The Hand Is Being Reworked by What It Holds
手在改造握它的人
Four Mechanisms of Recoil upon the User, and the Test They Compose
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.
"Outside SAE's range" does not mean "we happen to agree." It means SAE's criteria do not speak to that question. Saying nothing is not a judgement of value.
Volumes are ordered by thickness of interface with SAE, not by rank.
A note on this volume's form: it is not a merged volume in the usual sense. Volumes II, IV, V and VIII bring several authors together because they give reasons of one kind. This volume brings four mechanisms together because they are heterogeneous — a matrix, not a consolidation. The point of setting them side by side is precisely that they do not reduce to one another.
Abstract
Kant, Arendt, Machiavelli and SAE each supply one mechanism for a single fact: a means reworks the party that wields it. The four operate at different sites and on different timescales, and none contains any other — each has cases the other three do not reach.
This volume does not claim that the four "do not overlap," nor that SAE's is "the narrowest." What it shows is non-containment, not disjointness; and the four have heterogeneous domains, so there is no common scale on which to rank their breadth.
What is retained is one specific blind spot: §12 of the SAE War Treatise binds disqualification to the existence of an information gap, so an arrangement conducted openly, with the executor fully informed, in which the person's standing as an end is nonetheless cancelled, draws no verdict from that criterion.
Three of the four can be taken up. Machiavelli's supplies what SAE lacks — a carrier-level mechanism: not a party at war sliding at the level of ideas, but the existence of a group whose livelihood is bound to war, a group requiring no one to err in thought. Arendt's reason lies closest to SAE's, the difference being that she does not treat the cannot-not; her conceptual distinction between violence and power stands, and what cannot be transferred directly is the empirical diagnostic derived from it, which needs an added mediating distinction. Kant's supplies the temporal dimension, and in 1797 he pushed it to the level of a criterion.
Together the four compose four angles of test for a list SAE could not have produced alone. But a prior gate must be passed before the four questions: does this arrangement cancel any particular person's standing as an independent end? The four questions screen what a means does to its user; they do not screen what an arrangement does to a particular person. The substitution case set out in §2.2 passes all four while the child's standing is nonetheless cancelled — that is the evidence for the gate's necessity.
And the verdict is asymmetric: failing any one item yields a determination; passing all of them yields only "no structural consequence was found on the dimensions covered." A limited screen issues no certificate.
A closing judgement about insight itself: Machiavelli, in one and the same book, both diagnosed the mechanism completely and advocated destroying cities accustomed to freedom. Insight is not enough to make a man stop.
Sources and scope
Materials: Kant, Perpetual Peace, Sixth Preliminary Article (reason six), and Doctrine of Right §57 (AA VI:347–348); Machiavelli, Art of War Book I and The Prince ch. 5; Arendt, On Violence (1970); SAE War Treatise §7, §11, §12. Verification status is given in the appendix.
Arendt's On Violence remains in copyright. Her positions are paraphrased throughout; no passage is quoted.
Not treated here: the operational detail of any military technique. The examples in §5 serve only to test whether the criteria are usable; they do not constitute a complete assessment of any technique — that would require execution-layer knowledge.
Saying nothing is not a judgement of value.
1 · Four mechanisms
Setting them side by side is the main work of this volume. Taken singly, each looks like common sense; only together do they show that they are not governing the same thing.
1.1 Kant: the habit is carried over into peace
The Sixth Preliminary Article of Perpetual Peace prohibits assassins, poisoners, breach of terms of surrender, and instigation of treason. He gives six reasons; the least often cited is the sixth:
The methods just mentioned necessarily lead to this result — for these infernal arts, being in themselves base, would not, if adopted, be for long confined to the sphere of war.
Take, for example, the use of spies (uti exploratoribus). Here only the dishonesty of others is made use of; but vices such as these, when once encouraged, cannot in the nature of things be stamped out, and would be carried over into the state of peace, where their presence would be utterly destructive to the purpose of that state.
Site: the user's own community after the war. Mechanism: irreversibility of habit — once a practice is institutionally encouraged, it does not stop on the day the armistice is signed.
Note why he names espionage. He writes it himself: "here only the dishonesty of others is made use of." He knows this item looks the most harmless of its family, and that is exactly why he chooses it. If even this one spills over, the heavier ones need no argument.
