Divine Dictatorship: The Path Towards All Answers

Divine Dictatorship

Barely anyone that walks across a bridge thinks of it as placing a bet. Strictly speaking, though, it is one: we’re betting it doesn’t break. In the same way, trusting a friend means betting on their track record, and taking a medication means betting on the trials that supported its benefit. The same logic extends to every belief we hold, because each one amounts to a bet about how the world works, only most of these bets are backed so well that we have stopped experiencing them as bets at all.

Faith, then, involves a wager like every other belief, and what sets it apart is how it arrives at the odds. Instead of letting evidence decide, it hands the decision to an authority nobody may question, much as a dictatorship hands every important decision to a ruler who is dangerous to question. The comparison may sound harsh, yet faith’s defenders commonly draw it themselves, at least implicitly, whenever they argue that a framework which makes people kinder needs no consistency in its checks.

We don’t need faith to believe what the evidence already supports. Nobody requires it to expect the sun to rise tomorrow or to trust that a dropped glass will fall, and believing that a well-tested medication will usually work calls for it just as little, as observation is enough for all of these. Faith only becomes necessary in the space between what the evidence supports and what a person wishes to hold as certain, and what it does within that space is remarkably specific: it reweighs the odds.

Whenever a believer weighs the probable against the improbable, faith slips a hidden variable into the comparison, one that cannot be disproven. Once that variable enters the calculation, the ordinary rules of likelihood become subordinate to it. An outcome that would be wildly unlikely by every observable measure may turn entirely probable, provided a higher power is orchestrating it. In this sense, faith is the act of looking at poor odds and deciding that an invisible presence has changed them. That is also what makes it so hard to argue with, since any evidence about the odds can be answered by pointing to the variable that overrides them, while the variable itself can never be measured.

Most of our daily reasoning rests on a principle so basic that we rarely bother to name it, which is that the obligation to provide evidence lies with whoever makes a claim and not with whoever doubts it. Say someone insists that, at this very moment, invisible cows with tiny wings are hovering in the stratosphere. You’d dismiss it without a second thought and feel no duty to disprove it, because it collides with everything we know, and nobody would expect you to launch a weather balloon before declining to believe in flying cattle.

Yet, the more abstract a claim becomes, the more willing we tend to be to suspend this principle, especially when the claim fills a gap in our understanding. Unlike the winged cow, a metaphysical claim steps outside the realm where observable reality could contradict it, and that is exactly where its appeal comes from. It offers certainty in a chaotic world and can give the believer the flattering sense of possessing knowledge about absolute truths that others fail to grasp. (Detailed explanation: As explained in Converging Conversation, conspiracy beliefs seem driven by needs for understanding and control (Douglas et al., 2017) and tend to go with feeling one holds rare knowledge others lack (Lantian et al., 2017).)

How much we depend on the burden of proof becomes obvious once we picture a world that reversed it. Say a neighbor accuses you of a brutal murder without a shred of evidence, and when asked where the body is, he explains that the victim was an unregistered person whose remains you dissolved completely. In that world, you are presumed guilty unless you can produce a verified, second-by-second record of your entire life. Even then, any gap in it, such as the minutes you spent in the shower, could be turned against you, since the accuser might claim you have secret abilities that let you commit the crime unseen. Had the shower been recorded as well, he could simply insist that you used a clone.

Since nobody can prove a negative against an imagination unconstrained by evidence, a system that accepts claims without proof either dissolves into chaos or hands absolute power to whoever gets to decide which accusations count. After all, the presumption of innocence, which protects every one of us in court, is simply the burden of proof applied to our freedom.

Not everyone accepts this standard for every belief. Some argue that belief in God can be reasonable without any argument, much as we trust our memory or assume that other people have minds. However, the trust we place in memory and in other minds is shared by every human being and corrected every day, as others confirm or dispute what we remember. A conviction that nobody else can correct therefore has a weaker claim to the same trust. Nor is the burden of proof a dogma of its own, since it is a tenet we keep because decisions made on it fail less often than decisions made without it. That record is open for anyone to check, and should it ever turn, the tenet ought to be dropped.

Still, the burden of proof doesn’t let us declare eternally that no higher power exists. Since almost no claim can be ruled out completely, what matters is whether our conviction is proportional to the evidence, a proportion faith tends to reverse by granting certainty to the claims with the least support. We might also comfortably suspend our critical ways whenever a claim carries the weight of tradition or the agreement of many people around us, even though neither has made the claim itself any more likely. In doing so, we raise subjective experiences nobody can verify (whether they arose in clarity, in a dream, or under the influence of a substance) above observations that others can check, much as we place anecdotes above patterns and comforting narratives above regularities.

Many believers consider their conviction proportional all the same and point to evidence of a kind, such as physical constants that seem finely tuned for life, the question of why anything exists at all, or a personal experience of the divine. (Detailed explanation: Physicists still debate how much of this apparent fine-tuning holds up: some argue that the values permitting life fall within extremely narrow ranges (Barnes, 2012), while others find that some constants could vary by orders of magnitude and still allow stars and habitable planets (Adams, 2019). Proposed explanations range from deeper laws not yet discovered to a vast number of universes among which ours would be one of the few able to host observers.) Even granted in full, these arguments only reach so far. A finely tuned universe, if that is what we live in, says nothing about which deity tuned it or whether that deity hears prayers, let alone which scripture speaks for it. Answering why anything exists with a creator merely moves the question one step back without settling it, unless the creator is declared exempt from it.

Experiences of the divine, however powerful, can’t be checked by anyone else, and they point toward as many different gods as there are traditions. At most, then, such arguments support a distant first cause, which still lies a long way from the detailed doctrines most faiths ask their members to hold.

This leaves us with a curious paradox. Logically, a belief that nothing could ever confirm should be held no more firmly than its other reasons allow, yet in practice the least falsifiable beliefs are often the ones held with the greatest conviction. Because no observation can shake them, their holders may come to feel that nothing ever will. The spectrum ends up turned upside down, with the claims we could verify most easily treated as provisional and the claims we could never verify treated as settled. (Detailed explanation: When religious participants read that science might one day settle whether God exists, they reported weaker conviction than when told the question lay beyond science, and when a scientific discovery was framed as a threat to their faith, they rated untestable reasons for it, such as morality, as more important. Supporters and opponents of same-sex marriage behaved alike, treating the issue as a matter of fact when data favored their side and as a matter of opinion when it did not (Friesen et al., 2015).) Proportionality cuts the other way as well, which keeps it from hardening into a dogma of its own. Unusual events do happen, and dismissing every anomaly would be just as inaccurate as treating every anomaly as proof, so what matters is how much weight we give them and what we derive from them.

If someone survived a lightning strike, that hardly makes it reasonable for you to stroll through a thunderstorm holding a metal pole, hoping to be just as lucky. Yet, the mere possibility of lightning shouldn’t paralyze you either. If the storm has long since passed and the thunder has been silent for half an hour, letting such a remote chance stop you from helping a friend in need would be questionable, to say the least. (Detailed explanation: Lightning can strike as far as 10 miles (16 km) from the rain, so the US National Weather Service warns that anyone who can hear thunder, even a distant rumble, is within reach, and advises waiting 30 minutes after the last thunder. About nine in ten people struck by lightning survive, though many with lasting injuries (National Weather Service, 2018).) Since our time and energy are limited, it is wiser to prepare for what is probable than to pray for what is improbable, just as it would be unwise to let the improbable dictate our fears.

That same balance applies to personal experiences that seem to defy explanation, whether a striking coincidence or a moment that feels like a sign. However meaningful they feel, you ought to treat such experiences the way you would treat a weather forecast that predicts an unlikely storm, as a signal worth noticing and perhaps worth preparing for. A forecast like that prompts you to pack an umbrella, while nobody would take it as a reason to abandon meteorology. If you read unusual experiences through the lens of preparedness instead of conviction, you can stay alert without surrendering your judgment, at least as long as the degree of preparation stays proportional to the cost of being wrong.

The important part of all of this is that we tend to not be aware of all the consequences seemingly trivial things can have. After all, if everything is a wager, why not believe in God, since if God exists the believer gains salvation, and if not, they lose comparatively little? In its everyday form, the same logic holds that a small ritual, whether a prayer or knocking on wood, costs only a moment, while the risk of ignoring it (eternal punishment or catastrophic misfortune) seems enormous, so why not play it safe?

The reasoning looks prudent until we try to apply it consistently. If every claim that promises reward or threatens punishment called for the same precaution, we couldn’t leave the house without a rabbit’s foot in our pocket and salt thrown over our shoulder. We would also have to avoid every crack in the pavement on the way while reciting protective words for thousands of mutually contradictory gods and spirits from every culture in history. Taken to its conclusion, we would have to wear a crash helmet on a daily stroll to protect against falling flowerpots or debris dropped from a passing plane, much like how many people drive to work every day without hesitation, yet feel terrified of flying, despite planes being statistically safer. The wager only seems acceptable because we apply it to a single tradition, usually the one we grew up with, and ignore all the others without a second thought.

There’s a deeper problem underneath the wager, too. A gambler in a casino at least knows that the roulette wheel exists and (hopefully) what the odds are, whereas with unverifiable claims we are betting on a game without knowing whether the wheel is real or what its rules might be. One interpretation promises reward for devotion and another condemns the same devotion as misguided. Under such conditions, taking part carries no calculable advantage over abstaining, since participation might be exactly what a different, equally unverifiable authority would punish.

If both paths offer the same unknowable chance of being right, the rational choice is the one that costs less when it turns out to be wrong. Participation demands time and attention, if not way more, whereas abstention demands far less.

Our everyday behavior shows how inconsistently we apply our fears. We walk down streets without body armor and sleep in homes that aren’t fortified against intruders, and we eat food prepared by strangers without ever testing it for poison. We accept these risks because we judge them small enough to live with, and, unlike supernatural threats, they exist within a reality where we can observe them and confirm whether they are present or absent.

A falling flowerpot or roof tile is a remote danger, yet it rests on gravity and loose masonry, mechanisms we can see and understand. Structuring an entire life around a threat for which no such mechanism can be observed, while shrugging off risks that are measurably real, therefore hardly counts as caution, as it misaligns the judgment we rely on everywhere else.

Faith also draws confidence from repetition. Whenever several traditions tell the same story or honor the same figure, their agreement feels like independent confirmation, as if the event or person must be historically settled and only the interpretation were left open to dispute. Agreement, however, only counts as evidence when the sources are independent of one another, because a rumor repeated by ten people who all heard it from the same person is still a single rumor. Science treats this distinction as essential, since a finding replicated by independent teams in different countries carries far more weight than the same finding cited over and over, as each independent attempt could have failed on its own.

The great flood is a good example. The story of a devastating deluge that a chosen few survive appears in Mesopotamian texts such as the Atrahasis epic and the Epic of Gilgamesh, both of which predate the biblical account, and the motif recurs in Genesis and later in the Qur’an. (Detailed explanation: Merely hearing a statement repeatedly makes it seem more true [illusory truth effect] (Dechêne et al., 2010). Independent replication is the opposite test, and, as noted in The Part-Time Religion of the World, only about a third of 100 published psychology studies produced significant results again when independent teams repeated them (Open Science Collaboration, 2015). The oldest surviving copy of the Babylonian Atrahasis epic, in which the gods send a flood that one man survives in a boat, was written down in the 17th century BCE, about a thousand years before the Genesis account took shape; a later version of the same story was worked into the eleventh tablet of the Epic of Gilgamesh (Lambert & Millard, 1969).) One plausible explanation is that the earliest civilizations grew up along rivers, where catastrophic floods were a recurring and terrifying reality, and that the story then traveled with trade, conquest, and migration, to be adapted by each culture that inherited it. Figures such as Abraham or the angel Gabriel likewise appear across the Jewish, Christian, and Islamic traditions, which grew from one another and therefore inherited many of the same characters. Their shared presence tells us a great deal about the lineage of these stories and very little about their historical accuracy. That does not prove the stories false, only that their agreement can’t count as evidence that they are true. Recurring motifs work in much the same way. Miraculous births and divine parentage appear across countless mythologies, as do journeys into death and back and heroes who sacrifice themselves for their people, often as a way of marking a figure as more than human. Such motifs likely spread because they work as stories, evoking awe and explaining suffering while promising that the world is ordered by meaning. Their persistence tells us what human beings need from a narrative while saying nothing about whether the narrative happened.

Demanding evidence for every conviction does, however, raise a serious question about what becomes of human rights and worth or justice. Not one of these can be measured in a laboratory, yet they form the foundations of what most of us deem the societies worth living in, so does consistency require us to discard them along with faith? I would argue it does not, because they are a different kind of claim. A statement that a particular being exists, that a historical event occurred, or that a ritual changes outcomes in the world is a claim about what exists. Even when it is placed beyond any test, it still asserts something about reality and therefore carries the burden of proof.

