Four Permissions
四种许可
Four Types of Condition Licensing War: Attitude, Purpose, Norm, Structure
About this series
The SAE War Treatise treats only the 15DD side of war. The war philosophy of the 14DD side never appears there in full form — it enters only as the object being constrained, or as the direction of a slide.
This series lets the other side speak. It reads historical works of war philosophy and asks where they run alongside SAE, where they diverge, and where they must be criticised. It is not a study of those authors. It uses them to show what SAE is.
Four working rules govern every volume:
- Where we agree, the point is why we agree — where SAE's reason and theirs part company, and what each reason therefore does and does not reach.
- Where we disagree, we set out SAE's own position rather than correcting theirs.
- Where a text touches the ground — treating a person merely as a means, colonisation, enslavement, using formal legitimacy to cover domination — criticism is required.
- Where the question lies outside SAE's range, we say nothing.
Volumes are ordered by thickness of interface with SAE, not by rank.
A note on Chinese terms: Laozi (the Daodejing), the Sima Fa (Marshal Rangju's Methods, a Warring States military-ritual text), and the Mozi are cited from standard Chinese editions; translations are the author's own. Bude-yi (不得已, "having no alternative") and bude-bu (不得不, "cannot-not") are kept distinct throughout; §5 turns on the difference.
Abstract
Laozi, the Sima Fa, the Mozi and SAE overlap at one place: all four refuse to treat war as an ordinary instrument of policy, and each establishes a condition of permission — respectively by attitude, purpose, norm, and structure.
One stronger reading must be excluded at once: the four do not share a "no other road is open" precondition. The Sima Fa's "to stop war by war, war is permissible" is a justificatory condition; the Mozi's distinction between gong (attack) and zhu (chastisement) is a qualifying condition; neither is a last-resort condition. Only Laozi and SAE clearly hold one.
The four conditions mark out different ranges. An attitude-type reason can lead to not fighting at all; a structural one cannot. A purpose-type reason supplies war with a legitimate result to be achieved, and therefore permits, structurally, enlarging the war for the sake of that result. The Sima Fa places just punitive war and post-war political rectification within one and the same normative architecture; this volume holds that the juxtaposition has a structural reason, but the text does not assert that inference, so it is marked as interpretation. A norm-type reason requires a position from which "righteousness" can be adjudicated.
The difference between bude-yi and bude-bu is that the former is a judgement about the agent's situation and can be sincerely mistaken; the latter requires structural conditions verifiable from outside.
The Mozi's Preparing the City Gates makes the internal condition of a community a necessary condition of defensibility, and its "the ruler must honour and employ him" corresponds to one of the four gates of SAE's Theorem One. The Mohists' defence of besieged cities is a genuine historical test of the gate requiring an intervening party to have interests of its own, and this volume issues a determination. The halting of the attack on Song recorded in Gongshu is not of the same type as Sunzi's "attack the enemy's plans."
Finally, Mohism serves as a counterexample to a general proposition about how constraints are generated — a proposition stated in full in §7 — and forces its rewriting.
Sources and scope
Materials: Laozi chs. 30 and 31; Sima Fa, "Benevolence as the Root"; Mozi, "Against Offensive Warfare" (upper, middle, lower), "Gongshu," "Preparing the City Gates," "Lu's Questions"; SAE War Treatise §8, §10, §11, §18, and Theorems One and Three. Verification status is given in the appendix.
Not treated here: the siege engineering and fortification technique of the Mozi's later chapters; the metaphysics of Laozi; the military-institutional content of the Sima Fa's remaining chapters. These belong either to the execution layer or outside this series' problem domain.
Saying nothing is not a judgement of value.
1 · How thick the overlap is, and what it is not
Laozi, ch. 31:
Weapons are instruments of ill omen, not the instruments of a gentleman; he uses them only when he has no alternative (bude-yi), and prizes calm restraint above all.
Chapter 30 runs the same way: one who assists a ruler by the Way does not coerce the world with arms. Such affairs tend to rebound.
Sima Fa, "Benevolence as the Root":
Thus: if killing people brings peace to people, killing is permissible; if attacking a state out of love for its people, attacking is permissible; if stopping war by war, then even war is permissible.
Mozi, "Against Offensive Warfare," lower, answering the challenge that Yu, Tang and King Wu all waged war and were nonetheless sage kings:
You have not examined the class of my words, nor understood their reason. Those were not what is called gong (attack); they are called zhu (chastisement).
