Self-as-an-End
SAE The Second Enlightenment · Paper 7

At the Door
门口

Han Qin (秦汉) · Independent Researcher · 2026
DOI: 10.5281/zenodo.24030415 · Full PDF on Zenodo · CC BY 4.0
Abstract

This series takes up the half of "a person is an end" in which oneself is an end, and goes as far as the door to "others are ends too." This paper examines the door. From oneself being an end to others being ends too, the first step is necessary, yet not sufficient: a person who thinks for himself can go on to take any other person whatever as an end, or can recognize only his own group. What the door looks like: the scale in one's own hand stops before another person's share; not reading another's mind, not first demanding proof of worth from him, not legislating his law for him, not crossing his line. This step needs each person's own judgment, and judgment cannot be replaced. Rules are being written more and more into programs and models; whoever decides on a deployment settles in advance a conditional procedure and lets it run in scenarios to come, and he bears the whole of this judgment of deployment. The norm cannot support counting heads; this is the floor. A rule can at most allow it as a last resort; even when it is allowed, those responsible are still held to account: they can be questioned, and they bear responsibility for the consequences. This series parts from Parfit's response to "not counting the numbers," and from Taurek's prior agreement on shared rescue resources, because of one commitment only: a person has no equivalent.

This is an independent English rewrite, not a translation, of the Chinese final text v1.0. Terminology and framing are recomposed for an English-reading audience; the philosophical content is the same.

Introduction

Outside the door there is always another person. The papers so far have all written the half of "a person is an end" in which oneself is an end, and the starting point was this: a person has no equivalent, and the share a person cannot let go of is not weighed for him on someone else's scale (Prequel, §1). Between oneself being an end and others being ends too there is a door, and this series goes as far as the door (Prequel, §6).

This paper examines the door. It has two things to take up. One is the door itself: how the step between oneself being an end and others being ends too is taken, and what the door looks like. The other is what taking this step depends on: each person's own judgment. Judgment cannot be replaced; and today more and more judgments are being written in advance into programs and models, which is where this series feels the urgency (Prequel, §3). So this paper also takes up who bears that moment once rules are written into programs.

§1 From Oneself Being an End to Others Being Ends Too

Oneself being an end is the step before others being ends too. Without this share as its ground, the recognition that "others are ends too" has nowhere to grow (Paper 2, §1): a person who has handed even his own share over to another's scale will mostly see, when he looks at others, only the readings on the scale.

Yet this step is only necessary, not sufficient. That oneself is an end does not imply that others are ends too. Paper 1, §4 describes the two roads before a person who thinks for himself. One goes as far as "others are ends too," any other person whatever. The other goes only as far as one's own group: only this group are recognized as ends, the members recognize one another, and those outside do not count. Such a group can think and can calculate, and its members hold one another up; it can stand only by keeping those outside pressed down, and in an environment that leans toward the kingdom of ends, it does not last. The kingdom of ends is Kant's term: everyone treats himself and others as ends, never merely as means (Kant 1785, AA 4:433). It is also more dangerous than a person who harms others for himself alone: a group of people holding one another up can write a judgment into an institution and carry it out year after year (Paper 1, §4).

So to get from this end to the other, one more thing is needed in between: each person's own judgment. The person in front of me, his share is his, and is not weighed by my scale; this recognition no rule can make for a person, and no share of "oneself is an end" brings it with it automatically. §3 takes this up.

On this series' reading, Smith stands at the end where the setting out begins. He wrote of exchange: each for his own ends, and each getting what he wants in the exchange (Paper 1, §3; Smith 1776). At the opening of The Theory of Moral Sentiments he also said that, however selfish a person may be supposed to be, there are evidently principles in his nature that make him concerned with the fortunes of others (Smith 1759). In reality the two halves are always mixed together, and this series accepts that too. Nothing is said here about how far Smith went, only about where the setting out begins: his side on exchange is taken as the starting point, and from here the way leads to the door.