Human rights, as we understand them here, need to make no such assertion, because they do not claim to describe a hidden feature of the universe waiting to be discovered. They are agreements about how we choose to live together, derived through reasoning and shared experience. As explored earlier with the social contract and the conditioning of equal worth, their value lies in what they make possible. Equal worth is often taught as a taboo, and in childhood it may be held with something close to faith. Its reasons, however, can be stated and examined, even where institutions are wise not to reopen it case by case. We keep that wall standing because we can say why it stands, and asking about it is never (technically) forbidden.

Such agreements may not be measurable, yet they can be examined. We can ask what a principle protects and compare the lives of people in societies that uphold it with the lives of those in societies that don’t, refining its application as our understanding grows. The right to vote, once restricted by gender or skin color, was expanded through exactly this kind of examination and through the struggles it justified, as societies came to recognize that their earlier boundaries had served less utility or would be deemed inconsistent. Rejecting unfalsifiable claims about reality therefore does not require us to abandon ethics, only to be honest about what kind of claim an ethic is.

That same distinction exposes a common misunderstanding about which side of the debate can rightly be called open-minded. Accepting faith and spirituality is routinely presented as the more open stance, set against a supposedly narrow, materialist worldview. Yet, favoring the improbable over the probable means being open only to the least likely explanations while closing oneself off to the most realistic ones, and even that openness tends to be selective.

Believers who grant that other paths may lead somewhere still usually treat their own specific claims as true and those of others as partial at best, although both often rest on the same kind of evidence. Those who accept nearly everything, from crystals to past lives to each new spiritual trend, show in turn the absence of a filter or at least a more unreliable one, namely the bullshit radar that protects us from being exploited. (Detailed explanation: In a 2021 survey, 58 percent of American Christians said that many religions can lead to eternal life, while about three in ten said their own was the one true faith (Pew Research Center, 2021). People who rate randomly assembled buzzword sentences as profound [bullshit receptivity] also tend to report more religious and paranormal belief, more conspiracy belief, and more trust in alternative medicine (Pennycook et al., 2015).) Properly understood, open-mindedness is the willingness to change one’s mind when the evidence changes or a flaw in one’s reasoning has been sighted, applied evenly, including to the beliefs we hold most dear.

Faith also has less visible consequences, which commonly show up in how we approach our own ambitions. A popular view of self-help or self-improvement encourages people to believe in their dreams so completely that doubt itself becomes the enemy, and if the goal hasn’t materialized yet, the advice goes, one simply needs more tries, more patience, or stronger faith. Structurally, this is the reasoning of a gambler who keeps feeding a machine because the win must surely be close.

Real improvement relies on something far less romantic, namely realistic standards, which means measuring your progress against your own past instead of a fantasy and understanding why you advanced or stalled, so that the next attempt can be better informed. Chasing a far-fetched ideal that doesn’t fit our nature or circumstances, while refusing to read the feedback reality provides, tends to widen the gap between expectation and outcome until disappointment becomes chronic. (Detailed explanation: In small experiments, people asked to imagine a desired future as if it had already come true showed less energy than those who questioned that future or imagined it neutrally, and in one study they reported accomplishing less over the following week (Kappes & Oettingen, 2011). In earlier studies that followed people over time, positive fantasies predicted fewer job offers, weaker exam results, and slower recovery after hip surgery, while positive expectations based on experience predicted the opposite (Oettingen & Mayer, 2002). Unrealistic expectations about how fast, easy, and rewarding self-change will be feed a cycle of failure and renewed attempts [false hope syndrome] (Polivy & Herman, 2002).)

Ambitious goals aren’t a mistake in themselves, though. A distant aim can organize a life precisely because we know we may never reach it, and realistic optimism lets us lean on hope where its benefits are clear and its downside is limited. What matters is what the hope is allowed to do. Hope that listens to feedback can adjust its path and scale its expectations, growing stronger with every lesson, while faith in a goal that refuses to listen can only grow more insistent and hears less the louder it becomes.

Shortcut of Sin

Few of the tools religion has given humanity have proven as efficient as the concept of sin. Calling an action a sin, or a person a sinner, delivers a complete moral verdict in a single word. In everyday use at least, nobody has to understand the circumstances or weigh competing values, let alone consider how the situation looks from the other person’s perspective, because the label does that work for them. With very little mental effort, the label sorts complex human behavior into right and wrong, and it does so with the authority of a source nobody may question.

There were good reasons why such a shortcut once made practical sense. Yet, the simplicity that made a list of sins workable also made it blunt, because a rule that can’t see context will inevitably condemn some actions that would have been entirely reasonable given the circumstances or a different set of values. Applied this way, the label stops serving the mutual benefit it was meant to protect and ends up serving whoever controls its definition.

A list also creates loopholes, however broad its headings may be. If morality consists of avoiding the items written down, everything left unwritten becomes permissible by default. A person can then condemn the listed failings of others with great conviction while exploiting the gaps themselves, practicing greed, cruelty, or indifference in forms the list never anticipated and still considering themselves righteous. In this sense, sin is the original version of a checklist, refined over centuries.

The cost of this shortcut has grown along with the complexity of our world. In earlier eras, the mental effort of deep moral reflection may have been a burden few could afford, while the harms a list could miss were fewer and closer to home. Many of the harms we take part in most regularly appear on no ancient list, and the churches that have begun to name them as sins did so late.

Through our consumption we fund environmental destruction, and for the sake of convenience we support exploitative supply chains. We may vote for policies we know will burden our children, all while sleeping soundly, because we have committed none of the sins we were taught to fear. Judging by our consumer habits and the outrage we reward, most of us in wealthy countries have become sinners of comfort, harming one another in ways the old rules never imagined while keeping every rule they did write down. (Detailed explanation: Catholic teaching counts sins “in thought, word, deed, or omission” and treats a sin as mortal only when grave matter is committed with full knowledge and deliberate consent (Catholic Church, 1997). In 2015, Pope Francis quoted Patriarch Bartholomew’s words that to contaminate the earth’s waters, land, air, and life “these are sins” (Francis, 2015).)

Another consequence is less visible, yet perhaps more corrosive. Once morality is experienced as a set of requirements for salvation, it can acquire a ceiling, even though most traditions officially call believers to ever greater holiness. A believer who feels they have cleared the bar by avoiding the major sins, donating regularly, attending the required gatherings, and performing the prescribed duties may in practice find little reason to strive for more. Their goal shifts, perhaps unnoticed, from becoming as good as possible to becoming good enough.

Why, then, do we crave a moral authority in the first place? Part of the answer is that it spares us the unsettling task of deciding for ourselves what makes a person good, since an external authority hands us a fixed measuring stick that can be applied to anyone, ourselves included, and that stability and loss of responsibility are comforting. Yet, if you reflect honestly on how your own judgment of people has changed over the years, you will likely notice that your idea of a good person today looks quite different from the one you held as a child or teenager. You came to value qualities you once overlooked and to see through qualities you once admired.

The difference between a borrowed rule and an examined principle is easier to see with one of my own. Here it is: I help others without expecting a favor in return, as long as I feel that most of those I help appreciate it, because it is a standard I would accept being held to in their position. This principle integrates emotion instead of fighting it, because my willingness to help is tied to my own experience of appreciation, and it rests on reasoning I can state and examine. It also accommodates real limitations, such as depression or physical exhaustion, without breeding guilt, as it was built in the knowledge that my capacity fluctuates.

Compare this with a directive distilled to its core: help others because God expects it of you. On a good day the command may produce similar behavior, but it rests on external authority. Where it aligns with a person’s feelings and values they may experience harmony, and where it doesn’t, conformity is demanded regardless of their circumstances or state of mind. A thoughtful believer may well give a similar reason for helping. The difference only shows once the reason and the command come apart, since I can drop or adjust a rule whose reason no longer holds, while a command keeps its force either way, regardless of its long-term utility, not only for the person but also for others.

The more thoughtful version, help others because God knows it will bring good, doesn’t escape this problem either. This version turns the principle into a form of discipline whose value depends entirely on a promised outcome the believer can’t evaluate from their own perspective. Whenever the good fails to appear in their observable life, the reward can always be deferred to an afterlife, which keeps the directive safe from any test.

That same rigidity appears whenever a framework prescribes a single path as universally good. Many traditions extol parenthood as a blessing and a duty for every married couple, yet whether it serves a person well depends on their circumstances, such as their mental health and economic stability as well as their capacity to care for a child. Some find in parenthood the most fulfilling chapter of their lives, while for others it would harm both themselves and the child they bring into the world.

Those who choose not to become parents may still contribute enormously to the next generation, whether as teachers, mentors, researchers, innovators or caregivers, or simply as the aunt or uncle a child can always turn to. A universal prescription rooted in an unquestionable imperative is blind to all of these differences. It pushes people into choices that the very goal it claims to serve, the flourishing of families and children, would advise against. (Detailed explanation: Whether parents are happier than non-parents depends heavily on circumstances, such as age, marital status, and the age of the children (Nelson et al., 2014). Across 22 wealthy countries, the happiness gap between parents and non-parents shrank or vanished where paid leave and childcare made work and family easier to combine (Glass et al., 2016).)

History shows how rarely moral norms stay fixed for long. Marriage has meant a strategic alliance between families in one era and a partnership of personal choice in another, while gender roles have ranged from rigid separations of duty to far more fluid arrangements. Justice itself has moved from the logic of an eye for an eye toward principles of rehabilitation and restoration, and what one society considered unquestionable truth, another treated as negotiable custom.

Some anchors do appear nearly everywhere, such as helping family, returning favors, and respecting what belongs to others, but beyond these the common ground is likely narrower and more flexible than we tend to assume. (Detailed explanation: A study of ethnographic records from 60 societies found that seven cooperative behaviors (helping family, helping one’s group, returning favors, bravery, deference to superiors, dividing disputed resources fairly, and respecting others’ property) were judged morally good in all but one of 962 observations (Curry et al., 2019). Marriage, by contrast, was for most of history a way to forge alliances and pool labor and property, and the ideal of marrying for love took hold only in the last two centuries (Coontz, 2005).) A list of sins frozen at one point in history must therefore either be reinterpreted endlessly or be enforced against a reality it no longer fits. Think of the person who steals bread to keep a child from starving, where a rule applied without regard to its purpose ends up demanding exactly the suffering it was meant to prevent.

Belonging to a community of faith can also offer a subtler kind of moral relief. When a congregation runs a food bank or organizes relief after a disaster, its members can absorb a share of that goodness simply by belonging. In the mind of the individual, the institution’s good deeds become partly their own, even if their personal contribution amounts to little more than attendance. This is a kind of moral free-riding, a close relative of moral licensing, in which visible association with a good cause buys a sense of virtue without the cost of virtuous action. (Detailed explanation: When people learned that members of their own group had acted without prejudice, they felt freer to voice prejudiced views themselves [vicarious moral licensing] (Kouchaki, 2011). As noted in Vicious Virtuous, moral licensing in general appears smaller than first reported and may show up mainly when others are watching (Kuper & Bott, 2019; Rotella et al., 2026).)

Time adds its own weight, through the same investment loop that keeps a fan buying albums. Someone who has spent several hours every week for decades, even if it is only praying before going to bed, has invested a (substantial) part of their life and thoughts. The larger that investment grows, the harder it becomes for the ego to consider that some of it may have produced less than promised, or could even be seen as outright comical once transferred onto a custom that is less socially acceptable. A gathering that spends an hour describing the importance of compassion could spend part of that hour practicing it, explaining the values while living them. Many communities do exactly this, and their kitchens, shelters, and volunteer networks have earned recognition. Where the practice consists mainly of words, however, the time risks becoming a substitute for the action it describes.

To add a personal anecdote: my mother is involved in food sharing and, together with others, once organized a large event on the city’s main square. While we were setting everything up, hauling tables, heavy crates, pallets, and drinks, a group of Jehovah’s Witnesses arrived and set up their stall nearby. Since nobody but us was there yet, they had no one to talk to and spent the time chatting among themselves or looking at their phones, yet not once did they offer to lend a hand. The irony was hard to miss. It is, of course, a single anecdote, yet it fits a pattern I have noticed many times before: people who do a great deal for their church while ordinary good deeds seem to interest them very little unless the church’s name is attached, or believers who preach openness, acceptance, sharing and community and still turn out to be among the greediest and most prejudiced people I have met.

Petitionary prayer raises another question, since asking a higher power to heal a friend or change an outcome is, at its core, a request made in the hope that someone else will act. As long as it accompanies effort, it may provide comfort and focus, but once it replaces effort, it starts to resemble a complaint addressed upward, a statement of dissatisfaction that never turns into action and therefore does little to earn us the right to complain. If a benevolent deity existed and cared about the world, it would be hard, if not impossible, to imagine it preferring a thousand requests over a single act that improves someone’s life. (Detailed explanation: Drawing comfort from faith in hard times is linked to better adjustment, as noted in The Part-Time Religion of the World (Ano & Vasconcelles, 2005), but prayer by strangers did not help patients recover: in a trial with about 1,800 heart-bypass patients, complications were equally common with or without prayer, and somewhat more common among those told for certain that they were being prayed for (Benson et al., 2006), and a Cochrane review of ten trials with about 7,600 patients found no clear effect of intercessory prayer on death, clinical state, or readmission (Roberts et al., 2009).)

