‘Talent borrows, genius steals.’ Do you like it? I just came up with it. No, honestly. Any resemblance to the work of anyone else is purely coincidental.

The idea that taking someone else’s words and passing them off as one’s own constitutes a form of theft goes back to antiquity. Aeschines, one of Socrates’s disciples, was said to have read out dialogues appropriated from his master, to which one philosophically informed heckler blurted out: ‘Oh! you thief; where did you get that?’

But, as I learned from Roger Kreuz’s Strikingly Similar, it was the Roman poet Martial who gave us our modern word for this crime. Plagiarius means kidnapper. So when Martial accuses the poet Fidentinus of being a plagiarist, the sense is worse than theft: it is as though one’s children have been carried away and raised by someone else.

What might it mean to think of our words as though they were our property or even our children? The Argentinian writer Jorge Luis Borges tells the story of Pierre Menard, a Frenchman in the late 19th century, who made it his project to rewrite Don Quixote word for word. Easy you’d say: just copy it out. But Menard’s idea was to live his life – 300 years after Cervantes – in such a way that it would be the natural expression of himself to write exactly those words as if they were his own, written for the first time. Borges gives us an example:

It is a revelation to compare Menard’s Don Quixote with Cervantes’s. The latter, for example, wrote (Part one, Chapter nine): ‘… truth, whose mother is history, rival of time, depository of deeds, witness of the past, exemplar and adviser to the present, the future’s counselor’.Written in the 17th century by the ‘lay genius’ Cervantes, this enumeration is a mere rhetorical praise of history.

Menard, on the other hand, writes: ‘… truth, whose mother is history, rival of time, depository of deeds, witness of the past, exemplar and adviser to the present, the future’s counselor.’ History, the mother of truth: the idea is astounding. Menard, a contemporary of William James, does not define history as an inquiry into reality but as its origin.

And so on. Like all of Borges’s remarkable stories, it is short and playful, but also opens out into vast philosophical implications about what texts really are. Is it plagiarism if Menard is truly expressing himself? And isn’t every text unique to the moment in which we read it? As Roland Barthes might put it, can plagiarism really exist after the Death of the Author?

These, however, are not problems that trouble Kreuz, whose principal method, he tells us somewhat depressingly, was to type ‘plagiarism’ into the search function of the New York Times online archive and tally what he found there. What we have, then, is a series of anecdotes, each a page or so long, arranged under headings such as ‘Unconscious Plagiarism’, ‘Plagiarism and Politics’, ‘Unrepentant Plagiarists’, etc. We learn that the largest settlement ($5.4 million) ever paid out for copyright infringement in the music industry was by Michael Bolton and his record label over the similarity of ‘Love is a Wonderful Thing’ to an Isley Brothers song of the same name. And that Joe Biden’s political speeches throughout his career had a tendency to include phrases lifted from the speeches of his heroes. And that, having woken up with the tune to ‘Yesterday’ in his head, Paul McCartney spent a month humming it to anyone who would listen to make sure he hadn’t heard it somewhere before. ‘It became like handing something in to the police,’ he said. ‘I thought if no one claimed it after a few weeks, then I could have it.’

Kreuz’s style is engaging, and he has a nice way of telling a story. The trouble is that, while the stories pile up, the next part never arrives. One is constantly left wondering: where is the analysis, the complexity? Where is Kreuz’s own thinking? Strikingly Similar is exhausting in the way that a banquet of hors d’oeuvres might be. One keeps being served morsels that are in themselves interesting and tasty, but about halfway in you realise that there’s no main course on its way and despair sets in. I looked again at the publisher’s press release. It promised ‘the first in-depth history of plagiarism’. Unfortunately, depth is exactly what’s missing. If it were one of my students’ essays, the margins would be full of me scribbling: ‘So what?’

In fact let’s stick with that idea of the student essay. Why might an in-depth looking at plagiarism be an important book right now? Because for the last three years – since 30 November 2022, to be precise, when ChatGPT 3.5 went live – the whole basis of how we assess work in schools and universities has had the rug pulled from under it.

The crisis is an existential one. Unless the student is especially stupid or lazy or brazen, it’s extremely difficult to catch work that has been written with the aid of AI. Are we then wasting our time attempting to prosecute this? Or should we – and this is increasingly the preference of university managers, who are not immune to being stupid, lazy or brazen themselves – find ‘ways of working with AI’? To many, this sounds like an effort to eliminate plagiarism simply by redefining what the term means. To take my own discipline, English literature, what is it that we aim to teach if not the ability to read insightfully and to structure those insights in writing that is stylish and expressive? At what point in that process is there room for a mechanical servant?

Meanwhile, in case you weren’t aware of it, a $1.5 billion class action has been raised against the American AI company Anthropic. If you published a book at any point in history before July 2021 there’s a good chance you can join the lawsuit as a claimant.

At the end of Something Similar is a four-page breeze through the recent history of ChatGPT and its ilk. It reads as an afterthought. Kreuz – who might reasonably regard himself as a leading expert on the legal and ethical implications of all this – offers this brief opinion: ‘The training of large language models with copyrighted text may or may not be deemed problematic.’ Thanks for that, but it’s simply not enough.

Somewhere, between the enormo-scraping of Large Language Models and the writers – professionals as well as students – who use these tools to produce work which they will claim as their own, something is happening that looks very much like plagiarism. That ancient distinction between theft and kidnapping, between taking someone’s property and passing someone else’s children off as your own, feels like it might be useful. The matter might be complicated, but that’s why an academic such as Kreuz could be the careful and thorough explainer that we need. If only he would push harder at the questions he raises and not just tell us stories.

Plagiarism is big news now, and Strikingly Similar trades on that: the subtitle promises ‘Chaucer to Chatbots’. Unfortunately, Kreuz drops the ball. Imagine a travel writer perched in one of the Titanic’s lifeboats, filing their report on the deliciousness of the food and the jauntiness of the band. The incident with the iceberg may or may not be deemed problematic.