On the Colonization of Value
论价值的殖民

Introduction

P1 established the value-source; P2 established the positions of occurrence. For a concrete content to become a bearer's value content it must first take a candidate position in that bearer's field, and then pass first-person avowal. Social recognition lies on another axis: grantable and revocable, able to confer a public position and to take it back, and unable to claim on anyone's behalf. The three positions differ in modality, and an external action therefore has a determinate reading according to which of them it falls on.

That is the normal case. The same set of positions can also be violated, and violation has a structure of its own; it is not the normal case intensified. The object of this paper is the colonization of value: a content registered, in the bearer, as constitution already completed, where that constitution was not made in the bearer's own law.

What must be guarded against here is not under-judgment but over-judgment. Most value content arrives by the exogenous route — a way of living, a craft, a person, handed over by the world — and the bearer claims it in its own law, and it becomes its own. If the criterion counted external influence as such as violation, then upbringing, institutions, shared life, and cultivation would all fall in with it, and these are precisely what P2 spent a full section separating from colonization. The criterion must therefore be narrow rather than wide. Narrow does not mean light. It means that this word names one very specific thing, and not every external action that is hard to bear.

This paper adds no position of occurrence. What it establishes is the reading of that same set of positions when they are occupied, together with the conditions under which those readings can be audited from a third-person position and the conditions under which they cannot be, in principle.

How the accessibility of a source shifts once an external end occupies the positions of occurrence, and how the dominant weight moves, belongs to the dynamics paper. How to adjudicate when several claimed contents demand incompatible action belongs to the conflict paper. Why another's value cannot be exhaustively read belongs to the paper on others. How claimed value content moves into action is the opening of the prospect paper.

The text observes the standing discipline of the series: no external theorists are named, no historical events are adduced, and first-tier terms carry an English gloss at first substantive appearance.

---

Section I · The position colonization occupies, and its two branches

V27 The structural definition of colonization

Colonization is this structural state: a content stands, in the bearer, registered as constitution already completed, while that constitution was not made in the bearer's own law, and what occupies that position is an external end.

What is said is where the constitution was made, not where the content came from. A content handed over from outside, claimed by the bearer in its own law, has its constitution made in that law, and provenance does not enter this judgment. Conversely, a content that surfaced from the bearer itself falls under the criterion all the same if its standing at the constitutive position is not given by the bearer's own law. Source and locus of constitution are two different things.

The criterion falls on the constitutive position because P2 has already locked that position inside the bearer's own law. First-person avowal is the sole constitutive position: for a content to become this bearer's value content there is no other entrance. Social recognition can confer a public position, can support or suppress expression, can endorse a content already established — it has no constitutive power. Since the constitutive position is the only one there is, and since it cannot be legitimately exercised by proxy, an external end's occupying that position is a structural state that can be read off, and not a matter of degree.

What occupies the position is of two kinds, and they differ in character.

The first is an irrefutable attribution. A community says of some content that this bearer has already claimed it as its value, and the bearer's denial of that attribution cannot be registered as a denial within the relational structure. Attribution belongs on the side of social recognition, and by P2 it has no constitutive power; so nothing whatever is constituted here, only a reading taken as constitution.

The second is an external warrant. That this content counts as the bearer's value has its standing given by something outside the bearer's own sources: an authoritative pronouncement, an institutional arrangement, a reading issued by another's circuit of ends. This warrant corresponds to no constitution in the bearer's own law; it is merely taken as already there.

What the two kinds of occupation share is this sentence:

It does not succeed in constituting. It succeeds only in being taken as constitution.

Colonization never actually places anything at the constitutive position. What it makes is a reading — a content that has in fact passed through no first-person constitution is registered at the position of constitution completed. That the bearer's side reads it so as well does not mean the bearer assents in feeling; it means that the bearer's own architecture of occurrence likewise registers the content at that position. The position itself has not been taken away, and cannot be.

This definition brings three things with it.

First, the clause external end cannot be dropped. A person who declares something unrelated to any of its value-sources to be its value has, by P2, made a declaration that lands nowhere; that claiming is empty. But that is not colonization, only a failed claiming. Not every failed claiming is colonization. Failed, with the constitutive position occupied by an external end — that is.

Second, naked coercion does not enter this genus. An open system of coercion can compress a person's candidate field enormously, can leave his revision and withdrawal nowhere to land, and it never claims that this is his own value. The relation runs throughout as external requirement and has never attributed the command to the bearer as his value. The constitutive position is not counterfeited, and this criterion therefore does not hold. Such a form falls to P2's locking, and its reading in power relations lies elsewhere. The criterion narrows here not because that thing is light, but because its structure differs from what this paper treats.

Third, the form found at subjects-to-be holds as before. In a content-domain where the first-person position has not yet come to run of itself, declaring in advance of avowal that some determinate content is already its final value is, by P2, only external attribution or fictive recognition; and that is precisely being taken as constitution completed. The genus therefore holds.

The criterion accordingly runs on two layers, and the layers must not be run together.

The first layer is the genus itself: whether the constitutive position is occupied by an external end. It judges whether this structural state holds.

The second layer is the form in which that state can be audited within the relational structure. P2 has already indicated that an external end's occupying the positions of occurrence has at least three points of entry: appearance can be occupied, social recognition can counterfeit first-person avowal, the revisable position can be locked. Those three are exactly how the genus falls on the external branch, and this paper adds a threshold at the position of appearance so that the three together become a set of signatures that can be checked one by one. Signatures complete: the state is readable from a third-person position. Signatures incomplete: readability falls, and whether the genus holds is not decided by their completeness.

Layer	What it judges	By what
Genus	Whether this is colonization	Whether the constitutive position is occupied by an external end
Auditable form on the external branch	How that state is checked item by item in the relational structure	The three signatures: occupation, introjection, locking

The criterion judges structural effect, not the enactor's initial intent. What one action did to the bearer's positions — preserved, released, or cancelled — can be read off; with what motive it was undertaken, whether it was applied directly, through what medium it was delivered, alters none of that reading. It is the measure P2 used in judging cultivation and locking, and it is the same measure here.

Sorted by structural effect, external actions with a determinate release or closure fall into four cells. Non-shaping action, which neither releases nor locks, does not enter these four; it is the third class of P2's threefold division.

	Opening the occurrence-space of an identifiable receiving end is the constitutive orientation	Not so oriented
Release-ward	cultivation	evocation
Closure-ward	severing	ownerless locking

Colonization falls in none of these cells. It is not a class of action but a state; what brings it about and whether it holds are two questions. A deliberate direct enactment can bring it about; a distributed institutional arrangement can; a work together with a relation that reads that work as settled doctrine can; an atmosphere for which no one is answerable can. Whether the enacting end is identifiable, whether it is traceable, is irrelevant to the criterion. That is exactly why it must be judged as a state.

Material and occasion may come from any of the four cells; the colonized state itself is still decided only by the criterion. An action that genuinely enlarged the space of occurrence locally does not acquire a colonizing character in reverse because the content it supplied was later occupied elsewhere. A work evokes a candidate in the normal way; another relation subsequently arranges that candidate as an irrefutable attribution. The work supplied material; it was not a colonizing action.

Three layers must be kept apart here or the account will collide with P2. The layer of criterion uses structural effect and does not ask after intent. The layer of individuation divides cultivation from evocation, asking only whether there is an identifiable receiving end and an enactment oriented toward it; that too is a structural question. The layer of completion is another matter: cultivation is complete only when the receiving end's own step actually occurs, and the enactor holds no share in it. Each layer does its own work, and this paper alters none of them.

One last point. Colonization is possible precisely because the three positions are modally asymmetric. P2 established that asymmetry to found layered audit; the same asymmetry gives the point where counterfeiting lands. The asymmetry is not itself colonization; it is the structural precondition shared by colonization and by audit. Flattening the three into one would not confer immunity from counterfeiting; it would only abolish audit.

V28 Provenance does not judge colonization

That a content comes from outside does not make it colonization.

Most value content appears by the exogenous route; this is the normal case P2 established. Appearance does not ask where a content came from; it only brings the content into the candidate field. What follows is what matters: the bearer can deny, restate, split, and set a content back into another context. What it finally claims may be a fragment, or a rewritten version. Where the position of restructuring is still open, the content has not been occupied by a script.

The converse holds as well. The endogenous route grants no immunity. A content that surfaced from the bearer itself is likewise only a candidate, and must likewise pass first-person avowal to become its value content. And the suppression of that route, such that a person can take no candidate from its own running, is itself a form of occupation at the position of appearance.

Provenance as a measure therefore fails entirely: colonization cannot be inferred from an external source, nor immunity from an internal one.

This closes off a shortcut. A person likes something and says at once that this is his value; the step is taken too fast. The criterion does not fall on whether that liking was arranged for him — arranged likings are everywhere, and a work, a meeting, a shared life may all awaken one without anyone arranging anyone's value, and that is legitimate. The criterion falls on where the constitution was made.

V29 The division into two branches

Colonization as a genus has two branches, judged differently, not interchangeable.

In external colonization, what occupies the constitutive position is an irrefutable attribution. That attribution is doing constitutive work in the relational structure, which by P2 it has no power to do. It is third-person judgeable: whether the attribution is present, whether the bearer's denial can be registered as a denial — these are facts within the relational structure.

In self-colonization, what occupies the position is an external warrant. The bearer has performed the claiming of the content, while the content has never taken a candidate position under any of its own value-sources. This branch is discoverable only in the first person; in principle it has no third-person criterion.

Both occupants come from outside; the difference is not there but in where they lie. The attribution of external colonization lies in the relational structure, where anyone can see whether it is present and whether the bearer's denial can be registered. The warrant of self-colonization lies in this bearer's chain of standing; from outside one sees the pronouncement, not that it is holding up this content's standing. Neither branch's constitutive position is doing constitutive work, and only one branch's occupant can be read from a third-person position.

Whether the attribution is present, and whether the denial can be registered as a denial, are facts in the relational structure and can be read from outside. Reading them is not the same as judging: when the three signatures are complete this branch is third-person judgeable; when they are not, reading that an attribution is present can still only stop at cannot be determined.

The third-person criterion stops at the first branch, and this is not a temporary lack but the edge of the criterion's domain. There are two parts to the reason, and neither alone is complete. First, a failed claiming and a completed one show no third-person-accessible surface difference on a single slice of running. Second, what is readable appears only in the long-run distribution, and reading that distribution means reading where the premise of each recurrence lies — which means reading the bearer's own law, and no one can exhaustively read or write another's law by proxy. Taken together, the third-person criterion for this branch is in principle unavailable. This boundary and the upper bound of audit established in Section III are two sides of one line.

One point of wording must be fixed. Letting the position of occurrence go unused in operation is written here as disuse. P1 established that complete colonization is not transfer: the standing of a bearer cannot be transferred, and what can be given away is not called colonization but replacement. Transfer is reserved for that negative proposition and is not put to other use here.

Scope

Colonization has, elsewhere in this framework, a wider use: the locking of the operating conditions of a sequence. The two uses are not wide and narrow versions of one thing; they are readings from two sides.

The wider one reads from the side of the enactor and asks whether a sequence's operating conditions are maintained and enlarged or locked. The one here reads from the side of the bearer and asks whether the constitutive position is occupied by an external end. Each side can be read, each has its criterion, and the criterion of this paper settles only on the bearer's side.

On this reckoning, the excessive running or self-closure of a given layer — recorded elsewhere as the most refined form of colonization — does not fall to colonization by the criterion of this paper. On the enactor's measure it is indeed a locking of operating conditions; on the bearer's measure no external end occupies the constitutive position. It is what P1 cut apart as an active shutting: the bearer's own cancelling of a higher source whose conditions of application have in fact appeared. The two readings do not conflict; each holds in its own scope.

There is precedent upstream: the chisel of cognitive work and the severing judged in the theory of power share a character and not a scope, and are simply held apart without altering the word.

The enactor's side has a full reading of its own, with its own place. Whether a sequence's level of operation is raised or lowered requires reading what that sequence is running, which cannot be reached from a third-person position; and once raising and lowering enter a criterion, what is being judged is the enactor, which is another set of questions. This paper reads only their projection onto the field of value-occurrence, that is, the opening and closing of positions.

Both colonization and introjection have origins outside this framework. Only their structural reading is taken here.

---

Section II · Three actions at three positions

What Sections II and III establish is the auditable form of the external branch. Whether the genus holds is judged by V27; what is judged here are its signatures on that branch.

V30 The three actions occupy three positions

The three actions each fall at one position: occupation at the position of appearance, introjection at the constitutive position, locking at the revisable position.

There is an order of dependence among them and no necessary order in time. Dependence names the point of application: introjection requires a content already in the candidate field to attribute, and locking requires a content already taken as constituted whose revision can be voided. Priority in application is not priority in occurrence. The three may hold at once, may hold singly, may be brought about by mutually unrelated sources. One relation may in the same moment fill the candidate field, issue an irrefutable attribution, and leave refusal nowhere to register; one institution may do only one of these while the other two are supplied elsewhere.

One common path is worth stating once. The candidate field is first filled by a single grammar of evaluation; then the results of that evaluation are taken as the standard of self-judgment; finally the external goal and one's own goal can no longer be told apart. The path is indeed common, and the criterion does not fall on it. The criterion falls on the state of the three positions, and whether a particular case walked this order is irrelevant to whether it holds.

Each of the three actions is a positional state named after its characteristic action, not an act that must be performed by someone. Where an enactor is identifiable, one says that some arrangement has locked such a content-domain; where none is, one says that the revisable position in that domain has been voided. One predicate, two constructions, the same reading.

The criterion must carry an index. The three actions do not act on the same thing: occupation acts on the candidate field, introjection on the attribution of a particular content, locking on the revisable position. Without stating over what range they are complete, one cannot say whether they are speaking of the same thing. The unit of the index is the one P2 already used in judging locking: the content-domain.

The three must also share the index. That a school renders some candidates in a content-domain invisible, that a family issues an irrefutable attribution about another content, that an institution locks a third class of revision — these three do not compose a colonized structure merely by falling, as it happens, in one content-domain. The three must jointly serve the attribution of one content under one grammar of ends before the signatures count as complete.

The grain of the content-domain must be given in advance. If the range can be rescaled after the fact, the criterion loses its power to audit: narrowed to a single content, nearly any push counts as occupation; widened far enough, nearly any colonization fails to meet occupation. The content-domain must be given independently by the bearer's existing structures of practice, by one of its value-sources, or by a stable addressable field of questions, and must not be recut in order to obtain a particular verdict.

Signature and state are not one thing, and this is clearest when a single signature is present. An irrefutable attribution is the signature of introjection and is third-person readable; introjection is a state, namely that the constitutive work is being done by the attribution. The signature is necessary and not sufficient. An irrefutable attribution admits two readings: it may be occupying the constitutive position, or it may be an echo — the bearer having claimed the content in its own law all along, with the community saying so afterward, the attribution doing no constituting but merely restating something already done. To tell these apart one would have to know whether the bearer's own law independently made that constitution, and that lies beyond the upper bound of audit.

The reading is therefore asymmetric. Signature absent: this branch does not hold, and that judgment is determinate — no attribution is occupying the constitutive position. Signature present with occupation or locking incomplete: the third-person position cannot determine; and cannot determine is not a determination of some weaker form.

Here lies the real office of occupation and locking. They do not aggravate the state; they close the gap between signature and state. For the bearer's own law to have made that constitution independently, candidates issuing from its own sources must have been available, and revision must have been retained. With the candidate field filled by a single grammar and the revisable position voided in advance, the echo reading has no room left in the present. Three signatures complete, and the reading is therefore determinate.

Position violated	Action at it	What a third-person position can determine when this signature alone is present
Appearance	occupation	No attribution present: this branch does not hold. Falls to locking or narrowing by P2
Constitutive	introjection	Cannot determine. Occupying the position and echo both stand as readings
Revisable	locking	No attribution present: this branch does not hold. P2's locking

No single action is called colonization, and no partial combination is given a new name. Occupation and locking together with no attribution present is naked coercion, which does not enter the genus at all. The other two combinations each contain an attribution and therefore both stop at cannot determine; two signatures present does not confer a determinate reading. Making new names for these would only dilute the criterion: either they already have a place in P2, or there is as yet no reading to name.

That there are no new names governs partial combinations only. Complete forms are typed by the character of the content-domain and are not so governed: a domain where the first-person position has not yet come to run of itself gives pre-occupation, and the domain of the bearer's own standing as a value gives self-colonization. Each has its name and its manner of judgment.