To be clear, my critique targets the framework and not (necessarily) the many believers who do real good. What I question is how much a good deed says about a person when its motive is the fear of hell or the hope of salvation, compared with the same deed done for reasons that person has self-inquired. The evidence suggests that the motive is often more mundane still, as much of this help seems to work less like a conclusion drawn from reflection and more like a contribution that secures one’s place in a community. (Detailed explanation: Across 53 countries, people who attend services often do more volunteer work, including for secular organizations, although in devout countries the gap between churchgoers and everyone else nearly disappears (Ruiter & De Graaf, 2006). In the United States, the difference seems to come less from belief than from belonging: how many close friends people have in their congregation predicts their generosity better than their theology does (Putnam & Campbell, 2010).)

Goodness that rests on an authority, without reasons the person understands, is only as reliable as the direction in which that authority happens to point. Under a different leader or a different interpretation, the same framework that moves one person to feed the hungry can move another to harm and destroy. A follower who never learned to examine the reasons has no internal compass with which to tell the difference.

Pandora’s Box of Fears

Rules backed by reasons can persuade. Rules without reasons need something else to hold them in place, and throughout history that something has very often been fear.

Hope and fear tend to arrive together. We usually cling to hope because we fear the alternative, and the more desperately we hope for an outcome, the more we dread its absence. Faith is commonly celebrated as a source of hope, and for many people it is one. Yet, we talk far less often about how much fear it can introduce along the way, fears that didn’t exist before the faith arrived and that the faith then offers to relieve.

Every worldview must deal with the unknown, and science does so within a structured method. Whenever it cannot yet explain something, it neither rushes to fill the gap with a definitive answer nor treats every conceivable explanation as equally likely. Instead, it narrows the space of possibilities by systematically ruling out what doesn’t hold up, until the remaining uncertainty becomes smaller and better defined. A framework that admits supernatural explanations without evidence lacks this filter. Once an invisible cause is allowed to explain one event, there is no way to keep invisible causes from explaining others that is neither arbitrary nor inconsistent, and the space of what might be true, or of what might be threatening, expands without limit.

On top of that, a framework that teaches certainty without evidence gradually blurs the line between reasonable doubt and irrational fear. Once inner conviction counts as sufficient reason to believe, external verification starts to seem secondary, and the weight of evidence can shrink exactly when it is needed most.

Faith can thus work like a vulnerability in the mind’s immune system. A single unfounded conviction may cause no visible harm on its own, much as a weakened immune system doesn’t make us ill by itself, yet it changes how well we can defend ourselves against whatever comes next. If it is socially accepted to hold firm conviction in a miraculous birth because an ancient text describes it, it becomes hard to find a consistent argument against firm conviction in a conspiracy theory because a viral video describes it. (Detailed explanation: In a Canadian national survey, stronger religious belief went with more paranormal belief, but more frequent church attendance went with less (Orenstein, 2002); a reanalysis found that the link between religious and paranormal belief held only among people who rarely attended services (McKinnon, 2003). Across studies, mostly from the United States, more religious people endorsed conspiracy theories slightly more often, but the link shrank considerably once political orientation was taken into account, and in German and Polish data religiosity was unrelated or even slightly negatively related to a general conspiracy mindset (Frenken et al., 2023). Short videos that teach people to recognize manipulation techniques make them better at spotting it (Roozenbeek et al., 2022).)

What both ask of us is to decouple truth from proof. The mechanism that lets us feel certain without evidence is the same one that populism and scams rely on. A habit of demanding reasons, applied consistently, therefore remains one of the few reliable protections against all of them.

Such convictions are rarely held in isolation either. Beliefs are contagious, and people who share an unfounded conviction tend to validate one another, especially when nobody around them disagrees, each person’s certainty reinforcing the others’ until the belief feels far more solid than any of them could justify alone. As in any echo chamber, the comfort of agreement can easily stand in for the work of verification, until actions follow the belief long before anyone has checked whether it has earned them.

The relationship between hope and fear shows most clearly in faith’s oldest promise, an answer to what happens after death, which most major traditions offer in one form or another. Some describe a cycle of rebirth and others a final destination of reward, judgment or punishment, while almost all of them speak to the greatest unknown any of us will face. A scientific view, by contrast, can only offer its most probable conclusion, which is that death most likely resembles the state before birth, an absence and not an experience. A state we will never experience cannot harm us.

This is also where the promise shows its other side. Many traditions offer comfort in the face of death, yet some of the most influential ones also introduce a new and far greater fear, the possibility of hell, a fate worse than nonexistence. Hell works both as a warning and as a mechanism for compliance, and even where it is rarely emphasized, its presence may lower the barrier to accepting unfounded fears. This fear is also hard to separate from the comfort it accompanies. Many believers keep heaven and let hell fade, yet within the doctrines that teach both, it is the fear of eternal suffering that gives the promise of salvation its urgency.

In that sense, this Pandora’s box of fears looks more like a structural feature than an accidental byproduct, as it amplifies the perceived value of what the faith offers by first creating the problem only the faith can solve. Hell is, in effect, the same self-created problem that can surround the search for a meaning of life, just with eternity attached.

Commercial culture has learned the same trick, and a great deal of modern marketing works by selling fixes for problems we did not know we had, from insecurities manufactured by an advertisement to health anxieties invented for the sake of a supplement. The structure itself is ancient, convincing people that they are in danger and then offering the only reliable solution. A doctrine of eternal punishment followed by a path to salvation may be the most enduring version of this model ever devised, since both the danger and the protection lie beyond anything the customer could ever verify. (Detailed explanation: In the United States, 73 percent of adults say they believe in heaven but only 62 percent in hell (Pew Research Center, 2021). Across dozens of countries, belief in hell went with lower happiness and life satisfaction, and an experiment suggests this is mainly because thinking about hell stirs negative emotions (Shariff & Aknin, 2014). Yet across 67 countries, a higher share of people believing in hell also went with lower national crime rates, while belief in heaven went with higher ones, though such data are only correlational (Shariff & Rhemtulla, 2012). Critics call the practice of widening the boundaries of illness to grow markets for treatment [disease mongering] (Moynihan et al., 2002).)

This also exposes a curious double standard. A devout adult might gently tease a child for being afraid of the dark, or laugh at a friend who believes their house is haunted, while organizing their own moral life around the fear of an invisible judge. Structurally, the two fears are the same, as both concern unseen forces that can’t be observed or ruled out and both shape behavior through dread of what might be waiting in the dark. What separates them is the number of people who share the fear and how long they have shared it, and not the quality of the evidence.

Remember the child at the top of the cellar stairs, filling the dark with monsters? Calling those monsters sacred wouldn’t make them any more verifiable.

Even the more comforting half of the promise looks different on closer inspection. Descriptions of heaven tend to emphasize endless peace and joy and the absence of suffering, yet much like the perfect day on an endless loop, a state without contrast does not stay blissful for long, so an eternity of uninterrupted happiness would, by the mechanics of our own minds, flatten into monotony.

Can anyone die in heaven, or leave it, and if not, why is nonexistence, which at least ends all suffering, so rarely considered a form of peace in its own right? Traditions built around rebirth take the opposite view. For Buddhism, endless rebirth is itself the problem, and the highest goal is release from the cycle altogether, which may be the closest a major religion comes to treating an ending as peace.

Some older visions of the afterlife seem to grasp the importance of contrast more intuitively. In Norse mythology, the warriors gathered in Valhalla spend their days fighting and their nights feasting, healed after every battle so they can face the next, all in preparation for a final battle at the end of the world. (Detailed explanation: We quickly get used to lasting pleasures [hedonic adaptation] (Frederick & Loewenstein, 1999), and the philosopher Bernard Williams argued that an endless life would sooner or later become unbearably tedious (Williams, 1973). The Buddhist goal of release [nirvana] literally means a “blowing out” of greed, hatred, and delusion, though Buddhist teaching rejects describing it as mere annihilation (Harvey, 2012). In Snorri Sturluson’s thirteenth-century Prose Edda, the fallen warriors of Valhalla [einherjar] go out each day to fight and fell one another, then ride home to drink together, a host that will still seem too small when the wolf Fenrir comes at the end of the world [Ragnarök] (Sturluson, 1916).) Whatever one thinks of its values, this afterlife includes struggle and variation, and even an ending, which may make it more believable than a heaven of unbroken calm as a place where human beings could stay content. The more useful conclusion, however, points back to this life, because the only existence we can verify is the one we are living.

Doorway to Delusion

Unfounded convictions also rarely stay single, because a claim that can’t stand on its own evidence usually needs other claims to hold it up, and each of those tends to need further support in turn. To protect the original belief, a person may start dismissing or reinterpreting contradictory information, and the further this chain extends, the further the believer drifts from what would be realistic. Over time, something subtler happens as well, since growing used to contradiction dulls our sensitivity to it. Someone who has learned to accept that a framework can be both perfectly clear and endlessly in need of interpretation, or both all-loving and willing to condemn, may find it easier to overlook the gap between a person’s words and their actions in everyday life, which is the gap manipulators exploit.

Outside religion, the pattern is easy to see. Consider someone convinced that a secret method exists for getting rich quickly, one that bypasses the usual requirements of persistence and risk. Sustaining that conviction forces them to downplay basic economic realities and to persuade themselves that the odds are stacked in their favor. When they run into evidence of failure, such as people around them losing money in the same scheme, they may resort to wishful thinking or blame the victims for lacking resolve or understanding. Each defense deepens their commitment, and the stronger that commitment grows, the more vulnerable they become to reckless decisions and to the next scam promising what the last one failed to deliver.

A similar chain can form around the idea of destined love. A person determined to see their partner as perfect or predestined may explain away every red flag as a misunderstanding or a challenge meant to be overcome. Instead of addressing real problems or incompatibilities, they hold on to the conviction that true love conquers all and that time will resolve what they refuse to confront, hoping for change without changing anything themselves. Every new rationalization makes the next one easier, until the insight that could have led to a healthier choice is filtered out before it can even arrive.

At the scale of a society, normalizing unfalsifiable frameworks carries at least four costs. The first is dogmatism, because there is little incentive to stay open-minded or to consider alternatives when no evidence could ever change a belief. The second is arbitrariness, as competing unfalsifiable claims tend to look equally valid, or equally invalid, so people choose among them for reasons that have little to do with truth, such as upbringing or emotional appeal. Once chosen, such beliefs often become entrenched in ways that make them hard to abandon even after they stop serving the person who holds them. The third cost is a lack of predictive power, since a claim that cannot be tested cannot guide a decision with any measurable reliability, which deprives it of the practical value evidence-based models provide in things like medicine as well as our daily life. The fourth cost lies in misallocated resources and misplaced confidence in unverified claims, hence often in validating other people’s misinformed claims, along with the gradual marginalization of credible information.

Together, these costs create fertile ground for exploitation. It’s no coincidence that scams flourish in faith-like spaces, from paid horoscope readings and spiritual healers to preachers who promise blessings in exchange for donations and movements that demand ever greater commitment from their members. What makes such schemes almost impossible to catch from the inside is the defense that guards any protected premise, crediting success to the system and failure to the follower, so the seller can never be wrong. If the treatment doesn’t work, the patient’s energy was blocked, and if the blessing doesn’t arrive, the believer didn’t believe enough. You can expose this premise with the same question as any other: what would have to happen for this person to admit that their method does not work?

Those hit hardest are often those with the fewest defenses. Children are built to trust the adults around them and haven’t yet developed the tools to examine what they are told. Older people may face isolation, declining health, grief, or the approaching certainty of death, all of which can make promises of healing, protection, or reunion especially compelling, and scammers have long known to target them. (Detailed explanation: The American prosperity gospel teaches that faith, positive declarations, and generous giving to the ministry [seed faith] will be repaid with health and wealth (Bowler, 2013). In 2024, Americans aged 60 and over filed about 147,000 internet-fraud complaints with the FBI and reported nearly $4.9 billion in losses, more than any other age group (Federal Bureau of Investigation, 2025). Older adults report losing money to fraud less often than younger adults, but when they do, their losses tend to be far larger (Federal Trade Commission, 2025).) A society that treats unfounded conviction as a private matter of little consequence leaves these groups to fend for themselves.

The placebo effect adds a particularly deceptive layer on top. Because a placebo response depends on belief and not on the specific method being used, it can lend apparent validation to (almost) anything. A healer whose technique has no plausible mechanism will still see some patients improve, partly through expectation and partly because many conditions improve on their own over time. Without a controlled comparison, those improvements look like proof and the testimonials pile up. This is why the placebo effect is at once a modest but real argument for taking hope seriously and one of the easiest tools for selling what doesn’t work.