§2 What the Door Looks Like

On the other side of the door is another person's share. Arriving at the door, there is first one thing that stops: the scale in one's own hand stops before the other person's share. The scale is still there, and what has a price is weighed as before: the work he does, the price he asks, the few hours this task needs of him, can all be weighed (Paper 1, §2). What stops is the scale's reaching into his share. Any other person whatever has no equivalent either.

Each of the earlier papers describes one aspect of what stopping looks like; taken together, these aspects show what the door looks like.

  • Not reading another's mind. Looking at what he did and what he said aloud, without deciding for him what he has avowed within (Paper 3).
  • Not first demanding proof of worth from another. His share does not have to be converted first into some use for someone before it counts (Paper 6).
  • Not legislating his law for another. Another person's share is not given by anyone else and can only be set by that person himself (Action Theory, Paper 2, D11; that proposition is stated on the premise of "still taking others as ends too," and the door borrows it to name the step at which one stops on arriving here, not to derive from it that others are ends too).
  • Not crossing another's line. Not putting one's own ends in place of his law, and not crossing his line to violate him (Paper 2, §§4, 5).

These four "nots" set out the boundaries of this recognition that cannot be cancelled in action. Others who see these boundaries kept still cannot certify, on a person's behalf, that "he has recognized any other person whatever as an end"; that recognition still lies with him.

These four "nots" are addressed to a person who can decide for himself. Suppose an elderly man can no longer decide for himself in certain matters, and someone handles those matters for him: that is a matter for another place, for rules to set; it needs grounds, its range is drawn by institutions, and it can be questioned; no rule decides a person's own affairs for him when he can decide them himself (A Way Out, a Way Forward, §3, E41).

Suppose a man of thirty quits a secure job to learn watch repair. His father says, "You just want to run away for a while." This is reading his mind for him. He goes on to ask, "How much can you earn a month at that?" Asking how much he would earn could be an ordinary question; but if the father makes earning that much the condition on which he will accept the matter, so that if it cannot earn that much it does not count, this is first demanding proof of worth from him. Then: "Get your proper job sorted first, then we can talk about this." This sentence can be no more than an opinion; but if the father makes it a condition, so that unless the son goes back to work the help already promised will be withdrawn, in order to make him take the road the father has designated instead, this is legislating his law for him. Each of these three sentences reaches deeper into the son's share than the one before.

On this side of the door, there is in fact still a great deal the father can do. He can say he is worried, say he disapproves, speak truthfully about the hardships he has seen in learning a trade, and also explain what help he is willing to give. He can also stop subsidizing his son's rent: to recognize a person is not to be bound to provide for him without limit; it is his own money, though his own money does not automatically vouch for how it is used either. Merely not adding a further sum is not enough to decide that he does not recognize his son; support already agreed on, the entanglement actually caused, and whether he used resources to decide his son's road for him, are checked separately against the facts (Paper 6, §3; A Way Out, a Way Forward, §3, E41). The scale stops; the person has not walked away. What stops is only the scale.

Doing these things only for one's own group does not yet count as having reached the door. On the road described in §1, members of the group may well not read one another's minds, and may not demand proof from one another; what belongs to the door is doing these things for any other person whatever, without first asking whether he is one of one's own group. The other person beyond the door is any other person whatever, not humanity as a whole, and not only the few one cares about; care for the people one cares about often comes earlier, and it is not yet this recognition, yet it is not lower than it either; this framework continues this layer in Action Theory.

The door is made of a few "nots." These "nots" are all stoppings on one's own side, not accomplishing anything for others. Stopping is not only keeping still, either: if one's own hand has already sealed off another's road, taking the hand away is also part of stopping. Past the door, how one treats others and what one does for them is another matter (§6). Nor can these things be done by following a checklist: which share is his, where the line lies, whether the question in front of one is asking about a matter or weighing a person, has to be judged by oneself each time. This brings us to the next section.

