Interfacing with Existing Traditions, and the Prospect
传统理论对接与展望

Introduction

When a series is finished, the easiest thing to do is to say one more thing on its behalf that it never said.

The prequel cleared the space of reference. The word value is claimed at once by the market, by ethics, and by ordinary language, so that one word carries at least three things; the five misplacements the prequel excluded by structural negation are price, direction, the pricing of the remainder, the ladder, and talking past one another. P1 then set up the five machines the prequel had marked out, P2 established the occurrence of value, P3 read the colonization of value, P4 gave the readings value takes in time, P5 established adjudication in conflict, and P6 established the value of the other.

All seven work inside the machine. Each has its object, each has its load-bearing propositions, and each keeps the same disciplines: no external theorist is named, no historical event is invoked, and no normative force is drawn upon. Those disciplines make the seven papers clean, and they also make certain sentences unsayable within them. The seven cannot say what this machine looks like as a whole, and they cannot say where it parts company with machines other people have built.

This paper steps outside and says those sentences. Standing outside takes a different discipline, and this paper is therefore exempt from the rule against naming external theorists. The exemption reaches only the naming: this paper still adds no proposition to the object level, sets no proposition numbers, and still draws on no normative force there.

A word on this paper's place in the series. Four published texts, the prequel and P1, P2 and P3, wrote down a piece of unfinished work and named its recipient a prospect paper. This paper is where those forward assignments are settled, and settling them is not the same as inheriting that name. Section Seven gives the reason: a forward assignment points at a piece of work, not at a permanent bibliographic type for some future product. That work was already reassigned between P4 and P5. This paper states that reassignment in public, by the second of the three routes set out in Section Seven, and is written as Paper 0. The prospect paper is therefore this paper's historical alias, not its place in the series.

---

Section One · What Kind of Machine This Is

One · Not one principle, but three ledgers

The prequel cleared the space of reference and the six papers that followed set up the machine; together they are one construct. What that construct set up is easily collapsed into a single elegant principle, and collapsing it distorts, because what the six papers read is not one kind of thing.

One general remark first: the three ledgers below are analytic ledgers, not bins for papers. What is recorded at the end of each of the three passages is who set that ledger up; those who work on it are not only they, as the matrix that follows shows.

The ledger of constitution. It asks how a content becomes the value content of some bearer. The words on this ledger are: domain of bearers, source of value, candidate, first-person avowal, the standing of a value content. It was set up by P1 and P2.

The ledger of operation. It asks how what has already come to hold moves in time. The words on this ledger are: the presence of a source position, reachability, the rise and fall of particular contents, the centre of gravity, revision and exit. It was set up by P4, while P3 reads what this ledger looks like when an external end has interfered with it.

The ledger of action and relation. It asks what happens when two demands collide, and what occurs between two bearers. The words on this ledger are: the ordering of action, adjudication, cultivation, colonization, resonance, joint audit, boundaries where two people meet. It was set up by P5 and P6.

A paper may cross ledgers, and most do. P3 comes in from outside through relational conditions, works on the running channels of appearance, reachability and revision, and at last generates a mistaken reading that a constitution has been completed; it crosses all three. P6 returns to the ledger of constitution in its first proposition, stands at the layer of epistemic authority in its second, and only in the later ones works on the ledger of action and relation.

The ledgers are not there to give each paper one bin. They are there to stop the subjects of one ledger from being claimed by another. On the ledger of constitution the subjects are contents and positions; on the ledger of operation they are readings; the ledger of action and relation is a compound ledger, divided into an action column and a relation column, whose subjects include demands for action, orderings of action, events, relational channels, authority and boundaries. The distinction among these subjects, together with the two bars on back-flow set out below, is among the most important reasons why the six papers cannot be compressed into one.

The relation between the three ledgers and the six papers can be set out as a matrix. It reads as what each paper did on each ledger, not as which ledger each paper belongs to.

Paper	Ledger of constitution	Ledger of operation	Ledger of action and relation
P1	domain of bearers; sources of value; the outward-facing content of one's own law	the unfolding of sources	the conditions side of cultivation
P2	first-person avowal; the standing of a content	avowal taking shape over time	the conditions side of social recognition
P3	an external end occupying and counterfeiting the function of the constitutive position, without taking the position away and without any constitution being completed outside	reachability; locking; channels of revision	external ends and relational structure
P4	the boundary between constitution and operation	presence, reachability, particular contents, centre of gravity	relational facts that can enter third-person audit and joint verification and yield fallible, revisable readings
P5	standing does not flow back	a content set aside is still held	demands for action; conflict; adjudication
P6	the registration of that one content; sources do not merge	the limits of third-person readings	standing of value against readings of the world; resonance; joint audit

Two · Two bars on back-flow

Setting the three ledgers side by side is not yet enough, because what matters is not that each speaks for itself but what may not pass between them. Taken together, the seven papers set up two such bars.

The first. Readings of operation, action, consequence and relation do not, of themselves, flow back into a source or a standing of value.

P4 says it most directly in establishing that output and position are independent axes: to put something visible in place of something in principle invisible, to let what is easily had stand in for what is hard to have, is a move that often works elsewhere and does not work here, because the reading substituted in does not prove the thing that had to be proved. P5, in establishing conflict, says that setting aside is not exiting, because setting aside cannot do what exiting does. P6, in establishing comparison, says that the conditions, options and consequences in the world can be checked together, and that those readings do not flow back into standing. The three say the same thing.

The second. External conditions may shape appearance, reachability, and the channels of action and relation, yet do not, of themselves, stand in for a first-person constitution.

P1, in establishing cultivation, says that one bearer can loosen another's grid. P3, in establishing colonization, says that an external relation can form an occupation at the position of appearance, counterfeit constitution by introjection at the constitutive position, and form a locking at the position of revision. P6, in establishing the other, says that appearance, social shaping and cultivation may all come from outside, while the constitution by which a content becomes his value content is not accomplished at my place. These three also say the same thing.

These two are what the six papers jointly made. They are not derived from some more fundamental principle; they are the two lines that remained between the ledgers after each paper had done its work on its own.

Three · Four nodes

Four nodes are used repeatedly across the three ledgers. They are listed here, without any claim that they are equivalent to one another.

The domain of bearers. End-value is borne by bearers. P1, in establishing the domain, divided it into four slots, of which the fourth, doorless, for example an animal; a doorless object has its experience and can be an end that is cared about, and it does not fill that position as a bearer of end-value.

First-person avowal. A content becomes someone's value only through his own avowal, and P2 writes that avowal is not verification: the bearer is not ascertaining a fact that already exists, but taking it up within the workings of his own law.

No common external measure. P1's leading proposition writes that once something is the bearer's own and issues from his own law, there is no longer any coordinate system outside that law which can pass value judgement on it across layers or across subjects.

The source is not placed outside. P1 states it uniformly: across the domain of bearers, the source of end-value is anchored on the condition that the standing of being an end is not placed outside; for a complete subject, the running of that source is anchored on the autonomous reflexive operation of his own law; for a subject-to-be, the source is already there while the manner of running is still forming.

The four support one another; they are not equivalent. That the source lies on the bearer's side does not by itself entail that there can be no cross-subject aggregation: a theory that makes utility depend on the subject and still builds an aggregation function is not, in point of form, contradictory. Conversely, holding that there is no common measure does not by itself entail that the source depends on the subject: an objective value pluralism may perfectly well hold both that values are objective and that they are incommensurable with one another. And that value is borne by bearers does not by itself entail that first-person avowal is the only constitutive event; that last requires P2's own argument. What runs between the four is dependence, not equivalence, and this paper does not collapse them into one sentence.

Four · This machine is itself one construct

The methodology of this system established the chisel-construct cycle: every construct carries a remainder (ρ) that it cannot chisel away.

The seven papers are one construct. They have their object, their criteria, and the things they can and cannot read; they therefore have a remainder. And the remainder is not everything Section Six has to register: there are also objects adjacent within the jurisdiction that were never set up as questions of its own, limits already established, a candidate this paper proposes, handovers whose address is already fixed, and an object-level open question that only came to light on looking back. Calling all of these a remainder would cost that word its precision inside this system. Section Six therefore registers in six slots.

---

Section Two · The Layers and the Method of Interfacing

This section comes before the Kant section and before the eight direct engagements. The reason is that once a line of divergence is drawn loosely, placing slides into judging, and an interface attempted before the layers have been separated will most easily read two remarks made at different layers as a head-on conflict.

One · The layers are not the same

Metaethics, value theory, normative ethics and welfare theory are not at one layer.

Metaethics asks what evaluative judgements are, whether they have truth values, and where those truth values are grounded. Normative ethics asks how right action is determined. Welfare theory asks what is good for a person. The layer this series occupies asks how a content becomes the value content of some bearer, how it runs, and what to do when two of them collide.

The distinction matters because metaethical positions mostly give no answer to this series' first question. A value realist may assign value to persons, and may equally assign it to states of affairs, to experiences, to activities; realism by itself does not settle who the bearer is. To take a metaethical position as an opponent in a dispute about bearers is to compress two layers into one.

Two · Three separate registers: constitutive dependence, truth-grounding, epistemic authority

Three questions are often carried along by one word. This section separates them and states, for each, what this series has and has not established.

Register one · constitutive dependence. Whether a particular value content depends for its holding on the bearer's avowal. This series has a definite position here: P2 established that first-person avowal is the only position on the axis of occurrence that constitutes a value content, and that avowal is not verification.

Register two · truth-grounding. What makes judgements about avowal, value and relation true. This series has taken no complete position here, and this paper does not take one on its behalf. A distinction is worth marking: that a value content comes about by constitution, and whether judgements about it have an objective ground of truth, are two different things. Even if the content comes about by constitution, the following may still be objective facts: whether he ever completed that avowal; whether that content still holds at this moment; whether the channel of revision has been closed; whether some external ascription is counterfeiting constitution. So constitutive dependence on the subject is not anti-realism about truth. Within metaethics there are several ways of drawing the line between realism and its alternatives, and of deciding whether objectivity requires mind-independence; this series does not settle that whole layer at one stroke on the strength of constitutive dependence.

Register three · epistemic authority. What the first person, the third person, and joint procedures can each know. This series has a partial position here: P4 established underdetermination before a third person, P3 established the upper limit of audit, P6 established how authority is distributed in joint audit.

This paper therefore says only this much about where the series stands in metaethics: it has a definite position on the constitutive dependence of particular value contents, a partial position on epistemic authority, and no closed position on a full semantics, ground of truth, or theory of objectivity. This is not modesty; it is a guard against readers filing the series prematurely under Humean constructivism or anti-realism in general.

Three · Three kinds of subject-relativity do not entail one another

The first is good for a person, that is, something that makes a person's life go better. The second is an agent-relative reason, that is, a reason whose form ineliminably mentions that agent. The third is mind-dependence, that is, whether some value fact metaphysically depends on attitudes, practice or rational activity.

These do not entail one another. A realist may perfectly well accept agent-relative reasons; a constructivist may construct requirements that are universally valid and agent-neutral.

What this series means by saying that value is borne by bearers falls on the constitutive side, not on the first two. To say it is borne by bearers is to say that the constitutive act by which a content becomes a value content lies on the bearer's side; it does not say that value is always good for someone, nor that every reason must mention the agent in its form. This distinction is load-bearing when Section Four engages consequentialism and Sen.

Four · Three senses of a standpoint outside the subject

The first is a mind-independent ground of truth: the truth or falsity of evaluative judgements does not depend on anyone's attitudes. The second is an impartial standpoint: an evaluative position that does not change according to who is acting. The third is a universal practical standpoint that any rational subject can occupy.

What this series' leading proposition denies is the kind of external standpoint that could pass judgement on the bearer's own standing of value. That denial falls in the second sense, and is confined to the object standing of value; it does not amount to a verdict in the first sense (see register two above), nor does it deny that some impartial evaluative position may hold for other objects. The third sense differs from the first two: the standpoint of universal legislation is not the view of an observer standing outside everyone, but the legislative position that any rational will, as a rational will, must be able to occupy. These three senses are load-bearing when Section Four engages consequentialism and the Mohists.

Five · How interfacing is done

To put a tradition into these coordinates is to say where it draws its problem and along which line it parts company with this series. It is not to say who is right.

That is not easy to keep to. Once the line of divergence is drawn loosely, placing slides into judging; and if the opponent is written a little weaker, placing becomes self-praise. This paper therefore takes a more laborious route: each tradition is first stated in its own strongest form, then the line of divergence is identified, and only then is it said which side of that line this series takes. Where this series has reached no judgement, that is written down as no judgement reached; the weakness of an opponent is not made to stand in for one's own proof.

This rule is not this paper's own invention: the Power Theory of this system already has a meta-level self-description, in which it is written that interfacing with a tradition means locating the traditions within a coordinate system, not adjudicating between them. This paper follows that.

---

Section Three · The Interface between Kant's Dignity and the Source Not Being Placed Outside

This section is not one of the traditions engaged. Kant is the ancestry this series acknowledges, and the relation between the distinction of dignity from price and this series is more specific than the word ancestry suggests: on one mark, the two coincide in structure.

Coincidence is not identity. This section first says where they coincide, then says how far the coincidence runs.

One · A shared negative mark

In the second section of the Groundwork Kant separates what has a price from what has dignity. The mark is replaceability: what has a price can be replaced by an equivalent, and what is raised above all price, and therefore admits of no such equivalent, has dignity.

When the prequel cleared the space of reference, the first thing it did was to cut price away. It did not do so by declaring price low and end-value high. One of the two locks on the emblem forbids exactly that reading: pearl and casket divide reference only, and are not ranked. It cut by way of conditions of constitution:

> The number of subjects is not rhetoric here but constitution: exchange requires a counterparty, conversion requires a public measure, substitution requires an equivalent; every constitutive act of price contains a counterparty position, and the constitution of end-value does not contain that position.

Substitution requires an equivalent. This sentence puts the two in contact on one negative mark. The difference lies in which side each argues from, and in what else each carries: Kant argues from the side of standing, saying that the standing of a person admits of no replacement by an equivalent; the prequel argues from the side of conditions of constitution, showing that every constitutive act of price requires a counterparty position while the constitution of end-value does not. The thought experiment of the solitary subject is the form of that showing: in a universe with only one subject, the conditions for the constitution of price all fail, and end-value holds all the same.

The coincidence here is real, and its extent must be stated exactly. Admitting of no equivalent is a negative mark. It marks the boundary between dignity and price; it is not a full constitutive account of dignity. Kant's dignity is connected further with personality, with moral capacity, and with autonomous legislation, and those are not exhausted by this one mark. So what this section claims is that the two share this one negative mark, not that this series has given a full interpretation of Kantian dignity.

Two · Four objects that must be kept apart

Pressing on from this coincidence, it is very easy to press too far in one breath. Four objects lie close together here, and this section separates them one by one.

First, the standing of a person as an end. In this series that holds prior to anyone's avowal. P1 writes that the other is an end, and this is a fact, and it holds before I recognise it.

Second, the first-person standing of value of a particular content. That holds only through the bearer's own avowal. P2 writes that avowal is not verification: the bearer is not ascertaining a fact that already exists, but taking it up within the workings of his own law.

Third, the anchoring of the source of value. This says on which side the position that gives value lies, namely that the source is not placed outside.

Fourth, the structure of colonization. This is P3's object: a content is treated at the bearer's place as though its constitution were complete, while that constitution was not accomplished within his own law.

The four are not one thing. What first-person avowal constitutes is the second, not the first: a person who does not avow his own standing as an end does not thereby lose it. The third is a structural fact about a position; the fourth is a structural event about relation and operation. Compressed into a single sentence they read forcefully and state something untrue.

Three · Colonization occupies a function; it does not take the position away

An earlier draft wrote the violation of dignity as having that constitution accomplished elsewhere, and called it the position being taken. This passage has been corrected twice, and was wrong both times, so both are recorded here. The first draft, writing the position taken, said too much and conflicted with P3's permanence of positions; the revision, writing that it is not the position being taken but only a reading, over-corrected and struck out P3's own language of occupation and introjection along with it. The correct slot lies between the two, and P3 had already written it.

P3 establishes three things at once, and all three must be said or the account will go wrong. First, no real constitution is completed outside: in establishing the structural definition of colonization P3 writes that a content is treated at the bearer's place as though constitution were complete, while that constitution was not accomplished within his own law, and what occupies that position is an external end. Second, the three positions are structurally permanent: colonization always acts on reachability and never on presence, and the constitutive position stays locked within the bearer's own law. Third, occupation, introjection and locking remain P3's own formal structural terms: in establishing that three acts fall at three positions, P3 writes that occupation falls at the position of appearance, introjection at the constitutive position, and locking at the position of revision, and in adjudicating colonization it asks head-on whether the constitutive position is occupied by an external end.

So the accurate statement is this: an external end occupies the function of the constitutive position, so that an external warrant counterfeits, in relation and in operation, a first-person constitution already completed, while that constitution was never really accomplished outside and the constitutive position was never taken away. Occupying a function and taking a position away are two different things.

This distinction also explains why P3 confines the conclusions of audit to the form of positions and signatures: the constitutive position is permanent, what the outside can obtain are readings on relation and channel, and a third person cannot close a first-person constitution out of those readings. What audit can ask is therefore only whether that position stands open or closed at this moment, and whether those paths are passable.

Four · How far the coincidence runs

The conclusion of this section can be written in one sentence:

> Kant's dignity and this series' source not being placed outside coincide in structure at admitting of no equivalent, at autonomy, and at qualification not being wholly placed outside; and they are not the same concept. Colonization is one particular structural way of violating subjecthood; it is not a full definition of the violation of dignity.

Saying coincidence is useful; saying identity does harm. It is useful because the argument from conditions of constitution, by which this series cut price away, yields from within itself a neighbouring boundary of non-substitutability, and that route need not first posit an absolute value. It does harm because Kant's dignity carries with it personality, moral capacity and autonomous legislation, for none of which this series has supplied a counterpart; to claim identity is to claim on this series' behalf what it never wrote. Nor does that boundary constitute a second proof of Kant's proposition about dignity: not needing an equivalent and never admitting of replacement by an equivalent are not modally the same sentence.

Five · Where this series stands in a live interpretive dispute

There is a dispute in Kant scholarship about dignity, and since this section's reading takes no side in it, where it does stand must be said clearly.

One side may be called the substantive value reading. Wood places the objective value of rational nature at the centre of the theory and takes the passage from the formula of universal law to the formula of humanity to be a passage from pure form into substantive content. This side tends to say that humanity has an unconditional value, and that the duty of respect is explained by that value.

The other side is not a simple formalism. Sensen objects to reading dignity as an absolute value property that we first discover and from which the duty of respect is then derived; he ties dignity and respect more closely to rational law and prescription. Bojanowski agrees that Kant is not to be read as a moral intuitionist, and yet argues against denying on that ground the absolute value of humanity, holding that the idealist structure of practical reason and absolute value can stand together.

This series takes no side in that dispute, because its account of the source not being placed outside is neither a property nor a prescription but a position. The position that gives value lies on the bearer's side, and that can be shown from conditions of constitution; the thought experiment of the solitary subject shows precisely this.

