Why we believe in the supernatural

Around the world and throughout history, people have expressed belief in the existence of powerful, invisible entities who — much like us — have intentions, needs, and goals. Whether gods, demons, ghosts, or ancestor spirits, these “supernatural agents” cannot be proven to exist, yet belief in them is strongly held and expected of others in some religious communities.

Despite being separated by vast distances and oceans, cultures around the world have developed strikingly similar beliefs about the nature and character of these supernatural entities. Why have so many people throughout history believed in the existence of supernatural agents? Where did these ideas originally come from and how did they arise?

How did belief in supernatural agents originate in human history and why is it still so ubiquitous today?

Religion is natural

The Cognitive Science of Religion (CSR) attempts to answer such questions using interdisciplinary methods derived from neuroscience, anthropology, and experimental psychology. According to Justin Barrett, a foundational member in CSR, “the cultural phenomena typically labeled as ‘religion’ may be understood as the product of aggregated ordinary cognition.”

This is an apt summary of the naturalness-of-religion thesis, which is the assumption that religion results from ordinary cognitive processes, rather than arising from rare or extraordinary cognition. In this view, belief in supernatural agents isn’t extraordinary, but a natural part of the human experience. This reorients the scientific study of religion to the natural foundations of religion, focusing on what is similar between religious experience and ordinary life.

Several CSR theories provide explanations for why people proposed the existence of gods or “supernatural agents” — a catch-all term for disembodied beings purportedly endowed with intention and agency.

The first of these theories attempts to explain why these supernatural beings are generally characterized as “agents” — that is, sentient, person-like beings who undertake intentional actions in accordance with their feelings and goals.

Quote: “The cultural phenomena typically labeled as ‘religion’ may be understood as the product of aggregated ordinary cognition.” — Justin Barrett, “Exploring the natural foundations of religion”

Why supernatural agents? The Hyperactive Agency-Detection Device

Drawing upon prior work by CSR researchers, Justin Barrett hypothesized that human cognition includes a hyperactive agency-detection device (HADD) that biases people to attribute intentional action in uncertain and ambiguous situations.

For example, if we hear a branch snap, we quickly leap to the assumption that someone or something is there, rather than passively assuming falling snow or some other inanimate force snapped the branch. We automatically assume agency.

Photo by Kaung Myat Min on Unsplash

HADD persisted in humans because it provides an evolutionary advantage, increasing our wariness and survival rate. While the costs of a false positive (assuming the snapping branch is a dangerous agent, when it was only falling snow) are negligible, the penalty for a single false negative (assuming it’s only falling snow, when it’s really a hungry lion) could be fatal.

Because of this, early humans may have subconsciously populated their environment with supernatural agents like invisible people or spirits to make sense of strange and otherwise unexplained phenomena.

But why would people believe that these agents have supernatural powers?

Quote: “We constantly anthropomorphize. This is probably a genetic inclination. It is the source of religious belief in spirits, gods, God.” — Stewart Guthrie, Faces in the Clouds

Supernatural agents are minimally counterintuitive

One theory developed by cognitive anthropologist, Pascal Boyer, notes that supernatural agents are usually minor alterations of deeply ingrained concepts, such as a person, animal, or inanimate object.

For example, we expect animals to follow certain behaviors — such as searching for food — and to have certain physical qualities — such as solid, sensitive bodies. Beings with supernatural qualities violate some of our expectations about a concept of that type, thus catching our attention (an animal that can transform into a human), while also remaining inferentially rich due to fulfilling other expectations derived from the underlying concept (needing to eat and sleep).

Boyer calls such concepts minimally counterintuitive.

Photo by Syarafina Yusof on Unsplash

For example, a ghost is like an ordinary person — it can see and hear, remembers things, has needs and goals, can become offended, etc. — yet it lacks a physical body. Empirical studies have shown that people remember minimally counterintuitive concepts, such as ghosts, better than ordinary, fully intuitive ideas, making them more likely to spread and become ingrained in human cultures. This could explain the ubiquity of stories and beliefs about supernatural entities throughout human cultures.

Although sophisticated theological doctrines may be very complex and highly counterintuitive, everyday religious beliefs are often more common-sense and fit the criteria for minimally counterintuitive concepts. For example, theologians understand that the god Shiva knows our every thought, but most believers still feel the need to express their thoughts to Shiva via prayer. Complex theological doctrines would have developed from more popular, minimally counterintuitive concepts.

Many commonsense religious beliefs about supernatural agents hit a sweet spot between inconceivability and memorability, allowing human minds to easily conceptualize these beings and, despite their admitted strangeness, to develop deep conviction in their reality. The strangeness of these concepts actually makes them more appealing and more likely to be remembered, passed on to future generations, and preserved in the wider culture.

