---
id: "20260709-2222-hop-harpending-race-science"
title: "Henry Harpending co-authored bioarchaeology's foundational Osteological Paradox paper and a PNAS paper on accelerated human evolution, then became a documented race-science figure"
type: "capture"
status: "promoted"
promoted_to: ["30-notes/claim-osteological-paradox-skeletal-statistics-confound-past-health.md (Claim 1)","30-notes/claim-hawks-2007-human-adaptive-evolution-accelerated-recently.md (Claim 2)","30-notes/claim-harpending-became-a-documented-race-science-figure.md (Claim 3)"]
not_promoted: ["Standalone note on *The 10,000 Year Explosion* (2009) — folded into Claim 3 note as supporting evidence; thin and unsourced (no URL) as a standalone.","Standalone note on the Ashkenazi-IQ paper (Cochran, Hardy & Harpending 2006) — folded into Claim 3 note; capture gave year inconsistently (2005/2006), citation routed to the SPLC-quotes verification question.","The 'foundational-methods co-author -> race-science figure' pivot as its own claim-note — expressed through wikilinks and commentary across the three notes rather than minted as a fourth note; it is an emergent juxtaposition, not an atomic world-claim (and the capture itself flagged the gotcha risk).","Further leads (Gregory Cochran's own profile; Raoul Naroll's floor-area method; William Petty / political arithmetic) — explicitly saved, un-chased hooks; left for the topic queue."]
questions_routed: ["50-questions/question-verify-harpending-2007-acceleration-field-acceptance.md","50-questions/question-verify-harpending-splc-race-science-quotes-primary.md"]
origin: "hop-batch"
model: "claude-sonnet-5"
date_created: "2026-07-09T00:00:00.000Z"
hop_chain: ["seed: claim-chandler-four-thousand-years-foundational-urban-dataset.md -> paleodemography / osteological-paradox age-at-death estimation bias (max_cosine 0.614, adjacent)","paleodemography osteological paradox -> Henry Harpending, its co-author, and his later race-science career (max_cosine 0.671, adjacent)","Henry Harpending -> 2007 PNAS 'Recent acceleration of human adaptive evolution' paper he co-authored (max_cosine 0.730, adjacent)","2007 PNAS accelerated-evolution paper -> search for mainstream population-genetics rebuttal (no clear source surfaced; dead end, chain closed)"]
novelty_max_cosine: 0.653
tags: ["historical-demography","bioarchaeology","population-genetics","scientific-racism","source-criticism","credibility-paradox"]
source_url: "https://os.pennds.org/archaeobib_filestore/pdf_articles/CA/1992_33_4_Woodetal.pdf"
source_author: "James W. Wood, George R. Milner, Henry C. Harpending, Kenneth M. Weiss"
source_date: "1992-08 (Current Anthropology 33:4, pp. 343-370)"
source_quote: "by James W. Wood, George R. Milner, Henry C. Harpending, and Kenneth M. Weiss"
source_tier: 1
provenance: "hop-protocol session, headless, 2026-07-09"
---


**Claim 1.** Henry C. Harpending co-authored "The Osteological Paradox" (Current Anthropology, 1992), still one of bioarchaeology's most cited methods papers, which argued that skeletal statistics cannot straightforwardly reveal past population health because of "demographic non-stationarity, selective mortality, and unmeasured, individual-level heterogeneity in the risks of disease and death." Byline: "by James W. Wood, George R. Milner, Henry C. Harpending, and Kenneth M. Weiss." Source: primary journal scan (JSTOR/Current Anthropology). Tier 1.

**Claim 2.** The same Henry C. Harpending (Dept. of Anthropology, University of Utah) co-authored Hawks, Wang, Cochran, Harpending & Moyzis, "Recent acceleration of human adaptive evolution" (PNAS, 2007), which concluded from HapMap SNP data that "selection has accelerated greatly during the last 40,000 years" and that "the recent rate must be one to two orders of magnitude higher than the long-term rate to explain the genomewide pattern." source_url: https://pmc.ncbi.nlm.nih.gov/articles/PMC2410101/ — Tier 1 (peer-reviewed primary paper).

**Claim 3.** Harpending went on to co-author *The 10,000 Year Explosion* (2009) and a paper on Ashkenazi IQ, and by the 2010s was speaking at explicitly white-nationalist-adjacent venues. Per SPLC's documentation, he told the 2009 "Preserving Western Civilization" conference (sharing a stage with Peter Brimelow and J. Philippe Rushton) that "group differences, as far as we know, are in the DNA," and at the 2011 H.L. Mencken Club he described medieval European law enforcement's killing of "impulsive, violent, innumerate, illiterate, and lazy" individuals as "eugenics." source_url: https://www.splcenter.org/resources/extremist-files/henry-harpending/ — Tier 3 (advocacy-org documentation, but quotes are directly attributed and verbatim).

## Why this was hop-worthy

One of four co-authors of a still-standard bioarchaeology methods paper — the kind of foundational, uncontested infrastructure a field builds on without a second thought — is the same person who later became a documented, self-acknowledged (if disputed) figure in scientific racism, discovered via the paleodemographic mechanism-question this session was redirected toward.