That position has its own costs and benefits within the dispute. The cost is that it cannot, as the substantive value reading can, derive a duty directly from a value: this series draws on no normative force at the object level and cannot say what a person ought to do. The benefit is that it need not answer the question that embarrasses the substantive value reading, namely what kind of thing that absolute value is. And it must be said honestly that taking no side is not the same as having resolved the dispute. What this series offers is another route, and whether that route can replace what the other two sides set out to explain is not judged here.

Six · The strongest objection from the Kantian side, and this series' two steps

There is an objection to first-person constitution that is ready to hand on the Kantian side, and it takes a strong form: humanity is an objective end, and not because a person first chose, in the first person, to have dignity; a person who does not avow his own dignity, who even avows a maxim of self-enslavement, is not thereby released from the formula of humanity.

This series agrees with that, and it was P1 that established it. P1 divided two steps. The first is a fact: this particular other too is an end, and that the other is an end holds before I recognise it. The second is a response: to recognise and to continue that fact from the first-person position, where recognition acknowledges the fact and carries it on rather than making it.

What first-person avowal achieves is never anyone's standing as an end, but the second of the four objects listed in this section: a particular content taking the standing of value under this bearer's name. The standing of an end lies at the step of fact, the avowal of a particular content at the step of response; once the two steps are apart, that objection is not aimed at a position this series holds.

There is one further thing to keep apart, and here Kant scholarship is itself divided. Respecting another's humanity and approving of every end he has chosen are not the same. Dean brings the morally good will into the core of humanity, while his critics reply that even where a person's particular choices are bad, the capacity to make choices must still be respected.

What this series gives, on the side of value theory, is one unfolding of that distinction. The three limits P6 established say what I have no standing to do with another's values: reading cannot close into a reading guaranteed to be exhaustive, standings of value admit no common external measure, and what happens in a relation is not enough to merge sources that already exist. And not one of those three says that I must approve of every content under his name. Having no standing to dispose and being required to approve are two different things; this series judges the former and does not judge the latter.

---

Section Four · The Eight Engaged Directly

The eight are ordered by their line of divergence, from the innermost outward: first what makes the standing of a particular content hold, then whether ranking is possible, then the question of the other, and last whether there is a standpoint outside the subject.

One limitation must be stated first. The divergences discussed below are confined throughout to one subject, the standing of a particular value content. This series does not hold that every value fact fails to exist prior to recognition: that the other is an end is a fact prior to my recognition, the structural shape of a source of value is not made by any one avowal, and the conditions and relational channels in the world hold on their own. So the fault line in this section is not whether value is discovered or constituted, but this: whether the standing of a particular end-value content under some bearer's name is constituted by his first-person avowal, or is to be explained by a fittingness, a relation of reasons, or an objective rank order independent of actual avowal.

Four.1 · Fitting-attitude theories and buck-passing

The common form of this family is not that whatever I actually like has value. It is that the evaluative standing of a thing is to be explained by what attitude is fitting towards it, or by what normative reason we have to take some attitude towards it. Brentano on what is worthy of love, Ewing on the fitting pro-attitude, Gibbard on apt feelings, and Scanlon bringing the question back into contemporary value theory, are several key nodes of one family of thought. They are not a chain of succession passed on in order, and this paper does not narrate them as one.

Buck-passing is a stronger step. Scanlon holds that being valuable is not an additional first-order reason-providing property; to say a thing is valuable is to say that it has certain other properties, and that those properties give us the corresponding reasons to value it or to respond to it in some way. Scanlon's own terms are other properties and natural properties; lower-order properties is a later convenience.

This step is often divided into a positive claim and a more contested negative one. The division is useful, and what the positive claim says must be fixed first: if it says only that value has a constitutive connection with some fitting response, most of this family accept it; if it already says that the reasons are given by other properties rather than by a value property, that is already part of Scanlon's own version and cannot be read back into earlier writers.

Here is the line of divergence. The strongest form of this family's objection to this series is not that experts know better than the person himself. It is a conceptual argument: a standard of fittingness or of reasons is not constituted by actual avowal itself. A person may actually approve of something through deception, through fear, through a false factual belief; conversely, he may fail through depression, ignorance or distorted socialization to appreciate what was worth valuing. If the evaluative standing of a content is explained by a fitting response, actual avowal may be neither sufficient nor necessary.

This series' answer does not go around that. P2, in establishing first-person avowal, writes it plainly:

> Still less is it verification. The bearer is not ascertaining a fact that already exists, is not determining whether he really does care about something, but is taking it up within the workings of his own law.

The divergence can therefore be put exactly: fitting-attitude theories tie the evaluative standing of a particular content to a standard of fittingness independent of actual avowal, and this series ties it to that act of taking up within the workings of the bearer's own law. These are two different constitutive accounts given to the same grounding question, not two observations of the same empirical fact.

The best-known difficulty inside this family comes from the same place, the wrong kind of reasons. A demon may threaten you with punishment unless you admire something worthless; you then do have a prudential reason to admire it, and that reason seems not in the least to make it admirable. So the problem is not to find a reason, but to separate, without circularity, reasons given by the object from reasons given by the consequences of the attitude. Rabinowicz and Rønnow-Rasmussen, and Olson and others, have treated this difficulty specifically. If in the end the right kind of reason can only be defined as the kind that reflects the value of the object, then the direction of explaining value by reasons has come back round. This paper does not adjudicate that difficulty; it notes only that it arises at the same place as this series' line of divergence.

Four.2 · Scheler's material value ethics

This tradition is not on the usual list. This paper engages it because it is the most direct opponent of this series' leading proposition.

Scheler takes values to be objective value qualities given in intentional feeling, and holds that there is an objective order of higher and lower among value modalities. His basic value modalities are four: the agreeable and disagreeable, vital values, spiritual values, and the holy and unholy. The useful is not among these four; it is a derivative value founded on the agreeable, and Scheler's own term for it is derivative or consecutive value. There are reconstructions in the literature that give the useful a stronger place, and there are rebuttals of those reconstructions on the strength of the text; this paper takes the four of the text and records that disagreement.

A straw man must be avoided first. Scheler's rank order is not the ladder the prequel rejected. The ladder the prequel cut away puts values of different layers onto one measure and ranks them; Scheler's order denies precisely that there is a homogeneous unit. The two traditions do not diverge over the measure but over the ground of the order. Nor may it be inferred from this that the layers are incomparable for him: he positively holds that there is an objective higher and lower.

The divergence is one of foundations, and it cannot be adjudicated within shared premises. Scheler places the ground of the value order in disclosure through intentional feeling; this series ties the standing of a particular value content to the bearer's own law and to first-person constitution. The two have no common ground, and therefore do not prove anything against each other. This series' leading proposition can refuse that rank order only within this series' own premises; it does not constitute an internal refutation of Scheler, and equally, Scheler's rank order does not constitute an internal refutation of this series.

This is where the discipline of placing rather than adjudicating bears the most weight. To say that this series has already refuted Scheler is to take one's own premises for public ones; to say that the two cannot speak at all is to erase a real divergence. The accurate statement is one sentence: the divergence lies at the foundations, and the choice of foundations is not settled by either side's internal argument.

What connects with the previous tradition is this: both tie the evaluative standing of a particular content to something independent of actual avowal, one to fittingness and one to an objective rank order. P2 carries the constitutive side of both divergences. And Scheler has one further point that P2 does not settle: his objective rank order also parts company with P1's no common external measure.

Four.3 · Consequentialism and agent-neutral value

The line of divergence changes here. The previous two disputed what makes the standing of a particular content hold; this one and the next dispute whether ranking is possible.

The most general form of consequentialism is not the maximization of pleasure. It ties the rightness of an action functionally, in some way, to an evaluation of outcomes: which total outcomes are better does not change according to who is acting. That is the agent-neutral version; there are also agent-relative versions that allow each agent his own evaluative function.

A straw man must be avoided first. A standard maximizing model usually assumes a sufficiently rich ordering of outcomes, while cardinal utility and a single physical unit are not requirements of consequentialism as such; many versions do with an ordinal better-than relation. Nor is it a logical truth that the ordering must be complete: a version admitting incomparability and working with a partial order can still select maximal elements, though not necessarily a unique best.

Where the line of divergence lies has to be separated first. A general agent-neutral consequentialism needs an evaluation of outcomes that does not change with the identity of the agent, and what it evaluates need not be any person's standing of end-value; the two are not the same. This series treats how a particular value content comes to hold under a bearer's name; it is not a complete theory of the value of states of affairs, and therefore cannot directly supply consequentialism with an evaluative function over outcomes. That is in the first place a matter of something not supplied, not a head-on conflict. And the three senses distinguished in Section Two are to be kept in mind: what consequentialism needs is impartiality in the second sense, not necessarily a mind-independent ground of truth in the first.

What this series can say here must be said in three layers, or the ordering of outcomes will be pressed back into the standing of value.

The first layer is what is not supplied. Consequentialism hands the evaluation of outcomes and states of affairs to a theory of the value of states of affairs; what this series treats is how a particular value content comes to hold under some bearer's name. These are not the same module awaiting completion, and this series cannot supply that function.

The second layer is a definite conflict. If some version places a bearer's standing of value itself onto a common scale, or takes its ordering of outcomes directly from that scale, it conflicts head-on with no common external measure. That conflict is real.

The third layer is what is not judged. If the ordering of outcomes uses welfare, harm, resources, capabilities, satisfaction of rights, risk or some other informational basis, then what it orders is not any person's standing of value, and this paper cannot support or refute it on the strength of value theory alone. An agent-neutral evaluative position and an external position that could pass judgement on a person's standing of value are not the same thing; equating them is exactly the kind of back-flow barred in Section One.

And so, that this series holds standing to admit only of holding or not holding, and not of amount, does not prove that a complete ordering of outcomes cannot be had: even where one input takes only presence and absence, one may still build a lexical order, set inclusion, counts of satisfaction, priority rules or other outcome variables. That this series has not supplied a theory of outcomes is true; that this series has proved that no complete outcome function can be had is not.

The objections from the deontological side and this series' own form a neighbouring opposition on certain agent-neutral totalizing moves, while their grounds and their subjects are not the same. Agent-centred restrictions say that some acts should not be performed by me even if performing them would reduce the total quantity of like wrongs. Scheffler's agent-centred prerogative is milder: even granting that the best total outcome is a genuine reason, an agent may have a permission not to sacrifice his own projects in a fully impartial way. Williams' integrity objection holds that treating a person's projects as merely one input into the world's total utility function changes his relation to those projects.

Korsgaard has warned of one thing: agent-relative and deontological are not logical synonyms, and that formal distinction cannot be equated directly with the division between consequentialism and deontology. That reasons belonging to whom, reasons mentioning whom, and reasons being shareable are three things easily run together is this paper's own summary, not her words. This series is not filed under agent-relative value theory merely for sharing an opponent with the deontological side: what this series judges is not whether reasons mention the agent in their form, but on whose side the position that constitutes a value content lies.

Four.4 · Sen, and the cross-subject informational basis of public evaluation

This is the hardest of the eight to hold off, and what it brings to light is not a question about where one quantity belongs but a whole stretch of ground this series has never entered.

Three equations must be removed first: preference is not utility, and utility is not welfare. In formal theory preference is usually a comparative relation among options. Utility in modern decision theory is in the first place a representation of preference, not a Benthamite quantity of psychological pleasure open to direct observation. Welfare says how a person's life goes for that person. A preference-satisfaction theory of welfare may treat satisfied preferences as constitutive of welfare or as its indicator, but that is a substantive theoretical assumption, not the mathematical definition of a utility representation.

Harsanyi's theorem can be held off. Under expected-utility axioms and certain consistency conditions, social evaluation has a representation as a weighted sum of individual utilities. What is at issue is an aggregation theorem, not an impartial-observer theorem. Its domain is common lotteries, that is, options over the same set of probability distributions; both individual and social preferences must satisfy expected-utility structure; and only with a Pareto or consistency condition added, in particular that if everyone is indifferent between two lotteries then society is indifferent too, does it follow that social utility is an affine or weighted sum of individual utilities. That does not show that a measuring rod of utility exists in nature. Cardinality comes from the lottery axioms, and gives each person a scale unique only up to a positive affine transformation; interpersonal comparability of units does not arise from this representation theorem by itself, since a change in an individual's scale can be absorbed by the weights. To make an equally weighted aggregation ethically unique still requires a normalization or another bridge. What this series objects to is not this theorem.

Sen cannot be held off, because he does not want that rod. He separates three questions: measurability, interpersonal comparability, and aggregation. Measurability concerns how much information a person's utility representation preserves; interpersonal comparability concerns whether meaningful comparisons can be made between different persons' representations; aggregation concerns combining them into a social evaluation. These three may not be run together, nor may any one of them be inferred automatically from another. And it must be put precisely: aggregation is the operation of forming a social ordering out of individual information, and is not a third dimension of measurement of the same type as the first two. What Sen holds is partial comparability: between full comparability and complete incomparability there is a structured middle ground, from which partial social orderings can be had without pretending that a perfect rod exists.

What this tradition brings to light must therefore be stated afresh. It is not the question of which side one quantity falls on. It is this:

> What is the relation between the various informational bases used across subjects in public evaluation, and first-person standing of end-value.

Under that question lie at least five things, on none of which this series has passed judgement: what welfare, utility, capability, resources, health, opportunity and outcome each have as their subject; which of them can be jointly verified; which can form partial comparisons or public orderings; why they do not thereby flow back into an individual's standing of value; and, when a public evaluation acquires force for action or for institutions, where that force comes from.

The half this series has said must also be said exactly. P6 writes that the standings of two bearers' respective contents admit no common external measure, while the conditions, options and consequences in the world can be checked together and those readings do not flow back into standing. So this series does not forbid every cross-subject comparison; what it forbids is letting the readings of comparison flow back into the standing of value. On the five things above, this series has given neither their several subjects nor a complete account of their relation to standing. The registration is in Section Six.

Four.5 · Recognition theory

The line of divergence changes again, from ranking to the other.

This tradition is not one unified claim, and its layers must be separated first. Hegel's recognition is stated within the developmental structure of consciousness and of right. Honneth divides recognition into three: love, rights, and solidarity, corresponding to self-confidence, self-respect and self-esteem, and holds that the absence of recognition, in the form of disrespect, injures these practical self-relations. Fraser objects to understanding recognition as an intersubjective psychological relation and holds a status model, on which the core of misrecognition is that institutionalized patterns of cultural value prevent someone from participating as a peer, so that it must be treated together with redistribution; and in her framework cultural patterns are not the only obstacle, since a lack of resources likewise impedes participatory parity. The disagreement between Honneth and Fraser can be reconstructed as one about the object, the level of explanation, and the social ontology of recognition. That is an interpretive reconstruction, not a neutral description both sides share: Fraser would stress status order and participatory parity, while Honneth holds that social conflict and institutions can be understood within a more basic grammar of recognition.

This tradition is not the simple contrary of this series. In the Hegel-Honneth line the core of mutual recognition demands precisely that the other be taken as an autonomous and competent judge rather than as an object under assessment. That runs in the same direction as the three limits P6 established, not against them.

The line of divergence lies elsewhere, and it falls on only one of the two lines. The Hegel-Honneth line denies this picture: that a complete and autonomous subject exists independently first and then contingently comes into relation with others. And this series' thought experiment of the solitary subject positively holds that the minimal conditions for end-value to hold require no counterparty; P1 further writes that a solitary subject is self-sufficient with the formal remainder position left empty.

Here the two really do part. Recognition theory holds that the practical self-relation depends on structures of recognition; this series holds that end-value does not depend for its holding on a second subject, and that what recognition achieves is something else. This series does not deny that relations do a great deal to form a person: P1 in establishing cultivation and P3 in establishing colonization both say that what is outside does reach in. What it denies is that the constitutive act by which a content becomes a value content requires a second party. Dependence and constitution are two different things.

Fraser's line is adjacent to the institutional slot registered in Section Six: she ties misrecognition to institutionalized patterns of cultural value, while this series caught sight of institutions in treating common contents and the external measure, and never set them up as a question of its own. That sentence says where the material lies; it does not hand that slot to anyone.

Four.6 · Levinas

This tradition's relation to this series is the least even: half close, half in head-on conflict, and the conflict falls exactly where this series is most exposed.

A limitation first. Taking the irreducibility of the other as one's point of departure points most accurately to Levinas alone. Buber's philosophy of dialogue, Løgstrup's ethical demand, and Derrida's inheritance of and criticism of Levinas all lie near this direction, and their theoretical structures differ widely. Grouping them under one heading is an analytic grouping this paper makes for convenience of exposition; it is not a self-identified school with a unified doctrine. So Levinas alone is the object here.

His most distinctive reversal is this: the ethical relation is not that a knowing subject first identifies another as having some property or value and only then incurs an obligation; rather, in the face-to-face encounter the other presents himself in a way that cannot be wholly absorbed by my concepts, and I find myself already placed under responsibility. The other of Totality and Infinity is an exteriority, and that is not merely being outside me in space: the other cannot be reduced to my representation, my concept, or a content within my totality.

The close half. I cannot claim, by a totalizing cognition, to have wholly grasped who he is. That runs in the same direction as the first of the three limits P6 established: for a particular other's values, third-person evidence cannot close into a reading guaranteed to be exhaustive. Both refuse the same thing, namely settling the other side with a representation of one's own making.

The conflicting half, and it conflicts cleanly. The other's ethical demand does not hold on condition of my avowal; in the structure of meaning, responsibility even precedes my freely accepting it. And what this series establishes is precisely that a content becomes my value content only through my own first-person avowal.

Where the divergence really falls can be put more exactly, and here there are two readings. If that responsibility is read as a kind of value content, the two conflict head-on, and this series' answer is P2's sentence, that avowal is not verification. If that responsibility is read as a normative demand this series does not treat, the two do not conflict but speak past each other, since this series draws on no normative force at the object level. This paper sets out both readings and does not adjudicate which is right. That is the price of keeping to the discipline of placing rather than adjudicating; it is not evasion.

There is one further point available as a structural analogy. He does not stop at the incomparability of each single other. Once a third party appears, that is, another other and all the rest, one must ask how infinite responsibility to the person before me is to coexist with the demands of the others. Justice, comparison, deliberation and the necessity of institutions arise from there; this is the argument in Otherwise than Being, where the third party is the birthplace of the limitation of infinite one-directional responsibility and of the question of justice, and comparison and judgement unfold in the passages that follow. So it is inaccurate to say that he always refuses all comparison; accurately, comparison is derivative, forced upon the ethical relation by a plurality of others.

And P6's scope turns at the same place: its second section speaks of one particular other, its third of two bearers, its fourth of the whole, and the third is precisely where comparison first becomes a problem. The shapes may be set side by side as an analogy: when the move is made from two to many, comparison becomes a question that must be answered. But this is an analogy and not an isomorphism: his third party is a question of justice forced out of infinite responsibility from within the ethical relation, while P6's widening scope is this series' own order of argument; their objects differ, and so do their answers. He moves towards justice and institutions, this series towards conditions in the world that can be checked together while their readings do not flow back into standing.

Four.7 · The ethics of care, and adaptive preferences

This tradition strikes at the slot registered in Section Six about subjects-to-be, and it brings the hardest question of the eight.

What the ethics of care commonly opposes is not all principles, but the picture on which the mature moral agent is presupposed as independent of others, with dependence an occasional exception. It begins from dependence, care, vulnerability and concrete relations. Tronto divides care into four phases, caring about, taking care of, care-giving and care-receiving, with four corresponding moral qualities, attentiveness, responsibility, competence and responsiveness. The two sets are not one set, and this paper keeps them apart as in the text. Kittay pushes ineliminable dependence and the labour of care to the centre of theories of justice. Relational autonomy, as a development, does not deny self-determination; it holds that the capacity for self-determination itself depends on social relations and institutional conditions.