This is also why the bluntness of modern medicine, which can feel cold and discouraging and may even cost a patient some helpful hope, ought to be read differently. (Detailed explanation: Across trials that compared a placebo with no treatment at all, placebos had no important effects on objective outcomes and only small effects on patient-reported ones such as pain and nausea (Hróbjartsson & Gøtzsche, 2010), so much of the improvement seen in placebo groups comes from elsewhere, such as the natural course of the illness. Telling patients about possible side effects can make them more likely to experience them [nocebo effect], which creates a real tension between full disclosure and patient welfare (Wells & Kaptchuk, 2012).) That transparency is also evidence of intent. A system whose primary goal is health has reasons to tell patients the truth even when it disappoints them, while a practitioner whose primary goal is a sale or a devoted follower has reasons to offer whatever comfort keeps the patient coming back. And if we look at open-label placebos, they suggest that so much of the comfort dishonesty offers can likely be had without it. (Detailed explanation: As explained in Constructive Confrontation, patients knowingly given an inert pill [open-label placebo] reported more improvement than those given no treatment, mainly in self-reported symptoms (von Wernsdorff et al., 2021; Fendel et al., 2025).)

In many countries, membership and attendance at traditional religious institutions have declined over recent decades, yet the need that dogma serves has not gone away and has only fragmented. Feeds that learn what we like, combined with long stretches of isolation, may have made it easier for many people to assemble their own hyper-personalized systems of faith. They stitch together fragments of wellness culture and astrology with political narratives, health claims, and conspiracy theories into a worldview tailored to their particular fears and hopes. Each element can then be reinforced by a feed that shows them more of the same and by an online community that shares their particular mix. Trying to correct a single claim within such a system can feel like trying to convert someone to a different religion, as no claim is ever held alone. (Detailed explanation: In 2020, only 47 percent of Americans belonged to a church, synagogue, or mosque, down from 70 percent in 1999 (Jones, 2021), although the share who identify as Christian, about 62 percent, has held roughly steady since about 2019 (Pew Research Center, 2025). Sociologists have named the blend of New Age spirituality and conspiracy thinking [conspirituality] (Ward & Voas, 2011). Whether feeds are what drives it is less clear: on Facebook in 2020, cutting back content from like-minded sources for three months left users’ political attitudes measurably unchanged (Nyhan et al., 2023).)

Naturally, these personalized frameworks can be harder to examine than classical dogma. A traditional religion at least has a text and a history, along with a defined set of claims that can be studied and questioned, while a personalized belief system including faith has none of these. Its premises shift from week to week, drawn from whatever the feed delivered most recently, and when one element is challenged, its holder can jump to another without ever having to defend the whole.

Much of this new spirituality also works as an aesthetic and an identity. Crystals, tarot decks, birth charts, and manifestation journals can be beautiful objects, and the communities around them can offer belonging and a sense of individuality. A pleasing surface, however, can lower our defenses, and a market has grown that sells the look of spiritual depth as readily as any other lifestyle. Enjoying these things does no harm in itself. The danger begins where the aesthetic influences its surroundings negatively, when a horoscope helps to decide whether to take a job or a stone replaces the visit to a doctor. (Detailed explanation: About 30 percent of American adults consult astrology, tarot cards, or fortune tellers at least once a year, most of them mainly for fun; only 1 percent say they rely on them a lot when making major decisions (Pew Research Center, 2025). As noted in The Part-Time Religion of the World, cancer patients who chose alternative medicine alone over conventional treatment had about 2.5 times the risk of death (Johnson et al., 2018).)

All of this leads to an uncomfortable question most of us would prefer not to ask: if holding firm conviction in claims nobody can test is socially normal, maybe even at times admirable, where exactly do we draw the line between faith and delusion?

Broadly speaking, psychiatry describes a delusion as a fixed belief that does not change in light of conflicting evidence, a description that, taken on its own, might seem to fit a great many convictions held by perfectly healthy people. Psychiatric definitions have long anticipated this problem and addressed it directly by traditionally excluding beliefs that are ordinarily accepted within a person’s culture or religious community. A conviction held within such a community, let alone one shared by millions and handed down over centuries, is therefore by definition not a delusion, however little evidence supports it. (Detailed explanation: The DSM defines a delusion as a false belief about external reality that is firmly held despite clear evidence to the contrary, and excludes beliefs ordinarily accepted by other members of a person’s culture or subculture, naming religious faith as the example (American Psychiatric Association, 2013; American Psychiatric Association, 2022). Researchers who study this boundary call the cultural exemption controversial but clinically useful, and note that religious themes are common in delusions themselves (McKay & Ross, 2021).)

This exemption calls for some important thoughts, as it reveals how the line is actually drawn and questions if we should do the same. Here are two convictions. One person believes that an omnipotent, invisible being watches over them, answers their prayers, and will judge them after death. Another believes that a coordinated group of strangers is secretly following them, monitoring their communications and arranging small harassments throughout their day, a conviction some people describe as being gang-stalked.

Neither conviction has support from the standpoint of verifiable evidence, and one could even argue that the second is the more plausible of the two. People can, in principle, follow and harass one another, while the first requires a being whose existence lies beyond anything we can observe. Yet, we treat the two very differently, respecting the first as faith, whereas psychiatrists generally regard the second as a persecutory delusion.

By the formal definition, as the exemption makes plain, the difference lies in who else holds the belief. In practice, however, what brings a belief to clinical attention is how it works in the life of the person who holds it. Religious faith is typically shared and embedded in a community, and for many believers it is compatible with daily functioning, if not supportive of it, as it offers comfort and structure as well as a sense of belonging. The belief in being persecuted by a hidden group, by contrast, is typically isolating and terrifying, and it tends to bring severe distress along with a gradual collapse of relationships and work. What brings it to a clinic is that it demonstrably harms the person holding it, and recognizing it as a disorder is what opens a path to help.

In many ways, I would argue, this pragmatic line is the right one for medicine to draw (for now). Clinical categories exist to guide treatment, and since time and expertise are limited, clinicians must focus on where suffering is most acute and where intervention can actually help, a purpose that a category applied according to harm and functioning fulfills well. Such a category also reflects how traits exist in reality. Most of them come in degrees instead of being simply present or absent, and whether a given degree helps or harms depends (heavily) on context. For someone recovering from trauma, an unverifiable belief can be a powerful tool, while a belief that is, on paper, more plausible can trap someone else in fear. (Detailed explanation: In one study, members of new religious movements endorsed delusion-like beliefs about as often as psychiatric inpatients with delusions, yet were less distressed by them (Peters et al., 1999). By contrast, when researchers compared 128 people who believed they were being stalked by organized groups with 128 people stalked by a single individual, all of the group-stalking accounts were judged delusional, and those complainants reported more depressive and post-traumatic symptoms and worse functioning (Sheridan & James, 2015). Unusual experiences such as hearing voices or holding paranoid beliefs are not confined to psychiatric patients: across 61 population cohorts, about 7 percent of people reported such experiences, most of them transient, and only a small minority went on to develop a psychotic disorder [psychosis continuum] (Linscott & van Os, 2013).) In practice, then, medicine mostly asks whether a conviction is impairing a life and leaves aside whether it is true.

The danger begins when we mistake this practical line for a verdict about truth. A society that treats a belief as acceptable because many people hold it, because it has been around for a long time, or because it rarely produces visible breakdowns, implicitly suggests that the belief is reasonable as well. Widespread acceptance, however, has never guaranteed accuracy, let alone fairness. Slavery was once considered natural across much of the world and was defended with the full weight of tradition.

A less obvious unfairness is at work as well. Now put yourself in the shoes of someone whose conviction has been diagnosed as delusional. From the inside, your belief may feel just as certain and meaningful as any religious conviction held by the people around you, and as far as you can see, little more than the number of fellow believers and the length of the belief’s history separates the two. That is hardly a satisfying answer, and it may even make treatment harder, since you could reasonably point to the surrounding culture’s tolerance for unverifiable certainty as evidence that your own certainty warrants the same respect.

The internet has blurred the distinction even further. People who share unusual convictions, including the belief in being gang-stalked, can now find one another online, where they form communities, validate each other’s experiences, and thereby acquire the very social support that once distinguished faith from delusion. (Detailed explanation: About a quarter of self-published accounts of gang-stalking described finding support from fellow “targeted individuals” online (Sheridan et al., 2020), and a psychiatrist notes that such online subcultures strain the DSM’s cultural exemption, raising the question of how large an accepting group must be before a belief stops counting as delusional (Pierre, 2020).)

To be clear, believers are not (necessarily) mentally ill; my point concerns the structure of a belief rather than the people who hold it. A faith held by a thoughtful, kind, and well-functioning person shares one structural feature with a delusion, namely a conviction that is not meant to yield to evidence, without sharing its harm. Recognizing this shared feature still matters, because it explains why a culture that celebrates unverifiable certainty in one domain has no consistent argument against it in others. Once the structure itself is admired, the only question left is how much damage a particular instance causes, and by the time that damage becomes visible, it is often too late to prevent it.

Focusing on visible damage comes with a blind spot of its own, however. A purely clinical lens, concerned only with whether a belief disrupts a person’s functioning, treats irrational convictions as harmless until they cross a threshold of obvious impairment. Yet, long before that point, such convictions shape how people interpret information, including which cues they notice and which they ignore, and thereby steer behavior in ways that reinforce the very beliefs they grew from. Over a lifetime, the costs pile up in less visible forms, such as lost time and missed opportunities and stunted growth. Across a society, the same convictions can make people more open to manipulation and sharpen the divisions between in-groups and out-groups, while eroding the shared habits of evidence and doubt that democratic decision-making depends on.

As science has advanced, faith has retreated from many of its testable claims, and what that retreat leaves behind says a lot. Where sacred texts once made claims that could be tested, many were contradicted. Although some readers had treated such passages as figurative long before, and others have kept the literal reading regardless, many adherents now read the contradicted passages as metaphor while still treating the untestable ones as truth. (Detailed explanation: Non-literal readings of the creation story are older than modern science: Augustine offered one in the early fifth century (De Cruz, 2022). Literal readings have not disappeared either: in 2024, 37 percent of Americans said God created humans in their present form within the last 10,000 years, including 61 percent of weekly churchgoers (Brenan, 2024).) The result is a selective acceptance whose rule, where one is stated at all, amounts to letting evidence decide only those questions it can reach. Two claims from the same source, resting on the same authority, end up treated in opposite ways, one kept because it can never be checked and the other set aside because it was.

This selectivity tells us something in itself. If a text’s authority were sufficient reason to believe its claims, that authority would apply equally to all of them, and if evidence is the reason to set some claims aside, then evidence should also determine how much confidence we place in the ones that remain. Holding both standards at once, evidence for the claims that were tested and authority for those that can’t be, produces exactly the kind of inconsistency this chapter has been tracing. Over time, a framework that survives this way gets confined ever more tightly to the realm of the unverifiable, where it stays without ever having proven itself, simply because it is the only place left where it cannot lose.

Much of faith’s enduring power, then, comes from a single promise, that the uncertainties of life can be closed and the gaps in our understanding have already been filled. That promise is understandable, because uncertainty can be frightening. Yet, it rests on an assumption I would question: that uncertainty is a defect that needs to be cured, when it might just as well be a condition we can learn to live with, and perhaps even value.

Gift of Gap

We tend to suffer from the agony of the gap, the restless discomfort that sets in when one direction ends and the next has not yet taken shape. Since a prediction engine built for survival registers an uncalculated future as danger, it pushes us to fill that void as quickly as possible, and faith offers one of the most complete ways to do so. Its promise is that the gaps have already been closed: the future rests in someone’s hands and suffering has a reason, death leads somewhere, and the questions that keep us awake at night have answers we only need to accept. (Detailed explanation: Feeling uncertain about oneself tends to strengthen identification with groups that offer a clear identity and worldview, and religion is especially well suited to provide one [uncertainty-identity theory] (Hogg et al., 2010). The narrower idea that losing personal control directly raises belief in a controlling God has, as noted in The Part-Time Religion of the World, failed a large preregistered replication (Hoogeveen et al., 2018).) What faith rarely invites us to consider is whether the gap itself might be worth keeping.

To begin with, questioning does not create uncertainty, which has always been there. When we ask how something works and find that the answer is more complicated than we assumed, that complexity existed all along, and our question merely brought it into view. The same holds for the world as a whole, which, to many who follow the daily news, feels as if it is getting worse, even though much of what seems like new chaos is likely old chaos made visible by better information. Avoiding reality does not change it, while facing the unknown removes the illusion of certainty we were leaning on, and with it the false sense of security that kept us from preparing.

Understanding works in a similar way, as every answer we find tends to raise new questions once a deeper look reveals factors and connections we couldn’t see before. Take the discovery that a disease is caused by a microorganism, which immediately makes us want to know how it spreads and why some people resist it, as well as how it evolves. Knowledge, in that sense, is more like an expanding frontier and never quite a finished building, since every discovery sharpens our view while widening the horizon of what remains to be explored. That can feel discouraging to anyone hoping that enough learning will eventually deliver complete certainty, while for those who enjoy discovery, it is a promise that there will always be something left to find.

