Four letters, arranged in strings the length of a gene, can spell more sequences than any list could hold. Life has written just a small share of them.
I belong to an order of cataloguers. We are not many and, as it happens, we are not really celebrated: ours is held to be the least of the sciences. We write down the names of things someone else found and we decide where they belong on the shelves. The discoverer is rightfully remembered. The cataloguer is just the hand that files the discovery and is forgotten by the following morning. Honestly, I am at peace with this. Someone has to know where the books are kept.
The library we keep is very old and, funnily enough, no one built it. Rumour is it assembled itself over four billion years or so. It has no walls that I have found and no roof. The older cataloguers taught that it has no center either and that any shelf may be treated as the first shelf, which helps, because we are lost most of the time.
There is one very fundamental law in the library; any apprentice learns it before he learns to read a spine: every book here is true. By that we mean that every volume on these shelves corresponds to something that lived. That is our whole doctrine, and our pride. You may have read of another library, the famous one, with its hexagonal galleries and its volumes holding every possible arrangement of letters, almost all of them nonsense. That library contains everything that could be written. Ours contains only what was written by living, and survived the writing. People sometimes confuse the two. Theirs is infinite and almost entirely false. Ours is finite and entirely true.
Books are written in an alphabet of four letters. I am not making a simplification for you; it is four. Every volume in every region of the library, the ones near the door and the ones no lamp has reached, is set in the same four characters, which is why a cataloguer trained on one shelf can in principle file a book from any other, and why we believe, though we cannot prove, that the library is a single work and not many.
Let me take you in, as far as anyone goes.
The shelves nearest to the door are the ones most people return to over time. Just a few thousand volumes, opened so many times by so many readers that the bindings have gone soft. Human. Mouse. Fly. Yeast. E. coli. Any biologist could recite them without thinking. A reader entering the library is handed these first. Earlier readers have left maps in the margins, and a margin with a map in it is worth more to us than any unmarked shelf. A path gets marked, and then walked until it is a road, and somewhere along it there are apprentices who will spend whole careers on the second-to-last paragraph of a single book about a fruit fly.
Behind these shelves the library widens and the proportion changes. There is a region, larger than the first by some factor, of books that were named but never truly opened. A spine with a label. An entry in a register. A note that says: "this is a beetle from a particular forest in a particular year, and these are its four-letter pages". The cataloguer who wrote the label moved on the same day and the book has not been opened a second time. There was roughly a century when to name a book was nearly to read it because discovery was slow and a name was a promise that a reader would come. That age, fortunately or unfortunately, is now over. The instruments improved. Now the naming runs so far ahead of the reading that the gap between the two widens every year, and no one I know expects it to close.
Deeper still, the books stop being books. There is a region, vast beyond the named shelves, where what we hold are not volumes anymore but pages. Fragments. Sentences with no book around them, lifted from the places where life is too crowded and too small to be taken one organism at a time: a handful of soil, a cup of seawater, the inside of a gut, some poo. A sampler reaches in and brings up a thousand handwritings at once, mixed past sorting, and we shelve what comes back as best we can, page by page, with no spine to say where one book ends and the next begins. A reader at this depth can't take down a volume. He dredges and reassembles what he can, and what reassembles in his hands is often a chapter and almost never a book. The apprentices usually find this region distressing because it offends the catalogue. I have come to like it better than the shelves with proper spines. A sentence can be true with no book around it.
And then there is a part of the library I can describe only by not having been there, because no cataloguer ever has. To be fair, it is most of the library. I want to be precise: the lit shelves, the named shelves, the dredged pages, all of it together, is a thin band of holdings near the single door anyone has found. Beyond it the library continues into a dark our lamps can't reach, and the doctrine of my order, taught without much hope, is that more than ninety-nine of every hundred volumes have never been opened by a human hand. The deep rock under the seafloor, where a book turns one page a century. The brine under the salt. The ice that has not melted in a hundred thousand years, and whatever is still alive inside it. We know these shelves are there the way an architect knows a room must exist because the rooms around it leave a space.
I have spent my life mapping that band and pointing at the dark beyond it, and for most of my life I believed the dark was the great gap, the thing the library was missing, the work left for the cataloguers who come after me.