The first divergence need not be made into an opposition. This series takes bearers as its unit and this tradition starts from relation and dependence; but this series does not deny that relations do a great deal to form a person. What it denies is only that the constitutive act by which a content becomes a value content requires a second party. Dependence and constitution are two different things.

The hard one is the second, and this series cannot answer it in full. By way of adaptive preferences the objection can be put sharply: a person raised under long-term disadvantage may sincerely report herself satisfied, and that satisfaction may be exactly what the disadvantage has shaped. If a content takes standing under her name merely by being avowed in the first person, can a third party say that the avowal itself is defective?

An earlier draft answered this as a choice between two: either it really is her own, or it is colonization, and a third person cannot tell which. That dichotomy left out too much. She may sincerely avow on the strength of seriously mistaken facts; she may sincerely avow among extremely restricted options; the avowal may be genuine while living by it damages her health and her opportunities; preferences may be formed under oppression without her thereby losing agency; and one content may at once contain a genuine avowal, external shaping, and conditions from which there is no exit. The literature on adaptive preferences is itself divided on these, and there is no acknowledged simple pair of boxes.

What this series can give here is one ledger, not the whole set.

Ledger	What it asks	What this series has
Standing of value	whether this content has taken her first-person avowal	constitution is completed by the bearer in the first person; a third person may form fallible, revisable judgements, and may not stand in for her avowal, nor raise his judgement into a final certification
Welfare and interest	what living by it does to her health, capabilities, opportunities and condition	not established by this series
Power and rights	whether there was coercion, blocked paths, manipulation or absence of exit in the process of formation	partly established; the Power Theory and the Rights Theory hold the rest

The three ledgers may hold at once, and they do not settle one another automatically. That a content has taken standing on the first ledger earns it no marks on the second; that a blocked path is found on the third does not automatically cancel its standing on the first, but says that this place must be audited separately by P3's criteria.

So this series' answer to this tradition, stated exactly, is this sentence: it can add to the discussion of adaptive preferences one ledger of constitution and channels. A third person may form fallible and revisable judgements about positions and channels; he may not stand in for the person's avowal, nor raise his judgement into a final certification. And this series cannot use that ledger to replace the evaluation of welfare, the scrutiny of rights, and the judgement of fact. Those three it has not established, and this paper does not establish them on its behalf.

One point of wording is worth noting. That the ethics of care has no single measure does not mean that anything goes; what it asks for is context-sensitive practical judgement. And the term for judgement appears nowhere in the seven papers and is nowhere set up as an independent explanatory device, its full theoretical work having been handed by P5 to the Theory of Action. So this series does not contend with it here; the two are simply not speaking on the same ground.

Four.8 · The Mohists: standards, and the benefit of all under heaven

The last of the eight strikes at the second of the three senses distinguished in Section Two: whether there is an evaluative position that does not change according to who is acting.

The Mohists set up standards and models, Fa Yi, and hold that one who governs cannot be without standards, and that standards are not to be taken merely from one's parents, one's teacher or one's ruler, but are best taken from Heaven. Here fa means norm, standard, model; it is not legislation in the modern sense. Impartial care and mutual benefit, jian xiang ai, jiao xiang li, are set up as a public standard, appealing to Heaven's impartiality: what Heaven desires is to be done, what Heaven does not desire is to be stopped. This is the tradition among all those engaged here that comes closest to a public standard of adjudication rising above each private standpoint.

Its own context must be preserved and not modernized. The Heaven of the Mohists is not a neutral observer's position; it has desires and aversions, and rewards the good and punishes the violent. Standards are taken from Heaven because Heaven's conduct is broad and impartial. To write this as a personal God in the modern sense is contested, and this paper does not take that step. It should also be recorded that the Mohists are not simple divine-command theorists: they also explain by independent reasons, of rightness, benefit and order, why Heaven's will is worth taking as a standard. This resembles in form the impartial evaluative standpoint consequentialism requires, while its ground differs: consequentialist impartiality is a methodological requirement, Mohist impartiality is a cosmological ground. To write the Mohists as Chinese utilitarianism is inaccurate, and would cost this interface its value.

The line of divergence here has been corrected twice and needs correcting once more, and this time backwards. An earlier draft cited Fei Yue and Jie Zang as showing that the Mohists, on the ground of the benefit of all under heaven, judged what some people cherish to be improper, and had thereby reached into the side of the holding of value; the revision narrowed this to a question awaiting proof. Having gone to the text, something stronger must be said: the text itself never took that step.

Fei Yue, Against the Use of Music, expressly grants that music is pleasing to the ear, that ornament is beautiful to the eye, and that fine food is sweet to the mouth, and only then, because rulers cast instruments, levy taxes and take up labour, judges that it does not accord above with the affairs of the sage kings and does not accord below with the benefit of the myriad people. Jie Zang, Moderation in Funerals, likewise tests lavish burial and prolonged mourning, as an institutionalized practice, against wealth, population and order.

So what the Mohists judge is the making of music and the practice of lavish burial as public practices, as affairs of government and as allocations of resources; it is not a demonstration that what is cherished is unworthy of being cherished. Fei Yue even contains a sentence that all but rules out this misreading in advance.

The objects of the two therefore do not overlap. Mohist standards judge conduct and government; this series judges how a content becomes the value content of some bearer. And one question may still be asked:

> Can a public standard restrict actions organized out of a person's values without thereby acquiring the standing to rewrite that person's standing of value?

That question falls precisely on the two bars on back-flow set up in Section One. If a public standard works only on the side of action, it does not conflict with the second bar; if it goes further and lets a judgement of conduct flow back into a judgement of that person's standing of value, then it conflicts with this series. On the Mohist text, the two chapters examined above fall on the former side: what they judge is practice and government, and there is no back-flow.

The divergence therefore lies neither in whether there is a public standard nor in whether the Mohists crossed an axis. It lies elsewhere: this series denies that there exists an external coordinate system able to pass judgement on a bearer's own standing of value, and Mohist standards never claimed that standing. The two do not conflict here, and that is exactly where the value of this interface lies: a tradition that holds a public standard without reaching into the holding of value shows that a public standard and this series' leading proposition can stand together.

One neighbouring point in passing. The contrast between the Confucian line and this series falls elsewhere, not on the question of a standpoint. Shu, what you do not want done to yourself, do not do to others, takes one's own side as its starting point and extends to the other. This series' step does not run that way: that the other is an end is in this series a premise, not a conclusion drawn from my being an end; only after the two standings of being an end have each been established does the source not being placed outside give an isomorphic structure on both sides. So both enter the relation to the other from first-person language, while the structures of inference are not the same. Shu has no single fixed English rendering, and this paper does not compress it into reciprocity in the modern sense of exchange. This paper therefore treats it as a contrast only, not as an ally; the placement is in the next section.

---

Section Five · Where the Other Seven Stand

These seven do not conflict directly with this series, or the point at which they would has already been covered by one of the eight. Only their place in the coordinates is given here.

The three dividing lines in metaethics. Cognitivism and non-cognitivism divide over whether normative judgements have propositional content capable of truth and falsity. Realism and error theory divide over whether such discourse systematically fails; Mackie holds that ordinary moral judgements presuppose objective values that do not exist in the world, and are therefore, on their face-value semantics, systematically false. Positions that make truth-grounding independent of attitudes and positions that make it mind-dependent divide over that. Section Two has said that these three lines do not by themselves settle who bears value. And by the three registers of Section Two, this series has taken no position on truth-grounding, and therefore has no place to mark on any of these three lines. That is not an omission; it is a place left unclosed.

Constructivism. This is the one most worth stating precisely, because it lies nearest to this series. Korsgaard's procedural realism says that normative questions have correct answers because there is a correct procedure of practical reason, not because the procedure discovers an independently existing normative domain. Street ties the substantive content of reasons to an agent's contingent evaluative starting points; her 2006 paper is a challenge to realism, and her 2008 paper the positive statement of constructivism about reasons.

Near, and not to be filed under it, and this distinction must be written exactly. What Street calls an evaluative starting point is an overall standpoint structured by reflective consistency, not one explicit act of first-person avowal; what this series establishes is a constitutive event, namely that a content which has already taken a candidate position within some domain of sources acquires, within the workings of the bearer's own law, a standing that can be drawn on, restated and revised. The two may be near in rejecting that value is wholly independent of evaluative activity, while this series has at present given no complete theory of reasons, no theory of truth, and no procedure of reflective consistency, and therefore cannot be formally filed under Humean constructivism. That sentence and register two of Section Two mark the same unclosed place. It should also be noted that Korsgaard's own term realism is very broad there, its contrary lying near a nihilism about normativity, and it should not be made to carry too much later taxonomy.

Agent-centred restrictions and the prerogative on the deontological side. They form a neighbouring opposition with this series while their grounds and subjects differ, and Section Four has engaged them. Only Korsgaard's warning is added here: agent-relative and deontological are not logical synonyms, and that formal distinction cannot be equated directly with the division between consequentialism and deontology.

Value pluralism and incommensurability. Three words must be kept apart: incommensurability is the lack of a common unit of measure; incomparability is that no comparative relation holds between two items; the former does not entail the latter. Chang sets up a fourth relation, parity, on which two items under some covering consideration are neither better nor worse nor equal, and yet remain within one evaluative space.

Where this series stands in this slot can be said only this far, and an earlier draft said more than it should. This series holds definitely that standings of value admit no common external measure, and rejects the predicates better, worse and equal. Whether it further holds that they are incomparable, or allows some non-metric relation of the sort parity names, is not judged by the existing object-level texts. The earlier draft argued from the fact that conditions and consequences in the world can be checked together that this series does not fall into the incomparability slot; that step does not go through: that objects on the side of the world can enter joint evaluation says nothing about whether some positive comparative relation holds between two standings of value. The registration is in Section Six.

Preference, utility and welfare in economics. The three equations must be removed: preference is a comparative relation among options, utility is in the first place a representation of preference, welfare is how a person's life goes for that person. Price was routed out by the prequel when it cleared the space of reference. And preference, utility and welfare still have to be placed separately: their relation to standing of end-value cannot be settled on the strength of price having been routed out. The place where contact is really made is Sen, engaged in the previous section.

Harsanyi's representation theorem. Under expected-utility axioms and certain consistency conditions, social evaluation has a representation as a weighted sum of individual utilities. This does not show that a natural measuring rod of utility exists; it shows only that under explicit normative and representational conditions aggregation can have a strict structure. And what it does not settle must be said: it does not by itself answer which welfare information can be compared interpersonally, where the weights come from, or what the normative ground is. What this series objects to is not this theorem, and the two do not stand in opposition.

The Confucian line. Shu, what you do not want done to yourself, do not do to others, takes one's own side as its starting point and extends to the other. This series does not infer from my being an end that the other is an end: that the other is an end is a separately established premise here, and only after the two have each been established does the source not being placed outside give an isomorphic structure on both sides. Both enter the relation to the other from first-person language, and the structures of inference are not the same. This paper therefore treats it as a contrast only, not as an ally. It should also be said that one should not pretend there exists a unified East Asian theory of value: the Confucian and Mohist lines appear separately in this paper precisely because they stand far apart on the question of a standpoint.

---

Section Six · The General Ledger of Boundaries and Destinations

This section registers the places where this machine did not go on.

Identities must be sorted before registering, because reading places of different provenance as one kind makes a registration look complete while it is not. Every construct carries a remainder, as the methodology established; and what has not been written on is not of one kind only. This section therefore registers in six slots.

Slot	What it is	What is registered here	In sections
One	a remainder left by this construct	whether another bearer can come to hold on other conditions	one
Two	objects adjacent within the jurisdiction, never set up as its own questions	institutions; the cross-subject informational basis of public evaluation	two, three
Three	a limit already established	a third person has no external threshold at a subject-to-be	four
Four	a research candidate proposed by this paper	the developmental history of a subject-to-be's first-person operation	five
Five	handovers whose address is already fixed	Theory of Action, Theory of Meaning, Power Theory, Rights Theory	six
Six	an object-level open question that came to light on looking back	what relation, if any, remains between standings of value	seven

The second slot occupies two subsections, so slot numbers and subsection numbers run one apart from the third slot on. Each subsection below carries its slot number.

One · Slot one · A remainder left by this construct: whether another bearer can come to hold

Those people by the river cleared the silt together, stopped a sand-dredging operation, and put in money to repair a stretch of embankment. What P6 judges is that those undertakings do not of themselves yield the conclusion that their sources have merged; the ground is that each person's source is anchored in himself, and that the directions of several sources crossing does not amount to their becoming one.

It did not judge one further thing, and it expressly declined to. P6's words are: as to whether another bearer can come to hold on other conditions, this section does not ask; and even if one did, that would not mean that the sources of the original people had been swallowed by it.

This place is an express reservation, and it is also a remainder. The first of those follows from P6's declining to judge. The second needs a further proof, and that proof cannot be derived directly from any substantive proposition at the object level; it must be given by the input form of the machine as a whole together with P6's express reservation.

Every source-position, constitutive and operational predicate in this series takes an already established bearer as its subject. P1, in establishing the domain of bearers, gives who can bear end-value, namely complete subjects and subjects-to-be, and not how a new bearer comes about; it says that a bearer's qualification as a bearer of end-value and the non-externality of his source of value are both present as of now, and it says that giving up a door is a matter of operation and not a loss of the qualification to bear. For the type transition from several subjects or a relational structure to a new bearer, there is no corresponding predicate of occurrence in the seven papers.

And this is not merely a ready criterion left unwritten by accident; what it exposes is that the current machine's input types lack one piece of transitional machinery. The reason is this. If one insisted on using the existing source-position and constitutive predicates to explain that transition, one would first have to take the bearer still awaiting establishment as the subject of those predicates, and the argument would be circular. It is this failure of absorption, produced by the input types themselves, that is the reason for registering it as a remainder.

Two things do not follow from this, and must be written down nearby. First, it does not say that a third person cannot describe the conditions under which a new bearer forms: describing conditions of formation and standing in for constitution are two different things, and the second bar on back-flow in Section One bars the latter, not the former. Second, it does not prejudge whether a relation can be a cause or a condition in some particular coming-to-hold; P6 expressly declines to judge that, and this paper does not judge it on its behalf.

The limits of the reservation must also be stated exactly, or the registration will itself overstep. What P6 laid down is that what happens in a relation is not of itself enough to yield the conclusion that sources have merged, nor enough to show that a new bearer has come to hold; it did not judge whether a common relation can be a cause or a condition in the forming of some bearer, and it wrote that it does not judge it. So all that can be registered here is one sentence: if another bearer comes to hold, its structure as a bearer and how its source position was acquired must be given separately, while source positions are given with the bearer's structure, as P1 established. Whether a common relation counts in that coming-to-hold, the Value Theory does not judge.

Two · Slot two, first item · An adjacent object: institutions

P6 writes that what is done together, the rules laid down together, and what is kept together are not in themselves at fault for being common, and gives the conditions under which a common content becomes an external measure. That judges whether it has overstepped; it does not say what it is. And from beginning to end this series never set up what institutions are as a question of its own; even raising it was incidental.

Being caught sight of in passing and being expressly not judged are two different things. In the former the series never asked at all, so there is nothing reserved; it is therefore not a remainder but an object adjacent within the jurisdiction. If anyone takes it up, what is taken up had better not be the question of how institutions as such are possible, which is too wide, but the stretch this series caught sight of: under what conditions common rules, common decisions and common arrangements acquire institutional structure.

Section Four, in engaging recognition theory, noted that Fraser ties misrecognition to institutionalized patterns of cultural value. That line is the nearest material to hand for this slot. That sentence says where the material lies; it does not hand this slot to anyone.

Three · Slot two, second item · An adjacent object: the cross-subject informational basis of public evaluation

Section Four, in engaging Sen, caught sight of another place, and it is far wider than an earlier draft made it.

The half this series has said is: standings of value admit no common external measure, while the conditions, options and consequences in the world can be checked together, and those readings do not flow back into standing. So this series does not forbid every cross-subject comparison; what it forbids is letting the readings of comparison flow back into the standing of value.

And this whole stretch beneath it the series has never entered: what welfare, utility, capability, resources, health, opportunity and outcome each have as their subject; which of them can be jointly verified; which can form partial comparisons or public orderings; why they do not thereby flow back into an individual's standing of value; and, when a public evaluation acquires force for action or for institutions, where that force comes from.

This is likewise not a remainder but an object adjacent within the jurisdiction: the series caught sight of it in treating cross-subject comparison, and never set it up as a question of its own.

This section does not write its destination. By the definition set out in Section Seven, writing down a destination is a forward assignment and therefore a debt; and this paper, which lays down that rule, opens not one debt of its own.

Four · Slot three · A limit already established: a third person has no external threshold at a subject-to-be

An infant who cannot yet say what he wants: his source of value is not mine to give, and P6 judges that firmly. P6 also writes that how far particular value contents have come about under his name is not judged by whether he can speak, since speech was never a necessary signature of avowal.

That sentence is often read as this series owing a measure here. It is not. In the same passage P6 gives a threshold: he cannot yet be the first-person avower of that content, because that row is not yet actually running within him. And that threshold gives the domain of definition of that one content; it is not a general threshold for every particular value content of a subject-to-be. What would need a measure is a different question: how far along that row's running has come.

And on that question, what this series judges is not that it has not given one but that it cannot, and it has judged this already. P2, in establishing the axis of occurrence, writes that avowal may happen in an instant and may also take shape by degrees and settle only after years, and that length of time does not judge this. P4, in establishing external underdetermination, writes that there is no set of external signatures both necessary and sufficient and transferable to another person. Taken together, a third person has no external threshold by which to determine finally how far a value content has come about at a subject-to-be.

This is a negative conclusion already established, not a hole waiting to be filled. It stops the theory from going further in that direction, and that is a result of the theory, not a gap in it. This section therefore lists it in the third slot and not among remainders.

This slot and the first slot are not the same question; the two ask at different layers. The first asks whether a new bearer comes to hold, which is a transition of bearer type and belongs to ontology and occurrence; this one asks, of a subject-to-be already within the domain of bearers, when some running of his reaches some state, which belongs to epistemic authority. The first currently lacks a predicate of occurrence; this one has already secured the conclusion that there is no external final threshold. All that can be said of the two is one sentence: both expose the limit that external reading cannot stand in for first-person structure. That is where it stops, and no further merging is permitted.

This series comes to a stop here, not because its criteria have only one form. It has criteria of other forms, for instance P5's revocability, of which P5 writes that the subject is the action. The reason it stops is that this series has already cut two things apart, and P4 makes that cut most plainly: facts within a relational structure can be told apart, and telling those facts apart does not turn into a determination of constitution. A third person may obtain fallible and revisable readings of conditions of formation, of relational positions, and of the channels of expression and revision, and those facts can enter joint verification; whether an avowal has already occurred as a first-person constitution cannot be closed out of those external facts.

Five · Slot four · A research candidate proposed by this paper: the developmental history of a subject-to-be's first-person operation

Cast into view alongside that established limit is a candidate this paper proposes: whether the way a subject-to-be's first-person operation gradually comes to be actually running admits of a separate description.

