What Even Is a Religion? (And Who Gets to Decide?)

Imagine you’re at a museum, and you walk past three different exhibits. The first one shows a group of people gathered in a building with a cross, singing hymns. The second shows a statue with many arms, surrounded by offerings of flowers and food. The third has a person sitting cross-legged, eyes closed, breathing slowly, with no visible god or statue at all.

You’d probably call all three of these “religions.” That seems obvious.

But then imagine walking past a fourth exhibit: a group of people standing in line at a store, waiting to buy a new video game that’s just been released. They’ve been there for hours. Some slept outside the store. They’re all wearing T-shirts with the game’s logo. They’re chanting slogans from it. They look absolutely devoted.

Is that a religion too?

Most people would say no. But why not? What’s the difference between the person sitting quietly meditating and the person camping out for a video game? Both are doing something they deeply care about. Both are part of a community. Both have rituals. Both have beliefs about what matters most.

This is the puzzle at the heart of the concept of “religion.” It seems like a simple word, but when philosophers and anthropologists try to pin down exactly what it means, things get surprisingly messy. The question isn’t just “what counts as a religion?” It’s also: did the word religion even exist before modern Europeans invented it? And if we use it to describe people from other times and places, are we distorting their lives to fit our categories?

How the Word “Religion” Got Stretched

The word “religion” didn’t start out meaning “a type of cultural practice like Christianity or Islam.” The original Latin word religio meant something closer to “scrupulousness” or “carefulness” — the feeling you get when you’ve made a promise you need to keep, or when you’re worried about a taboo. You could be religio about your duties to the gods, but you could also be religio about your duties to your family or your community. It was a feeling, not a category.

For a long time, that’s what the word meant: devotion, not a social group. In the Middle Ages, when Christians created monastic orders (groups of monks or nuns living under a rule), they called each such order a religio. But there was still no general concept called “religion” that included Christianity, Judaism, and the beliefs of people in faraway lands.

That changed in the 1500s and 1600s. As European Christians split into warring factions (Protestants vs. Catholics vs. various smaller groups), some thinkers started looking for what all these factions had in common. They wanted to find the shared core of all the Christian groups. But then they went further, and started asking: what do all the world’s belief systems have in common?

In the 1600s, a man named Edward Herbert argued that every religion shares five basic beliefs: that there’s a supreme deity, that this deity should be worshipped, that virtue matters, that you should repent when you do wrong, and that you’ll be rewarded or punished in this life or the next. You can see what Herbert was doing: he was taking his own version of Christianity and treating it as the model for “religion” everywhere. If a group didn’t believe in a supreme deity, Herbert had to argue that the multiple gods they worshipped were really just aspects of the one true God. Otherwise, he’d have to say those people didn’t have a religion at all.

This approach — stretching the concept to include everything you want to include — kept happening. In the 1800s, the anthropologist Edward Tylor proposed a much simpler definition: “belief in spiritual beings.” This was broader than Herbert’s version. It could include people who worshipped many gods, or spirits, or ancestors. It didn’t force everyone into a Christian-shaped box. Tylor actually wanted to defend non-European peoples from being dismissed as “primitive” or “superstitious.” He argued that they were using the same intelligence as anyone else to ask deep questions about life and death. They just didn’t have microscopes.

In the 1900s, the definition stretched even further. Some thinkers argued that religion doesn’t have to involve personal beings at all. It could also involve impersonal forces or orders — like the Dao in Daoism, the law of karma in Buddhism, or the “unseen order” of the universe that William James wrote about. By this definition, a Buddhist who doesn’t believe in any gods at all could still be said to have a religion, because their practice is about aligning themselves with the way reality fundamentally works.

So the history of the concept looks like three expanding circles: first, you need a supreme God; then, you can have many gods or spirits; then, you don’t even need a person-like being — just some kind of ultimate reality or order. Today, most people operate with this third, biggest circle. No one argues about whether Buddhism is a religion anymore. It just is. But the fact that the definition keeps expanding raises an uncomfortable question: is it expanding because we’re getting a better understanding of what religion really is, or is it expanding because we want to be nice and include everyone?

