José Vergara: For all his iconoclastic views, Vladimir Nabokov, at heart, believed in human connection and collaboration. As Sara Karpukhin points out using a moment from the novel Pnin, individual style—something developed over time and in dialogue with others—mattered a great deal to him. It wouldn’t be enough to simply reproduce with machine-like efficiency the voice of a lecturer for students. Instead, it’s the nuances, quirks, and even errors that define voices, including those of instructors, and that Nabokov appreciated in how we learn from one another.
What would Nabokov make of an AI-narrated volume devoted to teaching his works?
As we come to terms with how the world of education, from top to bottom, is using AI (or not), the anxieties are ever-present. When it comes to writing, at least, AI presents the opposite of the anxiety of influence; it’s the anxiety of no-influence—an assemblage of more or less accurate statements that lacks traces of the difficult but necessary process of working through others’ ideas to generate one’s own. And yet, it’s important to consider what AI opens up in terms of accessibility and reach.
What then would Nabokov make of an AI-narrated version of a volume devoted to new approaches to teaching his works? In the following conversation, my co-editor Sara Karpukhin, Amherst College Press (ACP) Acquisitions Editor Hannah Brooks-Motl, and I discuss the audio version of our book Reimagining Nabokov: Pedagogies for the 21st Century, the logic and process behind its production, and the questions, worries, and possibilities it engenders.
JV: Hannah, could you tell us a bit about how our book became an audiobook? How common is this in the world of university presses?
Hannah Brooks-Motl: We’ve been in the middle of a so-called audiobook boom for quite some time (since 2016, according to the Wall Street Journal), with publishers of all kinds getting in on the act—even university presses. University press (UP) audiobooks are typically read by professional narrators, may be produced in conjunction with media studios, and are available for purchase or through various subscription platforms. Audio increases a book’s possible reach and potential sales, but producing audiobooks is time consuming and expensive so it’s no wonder presses carefully select the titles they’re willing to invest resources in: even major UPs like Princeton, Yale, and Chicago have just a handful of titles available as audio.
JV: And what are the motivations for producing audiobooks, in general and with the help of AI?
HBM: As a fully open access (OA) press, Amherst College Press isn’t looking to sell audiobooks, so why would we create them in the first place? At the Association of University Presses conference in 2021, disability activist Leah Lakshmi Piepzna enjoined UPs to think about audio as an accessibility issue: audio is one way readers with print disabilities can access books, often in conjunction with a print or digital copy. According to Jon McGlone, Fulcrum’s accessibility specialist, audiobooks can also “help readers with learning disabilities process information more effectively and help readers with physical or mobility disabilities who have trouble holding a physical book or using a computer/device to read a digital book.” We have committed to ensuring our entire catalog is available in audiobook format, including Reimagining Nabokov.
ACP doesn’t charge fees to authors, so we seek solutions that allow us to sustain our commitment to diamond OA. Without endless resources to pour into production, and in conjunction with our publishing partner Fulcrum, we identified the AI-narration offered by Google Play as one way to make our books accessible. Because the vast majority of our titles exist as EPUBs, the conversion process is relatively straightforward and automated: simply put, AI narration allows us to honor our commitments to accessibility broadly. And yet we are conscious of the potential harm that AI-narrated books could do to the human workers whose livelihoods are based in narrating audiobooks, even as engaging their services for scholarly books with limited audiences may be cost-prohibitive for most non-profit scholarly presses.
JV: It strikes me that this line of thinking is in keeping with the subtitle of our book: Pedagogies for the 21st Century. We’re attempting to make use of what resources (technologies, ideas, interpretations, ways of engagement) we have in the present moment to imagine an improved future for our students and colleagues in their engagement with Nabokov. Naturally, predicting the future is tricky business.
Any and all uses of AI will require careful cost-benefit analysis by publishers, authors, and audiences.
HBM: This is just one of many possible tradeoffs, of course—some of which we may not even be aware of yet. No doubt any and all uses of AI will require careful cost-benefit analysis by publishers, authors, and audiences whether they’re reading, listening, or engaging books in some other way. While we can’t necessarily foresee all the harm AI may do, using it to create audiobooks of our titles seemed one small way we might harness it for good.
I’m curious, José and Sara, about your reactions to this encounter between AI and your book? It does sort of beg a thought experiment about what Nabokov would make of AI.
