Core distinction
Moral psychology asks how human beings form, express, and revise moral judgments. Metaethics asks what those judgments mean, whether any are true, what could make them true, and how moral knowledge is possible. Evidence about the first constrains theories of the second, but does not by itself decide among them.
The Righteous Mind, Conclusion — Haidt endorses pluralism; his data do not prove it
Haidt ends by warning against moral monists and invoking Isaiah Berlin’s value pluralism. That is a philosophical conclusion layered onto the book’s psychology. Moral Foundations Theory can show that several kinds of concern recur, vary in salience, and organize disagreement. It cannot by itself establish that each concern tracks a genuine irreducible value, that no master principle can rank them, or that Berlin’s account of the human moral horizon is true.
The same separation blocks the opposite mistake. A finite plurality of genuine values would not imply relativism, equal validity, or exemption from cross-cultural criticism. One can understand the value another person pursues while rejecting the person’s facts, hierarchy, means, or treatment of others. → permanent pluralism distinction · Moral monism and Berlin, PDF p. 353 (file not included in this release) · Conclusion endnotes 1–2, PDF p. 433 (file not included in this release) · Conclusion source audit
The Righteous Mind, Chapter 6 — three tempting inferences that still do not follow
Chapter 6 adds three important examples of the descriptive–normative gap. First, finding several recurring moral concerns does not prove normative value pluralism or disprove a monistic theory capable of ranking them. Second, proposing an evolved origin for a judgment does not establish either its truth or its falsity; an evolutionary genealogy needs a separate debunking or vindicating argument. Third, a founder’s temperament or possible diagnosis cannot decide whether the founder’s ethical arguments are valid.
Haidt’s secure contribution is narrower: a descriptively adequate moral psychology should observe a plural range of motives rather than deduce human nature from one preferred rule. The validity, authority, and ranking of those motives remain independent philosophical questions. Plural moral psychology, PDF pp. 137–140 (file not included in this release) · Ad hominem rejected and evolutionary method, PDF pp. 147–149 (file not included in this release) · Chapter 6 source audit
The Righteous Mind, Chapter 7 — repugnance identifies an appraisal, not its truth
Haidt uses the Meiwes–Brandes killing to make Sanctity/degradation vivid and Leon Kass’s “wisdom of repugnance” to argue that disgust can warn us before an articulable harm is visible. Psychologically, the case shows that consent and direct welfare do not exhaust many people’s moral concepts: bodily dignity, sacred life, pollution, and desecration also organize judgment.
None of that settles the normative or metaethical issue. A disgust response may register a morally relevant feature, an unarticulated empirical risk, a learned taboo, or prejudice. Haidt’s premise that the case caused no direct material or utilitarian harm is itself contestable: Brandes died, capacity and conditions of consent matter, homicide law protects public goods, and third-party effects need analysis. The correct sequence is to treat repugnance as a prompt for inquiry, then state and defend the reasons it may be tracking. The endnote’s later murder conviction is a legal fact, not by itself a metaethical proof. Meiwes, autonomy, and degradation, PDF pp. 175–177 (file not included in this release) · Kass and repugnance, PDF pp. 182–183 (file not included in this release) · Retrial and murder conviction, endnote 36, PDF p. 391 (file not included in this release) · Chapter 7 source audit
The questions live at different levels
| Level | Central question | Example |
|---|---|---|
| Psychological | What causes the judgment? | Disgust, empathy, coalition, deliberation, or social learning |
| Semantic | What does the judgment express? | A belief, prescription, plan, emotion, or hybrid state |
| Ontological | Are there moral facts, and what are they? | Natural facts, non-natural facts, constructed standards, or none |
| Epistemic | How could we be justified? | Perception, intuition, inference, coherence, testimony, or practice |
| Normative | What ought an agent do? | Respect a constraint, maximize value, keep a contract, cultivate virtue |
Confusing these levels creates several recurring invalid inferences.
Four inferences that do not follow
“Judgment is intuitive, therefore it is irrational”
Fast cognition can register expertise or morally relevant features, just as it can encode bias. Speed and emotional force identify a causal process, not its reliability. The proper question is whether the process tracks reasons under the relevant conditions.
“Moral reactions evolved, therefore no moral claims are true”
An evolutionary explanation can support a debunking argument only with additional premises: that the process is insensitive to truth or reasons, that no corrective method exists, and that an error account explains the evidence better. Evolutionary origins alone do not refute arithmetic, perception, or morality.
“People disagree, therefore morality is subjective”
Disagreement is evidence about epistemic difficulty, ambiguity, incentives, culture, and possible underdetermination. It is compatible with realism, constructivism, relativism, error theory, and non-cognitivism. The pattern and source of disagreement matter more than its mere existence.
