At a book launch a few years ago someone in the audience asked the author the question: who did you write it for? She answered immediately; she wrote it for herself. The room applauded, warmly, and she was right to say it, because it was true and because the culture around books has trained all of us to admire that answer above the available alternatives. After all, should we be writing books for anyone but ourselves?

I applauded too, and then drove home mildly irritated, and it took me most of the R27, all the way past the pseudo-brutalist pyramids of Dolphin Beach Hotel, to work out why. If the answer is true without remainder – if the book really was written for its writer and for nobody else – then forty people had just surrendered a Thursday evening to celebrate a private act, a bookseller had cleared a table for it, and potentially years of other people’s work had been a courtesy extended to a woman who did not require it. Everyone in that room except the author was excess weight.

That is, of course, not what she meant. But it is the right question, and almost never answered seriously.

My September essay, On Authority, is about where an author’s authority ends and a publisher’s begins. That essay will be about jurisdiction: who decides, and on what grounds. This one is about purpose, which is the prior problem. Before you can settle who gets to decide, it is useful to know what the whole arrangement is for.

The arrangement, this nebulous thing known as a book, or a story, or a narrative, or a project, or a piece, has three parties in it. The writer, the publisher, and the reader. Each is necessary, and none is sufficient on their own (if you are a self-published author, you occupy the role of writer and publisher – one cannot even self-publish a book without the involvement of the pesky, pernicious publisher. Even if it is you). And they are never in the same room at the moment that matters, which is why the relationship between them is so persistently badly described – usually as a hierarchy, occasionally as a conflict, and hardly ever as the thing it actually is: a good ol’ ménage à trois.

The most useful thinking I have found on arrangements of this kind does not come from publishing. It comes from a tradition that transmitted an enormous body of text for four and a half centuries without producing a single manuscript.

The chariot

I’ve been promising a Buddhist tangent for months now. That promise is made manifest today, dear reader! Come on a journey with me.

In the Milindapañha, the Bactrian Greek king Menander (Milinda in the Pali) puts a series of hard questions to the monk Nāgasena, who answers them with the simile of the chariot[1].

The King arrived by chariot. “But what is a chariot?” Nāgasena asked. “Is it the wheels, or the axles, or the reigns, or the frame, or the seat, or the draught pole? Is it a combination of those elements? Or is it found outside those elements?”

The King answered no to each question. “Then there is no chariot!” Nāgasena said.

Now the King acknowledged the designation “chariot” depended on these constituent parts, but that “chariot” itself is a concept, or a mere name.

Now do it with a book. Is the book the manuscript? The file that will one day be Khala, the novel I’m writing, exists, and existing (currently) is the only thing it does; a manuscript is a book the way a score is a symphony – it is, and it isn’t. Is the book the printed object, then? A pallet of unsold copies is not literature. It is stock, and it depreciates. Is it the reading? A reading with nothing to read is a person sitting looking at ink blots pressed onto smushed tree pulp.

The book is none of the three and does not exist without all three. It is not an object at all. It is a relation, and it is live only for the few hours a stranger spends inside it. Everything the industry does is either preparation for those hours or clearing up afterwards.

Which produces a conclusion that is unflattering to everyone, so I will put it plainly: nobody owns a book. You cannot own a relation. You can only hold up your end of one.

Four hundred and fifty-four years

The Buddha wrote nothing down. By Theravāda tradition the Pali Canon was not committed to writing until 29 BCE, four hundred and fifty-four years after his death[2], when some five hundred monks gathered at the Aluvihare cave temple in Matale, in what is now Sri Lanka, and scratched it into palm leaves with metal styluses.

The four and a half centuries before that are the interesting part. The monks were organised into groups, each responsible for memorising a specific portion of the canon. Communal recitation worked as error correction: when one reciter drifted, the group caught it. And the texts themselves were engineered for the task – the relentless repetition that modern readers find tedious in the suttas is not weak writing, it is redundancy built into the medium, a memory technology with checksums. It was a distribution system, and a formidably robust one.

What eventually forced the writing was risk. Famine and war under Vaṭṭagāmaṇi threatened the reciters themselves, and if you kill enough monks you lose the text. The manuscript of the Pali Canon was the contingency plan for the failure of the mechanism.

