More thoughts about Turnitin
posted by Alfred Yen
Thanks to all who have posted interesting comments about Turnitin. I particularly appreciate the thought behind the comments arguing that Turnitin has no fair use defense, but I haven’t changed my mind. Let me explain why I still think it’s fair use in the limited space this forum permits.
At the outset, I’ll concede factors 2 and 3 for the plaintiffs, although in at least some cases the nature of the copyrighted work may be sufficiently factual (i.e. academic research paper) that 3 isn’t the slam dunk that some have suggested. That having been said, remember that factors 2 and 3 alone are not enough to swing fair use for a plaintiff, as we learned in the Sony case, where factors 1 and 4 insulated widespread wholesale copying of fictional works.
It’s true that Turnitin is engaged in a commercial endeavor, but that alone doesn’t swing a fair use analysis either. In Campbell v. Acuff-Rose, the Supreme Court understood that people create parodies for commercial purposes, but still reversed the lower court’s pro-plaintiff decision. In doing so, the Court astutely observed that commercial use is a factor that weighs against fair use, but that a great deal depends on whether the defendant’s use substitutes for the plaintiff’s. Although the Turnitin’s use is commercial, its purpose in doing so is to produce critical information about the scholarly bona fides of the plaintiff and others, and not to create copies that displace the plaintiff’s work in the marketplace.
Now let’s consider the markets that the defendant’s use might affect. First, it may diminish the plaintiff’s ability to sell papers to people who want to plagiarize. Second, the defendant may diminish the plaintiff’s ability to sell his paper to outfits like Turnitin for purposes of trying to catch plagiarists.
As for the first market, Turnitin’s use does not substitute for the plaintiff’s work. Yes, Turnitin’s use affects that market, but that is because Turnitin produces information that exposes the plaintiff and the plaintiff’s customer as dishonest academics. That’s criticism and comment, and it does not count as affecting the marketplace for the copyrighted work. Campbell stands for the proposition that if people don’t want the plaintiff’s product because of something they learned from the defendant’s alleged infringement, that is not an effect on the market for the copyrighted work. To quote the Court, this is the “distinction between potentially remediable displacement and unremediable disparagement.”
As for the sale of the plaintiff’s work for detection of plagiarism, the effect on the marketplace is small at best. If the plaintiff wants to exploit a market for sale of manuscripts to people who want to plagiarize, the plaintiff will not sell into any market for catching plagiarists. Additionally, the plaintiff cannot claim that he writes specifically for the detection of plagiarism because there’s no demand for a work written for that purpose. A work is valuable for detecting cheaters only if it’s already a published work that someone might find and use, or if it has been sold to plagiarists. There is no meaningful independent demand for works used to catch plagiarists. Thus, its elimination would have no effect on the production of copyrighted works. Accordingly, the effect on this market, even if considered an effect on the market for purposes of copyright, is too small to swing factor 4 to the plaintiff.
Accordingly, I conclude that the defendant’s case in factors 1 and 4 is strong enough to overcome the plaintiff’s strength in 2 and 3.