“Moral language motivates, therefore it cannot express beliefs”
Non-cognitivists use motivation to explain moral language as attitude or prescription; cognitivists can reply that beliefs interact with desires, identities, or rational commitments. The empirical connection between judgment and motivation informs this debate but does not settle the semantics unaided.
How psychology legitimately bears on metaethics
Psychology can still matter in at least five ways:
- Reliability: it identifies environments in which judgment is predictably distorted.
- Genealogy: it reveals why certain concerns and concepts are psychologically available.
- Feasibility: it tests whether a normative theory demands forms of agency people can sustain.
- Disagreement: it helps distinguish informational conflict from value, identity, or coalition conflict.
- Debunking: in conjunction with philosophical premises, it can challenge the warrant of a class of intuitions.
Metaethics constrains psychology in return. Calling a response a “moral judgment” presupposes some account of judgment, normativity, and meaning. A purely functional claim that morality sustains cooperation also cannot establish that cooperation is morally good without a further normative premise.
Application to Haidt and moral-foundations theory
Haidt is strongest when explaining why moral disagreement persists and why argument often follows intuition. The framework supports humility and better communication. It is weaker when readers convert descriptive foundations into equal moral entitlements or infer that because a concern is evolved and widespread it is justified.
Haidt, Chapter 1 — what cultural variation establishes
Haidt uses two research programs to widen the descriptive map of morality: Shweder, Mahapatra, and Miller’s 600 interviews in Chicago and Orissa, and Haidt, Koller, and Dias’s 360 interviews across Philadelphia, Porto Alegre, and Recife. The second study directly asked whether anyone was harmed; Haidt reports that group differences in moralizing disgust and disrespect became larger after respondents who perceived harm were removed.
This supports the claim that the experienced moral domain can extend beyond conscious harm judgments and vary with culture, class, and age. It does not show that morality is relative, that every culturally enforced duty is justified, or that no cross-cultural moral criticism is possible. Those conclusions require philosophical premises the interviews did not test. Shweder study, PDF pp. 31–35 (file not included in this release) · Haidt study, PDF pp. 36–41 (file not included in this release) · Chapter 1 evidence audit
“Morality binds and blinds” is therefore a psychological hypothesis with normative implications, not a complete moral theory. Binding can enable trust and sacrifice; it can also intensify conformity, exclusion, and collective wrongdoing.
Haidt, Chapter 2 — a causal model is not a theory of moral truth
The social-intuitionist model separates six causal routes: intuition commonly produces judgment; judgment commonly recruits reasons; other people’s reasons and social responses can alter intuition; and deliberate or privately reflective revision remains possible. This helps explain how a verdict forms and changes. It does not tell us whether the verdict is true, what would make it true, or how much evidential authority its genealogy deserves.
Two source limits reinforce that separation. Haidt’s cognitive-load null results and thirty-person dumbfounding interviews were unpublished, and he acknowledged having no daily-life frequency estimate for the private-reasoning pathways. The model is a useful psychological framework, not a measured universal law or a metaethical conclusion. Null studies and interviews, PDF pp. 54–59 (file not included in this release) · Endnotes 26 and 30, PDF pp. 367–368 (file not included in this release) · Endnote 44, PDF p. 369 (file not included in this release) · Permanent model note
Haidt, Chapter 3 — causal timing and neural location still do not settle moral truth
Chapter 3 argues that rapid affect, bodily state, social cueing, early social evaluation, and emotion-associated brain activity can precede explicit moral reasons. Even if every causal claim were robust, the result would identify how a judgment arose, not whether it is true, justified, or action-guiding. A disgust response could detect degradation or import an irrelevant odor; a deontological trolley response could protect a person or reflect framing; a utilitarian response could compare welfare or rationalize emotional distance.
The source’s own qualifications reinforce the distinction. Its foul-odor comparisons were individually mixed despite a significant combined analysis, and the original finding that utilitarian trolley answers took longer was later challenged as a story-selection artifact. Descriptive evidence can reveal possible unreliability and motivate debunking arguments, but the bridge to a metaethical conclusion still requires premises about truth-tracking, relevance, and available correction. Bodily influence, PDF pp. 80-81 (file not included in this release) · Moral neuroscience, PDF pp. 86-88 (file not included in this release) · Odor endnote, PDF p. 372 (file not included in this release) · Reaction-time correction, PDF p. 374 (file not included in this release) · Permanent concept note
Naturalistic shortcut
“This trait evolved,” “most cultures value it,” and “it supports cooperation” are not yet conclusions that the trait or norm is right. Each can supply a premise, but the evaluative bridge must be stated and defended.