SAE, Theorem One: cannot-not. The carrier of cultivation has been lost; cultivation in its mild forms can no longer continue.
Proposition 4.1 All four refuse to treat war as an ordinary instrument of policy, and each establishes a condition of permission — by attitude, purpose, norm, and structure respectively.
The overlap is real, and narrower than it looks.
What the four share: none accepts "advantageous, therefore permissible"; all require an additional gate before war; and all place that gate high in their own systems. Laozi puts it in the two chapters on arms, the Sima Fa in the opening passage, Mozi argues it across three chapters, SAE places it in Theorem One.
What the four do not share: the character of that gate.
Reading all four gates as one and the same "no other road is open" precondition does not hold.
- Laozi: uses them only when he has no alternative — this is a last-resort condition.
- SAE: the carrier of cultivation lost, mild forms no longer possible — also a last-resort condition.
- The Sima Fa: killing that brings peace is permissible; stopping war by war makes war permissible — this is a justificatory condition. It does not ask whether other roads have been walked; it asks what this war serves. A war serving the end of war is permitted on this reason even if other roads remain open.
- The Mozi: those were not gong but zhu — this is a qualifying condition. It does not ask whether other roads have been walked; it asks which class this resort to arms belongs to. Zhu does not require first showing that the alternatives to zhu have been exhausted.
So only Laozi and SAE clearly hold a last-resort condition.
Proposition 4.2 The four conditions of permission fall into four types: attitude (Laozi), purpose (Sima Fa), norm (Mozi), structure (SAE).
And this fourfold division is precisely the device that prevents the misreading above: it first says the four are not of one kind, and then examines what range each marks out. Press the four into a shared premise at the start, and every subsequent section will be supplying reasons for a premise that does not exist.
2 · The attitude type: Laozi
2.1 The form of the reason
Laozi's bude-yi hangs on a judgement about weapons themselves: arms are instruments of ill omen.
Fine weapons are instruments of ill omen; all creatures loathe them. Therefore one who has the Way does not dwell with them.
The form of the argument is: this thing has a bad character in itself, all creatures loathe it, and so one who has the Way does not dwell with it. That being so, to use it can only be a matter of having no alternative.
This is a judgement about the character of an object, and its upshot is an attitude — calm restraint above all, victory is not to be admired, a victory in war is to be received with funeral rites.
2.2 The difference in range: one exit
Proposition 4.3 An attitude-type reason can lead to not fighting at all; a structural one cannot.
Laozi's reason leaves an exit, and he himself walks toward it: since arms are of ill omen, the best course is not to use them at all. Chapter 30's "does not coerce the world with arms" and chapter 31's "one who has the Way does not dwell with them" both point that way.
SAE's cannot-not closes that exit. Once the structural condition is met, not fighting is not one option among others — it merely transfers the cost elsewhere, and usually to those least able to bear it.
This difference is not one of degree. Laozi's criterion permits a person to say "I will not take part in this," and supplies a complete reason for the choice. SAE's does not, because it asks not "shall I take part" but "has this situation any other form left."
2.3 One item Laozi has and SAE lacks
Chapter 30: "One who assists a ruler by the Way does not coerce the world with arms. Such affairs tend to rebound."
Affairs of this kind readily come back upon you.
This is a judgement about recoil, and its position is worth noting: it sits in the judgement about going to war, not in the judgement about means.
SAE's treatment of recoil is concentrated at the level of means — §12 on how deception suppresses one's own side's subjecthood; and the four recoil mechanisms set side by side in Volume III of this series (Kant on the spillover of vice, Arendt on violence consuming power, Machiavelli on the carrier-level interest group, and SAE's §12) all likewise operate on means. Theorem One contains no such item. It asks whether this is a cannot-not; it does not ask how the thing will come back upon us.
This item can be taken up, but with one difficulty: adding "will it rebound" to Theorem One would give the theorem a component of consequence-calculation, and that is exactly what Proposition 4.7 is concerned to keep apart from it. How to take it up is left open; this volume does not settle it.
2.4 "Victory is not to be admired": a second overlap, again with a different reason
Laozi, ch. 31, latter half:
Victory is not to be admired; one who admires it delights in killing men. … Where men have been killed in great number, let it be mourned with grief; a victory in war is to be received with the rites of mourning.
§8 of the treatise agrees with this conclusion, but reaches it by another road.
