Early in my freshman year of college, a speeding car struck my twin brother, Jim, on a street near our campus. These were pre-cellphone days, but I happened to be in my dorm room when the call came in, so I got to ride with my brother in the ambulance. Our sister, Alice, who was in the year ahead of us, soon arrived at the hospital.

Shortly after the orderlies wheeled Jim away to be intubated, an intensive-care doctor explained to me and Alice that our brother was suffering from acute respiratory failure. This man, whom we’d never seen before, casually added that Jim was unlikely to make it to morning. Then he continued on his rounds. The first thing we did, once he’d left, was pray.

We’d been raised in a devout Catholic home, attending Mass every Sunday and on holy days of obligation, saying grace before meals, prayers before bed, and rosaries on long car rides, constantly adding sick or troubled loved ones to our intentions list. At the hospital, praying together was a distraction, but it was also an act that we believed to have some power to help our brother live through the night.

As it happens, he did live through it. His recovery was long—months stretching into years—but ultimately complete. I thanked God for that. But the memory of that first night, when I thought I was losing him forever, stayed with me. The recognition of radical human vulnerability pushes some people toward belief, but for me it had the opposite effect. On campus that spring, I started skipping Mass. This proved to be the initial step on a path that eventually led to my rejection of the faith in which I’d been raised. An answered prayer made me an atheist.

In many ways, those years—the turn of the twenty‑first century—were an ideal time to be a budding unbeliever. In 2004, an unknown writer named Sam Harris published “The End of Faith,” a short polemic on the existential threat that religion posed to Western civilization. In rapid succession, Richard Dawkins’s “The God Delusion” (2006), Daniel Dennett’s “Breaking the Spell” (2006), and Christopher Hitchens’s “God Is Not Great” (2007) followed Harris’s book onto best-seller lists, and the so-called Four Horsemen became the public face of a resurgent New Atheism. But I quickly discovered that I was not the audience for these books. I wasn’t looking to talk my way out of a belief in God—I was already out. I wanted to know what to believe in instead.

If I was still in search of beliefs, many atheists would object, I hadn’t really gotten over my religious upbringing. A good atheist deals not in faith but in facts, not in belief but in knowledge. Yet I could find no obvious factual, knowledge-based answer to the question that was most pressing to me: How am I to live?

I don’t mean to suggest that the New Atheists had no moral sense. On the contrary, they were largely fuelled by moral outrage at the needless suffering religion caused. But the nature of morality was seemingly the only thing about which they did not care to argue. They thought it simply self‑evident that we desire pleasure over pain for ourselves, and that any decent person wished the same for others. One of religion’s greatest harms, they believed, was that it turned people away from this basic intuition. Of the Four Horsemen, only Harris aspired to a “science of good and evil” which could subject moral claims to the same rational scrutiny as all other claims, but his chapter on the topic quickly devolves into an argument about the indefensibility of pacifism and the moral necessity of government torture. (It was a strange time.)

Anyway, I wasn’t really looking for practical guidance. To ask “How am I to live?” is to inquire as to not just what is right but what is good. It is to ask not just “What should I do?” but “How should I be?” The most generous interpretation of the New Atheist view on this question is that people ought to have the freedom to decide for themselves. On that, I agreed completely, but that left me right where I’d started, still in need of an answer.

Setting down the popular polemics of the day, I began to read modern philosophy, which I understood to be the primary means by which humans have sought secular answers to life’s questions. I read the philosophers most frequently cited as models by modern‑day atheists—John Locke, David Hume, John Stuart Mill—as well as those whom meaning‑hungry young people habitually embrace as secular gurus: Friedrich Nietzsche, Ludwig Wittgenstein, Albert Camus. But I also read philosophers who are mostly read just by other philosophers.

Even when I was struggling with the most challenging of these works, the reading felt urgent to me. I wasn’t submitting papers or getting grades; I wasn’t looking to earn a degree or to pursue a career. I wasn’t even trying to impress people at literary parties. (For that, I had thousand‑page postmodern novels.) I was just trying to figure things out. Immanuel Kant’s three “critiques” are often cited as the works that first made philosophy inaccessible to nonspecialists, but in Kant’s opinion he was addressing very straightforward questions—What can I know?, What must I do?, and What may I hope? I was decidedly a nonspecialist, and these were the questions I wanted answered.

Among other things, this reading taught me that atheists do hold beliefs, not just about morals and ethics but about how the world actually is and how humans fit into it. Of course, not all atheists hold the same beliefs—just as not all theists do—but I found that modern atheist belief tends to cluster into two broad traditions.

The most prevalent atheist world view goes by many names—empiricism, positivism, physicalism, naturalism—but the term that best captures the fullness of its present‑day iteration, as I see it, is scientific materialism. Roughly speaking, this view holds that the material world is all that exists, that humans can know this world through sense perception, that the methods of science allow us to convert the raw data of these perceptions into general principles, and that these principles can be both tested and put to practical use by making predictions about future events.