1.2 Arendt: violence destroys power and does not create it
Arendt defines power as the capacity to act in concert — it belongs to a group and exists only while the group holds together. Violence is instrumental: it requires an external end to justify it, and its efficacy grows with the improvement of the implement, independently of the number of those who hold it.
From this she draws her judgement: violence can destroy power, but it can never create it. A regime that depends on violence to rule has thereby shown that it has already lost power.
Site: power itself. Timescale: immediate — not "it degrades with use," but it consumes at the moment of use, since resorting to violence just is the showing that the capacity to act in concert no longer suffices.
1.3 Machiavelli: those who live by war must wish that peace not exist
In Art of War Book I, Fabrizio Colonna argues that a well-ordered republic never permits any of its citizens or subjects to practise war as a trade.
What he gives is a chain, not a sentence:
- The trade has no ordinary livelihood in peacetime, so those who depend on it must wish that peace not exist (desiderare che la pace non sia).
- That incentive produces predation: robbery, fraud, violence and killing are structurally attached to the trade.
- Commanders' standing and income depend on war, giving them an incentive to deceive their employer in order to prolong it.
- When peace comes and such men are discharged, they readily become armed adventurers living by plunder.
- Hence a constitutional trilemma: a prince maintaining professional soldiers in peacetime faces three bad options — keep fighting, maintain them indefinitely, or face a body of armed men whose livelihood has vanished.
- Loyalty attaches to the commander rather than to the commonwealth.
Site: an interest group within the social structure. Timescale: long.
Two qualifications. The speaker is Fabrizio, not Machiavelli in his own voice; and what is forbidden is private livelihood, not knowledge of war — The Prince ch. 14 says in the author's own voice that a prince should make the art of war his principal study, and the two do not conflict.
1.4 SAE: the defect of gui lies in the suppression of one's own side
§12 of the SAE War Treatise holds that gui — deception, the lowest rung in its hierarchy of means — is disqualified not because it deceives the enemy but because maintaining an information gap requires that one's own executors be used as instruments. If the executor knows the whole picture, the gap cannot be held; so to use gui one must first keep one's own people in the dark.
Site: the executor's standing as a subject, on one's own side. Timescale: during the war.
One registration is required here: the four-rung hierarchy of means used in this section (mou, yinmou, yangmou, gui) previously drew part of its warrant from an attribution that has since been withdrawn — that Sunzi had already drawn a comparable distinction. The re-argument of the hierarchy itself does not belong to this series; it is a matter for §12 of the treatise. This volume uses it while that argument remains incomplete.
2 · Where the four differ
2.1 Side by side
| Site | Timescale | Depends on what to hold | |
|---|---|---|---|
| Kant | the ethos of one's own community | after the war | irreversibility of habit |
| Arendt | power itself | immediate | power = capacity to act in concert |
| Machiavelli | an interest group in the social structure | long | livelihood bound to war |
| SAE §12 | the executor's standing as a subject | during the war | an information gap requires an uninformed executor |
Proposition 3.1 The four mechanisms point at one fact — a means reworks the party that wields it — while differing in site, timescale, and conditions of application.
2.2 Non-containment, and one specific blind spot
Proposition 3.2 The four do not contain one another — each has cases the other three do not reach — and therefore none can replace the rest.
What this volume shows is non-containment, not disjointness — "their coverage does not overlap" does not hold. A single technique may perfectly well depend on an information gap, build a professional carrier, spill over into peace, and substitute for concerted action, all at once — the four are not a partition but four angles of test with overlapping regions and regions of their own.
And the ranking "SAE's is the narrowest" likewise does not hold: the four have heterogeneous domains (ethos, power, interest group, executor's standing), and there is no common scale on which to rank their breadth.
Each one's blind region can still be named:
- Kant's cannot test the immediate. A means genuinely used once, genuinely forming no ethos, draws no verdict from it — and "used only once" is something a party can sincerely believe.
- Arendt's cannot test means that do not employ violence. Deception, bribery, blockade, and information manipulation lie outside her definition — and those are among the most used today.
- Machiavelli's cannot test cases without a professional group. An army raised temporarily and disbanded after the war produces no one who lives by war.
