What this covers

Psychologist Molly Crockett and host Sean Carroll examine what happens in the mind when humans make moral judgments and why those judgments matter so much to how we see ourselves. Rather than treating morality as a restraint on selfish impulses, Crockett presents humans as fundamentally social learners equipped with basic empathy and altruism, whose moral beliefs are shaped by cultural observation and feedback. The conversation traces where moral intuitions originate, why we sometimes deceive ourselves about our own motives when condemning others, and how our sense of personal identity is more thoroughly anchored to moral character than to memory itself.

The discussion moves across several distinct territories. Crockett explains how costly signaling operates in punishment—people spend resources to punish unfair behavior even when anonymous, then retrofit the act with justifications about deterrence. She examines normativity across cultures during the pandemic, showing that people trusted leaders endorsing impartial resource distribution but rejected those endorsing instrumental harm. A central concern emerges around how social media algorithms intervene on cultural evolution by amplifying outrage and villain narratives, reshaping what new generations learn about right and wrong. Crockett argues this creates a mismatch: our moral apparatus excels at identifying individual villains and victims but leaves us poorly equipped to assign responsibility to systems that constrain choices. The conversation also touches on why deontological reasoning is rationally defensible in social terms, how mass gatherings amplify moral narratives, and why the speed of social media compresses moral learning that historically unfolded over decades.

Sharpest takeaway

Crockett argues that human morality is not a constraint on inherently selfish impulses but a culturally-learned overlay on basic prosocial sentiments, and that the modern attention economy—via algorithms intervening on social learning—is reshaping our moral cognition in ways that obscure systemic causes of harm in favor of finding individual villains.

  • Humans are fundamentally social learners with basic empathy/altruism, not selfish creatures needing morality as a check
  • Algorithms reward outrage and villain-narratives, intervening on cultural evolution of norms in a historically new way
  • Moral cognition is built to identify victims and villains, leaving us unequipped to assign responsibility to systems

The argument · threads8 threads · 33 claims
0.76

Humans are primarily social learners whose moral beliefs are culturally constructed, not innate impulses restrained by reason.

5 pointscentrality 5/5
  • What is most distinctive about humans is not morality or norms per se, but an unparalleled capacity to learn socially—through observation and feedback—and to transmit knowledge culturally from generation to generation, which underpins our norm-focused nature because social cohesion aids survival.

    “if you ask scholars of human nature what is most distinctive about humans, it's not necessarily morality or norms, but rather our unbelievable capacity to learn socially”

  • Rather than 'morality,' the more useful frame is norm psychology or normativity—how people think about what is appropriate or inappropriate, allowed or disallowed within a society—because these are social constructions with tremendous within- and across-culture variation built atop basic domain-general building blocks like empathy and care traceable to other mammals.

    “I am shifting more towards favoring the term norm psychology or normativity, which concerns how people think about what is appropriate or inappropriate, what is allowed or disallowed within a society. And of course, these are all social constructions.”

  • Because humans are prodigious social learners and culture shapes behavior as much as or more than nature, education is the best tool for moral change—each new generation offers an opportunity to give people better tools to think about themselves, others, and morality, though this is undermined by underfunding and by discourse systems optimizing for profit rather than well-being.

    “education is I think the best tool we have... with every new generation, there are opportunities to give people the tools and resources to think about themselves and others and problems in the world and morality in ways that might be better than what we have now.”

  • Very young children believe you are equally obligated to help a stranger as a friend or family member, and the favoritism toward close others (parochial morality) develops with age—it is learned from culture rather than something we are born with, per work by Julia Marshall and Paul Bloom.

    “very, very, very young children think you are equally obligated to help a stranger versus a friend or family member in need... This favoritism towards people who are more close to us isn't something we're born with, it's something that we learn from our culture.”

  • People are not fundamentally a bundle of terrible impulses that morality sits on top of to restrain, nor are they fundamentally angelic; they are fundamentally social learners whose beliefs about their group and culture shape deliberate conduct, working in concert with basic instincts that include empathy and altruism.

    “What we are fundamentally is social learners... which is on top of or in concert with basic instincts that we have, which include a basic instinct towards empathy and altruism.”

0.72

Profit-driven algorithmic systems reshape moral learning by amplifying outrage and extreme content, compressing moral change and rewarding particular narratives.

4 pointscentrality 5/5
  • Social media algorithms are designed first and foremost to grab attention and maximize time online to make money, and by definition the content that captures attention is extreme and outrage-provoking, even when that information is far-fetched or unrepresentative—creating discourse and meta-discourse divorced from things that actually affect our lives.

    “the way that social media algorithms are designed, which is first and foremost to grab our attention in order to keep us online for longer, in order to make the most money. And by definition, the stuff that captures our attention is information that's really extreme, that provokes outrage”

  • Morality can change at cultural scale through representation (we conform by observing what others do) and reinforcement learning (rewards shape behavior); what is historically new is that profit-driven algorithms intervene on these social learning processes—elevating certain villain archetypes and rewarding outrage with likes/retweets—thereby reshaping what new generations learn about right and wrong.