Buddhism takes refuge in three things: Buddha, dharma, and sangha – the teacher, the teaching, and the community that carries it. The three are taken together and nobody ranks them, because the first two do not survive the loss of the third. A teacher without a community cannot teach. A teaching without a community is a private experience with a very short half-life.

I am not going to claim that a small press in Cape Town is a monastic order; our hours are worse and our vows are commercial. I am claiming that the structural problem is identical, and that it was solved once, at scale, by people who understood that transmission is not a service appended to the work after the work is finished. It is a constituent of the work.

What each party actually holds

The writer holds the intention. Only the writer knows what the book is trying to do – not what it is about, which anybody can work out, but what it is for, which no editor can supply and no reader can recover if it was never there in the first place. This is the one thing in the arrangement that cannot be outsourced, and it is the reason a writer’s no, when it is a real no, has to be final.

The publisher holds the distance. We know what lies between a finished manuscript and a stranger’s hands, because measuring that distance is the whole of the job: the price, the shelf, the metadata, the bookseller with ninety seconds and forty other titles to place, the fact that a book which breaks even at 821 copies is being printed in tranches of three hundred. The writer is not obliged to find any of this interesting. The writer is obliged to accept that somebody has to.

The reader holds the meaning, and holds it completely. They will not be consulting us about it. They will misread the ending, adore the minor character we nearly cut, and quote a line that the author has no recollection of writing. Note, too, what the reader does not hold: any obligation whatsoever. They signed nothing. They may abandon the book on page thirty and owe no one an explanation. Two parties to this arrangement carry duties. The third carries none, and the entire apparatus exists for the third. That asymmetry is not a defect in the design. It is the design.

Skilful means

Which brings us to compromise, and to who makes it. In the Lotus Sutra a wealthy man comes home to find his house on fire and his children playing inside, too absorbed to grasp what fire is or to heed him when he calls. So he tells them there are carts waiting outside. The children run for the carts, and live. The Buddha then asks whether the father lied. This is the parable of the burning house, and the doctrine it dramatises is upāya: skilful means, the teaching shaped to the capacity of the person being taught.

It is also, unmistakably, a description of a marketing department. The cover the author did not want and can now no longer imagine the book without. The blurb that flattens a hundred and forty thousand words into a promise about a stolen child and a drowned god. The comparison title that is not quite right but is close enough to place the book in a stranger’s mind in the four seconds available. The tag that says ‘for fans of’. These are carts. They are not the book and they are not lies about the book. They are the devices by which somebody who cannot yet know what the book is crosses the threshold to find out.

And upāya has a hard limit, which is intention. The father is exonerated because the children get what they were promised and better. For us, if the book inside is not the book on the cover then it is not skilful means, it is fraud.

So who compromises? Not the reader, who cannot. A reader can only decline. Every compromise in publishing is therefore struck between two parties on behalf of an absent third whom neither of them has met, which is an unnerving way to run anything and accounts for most of our disagreements. The discipline I try to hold to is this: argue about the mechanism, not about the reader’s taste. I do not know what the amorphous, profoundly ambiguous though oft-invoked “South African reader” wants. I do know what happens when a book’s spine cannot be read from a metre away, when a title cannot be found by somebody who half-remembers it, when a first chapter dumps twenty proper nouns within two pages. These are specific things that are testable, verifiable, and solvable. “Readers won’t like it” is a claim about a person who does not exist, that means nothing.

What the whole thing is for

If the writing were the point, the writing would be the end of it. It isn’t, and every writer knows it isn’t, which is why writers want to have their work published. Wanting to be read is not vanity appended to the work. I feel tremendously strongly about this (enough to use two adverbs in a row): To be read is the work’s telos, its ultimate end or purpose, and it is a worthy one. A novel is a very elaborate machine for putting something from one person’s head into the head of somebody they will never meet. Publishing is the mechanism of achieving the work’s telos.

This is worth stating plainly, because publishing measures everything except the event it exists to produce. We measure unit sales and royalties, stock turn and profit margins, trends and reach, a cornucopia of data spilling forth from sales reports and industry surveys and networking events. Not one of those things can see a person reading. We measure the last observable thing before the event and then call it the event, because the event itself takes place in a room we are not in, generates no data, and pays nobody. The event, reading a book, remains a stubbornly arcane activity that, at least for now, largely resists the insidious optic fibre tendrils of data analysis and Big Tech.

