Strategic Altruism: The Machiavellian Roots of Human Kindness

Nice Nazis, friendly factory farmers, and the evolution of sincere but selective moral instincts

Daniel Williams is a Lecturer in Philosophy at the University of Sussex and an Associate Fellow at the Leverhulme Centre for the Future of Intelligence (CFI) at the University of Cambridge. He writes at www.conspicuouscognition.com. You can find out about his research at www.danwilliamsphilosophy.com.

---

---

“It is good to appear clement, trustworthy, humane, religious, and honest, and also to be so, but always with the mind so disposed that, when the occasion arises not to be so, you can become the opposite.”

So advises Niccolò Machiavelli in The Prince, a sixteenth-century political treatise that provides guidance to ambitious rulers. As with much of The Prince, the recommendation is, well, Machiavellian. It is brutal, amoral, and cynical. It is also insightful about features of human nature we would all prefer to ignore.

This essay explores the evolutionary underpinnings of human niceness. I will argue that much of our social behaviour is rooted in reputation management and that the subtle incentives of reputation management explain why human altruism is both sincere and strategic. Our moral instincts are much more Machiavellian than we would like to admit.

A roadmap:

• Section 1 explains how niceness can evolve.

• Section 2 explains how “social selection”—roughly, social competition and reputation management—shaped the evolution of our species.

• Section 3 explains why many people think the role of social selection in human evolution undermines a cynical view of human nature.

• Section 4 explains why this is wrong.

• Section 5 clarifies in what ways humans are—and are not—Machiavellian.

1. The evolution of niceness

Humans can be exceptionally kind, friendly, generous, trustworthy, and principled. These traits are surprising from an evolutionary perspective. We are apes. Like other animals, we evolved through a competitive process of natural selection. Roughly, this process favours traits that help organisms propagate their genes. Altruism, morality, principles, and so on do not seem to help with that goal; they seem to hinder it. Organisms that aid other organisms should be outcompeted by organisms that look after themselves. Those who care about abstract ideals—justice, fairness, equality, and so on—should fare especially badly.

If so, why is there so much niceness in the world? Humans—at least some humans, at least sometimes—provide an obvious example, but not the only one. Nature is teeming with friendliness, care, mutual aid, and sympathy.

Altruism

One evolutionary explanation of altruism is kin selection, a process in which organisms spread their genes by aiding relatives who share copies of those genes. Kin selection explains most of the spectacular examples of altruism in non-human animals. It also explains why humans are instinctively nepotistic. However, it cannot explain why we are so often friendly and fair-minded towards non-relatives.

Another idea—one endorsed by Darwin himself—is group selection. In this process, selection operates not just at the level of individuals but at the level of groups, and altruism evolves because altruistic groups outcompete selfish ones. I doubt group selection played a role in human evolution (see here and here). Although we are undeniably groupish, this is because individual ancestors with a coalitional psychology outcompeted their less group-minded cousins, not because groups themselves were subject to natural selection.

Kin selection and group selection generate biologically altruistic tendencies: dispositions to benefit others (i.e., genetic relatives or group members) at a personal cost. However, much of the cooperation we see in nature is “mutualistic”, not altruistic: it is organisms helping themselves by helping others.

Mutualism

Sometimes, mutualism arises from interdependence: organisms assist other organisms because their fitness depends on the welfare of those organisms. Theorists such as Michael Tomasello speculate—plausibly, in my view—that ecological and social changes forced our ancestors to become highly interdependent, initiating our evolutionary journey towards the extremely high levels of cooperation we achieve today.

Mutualism also results from reciprocity, which can be direct (“you scratch my back, I scratch yours”) and indirect (“you scratch my back and someone else will scratch yours”). Both forms are central to human social life and are typically scaffolded by social processes involving norms, reputations, and gossip. Reciprocity is distinct from interdependence but typically creates interdependence.

One important form of reciprocity involves status. Humans crave and compete for status. These status games can be understood as an economy in which people exchange respect and admiration for access to the benefits that impressive and virtuous people provide. This explains why human virtue is often competitive.

2. Social Selection

The ubiquity of cooperation in the biological world puzzled Darwin, but so did the peacock’s tail.

If you picture evolution as a process favouring the “survival of the fittest”, the peacock’s tail poses an obvious problem. The lavish and elaborate tail feathers do not appear to accomplish anything useful. Worse, they require energy to grow and maintain, and they increase the birds’ vulnerability to predators.

To solve this problem, Darwin proposed the concept of sexual selection, which can give rise to traits not advantageous for survival if they are sexy enough. If peahens preferentially mate with peacocks displaying the most elaborate and impressive tails, evolution will favour those tails.