    “what we have now that we've never had before is algorithms intervening on social learning processes and cultural evolution in a way that I think is new.”

  • The speed of social media compresses moral learning that historically occurred over decades into days, removing time to digest and reflect, and producing a backlash where rapid new proposals about treating each other better are dismissed as illegitimate—'just woke cancel culture'—simply because they are unfamiliar.

    “we are used to social or moral progress or change happening over a course of decades, and what we are seeing now is a much, much faster timescale. And I think that there are some people who are just responding to this as if this is illegitimate.”

  • Moral narratives function as an epistemic power move: they impose the narrator's preferred structure of causation and moral responsibility onto an audience—often without the audience's awareness—as seen in greenwashing and reputation-laundering (e.g., Sackler donations, fossil fuel PR) that direct moral judgment away from the actual consequences of choices.

    “moral narratives are essentially a a kind of, an epistemic power move, because what they do is they impose the narrator's preferred structure of causation and moral responsibility into the audience, and in many times the audience doesn't even know that that is happening.”

0.58

Moral punishment serves as costly signaling of trustworthiness rather than effective behavior change, driven by masked revenge motives.

4 pointscentrality 4/5
  • Although revenge and retribution feel sweet and satisfying, they are not clearly helpful for changing offenders' behavior in the long term, and data show that holding on to grudges and revenge is actually harmful to one's own well-being.

    “there's this idea that revenge is sweet and it's satisfying, but actually there's data is showing that if you hold on to grudges and revenge, that's harmful to your well-being as well.”

  • Moral judgments and punishments act as a costly signal to others that the condemner would not themselves behave that way, so that condemning other people makes one be perceived as more trustworthy (per work by Jillian Jordan and colleagues).

    “when you condemn other people, you are seen as more trustworthy, so there is a way in which moral judgments and punishments act as a signal to other people that you yourself would not behave that way.”

  • In economic-game experiments where punishment is anonymous and the target never learns they were punished, people still pay money to reduce the payoff of someone who behaved unfairly, yet afterward report they were 'teaching a lesson'—revealing a genuine taste for revenge masked by a deterrence justification.

    “we gave people an opportunity to punish someone and they would never know that they had been punished... they will pay money to reduce the pay-off of somebody who's behaved unfairly towards them, even if they never find out. But then when you ask them... a lot of people are like, oh, I wanted to teach a lesson, even though in some of these cases, there is no lesson being transmitted.”

  • Most behaviors are multiply determined by several motives, but when we explain a choice afterward, social forces incentivize us to pick the most socially desirable reason, which over time creates self-serving narratives—as in punishment, where people claim they wanted to 'teach a lesson' even when no lesson could possibly be transmitted.

    “most of our behaviors are multiply determined... And then when we explain that choice afterwards, there are social forces that incentivize us to pick the reason that's the most socially desirable to explain our behavior.”

0.54

Mass gatherings amplify narratives about human nature and expand moral circles through time spent collectively, effects independent of substance use.

3 pointscentrality 4/5
  • At multi-day mass gatherings like Burning Man, the longer people stay, the more transformed they feel and the more their moral circle expands to encompass all of humanity rather than just near-and-dear others.

    “with every passing day that people spend at an event like Burning Man, the more they feel transformed and the more expanded their moral circle is to encompass everyone.”

  • While psychedelic substance use predicts reporting transformative experiences at mass gatherings, controlling for substance use shows that over and above any substance, more time spent at collective gatherings independently produces feelings of transformation and positive emotion.

    “Over and above any type of substance use, we find that the more time people spend that these collective gathering events, they feel transformed and positive emotions and so on.”

  • Mass gatherings act as amplifiers of narratives about human nature: whatever story is being promoted about how we treat each other—whether humanity is basically good or basically evil—the gathering intensifies it, which explains both the positive effects and the dark side (mobs, Nazi rallies).

    “I think of mass gatherings as amplifiers. So whatever the story that's being told, the culture of a mass gathering, in terms of how we treat each other... the mass gathering is going to amplify it”

0.54

People lack conceptual tools to assign responsibility to systems rather than individuals, leaving them epistemically trapped in villain-seeking cycles.

3 pointscentrality 4/5
  • Our moral cognitive apparatus is set up to identify victims and villains and to judge the rightness or wrongness of individual actions, leaving us without a good conceptual system for assigning responsibility to systems that constrain people's choices such that they cause harm—a gap exemplified by pandemic discourse that blamed individuals for not masking while ignoring that society offered no good choices.

    “our whole moral cognitive apparatus seems to be really set up to identify victims and villains... but what about systems? What about systems that constrain people's choices such that they can't help but act in ways that cause harm... like how do you blame a system?”

  • Although people believe they are reasoning toward flourishing for themselves and others, much of what feels like reasoning is actually motivated, shaped invisibly by systems of power that trace back generations and are in many ways designed not to be seen.

    “a lot of times we think we're reasoning, but actually, it's motivated, right?”

  • We hold three incompatible beliefs—there are good guys and bad guys, you must be one of the good guys, and yet suffering happens sometimes because of you—and our defense mechanisms resolve this by externalizing blame, leaving us epistemically under-resourced and stuck in a vicious cycle of seeking villains that creates intractable interpersonal and political conflicts.

