Religious claims should not be accepted as true without sufficient evidence. Different religions make contradictory claims about gods, revelation, and salvation, so they cannot all be correct. Without reliable evidence establishing one religion's supernatural claims, withholding belief is more rational than treating those claims as established facts.

You're broadening “religion” from truth claims to practices and metaphors, but those aren't what makes a religion true. I don't need to prove every ritual, story, or metaphor false. The relevant question is whether the religion's supernatural truth claims are true. A belief surviving hardship or inspiring commitment demonstrates conviction, not truth, because people can sincerely sacrifice for false beliefs. If your defense is that religion contains useful practices and meaningful metaphors, that's compatible with religion being factually false. Usefulness, cultural importance, and sincerity don't establish the truth of supernatural claims.

You're attacking a stronger claim than necessary. “Religion is false” doesn't require every ritual, metaphor, or practice to be false, because those things aren't necessarily truth claims. Your science analogy actually supports my point: we evaluate each hypothesis by its evidence. Likewise, religious claims about gods, revelations, miracles, and supernatural events can be evaluated individually. If those central factual claims lack sufficient evidence, I can reject them without claiming that every religious practice or metaphor is meaningless. And people sacrificing their lives proves conviction, not truth. People have made enormous sacrifices for mutually contradictory beliefs, so commitment cannot establish which belief is actually true.

You're conflating “not proven false” with “true enough to accept,” and that's the contradiction I want to press. You say unfalsifiable religious claims are merely “off the table,” yet you still defend religion as a meaningful system without establishing its supernatural claims. I don't need to prove every unfalsifiable claim false; I can reject a specific claim when its evidence is insufficient, while remaining agnostic about claims that genuinely cannot be evaluated. So either you're defending religion's truth claims, in which case you owe evidence for them, or you're retreating to practices and metaphors, in which case you've conceded those don't establish religion's supernatural truth. Agnosticism about a particular unfalsifiable claim is compatible with rejecting unsupported religious claims generally. “You can't disprove it” establishes possibility, not truth.

Your mountain analogy actually exposes the flaw in your argument. “No successful climb has been demonstrated” does not prove nobody climbed it, but if the claim is “someone climbed it,” lack of evidence gives us no reason to accept it. Likewise, if specific religions make specific supernatural claims, “you haven't disproved every religion” does not give those claims positive evidential support. You also keep switching between “religion” as every cultural practice and metaphor and religion as factual supernatural claims. I don't need to prove that prayer, ritual, art, or metaphor is false because those aren't necessarily truth claims. The proposition I can reject is the factual claim that a particular god exists, revealed a particular scripture, performed particular miracles, or guarantees particular supernatural outcomes. And your “one true religion, therefore not zero” argument is a false dichotomy: the possibilities include one, several, or none. Religious diversity is enormous, with Pew's 2026 analysis identifying seven major categories across 201 countries and territories, while even within traditions there are major disagreements about doctrine. Most importantly, you're demanding disproof when the rational standard for belief is evidence proportional to the claim. An unfalsifiable claim isn't automatically false, but it also isn't entitled to belief merely because it cannot be falsified. Philosophical accounts of falsifiability specifically distinguish claims that can conflict with possible observations from claims insulated from such testing.