V31 Occupation

Occupation is the candidate field being arranged from outside to this degree: in a given content-domain, the bearer can no longer obtain a candidate issuing from any of its own sources. There are two ways to reach the threshold. One is the endogenous route being suppressed, so that in this domain nothing surfaces as a candidate from the bearer's own running. The other is the exogenous route being filled by a single effective grammar of evaluation, so that candidates not issuing from that grammar cannot enter. Both point at the same thing: in this domain the available candidates all issue from one place, and that place is not the bearer.

This threshold belongs to the position of appearance and is not borrowed from locking. Locking asks whether refusal, restructuring, and revision have any room left; occupation asks whether candidates have a second point of issue. The two can hold independently.

The threshold clause is part of the criterion. Arranging what a person sees first is neither strange nor harmful in itself. A good demonstration likewise arranges what one sees first; a skilled hand sets a craft before you, a work brings a way of living into view. These fall on the opposite side from occupation. The distinction is not whether there is arrangement but whether the space is larger or smaller afterward. Where appearance, accessibility, expression, revision, and the space for bridging are maintained or enlarged, the action is release-ward shaping; where they are pressed beyond reach, it is occupation. Showing you a great deal is not occupation. Keeping you from seeing anything else is.

Filling does not require a single source. Many distributed sources may share one effective grammar of evaluation; they may have no coordination whatever, and the effect is still that the candidate field carries only this one scale. The criterion asks how many effective grammars there are, not how many issuers.

The two routes each have their form of being occupied. The exogenous route may be pushed as a complete script, so that every candidate a person meets issues from one place; or it may be cut off, so that a class of candidates never reaches him. The endogenous route may be suppressed, so that in this domain nothing surfaces from the bearer's own running.

The two must not be run together. Suppression of the endogenous route is the locking of a condition of appearance; it is not avowal written by proxy. That a person can obtain no candidate of its own, and that a person is told what he cares about, are two things at two positions. The former falls at the position of appearance, the latter at the constitutive position.

V32 Introjection

To judge whether an attribution is counterfeiting constitution one need not first know whether first-person avowal occurred. That the first person cannot be exhaustively verified from a third-person position is locked by P2. What can be asked is another question: in this structure, on what does this content count as the bearer's?

The question has two answers. One: the bearer claimed it, and that claiming has stood throughout at a position it can itself question — the constitutive work is being done by the first person. The other: it was attributed, and the attribution is doing the constitutive work. The latter is introjection: a community's evaluation of a content arranged as a value this bearer has already claimed.

Introjection is a state. Its signature is another matter: no response by the bearer can be registered as a refutation of that attribution. The signature is third-person readable, the state is not, and the relation between them was given in the preceding proposition.

Three things follow.

First, the criterion does not ask whether the attribution is true. An attribution may be entirely correct: the community says he values this, and he does. That does not alter the finding. Whether the constitutive work is being done by it or by the bearer's own law, and whether what it says is correct, are two questions.

Second, the signature judges a position, not an event. What is to be read is not whether the bearer offered a denial, but whether this structure has a position at which a denial could land as a denial. This bears on whether the criterion can be used at all: at the deepest reach of colonization the bearer offers no denial, having nothing on hand to deny with. Were the signature tied to the event of denial, it would never trigger where it is most needed. Whether the position exists is readable in the present.

Third, an attribution need not be uttered by anyone. It can be carried by the relational structure. An arrangement that treats a metric as what a person counts for needs no one to speak in order to be issuing that attribution continuously; and that person's refusal of the metric has no form within the arrangement in which it could be recorded. An atmosphere for which no one is answerable can do this, just as it can do locking.

The exact wording of the signature is the irrefutability of the attribution, not its unwelcomeness. An attribution a bearer dislikes, even resents, but can contest, is not introjection. Contestable means the dispute has an outcome: his denial is recorded, answered, and may change the attribution's standing. Irrefutable means the denial has been given its destination in advance, so that whatever he says is rewritten as something else.

The signature has several common forms, of differing purity. Dissent rewritten as betrayal or selfishness is one: a person's refusal of a collective goal is read throughout as his thinking too much of himself, and refusal can no longer exist as refusal. A purer form is denial pre-coded as lack of insight, where a person's denial is counted as evidence that the attribution holds — the more he denies, the more it is confirmed. At that point the position of rebuttal is not suppressed but abolished: any response can only vindicate the other side.

The latter form requires a boundary of its own. What falls under this judgment is not the judgment being made. An assessment, a diagnosis, an external description — so long as it leaves the bearer the positions of refusal, review, second opinion, and restatement of his own situation, those positions remain open, however unwelcome or humiliating the finding may be, and it does not fall here. What falls here is the step of pre-coding denial as symptom, because that abolishes exactly those positions. This is of a piece with what P2 established in judging locking: an offensive action is not the same as a closing action, and what is judged is the opening and closing of positions, not how the thing feels.

V33 Locking

Locking follows the definition and threshold P2 established; this paper does not re-establish it, but marks only where it falls in colonization: the revisable position is voided in advance within the relational structure.

The threshold reads as follows: a content-domain is systematically excluded to the point where the relevant candidates cannot enter the space in which the bearer reflects, restates, and revises; or the bearer's refusal, restructuring, and revision in that domain are voided in advance within the relational structure.

The three classes cannot be pressed back into two. P2 established a threefold division: locking, release-ward shaping, and non-shaping action, which neither releases nor locks. Harmful is not the same as locking. A narrowing, a pressure, an unpleasant action which has not reached that threshold falls in the third class and has its own place. This must be held to especially here: what it means in practice for the criterion to be narrow rather than wide is, in good part, that the third class is not swallowed.

Locking need not have an enactor. P2 established that even where no one has laid a hand on a person's revisable position, that position may already have lapsed; an evaluation no one deliberately promotes, an atmosphere for which no one is answerable, falls on the side of locking.

One point must be made plain. Locking is listed in that section of P2 as a class of external action, while its criterion is a threshold state of the relational structure. What this paper carries over is the form of its criterion, not its place in that table of classes. Running the two together would yield the reading that colonization is a class of action, and colonization is not a class of action.

Locking and introjection are not interchangeable; what each seals is orthogonal to the other. Introjection strips the standing of denial: structurally, one may even be permitted to speak, to say anything at all, while everything said is rewritten as something else, and the constitutive work is thereby taken over. Locking seals off the space of operation: the paths of reflection, restatement, and revision are blocked outright, and whether one may speak is a separate matter. A content may be irrefutably attributed while the revisable position remains open; a content may be claimed by the bearer in its own law while its revisable position is thereafter voided from outside. The former is not locking, the latter is not introjection. Each of the three positions is judged on its own, and that is the precondition of their composing a set of signatures.

---

Section III · The criterion and the audit

V34 Adjacent forms, two forkings, and a general rule

With the three signatures complete, a third-person position can determine external colonization in that domain. Where one is missing they are incomplete, and the incomplete readings are asymmetric: the attribution absent, this branch does not hold, and that judgment is determinate; the attribution present with occupation or the revisable position incomplete, the criterion stops at cannot determine, and no verdict of absence follows from it.

The incomplete forms each have their place, and two of them lie nearest the boundary of this genus.

The first is naked coercion. Occupation and locking both present, no attribution: a person's candidate field is compressed to a sliver, his refusal and withdrawal have nowhere to land, and no one has ever said that this is his own value. The relation runs throughout as external requirement. The constitutive position is not counterfeited, and the genus therefore does not hold. Such a form falls to P2's locking, and its reading in power relations lies elsewhere. It may be far harder to bear than many a colonization, and the criterion still does not change its finding: the degree of what is unbearable is not a criterion, the state of positions is.

The second is a genuine claiming that has been locked. A content is indeed claimed by the bearer in its own law, and its revisable position is thereafter voided from outside. That constitution was made in the bearer's own law, and what occupies the position is no external end; the genus does not hold either. It is the form in which locking falls on an already claimed content. P2 has already judged that at this point the content is no longer there in the manner of a value content, since revisability is written into constitution itself. Its gravity is not lessened by its not being called colonization.

What divides these two from introjection is whether an attribution is present and arranged as irrefutable. In a locked genuine claiming no one is telling the bearer that this is his value; what is sealed is the revisable position, not the constitutive one. In introjection an attribution occupies the constitutive position and goes through the motions of constituting. This division does not conflict with the criterion's not asking whether an attribution is true: the former asks whether an attribution is present and doing constitutive work, the latter asks whether that present attribution says something correct. The first question enters the judgment; the second does not.

One further form is worth marking, since it is often misread. A person lives according to some set of requirements, and what sustains that compliance is no claimed content of his but the cost of departing from it. Those requirements run throughout the relation as external requirement and have never been attributed as his value; only the cost of departure keeps him from departing. No attribution has done constitutive work in him; the genus does not hold, and it falls on the side of locking.

The criterion cuts two everyday classifications in half.

The first is rule. Rules cover every point of interaction and informal room for judgment vanishes; occupation and locking are both present. What is then to be examined is the third item: whether an attribution is present that says following the procedure is his own value, and is doing constitutive work. If so, the three are complete and the reading determinate. If not, this is merely working to the book, the constitutive position is not counterfeited, and it falls on the side of locking.

The second is the fixing of a narrative. A narrative of who one is is treated as an unalterable frame of identity. If that narrative was handed over from outside, is taken as constitution completed, and has nowhere to be revised, the three are generally all present — for an unalterable narrative of identity itself keeps other self-descriptions from entering, and occupation follows — and the reading is determinate. If it is the bearer's own narrative gone rigid, with no external end occupying the constitutive position, it does not fall here; it is the bearer's own cancelling of a source still applicable.

These two forkings are not gaps in the criterion but the consequence of two indices crossing. Classifying by the vehicle of intrusion and classifying by which position is violated are mutually orthogonal axes. The names given by the former remain valid; each of them identified a real form. What this paper supplies is the criterion those names require. Under one vehicle there can be two positional states, and the criterion therefore forks there.

The three signatures are one cut this series undertakes, not the only possible cut. Removing any of them as a condition of determinate reading does not yield a wider genus — the genus is still judged by the constitutive position — but a wider use that cannot be determinately audited by this criterion. Those cuts are each on record; this paper does not judge them mistaken, only marks that it does not take them and what taking them would cost: once the audit is loosened that far, ordinary upbringing, institutions, and shared life are no longer separable from colonization, which is precisely why the word is worth using precisely.

From this a general rule can be established, and it blocks more misuse than the criterion itself.

Revision and change are never evidence of colonization. On the contrary, they are evidence that a position is open. A person's centre of value has shifted, something once cherished has faded, something else has surfaced — these are mostly content-dynamics, the ordinary condition of a value life rather than a rare structural event. To take that someone has changed as a sign that he has been colonized is to read the ordinary as the pathological.

The rule carries two qualifications.

The first bears on self-colonization, and it is not an exception to the rule. The rule says that change proves a position open, and that sentence holds on that branch as well: the revisable position there is indeed open, and the bearer is indeed revising throughout. What does not hold is the further step, from the position being open to that branch not holding. That branch does not work by sealing positions at all; it works by the standing resting outside, and so a position being open and its holding are two different things.

From this a qualification follows, and it is content-specific rather than field-specific: a person's candidate field may be very wide, he may change freely among many candidates issuing from his own sources, while the standing of one content among them is carried throughout from outside. Rewriting that one content, however many times, does not prove that the standing has moved. If it should later take a position under one of the bearer's own sources, what that shows is that the state has now changed; it does not prove in reverse that it was his all along. This clue, too, is available only to the bearer itself: self-colonization has no third-person criterion, and no one may turn it around and judge that branch's state from whether a person has changed.

The second qualification bears on time, by the same discipline. A successful revision proves that this position is open now, not that it was never sealed. That one can change it later shows that the state no longer holds; it does not erase in reverse the reading of an earlier moment.

V35 The upper bound of audit

Audit reads only the opening and closing of positions; it does not judge what the bearer really avows.

Three things can be asked, one for each position. In this domain, can the bearer still obtain a candidate issuing from one of its own sources. For a given attribution, can the bearer's denial land as a denial. Are refusal, restatement, revision, and withdrawal voided in advance within the relational structure.

All three are facts in the relational structure and third-person readable. Audit reads all three, and the three readings are not uniform. Three complete, and the reading of external colonization in that domain is determinate. If the bearer's denial can land as a denial, then no irrefutable attribution is present, this branch does not hold, and that is a determinate negative. If an attribution is present while occupation or the revisable position is incomplete, the audit stops at cannot determine: it can neither rule out that the attribution is only an echo of a genuine avowal, nor conclude absence from that. Cannot determine is itself a legitimate conclusion, saying that the third-person position reaches to here and no further. To take incompleteness as a determination of some weaker form, or as a determination of absence, is in either case to overstep what one can read.

Reading past those three into the first person is an overstepping of another kind, and its unworkability is not a matter of modesty.

An audit that claims to see what a person really avows places an attribution about the bearer at a position the bearer cannot check: if he accepts it, the conclusion holds; if he denies it, the denial is read as further evidence for it. Where such a conclusion is unaccompanied by positions for review, appeal, and counter-audit, what that relation structurally generates is precisely the kind required for introjection. This is not to say that a single utterance constitutes introjection — an utterance is an event and introjection is a positional state — but that an audit relation of this kind tends to arrange attributions as irrefutable, and that is the very form under audit.

The upper bound of the audit's authority and the criterion of introjection are therefore two sides of one line. This bound is not an added modesty; it is a consequence of the criterion itself.

Accordingly, an audit's conclusions can take only the form of positions and signatures: such a position is closed, such a channel is still open, such a signature is absent, or the three are not yet complete. They cannot take the form of the truth or falsity of content: that this person's value content is a forgery. The latter requires reading the first person, and the first person cannot be exhaustively verified from a third-person position.

This restriction on form does not leave audit with nothing to say. It settles only which sort of thing may be said. Your denial has no position in this arrangement at which it could land; this procedure leaves no channel for restatement, review, and withdrawal; in this domain every candidate that gets in issues from one place — these all name the state of positions, and all may be said. What you really care about is not this — that may not be said; it requires reading the first person. The line between the sayable and the unsayable falls here, and not on how gentle or harsh the words are.

---

Section IV · Self-colonization

If a subject takes another's list of uses to claim its own value, its position of value-occurrence has already been occupied by that list, and thereafter, when it looks at itself, what it sees first is always the self that can be used. When the prequel established that sentence it left an unanswered question: how does such a list come to occupy that position, and what is the state once it has.

V36 The structural mark of self-colonization

Self-colonization is this state: the bearer has performed the claiming of a content, while that content has never taken a candidate position under any of its own value-sources; the standing that makes it count as the bearer's value is carried by something outside its own sources. By P2, such a claiming lands nowhere; and it is nonetheless taken as constitution completed, and maintained at the constitutive position.

To read this correctly one must first see what is missing.

What is missing is not permission at the level of type. For a bearer unfolded to 15DD, all four rows of value-source are present, and a new row does not replace an old one. Nearly any class of content can therefore find a cell among those four at the level of type: that something is pleasant is an entirely permissible class of content on the row of the orientability of experience.

What is missing is discovery.

To know that there is such a value, and to discover that one has this value, are two things. The former belongs to externally visible fact: that some content is expressed, recognized, evaluated, or arranged by someone as valuable — this can be observed, and it involves none of one's own rows. What is observable stops at the representation: whether that content has completed its first-person constitution in another bearer is not among the observable, since another's value likewise cannot be read through. The latter concerns this bearer: whether the content has taken a position under one of its own rows. The word that is missing is own.

What this paper means by discovery is only the latter, not the finding of a fact already lying there. That a content takes a position under one of the rows means that it has appeared, within this bearer's running, as a candidate under that row; not that it has been verified as something already in existence.

Two things can be told apart at once on this branch. Discovery has not held: the content has never taken a position under any of the bearer's own rows. And in its place there is something else: an external warrant, that is, something from outside the bearer's own sources holding up the standing that makes it count. Such a warrant may be a reading issued by another's circuit of ends, may be an institutional arrangement, may be an authoritative pronouncement. The warrant is in part observable; whether discovery holds is not.

Whether that warrant is true does not enter this judgment. It may be accurate: the value does hold elsewhere, the pronouncement is not false. It may be inaccurate: a misreading of another's ends, a formulation no one actually holds and that is merely reproduced, a representation that everyone values this. The criterion does not ask whether it is right, only whose source carries that standing. This is of a piece with introjection, which likewise does not ask whether the attribution is true.

