What Religion Was Actually Doing, and What Replaced It

Start with a fact that reverses the assumption most readers of this chamber hold.

Religion is not declining globally. It is redistributing. Projections from the major demographic work on this find the religiously unaffiliated shrinking as a share of the world population over the coming decades, not growing — because secularisation is concentrated in rich, low-fertility societies, and the fastest-growing populations are in places where religious affiliation is near-universal.

So the story a reader in Tokyo or Copenhagen experiences as a one-way global process — the rules loosening, belief receding, everyone arriving eventually at the same place — is a local weather report mistaken for a climate. Where you are standing, the trend is real. Worldwide, it is not the trend.

That matters for this chamber because it removes a temptation. If secularisation were inevitable and universal, the religious societies described here would be earlier versions of the secular ones and there would be nothing to learn from them, only something to wait out. They are not earlier versions. They are different settlements, and several of them are growing.

Religious change in the last two decades has been faster and stranger than the simple secularisation story allows, and four movements are worth naming because each does something different to intimate life.

Latin America is converting. A historically overwhelming Catholic majority has been losing share to evangelical and Pentecostal churches for a generation, at a speed that has reorganised the politics of several countries. The Brazil row in this chamber’s matrix exists for this reason, and the lazy assumption — Catholic to evangelical means liberal to conservative — misreads both, since the new churches are frequently stricter on personal conduct and more socially mobile in who they admit to authority.

The rich democracies are disaffiliating. The share describing themselves as of no religion has risen sharply across Western Europe, North America and Australasia, driven mostly by generational replacement rather than by individuals changing their minds.

Iran is the sharpest and most contested case. Large online surveys have reported very high rates of disaffiliation from Shia Islam among Iranians — figures that would be astonishing if they generalise. The methodology is non-probability and self-selecting, in a country where answering honestly to a state-run survey is not safe, so the true number is unknowable in either direction. What can be said is narrower and still significant: state enforcement of religion appears, from everything observable, to have produced private withdrawal from it rather than the loyalty it was intended to secure.

And the Gulf states are running the reverse experiment: rapid liberalisation of what is permitted in public life — entertainment, women driving, foreign residents’ family law — with the religious basis of governance unchanged. The United Arab Emirates row in the matrix is about exactly this two-track arrangement.

Four directions at once. Which is why “religion is declining” is not a description of the world; it is a description of one quadrant of it.

Now the argument this essay exists for, and it runs against the expectation.

The secular account of religion’s role in intimate life is that it supplied rules — about sex before marriage, about contraception, about divorce, about who may marry whom — and that removing those rules increased freedom. That account is true. It is also the smallest of the four things religion was doing, and the only one anybody has replaced.

The second was a script. Not a rule about what is forbidden, but a shared answer to what happens next: at what age, in what order, marked how. A person did not have to invent the shape of a life, because the shape was supplied, and the supply was free.

The third was a community that noticed. Somewhere that registered your absence. People who turned up when somebody died, without being asked and without a group chat having to be organised, because turning up was what the structure did.

The fourth was a ceremony at thresholds. Birth, adulthood, marriage, illness, death. Occasions where a person is handed words to say when they have none of their own, and a room full of people is told what to do with their hands.

Rules were removed. Scripts, communities and ceremonies were not replaced. They were simply subtracted, and the subtraction is what a great deal of modern intimate difficulty is actually made of.

Consider what the loss of a shared script actually costs, because it is usually described as a gain and it is both.

A woman in a society with a script knows what is expected of her at thirty. She may hate it, she may defy it, she may pay for defying it — but she is oriented. She knows what she is agreeing to and what she is refusing.

A woman in a society without one has to invent the answer to every threshold question herself, repeatedly, without models, and while being told that the inventing is freedom. When to partner. Whether to. Whether the thing she has is a relationship. What she owes an ageing parent. Whether wanting a child at thirty-eight is a plan or a panic. There is no longer a default answer, which means there is also no longer a default answer to argue with, and arguing with a default is much easier than generating one from nothing.

This chamber has described the consequence from four other directions without naming the cause. The expectation-gap essay found people evaluating each other indefinitely with no shared sense of what a good-enough answer looks like. The arrangements essays found couples inside agreements nobody had ever stated. The mobility essay found people released from local expectation and then unable to say what they actually wanted instead.

All of those are downstream of the same subtraction: the script went, and the work it was doing did not.

And that work is unevenly distributed, which is the part that should offend a reader who cares about fairness. Inventing a life from first principles requires time, models, money and confidence. The people with the most of those do it best. The freedom is real, and it is priced.

There is a body of research that is inconvenient for the account most readers of this chamber hold, and it should be stated rather than left out.

Across a number of studies in several countries, people who attend religious services regularly report higher levels of social support and, on several measures, lower loneliness than those who do not. The effect appears to attach to attendance — to showing up in a room repeatedly — more than to belief.

The obvious caveats are real and must be stated. The research is correlational. People who attend may differ systematically from those who do not. The effect sizes vary and the measures are crude, and the same communities can be actively harmful to people they exclude, which the essays on Spain, the Philippines and Korea in this chamber have each documented.

But the narrow claim survives all of that, and it is the interesting one: repeated, unchosen, low-intensity contact with the same people over years does something that a secular life does not automatically provide. Not because secular people are worse at friendship. Because the structure that used to put them in a room together every week was removed, and nothing was built in its place with the same properties — obligatory, recurring, cross-generational, and not requiring anybody to organise it.

The Denmark essay found young people lonely in the society that solved almost everything solvable by policy. This is one of the things policy did not solve, because policy does not build the kind of structure that puts you next to the same person for twenty years without either of you having chosen it.

Now the mirror, and Japan is not the laggard in this one either.