Different Ways to Define What “Counts”

Philosophers and anthropologists have two main strategies for defining religion. They’re called “substantive” definitions and “functional” definitions. They lead to very different results.

A substantive definition draws the line based on what people believe in. Tylor’s “belief in spiritual beings” is a substantive definition. So is Herbert’s list. The idea is: religion is about a special kind of reality — gods, spirits, an ultimate order, something that goes beyond everyday experience. If a practice doesn’t involve that kind of belief, it’s not a religion, no matter how important it is in people’s lives.

A functional definition draws the line based on what role a practice plays in people’s lives. If something serves as your ultimate concern — whatever organizes your values, gives your life meaning, and creates a community around you — then it functions as a religion, even if it doesn’t involve any gods or spirits. The sociologist Émile Durkheim defined religion as whatever system of practices unites people into a single moral community. By this definition, nationalism can be a religion. So can capitalism. So can devotion to a sports team or a political movement. So can being a superfan of a celebrity or a video game.

This is where things get really interesting. A functional definition can include things that a substantive definition excludes, and vice versa. An atheist who has no belief in any supernatural being can still have an “ultimate concern” that structures their entire life — say, fighting for justice or building a successful career. By Tillich’s definition, that atheist has a religion. But by Tylor’s definition, they don’t.

On the other hand, a substantive definition can include things that a functional definition excludes. An ancient Roman who sacrifices a bird to the gods before a journey might, for Durkheim, be practicing magic rather than religion, because the ritual is individual and doesn’t build a community. The same action could be called “religious” by one definition and “not religious” by another.

Which one is right? There’s no answer. These definitions are tools. You pick the one that helps you study whatever you want to study. But the fact that different tools produce different results means that “is this a religion?” is never a simple factual question. It depends on what you mean by the word.

But Maybe “Religion” Doesn’t Exist at All

Here’s where it gets even stranger. Some scholars argue that the whole concept of “religion” is a modern European invention — a tool that was created during colonialism to sort and control other cultures — and that it doesn’t actually correspond to anything real in the world.

The argument goes like this: Before modern times, most cultures didn’t have a separate sphere called “religion.” Their beliefs about gods and spirits were woven into everything: politics, family, work, art, entertainment. There was no word for “religion” because there was no thing to separate out from the rest of life. The ancient Romans didn’t have a religion; they just did things that involved the gods, the same way they did things that involved their families or the state. It wasn’t a separate category.

And when Europeans started using the word “religion” to describe what other people did, they were forcing other cultures into a box that didn’t fit. They were saying, “Your practices around spirits are basically the same kind of thing as our Christianity” — which might not be true at all. Worse, they were using the concept to rank societies: some had “true religion,” others had “superstition,” and some had “no religion at all” (which was a justification for conquering them and “civilizing” them).

This critique has led some historians and anthropologists to argue that we should stop using the word “religion” altogether. Timothy Fitzgerald says the word “picks out nothing and clarifies nothing.” It’s a fiction, like the idea of biological races. If we keep using it, we’re just perpetuating illusions that were invented for colonial purposes.

Other scholars say this goes too far. Yes, the concept was invented by particular people in a particular time for particular purposes. But that doesn’t mean it can’t be a useful tool for comparing different cultures. Lots of concepts are invented — that doesn’t make them false. The concept of “democracy” was invented by ancient Greeks, but we can still use it to describe systems of government in other times and places, even if those systems don’t call themselves “democracy.”

And there’s evidence that people were developing concepts like “religion” long before European colonialism. The Roman historian Josephus, writing in the first century CE, distinguished between different “ways of worship” among Jews, Greeks, and Egyptians. Medieval Chinese writers developed a term (jiao) that grouped together Buddhism, Daoism, and Confucianism as comparable traditions. So the idea of “types of religion” might not be quite as modern and Western as the critics claim.