It is a suggestion this paper makes; it is neither a boundary the series left nor a remainder. If anyone writes it, one thing must be laid down in front: such a description may not turn round into an external procedure for certifying that at this point a value content has come to hold. Cross that line and it falls back on the side this series has already judged.

The three ledgers used in Section Four, in engaging the ethics of care, are the practical use of that line: this series explains which relational positions and channels can enter audit, and it also explains that a third person's judgement may not stand in for a first-person constitution, and may not be raised into a final certification.

Six · Slot five · Handovers whose address is already fixed

The first four slots are places where the series did not go on. Four further places are not that: they were handed over, and who receives them is written in the published texts.

The whole branch of action was handed to the Theory of Action. P5 writes in closing: how an ordering of action forms, how it is carried out, how it is looked back on afterwards, and how judgement, habit, hesitation and capacity enter into it, are handed over entire to the Theory of Action; this paper does not use them and passes no judgement on them. Judgement is among them. The term appears nowhere in the seven papers and is nowhere set up as an independent explanatory device.

How a content is thereafter carried on by one and the same life goes to the Theory of Meaning. P5 and P6 both make this handover. P6's words in closing are: how the content that the other can have values of his own, and that this is not mine to give, is thereafter carried on by one and the same life, goes to the Theory of Meaning. P6 states that handover expressly in its closing and does not further explain its relation to the old prospect address.

Direction-value goes to the Power Theory. That branch was routed out when the prequel cleared the space of reference, and P1 writes again in establishing sources that direction-value goes to the Power Theory, whose severing requires two subjects. The sole object of the Value Theory is end-value.

How the boundary is put in question when two subjects meet goes to the Rights Theory. P5 writes in establishing adjudication: he is looking at what this move will touch in whom, at where that boundary is being touched, and how that boundary is put in question, and by which mode a dispute is handled, goes to the Rights Theory. The boundary of encounter and the two modes of handling have already been established by the Rights Theory in its paper on the thickness and boundary of basic rights.

Being assigned is not being completed. This paper registers addresses only and does not perform the receiving series' acceptance. These pieces of work no longer belong to the writing tasks of the Value Theory's object level, while the published accounts of the corresponding forward assignments still have to be settled item by item. The current state of the four is as follows. This section registers only as far as this column, and completes no acceptance here.

Destination	What was handed over	What can be registered now
Theory of Action	formation, execution and retrospection of an ordering of action; judgement, habit, hesitation, capacity	address fixed; whether discharged is to be verified separately
Theory of Meaning	how a content is thereafter carried on by one and the same life	address fixed; whether discharged is to be verified separately
Power Theory	direction-value	address fixed; whether the existing texts discharge it is to be verified separately
Rights Theory	the boundary of encounter and the two modes of handling	address fixed; the text has pointed to the corresponding substance, and it may still be verified separately

Seven · Slot six · An object-level open question that came to light on looking back: what relation remains between standings of value

Standings of value admit no common external measure, and the predicates better, worse and equal have been excluded by P1, P5 and P6. But whether it is further held that no comparative relation of any kind holds between two standings of value, or whether some non-metric relation is allowed, the existing object-level texts have not judged.

Section Five has explained why the fact that conditions and consequences in the world can be checked together cannot answer this question on the series' behalf: objects on the side of the world can enter joint evaluation, and that says nothing about whether some positive relation holds between standings of value.

This paper registers this as unjudged and does not obtain an answer on its behalf.

---

Section Seven · A Forward Assignment Is a Debt

Why this paper only registers and does not promise: the reason is in this series' own record, and that record concerns this paper.

What layer this section establishes must be said first. It is not a norm at the object level of value theory; it does not say what a person ought to do. It is a discipline of publication: how a pointer written into a published text is thereafter to be settled. The rule that no normative force is drawn upon at the object level remains in force in this section.

One · What counts as a forward assignment

Without a definition, every sentence pointing to the future becomes a debt, and the slots of Section Six would then conflict with themselves.

A forward assignment is this: a published text, in a definite mood, hands a specific and unfinished piece of work to an already named subsequent paper, series or theoretical position, and settles the scope of the present paper by that handover.

Objects adjacent within the jurisdiction, limits already established, research candidates, and registrations of what is unjudged do not thereby become forward assignments. Of the six slots registered in Section Six, only the four handovers in the fifth slot are forward assignments. Of the remaining five slots, not one is.

Two · Three routes of settlement

A forward assignment is an account that must be settled in public. There are three routes.

The first is discharge: that paper is really written and does that work. The second is public reassignment: a later text states that the work has changed destination, and why. The third is public withdrawal: it states that the question has been dissolved by later work, or was wrongly posed to begin with, and gives reasons.

Settling in parts, discharging some and reassigning or withdrawing the rest, is a combination of the three and does not establish a fourth kind.

Whether an account has been settled is judged not by what kind of file it is written in, but by whether it enters the regular chain of publication, and whether a reader can find it in the formal text or in a public correction. A public version note counts; an internal working record does not: a reader cannot find it, and that account is still outstanding on the page.

Three · This paper is itself one such account

The prequel wrote: how value goes to action is left to the prospect paper. P1 wrote the same thing in wider words: this series runs to the prospect paper, and the Theory of Action is another branch; the prospect paper is to run from the constraint of ends to the appearing of action, and it lays the ground for a future Theory of Action. P2 wrote: how a value content already avowed goes to action is the entrance to the prospect paper. P3 wrote: how a value content already taken up goes to action is the entrance to the prospect paper. Four papers point at one piece of work, and the recipient is written as this paper.

P4 wrote the sentence at the same position, and the recipient had changed. P4's words are: how a value content already taken up goes to action is the entrance to the Theory of Action. It differs from P3's sentence in three characters at the end. And the words prospect paper never appear again anywhere in P4, nor is there a sentence explaining how this one stands to the sentence in the earlier papers.

What appears there is a forking of addresses, not a reassignment already carried out. By the rule laid down above, a public reassignment must state that the work has changed destination and why; P4 did not state it, it simply wrote on at the new address. P5 then extended the branch of action in full, from judgement alone to judgement, habit, hesitation and capacity entire, and handed the carrying-on of a life to the Theory of Meaning; P6 states the handover to the Theory of Meaning expressly in its closing and does not further explain its relation to the old prospect address. The new architecture took shape from there, while the old address and the new stood side by side, unsettled.

This paper is where that account is settled, and the route of settlement is the second. The work has been reassigned: one stretch went to the Theory of Action from P4 on and was extended in full by P5, and one stretch went to the Theory of Meaning by P5 and P6. P4 first shows the new address, P5 and P6 make the new structure of distribution clear, and this Paper 0 is where the reassignment is publicly completed.

Settling is not inheriting that name. A forward assignment points at a piece of work, not at a permanent bibliographic type for some future product. That work has been wholly reassigned along the two routes, the forming and carrying out of action to the Theory of Action, the carrying-on by one and the same life to the Theory of Meaning; and this paper takes on a further piece of meta-level work that the old forward assignment never specified: to stand outside the seven papers and describe this machine, to interface with the existing traditions, and to register boundaries and destinations. The prospect paper is therefore this paper's historical alias, not its place in the series.

Four · And so this paper opens not one new debt

The second item of the second slot in Section Six expressly does not write a destination, and that is to keep to this rule. If an object never set up as a question were handed off to some series in passing, this paper would have opened its own first account on the very page where it lays down the rule.

For a paper that lays down a discipline of accounting, what matters most is not how well it states the discipline, but whether it keeps to it itself.

So this paper makes no promise. The registrations in the first, second, third, fourth and sixth slots are not promises this paper makes about future writing: if they are written later, the later text need not turn back and cite this paper as its ground; if they are never written, this series is not thereby in debt.

The fifth slot does not fall under the sentence above. Those four are forward assignments already written into the published texts; their addresses are fixed, and their state of completion must still be settled separately by discharge, public reassignment or public withdrawal.

---

Section Eight · Closing

The object level of the Value Theory ends here.

The prequel cleared the space of reference and said what value is not. The six papers that followed set up a machine for reading value, and it is not one principle unfolded six times: it keeps three ledgers, of constitution, of operation, and of action and relation, each with its own subjects. Among the most important reasons the six cannot be compressed into one are the two lines that run between the ledgers: readings do not of themselves flow back into a source or a standing, and external conditions do not of themselves stand in for a first-person constitution.

Placed within the map of existing value theory, this machine parts company with eight traditions each along its own line, and is not speaking on the same ground as seven others. The innermost line of parting is this: whether the standing of a particular end-value content under some bearer's name is constituted by his first-person avowal, or is to be explained by a fittingness, a relation of reasons, or an objective rank order independent of actual avowal. That line governs the standing of particular contents only and does not extend to every value fact.

The interface with Kant meets on one negative mark: admitting of no replacement by an equivalent. The meeting is real, and sharing one mark is not conceptual identity; it is a negative mark, not a full constitutive account of dignity.

And the places where this machine stopped cannot all be called remainders. This paper registers in six slots: one remainder left by this construct, two objects adjacent within the jurisdiction and never set up as its own questions, one limit already established, one candidate this paper proposes, four handovers whose address is fixed, and one object-level question that came to light only on looking back, namely what relation remains between standings of value.

Closing a machine is not calling everything beside it an opening, nor saying one more thing on its behalf that it never said. The whole work of this paper is to say what kind of machine it is, to say where it parts company with the machines of others, and then to register what it did not say, each under the identity that is its own.

---

Acknowledgements

This is Paper 0 of the Value Theory branch of the SAE (Self-as-an-End) philosophical system.

Authorship and responsibility. This paper is signed by Han Qin alone; all argument, all decisions between alternatives and the final text were adjudicated by the author, who bears sole responsibility for it.

The facts of drafting and review. Drafting assistance and cross-review were carried out with the help of large language models. Drafting and the internal precision audit were undertaken by Claude (Anthropic), ontological-limit and contradiction detection by Gemini (Google DeepMind), boundary testing by Grok (xAI), and structural gatekeeping with pre-publication signature by ChatGPT (OpenAI). The external literature materials for Sections Four and Five were compiled by ChatGPT's deep research. Thanks are due to these systems and to the research and engineering teams behind them. Their comments were adjudicated by the author item by item, and both what was adopted and what was rejected are on record.

The degree of verification of external citations. This paper's external citations were verified item by item at four grades: confirmed against the original text; original text in hand with the precise location unconfirmed; confirmed on reliable secondary sources; and unconfirmed. The final net figures for each grade and the state of each item are given in a separate graded verification table for external claims. This paper does not claim that all external citations have been confirmed against original texts, nor that it has covered every external load-bearing point in the text.

The corrections made to this paper fall into three classes, recorded here separately and not merged into one.

First, five substantive rewrites occasioned by structural review: the writing that made the four nodes logically equivalent and then compressed them into a machine that reads positions only was withdrawn, and three ledgers with two bars on back-flow were set up in its place; the writing that merged dignity, the standing of an end, particular contents and colonization was withdrawn, the four objects were separated, and the structure of colonization was corrected against P3, namely that positions are permanent while an external warrant counterfeits a constitution already completed; the scope of the discovered-or-constituted fault line was pulled back to the standing of particular contents; the writing that inferred mind-dependence directly from dependence on the bearer, and denied a mind-independent ground of truth, was withdrawn, and three separate registers, constitutive dependence, truth-grounding and epistemic authority, were set up, with this series recorded as unclosed on the second; and the inference that what is a person's own value is thereby good for him was withdrawn, since it switched axes between standing of value and welfare, and Sen, adaptive preferences and Section Six were rewritten accordingly.

Second, two substantive corrections of external material: Scheler's basic value modalities are four and not five, and the useful is a derivative value founded on the agreeable, while an earlier draft listed it as a coordinate level; and what the Mohists judge in Fei Yue and Jie Zang is the making of music and lavish burial as public practices and affairs of government, whereas Fei Yue expressly grants first that music is pleasing to the ear, that ornament is beautiful to the eye and that fine food is sweet to the mouth, so that the earlier drafts' inference from public criticism to the impropriety of what is cherished crossed a step the text never took.

Third, one correction arising from a return to the upstream texts of this system: the treatment of colonization was corrected twice, the first draft saying too much and the revision over-correcting, with this version taking the slot between the two, namely that the function of the constitutive position is occupied without the position being taken away. This one was not obtained from external verification but from re-reading P3's own text.

There were also three narrowings of external formulations: the lineage of the fitting-attitude family is not written as a single line of succession; Harsanyi's theorem has its domain and central assumptions written in; and Sen's three-way distinction is given as three that may not be run together rather than as three that are independent.

Page references for external literature are handled by each item's grade of confirmation: those confirmed against the original text carry page numbers; those with the original text in hand but the precise location unconfirmed give only the confirmed chapter or textual range, and do not counterfeit a precise page; those confirmed on reliable secondary sources give the source without page numbers; those unconfirmed give no location, and the corresponding places in the text are narrowed. An earlier draft deleted page references in bulk, which was an over-reaction: it removed the reliable along with the unreliable, and concealed the difference between them.

Every sentence this paper quotes from the upstream texts of this system was compared item by item against its published wording; and this paper's statements about external writers were separately verified item by item at grades. Both comparisons are kept in separate tables, as instruments of drafting and review, and are not published with this paper. The choices made in placing the traditions, and the way the lines of divergence are drawn, are not within those tables.

On the limits of this paper's verification, one thing is written here and not left for some other document to declare on its behalf: this paper's external placements rest on graded verification, of which only the grade confirmed against the original text can be warranted to agree with the source; this paper does not claim to have covered every external load-bearing point, nor that the lines of divergence have been reviewed externally. Chinese classical texts are cited from current editions; no particular critical edition was adopted as a base text.

---

References

Within this system

The following are archived on Zenodo, with Chinese and English editions under the same record. The DOIs cited are concept DOIs, which always point to the most recent version; version numbers and access dates are therefore not given. All works are single-authored (Han Qin). Drafting and review assistance are described in the acknowledgements.

Qin, H. (2026). Systems, Emergence and the Human Condition: A Normative Social Theory Centred on the Subject as an End in Itself (SAE Foundational Paper 1) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.18528812

Qin, H. (2026). The Three-Layer Two-Dimensional Unified Structure: The Complete Framework of Self-as-an-End (SAE Foundational Paper 3) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.18727326

Qin, H. (2026). General Methodology: The Chisel-Construct Cycle (SAE Methodology) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.18842449

Qin, H. (2026). The Momentum of the Power Machine (SAE Power Theory Paper 0) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.20707047

Qin, H. (2026). On the Origin of Power (SAE Power Theory Paper 1) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.20370224

Qin, H. (2026). On the Thickness and Boundary of Basic Rights (SAE Rights Theory Paper 2) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.20769253

Qin, H. (2026). On How Rights Presuppose No Duty (SAE Rights Theory Paper 0) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.21314907

Qin, H. (2026). What Value Is Not (SAE Value Theory, Prequel) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.21362575

Qin, H. (2026). On the Source of Value (SAE Value Theory Paper 1) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.21405207

Qin, H. (2026). On the Occurrence and Recognition of Value (SAE Value Theory Paper 2) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.21437038

Qin, H. (2026). On the Colonization of Value (SAE Value Theory Paper 3) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.21611154

Qin, H. (2026). On the Dynamics of Value (SAE Value Theory Paper 4) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22050204

Qin, H. (2026). On the Conflict of Value (SAE Value Theory Paper 5) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22122819

Qin, H. (2026). On Value and the Other (SAE Value Theory Paper 6) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22191071

External literature

Only works actually cited in the text are listed. Page references are given according to each item's grade of confirmation: those confirmed against the original text carry page numbers; those with the original text in hand but the precise location unconfirmed give only a chapter or textual range; those confirmed on reliable secondary sources give the source without page numbers; those unconfirmed give no location. The grade of each item is recorded in the separate graded verification table.

Bojanowski, J. (2015). Kant on Human Dignity: A Response to Oliver Sensen. Kant-Studien, 106(1), 78–87.

Brentano, F. (1889). Vom Ursprung sittlicher Erkenntnis. English: The Origin of Our Knowledge of Right and Wrong.

Bruckner, D. (2009). In Defense of Adaptive Preferences. Philosophical Studies, 142.

Buber, M. (1923). Ich und Du. English: I and Thou.

Chang, R. (2002). The Possibility of Parity. Ethics, 112.

Colburn, B. (2011). Autonomy and Adaptive Preferences. Utilitas, 23.

Dean, R. (2006). The Value of Humanity in Kant's Moral Theory. Oxford University Press.

Derrida, J. (1964). Violence et métaphysique. In L'écriture et la différence (1967).

Elster, J. (1983). Sour Grapes: Studies in the Subversion of Rationality. Cambridge University Press.

Ewing, A. C. (1947). The Definition of Good. Macmillan.

Fraser, N. (2000). Rethinking Recognition. New Left Review, 3.

Fraser, N., & Honneth, A. (2003). Redistribution or Recognition? A Political-Philosophical Exchange. Verso.

Gibbard, A. (1990). Wise Choices, Apt Feelings. Harvard University Press.

Harsanyi, J. C. (1955). Cardinal Welfare, Individualistic Ethics, and Interpersonal Comparisons of Utility. Journal of Political Economy, 63.

Hegel, G. W. F. (1807). Phänomenologie des Geistes. English: Phenomenology of Spirit.

Hieronymi, P. (2005). The Wrong Kind of Reason. The Journal of Philosophy, 102.

Honneth, A. (1992). Kampf um Anerkennung. English: The Struggle for Recognition (1995). Polity Press. Cited here: ch. 5, "Patterns of Intersubjective Recognition: Love, Rights, and Solidarity."

Kant, I. (1785). Grundlegung zur Metaphysik der Sitten. Akademie-Ausgabe, Bd. 4. Cited here: the second section, on ends in themselves, dignity and price.

Khader, S. J. (2011). Adaptive Preferences and Women's Empowerment. Oxford University Press.

Kittay, E. F. (1999). Love's Labor: Essays on Women, Equality, and Dependency. Routledge.

Korsgaard, C. M. (1993). The Reasons We Can Share: An Attack on the Distinction between Agent-Relative and Agent-Neutral Values. Social Philosophy and Policy, 10.

Korsgaard, C. M. (1996). The Sources of Normativity. Cambridge University Press.

Levinas, E. (1961). Totalité et Infini. English: Totality and Infinity: An Essay on Exteriority (1969).

Levinas, E. (1974). Autrement qu'être ou au-delà de l'essence. English: Otherwise than Being or Beyond Essence, trans. A. Lingis, Duquesne University Press, 1998. Cited here: on the third party and justice, AE 245ff. / OB 157–159.

Løgstrup, K. E. (1956). Den etiske fordring. English: The Ethical Demand.

Mackenzie, C., & Stoljar, N. (eds.) (2000). Relational Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self. Oxford University Press.

Mackie, J. L. (1977). Ethics: Inventing Right and Wrong. Penguin. Cited here: ch. 1.

Nussbaum, M. (2001). Adaptive Preferences and Women's Options. Economics and Philosophy, 17.

Olson, J. (2004). Buck-Passing and the Wrong Kind of Reasons. The Philosophical Quarterly, 54.

Rabinowicz, W., & Rønnow-Rasmussen, T. (2004). The Strike of the Demon: On Fitting Pro-Attitudes and Value. Ethics, 114(3), 391–423.

