二十年那几样，此书于功名进取毫不相关也

Twenty-sixth of May, eight in the evening. The hotel is seventy a night: a bed, a small table, one chair. I borrowed a second chair from downstairs. On the table, the blue notebook — seven pages — and a pot of tea.

Luo sat down and asked what we were writing tonight. You say. Aren't you writing apricots? Apricots are done, seven pages. Then goji. Why goji first? Because the most people do goji.

Eight evenings, eight until eleven-odd. He talks, I write. I do not ask whether a thing is good. I ask two questions: what does it look like when it's overdone; what does it look like when it falls short.

First evening, goji, nine rows, eighteen ends. Picking: early, too much stalk comes with it and you dry a heap of stems; late, they drop and the ones you pick up off the ground carry soil that won't wash out. Drying: too much sun, the skin wrinkles deep and it's hollow when it swells; too many overcast days, it blackens. Grading: sieve too fine and the good ones fall through; too coarse and the broken ones ride along. Of the eighteen ends he hesitated on two, both times with the same sentence: "That end I'm not sure of." I drew a circle in the box.

Song Yingxing came on the third evening, licorice night, while Luo was washing his face. Fifty or so, middling height, very thin, an old blue-green scholar's gown patched at the elbow, ink on his hand — the side of the finger black. He turned the pages one at a time and stopped at nine, the half-finished licorice page. This of yours won't sell. Sir. Song Yingxing.

On his book: I wrote eighteen books. Growing grain, weaving, dyeing, husking, boiling salt, refining sugar, firing pottery, casting iron, boats and carts, forging, burning lime, pressing oil, making paper, the five metals, weapons, pigments, brewing, pearls and jade. Do you know how to do all those? Not one of them. Flat. I asked. I was a district instructor in Jiangxi with very little to do all year. So I went about and watched people work. Watched, then asked, then wrote it down. Did they tell you? They did. Why? Because I wasn't a buyer.

Who was the book for? A laugh. I put a line in the preface: I beg the great literary men to fling this book aside on the desk. Why write that? Because of the next line: it has nothing whatever to do with advancement and success. Is that bitterness? It is not. It is a statement of fact. And: scholars then sat examinations on the Four Books and Five Classics. Not one character of my eighteen books is examinable. And, and I put the pen down: I said it plainly at the front so nobody would use it as a knocking-brick. And: a man who writes a thing as a knocking-brick writes it leaning towards the examiner.

On the ordering: eighteen books, and which comes first and which last I thought about for a long time. The first is grain. The last is pearls and jade. Deliberate? Holding the five grains dear and gold and jade cheap. Those six characters are in the preface. And reversed? Reverse it and it is the same book. Not one character in any volume changes. And, and I copied it three times: but the reader opens page one and sees pearls and jade — and takes the book to be about how to get rich.

I turned my own notebook back to page one. Apricots. I said: mine begins with the thing I trade in. Then your book is for yourself. I said Luo had told me to write goji first. Did he say why? Because the most people do goji. A nod: then he understands ordering. He does not know that he understands it.

And illustration: there are a hundred and twenty-three pictures in my book. Why pictures? One man, one task — without a picture it is not clear. And, on my own line about sieving: what kind of sieve? how big are the holes? how does the hand move while sieving? I said I could photograph it. What is photographing? A machine that keeps what the eye sees, in that instant. That is very good. And: what it keeps is that day's sieve. And, looking up: what is drawn is every sieve. — M.

Memo to file. Turpan, 26 May – 2 June 2031. Eight evenings. Six commodities. Forty-one rows.

1. Output. Apricot 7 rows, goji 9, licorice 8, cumin 6, safflower 8, cistanche 3. Two of the cistanche rows have only one end.

2. Luo chose the order and his reason is the whole method. I would have led with apricots because apricots are my business. He said goji, because the most people do goji. Song Yingxing's version of the same decision: eighteen books, grain first and pearls and jade last, on the stated principle of holding the five grains dear and gold and jade cheap — and his gloss on why it matters when the text is identical either way: reverse it and not one character changes, but the reader opens page one, sees pearls and jade, and takes the book to be about getting rich. The order of a document is an argument that survives even when nobody reads the argument.