The concept of sexual selection embodies a profound idea: agents' preferences and choices introduce a fundamentally novel selection pressure—what Darwin called a “second agency” of evolution—into nature, one that can favour the emergence of traits that seem wasteful, pointless, and just plain weird relative to mundane, practical goals.

A century later, Mary Jane West-Eberhard pointed out that sexual selection is just one example of social selection. Although the precise meaning of “social selection” is not always clear, the basic idea is simple: sexual selection is just a special case of the more general fact that social preferences and competition—for example, competition for status, friendships, and alliances—introduce fundamentally novel Darwinian dynamics.

In my view, social selection in this broad sense is responsible for almost everything that is weird, wonderful, and paradoxical about our species. It is the secret of our evolutionary success—and also of our strangeness.

Partners, punishment, and prestige

One form of social selection concerns partner choice. We compete to attract attractive cooperation partners and alliances. This includes mates and spouses—as Geoffrey Miller points out, we are “courtship machines”, not just survival machines—but also friends and broader cooperative units like communities and subcultures.

Such competition is obvious in high school, where kids crave access to the cool cliques, but high school never ends. Human societies function as biological markets in which people compete to access the best mates, friends, and alliances. To succeed in such markets, people must seem impressive, trustworthy, and fair-minded.

A related form of social selection involves group punishment. Human societies are organised around norms: don’t lie, don’t steal, don’t covet your neighbour’s wife or ox or donkey, don’t insult your parents, don’t depict the Prophet Muhammad, and so on. Norm violations tend to be discovered because humans are naturally nosy and love to gossip. If they are discovered, they are punished. Even if the punishment is just a worse reputation, the consequences can be devastating. However, punishment can be much worse than a bad reputation; both in many places today and throughout our evolutionary past, human communities frequently kill their worst norm violators. Given this, being viewed as a good, norm-following group member is one of our most important adaptive challenges.

Finally, there are status games. Social competition in other great apes (e.g., chimps and bonobos) focuses on dominance hierarchies, a form of social rank rooted in fear, intimidation, and aggression. Humans have dominance hierarchies, but we also have prestige, a form of social rank rooted in the amount of respect, admiration, and deference people receive from others. Given this, much social competition in our species involves competition to be impressive, admired, and respected—more specifically, to be more impressive, admired, and respected than others. This generates traits and behaviours far more wasteful and extravagant than the peacock’s tail. It is the engine of art, science, and philosophy.

“Nature when she formed man for society,” wrote Adam Smith, “endowed him with an original desire to please and original aversion to offending his brethren.” Smith was describing social selection: to survive and thrive in human society, humans must crave social approval and fear disapproval. A century before, Thomas Hobbes landed on the same idea: He “that breaketh his Covenant,” he wrote, “cannot be received into Society… if he be left, or cast out of Society, he perisheth.”

3. Homo Puppy

Since Darwin introduced the theory of evolution by natural selection, many have worried that it implies a cynical theory of human nature. If humans evolved through a competitive process that favours those organisms best at spreading their genes, surely our fundamental motives and instincts must be selfish, at least when it comes to interacting with non-relatives.

The fact that humans are intensely cooperative does not seem to address this worry. If most human cooperation is mutualistic—if it is cooperation that evolves through Darwinian self-interest—our prosocial motives and emotions should be highly sensitive to the personal costs and benefits associated with cooperating. We should be strategic cooperators, cooperating only when we expect to personally benefit.

Many people appeal to social selection to combat these cynical worries. Roughly, they reason as follows. To survive and thrive in social environments, our ancestors had to cultivate reputations as friendly, fair-minded, generous, and trustworthy. However, the best way to cultivate such reputations is to be friendly, fair-minded, generous, and trustworthy. Given this, we—at least those of us who are not psychopaths—have strong, innate, and sincere altruistic motivations and emotions.

This reasoning is often tied to the hypothesis that our species “self-domesticated”. Just as peacocks evolved colourful and extravagant tails through the power of mate choice, our ancestors were tamed and transformed into cuddly, prosocial, empathetic do-gooders by social preferences. In his best-seller “Humankind”, Rutger Bregman uses this reasoning to argue for the “radical idea” that “most people, deep down, are pretty decent.” Throughout the book, he refers to human beings as “Homo puppy”.

There is an important grain of truth in this analysis. Traditional evolutionary theories of mutualistic cooperation struggle to explain why humans rarely seem like self-serving bean counters when helping others and forming relationships. For example, if cooperation results from interdependence, one would expect us to be completely indifferent towards those we do not depend on, which is rarely true. And if it results from reciprocity, one would expect us to carefully consider whether our helping behaviour will be compensated, which is also rarely true: we “cooperate without looking”—without explicitly weighing the personal costs and benefits associated with being nice and helpful.

