It seems unlikely that if you teach college writing, that you haven't been dealing with the impact of ChatGPT and other LLMs. I of course have had students using this next technology since Spring 2023. But the impact it's had on me and and my teaching is probably most starkly represented by this chart:
According to my syllabus policies, as well as our program policies, use of ChatGPT or similar technology is prohibited. Though I guess I can see how it might be useful in other courses, I don't see how it's useful in a writing course; my class (I would argue) is trying to use writing as a mode of thinking. We don't write to record what we already know, we write to come into the knowing of something.
The thing about ChatGPT use is that it's difficult to "prove" in some kind of "objective" sense—and this is the kind of thing academic integrity panels supposedly want. But the whole reason I am a professor of academic writing is that I supposedly have some kind of expertise in academic writing, and I do. It's an expertise honed by reading student writing for years, almost decades. I first taught an academic writing course in Fall 2008. Without methodically counting it all up, I would estimate that in that time I've taught sixty-five sections of first-year writing course, which means I've read the work of approximately thirteen hundred students, each of whom ought to write at a minimum of ten pages per semester, if we're going to be conservative. That means I've read at least (and certainly much more than) thirteen thousand pages of college-student papers in my life.
![]() |
image generated, of course, by ChatGPT |
But anyway, clearly lots of them are using it. (Though, admittedly, some are clearly not! I have a lot of fondness for the crappy C paper now.)
The thing you can actually prove, though, is the veracity of sources. That is the basis on which all of my charges were filed this semester. Of my fourteen cases, I think three were about sources that did not exist. This, for me, means failure of the course. If you don't see that a basic part of research writing is that the sources you cite have to actually exist, then I don't know that you really belong in my class at all. I can't teach you this. Sure, use ChatGPT to find sources (in my personal experiments I have not found it to be very good at this, but I am sure it could be), but then make sure they are real! And of course, if you're summarizing fabricated sources on, say, an annotated bibliography, you are lying, because how could you have read a source that doesn't exist?
All of my other charges were about fabricated quotations from real sources, students claiming that direct quotations existed that did not actually exist. My institution requires that all academic integrity filings be accompanied by a formal meeting that is witness by a neutral faculty member; to make up for all the colleagues that had to witness mine, I witness a lot of other people's. This was the thing most of them saw as well. Frustratingly, a lot of students had this weird defense: that they didn't know quotation marks were reserved for direct quotations. One might tempted to believe this, except that very few students were making this mistake over a year ago, and though I think that the teaching of writing at the pre-college level has probably got worse since I started teaching in 2008, I don't believe that just a couple years ago, teachers stopped explaining what quotations marks are for.
So, the only explanation is that they are using ChatGPT to "find quotations"... but again, not confirming the existence of the quotations. Some of my students have admitted this when confronted, others have come up with not very compelling explanations. I have seen this across the board, in both my research writing class and in my text-based humanities course. (I can see why students might think I might not know that they made up a quotation from a source they found that I've never read; I don't see why they don't realize I won't catch them when they make up quotes from stories we all read together!)
Thankfully (I guess) it doesn't matter. At my institution, fabricated quotations or sources fall under the category of Deception and misrepresentation; I don't have to prove the fabrications come from anywhere in particular, I just have to prove they don't exist.
In my research writing course, fabrications in the final paper mean failing the final paper. It seems to me there's a basic parameter of the assignment you've failed to reach. And failing the final paper means failing the course. It seems to me there's something this course was supposed to teach you that you just didn't learn, and thus you need to take it again, if you can't write at least a D-level research paper by the end of an entire semester. In the past, I haven't had to enforce this policy very much, but I did at least three or four times this semester.
On top of all this, one has to remember—these are just the students I caught fabricating. There are also all the students using LLMs that I recognized but couldn't "prove," and thus just dinged them on points (I have become fond of marking assignments with a "0" and writing, "This is so vague an AI could have written it"), and then all the students who are actually "good" at using LLMs to write and turning in work that seems human-written... even when it's not.
Even though ChatGPT has been around a couple years, it's clear that something is shifting in the way students use it. You can see that even just last semester, I only filed one charge (though I didn't teach research writing that time.) Many of my colleagues had similar experiences this semester in particular (though I don't think anyone filed as many charges as I did). As one of my colleagues has said, it seems like there's a shared ethos we used to assume the existence of that just doesn't exist anymore. And how can you teach people if that's the case? I can't teach someone that they want to be ethical.
So anyway, it's been a depressing semester. I even had two students file appeals that persisted beyond the end of the semester (and another just not respond to my communication attempts, which is for some reason an automatic appeal), which meant that it didn't stop even when the semester was over! But finally about two weeks ago, I heard about my last outstanding appeal.I can't just keep doing everything the same way, clearly. But more on that in another post.
No comments:
Post a Comment