JV: I was surprised when I learned about this AI-read audiobook version of Reimagining Nabokov. I wasn’t aware that it would receive one, and I hadn’t even considered the possibility. As someone who is AI agnostic at best, AI resistant at worst, I have to admit I had mixed feelings about it at first. As someone who is resistant to AI usage in my classroom, this audiobook felt a little hypocritical, perhaps. Learning more about the thought process that informs ACP’s decision to make all their books into audiobooks makes me appreciate it. I’m all for ensuring our work is accessible to as many readers as possible. In fact, that was one of the driving motivations behind Reimagining Nabokov. We wanted to reengage Nabokov’s texts with new generations in mind and share new ways of doing so. I also love the idea of our colleagues listening to the book while taking a walk or doing the dishes if it means that they can fit it into their schedules more easily. What will this wider reach mean not just for the book but for the teaching of Nabokov’s writing? This aspect of the book’s publication seems like a logical next step, particularly when you consider what Yuri Leving and I write about in Part 1 about digital innovations in teaching Nabokov.
Nabokov himself is oddly silent about robots in his fiction. I would think, though, that given his belief in the creative spark and the possibility of genius, AI would make him nervous. I’m reminded of King, Queen, Knave, in which an inventor’s mechanical mannequins might embody lives without real thought, purposeless automatons that reflected the rising political tensions in Germany at the time of the novel’s composition. I don’t think it would be a stretch to suggest that Nabokov would lament the loss of human imagination that comes with AI. Still, he might be charmed or bemused by the oddities that AI so often produce in text and image.
Sara Karpukhin: Thank you, Hannah, for explaining the thinking behind the decision to make our volume into an audiobook in the first place! I also agree with José in seeing the point of doing so. And knowing that Leah Lakshmi Piepzna was part of the inspiration makes me want to agree even more.
Technology makes accessibility issues visible. Disability scholars talk about this connection – and about commercial, intellectual, and pedagogical concerns it can generate. Speaking for myself, in my large-enrollment Nabokov class, it has become standard practice for students to ask for slides and lecture transcripts in the last two years or so, and even make them available before class sometimes. I try to provide them whenever I can.
There’s an anxiety, of course. I am worried about plagiarism and machine-written essays. I’m worried about my intellectual property. I also believe that the spontaneity of live conversation counts for a lot in education. In addition to what José remembered in King, Queen, Knave, I keep returning every year, with steadily growing panic, to the passage in Pnin where Nabokov satirizes precisely this uneasy union of efficiency and idealism in America’s educational system:
‘Phonograph records on every possible subject will be at the isolated student’s disposal…’ – ‘But the personality of the lecturer,’ said Margaret Thayer. ‘Surely’ that counts for something. – ‘It does not!’ shouted Hagen. ‘That is the tragedy! Who, for example, wants him’ – he pointed to radiant Pnin – ’who wants his personality? They will reject Timofey’s wonderful personality without a quaver. The world wants a machine, not a Timofey!’
This was written in the 1950s!
That anxiety aside, I also believe that whether I want it or not, AI is here to stay. Students and instructors already use AI, from ChatGPT to instant translation services to plain-vanilla search engines where AI has been heavily used for years now. In language instruction, we have corpus linguistics. I myself recommend to my students speech-to-text transcription engines that exist on the market now and that I suspect they may need for work even after they leave school. There is a growing profession of prompt writers now who help human experts get the most out of language models as ChatGPT, and sometimes I wonder if this is a skill that an educator may learn, too, so as to teach students how to benefit from this technology.
Honestly, I don’t think there’s an easy answer to the questions we are trying to answer. Ultimately, I think, for me, accessibility and pedagogy are both a “moral good” that the technology can serve. How to prevent the abuse and/or commercialization of the technology at the expense of individuals is the pressing question for the humanities in and outside of the classroom, I guess: How do I maintain trust with my students? How do I respond to their experience? What do I understand of the world they will inherit? What values are at risk of disappearance in that world? Work in progress.
There’s a lot of uncertainty about how AI will shape the work of scholars, designers, readers, reviewers.
