1 The controversy between deontologists and consequentialists rexamined

The antagonism between deontological and consequentialist approaches to ethics is one of the most persistent antagonisms in contemporary ethics. Indeed, the difference between deontological and consequentialist approaches might seem unbridgeable: Ever since Elizabeth Anscombe coined the term “consequentialism” in the 1950s (Anscombe 1958: 11), there has been a controversial debate about what exactly the crucial difference between “consequentialist” and “deontological” ethics is. Nevertheless, there seems to be sufficient agreement that in deontological ethics, the moral judgement of an action is based on intrinsic characteristics of the action itself, i.e. every action has some intrinsic characteristics that make the action morally right or wrong, regardless of its consequences. In contrast, consequentialism is usually thought to be characterised by the idea that the moral judgement of an action is based exclusively on the consequences, the action has for the realisation of certain goods, which are usually arranged in an evaluative hierarchy (e.g. life over property). If one takes the difference between deontological and consequentialist ethics on this theoretical level, then indeed it seems impossible to overcome the dichotomy: Evaluating an action solely on the basis of its consequences is obviously incompatible with evaluating it solely on the basis of its intrinsic characteristics. Within this descriptive framework, it seems indeed impossible to reconcile both positions.

Yet, whenever such a seemingly irreconcilable dichotomy arises within philosophy, it is usually helpful to ask oneself whether one is approaching the problem from the right angle and with the right conceptual framework. If one does so, it can sometimes be shown that a seemingly irreconcilable dichotomy is not a dichotomy at all if one approaches it from a different angle or describes it with a different conceptual framework. If we do this here, we can first of all see that the controversy between deontology and consequentialism is obviously motivated by a very specific question: namely, whether there are absolute, unconditionally valid negative duties that cannot be overridden by the fact that observing such a duty has a worse overall outcome than violating it. Or, to put it in other words: whether there are actions that are always and in every case forbidden to any rational agent.

This assumption is shown not least by the fact that Elizabeth Anscombe introduced the term “consequentialism” primarily to characterise an approach to ethics that rejects the very possibility of prohibitions, that have to be observed regardless of the consequences of observing them (Anscombe 1958: 8). Likewise, most, if not all classic contributions to the controversy precisely evolve around the problem of absolute negative duties, such as Samuel Scheffler’s “The Rejection of Consequentialism” (Scheffler 1994), Bernard Williams’ contribution to “Utilitarianism and is Critics” (Williams 1973), Thomas Nagel in “The View from Nowhere” (Nagel 1986: 164-185), Shelly Kagan’s “The Limits of Morality” (Kagan 1989), Francis M. Kamm’s “Intricate Ethics” (Kamm 2007) or Robert Nozick’s reflections on deontological “side-constraints” in “Anarchy, the State and Utopia” (Nozick 1974: 29–35).

Introducing the distinction between “agent-relativity” and “agent-neutrality” (Nagel 1986: 164-185) as a complementary characteristic to differentiate between deontological and consequentialist ethics only confirms this finding. It only became necessary because it turned out that unconditional negative duties cannot be based solely on the idea that intrinsic characteristics of actions form the basis for morally judging that action. For if one were to assume this alone, it would be impossible to explain why an action, by which an agent realises one instance of a certain kind of morally wrong action, should not be allowed if he could prevents several instances of the same kind of action to be carried out by other agents.Footnote 1 To illustrate this with an example based on Williams’ thought experiment of “Jim and the Indios” (Williams 1973: 98): Why should it be forbidden for an agent A to kill an innocent person C, if this murder can prevent agent B from killing five other innocent persons D, E, F, G and H? “Agent-relativity” tries to give an answer to this question: It simply means that a morally wrong action is not simply to be classified as morally wrong, regardless of who commits it, but that each individual agent has a moral duty not to commit it. The adequate linguistic expression of, for example, the prohibition of murder would therefore not be “murder is morally wrong and therefore no one should be murdered”, but “thou shalt not murder”. Thus, deontological ethics would not only be distinguished from consequentialist ethics by saying that some intrinsic characteristics of the action are the basis for its normative evaluation, but also by saying that normativity in deontological ethics has the form of agent-relative duties. This introduction of agent-relativity as an additional way to characterize deontological ethics only makes sense, if what “deontologists” actually want to defend, is the idea that there are unconditional or absolute duties that everyone has to observe.

