Last updated on October 31st, 2026 at 10:50 pm
The relations between science and religion in the modern world are calmer than the culture war suggests, but only if you ask the right people. Put the question to the general public and most will say the two are at odds. Put it to believers and scientists about their own lives and the fight mostly disappears.
After checking the main surveys, court-of-history episodes and philosophical arguments, my short answer is this: the conflict is real on a handful of specific questions, such as human origins, and largely imagined everywhere else.
The newest twist arrived in 2025 and 2026. Churches are no longer only defending themselves against scientific claims. They are now cross-examining the people who build artificial intelligence.
What the evidence says
Perception beats reality. In a Pew survey, 59% of Americans said science and religion often conflict, yet only 30% said their own beliefs conflict with science.
The Galileo story is half history, half myth. The Vatican admitted error in 1992, and John Paul II himself called the case a “kind of myth” about the Church and science.
Scientists are less hostile than you think. A multi-country study of more than 22,000 biologists and physicists found most scientists see science and religion as separate spheres, not rivals.
Four models cover most positions: conflict, independence, dialogue and integration. Each has serious defenders and serious critics.
Religion is not vanishing. The religiously unaffiliated reached 1.9 billion people, but Christians and Muslims still make up more than half the world between them.
The 21st-century fault line is technology. Two Vatican documents on AI, from January 2025 and May 2026, show religion questioning science’s applications rather than its findings.
Do science and religion really conflict?
Most Americans say yes, but far fewer say it about themselves. A Pew Research Center report built on a 2014 survey of 2,002 adults found that 59% of Americans say science and religion are often in conflict, while 38% say the two are mostly compatible.
Ask the same people about their own faith and the number drops. Just 30% said their personal religious beliefs conflict with science, and 68% said they do not.
That gap is the whole story in miniature. Cary Funk, who co-wrote the report, put it this way: “Our perceptions of others are often different than our perceptions of ourselves”. In other words, many of us imagine a war between two camps while living peacefully inside our own.
Surprisingly, the people who feel the war most strongly are often outside religion. Among religiously unaffiliated adults, 76% said science and religion often conflict, but only 16% of them said their own beliefs conflict with science. Weekly worshippers were less sure about a war: 50% of highly religious adults saw conflict, which is still a lot, but well below the unaffiliated figure.
Where do believers who do feel a clash find it? Among the 30% who see a conflict, the most common source is the creation of the universe and evolution. That detail matters, because it shows the dispute is narrower than the slogan “science versus religion” implies.
One honest caveat before we go on. These Pew numbers are more than a decade old, and I could not find a newer Pew survey that repeats the same question. Treat them as a well-built baseline, not a 2026 reading.
Where did the war story come from?
It mostly came from Galileo. In 1633, at age 69, he was forced by the Roman Inquisition to repent for defending a sun-centred cosmos. For centuries that scene stood for religion crushing science.
The Church’s own correction came much later, and it came slowly. Pope John Paul II asked a commission to re-examine the trial in 1979, at the centenary of Einstein’s birth. On 31 October 1992, he addressed the Pontifical Academy of Sciences and the commission’s findings were presented.
Press coverage at the time reported that the Church had admitted to erring for 359 years. The commission chairman, Cardinal Paul Poupard, described the judges’ mistake as a “subjective error of judgment”.
The pope’s reading of the affair was more interesting than the headline. He said Galileo’s accusers had taken a question of fact that belongs to scientific investigation and moved it into the realm of doctrine. He also argued that “The Galileo case has become a kind of myth”, a symbol of the Church’s supposed rejection of scientific progress.
Historians still argue over how far that myth should be trimmed. One history blog that leans on the Galileo scholar Maurice Finocchiaro’s book Retrying Galileo argues that the 1633 trial was a mainly political affair that barely touched on science. That is an argument, not a settled fact, so I would not repeat it without checking Finocchiaro directly.
What I can say with confidence is narrower: the 1992 address was the end of a process, not a sudden confession, and it shows an institution revising its own account over more than a decade.
I am writing this on 31 October 2026, exactly 34 years after that speech. The date feels fitting, because the question it raised is still open. Who decides where a scientific claim ends and a doctrinal one begins?
What are the main ways thinkers relate science and religion?
