As some readers may know, I like to make sport out of finding plagiarism. When grading a paper, sometimes you get a sense that a sentence or sentences seem out of place. In the most egregious cases, you might react by saying, “wow, I wish I could write this well!”. In a student paper, that’s a dead giveaway of plagiarism almost always. In other cases, the sentence sticks out for some other reason. Maybe it’s a stylistic or tonal change; the sentences don’t flow together. Or maybe it’s the use of specific types of language, using terms that need to be well-defined before any thoughtful writer would put them into the paper. Whatever the reason, I have a decent track record at uncovering plagiarism.
My crowning achievement (so far) in plagiarism detection came just yesterday, and this success makes my student paper plagiarism detection look like amateur hour. I had been asked to review a paper for a well-regarded journal in the field of human-technology interaction. The paper I reviewed was not particularly well-written. It was disorganized and highly jargonistic. In my comments for the review, this issue was the first one I addressed, and I scanned the paper looking for a problematic sentence to illustrate my point. I picked a sentence essentially at random–the first one that I spotted that fit the desired criteria–and pasted it into my review.
The sentence didn’t stand out to me then, but as I reread my comments before sending them to the editor, something about the sentence seemed off. It’s hard to say exactly what. The sentence used a lot of jargon (exactly why I put it in as an example), and that jargon was out of context. That is, the sentence used terms that should have been defined earlier and used in examples so that when the reader arrived at this sentence, he could tell what the sentence was trying to say. In short, it was hard to imagine the normal writing process (for good or bad writers) producing this kind of sentence.
A quick search on Google Scholar confirmed my suspicions. The sentence had been used, almost verbatim, on two separate occasions: in a book published in 2011 and in a journal article from 2012. Neither of these two sources shared a similar author. (And I cannot know the authors of the manuscript I was reviewing.) The journal article was what the authors of the manuscript had cited (though not quoted; if only they had used a direct quote, then there would be no issue at all!). There were minor differences between versions of the sentence. In the sentence in the manuscript, one adjective was followed by another, like this–“The user experiences top (extraordinary) productivity.” In the sentence in the book, the two adjectives were reversed, and the sentence contained some extra phrasing in the sentence that had been removed in the manuscript.
Minor variations aside, the sentence was still most certainly a case of plagiarism. But plagiarism of who? Let’s say that the authors of the manuscript were the authors of the book, which was published before the journal article (if we go by year of publication, anyway; that may not be an accurate representation of when the writing was done, as the book may have gone to print faster than the journal article). If that is the case, then why cite the journal article instead of the book? If the authors of the manuscript were the authors of the journal article, then they are guilty of plagiarism twice over. And if the authors of the manuscript are separate from both other sets of authors, then they are guilty of plagiarism, as are the authors of the journal article. In short, it’s quite a mess.
I submitted my comments on the review, including a scathing note to the authors about the plagiarism, and then I also emailed the editor. She got back to me right away to say she is talking with other editors to understand what the journal’s policies are in a case like this. Here’s what the policy should be. First, the article must immediately be rejected. Second, the departments and institutions that employ the authors must be informed. Third, if these institutions do not launch an investigation into the plagiarism (the outcome of which could find that the authors made a mistake and did not intend to plagiarize), then the journal itself should launch an investigation; pending the outcome of that investigation, these authors should be forbidden from publishing. Further, all their past work must undergo intensive scrutiny to determine if there are other instances of plagiarism. The investigation may go quickly. For example, if it is a case of self-plagiarism, then a simple warning is all that is required. But publishers cannot take the risk that these authors plagiarized numerous times in the paper or that they have plagiarized in the past. Going through the submitted manuscript to see if there are other instances of plagiarism in it is a good place to start. If none, then I’m inclined to believe this was an honest mistake.
As for me, Michael Braun, Plagiarism Hunter, I will stay on the chase. I wish I had a more systematic way of finding plagiarism, but it really is more instinct and hunch-based than anything. I’m sure I’ve missed instances of plagiarism in the past. And occasionally, I do searches on suspicious-sounding sentences in student papers that turn up nothing. When something catches your eye or sounds strange in your ear, it’s worth a quick search. Sometimes what you find will surprise you.