Recommended Reading

Hinman, L. (1994) Ethics Florida:Harcourt Brace College

Walker, Ralph C. S. (1978) The Arguments of the Philosophers: Kant, London:Routledge & Kegan Paul

Pojman, L. P. (1995) Ethics: Discovering Right and Wrong Belmont, CA:Wadsworth. Pp. 133-159

Rachels, J. (1999) The Elements of Moral Philosophy Boston:McGraw-Hill College. Pp. 122-142.

Immanuel Kant (1959) [1785] Foundations of the Metaphysics of Morals (tr. Lewis White Beck) Indianapolis:Bobbs-Merrill.

EthicsUpdates

Text of Kant

Immanuel Kant Groundwork of the Metaphysics of Morals in Perry/Bratman pp. 529 ff.. (Page references are to the standard edition)

Kant

Born 1724 in Königsberg. Father was a saddler. Professor at U. of Königsberg. Died 1804.

Introduction

Remember that I said there were two main classes into which theories that take ethics as talking about something real can be assigned: the teleological and the deontological. Today we shall look at the most important current version of deontological theory, which is Kant’s ethics.

The Inquisitive Murderer

One of the most famous examples that Kant uses in explaining the consequences of his view of ethics is the case of the inquiring murderer (from ‘On a Supposed Right to Lie from Altruistic Motives.’) It goes something like this. You are walking along the street when you see a person who seems to be in some distress. It turns out that they are being hunted by a cold-blooded killer who wants to do them in. They tell you that they’re going home to hide until the heat is off. No sooner have they rounded the corner than someone who is obviously a cold-blooded killer intent on mayhem appears. They ask you if you’ve seen their intended victim. They tell you that their current plan is to go to the victim’s home to see if they’re hiding there. What do you do?

I think most people here would say that you should certainly not cooperate with the intending murderer by telling them that they will indeed find the person they seek at home. Carry on. Probably you will even feel that you should not remain silent and allow events to continue in the current fashion, which you have every reason to believe will result in foul murder being done. In fact, most people will think that you should – at the very least – lie to the inquisitive murderer.

Kant, however, thinks that it’s always wrong to lie. Therefore you must tell the truth if you reply. This seems to most people to be an absurd conclusion. What possible reasoning could lead to this conclusion?

Duty

Derivation

Kant took as the root concept of ethics not goodness, as the utilitarians did, but obligation. Whereas the Utilitarians developed a complete idea of what constituted the good for humans, and from that were able to deduce just what were the salient considerations in making a moral choice, Kant developed an idea of what it meant to be obliged to do something, and from that deduced an entire moral theory. How does he do that?

Obligation is expressed by the word ‘ought’, but Kant is aware that there are two distinct ways of using this word. In the first place, we may say things like: if you want to get a good mark you ought to do your homework; if you want to win the championship you ought to practise hard, to get to Pac Fair you ought to take the number 6 bus. All these are the sorts of obligation that refer to the logical or practical necessities that are implied by the desire to achieve certain goals. Thus: if one has a goal that is desired, and one believes that the only way that that goal can be achieved is for a certain course of action to be followed, then that is what ought to be done. This sort of statement Kant called a Hypothetical Imperative. The strength of the ‘obligation’ in such an imperative was derived entirely from the strength of the desire that occurred in the antecedent of the hypothetical statement. It has no absolute strength of its own because by removing the desire the obligation is also removed.

Quite different from this form of imperative is the form that Kant sees as being characteristic of moral judgements. They do not say if this then one ought to do that: they say simply that one ought to do such and such. Thus: you ought to tell the truth, you ought to keep your promises, you ought not to steal. These are not hypothetical statements, they are categorical; and so Kant calls them Categorical Imperatives. Being other than hypothetical, however, means that there is a problem about what gives the ‘ought’ in them their power. Kant argues that whereas the obligatoriness of the hypothetical ‘ought’ is derived from the fact that we have desires, the obligatoriness of the categorical ‘ought’ is derived from the fact that we have reason. This is because any such categorical imperative is derivable from a single such imperative which is such that any rational creature is bound by it. Naturally, this root imperative is central to Kant’s philosophy; it has the following form (Kant, p. 421):

The Categorical Imperative:

Act only according to that maxim by which you can at the same time will that it should become a universal law.

