Writing is More than Writing Things Down: A Response to Joe Cruz and The New York Times

Author
Daniel Emery
Jessa Wood

In our writing-across-the-curriculum work with faculty across campuses, we often hear faculty speculate about the future of writing in the wake of the Generative AI (GenAI) revolution. According to the AI in Higher Education Global Survey 2026, most higher education faculty worldwide believe GenAI will bring “significant” or “transformative” change to the instructor's role (p. 43); the AAC&U found that 95% worry about students’ overreliance on AI tools. Similarly, most students worry that their programs are ‘outdated’ (AI in Higher Education Global Survey, 2026, p. 25) and are skeptical that faculty are prepared to guide their use of GenAI (p. 28).

In response to concerns about GenAI, some faculty are deemphasizing or abandoning out-of-class writing projects and online writing environments, returning to in-class Bluebook writing or instituting oral exams. The University of Chicago, long a champion of tradition in higher education, is actively proposing ‘device-free’ courses and ‘analog’ classrooms for undergraduate courses.

A reconstructed papyrus, inscribed by Oxyrhynchus in the 2nd Century CE, of Plato’s Phaedrus.

Discourse about higher ed’s move away from long-form writing has flourished in academic reporting in The Chronicle of Higher Education and Inside Higher Ed and popular publications like The Atlantic. Just a few days ago, The New York Times1 published the latest such analysis: an opinion piece by Joe Cruz, Professor and Chair in the Department of Philosophy at Williams College2, provocatively titled “I’m a College Professor. Writing Isn’t as Important as We Think.” Cruz’s stated aim is to challenge what he sees as a common but misplaced assumption among college faculty: that “students learn to think — critically, creatively, reflectively, tenaciously — only by writing.”

In this blog post, we’ll highlight both the insights and limitations of Cruz’s argument—and, by extension, broader conversations interrogating the role of writing in higher education. As practitioners of writing across the curriculum, we’re obviously interested parties, but we maintain that reading and writing offer opportunities for cognitive engagement and access to practices of knowledge production that higher ed shouldn’t abandon.

Unpacking Cruz: An Oral Exam vs. Traditional Essay Showdown (?)

Cruz’s dismissal of writing hinges on his success in instituting a ‘writing-free’ oral examination, a 45-minute one-on-one assessment designed to audit students' understanding and press their analytical claims. Students not only ‘prove’ their knowledge of the material in a way writing couldn’t guarantee, but their insights and creativity were, he says, even better than in their original written work.

Cruz is absolutely right that oral discourse is a powerful way to engage with ideas. He speaks eloquently to the value of the “messy, searching conversations” in his courses, rich with opportunities to make novel connections and develop understanding together. Discussions absolutely provide crucial opportunities to explore and develop ideas—and for some students, reducing the extra cognitive load of encoding their language through handwriting or typing can be essential for freeing up working memory to engage more deeply with content (one reason that speech-to-text is an important assistive technology).3 Likewise, the opportunity to engage one-on-one with a master interlocutor and subject matter expert with whom they had established personal rapport was no doubt a deep and meaningful learning experience for many of Cruz’s students.

At the same time, the value of discussion to a rigorous education in philosophy (or any other discipline) is not evidence that writing is irrelevant. In fact, a cynical reader might note that Cruz hasn’t removed reading and writing from his course.

Cruz describes his exams as opportunities to “audit [students’] understanding of their claims and push them further in their analyses.” Before conversing with Cruz one-on-one, students closely read texts and refined their thinking through messy, recursive writing, ultimately producing drafts that they built on in their conversations. (We recognize these conversations as a form of writing conferences.) We’d argue Cruz’s students’ success in their oral exams is more likely a testament to the power of combining carefully-structured reading, writing, and oral discourse opportunities than evidence of the merits of standalone oral exams.

In his conclusion, Cruz argues that outsourcing writing to AI may be less harmful than we imagine, as long as instructors assess students in other ways. Yet, in an earlier throwaway line, Cruz admits to allowing AI use only “in accordance with specific guidelines.” Clicking through to his course materials reveals four full pages that carefully structure students’ engagement with the tools to support learning.4 It seems that adopting oral exams hasn’t freed Cruz from the need to thoughtfully guide students’ engagement with GenAI to protect their learning. Indeed, his strict restrictions reveal that he sees several forms of GenAI use as detrimental to learning.