Four, according to the most influential map in the field. The physicist-theologian Ian Barbour proposed conflict, independence, dialogue and integration.
Under conflict, the two compete on the same turf and one must win. Independence says each has its own domain. Dialogue treats them as conversation partners. Integration goes further and tries to fold both into one picture.
The map is useful, yet it has critics. One article in the journal Zygon argues that Barbour’s four options carry hidden values, including a belief in steady progress from conflict toward dialogue and integration. I find that objection fair. Barbour is describing the field, but he is also quietly steering you toward the friendly end of it.
Here is how the best-known voices line up.
| Position | Voice | Core claim |
|---|---|---|
| Conflict | Richard Dawkins | Belief in God is a scientific question, and the evidence says almost certainly no |
| Independence | Stephen Jay Gould | Science covers facts, religion covers values and meaning |
| Dialogue and integration | John Polkinghorne | Science and theology are both honest searches for truth and can inform each other |
| Redrawn conflict | Alvin Plantinga | Real conflict is between science and naturalism, not science and theism |
Dawkins and the conflict view
Dawkins’s 2006 book The God Delusion argues that a supernatural creator almost certainly does not exist. Its sharpest claim is methodological. Dawkins maintains that God’s existence or non-existence is a scientific fact about the universe, discoverable in principle if not in practice.
Whether you agree or not, that claim matters because it rejects the polite truce. If God is a factual hypothesis, science is allowed to weigh in.
Gould and the truce
Gould published his answer in 1997 and expanded it in the 1999 book Rocks of Ages. His idea, non-overlapping magisteria, says science and religion occupy different areas of inquiry, fact versus values. A reviewer for The Irish Times summed up his slogan neatly: “science studies how the heavens go, religion how to go to heaven.”
Both flanks attacked it. Dawkins replied that religions make existence claims, and those are scientific claims.
A review in American Scientist argued that the membrane between the two domains is semipermeable, because science has plenty to say about the “what” behind religion’s “why” questions. A 2024 paper adds a twist: NOMA is usually filed under independence, yet Gould also allowed dialogue and integration at the level of the individual person.
Polkinghorne and the dialogue view
Polkinghorne makes a rare case study because he lived both lives. He was Professor of Mathematical Physics at Cambridge from 1968 to 1979, then resigned to train for the priesthood. He later won the 2002 Templeton Prize and died on 9 March 2021.
His obituarist in the Church Times noted that he argued critical realism was the right way to understand both science and theology. His credibility came from working at the highest level of physics, which made sceptical colleagues listen.
Plantinga and the redrawn conflict
Alvin Plantinga flips the usual question. In his own summary, he argues that there is no serious conflict between current science and Christian belief, and that evolutionary theory is neutral between guided and unguided evolution. The real clash, he says, is between science and naturalism, because naturalism combined with evolutionary theory is self-defeating.
His book’s short version: superficial conflict but deep concord with theism, and superficial concord but deep conflict with naturalism.
I think the first half of that argument is his strongest. Science, as a method, cannot tell you whether the process behind evolution was guided. That is a philosophical add-on, whichever side makes it. The second half is more contested.
A review by Richard Smith criticises his case for deep concord between science and theism and his frontal assault on naturalism, and other critics say calling naturalism a “quasi-religion” may hide the theological assumptions in his own critique.
Do scientists think science kills faith?
Mostly no. The best evidence comes from Rice University’s Religion Among Scientists in International Context project, described by its funder as the first empirical study of religion among scientists across regions, covering over 22,525 biologists and physicists. The eight regions were France, Hong Kong, India, Italy, Taiwan, Turkey, the United Kingdom and the United States.
The headline finding was blunt. Scientists do not think science is a secularizing influence, and most see religion and science as operating in separate spheres.
The detail is more interesting than the headline. In the UK, 65% of surveyed scientists said they were nonreligious, compared with 50% of the general population, yet 49% of the same UK scientists acknowledged basic truths in many religions.
In contrast, Indian scientists proved significantly more religious than British scientists. The lead researcher said this challenged the idea that scientists are universal carriers of secularization.
For me, this is the strongest single argument against the “science kills faith” story. If science itself corroded belief, you would expect the same effect everywhere. Instead, the scientists’ own religiosity tracks their national culture.
Is religion actually fading in the 21st century?
