Making the rounds in April was a higher-ed post for our time: “The honest B or C student, submitting essays filled with awkward constructions, malapropisms, [and] earnest, if failed or surface-y arguments, [is] a surprise hero of the present age.”
Recommended Stories
- San Francisco pulls plug on new grading system after backlash
- MIT becomes latest elite university to shut down DEI office
- Trump administration announces steps to fight student aid fraud after millions paid to ineligible recipients
So wrote professor Matt Dinan of Canada’s St. Thomas University, a liberal arts institution in Fredericton, New Brunswick. But so could have written every English professor to whom I’ve spoken in the last 2 1/2 years, from bustling Los Angeles to the remote Nebraska plains. Across the nation — indeed, across the computer-age world — the introduction of ChatGPT and other so-called large language models has scrambled classroom values and turned previous standards of excellence into markers of guilt. When, in 2007, I arrived on campus as an assistant professor, the verb-tense mistake and erroneous homonym were symptoms of blithe unteachability. Now, professors of my acquaintance concede that they are positively thrilling. As one cheeky Dinan respondent posted, “I feel scene.”
How did this happen? The short answer is that ChatGPT’s ability to mimic standard English prose has presented students with an opportunity to outsource their work. Whereas college essays once required an hourslong abstention from video games and pornography, today’s 19-year-olds can “complete” such assignments in a matter of seconds. Never mind that most chatbot “writing” is unreadably flavorless and sad. It is, on the level of the sentence, grammatically correct. As late as 2021, an error-free paper signaled that a professor was dealing with a rival intelligence, a young man or woman with whose ideas he could happily interact. In 2025, that mental shorthand no longer holds.
Consequently, professors now find themselves in a curious position. A clean student draft might be the output of a trained and lucid mind, or it might be a mere computerized belch. While the instructor will have his suspicions, he won’t be certain. Neither will he keep his job if, attempting to restore moral order, he hurls accusations without proof. At some point in the near-to-medium future, the college essay will cease to exist as an assignment category, or ChatGPT will become so ubiquitous that professors assume and allow its use. For the moment, though, a strange reversal has taken hold. What precision once signified, surface flaws now do. The comma splice, at one time an enemy, has become a white flag of peace. Professors see it approaching and rejoice.
Readers who know their educational history will recognize how deeply ironic this is. For much of the 20th century, John Dewey’s progressives battled conservatives over literacy instruction, with the former insisting on a “whole word” method that deemphasized letters’ actual sounds (i.e., phonics). According to this way of thinking, even flatly incorrect guesses on the part of children could count as “reading” so long as they captured a sentence’s communicative essence. Because this was nonsense, it eventually lost out to a phonics resurgence that began with the 1955 publication of Rudolf Flesch’s Why Johnny Can’t Read. Yet the Left’s bad ideas never quite die. By the 1980s, the “whole word” movement had infiltrated foreign-language instruction in the form of “communicative language teaching,” a pedagogy that stresses “real-life” communication rather than grammatical rules. By the 2010s, it had raised a clenched fist and become “code-meshing,” “critical grammar,” and “linguistic antiracism,” postmodern ideologies linked by the suspicion that the classroom grammar nazi is, in actual fact, a white supremacist and bigot.
What all of these developments have in common is their hostility toward linguistic standardization. To a certain kind of leftist education theorist, the very notion that one could make oneself clear but still be charged with an error is outrageous. After all, the whole point of language is to convey meaning. To the extent that grammatical rules facilitate that work, they are useful. Carried even an inch further, however, such regulations are snobbish, exclusionary, and profoundly unwoke.
Conservatives take a different tack, of course. We recognize that English is a magnificent cultural inheritance, the language of William Shakespeare, Jane Austen, Emily Dickinson, and the King James Bible. We grasp, too, that having a civilization requires the occasional subsumption of individual preferences into group norms. Most importantly, because we are far more religious than our progressive friends, we understand that the law exists for our good. The rules of English grammar have inspired more true freedom and creativity than linguistic anarchy ever will.
