Last updated on 22 Jun 2018
Some interesting things in the web lately. One is a new system that purports to find cases of plagiarism. Science reports on the new Dèjá Vu database [news summary here; both behind the Science paywall; check this and this if you don’t have access], which checks content of science papers for copied material. But there’s a controversy, of course, because many of the papers identified are review articles, which are supposed to have copied material. Also, papers cite each other and have to repeat the materials and methods when that is appropriate. So the database has added some categories: Distinct, which are unique papers; Sanctioned, in which reviews and similar papers are deemed appropriate copying; Update, in which previous papers are given updates; Unverified, which means what you might think. None of these are considered plagiarism (yet). And then there are papers that are considered Duplicate, which are inappropriate copying. Like all automated systems of referencing, this has problems and limitations. I really hope that employers don’t start using this as a metric the way other automated indices have been used.
While talking about accuracy, the lawsuit against Jared Diamond and the New Yorker, reported by Stinkyjournalism.org, is apparently a case of inadvertent dishonesty and lazy fact checking by the New Yorker, in which Diamond used real names to report mass murder in the Papuan highlands, based on stories told to him by his local driver, all of which turn out not to be true, and which could have been checked according to the report. [Hat tip to Columbia Journalism Report, via Bioephemera, who notes that Diamond is using the defence “It wasn’t science”]. Diamond is most likely at fault here for being naive, but the New Yorker is evidencing the incredible decline in reliability and accuracy even of the high-end media outlets. The slander may very well lead to tribal violence, as a carefully negotiated settlement between tribal groups who had a conflict (but not mass murder) is threatened, and individuals, including Diamond’s informant, may be personally at risk.
It pays to be critical of all sources.
“many of the papers identified are review articles, which are supposed to have copied material”
Review articles are supposed to review concepts discussed in original research articles, but they should not repeat the content of those articles. The real exceptions are those articles that are supposed to be updates of previously published articles. Whenever an author reuses content published elsewhere, it’s plagiarism (whether the content was published by another author or by that author).
Is the moral of the story “don’t trust Diamond”?
This is astonishing. This is the New Yorker, whose fact-checking is legendary. Perhaps a bit of a reminder, though, that Diamond is not an anthropologist nor a journalist. He’s a writer of popular books.