§3 Judgment Cannot Be Replaced

Paper 1, §2 sets out three layers. Above the three layers there is a root: a person is an end, and a person has no equivalent. The top layer is the norm, which gives principles that come from this root; what principles cannot settle is handed to rules; what rules cannot settle in full falls to the person present at the time. Downward, there is guidance and testing, and room left for the layer beneath; upward, there is no supplanting. A single judgment does not generalize into a rule or a norm; a rule cannot alter the norm either. Supplanting is overstepping. To supplant the norm with a rule is to speak of an exception as if it were a principle; worst of all is to supplant the norm directly with a single judgment: a person decides on his own that the person in front of him is not worth it, and treats this as the way people may simply be regarded.

The step at the door falls on the lowest layer. The few "nots" of the previous section are words at the level of the norm; but whether the question in front of one is asking about a matter or weighing a person, and how far his share reaches in this case, the norm cannot say and rules cannot write out in full. Kant said in the Critique of Pure Reason why they cannot be written out in full. A rule gives something general; to know whether the case in front of one falls under a given rule, one would need another rule to teach it; yet to use that rule, one would again first have to judge whether it applies to this case. Pushed down in this way, it never ends. So he said that judgment is a peculiar talent which cannot be taught, only practiced (Kant 1781/1787, A133/B172). He also said that examples are the go-cart of judgment (A134/B173–174).

Philosophy's famous trolley problem stops right at this place. Foot set out the example: a runaway tram, whose driver can only steer it from one narrow track onto another; on one track five men are working, on the other one (Foot 1967); Thomson discussed it in 1976 under the title "the trolley problem" (Thomson 1976). Paper 1 said that this series does not answer for anyone whether to pull the switch; it only says that a rule cannot answer for him. Whatever rule is written, what is written in is a procedure settled in advance, which cannot replace the judgment of that later moment; when the moment really comes, it is still the person present who judges, and he who bears it.

In reality there have been cases that came to this step. Germany's Aviation Security Act of 2005 contained a provision: if an aircraft was to be used against human lives, and shooting it down was the only means of averting this danger, armed force could be used against it. On 15 February 2006, Germany's Federal Constitutional Court held this provision incompatible with the Basic Law, and void. The reasons were on two levels: the Federation's use of the armed forces here exceeded the powers the Basic Law gives it; and so long as there are people on board who have nothing to do with the hijacking, shooting the aircraft down uses their deaths as a means of saving others; they are treated as objects, and deprived of their rights at the same time (BVerfG 2006, para. 124). The court rejected this grant of state authority; as for how a shooting down carried out nonetheless, and the order to shoot down, would be assessed under criminal law, the court said this did not need to be decided here (para. 130). The line drawn by the norm, and how responsibility is assigned to the particular people who act, part here: if that moment really came, the one judging would judge for himself, and the law would question him afterward.

The step at the door is like this too. No rule can recognize on a person's behalf that "the person in front of me, his share is his." Rules can guide him and test him; facts, reasons and feelings can all enter his judgment, but how it applies in this case, how to weigh and choose, and who takes it up, the one judging still has to do himself (Action Theory, Paper 1). If he recognizes it, that is his judgment; once judged, it is his to bear, and it can be questioned. That a judgment is his does not make it right (Action Theory, Paper 1, §5).

§4 Rules Written into Programs

In the examples of the previous section, there was still someone present at that moment. But today, more and more judgments are being written in advance into programs and models (Prequel, §3). The nearest example is autonomous driving.

In 2017, the Ethics Commission on Automated and Connected Driving, set up by Germany's Federal Ministry of Transport and Digital Infrastructure, delivered its report, listing twenty ethical rules. Rule 8 says that genuine dilemmas of life against life depend on the actual concrete situation at the time, and on behavior of those involved that is hard to predict; they "can thus not be clearly standardized," nor programmed so as to be ethically beyond question; technological systems cannot replace, or decide in advance, the decision of a responsible driver with the moral capacity to judge. The same rule also says that a driver who, in an emergency, kills one person to save one or more others acts unlawfully, but not necessarily culpably; legal judgments of this kind, made after the fact and taking the particular circumstances into account, cannot readily be turned into abstract, general judgments in advance, and so cannot be turned into corresponding programming either (Ethik-Kommission 2017, Rule 8).