- SAE's cannot test means that require no information gap. An arrangement conducted openly, with the executor fully informed, in which a person's standing as an end is nonetheless cancelled, draws no verdict from §12.
Corollary 3.2a §12's criterion is bound to the single variable of the information gap; there is therefore a class of arrangement it cannot reach: one with no information gap in which the standing as an end is nonetheless cancelled.
What is wanted here is the specific blind spot, not a comparative judgement of breadth such as "the narrowest of the four" (see Proposition 3.2).
§12 locates the defect of gui in "one's own people must be kept in the dark." That location is accurate — but it binds the criterion's reach to the information gap. A practice that cancels the executor's standing without depending on an information gap walks straight past it.
2.3 A worked case, set out in full
This case is given here in full because the later sections use it repeatedly.
Source: the Sugawara story as told in the chapter on loyalty in Nitobe Inazō's Bushido (1899). Sugawara no Michizane has been exiled; his political enemies mean to root out his line, and discover that his young son is hidden in a village school kept by an old retainer, Genzō. The order comes down: on such a day, produce the boy's head.
What occurs to Genzō is to find a substitute — and just then a new pupil arrives, of the same age as his lord's son and resembling him, brought by a mother of noble bearing.
Nitobe inserts a sentence here: the mother and the boy himself were equally aware of the resemblance; in the privacy of their home, both had already laid themselves upon the altar. On the appointed day, the official who comes to verify the head is the boy's own father; he checks the features one by one, unhurried, and pronounces the head genuine. That evening, crossing his own threshold, he says to his wife: "Rejoice, my wife, our darling son has proved of service to his lord."
There is no information gap anywhere in this — the parents knew, the boy knew, the teacher knew, and the official who verified the head was the boy's father. Everyone was informed, and the child's standing as an end was nonetheless cancelled: his entire function in the arrangement was that his face resembled another child's.
(The full treatment of this case is in Volume VIII of this series; here it serves only as an object of test.)
So §12 needs a supplementary criterion that does not depend on an information gap — and the four questions of §5 move in that direction, though as §5.3 records, they do not reach it either.
3 · Three that can be taken up
Under the working rules, what the other side has and SAE lacks must be marked, and what can be taken up should be. All three can, and each in a different way.
3.1 Machiavelli: what SAE lacks is the carrier level
Proposition 3.3 §7 of the treatise gives a judgement at the level of ideas; Machiavelli gives a mechanism at the level of the carrier, and SAE does not have the latter.
§7 holds that war rises from means to end. That is a diagnosis at the level of ideas: a party at war, in the course of things, comes to treat the means as the end. Its subject is "the party at war," and a party at war is something that holds ideas.
What Machiavelli gives is not an idea: there exists a group that eats by war. That group requires no one to err in thought. It need only act on its own interest, and war acquires a continuing impetus.
SAE genuinely lacks this, and it is harder than the judgement at the level of ideas, for two reasons.
First, it can be observed from outside. How many people in a society have their livelihood bound to a state of war is countable — the size of a standing army, employment in the arms industry, regional economies dependent on military orders. A slide at the level of ideas cannot be counted; a binding of livelihood can.
Second, it does not vary with intention. A wholly sincere decision-maker, without a trace of expansionist ambition, still faces Machiavelli's trilemma: what becomes of the men he maintains when peace arrives? The difficulty does not dissolve because his heart is good.
One verification finding makes this determination firmer: the six reasons Machiavelli gives are entirely matters of incentive structure and social position; not one is an efficiency reason. An earlier version of this series supposed that the force of the restriction lay in its "holding against efficiency" — that was wrong, and correcting it strengthens the argument: if the force came from holding against efficiency, it would be the outcome of a weighing and would vary with efficiency parameters; coming from a carrier-level diagnosis, it does not.
Two consequences of taking it up:
- §7 gains a checkable indicator: the proportion of the population whose livelihood is bound to a state of war.
- The list of non-transferable techniques gains an entry: if the working of a technique requires establishing a professional group that lives by it, it carries that structural consequence with it, whatever the user's intentions.
