance should be considered an “intervening variable” that makes it easier for human researchers to explain animal behavior. As illustrated in Figure 2.2, when individuals A and B enter a contest over a resource, their relative age, their body size, the social status of their mothers, their hormone levels, and their previous interactions (the independent variables) can lead to an interaction in which A supplants B, B grooms A, A attacks B, B is submissive to A, or A gets what he wants (the dependent variables). Hinde argued that, when researchers attempt to explain behavior, instead of analyzing the twenty-five different lines that connect the five independent variables to the five dependent ones, it’s easier to place an intervening variable in the middle, in this case dominance. Hinde believed that researchers make up a lot of intervening variables to explain behavior. According to him, emotions don’t exist in a concrete sense either. For example, as illustrated in Figure 2.2, anxiety is an intervening variable we can use to explain why lack of self-confidence and the absence of a supporting adult can lead a child to spend less time exploring and to seek social contact with other children in novel environments.
Figure 2.2. Dominance and anxiety as intervening variables.
Most primatologists these days disagree with Altmann and Hinde and accept that dominance is real; pretty much everyone agrees that anxiety is real too. Why else would so many psychiatrists prescribe Xanax to their patients? There is still disagreement, however, as to what dominance relationships really are. For example, according to primatologist Irwin Bernstein, dominance is a learned relationship. In this view, dominance relationships are established in the course of the initial encounters between two individuals by means of repeated fighting. Once fights have a clear winner and loser and dominance is established, signals are exchanged to inform others that previous events are remembered and the dominance relationship is acknowledged and respected. As we’ll see later, although previous experience is generally an important component of dominance, there are situations in which experience and learning don’t play a role in the establishment of dominance.
Primatologists recognized early on that—as explained earlier—dominance between two individuals can be expressed in different ways: priority of access, directionality of aggressive and submissive behavior, freedom of movement and spatial position, lopsided exchange of grooming, visual monitoring, and receipt of attention. Instead of simply recognizing that dominance is a phenomenon with multiple dimensions, some primatologists have argued that there are different types of dominance. For example, Frans de Waal has suggested that primates have two kinds of dominance relationships: real dominance, which explains who wins fights, and formal dominance, which explains why subordinates give submissive signals to dominants. Wilson and others have proposed that there are different types of dominance relationships depending on context. One type of dominance, referred to as absolute dominance, is independent from context. If individual A has absolute dominance over individual B, A always gets priority regardless of whether A and B are fighting over food, space, or a mate and regardless of where the fight takes place. Relative dominance, by contrast, is dependent on context. For example, A could be dominant over B when they both want the same food, whereas B could be dominant over A when they both want to mate with the same attractive female. One particular type of relative dominance is territoriality, where dominance depends on the location of the encounter. In animals that defend territories in which they build nests and forage, A may be dominant over B when their encounter takes place in A’s territory, and B is dominant over A when the encounter takes place in B’s territory. Another distinction was recently introduced by primatologist Rebecca Lewis, who proposed that a dominance relationship between two individuals should actually be called a power relationship and that we should distinguish between dominance, when power is based on force or the threat of force, and leverage, when power is based on resources that cannot be taken by force.11
What should we make of all these distinctions between different kinds of dominance? In my view, they are not justified and are a potential source of confusion and misunderstanding. The best tool to clear the muddy waters of primate dominance is game theory. Once we consider dominance using game theory, it becomes clear that not only are dominance relationships real, but dominance is always one and the same thing.
Of Bears, Hawks, and Doves
The use of game theory to examine dominance was pioneered by British evolutionary biologist John Maynard Smith in the early 1970s with the publication of his classic article, “The Logic of Animal Conflict.”12 There he looked at how two individuals decide whether to settle a disagreement by having a fight or by establishing dominance. Before I go on, let me clarify that even though evolutionary biologists use terms such as “decisions” and “logic,” they don’t presume any conscious rational thinking on the part of animals (or of people, for that matter). The decisions and the logic in question are the products of natural selection, which has given organisms predispositions to behave adaptively (in a way that increases the benefits and reduces the costs of their behavior) without necessarily engaging in complex thinking or being aware of the consequences of their actions.
To illustrate the game theory approach to dominance, let’s start with a simple hypothetical situation. Yogi Bear and Boo-Boo Bear (the famous cartoon characters created by Hanna-Barbera) meet in a forest one day and there discover an apple. They have never met before and have no reason to think they will meet again in the future. They both want the apple, but the apple cannot be shared. For the sake of simplicity, let’s assume that if Yogi and Boo-Boo fight over the apple, they have exactly the same probability of winning. The moment they see each other and the apple, each bear has the same two options: to fight or to let the other bear take the apple. If a fight is initiated, the winner will eat the apple and the loser will get nothing. If one bear yields, he will simply walk away, maybe to look for an apple somewhere else. If both bears yield, they will use peaceful signals, such as facial expressions or gestures, to negotiate who gets the apple and who doesn’t. In this situation, each bear has the same probability of getting the apple so that, were they to meet one hundred times, each bear would get the apple about fifty times.