Rule 9 says that in unavoidable accidents, any distinction by personal features such as age, gender, physical or mental constitution is strictly prohibited, and offsetting victims against one another is prohibited too. As for programming in general to reduce the number of personal injuries, the commission says this "may be justifiable"; the last sentence of the same rule adds: "Those parties involved in the generation of mobility risks must not sacrifice non-involved parties" (Rule 9).

The explanatory part of the report goes into more detail. If programming reduces the risk to every single road user in equal measure, general programming aimed at reducing the number of personal injuries could be justified, without breaching Article 1(1) of the Basic Law: before the victims can be identified, it is also in their interest. At the same time, the commission refuses to infer from this that human lives may be offset against one another; it classifies the killing or serious injury of people by automated vehicle systems as wrong without exception. Where the lives of several people are already directly threatened, it holds that a requirement to choose the course with the fewest deaths can be defended; however, it says, discussion of this part has not yet been fully completed, and consensus has not been reached on every point (Ethik-Kommission 2017, IV.1.6). The commission went this far.

One step further down is this series' own word: distributing risk in advance can be reckoned; the provision written in, once it lands at that moment on people who can be identified, the norm cannot support; a rule can at most allow it as a last resort, and even when it is allowed, those responsible are still held to account. The norm does not say counting is forbidden, and it also has a direction: where counting can be avoided, there is no counting (Paper 1, §2).

Fully autonomous driving pushes this to the limit. Paper 1, §2 looked at counting heads in three stages: beforehand, prevention; in the moment, judgment; afterward, questioning. Beforehand, the remit is prevention: braking in time, seeing clearly, driving slowly, so that the car never comes to the point where counting begins. What is reckoned at this stage is risk and cost, and no one is yet standing on the scale; what Schelling called the "statistical life," reckoning what a small reduction in risk is worth, belongs to this stage (Schelling 1968). Yet more and more rules will be written into programs and models, and at the layer of rules there will, in practice, be times when heads are counted.

There may be no one in the car able to take over the judgment on the spot. Whoever decides on a deployment has settled in advance a conditional procedure, and decided to let it run in scenarios that have not yet arisen one by one. The provision in it that counts heads, the norm cannot support; a rule can at most allow it as a last resort, and in a program it is the same: only when every other road is closed does it come into play. The judgment of deployment he has made is his to bear in full, the chief executive of the company, for instance, not whoever wrote the code; he cannot say "the program decided it" or "the standard told us to write it." What actually happens later will bring new judgments of checking, revision and halting operation; that one occasion beforehand did not settle the later accounts in advance. Once it is in place, it remains under oversight and open to questioning, and checks beforehand and questioning afterward cannot replace each other (A Way Out, a Way Forward, §4, E42).

Even when it is allowed, those responsible are still held to account. Being held to account has two layers: being questioned, and bearing responsibility. Suppose a car, caught between protecting the people inside it and a pedestrian outside, has been written to accelerate into the pedestrian, and the pedestrian dies. This is not merely a matter of answering a few questions afterward. The commission itself says that those involved in generating mobility risks must not sacrifice those not involved (Rule 9); whoever settled this provision and put it on the road is responsible for that life.

Questioning does not begin by assuming he was wrong: the norm does not say counting heads is wrong either. What is asked is: was prevention done, were the other roads truly closed, how was that provision settled, and what has happened since it was put in place? Beyond questioning there is also responsibility: the consequences of that moment are his to bear, and it does not end once the questions are asked. However many more were saved, the one who was not saved is not wiped off the account.

Whose decision, his to bear. Those who write the code and run the tests follow it, and bear their own step. Their step does not thin out the share of the one who decided; nor is the decider's share spread thinner because many people below him follow it. To say "the company decided it," like saying "the market decided it," does not stand in for him: whatever is said to be done by a whole has to be traced back to the people who make the decisions at the key junctures (Power Theory, Paper 2). How much the company, as a legal person, pays in compensation or in fines is a matter for the layer of rules and the institutional side; what is spoken of here is on whom that judgment falls.

