Chapter V
The mind that split open
Language, gossip, dreams, trance, and the curious fact that children are born believers.
In 2004, the psychologists Jesse Bering and David Bjorklund staged a puppet show for children in Florida. A little mouse puppet wandered through a forest, feeling hungry and thinking about going home. Then an alligator appeared and ate it. "Well," the children were told, "the mouse is not alive anymore." Then the researchers asked questions. Does the mouse still need to eat? Does he still need to drink? Almost all the children, even the youngest, knew that the answer was no. The body was finished. But then they were asked other questions. Is the mouse still hungry? Does he still love his mom? Does he still want to go home? Here, the answers changed. Many of the younger children said yes. The body had stopped, but the mind, in their intuition, went on.
The study is small and its interpretation debated, but it points toward something that decades of developmental research have supported. Human beings seem to be born with a set of intuitions that make religious ideas easy to think. Children readily separate minds from bodies. They see purpose everywhere; the psychologist Deborah Kelemen has shown that young children prefer explanations such as "rocks are pointy so that animals can scratch on them" over physical explanations, a tendency she called promiscuous teleology. They attribute feelings to objects, and they are quick to imagine invisible friends. The Swiss psychologist Jean Piaget, in the 1920s, called this childhood animism and assumed children would grow out of it. More recent researchers suspect that we never really do. We simply learn to override it.
This chapter is about the mind that holds these intuitions: how it evolved, what made it capable of imagining spirits and the dead, and why it began, at some point in its long history, to organize those imaginings into shared religious worlds.
The body had stopped, but the mind, in the children's intuition, went on.
The cathedral of the mind
In 1996, the British archaeologist Steven Mithen published The Prehistory of the Mind, one of the most ambitious attempts to connect archaeology with cognitive science. He proposed that the mind of early humans was like a cathedral with several separate chapels. One chapel held social intelligence, the capacity to understand other people. Another held natural history intelligence, knowledge of animals, plants, and landscapes. A third held technical intelligence, the skill of making tools. In earlier hominins, and perhaps in Neanderthals, Mithen argued, these chapels were largely sealed off from one another. A Neanderthal could be a brilliant toolmaker and a subtle social operator and a superb tracker, but could not easily combine these domains.
What happened in modern humans, Mithen suggested, was the opening of doors and windows between the chapels, a development he called cognitive fluidity. Once social intelligence could flow into natural history, animals could be thought of as people, and people as animals: totemism, animism, and the lion-man became possible. Once technical intelligence could flow into social intelligence, objects could carry social meanings: beads, ornaments, symbols. Once natural history could flow into technical intelligence, tools could be designed for particular animals. Art, religion, and science, in this model, were all products of the same mental revolution.
Mithen's model has been criticized as too neat, and the evidence that Neanderthals made ornaments and used pigments undermines the sharpness of his contrast between them and us. But the core idea has proved remarkably fertile. Religion, from this point of view, is not a separate faculty. It is what happens when a mind built to think about people starts thinking about everything else in the same terms.
The social brain and the problem of the crowd
Why did the human brain become so large and so socially skilled in the first place? The British evolutionary psychologist Robin Dunbar proposed an influential answer in the 1990s, the social brain hypothesis. Across primates, he found, the size of the neocortex relative to the rest of the brain is correlated with the size of the social group a species typically lives in. Larger groups require more brain power to keep track of who is allied with whom, who owes what, and who can be trusted. Extrapolating from primate data, Dunbar predicted that humans should naturally maintain stable relationships with about 150 people, a figure that has become famous as Dunbar's number, and which corresponds surprisingly well to the size of many hunter-gatherer communities, Neolithic villages, and military units.
Primates maintain their relationships by grooming, picking through each other's fur, which releases endorphins and builds trust. But grooming is time-consuming and can only be done one-on-one. As human groups grew larger, Dunbar argued, grooming could no longer hold them together. Language, he suggested, evolved partly as a form of vocal grooming, allowing people to bond with several others at once and, crucially, to exchange information about people who were not present: gossip. Much of human conversation, then and now, is about other people.
But gossip has a limit too. For larger groups, and for the periodic gatherings of several bands, something more was needed. Dunbar and his colleagues have argued that this is where ritual came in. Laughter, singing, dancing, and rhythmic movement in synchrony all trigger endorphin release, producing a feeling of warmth and connection among many people at once. Experiments show that people who sing, dance, or row together in synchrony report stronger bonds and have higher pain thresholds afterward, a marker of endorphin release. Religious ritual, in this light, is a technology for bonding groups larger than grooming or conversation can manage. It is Durkheim's collective effervescence, explained through brain chemistry.