Not in the global total, though the mix is changing fast. Pew’s 2025 report on 2010 to 2020 found that Muslims made up 25.6% of the world and Christians 28.8%. The religiously unaffiliated grew by 270 million to 1.9 billion, or 24.2% of the world, which made them the third-largest group.
Do not read “unaffiliated” as “atheist.” A separate Pew analysis found that many religiously unaffiliated adults still believe in God, including 77% in South Africa and 92% in Brazil. Leaving organised religion and leaving belief are different moves.
The American numbers on human origins show the same slow drift. A Gallup poll of 1,024 adults in May 2024 found 37% say God created humans in their present form, 34% say humans evolved with God’s guidance, and 24% say humans evolved with no divine role. Gallup noted that creationism and God-guided evolution peaked at 47% and 40% in 1999 and have trended downward since.
Views on Human Origins
Percentage of U.S. adults, May 2024. Response labels are paraphrased.
- God created humans in their present form — 37%
- Evolution guided by God — 34%
- Evolution without God’s involvement — 24%
- No opinion — 5%
The creation response specifies creation within approximately the past 10,000 years. These results measure public beliefs.
Put plainly, the middle position is large, and the strict secular view is growing from a low base.
Why does it shift? Charles Taylor’s warning
It is tempting to say science caused all of this. The philosopher Charles Taylor disagreed. In A Secular Age (2007), he distinguishes three senses of secularity, and a summary of his book says he critiques the simple story that religion declined because science and rationality rose.
One admiring reviewer reads Taylor as arguing that it is not “science” or “Darwinism” that drives the change, but a moral narrative that was already present in Christianity.
I find that persuasive, though I would not call it proven. It suggests that when someone leaves a church, the lab coat is rarely the reason. The more common cause is a story about human dignity, freedom and self-direction that no longer needs God as its anchor.
The original version of this article also leaned on Martha Nussbaum’s The Fragility of Goodness as a science-and-religion source. I could not verify that the book discusses that relationship, so I have not carried the claim over.
Where do science and religion still collide?
In three places, and each one is narrower than the slogan.
Human origins. This is where the survey conflict lives, and it is a conflict over interpretation more than data. Plantinga’s point stands that evolutionary biology as a science does not decide whether a divine hand guided the process. Dawkins’s counterpoint also stands: for strict literalist readings of scripture, the evidence leaves little room.
Cosmic beginnings. The Big Bang is a good test case, because its first formulation came from a priest. Georges Lemaître was ordained in 1923 and proposed the “hypothesis of the primeval atom,” now regarded as the first formulation of Big Bang theory. He also first derived the relationship now called Hubble’s law and published the first estimate of the Hubble constant in 1927, two years before Hubble.
In 2018, the International Astronomical Union voted to rename it the Hubble-Lemaître law.
The religious reaction was not always wise. Pope Pius XII proclaimed in 1952 that the Big Bang affirmed a transcendental creator, a move that blurred the very line Galileo’s case had taught the Church to respect. A theory that is proven wrong would, by that logic, wound the faith. Science has no obligation to protect anyone’s theology, and the lesson cuts both ways.
Environmental policy. Here religion stood with the scientists. Pope Francis’s 2015 encyclical Laudato Si’ backed scientists who say climate change is real and mostly caused by humans. The United Nations Environment Programme welcomed it, saying our response to environmental degradation cannot be defined only by science, technology or economics but is also a moral imperative.
That is a clear example of the dialogue model: science supplies the diagnosis, religion supplies moral urgency.
Collins offers a living example of the same pattern in genetics. Francis Collins led the Human Genome Project, and in his 2006 bestseller The Language of God he argued that science is compatible with faith, properly understood. He later founded the BioLogos organisation in November 2007 to improve the science-and-faith conversation in America.
Who is questioning whom in the age of AI?
Here is the genuinely new development. For most of the modern period, scientists asked religion to justify itself. In 2025 and 2026, the Catholic Church asked the technology sector to do the same.
In January 2025, the Vatican released Antiqua et Nova, a note on AI and human intelligence. It was approved by Pope Francis on 14 January 2025 and published on 28 January.
The note argues that AI should be used only as a tool to complement human intelligence, not to replace its richness, and it calls autonomous lethal weapons a cause for grave ethical concern. It also says machine learning is fundamentally different from the developmental growth of human intelligence.