Source routes
- The Righteous Mind and The Psychology of Morality supply descriptive mechanisms.
- Moral Cognitivism vs. Non-Cognitivism supplies the semantic dispute.
- Moral Realism vs. Anti-Realism supplies the ontological and epistemic dispute.
Downstream use
- Argument map: How Should Moral Judgment Combine Intuition, Reason, and Consequences?
- Position: Morality
- Writing: Your Moral Intuitions Need a Constitution
Captured moral-psychology research notes
Capture provenance — Random Notes on Moral Phil research
Original location:
90 Notes/Random Notes on Moral Phil research.md. Filed on 2026-10-06. The material below is preserved from the capture; relocation does not independently verify its claims or attributions.
Short answer: Jonathan Haidt isn’t staking out a classic metaethical position (cognitivism vs. non-cognitivism). He’s a moral psychologist. His published work—especially the Social Intuitionist Model and Moral Foundations Theory—aims to explain how people form moral judgments (mostly via quick intuitions, with reasoning often post-hoc), not what moral claims mean or whether moral facts exist. That empirical program is compatible with several metaethical views, so he generally doesn’t pick one. (its.caltech.edu, ScienceDirect, moralfoundations.org)
Here’s the clean way to place him:
- Primary stance (descriptive): psychological intuitionism/sentimentalism.
Moral judgments are typically caused by fast, affect-laden intuitions; reasoning tends to justify or persuade after the fact. That’s a psychological thesis, not a semantic one. (its.caltech.edu) - What that does—and doesn’t—imply about metaethics:
- It aligns naturally with Humean/sentimentalist pictures (and can be made to play nicely with sophisticated expressivism/quasi-realism), but it doesn’t force them. A cognitivist could accept Haidt’s psychology and still say moral sentences state truth-apt propositions; they’d just add a story about why belief is so closely tied to emotion/motivation. (Stanford Encyclopedia of Philosophy)
- So: Haidt’s view does not settle whether moral claims are truth-apt (cognitivism) or primarily attitude-expressing (non-cognitivism). It underlines that intuitive emotion is central to moral judgment formation. (its.caltech.edu)
- Normative framework:
Haidt doesn’t offer a prescriptive moral theory (like utilitarianism or Kantianism). He proposes a pluralist, descriptive map of common “moral foundations” (Care/Harm, Fairness/Cheating, Loyalty/Betrayal, Authority/Subversion, Sanctity/Degradation, Liberty/Oppression) and shows how different groups weight them. That’s an explanatory framework, not a verdict-giving moral theory. (moralfoundations.org, Wikipedia, The UK in the World Values Survey)
If you want a one-liner to use in debate:
“Haidt’s work is about how we form moral judgments—fast, intuitive, socially shaped—not about what moral words mean or whether there are stance-independent moral truths; so his psychology doesn’t commit him to either cognitivism or non-cognitivism.” (its.caltech.edu)
If you’d like, I can add a quick table showing how Haidt’s psychology can be integrated with each major metaethical view (realism, expressivism, constructivism) and what each would say about “ought” and “wrong.”
Short version: among (mostly Anglophone) professional philosophers, the majority are moral realists and cognitivists.
- Moral realism vs anti-realism (2020 PhilPapers Survey):
~60% accept/lean moral realism; ~26% accept/lean anti-realism (remainder other/undecided). This is up slightly from 2009 (~56% realism). (survey2020.philpeople.org) - Cognitivism vs non-cognitivism (2020):
Roughly 63–69% cognitivism vs 17–21% non-cognitivism (differences reflect inclusive vs exclusive counting across subpopulations). (survey2020.philpeople.org) - If you force a finer-grained metaethical choice (2020):
Among options naturalist realism, non-naturalism, constructivism, expressivism, error theory, responses (inclusive) cluster as:
~32% naturalist realism, ~27% non-naturalism, ~21% constructivism, ~11% expressivism, ~5% error theory(with some “other/undecided” and multi-select behavior). This breakdown dovetails with the headline “realism + cognitivism” pattern. (survey2020.philpeople.org) - Method note / source: These figures come from the 2020 PhilPapers Survey (≈2,000 philosophers), with a peer-reviewed write-up by Bourget & Chalmers/Bourget & Levinstein (methodology and tables). (Michigan Publishing)
If you want, I can turn this into a one-screen cheat sheet (definitions + the dominant positions + quick arguments for/against each).
“Pro tanto” indicates a reason or obligation with some weight, but not necessarily decisive, while “all things considered” signifies a final, overall judgment that weighs all relevant factors to determine the decisive truth or course of action.In essence, a pro tanto reason can be overridden by stronger opposing reasons, but an all-things-considered conclusion is the ultimate verdict after all factors have been considered and balanced.