John McWhorter - After Claudine Gay and Neri Oxman, the search for copied language is out of hand
|
By John McWhorter
In December, a group of outside scholars appointed by a Harvard board was roundly criticized for describing the plagiarism that ultimately contributed to former President Claudine Gay’s resignation as “duplicative language.” This description was seen by many as an effort to minimize Gay’s transgression. And it was. But I think the board was on to something useful nevertheless. The term “plagiarism” is overstretched.
Ironically, Bill Ackman, the billionaire hedge fund manager who worked so hard to push Claudine Gay out of her job, would now seem to agree. In a twist so uncanny you couldn’t have written it any better, Ackman’s wife, Neri Oxman, a former M.I.T. professor, appears to have lifted chunks of her dissertation from other sources, including Wikipedia.
In the blink of an eye after these revelations, Ackman acquired an exquisite sensitivity to the difference between real plagiarism and the other, accidental-word-copying kind. Yet the difference he suddenly understands is one that anyone can. To think that neither Gay nor Oxman “really” plagiarized, or to believe that the sanction for such errors should be less severe, is an entirely reasonable point of view.
But here in this world, in this language, the term “plagiarism” still covers both “real” plagiarism — the theft of another person’s ideas — and the use, perhaps inadvertent, of another person’s language. For that reason, I continue to think Gay was correct to step down — especially given that Harvard, like many universities, explicitly defines plagiarism for undergraduates in the “old” way. If accidental cutting and pasting could theoretically get a sophomore suspended, repeated instances of the same should lead an administrator to step down. Meanwhile, given his support for the attacks on Gay, Ackman’s defense of his wife is a hot mess: He should knock off the sputtering and just eat crow.
Leave it to a linguist to say this, but we need another word. In this case, we need a word for the relatively minor, “duplicative language” version of plagiarism.
To present someone else’s ideas as one’s own is unquestionably wrong, in academia and elsewhere. However, to cite boilerplate statements — the assumptions basic to a field, for instance — word for word, or close to it, without citing the person who typed the words originally is something different, and vastly less egregious. I would argue, in fact, that there may be nothing wrong with it at all, in particular when it is done accidentally.
For example, let me reach for a book that happens to be close at hand as I write: “Writing With Style” by The Economist’s Lane Greene. (It’s a lovely book, by the way. I vigorously recommend it as an alternative to Strunk and White’s hopelessly arbitrary “The Elements of Style.”) I find this passage at the start of one of the chapters: “In recent years researchers in artificial intelligence have unveiled systems that seem to ‘write’ without any human involvement. The best of these churn out remarkably convincing prose.”
This is a simple statement of fact, provided as background for the meat of the chapter. It’s not a notable idea, and it’s not written with meaningful style. But if that sentence were to appear in a book of mine, even decades later, precisely as written or with just a couple of words changed, I’d be guilty of plagiarism. However, I’d be fine if I just reworded the thought minimally as: “Artificial intelligence researchers have recently developed algorithms that seem to ‘write’ by themselves, with the most advanced of them easily generating text that is uncannily similar to what a human would write.”
A few phrases flipped and a few words changed, but the precise same content — so what purpose would my minimal rewording have served? It would seem to be a kind of politesse at best, prioritizing form over content. Of course, there are instances where the form should be elevated over content. For example, where fiction is concerned, we would probably consider word-for-word lifting unacceptable regardless of intent or of what the words “meant”: In that genre, the composition of sentences is a primary component of the art. And to be sure, there are plenty of gray zones between fine writing and technical prose that would be less easy to adjudicate. But gray zones don’t justify neglecting a clear issue.
Most of the “plagiarism” examples that have caught up Gay and Oxman and others of late are not anything that aspires to artistic creation or the presentation of a new idea. They are boilerplate statements in books or academic papers whose reuse is only uncovered through the use of high-tech language searches. (I shudder to think how much time and money people are going to start wasting trying to smoke out the cut-and-paste plagiarisms of people they don’t like.)
I, for one, would be quite OK with someone else lifting boilerplate statements from my work. It would deprive me of nothing and would at least suggest they had read it! And I do not appear to be alone, given that some scholars said, during Gaygate, that they did not consider themselves wronged in having had blocks of their text lifted. In an alternative universe in which we had two words for “plagiarism,” that entire discussion could have been clearer and more efficient.
Right now, I am finishing a chapter for a volume in which a large number of linguists have been asked to contribute an article covering a certain sub-area or a summary of their prior work on it. Mine is the summary kind. (This is a type of article you start being asked to do at a certain stage in your academic career.) It’s been a slog, because of course one is not supposed to “plagiarize” even one’s own work. I could easily get this one done in an afternoon by just cutting and pasting paragraphs from my earlier work — or paying someone to do it for me — and then tacking on a few newer citations plus an introduction and conclusion.
But instead, I must reword each sentence at least somewhat — much as I did above with that passage from “Writing With Style.” All of this for words I wrote as far back as 30 years ago. (And because these were academic books and articles, it’s not as if most of this material ever exactly got around.) It is, if you ask me, an empty formality.
Cutting and pasting is not the same as stealing ideas. “Plagiarism,” as a term, should be restricted to the latter. That means we need a new term for the former. There is no reason the new term has to be a formal one derived from Latin like “plagiarism” — or “duplicative language” for that matter. And in fact perhaps it should not be. Latinate words tend to look and feel more intimidating, handy for things you get in trouble for. Our new term could be less menacing, in line with referring to something that should be sanctioned less, if at all. Perhaps we already have the term: “cutting and pasting” — as distinct from, rather than a form of, plagiarism.
In the next newsletter, I will continue these thoughts on the use, misuse and evolution of language.
|
Replies are disabled on threads older than 7 days.
|
|