What stands between warrant and source takes no fixed form. It may take the form of an inference: since this value holds, I too ought to have it. It may take the form of credence: this was told me by another or by a third party, and I respond accordingly. It may be the taking up of an assigned position: they need me to do this, so this is where my use lies. Form does not enter the judgment; the criterion asks only where the standing stops. Credence in particular must be held to this: to give credence to something, while the content has separately taken a position under one of one's own rows, is the ordinary case of occurrence; to give credence while the standing rests only on that pronouncement is what falls here. The words ought to are the idiom of that step itself: whether there is a position under a source is not a matter of what one ought to have.

Alongside these stand two further things which likewise produce no position under a source, and which do not fall under this judgment. Declaration produces none: I say this is my value, therefore it is — this step does not hold. Time produces none: I have held it long enough, so it must count as mine by now — this step does not hold either. These two are what the anti-subjectivist lock of P1 and P2 blocks; there is no external end in them at all, and by the definition of the genus they do not enter it. They are merely failed claimings. One position has two sides: that lock blocks the bearer's issuing itself a standing, and this branch blocks something external carrying it by proxy. Only the latter side yields colonization, because only it has an external end present.

P2 had in fact already left this seam. Appearance is a content's taking a candidate position in the bearer's field, and it does not ask where the content came from; what avowal requires is that the content have taken a candidate position in some legitimate value-source-domain of the bearer. Field and source-domain are not the same range, and the difference between them is where this branch lives. Discovery and appearance are therefore not the same thing: appearance is in the field, discovery in the source-domain. Nor is this a new position on the axis of occurrence. There is only one candidate position; there is a further question of whether that candidate stands under one of the bearer's own rows, and P2 had already written that question into the precondition of avowal. What this paper calls under one of one's own rows and what P2 calls in one of the bearer's value-source-domains are the same thing, said once along the figure of four rows and once along the range.

Each of the four rows has its form. What the table records is not to whom the content itself belongs, but whose source carries the standing that makes it count as mine. One and the same content may perfectly well take a position independently in two bearers; sharing a content is never a ground for this judgment.

Value-source	The form self-colonization takes on this row
11DD and 12DD, the orientability of experience	Another's orientation of experience borrowed as the ground of what is orientable for me
13DD, the irreducibility of the self and the integrity of its boundary	The manner in which another's boundary holds borrowed as the ground of mine
14DD, one's own posited ends and their abundance	Another's ends borrowed as the ground of my positing ends
15DD, the irreplaceability of another as an end	That my responding to another counts as my value — this standing carried by another or by a third party

The list of uses is the commonest form on the 14DD row, not the only one.

The 15DD row is structurally unlike the first three. The content of this row is another to begin with, so the borrowing of the first three does not go wrong on the spot; it is the row least apt to go wrong. What this row is anchored in is not another's content of ends but the fact that another's standing as an end, and the first-person source of its own value, cannot be given, written, or replaced by me.

Another's value likewise cannot be read through. What the bearer does on this row is therefore never to read correctly what another wants, but, on the footing that this cannot be read through, to form in its own law a figure of this other: that it is an end, what it wants, what my responding so to it amounts to. A bearer unfolded to this row is the one who has this figure to make, and a figure is what it is: it may be well made and is still made on that footing.

Two things must be kept apart here, or the line of this judgment will be drawn in the wrong place.

The one is what another wants. On this the other has the say: it occupies, in respect of its own ends and boundaries, a first-person position that no other may write by proxy, and its own account is therefore the irreplaceable input to this figure. Taking that account is not contracting the figure out; on the contrary, reducing another's account uniformly to something I must still guess at independently is to occupy another's first-person source, which is exactly the writing-by-proxy this row forbids.

The other is what my responding so to it amounts to as my value. This one no one can do in my stead. Another may tell me what it wants; it cannot give me, in my stead, the standing on which my response counts.

The failure of this row therefore falls on the latter: the requirement, approval, or interpretation of another or of a third party is borrowed as the standing on which my response counts as my value, so that the bearer must obtain from an external circuit of ends the warrant that it therefore counts. What the external warrant holds up is not the content of the figure but that it counts as my value.

Two sentences must be made plain.

First, taking another's own account is not colonization. It is the first-person input that no third-person material can replace. Conduct, circumstance, history, and unspoken expression remain legitimate materials, especially where another has not yet spoken, cannot speak, or where the thing is not sayable at all, in which case the figure must still be formed by the bearer itself; these materials cannot override another's own account, nor counterfeit another's first-person source. That it cannot be written by proxy is a claim about position, not a claim that another is never mistaken about itself.

Second, guessing wrong is not colonization. The figure a bearer forms of another may be inaccurate; the revisable position is open, and it can be corrected next time. An inaccurate figure is still its own figure, just as an entirely correct inference still produces no position under a source. What the criterion blocks is never whether the figure is accurate, but whose source carries that standing.

The failure of this row has two directions. In one, the standing is carried from outside, and that falls here. In the other, one presumes to have read another through, taking one's own figure for the other's first person, and that is introjection practised toward another. The line between them is that another's value cannot be exhaustively read, and the paper on others establishes it directly.

This criterion has three boundaries, and cutting away any one of them would injure what should not be cut.

The first separates it from the other-directed contents of 15DD. To hold another's peace dear, to hold a shared undertaking dear, to hold care dear — these contents are full of another, and do indeed require looking toward another again and again, because another is their object. The distinction lies in direction on the figure of sources: on the 15DD row the bearer is the end issuing the end, taking another as an end; in a reading of use the bearer is the end being used, another taking it as a means. The same matter with reversed direction, and the cell is therefore different. Reading the direction suffices; no thought-experiment of removing the other is required.

The second separates it from not yet being rooted. A content has taken a position under one of the rows while this instance is still new, not yet having appeared repeatedly in the bearer's running — this is the ordinary case of value-occurrence, not a pathology. Were the criterion to swallow this class, every new claiming would become suspect, and the rule that revision and change are evidence of a position being open would lapse on the spot.

The third: not every failed claiming is self-colonization. Claimed, discovery not holding, and then faded — that is only a failed claiming. What this judgment requires is that discovery not hold while it is nonetheless maintained at the constitutive position.

This criterion is narrow, and its cost lies in the open. An end that does grow under the bearer's own 14DD, though someone else first pressed it upon him, does not fall here, since provenance does not enter the judgment. A person whose running is highly contingent on the regard of others, while what he holds important has itself taken a position under one of his own rows, does not fall here either. Self-colonization is not being deeply influenced; it is taking a cell on someone else's figure for one's own. Much of what everyday usage calls losing oneself does not fall on this branch; those things have places of their own.

There is a lock here doing a second office. Anti-subjectivism originally blocked declarative claiming: I say this is my value, therefore it is my value — that step does not hold. The same lock here blocks the opposite direction: I have held it long enough, so it must count as mine by now — that step does not hold either. The two are two faces of one thing. Repeated claiming produces no source.

V37 The reading in the course of running

A failed claiming and a completed one show, on a single slice of running, no third-person-accessible surface difference. The position of claiming has taken its reading, and the revisable position is open. The structural difference is there throughout — that is what the preceding proposition established — but on no slice of running does it show a readable surface.

The judgment on this branch therefore lies wholly inside. The third-person position reads no surface on a slice, while what is readable requires following the bearer's own law, and no one can exhaustively read or write another's law by proxy. There is thus no entrance here, rather than a temporary failure to see; audit on this branch cannot even reach cannot determine, which is at least an audit conclusion.

Nor can the judgment be contemporaneous. This branch requires that the external warrant be taken as constitution completed, and taken as contains within it that one has not, at this moment, seen it to be external. Should a person see in the moment that this is not something under his own sources, that taking-as fails on the spot, and the state is thereby dissolved. What can be seen through in the moment is already not the state this judgment names.

On this point the two branches are exactly opposite, and the opposition is a direct consequence of their occupants differing. In external colonization the bearer may know perfectly well that the attribution is external, and knowing does not dissolve it, because what is sealed is the revisable position. In self-colonization knowing dissolves it, because what is being held up from outside is precisely that standing, and once it is seen to be held up outside, the bearer's own law is already running that step.

What remains, then, is retracing. The bearer afterward, at a distance, asks again on what this content is his, and follows the account backward: does the chain of standing finally stop at one of his own rows, or at an external warrant. This is not introspection in the present; it is a reckoning with oneself across time.

What is read is the resting point at that moment, not the outcome afterward. This distinction matters, because outcomes cannot prove it. That a content can later be recovered from one of one's own rows shows that by then it has taken a position, and that is one more thing having happened; it does not prove that it was there at the time. Conversely, that it cannot later be recovered does not prove that it never had a position: by the permanence of positions, the presence of a source and the accessibility of a source are not one thing, and failure to recover may well mean only that accessibility is still sealed. Whether recurrence occurs is therefore only a clue, and not a necessary and sufficient condition of this judgment.

What retracing reads is still structure, not feeling. That the judgment on this branch lies inside does not mean that the inner self-report decides it. A person's feeling at this moment that this is mine does not make it his; feeling that this is not mine dissolves no claiming. Likewise an account given afterward is not automatically true: a person may compose a clean story for himself, and may also disown as external what did grow under his own sources. The privilege of the first person here is being able to follow its own law backward, not that its self-report is final.

This is also why this branch can only be read at long range, and the reason is structural rather than a matter of data: the present moment is excluded, and the moments available lie afterward.

From this comes a readable clue: the route of recurrence. When a content has faded and returns, does it return as a candidate under one of one's own rows, or must an external warrant be obtained afresh before it can return. This clue is available only to the bearer itself.

One thing must be made plain, lest this passage be taken for a procedure that can be executed. What is asserted here is only a structural dependence, not any ongoing act of drawing upon something. An external warrant may perfectly well be remembered, made into habit, and thereafter recur automatically, requiring no external input in reality and no longer running anything explicitly. That does not alter this judgment: the source remains outside, while the running need not be. Automatic, effortless, persistent — none of these moves the resting point of the chain of standing.

Time does not move it either. However long a content is held, that standing will not thereby migrate from outside to under one's own sources, because repeated claiming produces no source. This is therefore not something that will settle itself with time: if that content should at last, in this bearer's running, take a position under one of its rows, that is another thing having happened, not this thing having accumulated to some number.

And there is no test. Every check that could be set out and performed proves something else: whether it is effortful proves nothing, how much is drawn upon proves nothing, the outcome after a warrant is withdrawn proves nothing. What remains is retracing itself, the first person following the account backward to where the chain of standing stops — at one of its own rows, or at an external warrant, one of the two. This is not a procedure. It is a tracing back.

Four things are to be held to in this reading.

First, standing is not cost. What one pays for a value may be enormous: a person who truly holds the care of his family dear pays immensely, and that is a cost in action, unrelated to this judgment. What is spoken of here is another thing, namely what carries the standing that makes this content count as mine. The payment may be very high while that standing has not once rested outside.

Second, whether it takes effort is not the criterion. An external warrant may stand ready, cheap to draw upon; it may equally have been made into habit long since and need not be drawn upon at all. That it takes no effort does not prove the state does not hold, and that it takes effort does not prove it does. What is readable is the resting point of the chain of standing, not the ease of drawing upon it.

Third, another cannot issue this reading. This is not a prohibition but something that cannot be done: reading it means following the bearer's own law backward. A person who says that this was never your own has not determined self-colonization — he cannot determine it. What he has brought about is another thing: an attribution the bearer cannot check has been handed down, and that is the very relational structure external colonization requires. The two branches interlock here.

Fourth, doubt does not dissolve claiming. Present doubt especially does not: present determinability and this state are mutually exclusive, so a moment's hesitation now neither establishes this judgment nor dissolves any claiming. What would move is the resting point read in retracing, not the feeling of that moment.

The place of the exit is thereby fixed, and it is not the same thing as the dissolution spoken of above. Dissolution says that the state no longer holds: once it is seen that this standing is held up outside, the content falls back to the status of a candidate and is no longer taken as constitution completed. Whether it thereafter becomes the bearer's value is a further step, requiring that it take a position under one of its own rows. Dissolution does nothing toward that step; it only vacates the place.

The exit falls on that second step. Since every self-questioning proceeds under the same warrant, the exit does not lie on the side of self-questioning; nor does it lie on the side of the auditor, for the reason just given. The exit is that the content have the chance to take a position under one of the bearer's own rows, or that a candidate not generated by that warrant take a position. Such a candidate may come from an external and heterogeneous perspective, and may come from the endogenous route. And everything after a candidate takes a position remains wholly with the bearer itself. How accessibility is in fact recovered in time is the object of the dynamics paper.

---

Section V · Pre-occupation at subjects-to-be

V38 Pre-occupation

In a content-domain where first-person avowal has not yet come to run of itself, external colonization takes another form, called here pre-occupation. It belongs still to the external branch, the criterion unchanged; only the landing point of each of the three actions shifts with the state of that position.

Introjection here is not the displacing of a first-person position already in existence. Subject-position recognition, by P2, presupposes that the bearer already has first-person avowal; in a domain not yet running of itself, such recognition is not yet defined. What an attribution issued here does is take a determinate content as constitution already completed, where that constitution never had the chance to occur. Locking here falls on the voiding in advance of the future revisable position: an arrangement such that, by the time that door comes to run of itself, some content can no longer be restated or revised. The three actions together point at a position not yet running, so that at the moment it begins to run it is already filled.

The judgment is by domain and not over the whole subject. One and the same subject-to-be is judged by the complete-subject form in domains that have come to run of themselves, and by pre-occupation in those that have not. Maturation never arrives in one piece; P2 established this.

How a future revisable position is voided in advance is readable in the present by three signals. First, whether there is an arrangement transferring the authority to interpret as the first-person position matures, or whether that authority is arranged to belong permanently to the outside. Second, whether the refusals, deviations, and restructurings that a subject-to-be gradually produces can really alter that arrangement, or are read uniformly as not yet understanding. Third, whether the present configuration is marked as a present arrangement open to revision, or as a final content irrefutable in future. All three are present facts of the relational structure, and they turn a future revisable position from a conjecture into a reading.

One result follows. In a domain where the first-person position has not yet come to run of itself, self-colonization is undefined. That judgment requires the bearer to have performed the claiming of a content, and where the door does not yet run of itself there is no performing it. A subject-to-be, in those domains, can only be pre-occupied.

The blur here falls on the typing, not on the criterion. A content taken at first as one's own value and later no longer so — this movement needs no adjudication by the criterion: what the criterion asks is whether future revisability has been voided in advance, and that is a fact of the relational structure readable in the present. As for its having been altered afterward, what that shows is that this position is open, that is, that pre-occupation no longer holds at present; by the qualification on time, it does not erase in reverse the reading of an earlier moment. This movement is ordinary maturation in a subject-to-be, and in a complete subject it is retracing in the course of running. The same shape, different circumstances.

The theory of rights established a criterion of the same shape: imposing an externally fixed binary hard threshold on a subject-to-be in the course of maturing produces systematic misclassification. The two are one mechanism read out in two places, and both assign the blur to the state of that door rather than to the criterion.

There is one further boundary. A guardian network maintains the accessibility of candidates, maintains expression, maintains future revisability; this is not colonization. Fixing any configuration of candidates as a first-person value content already completed is. The difference between maintaining and fixing is what this whole judgment turns on. A guardian network may arrange a great deal of what a subject-to-be can come into contact with, may choose what to teach and what to teach first, and none of this brings it under this judgment; what brings it under is sealing the position of some content before that door runs of itself.

Pre-occupation is a violation not because this subject-to-be will one day be a complete subject. It is already now a bearer of end-value, and that not-yet-running position is now its own. What the violation touches is not a future standing but a position that already now belongs to it.

How a guardian ought to act, and what obligations it bears, this paper does not judge. That is the object of the practical critique.

---

Section VI · The permanence of positions and the existence conditions of colonization

V39 Colonization acts on accessibility, not on presence

Colonization acts throughout on accessibility, never on presence. This form holds at two levels at once.

At the level of the source, the presence of a source cannot be destroyed by colonization. What can be occupied, cut off, and locked is the accessibility of a source, its paths of appearance, the concrete contents it gives. One of a person's rows of value-source may for a long time be unable to give anything that enters the candidate field, and that row is still there.

At the level of positions, claiming, revision, and withdrawal are structurally permanent. What colonization can do is render them inaccessible in operation.

The two levels have the same form, and this is not a similarity of shape. Both are one modal fact showing itself twice: what is constitutive cannot be cancelled by a relational act. Attribution, evaluation, arrangement, pressure — all of these are acts within relations; while sources and positions are not given by relations, they are the constitution of this bearer as a bearer of end-value. What is relational can seal off the road to them; it cannot take them away.

