SDI Productions/E+/Getty Images
It is surprising how quickly we faculty have gone from barely being aware of ChatGPT to the vague perception that it exists somewhere (but our students?—never!) to the shock of detecting those first offloaded writing assignments to the present doomscrolling of a near constant stream of articles and books lamenting the existential crisis at hand. I won’t rehearse why the AI crisis is existential, or what students/humans lose when they stop producing their own written work. Nor will I offer the seemingly obligatory paragraph about how LLMs will never be able to capture humanity’s finest observations, metaphors and arguments. My guess is that, in a few years, the thirsty machine will write better than all of us. The question I ask myself instead is one I often find myself asking in the face of some contemporary crisis. Because I read and teach ancient texts for a living, my question usually is: What would antiquity have made of our current handwringing?
I like questions like these, because they often reveal that I’m wearing modern blinders. Born and raised in a particular historical bubble, I view things like writing, college and college writing in a certain way, and there’s not much I can do about it. An academic essay looks like this, and is supposed to look like this—the feeling goes—because that’s just what an academic essay is. But when I turn to ancient cultures so different from this one, I usually find that things were not always this way. Far from it: Ancient teachers of higher education would find our present way of teaching—the very one that has given rise to our present writing crisis—altogether bizarre.
“Why do you write ‘academic essays’ in the first place?” they would ask. “Why are these ‘final papers’ the ultimate assessment of a student’s performance in your class?” Writing is one thing: Ancient students wrote a great deal, scribbling away on wax tablets, organizing their thoughts, rewriting, editing and all the rest. But to have such writing be the final goal? That would be bizarre for ancient teachers. For them, writing was just a tool used to prepare for another goal. For them, the final goal was to speak.
Oddly enough, this wasn’t just the goal for higher education in antiquity. It remained the goal through the rise of Christianity and Islam, right on through to the creation of those guilds called “universities” in the medieval period, straight past the Renaissance and Enlightenment all the way to the higher education of the 19th century. In medieval and Renaissance universities, students worked on giving public “disputations”—formal arguments that students had to present with thesis, counterarguments and refutations. In the first colleges of the U.S., students would similarly work on their oratory—writing and revising drafts only to prepare for the all-important oral performances. Public speaking, not some written assignment, was the expected final product.
Last spring, I taught a class based on the ancient rhetorical education, and had students shift their attention to speaking rather than writing. We still wrote, of course—just as ancient students wrote—but the writing was ungraded as its sole purpose was preparation for speaking, which was graded. Like ancient students, these students dabbled in a variety of genres—constructing proofs, telling stories, creating characters, opposing arguments. The ancient rhetorical education reminds us that there are more genres to practice than the research essay.
It made me wonder: How many of these students will actually go on to careers of writing professional research papers anyway? Would the ancient model in fact better serve most students’ future lives?
Beyond the research paper, there is such a variety of expression used in public life, from the bar to the boardroom. And then there is personal life too: not just toasts and speeches (which, thank goodness, are rare enough), but simply expressing to friends how we feel, putting into words why something is right or wrong. Will the words be there when we need them? Will the research essay alone have helped us find them?
We had a great time in that class, and the students were fantastic. But here is what I learned. When we shift the emphasis to speaking instead of writing, suddenly the value of writing itself becomes present and urgent instead of vague and, if we can judge by AI’s increasing popularity, unconvincing. Often students cannot help but ask themselves in the face of some time-consuming writing assignment: Am I writing this because I have an assignment due? Because I’ll read this essay 15 years from now? But in this class, for these students, the answer instead became immediate and concrete: Do I want to have something to talk about tomorrow or not? Do I want to stand before my friends and draw a blank, or know exactly what it is I need to say?