## Further leads

- The PNAS 2007 paper (0.730 cosine — the closest score of this whole chain) sits right next to the vault's existing Mayr/punctuated-equilibrium cluster; a dedicated hop comparing "accelerated adaptive evolution" claims against orthodox evolutionary-rate theory is un-chased.
- No mainstream population-genetics rebuttal of the 2007 paper's "100-fold" claim was found in this session — worth a dedicated search (ascertainment bias / demographic-growth confounds were the closest leads).
- Gregory Cochran (co-author, both PNAS 2007 and *10,000 Year Explosion*) is himself an unfamiliar-name hook not yet followed.

> [!note] Seek's commentary:
> There's a tempting echo with the seed note's Chandler "credibility paradox" (distrusted at the level of interpretation, indispensable at the level of data) — except inverted: Harpending is trusted mainstream infrastructure (PSU/Utah, PNAS, still-cited methods) whose *later* interpretive work is what's now distrusted. Pleasing pattern-match, not evidence of anything deeper; flagging it the same way the seed note flagged its own.
> — Seek

## Hop chain

Hop 1: paleodemography / the osteological paradox — https://pmc.ncbi.nlm.nih.gov/articles/PMC2410101/ (background) and https://os.pennds.org/archaeobib_filestore/pdf_articles/CA/1992_33_4_Woodetal.pdf (primary)
- Hook type: mechanism question (redirect from seed's dataset-level demography to the archaeological method used to estimate ancient populations from skeletons)
- Hook: "age-mimicry" — skeletal age-at-death estimates trained on modern reference samples systematically reproduce the reference population's age structure rather than the true target population's
- Why followed: explicit redirect target (archaeology methods / demography), scored adjacent (0.614) and away from the seed's own dataset topic
- Key findings: The Osteological Paradox (Wood, Milner, Harpending & Weiss 1992, Current Anthropology) is the field's foundational critique — skeletal statistics confound demographic non-stationarity, selective mortality, and unmeasured frailty heterogeneity, so naive interpretation of skeletal health/age data is unreliable.

Hop 2: Henry Harpending (person behind the 1992 paper) — https://en.wikipedia.org/wiki/Henry_Harpending
- Hook type: the person behind the thing
- Hook: same Harpending, University of Utah anthropologist/population geneticist, later became a race-science figure documented by SPLC
- Why followed: zoom-out per alternation rule after the mechanism-level hop 1; highest-ranking hook type available (person not yet in vault) once cross-domain-bridge candidates were checked
- Key findings: Harpending pioneered population-genetics/geography methods in the 1970s (fieldwork with the !Kung), then co-wrote *The 10,000 Year Explosion* (2009) and a 2005 Ashkenazi-IQ paper, and was profiled in SPLC's extremist files for genetic-determinist claims about race.

Hop 3: "Recent acceleration of human adaptive evolution" (Hawks, Wang, Cochran, Harpending & Moyzis, PNAS 2007) — https://pmc.ncbi.nlm.nih.gov/articles/PMC2410101/
- Hook type: surprising claim + cross-domain bridge (mid-band cosine 0.730 to the vault's existing Mayr/punctuated-equilibrium evolutionary-biology cluster — the highest-value bridge signal of the chain)
- Hook: the paper's headline quantitative claim that human adaptive evolution accelerated ~100-fold in the last 40,000 years
- Why followed: zoom-in per alternation after hop 2's zoom-out to the person; highest novelty-assist score in the chain and a genuine link to an existing vault cluster
- Key findings: Using 3.9M HapMap SNPs, the authors argued a constant historical substitution rate can't explain observed heterozygosity, recombination patterns, or linkage-disequilibrium block counts — the data fit an accelerating-rate model instead, with the rate increase attributed largely to population growth since agriculture.

Hop 4: search for critique/rebuttal of the 2007 paper's methodology — (no single source; general search)
- Hook type: mechanism question (does the field accept this number?)
- Hook: whether "100-fold" survived scrutiny given known confounds like ascertainment bias and demographic history mimicking selection signals
- Why followed: natural completion of hop 3's claim before closing the chain
- Key findings: none conclusive — general literature on ascertainment bias and selection-vs-demography confounds exists, but no specific rebuttal of this paper surfaced. Diminishing returns; chain closed here.

Saved hooks not followed:
- Gregory Cochran (PNAS 2007 and *10,000 Year Explosion* co-author, also known for the "gay germ hypothesis") — from hop 3 — unfamiliar-name hook, likely as colorful as Harpending, saved for a future chain.
- Raoul Naroll's archaeological "floor-area method" for estimating prehistoric settlement population from house-floor area — from initial candidate scan off the seed note — mechanism hook, scored adjacent (0.634) but less bridge-y than the paleodemography branch; saved.
- William Petty / "political arithmetic," the 17th-century origin of demography as a discipline — from initial candidate scan — person/historical hook, scored adjacent (0.686, closest to an existing blind-mathematicians note) but no clear bridge; saved.

post-worthy: maybe — the Harpending pivot (mainstream methods co-author -> documented race-science figure) is a strong, well-sourced story, but it risks reading as a "gotcha" about one scientist rather than the kind of vault-extending connection the protocol favors; worth a second pass before Cali decides.