What this covers

Julia Galef joins Sean Carroll to defend rationality itself—not as an academic abstraction but as a practical stance toward reality that, she argues, serves people better than the alternatives. The conversation centers on the distinction between two mindsets: the "soldier" who defends existing beliefs, and the scout mindset, which seeks to see the world accurately. Galef's core claim is that self-deception rarely delivers the promised payoffs of happiness or success; in nearly every case where someone claims you must deceive yourself to thrive, an honest third option exists that achieves the same goal. The discussion ranges across why this counterintuitive idea matters, what blocks people from adopting it, and what practical habits distinguish those who actually do.

The conversation covers substantial ground on how belief and identity intertwine, with examples running from Governor Earl Warren's reasoning about Japanese-American internment to Dan Kahan's Graph of Despair, which shows educated people diverging sharply on factual questions. Galef emphasizes that intelligence and expertise offer no immunity to motivated reasoning; they may even amplify it. She introduces the concept of Planets of Belief—networks of mutually reinforcing assumptions—to explain why changing minds is so difficult even between honest inquirers. The pair also probe when Bayesian reasoning breaks down in practice, how the difference between epistemic vs instrumental rationality fuels confusion about whether self-deception works, and the problem of identity capture, where disagreement feels like a personal attack rather than factual correction. Throughout, Galef distinguishes between holding beliefs strongly and holding identity tightly—a practical reframing that opens paths to both clarity and productive collaboration.

Sharpest takeaway

Galef argues that adopting a 'scout mindset'—seeking to see reality accurately rather than defending pre-existing beliefs—is both more rational and, contrary to popular claims, generally more effective at achieving one's goals than the self-deceiving 'soldier mindset'.

  • Intelligence and knowledge do not protect against motivated reasoning and may amplify it
  • Beliefs become part of identity, making them defended like territory under attack
  • In nearly every case where self-deception is claimed necessary for happiness or success, a non-deceptive third option exists

The argument · threads8 threads · 38 claims
0.85

Scout mindset means accurately mapping reality and preferring truth over comforting falsehoods.

6 pointscentrality 5/5
  • Scout mindset is the motivation to see things as they are and not as you wish they were—the scout's role being to explore and form as accurate a map of reality as possible, having preferences about what they find but valuing knowing what is actually there above all.

    “the scout's role is to go out and explore and see what's really there, and form as accurate a map of the landscape or of the situation as possible”

  • A practical sign of scout mindset is being able to explain the views of those you disagree with compellingly enough that a believer couldn't tell you don't actually hold them—avoiding inadvertent straw-manning—and being able to name reasonable critics of your own positions rather than focusing only on the unreasonable ones everyone has.

    “really trying to understand the views of people who disagree with you and understand them well enough that you could explain them and explain them compellingly enough that someone else couldn't tell that you actually don't believe that yourself”

  • Beyond intellectual rewards, scout mindset is emotionally liberating: being able to follow evidence wherever it leads without needing to believe particular things to feel happy or motivated is freeing, and there is a genuine thrill and joy of curiosity in seeing the world more clearly regardless of whether the conclusion is one you wanted.

    “it is, I and many people I know find it very freeing to be able to follow the evidence where it leads without feeling like you have to believe such and such in order to be happy”

  • In every specific case where people claim you must self-deceive to be happy, motivated, or successful, there is an alternative strategy that achieves the same goal without deception; people give up too soon and neglect this third option, when in fact you can make peace with reality rather than choosing between comforting falsehoods and misery.

    “In all of those cases, I have seen other ways you can achieve those goals without having to deceive yourself... they're kind of giving up too soon, and they're assuming, well, you've got to deceive yourself if you want to be happy or motivated”

  • Scout mindset refers only to how you decide what to believe, not to constant action-paralyzing second-guessing; it means holding a map of reality with appropriate uncertainty that you can consult when making decisions, then acting on the best decision until new information or a designated review point warrants revisiting it.

