Even the great Sherlock Holmes might have been confounded by the latest conundrum surrounding Artificial Intelligence. There are allegations that this year’s Commonwealth Short Story Prize has been awarded to a story that was written using AI. But the scrutiny of Jamir Nazir’s “The Serpent in the Grove” reveals something more insidious than the possibility of AI-generated fiction winning a prestigious literary competition. It shows how profoundly mistrusting readers have become of language itself. Readers dissecting Nazir’s story pointed to repeated constructions, elaborate metaphors, patterned phrasing and — the much maligned — em dashes as evidence of possible machine authorship, moving far away from the story’s emotional or literary qualities.
This scepticism obscures a more serious transformation in contemporary writing. The suspicion around AI-generated prose has created a climate in which readers increasingly question the authenticity of elaboration, density and repeated patterns, even though these linguistic quirks have existed for centuries. Many of the features identified as ‘AI tics’ in Nazir’s story are familiar literary devices — repetition, parallel syntax, ornamental imagery and recursive phrasing belong to traditions extending from William Faulkner and Toni Morrison to Arundhati Roy and Jamaica Kincaid. Large Language Models absorbed those rhythms through training data, making it impossible to differentiate between man and machine at times.
This has given birth to a peculiar anxiety. Readers now read literature for meaning while scanning for signs of fabrication. A metaphor that lingers too long can trigger suspicion: one wonders what the sceptical reader would make of Thomas Hardy’s pages-long metaphor on Egdon Heath. A sentence that winds its way through many subordinate clauses — think Marcel Proust — risks sounding algorithmic. This has implications beyond a disputed literary prize. Writers are attempting to adapt themselves to these anxieties. One writer told the Wall Street Journal that she “use[s] aggressively casual language, like, ‘hey yo, for real,’ or drop[s] a bunch of exclamation points” to sound human. The same article has writers confessing that Walt Whitman’s beloved trick of repetition has become too risky; even those following Strunk & White’s The Elements of Style to write prose that appears too polished have been accused of using AI. Writers are thus beginning to flatten language or debasing it pre-emptively, removing idiosyncrasies before anyone else can accuse the prose of sounding machine-made.
AI reproduces literary styles with astonishing fluency and readers exposed constantly to AI-generated prose begin associating stylisation itself with fabrication. The resultant shift is subtle but profound — language starts drifting towards safer territory: cleaner syntax, flatter rhythms, colloquial vocabulary and emotionally controlled prose that cannot easily trigger suspicion. Yet, the literary styles now viewed suspiciously are precisely the traditions that taught generations of readers how language could move beyond communicating information to evoking feelings. Toni Morrison’s Beloved depends upon repetition because trauma itself is repetitive. Arundhati Roy builds emotional pressure through recurring images and verbal echoes because memory in The God of Small Things unfolds through association rather than chronologically. AI did not invent these techniques; it borrowed them.
This atmosphere of disbelief also robs readers of the pleasures of reading. Literature’s magnetic pull relies on the belief that a kindred consciousness is speaking to the reader intimately across time and space. AI disrupts that faith by forcing readers to question whether there is such a literary consciousness at all. Once that doubt takes hold, the emotional contract between the reader and the writer is broken forever.
