📖 Résumé de lecture · non citable

Moral Tribes - Greene (2013) (EN)

Résumé de lecture, chapitre par chapitre — rapporte ce que le livre dit.

Reading summary, part by part

🔬 Built from the reading dossier, not from the book. Sources:
_dossiers/Moral Tribes - Greene (2013) (EN) — DOSSIER.md (150 candidates of 1 460) and
… — SONDES.md. ⚠️ Not a claim of coverage: 1 304 candidates were cut by the cap.

⚠️ Chiffres révisés le 2026-08-25. Le dossier a été régénéré avec dossier_lecture.py v1.4, qui corrige un défaut : la v1.3 ne scannait pas le texte situé avant le premier en-tête détecté. Pour ce livre la tête représentait 299 mots sur 162 418, soit moins de 1 % — l'écart est marginal et aucune citation de ce résumé n'est affectée (toutes re-vérifiées). ⛔ Sur un autre livre du corpus (Pocket Guide, Porges 2017) la même tête valait 96 % du volume.

---

Identity

Contents, as read — Introduction: The Tragedy of Commonsense Morality · I. Moral Problems — 1. The Tragedy of the Commons · 2. Moral Machinery · 3. Strife on the New Pastures · II. Morality Fast and Slow — 4. Trolleyology · 5. Efficiency, Flexibility, and the Dual-Process Brain · III. Common Currency — 6. A Splendid Idea · 7. In Search of Common Currency · 8. Common Currency Found · IV. Moral Convictions — 9. Alarming Acts · 10. Justice and Fairness · V. Moral Solutions — 11. Deep Pragmatism · 12. Beyond Point-and-Shoot Morality: Six Rules for Modern Herders.

⚠️ Extraction artefact, declared: the contents line reads "Eff iciency" in the file — a broken word from the conversion, not a spelling in the book. Restored above; noted so the file and this summary can be reconciled.

---

⛔ A finding about the dossier itself, before any content

This .txt contains the body and the endnotes, and both are structured as PART I … PART V. The dossier's section detection matched both, producing 15 "sections" where the book has 5. Consequence, measured: a substantial share of the top-scoring candidates are drawn from the endnotes — bibliographic strings ("Bentham (1830). 268 favored by many: Kant (1785/2002)") that score high on markers and mean nothing as prose.

⚠️ This is not a defect the tool hid — the section names are printed, so the notes are visible as their source. But it is the reason this summary quotes far fewer of the top-ranked candidates than the Haidt one, and the reason a future run on this book should restrict itself to the first 738 524 characters.

⛔ And it has already contaminated the corpus. Two of the four Greene quotations used in canon 42 are located in the endnotes, not in the body:

⚠️ They are Greene's words either way — nothing is fabricated, and the canon is not wrong. But a sentence written in a note answering an anticipated objection does not carry the same weight as a sentence in the argument, and the canon does not say which is which. Decision for the owner: annotate canon 42, or leave it. Not done here.

---

Introduction and Part I — the problem

The frame is a parable: herders sharing a commons. The Tragedy of the Commons is Me versus Us — solved, biologically, by cooperation. The book's own problem is the second one, the Tragedy of Commonsense Morality: Us versus Them, tribes each equipped with a working morality, in conflict precisely because each morality works.

Part I supplies the machinery. Two recap sentences carry it:

Verified quotations

"In sum, we are a caring species, albeit in a limited way, and we probably inherited at least some
of our caring capacity from our primate ancestors"

Verified quotations

"In sum, our brains are wired for tribalism."

⛔ And the immediate self-limitation, which must travel with it:

Verified quotations

"being wired for tribalism does not mean being hardwired for tribalism"

⚠️ Wired but not hardwired is the author's distinction, in his own sentence. Any reuse that drops the second clause converts a claim about tendency into a claim about destiny — and the book explicitly refuses that.

---

Part II — dual process

Trolleyology and the "point-and-shoot" model: automatic settings (fast, efficient, inflexible) and manual mode (slow, flexible, costly). This is the part of the book that rests most directly on Greene's own experimental programme, and it is where the corpus's RT cautions bite hardest — a dual-process model of moral judgment is not the same object as dual-process accounts elsewhere in cognition, and the corpus should not let one inherit the other's evidential standing.