§5 The Objections from the Other Side

Against this paper, and against what Paper 1 said about counting heads, the strongest objections come from two places: one in philosophy, one at the scene of emergency care.

The one in philosophy begins with Taurek. He wrote a paper whose title is itself a question: "Should the Numbers Count?" He set out an example: a drug, all given to one person called David, can save him; divided among five others, a fifth each, it can save those five. He said he did not feel obliged to save the five; he could toss a coin, giving each person an equal chance. What matters, he said, is the loss each person himself suffers: five people each losing a life does not add up to any one person losing five times as much (Taurek 1977). He and this series go by a similar road: neither adds people up. But this series does not decide for the person present what method to use: tossing a coin is a judgment, counting is a judgment too, and both are borne by whoever judges.

The next year, Parfit responded to him. Parfit said that five people each losing their lives is worse than one person losing his; giving equal weight to each person does not require giving each an equal chance; counting everyone in leads one to save the larger group (Parfit 1978). Brought here, the objection runs like this: you say counting heads is putting people on the scale; but not counting is failing to take the extra four people seriously. He can ask one step further: why should taking five equal pleas for rescue together amount to putting a price on people?

This objection is strong. This series does not say that Parfit is wrong; it parts from him only at this step, and the place where it parts is "taking together." Five pleas for rescue, looked at one by one: each is one person's whole life. On this reading, to say that five weigh more than one, there first has to be a scale on which one life and another can be placed, to weigh out five as more than one. That a person has no equivalent says precisely that there is no such scale: a life is not offset against other lives; one does not offset one, and five do not offset one either (Prequel, §1). Taurek's saying that the losses of the five do not add up to any one person losing five times as much also stops here. So the level of the norm cannot support counting heads: it does not say that counting is right, nor does it give "saving the larger group" a ground at the level of the norm. This is the floor.

This is not to say that not counting is right. Not counting is also a judgment, and also has to be borne; the norm vouches for neither side. When it truly comes to the moment where every other road is closed, and the person present counts and decides by heads, that act is his: he bears it, and is held to account afterward; however many more were saved, the one who was not saved is not wiped off the account.

Taurek went one step further, closer to this paper. He imagined a rescue resource set up in everyone's name, taking a coast guard as his example: when it is set up, no one knows whether he will end up on the larger side or the smaller; an agreement to save the larger group, giving each person the greatest chance of being saved, is something everyone can accept in advance. The person in charge of the resource rescues according to the agreement, and his duty comes from that agreement; explaining his duty in this way does not require anyone to accept that he himself, or his life, is worth less, from some impersonal and objective standpoint, than two, three, or even three hundred others. He also set a limit on this step: an arrangement that consistently gives less to one group cannot be described as giving equal weight to each person (Taurek 1977, 310–316).

This step lies beforehand. What is agreed beforehand is how risk is distributed; at that time no one is yet standing on the scale, and this stage can be reckoned (§4). But once the agreement reaches that moment and is carried out, what falls is an identifiable person. Here too this series parts from him: what the agreement can do is allow; it cannot make counting heads something the norm supports, nor does it produce a duty to count. The layer of rules is the same: at most it allows counting as a last resort, and does not support it. Once it is allowed, the person who counts according to it bears that act of counting, and the person who set the rule bears the rule he set; both are held to account; the rule written there cannot answer for them.

This series parts from Parfit and from Taurek because of one commitment only: a person has no equivalent, and a life is not offset against other lives (Prequel, §1). This commitment is not derived from anywhere else; it is the starting point of this series. It has a cost. The person present who counts gets no "this is right" from the norm to share his burden; whoever sets up a protocol that saves more people with the resources given can say only that it is allowed as a last resort, that it is questioned and borne, and does not get the layer "it is right"; the person who does not count does not get it either. At this point the norm gives only a floor and a direction, not an answer for the person present. The ground that Parfit and Taurek each gave, this series does not take over.