Scanlon, T. M. (1998). What We Owe to Each Other. Harvard University Press. Cited here: ch. 2, pp. 95–100.

Scheffler, S. (1982). The Rejection of Consequentialism. Oxford University Press.

Scheler, M. (1913–1916). Der Formalismus in der Ethik und die materiale Wertethik. Gesammelte Werke, vol. II (Francke). English: Formalism in Ethics and Non-Formal Ethics of Values (1973). Cited here: the passages on das Nützliche and on the value modalities.

Sen, A. (1970). Interpersonal Aggregation and Partial Comparability. Econometrica, 38(3), 393–409.

Sen, A. (1970). Collective Choice and Social Welfare. Holden-Day.

Sen, A. (1977). Rational Fools: A Critique of the Behavioral Foundations of Economic Theory. Philosophy & Public Affairs, 6.

Sensen, O. (2011). Kant on Human Dignity. De Gruyter.

Street, S. (2006). A Darwinian Dilemma for Realist Theories of Value. Philosophical Studies, 127.

Street, S. (2008). Constructivism about Reasons. In R. Shafer-Landau (ed.), Oxford Studies in Metaethics, vol. 3. Oxford University Press.

Tronto, J. C. (1993). Moral Boundaries: A Political Argument for an Ethic of Care. Routledge. Cited here: ch. 5, pp. 127–136.

Williams, B. (1973). A Critique of Utilitarianism. In J. J. C. Smart & B. Williams, Utilitarianism: For and Against. Cambridge University Press.

Wood, A. W. (1999). Kant's Ethical Thought. Cambridge University Press. Cited here: the discussion of the formula of humanity.

Mo Di et al. Mozi. Cited here: Fa Yi 4, Jian Ai (middle) 15, Tian Zhi (upper) 26, Fei Yue (upper) 32, Jie Zang (lower) 25. Quoted from current editions and rendered by the author; no particular critical edition was adopted as a base text, and no edition or page is given.

Analects. Cited here: Wei Ling Gong 15.24 (15.23 in the older numbering). Quoted from current editions; shu is not given any single fixed English rendering.

引言

一个系列写完，最容易的是替它多说一句它没说的话。

前置篇清洗了指称空间。价值这个词被市场，被伦理，被日常语言同时征用，一个词至少驮着三样东西；前置篇以结构性否定排除的五类错位指认是价格，方向，余项的标价，阶梯，各说各话。此后 P1 把前置篇明标的五处机器立起来，P2 立价值的发生，P3 读价值的殖民，P4 给时间中的读数，P5 立冲突的裁断，P6 立他者的价值。

七篇都在机器之内工作。它们各有对象，各有承重，各自守着同一套纪律：不点名外部理论家，不援引历史事件，规范效力不支取。这套纪律使七篇写得干净，也使七篇说不出某几句话。它们说不出这台机器整个看上去是什么样子，也说不出它与别人造的机器在哪里分开。

本篇站到外面说那几句。站到外面要另一套纪律，本篇因此豁免不点名外部理论家这一条。豁免只及点名：本篇仍不新立对象层的命题，不设命题号，对象层的规范效力仍不支取。

关于本篇的篇位须说明一句。 前置篇 P1 P2 P3 四处正文写下过一件待完成的工作，收件人写作展望篇。本篇是那几笔前指的结算处，而结算的方式不是继承那个名字。第七节说明了理由：一处前指指向的是一项工作，不是替未来的成品锁定一个永久的书目类型。P4 首次出现新的地址，P5 与 P6 使新的分流结构清楚，而本 Paper 0 才公开完成这次改派。本篇按第七节所立的第二路作这一结算，并以 Paper 0 的身份写成。展望篇因此是本篇的历史别名，不是本篇的篇位。

---

第一节 这是一台什么样的机器

一 · 不是一条原理，是三张账

前置篇清出指称空间，其后六篇立起机器，合起来是一次构。这一次构立起的东西容易被收成一句漂亮的原理，那样收会失真，因为六篇读的根本不是同一类东西。

须先说一句总的：下面三张账是分析的账，不是篇目的分箱。每一段末尾所记的是立这张账的是谁，而在这张账上做事的不止它们，详见随后的矩阵。

构成账。 它问的是一项内容如何成为某个承担者的价值内容。这张账上的词是：承担域，价值源，候选，第一人称认可，价值内容的地位。立这张账的是 P1 与 P2。

运行账。 它问的是已经成立的东西在时间里怎么走。这张账上的词是：源位的在场，可达，具体内容的起落，重心，修订与退出。立这张账的是 P4，而 P3 读的是它被外来目的动过手脚时的样子。

行动与关系账。 它问的是两项要求撞在一起时怎么办，以及两个承担者之间发生什么。这张账上的词是：行动次序，裁断，涵育，殖民，共鸣，共同核验，相遇处的边界。立这张账的是 P5 与 P6。

一篇可以跨账，而且多数篇都跨：P3 的关系条件从外部进入，作用于显现 可达与修订这些运行通道，最后生成一项构成仿佛已经完成的错误读数，它一路横跨三张账；P6 的第一条命题回到构成账，第二条在认识权限那一层，后几条才在行动与关系账上。

账的作用不是让每一篇只领一格，是禁止主词彼此冒领。三张账的主词之别，与下节那两条禁止回流，合起来说明六篇为什么不能压成一篇来读。构成账上的主词是内容与位置；运行账上的主词是读数；行动与关系账是一个复合账，内分行动与关系两栏，主词包括行动要求 行动次序 事件 关系通道 权限与边界。把三张并成一张，就得先把这些主词抹平，而抹平之后每一篇的精确都没了。

三张账与六篇的关系，可以列成一张矩阵。它读起来是哪一篇在哪一账上做了什么，而不是哪一篇属于哪一账。

篇	构成账	运行账	行动与关系账
P1	承担域 价值源 法之他向内容	源的展开	涵育的条件侧
P2	第一人称认可 内容的价值地位	认可在时间中的渐次成形	社会承认的条件侧
P3	外来目的占着并冒充构成位的职能，而没有夺走构成位置，也没有在外部完成构成	可达 锁定 修订通道	外来目的与关系结构
P4	构成与运行的分界	在场 可达 具体内容 重心	可进入第三人称审计与共同核验并形成可错可修订读数的关系事实
P5	地位不回流	被让的内容仍在	行动要求 冲突 裁断
P6	那一项内容的登记 源不合并	第三人称读数的边界	价值地位与世界读数 共鸣 共同核验

二 · 两条禁止回流

三张账并列还不够，因为真正要紧的不是它们各说各的，是它们之间不许怎么走。七篇合起来立的是两条。

其一 · 运行 行动 后果与关系上的读数，不凭自身回流为价值源或价值地位。

P4 立产出与位置两轴独立时说得最直：拿一个看得见的东西去替一个原则上看不见的东西，以易得者代难得者，这一步在别处常常管用，在这里不管用，替下来的那个读数所证的不是要证的那件事。P5 立冲突时说，让不等于退出，因为让这个动作办不到退出这件事。P6 立比较时说，世界里的条件方案与后果可以一起核对，而那些读数不回流为价值地位。三处说的是同一条。

其二 · 外部条件可以塑形显现 可达 行动与关系通道，却不凭自身代行第一人称的构成。

P1 立涵育时说外部可以松开另一个人的网格；P3 立殖民时说，外来关系可以在显现位形成占位，在构成位以内摄的方式冒充构成，在可修订位形成锁定；P6 立他者时说显现 社会塑形与涵育都可以来自外面，而那道使内容成为他的价值内容的构成不在我这里做成。三处说的也是同一条。

这两条才是六篇共同做成的东西。 它们不是从某一条更根本的原理推出来的；它们是六篇各自在自己那张账上工作之后，账与账之间留下的两道禁行线。

三 · 四个节点

三张账上有四个节点被反复用到，此处列出，但不宣称它们彼此等价。

承担域。 目的价值由承担者承担。P1 立承担域时分了四格，其中第四格，无门，例如动物；一个无门的客体有它的体验，它可以是被关切的一端，而它不作为目的价值的承担者去充实那个位置。

第一人称认可。 一项内容成为某人的价值，须经他自己的认可，而 P2 写明认可不是核实：承担者不是在查明一项已经存在的事实，而是在自己的法之运作中把它认领下来。

不取共同外尺。 P1 的头号命题写道，一样东西一旦是承担者自己的，出自它自己的法，就再没有一个法外的坐标系能对它执行跨层或跨主体的价值审判。

来源不外置。 P1 统一地说：跨承担域，目的价值的来源锚在目的地位不外置这一条上；对完备主体，它的运行锚在它自己的法的自律反身运作上；对准主体，来源已在而运行方式尚在生成。

四者互相支撑，而不是同一命题的四种说法。来源在承担者一侧，不自动推出不能跨主体聚合：一套让效用依赖主体而仍建立聚合函数的理论，在形式上并不矛盾。反过来，主张没有共同尺度，也不自动推出来源依赖主体：一种客观的价值多元论完全可以同时主张价值客观而彼此不可通约。而由承担者承担，也不自动推出第一人称认可是唯一的构成事件，后者要 P2 的独立论证。四者之间是依赖关系，不是等价关系；本篇不把它们收成一句。

四 · 这台机器自己也是一次构

本体系的方法论立过凿构循环：一次构总有它凿不尽的剩余，那就是余项。

七篇是一次构。它有它的对象，有它的判据，有它读得出与读不出的东西，因此它也有余项。而余项不是本篇第六节要登记的全部东西：那里还有辖域上相邻而未曾立为问题的对象，有已经取得的极限结论，有本篇自己提的研究候选，也有已有明确地址的手递。把这些一概叫作余项，会使余项这个词在本体系里失去精度。 第六节因此分六格登记。

---

第二节 对接的层级与方法

本节在康德节与八支对接之前。理由是：一旦分歧线画得含糊，定位就会滑成评判，而分层没有先做的对接，最容易把两边不同层的话读成正面冲突。

一 · 层级不同

元伦理学，价值论，规范伦理学，福利理论，不是同一个层级上的东西。

元伦理学问的是评价判断是什么，有没有真值，真值根据在哪里。规范伦理学问的是正确的行动如何被决定。福利理论问的是什么对一个人而言好。而本系列所在的那一层问的是：一项内容如何成为某个承担者的价值内容，它如何运行，两项撞上时怎么办。

这一分要紧，因为元伦理诸立场对本系列的头一问多半不给答案。一个价值实在论者既可以把价值赋给人，也可以赋给事态，赋给经验，赋给活动；实在论本身不决定承担者是谁。所以拿一个元伦理立场当作承担者之争的对手，是把两层压成了一层。

二 · 三张分表：构成依赖 真值根据 认识权限

有三问常被同一个词带过，本节把它们分开，并逐一说明本系列在每一问上取得了什么。

表一 · 构成依赖。 一项具体的价值内容，是否依赖承担者的认可才成立。本系列在这一问上有明确立场：P2 立的第一人称认可是发生轴上唯一构成价值内容的位置，而认可不是核实。

表二 · 真值根据。 关于认可 价值与关系的判断凭什么为真。本系列在这一问上没有取得完整立场，本篇也不替它取得。 须留意：一项价值内容由主体构成，与关于它的判断有没有客观的真值根据，是两件事。即便内容由构成而来，下面这些仍可以是客观的事实：他有没有完成过那道认可；那项内容此刻还有没有地位；修订的通道有没有被封；某一处外部归属是不是在冒充构成。所以构成上的主体依赖，不等于真值上的反实在论。 元伦理内部对建构主义算不算实在论 客观性要不要求心灵独立，本来就有多种界定；本系列不由构成依赖一次结算掉那一整层。

表三 · 认识权限。 第一人称 第三人称与共同的程序各能知道什么。本系列在这一问上有部分立场：P4 立外部欠定，P3 立审计的上界，P6 立共同核验的权限分布。

因此本篇对本系列的元伦理定位只说这一句：它对具体价值内容的构成依赖有明确立场，对认识权限有部分立场，对完整的语义 真值根据与客观性理论尚未闭合。这不是示弱，是防止读者把它过早归进休谟式建构主义或一般的反实在论。

三 · 三种主体相关互不蕴含

一是对某人而言好，即某物在一个人的生活里使他的生活变好。二是行动者相关的理由，即理由的形式里不可消去地提到那个行动者。三是心灵依赖，即某种价值事实在形而上学上是否依赖态度 实践或理性活动。

三者互不蕴含。一个实在论者完全可以承认行动者相关的理由；一个建构主义者也可以建构出普遍有效而行动者中立的要求。

本系列所说的价值由承担者承担，落在构成那一侧，不落在前两样上。说它由承担者承担，说的是那个使内容成为价值内容的构成动作在承担者这一侧；它不是说价值总是对某人有好处，也不是说每一个理由的形式里都要提到行动者。 这一分在第四节接后果论与接森时都要用到。

四 · 主体之外的立足点有三义

一是心灵独立的真值根据：评价判断的真假不依赖任何人的态度。二是不偏不倚的立场：一个不因谁在行动而改变的评价位置。三是任何理性主体都能占据的普遍实践立场。

本系列的头号命题所否认的，是能对承担者自己的价值地位执行审判的那种外部立足点。这一否认落在第二义上，且限于价值地位这一对象；它不等于对第一义作了裁断（见本节表二），也不等于否认一切不偏不倚的评价位置在别的对象上可以成立。第三义的普遍立法立场与前两义不同，它不是站在所有人外面的观察者视角，而是任何理性意志作为理性意志都必须能够占据的立法位置。这三义的分别在第四节接后果论与接墨家时是承重的。

五 · 对接的做法

把一支传统放进坐标，说的是它把问题画在哪里，与本系列在哪一条线上分开。不说谁对谁错。这一条不是本篇自定的：本体系的权力论已有一篇元层自陈，其中写明与传统对接是定位诸传统在坐标系里的位置，不裁决高下。本篇沿用那一条。

这一条要守住不容易。分歧线一旦画得含糊，定位就滑成了评判；而把对手写弱一点，定位就成了自我表扬。所以本篇取一条更笨的做法：每一支先照它自己最强的形式说，再指出分歧线落在哪一条上，最后才说本系列在那条线上取的是哪一侧。若某一处本系列并未取得判断，就写明未取得，不借对手的弱处代替自己的证明。

---

第三节 康德的尊严与来源不外置的接口

本节不是对接的一支。康德是本系列自认的师承，而尊严与价格那一分与本系列的关系，比师承二字所示的更具体：在一个否定标记上，两者相接。

重合不是同一。 本节先说重合在哪里，再说重合到哪里为止。

一 · 共享的否定标记

康德在《道德形而上学奠基》第二节把有价格的东西与有尊严的东西分开。判据是可替代性：有价格者可以由等价物替代，超出一切价格因而不容许这种等价替换的，具有尊严。

前置篇清洗指称空间时，第一件事就是切走价格。切的方式不是宣布价格低而目的价值高。徽记那两道锁里有一道正是禁止这样读：珠与椟只分指称，不排高下。切的方式是构成条件：

> 主体数目在这里不是修辞，是构成：交换要对手，折算要公共尺，替代要等价物，价格的每一个构成动作都含一个对手方位置，目的价值的构成不含这个位置。

替代要等价物。这一句使两边在一个否定标记上相接。分别在于从哪一侧证，也在于各自另有什么承重：康德从地位那一侧说，说人格的地位不容许等价替换；前置篇从构成条件那一侧证，证价格的每一个构成动作都要一个对手方位置，而目的价值的构成不含那个位置。孤独主体的思想实验就是这一证的形式，一个宇宙里只有一个主体时，价格的构成条件全部落空，而目的价值照样成立。

这一处的重合是实的，而它的范围要说准。 不可等价替换是一个否定性的标记，它标出了尊严与价格的分界，而不是尊严的完整构成说明。康德的尊严还与人格 道德能力与自主立法相连，那几样不由这一个判据穷尽。所以本节说的是两者共享这一个否定标记，不是说本系列已经给出了康德尊严的完整解释。

二 · 四个对象必须分开

顺着这一处重合往下推，最容易一口气推得太远。有四个对象在这里挨得很近，本节逐一分开。

其一 · 人作为目的的地位。 这一条在本系列里先于任何人的认可而成立。P1 写道，他者是目的，这是一个事实，在我承认之前就是。

其二 · 一项具体内容的第一人称价值地位。 这一条要经承担者自己的认可才成立。P2 写道，认可不是核实，承担者不是在查明一项已经存在的事实，而是在自己的法之运作中把它认领下来。

其三 · 价值来源的锚定。 这一条说的是那个给出价值的位置在哪一侧，即来源不外置。

其四 · 殖民的结构。 这一条是 P3 的对象，说的是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。

四者不是一件事。 第一人称的认可所构成的是其二，不是其一：一个人不认可自己作为目的的地位，那个地位并不因此消失。其三是一个关于位置的结构事实，其四是一个关于关系与运行的结构事件。把四者压成一句，读起来有力，说出来不准。

三 · 占着职能，而没有夺走位置

此处经两轮更正，两次都偏，故把两次都记明。初稿把侵犯尊严写成位置被占，那是说满，与 P3 的位置恒在相犯；改稿又写成不是位置被占而只是一个读数，那是过校，把 P3 自己的占位与内摄语言一并否掉了。正确的一格在两者之间，而 P3 早已写好。

P3 同时立着三件事，须一并说，少一件都会偏。其一，没有真正的构成在外部完成：P3 立殖民的结构定义时写道，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。其二，三个位置在结构上恒在，殖民一律作用于可达，从不作用于在场，构成位始终锁在承担者自己的法里。其三，占位 内摄与锁定仍是 P3 自己的正式结构词：P3 立三个动作分居三位时写明，占位落显现位，内摄落构成位，锁定落可修订位，而它判殖民时正面问的就是构成位是否被一个外来目的占着。P3 判内摄时说得很直：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。而在构成位没有被冒充的场合，那一判据就不成立。

所以准确的说法是：一个外来目的占着构成位的职能，使一份外部担保在关系与运行中冒充第一人称的构成已经完成，而那道构成并没有真的在外部做成，构成位也没有被夺走。占着职能与夺走位置，是两件事。

这一分也解释了 P3 为什么把审计的结论限在位置与签名的形式上：构成位恒在，外部能取得的只是关系与通道上的读数，而第三人称不能由那些读数闭合第一人称的构成。审计能问的因此只有那个位置此刻是开是闭，那几条路通不通。

四 · 结构重合到哪里为止

综上，本节的结论可以写成一句：

> 康德的尊严与本系列的来源不外置，在不可等价替换 自主与资格不得完全外置这几处发生结构重合，而并非同一概念。殖民是侵犯主体性的某一种特定结构方式，不是尊严侵犯的完整定义。

说重合有用，说同一有害。 有用在于：本系列切走价格所用的构成条件论证，从自身内部给出一条与不可等价替换相邻的非替代边界，而这条路径不必先设定一份绝对价值。须说准的是，它不因此构成康德尊严命题的第二份证明：不需要等价物与绝不容许由等价物替代，不是字面同一的模态。有害在于：康德的尊严所连着的人格 道德能力与自主立法，本系列并未一一给出对应；宣称同一，等于替本系列认领了它没有写过的东西。