Religious ritual is a technology for bonding groups larger than conversation can hold together.
The ladder of intentions
Dunbar has also pointed to a subtler cognitive requirement for religion. Philosophers speak of orders of intentionality. If I believe something, that is first-order intentionality. If I believe that you believe something, that is second order. If I believe that you want me to believe that she intends something, that is fourth order. Most adults can manage about five orders before they lose track. Chimpanzees seem to manage about two.
A shared religion, Dunbar argues, requires the upper reaches of this ladder. For a group to share a belief in spirits, I must believe that you understand that we both believe that there are beings who intend certain things and who can know what we want. Religion is not just a belief. It is a belief about minds, held in common by other minds, each aware of the others. It depends on the most sophisticated social cognition our species possesses. If this is right, then religion in the full sense was not possible until the mind had developed the capacity for high-order intentionality, which may be one reason why its clear archaeological signs appear relatively late.
The gift of language
When did human beings begin to speak? The honest answer is that no one knows. Language leaves no fossils. Researchers have looked for indirect evidence: the shape of the hyoid bone in the throat, which in the Neanderthal from Kebara (whom we met in Chapter II) is almost identical to ours; the size of the canal through which nerves reach the tongue; the imprints of brain regions on the inside of skulls; and the gene FOXP2, involved in the fine motor control required for speech, whose modern human version Neanderthals also shared. Most of this evidence suggests that the physical capacity for speech is at least several hundred thousand years old, and may have been present in the common ancestor of Neanderthals and modern humans.
Whatever its date, language transformed the possibilities of religion. Without language, one can feel awe, fear, grief, and the presence of something beyond. With language, one can name it, describe it, tell stories about it, and pass those stories to others who have never had the experience. Language makes the absent present. It allows the dead to be spoken of, ancestors to be recited in genealogies, and the landscape to be filled with narrative. It allows rules to be stated: do not eat this animal, do not speak this name, do not enter this place. And it allows the invention of worlds that no one has ever seen, which is, among other things, what religion does.
The linguist Daniel Everett, who lived for many years with the Pirahã of the Brazilian Amazon, reported a people whose language and culture emphasized immediate experience, what had been directly witnessed, and who had little interest in creation myths or distant pasts. The Pirahã do encounter spirits, which they see and speak with directly. Everett's claims about their language are fiercely debated. But the case suggests that the particular shape of a people's religion may be intertwined with what their language and culture treat as worth saying.
Dreams: the first other world
Every human being dreams, and every known culture has theories about what dreams are. For most of human history, the dominant theory was that dreams are real experiences. In sleep, the soul travels. It visits places, meets other people, and encounters the dead. When you dream of your dead grandmother, you have met her. When you dream that a neighbour attacked you, you may have reason to be wary of that neighbour.
This was the heart of Tylor's theory of the origin of religion, and for all the criticism the theory has received, the centrality of dreams to hunter-gatherer religion is beyond doubt. Among many Aboriginal Australian groups, songs and ceremonies are received in dreams from ancestors. Among many Native American peoples, the vision quest, a solitary ordeal of fasting and exposure to induce a visionary dream, was the central religious experience of a young person's life. Among the Achuar of the Amazon, Descola found, people discussed their dreams every morning at dawn to decide how to act that day.
Modern neuroscience offers various accounts of why we dream. The Finnish psychologist Antti Revonsuo proposed the threat simulation theory: dreams evolved to rehearse responses to danger, which is why so many dreams involve being chased, falling, or failing. Others emphasize the role of dreams in consolidating memory and processing emotion. Whatever their function, dreams gave every human being, every night, the direct experience of another world, one in which the dead were alive, animals spoke, the body could fly, and the self could become someone else. It is hard to imagine a more powerful source for the conviction that ordinary reality is not the only reality.
Dreams gave every human being, every night, the direct experience of another world.
The doorways of trance
If dreams were the nightly other world, trance was the other world entered on purpose. As we saw in Chapter III, altered states of consciousness can be induced by many techniques: rhythmic drumming, dancing to exhaustion, hyperventilation, fasting, sleep deprivation, sensory deprivation in dark and silent places, pain, and psychoactive plants. The anthropologist Erika Bourguignon, in a 1973 survey of 488 societies, found that 90 percent had institutionalized forms of altered states of consciousness. Such experiences are not the exotic exception in human religious life. They are close to the norm.