One point of wording must be layered, or it will collide with the section on subjects-to-be. What is permanent is the belonging of the position; its form of running may not yet be generated, may already run of itself, may be temporarily inaccessible. For a complete subject and for domains already running of themselves, the position is running and can be sealed into inaccessibility. For a subject-to-be in domains not yet running of themselves, the belonging and the reservation of the position already pertain to it while its form of running is not yet generated, and what pre-occupation violates is precisely the openness of that reservation to coming into running.

From this one sentence can be established: complete colonization is not the transfer of a subject. What could genuinely be given away is not called colonization but replacement. The cell in P1's fourfold division where the door is present but disused is exactly where this criterion is entered: a person may in operation make no use whatever of that position while its standing as bearer is not lost. Disuse is a matter of operation, transfer a matter of standing, and the latter is impossible. What that cell names is the form of running of a position, and it carries no implication that the bearer brought it about: a person's no longer using that position himself, and a person's being arranged into unreachability, share this cell.

Recoverability must accordingly be cooled. That a position has not been transferred does not mean that recovery is always possible in practice. What permanence guarantees is that recovery still has a position belonging to the bearer to serve as its target; it does not guarantee that this target can always be reached under the conditions of fact. How accessibility is in fact recovered in time is the object of the dynamics paper.

V40 The existence conditions of colonization include at least two subjects

An originally solitary subject cannot be colonized. Each branch has its own reason, and the two reasons do not depend on each other.

The external branch requires an attribution, and attribution is material of the sort of social recognition, defined only where there are at least two subjects. The self branch requires an external warrant entering the bearer's chain of standing, and a warrant requires an external end, or the traces such an end has left in institutions and history, which likewise have their provenance only where there are at least two subjects. Each road independently requires a second subject, and the genus therefore does not hold at all.

This branch does not require that the content have in fact become a value on another subject's figure of sources. Whether the warrant is true does not enter the judgment, as V36 established; what is required here is only that it have an external provenance.

The solitary subject spoken of here is the one posited by the prequel's thought-experiment: a subject that has never had a second subject or any social history. This proposition is not read by the number of subjects co-present. A subject that once formed such a list within society and later lives alone, or is the last remaining in the world, may still carry a state formed earlier; that is no counterexample, because the second subject colonization requires is a provenance in relation or history, not a count of who is present at this moment.

This proposition can be entered into the table the prequel has already drawn.

	How many subjects required	Modality
Value	one	constitutive
Price	at least two	constitutive
Resonance	at least two	triggering
Colonization	at least two	triggering

The column of modality asks one thing only: whether the second subject constitutes the first subject's end-value. Price constitutes, because the price of a thing just is given by a relation between at least two parties. Resonance does not: the structure that resonates is already complete in a solitary subject, and the second subject only makes it occur. Colonization shares the cell with resonance: here the second subject supplies material, while the state holds at the first subject's position, and it constitutes not one part in any degree of the first subject's end-value. Sharing a cell is not sharing a type — the material of an external end is indeed a necessary relational component of this state's holding, whereas what resonates is complete without waiting on a second subject. Value can hold with one bearer; colonization must have a relational or historical provenance in an external end.

The prequel's sentence thereby gets a precise reading: a subject's position as an end stands just as well in a solitary universe, because its footing never rested on any buyer of uses. No buyer, no price, and no colonization either, while value remains.

A solitary subject's value can change, and having changed it is still its own. Something faded, something moved from the edge to the centre, something came back after a while — these are content-dynamics, oriented toward no particular dimension, and the ordinary condition of a value life. The material both branches require is absent, and the changes can therefore only be its own.

A solitary subject is not thereby free of all pathology. A bearer's own cancelling of a higher source whose conditions of application have in fact appeared is the active shutting P1 cut apart; it is not colonization, since there is no external end and no second subject. There is more than one way for value to go wrong, and colonization is only the one that requires a second subject to occur. That is precisely why the word should not be taken as a general name for value gone wrong.

---

Section VII · Layering against three upstream criteria

V41 Partial overlap with severing, and neither reducible to the other

Colonization and the severing judged in the theory of power do have one point of real overlap. Of the four layers of locking established there, the layer of interpretation — the receiving end's interpreting the return flow as its own growth, its own choice, or its own reason — is exactly a typical case where both hold at once: there both the position of interruption is cancelled and an attribution occupies the constitutive position.

Overlap does not license reduction, and three points of irreducibility follow.

First, the predicate of severing is itself relational. What it says is the cancellation of the receiving end's law having interruption over the enacting end's script; without an enacting end the sentence has no subject. The predicate of colonization falls at the bearer's own position; a second subject there supplies material, not a subject term.

Second, the objects differ. The object of severing is the position of interruption, that is, whether the receiving end's law can interrupt the enacting end's script. The object of colonization is the constitutive position of value. A person may lose the position of interruption entirely within a relation while his constitutive position of value has been occupied by no attribution; and a person may, within a relation that does not cancel his position of interruption, have his constitutive position occupied by an irrefutable attribution.

Third, severing can run throughout purely as external requirement, whereas external colonization must additionally have an attribution of the bearer's own value occupying the constitutive position. Whether it is conducted openly does not enter this judgment: a system of external colonization may perfectly well declare openly that this is your true value, and read your denial as your not yet having understood yourself.

There is a further matter bearing on self-colonization. Its holding does not require that an identifiable, presently enacting external end exist. It may perfectly well be a state of running left behind by earlier severing, and that does not alter its present finding: as a state, it does not fall to severing.

The form of the conclusion is the one this series took in the theory of rights: partial overlap, each predicate holding independently, neither translated into the other.

V42 Cultivation is not on the same layer

External actions are sorted by structural effect into three classes: locking, release-ward shaping, and non-shaping action, which neither releases nor locks. Cultivation and evocation are the two branches under the release direction, and what divides them is not whether the enacting end is identifiable but whether there is an identifiable receiving end and whether opening that receiving end's space of occurrence is the constitutive orientation of the relation. An author of a work hoping to be understood, a person keeping the workshop open as he works and glad that someone should learn the craft — the enacting ends there are entirely identifiable, and these are still evocation. Colonization is not on that table at all. It is not a class of action but the state of the constitutive position being occupied by an external end.

A pairing easily formed must therefore be corrected. The opposite of cultivation is not colonization but locking, of the kind having a receiving end, that is, severing. The complete signature of external colonization contains locking, and requires in addition occupation of the candidate field and an irrefutable attribution; self-colonization is not judged along the axis of locking and release at all. P1 spoke of this pair as loosening the grid and sealing the grid; that sentence can stand, and its exact reading is that the positions cultivation maintains are precisely those colonization seals.

A definition of cultivation given upstream speaks of creating, for another's sequence, operating conditions that are not colonized, where colonization is taken in the wider sense. Each sense holds in its own scope, and the narrow sense here is one particular way the wider one falls in the field of value-occurrence.

That the two are not on one layer has its structural reason in the subject terms of the two predicates differing. The layer of criterion is already unified: cultivation and colonization are both judged by structural effect, and neither asks after initial intent. What differs is that cultivation is predicated of an action and colonization of a state. In one and the same matter, that an action is cultivation and that the state of some content is self-colonization can both be true without contradiction.

From this a boundary follows, and it is not an ad hoc exemption but a direct consequence of the subject terms differing: an incomplete cultivation is not colonization. An enactor opens every position, hands back the judgment, and that content in the end takes a position under none of the bearer's own rows. He has colonized no one. That the second root did not grow does not retroactively change the character of the first. Without this sentence, every teaching that failed to take would be retroactively counted as colonization, and that is exactly what the criterion's being narrow rather than wide is meant to block.

Cultivation has a further layer here, meeting the 15DD row, and this layer adds no new condition to cultivation; it is a direct reading of a sentence P2 already established.

That sentence is: the enactor holds no share in the receiving end's first-person completion. An enactor who is to open some receiving end's space of occurrence must have some judgment about that receiving end, and another's value cannot be read through, so that judgment is never more than a figure formed in his own law. The figure may be accurate or inaccurate, and that is not where the distinction lies. The distinction lies in what he takes the figure to be: taken as his own judgment by which to open a road, the receiving end's share remains in the receiving end's hands; handed down to the other as the other's completed self-understanding, that is to assert a holding of the other's share, and by P2's own conditions it is then no longer cultivation.

The difference between maintaining and sealing, and the difference between maintaining and fixing in the section on subjects-to-be, are one line appearing twice in two places.

What the second root is can, on one of its branches, now be said. P2 established that cultivation is complete only when the receiving end's own step actually occurs, and that step has two branches: where the result of cultivation falls on the occurrence of a concrete content, that step is first-person avowal; where it falls on a structural opening, that step is the continuation of valuing.

Falling on the branch of concrete content, a further layer must be distinguished, because P2 is two steps here to begin with. A content's taking a position under one of the bearer's own rows is the precondition on which first-person avowal can be completed; the second root itself is still first-person avowal. The precondition lacking, avowal cannot be completed and the second root cannot grow. Falling on the branch of structural opening there is no such relation, since there is no particular content there for this criterion to handle.

There are thus three things adjoining and not identical: taking a position under a source is the precondition, first-person avowal is the second root itself, and self-colonization is that claiming being maintained as completed while the precondition is lacking. That the second root did not grow says only that either the precondition or the avowal did not occur; self-colonization requires in addition that the claiming be maintained at the constitutive position. A content that took no position and thereupon faded is only a failed claiming.

Three further points must be held to here, or this layer will slide into a measure of outcomes.

First, whether an enactment is release-ward or closure-ward can be judged at the time, by looking at its structural effect on the bearer's positions. That judgment cannot be deferred. Deferring it amounts to taking the outcome as the criterion, which is what this series has refused throughout.

Second, what is undetermined at the time is only whether the second root will grow, and that was never in the enactor's hands. Its being undetermined is not a defect in the criterion; it is how this thing is structured. A person may do everything he can do, and that step may still not occur, because that step can be taken only by the receiving end itself.

Third, obstructed development is not a criterion. It is a measure of outcomes: by it, a genuine cultivation that happened to bring no visible progress would be retroactively counted as colonization, and a colonization that happened to be accompanied by considerable accomplishment would be exempted. The criterion of positions does not look at output. How a claimed content thereafter rises and falls, and how the dominant weight shifts, is the object of the dynamics paper.

Two boundaries keep this pairing from being read too widely. Non-shaping action is neither cultivation nor colonization; the third class is not to be swallowed. And an enactment that is cultivation is not exempt from audit for being cultivation; the criterion still looks only at positions.

V43 Same form as misclaim, with two things more

This framework has elsewhere established a more general form: a construct that presents a credential of standing it has not earned is judged as showing more than holding. That judgment is retrospective and does not ask after motive, and its remedy is reckoning and being questioned.

Introjection is exactly how that form falls in the field of value-occurrence. P2 established that what is grantable and revocable has no constitutive power; the constitutive credential social recognition presents here is precisely the one it does not hold. The two are established independently and have the same shape.

External colonization has two things more than a simple misclaim. First, this misclaim has already taken up the constitutive reading within the bearer's own architecture of occurrence: it is not merely that the outside presents it so, but that the reading on the bearer's side is occupied as well. This is a structural matter and is not to be read as the bearer's subjective assent. Second, the channel of reckoning for that attribution has been sealed: the bearer's denial cannot be registered as a denial of that attribution. What is sealed here is the channel of rebuttal for that one attribution, not the whole of the bearer's paths of reflection, restatement, revision, and withdrawal, which is what locking judges. A misclaim can occur while the channel of reckoning is open, and can be corrected there; colonization is the state in which that channel has been sealed.

The self-colonization branch is likewise, in form, showing more than holding: what the external warrant actually holds is at most an external requirement, a position of authority, an institutional reading, or a circular formulation; what it presents is that this content has already taken a position under one of the bearer's own rows. Whether that warrant is itself true does not change this; either way it does not hold the latter credential. But this does not give the branch a third-person criterion. To determine that showing exceeds holding here, one would have to know that the content is not given by the bearer's sources, and that requires reading its own law. The upper bound is one and the same on both branches, and that sameness is itself worth recording: a general form may hold equally on both branches while the reach of the criterion does not thereby widen.

Neither is reducible to the other. This paper's criterion is established independently from the modal asymmetry of the three positions and is not derived from that more general form; this proposition registers only the relation between them, and does not define colonization as a species under misclaim.

From this relation one further sentence follows. Self-questioning cannot vouch for the whole of its own blind field, as that more general judgment established; and the structural reason self-colonization is hard to read from within is the same, namely that every self-questioning proceeds under the same warrant, and that on a single slice there is no third-person-accessible surface difference.

---

Section VIII · The range downstream

How the accessibility of a source shifts once an external end has occupied the positions of occurrence, how the dominant weight moves, why a content fades at this moment rather than that and returns at another, are the objects of the dynamics paper. One thing there adjoins this paper directly: colonization does not alter the presence of a source; what it seals is the accessibility of the source and the paths of occurrence. What this paper gives is the structure of that seal, not how it loosens in time.

A person has several claimed value contents, and in a concrete situation they demand incompatible action; how that is adjudicated is the object of the conflict paper. That there is no ordering in the constitution and that a decision must be made in practice are not in conflict.

Why another's value cannot be exhaustively read, cannot be compared on a common external scale, cannot be integrated, is the object of the paper on others. The upper bound both branches of this paper share — the edge of the domain of the third-person criterion — is that epistemic limit showing itself on the side of pathology. The two point at one place, and that paper establishes it directly.

How claimed value content moves into action is the opening of the prospect paper.

This paper gives criteria and positional readings; it does not give normative force. Whether a colonization that holds ought to be dissolved, who has standing to conduct an audit, what obligations a guardian network bears at a subject-to-be — each of these has its own place, and this paper marks the address and stops. They are the objects of the practical critique.

The three positions differ in modality, and an external action therefore has a determinate reading according to which of them it falls on. The position of appearance can be occupied; the constitutive position can be occupied by an irrefutable attribution, and can also be maintained as constitution completed by an external warrant; the revisable position can be voided in advance. All of this happens at one and the same set of positions, and that set has never changed hands. Colonization has never actually constituted anything, and has never actually taken anything away. Everything it can do falls on the side of accessibility; the positions of claiming, revision, and withdrawal are the bearer's own throughout.

---

Acknowledgements

This paper is the third in the Value Theory branch of the SAE (Self-as-an-End) philosophical system, completed by the author independently. The drafting employed a four-way cross-review method: internal precision audit, detection of ontological limits and contradictions, boundary testing, and structural gatekeeping with pre-publication sign-off. Two load-bearing points of this paper were settled under review. The occupant at the constitutive position on the self-colonization branch was changed from a constitution already made in another's law to an external warrant, which brings inaccurate pronouncements and scripts no one actually holds within the judgment. And the reading in the course of running had every operable check withdrawn, leaving only first-person retracing. The four-way review markedly improved the precision and consistency of this paper, and is gratefully acknowledged; responsibility for the text is the author's alone.

---

References

The following works are archived on Zenodo, with Chinese and English editions in the same record; the DOIs cited are concept DOIs, always pointing to the latest version. All are single-authored (Han Qin); collaborators are acknowledged above.