Laozi's reason is that arms are of ill omen: the thing itself is not good, so its success is not good either.
SAE's reason is a separation of two dimensions: military victory is one dimension, soundness in principle another. Winning does not amount to soundness — not losing is not the same as winning.
The difference in range: on Laozi's reason, no victory in any war is admirable, because arms are of ill omen in themselves. On SAE's, the completion of a defensive war sound in principle is not something that cannot be acknowledged — what cannot be acknowledged is the inference that victory itself constitutes legitimacy.
So the two block different things: Laozi blocks the glorification of war; SAE blocks the settling of principle by outcome. The former is broader; the latter is more specific — one blocks widely, the other blocks squarely against a particular inference.
3 · The purpose type: the Sima Fa
3.1 The form of the reason
Thus: if killing people brings peace to people, killing is permissible; if attacking a state out of love for its people, attacking is permissible; if stopping war by war, then even war is permissible.
Three sentences with one structure: the act itself (killing, attacking, warring) is not legitimate, but if it serves a legitimate result (peace, love of the people, the end of war), it is permitted.
This places war within a sequence of means: there is an end to be attained, war is one available means, and it is therefore permitted within that scope.
3.2 The structural consequence
Proposition 4.4 A purpose-type reason supplies war with a legitimate result to be attained, and therefore permits, structurally, enlarging the war for the sake of that result.
Argument: if the permission for war derives from the result it serves, then while that result is not yet attained, the war has not yet completed its legitimacy. "Continuing the war in order to stop war" is coherent within this reason, not an abuse of it.
And this is not a consequence pushed onto the text from outside. The Sima Fa writes it out.
3.3 The extension the Sima Fa gives itself
In the same chapter, the disposition following the chastisement of the guilty:
On entering the territory of the guilty: do no violence to the spirits; hold no hunts; damage no earthworks; burn no houses; fell no forests; take no livestock, grain or implements. Seeing the old and the young, escort them home unharmed; even meeting the able-bodied, if they do not resist, do not treat them as enemies. If they have been wounded, give them medicine and send them back. The guilty having been chastised, the king and the feudal lords shall set that state in order, raise up the worthy and establish the enlightened, and restore its proper offices.
The first part is a remarkably detailed code of conduct in war and occupation — no desecration of shrines, no hunting, no destruction of waterworks, no burning of houses, no felling of timber, no seizure of livestock, grain or implements; the old and young escorted home unharmed; the able-bodied not treated as enemies if they do not resist; the wounded given medicine and returned.
That code is real, and finer than most traditions of its era. It should not be passed over.
But the last clause is another matter: after the guilty have been chastised, the king and the lords are to set that state in order, select the worthy, install an enlightened ruler, and restore its offices.
Proposition 4.5 The Sima Fa, "Benevolence as the Root," places just punitive war and post-war political rectification within one and the same normative architecture.
This is a textual fact: one chapter, one sequence of norms — first the prohibitions governing conduct on entering enemy territory, then, immediately, the rectification of that state by the king and the lords once the guilty have been chastised. The two are contiguous in the text.
The next step is this volume's interpretation, and is marked as such:
If the permission for war derives from the result "stopping war," then securing that result belongs to the war. And a state just chastised will generate fresh disorder if its internal order is not rebuilt — so on this reason, rebuilding is entailed.
The text does not assert that inference. The Sima Fa nowhere says "because the purpose is to stop war, political rectification is necessary"; it merely places the two together. This volume holds that the juxtaposition has a structural reason, and that holding is interpretation, not quotation.
A note on the state of the text: the five surviving chapters of the Sima Fa are the remnant of a larger ancient corpus, formed principally in the scholarly milieu of Warring States Qi, though preserving an earlier stratum of military ritual. One should therefore not draw a strong claim about authorial logic from the co-presence of two propositions in one chapter. This volume's interpretation is offered as one reading, not as the necessary sense of the text.
But even taking only the layer of textual fact, the contrast with SAE stands: §18 prohibits precisely the act of "setting that state in order, raising up the worthy and establishing the enlightened," and the Sima Fa writes it into the sequence of norms for a just war. Two normative systems giving contrary provisions at one and the same place — that does not depend on any interpretation.
3.4 Where Theorem Three stands here
Theorem Three requires that war be directed toward ending. But "ending" is not the same as "attaining a legitimate result."
The content of ending is: domination broken, and the possibility restored for the other party to organise itself.
