What Was Religion Doing All This Time?

Imagine two people lose a parent.

One belongs to a religious community.

People arrive with food. Someone organizes the funeral. Familiar words are available when original words are impossible. The community visits, prays, sits quietly and notices months later that grief has not obeyed the administrative schedule.

The other person receives compassionate-leave paperwork, several heart emojis and a recommendation for a mindfulness app.

The contrast is unfair.

Secular friends can be extraordinary. Religious communities can be absent, judgmental or cruel. An app may genuinely help, while a casserole has known limitations.

Still, the scene points to a question.

Modern societies have become skilled at asking whether religion is true.

Have we become less skilled at noticing everything religion was doing?

“Religion” is not one thing

The word covers different doctrines, institutions, histories and experiences.

It can mean private belief, communal worship, inherited identity, ritual practice, political authority, mystical experience or the reason two relatives stop speaking after a wedding.

Any sentence beginning “Religion causes…” should therefore produce immediate caution.

I am not trying to prove a doctrine or construct one universal account.

I want to look at religion as a bundle of social technologies:

• meaning;
• community;
• moral formation;
• ritual;
• obligation;
• memory;
• authority;
• and coordination.

The bundle can support people.

It can control them.

The uncomfortable possibility is that the benefits and dangers emerge from the same machinery.

Religion makes community recurrent

Many modern relationships are selected through compatibility.

We choose friends who share interests, age, education or a willingness to discuss the same television programme with unreasonable seriousness.

A congregation can be different.

It creates repeated contact among people who may not otherwise choose one another. It gives them roles, obligations, shared time and a reason to notice absence.

That recurrence matters.

Prospective studies have found associations between religious-service attendance and outcomes including lower depression and lower mortality in some populations. One large study followed US nurses over time and found frequent attendance was associated with lower subsequent depression risk; another found an association with lower mortality. These are observational findings from a particular population, not randomized evidence that attendance causes health. Selection, social support, health behaviour and reverse causation remain difficult to eliminate. (Li et al., 2016; Li et al., 2016)

Newer causal methods add useful restraint. A 2026 target-trial emulation using a large New Zealand panel found selective rather than universal effects of attendance across wellbeing measures, with gains concentrated in some coordination-related domains. (Bulbulia et al., 2026)

The honest summary is not:

Religion makes people healthy.

It is:

Recurrent communal participation may provide forms of support and coordination worth understanding, but attendance is not a clean treatment and the effects are not uniform.

Religion asks who you should become when nobody is watching

Law regulates a narrow band of behaviour.

It can punish theft.

It has limited capacity to make someone generous, faithful, forgiving or honest in a room with no witnesses.

Religious traditions often extend moral life into private conduct:

What kind of person should I become?

What do I owe a stranger?

What does restraint look like when violation is profitable?

Belief in supernatural monitoring or punishment may support cooperation in some contexts. A cross-cultural behavioural study across eight societies found that belief in moralistic, knowledgeable and punitive gods predicted more impartial behaviour toward distant co-religionists in economic games. (Purzycki et al., 2016)

Again, the conclusion should remain narrow.

The study does not show that religion invented morality, that non-believers lack it or that cooperation extends equally to outsiders.

It suggests one way sacred accountability may help widen a circle of cooperation.

It can also draw the circle sharply.

Religion coordinates life before the form arrives

Religious institutions have often participated in marriage, education, charity, burial, dispute resolution, family obligation and political legitimacy.

That history is neither uniformly benevolent nor easily summarized across civilizations.

The functional point is simpler: a religious system can combine expectations that modern societies often distribute among separate organizations.

It can tell a person:

• what this suffering means;
• who will show up;
• what duties follow;
• which ritual marks the transition;
• what the community remembers;
• and who has authority to interpret all of it.

Integration is powerful.

The bereaved person does not have to assemble a grief infrastructure from six providers while grieving.

The structure arrives before the need.

Meaning is not merely emotional first aid

Religion places death, sacrifice, injustice and hope inside a larger narrative.

That can make suffering bearable.

It can also make avoidable suffering seem necessary.

“There is meaning here” can support endurance.

It can become:

Do not question the condition causing the pain.

This is the control problem.

The same authority that stabilizes a moral community can prevent reform.

The same ritual that preserves memory can enforce conformity.

The same dependence that ensures nobody grieves alone can make departure socially catastrophic.

The same sacred boundary that protects a group can exclude, shame or punish.

Religion’s failures are not an unrelated bug attached to its power.

Some may be the shadow of the same coordination mechanisms.

Modernity unbundled the functions

Government supplies law.

Therapy supplies structured emotional support.

Schools supply education.

Charities coordinate service.

Clubs provide community.

Philosophy offers meaning.

Apps provide meditation, reminders and—if notifications are enabled—minor eschatology.

This unbundling creates freedom.

People can reject a doctrine without losing every relationship, service and explanation at once. Competing institutions limit concentrated authority.

But the replacements may be fragile because they are elective, specialized and age-segregated.

A running club may provide friendship but not burial.

Therapy can help interpret suffering but does not ordinarily organize neighbours to deliver food.

A philosophy can offer meaning but may not notice when you stop attending.

The missing ingredient may be obligation.

Modern institutions are good at serving preferences.

Belonging sometimes requires people to remain present after preference has left the room.

The secular imitation problem

Could we preserve the support without the doctrine and the belonging without the control?

Perhaps.

But any serious replacement would need to answer practical questions:

• How does membership become recurrent?
• What do members owe one another?
• Which rituals survive more than one enthusiastic founder?
• How are children and elders included?
• What limits authority?
• What happens when a member becomes inconvenient?
• Which values are non-negotiable?

Once a community has shared ritual, moral expectation, intergenerational memory, authority and things treated as sacred, it may begin to resemble the category it was designed to replace.

That does not prove religion is inevitable.

It suggests that its functions are harder to reproduce than its critics sometimes assume.

What I still cannot figure out

Individual freedom protects people from coercive communities.

Durable belonging requires claims on the individual that preference alone may not sustain.

Can a society preserve freedom, belonging and moral responsibility without placing anything beyond individual preference—or does every stable society eventually make something sacred?

Until the next strange question,

Osagie

My current hunch: Religion’s durability comes partly from bundling meaning, obligation and recurrent community in a way modern substitutes rarely match.

Most likely reason I am wrong: This functional account may romanticize religious institutions, understate secular forms of durable care and treat harms caused by coercion as an acceptable cost rather than a central verdict.