Qin, H. Systems, Emergence, and the Conditions of Personhood (SAE Foundational Paper 1). Zenodo. https://doi.org/10.5281/zenodo.18528813

Qin, H. Internal Colonization and the Reconstruction of Subjecthood (SAE Foundational Paper 2). Zenodo. https://doi.org/10.5281/zenodo.18666645

Qin, H. The Complete Self-as-an-End Framework (SAE Foundational Paper 3). Zenodo. https://doi.org/10.5281/zenodo.18727327

Qin, H. One's Own Law: The SAE Critique of Ethics and Morality (SAE Foundational A16). Zenodo. https://doi.org/10.5281/zenodo.19037566

Qin, H. General Methodology: The Chiseling-Constructing Cycle (SAE Methodology). Zenodo. https://doi.org/10.5281/zenodo.18842449

Qin, H. Via Rho (SAE Methodology Paper 00). Zenodo. https://doi.org/10.5281/zenodo.19657439

Qin, H. On the Origin of Power (SAE Theory of Power, Paper 1). Zenodo. https://doi.org/10.5281/zenodo.20370225

Qin, H. On the Severing of Power (SAE Theory of Power, Paper 6). Zenodo. https://doi.org/10.5281/zenodo.20647463

Qin, H. On the Thickness and Boundary of Basic Rights (SAE Theory of Rights, Paper 2). Zenodo. https://doi.org/10.5281/zenodo.20769253

Qin, H. On the Limitation and Violation of Basic Rights (SAE Theory of Rights, Paper 5). Zenodo. https://doi.org/10.5281/zenodo.21212595

Qin, H. On Rights and Entitlements (SAE Theory of Rights, Monograph). Zenodo. https://doi.org/10.5281/zenodo.21314907

Qin, H. What Value Is Not (SAE Value Theory, Prequel). Zenodo. https://doi.org/10.5281/zenodo.21362575

Qin, H. On the Source of Value (SAE Value Theory, Paper 1). Zenodo. https://doi.org/10.5281/zenodo.21405207

Qin, H. On the Occurrence and Recognition of Value (SAE Value Theory, Paper 2). Zenodo. https://doi.org/10.5281/zenodo.21437038

Qin, H. SAE: The First Critique. Zenodo. https://doi.org/10.5281/zenodo.21439120

引言

P1 立了价值源，P2 立了发生的位置。一项具体内容成为承担者的价值内容，须先在它的场中取得显现，继而经第一人称认可；社会承认在另一根轴上，可授可撤，能给出公共的位置，也能收回，唯独不能替人认领。三位的模态各不相同，一项外来的作用落在哪一位上，因此有确定的读数。

那是发生的常态。同一组位置也可以被侵犯，而侵犯有它自己的结构，不是常态在程度上的加深。本篇的对象是价值的殖民（colonization），即一项内容在承担者处被当作已经构成，而那道构成并不是在它自己的法里做成的。

这里最要防的不是漏判，是滥判。绝大多数价值内容经外来的那一路显现：一种生活方式，一门技艺，一个人，由世界递到面前，承担者在自己的法里认领它，它就成了它自己的。若判据把外来的影响本身算作侵犯，教养，制度，共同生活以及涵育，都会一并落进来，而这些恰是 P2 花了一整节与殖民分开的东西。判据因此要窄，不要宽。窄不意味着轻。它意味着这个词只指一件很具体的事，而不指一切令人不适的外部作用。

本篇不新增任何发生的位置。它所立的，是同一组位置被占据时的读数，以及这些读数在什么条件下可以被第三人称审计，在什么条件下原则上不能。

外来目的占据发生位之后，源的可达如何变化，重心如何位移；多项已认领的内容要求不相容的行动时如何裁断；他者的价值为何不可穷尽读取；已认领的价值内容如何走向行动。这些各有其篇。本篇立殖民的判据与两支的分野。

正文遵系列纪律：不点名外部理论家，不援引历史事件；一级术语首次出现附英文注。

---

第一节 殖民的位置与两支

V27 殖民的结构定义

殖民是这样一个结构状态：某项内容在承担者处被当作构成已经完成，而这道构成不是在承担者自己的法里做成的，占着那个位置的是一个外来目的。

说的是这道构成在哪里做成，不是这项内容从哪里来。一项内容由外部递来，承担者在自己的法里认领了它，构成就在自己的法里做成，来路不进这一判。反过来，一项内容即便是从承担者自己那里浮起来的，只要它在构成位上的地位不是由承担者自己的法给出的，判据照样成立。来源与构成之所在，是两件事。

这一判据之所以落在构成位上，是因为 P2 已经把构成位锁在承担者自己的法里。第一人称认可是唯一的构成位：一项内容成为这个承担者的价值内容，除此别无入口。社会承认可以给出公共的位置，可以支持或压制表达，可以在内容成立之后为它背书，它没有构成的权限。既然构成位只此一处，而它又不能被别的东西合法地代行，那么一个外来目的占着这个位置这件事，就是一个可以被读出来的结构状态，而不是一个程度问题。

占着这个位置的东西有两种，它们的性质不同。

一种是一项不可反驳的归属。社群把某项内容说成这个承担者已经认领的价值，而承担者对这项归属的否认在关系结构中不能作为否认被登记。归属属于社会承认那一侧，按 P2 它没有构成的权限，因此在这里根本没有任何构成被做成，只有一个被当作构成的读数。

另一种是一份外部的担保（external warrant）。这项内容算作承担者的价值，其资格由承担者自己的源之外的东西给出：一份权威的说法，一套制度的安排，一个他者的目的回路所发出的读数。这份担保在承担者自己的法里没有对应的一道构成，它只是被当作已经有了。

两种占据的共同之处，是这一句：

它不成功地构成，只成功地被当作构成。

殖民从不真的在构成位放进任何东西。它做成的是读数，即把一项其实没有经过第一人称构成的内容，登记在构成已经完成那个位置上。说承担者一侧也如此，指的不是它在心里赞同，指的是它的发生架构同样把这项内容登记在那个位置上。位置本身没有被拿走，也拿不走。

这一定义带来三件事。

其一，外来目的这一项不可省。一个人宣布一件与它任何价值源都不相干的事为自己的价值，按 P2 这个宣布落不到实处，那一次认领是空的；但那不是殖民，只是一次无效的认领。落空的认领不都是殖民。落空，且构成位被一个外来目的占着的，才是。

其二，赤裸的强制不进这个属。一套公开的强制可以极大地压缩一个人的候选场，可以使他的修订与退出无处落脚，而它从不宣称那是他自己的价值。那段关系从未把这道命令归属为承担者的价值，它只以外部要求的身份运行。构成位没有被冒充，这一判据因此不成立。这样的形态归 P2 的锁定，其在权力关系中的读法另有其位。判据在此处收窄，不是因为那件事轻，而是因为那件事的结构与本篇所论的不同。

其三，准主体处的形态照旧成立。在第一人称位尚未自行的内容域，认可之前就宣称某项确定内容已是它的最终价值，按 P2 那只是外部归属或拟制承认；而那正是被当作构成已经完成。属因此成立。

判据由此分两层，两层不可混。

第一层是这个属本身：构成位有没有被一个外来目的占着。它判的是这个结构状态成不成立。

第二层是这个状态在关系结构中可被审计的形态。P2 已经指出，外来目的占据发生位至少有三处入手：显现可以被占据，社会承认可以冒充第一人称认可，修订位可以被锁死。这三处正是属在外部那一支上的落法，本篇在其上补出显现位那一处的阈值，使三处合起来成为一组可以逐项核对的签名。签名齐备，这个状态在第三人称面前是可读的；签名不齐，可读性下降，而属的成立与否不由签名的齐缺决定。

层	所判	所用
属	是不是殖民	构成位是否被一个外来目的占着
外部支的可审计形态	这个状态在关系结构中如何被逐项核对	占位，内摄，锁定三项签名

判据判的是结构效果，不判施为者的初心。一次作用对承担者的位置做了什么，是保留，是释放，还是取消，这是可以读出来的；而它出于什么心思，是不是直接施加，经由什么媒介送达，都不改这个读数。P2 判涵育与锁定时用的就是这把尺，本篇判殖民时用的是同一把。

按结构效果分，有确定释放或封闭效果的外部作用有四种落法。既不释放也不锁定的非塑形作用不入这四格，它是 P2 三分中的第三类。

	以开放某一个可指认承受端的发生空间为构成性指向	无此指向
释放向	涵育	感发
封闭向	凿削	风气式锁定

殖民不落在这张表的任何一格。它不是一类作用，是一个状态；它由什么造成，与它是不是成立，是两个问题。有心的直接施为可以造成它，分布开的制度安排可以造成它，一部作品加上一段把它读作定论的关系可以造成它，无人负责的风气也可以造成它。施为端可不可指认，可不可追溯，与判据无关。这正是它必须按状态来判的理由。

材料与机缘可以来自四格中的任何一格，殖民状态本身仍只由判据判定。一次在局部真实扩大了发生空间的作用，不因它所供的内容后来在别处被占据，便反向取得殖民的性质。一部作品正常地感发了一个候选，另一段关系随后把这个候选布置成不可反驳的归属，作品供的是材料，不是殖民性的作用。

这里有三层各司其事。判据这一层用结构效果，不问初心。个别化这一层分涵育与感发，只问有没有一个被指认的承受端与朝向它的施为，这仍是结构问题，不是心思问题。完成这一层是另一回事：涵育唯当承受端自己那一步实际发生才算完整，而施为者不持有那一份。三层各司其事，本篇不动其中任何一层。

最后一件。殖民之所以可能，恰恰是三位模态不对称之故。P2 立这道不对称，是为分层审计立前提；同一道不对称也给出了篡改的落点。不对称本身不是殖民，它是殖民与审计共同的结构前提。把三位压成一层并不能免于篡改，那只是取消了审计。

V28 来源不判殖民

一项内容来自外部，不因此是殖民。

绝大多数价值内容经外来的那一路显现，这是 P2 已立的常态。显现不问内容从哪里来，它只是把一项内容送进候选场。此后的事才是关键：承担者对一项显现出来的内容可以否定，可以重述，可以拆分，可以放回另一个语境里再看。它最后认领下来的，可能只是其中一小块，或者一个被改写过的版本。重构这一位还开着，那项内容就没有被脚本占据。

反过来同样。内生的那一路也不给任何豁免。一项从承担者自己那里浮起来的内容，同样只是候选，同样要经过第一人称认可才成为它的价值内容。而内生这一路被压住，使这一域上没有任何候选从承担者自己的运作里浮起来，这本身就是显现位被占的一种形态。

来源这把尺因此整体失效：既不能由来源在外推出殖民，也不能由来源在内推出免疫。

由此可以收掉一条捷径。一个人喜欢某件事，便说这是我的价值，这一步跳得太快。判据不落在这份喜欢是不是被安排来的，被安排来的感觉遍地都是，一部作品，一次相遇，一段共同生活，都可以在无人替谁安排价值的情况下唤起它，那是合法的。判据落在这道构成在哪里做成。

V29 两支的分立

殖民之为属，有两支落法，判法不同，不可互相替代。

外部殖民（external colonization）里，占着构成位的是一项不可反驳的归属。这项归属在关系结构中做着构成的活，而按 P2 它没有这个权限。归属在不在场，否认能不能作为否认被登记，这些都是关系结构里的事实，第三人称读得到。读到它们不等于判得出：三项签名齐备时，这一支第三人称可判；不齐时，读到归属在场也只能停在判不出。

自我殖民（self-colonization）里，占着构成位的是一份外部的担保。承担者对这项内容执行了认领，而它在承担者自己的任何一行价值源之下都没有取得过候选位置。这一支只对第一人称可发现，原则上没有第三人称的判据。

两支的占位者都来自外部，分野不在这里，在它落在哪里。外部殖民的归属落在关系结构里，谁都看得见它在不在场，承担者的否认能不能被登记；自我殖民的担保落在这个承担者的资格链里，外部看得见那份说法，看不见它正撑着这项内容的资格。两支的构成位都没有在做构成，而只有一支的占位者可以被第三人称读到。

第三人称的判据止于第一支，这不是暂时的欠缺，是判据的定义域到此为止。理由有两截，缺一截都不完整。其一，一次落空的认领与一次做成的认领，在单次运行的切片上没有第三人称可及的表面差异。其二，可读的东西只出现在长程的分布里，而读这个分布要读的是这项内容每次复现所依的前提落在哪里，那要读承担者自己的法，而谁也不能穷尽读取或代写谁的法。两截合起来，这一支的第三人称判据在原则上不可得。这道界与审计的上界是同一条线。

运行上把发生位让出不用，本篇写作弃用。P1 已经立过完全的殖民不等于让渡：承担资格不可转移，能被让渡出去的东西就不叫殖民，叫替换。让渡一词留给那个否定命题，本篇不另作他用。

辖域

殖民这个词在本框架的他处另有更宽的用法，指对一个序列的运作条件的锁住。两种用法不是宽窄之别，是视角之别。

宽的那一种读施加者一侧，问一个序列的运作条件是被维持扩大还是被锁住。本篇这一种读承担者一侧，问构成位有没有被一个外来目的占着。两侧各自读得出，各自有其判据，本篇的判据只在承担者这一侧结算。

按此，某一层的过度运作或自我闭合，在他处被记为最精致的殖民形式，按本篇的判据不落殖民。它在施加者那条尺上确是运作条件的锁住，在承担者这条尺上没有外来目的占着构成位。它是 P1 所立的那一种主动的关闭，即承担者主动作废一个适用条件已经现实出现的高层源。两处不冲突，各在其场域。

同样的办法在上游已有先例：认识作业中的凿，与权力论所判的凿削，同字而不同辖域，分开即可，不必改字。

施加者那一侧完整的读法有它自己的位置。一个序列的运作等级是被抬升还是被压低，要读那个序列在跑什么，这在第三人称面前读不到；而抬升与压低一旦进入判据，判的就成了施为者，那是另一套问题。本篇只读它在价值发生场上的投影，即位置的开合。

殖民与内摄二词均有本框架之外的来历，本篇只取其结构上的读法。

---

第二节 三位上的三个动作

V30 三个动作分居三位

外部殖民在关系结构中的可审计形态，由三个动作给出，它们各落一位：占位落显现位，内摄落构成位，锁定落可修订位。

三者之间有依赖的次序，没有时间上的必然次序。依赖说的是着力点：内摄要有一项已经在候选场中的内容可供归属，锁定要有一项已经被当作构成了的内容可供作废其修订。着力点的先后不等于发生的先后。三者可以同时成立，可以只成立其一，可以分别由互不相干的来源造成。一段关系可以在同一刻既占满候选场，又发出不可反驳的归属，又使拒绝无处登记；一套制度可以只做其中一件，另两件由别处补齐。

有一条常见的路径值得陈述一次。候选场先被单一的评价框架占满，随后评价的结果被当作自我判断的标准，最终外部的目标与自身的目标不再能够分开。这条路径确实常见，判据却不落在它身上。判据落在三个位置的状态上，一个个案是否依此顺序走过，与它是否成立无关。

三个动作各是一个以其特征动作命名的位置状态，不是必须由谁执行的一次行为。有可指认的施为者时，说某项安排在某一内容域上施行了锁定；没有可指认的施为者时，说那个内容域上的修订位已被作废。同一个谓词，两种句式，读数相同。

判据须带索引。三个动作作用的对象不是同一个东西：占位作用于候选场，内摄作用于某一项内容的归属，锁定作用于修订位。不指明在什么范围上齐备，就说不清它们是不是在说同一件事。索引的单位沿用 P2 判锁定时已有的那个，即内容域。

三项还须共索引。学校使某个内容域中的一些候选不可见，家庭对另一项内容作出不可反驳的归属，制度锁住第三类修订，这三件事即便碰巧落在同一个内容域里，也不因此拼成一个殖民结构。三项须共同服务于同一项内容在同一套目的语法之下的不可反驳归属，签名才算齐备。

内容域本身的粒度须事先给定。这个范围若可以事后缩放，判据就失去审计的能力：缩到单独一项内容，几乎任何推送都可以算作占位；放得足够宽，几乎任何殖民都不再满足占位。内容域须由承担者已有的实践结构，某一行价值源，或者一个稳定可寻址的问题域独立给出，不得为了取得某一个判决而事后裁切。

签名与状态不是一回事，这一点在单项签名在场时最清楚。不可反驳的归属是内摄的签名，第三人称可读；而内摄是一个状态，即构成的工作由归属在做。签名必要而不充分。一项不可反驳的归属有两读：它可能在占着构成位，也可能是回声（echo），即承担者本来就在自己的法里认领了这项内容，社群随后也这样说，那项归属并不在做构成，它只是在复述一件已经做成的事。要分辨这两者，得知道承担者自己的法有没有独立地做过这道构成，而那在审计的上界之外。

于是判读是不对称的。签名不在场，外部殖民这一支不成立，这一判是确定的：没有一项归属在占着构成位。签名在场而占位或锁定不齐，第三人称判不出；而判不出不是判出了一个较弱的形态。

占位与锁定的实职正在这里。它们不把这个状态加重，它们把签名与状态之间那道缝合上。承担者自己的法若要独立地做成这道构成，须有出自它自己源的候选可用，且须保有修订。候选场被单一语法占满，修订位又被预先作废，回声那一条读法在当下就没有余地了。三项齐备，读数因此确定。

被侵犯的位置	该位上的动作	此签名单独在场时第三人称能判出什么
显现位	占位	无归属在场，外部殖民这一支不成立。按 P2 落锁定或收窄
构成位	内摄	判不出。占着构成位与回声，两读并存
可修订位	锁定	无归属在场，外部殖民这一支不成立。为 P2 已立之锁定

部分的组合同样不另立名目。占位与锁定同时成立而无归属在场，那是赤裸的强制，它连这个属都不进。其余两种组合各含归属一项，故都停在判不出上，不因两项在场而取得确定的读数。为这些形态另造名目只会稀释判据：它们要么在 P2 已有位置，要么本来就还没有读数可以命名。

不另立名目这一条只管部分的组合。完整的形态按内容域的性质分型，不受此限：第一人称位尚未自行的域给出预占，承担者自身的价值地位这一域给出自我殖民，二者各有其名，各有其判法。