It does not include making the result. A situation in which domination has been broken and the other party then falls into disorder is, by Theorem Three, ended; by the Sima Fa, incomplete.
The difference can be stated precisely: the Sima Fa's terminus lies in the state of affairs — is that country now settled; SAE's lies in the relation — has the domination been broken, has the other party recovered the position from which it decides for itself.
A state of affairs can be produced by the victor. A relation cannot.
3.5 One correspondence across traditions
In Doctrine of Right §60 (1797), Kant permits causing an "unjust enemy" to adopt a constitution unfavourable by its nature to war — the same act.
And the two reasons are entirely different: Kant's runs through Recht (the character of that state's constitution is a property identifiable from outside), the Sima Fa's through the stopping of war (the result must be secured).
Two wholly different paths landing at the same place shows this is not the deviation of one tradition. It is the common exit of the practice of setting a legitimate goal for war — once legitimacy derives from the result to be attained, securing the result belongs to the war, and the most direct way to secure it is to build the order on the other party's behalf.
(Kant's provision is treated in full in Volume I of this series.)
4 · The norm type: the Mozi
4.1 The form of the reason
The argument of "Against Offensive Warfare," upper, is an analogy:
Suppose a man enters another's orchard and steals his peaches and plums. Hearing of it, all condemn him… As for killing an innocent man, stripping him of his clothes, and taking his spear and sword — this is more unrighteous still than entering a stable and taking another's horse or ox. Why? Because the injury to others is greater… In such cases the gentlemen of the world all know to condemn it and call it unrighteous. Yet when it comes to the great unrighteousness of attacking a state, they do not know to condemn it; they follow along and praise it and call it righteous. Can this be called knowing the distinction between righteous and unrighteous?
The structure is clear: unrighteousness increases with the injury done; stealing fruit, stealing livestock, killing the innocent — each graver, and all commonly held unrighteous; while attacking a state, which injures most, is called righteous. This is an inconsistency in logic, not a divergence in values.
Mozi presses it to its extreme: killing one man is one capital crime, killing a hundred is a hundred; while attacking a state kills by the ten thousand and is praised as righteous — "seeing a little black, one calls it black; seeing much black, one calls it white." That is not a difference in values but an inability to tell black from white.
4.2 That adjudicating position
But in "Against Offensive Warfare," lower, answering the challenge, Mozi introduces a division:
Those were not what is called gong; they are called zhu.
The division is clean, and it places the whole weight on one thing: who adjudicates whether this is gong or zhu.
Mozi's criterion is "to promote what benefits the world and eliminate what harms it" — a substantive standard requiring someone to certify it. And the instances he cites for the three sage kings all appeal to the mandate of Heaven: the Miao were in great disorder, Heaven decreed their destruction; Xia's virtue was in great disorder, Heaven then commanded Tang; Heaven commanded King Wen of Zhou to chastise Yin.
In his setting, the adjudicator is Heaven, and Heaven's will is displayed through portents — a solar prodigy at night, three days of blood-rain, a dragon born in the temple, a red bird bearing a jade tablet.
Proposition 4.6 A norm-type reason depends on a position from which "righteousness" can be adjudicated.
This is isomorphic with just war theory — that framework likewise requires a position from which to determine "who is the just party," and among states there exists no adjudicator all parties acknowledge. That structure is treated in full in Volume V of this series; two points only are recorded here:
First, SAE does not establish that position. Theorem One does not ask whether this war is gong or zhu; it asks after structural conditions verifiable from outside.
Second, what Mozi has and SAE does not must be acknowledged. The analogical argument of "Against Offensive Warfare," upper, does not depend on the mandate of Heaven — it is a pure consistency argument, and anyone who grants that greater injury is greater unrighteousness must accept its conclusion. Its force is independent of the adjudicating position, and it engages precisely the class of opponent SAE finds hardest: one who justifies aggression by saying his side is righteous. Mozi's method is not to argue about righteousness with him, but to point out that his measure differs between small matters and large.
SAE can use this, and SAE currently has no equally forceful form of it.
5 · Bude-yi and bude-bu
Proposition 4.7 Bude-yi is a judgement about the agent's situation; bude-bu requires structural conditions verifiable from outside.
Among the four, Laozi comes closest, since bude-yi does structurally point at "there is no other way," not at "unwillingly." But the centres of gravity differ:
- Bude-yi centres on the agent — he was forced to it, he had no other choice, he did it and took no pleasure in it. Laozi's "calm restraint above all" and "victory is not to be admired" say exactly this.
