Weeding the Shelves: What to Keep, What to Let Go, and What a Library Is Actually For

A personal library is not really a collection. It is a set of decisions you made and then stopped revisiting. Books arrive one at a time, each individually justified, and nothing ever leaves, and after a decade you are living with the accumulated taste of several people you used to be.

That is not automatically a problem. But the part of this subject nobody writes about is the second half of the transaction — when a book stops earning its place, how you tell, and what to do about it. Acquiring is the easy, pleasant, well-documented half. Letting go is where a library either becomes a working tool or turns into storage.

First, the unread books are not the problem

There is a standing piece of tidying advice that you should get rid of anything you have not read. It is wrong, and it is worth saying why before anything else.

Umberto Eco kept an enormous library and was reportedly used to visitors asking how many of the books he had read, a question he found beside the point. Nassim Nicholas Taleb gave the idea a name in The Black Swan: the antilibrary. The unread books are the useful ones. They are the visible edge of what you do not know, sitting where you will bump into them. A shelf of books you have already finished is a record. A shelf of books you have not is a set of open doors.

So the distinction to make is not read against unread. It is between an unread book that is still a live intention and an unread book that is an artefact of somebody you no longer are.

You can usually tell by looking at it and noticing what you feel. Curiosity means keep. A small flinch of obligation means it has stopped being a book and become a reproach, and a shelf of reproaches is one of the more reliable ways to read less, because sitting down to read starts to feel like facing a backlog.

The categories that survive

Over time I have found that the books that genuinely earn permanent space fall into a small number of groups, and they are not the ones I would have predicted.

Things I return to in pieces. Reference in the broad sense — a dictionary I like, a couple of technical books I actually open, a poetry anthology that gets pulled down and put back twenty times a year without ever being read through. These justify themselves through use, and use is easy to verify.

Things I have reread and expect to reread. Rereading is the honest test of a book's place in a life, and almost nobody rereads as much as they imagine they will. A book you have read twice has proven something. A book you intend to reread has proven nothing yet, which is fine, but it belongs in a different mental category from the first group.

Things that are objects as much as texts. An edition somebody gave you, a book with your own handwriting in the margins from fifteen years ago, a copy that was in your bag through a particular year. These are not being kept for the text and should not be assessed as if they were.

Things that hold a period in place. A run of books from a phase of reading — a year lost to Russian novels, a stretch of obsessive reading about one subject — can be worth keeping together as a unit even if no individual volume would survive on its own merits.

What to release, and the honest reasons

Duplicate coverage is the easiest. Four introductory books on the same subject, of which one is clearly the best. Keep the best.

The aspirational shelf is harder and more valuable. Everybody has some version of it: the difficult classics acquired in a mood of self-improvement, the professional titles from a career direction you did not take, the language course. Some of these are still live intentions. Most are not, and distinguishing them is a matter of being honest about the next twelve months rather than about your ideal self.

Then there are the books kept for what they say about you. This is not a shameful motive — a library is partly a self-portrait and always has been — but it is worth noticing when it is the only motive. A shelf arranged for a visitor who has never come is not doing much.

And there is the category people hold onto out of sunk cost: the expensive book that disappointed, the one you were given and did not like. The money is spent either way. The shelf space is the only thing still being paid.

Rereading is the real practice

If curation has a positive form rather than a subtractive one, it is rereading, and it is the thing that gets squeezed out by the pressure to keep up with what is new.

The argument for it is not nostalgia. A book you read at twenty-two and again at thirty-five is functionally two different books, because most of what you get from a text is generated by what you bring to it. The second reading is where you find out how much you have changed, which is information you cannot get any other way.

It also fixes the acquisition problem at its root. A reading life organised around returning to a small number of books produces a smaller library almost automatically, because the pressure to own everything you might one day want relaxes.

There is a related argument about medium worth mentioning carefully. Comparisons of reading on paper and on screens generally find a modest advantage for print on comprehension of longer and denser material, and researchers including Maryanne Wolf have argued that sustained deep reading is a different cognitive activity from skimming. I would not overstate this — the effects are not enormous and depend heavily on the task — but it is one reason a physical shelf still earns its footprint in a house where everything else has gone digital.

Giving them away properly

Where books go matters more than people expect, because the difficulty of letting go is mostly the fear that the book will be wasted.

The best outcome by a distance is giving a specific book to a specific person who will want it. It takes more effort than a donation bin and it is the only method that reliably feels good afterwards. Second best is somewhere the book will be resold or rehomed rather than pulped: a secondhand dealer, a hospital or school collection that has actually asked for books, a street stall. In Indian cities the pavement booksellers and the old Sunday book markets are a genuinely efficient recycling system, and a book that passes through one of them will find a reader faster than most charity channels.

Two practical notes. Books with your own annotations are usually unsellable and should go to a person rather than a shop, if they go anywhere. And lending is not letting go; a lent book is gone with a small chance of return, so lend accordingly and stop keeping score.

The point of the exercise

A library that is curated is not smaller for the sake of being smaller. It is a library where you can find the thing you are looking for, where the sight of the shelves suggests reading rather than debt, and where every book on it is there because of a decision you would still make.

That requires making the decision more than once. Not on a schedule, and not as an annual purge — those tend to be destructive, because purging is a mood — but at the natural moments, when a shelf is full and something has to give. The full shelf is the useful constraint. It forces the question that acquisition never asks: given that this stays only if something else goes, does it stay?