The Tso Chuan — The Manual Is Downstream of the Deed

The Cookbook Written After the Meal: Master Metaphor

Picture a cook who feeds a village for years on instinct — a pinch here, a hot pan there, dishes everyone loves and nobody has written down. Long after he is gone, a scholar tastes the surviving recipes, reverse-engineers them, and publishes a famous cookbook. Everyone learns to cook from the book. They cite the book. They argue about the book. And almost nobody remembers that the book is a transcript of meals that were cooked first, by people who never read a recipe in their lives.

That is the argument the Ch'ing scholar Li Yüan-ch'un made about Chinese military thought, and it is the heart of this page. The Tso Chuan — an old chronicle long misread as a mere commentary on the Spring and Autumn Annals — was rediscovered as the true progenitor of classical military theory.¹ The famous manuals, the Art of War and the Wu-tzu, came later. The doctrine was abstracted from battles already fought. Practice precedes theory. The cookbook is downstream of the meal.¹

What This Actually Is

For centuries the Tso Chuan was treated as a worthless repository of outdated tactics by Western-oriented critics during the Ming and Ch'ing.¹ Then, in an irony Sawyer enjoys, the same period saw it become the focus of intense interest precisely as the foundation of Chinese military science.¹

A line of scholars built the case. In the Ming, Ts'eng Yi wrote Tso Strategies (Tso Lüeh). Ch'en Yü-mo compiled the comparative Tso-shih Ping-lüeh (Mr. Tso's Military Strategies), packed with cross-text commentary. Wei Hsi wrote the Ping Mou (Military Plans), dissecting thirty-two essential categories of mou — strategic plans — found inside the Tso Chuan.¹ And in the Ch'ing, centrally for this page, Li Yüan-ch'un compiled the Tso-shih Ping-fa (Mr. Tso's Art of War), which abstracts and comments on nearly two hundred military incidents drawn from the Tso.¹ That work was revived in recent decades in a People's Republic vernacular edition, Shallow Explanations of Tso's Art of War, putting the old argument in front of a broad modern audience.¹

All these scholars share one assumption: the origins and the very foundation of Chinese military science lie in these early battlefield tactics, and the lessons carry a timeless validity that can be studied and reused.¹

The Empty Words and the Attested Deed (The Internal Logic)

Li Yüan-ch'un's claim is sharp enough to quote against itself. Spring and Autumn commanders, he argues, used unorthodox (ch'i) tactics throughout the period — and since those events all preceded the historical Sun-tzu and Wu-tzu, the famous theoretical writings were in fact based upon the lived experience the Tso Chuan preserves.¹

Then comes the line that does the work: "their discussions are merely empty words, but what the Tso speaks about has been attested in experience."¹ Sun-tzu and Wu-tzu are abstractions. The Tso is the record of things that actually happened to actual armies. Li goes further — those who later excelled at war "knew that their tactics came out of Sun-tzu and Wu-tzu, but none realized they actually originated in the Tso."¹ The cookbook got the credit; the cooks got forgotten.

His evidence is concrete, and it runs on a single move: for every celebrated "first," find the earlier deed in the Tso. The military esteems the unorthodox, Li says, and among unorthodox measures none is more famous than T'ien Tan's fire oxen — yet the incendiary elephants in the Tso preceded it.¹ Nothing was more unorthodox than Tsung Ch'üeh fabricating the appearance of lions, or Ti Ch'ing's bronze mask — yet "cloaking horses precedes them in the Tso."¹ The pattern is relentless: whatever the manuals are praised for, the chronicle did it first.

The Essay and the Battle (Synergies & Handshakes)

This page is one of the two rival origin claims in the corpus, and it cuts against its sibling.

• The T'ai Kung — Unorthodox Progenitor — the rival story. The T'ai Kung legend gives the unorthodox a single founding hero; the Tso gives it no founder at all, only accumulated practice. One descends from a man, the other rises from many anonymous battlefields.
• The Zheng/Qi Duality — the doctrine Li says was abstracted from the Tso's battles rather than invented by Sun-tzu. This page is the claim that the grammar came after the speaking.
• Animal Delivery Systems in Warfare — Li's own lead example: the incendiary elephants of 506 BCE that prefigure T'ien Tan's fire oxen.
• Chinese Military Theory — Historical Development — the long arc in which the practice-before-theory claim is a structural argument about how the whole tradition formed.

Analytical Case Study: The Elephants Before the Oxen (506 BCE)

Li's cleanest example is a chain of two fires forty years apart, and it shows his method working.