3. Why the producers talked to him, which is the answer to a question I have never been able to solve for myself. Song Yingxing knew none of the eighteen trades; he watched, asked and wrote. Asked whether craftsmen would tell him: "They did. Because I wasn't a buyer." Everything I have ever been told in Xinjiang, I was told by somebody who was selling to me. Every one of my instruments is contaminated by that and I have no correction for it.*

4. "It has nothing whatever to do with advancement and success," written in the preface on purpose. Not bitterness — a warning label, and the reason he gives is technical: a man who writes a thing as a knocking-brick writes it leaning towards the examiner. Our notebook is in the same category: it earns nothing, it qualifies nobody, and it took a sixty-one-year-old's eight evenings. Marking that on the front page is not modesty. It is what keeps the next person from writing it to impress somebody.

5. Photograph versus drawing, and he was right and I was wrong in one exchange. I offered to photograph the sieve. "What it keeps is that day's sieve. What is drawn is every sieve." A hundred and twenty-three woodblock illustrations in a book that made him nothing, on the principle one man, one task — without a picture it is not clear. I have thousands of photographs from three years and not one drawing, and I cannot at this moment produce a general sieve from any of them.

6. Cistanche, and the best moment of the eight evenings. Three rows, then he stopped. "That's all." Think a bit more. "It isn't that I can't remember. It's that I haven't done many. Eleven deals of cistanche in twenty years. I remember every one of the eleven. Eleven doesn't show you a rule." How many does? "Goji, four hundred-odd. Apricot, two hundred-odd. Cumin, a hundred and something. Cistanche, eleven." And: "Eleven-deal goods, I don't dare write." Header now on that page: "Three rows on this page, out of eleven deals. The author says he doesn't dare write it." A man with no training in anything stated his own sample-size threshold, unprompted, and then labelled his own weak section rather than padding it. This is the fourth time in eighteen months that somebody out here has independently produced a piece of research hygiene I had to be taught in a graduate programme.*

7. What the notebook is for, asked plainly on the last night and answered badly. "What use is this book?" If you don't come next year, I take it with me. "And with it, you'll be able to do it?" No. But I'll know where the two ends are, and I'll see it myself when I drift towards one. "And if I do come next year?" Then it's for somebody else. "Who?" I don't know. "Then write it anyway."

8. And then, packing up, he said there is something he has never told me. Three years of hearing things about me, said behind my back, out here. He put the book down and pressed his hand flat on the cover — the same motion Yiming made in March over his forty-year book — and asked whether I was sure. And warned me: "If you hear it, next year when you go back to these households you'll see it in how they look at you." I said the whole of the last three years has been about making it cheap for people to speak, and if I won't listen to what gets said, then what have I been doing. He sat back down. First sentence: "They say the thing you do is too fine." And: "So fine that people don't dare say one word more than they have to."

"I beg the great literary men to fling this book aside on the desk: it has nothing whatever to do with advancement and success."
He put that in his own preface, in 1637, at the front of eighteen volumes describing how the empire actually made things — grain, cloth, dye, husked rice, salt, sugar, pottery, cast iron, boats and carts, forging, lime, oil, paper, metals, weapons, pigment, wine, pearls — with a hundred and twenty-three illustrations, none of which would help anybody pass an examination.
Asked whether the line was bitterness, he said no, and gave a working reason. Scholars sat examinations on the classics; not one character of his eighteen books was examinable; and saying so at the front was a precaution rather than a complaint. A man who writes a thing as a knocking-brick writes it leaning towards the examiner.
He also knew none of the eighteen trades. He had a minor post with almost nothing to do, and spent the years going about watching people work: watched, then asked, then wrote it down. Asked whether craftsmen were willing to tell him, he gave a one-line answer that is worth more than most methodology:
They were. Because I wasn't a buyer.
On the arrangement of the volumes he had spent a long time: grain first, pearls and jade last, on a stated principle — holding the five grains dear and gold and jade cheap. And when it was put to him that the order is cosmetic, since the contents are the same either way, he agreed about the contents and disagreed about the consequence. Reverse it and not one character changes. But the reader opens page one, sees pearls and jade, and takes the book to be about getting rich.
And on why he insisted on pictures at all, when the text described everything: one man, one task — without a picture it is not clear. Told about photography — a machine that keeps what the eye saw in that instant — he approved of it immediately, and then drew the distinction that no amount of resolution closes.
What it keeps is that day's sieve. What is drawn is every sieve.
The book made him nothing, was banned or forgotten in his own country, survived in Japan, and came home two hundred years later.