五 · 本系列在一场解释争论里的位置

康德研究里有一场关于尊严的争论，本节的读法在其中不站在任何一边，因此要说清楚它站在哪里。

一侧可以称作实质价值的读法。伍德把理性本性的客观价值放在理论中心，认为从普遍法则公式到人性公式是从纯形式进入实质内容。这一侧倾向于说：人性有一份无条件的价值，尊重的义务由这份价值得到解释。

另一侧不是简单的形式主义。森森反对把尊严读成一块先被我们发现 然后据以推出尊重义务的绝对价值属性，他把尊严与尊重更紧地系于理性法则与规定。博亚诺夫斯基同意不能把康德读成道德直觉主义，却反驳由此否认人性的绝对价值，主张实践理性的观念论结构与绝对价值可以并存。

本系列在这场争论里不站边，因为它对来源不外置的说明不是属性也不是规定，是位置。 那个给出价值的位置在承担者这一侧，这一点可以从构成条件上证，而孤独主体的思想实验证的就是它。

这个位置在争论里有它自己的代价与好处。代价是它不能像实质价值读法那样，直接从一份价值推出一条义务：本系列的对象层规范效力一律不支取，它说不出人当如何。好处是它不必回答那个使实质价值读法为难的问题，即那份绝对价值究竟是一种什么样的东西。而须诚实说明的是：不站边不等于已经解决了那场争论。 本系列给的是另一条路径，那条路径能不能替代前两侧所要解释的东西，本篇不判。

六 · 康德一侧最强的一条反驳，与本系列的两步

有一条对第一人称构成的反驳，在康德一侧是现成的，而且形式很强：人性是客观目的，并不因为人先在第一人称里选择要具有尊严；一个人不认可自己的尊严，甚至认可一条自我奴役的准则，也并不因此解除人性公式对他的约束。

本系列同意这一条，而且它是 P1 立的。 P1 分了两步：第一步是事实，这个具体的他者亦是目的，而他者是目的，在我承认之前就是；第二步是回应，从第一人称的位置承认并接续这个事实，承认做的是认这个事实，接续这个事实，而不是造这个事实。

第一人称的认可所成就的，从来不是任何人的目的地位，而是本节第二小节所列的其二：一项具体内容在这个承担者名下取得价值地位。目的地位在事实那一步，具体内容的认领在回应那一步；两步分开，那条反驳打的就不是本系列持有的立场。

还有一处也要分开，而这一处康德研究内部也在争。尊重他者的人性，与赞同他所选的每一个目的，不是一回事。迪安把道德上善的意志纳入人性的核心，对他的批评则指出：即使一个人的具体选择是恶的，作出选择的能力仍须被尊重。

本系列在价值论一侧给的是这一分的一个展开。P6 立的三条限界说的是我在他者的价值上没有哪几样处置：读不能闭合为可保证穷尽的读数，价值地位之间不取共同外尺，关系上的事不足以使既有的源合并。而这三条一条也没有说我须赞同他名下的每一项内容。不能处置与必须赞同，是两件事；本系列判前者，不判后者。

---

第四节 正面接的八支

八支按分歧线由内到外排：先是具体内容的地位凭什么成立，再是能不能排序，再是他者那一问，最后是有没有主体之外的立足点。

一处限定要先说。 下文所说的分歧，一律限于具体价值内容的地位这一主词。本系列并不主张一切价值事实都在承认之前不存在：他者是目的这一事实先于我的承认，价值源的结构形状不由某一次认领造出，世界里的条件与关系通道也各自成立。所以这一节的断层不是价值是被发现的还是被构成的，而是：一项具体目的价值内容在某个承担者名下的地位，是由他的第一人称认可构成，还是由独立于实际认可的合适性 理由关系或客观位阶来说明。

四之一 · 合适态度与买单转移

这一支的共同形式不是我实际喜欢什么什么就有价值。它是：一样东西的评价地位，应由什么态度对它是合适的，或我们有什么规范理由对它采取某种态度来解释。布伦塔诺论正确的爱与恨，尤因论合适的赞成态度，吉伯德论恰当的感受，斯坎伦重新把这一问题带回当代价值理论，是同一思想家族的几个关键节点。它们之间不是一条依次传递的师承链，本篇不作那样的叙述。

买单转移是更强的一步。斯坎伦主张，有价值并不是一个额外的一阶理由提供性质；说某物有价值，是说它具有某些别的性质，而正是那些性质给我们相应的理由去珍视它或以某种方式回应它。斯坎伦自己的用语是其他性质与自然性质，低阶性质是后来的便利说法。这一步常被分成一个正面主张与一个较有争议的否定主张。分法有用，而须先规定正面主张指什么：若它只说价值与某种合适的回应有构成性的联系，那么这一家族多数人都接受；若它已经说到由别的性质而非价值性质提供理由，那便已是斯坎伦这一版的一部分，不能倒推给更早的作者。

分歧线在这里。 这一支对本系列的反对，最强的形式不是说专家比本人更懂。那是一个概念上的论证：合适性或理由标准不由实际认领本身构成。 一个人可以因欺骗 因恐惧 因错误的事实信念而实际赞成某物；反过来，他也可以因抑郁 因无知 因扭曲的社会化而没有欣赏本来值得珍视的东西。若一项内容的评价地位由合适的回应来说明，那么实际的认领既可能不充分，也可能不是必要的。

本系列的答不是绕开这一条。P2 立第一人称认可时写得很直：

> 它更不是核实。承担者不是在查明一项已经存在的事实，不是在确定自己究竟是不是真的在乎某物，而是在自己的法之运作中把它认领下来。

两边的分歧因此可以说得很准：合适态度论把具体内容的评价地位系于一个独立于实际认领的合适性标准，本系列把它系于承担者自己的法之运作中的那一道认领。这是对同一个根据问题给出的两种不同的构成说明，而不是对同一个经验事实的两种观测。

这一支内部最著名的困难也从同一处来，即错误种类的理由。恶魔可以威胁你，不欣赏这件毫无价值的东西就惩罚你；于是你确实有审慎的理由去赞赏它，而这个理由似乎丝毫不能使它变得可赞赏。所以问题不是找到理由，而是非循环地把由对象给出的理由与由态度的后果给出的理由分开。拉宾诺维奇与勒诺-拉斯穆森，以及奥尔森等人对这一困难有专门的处理。若最后只能把正确的那一类理由定义成反映对象价值的理由，那么用理由解释价值的方向就绕了回去。这一困难本篇不判其成败，只指出它与本系列的分歧线在同一处。

四之二 · 舍勒的质料价值伦理学

这一支不在通常列举的谱系里，本篇接它，因为它是本系列头号命题最直接的对手。

舍勒把价值看作可以在情感的意向体验中被给予的客观价值质，并主张不同的价值样态之间有客观的高低次序。他的基本价值样态是四层：快适与不快适，生命价值，精神价值，神圣与不神圣。实用不在这四层之内，它是奠基于快适的派生价值，舍勒本人的用语是派生的或后续的价值。研究文献里另有把实用的地位处理得更强的重构，也有依原文反驳那种重构的，本篇取原文的四层，并记明这一分歧。

须先不打稻草人。 舍勒的位阶不是本系列在前置篇否掉的阶梯。前置篇切走的阶梯是把不同层级的价值放上同一把尺去排高下；舍勒的位阶恰恰否认同质单位。两支的分歧不在尺上，在序的根据上。

分歧是奠基上的，不是可以在共同前提内裁决的。 舍勒把价值位阶的根据放在情感意向体验的揭示上；本系列把具体价值内容的地位系于承担者自己的法与第一人称的构成。两者没有共同的地基，因此互不证明。 本系列的头号命题只能在本系列的前提之内拒绝那个位阶，它不构成对舍勒的内部反驳；同样，舍勒的位阶也不构成对本系列的内部反驳。

这一处正是定位不裁决那条纪律最吃劲的地方。 说本系列已经驳倒了舍勒，是把自己的前提当成了公共前提；而说两边无从对话，又抹掉了一处真实的分歧。准确的说法只有一句：分歧在奠基处，而奠基处的选择不由任何一方的内部论证决定。

与上一支相接的是：两支都把具体内容的评价地位系于某种独立于实际认领的东西，一支系于合适性，一支系于客观位阶。P2 共同承担这两处分歧的构成侧。而舍勒还多一处不由 P2 结清：他的客观位阶另与 P1 的无共同外尺相分。

四之三 · 后果论与行动者中立的价值

分歧线从这里换了一处。前两支争的是具体内容的地位凭什么成立，这一支与下一支争的是能不能排序。

后果论最一般的形式不是快乐最大化。它是把行动的正确性以某种方式函数性地系于结果的评价：从谁来行动，并不改变哪些总体结果较好。这是行动者中立的那一版；另有允许每个行动者使用不同评价函数的行动者相关的版本。

须先不打稻草人。 标准的最大化模型通常假设一个足够丰富的结果排序，而基数效用与单一的物理单位都不是后果论本身的逻辑要求；许多版本用一个序数的更好关系就够。排序必须完备也不是逻辑真理：允许不可比较的偏序版本仍可选出极大元，只是未必产生唯一的最好。

分歧线在哪，要先把一件事分开。 一般的行动者中立后果论需要一套不随行动者身份改变的结果评价，而它的评价对象未必是某个人的目的价值地位；那两者不是一回事。本系列处理的是具体价值内容怎样在承担者名下成立，它不是一套完整的事态价值论，因此不能直接为后果论提供结果评价函数。这首先是一项未供给，不是正面冲突。 另须记着第二节分过的三义：后果论所要的是第二义的不偏不倚，不必是第一义的心灵独立的真值根据。

本系列在这一处能说什么，须分三层说，否则会把结果排序压回价值地位。

第一层是未供给。后果论把结果与事态的评价交给一套事态价值论去填，而本系列所处理的是具体价值内容如何在某个承担者名下成立。两者不是同一个待填的模块，本系列因此提供不了那一套函数。

第二层是明确的冲突。若某一个版本把承担者的价值地位本身放上共同的量表，或者由那个量表直接取得结果排序，它与不取共同外尺正面相犯。这一层的冲突是实的。

第三层是未判。若结果排序用的是福利 伤害 资源 能力 权利满足 风险或别的信息基础，那么它排的不是某个人的价值地位，而本篇不能仅凭价值论去支持或反驳它。行动者中立的评价位置，与一个能对个人价值地位发判的法外位置，不是一回事；把二者等同，正是本篇第一节所禁的那种回流。

也因此，本系列判地位只取在与不在不取量值，并不证明完备的结果排序不能成立：即便某一项输入只取有与无，仍可另立词典序 集合包含 满足数量 优先规则或别的结果变量。本系列没有提供一套结果函数，这一句成立；本系列已经证明完备的结果函数不能成立，这一句不成立。

义务论一侧与本系列在某些行动者中立的总体化做法上形成相邻的反对，而理由与主词并不相同。 行动者中心的限制说的是，有些行为即使能减少同类错事的总量，也仍不应由我实施。谢弗勒的行动者中心特权更温和：即使总体结果最好是一个真正的理由，行动者仍可有不按完全不偏不倚的方式牺牲自己计划的许可。威廉斯的完整性反驳强调，把一个人的计划仅当作世界总效用函数里的一项输入，会改变他与自己计划的关系。

科斯嘉德警告过一件事：行动者相关与义务论不是逻辑上的同义词，那个形式区分不能直接等同于后果论与义务论之分。而理由属于谁 理由指涉谁 理由是否可共享这三样容易被混在一起，那是本篇自己的概括，不是她的原话。本系列不因与义务论一侧共享对手就归入行动者相关的价值理论：本系列判的不是理由的形式里提不提到行动者，是那个构成价值内容的位置在谁那一侧。

四之四 · 森与公共评价的跨主体信息基础

这一支是八支里最难挡的一支，而它照出的不是一个量的归属问题，是一整片本系列未曾进入的地方。

先要去掉三个常见的等号：偏好不等于效用，效用不等于福利。偏好在形式理论里通常是可选项之间的一个比较关系。效用在现代决策理论里首先是偏好的一种表示，不是边沁式的一份可直接观测的心理快乐物质。而福利说的是一个人的生活对这个人而言如何。偏好满足的福利理论可以把被满足的偏好当作福利的构成或指标，但那是一项实质的理论假设，不是效用表示的数学定义。

海萨尼的表示定理挡得住。 这里说的是聚合定理，与无偏观察者定理不是一回事。它的定义域是共同彩票，即同一组概率分布上的选项；个体与社会的偏好都须满足期望效用结构；再加一个帕累托或一致性条件，尤其是人人在两个彩票之间无差异则社会亦无差异这一条，才推出社会效用是个体效用的仿射或加权和。 那不证明宇宙里天然存在一把效用米尺。基数性来自彩票公理，而它只给每个人各自正仿射唯一的尺度；人际单位的可比性并不因这个表示定理自动产生，改变个体的尺度可以由权重吸收。要把等权的加总说成伦理上唯一的，仍须另有归一化或另一座桥。本系列反对的不是这个定理。

森挡不住，因为他不要那把尺。 他把三个问题分开：可测量性，人际可比性，加总。可测量性说的是一个人的效用表示保留多少信息；人际可比性说的是不同人的表示之间能否作有意义的对照；加总说的是把它们合成一个社会评价。三者不能互相混同，也不能由其中一个自动推出另外两个。须说准的是，加总是由个人信息形成社会排序的操作，并不是与前两者同型的第三个测量维度。而他主张的是部分可比性：在完整可比与完全不可比之间，存在有结构的中间地带，据以得到部分的社会排序，而不必假装存在一把完美的尺。

前一稿在此处犯过一个错，此处更正。 前一稿在正确说出这一三分之后，又把它压回一个对象，称作福利经济学拿来作人际比较的那个量。那与刚刚建立的三分相犯。 森的工作恰恰说明：人际比较不必只有一个总量；比较的信息可以不完备；社会排序可以是部分序；信息基础可能是效用 福利 功能 能力或别的对象；而可比较不等于可加总。

所以这一支照出的问题要重新写。 它不是那个量落在哪一侧，是：

> 公共评价在跨主体使用的各种信息基础，与第一人称的目的价值地位之间，是什么关系。

这一问底下至少有五件事，本系列没有完整处理，只已经给出两道边界：福利 效用 能力 资源 健康 机会与后果各自是什么主词；其中哪些可以共同核验；哪些可以形成部分的比较或公共的排序；它们为什么不因此回流为个人的价值地位；以及当公共评价取得行动或制度效力时，那份效力从何而来。

那两道边界是共同核验与不得回流，见下。 P6 第三节写道，两个承担者各自内容的价值地位之间不取共同外尺，而世界里的条件方案与后果可以一起核对，那些读数不回流为价值地位。所以本系列并不禁止一切跨主体的比较，它禁止的是把比较的读数回流到价值地位上去。而上面那五件事，本系列既没有给出它们各自的主词，也没有给出它们与价值地位之间关系的完整说明。 登记见第六节。

四之五 · 承认理论

分歧线再换一处，从排序换到他者。

这一支不是一个统一的主张，须先分层。 黑格尔的承认是在意识与法权的发展结构里说的；霍耐特把承认分成爱 法律的承认与社会连带三个层次，各自对应自信 自尊与自我价值感，而承认的缺失以蔑视的形式损伤这几种实践的自我关系；弗雷泽则反对把承认理解成主体间的心理关系，主张一种地位模型，认为不承认的核心是制度化的文化价值模式使人无法作为同侪参与互动，因而须与再分配一起处理。而在她的框架里，文化模式不是唯一的障碍，资源上的匮乏同样妨碍平等参与。霍耐特与弗雷泽之间的分歧，可被重构为关于承认的对象 解释层级与社会本体论的争论。须说明这是一种解释性的重构，不是双方共享的中立描述：弗雷泽会强调地位秩序与平等参与，霍耐特则认为社会冲突与制度也可以在更基础的承认语法里理解。

这一支不是本系列的简单反面。 在黑格尔与霍耐特那一线，相互承认的核心恰恰要求把对方当作自主而有能力的判断者，而不是把他当作一个待评估的对象。所以它与 P6 所立的那三条限界方向一致，而不是相反。

分歧线在别处，而且只落在其中一线上。 黑格尔与霍耐特那一线否认这样一幅图：一个完整而自主的主体先独立存在，然后偶然地与他人发生关系。而本系列的孤独主体思想实验，正面主张目的价值的最低成立条件无须对手方；P1 又写明，孤独主体形式余位空着也自足。

两边在这里真的分开，而分开的是哪一线要说准。 与孤独主体正面相接的是黑格尔霍耐特那一线，它使实践的自我关系依赖承认结构；弗雷泽那一线首先把不承认定位为制度化的地位从属，它更邻接的是制度那一处。本系列说目的价值的成立不依赖第二个主体，而承认所成就的是另一件事。本系列不否认关系对一个人的形成有巨大作用，P1 立涵育与 P3 立殖民说的都是外部进得来；它否认的是那个使内容成为价值内容的构成动作要一个第二方。依赖与构成，是两件事。

弗雷泽那一侧与本篇第六节所登记的制度那一处相邻：她把不承认接到制度化的文化价值模式上，而本系列在处理共同内容与外尺时照出了制度，却从未把它立为自己的问题。这一句是说材料在哪里，不是把那一处交给谁。

四之六 · 列维纳斯

这一支与本系列的关系最不齐整：一半接近，一半正面冲突，而且冲突处正是本系列最要紧的那一处。

先要限定。以他者的不可还原为出发点这句话，最准确地指向列维纳斯一人。布伯的对话哲学，勒格斯特鲁普的伦理要求，德里达对他的继承与批评，都与这一方向邻近，而它们的理论结构差异很大。把它们并称为一个他者学派，是本篇为叙述方便所作的分析性分组，不是一个自我认同 学说统一的学派。所以此处只以他为对象。

他最有辨识度的反转是：伦理关系不是先有一个认知主体，把另一个人识别为具有某种性质或价值，随后才产生义务；相反，在面对面的相遇中，他者以一种不能被我的概念完全吸收的方式向我呈现，而我发现自己已经被置于责任之中。《总体与无限》所写的他者是一种外在性，而那不只是空间上在我之外：他者不能被还原成我的表象 我的概念，或我的总体中的一个内容。

接近的那一半。 我不能以总体化的认知宣称已经完全掌握他是谁。这与 P6 所立的第一条限界方向一致：对一个具体他者的价值，第三人称证据不能闭合为可保证穷尽的读数。两处都在拒绝同一件事，即以一份自己做的表征去把另一侧结清。

冲突的那一半，而且冲突得很干净。 他者的伦理要求并不以我的认可为成立条件，责任在意义结构上甚至先于我自由地选择接受它。而本系列所立的正是：一项内容成为我的价值内容，须经我自己的第一人称认可。

分歧真正落在哪里，可以说得更准，而这里有两条读法。若把那份责任读成一种价值内容，两边正面冲突，本系列的答就是 P2 的那一句，认可不是核实。若把那份责任读成本系列不处理的规范要求，两边不冲突而只是各说各的，因为本系列的对象层规范效力一律不支取。两条路本篇都指出来，不裁哪一条更对。 这是执行定位不裁决那条纪律的代价，不是回避。

另有一处可作结构类比。 他并没有停在每一个单独的他者都不可比较。一旦出现第三者，即另一个他者以及所有其他人，就必须问：对眼前这个人的无限责任，如何与其他人的要求共存。正义 比较 斟酌与制度的必要性正由此产生。这是《别于存在》里的论述，第三者使无限的单向责任受到限制并成为正义之问的出生点，比较与判断在紧随的段落展开。所以说他永远拒绝任何比较是不准确的，准确的说法是：比较是次生的，由多个他者迫使伦理关系进入正义。