If one conceptualizes the point of contention behind the dichotomy of deontology and consequentialism in this way, it becomes apparent that both, consequentialism and deontology, each correspond to some widespread moral intuitions: Consequentialism obviously corresponds to the widespread moral intuition that most rights and duties can be weighed against each other with regard to the goods affected by an action. To give just one well-known example: If I can only save the life of a potential victim of a murderer by lying to the murderer, lying seems to be an appropriate and morally right, if not morally required action. It is even more evident in cases, in which I can only save the life of person A by stealing or destroying the property of person B. The idea that this kind of actions is justified obviously results from a weighing of the goods of life and property against each other. On the other hand, however, there is a moral intuition that is at least as widespread: the intuition that there are certain actions that a rational agent is always and in any possible circumstances forbidden to carry out, even if they would have a better outcome than an alternative action or the omission of the action. Actions that are often mentioned in this context include e.g. the killing of people outside of self-defence, torture, the infliction of extreme pain or deliberately unjust punishment. This intuition seems to be the actual motivation of deontological ethics.

What links the standard definitions of deontological and consequentialist ethics to the question of the possibility of absolute duties, is rather obvious: absolute duties are only possible if the moral evaluation of an action does not depend on variables that can be different for the same type of action under different circumstances. The result of a weighing of the consequences of an action against each other, however, depends on what consequences this action has in the individual case, and this in turn usually depends on the circumstances of the action. Deontological theories, as they are conceived by the standard definition, therefore do not necessarily have to assume the existence of absolute negative duties, but they can at least do so. Consequentialist theories, on the other hand, cannot assume the existence of such duties from the very outset.

If one assumes that the controversy between “deontologists” and “consequentialists” is actually motivated by a struggle between the two moral intuitions mentioned above, a possible reconciliation of these two prevalent models of ethics becomes conceivable. Of course, it still seems to be impossible to reconcile deontological and consequentialist ethics on the level of their usual theoretical conceptualization, i.e. as an approach that bases moral judgement on intrinsic characteristics of actions vs. an approach that bases moral judgement on the consequences of actions. But if these contrary definitions are only an attempt to conceptualize two equally strong, but different moral intuitions that might also be conceptualized in other ways, a theory that satisfies both intuitions seem to be at least possible. Such a theory would have to be able to distinguish between actions, for which a consequentialist weighing of goods and evils is legitimate and actions, for which such a weighing is not legitimate. In other words, it would have to be able to give reasons why certain acts–such as the prohibition of killing innocent people or the prohibition of torture–are impermissible even if a mere weighing of goods and interests would speak in favour of these actions. At the same time, it would have to explain why in other cases a violation of duties or rights can be legitimate or even required on the basis of a weighing of goods and evils. In this article, I want to propose such a theory.

To this end, I will first examine the most promising theoretical rationale of deontological ethics–a freedom-based approach–with the goal of determining whether it can integrate both moral intuitions as described above. The answer to this question will be negative, but it will open up the perspective of an alternative approach to the problem. This alternative approach has the character of a theory of intersubjective justification. After having sketched this theory, I want to show that and in what way a justification-theoretical approach to ethics allows the moral intuitions underlying consequentialist ethics on the one hand and deontological ethics on the other hand to be reconciled. This will result in a criterion to distinguish the realm of a legitimate application of consequentialism from a realm, in which the use of a consequentialist weighing of goods and evils is inappropriate due to the existence of absolute negative duties. In the last chapter before my conclusions, I will further expand on this criterion by discussing some examples.