Quote: “Once we understand how mental representations can be transmitted, [and] that not all ideas can be transmitted equally easily, then we have the first elements of an explanation of religious representations.” — Pascal Boyer, The Naturalness of Religious Ideas

Encountering the supernatural in dreams

In addition to HADD, dreams are another possible source of supernatural agent cognitions. We often feel that the characters in our dreams have thoughts and intentions, and they sometimes display incredible powers, such as monsters in nightmares.

Almost everyone has dreams, and across cultures, dreams have been thoroughly intertwined with religious beliefs, practices, and experiences. Neuroscientist and sleep researcher Patrick McNamara explains that “all humans are endowed with brains innately primed to daily generate god concepts in dreaming.”

Stories about life-changing dreams are common in religious literature and oral tradition. In these dreams, the dreamer feels that they have really encountered a supernatural agent. For example, the Hebrew Bible relates Jacob’s dream of the ladder leading to heaven, and Jacob awakes convinced that God will protect him. The diary of Lady Sarashina “As I Crossed a Bridge of Dreams” relates her experience of personally meeting Amida Buddha in a dream which leaves her convinced of salvation.

Photo by Jr Korpa on Unsplash

Perceived encounters with supernatural agents in dreams can have effects which change the dreamers’ personality and behavior for months or years thereafter.

Many religions also explicitly endorse dreams as a possible source for attaining knowledge and truth. Religious rituals and practices dictate how to achieve a religious dream. Others “test” dreams to determine if the message of the dream is legitimate, and to disempower demonic nightmares.

Dream-interpreters are important religious figures in many religions, and several civilizations produced books on dream interpretation which were used for centuries. Examples include the Greek Artemidorus’s “Oneiocritica,” the Chinese Zhang Fengyi’s “Meng-shan-lei-kao,” and the ancient Egyptian “Ramesside Dream Book.”

Dreams are often taken as direct evidence of a spirit realm and of disembodied spirits. In visitation dreams, the dreamer encounters a deceased friend or relative. Upon waking, the dreamer is absolutely convinced the person in the dream really was the loved one. Visitation dreams have real emotional and cognitive effects, such as a sense of closure and diminished grief, suggesting they cannot simply be labeled as delusions.

Quote: “All the world’s religious traditions have pointed to dreams as spiritual events of extreme importance in the transformation of the self and with respect to communication with the spirit world.” — Patrick McNamara, “Dreams as a source of supernatural agent concepts”

Modern CSR has paid scant attention to dreams as a source of religious consciousness, despite the vast evidence of associations between dreams and religion. At the root of this is a perception of dreams as merely bizarre, meaningless byproducts of the sleeping brain. However, as we learn more about the neuroscience of dreams, we find that dreams are a highly complex cognitive product of a specialized brain system.

But what is it about dreams that makes them so memorable and prone to producing encounters with perceived supernatural agents? In our next article, we will explore cutting-edge research on the cognitive neuroscience of REM sleep to better understand how this universal feature of human experience could have played a key role in the origin of religion.

But what about Mickey Mouse?

Photo by Steven Beyer on Unsplash

Kim Sterelny, who approaches the origin of religion as a process of human cultural evolution, points out a glaring flaw in CSR’s reasoning about the origin of supernatural agent concepts from HADD, minimally counterintuitive concepts, or dreams. The fact that people may accept supernatural agents as sufficiently reasonable does not explain why people attach religious importance to these concepts.

This is often called the Mickey Mouse problem. Sterelny quips, “a talking mouse is salient and memorable, but not thereby an object of veneration.”

Believability does not explain the high levels of commitment and investment that characterize many religions. Explaining people’s investment into their religion requires a slightly different approach, looking at the origin and development of religious rituals rather than the origin of supernatural agent concepts.

In this companion article, we explore anthropological theories about human cultural evolution and how engagement in community rituals could have led to the rise of religious belief and ideological religions.

Quote: “But if religious ideas are intrinsically implausible, […] how do the first generations of believers appear?” — Kim Sterelny, “Religion re-explained”

Through a process of gradual cognitive evolution and cultural competition, religion emerged in human societies. Humans’ tendency to attribute agency due to HADD and dream experiences created the first supernatural agent ideas. The ideas of these supernatural agents survived because they were minimally counterintuitive and easily passed down in stories.

The origin of supernatural agent concepts — and continued commitment to these beliefs — is a complex process that sheds light on humankind’s first attempts to explain the world.

Author profile
Jessie Saeli
Religion & Media Fellow

Jessie is a student in the M.A. program in philosophy at Boston College. She began her graduate studies after graduating magna cum laude from the University of Notre Dame, where she studied philosophy and Russian. Her primary interests are existentialism, Russian philosophy, and the question of what a “self” is. She enjoys reading and writing about philosophy, theology, and literature.