⚠️ This summary does not grade Part II. The dossier surfaced the argument, not the evidence base: the studies are cited by name in the endnotes, which this reading did not treat. Grading Part II requires reading the primary articles, not this book. Foyer 08 currently holds one primary article for seven deposits.

---

Part III — the common currency

The move from description to prescription, and the book knows it is making it:

Verified quotations

"one might hope that science will reveal the hidden essence of morality and, thus, not only
describe morality (as we did in chapters 2 , 3 , and 4 ), but also prescribe ."

The proposed metamorality is utilitarian — "deep pragmatism". Two statements about its status:

Verified quotations

"Utilitarianism is not, at the most fundamental level, a decision procedure ."

Verified quotations

"In sum, utilitarianism combines the Golden Rule’s impartiality with the common currency of human
experience."

⛔ The load-bearing empirical premise is declared a conjecture, and declared untested:

Verified quotations

"I hereby conjecture that there is not a tribe in the world whose members cannot feel the pull of
the utilitarian thinking described above"

with, in the same passage:

Verified quotations

"Experimental anthropologists are encouraged to go out and check"

⚠️ This is the single most important sentence in the book for this corpus. The universality of the utilitarian pull is what makes the metamorality common rather than one more tribe's morality — and the author labels it a conjecture and asks someone else to test it. ⛔ Any reuse presenting Greene's metamorality as scientifically grounded contradicts the author's own statement of its status.

He also concedes the limit of the axiomatic route:

Verified quotations

"I cannot prove that moral axioms of this kind do not exist, but it’s fair to say that we should
not count on their arriving anytime soon."

---

Part IV — moral convictions

The "gizmo" argument: our alarm at prototypical violence is a well-designed device, and precisely because it is a device, its operating characteristics are not moral truths.

Verified quotations

"it makes little sense to regard it as infallible and to elevate its operating characteristics
into moral principles"

⚠️ This is the argumentative core of the book, and it is a normative argument built on a descriptive claim — the same seam Haidt draws explicitly (see that summary). Greene crosses it deliberately: the intuitions that object to utilitarianism are inherited automatic settings, hence without normative authority in situations they did not evolve for. The corpus may report this argument; it may not report it as a finding.

One empirical claim in this part is stated with unusual force, and should be flagged rather than carried:

Verified quotations

"I emphasize that this is not a tentative finding.* After decades of research, the weak
relationship between wealth (above a modest level) and happiness is more like a law of human
nature."

⛔ The corpus holds nothing on this. The claim is from 2013, the literature it summarises has been actively contested since, and this book is not a source for it. ⚠️ "This is not a tentative finding" is a rhetorical intensifier, not an evidence level — noting it here precisely so it is not later mistaken for one.

---

Part V — moral solutions, and the disagreement with Haidt

Part V applies the metamorality to abortion and other live controversies, and contains the explicit break with the other pillar of this foyer:

Verified quotations

"According to Haidt, liberals should be more open to compromise with social conservatives. I
disagree."

and:

Verified quotations

"In sum, American social conservatives are not best described as people who place special value on
authority, sanctity, and loyalty, but rather as tribal loyalists"

⚠️ Read this as a disagreement inside the foyer, not as a result. Haidt describes six foundations and declines to rank them; Greene re-describes three of them as tribalism and ranks them. Both moves are argued, neither is measured. ⛔ The corpus must not resolve this dispute by preferring the author it finds more congenial. Where the two disagree, the corpus records the disagreement.

---

What this book is good for in this corpus

• The two tragedies — a clean conceptual distinction (Me/Us versus Us/Them), well stated.
• The dual-process model of moral judgment — as a model, with its primary articles still to be read.
• The author's own labelling of his central premise as a conjecture — a rare and usable piece of

epistemic honesty.

• A documented, explicit disagreement with Haidt — the two pillars of foyer 08 do not agree, and

say so.

What it must not be used for

• ⛔ Utilitarianism as a scientifically established metamorality — the author calls its premise a conjecture.
• ⛔ Wealth and happiness — flagged, ungraded, and 2013.
• ⛔ Any claim that Greene's neuroscience shows his normative conclusion. He argues it; he does not

claim the data prove it.

• ⛔ Citing an endnote as if it were the argument (see the finding above).

---

Provenance

⛔ Tool documentation: _method/OUTIL_dossier_lecture.md.