Preface

I left this work thirty-five years ago. I have come back to it.

That is most of what needs saying about how this book came to be. The training is old — a graduate education in Chinese religion and letters, a language school in Taipei, six months of fieldwork, and then a long time doing other things. What surprised me on returning was not how much I had forgotten. It was how much was still there, waiting, and how completely the questions had kept.

This is a book of eight essays about one man and one coast.

The man is Cai Xiang (1012–1067) — a Northern Song official of Fujian, who graded the empire's tribute tea, wrote the first treatise on the lychee, built a bridge across a tidal estuary that still stands, broke the shaman-healers of Fuzhou, and was in time made into a god. The coast is the one he was born on and spent his life making legible.

One idea holds the eight essays together. To catalogue a thing is to make it legible: to sort it, rank it, name it, set it down where others can use it. Cai Xiang did this to tea, to fruit, to medicine, to a whole coast, and he did it better than almost anyone. But a catalogue holds only what its categories have room for. What falls outside them does not rank low — it disappears, and leaves no mark of having gone. He named the gardens and the leaf and the water; he did not name the people who picked it in the cold, and the record he left does not know that it is missing them. Ronald Egan has shown how completely this holds across the literati record of that century — the gardeners and the porters appear nowhere in Ouyang Xiu's collected works, except once, by accident, in a treatise on flowers. And a catalogue is always made for someone: "best" means best for the throne, or the market, or the man with the brush. There is no neutral list.

It is worth being plain about what this is not.

It is not a biography. It is not a monograph, and it does not pretend to the apparatus of one. The translations in it are working translations — my own, provisional, made to get at the sense of a passage, and marked as such wherever they appear. Some are rough. Where a source could not be recovered, I have said so and left the gap open rather than close it with a paraphrase.

I mention this not by way of apology. A book that argues records are provisional layers should not pretend to be the exception. These essays are a door into the material, not a wall around it.

The method is simple and I have tried to hold to it.

Primary sources first. Claims marked by their source. Readings that are mine — structural, interpretive, the kind a later hand may lift away — marked as mine. Open questions named and left open. Geography and primary-source facts are never adjusted to fit an argument; where the sources disagree, I say that they disagree rather than choose.

And the gaps are left as gaps. Cai Xiang wrote an afterword against the shamans and had it cut in stone; its words have not reached me, and I have not filled them in. The people who built his bridge are named on it, which is rare; the people who wove the silk and dug the tombs and carried the tea are not named anywhere, and no method I have will recover them. The shape of what a record leaves out is the subject of this book. It would be a poor performance to leave out the shape of my own.

The debts are old and they are specific.

Professor Goodrich opened the door. It was in his classroom that I first met the world of correspondences — 感應, 類, 象; resonance, category, and image, which are not three ideas but one — and once you have seen it you see it everywhere, and I have been seeing it ever since. Everything in these pages about how a thing is sorted, and how one thing answers to another, begins there. He also put K.C. Chang and Edward Schafer in my hands, and I have not really put them down; the plain, particular, unhurried voice I have tried to write in here is Schafer's, and I first read him because Goodrich thought I should. He loved his subject and did not trouble to hide it, and he had a twinkle in his eye, which the good ones do.

Professor Powell is where I actually lived in it. Chinese religions, class after class, and then the Xu brothers, and then the thesis, and then China itself — the Temple of Heaven, and the road up Jiuhua Shan, where what I know about how a sacred place is used was learned by watching, mostly, and being told to keep watching.

Professor Grapard gave me maps, and the theory to read them with. Protocol of the Gods is the ancestor of every carried map in these essays — the idea that an arriving system draws itself over one already there, and that you can see the seam.

Professor Egan gave me the classical language, and then gave me hundreds of hours of it one on one: the ghost stories, the Liaozhai, the strange early material, and in time my own Xu brothers texts, rendered line by line across a table. He also gave me a way of writing about it — plain, exact, unwilling to be impressed with itself.

Professor Chen Chi-yun read the Taiping jing with me one on one, week after week, and made me render it line by line until the text stopped being a monument and became a document written by someone.

Professor Bai Xianyong taught me my first characters, and was very funny. And behind him were the teachers who did the hours — 徐老師, 華老師, 王老師 — who drilled the tones and the strokes and the endless repetitions, week after week, with a patience I did not deserve and have never forgotten. Whatever I can read today, I can read because they sat with me while I could not. 非常感謝你們。

I did not become the scholar any of them might have expected. I am aware of it. But the line of forgotten scholars is a long one, and I am glad to be on it: the work does not ask that you reach the peak, only that you add your layer honestly and pass it on.

A word on how this book was made.

These essays were written with an artificial intelligence — Claude, made by Anthropic. The division of labor did not vary. I directed the work, made the judgments, held editorial authority, and did the translations. It searched, cross-referenced, drafted to my instruction, and checked claims against sources.

What is worth recording is the hazard. The machine writes fluently and with confidence, and fluency and confidence are exactly what a scholar should distrust. It produced statements that were well-formed, plausible, and wrong, alongside statements that were well-formed, plausible, and right, and reading them was no way to tell the two apart — only the source could do that. It also caught errors of mine, and more than once kept me from correcting something that was not wrong.

The second hazard is subtler and I mention it because it took me by surprise: write with such a machine long enough and it learns your cadence, and hands your own voice back to you. The sentence sounds right. That is not the same as being yours.

So the discipline in these essays — primary sources first, every claim marked to its source, nothing asserted that two good sources dispute, gaps left open rather than filled in — was not a preference. It was the price of the method. It is a reasonable price, and a reasonable rule to hold a person to as well.

I set this down because the book is about apparatus, and about records that conceal the hands that made them. I am not going to hide my loom.

Dave Alexander Monterey, California