So: who is a book for? Not the writer, who is finished with it – whatever she says at the launch, she wrote it to be read, or she would never have pursued its publication. Not the publisher, who never owned it and whose entire function is to hand it over. It is for the person who was not in the room: the one the apparatus was built to reach, who owes us nothing, and who will decide in roughly fifteen seconds whether any of it worked.

So perhaps the answer is this: the Venn diagram of who a book belongs to, who it is for, is a circle, for better or for worse.

Here’s what I’ve been watching, listening to, and conspicuously not reading.

Series

Prime Video dumped all eight episodes of Ride or Die on us in July and we got through them in three sittings. Octavia Spencer and Hannah Waddingham as lifelong best friends, one of whom has been an international assassin for twenty years without ever mentioning it. The premise is doing nothing new whatsoever and it does not matter in the slightest, because the two of them are so palpably enjoying themselves that it becomes contagious. Five stars, and I will hear nothing against it.

For fans of: Killing Eve, women of a certain age with excellent knitwear and appalling secrets, Ted Lasso withdrawal

I mentioned a few issues ago that I was revisiting The Rings of Power with the craft brain switched on. I have now done both seasons properly, and I am going to plant my flag: five stars. Yes, the timeline is compressed to the point of violence. Yes, the purists are furious. But the second season’s Annatar plot is a marvellous, and infuriating, piece of dramatic irony – an entire civilisation being charmed by a man the audience knows is the devil – and Númenor remains one of the most convincing invented cities I have seen rendered in years. It is doing, at enormous expense, precisely what this issue’s essay is about: carrying something from one mind to a great many others, and accepting some losses in the carriage.

For fans of: Tolkien, spending nine figures on a hairstyle, watching a slow-motion car crash where everyone in the car is beautiful

Books

A reading moratorium this issue, and not a voluntary one. Submissions for the MENAGERIE anthology have been arriving in volume, and we are in final preparation on the last seven books Mirari publishes this year. My reading has therefore been reduced entirely to other people’s unpublished work, which is a genuine pleasure and is also, structurally, not reading. Normal service will resume next month with Naomi Ishiguro’s Rainshadow Orphans.

Movies

The MCU rewatch has finally deposited me at Avengers: Endgame, and I loved it without reservation. I cried three times (as usual). Whatever you make of the eleven years of setup, landing an ending of that size is a feat of structural engineering, and I say that as a man who spends his working life telling other people their endings do not work. Overwhelmed, entirely satisfied, and ready for the next phase.

Podcast

A new category, on the grounds that I have been in the car a great deal. Stephen Fry’s 7 Deadly Sins takes the seven in turn – pride, avarice (or greed), lust, envy, gluttony, anger (wrath), and sloth (PALE GAS, as per Mister Fry) – roughly half an hour each, researched, written and read by Fry himself. He handles them not as theology but as descriptions of ordinary behaviour, which is the only way they become interesting again after two thousand years of finger-wagging. An absolute masterpiece. It also happens to be a masterclass in the thing the essay above is preoccupied with: a man who knows precisely what he is trying to put into your head, and has thought carefully about the vehicle.

Art

This essay’s ‘unrelated art that moves me’ is Vilhelm Hammershøi's Interior, Strandgade 30 – one of many versions of the same room, this one painted in 1901 and hanging in the Städel Museum, Frankfurt am Main. Hammershøi painted the rooms of his Copenhagen apartment for eleven years, usually with his wife Ida somewhere in them, almost always with her back turned. Nothing happens. Nothing is explained. What holds me is how completely the picture hands the work over. He tells you nothing, so whatever the room means, you brought it in with you.

[1]The simile is older than the Milindapañha, which sits just outside the Pali Canon proper. It appears in the Vajirā Sutta, spoken by the nun Vajirā in the Buddha’s presence. Nāgasena credits her with it.

[2]The 454-year figure is the tradition’s own arithmetic, counted from a parinibbāna the Theravāda places in 483 BCE. Modern scholarship is not settled on the Buddha’s dates – most estimates fall somewhere between 480 and 400 BCE – so read ‘four and a half centuries’ as indicative, not prescriptive. The order of magnitude is not in dispute.