A large share of Japanese people describe themselves as having no religion. The same people visit a shrine at New Year, take a child at three and seven, marry in a chapel with a hired officiant, and are buried by a Buddhist temple. Government figures on religious affiliation, compiled from what organisations themselves report, have long summed to substantially more than the population of the country.

That is not hypocrisy and it is not confusion. It is a different configuration, and it is worth naming precisely: Japan kept the ceremonies and the thresholds while letting doctrine and exclusive affiliation go. Ritual without creed.

Which means Japan did not arrive late at secularisation. It arrived early at something the West is now drifting toward — people who do not believe, do not affiliate, and still want a room, a form of words and somebody official at the moments that matter.

And it means the Japanese loss is narrower than the Western one. The ceremony survived here. The script largely did not: the shape of a woman’s life that was supplied by the post-war settlement — marry by a certain age, leave work, raise children, care for parents — has weakened as an instruction while remaining fully alive as a judgement, which is arguably the worst available combination. Enough script left to be measured against, not enough to be oriented by.

We are not arguing for religion and we are not arguing against it. This house set that rule for the chamber before the first essay and it has held: a chamber that became a sermon in either direction would be useless to a reader who is actually religious and dishonest to one who is not.

What we think is true is narrower and, we hope, usable by both.

The rules were the least of it. Anybody who thinks the religious inheritance in their life is mainly a set of prohibitions about sex has not looked at what else was in the package, and is likely to be surprised by what goes missing when they discard the whole thing.

What actually goes missing is orientation, witness and ceremony. And of those three, the one this Library can do anything about is the first: a shared vocabulary for the questions that used to have supplied answers. That is not a small thing — the Spain essay showed a population acquiring a sentence in three years — but it is one third of what was lost, and we would rather say so than pretend an essay is a congregation.

The other two are harder and mostly not ours. Witness requires showing up in the same room over years, which nobody can arrange for you and which the modern arrangement of work and cities actively obstructs. Ceremony requires a form of words at thresholds, and many people improvise these now, badly, alone, at the exact moments when improvising is hardest.

If you are religious, none of the above is an argument against your practice, and a fair reading of this essay is that your tradition is carrying more of your life than the secular account credits. If you are not, the honest position is not that you have escaped something. It is that you have taken on a load that used to be carried collectively, and that noticing this is more useful than either celebrating or mourning it.

You do not owe tradition obedience. You do not owe modernity performance. You are allowed to discover what actually fits your life.

Three, and the first is the one this essay is most likely to be misused for.

This is not an argument that people should return to religion for the social benefits. Attending a service you do not believe in, to obtain the support of a community you are deceiving, is a worse arrangement than loneliness and is also unlikely to work, since the benefit appears to depend on the thing being real to the people around you.

Nor is it an argument that secularisation was a mistake. The rules that were removed were removed for good reasons, and the essays in this chamber on Spain, the Philippines and Korea document in detail what those rules cost women. Nothing here should be read as wishing them back.

And the global-redistribution fact should not be imported as a prediction. That the unaffiliated are projected to shrink as a share of the world says almost nothing about what will happen in Japan, and projections over decades are estimates with wide error, not a schedule.

What travels is the inventory: rules, script, witness, ceremony. It is worth applying to your own life honestly, in either direction, because most people discover they are still using more of the package than they thought — or missing more of it than they had language for.

The projection that the religiously unaffiliated will shrink as a share of world population comes from major demographic projection work and rests on fertility differentials and age structure. Projections decades out carry wide uncertainty and depend heavily on assumptions about switching and fertility convergence; it is a modelled estimate and this essay treats it as one.

The four regional movements — Latin American evangelical growth, rich-democracy disaffiliation, Iranian disaffiliation, and Gulf liberalisation under unchanged religious governance — are described in outline from survey and reporting work of very different quality. The Iranian figures in particular come from large non-probability online surveys in a country where honest answers to official surveys are unsafe, so the true level is unknowable in either direction and the essay says so in the text rather than only here.

The research associating religious service attendance with higher social support and lower loneliness is correlational, from several countries and varying measures, with real selection effects: people who attend differ from those who do not in ways that are hard to control for. The narrow claim the essay makes — that repeated unchosen contact does something a secular life does not automatically supply — is an interpretation of that literature, not a demonstrated causal effect.

Japanese religious-affiliation figures compiled from organisational self-reports exceeding the national population is a long-standing feature of the government statistical series and reflects overlapping counting rather than anything about individuals. The characterisation of Japanese practice as ritual without creed is a widely held reading and is still a characterisation.

The four-part inventory of what religion supplied — rules, script, witness, ceremony — is this essay’s own framing rather than a finding from the sociology of religion, which uses different and more careful categories. It is offered because it is useful, not because it is established.

And the whole essay speaks about societies. It says nothing about whether any particular belief is true, which is not a question this chamber is equipped to take up or has any business taking up.

The question worth leaving a reader with is not whether to believe anything.

It is which of the four things you are currently without, because they go missing separately and most people have not audited it. Many readers of this chamber have no rules they did not choose, which is an achievement. Fewer have a script, and the ones who do usually inherited it as a judgement rather than as an orientation. Very few have witness — the people who would notice an absence without being told. And almost nobody has ceremony: a form of words for the thresholds, which now arrive unmarked and have to be crossed without anybody saying anything.

This Library can help with one of those and is honest about which. What we can supply is language: a way of naming a question that used to come with its answer attached.

And the other thing we can supply is smaller and more specific than a congregation. An evening in which somebody is paying attention, in which the terms are yours and set in advance, in which saying what you want is the activity rather than the risk, and in which nothing is inferred from your silence. Not a ceremony and not a community. A few hours of being accompanied, which is the one part of the package that can be arranged for a Tuesday.