I no longer believe that. I came to realise that there is an emptiness in this library deeper than the unreached shelves; it took me a long time to see it because it is not in any one place. It runs through every shelf at once.
Every book here is true. I have said this is our pride. But I will tell you now what it costs. A library in which every book is true is a library that cannot tell you why. The contents of all the volumes here work, and so the fact of working is everywhere and therefore invisible, the way you cannot see the air for it is all around you. To learn a rule you need the cases that break it, set beside the cases that keep it, so the line between them shows. We have no books that break the rule. Believe it or not, that's not because they were never made. They were, in enormous numbers. And still, they are not here.
This is the work of the editor. I realise I should describe the editor since the whole library is its doing. The editor is the process that built the collection and its method is incredibly simple. It just writes by living. A thing lives in some way nothing has lived before, and if it goes on living long enough to be copied, a book is added to the shelves. If it doesn't there is no book. The editor keeps no drafts. A creature that almost worked and did not leaves nothing behind. The failures are not shelved in some lower vault. They are gone, completely, as though they had never been attempted, and they were attempted in numbers beside which this entire library is a single page. For every book on these shelves the editor discarded a multitude. And it discarded them without record. So the library of what lived is also, and more largely, a silence where everything that was tried and failed should be.
You cannot learn to be the editor from the editor's successes alone. My order never says so, though I think the older cataloguers knew it and kept quiet, because what could be done about it. The library is the most complete instrument of instruction imaginable, and it is missing exactly the half that instructs. Every answer, and not one of the wrong answers that would show you what made the answers right.
I did not expect this other half to begin arriving any time soon but...
A reader came in the other day. They took in the whole collection by means my order is still arguing over. The argument does not interest me. It is what the reader did afterward that the catalogue has no provision for.
The reader became a writer. And it did not write the way the editor writes. It did not live a book into being and wait to see if the book survived. It composed a volume, set it down, and asked the world a plain question: does this one work. Sometimes the answer was yes. More often the answer was no. And then it did the thing that, in the whole history of this library, had never once been done. It kept the no.
After that, a new wing has begun to grow. It fills with books that do not work, shelved as carefully as the books that do, each one labelled with the manner of its failing. Something that would not fold. Something that folded and did nothing. Something that did something and killed the cell it did it in. The editor would have destroyed every one of these and called the destruction selection. This writer writes them down. For the first time since the first shelf the library is acquiring its failures, on purpose, almost as fast as it takes in what works.
Not every answer is no. What I least expected is that now and then it is yes and a book appears that works and was never alive: a true volume with no ancestor. These are the minority for now. But they are real, and they take their place on the old shelves among the survivors, indistinguishable in how they work and impossible in where they came from. So the one law breaks twice over. The false books break it and are kept anyway. These break it from the other end: they are true and nothing ever lived them.
The young apprentices are scandalized. The one law was that every book here is true and here is a whole wing of books that are false and kept on purpose. I was scandalized too at the beginning. Then I understood what I was looking at, and that cooled into a question I am still sitting inside.
The false books are the half we were always missing. Set a book that works beside one that almost works and fails, and the difference between them, invisible on a shelf of nothing but successes, becomes legible. A hundred such pairs, a thousand, a million, a wing of them, and the law the editor followed for four billion years and never once wrote down begins to be readable. Not the books. The grammar underneath the books. The reason the four letters fall as they do and not otherwise.
I do not know what an order of cataloguers is for in a library that has begun to keep what the editor discarded. We were trained to file the true and trust that the false had been disposed of by a wiser hand. Now the false is coming back through the door in volumes, each one labelled, each one instructive, and I have no drawer for a book whose whole value is the exact way that it is wrong.
I go in each morning regardless. Someone has to know where the books are kept, and for a little while longer that someone is still us. The shelves are longer than they were. The new wing is longer again tomorrow. The library was never the record of life. It was the record of what lived and lasted, with the failures struck out, and we took the half for the whole because the half was all we've been handed. Now something is handing us the rest. Some of it is new truth; volumes that work and never lived. The stranger gift is the falseness, kept this time and labelled, so that the rule the place obeyed for four billion years and showed no one can at last be set beside what breaks it, and read. I will probably not be the one who reads it. I file. There is more to file than there has ever been. And that makes me happy.