I’m guessing that most readers of this essay are unfamiliar with the theory its title refers to. That’s a pity, because multilevel selection (MLS) theory, which was developed to explain sociable relations in all species, may also hold the key to understanding much of what we hold dear in our social relationships, including the importance of free speech—and how to protect it.
Charles Darwin and Alfred Russell Wallace were not the first to think about evolution as a grand explanatory theory. Darwin’s grandfather Erasmus, as well as Herbert Spencer and the anonymous author of Vestiges of the Natural History of Creation (published in 1844), were thinking about it that way too. Those writers (incorrectly) envisioned evolution as a force for progress, leading toward ever higher life forms and fulfilling humanity’s spiritual destiny. Although this was a scientific idea, it was uplifting in the same way that a religion such as Christianity can be uplifting. In fact, it was because these thinkers were working toward an uplifting worldview without relying upon religion that they were such an important part of what later came to be called the Enlightenment.
Before Darwin and Wallace developed their theory of evolution by natural selection, animal and plant breeders were well aware that organisms vary in nearly everything that can be measured about them and that traits in parents tend to “breed true” in their offspring, even if the mechanisms of inheritance were unknown. This is what enables the selection of favourable traits in our crops and domesticated animals. Darwin and Wallace added the insight, which seems obvious in retrospect, that the same ubiquitous variation will result in differences in survival and reproduction in the natural environments of all species, so that a form of “natural” selection will occur analogous to “artificial” selection.
Imagine what it must have been like to be Darwin or Wallace at the dawn of their discovery. Never had such a simple theory explained so much. Identity by descent. The fossil record. The geographical distribution of species. The mysteries of development. And the wonderful contrivances of animals and plants that adapted them to their environments. But there was one phenomenon that Darwin and Wallace could not explain with their theory. Since natural selection favours traits that cause some individuals to survive and reproduce better than others, how can it favour traits that cause individuals to act for the benefit of others, at their own expense? How can it explain just about everything that we call moral in human life, such as altruism, honesty, bravery and charity?
This was a gaping hole in the theory of natural selection—and one which made it the opposite of uplifting. Instead, it suggested a worldview that might be described as the evil twin of “Imagine,” John Lennon’s uplifting song envisioning social harmony. Imagine, if you will, these alternative lyrics:
Imagine there’s no goodness
It’s awful but you must try
Just individuals striving
To get the biggest slice of the pie
Imagine all the people
Livin’ for themselves …
This implied worldview is one reason why the theory of natural selection, though profoundly explanatory, was also profoundly disturbing. Fortunately, a little more thinking enabled evolutionary thinkers to soften the blow. It turns out that the theory of natural selection can explain how the behaviours that we associate with morality can evolve in any species, including our own. Let’s use the word prosocial to describe any behaviour oriented toward increasing the welfare of other individuals or of one’s social group. Not all such behaviours are self-sacrificial (altruistic), but they generally exact some cost—such as time, energy or risk—that less prosocial others in the group are not necessarily incurring. As a result, individuals who tend to be more prosocial usually have a fitness disadvantage compared to more self-serving individuals within their social group. However, groups with more prosocial individuals tend to robustly outcompete groups with fewer prosocial individuals. So prosocial behavioural traits that can be a disadvantage at the individual level can nevertheless evolve because of the fitness advantage of being in a group whose members are more prosocial towards each other compared to rival groups.
This is called group selection theory, although it would be better named two-level selection theory to acknowledge that selection is taking place both within and between groups.
It can be illustrated by an analogy to the game of Monopoly. That game as usually played is one-level selection—competition among individuals within a single group. The object is to drive everyone else bankrupt by buying up real estate. There is no room for cooperation beyond temporary alliances that are designed to reduce the number of competitors by eliminating a front-runner. The world of Monopoly is the world described by the evil twin of the song “Imagine.”
To bring in two-level selection, imagine that you are midway through a game of Monopoly when you learn that you are playing in a Monopoly tournament, where multiple groups are playing separate games, and the trophy goes to the group that collectively develops its real estate the fastest. That’s a real game-changer! You will immediately stop competing with the other players at your table, and start planning strategy with them. They are no longer your rivals; they have become your teammates. You will trade properties so that everyone in your group can start building houses and hotels. You will divide labour, making one person the banker. You will pool your money so that each person can afford to buy every property that they land on. There will be cheers, hugs and high fives if you win the tournament. All of these changes result from replacing competition among individuals with competition among groups.
That’s the good news. The bad news is that, while a Monopoly tournament like that allows for cooperation within teams, it doesn’t allow for cooperation among teams. To explain how that could evolve, we need to add another layer, making the tournament a competition among groups of groups. This how two-level selection theory expands into multi-level selection theory whenever social interactions take place in a nested hierarchy of groups.
