By Malcolm R. Sedgewick
When I first encountered Nada Gordon’s gallery of imaginary literary critics, I was less offended than perplexed. Gordon is a serious poet, or at least a poet who has spent several decades engaged in a sustained and sometimes interesting experiment with the possibilities of language. She has written books, participated in literary communities, and developed an aesthetic sufficiently distinctive to attract both admirers and detractors. Why, then, would she want to manufacture a collection of people who have never existed in order to have them write about her work? What satisfaction could possibly be derived from commissioning fictitious judgments of one’s own poetry, particularly when those judgments are produced by a machine that has no investment in their accuracy, significance, or consequences?
The enterprise currently comprises four critics. Mira K. Vell is an enthusiast for the grotesque and ornamental, whose fictional scholarly career has been devoted to the aesthetic vindication of excess. Adrian V. Thorne is a severe intellectual who finds Gordon’s poetry a particularly ingenious form of bad faith. River Solstice Vale, a San Francisco-based genderfluid ecopoet, objects to the ethics of Gordon’s appropriations and her supposed indifference to ecological and gender justice. Eleanor Whitcomb Hale, a former Iowa Poet Laureate with an implausibly extensive collection of literary honors, believes that Gordon’s work substitutes verbal ingenuity for the emotional honesty that poetry ought to provide.
Each critic has an elaborate professional biography, complete with imaginary publications and distinctions, and each has been furnished with a computer-generated photographic portrait. The essays have been published on Gordon’s blog, Otohime, with editorial notes acknowledging their fictional origins.
The acknowledgment is important, though it does considerably less work than Gordon may imagine. We are not being deceived into believing that these people exist. We are being invited to enjoy their simulated existence, to appreciate how efficiently a machine can reproduce the recognizable intellectual and social characteristics of various literary types. The results are superficially impressive. The critics have different vocabularies, different objects of resentment, and different ideas about what constitutes an artistic achievement. They sound, at first, rather like people one might actually encounter at a poetry reading, a faculty meeting, or a particularly unpleasant conference panel.
Yet after reading all four essays, I began to notice that their differences were largely matters of costume. Beneath the costumes, the same verbal machinery was operating with dispiriting regularity.
Consider Mira K. Vell, the most sympathetic of Gordon’s inventions. Vell admires the poet’s extravagance and proposes to understand it through a familiar academic vocabulary of ornament, embodiment, nonsense, and excess. She writes that the peculiar beauty of Gordon’s work arises where “the language of the world, in all its vulgarity and magnificence, becomes something at once recognizably inherited and unexpectedly strange.”
This is an agreeable formulation, and in another context it might serve as the beginning of a useful argument. But what has actually been established? The language is vulgar and magnificent, inherited and strange, familiar and transformed. These terms are sufficiently elastic to accommodate almost any experimental poem whose author has made extensive use of found materials. Vell’s critical intelligence consists largely in discovering flattering descriptions for procedures whose value she has already decided to affirm. She offers an elegant vocabulary of appreciation, but one looks in vain for a moment in which her aesthetic commitments are seriously tested by the work under consideration.
I am aware that a critic’s sympathy with an author does not invalidate her judgment. Some of the most penetrating criticism has emerged from an intimate understanding of the writer’s ambitions. But sympathy becomes intellectually uninteresting when it functions as a guarantee that whatever the poet does will be interpreted as an achievement. Vell’s essay proceeds with a confidence that is suspiciously independent of its evidence. She has been designed to admire Gordon, and admire her she does. The machine has learned how a certain kind of academic admiration sounds, and it produces that sound with remarkable consistency.
Adrian V. Thorne has ostensibly been constructed to provide the opposite judgment. His essay, “The Alibi of Excess,” is considerably more severe, and I confess that portions of it are better than I expected. Thorne argues that Gordon’s poetry possesses an extraordinary ability to convert its potential deficiencies into evidence of aesthetic accomplishment. His opening paragraph contains the observation that “prolixity becomes abundance; arbitrary association becomes linguistic freedom; compositional negligence becomes an openness to accident.” He subsequently claims that Gordon’s self-consciousness functions as an elaborate mechanism of exoneration: by acknowledging the questionable nature of her procedures, she seeks to relieve herself of the obligation to defend them.