- Bude-bu centres on the situation — whether those forms are still open, irrespective of what the agent wants or feels.
Corollary 4.7a Bude-yi can be sincerely mistaken; bude-bu cannot.
A man can sincerely feel he has no alternative. Fear, humiliation, misinformation, incomplete knowledge can all produce that feeling, and produce it without a trace of insincerity. Laozi's criterion has no guard against this, because it was never about the structure of a situation.
And in this matter, being sincerely mistaken is as fatal as being insincere.
This is SAE's practical reason for placing the criterion in structure rather than in attitude. Not that structure is finer than attitude; that feeling cannot be reviewed, and the cost of war is not borne by feelings.
But one cost must be acknowledged at once. Whether a structural condition can be verified is itself a question. "Whether cultivation in its mild forms can still continue" cannot be counted the way "our own casualties" can. So Corollary 4.7a states a difference in principle — an attitude-type criterion is unverifiable in principle, while a structural one is verifiable in principle and not always in practice. The difference is real, and smaller than it looks.
6 · The Mohists defending cities: a real test of Theorem One's four gates
The Mohists did not merely oppose offensive war; they organised armed parties to help other states hold their walls. This is the only material in pre-Qin China that can be brought to bear on the second face of Theorem One.
6.1 The material
"Gongshu": Gongshu Ban had built scaling ladders for Chu, to attack Song. Mozi set out from Qi and walked ten days and ten nights to Ying. He first persuaded Gongshu Ban, then the King of Chu. The king conceded the argument:
The king said: "Well said. Even so, Gongshu Ban has made me scaling ladders; I must take Song."
The argument did not stop the war.
There followed the famous simulated siege: Mozi unfastened his belt to make a wall and used slips of wood for engines; Gongshu Ban deployed nine changes of siege machinery, and Mozi repelled all nine — Gongshu Ban's engines were exhausted, and Mozi's defence had margin to spare.
Gongshu Ban said, "I know how to defeat you, but I will not say it." Mozi said, "I know how you would defeat me, and I too will not say it." The king asked the meaning, and Mozi answered:
Gongshu's intent is nothing more than to kill me. Kill me, and Song will be unable to hold, and can be attacked. But my disciple Qin Huali and three hundred others already hold my defensive implements upon the walls of Song, awaiting the invaders from Chu. Though you kill me, you cannot cut them off.
"Preparing the City Gates," opening, where Qin Huali asks how one may defend a small state, Mozi's answer is not about equipment:
Our walls and moats repaired, defensive implements ready, firewood and grain sufficient, superiors and inferiors in mutual affection, and further the aid of the neighbouring lords — by these one holds. And though the defender be skilled, if the ruler does not employ him, still one cannot hold. … Therefore the defender must be skilled and the ruler must honour and employ him; only then can one hold.
The second passage lists fourteen conditions for holding a city:
The general method of holding a city: thick and high walls; deep and broad moats; watchtowers in repair; defences well maintained; firewood and food sufficient for more than three months; the population numerous and select; officials and people in accord; many great ministers having rendered service to their lord; the ruler trustworthy and just, and the myriad people delighting in him without end. Failing that, the graves of fathers and mothers are there; failing that, the mountains, forests, marshes and lowlands are rich enough to profit by; failing that, the terrain is hard to attack and easy to hold; failing that, there is deep resentment against the enemy and great merit toward one's superiors; failing that, rewards are clear and credible and punishments severe enough to be feared. Where these fourteen are present … then the city may be held. Where not one of the fourteen is present, then even a skilled man cannot hold it.
6.2 Isomorphic with §11
Proposition 4.8 Preparing the City Gates makes the internal condition of a community a necessary condition of a city's defensibility, isomorphic with §11 of the treatise.
Of the fourteen, walls, moats, towers, implements, grain and terrain are material conditions; while officials and people in accord; the ruler trustworthy and just and the people delighting in him; the graves of the ancestors being there; deep resentment against the enemy and great merit toward superiors; rewards clear and punishments feared — these are conditions internal to the community.
And the conclusion is very strong: where not one of the fourteen is present, even the finest expert cannot hold the city. Lacking any one, the best defensive engineer will not save it.
This is "deepening cultivation is national strength" in Warring States form — whether a community can hold itself depends on its internal condition, not on how good a technician it has hired.