The place at the scene of emergency care is the "expectant" category in disaster triage. Many casualties arrive at once, there are not enough hands or drugs, and the triage protocol places in one category those whose hope of survival, given the resources at hand, is very small: they are given comfort and pain relief first, and hands and drugs go first to those with more hope. This category is set by resources: how it is drawn changes with the scale of the event, the resources at hand, and the training of the rescuers and their confidence in using the category; those placed in it are to be reassessed often, and looked at again as soon as resources arrive; placing someone in it does not settle the matter (Lerner et al. 2008, S31–S32). The objection runs like this: suppose there is reliable evidence that a protocol saves more people with the resources given; is there no ground at the level of the norm for making it the procedure everyone follows?

This is of the trolley-problem kind (Paper 1, §2). On what was said above, making it the procedure everyone follows is also only allowing it as a last resort; the protocol written there cannot stand in for that moment. The one who places this particular person in that category is the person doing triage on the spot; he judges, he bears it, and he is held to account afterward. Nor is he the only one held to account: why there were only so many hands and so many drugs that day is an account for those who allocated resources beforehand to answer, not something pushed onto the person present to carry alone.

The two objections land in the same place. The norm does not say counting is forbidden, but it cannot support counting heads; this is the floor; it has a direction: where counting can be avoided, there is no counting. A rule can at most allow it as a last resort: only when every other road is closed does it come into play. Whether other means were feasible is checked against the information available and the time at hand at that moment; a road that could have been found out earlier and was not cannot afterward be counted as ruled out with a word about "no time." If counting was done, that act belongs to the one who judged; he bears it, and is questioned afterward, and bears responsibility.

§6 Where It Is Handed Over

This series goes as far as the door. What lies beyond the door goes to three places.

Past the door, how a person treats others and how he acts are handed over to this framework's Action Theory (Prequel, §6). Action Theory starts from the far side of the door: that oneself is an end and others are ends too is, there, a premise, not something derived (Action Theory, Prequel). How not to legislate for others, how to do something together with others, how to protect those treated as means, and where judgment lies, have all been written there (Action Theory, Papers 6, 8 and 9, and Paper 0). The universal other of §2 is also continued there.

How institutions divide responsibility is handed over to the institutional side. How the various positions that set standards, write programs and put them on the road each bear their own part, how a company as a legal person pays compensation and fines, how those who allocate resources are questioned, how the slots each place leaves for reasons can leave more than a single slot for "what use is it?": all of this is there. The Political Economy series has written on accumulated power (Political Economy, Paper 0); dedicated papers on the institutional side will follow. This paper goes as far as on whom that judgment falls.

Striving for cultivation is handed over to the side that writes about what passes between people. Cultivation is leaving room and creating conditions for a person to grow by himself, not growing in his place (Prequel, §4). This series has taken only the half that is anti-colonization; the half that is cultivation is most visible between people. That side is against colonization too: between two people, once cultivation reaches into the other's share and grows in his place, it is no longer cultivation.

A person's own side, the institutional side, and the side that writes about what passes between people support one another, and none goes first (Prequel, §6). §4 of this paper is a place where several of them meet: who bears a rule once it is written into a program is said here; how the positions divide it belongs to the institutional side; how the one who writes it treats the people on the road whom he does not know belongs to Action Theory.

§7 Where This Paper Stops

This paper has written what the door looks like, not how to go through it. It does not answer for anyone whether to pull the switch, nor whether the person present is to count; it does not set how any particular car or triage protocol is to be written; the chief executive it mentions is a position, not any real person. How the various positions divide responsibility belongs to the institutional side; how one treats others past the door belongs to Action Theory; cultivation belongs to the side that writes about what passes between people.

