1. Intentions, Purposes and Plans of Action
Is it always clear when, or whether, the moral sphere has broken down and who is the guilty party when it does? These questions were not fully addressed in parts 2 and 3, where the examples of moral sphere breakdown were simple ones in which the guilty parties were easily identified (assailants, pirates, persecuting neighbors). Starting with simple cases before moving on to more complex ones is a common practice in ethical discussion.
But it is now time to move on to more complex issues and to develop the moral sphere theory of parts 2 and 3 in more detail. A good place to begin is by first addressing some questions that may have troubled you, even about the simpler cases of moral sphere breakdown discussed in parts 2 and 3. Addressing these questions will give us some clues about how to extend the discussion to more complex ethical issues.
In the example of the assault in the alley, for instance, someone playing the role of a devil’s advocate might object that it is merely a conflict between two different and competing ways of life. The assailant is thwarting the victim’s interests and purposes by force. But can we not also say the victim is thwarting the assailant’s purposes, if the victim forcibly resists the attack, as well he might? Let us call this “the symmetry objection.” It prompts us to ask questions like the following: On what grounds do we treat the parties differently in such cases, calling one guilty and the other innocent? On what grounds do we say one has broken the moral sphere and not the other?
These questions may seem hardly worth asking in cases like the assault, the pirates or the Gestapo. Isn’t it obvious, one might ask, who the guilty parties are in these cases? But the questions force us to think more carefully about the criteria for moral sphere breakdown in a general way that can be extended beyond the simplest cases. If we are justified in treating the parties differently in such conflicts, why is this so?
The first response that comes to mind in the assault case is a familiar one, though it turns out not to be quite adequate. It’s natural to think that the assailant is the one who broke the moral sphere because his actions initiated the conflict; and so he created the situation in which both parties could not be treated with openness. “He or she started it” is the schoolyard version of this response. We’re all familiar with that refrain from our school days, among siblings, and from raising children. The refrain is meant to identify the “guilty” party, the moral sphere breaker. And there is something right about it. But, while it points in the right direction, initiation of conflict cannot be the whole story in identifying the guilty party.
For one thing, initiation of conflict cannot be defined merely by asking who was the first to act in time or the first to act physically. Suppose the victim of the assault, suspecting the intentions of his assailant, had initiated physical contact by punching the assailant in the midsection and then running away, before the other man was able to make any move toward him. Though the intended victim acted first (“started it”) in this instance, we would not for that reason count him the guilty party, if his suspicions about the assailant’s intentions were well-founded.
What this shows is that we must attend to the intentions of agents as well as to their actions in order to adequately understand moral sphere breakdown. In the courtroom, as in everyday life, questions of intent enter the picture when trying to determine guilt or innocence, and for good reasons.
As we learn from philosophers of action, intentions embody our life-plans or plans of action of varying degrees of complexity expressed in our purposes or ends and the means by which we plan to carry out those purposes.[1] To focus on intentions therefore is to focus on the agents’ plans of action, of which their particular actions may be merely a part.[2] And I want to suggest that plans of action in general are the key to understanding moral sphere breakdown.
2. Moral Sphere-breaking Plans of Action
When the moral sphere breaks down, plans of action and ways of life intersect and conflict. To identify a guilty party, we have to find a difference in the plans of action of the parties that brought them into conflict. Suppose the plan of action of the man who was assaulted in the alley was to leave work that afternoon, do some shopping and go home to dinner. Nothing in this plan, or in his intentions or purposes generally that day, required or licensed him to impose his will on others, or to force others to do what he wanted, regardless of their desires or interests in the matter.
This was not the case for the assailant. His plan of action that day (“to assault and rob someone”) required him to force his will on some unsuspecting other person, whatever the desires, interests and purposes of that other person might be. So while both assailant and victim might have encountered obstacles and conflicts in the pursuit of their intentions or plans that day, there was a significant difference. The assailant’s plan of action was a “moral sphere-breaker” by the very nature of what he planned; his victim’s was not.
We might then say as a first step that moral sphere-breaking plans of action or ways of life are those that “require or license the agents acting on them to impose their wills on others, or make others do or undergo what the agents want, whatever the desires, interests, concerns or purposes of those others might be in the matter.”