    “if I am never the bad guy, and yet suffering is happening sometimes because of me, how do we deal with that? So I think that we're really... We are epistemically under-resourced to solve conflicts in our society”

0.44

People trust leaders who distribute resources impartially but distrust instrumental harm, revealing limits to utilitarian moral reasoning across cultures.

7 pointscentrality 2/5
  • The myth that people are basically selfish—homo economicus, where everything is explained by self-interest—is not grounded in science and is convenient for capitalism, and risks becoming a self-fulfilling prophecy that discourages collective action by suggesting there's no point caring about others if no one else does.

    “this idea that homo economicus, like people are basically selfish, everything can be explained in terms of self-interest, this is a narrative that is not grounded in science, and I will add, is very convenient for capitalism.”

  • There is a distinction between an in-group bias that appears universal and detectable from a young age, versus favoritism toward close versus distant others within a group, which is learned; the modern challenge is conceiving of all humanity as one giant in-group to address global problems like pandemics and climate change.

    “There is also separately an in-group bias that is naturally... That seems to be pretty universal and is detectable from a young age.”

  • Self-report has been popularly distrusted in the behavioral sciences, but this is an over-simplified view; self-report is actually underrated, and new computational methods that extract decision processes and compare them to self-reports (e.g., measuring introspective accuracy) can validate the link between subjective reports and objective behavior.

    “I think that self-report is underrated. And I think that it has been very popular in the behavioral sciences to not trust it, but I think that that is an over-simplified view”

  • The common claim in moral psychology that science is objective and value-free is mistaken; researchers' values infect which questions they choose to ask, so studying morality requires understanding how one's own values shape the science—for example, the goal of getting opposing political groups to 'meet in the middle' embeds a value that fails when one side is simply right.

    “our values infect the way that we choose to ask questions, the questions that we ask.”

  • Utilitarianism has two dimensions—impartial beneficence (everyone's welfare counts equally) and instrumental harm (harming one to save many)—and across 22 countries surveyed during the pandemic, people trusted leaders who endorsed impartially beneficent resource distribution but distrusted leaders who endorsed instrumental harm (e.g., rationing ventilators or imposing welfare-trading lockdowns).

    “across 22 countries that we surveyed during the pandemic, people prefer leaders who endorse impartially beneficent solutions... But when it comes to instrumental harm, it flips... people distrust utilitarian leaders who say it's okay to sacrifice the welfare one group of people in order to save others.”

  • Being excluded from the group was literally deadly in the ancestral environment because one could not survive without the group, and remains deadly today, with depression, suicide, and many health problems associated with loneliness and social isolation—making being seen as moral by others deeply important.

    “getting excluded from the group is deadly. In our ancestral environment, it was literally deadly... and nowadays it's still deadly, like depression and suicide and all sorts of health problems are associated with loneliness and social isolation.”

  • The incentive structure of academia is tilted more toward developing and promoting one's own pet theory and recruiting believers than toward actually solving problems.

    “the incentive structure of academia is tilted more towards coming up with your own pet theory and promoting it and getting lots of other people to believe in it than actually solving problems.”

0.40

Moral character is core to personal identity and can be lastingly transformed through experiences that dissolve the sense of self.

3 pointscentrality 3/5
  • Morality is more crucial to our sense of self than memory: in hypothetical studies, people are more likely to say someone has become 'a different person' when their morality has dramatically changed than when they have lost their memories, indicating moral character is the core of personal identity.

    “people are more likely to say, oh, that's a different person now if their morality has dramatically changed compared to if they've lost their memory.”

  • When people report gaining insight into the fundamental nature of the universe from psychedelics, this need not be the physicist's notion of cosmology—it may be a different, legitimate kind of understanding about how one fits into the universe, the meaning of one's life, and the point of it all.

    “what people mean when they say they understand the nature of the universe is more about how do I fit into this and what is the meaning of my life and what is the point of it all? So I think there's room for both.”

  • While drugs alter neurochemistry only acutely, psychedelics can produce an epistemically transformative experience—dissolving the sense of self and opening connectedness—that changes one's values and narrative of self lastingly, unlike SSRIs which must be taken indefinitely to maintain effects that may be overstated.

    “with psychotherapy assisted with a psychedelic substance, there can be changes to the narrative of the self and the way that you understand reality that have potential, I think, to carry on well beyond the time that you're high.”

0.32

Deontological moral reasoning is rationally adaptive because it signals trustworthiness and is preferred in intimate relationships.

1 pointcentrality 3/5
  • Deontological reasoning is highly rational in a social sense because deontological reasoners are seen as more trustworthy social partners; close relationships are premised on prioritization, so people overwhelmingly prefer deontological reasoners over utilitarians as spouses and best friends—challenging the view that deontology is merely an irrational intuition.

    “deontological reasons and decisions are seen as more trustworthy by social partners... most people don't want their spouse or their best friends to treat them just the same as any other person... people overwhelmingly prefer deontological reasoners over utilitarians.”