The evidence for psychoactive plants in deep prehistory is scattered and often speculative. Claims that painted figures in the Tassili n'Ajjer of the Sahara depict mushroom-headed beings remain unproven. More secure are the traces of fermentation. At Raqefet Cave in Israel, a team led by Li Liu reported in 2018 residues in stone mortars that they interpreted as evidence of beer-making by Natufian people about thirteen thousand years ago, at a site used for burying the dead. If confirmed, it would mean that fermented drink and funerals went together long before agriculture. Alcohol, of course, is a powerful way of dissolving the boundaries of ordinary consciousness, and feasting with it is among the most ancient forms of human ritual.
What altered states did, whatever their cause, was provide direct, personal, overwhelming evidence of powers and beings beyond the ordinary world. A person who has experienced being transformed into an animal, flying above the land, or meeting the dead in trance does not merely believe in the spirit world. He or she has been there. The authority of the shaman, and of religious experience more generally, rests on that claim of direct knowledge.
The voices in the head
No discussion of the primordial psyche would be complete without mentioning one of the strangest and most brilliant theories ever proposed about it. In 1976, the Princeton psychologist Julian Jaynes published The Origin of Consciousness in the Breakdown of the Bicameral Mind. Jaynes argued that, until as recently as about three thousand years ago, human beings did not have introspective consciousness as we know it. Their minds were "bicameral," divided into two chambers: one part, associated with the right hemisphere of the brain, generated commands in the form of auditory hallucinations, which the other part heard and obeyed as the voices of gods. The heroes of the Iliad, Jaynes pointed out, do not deliberate; gods tell them what to do. Consciousness, in his view, emerged only when the bicameral mind broke down under the stresses of large, complex societies, and the gods fell silent.
Almost no specialists accept Jaynes's theory in its strong form. The archaeological and linguistic evidence does not support the idea that people who built cities, kept accounts, and made laws lacked introspection. Yet the book remains widely read, and for good reason. It takes seriously the idea that the experience of hearing voices, of being addressed by an authority from outside the self, was once far more common and more central to human life than it is now. Research on auditory hallucinations shows that they are common in the general population and that their character varies by culture: the anthropologist Tanya Luhrmann found that people with schizophrenia in India and Ghana more often reported voices that were familiar and even helpful, while those in the United States more often reported hostile, violent voices. How a culture understands the mind shapes what the mind experiences. That insight, at least, Jaynes would have recognized.
Costly signals and the trust problem
There is one more reason why religious behaviour may have been favoured in human evolution, and it has to do with trust. Human cooperation beyond close kin is fragile. Anyone can claim to be a loyal member of the group; free riders who take without giving are a constant threat. How can you tell who is really committed?
The anthropologist Richard Sosis and others have argued that religious rituals function as costly signals. Anyone can say they are committed, but performing painful initiations, observing difficult taboos, spending days in ceremony, or crawling into dark caves costs time, comfort, and sometimes health. Only those who are truly committed will pay such costs. By paying them, they prove their commitment to others, and the group can extend them trust. Sosis found, in a study of nineteenth-century American communes, that religious communes survived much longer than secular ones, and that the more demanding the requirements, the longer the religious communes lasted.
This helps explain one of the most puzzling features of religious behaviour: its extravagance. Why crawl a kilometre into a cave to paint? Why bury children with ten thousand beads? Why drag multi-ton pillars to a hilltop? From a purely practical point of view, these acts are wasteful. From the point of view of costly signalling, their waste is the point. Only something that costs a great deal can prove that something else matters more.
From a practical point of view, religious acts are wasteful. From the point of view of signalling, their waste is the point.
The mind prepared
Let us gather the threads. By the time modern humans spread across the world, their minds contained a remarkable toolkit. They were prone to detect agents everywhere and to attribute minds to things that moved or responded. They intuitively separated mind from body and could imagine minds continuing after death. They saw purpose in the world. They had language to name and describe the unseen. They dreamed, nightly, of another world, and could reach that world deliberately in trance. They lived in groups whose cohesion depended on emotional bonding through synchronized ritual, and on costly displays of commitment. They could hold beliefs about beliefs, several layers deep, and share them.
None of these capacities is religion in itself. But together they make religion almost inevitable. The cognitive scientist of religion Robert McCauley has called religion "natural," in the sense that it builds on intuitions that come easily to the human mind, while science is "unnatural," requiring us to override those intuitions with difficult, counterintuitive reasoning. That is a humbling thought. It means that the primordial psyche is not a stage we have left behind. It is the default setting of the human mind, and every one of us begins life inside it.