Once you understand the logic of MLS theory, it becomes clear that what counts as prosocial at one level can become disruptively self-serving at higher levels, in both nature and humans life.
Self-preservation is a good thing—until it leads to self-dealing. Helping family and friends is a good thing—until it leads to nepotism and cronyism. Corporations maximizing profits for their shareholders is a good thing—until it leads to obscene heights of inequality. Nations growing their economies is a good thing—until it overheats the earth.
Before I suggest how MLS theory can illuminate issues related to free speech— and how it can even offer us some measure of the uplift that was offered by pre-Darwinian theories of evolution—I need to set the stage a bit more. The conditions under which higher level selection can occur are quite narrow, and often fail to occur. In both nature and human society, we should expect to find little islands of goodness in a sea that is filled with fellow creatures who are mostly striving to maximize their positional advantage.
Many people find it difficult to fully absorb the extent to which evolution can result in suffering and disorder when conditions that support prosocial behaviour are absent. Consider the kinds of scenarios that unfold in the absence of prosocial cooperation. Imagine that you are a woman who married a nice guy and are raising a family with him. Then a not-so-nice guy forces his way into your home, slays your husband and children in front of your eyes, and forces you to have his children. Or imagine that you decide to forcibly evict your next-door neighbours from their house so that your brood can move in after they reach marriageable age. These scenarios are analogous to the game of Monopoly. They show how life plays out when lower-level selection prevails against higher-level selection. They have played out countless times throughout human history, and still do, quite commonly, in some parts of the world. They also play out throughout the natural world. The idea that nature, if left to itself, would result in some kind of happy, harmonious balance is a vestige of pre-Darwinian Christian thought. Undisturbed nature is not the Garden of Eden, and MLS theory provides no warrant for expecting that it should be.
Yet there is a way to derive a uplifting message from MLS theory—by focusing on the islands of goodness that can be created as a result of higher-level selection for prosocial cooperation within and among groups. By uplifting, I mean not only spiritually uplifting—producing an animating psychological state oriented toward a common good—but also socially uplifting—reinforcing the conditions that enable prosocial behaviours to evolve in a Darwinian world. Those of us who have the good fortune to have access to islands of goodness can build them into larger islands, and we can create new islands for ourselves and others. Think of it as a grand social engineering project—not forcibly imposed on others through power or manipulation, but accomplished together. Done with, not done to, like an Amish barn raising. MLS theory suggests that this can be accomplished.
Compared to our closest primate relatives, our species is replete with prosociality, at least at the scale of small groups. Chimp communities exhibit a little cooperation and a lot of Monopoly-like competition. Naked aggression is over 100 times more frequent than in small-scale human communities. Cooperation is limited to immediate family members, dyadic friendships, and alliances formed to compete against other alliances within the community. The main context for community-wide cooperation is competition against adjacent communities, as MLS theory would predict.
Something happened during the evolution of our species that enabled members of small groups to become much more prosocial than chimps. To a large extent, that something was the addition of social controls. A bully who attempts to get her way in a chimp community often succeeds. But a bully who attempts to get her way in a small-scale human community often meets with vigorous pushback from other community members, which usually puts the bully back in her place. Dominating behaviour exists in human communities, but there is a collective version of it in addition to the individual version. When this collective domination operates strongly, the only way to achieve high status in the community is by contributing to the good of the group. Status must be earned by reputation, not by the exercise of raw power.
Just about everything that we associate with human morality can be understood from this perspective. Human prosocial behaviour (morality) has two dimensions. On the one hand, we establish norms and require each other to conform to them. On the other hand, we want to help others—motivated by emotions such as sympathy, empathy, friendship and love. This dual system makes sense: the compulsory normative framework makes it safe to behave in accordance with our desire to help others. Having a moral system protects the group from its members’ disruptive self-serving social strategies in the same way that the body’s immune system protects it from invasive disease. Just like our immune systems, our moral systems are always vigilant, often challenged, and sometimes overcome. But on an average day they do a pretty good job of keeping social life on an even keel.
How do these observations apply to the considerations of free speech? To be free is to do and say what one likes in pursuit of one’s welfare. People generally feel that they should be free from being pushed around, free to defend their rights, and free to defend the rights of others whom they think of as being in their group. People generally feel they should not be free to engage in Monopoly-like behaviour within their own group. That kind of behaviour would impair their group’s moral immune system, without which prosocial behaviours cannot safely be practised. Furthermore, the freedom to pursue one’s own welfare almost always requires the freedom to work cooperatively with others. Very little can be accomplished alone. The kind of freedom worth wanting is the elbow room to pursue one’s cooperative interests without destroying the infrastructure that makes cooperative social life possible.