2 A freedom-theoretical approach to deontological ethics

One promising approach that is able to theoretically catch up with the idea that certain acts are always and in every case impermissible is an approach based on freedom and the right to self-determination. According to this approach, the purpose of law is to distribute claims to freedom in a just and fair manner, which then accrue to individuals in the form of subjective rights. Such a consideration can be found, for example, in Kant’s so-called “principle of right” (Kant 1907 [1797]: 230) and in Rawls’ reformulation of the Kantian principle in his “theory of justice” (Rawls 1999: 266), but also in Fichte’s philosophy of right (Fichte 2000 [1796]). Regardless of the different ways to ground this ideas, all these approaches have in common that they do not understand subjective rights as claims to goods, but as claims to non-interference into one’s own legitimate sphere of freedom. The “right to life” would thus not be a right to ensure that the good “life” is not harmed and/or promoted, but a right to exclude other agents from disposing of one’s life. Similarly, the object of the right to bodily integrity would not be a good like freedom from pain or like the integrity of the body, etc., but the claim that nobody else but the holder of the right may decide what to do with his or her own body. Thus, according to the freedom-theoretical approach, it is not goods that are the object of rights and duties, but the indisposability of one’s own legitimate sphere of freedom with all the goods “therein” for others. The values of the various goods for the respective persons is no longer normatively relevant in such a concept; the only thing that counts for judging an action is whether or not the action interferes with the legitimate sphere of freedom of another person.

This is a fundamental departure from theory that is centered around goods and their values. It is replaced by a categorically different perspective: that of the intersubjective relationship between free and rational agents. The object of duties or rights are no longer goods, but individual spheres of freedom and thus claims to deny everybody else to dispose of the goods within them. Such rights can only be understood as specific kinds of relations between free and rational agents. Their perspective is what Darwall (2009) calls the “second-person standpoint”. What makes (ethical and legal) norms necessary here is the fact that the actions of one such an agent could interfere with the legitimate sphere of freedom of another such agent. Both, subjective rights and objective law, thus only become necessary in the first place because free and rational agents have the de facto ability to deliberately violate the legitimate spheres of freedom of other free and rational agents. It is evident that in such an approach, the difference between negative and positive duties is crucial, whereas it is irrelevant for a theory focused on goods and their value. The simple reason for this is that in standard-cases only doing something can violate the legitimate sphere of freedom of another agent, but not an omission.Footnote 2 If, however, the “meta-norm” of all subjective rights is to refrain from interfering with the legitimate sphere of freedom of other agents, negative duties always prevail over positive duties, as found as a principle in most deontological approaches to ethics.

This, however, implies that the violation of a negative duty would be impermissible even in those cases in which only a minor violation of a negative duty towards an agent A would be necessary to prevent serious impairments goods in one or more other agents. The reason for this is, that according to the freedom-theoretical approach an action is morally wrong und hence impermissible if and only if it violates the legitimate sphere of freedom of another person. Goods and interests, on the other side, are irrelevant in such an approach. Therefore, any action that violates the sphere of freedom of another person is morally wrong and impermissible. This, however, seems just as counterintuitive as the assumption of an unrestricted trade-off of all goods against each other. Likewise, a purely freedom-theoretical approach appears problematic insofar as it must systematically ignore the value that goods indisputably have for people. Asked what the meaning of the right to life is for them, most people in most situations would probably point out that life is an important, if not the most important good for them. A theory that systematically ignores this seems at least one-sided and in some ways incomplete. The question therefore still arises as to what a theory could be like that would allow both perspectives to be brought together.

3 The justification-theoretical approach

The results of the previous chapters can be summarised as follows: Ethical theorising with regard to the dichotomy between deontology and consequentialism, seems to face the dilemma of either having to give up the assumption that there are inviolable rights and/or duties, or the assumption that at least some violations of rights and/or duties can be justified. This dilemma can only be overcome in a theory that is able to give reasons why certain acts–such as the prohibition of killing innocent people or the prohibition of torture–are impermissible even if a mere weighing of goods would speak in favour of them. At the same time, however, it would also have to be able to explain why in other cases, a violation of negative duties can be permissible or even required on the basis of consequentialist considerations.