T'ien Tan's fire oxen are among the most famous unorthodox measures in all of Chinese military history — a desperate, carefully calculated stroke to break the long siege of the city of Chi-mo, oxen with blades and torches stampeded into the besiegers. Everyone knows the oxen.¹

Li's point is that the oxen were not first. In 506 BCE, the king of Ch'u faced Wu's onrushing forces, which had won several successive victories and penetrated all the way to the Ch'u capital of Ying. Immediately after crossing the Sui River, the fleeing king stampeded several elephants into his pursuers — whether by tying firebrands to their tails or using torches to enrage them — disrupting them enough to escape.¹ Same logic, same animal-as-weapon, same desperation-engineered-into-surprise, four decades before the celebrated oxen.

The case is the argument in miniature. The manuals get the glory for the unorthodox; the Tso records the deed that came first. And the historical detail matters: the Tso preserves how — the river crossing, the choice between firebrands and torches, the specific tactical moment of flight — in a way the abstract manuals never bother to.

Implementation Workflow: Reading Before Making

You want to do something genuinely new — a campaign, a piece of work, a move nobody has seen. The instinct is to clear your head and invent from scratch, to be original by refusing to study.

Li tells you the opposite, and he uses the image of writing an essay. You sit first with the antique works. You read them closely, not to copy them but to soak in how victory was actually calculated. You feel the means by which old battles were won, the specific texture of how surprise was built. Then — and this is the turn — you set the visible plots aside. You do not imitate the surface of the ancient stratagems; you "create unorthodox measures yourself."¹ The deep reading is not the enemy of originality. It is its precondition. "Without reading the ancient chapters and pouring over the ancient military books, it would be impossible."¹

And there is a residue Li leaves deliberately unexplained. Some of what works cannot be reduced to a rule at all — "what is not explainable is of course shen ch'i," the numinously or spiritually unorthodox.¹ So the workflow ends in honesty: read everything, imitate nothing, make something new — and accept that the best part of it will not be teachable, only enacted.

The Progenitor Failure (Diagnostic Signs)

The practice-before-theory insight fails when you forget which way the arrow points. The tells:

• You treat the manual as the source instead of the transcript. If you think the doctrine generated the battles, you will look for cleverness in the book that was never there — it was in the field, and the book only wrote it down afterward. Li's warning: those who excel "knew their tactics came from Sun-tzu and Wu-tzu, but none realized they originated in the Tso."¹
• You imitate the visible plot. Reading the ancients to copy their surface produces stale tactics that an enemy has already read in the same book. The point of the reading is to absorb the means of victory and then make something the book does not contain.
• You demand that everything be explainable. If you reject any measure that cannot be reduced to a stated rule, you have cut out shen ch'i — the unorthodox that works precisely because it cannot be codified. The manual cannot hold it, and a commander who only trusts the manual will never reach for it.

Evidence / Tensions / Open Questions

The crucial caveat is the one that complicates Li's whole edifice, and Sawyer flags it directly: the Tso Chuan itself never labels anything "unorthodox." The character ch'i appears in the Tso only inside personal names. The "unorthodox" tag was applied retroactively, from the T'ang dynasty onward, by later readers looking back through Sun-tzu's vocabulary at battles that did not describe themselves that way.¹

That is a real tension and it should not be smoothed over. Li says the Tso is the progenitor of the unorthodox; but the Tso never knew the word. So either the unorthodox was being done before it had a name — practice genuinely preceding theory, which is Li's strong claim — or the later tradition is reading its own categories backward into events that meant something else at the time. Both readings survive the evidence, and Sawyer keeps them both alive rather than picking one.

A bias flag: Li Yüan-ch'un has a stake. His whole project is to elevate the Tso against the Seven Military Classics, and a man arguing that his preferred text is the true source will naturally find precedents for everything in it. The retroactive-labeling problem is exactly the kind of thing a motivated advocate would gloss. Sawyer's own thesis — the depth and distinctiveness of a Chinese art of the unorthodox — is served by a deep, practice-rooted lineage, so the convergence of Li's motivation and Sawyer's runs in the same direction and deserves a skeptic's eye. The honest middle position: the deeds are attested even where the label is anachronistic, so practice-before-theory holds for the tactics even if "the unorthodox" as a self-conscious category really does begin with Sun-tzu.

Author Tensions & Convergences

Only Sawyer is cited, but the page lives on the friction between Li Yüan-ch'un's confident claim and Sawyer's textual caution, and Sawyer stages the friction deliberately. Li is an advocate, writing a preface to crown his text: the Tso is the progenitor, the manuals are empty words, the deeds were done before the doctrines were dreamed. Sawyer is a historian who admires the argument and then quietly notes that the Tso never once says ch'i outside a name. The convergence is that Sawyer does not use the labeling problem to dismiss Li — he uses it to sharpen the claim. The deeds are real and early; what is late is the category. That distinction is the page's whole yield, and neither man states it cleanly alone: Li overclaims (the Tso is the source of the unorthodox), Sawyer undercuts (the Tso never named it), and the truth sits in the gap — practice precedes theory at the level of the deed, while the self-aware concept of the unorthodox is genuinely Sun-tzu's invention, later cast backward over battles that were fought without it.