The life-plans of Hitler’s Mein Kampf and of the pirates who invaded Philadelphia were moral sphere-breaking in this sense: Their plans required them to impose their wills on others whatever the desires or interests of the others might be. The man assaulted in the alley may in fact resist the assailant when the trajectories of their life-plans cross. But it is the nature of the assailant’s plan that caused the conflict and moral sphere breakdown, if his intention or plan was indeed to assault the victim. So the assailant would be the guilty party if that was his intention, even if the victim acted first to preempt the attack. The assailant’s guilt, however, would reside not only in the action, but in the intention and plan of action that led to it.[3]
But this condition (“requiring or licensing agents to impose their wills on others…”) is only part of what we need to account for moral sphere-breaking plans of action and ways of life. A second reqirement must be added to deal with the symmetry objection. Before their plans of action intersect, the assailant’s plan is moral sphere-breaking in the above sense, since it requires him to impose his will on some other person that day, whatever the desires of the other person might be. By contrast, the assault victim’s plan before they intersect is not moral-sphere-breaking, since he plans only to go shopping and return home to dinner.
However, after their plans intersect, it is another story. If the assault victim resists the attack—or if third parties, such as passers-by or the police, intervene to subdue the assailant—they are also acting on plans of action that “require or license the agents acting on them to impose their wills” on some other person (in this case, on the assailant) “whatever the desires, interests, concerns or purposes of that other person might be in the matter.” Something else must account for the asymmetry in the plans of the assailant and those who resist him, after their plans intersect.
The clue needed for what this further something might be can be found in the discussion of the retreat of part 3. By taking an attitude of openness respect toward other ways of life, the retreatants had chosen not to impose their wills on others except as a last resort, when all ways of life could not be treated with openness, no matter what they did. Putting this point in another way, we may say that the retreatants accepted the following constraint on their plans of action as a result of their argument from openness: Imposing one’s will on others is permissible only when by doing so one is doing what one can do to maintain (to restore and preserve) a moral sphere in which all persons can be treated with openness respect by all others when one must depart from that ideal to some degree, no matter what one does.
The retreatants accepted such a constraint on their behavior, whereas the persecuting group that left the retreat and later aggressed against its neighbors accepted no such constraint. Failing to abide by this constraint provides the further condition needed to define geuinely moral sphere-breaking plans of action and ways of life. Putting the two conditions together, we can say that
Moral sphere-breaking plans of action or ways of life are those that (i) require or license agents acting on them to impose their wills on others, or make others do or undergo what they want, whatever the desires, interests, concerns or purposes of those others might be in the matter (ii) in situations where the agents are not doing what they can do to maintain a moral sphere in which all persons can be treated with openness respect by all others when they must depart from the ideal of treating all persons with openness respect to some degree, no matter what they do.
The plans of action of the assailant and the persecuting neighbors satisfy both these conditions and are therefore moral-sphere-breaking plans of action. The assailant, for example, is not in a situation in which he cannot treat all persons with openness respect, no matter what he does. For he could allow his victim and others to pursue their plans of action that day without interference, by simply not pursuing his plan to assault and rob someone.
In acting as he does, the assailant thus is “not doing what he can do to maintain a moral sphere in which all persons can be treated with openness respect by all others.” So his plan satisfies condition (ii) of the definition. Since his plan (to assault and rob someone) also satisfies condition (i) (requiring him “to impose his will on another person, whatever the desires or purposes of that other might be”), it is a moral sphere-breaking plan of action. Likewise, the persecuting neighbors satisfy both conditions when they engage in unprovoked aggression against their neighbors.
This is not the case, by contrast, for third parties such as passers-by or police, who intervene to subdue the assailant with the purpose of restoring the moral sphere to the degree that they can. By acting for this purpose, they are “doing what they can do to maintain a moral sphere in which all persons involved can be treated with openness respect by all others” when that sphere has broken down due to the actions of the assailant. The plans of action of these third parties, thus fail to satisfy condition (ii) for moral sphere-breaking plans and are not moral sphere-breaking. (Rather, we might say they are moral sphere-“restoring.”) And this is so, even though their plans require the interveners “to impose their wills on another [the assailant] whatever the desires or purposes of the other might be in the matter”—and so satisfy condition (i) of the definition.
What saves the agents from moral sphere-breaking in such cases is that they are acting with the purpose of doing their best to sustain an ideal in which all persons are treated with openness respect by all others, even as they must depart from that ideal to some degree because of the actions of another.