3.2 Arendt: closest in reason; the difference is the cannot-not
Proposition 3.4 Arendt's reason lies closest to SAE's, and the difference is that she does not treat the cannot-not.
She defines power as the capacity to act in concert, which corresponds closely to the structure of cultivation: a community's strength comes from its members' capacity to act together, not from how much coercion it can apply.
§11 of the treatise says almost the same thing: a subject with deeper cultivation is more likely to prevail, because cultivation just is the source of that capacity to act together.
The difference: her analysis stops at "violence cannot create power." That sentence is true, but it does not answer what to do when the other party is destroying your capacity to act in concert.
So this is not a matter of taking something up to fill a gap. It is the confirmation that two paths reached the same place, after which SAE walked a further stretch. Theorem One treats exactly that position.
3.3 Arendt: the conceptual distinction stands; the derived diagnostic does not transfer directly
Proposition 3.5 Arendt's conceptual distinction between violence and power may be retained; what cannot be transferred directly to the side of the dominated is the empirical diagnostic derived from it — "the presence of violence entails the absence of concerted power."
Registration of the divergence: cross-domain non-transferability — not conceptual failure.
The registration must be precise. "Her distinction does not hold on the side of the dominated" would be too strong — what needs adjusting is not the distinction itself but the mediating step missing when the derived diagnostic is transferred.
Her distinction: violence is instrumental, requires an external end to justify it, and grows in efficacy with the implement; power is the capacity to act in concert, belongs to a group, and exists only while the group holds together.
The diagnostic derived from it: the resort to violence shows that the capacity to act in concert no longer suffices.
That diagnostic is accurate in her own domain. Her book treats the relation between a regime and its own people — a regime that turns violence inward has indeed shown it cannot sustain itself by concerted action.
Transferred to the side of the dominated, a step is missing. A community whose carrier of cultivation is being destroyed takes up arms not because its capacity to act in concert is insufficient; precisely because that capacity is still there, and is being cleared away by an external force.
The mediating distinction to be added:
Violence that substitutes for concerted action — appearing where concerted action can no longer be sustained, as its replacement. Violence sustained by concerted action — itself a form of concerted action, whose organisation, mobilisation and persistence all depend on that capacity.
With this distinction added, her dichotomy is not merely preserved but sharpened: in the second case violence has not become power — power is still the capacity to act in concert, and violence is a form that capacity takes. The dichotomy is unbroken.
What SAE must preserve is the place of the second case, and Theorem One guards exactly that.
This is not pointing out an error on her part. She did not draw the distinction because it was not her question — her domain contains no category of "organised resistance by the dominated."
3.4 Kant: the temporal dimension, and he pushed it to a criterion
Proposition 3.6 What Kant gives in 1795 and in 1797 are two cross-sections of one thing, and the 1797 one is already a criterion rather than a diagnosis.
1795, reason six of the Sixth Article: the corruption of gui spills over into the state of peace and destroys the purpose of that state. This is a diagnosis — it describes something that will happen.
1797, Doctrine of Right §57 (AA VI:347–348): a state under attack may use any means of defence except those that would make its subjects unfit to be citizens — for then it would itself cease to be qualified, under international law, to exist as a person among states. The list of prohibitions that follows begins with using one's own subjects as spies.
This is already a criterion: it states a prohibition and gives the anchor by which to judge — the civic standing of one's own subjects.
Set beside §12, the complete shape has three parts, not two:
During the war (§12): to maintain an information gap, one's own people must be kept in the dark. The criterion (§57): whether a means may be used depends on whether it makes one's own subjects unfit to be citizens. After the war (1795, reason six): the practice does not vanish with the armistice; it has become how this community does things.
The middle part is what SAE lacks. SAE has the wartime diagnosis (§12) and a general judgement about cultivation, but it has never written "the standing of one's own executors" as an independent criterion on means. Kant did.
4 · One ground conflict, recorded but not developed here
Registration of the divergence: ground conflict — not internal failure, and not a mere "disagreement."
The registration requires a reason. Under the working rules, what touches the ground must be criticised; and under this series' registration discipline, a ground conflict should be named as such without detour. Heading this section "a disagreement" would be a mismatch of name and content.