So What Is Religion?

After all this, you might expect a tidy answer. But there isn’t one. Philosophers still disagree. What we have instead is a bunch of questions:

Should we define religion by what people believe in (gods, spirits, ultimate reality), or by what role it plays in their lives (giving meaning, creating community)?
Should the definition be narrow (only theistic traditions) or broad (anything that serves as an ultimate concern)?
Is “religion” a real thing that exists in the world, or a category we invented that distorts reality?
Can we use the word to describe ancient cultures that didn’t have the concept, or is that a kind of imperialism?

The way you answer these questions isn’t just an academic exercise. It affects real things: laws about religious freedom, debates about whether secular worldviews like atheism or nationalism should get the same protections as traditional religions, and how we understand people from different cultures.

Maybe the most honest thing to say is: “religion” is a word we use to talk about a set of human practices that involve things like gods, spirits, ultimate realities, rituals, communities, and deep meaning — but nobody agrees exactly where the boundaries are, and maybe they can’t be drawn neatly. The concept is like a net we throw over human life, trying to catch something important. But the net has holes, and what we catch depends on how we weave it.

Appendices

Key Terms

Term	What It Does in This Debate
Substantive definition	Defines religion by what people believe in (gods, spirits, ultimate reality)
Functional definition	Defines religion by what role a practice plays (giving meaning, creating community)
Monothetic definition	Says every religion must have ONE essential feature (or one fixed set)
Polythetic definition	Says religions share a “family resemblance” with no single feature required in all
Ultimate concern	Whatever matters most to a person and organizes their values (Tillich’s functional criterion)
Superempirical	A fancy word for “beyond what we can detect with our senses” — used to describe what religions refer to

Key People

Edward Herbert (1583–1648) — A British thinker who tried to find the five beliefs shared by all religions, basically using his own Christianity as the model.
Edward Tylor (1832–1917) — An anthropologist who defined religion minimally as “belief in spiritual beings,” partly to defend non-European peoples from being called primitive.
Émile Durkheim (1858–1917) — A sociologist who defined religion functionally as whatever unites people into a moral community, even if it involves no gods.
Paul Tillich (1886–1965) — A theologian who defined religion as whatever serves as your “ultimate concern,” meaning your deepest organizing value.
J.Z. Smith (1938–2017) — A scholar of religion who famously said “there is no data for religion” — meaning the category is created by scholars, not found in nature — but he still thought it was useful.
Timothy Fitzgerald — A contemporary critic who argues that “religion” picks out nothing real and should be abandoned as an analytical concept.

Things to Think About

Is being a fan of something (a band, a sports team, a video game) enough like what we call “religion” to be in the same category? If not, what’s the crucial difference?
If someone says “I don’t have a religion,” but they care deeply about justice and spend every weekend volunteering for a cause, does that mean they do have a religion by a functional definition? Should we tell them that?
The critics say the concept of “religion” was invented by Europeans to control other cultures. But the defenders say it can still be a useful tool for comparison. Both sides can’t be entirely right. Where do you land?
If a culture has no word for “religion” and doesn’t separate religious practices from other parts of life, does it still have a religion? Or does the category only exist once you name it?

Where This Shows Up

Legal cases: Courts have to decide what counts as a “religion” for purposes of religious freedom protections. Should a new belief system or a secular philosophy get the same legal treatment as Christianity or Islam?
School debates: Should schools teach “yoga” as exercise, or is it a religious practice? Should mindfulness meditation be allowed in class, or is it secretly Buddhist?
Political arguments: People sometimes call political movements like nationalism or Marxism “religions” to criticize them. Is that fair, or is it just name-calling?
Your own life: When you say someone is “obsessed” with something, or that a hobby is their “religion,” you’re actually doing what functional definitions do — treating whatever matters most to a person as a kind of religion.