Writing Is Not Merely Archiving

Cruz's penultimate paragraph reveals a common but deeply flawed assumption underlying his claims. Cruz suggests that writing exists merely to provide a durable record of thinking, implying that its presence in classes is at best an arbitrary convenience and, at worst, an inappropriate distraction from the real work of critical thinking. But research in writing studies—including research Cruz himself cites—emphasizes that thinking, writing, and learning happen simultaneously. Sometimes we have brilliant ideas and jot them down later, but the act of writing often shapes, refines, and transforms our thinking. Students learn a great deal when they scratch out a messy draft, get to their main idea on page three, and backtrack to revise. Furthermore, sharing our writing in progress with readers can create rich opportunities to refine our ideas, as Cruz and his students experienced aloud.

Writing is baked into academic knowledge-making processes, and for good reason. The process of idea generation, research, refinement, revision, reflection, and presentation is suffused with acts of writing, from calendar invitations and emails; to notes, drafts, and sketches; to probing questions from readers and advice from mentors; all the way down to decisions about the representation of significant figures and formatting options for captions. In many instances, writing lets us build arguments that are more complex and more rigorously evidenced than what we could accomplish through oral argumentation alone. Choices made in writing emerge from and reveal disciplinary values and ways of knowing.

In short, writing is not an optional add-on. Cruz’s students almost certainly bring more than just their great ideas and enthusiasm for learning through the classroom door: They may bring marked-up copies of assigned readings, information from the course syllabus or Canvas site, and volumes of prior knowledge and experience organized by years of reading and writing as well as speaking and listening. This experience can’t be ignored in our assessment of the success of his oral exams.  

Where Does This Leave Faculty?

Changes in writing technology can be challenging and even painful. We talk every day with faculty experiencing very legitimate frustration with the ways GenAI has changed their courses. Written artifacts that we once hoped would offer robust proof of students’ cognitive engagement no longer seem as secure an index of learning or engagement in a world of statistically programmed word predictors and text extruders.

When assessing a learning outcome requires being as certain as possible that students have committed certain details to memory, (more) secure assessments like bluebooks and oral exams are certainly worth considering. Nevertheless, as Cruz demonstrates, writing remains an excellent venue to make thinking legible, and faculty can encourage effective thinking and discourage overreliance on AI through their course policies, assignment design, and feedback practices.

What’s more, not everything we do in a college classroom is, or should be, aimed at summative assessment of students’ capacities. The grades aren’t ultimately as important as the opportunities we provide for learning and growth. So, don’t forget what is lost if we entirely surrender long-form, independent engagement with written text entirely. Most people in higher education have experienced the struggle and reward of closely reading and analyzing a text, or using writing to clarify our thinking and share our expertise. Depriving students of the chance to write something valuable, meaningful, and beautiful is a cynical and counterproductive way to resist GenAI. Not every student in your courses will choose to engage deeply in reading and writing—but they all deserve that powerful opportunity.

1Readers can access this and other NYTimes material through the University Libraries thanks to our intrepid librarian colleagues: As of September 21, the NYTimes is available for all UMN students and faculty to read for free through the University Libraries’ website.

2Professor Cruz is an excellent academic as well as a provocateur. He’s written authoritatively on topics in epistemology and cognition, including the encyclopedia entry for “Epistemology” in the Wiley Encyclopedia of Cognitive Science (2006). His forthcoming book, Making the Mind Legible, probes the history of explanation in cognitive science and will likely say a great deal about legibility as a dimension of knowledge and writing.

3For others, writing is more accessible than speaking because it provides the opportunity to think in less linear ways and engage in revision. 

4The restrictions he applies and the writing and thinking practices he encourages are all very good. Cruz consistently demands that students do their own thinking first and explicitly forbids the use of GenAI for topic generation or to supply the precise words a student might employ in an essay. He even goes so far as to exclude the citation of AI-generated text as reasonable evidence for an argument, demanding that students cite human authors.

Image Source: Papyrus Scroll of Plato's Phaedrus, enscrlbed by Oxyrhynchus, 4th C. C.E. Public domain via Wikimedia Commons.