4.1 The text
The Prince ch. 5 treats cities that formerly lived under their own laws. He considers three courses — destroy them; go and live there; let them live under their own laws, take tribute, and set up within them an oligarchy friendly to you — and then concludes:
He who becomes master of a city accustomed to freedom and does not destroy it, may expect to be destroyed by it; for in rebellion it always has the memory of its ancient liberty and privileges as a rallying point, which neither time nor benefits will ever cause it to forget.
He tried all three routes and conceded the other two were not secure.
The point of contact is unmistakable: what "destroy" is aimed at is precisely a community's capacity to reorganise itself — which is what the bridge clause constraining means, argued in §3 of this volume, exists to protect.
4.2 But this volume only records it, and the reasons must be stated
First, this volume's object is the user's side. The whole essay treats how a means reworks the hand that holds it — four mechanisms, four questions of test, all directed at the user. The problem in The Prince ch. 5 lies on the other party's side: what it cancels is the capacity of the conquered to reorganise, not anything of the conqueror's. This falls outside this volume's object.
Second, it is developed elsewhere. Volume II §4.2 treats it fully, including the location of the point of contact (he takes "holding" as a premise requiring no argument) and the qualification (this is not to say he is cruel).
Third, it must not be called internal failure here. Within the problem Machiavelli set himself — how to hold a newly conquered free city — his answer is coherent. The trouble lies in the premise of that problem, and a conflict of premises is a ground conflict, not a logical failure.
4.3 And this passage is itself the evidence for §6
A man who, in one and the same book, diagnosed the mechanism "those who live by war must wish for war" as completely as anyone has, advocated in another problem-domain the destruction of cities accustomed to freedom.
The two do not contradict, because they answer different questions. And that is exactly what §6 is about.
5 · The test the four compose
5.1 The list, and its criteria
SAE has a list no one has drawn up: the list of non-transferable techniques — which military techniques carry a structural consequence such that using them makes a party no longer a 15DD party.
This list can issue a determination on one side only, and that must be said first. A technique that trips any screening item can be determined to carry a structural consequence. A technique that passes them all can only be recorded as "this round of screening found none of the listed structural consequences" — which is not a certificate that a 15DD party may adopt it.
To write "which techniques a 15DD party may adopt" would exceed the strength of the screen. A limited screen issues no certificate; it can only record that nothing was found.
§12 supplied the first entry (gui), but by Corollary 3.2a that criterion is too narrow — it reaches only means that depend on an information gap.
Proposition 3.7 The four mechanisms set side by side yield four angles of test which together constitute the criteria for this list; SAE alone could not produce them.
5.1.1 But a ground gate must be passed before the four questions
Why this gate is necessary must be stated first.
Were the sequence "all four questions pass → declare structural neutrality," the substitution case of §2.2 would be let through — it passes all four while the child's standing is nonetheless cancelled (see §5.3).
That is the shape of "criticism was due and was not delivered." A screen for recoil on the user's side is run to completion and neutrality declared — while a ground conflict falls entirely outside the range of that screen.
The ground gate, to be passed before the four questions:
Does this arrangement cancel any particular person's standing as an independent end, or use them merely as a means?
Five trigger questions — if any is met, a candidate ground conflict must be registered, and reasons given for why it is not one:
- Is a particular person absorbed into the aggregate interest of a family, a state, a cause, or a community?
- Is their independent position to refuse, to withdraw, to revise, to answer, or to reorganise themselves cancelled?
- Are they used merely as a means — that is, does their own standing as an end not enter into consideration within the arrangement? (Substitutability is an external indicator of this condition, not its definition: a person who is not substitutable can equally be used merely as a means.)
- Does an external purpose acquire the authority to cancel their standing as a subject?
- Is consent, willingness, honour, duty, or community membership treated as an automatic exemption from the prohibition on using someone merely as a means?
Only what passes the ground gate proceeds to the four screening questions on recoil upon the user.
The division of labour between the two gates must be stated: the ground gate asks what this arrangement does to a particular person; the four questions ask what this means does to the party wielding it. Their objects differ, and neither can substitute for the other — treating the four questions as a complete test is precisely this error.
