NYU Press is withdrawing a book over a quotation that began life as someone else's paraphrase
New York University Press has pulled a 2025 monograph from sale and is destroying its remaining inventory because of a single quoted sentence. The interesting part is not that the quotation was wrong. It is that nobody invented it. The words started as a court brief's summary of a journal article, were adopted into a judge's ruling in italics with the note "emphasis added", and were then read as verbatim testimony and placed inside quotation marks by an author working from the ruling. Three defensible steps, no fabrication at any one of them, and a fake quotation at the end.
New York University Press has withdrawn Designer Science: A History of Intelligent Design in America, published in 2025, and told the complainant it was "pulling the current edition from the market" and "destroying all existing inventory". Retraction Watch reported the withdrawal on 28 August, quoting an NYU spokesperson who said the press intends to reissue the book once the passage is corrected. The book's page on the NYU Press site no longer returns content.
The cause was one sentence on page 132: the author, C.W. Howell, wrote that the biochemist Michael Behe "confessed on the stand" during the 2005 Kitzmiller v. Dover trial that the "plausibility of the argument for ID depends upon the extent to which one believes in the existence of God."
Behe never said those words, on the stand or anywhere else. What makes this worth a researcher's attention is how they came to be inside quotation marks anyway.
The chain
The provenance is documented, step by step, in the complaint sent to the publisher on 31 July, which reproduces the relevant pages side by side.
Step one, a real sentence. Behe published "Reply to My Critics" in Biology and Philosophy in 2001 (DOI 10.1023/A:1012268700496). The paper was entered as an exhibit at trial. On the page in question he wrote that "the argument is less plausible to those for whom God's existence is in question, and is much less plausible for those who deny God's existence." That sentence is real, it is his, and it is quoted accurately in the court's own ruling.
Step two, a summary of it. The plaintiffs' proposed findings of fact compressed that sentence into a characterisation: that Behe claims the plausibility of the argument for intelligent design depends on the extent to which one believes in the existence of God. This is a paraphrase, and it is presented as one. No quotation marks, a citation to the exhibit, nothing improper on its face.
Step three, italics. Judge John E. Jones adopted the paraphrase into his ruling, set it in italics, and appended the words "emphasis added". Emphasis is something you add to a quotation. The formatting carried an implication the underlying words did not earn.
Step four, quotation marks. Howell, working from the ruling, read the italics and the "emphasis added" as marking verbatim testimony, put the words inside quotation marks, and located the moment on the witness stand. He has confirmed this account. "After reviewing his notes earlier this month, C.W. Howell, the author, realized the passage was written by Judge John E. Jones about Behe," the NYU spokesperson told Retraction Watch. Howell said he welcomed the chance to correct it.
At no point did anyone fabricate anything. Each hop was individually defensible. The fabrication was emergent, produced by the loss of a single piece of metadata: the fact that the words were a summary rather than a transcription. Once that marker fell away, the formatting of the next document supplied a false one.
What did not catch it
A university press published this. It went through acquisition, review, copyediting and footnoting, and the passage carried a footnote pointing at a real document that really does contain those words. Every mechanical check passes. A citation checker following the footnote arrives at the ruling and finds the sentence exactly as printed.
What caught it was a reader who knew the subject well enough to be suspicious of the voice. Casey Luskin, of the Discovery Institute, was preparing a critical review and told Retraction Watch the line "did not pass the smell test" because it "did not sound like Dr. Behe". Luskin is an interested party, and his letter is an advocacy document written on behalf of an organisation that promotes the position the book criticises. That is worth stating plainly. It is also beside the point: the publisher and the author both examined the evidence and agreed.
Analysis
Two things follow, and neither is specific to this book or this subject.
The first is a practical rule with a long pedigree and poor compliance: verify a quotation against the source that generated the words, not the source you found them in. A quotation reached at second hand is an assertion about what a document says, and it inherits every error made between here and there. This case is unusual only in having a fully reconstructible chain. Most do not, which is a statement about visibility rather than frequency.
The second is about scale. The failure mode here is quoting from a document that is itself quoting, or summarising, another document. That is precisely the operation performed constantly by any system, human or machine, that drafts from retrieved passages rather than from primary sources. A language model given a court ruling has no more access to the distinction between a transcription and an adopted paraphrase than Howell did, and considerably less inclination to think a sentence sounds out of character. The mechanism is old. What is new is how cheap it has become to run it at volume, and how much of the resulting text will carry a footnote that resolves.
There is also a correction problem here that journals do not have. A retraction notice attaches to an article and travels with it. A monograph has no equivalent, which is why the remedy in this case was withdrawal, pulping and a reissue. Copies already sold keep the sentence, and so does anything that has already cited it.
What to watch
Whether the reissued edition carries a visible note explaining what was changed and why, rather than a silent correction. A quietly fixed second edition leaves the first edition's readers with no way to discover that the passage they relied on is gone.
Sources
- Retraction Watch, University press removes book about intelligent design for misquoting ID proponent, 28 August 2026
- Discovery Institute, letter to New York University Press, 31 July 2026, with reproduced pages
- Kitzmiller v. Dover Area School District, 400 F. Supp. 2d 707 (M.D. Pa. 2005), full ruling
- Plaintiffs' Proposed Findings of Fact and Conclusions of Law, ACLU of Pennsylvania
- M.J. Behe, Reply to My Critics: A Response to Reviews of Darwin's Black Box, Biology and Philosophy, 2001, DOI 10.1023/A:1012268700496