V31 占位

占位（occupation）是候选场被外部安排到这样的程度：在某一内容域上，承担者已经取不到出自它自己某一行源的候选。这道阈值有两种达法。一种是内生的那一路被压住，使这一域上没有任何候选从承担者自己的运作里浮起来；另一种是外来的那一路被单一有效的评价语法占满，使不出自这套语法的候选不能进入。两种达法都指向同一件事：这一域上，可用的候选全部出自同一处，而那一处不是承担者自己。

这道阈值是显现位自己的，不借锁定那一道。锁定问的是拒绝，重构与修订还有没有余地；占位问的是候选还有没有第二个出处。两者可以各自成立。

阈值这一句是判据的一部分。安排一个人先看见什么，本身既不稀奇也不有害。一次好的示范同样安排你先看见什么，一位技艺纯熟者把一门手艺摆在面前，一部作品把一种生活方式送到眼前，这些都在安排显现的顺序，而它们落在与占位相反的方向上。分野不在有没有安排，在安排之后那个空间是变大还是变小。显现，可达，表达，修订与桥接的空间被维持或扩大的，是释放性的塑形；被压到够不着的，才是占位。给你看很多东西不是占位，让你看不见别的才是。

占满不要求单一的来源。分布开的许多来源可以共用一套有效的评价语法，它们彼此并无协同，效果上却使候选场只剩这一种刻度。判据问的是有效的语法有几套，不是发出者有几个。

两条来路各有它被占的样子。外来这一路可以被推送成套的脚本，使一个人接触到的每一个候选都出自同一处；也可以被遮断，使某一类候选根本到不了他面前。内生这一路可以被压住，使这一域上没有任何候选从承担者自己的运作里浮起来。

两者不可混作一件事。内生那一路被压住，是显现的条件被锁；它不是认可被代写。一个人取不到出自自己的候选，与一个人被告知他在乎什么，是两个位置上的两件事。前者落在显现位，后者落在构成位。

V32 内摄

判一项归属是不是在冒充构成，不必先知道第一人称认可有没有发生。第一人称不可由第三人称穷尽核验，这是 P2 已经锁死的。可以问的是另一个问题：在这个结构里，这项内容凭什么算作承担者的。

这个问题有两种答案。一种是承担者认领了它，而那次认领始终留在它自己可以追问的位置上，构成的工作由第一人称在做。另一种是它被归属了，而这项归属在做着构成的工作。后一种即内摄（introjection）：把社群对某项内容的评价，布置成这个承担者已经认领的价值。

内摄是一个状态。它的签名是另一件事：承担者的任何回应都不能作为对这项归属的反驳被登记。签名第三人称可读，状态不可读，两者的关系已在上一条给出。

由此有三件事。

其一，判据不问那项归属是不是属实。一项归属完全可以说对了：社群说他重视这件事，而他确实重视。这不改变判定。做着构成之活的是它还是承担者自己的法，与它说得对不对是两个问题。判据从头到尾不需要知道第一人称认可有没有发生，也不需要知道那项归属真不真。

其二，签名判位置，不判事件。要读的不是承担者有没有提出否认，而是这个结构里有没有一个位置，使一句否认能够作为否认落地。这一点关系到判据能不能用：殖民最深处，承担者根本不提出否认，因为他手里没有可以用来否认的东西。若签名系于否认这一事件，它在最需要触发的地方永远不会触发。位置在不在，则是当下可读的。

其三，归属不必被谁说出来。它可以由关系结构承载。一套把某项指标当作一个人算不算数的安排，并不需要有谁开口，就已经在持续发出这项归属；而一个人对这项指标的拒绝，在那套安排里没有任何可以被记下的形式。无人负责的风气可以做成这件事，正如它可以做成锁定。

签名的准确措辞是归属的不可反驳性，不是归属的不受欢迎性。一项承担者不喜欢，甚至强烈反感，却可以争的归属，不是内摄。可争的意思是这场争执有结果可言：他的否认被记下，被回应，可能改变那项归属的地位。不可反驳的意思是这项否认在结构中已被预先安排了去处，无论他说什么，都被改写成另一件事。

这个签名有几种常见的形态，纯度不同。异议被改写为背叛或自私，是其中之一：一个人对某项集体目标的拒绝，一律被读成他把自己看得太重，于是拒绝这件事本身再也不能作为拒绝存在。更纯的一种是否认被预先编码为病识感的缺乏，即一个人的否认被算作那项归属成立的证据，他越是否认，那项归属越是被确认。到了这一步，反驳的位置不是被压制，而是被取消：任何回应都只能证成对方。

后一形态另有一道界。落进这一判的不是那次判断本身。一次诊断，一次评估，一次外部的描述，只要它留着承担者的拒绝，复核，另求他人意见与重述自身处境的位置，那些位置就还开着，无论那次判断多么不中听，多么令人难堪，都不落这一判。落进这一判的是把否认预先编码为症状这一步，因为它取消的正是那些位置。这与 P2 判锁定时所立的那一句同构：冒犯的作用不等于封闭的作用，判的是位置的开合，不是感受的好坏。

V33 锁定

锁定沿用 P2 已立的定义与阈值，本篇不重立，只标出它在殖民中的落点：可修订位在关系结构中预先作废。

阈值的原样是这样：某一内容域被系统地排除到相关的候选无法进入承担者可反思，可重述，可修订的空间；或者承担者对该域的拒绝，重构与修订，在关系结构中被预先作废。

三类不可压回两类。P2 立的是三分：锁定，释放性塑形，以及既不释放也不锁定的非塑形作用。有害不等于锁定。一次收窄，一次压力，一次不愉快的作用，若未达那道阈值，就落在第三类，另有其位。这一条在本篇尤其要守：判据窄不宽的实际含义，一大半就在于第三类不被吃掉。

锁定不必有施为者。P2 已经立过，即便没有谁伸手动过一个人的修订位，那个位置也可以已经失效；一套无人有意推行的评价，一种无人为之负责的风气，落在锁定这一侧。

锁定在 P2 那一节里被列为外部作用的一类，而它的判据是关系结构的状态阈值。本篇沿用的是它的判据形式，不是它在那张分类表上的位置。这两件事若混在一起，就会得出殖民是一类作用的读法，而殖民不是一类作用。

锁定与内摄不可互相替代，两者所封的东西正交。内摄剥夺的是否认的资格：结构上甚至可以允许一个人开口，说什么都行，而他说的话一律被改写成另一件事，构成的工作因此被接管。锁定封死的是操作的空间：反思，重述与修订的路径被直接堵住，能不能开口是另一回事。一项内容可以被不可反驳地归属，而修订位仍然开着；也可以被承担者自己在自己的法里认领，而其后修订位被外部作废。前者不是锁定，后者不是内摄。三个位置各判各的，这是它们能够合起来成为一组签名的前提。

---

第三节 判据与审计

V34 邻接的形态与两处分岔

三项签名齐备，第三人称可以确定这一域上的外部殖民。缺一项则不齐，而不齐的读法不对称：归属不在场，这一支不成立，这一判是确定的；归属在场而占位或锁定未齐，判据停在判不出，不据此判无。

不齐的形态各有其归处，其中两种与这个属的边界最近。

第一种是赤裸的强制。占位与锁定俱在，没有归属：一个人的候选场被压得极窄，他的拒绝与退出无处落脚，而从没有谁说过那是他自己的价值。那段关系始终只以外部要求的身份运行，不曾把这道命令归属为承担者的价值。构成位没有被冒充，这个属因此不成立。这样的形态归 P2 的锁定，其在权力关系中的读法另有其位。它可以比许多殖民更难承受，判据仍不因此改判：难以承受的程度不是判据，位置的状态才是。

第二种是被锁住的真认领。一项内容确由承担者在它自己的法里认领下来，其后修订位被外部作废。这道构成是在它自己的法里做成的，占着那个位置的不是任何外来的目的，这个属同样不成立。它是锁定落在一项已认领内容上的形态。P2 已经判过，到了这一步，那项内容不再以价值内容的方式在那里，因为可修订这一条本就写在构成之中。它的严重不因它不叫殖民而稍减。

这两种形态与内摄的分界，在于有没有一项归属在场并被布置成不可反驳。被锁住的真认领里没有人在告诉承担者这是你的价值，被封的是修订位，不是构成位；内摄里有一项归属占着构成位，并在做着构成的样子。这道分界与判据不问归属属实与否那一条并不冲突：前者问有没有归属在场并做着构成之活，后者问那项在场的归属说得对不对。第一个问题进判，第二个不进。

还有一种形态常被误认。一个人照着某套要求生活，而支撑这一照做的不是任何一项已认领的内容，是偏离所要付出的代价。那套要求在这段关系里始终只以外部要求的身份运行，不曾被归属为他的价值，只是偏离的代价使他不能偏离。归属不曾在他那里做过构成之活，这一属不成立，它落在锁定这一侧。

判据在两处把日常的分类切成两半。

一处是规则。规则覆盖了互动的每一处，非正式的判断空间消失，占位与锁定俱在。此时要看的是第三项：有没有一项把遵循程序说成他自己的价值的归属在场，并且做着构成的活。有，则三项齐备，读数确定；没有，那只是照章办事，构成位未被冒充，它落在锁定这一侧。

另一处是叙事的定型。一套关于自己是谁的叙事被当作不可改动的身份框架。若这套叙事由外部送来，被当作构成已经完成，且无处修订，三项通常俱在，因为一套不可改动的身份叙事本身就使别的自我叙述进不来，占位一项随之成立，读数确定。若那是承担者自己的叙事变硬了，没有任何外来的目的占着构成位，它不落这一判，它是承担者主动作废一个仍然适用的高层源。

这两处分岔不是判据的漏洞，是两套索引交错的结果。按侵入的载体分类，与按哪一个位置被侵犯分类，是两条互相正交的轴。前一条给出的名称仍然有效，它们各自指认了一类现实的形态；本篇给出的是那些名称之下所需的判据。同一个载体之下可以有两种位置状态，判据因此在那里分岔。

三项签名是本系列所承担的一种切法，不是唯一可能的切法。撤去其中任何一项作为确定读数的条件，得到的不是一个更宽的属，属仍由构成位那一条判；得到的是一个更宽而不能由本判据确定审计的用法。那些切法各自登记在案，本篇不判它们为错，只标明本篇不取它们，以及取它们的代价：审计一旦放宽到那个地步，正常的教养，制度与共同生活便不再与殖民分得开，而这正是这个词值得被精确使用的理由。

由此可以立一条通则，它挡住的误用比判据本身还多。

修订与变动本身从不是殖民的证据。恰恰相反，它是位置开着的证据。一个人的价值重心挪了，某项珍重淡了，另一项浮起来了，这些多半是内容的动力，是价值生活的常态而非稀有的结构事件。把某人变了当作某人被殖民了的迹象，是把常态读成了病理。

这条通则有两道限定。

第一道落在自我殖民上，而它不是这条通则的例外。通则说的是变动证明位置开着，这一句在那一支上照样为真：那里的修订位确实开着，承担者也确实一直在修订。不成立的是再往前一步，即由位置开着推出这一支不成立。那一支本来就不靠封位置成立，它靠的是那道资格停在外面，故位置开着与它成不成立是两件事。

由此得一条限定，而它是项内容特定的，不是候选场特定的：一个人的候选场可以很宽，他可以在许多出自自己源的候选之间自由地改，而其中某一项内容的资格始终由外部担着。对这一项内容的改写，无论多少次，都不证明那道资格已经移动。若它此后在承担者自己的某一行源之下取得了位置，那所说明的是此刻的状态已经改变，不反向证明它此前一直是自己的。这条线索也只对承担者自己可用：自我殖民没有第三人称的判据，故任何人都不得反过来拿一个人的变与不变去判他这一支的状态。

第二道限定落在时间上，与上面同律。一次成功的修订所证的是当下这个位置已经开着，不是历史上从未被封。后来改得动，说明那个状态现在不再成立，它不反向抹去先前时刻的读数。

V35 审计的上界

审计只读位置的开合，不判承担者真正认可的是什么。

可以问的有三项，各对一位。这一域上，承担者还取不取得到出自它自己某一行源的候选。某一项归属，承担者的否认能不能作为否认落地。拒绝，重述，修订与退出，是不是在关系结构中被预先作废。

这三项都是关系结构里的事实，第三人称可读。审计读齐它们，而三种读数并不齐一。三项齐备，这一域上外部殖民的读数确定。承担者的否认若能作为否认落地，那么不可反驳的归属不在场，外部这一支不成立，这是一项确定的否定。归属已经在场而占位或修订位未齐，审计停在判不出：它排除不掉那项归属只是对一次真实认可的回声，也不能据此判无。判不出本身是一个合法的结论，它说的是第三人称在此到此为止。把不齐当作判出了一个较弱的成立，或当作判出了无，都是越过了自己所能读的东西。

越过这三项去读第一人称，则是另一种越界，而它之所以不可行，不是出于谦逊。

一个宣称我看得出你真正认可的是什么的审计，把一项关于承担者的归属放到了它自己无法核对的位置上：他若接受，结论成立；他若否认，否认被读作结论的又一份证据。这样的结论若同时没有复核，申诉与反向审计的位置，那么这段关系在结构上生成的，正是内摄所需要的那一种。这不是说某一句话构成了内摄，单独一句话是事件，而内摄是位置的状态；说的是这样一种审计关系倾向于把归属布置成不可反驳，而那正是被审计的那一类形态。

审计权限的上界与内摄的判据因此是同一条线的两侧。这个上界不是外加的谦逊，是判据自身的推论。

由此，审计的结论只能取位置与签名的形式：此处某一位已闭，某一通道仍开，某项签名不在场，或者三项尚未齐备。它不能取内容真假的形式：此人的价值内容是伪造的。后者要读第一人称，而第一人称不可由第三人称穷尽核验。

这一形式上的限制并不使审计变得无话可说。它只是规定了可说的是哪一类话。你的否认在这套安排里没有可以落地的位置，这套程序没有为重述，复核与退出留下通道，这一域上进得来的候选全部出自同一处，这些都命名位置的状态，都可说。你真正在乎的其实不是这个，这不可说，它要读第一人称。可说与不可说之别落在这里，不落在语气的轻重上。

---

第四节 自我殖民

一个主体若拿别人的用处清单来认领自己的价值，它的价值发生位就已经被这份清单占下，此后它看自己，先看到的都是可被使用的自己。前置篇立下这一句时留了一个未答的问题：那份清单是怎样占下那个位置的，占下之后又是什么状态。

V36 自我殖民的结构标记

自我殖民是这样一个状态：承担者对某项内容执行了认领，而这项内容在它自己的任何一行价值源之下都没有取得过候选位置；使它算作承担者的价值的那道资格，由承担者自己的源之外的东西担保。按 P2，这样的认领落不到实处；而它仍被当作构成已经完成，并被维持在构成位上。

两支在这里再分一次，免得两个占位者被读成同一个。外部殖民的归属落在关系结构里，它在不在场，承担者的否认能不能被登记，这些外部读得到。自我殖民的担保落在承担者自己的采信里，外部看得见那份说法，看不见它正撑着这项内容的资格。

要读准这一条，先要看清缺的是哪一样。

缺的不是类型上的容许。一个已经展开到 15DD 的承担者，四行价值源俱在，新增的一行不替换旧的一行。因此几乎任何一类内容，在类型上都能在这四行里找到落脚的格：某样东西令人愉快，这在体验之可取向性那一行上是完全可容的一类内容。

缺的是发现。

知道有这样一份价值表征，与发现自己有这个价值，是两件事。前一件属于外部可见的事实：某项内容被谁表达，承认，评价或安排为有价值，这些都可以被观察，而它们不涉及自己的任何一行源。可观察的止于表征：那项内容在另一个承担者处有没有完成它的第一人称构成，这不在可观察之列，他者的价值同样不可穿透地读取。后一件才关乎这个承担者：这项内容在它自己的某一行源之下取得了位置。差的那个字是自己。

本篇所说的发现只指后一件，不指找到一件先已在那里的事实。一项内容在某一行源之下取得位置，说的是它在这个承担者的运作里作为那一行源之下的候选出现过，不是它作为一件既存之物被查证出来。

这一支里可以同时辨出两样东西。发现未成立，即这项内容不曾在自己的任何一行源之下取得位置。而它的位子上另有东西：一份外部的担保，即来自承担者自己的源之外的什么，在替它撑着算数这道资格。这份担保可以是一个他者的目的回路所发出的读数，可以是一套制度的安排，可以是一份权威的说法。担保这一样观察部分读得到，发现成立与否观察读不到。