Faith, however, interrupts this process in a subtle way. A life filled with ready-made answers leaves little room for a strategic vacuum, the deliberate pause in which unexamined doubts and lingering questions can surface, so that an honest audit of our lives becomes possible. When every quiet hour is already occupied by a script and every doubt already has an approved answer, the audit stops short of the framework itself. A framework does not even need to forbid self-examination, then, as long as it decides in advance where that examination ends.

The greatest uncertainty of all concerns death, and here the gap carries a gift that is easy to miss. If you knew exactly when each of the people you love would die, you would likely postpone the visits, conversations, and reconciliations you keep meaning to have, confident that there was still time until the date drew near. Since life can change at any moment, we are pushed, at least occasionally, to live in the present and to choose deliberately whom we spend our time with, as well as to say what matters while we still can.

Not knowing when the end will come is part of what gives the middle its urgency, although not everyone feels this gift in the same way. For some people, the unpredictability of life and death produces intense anxiety, which leads them to postpone or avoid deep connections out of fear of losing them. Here faith offers its most palliative function, turning an abrupt end into a transition and a void into a destination. (Detailed explanation: As explained in Delicious Dopamine, people turn toward emotionally meaningful relationships once endings feel near (Carstensen, 2006). Across 100 studies, religiosity was only weakly linked to lower death anxiety, and 10 of the 11 studies that tested for it found at least some support for an inverted U: the most religious and the least religious feared death least, and those in between feared it most (Jong et al., 2018).) That comfort is understandable, and I will come back to what can take its place. Still, the alternative to unbearable uncertainty doesn’t have to be false certainty, as it can just as well be a different relationship to the uncertainty itself.

Uncertainty is just as much the root of anticipation, curiosity, and adventure as it is a source of fear. The brain’s reward system appears to respond strongly to outcomes that remain open, which may be why a weekly wait can be so rewarding, and why a cliffhanger, an unopened letter, or a first date can feel so alive. (Detailed explanation: As explained in Delicious Dopamine, dopamine neurons in monkeys ramp up while an uncertain reward is pending, though what that signal means is still debated (Fiorillo et al., 2003; Niv et al., 2005).) Since the earliest days of our species, the unknown has been a threat as well as a catalyst, pushing our ancestors to explore, adapt, and invent. Imagine a world in which everything was known and predetermined. There would be no surprise and little to learn from, because looking back loses its meaning when nothing could ever have turned out otherwise.

Uncertainty has a darker side too, of course, as it can serve as an excuse to postpone a decision, and fear of the unknown can also feed on itself.

Think of walking into an unfamiliar laboratory for the first time, surrounded by equipment you’ve never handled. Your nervousness makes your hands shake, which makes a mistake more likely, and the first visible mistake confirms the fear that started it, until your hands shake even more. A loop like this can turn a manageable situation into an overwhelming one, and the way out is rarely more certainty and far more often experience, the gradual accumulation of encounters with the unknown that teaches the nervous system that the stakes are survivable. (Detailed explanation: Anxiety pulls attention toward threat and away from the task at hand, which makes tasks more effortful and often impairs performance [attentional control theory] (Eysenck et al., 2007). Exposure-based therapies work in part by letting people discover that a feared outcome does not occur, building new learning that competes with the old fear [inhibitory learning] (Craske et al., 2014).)

Faith in fate can make this paralysis feel like wisdom. If everything that is meant to happen will happen, there is little reason to seek out new opportunities or take risks, and if something never happens, it simply wasn’t meant to be. Such a belief may soften disappointment, yet it also shrinks the range of options a person perceives, because an opportunity that is never pursued cannot fail, and a door that is never knocked on can always be explained as one destiny never intended to open. (Detailed explanation: In a US national survey of more than 6,000 adults, people who held fatalistic beliefs about cancer, such as that there is not much one can do to lower one’s risk, were less likely to exercise, to avoid smoking, and to eat enough fruits and vegetables, even after accounting for education and other factors (Niederdeppe & Levy, 2007).)

Faith in fate is easily confused with determinism, the view that every event, our own decisions included, follows from prior causes, yet the two point in opposite directions. Fate suggests that what is meant to happen will happen whatever we do, whereas determinism holds that it happens through what we do. Whether determinism is true, nobody can currently say with certainty, and either way it changes far less about our lives than it seems. You may have to redefine what “free” means, but you still have to weigh your options, make your choices and decide, as these actions still shape what happens next, since our actions are both caused and causally effective. Asking someone out shows this well. Their answer may already be shaped by preferences they held long before you spoke, yet only by asking do you find out, and only through that experience can you refine how you ask the next time. Whether you made the outcome happen or merely came to understand the present more clearly, the uncertainty was never your creation, only something you discover more of. Where determinism does change something is in how we judge others. Since we can’t rule it out, it costs little to take seriously that people are shaped by circumstances they never chose, which sharpens the perspective-taking of the anti-hypocrisy compass without removing the need for consequences, as consequences are themselves among the causes that shape what people do next.

There’s another reason why uncertainty may feel heavier today than it needs to. The world we live in can seem more unpredictable than ever, likely less because it has become more dangerous than because we understand it better and hear about every crisis as it unfolds. Many of the existential threats our ancestors faced, from plagues to famines and natural disasters, struck with little warning and left them with little recourse, while many of today’s threats are of our own making, be it environmental damage or political instability. That realization can feel overwhelming, yet it holds a hidden encouragement, since dangers we created are dangers we can influence. (Detailed explanation: One prominent, admittedly subjective estimate puts the risk of an existential catastrophe this century at about one in six, almost all of it from human-made sources, above all unaligned artificial intelligence and engineered pandemics (Ord, 2020).)

Then there’s how faith relates to mystery. The unknown tends to captivate us under two conditions, either when there is hope of eventually understanding it or when we feel no need to understand it because something greater is believed to be in control. In the first case, mystery invites us forward, while in the second it becomes a refuge, a place to set down responsibility, which is the relief that comes with saying “it is beyond me.” When neither condition holds, and the unknown offers neither the promise of understanding nor the comfort of a guiding hand, it stops being a mystery and turns into a blank space that leaves us alone with our choices and our limitations.

Much of faith’s appeal lies in keeping this void from ever appearing. The void can be faced as well, however, and those who learn to face it often discover that standing alone with one’s choices is simply another way of describing freedom.

Uncertainty also shapes how we relate to one another, and a measure of not knowing can be an advantage there too. We rarely know the full truth about the people around us, whether it concerns their political views and private biases, the questionable things they did years ago, or the convictions they keep to themselves. While it is tempting to assume that knowing more would always bring us closer, often it wouldn’t. If you learned every opinion and every past misstep of the colleagues, neighbors, and acquaintances you currently get along with, you might find yourself disliking many of them, even though nothing about how they actually treat you had changed.

This is part of why many people steer clear of politics at work, and why some employers explicitly discourage it. The separation between public and private life serves a function beyond politeness. People who must cooperate every day find it easier to do so when they judge one another by what they do, without first weighing what the other believes, and a space with fewer preconceptions leaves more room for honest interaction. Not knowing someone’s views also spares us the effort of suppressing irritation or nursing a grudge, energy that would otherwise be drained from our work and more important activities, as well as from the relationships that matter most.

Not knowing also protects us from our own fallibility. Because our values keep evolving, we may understand what we condemn today differently later, which is why a little restraint in judgment can preserve relationships we would later be grateful to have kept. (Detailed explanation: In US experiments, people discriminated against supporters of the opposing party in nonpolitical decisions more strongly than they discriminated by race [affective polarization] (Iyengar & Westwood, 2015). In a 2022 survey, 20 percent of US workers said colleagues had treated them poorly over their political views, although only 8 percent of organizations had issued guidelines on political talk (Society for Human Resource Management, 2022). And as explained in Projected Presumption, people of all ages expect their values to change far less over the next decade than people a decade older report having changed (Quoidbach et al., 2013).)

Limited knowledge can even change minds in ways no argument could. Picture a strongly conservative man who holds negative stereotypes about gay people and works alongside a colleague he enjoys and respects, only to learn months later that this colleague is gay. By then, the relationship has already been built on, so the stereotype has to compete with a nice person he knows. Much as in the citizens’ assemblies, contact between groups tends to shrink prejudice, especially, the theory holds, when people meet as equals and cooperate toward shared goals. (Detailed explanation: As noted in Cooperative Conservatism, contact between groups tends to reduce prejudice (Pettigrew & Tropp, 2006), though randomized studies find weaker effects for ethnic and racial prejudice and have rarely tested whether meeting as equals matters (Paluck et al., 2019). In a US national survey, heterosexuals who knew gay people, especially close friends or relatives who had told them directly, held more favorable attitudes, although part of this link may arise because gay people come out more readily to those they expect to accept them (Herek & Capitanio, 1996).) The sin label works in the opposite direction, delivering its verdict before the meeting, so that only the category is ever encountered and never the person. Many traditions insist that the label falls on acts and not on persons, yet a verdict on what someone is assumed to do still arrives before anyone has met them. Nobody should have to hide who they are in order to be accepted, because the burden of prejudice belongs to the prejudiced, and the example simply shows how a judgment formed through experience can outweigh one formed in advance.

This approach has its limits, however. Choosing not to pry differs from being deceived, and choosing not to know everything is reasonable only as long as our circumstances stay livable and our values stay intact. That choice gives us no reason to tolerate environments that harm us, or to support, through comfortable ignorance, things we would find troubling if we looked. Understanding someone’s beliefs doesn’t mean excusing harmful actions either, and empathy ought to be extended at a personal cost only when the trade-off is worth it.

That same understanding serves another purpose as well, as it helps us tell people whose values are truly irreconcilable with ours apart from people whose differences come from different experiences or priorities and who may still be capable of cooperation even without full agreement. Without that discernment, we risk treating potential allies as permanent threats, or investing in alliances that can never fully form.

Approaching others without immediate prejudice also brings a strategic gain, since it lets us learn why they believe what they believe, and that understanding is a prerequisite for changing anyone’s mind. Many prejudiced beliefs grow from secondhand narratives or from a complete lack of contact with the people being judged. Answering them with hostility or matching contempt only confirms them by casting us in the very role the prejudice predicted. Kindness and cooperation that contradict their expectations, by contrast, offer something their worldview hasn’t prepared them for, a contradiction they get to experience instead of merely hearing about. (Detailed explanation: In field experiments, canvassers who listened and exchanged personal stories without judgment modestly but durably reduced exclusionary attitudes toward immigrants and transgender people (Kalla & Broockman, 2020).)

To get along in a world full of people we can never fully know, we need a tool, and that tool is good faith, meaning provisional trust in decent intent and baseline competence. It is often mistaken for naivety, yet I would argue it is better understood as a strategic baseline. Treating every stranger as a threat consumes (enormous) mental energy and tends to produce the very conflicts it anticipates, as people who sense suspicion commonly respond with defensiveness. Good faith is realistic optimism applied to people.

Good faith also takes a kind of courage that is rarely recognized as such. Cynicism is a comfortable shield, since it takes no bravery to assume that everyone is out to exploit us, and the cynic is never disappointed because they never hoped. Extending trust while knowing you might get burned is far harder, and it is worthwhile only because the reward, real connection and reliable cooperation, justifies the risk. That is also why good faith ought to be calibrated and never unconditional. It makes sense for you to be a little naive where the potential downside is small or the potential reward justifies the danger, and to be far more careful where a single misjudgment could cause serious harm. (Detailed explanation: People tend to underestimate how trustworthy strangers are, partly because distrust is rarely put to the test: those who never trust never learn they were wrong (Fetchenhauer & Dunning, 2010). In large US and German samples, more cynical people went on to earn less, except in countries where prosocial behavior was rare and cynicism arguably warranted (Stavrova & Ehlebracht, 2016).)

Good faith is a starting line and should never turn into a life sentence. Both parties must maintain it, and the moment someone shows bad faith through malice, your obligation to keep extending it ends. Continuing to offer good faith to those who repeatedly exploit it removes the consequences of their behavior and shifts the cost onto everyone who plays by the rules, which makes it a subtle form of complicity. Good faith also demands mutual effort, since in a debate or a shared project it presumes that both sides are willing to earn their credibility by doing the reading and preparing their arguments. When one side refuses to pay that price and argues purely from vague impressions, continuing to engage them in good faith mostly wastes the time of the person who did.

When someone does let you down, the more useful question is whether your baseline was wrong for this context or whether this was an outlier, rather than concluding that you were stupid to trust them. A colleague who misses one deadline after years of reliability is an outlier, while a friend who has yet to return a single thing they borrowed suggests your baseline was set for someone else, at least when it comes to lending. Either answer beats self-reproach, as the first lets you keep trusting and the second tells you where to recalibrate, often in one specific area instead of for the whole person.

Applied well, good faith also changes how we handle the imperfect ways in which people express themselves. When someone phrases an argument poorly or makes a minor error, good faith means you engage with the strongest possible version of what they are trying to say. That spares both sides from wasting energy on semantic misunderstandings and keeps the conversation focused on the substance, often where the real disagreement lies. Such charity tends to be returned as well, because people who feel understood at their best are (way) more willing to examine their own position than people who feel caught at their worst. (Detailed explanation: As explained in Converging Conversation, people who were listened to attentively became less extreme and more willing to acknowledge contradictions in their own views (Itzchakov et al., 2017).)