This paper is also where the series stops. It set out from one question: is it worth it? Whether a thing is worth it can be asked; whether a person is worth it is a question that does not hold (Prequel, §1). In between it wrote about being for oneself, about reading structure and not reading minds, about how one's good comes to be written and how it comes to be called "worth it," and wrote one sentence: no longer asking for proof of worth first. Written this far, it has come to a door.

The door is there. Whether to go through it is each person's own judgment.

Propositions of This Paper

  • W40 Oneself being an end is the step before others being ends too: necessary, not sufficient. A person who thinks for himself has two roads: going on to any other person whatever being an end; or recognizing only his own group. The second can stand only by keeping those outside pressed down, does not last in an environment that leans toward the kingdom of ends, and is yet more dangerous than a person who harms others for himself alone.
  • W41 What the door looks like is the scale in one's own hand stopping before another person's share: not reading another's mind, not first demanding proof of worth from him, not legislating his law for him, not crossing his line. What has a price is weighed as before; the scale stops, the person has not walked away. The door is made of a few "nots," all of them stoppings on one's own side, not accomplishing anything for others; stopping also includes taking away a hand that has already sealed off another's road. These four "nots" are boundaries of this recognition that cannot be cancelled in action, and others who see them kept cannot certify on anyone's behalf that the recognition has been completed.
  • W42 The things at the door are done for any other person whatever, without first asking whether he is one of one's own group. The other person beyond the door is any other person whatever, not humanity as a whole, and not only those one cares about; any other person whatever has no equivalent either. These things are addressed to a person who can decide for himself; the matters he truly cannot decide for himself are for rules to set, and no rule decides a person's own affairs for him when he can decide them himself.
  • W43 The step at the door needs each person's own judgment, and judgment cannot be replaced. Rules can guide and test judgment, but the person judging still has to apply them, choose and take responsibility in the particular case; once judged, it is borne by whoever judged, and can be questioned. Supplanting upward is overstepping, and worst of all is supplanting the norm directly with a single judgment.
  • W44 When rules are written into programs, whoever decides on a deployment settles in advance a conditional procedure and lets it run in scenarios that have not yet arisen; this judgment of deployment is his to bear in full, and he cannot say "the program decided it" or "the standard told us to write it"; later checking, revision and halting of operation are further judgments. Beforehand, the remit is prevention; if counting heads is in it, the norm cannot support it, a rule can at most allow it as a last resort, and even when it is allowed, those responsible are still held to account: they can be questioned, and they bear responsibility; once it is in place, it remains under oversight and open to questioning. Those who follow it bear their own step, which does not thin out the share of the one who decided; whatever is said to be done by a whole is traced back to the person who made the decision.
  • W45 The strongest objections from the other side do not change this: the norm cannot support counting heads; this is the floor, and its direction is that where counting can be avoided, there is no counting; a rule can at most allow it as a last resort, and even when it is allowed, those responsible are still held to account. An agreement beforehand can distribute risk, and does not produce a duty to count; whether the other roads were open is checked against the information and time available at that moment. If counting was done, that act belongs to the one who judged, and is his to bear; however many more were saved, the one who was not saved is not wiped off the account. The expectant category is set by resources and reassessed as resources change; being held to account also asks those who allocated resources beforehand. This series parts from the other side because of one commitment only: a person has no equivalent; the cost is that at this point the norm gives only a floor and a direction, not an answer for the person present.

Acknowledgments

This paper is the seventh paper of the Second Enlightenment series of the SAE (Self-as-an-End) philosophical system, and is signed by Han Qin alone. All philosophical argument and all decisions between alternatives were finally adjudicated by the author, who bears sole responsibility for the final text.

The writing and review of this paper were assisted by Anthropic's Claude, OpenAI's ChatGPT, Google's Gemini, xAI's Grok and Meta's Muse. Thanks are due to these companies and to their research and engineering teams.