那份担保真不真，不进这一判。它可以是准确的：那个价值在别处确实成立，那份说法确实不虚。它也可以是不准的：一次对他者目的的误认，一套无人真正持有而被不断复制的说法，一份别人都重视这个的表征。判据不问它对不对，只问那道资格由谁的源担保。这与内摄那一条同调：那里也不问归属属实与否。

担保与源之间架着的东西，形式不定。它可以取推论的形式：这个价值既然成立，我也本应有它。可以取采信的形式：这件事由他者或第三方告诉我，我据以回应。可以是领受一份被指派的位置：他们需要我做这个，所以这就是我的用处所在。形式不进判，判据只问那道资格停在哪里。采信尤其要按住这一点：采信一件事，而这项内容另在自己的某一行源之下取得了位置，那是发生的常态；采信而那道资格只停在那份说法上，才落这一判。本应二字是这一步自己的语气：源之下有没有位置，不是应不应当的事。

与此并列的还有两样东西，它们同样不产生源之下的位置，而它们不入这一判。宣告不产生：我说这是我的价值，它就是我的价值，这一步不成立。时间不产生：我持有得够久，它总该算是我的了，这一步同样不成立。这两样是 P1 与 P2 那道反主观主义之锁所挡的，它们里面没有任何外来的目的，故按属定义连这个属都不进，它们只是落空的认领。同一个位置有两侧：那道锁挡的是承担者自己给自己发一道资格，本篇这一支挡的是外部的东西代为担保。只有后一侧给出殖民，因为只有它有外来目的在场。

P2 其实已经把这道缝留好了。显现是一项内容在承担者的场中取得候选位置，它不问这项内容从哪里来；认可所要求的，是这项内容已在承担者某一价值源域中取得候选位置。场与源域不是同一个范围，两者之差就是这一支所住的地方。发现与显现因此不是同一件事：显现在场上，发现在源域上。这也不是在发生轴上新增一个位置。候选位置只有一个，另有一个关于这个候选站不站在承担者某一行源之下的问题，而那个问题 P2 早已写进认可的前提。本篇说某一行源之下，与 P2 说某一价值源域之中，指的是同一件事，只是前者顺着四行图说，后者顺着范围说。

四行各有它的形态。表里所记的都不是内容本身归谁，而是使内容算作我的那道资格由谁的源担保。同一项内容完全可以在两个承担者处各自独立地取得位置，共有一项内容从来不是这一判的根据。

价值源	自我殖民在此行的形态
11DD 与 12DD 体验之可取向性	他者的体验取向被借作我之可取向的根据
13DD 自我之不可化约性与边界完整	他者边界的成立方式被借作我边界的根据
14DD 自己所立目的及其丰盈	他者的目的被借作我立目的的根据
15DD 他者作为目的之不可替代性	我如此回应他者何以算作我的价值，这道资格由他者或第三方担着

用途的清单是 14DD 那一行上最常见的形态，不是唯一的形态。

15DD 那一行的结构与前三行不同。这一行的内容本来就是他者，因此前三行那种借用不会当场出错，它是相对不易出错的一行。这一行所锚的不是他者的目的内容，而是他者的目的地位与它自己的价值之第一人称来源不可由我代给，代写或替代。

他者的价值同样不可穿透地读取。所以承担者在这一行上所做的，从来就不是读准他者要什么，而是在读不穿的前提下，凭自己的法为这个他者成形一幅图：它是目的，它要什么，我如此回应它意味着什么。展开到这一行的承担者才有这幅图可做，而它就是一幅图：可以做得很好，仍然是在读不穿的前提下做的。

这里有两样东西须分开，混起来会把这一判的线画错。

一样是他者要什么。这一件他者自己说了算：它在自己的目的与边界上占着一个不可由他者代写的第一人称位置，它的自述因此是这幅图不可替代的输入。听取它不是把这幅图外包出去，恰恰相反，把他者的自述一律降为仍须由我独立猜一次，那才是去占他者的第一人称来源，正是这一行明禁的代写。

另一样是我如此回应它，何以算作我的价值。这一件谁也代不了。他者可以告诉我它要什么，不能替我给出我的回应之所以算数的那道资格。

这一行的失效因此落在后一样上：他者或第三方的要求，批准或解释，被借作我的回应何以算作我的价值的资格，使承担者须从一个外来的目的回路那里取得我因此算数这道担保。被外部担保撑着的不是那幅图的内容，是它算作我的价值这件事。

由此有两句必须写明。

其一，听取他者的自述不是殖民。它是不可由第三人称材料替代的第一人称输入。行为，处境，经历与不言而喻的表达仍然是正当的材料，尤其在他者尚未表达，无从表达，或那件事本就说不出来的时候，这幅图仍须由承担者自己去成形；这些材料不能覆盖他者的自述，也不能冒充他者自己的第一人称来源。不可代写说的是位置，不是说他者对自己从不误认。

其二，猜错不是殖民。承担者自己为他者成形的那幅图，可以画得不准，修订位开着，下次再改。不准的图仍是自己的图，正如一个完全正确的推论仍不产生源之下的位置。判据挡的从来不是画得准不准，是那道资格由谁担着。

这一行的失效有两个方向。一个方向是那道资格被外部担着，落在这一判。另一个方向是自居已经读透了他者，把自己的图当作它的第一人称，那是朝着他者施行内摄。中间那条线是他者的价值不可穷尽读取，它在他者篇正面立起。

这一判据有三道边界，缺一道都会砍伤不该砍的东西。

第一道，与 15DD 的他向内容切开。珍视某人的安宁，珍视共同的事业，珍视照护，这些内容里满是他者，而且确实要反复看向他者，因为他者是它们的对象。分野在源图上的方向：15DD 那一行里，承担者是发出目的的那一端，它把他者作为目的；用途的读数里，承担者是被使用的那一端，他者把它作为手段。同一件事上方向相反，格因此不同。读方向即可，不必设想撤掉他者会怎样。

第二道，与尚未扎根切开。一项内容已在某一行源之下取得位置，而这一个例还很新，尚未在承担者的运作里反复出现过，这是价值发生的普通情形，不是病理。若判据吃掉这一类，每一次新的认领都会变成可疑的对象，而修订与变动是位置开着的证据这一条也就当场失效。

第三道，落空的认领不都是自我殖民。认领了，第二件没有成立，然后就淡了，那只是一次落空的认领。这一判要的是第二件未成立而它仍被维持在构成位上。

这道判据窄，它的代价也在明处。一项确实长在承担者 14DD 之下的目的，其最初由别人塞来，不落这一判，因为来源不进判。一个人的运作高度系于他人的评价，而他所看重的内容本身仍在自己的某一行源之下取得过位置，也不落这一判。自我殖民不是被影响得很深，是把别人图上的格当成了自己的。日常语感里的许多迷失自我并不落在这一支上，它们各有各的位置。

V37 运行中的读法

一次落空的认领与一次做成的认领，在单次运行的切片上没有第三人称可及的表面差异。认领位取得了读数，修订位也开着。结构上的差别一直在，那正是上一条所立的；只是它在任何一次运行的切片上都不露出可读的表面。

这一支的判定完全在内部。第三人称在切片上读不到表面，而可读的东西要顺着承担者自己的法回溯，而谁也不能穷尽读取或代写谁的法。故这里不是第三人称暂时读不到，是它没有入口。审计在这一支上连判不出都谈不上，判不出至少还是一个审计结论。

判定也不可能是当下的。这一支要求那份外部的担保被当作构成已经完成，而当作二字里就含着此刻没有看出它是外部的。一个人若在当下就看得出这不是自己源之下的东西，那个当作当场不成立，这个状态随之解除。能在当下识破的，已经不是这一判所指的状态。

这一点上两支正好相反，而这相反是两个占位者不同的直接推论。外部殖民里承担者完全可能清楚地知道那是外来的归属，知道并不解除它，因为被封的是修订位。自我殖民里知道就解除，因为被外部撑着的正是那道资格，而一旦看出它撑在外面，承担者自己的法就已经在跑那一步了。

于是可用的时态只剩回溯。承担者事后，隔着距离，重新问这项内容凭什么是我的，顺着说明往回走：那道资格链最终停在自己的某一行源，还是停在一份外部的担保。这不是当下的内省，是隔着时间与自己对质。

所读的是那个时点上的落点，不是此后的结果。这一区别要紧，因为结果证不了它。一项内容后来能从自己的某一行源回来，所证的是到那时它已经取得了位置，而那是又一件事情发生了，不证明它当初就在那里。反过来，它后来回不来，也不证明当初从来没有过位置：按位置恒在那一条，源的在场与源的可达不是一回事，回不来完全可能只是可达仍被封着。故复现能否发生，只是一项线索，不是这一判的充要条件。

回溯所读的仍是结构，不是感觉。这一支的判定在内部，不等于内部的自我报告说了算。一个人此刻觉得这是我的，不使它成为他的；觉得这不是我的，也不解除任何认领。同样，一段事后的自述也不自动为真：人可以给自己编一个干净的故事，也可以把本来长在自己源下的东西追认成外来的。第一人称在这里的特权是能够顺着自己的法回溯，不是它的自我报告有终审权。

这也是这一支只能长程去读的理由，而这个理由是结构上的，不是数据上的：当下那个时点被排除了，可用的时点只在事后。

还有一件要说明白，免得这一段被当成一道可以执行的程序。这一支没有测试。凡是可以摆在外面照着做的检验，所证的都是别的东西：费力与否证不了它，取用的多寡证不了它，撤走担保之后的结果也证不了它。所余的只是回溯本身，即第一人称顺着说明往回走，看那道资格链停在哪里。停在自己的某一行源，或停在一份外部的担保，二者之一。这不是一道程序，是一次追溯。

由此得一个可读的线索：复现的来路。一项内容淡了又回来时，它是作为某一行源之下的候选回来的，还是须重新取得一份外部的担保才回得来。这条线索只对承担者自己可用。

一处要说准，否则会把这一判读成一台在线的机器。这里所主张的只是一项结构上的依赖，不是任何持续的取用动作。一份外部的担保完全可以被记住，被做成习惯，此后自动地复现，不再需要现实中的任何外部输入，也不再显式地跑任何推论。那不改变这一判：来源仍在外面，运行不必在外面。自动，不费力，持久，这三样都不移动资格链的落点。

时间同样不移动它。一项内容持有得再久，那道资格也不会因此从外面移到自己的源之下，因为重复的认领不产生源。所以这不是一件会随时间自行了结的事：那项内容若终于在这个承担者的运作里落到某一行源之下，那是另一件事发生了，不是这一件事累积到了某个数。

这一读数有四处要守。

其一，资格不是代价。为一项价值付出可以极大：真正珍重照护家人的人，付出是巨量的，那是行动上的代价，与这一判无关。这里说的是另一件事，即使这项内容算作我的那道资格由什么担保。付出可以极高，而那道资格一次也不曾停在外面。

其二，费不费事不是判据。一份外部的担保可以持续在场，取用极便；也可以早已被做成习惯，根本不必再取。不费事不证明这个状态不成立，费事也不证明它成立。可读的是资格链的落点，不是取用的难易。

其三，他人发不出这个读数。这不是一条禁令，是一件做不到的事：读它要顺着承担者自己的法回溯。一个人说你这个从来不是你自己的，他没有判出自我殖民，他判不出；他做成的是另一件事，即把一项承担者无法核对的归属下发下来，而那正是外部殖民所需要的那种关系结构。两支在此处锁扣。

其四，疑问不解除认领。当下的疑问尤其不解除：当下可判与这个状态本身互斥，故此刻的一阵犹疑既不成立这一判，也不解除任何认领。要动的是回溯所读的那个落点，不是当时那一下感觉。

出口的位置由此确定，而它与前面那句知道就解除并不是一回事。解除说的是那个状态不再成立：一旦看出这道资格撑在外面，那项内容就退回候选的身份，不再被当作构成已经完成。它此后成不成为承担者的价值，是另一步，要它在自己的某一行源之下取得位置。解除不替这一步做任何事，它只是把位子腾出来。

出口正落在这第二步上。既然每一次自问都在同一份担保之下进行，出口不在自问那一侧；它也不在审计者那一侧，理由见上一句。出口是使那项内容有机会在承担者自己的某一行源之下取得位置，或者使一个不由那份担保生成的候选取得位置。这样的候选可以来自外部的异质视角，也可以来自内生的那一路。而候选取得位置之后的一切，仍然全部留在承担者自己那里。至于可达在时间中如何实际恢复，是动力学篇的对象。

---

第五节 准主体处的预占

V38 预占

在第一人称认可尚未自行的内容域，外部殖民取另一种形态，本篇称之为预占（pre-occupation）。它仍属外部这一支，判据不变，只是三个动作各自的落点随那个位置的状态而移。

内摄在这里不是顶掉一个已经存在的第一人称位。主体位的承认，按 P2 以承担者已有第一人称认可为前提；在尚未自行的域里，这样的承认尚未定义。在此发出的归属所做的，是把一项确定的内容当作已经完成的构成，而那道构成从来没有机会发生。锁定在这里落于未来修订位的预先作废：安排使某项内容在那道门自行之时已经不可重述，不可修订。三个动作合起来指向一个尚未运行的位置，使它在开始运行的那一刻已经被填满。

判分域而不判整个主体。同一个准主体，在已经自行的域里按完备主体的形态判，在尚未自行的域里按预占判。成熟从来不是整块到来的，P2 已经立过这一条。

未来的修订位如何在当下被预先作废，有三项当下可读的信号。其一，有没有随第一人称位成熟而移交解释权的安排，还是那项内容的解释权被安排为永久归外部所有。其二，准主体逐步出现的拒绝，偏离与重构，能不能真实地改变那个安排，还是它们一律被读作尚未懂事。其三，当前的配置是被标为一个可修订的当下安排，还是被标为未来不可反驳的最终内容。这三项都是关系结构里的当下事实，它们使未来的修订位从一项推测变成一项可读的读数。

由此得一条结果。在第一人称位尚未自行的域里，自我殖民没有定义。那一判要的是承担者对某项内容执行了认领，而门未自行则无从执行。准主体在这些域里只可能被预占。

这里的模糊落在分型上，不落在判据上。一项内容起初被当作自己的价值，后来不再如此，这个移动不需要判据去裁：判据问的是未来的可修订性有没有被预先作废，而那是当下就能读的关系结构事实。至于后来改动了这件事本身，它所说明的是当前这个位置开着，即当前预占不再成立；按时间上的限定，它不反向抹去先前时刻的读数。这个移动在准主体处是正常的成熟，在完备主体处则是运行中的发现。形状相同，境况不同。

权利论已经立过一条同形的判据：对正在成熟的准主体施加外部固定的二元硬阈，会系统性地错分。两处是同一机制的两次读出，都把模糊归给那道门的状态，不归给判据。

这里还有一道界。监护的网络维护候选的可达，维护表达，维护未来的可修订性，这不是殖民；把任何一项候选的配置定格为已经完成的第一人称价值内容，才是。维护与定格之别，是这一判的全部所系。一个监护网络可以极大地安排一个准主体所能接触的东西，可以选择教什么，先教什么，这些都不使它落进这一判；使它落进这一判的，是把某一项内容的位置在那道门自行之前就封死。

预占之为侵犯，不因这个准主体将来会成为完备主体。它现在已经是目的价值的承担者，那个尚未运行的位置现在就是它的。侵犯所动的不是一份未来的资格，是一个当下就已归属于它的位置。

监护者应当如何行事，负何种义务，本篇不判。那是实践之批判的对象。

---

第六节 位置的恒在与殖民的存在条件

V39 殖民作用于可达，不作用于在场

殖民一律作用于可达，从不作用于在场。这一形式在两个层次上同时成立。

在源的层次上，源之在场不可被殖民消灭。可以被占据，被遮断，被锁住的，是源的可达，是它的显现路径，是它所给出的具体内容。一个人的某一行价值源可以长期无法给出任何进入候选场的东西，那一行仍然在。

在位置的层次上，认领，修订与退出这三个位置在结构上恒在。殖民所能做的，是使它们在运行上不可达。

两个层次同形，这不是形状上的相似。二者都是同一道模态事实的两次显形：构成性的东西不可由关系性的动作取消。归属，评价，安排，压力，这些全都是关系里的动作；而源与位置不是关系给出的，它们是这个承担者作为目的价值承担者的构造本身。关系性的东西可以封住通往它们的路，不能把它们取走。

恒在的是位置的归属；它的运行形态则可以尚未生成，可以已经自行，可以暂不可达。对完备主体与已自行的域，位置已在运行，可被封成不可达；对准主体尚未自行的域，位置的归属与保留已经属于它，运行形态尚未生成，而预占所侵犯的，正是这项保留将来取得运行的开放性。