If you notice that you can no longer extend good faith to anyone at all, that has diagnostic value too. Such pervasive suspicion can of course reflect a truly hostile environment, but it can also be a symptom that your own boundaries have been crossed repeatedly without you enforcing any consequences. Since consequences teach the world how to treat us, a person who knows they can protect themselves finds it far easier to be generous.

That provisional trust is what living with uncertainty looks like between people, and it shows that we do not need false certainty to act with confidence and warmth. Since we could never be sure which goals are the best, or exactly when we have achieved them, it becomes essential to savor the present and, whenever we have the energy, to keep growing. The gap that faith rushes to fill is the same space in which many valuable things can take place. That is why being prepared for change seems far better to me than simply praying.

The longing to escape uncertainty does not stop at the individual either. Whole societies have traded their freedom for the promise that someone above them knows best, and some of the most enduring forms of power in history have drawn their authority from the same source as faith.

Benevolent Bottleneck

A common defense of faith holds that even if its claims can’t be verified, a belief system that leads people toward kindness is surely harmless, perhaps even beneficial. The argument sounds remarkably similar to a familiar defense of dictatorship, and, as with moral absolutism, the resemblance runs deeper than rhetoric. After all, if a ruler is wise and benevolent, why should it matter that nothing checks their power? In both cases, the answer is that a system without checks and balances can only be as good as whoever happens to hold the power, or whoever happens to interpret the doctrine.

A benevolent ruler can be succeeded by a cruel one, just as a compassionate interpretation can be replaced by a vengeful one, and the structures that allowed good to be done swiftly will then allow harm to be done just as swiftly, since nothing is (systematically) in place to stop it.

Dictatorships are commonly defended on the grounds of efficiency, because a single decisive leader, free from bureaucratic delay and opposition parties, can act quickly. In narrowly defined situations, where the problem is clear and the solution known, and where dissent is minimal, this can indeed be efficient and effective at the same time. Efficiency and effectiveness, however, are two different things. Efficiency, meaning doing things quickly or with little friction, only has value as long as it consistently produces sound decisions, and without them all that remains is speed without substance. A rushed decision that collapses under its own weight demands more time, more correction, more resources and more control than a slower, better one would have needed. (Detailed explanation: Across countries, democratizations raised GDP per capita by about 20 percent in the long run (Acemoglu et al., 2019), and autocracies are overrepresented among both the best and the worst economic performers, as one would expect if outcomes depend heavily on who happens to hold power (Almeida & Ferreira, 2002).)

Governing on a large scale rarely offers such narrow situations. The decisions of a society involve economic, cultural, legal, social, geographic, and environmental factors that interact in complex ways. Since no single human being, however capable and well-intentioned, can master all of them, a benevolent dictator who wants to decide well must consult experts in each domain. Even if such a ruler somehow had every answer, they would still need to sleep, clean, eat and rest, and they could only process one problem at a time, just as they could only see the accurate proximity of one location. That alone turns the single decision-maker into a bottleneck through which every important choice has to pass.

This leads to an irony: the more a dictator pursues competent decisions across many areas, the more their process starts to resemble the participatory systems they claim to improve on, distributing expertise and gathering input while refining choices through deliberation. In the end, the pursuit of effectiveness erodes the very efficiency that was supposed to justify concentrating power in the first place.

A second paradox follows from the first. A dictator’s vision of reality can only serve a society well if it benefits nearly everyone, yet a vision that benefits nearly everyone would hardly remain unique to one person, since most people would come to share it and could choose it themselves. Where a ruler’s view truly is good for all, the dictatorship is unnecessary, and where only force can impose it, that need alone suggests it isn’t.

The deeper fragility of such systems lies in what they fail to cultivate. Since their stability rests on faith in the leader instead of on the capability of the people, they rarely build independent thought or civic responsibility, both of which would invite questions the system would prefer to avoid. A wise ruler may compensate for these flaws without being able to resolve them. They might try to groom a successor, yet no lesson can substitute for the experience of being accountable for one’s choices, so without that experience any successor would have to be nearly flawless, an impossible standard.

The public, meanwhile, stays unpracticed in autonomy and unfamiliar with compromise, shielded from the hard lessons only responsibility can teach. Once the central authority weakens or disappears, such a system survives only to the extent that power had been shared behind the scenes with a party or an elite all along. Even then, the public, never allowed to learn how to steer, inherits a regime without gaining a voice in it. (Detailed explanation: Of 79 dictators who died in office between 1946 and 2014, the regime outlived its leader in 92 percent of cases, usually where a ruling party already shared power; regimes built around a single strongman were the most vulnerable (Kendall-Taylor & Frantz, 2015).)

Good leadership also requires humility, meaning the ability to recognize one’s limits and listen to dissent, as well as to change course when circumstances shift. An absolute ruler, by contrast, relies on projecting infallibility to keep the trust of their subjects unconditional. Since admitting a mistake would put at risk the perception of reliability on which their authority rests, rulers in rigid power structures tend to persist with flawed decisions long after the evidence has turned against them, preferring to reframe events over revising their course. It is the same reflex as that of the parent who answers a child’s question with “because I said so,” preserving their authority at the cost of trust, except that on the scale of a nation it can preserve a regime at the cost of its people. (Detailed explanation: Amartya Sen observed that no major famine has occurred in a functioning democracy with a relatively free press, since governments that face elections and public criticism cannot easily ignore one (Sen, 1999).)

The way such systems justify themselves holds a further irony. Authoritarian rule usually claims to serve a common good that only strong, centralized power can secure. However, if a population truly shared a unified set of values, the rationale for coercion would largely disappear. Harmony would then emerge from mutual understanding, and the state could shift its priorities from surveillance and control toward education and infrastructure, along with the cultivation of civic life. Leadership could also become more fluid and merit-based, reflecting competence and public trust more than lineage or loyalty.

Authoritarian systems resist this logic, because their legitimacy rests on raising a single vision above all others, and over time that vision must harden into an orthodoxy presented as inevitable. Their hold on power persists precisely because the population is not united.

An authority truly driven by a higher collective aim, and not by its own preservation, would make its power conditional by design. Like a scientific theory, it would welcome comparison and scrutiny, and should a better alternative emerge, it would support its own replacement. Most observable authoritarian structures fail this test, as they tend to deflect accountability by blaming their failures on external threats or others’ shortcomings, and keep their legitimacy by controlling the narrative, framing dissent as disloyalty and presenting systemic failures as unavoidable misfortunes.

The case for distributing power, then, rests on a simple observation about human beings: no individual, and no small elite, can be omniscient enough to foresee every consequence of their decisions, or omnipotent enough to secure their power and the competence of their successors indefinitely. Any being that could do so would no longer be human and would, for all practical purposes, be a god.

That may explain why so many faith-based systems describe their highest authority in these very terms, since omniscience and omnipotence, moral perfection and immortality are exactly the qualities a ruler would need for unquestionable, centralized authority to work. Attributing them to a divine being places faith’s authority beyond the human limitations that undermine every earthly dictatorship, and the infallibility a human ruler can only pretend to possess becomes, in the case of a god, a matter of definition. A single mind stops being a bottleneck once it knows everything at once, and succession stops being a problem when the ruler is immortal. This arguably makes the all-powerful deity the only form in which the fantasy of the benevolent dictator can remain valid.

Even this consistency breaks down once it meets reality, however, because in most traditions the divine doesn’t govern directly but acts through prophets, messengers, sacred texts, institutions, and the natural order, every one of which passes through human hands. The source keeps its perfection by being held at a distance from the fallibility of its interpreters, so that any error can be blamed on the messenger and never on the message. That solution, however, brings back the very problems the divine was supposed to solve. As soon as divine will must pass through human intermediaries, it becomes subject to interpretation and distortion as well as to the ambitions of those who claim to speak for it, which recreates the bottlenecks and corruption of every human hierarchy. Many traditions resolve this tension by appealing to a hidden order, suggesting that even the flawed mechanisms ultimately serve a perfect plan we cannot see, an appeal that raises a question of its own.

If divine will has to pass through prophets, texts, and institutions before it reaches us, then what reaches us looks more like a faint signal, filtered through many noisy channels, than a clear command from a sovereign. Every interpretation of that signal is provisional and open to error, and nobody can claim an authoritative understanding of it without asserting one. Even the humble position that the divine is unknowable in principle is itself an interpretation, and therefore just as fallible as the confident ones.

The practical conclusion is that if we cannot verify a hidden order, we cannot act on it with any reliability either. Hence, the frameworks that stay useful are those designed to work under uncertainty, such as participatory governance and empirical inquiry. Their value never depended on metaphysical certainty and lies instead in their resilience whenever certainty is unavailable, which for human beings is the case nearly all of the time.

An absolute moral authority also claims legitimacy through superior knowledge or power, yet each of these claims undermines the need for obedience, because an authority that truly understood human nature could shape our desires directly, making resistance unlikely and commandments unnecessary. Theists usually answer that such an authority wants obedience to be freely given and not engineered. (Detailed explanation: The classic theistic reply is that a world of free creatures who can choose wrongly may be better than one of creatures compelled to do right, so a deity could not simply shape our desires without abolishing that freedom [free will defense] (Plantinga, 1977).) Yet, a choice made under the threat of eternal punishment, guided by instructions that need constant interpretation, makes for a strange model of freedom.

If, on the other hand, its plan would unfold the same way whatever we chose, as fate and not as a chain of causes that includes our choices, obedience would change nothing. In the first case, the demand sits uneasily with the very freedom it claims to protect, and in the second it makes no difference at all. Either way, we are asked to obey without being able to see why it matters.

Beyond that, the demand presumes something no human being can deliver. An authority that enforces a fixed moral code in a fluid world implicitly requires a complete moral blueprint, a guide that prescribes the correct action and the correct ranking of values for every conceivable situation. Even if such a blueprint could be written, nobody could memorize it, let alone apply it accurately in the middle of real life. Even the most devoted followers would inevitably misstep, out of ordinary human limitation and not out of malice, which turns the instruction to get it right into a burden that can never be lifted.

That burden grows heavier still when meaningful feedback is absent. In most areas of life, we learn whether we acted well by observing the consequences, while in a framework that judges outcomes mainly through loyalty, even clear failures can be reframed as necessary. This cuts the link between action and result that makes learning possible in the first place. As accountability fades, the role of the individual narrows to compliance, and their satisfaction comes to depend on rewards earned through obedience more than through agency. If the authority also permits confusion and suffering as tests of conviction, obedience comes closer to submission to an unpredictable force. (Detailed explanation: Scripture often frames hardship as a trial: “We will certainly test you with a touch of fear and famine and loss of property, life, and crops” (Qur’an 2:155) (Khattab, n.d.), and believers are told to “consider it all joy” when facing trials, because the testing of faith produces endurance (James 1:2–3) (National Council of the Churches of Christ in the USA, 2021).) In any other relationship, we would call it wishful thinking to keep trusting under such conditions.

Measured by the standards we apply to every other form of authority, such as practical effectiveness and responsiveness to changing conditions, an exclusive claim to authority quickly weakens, because practicality invites comparison, and comparison invites alternatives, some of which may yield better results. To keep its status, an absolute authority must therefore shift the basis of its legitimacy toward something harder to pin down, be it revelation or a transcendent purpose. Stripped of those claims, it would become one model among many, no longer above scrutiny and judged by its performance like every other system.

Faith-based systems don’t even need a scheming ruler to work as mechanisms of control. Their structure alone is enough. The benefits they promise, such as comfort and certainty, shared ritual and unity, flourish most fully in an environment free of doubt, since doubt weakens both the conviction that makes the comfort work and the cohesion that holds the community together. A framework whose rewards depend on the absence of doubt will therefore tend to discourage doubt, even if nobody ever intends to control anyone, and over centuries this tendency made faith a natural ally of established power.

Once doctrines are presented as sacred and beyond question, the institutions aligned with them inherit some of that untouchability. The medieval and Renaissance papacy, which crowned emperors and contested the authority of kings, is one of the clearest examples of how spiritual and political power can merge.

A less visible tool of such authority is the trade-off presented as inevitable. When values are framed as parts of a rigid system, raising one can seem to require sacrificing another, and the framework rarely invites anyone to ask whether that sacrifice is truly necessary. A leader can claim, for instance, that maintaining order requires suppressing dissent, treating stability and freedom as mutually exclusive instead of searching for ways to secure both. What makes this reasoning appealing is its emotional immediacy, as it offers instant validation for the value one already prioritizes while sparing everyone the harder work of examining whether the conflict is real or merely convenient.

Throughout history, faith has also served power in a more direct way. The promise of heaven and ultimate redemption could make severe earthly suffering and deep inequality feel bearable, even meaningful, by deferring the reward for endurance until after death. Doctrines such as the divine right of kings, which held that monarchs answered to God and not to their subjects, could make resistance feel sinful as well as dangerous, while the injustices of rulers were left to divine judgment or to cosmic retribution in some distant future. This offers relief of pain while also being the reason nothing is being done about the source of it. Wars waged under religious banners, such as the crusades, involved a mix of sincere piety and the promised remission of sins along with political ambition and plunder that historians are still untangling. Yet, they show how readily the language of faith can be wrapped around very worldly aims.