There is a genuine critical question here. I have wondered myself whether certain strains of experimental poetry have developed an almost inexhaustible capacity to assimilate objections into their own self-descriptions. Thorne’s account of a writer who anticipates criticism so thoroughly that criticism becomes another form of artistic material deserves consideration. It would be foolish to dismiss the observation simply because it has been produced by a language model.
Nevertheless, the essay’s apparent sophistication conceals the same structural difficulty as Vell’s enthusiasm. Thorne knows his conclusion before he begins. Every instance of verbal invention becomes evidence of evasion, every admission of vanity becomes a further expression of vanity, and every moment of pleasure is interpreted as a distraction from the absence of necessity. The argument is elegantly sealed against contradiction. Gordon cannot escape the charge of bad faith because anything she might say in her defense can itself be characterized as another manifestation of bad faith.
One notices, moreover, a peculiar resemblance between the two critics’ rhetorical habits. Both move with extraordinary fluency from an observation about a specific phrase to a comprehensive account of the poet’s aesthetic. Both employ distinctions that appear rigorous while remaining conveniently resistant to verification. Both repeatedly convert the same limited repertoire of compositional characteristics into sweeping judgments about authorial intention.
The approving critic discovers abundance wherever she looks. The disapproving critic discovers evasion. Neither seems capable of being surprised.
River Solstice Vale provides an especially instructive example of the technology’s dependence upon recognizable ideological conventions. Their essay, “The Aesthetics of Extraction,” begins by declaring Gordon’s poetry “a particularly insidious strain of late-capitalist experimental writing” and proceeds to characterize her practice as “poetry as resource extraction. This is poetry as shopping. This is poetry as a kind of gleeful, compulsive hoarding.”
The terms have been selected with admirable efficiency. Extraction, consumption, capitalism, hoarding: we know immediately where we are, what kind of person is speaking, and what moral verdict will follow. Vale’s literary judgment is inseparable from a prefabricated vocabulary of political disapproval. The critic has scarcely begun to examine Gordon’s poems before the available charges have arranged themselves into a prosecutorial sequence.
At one point, discussing a reference to a Central Asian sheep in Gordon’s “Poem to My Enemies,” Vale asks, “What would happen if the sheep were allowed to remain a sheep?”
I stared at this sentence for some time.
The question has the cadence of an ethical revelation. Its very simplicity seems designed to embarrass anyone sufficiently frivolous to have found the animal’s name or associations interesting. But it is difficult to determine what would constitute a satisfactory answer. The sheep is, after all, already a sheep. Its appearance in a poem does not alter its biological existence, and the fact that a poet uses an animal as a linguistic or imaginative resource does not in itself establish an act of material exploitation. Vale’s argument depends upon treating representation as a form of possession, an assumption that would make a considerable portion of world literature morally suspect.
The manufactured critic then contrasts Gordon with Brenda Iijima, Juliana Spahr, and Cody-Rose Clevidence, all of whom are said to understand the ethical obligations that Gordon neglects. I have no quarrel with these poets, whose work deserves attention on its own terms. What interests me is the extraordinary neatness with which they are conscripted into the argument. Each is made to exemplify a virtue whose absence has already been discovered in Gordon. Their works become illustrations of correct political consciousness, and the critical task is reduced to identifying which poet has demonstrated the approved relationship to the world.
Vale’s biography is almost as revealing as the essay. We are told that their practice operates “at the intersection of trans ecological justice, abolitionist poetics, anti-extractive aesthetics, disability solidarity, decolonial land relations, and radically embodied multispecies care.” The accumulation is clearly intended as satire, but it illustrates a rather dismal conception of contemporary political writing: a collection of interchangeable phrases, each signaling moral commitment without requiring any particular intellectual substance. The joke depends upon our recognizing the language as ridiculous before we have considered what any of its terms might actually mean.
