Neil deGrasse Tyson Is Wrong About Beliefs

The episode dismantles the claim—popular among prominent scientists like Neil deGrasse Tyson—that they “don’t believe anything,” arguing it is either trivially empty, semantically confused, or demonstrably false.

• Tyson has repeatedly said “I don’t believe anything” in interviews, and a NASA scientist on NPR told the host to “stop saying belief when you are talking to scientists,” insisting scientists only put belief in hypotheses.
• The host argues this posture is a kind of performative rationalism: a semaphore signaling “I’m more logical than religious folk,” rather than a coherent epistemological position.

Belief, in standard analytic philosophy, is simply a propositional attitude—holding some proposition to be true—and does not imply certainty, dogmatism, or faith.

• The Stanford Encyclopedia of Philosophy defines belief this way; it is compatible with updating on new evidence (Bayesian epistemology is built on this).
• Saying “I believe electrons exist but I’ll update if evidence changes” is perfectly rational and is exactly what belief means in philosophy.
• Claiming to have “no beliefs” requires redefining the word in a non-standard way, which is itself unscientific because it smuggles in unstated redefinitions.

The “no belief” claim often equivocates between belief that (propositional) and faith in (trust-based or resistant to counter-evidence), but this conflation doesn’t hold up.

• Philosophers distinguish belief that P from faith in X (a trust relation); Kierkegaard and Aquinas both treated faith as something beyond mere propositional assent.
• Even if a scientist means “I don’t have faith,” that’s a different claim from “I don’t have beliefs”—and the latter is what they’re actually asserting.
• Richard Swinburne’s Thomist view of faith involves propositions not fully seen or demonstrable, volitional trust, and resilience to counter-evidence—none of which maps onto the standard meaning of belief.

Assertion itself presupposes belief, so saying “I don’t have beliefs” is self-undermining.

• The belief-norm assertion in philosophy of language holds that you can only assert P if you believe P; therefore, asserting “I have no beliefs” already commits you to believing that you have no beliefs.
• When a scientist says “I hypothesize that consciousness is a universal principle,” this is shorthand for “I believe there’s a non-zero chance that consciousness is a universal principle”—the belief is implicit in the assertion.
• Thomas Campbell, who spent three hours arguing that consciousness is fundamental and fear produces ego, was asked to reconcile this with his claim that “belief is a problem”—he could not, because his entire position is built on beliefs.

Bas van Fraassen’s constructive empiricism—often cited to support the “no belief” view—does not eliminate belief; it merely narrows its scope.

• Van Fraassen argues scientists should accept theories as empirically adequate without committing to their truth about unobservables, but acceptance still involves belief in empirical adequacy.
• The host spoke directly with van Fraassen, who confirmed: “I believe many things. I believe that I had a father. I believe I wrote a book. Could I be wrong? Certainly.”
• Van Fraassen’s point is about epistemic modesty, not the elimination of belief from scientific practice.

The scientific method itself rests on beliefs: that evidence constrains theory, that modus ponens is valid, that the method is reliable.

• Quine’s “web of belief” places beliefs at the center of all knowledge, including scientific knowledge.
• Even radical uncertainty or fallibilism doesn’t eliminate belief—it just describes a type of belief that is held tentatively and updated with evidence.
• Requiring 99.9% credence before calling something a belief doesn’t help, because everyday actions (crossing the street, eating food) already require that level of confidence—so the scientist is still operating on beliefs.

The only logically coherent way to say “I hold no beliefs” is in a category-error sense—where belief is simply the wrong type of property to apply—but this doesn’t apply to any substantive scientific or philosophical claim.

• Asking whether belief applies to the electric charge of the universe is a category error; asking whether belief applies to whether electrons exist is not.
• The host’s verdict: scientists who say “I don’t have beliefs” are confused about their own cognition, engaging in linguistic posturing that ironically signals irrationality rather than rationality.
• The honorable statement is: “I believe X, and I’m willing to change my mind if the evidence warrants it”—not “I don’t believe anything.”