The subtle incentives associated with reputation management can address this problem for standard theories of mutualistic cooperation. Stingy cooperators do not make attractive mates, spouses, friends, or group members. Because we depend on social approval and trust, we must appear like unconditional and authentic altruists, which is extremely difficult for miserly, self-serving cooperators to achieve. When reputations are involved, the best way to cooperate for personal benefit is to appear to cooperate for reasons that have nothing to do with personal benefit—because it is the “right thing to do”, or required by the demands of “morality”, or something like that.

As Randolph Nesse puts it,

“If you tell the leader of a church that you want to join because you would like to get help when you get sick, you will likely be told that you just don’t get it; members are expected to help others willingly from their hearts, not because they want to get something. The paradox is that those whose helping is motivated by commitment often get more help when they need it than those who negotiate explicit contracts.”

This is an important insight. It identifies the grain of truth in the idea that social selection explains why the most cynical theories of human nature are mistaken. However, this grain of truth is typically buried under a large heap of wishful thinking and utopian propaganda.

4. Machiavellian Social Selection

Return to the quote that began this essay:

“It is good to appear clement, trustworthy, humane, religious, and honest, and also to be so, but always with the mind so disposed that, when the occasion arises not to be so, you can become the opposite.”

Machiavelli’s point is this: To win a good reputation, you must often develop traits that others like and admire—to appear trustworthy, you must be trustworthy—but this sincerity must somehow co-exist with a sensitivity to the costs and benefits associated with these traits.

His advice was directed at rulers navigating the city-states of Renaissance Italy, but the insight applies whenever reputation management is the driver of friendly and fair-minded behaviour, which is the case with human beings. To survive and succeed in worlds in which reputation management is necessary, people must often cultivate socially valued dispositions in ways that are—somehow—both sincere and strategic.

In my view, this fact left its imprint on the ways in which social selection shaped human evolution, generating the complex and often paradoxical mixture of sincerity and strategy that characterises human altruism.

Flexible reputation management

First, although people tend to think of instincts towards uncalculating niceness, friendliness, fair-mindedness, and so on as parts of our innate endowment, I think it is more plausible that our innate endowment includes dispositions to develop such traits when they are rewarded. This is important because cultural contexts vary substantially in the degree to which niceness and friendliness are reputationally lucrative, and in the degree to which social competition favours dominance or prestige. As Marco del Giudice puts it, “Prosocial behaviors… and aggressive, coercive behaviors can be seen as different types of strategies for gaining status and resources.”

Some contexts are cold, harsh, and brutal. In those contexts, individuals often develop the cold, harsh, and brutal psychological profiles conducive to getting along and getting ahead. Stalinist Russia did not select for warmth and friendliness; puppy-like humans did not rise to the top. Or consider the despotic political regime in China: In 2011, a two-year-old girl was hit by a van in a hit-and-run in the southern city of Foshan. The incident was captured by CCTV, which also recorded 18 people walking or cycling past the dying girl, noticing her without stopping to help. I do not know for sure, but I doubt the same would happen in Amsterdam, and I suspect this difference is rooted in the kinds of dispositions favoured and disfavoured in these very different cultural and political contexts. Indeed, the very idea of caring about strangers seems to be strongly linked to reputational incentives that emerge in highly individualistic, fluid social ecologies in which people benefit from signalling “impersonal prosociality”.

More generally, the fact we are disposed to cultivate prosocial traits only when they are socially rewarded explains why human communities invest so much effort into monitoring, rewarding, and punishing human behaviour. If powerful instincts towards niceness and fair-mindedness were simply parts of our genetic inheritance (like our instincts for sex, fatty foods, and helping our children), there would be no need to invest so substantially in creating and enforcing incentives—systems of social rewards and punishments—that align self-interest with prosocial behaviour. It also explains why even uncalculating cooperation—cooperating without looking— seems to be sensitive to whether other people are looking.

Targeted reputation management

If friendliness and fair-mindedness result from reputation management, one would also expect them to be targeted at potential audiences whose social approval makes a difference to one’s survival and success. From a harsh, Darwinian perspective, there is little point in cultivating a reputation for niceness and generosity among people who cannot benefit you. To quote the title of Jonathan Glazer's recent film, such people fall outside the “zone of interest”. Depicting the loving family life of the commandant of Auschwitz, Rudolf Höss, the film cleverly uses physical space—the separate locations of their family home and the arena of mass genocide it is next to—to illustrate a general feature of human psychology: the extreme selectivity of feelings—sincere, deep-rooted feelings—of warmth, empathy, and generosity.

