Oliver Scott Curry, Daniel Austin Mullins, Harvey Whitehouse (2019), Current Anthropology 60(1), 47–69 (followed by invited peer commentaries) — DOI-10.1086/701478
Key takeaways
- In the ethnographic records of the 60-society HRAF Probability Sample, seven cooperative behaviours predicted by game theory are morally valenced positively in 961 of 962 observations (99.9%). These are helping kin, helping one’s group, reciprocating, bravery, deference to superiors, dividing disputed resources, and respecting prior possession. The authors treat them as candidate universal moral rules.
Seeded from the abstract and a skim, before the reading in NOTE-034. What follows is what the work says about itself.
The theory of morality-as-cooperation says morality is a set of biological and cultural solutions to recurring problems of cooperation, identified through non-zero-sum game theory. It predicts that seven specific cooperative behaviours will be judged morally good in every culture. The authors code the moral valence of these behaviours in the ethnographies of 60 societies. The valence is uniformly positive, most of these morals appear in most societies, and they occur equally often across world regions. The authors propose the seven as universal moral rules and the theory as a unified account of morality for anthropology.
Standing in the record
Filed from the survey of 2026-09-25 of work the anthology set aside as out of scope (tier B): 115 seconds of active reading over 7 sessions in the papers-feed tracker. It was filed Deferred, unread. NOTE-034 is the close reading of 2026-09-25, and it placed the work: Proposed — the clearest statement of morality-as-cooperation, with an unusually good commentary set (Bloom; Smith & Kurzban; Wong) that states the objections exactly. Its empirical test is weakly diagnostic: the codebook defines the behaviours in virtue terms, the authors’ footnote 3 recodes conflicts as cooperation-vs-cooperation, and the first-pass selection was not blind. It would be settled by blind coding of all moralized content in the same sample, testing whether non-cooperative domains (purity, sex, harm) are moralized as often.
When seeded, before the reading in NOTE-034, the entry gave a priority for a deeper reading: medium — 7 sessions but only 115 s. The core claim is simple and captured here, and the commentaries are where a deeper read would add value.
What the seeding asked a deeper reading to check (NOTE-034 is that reading):
- An empirical anchor for game-theoretic accounts of morality. It is relevant to the ethics cluster and to the cooperation thread (c34 ergodicity, c31/c32 collective behaviour).
- The test is lopsided: coding records only the valence of behaviours that appear, so absence of evidence of negative valence is weak evidence. Check how the paper handles the fact that most morals were “observed in most” but not all societies (Table 2 coverage), and what the commentaries criticize.
- Check how independent the coding was (who coded, inter-rater reliability) in the methods and supplement.
Access when seeded: Crossref (issued 2019-02-02) and the version of record from the Oxford Research Archive (ORA; 23 pp. main article plus a 6 pp. supplement dated May 2016), which I read: abstract, methods, results, discussion opening, conclusion, and the start of the commentaries (Paul Bloom et al.). The UChicago PDF and ResearchGate returned 403. Date conflict: Crossref gives 2019-02-02, while ORA lists publication 2019-02-08 and acceptance 2017-12-22. A ResearchGate record ID suggests an earlier 2016 version circulated, but that is unverified. The journal’s keywords were not seen, so the keywords above are my own.