What about freedom to speak? Multilevel selection dynamics govern social behaviours that require cooperation within and among groups. And speech acts are just that: social behaviours that influence—and are influenced by—others’ social behaviours, both within and across groups. Language itself is a symphony of cooperation, requiring us to maintain an inventory of symbols with shared meaning and transmit them across generations. Speech requires us to trust the members of our group and have at least some minimal sense of their perspectives and interests—when people become enemies, it is often said that they are no longer on speaking terms. Multilevel selection dynamics teach us that, when we speak to each other, we need elbow room to pursue our cooperative interests—but we also need to avoid speech that leads to Monopoly-like behaviours within our group, because that would erode the islands of cooperative social life upon which the phenomena of speech depend.
Speech also plays an indispensable part in establishing and maintaining the human moral systems on which both freedom and cooperation depend. Consider this episode, which was observed by anthropologist Hortense Powdermaker on the Melanesian island of Lesu:
There is much talk in the village because Tsengali’s pig has broken into Murri’s garden. Murri displays no particular anger but Tsengali is much annoyed because of all the talk that the incident has occasioned. So he announces that he will give the pig to Murri to stop the talk. But Murri tells him it would be foolish “to eat a pig for nothing.” Instead, Murri declares that the incident has ended, and that there should be no talk.
In the same way that the tiniest injury to our bodies, such as a splinter, results in a swarm of activity by our immune system, this tiny disruption of the social order—not even one caused by bad intentions—resulted in “much talk in the village”: an activation of the moral immune system, using the tools of speech. And, just as our immune system will escalate its defence of the body to the level that is necessary to restore health, we can be confident that the villagers’ response would have escalated beyond talk if Tsengali’s behaviour had continued to violate their moral norms.
The “talk” in this anecdote might be dismissed as mere gossip, but gossip and other forms of small talk are far more sophisticated and morally important than they are often given credit for being. Over twenty years ago, I reviewed the literature on gossip and conducted a series of experiments to explore how it functions in groups. I found that it is almost universally considered acceptable—and even a moral responsibility—for people to gossip truthfully about a norm-breaking event. By contrast, when people gossip falsely, or in a malicious or self-serving fashion, that is regarded as itself a norm-breaking event, which should be (truthfully) gossiped about and criticised. When it is important to know the facts of a particular matter, people who are engaging in truthful gossip are sensitive to the quality of the information, such as whether it was reported by an eyewitness or second hand, how trustworthy the source is and the number of sources. When the facts of the matter are seen as unimportant, it’s considered OK to tell whoppers—if only for entertainment. And what we say about them follows a different set of rules than what we say about us. Thus, gossip functions exactly as would be expected from a MLS perspective.
The speech that people tend to engage in spontaneously when they are talking in small face-to-face groups is not necessarily the same as the speech that arises when groups are larger or communication is not face to face. And speech that is adaptive in one setting can be dysfunctional in another. Just as our immune systems can tragically misfire in sanitised modern environments, attacking the bodies that they evolved (in more pathogen-rich environments) to protect, our moral immune systems can also tragically misfire—when speech acts that may have been adaptive in one environment take place in new contexts, such as the digital social environments that came into existence only a few decades ago.
This may help explain why, in these new environments, it often seems that our moral immune system is misfiring—attacking the very norms it was originally designed to protect. You can’t fully understand how a system is attacking itself unless you understand what it was originally designed to accomplish. MLS theory helps explain how the human moral immune system evolved, and why the way it functions is effective in many contexts. So it may also help explain what causes the moral immune system to go haywire, and what might be done to restore its functionality. I’m not saying that it will be easy to develop this explanation, or that I have specific solutions to offer here. But I want to introduce MLS theory to readers who may be unaware of it, and suggest how it may help us better understand free speech—both what it is designed to do, and why the current challenges to it have arisen.
If MLS theory is not yet widely known, it is in part because it runs counter to a competing idea: that all things social can be fully understood if they are seen as varieties of individual self-interest. This idea can be seen in neoclassical economic theory and in the methodological individualism principle that some social scientists espouse; it may also be perpetuated in the popular imagination by the concept of the selfish gene in my field of evolutionary biology. The idea that everything can be explained by individual self-interest seems to lead people to expect that pursuing it will also somehow end up benefiting society as a whole, as if led by Adam Smith’s invisible hand. These assumptions may lead some to believe that individual self-interest alone can fulfil the role of our moral immune system, and thus that protecting free speech is merely a matter of providing an open marketplace for the exchange of ideas. This perspective ignores the role of prosocial behaviours in protecting our moral systems, and the fragility of those behaviours in contexts that our moral immune systems did not evolve to operate in. MLS theory, by contrast, takes these dynamics into consideration. If we want to work towards the world that is envisioned in John Lennon’s “Imagine”—and avoid its evil twin—we should try to understand free speech from the viewpoint of MLS theory.