At this point, some game theory terminology will make it easier to discuss this scenario and apply it more broadly to other situations. An apple is a commodity that two individuals may want at the same time. In the language of game theory, the options to escalate the fight or to yield are called the Hawk and Dove strategies. A contest in which the two opponents have the same probability of winning is called a symmetrical contest.
Whether Yogi and Boo-Boo play the Hawk or the Dove strategy depends on the value of the apple (let’s call it Benefit, or B) and the cost of fighting (Cost, or C). The benefits are the intrinsic nutritional value of the apple and the potential of the apple to keep a bear from starving. (The same apple with the same caloric content is a lot more valuable to a bear that is about to die from starvation than to a bear that has just eaten ten pounds’ worth of honey.) The main cost of fighting is the risk of injury, even death. Fighting also has other secondary costs: one can burn a lot of energy, make noise that attracts enemies or predators, or miss a chance to look for other food or a mate. In the case of two individuals who are friends (although this doesn’t apply to Yogi and Boo-Boo), one possible cost is damage to the friendship and the potential loss of future benefits it may have brought. Game theory predicts th
at when the benefit of eating the apple is greater than the cost of fighting (which is the same as saying that their ratio is greater than 1, or B/C > 1), both bears will play Hawk and escalate the fight.
Figure 2.3. Yogi Bear and Boo-Boo Bear. Cartoon by Matthew Hunter.
When the cost is greater than the benefit (B/C 1), the situation is more complicated. Let’s assume that fifty pairs of bears meet at the same time and get into a potential conflict about fifty different apples. Some bears are expected to play Hawk and others to play Dove. It can be shown mathematically that in this imaginary population of one hundred bears playing the Hawk-Dove Game, the frequency of Hawks is equal to the ratio between Benefit and Cost (B/C), and the frequency of Doves is equal to 1 – B/C. For example, when B/C is 0.60, sixty out of one hundred bears will play Hawk and forty of them will play Dove. When the benefits are almost as high as the costs and B/C is very close to 1, most bears in the population will be Hawks and only a few will be Doves. When the benefits are negligible compared to the costs and B/C is close to 0, almost all bears will be Doves.
The Hawk-Dove Game is a simple model, but real life is never so simple. In real life, a truly symmetrical contest doesn’t exist. Instead, there are always differences, or asymmetries, between the two contestants so that one of them has a greater probability of winning a fight than the other. In practice, this means that when Yogi and Boo-Boo confront each other over the apple, both immediately recognize that if they have a fight, the probability of winning or losing would not be fifty-fifty. The bear who thinks he has a higher chance of winning is more likely to play Hawk and provoke a fight—for example, by threatening his opponent—while the bear who thinks he has a higher chance of losing is more likely to play Dove and give up the apple peacefully. The moment the two bears recognize that their chances of winning and losing are not the same and act accordingly, their dominance relationship is established: one individual is dominant and the other is subordinate.
Establishment of dominance can occur when two individuals meet for the first time—as soon as they look at each other—or it may take place after repeated encounters and some fighting. Sometimes it’s not immediately clear to the two individuals whether their chances of winning or losing the fight are the same; an empirical test is needed. If Yogi Bear wins ten consecutive fights against Boo-Boo Bear, when they meet for the eleventh time they may both recognize that Yogi has a better chance of winning the next fight than Boo-Boo does. At this point, Boo-Boo yields to Yogi, and Yogi is happy to eat the apple without beating Boo-Boo. When dominance is established, the fighting usually ends. But in some cases, the two bears continue to think that their chances of winning are the same, or each bear thinks that he has the upper hand. In this case, dominance is not established and the bears keep on fighting.
It’s important to understand that establishing dominance is advantageous to both contestants and always preferable to fighting. Resolving the dispute with dominance is good for Yogi, the dominant bear, because he gets the apple without paying the price of fighting. Resolving the dispute with dominance is also good for Boo-Boo, the subordinate bear, although not as good as it is for Yogi. By yielding to Yogi, Boo-Boo doesn’t get the apple, so he gets zero benefits. However, if Boo-Boo Bear had fought and lost, not only would he have lost the apple but he would also have paid a high price physically. By becoming subordinate, Boo-Boo reduces the costs he is likely to pay. So the advantage of establishing dominance for the subordinate is that he cuts his losses.