Cross-Domain Handshakes

The practice-before-theory pattern is not confined to ancient Chinese chronicles. It shows up wherever a living skill gets frozen into a codified system, and two outside connections sharpen what the history alone cannot.

• Cross-domain (Japanese strategy): In-Yo as Japanese Grammar — Lovret treats in and yo (yin and yang) as a grammar of combat, a structural language abstracted from generations of swordwork rather than invented at a desk. The precise parallel: a combat grammar is to lived sword practice exactly as the Art of War is to the battles in the Tso — the rules are a distillation of the doing, not the source of it. The insight neither domain gives alone: Li's claim that "their discussions are merely empty words, but what the Tso speaks about has been attested in experience" is the general law of which the in-yo grammar is one instance. Any combat doctrine — Chinese unorthodox, Japanese in-yo — is a transcript of accumulated practice that later students mistake for the practice's parent. Seeing in-yo as a grammar rather than a theory lets you read every military classic the same way: as a language people spoke fluently before anyone wrote down its rules.
• Behavioral-mechanics (codified tactics): Deception and Misdirection Tactics — a modern misdirection manual is a catalogue of named, repeatable moves, exactly as the Tso-shih Ping-fa abstracts nearly two hundred incidents into studiable lessons. The parallel is the tension between the codified tactic and the living deed it was distilled from. The insight neither domain gives alone: Li's warning that you must read the ancients and then "without imitating the visible aspects of ancient plots, create unorthodox measures yourself" is the answer to the central failure mode of any tactics catalogue — that a tactic, once written down and circulated, is read by your target in the same book and stops working. The Tso understood, two millennia before influence manuals, that the visible plot decays the moment it is named, and that shen ch'i — the inexplicable, uncodifiable stroke — is the only kind that survives publication. A misdirection handbook teaches you the moves; the Tso teaches you why the moves on the page are already half-dead, and why the live edge is always the thing the book cannot hold.

Hold the two handshakes together and the history page reaches something it could not alone: every doctrine — sword grammar, military classic, influence manual — is a transcript of practice, and every transcript has the same fatal property, that the moment a tactic is legible it is shared, and the moment it is shared it is anticipated. Shen ch'i, the uncodifiable residue, is not mysticism. It is the structural remainder left over after everything teachable has been written down and therefore neutralized.

The Live Edge

The Sharpest Implication If the doctrine is downstream of the deed, then studying the manual will never make you the equal of the people the manual was written about. They were doing the unorthodox before it had a name, in conditions the book strips out. Li's "empty words" cut at the heart of every how-to tradition: the codified version is always a lossy compression of a living practice, and the part that gets lost — shen ch'i, the inexplicable stroke — is exactly the part that wins. The implication is bracing for anyone who learns from books: the goal of reading the ancients is not to reach their level by absorbing their rules, because their rules are a transcript of a fluency the rules cannot transmit. The goal is to read deeply enough that you can finally do the thing the book is only a record of — and then make a move the book has never seen.

Generative Questions

• If the Tso never used the word ch'i and the "unorthodox" was read back into it from the T'ang onward, how much of any tradition's "ancient wisdom" is its descendants' vocabulary projected onto ancestors who meant something else — and can the original meaning ever be recovered once the later label sticks?
• Li says the means of victory can be studied but the best of it is shen ch'i, "not explainable." Is the inexplicable residue a real category of skill, or just the name a tradition gives to the parts of expertise it has not yet managed to codify?
• If every codified tactic decays the moment it is published and read by the target, is there a stable advantage in withholding your doctrine — and does that explain why so much of the Chinese tradition framed its best teachings as "secret"?

Connected Concepts

• The T'ai Kung — Unorthodox Progenitor — the rival origin claim: a single founding hero versus doctrine abstracted from many anonymous battles.
• The Zheng/Qi Duality — the doctrine Li says was abstracted from the Tso's lived battles rather than invented by Sun-tzu.
• Chinese Military Theory — Historical Development — the long arc; this page is the structural argument that the theory formed after the practice.
• Animal Delivery Systems in Warfare — Li's lead example, the 506 BCE incendiary elephants that prefigure T'ien Tan's fire oxen.
• Ambush and Feigned Retreat in the Spring and Autumn — the broader body of Spring and Autumn battlefield practice the Tso preserves and Li mines for precedents.

Footnotes