References

Group I lists works of this system, archived on Zenodo with Chinese and English versions under the same record, all single-authored (Han Qin); DOIs follow those listed at the head of each paper on the author's site. The text restates upstream work in plain words, with the title or the paper and the proposition number noted in parentheses at the place of use. The Prequel and Papers 1 to 7 in the text are the papers of this series. Kant's works are cited by volume and page of the Academy edition (AA), and the Critique of Pure Reason by the pages of the first (A) and second (B) editions; judgments are cited by paragraph number; the Ethics Commission's report is cited by rule number and by the section number of its explanatory part, and quoted in its official English edition.

I · Works of this system

Qin, H. (2026). On the Morphology of Power (SAE Power Theory, Paper 2) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.20455221

Qin, H. (2026). What Action Is Not (SAE Action Theory, Prequel) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22970747

Qin, H. (2026). Not Doing and Cannot-Not-Doing (SAE Action Theory, Paper 1) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22970752

Qin, H. (2026). Anti-Colonization (SAE Action Theory, Paper 2) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22970758

Qin, H. (2026). Not Legislating for Others (SAE Action Theory, Paper 6) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22970777

Qin, H. (2026). Finding One Another (SAE Action Theory, Paper 8) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22970783

Qin, H. (2026). One's Own Share of Strength, and Guarding (SAE Action Theory, Paper 9) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22970787

Qin, H. (2026). The Place of Judgment (SAE Action Theory, Paper 0) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.22974154

Qin, H. (2026). Political Economy and the Tradition: Accumulated Power (SAE Political Economy, Paper 0) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.23874158

Qin, H. (2026). A Way Out, a Way Forward (SAE Political Economy, Special Study) [Preprint]. Zenodo. https://doi.org/10.5281/zenodo.24021385

II · Other works

Bundesverfassungsgericht [Federal Constitutional Court]. (2006, February 15). Urteil des Ersten Senats, 1 BvR 357/05 (Luftsicherheitsgesetz). BVerfGE, 115, 118–166.

Ethik-Kommission Automatisiertes und Vernetztes Fahren [Ethics Commission on Automated and Connected Driving]. (2017). Bericht [Report]. Berlin: Bundesministerium für Verkehr und digitale Infrastruktur (published June 2017; official English edition of the same date). See also Deutscher Bundestag, Drucksache 18/13500 (the Federal Government's action plan, with the report as an annex).

Foot, P. (1967). The Problem of Abortion and the Doctrine of the Double Effect. Oxford Review, 5, 5–15.

Kant, I. (1781/1787). Kritik der reinen Vernunft. Riga: Hartknoch. (Cited by the pages of the first edition, A, and the second, B.)

Kant, I. (1785). Grundlegung zur Metaphysik der Sitten. Riga: Hartknoch. AA 4:385–463.

Kant, I. (1900–). Kant's gesammelte Schriften (Academy edition, cited in the text as AA). Berlin: Reimer / de Gruyter.

Lerner, E. B., Schwartz, R. B., Coule, P. L., Weinstein, E. S., Cone, D. C., Hunt, R. C., Sasser, S. M., Liu, J. M., Nudell, N. G., Wedmore, I. S., Hammond, J., Bulger, E. M., Salomone, J. P., Sanddal, T. L., Lord, G. C., Markenson, D., & O'Connor, R. E. (2008). Mass Casualty Triage: An Evaluation of the Data and Development of a Proposed National Guideline. Disaster Medicine and Public Health Preparedness, 2(Suppl 1), S25–S34.

Parfit, D. (1978). Innumerate Ethics. Philosophy & Public Affairs, 7(4), 285–301.

Schelling, T. C. (1968). The Life You Save May Be Your Own. In S. B. Chase (Ed.), Problems in Public Expenditure Analysis (pp. 127–162). Washington, DC: Brookings Institution.

Smith, A. (1759). The Theory of Moral Sentiments. London: A. Millar.

Smith, A. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations. London: W. Strahan & T. Cadell.

Taurek, J. M. (1977). Should the Numbers Count? Philosophy & Public Affairs, 6(4), 293–316.

Thomson, J. J. (1976). Killing, Letting Die, and the Trolley Problem. The Monist, 59(2), 204–217.