As world views go, scientific materialism has a lot to say for it. It tells us that humans are capable, without any supernatural aid, of coming to understand, and ultimately to master, all of reality. It tells us that the store of human knowledge is constantly increasing and continuously improving our material conditions. To this end, it points to the astonishing human progress that has occurred in the time of science’s reign. And it encourages us to enjoy the fruits of this progress as much as possible, since our life here on earth is the only one we’ll get.

Most people who subscribe to scientific materialism take it to be so obviously correct that it could not be denied by any rational person who truly understood it. But my reading showed me that this world view has its shortcomings. The most basic is perhaps inherent to any world view at all: it rests on a set of principles which often can’t be proven, even by the standards of proof the world view embraces. The general principle that all real knowledge is derived from sense perception of material facts cannot itself be derived from the perception of facts in the world, and thus can’t really be sanctioned by scientific materialism’s own methods. Indeed, no general principle can be. The very legitimacy of deriving general principles from the particulars of experience can never be established from experience without already having the principle in hand.

This so-called problem of induction was first identified not by any counter-Enlightenment reactionary but by the Scottish empiricist David Hume. Earlier empiricists like Locke and Francis Bacon believed that the physical sciences should still be grounded partly in metaphysical belief. Hume became one of modern atheism’s great intellectual heroes by rejecting this idea. But he didn’t substitute some other foundation in its place. Instead, he argued that we should simply do without foundations entirely, apart from the rather shaky ones of custom, habit, and expedience. That has been more or less the scientific-materialist answer to the problem ever since: scientific materialism just works.

If by “works” one means that it can be put to good use, this is unquestionably so. But, if we mean that it captures within its frame all the notable features of our experience, that’s a different matter. In fact, what materialism can’t adequately capture is experience itself. Consciousness is not material, not publicly available through sense perception, not subject to the kind of observation that scientific materialism takes as the hallmark of knowledge. By the standards of the materialist world view, it simply doesn’t exist. For me, this limitation proved fatal. I spent far too much time within the confines of my mind to accept a world view that told me whatever was going on in there wasn’t real.

Luckily, I’d by then come into contact with the other great family of modern atheist belief, which I eventually came to call romantic idealism. This is the atheism of Nietzsche and Martin Heidegger and their existentialist descendants, which begins in precisely the place where scientific materialism leaves off, with the will of the subjective, conscious agent. At its most extreme, romantic idealism treats each of us as willing our own world into being, creating the reality in which we live. Even when it does not go quite this far, it treats our subjective experience as the proper subject of knowledge, in fact the only thing we can ever be said to know.

Romantic idealism arose in the post‑Enlightenment era, and it grew in opposition to the principles of Enlightenment rationality as much as it did to religious authority. Although atheism is often associated with hyperrationality, this form of it is unapologetically irrational. In place of reason, observation, and scientific study, it valorizes emotion, imagination, and artistic creativity. The ethics of romantic idealism are an ethics of authenticity: the greatest good is not maximizing pleasure and minimizing pain but living in a way that is true to our subjective reality. The movement rejects religious belief not for being empirically false but for being a ready‑made and inherited response to existential problems that we must work out for ourselves. The appeal of this world view—particularly for a young person engaged in just such a working out—should be obvious, and I soon found myself in thrall to it.

Like scientific materialism, romantic idealism does not have a solid foundation in any provable universal truth. But it revels in this condition: it is the lack of any such foundation that makes it possible for each of us to construct our own truth. This relativism carries clear dangers. Since the time of Locke, empiricism has been closely linked with political liberalism, whereas romantic idealism is associated with rather darker political forces. Jean-Jacques Rousseau, one of the founders of Romanticism, was a great inspiration for the French Revolution’s Reign of Terror. He argued that liberalism’s supposed universal rights were covers for bourgeois self-interest. This argument was later developed at great length by Nietzsche, one of several thinkers in this tradition who inspired the rise of fascism.

A more basic problem with romantic idealism occurs on the personal level: building meaning from scratch turns out to be an incredibly difficult task. The romantic-idealist approach is fraught with fear and trembling, a fact it doesn’t deny. It is not a route to happiness; indeed, it seems to hold the goal of happiness in contempt. (“Mankind does not strive for happiness,” Nietzsche wrote. “Only the Englishman does that.”) Many romantic-idealist writers have been fascinated by the “problem” of suicide—the problem, in their view, being that there’s no good reason not to do it.

It was this element of romantic idealism which finally led to my rejection of it. I grew tired of being unhappy and anxious all the time, of constantly questioning whether life was worth the trouble. One cause of the feeling, for me, was that the materialists had it right on an important point: there is, indeed, a world outside our heads that cannot be ignored or overpowered through force of will, and denying this is a recipe for misery.

After nearly twenty years of searching unsuccessfully for a livable atheist world view, I began, in my mid-thirties, to entertain the possibility that atheism itself might be part of the problem. There were many steps from here to my eventual return to robust belief, but I started with the notion that for me the authentic life might be one of faith—one that recognized the existence of both the external material world and the internal ideational world and sought to reconcile them, and one that accepted an absolute foundation to things and attempted to understand, in some provisional and imperfect way, the nature of this foundation and what it wanted from me.

