When Americans bought Doctor Zhivago in 1958, they weren’t only buying a novel. They were buying into one of the defining publicity events of the Cold War: a banned book, a refused Nobel, the only rival superpower visibly rattled by fiction.

Part 1 of an occasional series examining The Atlantic’s “The End of Reading Is Here”

Rose Horowitch’s recent essay for The Atlantic, “The End of Reading Is Here,” runs to roughly 8,600 words. It opens at the Library of Alexandria, takes in Gutenberg, McLuhan, Postman, and Donald Trump’s syntax along the way, and arrives, eventually, at a diagnosis: America has become postliterate, and the age of reading may turn out to have been a brief interlude between the oral and the digital.

Losing the will to live already? Me too. The sky is always falling in the world of publishing

But there is a certain delicious irony in an 8,600-word obituary for long-form attention, published in a magazine whose own subscribers are demonstrably still reading long-form journalism. Horowitch even notices this herself, in passing, without ever quite reconciling it with her headline.

But the more serious problem with the essay isn’t its length. It’s the evidence she reaches for near the start, to establish that Americans once had both the appetite and the stamina for demanding prose and that that has somehow evaporated.

That evidence is, don’t laugh, Boris Pasternak’s Doctor Zhivago.

An elegant comparison, built on sand

Horowitch quotes one of Pasternak’s long, syntactically layered sentences from almost seventy years ago, then sets it against a passage from last year’s bestselling romantasy novel. The implication is clear enough: in 1958, ordinary American readers embraced complex literature; today, they don’t. Zhivago becomes Exhibit A for a vanished reading culture.

It’s a persuasive move on the page, if you know nothing about Pasternak and the book, and sincerely want to believe reading is in its death throes as Horowitch clearly does.

It is also, as anyone who knows the book’s actual publishing history will recognise, close to the least representative bestseller she could have chosen.

Doctor Zhivago was not simply a long, well-written Russian novel that happened to top the American charts.

This was 1958. Eisenhower was the US president, Khrushchev his Soviet counterpart, and McCarthy dead barely a year. The Bay of Pigs and the Cuban Missile Crisis was still to come. Reds were under every bed, and along came a piece of literature that played directly to that social and political environment.

Doctor Zhivago was banned by Soviet censors, smuggled out of the USSR, and first published in Italy before reaching English-language readers. Pasternak was awarded the Nobel Prize in 1958 and then forced by the Kremlin to decline it – a story that ran on front pages for months.

And behind the scenes, Western intelligence had recognised an opportunity: declassified files confirm the CIA ran a covert operation, cryptonym AEDINOSAUR, to get Russian-language copies of the novel into Soviet hands, including at the 1958 Brussels World’s Fair. The aim wasn’t literary appreciation. It was to demonstrate, to a domestic and international audience, that Moscow was frightened of a book.

None of that appears in Horowitch’s account. For someone so keen on classical history cosplay (see the opening of the Atlantic’s piece were Julius Caesar and Alexander the Great set the scene for Boris Pasternak) Horowitch is either ignorant of more recent history (a damning indictment of the US education system overall, if so) or chose to deliberately ignore it.

Horowitch brings Pasternak himself into the essay purely as a data point – complex novel, 1958, bestseller – with the context edited out. But the context is precisely what explains the sales figures.

When Americans bought Doctor Zhivago in 1958, they weren’t only buying a novel. They were buying into one of the defining publicity events of the Cold War: a banned book, a refused Nobel, the only rival superpower visibly rattled by fiction. It is difficult to imagine a bestseller less useful as a neutral baseline for “how demanding a book the average reader could handle.”

The prestige-purchase problem

Publishing has a name for what happened next, even if the term is more recent: the Hawking Index – the gap between books bought and books actually finished. A Brief History of Time sold in the millions and was read, by most measures, in rather smaller numbers. Infinite Jest and Capital in the Twenty-First Century invite the same suspicion. There is every reason to think Zhivago belongs in that company.

Owning it, in 1958, was a statement – about intellectual seriousness, about which side of the Cold War’s cultural argument you stood on. Whether the average buyer parsed every clause of Pasternak’s prose is a separate question entirely, and one the bestseller list cannot answer.