Examples are ubiquitous. Slavery is an ancient practice that was nearly universal until very recently, and throughout history, slave owners have combined powerful feelings of empathy, generosity, and fair-mindedness towards their social equals with psychopathic disregard for the slaves they exploit, rape, and kill. It is illegal to enslave human beings today, but most of us benefit and (through our consumer choices) financially support factory farming, the industrial-scale torture of tens of billions of animals every year. One could make a version of “The Zone of Interest” about a friendly factory farmer who loves his children and then goes to work to torture animals so that people can have cheap chicken nuggets, except for the fact that modern audiences generally do not see anything wrong with eating the chicken nuggets.

When idealistic social scientists and writers are forced to confront the extreme selectivity of human altruistic concern, they typically appeal to mysterious, non-rational forces like “ingroup/outgroup psychology” or “othering”. According to this perspective, humans are instinctively altruistic but only towards an “ingroup”. The challenge, then, is to expand this ingroup to include more people—and perhaps animals as well.

This analysis is shallow. What matters, fundamentally, is not an abstract distinction between ingroup and outgroup but rather which people’s judgements matter to one’s self-interest. Given that one’s “ingroup” is often just the name we give to communities where it’s important to achieve a good reputation, the two things often overlap. But they do not always overlap.

Further, the analysis provides no explanation of why people tailor their degree of niceness toward different people within their community. People might have a baseline level of friendliness toward members of their ingroup, but they also adjust the strength of their warmth, generosity, empathy, and friendliness to those individuals where they earn the greatest returns. This includes more attractive and higher-status people, as well as people they are motivated to develop and maintain strong relationships with.

If this more cynical view is correct, why do people ever care about strangers or non-human animals? This is a complicated question. The short answer is that they mostly do not. I will publish more essays (and a forthcoming book, “Why it’s OK to be cynical”) in the future that provide a much longer answer. For now, though, I will merely note that competitive altruism and virtue signalling seem to play a major role: For complex and contingent reasons, it became reputationally profitable within specific societies over the past several centuries to expand the domain of moral considerations to encompass all humans. The important thing about targeted reputation management is therefore not so much that people focus their altruism on audiences whose judgements affect their interests; what matters is which traits those audiences reward. If moral universalism makes people look good, we should expect such virtues to emerge. The same is true of veganism.

Dynamic reputation management

Machiavelli’s specific point in the above quote is that the incentives associated with reputation management can change, sometimes rapidly. When they do—when the costs of being virtuous come to outweigh their reputational benefits—people must change course, abandoning virtue. This is difficult to do if the virtues are sincere. Nevertheless, people often manage this complex balancing act. Although they cooperate without looking—without miserly accounting of personal costs and benefits—they cannot help noticing when significant changes in incentives occur. As Moshe Hoffman and colleagues put in an admirably cynical passage on love,

“[F]alling in or out of love depends on the distribution of temptations but not their immediate realizations, suggesting that people will fall out of love when there is a permanent change in alternative mating opportunities or relationship costs but not when there is a one-off temptation. For example, one may fall out of love with one’s partner after becoming unexpectedly successful.”

This is a brutal insight, but I suspect many people will recognise it as an accurate description of how relationships often form and end in the real world. It also illustrates a more general phenomenon: Yes, people are often sincerely prosocial, friendly, fair-minded, and generous, but they—or at least their brains—also dimly track how personally advantageous these traits are. When circumstances change, the sincere virtues can disappear.

5. How Machiavellian is Machiavellian social selection?

When psychologists use the term “Machiavellianism,” they refer to a specific personality trait characterised by low empathy, cynicism, and the manipulative use of other people. Along with narcissism and psychopathy, it makes up the so-called “dark triad” personality.

In this sense, most people are not Machiavellian. People can be extremely empathetic, friendly, nice, fair-minded, and trustworthy. They can be completely sincere when they say they view people as ends in themselves. Any theory of humanity that denies these plain facts and hence fails to distinguish psychologically normal humans from Machiavellian schemers is wrong.

Nevertheless, although one can be too cynical about human nature, one can also be insufficiently cynical. Humans are not angels, even if we sometimes present ourselves as such. We are apes that evolved through a competitive Darwinian process, and our motives and instincts must reflect the constraints of this process. Unconditional and un-strategic cooperators cannot evolve.

Freud was wrong about many things, but he got one major thing right: The human mind is extremely weird. It is characterised by conflict, division, and self-opacity. These features arise primarily from a clash between our animalistic motives—the kinds of goals one would expect competitive primates to have (sex, food, status, resources, etc.)—and our needs to win social approval, trust, and esteem. To navigate this clash, we have evolved to balance the demands and tensions of appearing good while being selfish. The result is a moral psychology that is, paradoxically, sincere and strategic. Machiavelli was right about that, even if most of us are not Machiavellian.

---

---