Rejecting a quotation from publication and erasing it from the record are two different decisions. This page is the record.
What this page is
Every paper this article rests on was downloaded and read locally, so nothing failed on wording. Four items were held back for other reasons: two because the source is not citable, one because the objection it carries does not apply to the Y chromosome, and one because the numbers were verified but the sentence containing them was never transcribed.
Status words.
- REFUTED — the attribution or wording is demonstrably wrong. A better copy will not help.
- UNVERIFIABLE — no route to the named source has been found yet. A statement about this project's reach, not about the claim.
- NOT CITABLE — the wording is genuine and correctly attributed, but the venue cannot carry the weight the article would put on it.
- WRONG ARGUMENT — the wording is genuine and the objection is real, but it does not apply to the question this article is about.
- PRINTED WITH A FLAG — the substance is verified, but something about the citation is incomplete, so the article prints it as a paraphrase and says so.
Nothing on this page was established with an automated summarising fetcher. Every check was made against downloaded bytes.
1. Swamidass on the fifty-fold gap
Wording: "He is correct that YEC and evolutionary models require totally different mutation rates. He predicts that mutation rates are far higher (about 50x higher) than evolutionary science would predict. That is a major point of agreement and common ground … The issue, however, is that he hasn't quite realized that the measured rates of mutation are closely aligned with evolutionary science, and about 50x lower rate than the YEC model predicts." And, on the mechanism: "he 'forgets' to subtract out the error when he computes the rates, which is what gives him the higher mutation rate (which is erroneous)."
Source: S. Joshua Swamidass, posts 2 and 6 in the Peaceful Science discourse thread "Testing Jeanson's Model: Y Chromosome Mutation Rates," opened 16 July 2020, 106 posts. Fetched as Discourse JSON with curl and stored in the source library, so the post text quoted here is the stored text.
Status: NOT CITABLE.
Why it was held back. The wording is genuine and correctly attributed. But a forum post is not a published result, and the claim it makes is exactly the claim this article's own computations establish independently: the filter-ladder table shows where the factor comes from, and the coverage-versus-rate test shows the causal premise fails inside Helgason's own data. Leaning on a forum post for a load-bearing claim, when a reproducible computation is available, would weaken the article rather than strengthen it.
What would change this. A peer-reviewed critique of Jeanson & Holland 2019 making the same point. Searches of Europe PMC, OpenAlex and the open web found none. That absence is itself reported in the article.
2. Duff on circular confirmation
Wording: "when your critics point out that you are equating two quantities that population genetics has well-established reasons to distinguish, you cannot simply respond by saying 'my predictions were confirmed' unless you have first addressed the objection. If the foundational methodology is flawed—if you are, in fact, treating a per-generation mutation rate as a substitution rate and then extrapolating backward—then any 'confirmation' built on that same flawed methodology is circular."
Source: Joel Duff, "How Dr. Jeanson's Rhetoric Substitutes Mockery for Methodology," Naturalis Historia, 17 March 2026. Downloaded and stored in the source library, read locally.
Status: WRONG ARGUMENT, and held back deliberately.
Why it was held back, and this is the most important entry on the page. The wording is genuine and Duff's objection is correct against Jeanson's mitochondrial clock, where it is decisive and where the Mitochondrial Eve article makes it at length. It does not transfer to the Y chromosome. Mitochondria are present in hundreds of copies per cell and show heteroplasmy, so a family study and a phylogeny really can be counting different quantities, which is why the two mitochondrial rates differ by roughly twentyfold. The Y is haploid and does not recombine, and there is no heteroplasmy: a new Y mutation is either transmitted or it is not, and if it is, it stays on that lineage's branch permanently. On the Y a pedigree rate and a substitution rate are the same quantity, which is exactly why the six published estimates agree to within a factor of 1.36.
Importing this objection into the Y article would have been a real error, and a knowledgeable reader would have used it to dismiss everything else. The Y argument fails somewhere else: at the filter choice, at the causal claim about coverage, and at the arithmetic.
What would change this. Nothing about the wording. If a future paper shows a mechanism on the Y that makes pedigree and substitution rates diverge, the objection would become relevant and could be cited then.
3. Thomson's "59,000 years"
Wording as it circulates: that Thomson and colleagues dated Y-chromosome Adam to 59,000 years, a figure quoted on the creationist side for a decade as evidence that the Y tree is young.
Source: Thomson, Pritchard, Shen, Oefner & Feldman, "Recent common ancestry of human Y chromosomes," PNAS 97(13):7360–7365, 2000. Read in full from the PMC copy in the source library.
Status: UNVERIFIABLE as a clean sentence, and not printed.
What was tried and what was found. The paper is on disk and was searched directly. It does say "the expected time to the most recent common ancestor is remarkably short, on the order of 50,000 years," and the 59,000 figure appears in a sentence that runs across a line break and into a comparison clause, so the fragment recorded in the research dossier, "Y chromosomes is of the order of 59,000 years which agrees closely with," is a partial sentence rather than a quotable one. The number is real. The quotation is not clean enough to print, and reconstructing it from a two-column extraction risks exactly the kind of defect this project audits for.
Why it was not chased further. The article does not need it. Thomson's estimate is calibrated on an assumed human-chimpanzee split of 5 million years, its own field has since moved it by a factor of two to six, and even its lowest figure is thirteen times the Flood. The point it would have supported, that mainstream rates were once set by an assumed ape divergence, is made instead from Cruciani 2011 and Mendez 2013, both of which say it explicitly and in clean, quotable sentences.
What would change this. A clean single-column extraction or the publisher PDF, if a future revision decides the paper is worth the words.
4. The Kong paternal-age slope
Substance: a slope of 2.01 de novo mutations per year of paternal age, and a divisor of 2.68 Gb of callable autosomal sequence. These two numbers are the entire input to this article's computation of patriarchal drive.
Source as used, before verification: both figures were taken from Mendez et al. 2013 and both were attributed to Kong et al. 2012.
Status: RESOLVED AND RELEASED, 6 September 2026, with one attribution corrected.
What was found. Kong 2012 was obtained by two routes: the NIH author-manuscript PDF, which was already on disk under an orphan key with no index.tsv row, and a fresh PMC3548427 fetch with bin/pmcget.py. The slope is Kong's, verbatim on author-manuscript p. 3: "The number of mutations increases with father's age (P = 3.6x10-19) with an estimated effect of 2.01 mutations per year (standard error (SE) = 0.17)." The 2.68 Gb divisor is not Kong's. Kong prints 2.63 billion bases of effective genome coverage on p. 5; 2.68 × 10⁹ is Mendez's own figure, described in Mendez as "the length of autosomal sequence covered in each of the 78 trios." The article now attributes each number to its own source and says that Kong's smaller divisor would make patriarchal drive 2% larger, so the choice made here is the one unfavourable to this article's conclusion.
One further caveat now printed. Kong's regression pools paternal and maternal counts, "as they could not be reliably separated without data from a grandchild," and mother's age correlates with father's at r = 0.83. So 2.01 is an upper bound on the paternal-only slope, which again biases the computation towards the young-earth argument.
The locator was also wrong. The Mendez footnote said "Methods." Mendez 2013 is an AJHG Report with no Methods section; the likelihood-method sentence is in body paragraph 2 and the 2.01 / 2.68 figures are in the paragraph immediately after Figure 1. Anchored against the raw PMC HTML: the figcaption element ends at "…used by Cruciani et al.6", and "Assuming that mutations in the X-degenerate portion…" opens the next paragraph.