Faith, however, has also fueled resistance. Religious conviction inspired many who fought against slavery, and clergy stood at the forefront of movements for civil rights, drawing on scripture to demand justice in the present instead of patience until the afterlife. (Detailed explanation: Faith stood on both sides: the Southern Baptist Convention, formed in 1845 in a split over whether slaveholders could serve as missionaries, formally apologized in 1995 for the role slavery played in its founding (Southern Baptist Convention, 1995), while the Southern Christian Leadership Conference, founded in 1957 with Martin Luther King Jr. as its leader, drew its strength from Black churches and ministers (Martin Luther King, n.d.).) Faith itself, then, does not decide which side of power it serves, which is precisely the problem. Science can be bent too, but its records let others catch the bending, while a framework whose authority rests on interpretation offers no such check, and hence no reliable protection against submission.

Traditional religion has lost influence in many societies, yet this mechanism has simply migrated, and where religion stays strong, it has often found a new partner in populism, which uses much of the same machinery. Populism offers a savior figure who promises to fix what ordinary politics can’t, demands loyalty instead of verification, and frames critics as enemies of the people, all while supplying the comfort of a simple story and leaving the harder realities unaddressed. Such movements tend to thrive on vagueness and on their followers’ conviction that their leader is uniquely untainted. This follows the same pattern as faith, with authority placed beyond question and failures blamed on enemies or on the insufficient devotion of followers, while the reward always sits somewhere just beyond the next election. (Detailed explanation: Between 2010 and 2020, Christians’ share of the world population fell from about 30.6 to 28.8 percent (Hackett et al., 2025). Political scientists commonly define populism as a worldview that divides society into “the pure people” and “the corrupt elite” (Mudde, 2004). Its relationship with religion cuts both ways: in Western Europe, churchgoers have been less likely to vote for radical-right parties, largely because of their ties to Christian Democratic and conservative parties (Arzheimer & Carter, 2009), whereas in the 2016 US election about 80 percent of white evangelical Christians voted for Donald Trump, according to exit polls (Martínez & Smith, 2016).)

Science offers a different model of authority, one that began by acknowledging human fallibility and diversity instead of attempting to transcend them. (Detailed explanation: The Royal Society, which first met in 1660, adopted the motto Nullius in verba, usually rendered as “take nobody’s word for it” (The Royal Society, n.d.).) Since it depends neither on the infallibility of a leader nor on the permanence of a doctrine, and works through a method that stays open to correction, its progress is cumulative and not tied to the integrity of any single person or institution. All it requires is a shared aim, the pursuit of understanding, around which people with different priorities can work together.

Its decisions are judged by their outcomes and, more importantly, by the reasoning behind them, so that mistakes can be corrected without destroying the legitimacy of the whole. Its strength lies in its ability to navigate uncertainty, not in certainty itself, as it steadily brings clarity where there was once only doubt. This model, however, comes with a cost many people would rather not hear about: a distributed system can only be as capable as the people taking part in it. Where a dictator demands obedience, a system without one demands the willingness to engage, as well as responsibility, from everyone it includes. If we ask people to step away from faith, we owe them more than a critique: an honest account of what it costs to leave, what can take its place and why it is the best bet for (nearly) everyone.

Hope without Heaven

A critique that stops at demolition leaves people standing in the rubble of what once sheltered them, and frustration with nowhere constructive to go frequently flows toward the nearest answer that promises what the old faith used to provide.

Leaving a faith rarely happens in a single moment, since for most people it begins as a slow erosion, perhaps with a question that won’t go away or a contradiction that no longer feels resolvable. Sometimes it is a prayer that has started to feel more like talking to an empty room than it used to. Doubts typically pile up in private long before anyone speaks them aloud, and some people never speak them at all. (Detailed explanation: In interviews with American atheists, losing faith unfolded as a process rather than a moment: questioning the religion one grew up with, then rejecting theism, and only later “coming out” as an atheist (Smith, 2011).) When the moment of clarity does arrive, it often brings relief and loss at once.

That loss can be considerable, especially where the faith was acquired in childhood, since leaving the conviction then often means losing much of the belonging that grew up around it. The weekly gatherings disappear from the calendar and the holidays lose their familiar meaning, while the community that once showed up with food in times of illness may grow distant, if not close its doors entirely. In some communities, belonging was conditional on conformity all along, so those who stop believing are ostracized, sometimes even by their own families. (Detailed explanation: In a survey of 796 American atheists, reported discrimination ranged from slander and coercion to social ostracism and the denial of opportunities, and it was more frequent among those who had grown up with stricter religious expectations in their families (Hammer et al., 2012). People who leave demanding, exclusive groups also report worse health than people who leave less demanding traditions (Scheitle & Adamczyk, 2010).) Where belonging works this way, care that flows through membership tends to stop where the membership ends, which means that for those who leave such communities, the very thing commonly celebrated as the great strength of faith turns out to have been one of its sharpest instruments of exclusion.

Painful as it is, this loss also brings some clarity. Some of the relationships that dissolve once the shared conviction disappears were held together mainly by that conviction, while others give way under pressure from the community even though love wasn’t lacking. Either way, they were conditional, even if neither side had realized it. The relationships that do survive have passed a test few bonds ever face, and since a connection that withstands honesty earns a kind of trust that untested peace cannot provide, that strain has its value. Leaving a faith can therefore shrink a person’s social world while making what remains of it (much) more reliable.

A second loss is harder to spot from the outside. Many believers credit their faith, or divine help, for the good things in their lives, such as the exam they passed or the illness they survived, and often for the strength they found in a difficult season as well. Once the faith fades, that credit has nowhere left to go, and for a while it can feel as if a source of support has been withdrawn. Self-esteem may dip during this time, as the person starts to wonder whether they were ever capable on their own, or whether they will manage without the help they believed they had. (Detailed explanation: In everyday explanations, believers credit God overwhelmingly for good outcomes rather than bad ones (Lupfer et al., 1992). Among church members, a style of coping that leaves problems to God [deferring] has been linked to lower self-esteem and a weaker sense of personal control (Pargament et al., 1988).)

Worked through, however, this crisis can become one of the most powerful sources of grounded confidence a person will ever have. As far as anyone can observe, every success they once attributed to divine help came from their own effort and the support of the people around them, along with the circumstances they faced, so the strength they drew on was theirs all along. Even if something akin to a god existed, the most faith could have supplied was the spark that let believers draw on that strength themselves. Leaving the faith removes that spark, yet the capacity it once helped release remains, and that capacity can learn to start without it. In many cases, the belief in a divine backup was likely a placeholder for the belief in oneself that the person had needed all along.

Leaving also forces the question of what to do with the past. Above all in the first flush of disillusionment, it is tempting to condemn everything that came before as wasted time, foolishness or deception. Yet hindsight has its hazards, since the person we once were made the best choices available to them with the perspective and environment they had at the time. Much of what was learned and experienced within a faith can still hold value. Sympathy, empathy, honesty, generosity and a sense of purpose do not lose their worth just because the reasons once given for them have changed. They can be re-examined and re-grounded in earthly reasoning, and they often come out stronger, as they are now understood and no longer merely inherited.

The same holds for experiences and relationships. The joy felt at a holiday gathering was real, as were the comfort of singing together and the friendships formed in a youth group, and so was the sense of being part of something larger, even if the framework that structured all of it no longer convinces. Much as the horns effect can recolor an entire relationship after a painful ending, disillusionment can recolor an entire upbringing, and resisting that temptation is simply a matter of accuracy. We do not have to erase the past in order to outgrow it, and the shards of gold we feared losing outside the package can often be carried out of it.

Between the old life and the new one lies a stretch of time that can be thoroughly disorienting. William Bridges, a consultant who wrote influentially about how people get through major life changes, described transitions in three phases, beginning with an ending, the point at which continuing as before is no longer viable. A “neutral zone” follows, an in-between period in which the old identity is gone while the new one has not yet formed, and only after that does a new beginning take shape. (Detailed explanation: Bridges, a former English professor turned consultant, set out the sequence of ending, neutral zone, and new beginning in Transitions (1980), drawing on his own life change and on anthropological accounts of rites of passage rather than on controlled studies (Bridges, 1980).) The neutral zone is uncomfortable because it is a version of that familiar gap where the prediction engine can no longer see the nice road ahead.

It may still be the most important phase. Those who rush through it can end up grabbing the first answer that promises to end the discomfort. They trade one framework that couldn’t be questioned for another, regardless of its degree of utility. (Detailed explanation: Feeling uncertain about who we are draws us toward groups that offer a clear identity and firm rules for how to think and act, which can make extreme groups especially attractive [uncertainty-identity theory] (Hogg, 2014); personal uncertainty can likewise push people toward more extreme conviction (McGregor et al., 2001).) Treated as a strategic vacuum, a period set aside for an honest audit, the neutral zone serves a far better purpose than when it is handled as an emergency to be ended as fast as possible. It is the time to decide, without haste, which parts of your past have truly ended and which can continue in a new form, as well as the time to accept the feelings that come with the change and to postpone major decisions until the ground feels steadier.

Not every departure needs a public announcement. Some people keep attending family gatherings and holiday services long after they have stopped believing, not out of hypocrisy but out of love for the people who still find meaning there. Honesty with oneself comes first and is not negotiable, since self-deception would only rebuild the old structure in a new shape, while honesty with others can be paced. Being honest doesn’t demand that we unload every conclusion at once, and since timing can matter as much as content, a degree of strategic restraint is legitimate here. This holds as long as the silence is a conscious choice made for reasons one can state and does not turn into a debt paid off in installments out of fear.

Nor does leaving have to mean a life without ritual or community, as much of what faith provided can be rebuilt on different foundations. Holiday meals can stay family traditions without any doctrine, and music can still be sung together. The weekly rhythm of gathering can be found in clubs, volunteer organizations, sports teams or circles of friends who choose to meet regularly. As the dense networks of clubs and associations in the Nordic countries show, communal belonging depends on people who show up for one another. That habit can be built without shared metaphysics, even if secular groups have yet to prove that they can fully match what a congregation provides. (Detailed explanation: In US survey data spanning four decades, people who had left the religion of their childhood reported worse health and lower well-being than both those who stayed and those who had never been religious, and the gap was fully explained by how rarely they still attended services (Fenelon & Danielsen, 2016). What seems to be lost in leaving, in other words, is mostly the community.)

That said, not everyone who stays in their faith does so out of habit, fear or love for their community. Many stay simply because faith seems to work for them, comforting them and helping them endure what would otherwise be unbearable, and studies also link regular participation in religious communities with slightly better well-being. (Detailed explanation: Across many countries, religious people report slightly higher well-being, largely through social support, respect, and a sense of meaning, but the advantage depends on circumstances: it is clear in societies with harsh living conditions, while in prosperous societies religious and nonreligious people report similar well-being (Diener et al., 2011).) If a belief makes a particular person’s life better, who are we to take it away, and why would it not be rational for them to keep it, whether or not it is true?

I’m willing to concede a lot of this. For some people, faith may indeed be the faster route to stability, offering in a single package what would otherwise take years of reflection and experimentation to build. Some may even find a peace through it that they would struggle to reach any other way. Speed and accessibility are real advantages. Efficiency, however, differs from reliability, as a route that arrives quickly is only valuable if it consistently arrives at the right destination and keeps us there. As before, we must ask whether a world in which everyone lived this way would be the world we want, and I think the answer is clearly no, for faith earns its value only from how well it works today, and it works today only because it lives off a world that doesn’t fully run on it. The believer who prays for a recovery usually does so in a hospital built on centuries of medical research, and the community that carries them through hard times operates within a society whose laws, infrastructure and safety nets were largely built by people testing what works rather than trusting what was handed down. Remove that host, and much of what made faith look effective would go with it.

The objection also runs into a familiar problem, since judging a belief by its outcome, and not by the reasoning behind it, is the same error that makes gambling look like a strategy. Every casino has its lucky winners, and some of them walk away considerably richer. Still, nobody concludes from their success that gambling is a sound financial plan, as we know their win was the fortunate outcome of an unreliable method and that for every winner, many others had to lose. If beliefs were accepted simply because they benefit some of the people who hold them, any belief, however irrational or harmful in other hands or applied to something else, could be defended in the same way.

I want to be fair, so let’s suppose that there is a particular person for whom faith truly does produce better outcomes than any evidence-based approach could have produced within their lifetime, both in terms of well-being, stability and emotional resilience and in the decisions their faith inspired. Suppose further that their faith causes no broader harm, or that this person, judging only their own life, considers such harm irrelevant to their choice. Under those conditions, one could argue that choosing faith would be rational for that individual. To me, this narrow exception does not weaken the case for reasoned conviction. It shows that a framework based on reasoning can at least entertain the possibility that faith is the better choice, which a faith-based framework, unable to entertain its own error, could never do in return.