The best way to conceive of such a theory starts from the freedom-theoretical approach, but modifies it in a significant point. As we have seen, the freedom-theoretical approach is characterised by the idea that the recognition of other human beings as persons immediately leads to their recognition as bearers of an inviolable right to freedom and self-determination that grounds all other subjective rights. This, however, seems to skip a level of recognition that is even more fundamental than the recognition as a being entitled to an inviolable sphere of freedom. This more fundamental level would be about recognizing each other as beings who have a claim to be subjected to only those actions that are justifiable to them by good reasons. This idea of justification plays a central role in various contemporary theories, like the ones of Rainer Forst (2012) and Thomas Scanlon (2000). Forst is primarily interested in laying the foundation for a political philosophy of justice. In this context, the “right to justification” forms a meta-norm for any theory of justice. The contents of ethics and law then only emerge via political discourses based on the “right to justification”. The approach, I want to advocate here, is therefore closer to Scanlon’s “contractualism”, according to which an action is morally permissible if and only if none of the persons concerned by the action can “reasonably reject” the action and/or the principle of the action (Scanlon 2000: 189-202). The problem of Scanlon’s theory, however, is that it ultimately fails to give a plausible criterion to distinguish between “reasonable” and “unreasonable” rejections (Pogge 2001). To do this, we would need to extend Scanlon’s contractualism to include a theory that allows the reasonableness of objections to be assessed in terms of necessary characteristics of embodied human subjectivity and its relation to goods and freedoms. I will outline part of such a theory in the following chapters.

Against this backdrop, I think that the approach in general gives us a good basis to tackle the dichotomy between deontological and consequentialist ethics. Like the freedom-theoretical approach, a justification-theoretical approach starts from the relationship between persons instead of starting from the relationship between persons and goods. Also, it is more fundamental and at the same time theoretically less demanding than freedom-theoretical approaches à la Kant, Rawls or Fichte: It is more fundamental in that it asks what the claim to be recognised as a bearer of subjective rights actually results from. But it is theoretically less demanding in the sense that it does not need to presuppose a strong metaphysical or transcendental concept of freedom. Nor is it committed to a particular theory of what is and is not normatively relevant in the relationship between persons. It can therefore take into account the relationship that persons have to the goods they seek, just as it can take into account aspects of freedom and justice. The justification-theoretical approach therefore might allow–under certain circumstances–to think of encroachments on the legitimate sphere of freedom of another person as ethically (or also legally) permissible. Indeed, they would be permissible precisely if they could be justified to the person concerned in a way that he or she could only reject if, for example, he or she rejected the principle of recognition itself or general principles of justice resulting from it.

In order to determine, in which cases such a justification is possible, we have to examine possible logics the justification. Since the basic intuition behind deontological approaches seems to be that certain actions are always and under all conditions impermissible, I will first examine whether the logics of possible justifications in doing differ from the logics of possible justification in in omissions.

4 Logics of justification

Justifying an action means responding to an actual or possible objection that one of those affected by the action could raise against it. Where no objection at all can be raised against the action, no justification is required. Justifications, in turn, can only consist in giving reasons for action, which usually refer to the intentions of the action. If these intentions are such that they are incompatible with the recognition of other persons as beings who have their own interests, purposes and their own ideas of a successful life, these the reasons are not reasons by which his or her decision can be justified. Justification then fails.

The constellation that I would like to examine in the following with regard to the underlying logics of justification is the constellation that an agent A would have to actively and intentionally inflict a grave evil E1 on one or more persons, i.e. violate a negative duty towards A, in order to prevent a grave evil E2 from happening to one or more other persons. The danger of E2 occurring can emanate both from processes that are not themselves acts, such as illness or the derailment of a train, as it can also emanate from the actions of third parties. It is precisely this kind of cases that is usually at the centre of controversy between deontologists and consequentialists. One reason for this is that the difference between the two approaches becomes clearest here: Since the goodness of an action in consequentialism depends only on what consequences actions have for the realisation or non-realisation of goods, it is not relevant for consequentialism whether the realisation or non-realisation was brought about by an action or an omission. Conversely, as seen above, deontological ethics is characterised by the assumption that certain forms of action are always and in every case forbidden, i.e. that at least some negative duties such as the prohibition of torture or the prohibition of killing innocent people are absolute.