HBM: I share all the above ambivalences! AI was a huge topic at this year’s Association of University Presses conference—there’s a lot of uncertainty about how it will shape our work, the work of scholars, designers, readers, reviewers… It’s unclear now at least how much space there is for the publishing industry (those of us who aren’t megaliths like Wiley, Taylor & Francis, Elsevier etc.) to be proactive rather than reactive.
JV: What do you two think: how did “Mike” do? In my opinion, he’s hit or miss when it comes to pronouncing Russian words and names in the book, but on the whole, he sounds pretty good!
HBM: The politics of AI voices are really fascinating. Google offers a whole suite of them, and it’s interesting to see which identity categories they’ve made explicit (gender, age, nationality) and which they haven’t. “Mike” is an American male between the ages of 31-45. I believe we could choose a different narrator, which might present a different listening experience—more or less Nabokovian I leave to the experts here. Liz Faber has written a great book on The Computer’s Voice that I recommend to anyone wanting to think about the psychoanalytic and feminist implications of this topic.
JV: Another question for you, Hannah, one that I do not intend as confrontational, I promise! How do these kinds of developments in technologies factor into contracts? Was this part of ours? I should know, shouldn’t I? You mentioned Wiley and Taylor & Francis earlier, which brought to my mind the recent news that they are selling data and content from their journals to companies to train AI models. In other words, going back to the question of the future, how is ACP factoring in authors’ preferences when it comes to AI and technologies that haven’t even been invented yet? For the record, and for the reasons we’ve discussed, I likely would support AI-narrated audiobooks in cases like ours.
HBM: This is a great question, José, and not at all confrontational; in fact, it’s the source of a lot of anxiety for UPs. One way AI is impacting contracting has more to do with presses’ concern about what authors might be doing with this tech. Presses are beginning to write disclosure clauses into contracts, for example, requiring authors to disclose any use of AI in the writing or research of their manuscript.
As for other impacts, it’s just a truth of publishing that most contracts take all an author’s rights, meaning a press can do whatever they want to with a book—including turning it into an audiobook (usually to sell; presses license these kinds of “sub” or subsidiary rights in order to increase a book’s revenue streams)—unless an author negotiates otherwise, which is why, yes, everyone should read their contracts!
ACP is a bit different in this respect because while an author does give us the right to publish in any format, we utilize Creative Commons licenses that are designed to lift the onerous permissions barriers keeping scholarship from widely circulating. Reimagining Nabokov was published under a CC-BY-NC license (chosen by you and Sara, I swear), which means that people can reproduce the text so long as they attribute the source and do not do so for commercial gain. In theory, CC licenses should protect works from being scraped by LLMs since to reuse the contents of a book those models would need to cite and not commercially profit from their use of the text. But that will likely take some lawsuits to figure out. Questions around copyright, publishing, and AI are going to be huge and gnarly ones for the foreseeable future.
SK: I have another reason to be grateful that José and I got to work with ACP then, as opposed to the megaliths! AI’s main problem seems to be the corporate control of it, with little to no public oversight, affecting pre-existing economic models of creative work and potentially leading to a diminishment of humanity in the ever-mediated world. The good news, I guess, is that market research shows skepticism. And the art world has come up with the Support Human Artists hashtag reminding the public of the costs. As a human academic, my work seems cut out for me.
Hannah Brooks-Motl is an editor at Amherst College Press where she acquires in a range of list areas, including Russian, East European, and Eurasian Studies. She earned an MFA from the University of Massachusetts-Amherst and PhD from the University of Chicago. Her fourth collection of poetry is forthcoming from the Song Cave in 2025.
Sara Karpukhin is Lecturer in Russian at the University of Wisconsin-Madison where she teaches undergraduate courses in the history of Russian culture, Vladimir Nabokov, and queer experiences of contemporary Russian speakers. She is the co-editor of Reimagining Nabokov: Pedagogies for the 21st Century.
José Vergara is Associate Professor of Russian on the Myra T. Cooley Lectureship in Russian Studies at Bryn Mawr College. He is the author of All Future Plunges to the Past: James Joyce in Russian Literature, co-editor of Reimagining Nabokov: Pedagogies for the 21st Century, and project director of Encyclopedia of the Dog: An Annotated Edition of Sasha Sokolov’s Between Dog and Wolf. His writing and interviews have appeared in Literary Hub, Asymptote, Words Without Borders, Music & Literature, and World Literature Today.