The difficulty lies in what this scenario requires. To know that faith will lead to better outcomes than every available alternative, a person would need reliable foresight about the course of their own life, covering the decisions they will face and the paths they didn’t take, along with the consequences of each. They would also need to know that the typical risks of faith won’t materialize for them, from manipulation and missed treatment to misdirected effort and devastating disappointment. Nobody has that kind of knowledge. At the moment of choosing, the only information available is that the doctrine itself can’t be checked and that much of what it gives seems to travel with the community more than with the creed.

Even if someone could reliably predict that believing would be good for them, that still wouldn’t show the belief to be true. It would give them a reason to want to believe, not a reason to think the belief is correct, and a choice based on a prediction lasts only as long as the prediction holds. If the benefits stopped, or if the person’s priorities changed, they would have to give the belief up. A commitment that lasts only as long as it keeps paying off is not really faith. It is a calculation.

What remains, then, is a fortunate alignment between conviction and benefit, valuable to the person who experiences it but nothing that can be recommended to others or scaled to a society. This is the vulnerability I postponed earlier, when describing improbable narratives as a load-bearing pillar for someone who finds reality too hard to face. Such narratives can indeed bear that load, but the pillar holds partly by luck, and no one can know in advance whether their luck will last.

The comfort of faith also carries a structural tension that the pragmatist’s objection tends to overlook. Faith generally delivers its greatest benefits when it is absolute and untroubled by doubt or hesitation, since unwavering conviction can provide a great deal of confidence. That same absoluteness, however, raises the stakes of failure, as the stronger the faith, the more it has to lose when the hoped-for outcome doesn’t arrive. Where the conviction does survive the blow, it usually does so by explaining the failure away. Absolute faith also encourages blind adherence, which can lead people to ignore evidence and suppress their own critical judgment while they persist on paths that harm them. Faith held alongside lingering doubts, on the other hand, may be the hardest position of all, as it sacrifices much of the comfort while keeping many of the risks.

Part of what helps seems to be certainty itself, religious or not, since confident nonbelievers appear to fare about as well as confident believers. That raises the fair question of whether a conviction held only as firmly as its evidence allows can offer the same steadiness. (Detailed explanation: A review of 14 studies found that greater certainty in one’s worldview, religious or not, goes with better psychological health (Weber et al., 2012), and one study found that confident believers and confident nonbelievers both fared better than those in between [curvilinear relationship] (Galen & Kloet, 2011), although this pattern has not appeared consistently across later samples and measures.)

Intertwined with hope, faith can also become a subtle form of procrastination. Instead of confronting a difficult reality and acting on it, a person may settle into a posture of waiting for an intervention or a miraculous turn of events. If a favorable outcome does arrive, it may bring less surprise than one might think, as it merely confirms what was already expected. As with the credit discussed earlier, some of the gratitude then flows to the faith instead of to the people and the effort that produced the outcome. (Detailed explanation: More religious people tend to report more gratitude in everyday life, not less (McCullough et al., 2002).) If the outcome is unfavorable or stays uncertain, the inner conflict intensifies, pitting hope against reason and raising doubts about the future as well as about the validity of the conviction itself. This waiting closely parallels how people sometimes avoid processing painful emotions, deferring the work through distraction, substances or denial. A temporary deferral can be understandable when other responsibilities demand attention, but habitual deferral traps a person in a kind of limbo, never fully healed and never fully free of what they keep avoiding.

A loss that clashes sharply with what we firmly expected tends to hit us harder than one whose possibility we accepted early on. (Detailed explanation: Outcomes feel stronger when they are unexpected: a surprising loss hurts more than an expected one, while an expected win brings less joy than a surprising one [decision affect theory] (Mellers et al., 1997; Shepperd & McNulty, 2002). After a death, feeling unprepared for it has been linked to complicated grief (Barry et al., 2002).) Consider a family convinced that their faith guarantees a loved one’s recovery. If the recovery does not come, they may face both the grief of the loss and the shattering of a conviction they had leaned on entirely, often along with agonizing questions about what they did wrong or why they were abandoned. (Detailed explanation: Among 345 patients with advanced cancer, those who leaned most heavily on religious coping were roughly three times as likely to receive mechanical ventilation or resuscitation in their last week of life (Phelps et al., 2009).) Acknowledging the seriousness of a situation leaves hope intact while letting reality shape our actions as long as there is still time to act, so that we meet whatever happens prepared.

Optimism over Obedience

Steadiness doesn’t require certainty, and realistic optimism shows how. It is the deliberate decision to adopt a hopeful outlook where we understand why it helps and the circumstances call for it, held with full awareness of the risks. It delivers much of what faith promises, from the motivation to begin and the endurance to continue to the higher likelihood of success that comes from believing an effort is worthwhile. (Detailed explanation: Expecting success is not the same as daydreaming about it: among job seekers, students, and surgical patients, positive expectations predicted more effort and better results, while positive fantasies predicted less (Oettingen & Mayer, 2002).) Yet it does not switch off the radar. In its absolute form, faith asks us to stop reading the evidence once we have committed, whereas realistic optimism keeps reading it, ready to adjust course the moment the road changes.

Applied to our own circumstances, realistic optimism amounts to a provisional assumption that effort can pay off, which in turn raises the chance that it will, much as good faith toward people raises the chance of cooperation. Pessimism has its narrow uses, such as rehearsing what could go wrong before a test, but as a habit it rations our effort and narrows our vision. (Detailed explanation: Pessimism has a narrow niche: anxious people who deliberately expect the worst and rehearse what could go wrong [defensive pessimism] tend to do better than they predict, and worse when encouraged to expect success (Norem & Cantor, 1986).)

If an optimistic attempt fails, we have at least tried and gathered information, and grown more capable in the process, while if we never try, the loss is guaranteed. Unlike faith, realistic optimism is also honest about its own nature, as all it claims is that, where the conditions favor it, acting as though things can work out while closely watching for evidence that they can’t is the strategy most likely to make them work out.

To make this tangible, consider someone struggling with social anxiety, the persistent fear of being judged or rejected. A faith-based framework might comfort them with the popular idea of karma, which says that if you are kind, kindness will return to you, in this life or a later one. A probabilistic framework might say something remarkably similar, namely that if you consistently engage with others warmly and with good intentions, the likelihood of positive social outcomes rises. In practice, both encourage the same behavior and offer resilience in the face of uncertainty.

They part ways once reality doesn’t cooperate. When kindness isn’t returned, karma can only counsel patience or suggest that older deeds, perhaps from a past life, are still bearing fruit, whereas the probabilistic version expects exceptions from the start. It already knows that a cheerful approach can feel out of place while someone else is grieving and that some people respond poorly whatever we do, just as it knows that a single rejection says little about the overall pattern. It can also take in what psychology and neuroscience reveal about anxiety and refine its advice as new evidence arrives, while keeping its partial truths even when one of its arguments fails, since it rests on many pillars rather than one.

One common objection remains. Not everyone has the time or training to reason everything through, the argument goes, and religion offers a ready-made moral template to those who don’t. The concern is understandable, yet it rests on questionable assumptions. What a ready-made template costs is rarely time, since the time most believers spend on it is modest, and the price is paid instead in the questions it closes and the errors it cannot correct. Also, the argument from scarcity has lost much of its force, as most people in wealthy modern societies have considerably more free time than the generations who lived on the edge of survival. And most importantly, reasoning is not a fixed cost we pay anew with every decision. A foundation of reasoning compounds over time, so careful thinking can gradually turn into intuition wherever we get honest feedback (which, in moral questions, often has to be built deliberately), until questions that once required effort start to answer themselves. In a society that taught these habits from early childhood, as naturally as it teaches language, weighing evidence could stop being a time-consuming chore and simply become the way people think. (Detailed explanation: In 2024, Americans spent on average about 8 minutes a day on religious and spiritual activities; fewer than one in ten took part on a given day, and those who did spent about an hour and a half (U.S. Bureau of Labor Statistics, 2025). Annual working hours of full-time workers in Britain, the United States, and Germany fell from roughly 2,750–3,300 in 1870 to about 1,450–1,900 by 2000 (Huberman & Minns, 2007). As explained in Waiting for Demise, practice turns into reliable intuition mainly where the environment is regular and gives clear feedback (Kahneman & Klein, 2009), and classroom training in critical thinking has so far shown modest average gains (Abrami et al., 2015).) How such an education might look is a question for Part 2.

Realistic optimism can even reclaim some of faith’s language, as long as it does so with open eyes. Then there are the moments in life that feel like fate, such as an unexpected opportunity or much-needed help in a dark period from people you didn’t expect it from, a chance encounter that changes everything or a sudden realization that reorders your priorities. We don’t need destiny to explain them, as most are some mixture of chance and preparedness. Still, deliberately calling such a moment fate can serve a useful purpose, since it marks the moment as rare and valuable, the kind of opportunity that may not come again, and so turns raw luck into momentum. Much like an open-label placebo, this lens may work even when we know exactly what it is.

This is not fate in the mystical sense, which holds that an outcome is destined regardless of what we do, as we saw when separating fate from determinism. It is a lens we choose, one that leaves our effort part of how things come to be.

Grief is the hardest test for any alternative to faith, since few of faith’s comforts run as deep as the promise of meeting the dead again. Without that promise, more remains than you might think, as the people you love continue in you, through memory as well as through the ways they shaped who you became. Their habits of speech and their values, their humor and the way they handled hardship, even the arguments you had with them, leave an imprint on your choices and your ways of thinking. The same holds for a beloved pet whose routines and personality became part of yours.

Part of them lives on in us regardless, because they helped shape the person who now carries them forward; acting in ways they would have appreciated only adds to that. But what remains of them in us is not them, and accepting that difference matters, whatever form our ongoing bond with them takes. (Detailed explanation: Most studies find that religious or spiritual beliefs go along with somewhat better adjustment to bereavement, although the evidence is methodologically weak (Becker et al., 2007). Resilience after loss is also far more common than once assumed: in many studies, half or more of bereaved people show little depression or distress (Bonanno, 2004). And grief research has moved away from the idea that healthy mourning means letting go: many people keep an ongoing inner bond with the dead [continuing bonds] (Klass et al., 1996), although whether such a bond helps or hinders adjustment seems to depend on its form and on the person (Stroebe et al., 2010).)

There’s also a less obvious comfort, one faith rarely mentions. When someone’s final years were marked more by pain than by joy, whether through illness, mental anguish or hardships that couldn’t be eased, the thought that their suffering has ended can bring real relief. (Detailed explanation: Among 217 family caregivers of people with dementia, 72 percent said the death came as a relief to them, more than 90 percent believed it was a relief to the patient, and caregivers’ depressive symptoms fell markedly within the following year (Schulz et al., 2003).) Unlike the judgment, purgatory or further rounds of rebirth that many traditions teach, nonexistence holds no further trials, and for those left behind it can even bring a peculiar kind of freedom. If we struggle in the years that follow, we don’t need to worry that the people we lost are watching us suffer, feeling helpless or disappointed. If, as the evidence suggests, they are simply gone, they are beyond the reach of suffering in the most complete sense available, which can let us focus on finding our own peace.

Facing mortality honestly also changes how we spend the time we still have, which is why grief without heaven still carries meaning, as it turns us toward the living.

If the dead don’t watch over us, the question remains what, if anything, deserves our reverence. Consider how humanity has related to the one life-giver it can actually observe. If anything in the physical world can be called the source of life on Earth, it is the Sun, and many cultures have worshipped it, some to this day. Our greatest progress, however, came from understanding it, as we learned to track its movements for our calendars and to harness its energy for our homes. Even toward the one massive life-giver we can verify, then, the path forward has been understanding and growing independence, not obedience, so there is little reason to expect a different path toward life-givers we can’t verify at all. What deserves something like the reverence once offered to the Sun is not a deity but humanity itself, or more specifically those within it who chose to pursue what is reliable. They are not gods, but taken together they are something like titans, and nearly everything we rely on today was built on their work.

A faith demands that reality bend to it, and whenever reality refuses, the doctrine retreats into reinterpretation while the person who leaned on it absorbs the blow. A self-inquired conviction, by contrast, bends to reality and grows stronger each time it does, which keeps it from hardening into a prison built from the beliefs of a past self.

Such a conviction must also be willing to face its own worst-case scenario, and I’d be a hypocrite if I asked this of believers without applying it here. What if this entire argument is wrong and a god does exist? We cannot know what such a being would want, any more than believers can. If such a god is benevolent in anything like our sense of the word, however, it is hard to imagine it condemning a life spent acting with compassion and improving the conditions of others. A god that is not benevolent raises once more the question of whether it deserves reverence, if not consideration at all. If that god rewards only the correct belief, we are back at the old wager, with nothing to tell us which god’s rules to follow. Should evidence for such a being ever emerge, a framework built on reasoning would update. Knowing what kind of conviction we want, however, doesn’t yet tell us how to build one. It has to be forged through reasoning and tested through action, and then maintained through the daily friction of living by it.