I am not defending the rhetorical habits of contemporary activist poetry, which can be every bit as formulaic as those of any other literary movement. I am suggesting that parody ought to do more than reproduce its subject’s most readily identifiable verbal characteristics. Vale is a bundle of prejudices supplied with an ideological vocabulary. No comparable effort has been made to imagine the intellectual experiences through which those prejudices might have developed.
The fourth critic, Eleanor Whitcomb Hale, is perhaps the most depressing creation of all. Hale is an extravagantly celebrated American poet who believes, with a confidence apparently undisturbed by fifty years of literary argument, that poetry exists to communicate authentic human feeling. Her essay opens with a declaration that the poem must begin in courage, specifically “the courage to tell the truth.” She worries that Gordon’s facility with language has become an obstacle to emotional honesty, and she repeatedly urges the poet to abandon her elaborate verbal performances in favor of sincerity, vulnerability, and the ordinary dignity of human experience.
Hale’s conception of authenticity is so mechanically conventional that one begins to suspect deliberate cruelty on the part of her creator. She asks Gordon to remain with her tears, to describe a loss without decorating it, and to discover that the deepest achievement of poetry is to remind us that we are not alone. Eventually she arrives at the inevitable conclusion that somewhere beneath Gordon’s verbal extravagance there exists “a quieter poet. A braver poet.”
I would be interested to know how many thousands of workshop comments have contained some variation on this sentiment. The idea that the poet’s true self lies beneath the artifices of her style has become one of the more durable pieties of American literary culture. That it can now be reproduced almost instantaneously by a machine ought to give its habitual practitioners pause.
Hale’s discussion of Gordon’s “Grandmother’s Hands” is especially revealing. Confronted with a poem that transforms the grandmother’s hands into various fantastic and grotesque entities, she asks what the grandmother actually did with those hands. Did she knead bread? Did she braid a child’s hair? Did she hold the poet when she was frightened?
The questions are so perfectly representative of a certain pedagogical sentimentalism that one can almost hear the sympathetic murmur of the workshop. The grandmother must be retrieved from the poet’s imaginative transformations and restored to the familiar activities that establish her value as an object of familial affection. The hands must resume their proper duties in the kitchen and nursery. Anything else is treated as an evasion of grief.
I find Hale’s argument intellectually impoverished, but the essay’s poverty does not interest me nearly as much as the manner of its production. Gordon has asked a machine to manufacture a critic who misunderstands her work in a particularly gratifying way. The critic obliges. Her misunderstanding has been specified in advance, elaborated into a plausible essay, and decorated with a biography that makes her a former Poet Laureate of Iowa, a MacArthur Fellow, a National Book Award winner, and the recipient of enough additional honors to render ordinary professional achievement almost embarrassing.
She is then given a photographic portrait.
This is where the project acquires its most disturbing character.
Hale’s photograph depicts a benevolent older woman looking directly into the camera, her cheek resting in her hand, her expression communicating an almost oppressive solicitude. She appears to be listening with patient concern to a confession whose emotional significance she has already understood. The pose is familiar from the author photographs of countless books about healing, resilience, memory, and the redemptive power of storytelling.
Gordon apparently devoted considerable attention to the image, adjusting the fictional woman’s appearance until she achieved the desired combination of ordinariness and excessive compassion.
I find this deeply hideous. The technological accomplishment is substantial, but the achievement itself is an impoverishment. A human face has become a collection of legible psychological attributes that can be requested, adjusted, and discarded according to the user’s satisfaction. The wrinkles, expression, clothing, and arrangement of the hand suggest the presence of a life whose experiences have produced these visible effects. There is no such life. The woman has never listened to anyone, written a poem, taught a class, experienced grief, or received an award. Her compassion is a visual convention detached from the experience of compassion.