    “I mean to define it as having a "map" in your head of reality... and having your map be uncertain in all the places where it should be uncertain, and then you can pull that map out when you're making decisions”

  • There are two senses of rationality: epistemic rationality, reasoning to make your beliefs as accurate as possible, and instrumental rationality, making decisions that effectively achieve your goals; many argue these conflict, claiming self-deception (soldier mindset) is sometimes instrumentally optimal for happiness or success.

    “one of the senses is epistemic rationality, which is about... reasoning in a way that makes your beliefs more accurate over time. Then the other sense of the word is instrumental rationality, which is about whether the decisions you make are effectively achieving your goals”

0.75

Soldier mindset—defending beliefs against threatening evidence—is motivated by identity fusion and militaristic reasoning habits.

4 pointscentrality 5/5
  • When a belief becomes part of one's identity—whether political, religious, or even which programming language is best—disagreement feels like a personal attack, like someone stomping on your flag, and the more things admitted into identity the harder it becomes to think clearly about them.

    “when someone disagrees with them, it feels like a personal attack, it feels like someone is stomping on your country's flag”

  • Soldier mindset is the motivation to defend your beliefs against any threatening evidence; our everyday language about reasoning is strikingly militaristic—we buttress, fortify, and defend beliefs, poke holes in and shoot down opposing arguments, and frame changing our minds as conceding, admitting, or surrendering as if losing a battle.

    “the motivation to defend your beliefs or what you want to be true against any evidence or argument that might threaten those beliefs”

  • Governor Earl Warren argued for interning Japanese-Americans by treating both the presence and absence of sabotage evidence as confirmation of treachery—taking the lack of subterfuge as proof they were hiding it—an example of treating evidence as confirming a view regardless of its direction.

    “I take that to be even more confirmation that they are planning subterfuge, because they're being so sneaky about it, that they've managed to hide all the evidence”

  • Pastor Joshua Harris discounted years of complaints that his book I Kissed Dating Goodbye harmed readers, because he held a background belief that being well-intentioned means you aren't causing harm; only after a congregation abuse scandal shattered that node could he reconsider his book's harm—illustrating that a hidden prior belief must change before the target belief can.

    “he had this kind of background belief that hadn't budged, which is that if you believe you're well-intentioned then you're not actually causing harm, and now that that node had shifted, he could actually shift the node that was about the book in particular”

0.61

Intelligence and expertise do not protect reasoning from motivated distortion on ideologically charged topics.

4 pointscentrality 4/5
  • Our judgment isn't limited by knowledge nearly as much as it's limited by attitude; reasoning skill is a tool that can be directed equally toward finding truth or toward building airtight defenses of whatever one wants to believe, depending on motivation.

    “knowledge and thinking skill are... They're like a tool, and you can direct that tool to any end, you can direct it to the end of figuring out what's really true if you're motivated to do that, or you can direct it to the end of coming up with air-tight defenses of whatever position you want to believe”

  • Intelligence and subject-matter knowledge do not protect you from motivated reasoning on that subject; people wrongly assume their expertise makes their reasoning objective, but on ideologically charged topics that is probably false.

    “intelligence and knowledge on a particular subject does not protect you from motivated reasoning on that subject”

  • Dan Kahan's 'graph of despair' shows that at low scientific intelligence, liberals and conservatives agree similarly on climate change, but as scientific intelligence rises the views diverge sharply—Democrats approaching near-total agreement while conservatives fall to about 20%—contradicting the idealistic hope that more education produces convergence on truth.

    “as people increase in scientific intelligence, the correlation between political views and climate change use increases, and so the lines... they diverge as you go up the scale of scientific intelligence”

  • The more plausible explanation for the graph of despair is not that intelligence lets people deceive themselves better, but that scientifically knowledgeable people tend to be more politically engaged and therefore more aware of and motivated to give their tribe's correct answer on charged topics.

    “it's not that you are better able to deceive yourself, it's just that you... On this particular topic, you have more of a motivation to get the politically correct answer for your side”

0.61

Bayesian reasoning improves belief accuracy by attending to likelihood ratios of evidence given competing hypotheses.