3. Treating as Ends and as Means
The means by which agents may “impose their wills on others,” or “make others do or undergo what they want” can take many forms. Hence moral sphere-breaking plans of action can take many forms. Physical force and intimidation are two of the most common ways of imposing one’s will on others, but manipulation and deception are others.
Consider an elderly husband and wife who are persuaded by a con man to invest their life savings in a fraudulent scheme. They consent to invest their money and go along with the plan (though they would not have consented if they had known what the man’s plan really was). But even though the couple consents in this case, the con man’s life-plan is the “moral sphere-breaker” by the above account and he is the guilty party. For his plan requires him to “impose his will” on the victims, and “make them undergo what he wants” (to be defrauded of their life savings) “whatever their desires or interests may be in the matter.” By so acting, therefore, he is “not doing what he can do to maintain a moral sphere in which all persons can be treated with openness respect.”
In more familiar language, we can say that the con man’s intent in imposing his will in this way is to use his victims as a means to promote his own ends, regardless of the desires or interests of his victims.
This way of putting the matter is revealing. For this language of “means” and “ends” is familiar to those knowledgeable about the history of ethics from its prominent role in one of the formulations of Immanuel Kant’s well-known Categorical Imperative, the so-called Formula of Humanity. “Act so that you treat humanity, whether in your own person or in that of another, always at the same time as an end and never as a means only.”[4] Those who engage in moral sphere-breaking plans of action violate a similar principle as interpreted in the previous paragraph. They treat some other persons as mere means to their own ends, as does the assailant or the con man.
This language of persons treating others “as mere means to their own ends” is also common in everyday life, quite apart from its role in Kant’s well-known formula. So it is instructive to see how this language might be adapted to the moral sphere theory of parts 2 and 3. The results of doing so will not be the same as Kant’s Formula of Humanity, as we’ll see. But the differences from Kant’s formula, as well as the similarities, are revealing and instructive.
Kant’s idea is that humans are to be treated as “ends in themselves” and not used as “means only” because they are rational beings or persons capable of acting on principles that they legislate for themselves. As self-legislating or autonomous agents, they are capable of freely choosing their own ends and the principles on which they shall act. To treat their “humanity” as an end is therefore to respect their rational nature and their “capacity to determine [their own] ends through rational choice.”[5]
There are some obvious similarities between this Kantian idea of “treating persons as ends” and the idea of “treating persons with openness respect” in the moral sphere theory (“allowing persons to freely choose and pursue their plans of action and ways of life without interference from or subordination by others”). So it is tempting to think that “treating persons as ends” might simply be interpreted in the moral sphere theory as “treating persons with openness respect.”
But this would be a mistake for several reasons. First, Kant’s principle says that one should act so that one always “treats the humanity of all persons as ends.” And, as we have seen, it is not possible to treat all persons with openness respect at all times. When the moral sphere breaks down, it is not possible to act so that all persons are “allowed to pursue their chosen plans of action and ways of life without interference,” no matter what one does.
As a result, always treating all persons with openness respect is not the final goal or end of this moral sphere theory. Treating persons with openness respect is something we do only provisionally, up to the point where they would break the moral sphere. And the purpose of doing so is to place on persons themselves the responsibility of showing by how they plan to act and live in relation to others that they are worthy of being treated with such openness respect by other persons and by ourselves. Those who break the moral sphere show they are not so worthy. In other words, the idea is to treat persons with openness respect to the degree that we can, while at the same time trying to maintain to the extent possible a moral sphere in which all persons can be treated with similar openness respect by all others.
To place on persons themselves the responsibility of showing their worth in this manner by how they act and plan to act (and thus on the ends they choose to act upon) is what it would mean on this theory to “treat persons as ends,” that is, as beings capable of “freely choosing their own ends.”
Putting these points together, we get the following definition.
To treat persons as ends is (i) to treat them with openness respect (to act towards them in such manner that they are able to pursue their plans of action and ways of life without interference or subordination) (ii) to the degree that one can do this while maintaining to the extent possible a moral sphere in which all persons can be treated with openness respect by all others, and (iii) to act in this way in order to place on persons themselves the responsibility of showing by how they act and live and plan to act and live in relation to others that they are worthy of being treated with openness respect by all others, including oneself.