For the sake of simplicity, I would now like to consider an abstracted constellation of actions in which only three persons are involved: Miller; Smith, to whom an evil E1 would have to be actively inflicted in order to prevent a more serious evil E2 from happening to Jones; and finally Jones, to whom evil E2 would happen, if Miller did not inflict evil E1 on Smith.Footnote 3 Miller obviously has two options in this constellation of cases:

  1. 1.

    Miller refrains from inflicting evil E1 on Smith. As a consequence, E2 happens to Jones.

  2. 2.

    Miller inflicts evil E1 on Smith and thus prevents E2 from happening to Jones.

Let us consider what objections Smith and what objections Jones could raise against Miller’s action in each case and, accordingly, what Miller would have to justify in each case: In the first case, Jones can accuse Miller of allowing a serious evil happening to him. In the second case, Smith can accuse Miller of having inflicted a serious evil upon him. Now, from a purely consequentialist point of view, the only relevant factor to decide, which option Miller should take, would be which of the two evils is greater, i.e. Miller would have to choose the action that leads to the realisation of the lesser of the two evils. Thus, the justification for this choice vis-à-vis Smith would be precisely that the evil that befalls Smith is lesser than the evil that would befall the Jones if Miller had chosen the other option.

However, already the language we use to describe these options shows that something else is at play here and that what is at play, is relevant from the point of view of a theory of justification. In the first case, Jones suffers an evil “because of” Miller’s omission. This implies that the evil that Jones suffers has its cause in a process that could be influenced by Miller, but otherwise leads to that suffering via some already on-going causal chain of events. In the second case, however, Miller actively inflicts an evil on another person. If we include this difference in the logic of justification, it becomes apparent that a very relevant point is obscured by recourse to a mere weighing of goods and evils. In case (1), Jones can indeed accuse Miller of suffering an evil “because of” Miller’s decision, and Smith can accuse Miller of the same in case (2). In this respect, the two cases are indeed symmetrical. In case (1), however, Smith can accuse Miller of something more than just the fact that he suffers an evil. Smith can accuse Miller of including him in a strategic calculation of weighing goods and evils in the first place. Jones cannot raise this objection against Miller, since the evil that befalls Jones does not have its cause in the fact that Miller employs such a strategic calculation.

If Miller did not exist, could not intervene, or did not employ such a strategic calculation, then evil E2 that befalls Jones would have literally nothing to do with the possible evil E1 that Miller could inflict upon Smith. The suffering of an evil by Jones has its cause not in the fact that Miller uses a strategic calculation of weighing goods and evils to decide what to do. Rather, it has its cause in an already on-going causal chain of events, e.g. a disease, which takes place independently of Miller’s current actions.Footnote 4 Thus, an evil would befall Jones even, if there were no Miller present at this moment to weigh goods and evils. In contrast, Smith only suffers an evil if Miller exists, includes Smith in a strategic calculation of weighing goods and evils, and acts on the basis of this calculation. It is Miller’s weighing goods and evils alone that establishes a connection between evil E2 that would happen to Jones, if Miller wouldn’t act and evil E1 that Miller could inflict on Smith in order to prevent E2 from happening to Jones. Thus, if Miller inflicts E1 on Smith, he must not only justify the mere result of his weighing of goods and evils to Smith. Rather, he must also justify the act of including Smith in a calculation of weighing goods and evils in the first place. This means, that within the logics of justification, there is always and inevitably a fundamental asymmetry between the violation of positive duties and the violation of negative duties. In case (2), Miller therefore has to justify something to Smith that he does not have to justify to Jones in case (1), namely the inclusion of Smith in a strategic calculation of weighing goods and evils.