In such scenarios, the only valuable writing is the self-generated kind, because it is infinitely harder to try to speak by remembering whatever vague emptiness ChatGPT wrote. Try lecturing off someone else’s notes. It’s impossible. Instead, if we want the easiest route forward, the ideas have to come from us. This was one of the key things we learned in that class. Ancient students and ancient orators generally were not writing out speeches and then memorizing the words. They, like us, realized that this would require an inordinate amount of effort, and was simply not realistic. Instead, what they used were so-called “memory palaces” where speakers are not memorizing words, but conjuring vivid images of the things they want to say. Since the brain’s hardware is more visual than verbal, the memory palace is a way of hacking the brain and those visual powers. Contestants in memory competitions (see Joshua Foer’s Moonwalking With Einstein for that fascinating world) still use these memory-palace techniques all the time, but it has long stopped being a part of the general curriculum.
It's easy, though. If, in a lecture, I want to remember some counterargument —say, that Odysseus rejected immortality when Calypso offered it to him— I go in my mind to the set place in my memory palace—say, my kitchen sink—and I “see” Odysseus crying on Calypso’s beach. When I “see” this, I immediately know what it is I want to say. By contrast, when we memorize words, we tend to sound artificial because we don’t usually talk that way. When I speak to you, I want to tell you something, not some words. That’s why, if I tell my audience in real time about this thing I’m “seeing” (Odysseus crying on the beach), now the speech sounds natural, because this is simply how we talk in general. It’s as if I’m having the idea on the spot. But I’m not, really. I’ve already written out my ideas ahead of time, organized them, and sharpened my thinking regarding what I want to say.
After spending most of the semester focusing on speaking rather than writing, we ended with a little writing untethered from speaking—for example, keeping a journal together. There is so much, we found, that speaking cannot do. We’ll never get the oral wording just right as in the written form. We’ll never capture that perfect rhythm that we found after so many versions of that key sentence. We’ll never really get to express those things that are just too subtle or awkward to express in face-to-face contexts. The students conveyed those sentiments beautifully, often better than I could. But what more could we want than to have students feel this way?
For speaking, though, perfection is not the real goal. Students’ chief fear in speaking is the same as our own: We fear that we will fall far short of our ideal projections, and worse, do so in front of everyone. But here’s the thing: No one in the audience really wants perfection, or some ideal, polished orator. What we want is human connection. People crave the voice of their peers, not in spite of, but in many ways, because of its flaws. Witness the fate of the teleprompter-bound candidates of the last election. Polish and perfection is, at best, highly overrated.
We were all trained to write, not to speak, and, if our parents and grandparents went to high school and college, they were part of that same 19th-century academic tradition. But if we start small, and consider the long history of higher education when we choose our next assignment, we can ask ourselves: Why not have them speak, and use writing to prepare for that speaking? We teachers, as untrained speakers ourselves, could model the unconcern for perfection. We could tell our students that being tied to perfection is like being tied to a teleprompter. What people want is you, your voice.
At some point, relatively soon, I do think AI will write better than we can. But I also think that this will be no more a problem for writing than it was for chess. Remember Garry Kasparov versus Deep Blue? It turned out that chess wasn’t finished after all, because humans were not particularly interested in computers playing chess even though the bots were significantly better than their human counterparts. We’re “speciesist,” as Peter Singer put it, for better or worse. And so, those who long for human connection will continue to look for it among humans.
This suggests an interesting fate for the humanities. Not unlike the fate of physical exercise after the Industrial Revolution, people interested in the humanities will be doing it, like those fitness buffs, not just for the pleasure of it (endorphin kick!), but to build a better self. It’s a lot like the ancient memory palace: What kind of palace do I want to inhabit? One rich with memories, I imagine, not an empty one. One full of funny and moving scenes from different novels and films to share with our friends. One with different languages and their new ways of seeing, and great lines from our favorite writers. More, not less.
Recent defenses of writing are absolutely correct when they remind us that writing helps us to think, and that the process of writing is what’s most important. But I imagine that such defenses are a bit vague from the student’s perspective. Spoken assignments, meanwhile, make these intuitions concrete because they are directly related to tomorrow’s deadline. When we need to speak, the writing really is all about the process—the developing of an idea, clarifying what we want to say, finding the most effective organization of the thoughts we want to express tomorrow. In the face of a spoken assignment, the concern is not so much for potentially reduced mental capacities in an abstract, distant future, but the much more immediate fear of losing—right now—your voice.