With this definition of “treating persons as ends” in hand, it is then natural to interpret “treating persons as means only” or as “mere means” in terms of “engaging in moral sphere-breaking plans of action” or “breaking the moral sphere.” Such an interpretation has the advantage of allowing us to make an important distinction, in the spirit of Kant, between “treating persons as means” (simply) and “treating them as means only” or “as a mere means.”
Treating persons as means simply would involve satisfying condition (i) of the definition of moral sphere-breaking plans of action of the previous section:
To treat persons as means is to pursue plans of action or ways of life that require or license one to impose one’s wills on others, or make others do or undergo what one wants, whatever the desires, interests, concerns or purposes of those others might be in the matter.
Treating persons “as means only” or “as mere means” would then involve satisfying both conditions (i) and (ii) of moral sphere-breaking plans and would be equivalent to “engaging in moral sphere-breaking plans of action” or “breaking the moral sphere”:
To treat persons as means only or as mere means (=to engage in moral sphere-breaking plans of action or to break the moral sphere) is (i) to treat persons as means in the above sense (ii) in situations where one is not doing what one can do to maintain a moral sphere in which all persons can be treated with openness respect by all others.
It might then be possible to treat persons as means (simply) without necessarily engaging in moral sphere-breaking plans of action. Such is the case of the police or the retreatants, when they attempt to thwart the plans of assailants or persecuting neighbors. They “impose their wills” on the assailant and persecutors they are thwarting and thus treat them as means. But they do not treat them as means only or as mere means, so long as by so acting, they are “doing what they can do to maintain to the degree possible a moral sphere in which all persons can be treated with openness respect by all others.” Their actions are moral-sphere restoring rather than moral-sphere breaking.
4. Kant and the Supposed Right to Lie for Altruistic Motives
I am not claiming these are the meanings Kant gives to “treating persons as ends” or “as means only” in his Formula of Humanity. Indeed, there are reasons for thinking the meanings are not Kant’s, despite similarities. The main similarity is that on both views one respects humans “as self-legislating agents capable of choosing their own ends and the principles on which they will act.”
But this sort of respect in the moral sphere theory (“openness respect”) is not unlimited. For its purpose is to allow persons to show, by the ends and principles they actually choose to act upon, whether they are ultimately worthy of being accorded such openness respect by others. Having a rational nature or being a self-legislating or autonomous agent alone is not enough to make persons worthy of such respect. It depends on how persons exercise their rational nature and autonomy—on what ends or purposes they choose to pursue and on what principles and plans they choose to act.
As a result, the requirement to “treat all persons as ends and no one as means only,” so interpreted in terms of the moral sphere, allows exceptions to common moral rules, as we’ve seen, “Don’t lie,” “Don’t kill,” “Don’t coerce,” and the like. When assailants, pirates, Gestapo, and persecuting neighbors, treat others as “mere means”—when they “engage in moral-sphere-breaking plans of action”—they make it impossible for others to allow them to pursue their ways of life without interference while allowing all others to pursue their ways of life as well. They thereby show themselves unworthy in this instance of the provisional openness respect accorded to them.
But note that, if interpreted in this way, the general requirement to “treat all persons as ends and no one as means only” would not itself have exceptions. It would be a universal requirement. The point is rather that this general principle would entail that common moral rules, such as “Don’t lie” and “Don’t coerce” sometimes have exceptions.
By contrast, whether Kant’s Formula of Humanity and other versions of his Categorical Imperative allow exceptions to such common moral rules is a much disputed issue in Kant scholarship. There are certainly places in Kant’s writings where he seems to be arguing that exceptions to such common moral rules are not allowed. This is most clear in his controversial essay “On the Supposed Right to Lie from Altruistic Motives.”[6] It is instructive to compare what Kant says in this essay about treating all persons as ends with the interpretation of that requirement given here.