5 External vs. internal goods

This does not already imply that such a justification is impossible. But we have to ask, under which conditions it may eventually succeed and under which conditions it cannot. In order to answer this question, we have to take a closer look at the very nature of “weighing goods and evils”. In doing so, we notice that “weighing goods and evils” introduces a moment of external comparison between goods and evils–usually in different persons–into Miller’s reasons for actions, and thus into the reasons, which can serve as justifications for his actions. For some goods, however, it is precisely this external comparison that contradicts the way we usually understand our relationship to these goods. Let’s take the good of life as an example. For the person in question, life is not a good to which he or she stands in an external relationship that would be accessible to comparison and offsetting. A person does not “have” life, in the way in which we possess something, and is not related to life as something distinct from him or her. Rather, her life is nothing other than her very existence and insofar not conceivable as separate from her. Likewise, bodily or mental integrity are nothing that could be distinguished from a person in such a way that she would stand in an external relationship to them. I myself am nothing else than my life, my body and my psyche. A violation of bodily integrity–for example through the loss of an extremity or through the infliction of extreme pain–or of psychological integrity–for example through torture or mobbing–is therefore fundamentally different from a violation of a right or a duty, that relates to an external good which can be distinguished from me as a person. In contrast, the paradigmatic case of a right that relates to goods that can be distinguished from myself, my life, my body and my psyche as being external is the right to property, as far as it concerns external objects, but also, for example, the right to freedom of movement or the freedom of assembly. It is this externality that makes them offsettable against other goods or the same goods in other persons. Goods that are indistinguishable from the person seeking them, on the other hand, they are not commensurable and offsettable with the same goods in other persons in the way the consequentialist weighing of goods and evils requires.

Hence, if Miller bases his reasons for actions on such a weighing of goods and evils, and if the goods in question are inseparable from Smith as a person, Smith cannot only reproach Miller for having inflicted an evil on him, but he can additionally reproach Miller for the fact that the reason for this action is a consideration that is categorically incompatible with the special character of these goods. As we have seen, if Miller inflicts an evil upon Smith in general, he would not only have to justify his inflicting the evil, but also the fact that he includes something, that is a good for Smith, into a strategic calculus of weighing goods and evils. Now, it seems, that there are certain goods, to which this kind of calculus is fundamentally inappropriate, because they are goods that cannot be subdued to external comparisons, since they coincide with the person, for whom they are goods. If such goods exist, it follows that any justification for an action impairing or destroying such a good, is impossible, because the only justification, the agent could give, would include a kind of comparison that is categorically incompatible with the very nature of that good.

Now, one could object that all of this also applies to Jones, if E2 is of the kind that impairs or destroys Jones’ life, bodily and/or mental integrity, maybe in an even more severe way than E1 impairs these goods in Smith. Jones’ own life, bodily and mental integrity are just as indistinguishable from him as person, as Smith’s life, bodily and mental integrity are indistinguishable from Smith. In both cases, the goods coincide with the person’s existence as a bodily-mental being. This is true, but it does not provide an argument for employing a strategic calculus of weighing of goods and evils to justify Miller’s eventual infliction of E1 upon Smith. For what Miller cannot justify in the case of inflicting E1 upon Smith actions is not simply that E1 affects an intrinsic good. Rather, what Miller cannot justify is the fact that the reason for his action is a comparative calculus, that is incompatible with the specific character of non-external goods, simply because it is comparative. Since E2 happens to Smith independently from Miller’s reasons for actions, the same is not true, if Miller only omits to avert E2 from Jones.

This allows for a further clarification of my argument for absolute rights and/or duties: My argument is not solely based on the fact that the violation of these rights and/or duties is based on a strategic calculus of weighing goods and evils, which has to be justified to the person(s) concerned in addition the fact that those violations impair any goods. In most cases, it is still possible to justify the employment of such a strategic calculus. Justification is only impossible, if a second element comes into play, namely, if the goods that are impaired on the grounds of a consequentialist calculus are of such nature, that they are in way internal to the person(s) concerned that is incompatible with an external weighing of goods and evils. Both elements must come together in order to explain why some rights and/or duties are absolute, whereas most are not. If we would only refer to the non-externality of certain goods, we would end up in a stalemate, because both, Smith and Jones, could appeal to that non-externality, regardless of whether the good in question is impaired by an action or an omission from Miller’s side. If we would only refer to the fact, that Miller has to justify something more to Smith than to Jones, namely that he includes Smith into a strategic calculus of weighing goods and evils, we would not have a criterion that determines, in which cases the employment of such a calculus can be justified and in which cases it cannot.