Run the §2.2 case through both: it passes all four (no information gap, no professional group, no violence, no ethos formed), and it trips the ground gate's first, third and fifth questions — the child is absorbed into the family's aggregate interest; his own standing as an end does not enter consideration (his entire function is that his face resembles another child's); and "consciously and willingly" is treated as an exemption from the prohibition on using someone merely as a means. With the gate in place, that case is stopped where it stands.
5.1.2 The four questions
| Question | Source |
|---|---|
| Does this technique require one's own executors to be uninformed? | SAE §12 |
| Does it require establishing a group that lives by it? | Machiavelli |
| Will its practice remain how this community does things after the armistice? | Kant |
| Does it substitute for concerted action rather than growing out of it? | Arendt |
The wording of the verdict must match the strength of the screen; no universal may be issued here.
Failing any one: the technique carries a structural consequence. This side admits determination — one hit suffices.
Passing all: only "on the dimensions covered by the ground gate and the four questions, no structural consequence was found."
Not "it is structurally neutral." The reason is given by this section itself: the ground gate's five trigger questions and the four screening questions are both non-exhaustive — the five were generalised from three known failures to criticise, and cannot stop a failure of a different shape. Two non-exhaustive screens both passing cannot support an unqualified determination.
This is consistent with §5.1.1: that section shows the four questions cannot screen for ground conflict and therefore adds a gate; and having added a gate that is likewise non-exhaustive, the conclusion can still only be qualified. To write "it is structurally neutral" would collide with that section's own argument.
5.2 Three examples
These examples serve only to test whether the criteria are usable; they do not constitute a complete assessment of any technique.
Wartime secrecy — the first question turns on depth. Secrecy such that the executor cannot judge what he is doing: fails. Secrecy that withholds only the timing and route from outsiders: passes. This entry could not previously be judged at all, and the four questions make it judgeable, because the first question breaks the vague term "secrecy" into a specific question about the executor.
General mobilisation — this cannot be determined to fail the second question. General mobilisation does not automatically establish a group that lives by war: a general levy may be a temporary participation by the whole population, disbanded when the war ends — which is precisely the citizen militia Machiavelli recommends, not the professional soldiery he opposes. Whether the second question is tripped depends on whether the mobilisation produced a standing group with nowhere to go afterwards, and that must be checked case by case; the phrase "general mobilisation" does not settle it.
Nor can the fourth question be settled by "conscript versus volunteer." That conscription substitutes for concerted action while volunteering grows out of it — that correspondence does not hold: a voluntary private force may equally substitute for public authority, and a conscription may equally issue from a legitimate common decision. The fourth question asks where the technique's efficacy comes from, not how participants were enrolled.
Cognitive warfare — the third question is triggered, but triggering is not determination. The capacity for information manipulation it establishes does persist after the war, and "the capacity persists" shows only that a risk exists; it does not show that the practice must continue. Kant's reason (a vice once encouraged cannot in the nature of things be stamped out) is his claim; this volume cites it as the source of the question and does not use it to declare the technique failed.
5.3 One limit
The four questions do not exhaust the field.
The substitution case of §2.2 — everyone informed, no information gap, no professional group, no violence, and no ethos formed — passes all four while the child's standing as an end is nonetheless cancelled.
So the four questions as they stand do not reach this class.
But one step further must not be taken. To say "this would require the second formulation itself to appear, not these four derived tests of it" would be refuted by this volume's own candidate two sections later.
The candidate proposed in §7.1 — replacing "is the executor informed" with "is the executor in a position to refuse" — is itself a derived test, and it does reach that case: the child had grown up under the lever of shame, and was not in a position to refuse.
So the accurate statement is: the four questions as they stand do not reach this class; whether a fifth question can supply it, see §7.1.
And the limit still stands as the current boundary of this volume's result — the ground gate of §5.1.1 exists for exactly this reason. This section is preserved unaltered, because it is the evidence for why that gate is necessary; the gap itself now has a treatment.
6 · A judgement about insight itself
Proposition 3.8 The insight that a means recoils upon its user is not sufficient to make the one who holds it stop.
The evidence is in one and the same book. In the Art of War Machiavelli gives the fullest version of this mechanism anyone has given — six reasons, from livelihood to incentive to constitutional trilemma, set out layer by layer. And in The Prince ch. 5 he advocates destroying cities accustomed to freedom.