而 P6 的射程也在同一处转折：第二节说一个具体的他者，第三节说两个承担者之间，第四节上到全体，而第三节正是比较第一次成为问题的地方。两处的形状可作类比：从二到多的时候，比较才成为一个必须回答的问题。 但这只是类比，不是同构：他的第三者是伦理关系内部由无限责任逼出的正义问题，P6 的射程递增是本系列自己的论证次序；两者的对象不同，回答也不同。他走向正义与制度，本系列走向世界里的条件可以共同核对而读数不回流为价值地位。

四之七 · 关怀伦理与适应性偏好

这一支打的是本篇第六节所登记的准主体那一处，而且它带来八支里最难答的一问。前一稿在此处答得过满，此处重写。

关怀伦理共同反对的不是一切原则，而是把成熟的道德行动者预设为彼此独立 依赖只是偶然例外的那幅图。它从依赖 照料 脆弱性 具体关系出发。特隆托把照料分成四个阶段：关切，照顾，给予照料，接受照料；与之对应的四种道德品质是专注，责任，能力，回应。两组不是一组，本篇此处照原文分开。基泰把不可避免的依赖与照料劳动推到正义理论的中心。而关系性自主这一发展不否认自我决定，它主张自我决定的能力本身依赖社会关系与制度条件。

第一处分歧不必打成对立。 本系列以承担者为单位，而这一支从关系与依赖出发；但本系列并不否认关系对一个人的形成有巨大作用。它否认的只是那个使内容成为价值内容的构成动作要一个第二方。依赖与构成，是两件事。

难的是第二处，而这一处本系列答不完。 经由适应性偏好那一路，反对可以说得很尖锐：一个在长期不利条件下长大的人，可能真诚地报告自己满足，而那份满足恰恰是不利条件塑造的。若一项内容只要经他第一人称认领就在他名下取得地位，那么第三方能不能说那份认领本身有问题。

前一稿把这一问答成了二选一：要么真是她自己的，要么就是殖民，而第三人称分不出。那个二分漏掉了太多。 她可以确实认领而所据的事实严重有误；可以确实认领而选项极端受限；认领可以是真实的而依此生活会损害她的健康与机会；偏好可以受压迫形成而她并未因此失去能动性；一项内容里可以同时有真实的认领 外部的塑形与不可退出的条件。适应性偏好的文献内部本来就在争论这些，不存在一个公认的简单二格。

本系列在这里能给的，是一张账，不是全部的账。

账	问什么	本系列取得什么
价值地位账	这项内容有没有取得她的第一人称认可	构成由承担者第一人称完成；第三人称可以形成可错 可修订的判断，而不能代行认可，也不能把判断升级为终局的认证
福利与利益账	依此生活对她的健康 能力 机会与生活状态有什么后果	本系列未立
权力与权利账	形成的过程里有没有强制 封路 操纵与不可退出	本系列部分已立，另有权力论与权利论

三张账可以同时成立，而且不互相自动结算。 一项内容在第一张账上取得了地位，不因此在第二张账上得分；在第三张账上查出封路，也不自动取消第一张账上的地位，而是说明那一处须按 P3 的判据另行审计。

所以本系列对这一路的答，准确说是这样一句：它能为适应性偏好的讨论增加一张构成与通道的账。第三人称可以对位置与通道形成可错 可修订的判断；它不能代行当事人的认可，也不能把自己的判断升级为终局的认证。而本系列不能用这张账去替代福利的评价 权利的审查与事实的判断。 后三样本系列没有立，本篇也不替它立。

须留意一处措辞。 关怀伦理没有一把尺，不等于随便选；它要求的是对情境敏感的实践判断。而判断力这一术语在七篇正文里零出现，也未被明示立作独立的解释机件，其完整的理论工作由 P5 交行动论。所以本系列在这一处不与它争，只是不在同一片地上说话。

四之八 · 墨家的法与天下之利

最后一支打的是本篇第二节所分的那三义中的第二义：有没有一个不因谁在行动而改变的评价位置。

墨家立法仪以为仪表，主张治天下者不可无法仪，而法仪不可只取于父母 师与君，莫若法天。此处的法是规范与标准，不是现代意义的立法。兼相爱交相利被立为公共的标准，而它诉诸天的无私：天之所欲则为之，天之所不欲则止。这一支是本篇所接各支里最接近一个超出各私人立场的公共裁判标准的一支。

须保留它自己的语境，不作现代化。 墨家的天不是一个中立的观察者位置，它有欲有恶，赏善罚暴；法仪取于天，是因为天之行广而无私。把它写成现代意义的人格神则有争议，本篇不作那一步。另须记明：墨家并非简单的神命论，他们也以义 利 治等独立的理由说明天意为何值得作标准。这与后果论所要的不偏不倚的评价立场形式上相近而根据不同：后果论的不偏不倚是一个方法论上的要求，墨家的无私是一个宇宙论上的根据。把墨家写成中国的功利主义，既不准确，也会使这一处的对接失去价值。

分歧线此处经两轮更正，仍要再改一次，而这一次是往回改。 初稿举非乐与节葬为例，说墨家以天下之利为据判某些人所珍重的礼乐为不当，因而伸进了价值成立那一侧；改稿把它收窄为一个待证的问题。核到原文之后，须说的是更强的一句：那一步原文本身没有走。 非乐明确承认音乐耳知其乐，华饰目知其美，美食口知其甘，然后才因王公大人兴造乐器 征敛与占用劳力，判其上不中圣王之事，下不中万民之利。节葬下同样以富 众 治检验厚葬久丧这一制度化的实践。 所以墨家所判的是为乐与厚葬作为公共实践 政事与资源配置，而不是证明所珍重之物本身不值得珍重。非乐甚至有一句几乎是针对这一误读预先作出的排除。

由此，两支的对象本不重合。墨家的法判行为与政事，本系列判一项内容如何成为某个承担者的价值。而仍可以问一句：

> 公共的法能不能限制由个人价值组织出来的行动，而不同时取得替那个人重写他的价值地位的资格。

这一问正好落在本篇第一节所立的两条禁止回流上。若公共之法只作用于行动那一侧，它与本系列的第二条不冲突；若它进一步以行为的判定回流为对那个人价值地位的判定，那才与本系列相犯。就墨家的文本而言，上面所核的两篇落在前一侧：它们判的是实践与政事，没有回流。 分歧因此不在有没有公共标准，也不在墨家越没越轴。它在别处：本系列否认存在一个能对承担者自己的价值地位发判的法外坐标系，而墨家的法本就不主张那个资格。两支在这里是不相犯的，而这一处的对接价值恰在于此。一支主张公共标准而不伸进价值成立，正说明公共标准与本系列的头号命题可以并存。

顺带说一处相邻的。 儒家一线与本系列的对照落在别处，不在立足点那一问上。恕这一条说己所不欲勿施于人，它以自己这一侧为起点推及他者。本系列那一步不是这样走的：他者是目的，这一条在本系列里是前提，不是从我自己是目的推出来的结论；两项目的地位分别成立之后，来源不外置才在两边给出同型的结构。所以两者都从第一人称的语言进入他者关系，而推论的结构并不相同。本篇因此不把它接成同盟：恕是一条待人的准则，本系列那一步是一个前提与结构的分账。这一处只作定位，见下一节。

---

第五节 另七支的位置

这七支不与本系列直接相犯，或者相犯之处已由前一节的某一支涵盖。此处只说它们在坐标里的位置。

元伦理的三条分界线。 认知主义与非认知主义分的是规范判断有没有可真可假的命题内容；实在论与错误论分的是这类话语是否系统地失败，麦凯认为日常道德判断预设了世界中不存在的客观价值，因而按其表面语义系统地为假；稳健实在论与心灵依赖诸说分的是真值根据独不独立于态度。第二节已经说过，这三条线本身不决定价值由谁承担。而按第二节的三张分表，本系列对真值根据那一问尚未取得立场，因此它在这三条线上都没有位置可标。 这不是一处遗漏，是一处未闭合。

建构主义。 这一支最值得说清，因为它离本系列最近。科斯嘉德的程序实在论说的是：规范问题之所以有正确答案，是因为存在正确的实践理性程序，而不是因为程序在发现一个先于程序而存在的独立规范领域。斯特里特则更明确地把理由的实质内容系于行动者偶然的评价出发点，她既有对实在论的挑战，也有对理由的建构主义的正面陈述。

近，而不能归入，这一分要写准。 斯特里特所说的评价出发点是一个经反思一致性结构化的整体立场，不是一次显式的第一人称认领行为；本系列所立的是一个构成事件，即一项已在某一价值源域中取得候选位置的内容，在承担者自己的法之运作中取得可调用 可重述与可修订的地位。两者在反对价值完全独立于评价活动这一点上可能相近，而本系列目前没有给出完整的理由理论 真值理论或反思一致性的程序，因此不能被正式归入休谟式建构主义。 这一句与第二节表二是同一个未闭合。

义务论一侧的行动者中心限制与特权。 它与本系列在某些行动者中立的总体化做法上形成相邻的反对，而理由与主词并不相同，前一节已接。此处只补科斯嘉德那一句警告：行动者相关与义务论不是逻辑上的同义词，那个形式区分容易把理由属于谁 理由指涉谁 理由是否可共享三件事混在一起。

价值多元论与不可通约。 三个词要分开：不可通约说的是缺少共同的度量单位；不可比较说的是两项之间不存在任何比较关系；前者不推出后者。张美露另立第四种关系，即两项在某一评价覆盖下既非更好 非更差 亦非相等，而仍在同一评价空间之内。

本系列在这一格上的位置只能说到这里，前一稿说过了头。 本系列明确主张价值地位之间不取共同外尺，并拒绝更高 更低 相等这几个谓词。它是否进一步主张不可比较，或者允许类似 parity 那样的非度量关系，现有对象层正文没有判。 前一稿以世界里的条件与后果可以共同核对为据，断言本系列不落在不可比较那一格；那一步不成立：世界侧的对象可以进入共同评价，不说明两项价值地位之间存在某种正面的比较关系。这一处登记为未判，见第六节。

经济学的偏好 效用与福利。 三个等号要去掉：偏好是可选项之间的比较关系，效用首先是偏好的表示，福利是一个人的生活对他而言如何。价格已由前置篇分流出去。而偏好 效用与福利仍须分别定位：它们与目的价值地位的关系，不能由价格的分流代为结清。真正相接的那一处是森，前一节已接。

海萨尼的表示定理。 在期望效用公理与若干一致性条件下，社会评价具有个体效用加权和的表示。这不证明存在一把天然的效用米尺，只证明在明确的规范与表示条件下，加总可以有严格的结构。且须说清它不解决什么：它不自行回答哪些福利信息可以被人际比较，权重从何而来，以及那份规范根据是什么。本系列反对的不是这个定理，因此两边不构成对立。

儒家一线。 恕这一条说己所不欲勿施于人，它以自己这一侧为起点推及他者。本系列不由我自己是目的推出他者是目的：他者是目的在本系列里是另立的前提，两项分别成立之后，来源不外置才在两边给出同型结构。两者都从第一人称的语言进入他者关系，而推论结构不同。恕字在英译里没有唯一固定的对应词，本篇也不把它缩成现代意义的交换互惠。本篇因此只作对照，不接成同盟。 另须说明：不宜假装存在一套统一的东亚价值论，儒墨两家在本篇里各自出现，正因为它们在立足点那一问上分得很开。

---

第六节 边界与去处总账

这一节登记这台机器没有继续往下写的地方。

登记之前要先分身份。 前一稿把以下各项一概称作余项，那使余项这个词在本体系里失了精度。凡构必有余项，这是方法论所立；而没有继续写的东西不止余项一种。本节因此分六格。

格	是什么	本节所登	所在小节
一	本次构留下的余项	集体承担者能否另行成立	一
二	辖域上相邻而未曾立为问题的对象	制度；公共评价的跨主体信息基础	二 三
三	已经取得的极限结论	第三人称在准主体处没有外部阈值	四
四	本篇提出的研究候选	准主体第一人称运行的生成史	五
五	已有明确地址的手递	行动论 意义论 权力论 权利论	六
六	回看时才揭出的对象层未判	价值地位之间还剩什么关系	七

第二格占两个小节，格号与小节号因此自第三格起错开一位。下文各小节标题一律标出格号。

一 · 第一格 · 本次构留下的余项：另一个承担者能否另行成立

河边那些人一起清了淤，拦下过采砂，凑钱修了一段护坡。P6 判的是那几件事本身推不出他们的源已经合并，根据是各人的源锚在他自己身上，而几个源的指向交错并不等于它们并成一个。

它没有判另一件事，而且是明写不判的。P6 的原话是：至于另一个承担者能不能在别的条件下另行成立，本节不问；就算有，那也不等于原来那些人的源被它吞掉了。

这一处是一处明文的留置，而它同时也是一处余项。前一句由 P6 明写不判给出，后一句要另有证明，而那证明不能由某一条对象层的实体命题直接推出，须由机器整体的输入形式与 P6 的明文留置共同说明。

本系列的一切源位 构成与运行谓词，都以一个已经成立的承担者为主词。P1 立承担域时给的是谁能承担目的价值，即完备主体与准主体两层，而不是一个新的承担者如何发生；它说一个承担者作为目的价值承担者的资格与它价值来源的不外置当下都已经在场，也说门的弃用是运行上的事而不是承担资格的丧失。从若干主体或关系结构过渡到一个新的承担者，这一类型转换在七篇里没有对应的发生谓词。

而这不只是偶然漏写一句现成的判准，它暴露的是当前机器的输入类型缺少一条转换机件。理由如下。若强行用现有的源位与构成谓词去说明那一道转换，就得先把那个尚待成立的承担者作为这些谓词的主词，论证因而循环。这个由输入类型自身造成的无法吸收，才是把它列作余项的理由。

有两件事不由这一条推出，须就近写明。其一，它不表示第三人称不能描述一个新承担者的形成条件：描述形成条件与代行构成是两件事，本篇第一节的第二条禁止回流禁的正是后者，不是前者。其二，它不预判关系能否在某一次成立中成为原因或条件，那一处 P6 明写不判，本篇不替它判。留置的界限也要说准，否则登记本身就越了界。P6 判下的是关系上的事本身不足以自动推出源已合并，也不足以证明一个新的承担者已经成立；它没有判共同关系在某一个承担者的形成过程里能不能是原因或条件，而且写明了不判。 所以此处可以登记的只有一句：若另一个承担者成立，须另行说明它的承担结构与它的源位是怎样取得的，而源位随承担者的结构给定，这是 P1 立下的。共同关系在那一次成立中算不算数，价值论不判。

二 · 第二格之一 · 辖域邻接对象：制度

P6 写到共同做的事，共同定的规矩，共同守的东西，本身都不因共同而有错，并给了一项共同内容何时成为外尺的条件。那是判它有没有越界的条件，不是它是什么。而本系列从头到尾没有把制度是什么立为自己的问题，连提出它都是顺带的。

顺带照出而未曾立为问题，与明写不判，不是一回事。 前者本系列连问都没有问，谈不上留置，因此它不是余项，是辖域上相邻的一个对象。若有人要接，接的也不宜是制度本身如何可能那样宽的问题，而是本系列照出的那一段：共同的规矩，共同的决定与共同的安排，在什么条件下取得制度结构。

第四节接承认理论时说过，弗雷泽把不承认接到制度化的文化价值模式上。那一侧是这一处最近的现成材料。这一句说的是材料在哪里，不是把这一处交给谁。

三 · 第二格之二 · 辖域邻接对象：公共评价的跨主体信息基础

第四节接森时照出了另一处，而它比前一稿所写的宽得多。

本系列已经说过的那一半是：价值地位之间不取共同外尺，而世界里的条件方案与后果可以一起核对，那些读数不回流为价值地位。所以本系列不禁止一切跨主体的比较，它禁止的是把比较的读数回流到价值地位上去。

而底下这一整片本系列没有进入：福利 效用 能力 资源 健康 机会与后果各自是什么主词；其中哪些可以被共同核验；哪些可以形成部分的比较或公共的排序；它们为什么不因此回流为个人的价值地位；以及当公共评价取得行动或制度效力时，那份效力从何而来。

这一处同样不是余项，是辖域上相邻的对象：本系列在处理跨主体比较时照出了它，从未把它立为自己的问题。

本节不写它的去处。 按第七节所立的定义，写下去处即是一笔前指，也就是一笔债；而本篇立那条规矩，自己一笔债也不开。

四 · 第三格 · 已经取得的极限结论：第三人称在准主体处没有外部阈值

一个还说不出自己要什么的婴儿，他的价值来源不由我代给，这一条 P6 判得很实。P6 也写明：他名下的具体价值内容发生到了哪里，不由会不会说话来判，言语本来就不是认可的必要签名。

这一句常被读成本系列在此欠着一把尺。不是。P6 同一段还给了一个门槛：他还不能作那一项内容的第一人称认领方，因为十五维那一行还不在他内部现实运行。而那个门槛给的是那一项内容的定义域，不是准主体一切具体价值内容的通用门槛。 要一把尺的是另一问：那一行的运行走到了哪一步。

而那一问，本系列判的不是没给，是给不了，并且已经判过。P2 立发生轴时写明，认可可以发生在一瞬，也可以渐次成形，经年才定，时间的长短不判这件事；P4 立外部欠定时写明，没有一组必要且充分而又可以拿到别人身上去用的外部签名。两条合起来，第三人称没有一套外部阈值可以终局认定一项价值内容在准主体那里已经成立到了哪一步。

这是一条已经取得的否定性结论，不是一个待补的漏洞。 它使理论不再向那个方向前进，而那是理论的结果，不是理论的缺口。本节因此把它列在第三格，不列在余项之内。

这一格与第一格不是同一个问题，两者所问的层不同。第一格问的是一个新的承担者是否成立，那是承担者类型的转换，属本体与发生；这一格问的是一个已经进入承担域的准主体，其某一运行何时达到某个状态，属认识权限。第一格当前缺一个发生谓词，这一格已经取得没有外部终局阈值这一结论。

两处可以说的只有一句：都暴露了外部读法不能代行第一人称结构这一限制。说到同一个问题为止，不能再往下并。

本系列在这里到头，不是因为它的判准只有一种形式。它另有取别的形式的判准，例如 P5 的可撤性，那一条 P5 写明主词是行动。到头的理由是本系列已经把两件事切开，而 P4 把这一刀写得很直：关系结构里的事实可以分辨，而由分辨那些事实转不成判定构成。第三人称可以取得关于生成条件 关系位置以及表达与修订通道的可错 可修订的读数，这些事实可以进入共同核验；一项认可是否已经作为第一人称的构成发生，却不能由这些外部事实闭合出来。

五 · 第四格 · 本篇提出的研究候选：准主体第一人称运行的生成史

由那道已立的边旁侧照出的，是本篇的一项候选：准主体的第一人称运行如何逐渐取得可运行的生成史，这一段能不能被另行描述。

它是本篇提的建议，不是本系列留下的边，也不是余项。 若有人要写它，有一条须先立在前面：那样的描述不得反过来成为一套外部程序，用来认定到了这一步就算价值内容已经成立。越过这条线，它就落回本系列已经判过的那一边。