The fact that both elements must come together, also implies that my argument is not simply identical to a prohibition of treating Smith as a “mere means to an end” or of a claim to respect Smith’s right to self-determination. This becomes evident if we compare my proposal with an argument, made by Warren S. Quinn (1989), that might look similar at first glance. For Quinn too, the morally relevant difference between action and omission in cases like the ones examined here relates to the fact that the agent subjects his victims to his strategy of realizing a certain goal:

In these cases, but not in their indirect counterparts, the victims are made to play a role in the service of the agent’s goal that is not (or may not be) morally required of them. And this aspect of direct agency adds its own negative moral force‑a force over and above that provided by the fact of harming or failing to prevent harm. (Quinn 1989: 349 f.)

Quinn’s argument is obviously inspired by Kant, but in the form we find it in Quinn, it still has a number of open questions. On the one hand, Quinn’s reasoning for why it is morally impermissible to instrumentalize persons in such a way remains relatively vague. Quinn merely states en passant that such action is problematic because it overrides the will of the victim and does not leave her the choice of whether or not to sacrifice herself in order to save others, in short, because it violates the right to self-determination of the victim (Quinn 1989: 350). This would simply correspond to the freedom-theoretical justification of the primacy of negative right duties, as outlined in chapter III. It, however, leads to the problem that I have also outlined there: If what makes such actions morally wrong is the fact that they violate the right to self-determination, the freedom-theoretical approach would prima facie lead to a theory, according to which any violation of a negative duty without consent is morally impermissible. This consequence could only be avoided, if there were a way to distinguish between violable and inviolable negative rights within the theory itself. However, if the wrong-making property of a violation of a negative right is the fact, that it violates a negative right, it seems to be impossible, to make such a distinction within the theory and based on the principles of the theory alone, because it would contradict its basic assumptions. Quinn himself argues that in cases where the harm inflicted is “limited”, the prohibition to instrumentalize other persons may be overridden: “Sometimes these additional rights may be justifiably infringed, especially when the prior right is not terribly important and the harm is limited […]” (Quinn 1989: 351). This statement, however, is obviously not covered by his own theory. For, if what is wrong about intrumentalization is the fact that it overrides the victims right to self-determination, any action that overrides the victim’s self-determination is morally wrong. In order to justify the infringement of some negative rights, while maintaining the prohibition against intrumentalization for other negative rights, we have follow the lines of justification-theoretical approach I have proposed here.

6 Examples and applications

As we have seen, the differentiation between goods that are inseparable from persons and those that are external to them, allows for a differentiation between goods that are accessible to a consequentialist weighing of goods and evils and those that are to be excluded from it. This makes it possible to distinguish between two different categories of rights and duties, which also mark the respective legitimate realms of application of consequentialist and deontological ethics. I would like to further clarify this in the following by going more into detail on some of the goods already mentioned and the rights and duties corresponding to them. The right to property (or in terms of duties: the prohibition of theft or damage to another person’s property) is the paradigmatic case of a right that protects external goods. Accordingly, a typical example of a permissible weighing of goods would be if Miller appropriated Smith’s boat in order to save Jones, who was about to drown at sea.