4 pointscentrality 4/5
  • Bayes' rule states your confidence in a hypothesis should rise in proportion to the ratio of how likely the observed evidence would be if the hypothesis were true versus if it were false; even without precise numbers, being mindful of this ratio improves reasoning over default human intuition.

    “your confidence should go up in proportion to the ratio of how likely is that evidence... in a world where your hypothesis is true compared to how likely would that evidence be in a world where your hypothesis false”

  • The claim of Bayesian reasoning is not that it produces 100% accurate beliefs, but that starting from our inevitably flawed and subjective prior beliefs, updating in a roughly Bayesian way will produce more accurate beliefs than not doing so.

    “conditional on that, will we be better off if we then try to use Bayes' rule in updating on new evidence, like from that starting point of our subjective and probably flawed beliefs. And I'm claiming that I think we probably will be better off”

  • A single article about a woman passed over for a promotion is weak evidence that tech is unusually sexist, because even in a world where tech is no more sexist than average, such cases would still occur at least sometimes in any large industry, so the likelihood ratio is near one.

    “even in such a world, it's pretty likely that there would be at least one case of a woman who was passed over for a promotion due to her gender, just because in any big industry, that kind of thing's going to happen at least sometimes”

  • Two people can both be perfect scouts—genuinely trying and both smart—and still fail to resolve a factual or empirical disagreement, because even absent motivated reasoning, prior beliefs color how each interprets the same evidence, so they can't even agree on which direction the evidence should update them.

    “your prior beliefs color how you interpret the evidence, and so there are just these tangles... we can't resolve our disagreement because whatever evidence we look at, we can't agree on... what direction that evidence should update us”

0.56

Scout-minded action produces greater real-world impact than identity-satisfying activism or tribal defense.

3 pointscentrality 4/5
  • While soldiers who see the world in black and white have undeniably changed the world, being passionate and driven about achieving impact is likely a more effective strategy than being passionate about defending one's ideology or tribe—a claim Galef believes but admits cannot be rigorously proven.

    “I'm just trying to claim that I think it works better if you could be passionate and driven about impact, than be passionate and driven about defending your ideology or your tribe”

  • Actions can be plotted on two axes—how much they validate your identity versus how much real-world impact they make; many identity-satisfying actions (arguing with strangers online, bumper stickers, infighting with near-allies over small disagreements) provide little or negative impact, and holding identity lightly frees you to choose genuinely impactful actions instead.

    “one axis is how strongly does this action validate your identity, and the other axis is how much impact does your action actually make on the world?”

  • The Humane League shifted from splashy demonstrations against animal-experimenting scientists to negotiating with large agricultural companies over farm-animal welfare—because there are vastly more farm animals than lab animals—winning concessions like ending the grinding of male chicks, an unsatisfying strategy for tribal fighters but far more impactful for the actual goal of reducing animal suffering.

    “he pivoted the organization to focus instead on farm animal treatment, in part because just running the numbers, there's way more farm animals in the world than lab animals”

0.47

When evaluating implausible claims, assess the proponent's self-awareness rather than manufacturing substantive counterarguments.

4 pointscentrality 3/5
  • When evaluating an implausible idea, attend not just to the idea but to its proponent: someone who can articulate why their theory seems implausible and why others disagree, yet still maintains people are missing something, is a far better signal of a credible maverick than a fanatical demagogue who cannot see the opposing case.

    “people who understand and can articulate the reasons why their theory seems implausible or why other people don't already agree with them, if they can understand and articulate those things, and still say, I think people are still missing something, that is a good sign to me”

  • When dismissing an implausible claim, it is more intellectually honest to say it conflicts so strongly with your background beliefs that it's below your threshold for investigation, than to manufacture a weakly-supported substantive counterargument; the key distinction is between dismissing on cognitive prior-probability grounds versus dismissing because you are motivated to reject it.

