Read vs Listen: Yesteryear

One voice, two strands, and a line on the page between what the narrator thinks and what she says out loud.

Confidence 78%, which puts it in the leaning band. Bands run from close call at 50 to 64, through leaning at 65 to 79 and clear at 80 to 94, up to clear cut at 95 and above.

What this book actually is

A debut novel in three parts plus an epilogue, told in the first person throughout by one narrating character, Natalie Heller Mills. Two strands run through it, one in the twenty-first century and one in 1855, and Publishers Weekly describes the action ping-ponging back and forth between them. The separation is done with verb tense: the modern strand in the past, the 1855 strand in the present. No source found shows a dated chapter heading or an era label anywhere in the book, and one reader tracking the pattern notes at least one chapter that breaks it.

Chapters are short. Burke told The Rumpus that most of them are probably fifteen hundred words or less. A study guide puts Part 2 at 57 chapters, which is a single uncorroborated source, and whether the parts and epilogue carry titles, or whether chapter numbers restart with each part, is not documented anywhere we could check.

Knopf published it in the US on 7 April 2026 and 4th Estate took it in the UK after an eleven-way auction, publishing the hardback on 9 April. HarperCollins UK's product page carries a Sunday Times number one bestseller line, it was a Good Morning America Book Club pick for April 2026, and it sat at number three on the New York Times Combined Print and E-Book Fiction list dated 16 August 2026, in its seventeenth week. Page counts do not agree: 400 is listed against the UK hardback and 391 against the UK and export paperback, with both figures appearing against the US hardcover.

On the screen adaptation, keep expectations at headline level. Amazon MGM Studios won the film rights in July 2024 with Anne Hathaway attached to star and produce, and Hannah Friedman was reported in June 2025 as writing the screenplay. Those trade pieces returned paywall errors, so the article text could not be read and only the studio, the star and the screenwriter are being repeated here. No director and no release date were found. Readers file content warnings for sexual assault, child abuse, violence and domestic abuse, and one reviewer flags heavy swearing and on-page sex. Those apply to the book, so they apply to both formats equally.

The case for reading it

Three things, and the third is the one that decides it.

The book runs on two eras that alternate rather than proceed, and the past-tense strand is not a single block either, taking in university material around 2010 alongside the recent present. A page lets you glance back a paragraph to check where you are. It also lets you notice the chapter that breaks the tense convention, rather than simply feeling briefly lost.

Then there is the surface of the prose. Sentences stop unfinished, scenes end abruptly, and misquoted Scripture is inserted among fragments of thought and speech as a deliberate device. Spotting that a verse is slightly wrong depends on being able to stop and look at it again, which is a thing print makes easy and continuous audio does not.

The decisive one is typographic. Italics hold Natalie's interior monologue apart from what she actually says aloud, and one detailed reading of the structure reports that this boundary carries plot weight later on rather than being a styling choice. Print draws that line for you on every page. In a recording it exists only if the narrator draws it, and no review found says whether she does.

The case for listening to it

Stronger than the count suggests. This is one voice for one narrator with a small, ordinary cast of names, which is the configuration audio handles best, and Rebecca Lowman reads all of it with nothing running alongside her. The chapters are short enough to stop and start around, at roughly fifteen hundred words each. Burke has said the casting was aimed at finding someone who could carry the novel's interiority, which is a fair description of what the book asks of a reader.

The point that matters most is that the thing normally lost on audio is not lost here. Because the two eras are separated by tense rather than by a date at the top of a chapter, a listener hears the change in the verb at the point a reader sees it. What holds the audio back is what is absent rather than what is wrong with it: no awards or shortlistings were found, no AudioFile review was located, and the reviewers who have written about the narration disagree with each other about it.

The audiobook in detail

Random House Audio, 2026, 13 hours 47 minutes, audiobook ISBN 9798217174331. That is about 9 hours 11 minutes at one and a half times speed, a figure we have worked out rather than one anybody publishes. The exact release date, as distinct from the year, is not stated on any listing we checked. Audible UK lists the publisher as Penguin Random House LLC while UK print rights sit with 4th Estate; narrator and duration match across the two English listings, which is consistent with one recording licensed on both sides, but no source actually says so. Production credits are Melissa Bendixen and Alanna McAuliffe producing, Phoebe Neidl editing and Neil Baczek engineering. There is also a separate Spanish recording, read by Sarah Muñiz Hanif and translated by Mar García Puig for AdN, Editorial Grupo Anaya, with no duration published on the listing we saw. One reviewer mentions interjections of interviews or commentary about the children within the text; how those are set on the page, and what either format does with them, is not stated anywhere, and it is a single unconfirmed source.

Who should ignore this verdict

Anyone who samples it first. The whole read case rests on one typographic device, and if Lowman marks the shift into interior monologue clearly in the opening hour, the gap between the formats more or less closes. Anyone who reads in short bursts, because chapters of fifteen hundred words are built for a commute. And anyone who follows a shifting timeline more easily by ear than by eye, since here the shift is a change of tense rather than a heading, and hearing it happen in the verb may well beat spotting it in print.

Related verdicts

Questions people ask