Another case, which I want to discuss, is more complex. It is based on a thought experiment that Derek Parfit brought into play in a critical discussion of Scanlon’s theory of “reasonable rejection” (Parfit 2003: 169 f.). In this case, the life of a person named Black can only be saved if the rescuers perform an action that causes another person–let’s call him White–to lose his arm. Parfit argues that in this case, the principle of “reasonable rejection” must come to exactly the same result as consequentialism. The reason for this, he argues, is that Black could reasonably reject any rule of action that contradicts the “greater burden principle”, i.e. the principle saying: “We are permitted to impose a burden on someone if that is our only way of saving someone else from a much greater burden” (Parfit 2003: 169). Scanlon, on the other hand, is apparently of the opinion that duties to help have less overall weight than negative duties, as his comments on duties to help make rather clear (Scanlon 2000: 223-225). Parfit, however, is right against Scanlon, that such a priority of negative duties over positive duties cannot be justified within the framework of Scanlon’s own theory. Of course, Scanlon could argue against Parfit that the person who would lose an arm could reasonably reject the greater burden principle precisely because of the harm it permits to impose on him. But then the principle of “reasonable rejection” would simply lead to a stalemate, since Black, on the other hand, could reasonably reject a principle that rejects the “greater burden principle”, because a principle that rejects the “greater burden principle” would imply that a more serious harm happens to him.

Such a stalemate can only be avoided if one reads the idea of “reasonable rejection” in such a way that the “reasonable rejection” of a principle according to which an evil would occur as a result of an action outweighs the “reasonable rejection” of a principle from which follows that an evil occurs as a result of an omission. But this would require a theory that explains why and in what way the difference between doing and allowing an evil would be morally relevant. Such a theory, however, can obviously not be derived from the principle of “reasonable rejection” alone, as Parfit convincingly shows. The approach I have proposed above, however, offers such a theory. According to it, White could reasonably reject the “greater burden principle” by pointing out that, in the context of interpersonal justificatory relations, the operation of weighing goods and evils is at odds with the particular character that the good of bodily integrity has, i.e. that it is inseparable from him as person and that it therefore is incommensurable with any other goods in other persons. On the other hand, Black could not reasonably reject the principle that disallows White’s arm to be chopped off in order to save Black, precisely because that principle would be based on the justification-theoretical arguments developed here.

7 Conclusion

The starting point of my reflections was the fact that there are two strong moral intuitions with regard to the question, whether there are any absolute duties. Both intuitions are shared by many people, but do not seem to be theoretically reconcilable: On the one hand, the intuition that some types of actions are always morally impermissible, regardless of how a weighing of goods and evils would turn out in the concrete case, and on the other hand, the intuition that a weighing of goods and evils seems to be morally permissible, if not even required, with regard to most duties. Both intuitions do not seem to be theoretically reconcilable, because only approaches that emphasise the right to self-determination and freedom seem to make it possible to think of certain duties and rights as imponderable at all. These approaches, however, would imply that all negative duties would be inaccessible to a consequentialist weighing. In contrast, approaches that are not based on freedom, self-determination and the logic of negative subjective rights, but on the idea of weighing goods and evils, seem to have to allow for a weighing of goods in any case and in all respects. Thus, they cannot integrate something like inviolable rights or absolute constraints into their theoretical framework.

As a way out of this aporia, I proposed a justification-theoretical approach, according to which actions and/or their principle are permissible if they can be rationally justified to the persons concerned. In the analysis of the logics of possible justification, it then became apparent that the agent, in the case that he violates a negative duty on the basis of a weighing of goods, must not only justify to the victim the impairment of the good in question, but also and especially that his reasons for action include the victim in a strategic calculus in the first place. Such a justification proves to be impossible precisely if the goods in question are, by their very nature, incompatible with the logic of an external comparison of goods. This is the case if the goods themselves do not stand in an external relationship to the person, for which they are goods, but rather coincide with the person, in the form of her body, her psyche and her existence as a person. If this is not the case, a violation of a negative right based on a weighing of goods and evils, can eventually be justified. If it is not the case, justification is impossible and hence, the action categorically impermissible. This means that a justification-theoretical approach can integrate both the basic “deontological” intuition that some actions are always and in every case impermissible, and the basic “consequentialist” intuition that for most cases, a weighing of goods and evils constitutes an appropriate and permissible reason for action.