I understand that theatrical impersonation, portraiture, caricature, and literary invention have always involved the fabrication of character. I am not proposing that the creation of imaginary people constitutes some unprecedented moral crime. What seems new, and particularly dispiriting, is the indifference to the labor of invention itself. The imaginary person can now be generated through a succession of requests, each correcting an insufficiently convincing detail. The process begins to resemble the selection of upholstery. A little less attractive, please. More earnest. Less tired. Move the hand. Adjust the expression. Make the subject look more sympathetic.
At what point does the pleasure of creation become indistinguishable from the pleasure of ordering something?
There is a further irony in Gordon’s choice of subject. Her own poetry has frequently drawn upon the unpredictable peculiarities of actual language circulating in the world. Whatever one thinks of Flarf and its related practices, the poet’s encounter with found material offered at least the possibility of resistance. The language had an existence prior to the poem, having been produced by people whose motives, circumstances, and identities were not necessarily known to the writer. It could surprise, frustrate, or embarrass the person who appropriated it.
The synthetic critic offers no comparable resistance. Even disagreement is delivered as a service.
Thorne may accuse Gordon of aesthetic vanity, but he cannot refuse to participate in the performance. Vale may condemn her political irresponsibility, but they will not issue an unsolicited statement repudiating the use of their likeness. Hale may find Gordon insufficiently courageous, but she cannot challenge the representation of her own literary values. Every objection has been anticipated by the person who commissioned it.
There is a revealing moment in Thorne’s essay when he accuses Gordon of constructing an aesthetic system capable of transforming its deficiencies into the evidence of its own sophistication. The charge seems considerably more applicable to the gallery of critics than to the poetry it purports to examine. Whatever objection one raises against the project can simply become the basis of another installment. A reader finds the criticism too flattering? Produce a hostile critic. The hostility seems excessively academic? Produce an angry activist. The activist seems predictable? Produce a sentimental mainstream poet. The entire enterprise appears narcissistic and mechanically generated? Commission a critic who condemns narcissism and mechanical generation.
One can continue this procedure indefinitely, and I suspect Gordon intends to do precisely that.
I can already imagine the next critic. He will be a middle-aged literary essayist, perhaps educated at Yale, with a reputation for irascibility and a somewhat antiquated attachment to the independence of human judgment. He will have written several books lamenting the deterioration of literary culture. He will dislike artificial intelligence, resent the speed with which it manufactures seemingly competent prose, and regard this project as an especially depressing illustration of contemporary intellectual frivolity. His biography will mention an obscure literary quarterly that ceased publication some years ago. His photograph will probably be unflattering.
He may even be made to employ the expression “dead in the water.”
And when his essay appears, readers will be invited to appreciate the delicious irony of an AI-generated critic denouncing AI-generated criticism.
I cannot imagine a less interesting joke.
Its mechanism is obvious, and its success requires nothing beyond the continuation of the process it supposedly exposes. The machine has been asked to produce an argument against its own use, and it has complied. We are then expected to regard the compliance as a sophisticated commentary upon the instability of authorship.
But a system’s ability to imitate an objection does not constitute an answer to that objection.
The question remains whether anything resembling literary judgment is taking place. I would suggest that it is not. The gallery supplies the experience of being criticized without the inconvenience of encountering an independent critic. Its inhabitants have no intellectual commitments beyond those assigned to them, no reputations that might suffer from their errors, and no capacity to revise their judgments in response to experiences their creator has not anticipated.
Actual criticism is considerably less convenient. A critic may misunderstand a writer for reasons the writer cannot predict. She may recognize something the writer has failed to see. She may be indifferent to the work, refuse to review it, or produce an essay whose implications the author finds genuinely intolerable. These possibilities arise from the elementary fact that the critic has a life and a mind independent of the author.
I do not wish to romanticize the literary profession. Much published criticism is mediocre, derivative, self-serving, or plainly incompetent. I have spent a considerable portion of my career saying so. Nor do I deny that a language model may occasionally produce a useful observation, a felicitous formulation, or even a line of argument worth pursuing. Thorne’s remarks about self-consciousness and aesthetic exoneration suggest that such moments are possible.