6.3 "The ruler must honour and employ him" and the gate of genuine rising
Proposition 4.9 "The defender must be skilled and the ruler must honour and employ him; only then can one hold" corresponds to the genuine rising gate on the second face of Theorem One.
The force of the sentence lies in its negative form: however good the defender, if the ruler does not employ him, the city still cannot be held. Mozi is not discussing courtesy; he is stating an operational fact — the party being defended must itself want to defend.
And the practice recorded in "Gongshu" matches: Mozi went first to dissuade the attacker, while sending disciples to assist Song's defence, not to take it over. Nowhere in the chapter does he demand that Song hand over command, and nowhere does he establish anything in Song.
6.4 The gate of the intervening party's own interests: this volume's determination
Among the four gates on the second face of Theorem One, the hardest is that an intervening party must have interests of its own — the gate exists to prevent deciding on others' behalf whether they should resist.
What were the Mohists' own interests? They were not men of Song. "Gongshu" adds one detail at its close:
Mozi returned, and passing through Song it rained; he took shelter in a gateway, and the gatekeeper would not let him in.
The state he had saved did not know him, and would not let him in out of the rain. That detail rules out any motive of reward.
Proposition 4.10 The Mohists' defence of cities satisfies the gate of the intervening party's own interests, but their interest lies at the level of a community's continued existence, not at the level of benefit.
Argument: Mohism was a community with a definite doctrine, a definite organisation, and a definite way of life. That community's continuation required one condition — that there exist a number of independent small states, some of whose rulers were willing to listen and to employ them.
In a world annexed to the end, that condition disappears. Mozi calculated this himself in "Against Offensive Warfare," middle: of the states enfeoffed in antiquity, those destroyed by offensive war are beyond counting; Ju in the east was a very small state, wedged between great ones, and in the end perished. In a world where the small states have all been swallowed, Mohism has nowhere to go.
So stopping attacks is not altruism; it is guarding one's own carrier — only that carrier is not property or territory but "a world in which small states still exist and someone is still willing to hear a doctrine."
The consequence of this determination is that the sense of this gate is broader than first supposed. It does not require the intervening party to have material gains or losses at stake, only that the intervening party's carrier of cultivation genuinely depends on the outcome of the intervention.
The contrary possibility must be stated. Read more strictly — requiring a direct, identifiable common interest between intervening and intervened — the Mohists do not satisfy it, and Theorem One would then determine that the practice closest to 15DD in pre-Qin China does not hold. This volume adopts the broad reading, on the ground that the strict reading would exclude all cross-community assistance, which is not what this gate exists to prevent. What it exists to prevent is deciding on others' behalf whether they should resist — and the Mohists left precisely that decision to the other side (Proposition 4.9).
6.5 The halting of the attack in "Gongshu" is not of Sunzi's type
Proposition 4.11 The halting of the attack recorded in "Gongshu" is not of the same type as Sunzi's "attacking the enemy's plans."
Sunzi's fa mou is attacking the plan — destroying the other side's design so that it cannot stand.
In "Gongshu," the king's plan was not destroyed: having conceded the argument, he still said "I must take Song." What changed the outcome were two things: the result of the simulated siege, nine attacks repelled nine times; and the fact that three hundred men were already on Song's walls with defensive implements.
This is making the attack fail on cost, not making the plan fail.
And its precondition is that the defence was already in place. Had Qin Huali's three hundred not been on Song's walls, neither Mozi's advocacy nor the sand-table exercise would have stopped the king — Mozi says so plainly in the text: kill me, and Song cannot hold, and can be attacked.
This is a historical specimen of §10's "not fighting is also preparing", and the specimen gives a more specific form: preparation is part of what makes fa mou work, not the next step after fa mou has failed.
Sunzi's four-rung sequence (attack the plan, attack alliances, attack the army, attack cities) implies a temporal order — first plans, failing that alliances, failing that arms. In "Gongshu," preparation and persuasion proceed simultaneously, and preparation is the condition of persuasion's working.
This bears directly on Theorem One's "the exhaustion of fa mou and fa jiao": if preparation is the condition of fa mou's working, then "exhausting fa mou" cannot be understood as "saying everything one can say while unprepared."
And this is a concrete instance of a more general dependency:
Theorem One's threshold includes "fa mou and fa jiao exhausted," and determining whether that is satisfied requires knowing what options fa mou and fa jiao comprise and under what conditions each works — which is execution-layer knowledge, and SAE explicitly does not do the execution layer.
