Why the Classics Are Usually a Good Bet

If you are looking for a worthwhile book and know nothing else, a classic is usually a rational place to start.

This is not because old books are automatically good. Most old books are gone.

That fact is part of the argument.

Survival is information

Far more books once existed, circulated, mattered to somebody, and disappeared than survive as classics now.

A work that remains readable after centuries has usually passed through repeated acts of preservation. Someone copied it. Someone kept the copy. Someone taught it, quoted it, translated it, printed it, edited it, argued about it, imitated it, or decided that another generation should still have access to it.

Those decisions were not made by one committee at one moment. They accumulated across very different historical settings.

This is the useful sense in which a classic has survived a winnowing test of time.

If a new novel appeared yesterday, we know almost nothing yet about whether readers will still find it worth serious attention in fifty years. If people have voluntarily returned to a book for five hundred or two thousand years, that history does not settle the question of quality, but it should change our expectations.

The filter is not fair

The phrase “test of time” can become silly if it implies that history conducts a clean competition and the best books simply win.

Textual survival is much messier.

Works vanish because manuscripts burn, decay, are discarded, are never recopied, survive in the wrong place, fall out of school curricula, lose institutional patrons, or happen to exist on fragile material. Languages disappear. Libraries are destroyed. Religious and political institutions preserve some authors and neglect others.

The Oxford Handbook of Byzantine Literature's discussion of manuscript transmission, for example, stresses material causes, library context, educational canons, and institutional priorities as factors in what survived.

A recent study modeling manuscript traditions likewise emphasizes both cultural selection and historical contingency. Its central point is sobering: textual transmission is marked by very large losses, and drift as well as deliberate human choice helps determine what reaches us. The 2026 PNAS Nexus paper is available here.

So no, survival does not prove greatness.

But imperfect evidence is still evidence

The fact that a filter is biased does not mean it contains no information.

Suppose a work has been copied in many periods, translated into many languages, assigned by teachers with very different assumptions, borrowed for later art, attacked by critics, defended by admirers, adapted to new media, and still read by people who are no longer required to read it.

That history is evidence of unusual durability.

It may be durable because the work is beautiful. It may contain characters or arguments that remain difficult to exhaust. It may have become foundational to later literature. It may be historically indispensable. It may combine several of these.

Those are different kinds of value and should be distinguished rather than flattened into the word “important.”

A classic can be overrated and still worth reading

The canon is not a list of books you are morally required to admire.

Some canonical books will seem mediocre. Some will matter more as historical causes than as aesthetic achievements. Some reputations are sustained by institutions long after ordinary readers would have abandoned them. Some excellent works were excluded or lost for reasons unrelated to quality.

Reading the classics seriously includes the possibility of concluding that a famous book is not very good.

But there is a difference between testing a reputation and assuming that the reputation itself contains no evidence.

Centuries of attention create a reasonable presumption that there is something there worth inspecting.

The cost of trying has collapsed

There is also a practical reason classics are such a good bet now: many of them are in the public domain.

For most of literary history, encountering a particular old book required access to a manuscript, a library, an expensive printed edition, or at least a bookseller who happened to stock it. Today an enormous portion of the surviving canon can be opened immediately at little or no cost.

That changes the decision.

You do not have to believe that everyone should read Xenophon, Austen, Frederick Douglass, Milton, Sappho, Darwin, or Dostoevsky. You only have to notice that the price of testing whether one of them speaks to you is now often a few minutes.

Use the canon as a recommendation engine, not a commandment

When choosing among ten thousand recent books, the signals are noisy. Marketing budgets, launch timing, social media, review cycles, celebrity, and current fashion all push attention around.

Older works have their own distortions, but they also have a signal unavailable to a new release: continued life.

That is why a reader who wants something likely to reward attention can sensibly ask what has survived before asking what is currently being promoted.

You can still quit the book. You can dislike it. You can decide that history preserved the wrong thing.

The canon is not a command to admire. It is a recommendation produced, imperfectly, by generations of readers.

That is more modest than reverence and more intelligent than pretending two thousand years of attention tell us nothing.