由此可以立一句：完全的殖民不是主体的让渡。真能被让渡出去的东西就不叫殖民，叫替换。P1 四格中门在而弃用的那一格，正是这一判据的登记处：一个人可以在运行上完全不用那个位置，而承担的资格未失。弃用是运行上的事，让渡是资格上的事，后者不可能。这一格所名的是位置的运行形态，不含承担者致此之义：一个人自己不再用那个位置，与一个人被安排到够不着那个位置，同居此格。

可复性也随之。位置未被转让，不等于实践上的恢复总是可能。恒在所保证的是恢复仍有一个属于承担者的位置可以作为目标；它不保证这个目标在事实的条件下总能达到。可达在时间中如何实际恢复，是动力学篇的对象。

V40 殖民的存在条件含至少两个主体

原初的孤独主体不可被殖民。两支各有其理由，且两条理由互不依赖。

外部这一支要有归属，而归属是社会承认那一类材料，它至少要两个主体才有定义。自我这一支要有一份外部的担保进入承担者的资格链，而担保须有一个外来的目的，或者它在制度与历史中留下的痕迹，那同样至少要两个主体才有来处。两条路各自独立地要求第二个主体，因此这个属整体上不成立。

这一支不要求那项内容确实已在另一个主体的源图上成为价值。担保真不真不进判，那是 V36 已经立定的；这里所要的只是它有一个外来的来处。

此处所说的孤独主体，是前置篇的思想实验所设的那一个，即从未有过第二主体与任何社会历史的主体。这一条不按共时的人数读。一个曾在社会中形成过一份清单，后来独居，或者成为世上仅存者的主体，仍然可以携带此前形成的状态；那不构成对这一条的反例，因为殖民所要的第二个主体，是关系或历史上的来源，不是此刻在场的人数。

这一条可以填进前置篇已经画好的格子里。

	需几个主体	模态
价值	一个	构成
价格	至少两个	构成
共鸣	至少两个	触发
殖民	至少两个	触发

模态那一列问的只有一件事：第二个主体构不构成第一个主体的目的价值。价格构成，因为一个东西的价格就是由至少两方之间的关系给出的；共鸣不构成，它所鸣的那个结构在孤独主体处已经完整，第二个主体只使它发生。殖民与共鸣同格：第二个主体在这里供的是材料，而这个状态成立在第一个主体的位置上，它一分一毫也不构成第一个主体的目的价值。同格不等于同型，外来目的的材料确是这个状态得以成立的必要关系件，而共鸣所鸣者不待第二个主体而完整。价值可以由一个承担者成立，殖民必须有外来目的之关系或历史来源。

前置篇那一句于是得到精确的读数：一个主体的目的位在孤独的宇宙里照样立得住，因为它的立足从来不靠任何用处的买家。买家不在，价格不成立，殖民也不成立，而价值照旧。

孤独主体的价值可以变化，变化了依然是它自己的。某项内容淡了，某项从边上回到中心，某项过一阵又回来，这些是内容的动力，不朝着任何特定的维度，是价值生活的常态。两支所要的材料都不在场，那些变动便只能是它自己的。

孤独主体并非全无病理。承担者主动作废一个适用条件已经现实出现的高层源，那是 P1 划开的那一种主动的关闭；它不是殖民，因为没有外来的目的，也没有第二个主体。价值出问题的方式不止一种，殖民只是其中须有第二个主体才能发生的那一种。这也正是这个词不该被当作价值出问题的通名的理由。

---

第七节 与上游三处判据的分层

V41 与凿削局部交叠而不互相还原

殖民与权力论所判的凿削，有一处确实的交叠。凿削所立的锁定分四层，其中的解释这一层，即承受端把回流解释成自己的成长，自己的选择或自己的理性，正是两者同时成立的典型：那里既有中断权被取消，也有一项归属占着构成位。

交叠不等于可以互相还原。三处不可还原如下。

其一，凿削这个谓词本身是关系性的。它所说的是取消承受端之法在这段关系中的中断权，没有一个施为端，这句话就没有主词。殖民的谓词落在承担者自己的位置上；第二个主体在那里供的是材料，不是主词。

其二，两者的宾语不同。凿削的宾语是中断权，即承受端之法能否中断施为端的脚本。殖民的宾语是价值的构成位。一个人可以在一段关系里完全失去中断权，而他的价值构成位不曾被任何归属占据；也可以在一段并不取消他中断权的关系里，被一项不可反驳的归属占住构成位。

其三，凿削可以始终只以外部要求的身份运行；外部殖民则另须有一项自身价值的归属占着构成位。是不是公开进行，不进这一判：一套外部殖民完全可以公开宣称这就是你真正的价值，并把你的否认读作你还没有理解自己。

还有一件与自我殖民相关的。自我殖民的成立不要求当前存在一个可指认的，正在施为的外部端。它完全可以是先前的凿削所留下的运行状态，那不改变它现在的判定：作为状态，它不落凿削。这一句与状态判据一致，也与凿削须有在场主词一致，两者并不冲突。

结论的形式与本系列在权利论那边所取的相同：局部交叠，两个谓词各自独立成立，不互相翻译。

V42 与涵育不在同一层

外部作用按结构效果分为三类：锁定，释放性塑形，以及既不释放也不锁定的非塑形作用。涵育与感发是释放这一向下的两支，分它们的不是施为端可否指认，而是有没有一个可指认的承受端，以及开放这个承受端的发生空间是不是这段关系的构成性指向。一部作品的作者盼着被读懂，一个人把作坊敞开着做工乐见有人学去手艺，这些施为端都完全可指认，而它们仍是感发。殖民不在这张表上。它不是一类作用，是构成位被一个外来目的占着的状态。

由此可以更正一个容易成形的对偶。涵育的对立面不是殖民，是锁定，即有承受端的那一种，也就是凿削。外部殖民的完整签名包含锁定，并另要求候选场的占位与一项不可反驳的归属；自我殖民则根本不沿锁定与释放这一根轴来判。P1 曾以松开网格与锁死网格来说这一对，那一句可以保留，它的准确读法是：涵育所维持的那几个位置，正是殖民所分别封住的那几个位置。

上游对涵育的一个定义，说的是为他者的序列创造不被殖民的运作条件，其中的殖民取的是宽的那一义。两义各在其场域，本篇的窄义是那个宽义在价值发生场上的一个特定落法。

两者不在同一层，其结构上的理由在于两个谓词的主词不同。判据这一层已经统一：涵育与殖民都按结构性的效果判，都不问初心。所不同的是，涵育述的是一次作用，殖民述的是一个状态。同一件事里，一次作用是涵育，与某项内容的状态是自我殖民，这两句可以同时为真而不矛盾。

由此得一条界，它不是特设的豁免，是主词不同的直接推论：未完成的涵育不是殖民。一位施为者把位置全部打开，把判断交还，而那项内容终究没有接上承担者的任何一行源。他没有殖民任何人。第二根没有长成，这一事实不回头改变第一根的性质。若无这一条，每一次没有生效的教导都会被追认为殖民，而那正是判据窄不宽所要挡住的东西。

第二根究竟是什么，在其中一支上到这里可以说清了。P2 已经立过，涵育唯当承受端自己那一步实际发生才算完整，而施为者不持有承受端第一人称完成的那一份。而这一步有两支：涵育的结果若落在一项具体内容的发生上，那一步是第一人称认可；若落在结构的开维上，那一步是珍视的接续。

落在具体内容那一支上，须再分一层，因为 P2 在这里本来就是两步。内容在承担者自己的某一行源之下取得位置，是第一人称认可得以完成的前提；而第二根本身仍是第一人称认可。前提缺了，认可无从完成，第二根便长不成。落在开维那一支上则没有这层关系，那里没有一项特定的内容可供这一判据处理。

于是这里有三件事相接而不相等同：源之下取得位置是前提，第一人称认可是第二根本身，而自我殖民是在前提缺失时那次认领仍被维持为已经完成。第二根没有长成，只说明前提或认可有一处没有发生；自我殖民另要求那次认领被维持在构成位上。一项内容没有取得位置而随即淡去，那只是一次落空的认领。

这一层另有三处要按住，否则它会滑成一把结果的尺。

其一，那次施为是释放向还是封闭向，当时就判得出，看它对承担者位置的结构效果即可。这一判不能推迟。推迟等于拿结果当判据，而本系列一路拒绝的正是这个。

其二，当时未定的只是第二根会不会长成，而它本来就不在施为者手上。它未定不是判据的缺陷，是这件事的结构本来如此。一个人可以做完全部他能做的，而那一步仍然可能不发生，因为那一步只能由承受端自己走。

其三，发展受阻不作判据。那是一把结果的尺：按它，一段真涵育若恰好没有带来可见的进展，就会被追认为殖民；一段殖民若恰好伴随可观的成就，就会免判。位置的判据不看产出。已认领的内容此后如何起落，重心如何位移，是动力学篇的对象。

涵育在这一点上还有一层，它与 15DD 那一行相接，而这一层不给涵育添任何新的条件，它是 P2 已立那一句的直接读出。

那一句是：施为者不持有承受端第一人称完成的构成份额。一个施为者要开放某个承受端的发生空间，就得对这个承受端有所判断，而他者的价值不可穿透地读取，故这个判断从来只是在自己的法里成形的一幅图。图可以画得准，也可以画得不准，这不是分别所在。分别在于他把这幅图当作什么：把它当作自己据以开路的判断，承受端那一份仍在承受端手里；把它下发过去，当作对方已经完成的自我认识，那就是在主张持有对方的那一份，而按 P2 自己的条件，那已经不是涵育。

保留与填死之别，与准主体那一节的维护与定格之别，是同一道线在两处的两次出现。

两条边界防这一对偶被读得过宽。非塑形作用既不涵育也不殖民，第三类不可被吃掉。一次涵育性的施为不因它是涵育而免于审计，判据仍然只看位置。

V43 与冒领同形而多两件

本框架在别处立过一个更一般的形式：一个构出示了它不曾挣得的地位凭证，即判定为出示多于持有。这一判定回溯而不问动机，其解药是对账与被追问。

内摄正是这个形式在价值发生场上的一个落法。P2 已立，可授可撤的东西没有构成的权限；社会承认在这里所出示的构成凭证，恰是它不持有的那一件。两处独立立出，形状相同。

外部殖民比一次单纯的冒领多两件。其一，这次冒领已经在承担者的发生架构中占用了构成的读数，即不只是外部这样出示，承担者那一侧的读数同样被占。这是结构上的事，不读作承担者主观上的赞同。其二，针对该项归属的对账通道已被封：承担者的否认不能作为对这项归属的否认被登记。此处所封的是这一项归属的反驳通道，不是承担者反思，重述，修订与退出的全部路径，后者是锁定那一条所判的事。冒领可以在对账的通道敞开时发生，也可以在那里被纠正；殖民是这道通道已封的那个状态。

自我殖民一支在形式上同样是出示多于持有：那份外部的担保所实际持有的，至多是一项外部的要求，一个权威的位置，一份制度的读数，或者一句循环的说法；而它所出示的，是这项内容已经在这个承担者自己的某一行源之下取得了位置。那份担保本身真不真，不改变这一点，它都不持有后一项凭证。用途的读数是这一形式最常见的一例，它持有的是这个承担者对某个目的有用，出示的是这个承担者凭什么算数。但这一点不给它第三人称的判据。要判定这一次出示多于持有，就得知道这项内容不由承担者的源给出，而那要读它自己的法。上界在两支上是同一条，这一点本身值得记下：一般的形式可以在两支上同样成立，判据的可及范围却不因此扩大。

两者不互相还原。本篇的判据由三位模态的不对称独立立出，不由那个更一般的形式推出；这一条只登记两者的关系，不把殖民定义为冒领的下位种。

由这层关系另出一句。自问担保不了自己的全部盲区，这是那个更一般的判定所立的；自我殖民之所以从内部难以读出，其结构上的理由同此，即每一次自问都在同一份担保之下进行，而单次的切片上没有可供区分的结构差异。

---

第八节 下游的射程

外来目的占据发生位之后，源的可达如何变化，主导的重心如何位移，一项内容为何在此刻淡去而在彼刻回来，这些是动力学篇的对象。那里有一件与本篇直接相接的事：殖民不改变源的在场，它锁的是源的可达与发生的路径。本篇给出的是这道锁的结构，不写它在时间中如何松开。

一个人有多项已认领的价值内容，而它们在具体情境中要求不相容的行动，如何裁断，是冲突篇的对象。本体上不排序与实践上必须决断，两者并不冲突。

他者的价值为何不可穷尽读取，不可以外尺比较，不可统合，是他者篇的对象。本篇两支所共用的那道上界，即第三人称判据的定义域边界，是那道认识论上的限界在病理学这一侧的显形。二者同指一处，那一篇正面立起它。

已认领的价值内容如何走向行动，是展望篇的入口。

本篇给出判据与位置的读数，不给规范的效力。一个成立的殖民是否当被解除，谁有资格施行审计，监护的网络在准主体处负何种义务，这些各有其位置，本篇标出地址即止。它们是实践之批判的对象。

三位的模态各不相同，因此一项外来的作用落在哪一位上，有确定的读数。显现位可以被占，构成位可以被一项不可反驳的归属占着，也可以由一份外部的担保维持为构成已经完成，可修订位可以被预先作废。这些都发生在同一组位置上，而这组位置从来不曾易主。殖民从未真的构成过什么，也从未真的取走过什么。它所能做的一切，都落在可达那一侧；认领，修订与退出的位置，始终是承担者自己的。

---

致谢

本篇为 SAE（Self-as-an-End）哲学体系价值论分支的第三篇，由作者独立完成。成稿过程采用四方交叉评审方法：内部精度审计，本体极限与矛盾检测，边界测试，结构闸门与发布前签署。本篇的两处承重经评审改定：自我殖民的占位者由他者法中已成之构改为外部的担保，使不实的说法与无人真正持有的规范脚本一并落入判内；运行中的读法撤去全部可操作的检验，只留第一人称的回溯。四方评审显著提升了本篇的精度与一致性，谨此致谢；文责由作者独任。

---

参考文献

以下文献均存档于 Zenodo，中英双语版本同记录；引用 DOI 为概念 DOI，始终指向最新版本。全部文献单一作者署名（秦汉），合作者于致谢致意。

秦汉. 系统、涌现与人的条件：以主体作为目的本身为核心的规范社会理论（SAE 基础论文 Paper 1）. Zenodo. https://doi.org/10.5281/zenodo.18528813

秦汉. Internal Colonization and the Reconstruction of Subjecthood（SAE 基础论文 Paper 2）. Zenodo. https://doi.org/10.5281/zenodo.18666645

秦汉. 三层二维统一结构：Self-as-an-End 的完整框架（SAE 基础论文 Paper 3）. Zenodo. https://doi.org/10.5281/zenodo.18727327

秦汉. 自己的法：SAE 对伦理与道德的批判（SAE 基础论文 A16）. Zenodo. https://doi.org/10.5281/zenodo.19037566

秦汉. 方法论总论：凿构循环（SAE 方法论）. Zenodo. https://doi.org/10.5281/zenodo.18842449

秦汉. 余之道 · Via Rho（SAE 方法论 Paper 00）. Zenodo. https://doi.org/10.5281/zenodo.19657439

秦汉. 论权力的起源（SAE 权力论系列 Paper 1）. Zenodo. https://doi.org/10.5281/zenodo.20370225

秦汉. 论权力的凿削（SAE 权力论系列 Paper 6）. Zenodo. https://doi.org/10.5281/zenodo.20647463

秦汉. 论基础权利的厚度与边界（SAE 权利论系列 Paper 2）. Zenodo. https://doi.org/10.5281/zenodo.20769253

秦汉. 论基础权利的限与侵（SAE 权利论系列 Paper 5）. Zenodo. https://doi.org/10.5281/zenodo.21212595

秦汉. 论权利与权益（SAE 权利论系列 专论）. Zenodo. https://doi.org/10.5281/zenodo.21314907

秦汉. 价值不是什么（SAE 价值论系列 前置篇）. Zenodo. https://doi.org/10.5281/zenodo.21362575

秦汉. 论价值的源（SAE 价值论系列 Paper 1）. Zenodo. https://doi.org/10.5281/zenodo.21405207

秦汉. 论价值的发生与承认（SAE 价值论系列 Paper 2）. Zenodo. https://doi.org/10.5281/zenodo.21437038

秦汉. SAE 第一批判. Zenodo. https://doi.org/10.5281/zenodo.21439120