Yesterday I was feeling sorry for myself with regard to Turnitin and the like. I ended up having an interesting discussion with @LanceStrate, @mattthomas, and @KelliMarshall about the ethics surrounding plagiarism detection service. It got me to thinking about why it bothers me.

My gut feeling is this: Turnitin, SafeAssign et al make big bucks off of their database. More papers scanned means a bigger database; bigger database means (in theory) better plagiarism detection; better detection means (in theory) more value and more profit. Forcing students to relinquish their papers to this machine feels exploitative.

John Stuart Mill – Awesome Guy | cc licensed flickr photo shared by netNicholls

But I wonder why this bothers me. I have no problem feeding different kinds of information-gathering machines. Take Google. I use Gmail, Google Reader, Google Calendar, and google.com extensively. The more I use these services, the more information they gather about my online activities; bigger database means better ad targeting; better targeting means more value and more profit. My “stuff” – information about me, writing I produce, records of my activity, etc. – is not sacrosanct. I’m willing to give it up in some cases.

So what’s the difference? Most obviously, I am choosing to use Google’s products in a way that students are not asking to use Turnitin. I will grant that there are different levels of “forcedness”, as @LanceStrate points out. Students can opt out of a class, or out of school in general. And if instructors make the Turnitin requirement explicit in the syllabus on the first day of class (or earlier), students will be reasonably well-informed about what they will be “forced” to do. But no matter how you conceive of the spectrum of requirement, the fact remains that my use of Google is far freer than students’ use of Turnitin.

That a professor requires students to do certain things that they wouldn’t otherwise do is not, in itself, an indictment of the requirement. I doubt that my own students would write about the Nicomachean Ethics if their grade didn’t depend on it. But, in this case, I as an instructor am obligated to exercise my power in a responsible way. (Heavy is the head that wears the crown.) Requirements should not be arbitrary, but should serve the goals of the class and the best interest of the students. Requiring a paper on Aristotle has negative effects on students – it takes away from the time and energy they could be spending on other things that are valuable to them – and it’s my responsibility to ensure that these negative effects are outweighed by the benefits bestowed by such an assignment. A well thought-out term paper assignment will, in the long run, have positive utility for the student.

Is the same true for plagiarism detection? Are the negative effects of such technologies (being forced to enrich a corporate entity, losing control over one’s intellectual property, feeling a presumption of one’s own guilt in the absence of supporting evidence) outweighed by some benefits? It’s at this point in the thought process that the pedagogical implications of Turnitin should be considered.

• Is Turnitin good at detecting plagiarism? My experience says: Not really. While Google’s database doesn’t include as many student papers as Turnitin’s, Turnitin is in turn pretty awful at identifying plagiarism from the open web. Thoughtful reading and Googling has been more effective for me. I’d like to see data on the larger trends, though – for example, what percentage of student copying comes from the open web (Google’s domain) versus for-sale paper databases.
• How much harm does “plagiarism” really do? This is really the more important question. Even if it turns out that Turnitin is very, very good at plagiarism detection, there is very little benefit from the software’s use if it turns out that plagiarism, as defined, isn’t really that harmful. This question is tough to answer, though. For one thing, there are lots of different kinds of plagiarism, certain kinds of which are more harmful than others. A student who copies a paper wholesale from Wikipedia is doing more harm than one who synthesizes a coherent paper from a bunch of different sources, or one who fails to cite a paraphrased argument. Surely the second and third students are getting more out of the assignment than the first. Furthermore, I have an untested gut feeling that the most harmful types of plagiarism – where a student steals wholesale – are easier to detect without using Turnitin, since they’re more likely not to be even approximately in the student’s voice or level of expertise. If this is right, then it might be the case that Turnitin is most necessary for the least harmful varieties of “plagiarism” – varieties whose ethical implications, some might argue, ought to be reassessed in light of how new technologies are affecting knowledge creation. (Too big a topic to address here, but you get the idea.)
• Are there less troubling alternatives to Turnitin? Let’s grant that Turnitin is very good at detecting plagiarism, and that plagiarism is hugely pernicious. All things being equal, if we could avoid plagiarism by means that have less of a downside, we should choose those other means. In my experience (again, I have no comprehensive data to back this up), the answer is yes, there are far better ways. @KelliMarshall suggests assigning unique paper prompts, making plagiarism more difficult. I’ve found that the scaffolding of assignments – such that students write early, write often, and write in a low-stakes milieu – is extremely effective at lowering the tempation to plagiarize. To be more specific: When students are writing in journals or blogs – spaces where they are not harshly graded – and when their formal assignments allow students to pull from and build upon the ideas that they’ve already put to paper(/bits), cheating simply doesn’t happen very often. That initial moment – when a student sits down at the computer the night before the due date, not having written a single word, not knowing where to start, and copying out of desparation – is averted altogether. In the semesters I’ve used blogs and structured assignments in this way, I’ve had to deal with plagiarism maybe once per semester (out of 70+ students writing hundreds of papers). Another thing that’s worked really well for me is having frank discussions with students about why plagiarism is so demonized in academia in the first place (perhaps this conversation is a little more justified in an Ethics course). When they understand the motivations, and are not simply handed seemingly (and perhaps actually?) arbitrary rules about the Evils Of Plagiarism, they’re more likely to grok.

On balance, then, it seems to me that there is very little, if anything, to be gained from Turnitin et al that cannot be gained through other, less harmful means. Now I have to work up the guts to start sending links to this post whenever a faculty member asks me how to do plagiarism detection! But I suppose my lack of intestinal fortitude is a topic for another blog post.