In educated, liberal circles today, the conventional wisdom about science and religion is that they are compatible. Each belongs to a different sphere. In one popular formulation, natural science produces naturalistic explanations of natural phenomena, while the sphere of religion is knowledge about an entirely different realm, the supernatural. Or, possibly, science is about a narrow subset of reality that can be empirically probed by a “scientific method,” while other kinds of knowledge, such as morality, are in the domain of religion.

None of this is actually true. Science and religion are both ambitious enterprises with a habit of overstepping boundaries. But the notion of separate spheres is politically very convenient as a way of diffusing tensions between these ambitious enterprises. There is always a temptation to discover that what is politically convenient is inscribed into the nature of things. It would be easier to keep the peace if the very nature of science, properly done, prevented science from offending religious sensibilities. It would be best if the nature of religion, properly understood, prevented it from interfering with investigations of nature. If we believe this to be the case, our behavior would be channeled in politically safe directions. And yet, the notion of separate spheres is almost certainly mistaken.

Nonetheless, let me explore how a defense of separate spheres on political grounds might work. Perhaps publicly, scientists and others interested in the continued flourishing of the scientific enterprise should endorse the compatibility of science and religion. If they find themselves unable to speak in favor of such an intellectually weak position, they should then shut up and let others be the public face for science. We want Francis Collins rather than Richard Dawkins to represent science to the public, not in spite of but because of the fact that Collins has produced some seriously bad arguments for religious belief.

I will make two main assumptions. Both come down to expecting that current trends will continue.

In science, I expect that a broadly naturalistic approach will continue to be successful. Supernatural agents will continue to be out of place in our best explanations. If we ever obtain solid evidence for miracles, psychic powers, quantum healing, or intelligent design, the issue of compatibility will become moot. Science will not just be compatible with, but supportive of, the notion of supernatural agency. I very much doubt any of this will ever happen.

In religion, I expect the continued success of the more doctrinally conservative, strongly supernaturalistic, and populist forms of faith. This will translate into increased political power, putting more pressure on science to adapt to a political environment shaped by religion. If secularization turns out to apply to regions other than Western Europe with its unique history, or if fundamentalist and charismatic styles of religion suddenly go into retreat in the rest of the world, the pressure will come off and the issue will again become moot. I see no reason to think anything like this is going to happen, though I acknowledge that in observing social developments, it would be surprising not to run into surprises.

In that case, science will remain a thoroughly secular enterprise, where religion does not enjoy much influence outside of private convictions. Science and technology are examples of where Enlightenment hopes for human progress have been most unambiguously realized. Therefore secular liberals, including liberal religionists, will continue to have plenty of motivation to keep science healthy. Religious nonbelievers will have an extra motivation to favor a healthy scientific enterprise, since the results of science will probably continue to intellectually support nonbelief.

But a world where the modern forms of conservative religion add to their already impressive political successes is one where scientific institutions will have to tread more carefully. In that case, preserving the secular ambience and skeptical flavor of science will call for not attracting public attention to this skepticism. In conditions where secularism as a more general political position has been defeated, defenses of pockets of secularity will require not overly publicizing this secularity, let alone nonbelief. Presenting a quasi-official scientific view of separate spheres will be useful, particularly if this is, as is usually the case, a liberal theological position. The liberal bit might still arouse suspicion, but at least science will be more positively associated with faith.

The nature of this defeat is also important. Conservative religion has succeeded democratically, as a popular and populist force. Ordinary citizens, from all kinds of backgrounds and social classes, can be mobilized on the basis of piety. For most people, public morality is inseparable from religion. Some variety of acceptable faith is a requirement for political leaders and a legitimator of political regimes. Aside from some anomalous and minority cases such as Europeans, religion is the default mode for how humans do politics. And the most successful faiths are the more miracle-mongering, revelation-peddling, intensely supernatural sorts of religions.

Pious people—modern, politically active and technologically savvy people; not peasants—will naturally demand that publicly supported activities, such as science and education, support their religious convictions. At the least, they will not accept science and education undermining their faith. An impression of neutrality might be acceptable, especially since modern religious environment are pluralistic. And even an irredeemably elitist enterprise such as science can be tolerated in a popular democracy, as long as it has no great political or cultural power. Science will presumably continue to make itself useful for commercial and military endeavors, which gives it some more insulation from cultural politics. The persistent confusion of science with technology will also help, since people favor technology and this rubs off on science. But even so, science will be in a precarious position if it is perceived to harbor nonbelief. Conservatives, and especially religious conservatives, have already shown that they are quite willing to attack science if science appears to give political or religious offense. In a position of weakness, science will have to be much more careful about not giving offense, or at least not publicizing it.

In fact, the continued success of conservative religion will probably result in science having to retreat further than what we are accustomed to today. For example, I expect that creationism will eventually appear in secondary education. There is, indeed, no legitimate way to stop this—education is a political matter. What we might still be able to do is keep creationism out of mainstream natural science, though not applied science. It will not be a disaster; we will learn to live with it. After all, creationism is already a considerable popular success, which has very little effect on how the scientific community conducts its business. Science education is already something of a failure, especially in terms of public science literacy. Let’s be honest, students hate science, and the vast majority are in school to be warehoused or to be molded into future corporate drones anyway. We still have more than enough talent coming to us to keep science functioning very well.

The notion of separate spheres has already been doing good political work for a very long time in the United States, helping isolate science from religious pressure. It is already the conventional wisdom. So scientists will not have to learn anything new, never mind spending a lot of effort adapting to post-secular politics. In conditions where secularism has been more thoroughly defeated, it may well serve to limit the damage.