Roughly a decade has now passed since my return to theistic belief. Barring some extreme change, this second period of faith will soon have lasted longer than my years of atheist wandering. Yet those years will always be with me, and I’m grateful for them. They have made me understand the world around me, one lately marked by sorrow and despair, in a radically different way than I would have otherwise.

The signal predicament of our era is the global rise of illiberalism and intolerance. Secular liberals who have observed these forces with quite justified horror have often linked them to a familiar enemy: religion. In their view, Christian nationalism is one of the primary ideologies behind right‑wing illiberalism. Meanwhile, they tend to see the strain of left‑wing thought which rejects universalist liberal principles as equally under the sway of a kind of faith—the “cult” or “Church” of identity politics. (Richard Dawkins, of “The God Delusion,” speaks for many of them when he declared “woke” to be a “latter‑day Torquemadism . . . with its own religiously enforced dogma.”)

“Religion” is a famously malleable sociological category, but, if we stick to the criterion of theistic belief, the argument that modern‑day illiberalism is primarily a religious movement does not really hold up. Perhaps the crudest way to make the point would be simply to note that the rise of illiberalism has gone hand in hand with a decline of theistic belief and religious practice—both in the United States and around the world. The avatar of American illiberalism, Donald Trump, is the first President in generations who does not even pretend to be influenced or motivated by Christian faith. Trump is best understood as our first Nietzschean President, a man who explicitly embraces the will to power as the ultimate value, a force to which even the truth must give way.

Many of the young and highly educated cohorts who populate the portions of the left most suspicious of universal liberal values are also among those least likely to identify as religious believers. In a different manner than Trump, they hold that so-called objective “truths” are the expression of power dynamics—tools used by the élite to oppress the marginalized. In place of these truths, they champion the importance of identity and authenticity. In other words, large groups of both the left and the right have become romantic idealists, and they have come to pose the same challenge to liberalism and scientific rationalism that romantic idealism has always posed to these traditions.

Meanwhile, the failure of these traditions to respond adequately to the challenge is bound up with the problem identified by their earliest proponents: they have a very hard time articulating their foundational justification. When liberalism runs smoothly, it does a remarkable job delivering the goods it promises. For most people, this is a sufficient achievement to quiet any worries about its philosophical underpinnings. But when many people within liberal societies do not feel that the system is working, when the practical case for liberalism comes into question, secular liberals don’t have much else to go on.

For early liberals like Locke, it seemed obvious that liberalism—like empiricism—needed to be grounded in faith, even as it sought to enshrine tolerance for different varieties of belief. For Locke, the fact that we have immortal souls subject to eternal punishment and reward means that it’s irrational to submit to a government that denies God’s will as we understand it, no matter how much coercive power that government has. At the same time, Locke had the empiricist’s healthy suspicion that we could never have metaphysical certainty about what the Creator’s will was, which meant that no person should impose his answer to that question on another. It is for these reasons that faith must be treated as a matter of personal conscience, but also more generally that a regime grounded in a social contract must be one that respects individual freedoms. Our status as creatures of God confers on us certain rights that can’t be handed over as part of the social contract, rights that are at once natural and inalienable.

Many of the Founding Fathers held this same view. But, after Hume, liberals came increasingly to find even the barest invocation of metaphysical principles to be an embarrassment. The great post-Humean liberal theorist Jeremy Bentham called the idea of natural rights “nonsense,” and the idea of inalienable natural rights “nonsense on stilts.” Some liberals have tried to hold on to such abstract concepts without the metaphysical framework in which they make sense, but many eventually came to the Humean conclusion that liberalism could simply do without foundations, so long as it got the job done.

My own passage into and back out of unbelief—one marked by a close reading of works that earlier illiberal societies had attempted to suppress on religious grounds—has strengthened my liberal commitments. But it’s also made me acutely aware of liberalism’s very real limitations. As a means for allowing people with different conceptions of the good to live together fruitfully and peacefully, liberalism seems unmatched in human history. As a means of generating its own conceptions of goodness that feel compelling to most human beings, its record is quite mixed, perhaps because that’s not what it was ever designed to do.

What’s more, when liberals treat some version of scientific materialism as so self-evidently true that it must serve as the default context for public discourse; when they make allegiance to “reason” and “evidence,” as they define these terms, the price of admission into such discourse; and when they attempt to banish metaphysical or spiritual or even frankly religious talk from our politics and our culture, they are not practicing liberalism as its greatest exemplars understood it. They are eliminating from our shared vocabulary many of the concepts on which any justification for liberalism beyond the purely practical would have to depend.

I am not suggesting that the solution to our problems is for secular liberals to find God. But I am suggesting that religious believers be considered natural allies in the fight against irrational illiberalism, rather than its primary cause. This need not mean abandoning secularism, and it certainly doesn’t mean abandoning liberalism. It means, perhaps, seeing liberalism as so many liberals wish believers would see their faith: not as the expression of a universal truth to which every person must eventually submit but as a human construction—one of the finest we have ever made—worth defending even when it is helpless to defend itself, yet capable of being swept away by the same hands that built it in the first place. ♦

This is drawn from “Why I Am Not an Atheist: The Confessions of a Skeptical Believer.”