One might assume, given the overwhelming leftist tilt of campuses, that professors have never cared much whether students can write a decent sentence. My 20 years’ experience suggests otherwise. No, not every business or psychology instructor is underlining dangling modifiers in red ink. But clean papers generally get better marks than sloppy ones, all other things being equal. This is the case for a number of reasons, not the least of which is summed up in Robert Conquest’s insight that “everyone is conservative about what he knows best.” One needn’t be a curmudgeon, however, to have noticed that error-free writing does something that incompetent prose can’t: It permits a full and untroubled entry into the intellectual world of its creator.
Furthermore, careful writing is so often paired with careful thinking that the two are difficult to separate, perhaps especially on the undergraduate level. Indeed, if college faculty were forced to name the single most damaging consequence of ChatGPT’s ubiquity, it might well be the introduction of an artificial surface “correctness” that bears no relationship to correct thought. Computers cannot think. Their compositions are a cheap parlor trick. Though the majority of professors surely nodded along when their “grammar-is-racist” colleagues held forth, in their hearts, they knew better. Now that the much-relied-upon connection between writing and reasoning has been severed, probably forever, they will dearly miss it.
This brings us to the present moment and the odd turnabout that has taken place on campus. As discussed above, theory-obsessed School of Education weirdos have been chanting student-literacy heresies for generations. That they now welcome our new robot overlords should come as no surprise. (The American School District Panel: “Teacher-training organizations … should design AI professional development that explicitly starts with trust-building.”) Dispositional conservatives in other departments, meanwhile, are far likelier to hold the line against ChatGPT and reward students who don’t use it, even to the extent of celebrating the occasional human-generated mistake. In sum, professors who never liked proper grammar will now be getting more of it, courtesy of the machines. Those who have always loved it will be cheering for student errors, if only because such slipups are a sign that a real person has been at work. It would appear once again that God has a sense of humor.
Needless to say, vanishingly few professors got into the higher-ed business to teach robots. Yet this bare humanistic standard is only now under assault. Clinging to it, one discovers, to one’s astonishment, that the ed-school Left was never entirely wrong about the relative merits of correctness and communication. There has been, all this time, some truth to the notion that errors have their own peculiar value.
The bookshelf has taught this lesson for centuries. Wuthering Heights, among the most beloved English novels, is — how to put this delicately? — idiosyncratically punctuated. Few would read William Faulkner for grammatical instruction, but his work is as ornately beautiful as anything produced by an American. Elizabeth Bishop’s poem The Man-Moth has delighted readers for 90 years and wouldn’t exist had not a certain Manhattan newspaper misprinted “mammoth.” Perhaps nine copies remain of the 1631 “Wicked” Bible, in which the typo, “Thou shalt commit adultery,” neatly illustrates the fallibility of man.
Though the principle can be taken too far, it is also true that student mistakes open a door to authentic intellectual engagement. Every writer knows that repairing a bad sentence often means improving a poor idea or hitting upon a better one. While the vast majority of students barely glance at red marks, in my experience, those who do may end up better off for their trouble. Rather than signaling scholarly achievement, flawless writing may have been disguising or smoothing over unpersuasive arguments all along. The rough stylist benefits from no such camouflage.
It is just possible, if one squints, to apply this thinking to politics. Following the second Trump administration, we are likely in for four or eight years of Gov. Gavin Newsom’s (D-CA) soulless technocratic cool. Will we miss Trump’s blundering, his carelessness, his almost magical ability to miswrite executive orders, thus inviting judicial review? From the conservative-partisan perspective, the answer is a hard “no.” Give us a dead-eyed policy bot of our very own. From a simply human perspective, I’m not so sure. A president with a pulse can be a reassuring thing, even if it sometimes misfires.
There is a final irony as well. To adapt T.S. Eliot, artificial intelligence is further proof that mankind “shall not cease from exploration.” How splendid if we really do “arrive where we started” and know ourselves for the first time: imperfect, clumsy, fully human, and all the better for it.
Graham Hillard is editor at the James G. Martin Center for Academic Renewal and a Washington Examiner magazine contributing writer.