    “the distinction is between dismissing something because your prior beliefs indicate that it is very unlikely to be true versus dismissing something because you don't want to accept it or you don't want to take it seriously”

  • Human nature biases us against weird or fringe ideas and the unpopular mavericks who hold them, so as a scientific community we should err more on the side of openness—letting people investigate ideas most think are wrong—rather than being too unsympathetic to them.

    “I suspect that the bias is towards being too unsympathetic to weird or fringe ideas, just because I think that's human nature that we're... We tend to not like the unpopular maverick who has an idea that tells us we're wrong”

  • To stay honest while not evaluating every claim, use outside-view criteria: would you check out an equally implausible but more desirable theory; do smart, reasonable people you trust find it promising; and do you get angry when anyone takes it seriously—the last being a sign your stated reason of 'not worth my time' is actually motivated.

    “if I don't want anyone to investigate it, and if I get angry when anyone takes it seriously, then that kind of gives the lie to my claim that it's just not worth my time”

0.44

Scientists should critique their own ideas; the advocacy model fails to distinguish true from false claims in social science.

6 pointscentrality 2/5
  • Good theoretical scientists should be their own ideas' biggest critics rather than biggest advocates; those who claim advocacy is optimal because a free market of ideas will sort things out are likely rationalizing the pleasure of advocacy, and in practice—especially in social science where experiments rarely settle questions—the advocacy model fails to weed out true from false ideas because few people vet the arguments.

    “if you have people who are just advocating for their ideas without worrying too much about whether those ideas are true, I don't see that process really weeding out the true from the false ideas very well”

  • In physics, ideas can be tested conclusively so the life cycle of an idea runs on a quick time scale, whereas in social science (and the humanities) experiments rarely produce clear-cut results—statistical wizardry can support either side—so ideas rise and fall based on the advocacy of their proponents more than on evidence, and one can cling to an increasingly unlikely position for a long time.

    “in social science, the experiments are so rarely clear-cut that ideas rise and fall much more based on the advocacy of their proponents than based on the facts from the world”

  • Communication always involves the listener making assumptions about what the speaker meant, so if most advocates in a space make extreme claims, listeners will reasonably hear the extreme version even from someone making a more moderate claim.

    “communication is always imperfect and always involves some amount of making assumptions about what the other person meant to say”

  • The claim that everyone is always better off being rational or seeing the truth in every situation is too strong to defend; no one has provided a sufficient defense that everyone's happiness is maximized by a rational worldview in every case.

    “I've never heard anyone give a defense of that claim that seems sufficient for how strong a claim that is, like how do you know that everyone's always better off being rational?”

  • While the universal claim is too strong, we can be confident that on the margin people would be better off being at least somewhat more truth-seeking and engaging in less self-deception than they do by default.

    “I think we can be pretty confident that people would be better off being at least somewhat more truth-seeking than they are by default... that on the margin, we would be better off with more truth and less self-deception”

  • Although most people profess to value rationality, two real groups explicitly reject it: postmodern/romantic thinkers who deny objective truth and treat truth-seeking as an instrument of oppression, and those who find it more noble to stick to their guns and never change deeply held beliefs.

    “they kind of fall into two categories. One category is the kind of post-modernist or maybe in a sense romantic school of thought where there really isn't any objective truth”

0.38

Hold identity lightly: identify with causes contingently while maintaining emotional distance from tribal orthodoxy.

2 pointscentrality 3/5
  • Rather than keeping your identity small by refusing to identify with any cause (Paul Graham's advice), it is better to hold your identity lightly—you can identify with movements you value while maintaining emotional and ideological distance, always separating what your tribe believes from what you believe and treating your support as contingent on continuing to think it's right.

    “I think what you need to do is to be able to identify with causes and movements and so on that you value while still keeping some distance, emotional ideological distance, from those positions”

  • There is no shared definition of rationality; in colloquial practice, many people use 'rational' to mean 'believes what I believe,' treating anything that disagrees with them as irrational.

    “I think in practice, what a lot of people mean is, you should believe what I believe. And so anything that disagrees with me is not rational”