A layer of principle may declare that it does not do the execution layer. It may not declare that it does not depend on one.
The full treatment of that dependency is in Volume VI of this series (on Sunzi).
7 · A proposition about how constraints are generated, and Mohism's test of it
The proposition tested in this section is stated here in full, because it comes from another volume of this series and everything in this section depends on it.
The proposition tested (original form, universal):
Constraints on the enemy-facing side can be generated by reciprocal equilibrium — that one does not strike a state in mourning, does not exploit a natural disaster, drums only once the enemy has formed his ranks: such rules can arise and persist because the other side observes them too, and a violator will meet the same treatment next time. This is an equilibrium, and requires no party to hold any support at the level of principle.
Whereas on one's own side there is no counterparty to enforce. Who redeems the soldier's standing to be informed? Who makes a commander pay for violating it? In a one-directional structure where reward, punishment, life and death all issue from the commander, no mechanism can stabilise such a constraint. Such constraints therefore lack the conditions of their generation.
Provenance and evidential basis: the proposition comes from Volume VI of this series (on Sunzi), where the material was a single strand of Chinese military thought — the Sunzi with its commentators, and the Sima Fa's "Benevolence as the Root." The Sima Fa's well-known self-limitations (not striking a state in mourning, not exploiting disaster, drumming only after ranks are formed, pursuing no more than a hundred paces, attacking a state out of love for its people) all fall on the enemy-facing side, and those two books show no corresponding provision on the standing of one's own soldiers.
That proposition at the time flagged Mohism as a possible counterexample, for this reason: the Mohists were a transnational organisation, and their internal discipline was enforced by no state's military law — so if they did hold a self-limitation on their own side, and its support was not reciprocity, the universal fails.
This section is that test.
Proposition 4.12 Mohism is a counterexample to the original universal form of that proposition, and forces the rewriting of its domain and its table of mechanisms.
To say "Mohism is not a counterexample, only an addition" does not hold — if the original proposition is universal (no generative mechanism on one's own side), then a single sample with a non-reciprocal constraint on that side is a counterexample.
7.1 The material of the counterexample
"Lu's Questions" records: Mozi sent his disciple Sheng Chuo to serve Xiangzi Niu; Xiangzi Niu invaded Lu three times and Sheng Chuo followed him three times. Hearing of it, Mozi sent Gao Sunzi to request his withdrawal — he recalled his own student. His reason:
To speak of righteousness and not act on it is knowingly to transgress. It is not that Chuo did not know; his salary outweighed his righteousness.
This is a discipline on a member of one's own side, and its support is not reciprocity — Sheng Chuo could not punish Mozi, and Mozi had no need to fear retaliation from Sheng Chuo.
7.2 But its support is not "humanity is an end" either
This distinction must be stated, because it is the easiest thing in the section to misread.
What stabilised Mohist discipline was a whole bundle of institutional conditions, not the standing of a norm on its own:
- Rules acknowledged as constitutive of the community (not to keep them is not to be a Mohist)
- An internal authority and review mechanism (the office of the Grand Master)
- Real consequences for violation (recall, expulsion)
- Exit a real and not merely nominal possibility (voluntary entry; one may leave)
- A channel for appeal and reinterpretation (the Mohist tradition of disputation)
The character of this bundle must be stated.
These five are a set of candidate conditions, generalised from Mohism as a single sample, untested for necessity or sufficiency. This volume does not claim they are conditions any own-side constraint must have.
And "Mohism has all five" is itself overstated: the Grand Master and expulsion have textual support ("Lu's Questions," the recall of Sheng Chuo); "rules acknowledged as constitutive" is supported by "to speak of righteousness and not act on it is knowingly to transgress"; while "a channel for appeal and reinterpretation" comes from a general impression of the Mohist tradition of disputation, for which this volume has no specific citation.
What can be said is: Mohism possessed at least several of these, and the support of its discipline was plainly not reciprocity.
Compressing the mechanism into "voluntary community membership" is too simple.
7.3 A distinction that must be carried through
Why a norm holds and how a norm comes to be stably enforced within an organisation are two things.
- "Humanity is an end" does not fail because institutions are inadequate. In a community without a Grand Master, treating a person merely as a means is still not legitimate.
- That the office of Grand Master could stabilise discipline does not show that discipline was legitimate. That an organisation can make its members keep the rules, and whether those rules are right, are two questions.
