³Õºº¾ãÀÖ²¿

I have taught philosophy of religion for many years, and more times than I can count I have had students ask, “Where is the proof that God exists?”

In my classes, I analyze arguments by putting them in the form of numbered steps:

Atheism Is Reasonable
(1) It is not certain that God exists. [Premise]
(2) Therefore, it is reasonable not to believe that God exists. [1]
(The “[1]” indicates that the claim in (2) is taken to follow from the claim in line (1).)

One sees similar arguments on non-religious topics. For example, faced with the prospect of catastrophic consequences from climate change, the first line of defense is often to insist that the science is not certain.

Raising doubts about science has a long history. Despite separation of church and state, many have proposed teaching creationism or “intelligent design” in high school classrooms. These proposals are usually offered on the basis of a seemingly modest claim: the theory of evolution is just that, a theory, so we should teach students both evolution and creationism and let the students make up their own minds. In other words: we aren’t certain that evolution is true, so it is reasonable to teach the creationist alternative.

All of these are terrible arguments. The conclusion simply does not follow from the premise—in philosophical terms, none of the arguments are valid. To make the conclusion logically follow, we would need an intermediate premise, something like this: For any claim, if the claim is not known with certainty, then it is reasonable not to believe that claim. Now the argument is valid, meaning that the conclusion does logically follow from the premises.

According to Merriam-Webster, something is certain if it is “known or proved to be true: indisputable.” It is foolish to insist that facts about God or evolution are proven or indisputable in this sense. Instead of answering the rhetorical question with overconfident bluster, the better reply is to reject the intermediate premise. Very little is known with certainty, but that does not mean that it is reasonable to refuse to believe anything that is not proven.

Consider a simple example: Was Ronald Reagan ever president? Of course—many of us are old enough to remember his elections. We can also read historical accounts about his two terms. But can we prove that he was president?

Suppose someone were to claim that, after finishing his acting career and serving as governor of California, Reagan died, but a similar-looking person was sent onto the national scene. Of course, the suggestion sounds and is absurd (though I have heard equally absurd claims about Paul McCartney). But can we prove that this didn’t happen? Is it indisputable? No.

So we could propose the following argument:

Ronald Reagan Was Never President
(1) It is not known with certainty that Ronald Reagan was president. [Premise]
(2) It is reasonable not to believe anything that is not known with certainty. [Premise]
(3) Therefore, it is reasonable not to believe that Ronald Reagan was president. [1,2]

The conclusion is silly. Whatever else you think about Ronald Reagan, you should believe he was president. And yet the conclusion does follow from its premises, and the first premise is true. The lesson is that the intermediate premise is false: just because a claim is not known with certainty does not mean that it is reasonable not to believe that claim. We do not know with absolute certainty that Ronald Reagan was president, but you are being unreasonable if you fail to believe it.

However, it was the same intermediate premise that implicitly did the argumentative work in the arguments about evolution or the existence of God. So those arguments are also equally bad, despite being suggested with great sincerity and fervor by ³Õºº¾ãÀÖ²¿ students and even elected representatives.

We could avoid such bad arguments if we simply stopped talking about certainty or proof, at least in contexts like these. It is probably okay to feel certain that your mother loves you, and it is fine to talk about mathematical proofs. It would also be fine if, instead of simply asking what we can know with certainty, we spoke very carefully about degrees of certainty.

Exactly how certain should we be about matters like evolution or God, or climate change? What degree of certainty is required before we can say that withholding belief is not reasonable? Those are hard questions with no simple answers. But loose talk of certainty or proof gives us the illusion that we can replace the hard questions with easy ones. That just leads to bad arguments and, ironically, often leads you to feel certain you are right when you are not.

Scott Sehon is the Joseph E. Merrill Professor of Philosophy. He earned his AB at Harvard and his PhD at Princeton. Much of his scholarship concerns the philosophy of mind and action, especially free will.

This story first appeared in the Spring 2025 issue of ³Õºº¾ãÀÖ²¿ Magazine. Manage your subscription and see other stories from the magazine on the ³Õºº¾ãÀÖ²¿ Magazine website.