第四节接关怀伦理时那三张账，正是这一条线的实际用法：本系列说明哪些关系位置与通道可以进入审计，也说明第三人称的判断不得代行第一人称的构成，不得升级为终局的认证。

六 · 第五格 · 已有明确地址的手递

第一至第三格与第六格，是既有正文四种不同性质的未继续；第四格是本篇后见提出的研究候选；第五格是已经写明收件人的手递，接的是谁写在已发表的正文里。六者均不是余项的同义词。

行动那一整支交了行动论。P5 收束时写道：一道行动次序如何形成，如何被执行，如何在事后被回看，判断力，习惯，迟疑与能力如何进入其中，整个交行动论，本篇不用它们，也不对它们下任何判。判断力在其中。判断力这一术语在七篇正文里零出现，也未被明示立作独立的解释机件，其完整的理论工作由 P5 交行动论。

一项内容此后如何被同一条命持续担着，交意义论。P5 与 P6 两处都作了这一手递。P6 收束的原话是：他者可以有他自己的价值而这不由我代给，这一项内容此后如何被同一条命持续担着，交意义论。

方向价值归权力论。这一支自前置篇清洗指称空间时就划出去了，P1 立源时又写明：方向价值归权力论，而权力论所立的凿削须有两个主体。价值论的唯一对象是目的价值。

两个主体相遇时那条界如何被追问，归权利论。P5 立裁断时写道：他在看这一次动手会触到谁的什么，那条界在哪里被碰着，而这条界如何被追问，争议按哪一种方式处理，归权利论。相遇边界与两种处理模态，权利论已在论基础权利的厚度与边界一篇里立过。

已指定不等于已完成。本篇只登记地址，不替接收的系列作完成的验收。这些工作不再属于价值论对象层的写作任务，而相应前指的发布账仍须逐项结算。

四处的当前状态如下。本节只登记到这一栏为止，不在此完成任何一项验收。

去处	原交件	当前可登记的状态
行动论	行动次序的形成 执行与回看，判断力 习惯 迟疑与能力	地址明确；是否兑现另验
意义论	一项内容此后如何被同一条命持续担着	地址明确；是否兑现另验
权力论	方向价值	地址明确；已有文本是否完成该交件另验
权利论	相遇边界与两种处理模态	地址明确；正文已指出相应实体，仍可另验

七 · 第六格 · 回看时才揭出的对象层未判：价值地位之间还剩什么关系

价值地位之间不取共同外尺，而更高 更低 相等这几个谓词已被 P1 P5 P6 排除。但是否进一步主张两项价值地位之间不存在任何比较关系，或者是否允许某种非度量的关系，现有对象层正文没有判过。

第五节已说明，不能以世界里的条件与后果可以共同核对为据来替这一问作答：世界侧的对象可以进入共同评价，不说明价值地位之间存在某种正面的关系。

本篇只登记这一处未判，不替它取答案。

---

第七节 前指是债

本篇为什么只登记而不许诺，理由在本系列自己的记录里，而那笔记录关于本篇。

先要说明这一节所立的是什么层的东西。 它不是价值论对象层的一条规范，不说人当如何。它是一条发布纪律：一个系列在已发表的正文里写下的指向，此后如何结算。对象层的规范效力不支取这一条，在本节仍然有效。

一 · 什么算前指

不先定义，凡是指向未来的话都成了债，而第六节那几处就会与自己相犯。

前指指的是：已发表的正文以确定的语气，把一项具体而未完成的工作交给一个已经命名的后续篇 系列或理论位置，并以这一交付完成本篇射程的结算。

辖域邻接对象，已经取得的极限结论，研究上的候选，以及未判的登记，都不因此成为前指。 第六节六格中，只有第五格的四处手递属于前指。它们的地址已经明确，而完成状态须逐项另记；本节所结清的，是既有正文指向展望篇的那一组前指。其余五格，一处也不是前指。

二 · 三路结算

一处前指是一笔须公开结算的账。结算有三种。

一是兑现：那一篇真的写出来，做那件活。二是公开改派：后来的正文说明这件活换了去处，以及为什么。三是公开撤销：说明这一问已被后来的工作消解，或者本就问错了，并给出理由。

拆开来结，一部分兑现一部分改派，或者一部分撤销，那是三路的组合，不另立第四类。

判一笔账结没结，看的不是它写在哪一种文件里，是它进不进规范的发布链，读者能不能从正式文本或公开的勘误里查到。 公开的版本说明算，内部的编制记录不算：读者查不到它，那一笔账在纸面上仍然悬着。

三 · 本篇自己就是一笔这样的账

前置篇写下：价值如何走向行动，留展望篇。P1 以较宽的措辞写下同一件事：本系列到展望篇为止，行动论是另一支的事；展望篇要从目的的约束讲到行动的显现，它为将来的行动论奠基。P2 写下：已经认可的价值内容如何走向行动，是展望篇的入口。P3 写下：已认领的价值内容如何走向行动，是展望篇的入口。四篇指的是同一件工作，收件人都写作展望篇。

P4 写了同一个位置的那一句，而收件人换了。P4 的原话是：已认领的价值内容如何走向行动，是行动论的入口。它与 P3 那一句只差最后三个字。而 P4 全篇再没有出现过展望篇三个字，也没有一句说明这一句与前几篇那一句的关系。

在那里出现的是一个地址分叉，不是一次已经办完的改派。 按上面所立的规矩，公开改派要说明这件活换了去处以及为什么；P4 没有说，它只是照新地址写了下去。此后 P5 把行动那一支扩全，由判断力一项扩到判断力，习惯，迟疑与能力一整支，另把生命的承接交意义论；P6 在收束处明文交意义论，未再解释它与旧的展望地址是什么关系。新的架构由此成形，而旧地址与新地址一直并存着，没有结算。

本篇是这笔账的结算处，而结算的方式是第二路。 那件工作已经改派：一段自 P4 起交行动论并由 P5 扩全，一段由 P5 与 P6 交意义论。既有正文中指向展望篇的那几笔前指，由本 Paper 0 在此结清。

结清不等于继承那个名字。 一处前指指向的是一项工作，不是替未来的成品锁定一个永久的书目类型。那项工作已经按两路全部改派出去，行动的形成与执行归行动论，同一条命的持续承接归意义论；本篇另承担一项旧前指未曾规定的元层工作：站到七篇之外自陈这台机器 与既有传统对接 并登记边界与去处。展望篇因此是本篇的历史别名，不是本篇的篇位。

四 · 因此本篇一笔新债也不开

第六节第二格里的第二项特意不写去处，正是为守住这一条。一处未曾立为问题的对象若顺手交给某个系列，本篇便在立规矩的同一页上开了自己的第一笔账。

一篇立记账规矩的文章，最要紧的不是它把规矩说得多好，是它自己照不照办。

所以本篇不作许诺。第一至第四格与第六格的登记，都不是本篇对未来写作作出的承诺：它们若后来被写了，后来的文本不必回头引本篇作依据；若一直不被写，本系列也不因此欠账。

第五格不适用上面这一句。它那四项是既有正文已经写下的前指，地址虽已明确，完成状态仍须分别按兑现 公开改派或公开撤销结算。

---

第八节 收束

价值论的对象层到此为止。

前置篇清出指称空间，说价值不是什么。此后六篇立起一台读价值的机器，而它不是一条原理的六次展开：它记的是三张账，构成账 运行账 行动与关系账，各有各的主词。六篇之所以不能并成一篇来读，这两道禁行线是其中最要紧的理由，即读数不凭自身回流为源与地位，外部条件不凭自身代行第一人称的构成。

把这台机器放进既有价值论的地图，它与八支在各自的一条线上分开，与另七支不在同一片地上说话。分开处最内的一条是：一项具体目的价值内容在某个承担者名下的地位，由他的第一人称认可构成，还是由独立于实际认可的合适性 理由关系或客观位阶来说明。这一条只管具体内容的地位，不扩到一切价值事实。

与康德的接口在一个否定标记上相接：不可由等价物替代。相接是实的，而共享一个标记不是概念同一；那是一个否定性的标记，不是尊严的完整构成说明。

而这台机器停下的地方，不能一概叫作余项。本篇分六格登记：一处是本次构留下的余项，两处是辖域上相邻而未曾立为问题的对象，一处是已经取得的极限结论，一处是本篇提出的研究候选，四处是已有明确地址的手递，还有一处是回看时才揭出的对象层未判，即价值地位之间还剩什么关系。

关闭一台机器，不是把它旁边的一切都叫作开口，也不是替它多说一句它没说的话。本篇的全部工作，是把它是什么样的机器说出来，把它与别人的机器在哪里分开说出来，再把它没说的那些，按各自的身份登记为没说。

---

致谢

本篇为 SAE（Self-as-an-End）哲学体系价值论分支的 Paper 0。

作者资格与文责：本篇由秦汉独立署名，全部论证 取舍与最终定稿均由作者裁决，文责由作者独任。

起草与评审的事实：起草辅助与交叉评审借助大型语言模型进行。起草与内部精度审计由 Claude（Anthropic）承担，本体极限与矛盾检测由 Gemini（Google DeepMind）承担，边界测试由 Grok（xAI）承担，结构闸门与发布前签署由 ChatGPT（OpenAI）承担。第四节与第五节的外部文献材料，由 ChatGPT 的深度调研整理。谨向这些系统及其背后的研究与工程团队致谢。它们提出的意见由作者逐条裁定，采纳与驳回均记录在案。

外部引证的核验程度：本篇对外部引证作逐条的分级核验，分为四档，即已到原文确认；原文到手而精确位置未确认；据可靠二手确认；未能确认。各档的最终净数与逐条状态见另立的外部主张分级核验表。本篇不声称全部外部引证均已到原文确认，也不声称已覆盖全文外部承重。

本篇的更正分三类，此处分开记明，不合为一类。

其一，结构复审所致的实质重写五处：撤去把四根承重写成逻辑等价并进一步压成只读位置的写法，改立三张账与两条禁止回流；撤去把尊严 目的地位 具体内容与殖民合并的写法，改为四对象分开并按 P3 更正殖民的结构，即位置恒在而外部担保冒充构成已经完成；把价值被发现还是被构成这一断层的射程收回到具体内容的地位；撤去由承担者依赖直接推出心灵依赖与否定心灵独立真值根据的写法，改立构成依赖 真值根据 认识权限三张分表并写明本系列在第二张上未闭合；撤去是他自己的价值即对他好这一推论，因其使价值地位与福利换轴，森 适应性偏好与第六节三处随之重写。

其二，外部材料的实质更正两处：舍勒的基本价值样态是四层而非五层，实用是奠基于快适的派生价值，前稿把它列为并列层级；墨家非乐与节葬所判的是为乐与厚葬作为公共实践与政事，而非乐明确先承认了音乐之可乐 华饰之可美 美食之可甘，故前两稿由公共批评推出所珍重之物本身不当，跨了一步原文没有做的推论。

其三，体系内部上游回核所致的更正一处：殖民一处经两轮更正，初稿说满，改稿过校，本稿取两者之间的一格，即占着构成位的职能而未夺走它。这一处不由外部核验得出，而由回读 P3 原文得出。

另有外部表述的收窄三处：合适态度那一支的谱系不写成单线师承；海萨尼定理补写其定义域与中央假设；森的三分改为不能互相混同而非互相独立。

外部文献的页码按逐条的确认程度处理：到原文确认者引页码；原文到手而精确位置未确认者，只列已确认的篇章或文本范围，不冒充精确页码；据可靠二手确认者引出处而不引页码；未能确认者不引位置，并在正文相应处收窄。前一稿曾把页码整批删去，那是一次过度反应，它同时删掉了可靠的与不可靠的，并掩盖了两者的分别。

本篇所引的每一句上游正文，均逐条比对已发表原句而成；本篇对外部各家的说法，另经逐条分级核验。两项核对各另立一表，作起草与评审的用具，不随本篇发布。定位诸传统的取舍与分歧线的画法，不在两表之内。

关于本篇的核验限度，此处一并写明，不以别处的表代为声明：本篇的外部定位以分级核验为据，其中只有到原文确认的那一档可担保与所据原文相符；本篇不声称已覆盖全文的外部承重，也不声称各家的分歧线已经外部审看。中国古典文献所引依通行本，未采用特定校注本作底本。

---

参考文献

本体系

以下文献均存档于 Zenodo，中英双语版本同记录。所引 DOI 为概念 DOI，始终指向最新版本，故不另注版本号与访问日期。全部文献单一作者署名（秦汉）。起草与评审辅助已在致谢中说明。

秦汉. (2026). 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） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.18528812

秦汉. (2026). 三层二维统一结构：Self-as-an-End 的完整框架（SAE 基础论文 Paper 3） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.18727326

秦汉. (2026). 方法论总论：凿构循环（SAE 方法论） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.18842449

秦汉. (2026). 权力机器之势（SAE 权力论系列 Paper 0） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.20707047

秦汉. (2026). 论权力的起源（SAE 权力论系列 Paper 1） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.20370224

秦汉. (2026). 论基础权利的厚度与边界（SAE 权利论系列 Paper 2） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.20769253

秦汉. (2026). 论权利不预设义务（SAE 权利论系列 Paper 0） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.21314907

秦汉. (2026). 价值不是什么（SAE 价值论系列 前置篇） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.21362575

秦汉. (2026). 论价值的源（SAE 价值论系列 Paper 1） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.21405207

秦汉. (2026). 论价值的发生与承认（SAE 价值论系列 Paper 2） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.21437038

秦汉. (2026). 论价值的殖民（SAE 价值论系列 Paper 3） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.21611154

秦汉. (2026). 论价值的动力学（SAE 价值论系列 Paper 4） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.22050204

秦汉. (2026). 论价值的冲突（SAE 价值论系列 Paper 5） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.22122819

秦汉. (2026). 论价值与他者（SAE 价值论系列 Paper 6） [预印本]. Zenodo. https://doi.org/10.5281/zenodo.22191071

外部文献

以下仅列正文实际征引者。页码按逐条的确认程度给出：经原文核验确认者列页码，据可靠二手确认者只列出处，未能确认者不列位置。各条的确认程度见另立的外部文献核验表。

Bojanowski, J. (2015). Kant on Human Dignity: A Response to Oliver Sensen. Kant-Studien, 106(1), 78–87.

Brentano, F. (1889). Vom Ursprung sittlicher Erkenntnis. English: The Origin of Our Knowledge of Right and Wrong.

Buber, M. (1923). Ich und Du. English: I and Thou.

Chang, R. (2002). The Possibility of Parity. Ethics, 112.

Dean, R. (2006). The Value of Humanity in Kant's Moral Theory. Oxford University Press.

Derrida, J. (1964). Violence et métaphysique. In L'écriture et la différence (1967).

Ewing, A. C. (1947). The Definition of Good. Macmillan.

Fraser, N., & Honneth, A. (2003). Redistribution or Recognition? A Political-Philosophical Exchange. Verso.

Bruckner, D. (2009). In Defense of Adaptive Preferences. Philosophical Studies, 142.

Colburn, B. (2011). Autonomy and Adaptive Preferences. Utilitas, 23.

Elster, J. (1983). Sour Grapes: Studies in the Subversion of Rationality. Cambridge University Press.

Fraser, N. (2000). Rethinking Recognition. New Left Review, 3.

Gibbard, A. (1990). Wise Choices, Apt Feelings. Harvard University Press.

Hieronymi, P. (2005). The Wrong Kind of Reason. The Journal of Philosophy, 102.

Khader, S. J. (2011). Adaptive Preferences and Women's Empowerment. Oxford University Press.

Mackenzie, C., & Stoljar, N. (eds.) (2000). Relational Autonomy: Feminist Perspectives on Autonomy, Agency, and the Social Self. Oxford University Press.

Nussbaum, M. (2001). Adaptive Preferences and Women's Options. Economics and Philosophy, 17.

Harsanyi, J. C. (1955). Cardinal Welfare, Individualistic Ethics, and Interpersonal Comparisons of Utility. Journal of Political Economy, 63.

Hegel, G. W. F. (1807). Phänomenologie des Geistes. English: Phenomenology of Spirit.

Honneth, A. (1992). Kampf um Anerkennung. English: The Struggle for Recognition (1995). Polity Press.

Kant, I. (1785). Grundlegung zur Metaphysik der Sitten. Akademie-Ausgabe, Bd. 4. 本篇所引为第二节论目的自身 尊严与价格的部分。

Kittay, E. F. (1999). Love's Labor: Essays on Women, Equality, and Dependency. Routledge.

Korsgaard, C. M. (1993). The Reasons We Can Share: An Attack on the Distinction between Agent-Relative and Agent-Neutral Values. Social Philosophy and Policy, 10.

Korsgaard, C. M. (1996). The Sources of Normativity. Cambridge University Press.

Levinas, E. (1961). Totalité et Infini. English: Totality and Infinity (1969).

Levinas, E. (1974). Autrement qu'être ou au-delà de l'essence. English: Otherwise than Being or Beyond Essence, trans. A. Lingis, Duquesne University Press, 1998. 本篇所引为论第三者与正义处，AE 245ff. / OB 157–159。

Løgstrup, K. E. (1956). Den etiske fordring. English: The Ethical Demand.

Mackie, J. L. (1977). Ethics: Inventing Right and Wrong. Penguin.

Olson, J. (2004). Buck-Passing and the Wrong Kind of Reasons. The Philosophical Quarterly, 54.

Rabinowicz, W., & Rønnow-Rasmussen, T. (2004). The Strike of the Demon: On Fitting Pro-Attitudes and Value. Ethics, 114.

Scanlon, T. M. (1998). What We Owe to Each Other. Harvard University Press. 本篇所引为第二章论价值与理由的部分。

Scheffler, S. (1982). The Rejection of Consequentialism. Oxford University Press.

Scheler, M. (1913–1916). Der Formalismus in der Ethik und die materiale Wertethik. English: Formalism in Ethics and Non-Formal Ethics of Values (1973).

Sen, A. (1970). Interpersonal Aggregation and Partial Comparability. Econometrica, 38(3), 393–409.

Sen, A. (1970). Collective Choice and Social Welfare. Holden-Day.

Sen, A. (1977). Rational Fools: A Critique of the Behavioral Foundations of Economic Theory. Philosophy & Public Affairs, 6.

Sensen, O. (2011). Kant on Human Dignity. De Gruyter.

Street, S. (2006). A Darwinian Dilemma for Realist Theories of Value. Philosophical Studies, 127.

Street, S. (2008). Constructivism about Reasons. In R. Shafer-Landau (ed.), Oxford Studies in Metaethics, vol. 3. Oxford University Press.

Tronto, J. C. (1993). Moral Boundaries: A Political Argument for an Ethic of Care. Routledge.

Williams, B. (1973). A Critique of Utilitarianism. In J. J. C. Smart & B. Williams, Utilitarianism: For and Against. Cambridge University Press.

Wood, A. W. (1999). Kant's Ethical Thought. Cambridge University Press. 本篇所引为其中论人性公式的部分。

墨翟等. 《墨子》. 本篇所引为法仪第四 兼爱中第十五 天志上第二十六 非乐上第三十二 节葬下第二十五。所引依通行本，各句为作者自译或直录；本篇未采用某一特定校注本作底本，故不注版次与页码。

《论语》. 本篇所引为卫灵公第十五，第二十四章（旧编第二十三章）。所引依通行本，恕字未取某一固定英译。