So what this section gives is a hypothesis about institutional stability, not a judgement about the source of norms.
7.4 The rewritten form
Corollary 4.12a (revised) The original universal form of the proposition is overturned; the rewritten form is:
Self-limitations on the enemy-facing side can be stabilised by reciprocity; self-limitations on one's own side cannot depend on the other party's retaliation, and must be stabilised otherwise — by internal rights, procedures, membership structures, or institutional review. Reciprocity is not the only mechanism by which constraints are generated.
This form has a real boundary of application. Mohists could leave Mohism; soldiers cannot leave an army. The five conditions of §7.2 all fail in a conscript army — membership is not voluntary, the cost of exit is death, there is no internal review, and there is no channel of appeal.
So after the revision, a conscripted army still falls on the side of "the conditions of generation are absent."
And the corollary attached to the original proposition is unaffected:
A constraint hung on reciprocity is effective only on the face where reciprocity obtains; a criterion hung on "humanity is an end" takes effect without requiring the other party to be capable of retaliation.
That corollary is the practical reason for placing a criterion on the ground rather than on custom. Custom disappears where reciprocity disappears — the collapse of ritual restraint in the Warring States is one example compatible with this — and on one's own side there was never any reciprocity to be had.
8 · Left unresolved in this volume
8.1 Laozi's "such affairs tend to rebound" is a recoil mechanism situated in the judgement about going to war, and Theorem One has no counterpart (SAE's recoil mechanisms all sit at the level of means; see Volume III). Whether to take it up is undecided. The difficulty: taking it up would add a component of consequence-calculation to Theorem One, and Proposition 4.7 exists precisely to keep the two apart.
8.2 Proposition 4.10 adopts a broad reading of the gate of the intervening party's own interests (that a carrier of cultivation depends on the outcome suffices; material gain or loss is not required). The boundary of that broad reading is undetermined: what counts as "depending on"? Does a community that merely cares, in thought, about another's fate satisfy it? This volume does not answer.
8.3 Proposition 4.11 finds that preparation is a condition of fa mou's working, so "exhausting fa mou" cannot mean "saying everything one can say while unprepared." This deepens Theorem One's threshold determination into a dependency on execution-layer knowledge, treated in Volume VI.
8.4 The consistency argument of "Against Offensive Warfare," upper (seeing a little black one calls it black; seeing much black one calls it white) engages the class of opponent who justifies aggression by claiming righteousness more directly than any form SAE currently has. Whether SAE should adopt an argument of the same type has not been considered.
Appendix · Sources and verification status
| Material | Edition | Location | Status |
|---|---|---|---|
| Laozi | Chinese text (Gutenberg) | chs. 30, 31 | primary source verified |
| Sima Fa, "Benevolence as the Root" | Standard text, cross-checked across sources | ch. 1 | primary source verified (received text reads 修正其國, "set that state in order") |
| Mozi | Chinese text (Gutenberg) | Against Offensive Warfare (upper, middle, lower); Gongshu; Preparing the City Gates; Lu's Questions | primary source verified |
| Kant, Doctrine of Right §60 | Akademie edition, German | VI:349–350 | primary source verified; see Volume I |
| SAE War Treatise | Final version | §8, §10, §11, §18, Theorems One and Three | primary |
Translations from the Chinese are the author's own.
Two notes:
- The received text of the Sima Fa reads 修正其國. The five surviving chapters are the remnant of a larger ancient corpus, formed principally in the Warring States scholarly milieu of Qi, though preserving an earlier stratum of military ritual.
- On Proposition 4.5: that the chapter places just punitive war and post-war political rectification within one normative architecture is a textual fact; that the latter is a necessary extension of the former is this volume's interpretation, and is marked as such in the body.
A companion note, Known Limits, records the items this series has left unresolved.
On this series' method
This volume, like the rest of the series, retains its own correction records in the body text: withdrawn propositions, narrowed claims, and reductions are marked where they occurred rather than silently repaired.
They are retained for a reason: a framework that requires "cannot-not be questioned" must be the first thing questioned; and a text that shows no sign of ever having erred gives its reader no way to judge how its determinations were reached.
In this volume the point bears on the title itself. A title such as Four Ways of Having Nowhere Left to Go would presuppose that all four traditions share a last-resort condition — and §1 shows they do not. What a volume is called must answer to what it can establish.