Cutting one’s losses? That’s it? Yes, the truth is that subordination sucks, and I wouldn’t recommend it to anyone. There are two ways, however, in which Boo-Boo might further benefit from yielding to Yogi. First, if Yogi is a generous bear and the contest with Boo-Boo is over a commodity that can be shared—say, an apple pie—he may let Boo-Boo eat a small piece of the prize or give Boo-Boo something else as a token of appreciation for conceding. The other advantage of subordination is that by yielding to Yogi that day and other times in the future, Boo-Boo saves ammunition. He can safely wait things out until one day he is able to successfully challenge Yogi and reverse their dominance relationship. In fact, it would be very bad if Boo-Boo didn’t even try once to challenge Yogi. Patience is a virtue for a subordinate, but resignation may be the kiss of death. To understand the factors that might lead to a dominance reversal, we must delve further into these asymmetries that determine dominance.
There are two kinds of asymmetries between two individuals engaged in a contest. One type of asymmetry has to do with individual attributes, such as physical characteristics. Yogi is taller, heavier, stronger, healthier, and more mature than Boo-Boo. Characteristics that make an individual more likely to win or lose a fight are called resource holding potential, or RHP. Two individuals can recognize differences in their RHP as soon as they meet for the first time. When Yogi and Boo-Boo first met, both immediately noticed that Yogi was twice as big as Boo-Boo. Hence, they immediately established dominance. Yogi played Hawk and threatened to fight; Boo-Boo played Dove and acted submissively. Thus, a history of previous interactions is not necessary to establish dominance, and dominance is not necessarily a learned relationship.
In the experiment I described in Chapter 1, I mentioned that in the pairs of monkeys that had never met before the two females appeared to be sizing each other up. In some of these pairs, one of the two monkeys “smiled” submissively to the other and started grooming her. The two monkeys probably recognized that there were asymmetries in RHP, perhaps in size, and the individual who perceived herself to be inferior showed a sign of fear and submission, the bared-teeth display. The immediate establishment of dominance allowed the grooming exchange to begin. As expected, the exchange of grooming was heavily lopsided, with the subordinate monkey doing most of the work and the dominant failing to reciprocate. In other unfamiliar pairs, no clear asymmetry in RHP was communicated or recognized. As a result, dominance was not established and little or no grooming was exchanged. Both monkeys “defected” and looked stressed out when their hour together in the cage was finally over, presumably the price they paid for their inability to establish dominance—just like two spouses who are both unwilling to give in and pay the price for the stability and peace of their marriage.
The outcome of a fight between two individuals can be determined not only by differences in size and physical strength but also by differences in their willingness to use their bodies and take risks. Differences in motivation may arise from differences in the value of the commodity being contested: if the commodity is more valuable to one opponent, then this individual will be willing to fight harder than the other. In game theory language, an asymmetry in the value of the commodity to the two opponents is called an asymmetry in payoffs; the individual who has more to gain from winning is also more motivated to fight.
Most, if not all, kinds of context-dependent dominance reflect asymmetries in payoffs. In territorial species, territory holders or residents are more motivated to fight against intruders who have invaded the territory because they have more at stake. Their territory contains their nest, their food, and often their mates and offspring. Losing one’s territory may mean losing everything. The intruder has a lot less to lose; if things don’t work out here, he’ll simply invade the territory next door. Once again, decisions to fight or yield don’t require mental calculations about the costs and benefits of particular moves. For example, residents’ threatening to fight or escalating a fight against an intruder may simply be an emotional response to the situation. The sight of an intruder trespassing into his territory may enrage the resident, and this emotion enhances his aggressive motivation. Conversely, being in an unfamiliar place induces fear in the intruder and increases the probability of submission.
Figure 2.4. Bared-teeth display shown by a crested black macaque, Macaca nigra. Photo courtesy of Dr. Irwin Bernstein.
By this point it should be evident that there aren’t multiple kinds of dominance after all. Dominance is always on
e and the same, but dominance interactions may be different owing to asymmetries in RHP or in payoffs between contestants. Game theorists have developed more complex versions of the Hawk-Dove Game to take these asymmetries into account. As with the simple Hawk-Dove Game, interaction outcomes are predicted by the value of the commodity and the cost of fighting. Clearly, the usefulness of these models to explain the behavior of individuals in real-life situations depends on how accurately one can measure the exact value of B and C to the two opponents. Multiple asymmetries between the two opponents can make the measurement of B and C very tricky. Moreover, there is a further complication. For the sake of simplicity, the game theory models of asymmetric contests assume that two opponents always have accurate information about their asymmetries and relative probabilities of winning or losing a fight. Again, real life is hardly so simple, and the information in question is not always accurate.
Games Primates Play: An Undercover Investigation of the Evolution and Economics of Human Relationships Page 6