But the occurrence of an intelligent sentence does not settle the question of what kind of activity has produced it. Literary criticism derives its significance partly from the commitment of a person to a judgment and from that person’s willingness to expose the judgment to disagreement. The synthetic critic performs the outward signs of such commitment while remaining constitutionally incapable of possessing it.
I suspect Gordon would reply that her gallery is a theatrical experiment, that the critics are deliberately exaggerated representations of recognizable literary attitudes, and that the project is less concerned with receiving genuine judgments than with exploring the rhetoric of judgment itself.
This is a defensible description of what she is doing. It is also one that considerably reduces the project’s claim to novelty. Writers have been inventing satirical interlocutors, fictitious reviewers, and elaborate critical personae for centuries. The difference here is that much of the labor of characterization and composition has been delegated to a system capable of producing a serviceable approximation of almost any familiar discursive style.
I remain unconvinced that this delegation has generated anything of substantial literary interest. Gordon’s project reveals how readily certain kinds of criticism can be imitated, but it tells us surprisingly little about the critics beyond the prejudices with which they were initially supplied. Its satirical targets are already familiar, and its chief source of pleasure appears to lie in the rapid reproduction of their most recognizable characteristics.
What troubles me, finally, is the possibility that the gallery represents a retreat from precisely the encounters that have made Gordon’s poetry worth reading. Her work, at its best, has engaged with the unruly linguistic world beyond the poet’s immediate control. Here she has constructed a literary environment in which even antagonism arrives according to specification. Everyone has something to say about her, everyone can be induced to say it at whatever length she desires, and nobody can decline the invitation.
Perhaps I am taking the enterprise too seriously. Perhaps its creator regards the whole business as an elaborate joke, and my objections merely demonstrate the familiar inability of the literary conservative to recognize contemporary play.
I have heard that defense before, and I expect to hear it again.
Yet I keep returning to the question with which I began. Why would a poet who has spent decades developing a distinctive sensibility wish to devote so much energy to manufacturing imaginary responses to her own work? There are actual readers in the world, some of whom might have something unexpected to say about it. There are actual critics, however reluctant or distracted, whose judgments cannot be revised by the author. There is even the possibility that no one will be interested enough to respond.
That last possibility seems to me especially important. The silence of another person is an experience a writer must sometimes endure. It is one of the conditions of publishing work in a world populated by people whose attention belongs to them.
The machine offers an apparently inexhaustible substitute for that silence. It will always produce another critic, another opinion, another elaborately credentialed personality, another sympathetic or hostile face. There need never be a moment when the writer confronts the possibility that no one wishes to speak.
I find that prospect considerably more depressing than the occasional stupidity of an actual literary review.
Gordon’s gallery is dead in the water. Its fifth critic, should she choose to manufacture him, will be no more capable of rescuing it than the first four.
And if his objections are subsequently incorporated into the entertainment, I hope readers will at least recognize how little that accomplishment proves.
About the author
Malcolm R. Sedgewick is an essayist, literary critic, and former editor of the independent literary quarterly The Human Measure, which ceased publication in 2017. He is the author of The Trouble with Play: Irony and the Evacuation of Meaning (2006), The Vanishing Reader: On Attention, Difficulty, and the Literary Life (2018), and Against the Machine: Literature and the Abdication of Judgment (2025). His essays and reviews have appeared in Harper’s, The New York Review of Books, The New Republic, and The Hudson Review. He received his Ph.D. in English from Yale University and has taught at several New England colleges. A longtime critic of technological solutionism and the commercial degradation of literary culture, he lives in Northampton, Massachusetts. He does not maintain a social media presence.
Editorial note: Malcolm R. Sedgewick is a fictional literary critic. This essay was developed through a collaborative dialogue between Nada Gordon and ChatGPT, using the same artificial intelligence technology its purported author so vehemently condemns.