In an insightful article on this essay of Kant’s on lying, the well-known Kant scholar, Christine Korsgaard (1998) concedes that it does seem to be Kant’s meaning in this essay that the Formula of Humanity allows no exceptions on lying. Kant introduces a famous example in the essay which is an earlier version of our Gestapo example. A man fleeing a murderer is hiding in your home, when the murderer comes to the door and asks whether the man is hiding there. Kant answers (to the dismay of Korsgaard and other sympathetic readers of Kant) that you are obligated not to lie (for “altruistic motives”) even in such a case. In defense of this conclusion, Kant says:
After you have honestly answered the murderer’s question about whether his intended victim is at home, it may be that he [the intended victim] has slipped out so that he does not come in the way of the murderer and thus that the murder may not be committed. But if you had lied and said he was not at home when he had really gone out without your knowing it, and if the murderer had then met him as he went away and murdered him, you might justly be accused as the cause of his death. For if you had told the truth as far as you knew it, perhaps the murderer might have been apprehended by the neighbors while he searched the house and thus the deed might have been prevented. (1976: 348)
As Korsgaard says, what Kant seems to be saying in this passage is that if you lie and the murder is committed anyway as a result of the lie because of a series of unlikely events, then you are partly responsible and share the blame due to your lie. But if you abide by the requirement never to lie even in this instance, things may turn out well after all by a series of coincidences (however unlikely). And if things do not turn out well, and the murder is committed, you have at least done your duty by telling the truth and so are not to blame for the bad consequences.
Korsgaard is deeply troubled by Kant’s reasoning in this essay and she thinks his view must be modified in some way to avoid his conclusion. She wisely comments that, on Kant’s approach in the essay,
Your share of the way the world is is well-defined and limited, and if you act as you ought, bad outcomes are not your responsibility. The trouble is that in cases such as that of the murderer at the door it seems grotesque simply to say that I have done my part by telling the truth and the bad results are not my responsibility. (p. 300)
This feature of Kant’s moral theory has troubled many readers of Kant, as it troubles Korsgaard. Defenders of Kant have tried to respond to it in various ways. It is hard to believe that you are not in any way responsible for the outcome, if you tell the truth to the murderer at the door and the most likely result (the murder) does in fact take place. Yet, Korsgaard concedes that Kant’s Formula of Humanity can lead to this conclusion because, as she says, “the formula of humanity gives us reason to believe that all lies are wrong.” (p. 293)
Why is this the case? Well, the formula says that you should “Act so that you treat humanity, whether in your own person or that of another, always as an end and never as a means only.” And, as Korsgaard notes, to treat the “humanity” of persons as an end is to respect their capacity to freely choose their own ends. When you lie to others or coerce them in order to realize some end or purpose of yours, you make it impossible for them to freely choose whether or not to “assent to” and go along with your purposes.
If the husband and wife who were conned into investing in a fraudulent scheme had known of the con man’s purpose, they would not have assented. But he deceived them, thereby taking away their freedom to choose whether or not to go along with his true purpose. He thereby treated their humanity (their capacity to determine their own ends through rational choice) as mere means to his ends. And, as Korsgaard notes, the Formula of Humanity seems to say that you can never do this: “Act so that you treat humanity, whether in your own person or that of another, always as an end and never as a means only.” And this seems to imply that one must tell the truth to the murderer as well, or one will not be treating his humanity “as an end” but “as a means only.”
What has gone wrong in this reasoning may, I believe, be explained as follows. Korsgaard and other contemporary Kantians plausibly argue that the basic theory of value underlying Kant’s Formula of Humanity (and perhaps all Kant’s formulations of the Categorical Imperative) is, as she puts it, that “your rational nature is the source of justifying power of your reasons and so of the goodness of your ends.”(p. 293) But, even if we grant this claim (as many of Kant’s critics would not), so that every good end is so because it is a chosen end of some rational being, the converse would not follow (as Korsgaard and other Kantians would surely agree). That is, it would not follow that every chosen end of a rational agent is thereby a good end.
Rational nature and free choice may indeed be good things in general, as Kant holds. But the problem is that persons can exercise their rational nature and free choice to choose bad ends as well as good ends. So it does not follow (even if we were to grant that rational nature is “the source of the goodness of ends”), that we must treat persons as ends in themselves by virtue of their rational nature and capacity for choice irrespective of the plans of action and purposes they actually choose to pursue by employing their rational nature.
To tell the truth to the murderer at the door out of respect for his humanity or rational nature, even though he has used that rational nature to choose an evil end, is grotesque, as Korsgaard says. It is, one might say, a kind of “idolatry of reason.”
5. The Ends Principle: Ideal and Non-ideal Theory
Korsgaard is unhappy with this feature of Kant’s view for good reasons. She therefore suggests that Kant’s view must be modified somehow by adding special principles for “dealing with evil” (p. 282). To do this, she suggests we might appeal to a distinction, borrowed from John Rawls, between ideal theory and non-ideal theory in ethics. Kant’s Formula of Humanity would tells us how to behave in an ideal world in which no one is choosing evil ends. But in the real world where people do sometimes choose evil ends, other principles would be needed —a non-ideal theory—to tell us how to behave.