The two do not contradict, because they answer different questions: the former asks how a republic should arrange its internal affairs, the latter how to hold a newly conquered city. He is coherent within each.
And that is the content of this proposition:
Insight tells you what will happen; a criterion tells you what you may not do.
Between them lies a stretch of road, and that stretch is walked on the ground, not on the depth of the insight.
A man who sees a mechanism with perfect clarity may, in another problem-domain, do what that mechanism forbids, so long as that domain contains no criterion linking the two.
This is a qualification on the four questions of §5: the criteria can tell a user whether a technique carries a structural consequence. They cannot make the user care about that consequence.
7 · Left unresolved in this volume
7.1 Corollary 3.2a finds that §12's criterion is bound to the single variable of the information gap, and the substitution case of §2.2 is a counterexample (everyone informed, standing nonetheless cancelled). Whether §12 needs a supplementary criterion not depending on an information gap is undecided.
Candidate: add a fifth derived test alongside the four — is the executor in a position to refuse? The candidate already has a concrete object of test: it must reach the §2.2 case, and it does (the child grew up under the lever of shame and was not in a position to refuse).
Its standing must be stated: this is a derived test, not a replacement for the ground gate (see §5.3). One who can refuse and does not may still be used merely as a means, so the ground gate cannot be displaced by it under any circumstances.
What remains undecided: "in a position to refuse" is harder to verify than "informed." The attribution of that difficulty is treated in this series' working note One Axis, Two Directions, §6: "whether this criterion can be broken into publicly contestable sub-propositions" is a question about the form of a criterion and stays within this series' range; "whether it can be broken down in wartime" is an execution judgement and lies outside it. ("Those who bear the consequences" is not a single subject; that phrasing was withdrawn: an execution decision is made by actors with the actual information and institutional responsibility; those bearing the risk retain a position from which to speak and refuse; outside critics retain standing to question.)
7.2 Proposition 3.6 finds SAE lacks the middle part — writing "the standing of one's own executors" as an independent criterion on means. Whether to add it on Kant's model is not settled.
The difficulty: Kant anchors his to civic standing (a status in Recht), while SAE's counterpart would have to be anchored to what? "Still exists as a subject of cultivation" is a broader formulation, but it cannot be identified from outside the way "civic standing" can.
7.3 The gap identified in §5.3 should not be supplied by generalising the fourth question. That gap is handled by the ground gate of §5.1.1, and the gate's object differs from that of the four questions (the former asks what is done to a particular person, the latter what is done to the user). Merging the two classes again would contradict §5.1.1.
7.4 The carrier-level indicator taken up in Proposition 3.3 — the proportion of the population whose livelihood is bound to a state of war — has not been tested for operability. Does a peacetime standing army count? Does a region whose employment is dominated by the arms industry? The boundary is undetermined.
Appendix · Sources and verification status
| Material | Edition | Location | Status |
|---|---|---|---|
| Kant, Perpetual Peace | Mary Campbell Smith, trans. (1903) | Sixth Preliminary Article, reason six | primary source verified |
| Kant, Doctrine of Right | Akademie edition, German | §57 (VI:347–348) | primary source verified |
| Machiavelli, Art of War | Italian original checked; English: Farneworth / Lynch | Book I | primary source verified |
| Machiavelli, The Prince | English translation (Gutenberg #1232) | ch. 5 | primary source verified |
| Nitobe Inazō, Bushido: The Soul of Japan | 1899, English original (Gutenberg #12096) | ch. on loyalty (the Sugawara story) | primary source verified |
| Arendt, On Violence | 1970, in copyright | — | paraphrased; not quoted |
| SAE War Treatise | Final version | §7, §11, §12 | primary |
Translations from the Chinese are the author's own.
One note: the six reasons given in the Art of War are entirely matters of incentive structure and social position, with no efficiency reason among them. This volume found no claim in that text that career soldiers are superior as a class, and therefore does not argue the force of the restriction from its "holding against efficiency."
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.
§5.1.1 of this volume is the plainest instance. The ground gate exists because a screen for recoil on the user's side can be run to completion and neutrality declared, while a case passing that screen has a child's standing cancelled. The failure is recorded there rather than smoothed over, because it is the evidence for the gate.