In the summer of 1952, the Encyclopaedia Britannica published “Great Books of the Western World,” fifty-four volumes, leather-bound, the kind of set designed to sit impressively on a shelf and communicate, wordlessly, that the owner had serious intellectual intentions. Mortimer Adler, the philosopher who organized the project alongside the University of Chicago’s president Robert Maynard Hutchins, had spent years arguing that there was a specific body of literature every educated American ought to absorb before they died. He called it “the great conversation.” These were the texts, he believed, that contained the essential arguments of Western civilization.
It sold reasonably well. You can still find copies in used bookstores, the spines often uncracked, pages uncut. The set communicated serious intentions. For most of its owners, it communicated exactly those intentions and nothing further.
I’ve been thinking about Adler because I’ve been asked to write a list of books to read before you die, which sounds straightforward until you sit with it. The real question isn’t which books. The real question is: who decided which books belong on the list? How did a list become the answer to a question nobody precisely asked?
There’s a lineage worth tracing.
Adler and Hutchins didn’t invent the canonical reading list. They commercialized it. A Columbia instructor named John Erskine had been running a great-books seminar since the early 1920s, and the idea followed the path of institutional imitation: Columbia was doing it, so Chicago adopted it, so Encyclopaedia Britannica packaged it, so the aspirational middle class bought it. By 1952, the great conversation had a price tag and a mailman who would deliver it to your door.
Adler was honest, in his way, about what he was selecting. He wasn’t claiming these books were objectively great by some natural law. He was claiming they were great by the standard of Western civilization, which he believed was worth transmitting. This is a defensible position. But look at the authors in those fifty-four volumes and you’ll see exactly which civilization he had in mind: Western European, overwhelmingly male, almost entirely dead. No women until Virginia Woolf was added to the revised edition in 1990. Nothing from outside the European tradition. The great conversation, it turned out, was a conversation among a fairly specific group of people who had been dead long enough to be admired safely.
Harold Bloom made the argument more explicit and more combative when he published “The Western Canon” in 1994. He knew the canon was a selection, not a discovery. His defense was essentially Darwinian: these books survived because readers kept returning to them, generation after generation, and the ones that didn’t survive deserved to disappear. This is circular in a way Bloom acknowledged without conceding. The books that survived are the ones that got taught. The ones that got taught are the ones that already had institutional momentum. Bloom would say the momentum is evidence of value. His critics would say it’s evidence of who ran the institutions.
Both things are probably true. That’s how these patterns tend to work: one set of decisions, made by a small group in a specific decade, borrows the authority of an institution, repeats until it feels like common sense, and arrives at you looking like a fact about the world rather than a choice someone made. The difference between a canon and a preference is mostly repetition and a university press.
Here’s the problem with “books to read before you die” as a category.
It carries the implicit claim that there is a correct set, that someone has identified it, and that your job is to work through it before the clock runs out. Forty years of reading has left me confident about one thing: that’s not how it works. Reading isn’t certification. You don’t become educated by finishing a list any more than you become physically fit by owning a gym membership. What lists do, at their best, is give you an entry point. At their worst, they give you the right to stop looking.
What I can offer is something more honest and considerably more personal: the books that actually changed how I think. Not the ones I was supposed to admire. Not the ones on the shelf whose spines I’ve read and whose pages I haven’t. The ones I was still thinking about three weeks later, and six months later, and in some cases twenty years later.
Thomas Kuhn published “The Structure of Scientific Revolutions” in 1962, and it remains one of the most cited academic books of the twentieth century, which means nearly everyone has encountered the phrase “paradigm shift” and somewhat fewer have read the book that gave us that phrase. Kuhn’s argument, stripped to its load-bearing structure: science doesn’t advance through steady accumulation. It advances through rupture. A scientific community builds a shared framework of assumptions, methods, and problems worth solving. Within that framework, it does normal science: extending the model, filling gaps, solving what Kuhn called puzzles. But every framework produces anomalies, results that don’t fit. For a long time, the community ignores them or explains them away, because the framework is working and revision is expensive. When anomalies accumulate past a certain threshold, the whole thing breaks, and a new framework replaces it.
I read this at twenty-four and understood it as a book about science. By thirty, I understood it as a book about how any community of experts maintains its consensus. By forty, I understood it was about me.
The thing Kuhn makes clear is that the people inside a paradigm can’t easily see its edges. The assumptions that structure your thinking are precisely the ones you don’t examine, because they’re the ones you’re thinking with. This isn’t a comfortable observation. It is, by a comfortable margin, the most useful one I’ve ever encountered.