Korsgaard does not attempt to give a systematic account of what such a non-ideal theory might look like beyond suggesting the need for it to supplement Kant’s theory. But I do think she is in the right track in suggesting it. Though I also think that pursuing this suggestion would take us farther away from a distinctively Kantian ethics than she and other Kantians might like.
For note that just such a distinction between ideal and non-ideal theory in ethics follows naturally from the moral sphere theory defended here. We can see this by substituting for Kant’s Formula of Humanity an alternative principle that makes use of the definitions of “treating persons as ends” and “as means only” spelled out above in section 3. The resulting principle would not be the same as Kant’s Formula of Humanity, though it would make use of similar language. To avoid confusion, therefore, I will give this alternative principle its own name and will call it hereafter the
Ends Principle (EP): “Treat all persons as ends in every situation, and no one as means only (or as mere means).”
The meaning of this principle would then be spelled out in terms of the definitions of treating “as ends” and “as means only” of section 3, which I repeat here for convenience.
To treat persons as ends is (i) to treat them with openness respect (to act towards them in such manner that they are able to pursue their plans of action and ways of life without interference or subordination) (ii) to the degree that one can do so while maintaining to the extent possible a moral sphere in which all persons can be treated with openness respect by all others, and (iii) to act in this way in order to place on persons themselves the responsibility of showing by how they plan to act and live in relation to others that they are worthy of being treated with openness respect by all others, including oneself.
To treat persons as means only or as mere means (=to engage in moral sphere-breaking plans of action or to break the moral sphere) is (i) to pursue plans of action or ways of life that require or license one to impose one’s wills on others, or make others do or undergo what one wants, whatever the desires, interests, concerns or purposes of those others might be in the matter (ii) in situations where one is not doing what one can do to maintain a moral sphere in which all persons can be treated with openness respect by all others.
When “treating as ends” and “as means only” are thus interpreted, the above Ends Principle then allows naturally for a distinction between ideal and non-ideal ethical theory along the following lines:
Ideal Theory applies when the moral sphere obtains (“inside” the moral sphere) and would require that we “treat all persons with openness respect” (allowing them to pursue their ways of life without interference or subordination) and therefore that we not “break that moral sphere” which does obtain by “engaging in moral sphere-breaking plans of action.”
Non-ideal Theory applies when the moral sphere has broken down and requires that we do what we can to maintain a moral sphere in which all persons can be treated with openness respect by all others to the degree possible, restoring it when it has broken down and departing as little as possible from the ideal of treating all persons with openness respect, when we must depart from that ideal to some degree, no matter what we do.
When viewed in this way, the Ends Principle (“Treat all persons as ends in every situation and no one as means only”) would be an analogue of (though only an analogue of) Kant’s Formula of Humanity. And the moral sphere (the sphere in which everyone can treat everyone else with openness respect) would be an analogue (though only an analogue) of what Kant calls an ideal Kingdom of Ends and which he describes as “a systematic union of different rational beings under common laws.”
Moreover, it would be by making efforts to maintain this ideal moral sphere to the degree possible when it has broken down that we would learn how to act in non-ideal circumstances. Principles of non-ideal theory would require us to depart as little as possible from conditions in which all persons can treated with openness respect by all others when we must depart from such conditions to some degree.[7]
But what might the specific ethical requirements of non-ideal theory be, if viewed in this way? In the next section, we consider that question.
[1] As emphasized by recent philosophers of action such as Michael Bratman (1987) and Alfred Mele (1992).
[2] Viewed from this perspective, different ways of life or forms of life—notions that played a crucial role in the argument of parts 2 and 3—may be viewed as the most comprehensive life-plans or plans of action
[3] Of course, the “if” here is an important one. Suppose the man we’ve been calling the “victim” got it wrong when he initiated the conflict by punching the other man in the stomach. The other man we’ve been calling the assailant was in fact just innocently walking home through the alley and had no plan to assault and rob anyone.
[4] Kant 1959: 47.
[5] Korsgaard, p. 287
[6] Kant 1976.
[7] Korsgaard makes a similar suggestion (1998: 297) and again I believe she is right.