A translation problem, not a cognition problem

There’s a second flaw in the comparison, and it has nothing to do with politics. Doctor Zhivago is a Russian novel, translated into English in the late 1950s by Max Hayward and Manya Harari.

Full disclosure: my Russian runs to a handful of everyday tourism phrase book sentences (“Excuse me but I have diarrhoea. Where can I buy bicycle clips?”) and slightly more reading words from frequent visits last century, so I asked Claude about the linguistics issues here. Claude explained (you can tell by the em-dashes, right?) “Russian syntax runs naturally to long, multi-clausal sentences, and mid-century translation convention — along with mid-century editorial taste more broadly — favoured a more formal, expansive English than most trade editors would permit today.”

In other words, setting a 1958 translation of Russian sentence structure against a 2025 American YA sentence, as Horowitch does, isn’t measuring reading stamina across generations. It’s measuring the difference between translation conventions, editorial fashions, and genre expectations – variables that would produce the same contrast even if reading ability hadn’t moved an inch.

But the wider point holds beyond translation. Sentence length has never tracked cognitive seriousness particularly well.

Hemingway built a career on short sentences – his minimalist style is industry legend – and nobody has ever accused him of writing lightweight books.

Orwell prized plainness. No-one reads 1984 or Animal Farm to marvel over the language.

Steinbeck’s Of Mice and Men – for decades the most widely set text in British secondary-school English classrooms, not least because its brevity made it teachable – would, published today under Horowitch’s logic, presumably be cited as evidence of decline rather than as the tightly controlled piece of craft it is.

Kazuo Ishiguro writes some of the most quietly devastating prose in contemporary fiction in sentences a ten-year-old could parse.

Cormac McCarthy will give you a page-long sentence and then three words. Sentence length is a noisy, unreliable proxy for intellectual ambition, and it always has been.

What the rest of the 1958 list actually looked like

If we want to know what American readers were actually drawn to in 1958, the more useful exercise is to look past Zhivago to the rest of that year’s bestseller list: Leon Uris’s Exodus, Allen Drury’s Advise and Consent – big, dialogue-driven, politically charged epics and thrillers.

The 1950s reading public clearly was not, on this evidence, uniformly hungry for lyrical, syntactically demanding prose. It was hungry for scale, plot, and Cold War anxiety, rendered in a mix of styles from the ornate to the plain. That’s a stylistically mixed list – not so different from bestseller lists today.

The real question, and a better one to ask

None of this is to say Horowitch’s underlying concerns are baseless. Reading-comprehension scores among younger cohorts have absolutely declined; attention spans measured on hand-held screens clearly have shortened; these are real and worth taking seriously.

But serious claims need evidence that can bear the weight put on it, and a single, historically singular outlier – a book whose sales were substantially manufactured by a superpower’s intelligence agency and a Nobel controversy – fails abysmally to do that work.

The better – and more relevant – question isn’t whether Americans once read Doctor Zhivago. It’s why they did. Or at least, why they bought it. And once you ask that question properly, the answer has almost nothing to do with sentence length, and everything to do with the Cold War being, among other things, the greatest unplanned book marketing campaign in publishing history.

Reading hasn’t ended. Utter clickbait nonsense. As later instalments in this series will explore – through fantasy doorstoppers that dwarf Zhivago in length, through prestige television that behaves more like the nineteenth-century serial novel than 1950s TV ever did, through a look at the audiobook scene, and through what BookTok and Netflix adaptations are actually doing to print sales – what’s changed is where reading happens, how reading is defined, and who’s counted as doing it.

That’s a seriously interesting story. It just isn’t the one The Atlantic‘s headline is selling.

This is Part 1 of an occasional TNPS series examining the evidence behind “The End of Reading Is Here.” Part 2 will look at whether sentence length has ever been a reliable measure of literary seriousness.

As ever, I’ll make no commitment to a timetable. The Gambian power crisis has eased, but the monsoon season is arriving, and TNPS will continue to be haphazard in delivery, but always ready to call out the contradictions in the industry’s many narratives.

This post first appeared in the TNPS LinkedIn Analysis Newsletter.