Jane Jacobs published “The Death and Life of Great American Cities” in 1961. She wasn’t a credentialed urban planner. She was a magazine writer and a Greenwich Village resident who had been paying close attention to what the experts were getting wrong. In 458 pages she dismantled the theoretical foundation of American urban planning and replaced it with something that held up for the next sixty years.
Her argument: the experts building cities understand diagrams of cities, not cities. The features that look disorderly to the trained planner, the mixed uses, the short blocks, the old buildings, the sidewalk life, aren’t failures of design. They’re the mechanisms by which cities actually function. Strip them out in the name of order and you get ordered emptiness.
I came to this book in my late thirties, through a zoning meeting I was covering that had nothing to do with Jacobs. A planner mentioned her almost in passing, as someone everyone in the field had read and was now somewhat embarrassed to fully admit was right. I found the book the following week and didn’t put it down.
It changed how I see every street I walk on. That’s the real test, I think. Not whether you agree with a book, though in this case I mostly do. Whether, after reading it, you see something in the world that you couldn’t see before. Jacobs reorganized my perception of the built environment, permanently, in a way I hadn’t asked for and can’t undo. My own book about neighborhoods, published a few years ago, exists in large part because of the questions she asked first.
Studs Terkel’s “Working,” published in 1974, is nearly six hundred pages of people describing their jobs. A stonemason. A waitress. A strip miner. A firefighter. A parking lot attendant. A model. Terkel interviewed them in the early 1970s and let them talk, and what he captured is something no economist or sociologist had managed before or since: the actual texture of work, the dignity and the monotony and the fury and the small satisfactions, in the words of the people doing it.
There’s no central argument, no thesis to extract and carry around. That’s the point. The book refuses to argue its way to a conclusion. It listens. And through listening it makes a claim that I find irrefutable: everyone has a relationship with their work that’s complicated in ways they’ve almost never been asked to describe, and the asking turns out to matter.
I go back to “Working” more than almost any other book I own. I go back to it when I’m writing about people rather than ideas, because it reminds me what I’m supposed to be listening for.
Daniel Kahneman’s “Thinking, Fast and Slow” was published in 2011, when I was forty-four and thought I knew a reasonable amount about cognitive bias. I’d been writing about behavioral economics for years. I’d read most of the primary literature. I’d assigned Kahneman and Tversky papers in the courses I taught.
Then I read the book and caught myself, on page 71, committing the exact error Kahneman was describing. Anchoring. That morning I’d been in a conversation about renovation costs, and the first number mentioned had been wildly inflated, and I’d spent the conversation negotiating down from that anchor rather than asking what the work actually cost. I put the book down. I sat with that for a moment. I picked it up again.
This connects to what the research on experienced thinking shows: knowing that a bias exists doesn’t make you immune to it. The best books on how thinking works don’t teach you to think correctly. They teach you to catch yourself thinking incorrectly, in real time, and that’s a different and more durable skill. Kahneman’s book kept doing that to me for months after I finished it. It still does, which is something I can’t say about many books I thought were important when I read them.
I’ll give you one more.
James Watson’s “The Double Helix,” published in 1968, is the most honest account of how a major scientific discovery actually happened, which makes it somewhat uncomfortable for anyone who has invested in the cleaner mythology of science as a pristine pursuit of truth. Watson isn’t entirely likable in this book. The credit goes to him and Crick in ways that don’t fully account for Rosalind Franklin, whose X-ray crystallography data was critical to the model. Watson basically admits this. He writes about it the way someone writes about something they know wasn’t quite right and has decided to be candid about rather than deny. This isn’t exculpatory. It is, however, honest in a way scientific memoirs almost never are, and the honesty is precisely why you should read it. Science happens inside human institutions, with human politics and human vanity, and a book that shows you this clearly is doing you a service no textbook will.
My list isn’t Adler’s list. It isn’t Bloom’s list. That’s not perversity. It’s the recognition that any list curated by any set of people, in any decade, for any set of reasons, reflects the curators as much as the books. If you want another starting point from someone who spent more time than I did watching actual readers make actual choices, Arthur Dandridge’s recommendations are worth your time. His list differs from mine in ways that are instructive rather than contradictory, which is how good reading lists tend to work.
The most useful list is the one you build yourself, a book at a time, by noticing what startles you and following that feeling to its source. You’re not trying to finish the list before you die. You’re trying to keep making it longer. If you’re still adding to it, you’re probably still thinking. That seems like enough.
I’m fifty-nine. My list runs to somewhere around four hundred books, and I’m fairly certain I’ve got at least a third of it wrong. I keep reading. It keeps being worth it.
Warren Holt writes about ideas and culture for The Sunday Evening Review. He is the author of two books, including “The First Answer Is Usually Wrong.”