Then came a bigger step. Pope Leo XIV signed his first encyclical, Magnifica Humanitas, on 15 May 2026, and the Vatican released it on 25 May. He called for robust regulation of AI and for developers to work for the common good rather than profit. At the launch, he summed up his stance in six words: “Artificial intelligence needs to be disarmed.”
Religion News Service reported that the document warns AI risks widening inequality, weakening democracy and undermining what it means to be human.
Does the public agree? On faith specifically, largely yes. A June 2025 Pew survey of 5,023 Americans found that 73% say AI should play no role at all in advising people about their faith in God. The American Bible Society’s 2024 survey of about 2,500 people found that 68% do not think AI could help with their spiritual practices.
Notice what has changed. The debate is no longer “does the theory of evolution disprove scripture?” It is “who is responsible when a machine imitates a human face and voice?” That is an ethical and philosophical question, which is exactly the territory Gould reserved for religion, and also exactly where Barbour’s dialogue model has the most to do.
Where do I land after checking all this?
My own view, which you are free to dispute: the most defensible position is a modest one. Science is excellent at establishing how the material world works, and its findings should win when they conflict with a factual claim about that world. Religion, philosophy and ethics are where questions about meaning, obligation and the limits of technology get argued. The hard cases live on the border, and the border is blurrier than Gould wanted.
That position has real opponents.
A committed naturalist in Dawkins’s camp would say I am too generous, because religious claims about miracles and creation are factual and deserve factual scrutiny.
A committed believer in Plantinga’s camp would say I am too quick to grant naturalism the status of a neutral default. Both objections are serious, and I do not think either has been defeated.
Five lessons this debate teaches
My most important takeaway is that “science versus religion” is usually the wrong question. The better question is which specific claim is in dispute and what kind of evidence could settle it. The problem it solves is polarisation: people choose a team, then defend every claim the team makes.
1. Test the claim, not the camp. The Pew gap shows people judge the other side by stereotype. Before you take a side, name the exact statement in dispute, such as the age of the Earth or the cause of warming.
2. Ask what kind of question it is. Gould’s how-versus-why split helps, but Dawkins is right that some religious claims are factual. Sort your question first, then pick your tools.
3. Distrust your own side’s origin story. The Galileo case is a useful warning. Even John Paul II called the popular version a myth, and the full record is messier than either camp’s retelling.
4. Separate the finding from the moral conclusion. Laudato Si’ shows how a religious leader can accept the climate science and then argue for ethical action. You can accept a result without accepting every policy attached to it.
5. Use both lenses on new technology. AI needs data about what it does and a moral account of what it is for. The 2025 Pew result suggests most people already want human guidance, not a machine’s, on questions of faith and meaning.
Frequently asked questions
Are science and religion compatible?
It depends on the religion, the claim and the person asking. Many scientists say yes. In the Rice study, most scientists said they see science and religion as operating in separate spheres. Specific clashes, such as literal readings of creation, remain.
Did the Catholic Church apologise to Galileo?
Sort of. In 1992, John Paul II accepted a commission’s report and the Vatican acknowledged that Galileo’s judges had erred. Whether to call it an apology is a matter of dispute, since historians note the process ran from 1979 and the language was careful.
What are Barbour’s four models?
Conflict, independence, dialogue and integration. They describe whether science and religion compete, stay separate, converse or merge.
Is the world becoming less religious?
In some countries, yes, but globally the picture is mixed. Christians still made up 28.8% and Muslims 25.6% of the world in Pew’s 2025 report, and the unaffiliated hit 24.2%.
Conclusion
The relations between science and religion are not a single war, and they never were. They are a set of specific disputes, a few real and many inflated, wrapped in a story that people on both sides find convenient.
The data I could verify says the public imagines more conflict than they live, scientists mostly do not see science as a threat to belief, and the sharpest new debates are about technology rather than theology.
If you read one idea away from this piece, make it this: when someone tells you science and religion cannot coexist, ask them which claim they mean. The answer usually shrinks the argument to something you can actually weigh.
Anyone interested in philosophy, history of ideas, or the ethics of AI will find that exercise worth the effort, and I would recommend starting with Barbour’s four models and Gould’s essay, then reading Plantinga and Dawkins against each other.
