Rejecting a quotation from publication and erasing it from the record are two different decisions. This page is the record.
What this page is
Five items met in the research were not printed in the article as they circulate. None has been deleted. This page is for the next researcher: it gives the wording, the claimed source, the route tried and what would overturn the verdict.
Status words.
- MISUSED: the text is real and correctly attributed to its author, but the way it is usually quoted changes what the author meant.
- UNVERIFIABLE: no route to the named source has been found yet. That is a statement about this project's reach, not about the quotation.
Nothing here was established with Claude's WebFetch tool. Every quotation was read off downloaded bytes or archive.org search-inside paragraphs, as stated in each entry.
1. Dixon's "4,000 years BP or 40,000 years BP", applied to Indo-European
Wording as it circulates. Gray and Atkinson open their 2003 Nature paper with it:
"Dixon epitomizes this view with his assertion that, based on the linguistic data, the age of Indo-European 'could be anything – 4,000 years BP or 40,000 years BP are both perfectly possible (as is any date in between)'."
Claimed source. R. M. W. Dixon, The Rise and Fall of Languages (Cambridge University Press, 1997).
What was found. The words are Dixon's, but they are about the hypothetical ancestor of all Australian languages, not about Indo-European. Read by archive.org search-inside of risefalloflangua0000dixo (printdisabled), bin/iapdsearch.py '"40,000 years"', with leaf = page + 8 from BookReaderJSIA.php (leaf 45 = p. 37, leaf 160 = p. 152). Page 48:
"While working on the languages of Australia, over the past 30 years, I have at times experimented with the hypothesis that the languages are related as one family and then looked for a family tree model (an alternative scenario is discussed below). The question that then presented itself was: what could have been the date of the proto-language? After careful consideration, it seemed that the only honest answer is: it could be anything - 4,000 years BP or 40,000 years BP are both perfectly possible (as is any date in between)."
Dixon's statements about Indo-European are on p. 47 and the top of p. 48, and the article quotes those instead. The project's own first dossier repeated Gray and Atkinson's attribution before the check.
Status: MISUSED. The article prints Gray and Atkinson's sentence only to correct it, beside the Australian context and Dixon's real Indo-European statements.
What would change the verdict. Nothing: the page is read. A Dixon statement elsewhere applying the same range to Indo-European would make Gray and Atkinson's wording fair, and none was found.
2. Hodge's Babel date of 2242 BC
Wording as it circulates. "The Tower of Babel (2242 BC)", given as Bodie Hodge's date.
Claimed source. A sermon summary of a talk by Hodge on sermons.logos.com, found by web search. Hodge's book Tower of Babel (Master Books, 2013) was expected to state it.
What was tried. The book is not on archive.org by title or creator search. Google Books volume sBE99WGQvpYC was searched through bin/google-books-snippets.sh; it returned wording without page numbers for other passages, then stopped returning results under a rate limit. The date was not seen in the book's own words.
Status: UNVERIFIABLE. The article does not attribute 2242 BC to Hodge. It gives 2242 BC as the date Answers in Genesis prints, which is Ussher's dispersion.
What would change the verdict. A page-locked reading of Hodge 2013: a library copy through Evergreen or interlibrary loan, HathiTrust full-text search, or Google Books after the limit clears, with queries "2242", "14 language families" and "constructed language". That reading would also supply the page numbers the article's Hodge footnotes lack.
3. Roger Lewin in New Scientist, as quoted by Taylor
Wording as it circulates. Quoted by Charles Taylor in "The origin of language", Journal of Creation 11(1), 1997:
"'unlike biological species, languages change at an astonishing rate, as anyone who has struggled with Chaucer will attest. As a result, most historical linguists agree that going back more than 5,000 to 7,000 years is a futile enterprise.'"
Claimed source. Taylor's note 5: R. Lewin, "Ancestral voices at war", New Scientist 128(1722):25, 1990.
What was tried. The New Scientist issue is not in the source library and was not reached. The wording above is verbatim from Taylor's web text, read by curl.
Status: UNVERIFIABLE at source. The article does not quote it. The point it makes, that the comparative method's horizon is a few thousand years, is cited to Nichols, Crowley, Hammarström and Campbell and Poser instead.
What would change the verdict. The New Scientist archive page or a scan of the 1990 issue, to confirm the wording and whether the second sentence is Lewin's own or his report of others.
4. A commentator on Bergsland and Vogt, Current Anthropology 1962, p. 147
Wording. From the published comments on Bergsland and Vogt's paper, archive.org microfilm sim_current-anthropology_1962-04_3_2, leaf 38, p. 147, read with bin/iapdsearch.py:
"now that Bergsland and Vogt have adduced their well controlled counter-examples, the most devastating of which is the comparison of Old Norse (L1) and its lineal descendant Modern Icelandic (L2). As computed in glottochronology, time elapsed between these instances of L1 and L2 is a century or two or, by a variant method of counting, zero ... But the time elapsed between L1 and L2 is a thousand years according to historiographic ways of measuring time."
Claimed source. One of the commentators printed after the article: Dell Hymes, Isidore Dyen, Sarah Gudschinsky, Robert Lees and others.
What was found. The paragraph was returned without its heading, so the commentator's name was not read off the page.
Status: UNVERIFIABLE as to author. The article does not use it; Bergsland and Vogt's own figures make the same point.
What would change the verdict. A sweep of leaves 37 and 38 of the microfilm item for the commentator's name and address, which head each comment.
5. Four passages of Hodge's Tower of Babel (2013), read once
Wording, as recorded from Google Books search-inside snippets by the research agent of 2026-10-10 (research/rates-attestations-creationist.md, entry hodge-2013):
"From Japheth (Genesis 10:2-5), there came 14 language families; from Ham (Genesis 10:6-20), there came 39; from Shem (Genesis 10:22-31), there came at least 25 (excluding Peleg and other children who may have just been born)."
"language families resulted from God's confusion of the language at Babel. Since that time, the original language families have grown and changed into the abundant number of languages today."
"there were at least four languages, and likely five Greek tongues came out of Babel, each being unique."
"bringing a multitude of language families to the Americas. So some isolates may reflect similarities to certain language families in other parts of the world. But we should also consider the constructed language approach."
Claimed source. Bodie Hodge, Tower of Babel: The Cultural History of Our Ancestors (Master Books, 2013), Google Books volume sBE99WGQvpYC. No page numbers were returned.
What was tried by the verifier, 2026-10-10. bin/google-books-snippets.sh with '"minimum number of language families"', '"there came 39"', '"14 language families"' and 'Japheth': 0 results each; a plain Google Books search URL was redirected to Google's "sorry" page, and the Books API answered 429 (quota exhausted). Web search for the 14/39/25 wording found nothing. The book is not on archive.org, and no library ebook listing (hoopla, Libby) was found. A second route was found for one passage only: Goodreads' quotation page for the book (goodreads.com/quotes/11658125, fetched with curl) prints "So I suggest that the 78 is a minimum number of language families (that is a tally of those listed in Genesis 10 subtracting Noah, Shem, Ham, Japheth, and Peleg)." That sentence stays in the article, cut to the words both routes share.
Status: UNVERIFIABLE (single route). Nothing suggests the four passages are wrong; they were read once from Google's own snippet bytes and could not be read a second time. The article now states Hodge's position from the twice-read sentence and does not print these four.
What would change the verdict. A page-locked reading of the book (a print copy through Evergreen or interlibrary loan, or Google Books once its limit clears), which would restore all four with page numbers.