In this principle a ‘maxim’ is the general subjective rule of the particular action that you are taking, and a ‘law’ is a principle that passes the test of universalizability.

Application

How does Kant envisage the application of this principle? Well, consider the case of the inquisitive murderer. You may well think that there are good reasons for you to lie, but Kant thinks that you may not lie. Why? Suppose that you did lie. What is the maxim of your action? (This question actually suggests a difficulty for Kant’s formulation of his principle which we will look at a bit later.) Let us suppose for now that the maxim is ‘I should tell a lie.’ Kant then asks us to consider whether we could rationally will that maxim to be a universal law. As a universal law that maxim would be transformed into the general principle ‘Everyone should lie’. Could we will that ‘everyone should lie’ be a universal law? It would seem not, and not just because if that were a universal law then we would be unable to trust the communications of any of our fellows, society would become impossible. That would be merely the Utilitarian’s objection. More importantly, and crucially for Kant, the concept of a lie depends upon the concept of an assumption of truth-telling, and if it is a universal law that everyone should lie then there could be no assumption of truth-telling and therefore there could be no such concept as a lie, and therefore there could be no universal law that requires us to lie (because it is telling us to do something for which no reality corresponds to the name) and so – by this reductio ad absurdum – there can be no such thing as a universal duty to lie. Thus the act of lying, in any circumstance, violates the categorical imperative.

Something similar happens with breaking promises. If you break a promise to repay a debt, for example, then you are acting according to the maxim: ‘I should break my promise.’ This would be universalised as ‘everyone should break their promises.’ But if this were a universal law then there could be no concept of an expectation that people will honour their solemn undertakings, and thus there could be no concept of a promise, and therefore there could be no sense to a universal law to break promises (because there is nothing like a promise), and there can therefore be no such thing as a universal law to break promises. So by this reductio there is no way to rationally will that your maxim be the basis of a universal law. QED.

Can you construct a similar argument to show that you ought not to steal?

Problems with the Categorical Imperative

1.         The Indeterminacy of Maxims

One of the principal difficulties with the CI criterion as a guide to moral behaviour is that we don’t have a really good way of deciding what the maxim of our action is going to be. Note that any behaviour has a wide variety of descriptions to which it answers and it is this description – particularly as it is presented to the consciousness of the intending actor – that constitutes the maxim of the action. In the case of the inquisitive murderer, perhaps our action is not best described as telling a lie. Perhaps what we’re doing there is protecting the innocent from oppression. If we think that we should help innocent people escape unlawful death, and we think that this is the motive of our action then we make that to be the maxim of our act. It’s not at all clear that this couldn’t be consistently willed to be a universal law – ‘Everyone should protect innocent people from oppression’.

2.         The Plurality of Duties

It is often argued that it is entirely unlikely that any single principle could be the foundation of a complete moral system. The variety of moral choices is too great for all of our intuitive judgements to be consistent with the judgements that can be deduced from a single principle. The case of the inquisitive murderer is a possible case in point. There are two reasonable principles: that one should not allow the killing of innocent people; and that one should not tell lies. In the case in question there is an unavoidable clash between the two. You cannot honour one without violating the other. Of course, it is open to the Kantian to deny that there are always such potential clashes with any absolute principle – and I don’t know of any conceptual argument that would make the contrary case – but the prima facie evidence is that these conflicts do occur (certainly they occur with all interpretations of the CI) and it is up to the Kantian to show that the conflicts can be avoided, and then to show that they are avoided.

From another point of view we can look at the evidence of conflicts of duties as evidence for the non-absoluteness of moral laws. The reason, after all, that the conflict is felt to be a problem is that neither of the conflicting principles can be modified. But this may not be a fatal flaw in the Kantian scheme; it may be possible to describe your actions in such a way that the maxim of your act contains within itself sufficient flexibility that it can account for all the reasonable exceptions. For example, the bystander in the case of the inquisitive murderer could operate under the maxim, ‘I can lie to people if they are going to use my information to do wrong.’ But what if the people you are going to lie to are the legal authorities. Surely you aren’t going to defend lying to the legal authorities? Wouldn’t part of the consequence of their being legal authorities be that they have a right to your acquiescence in their actions? Isn’t that what an authority is? Wouldn’t the universal law that your maxim would lead to remove all concept of authority? And so the debate could continue.

3.         The Underdetermination of Duties

One more difficulty that we might mention is that it is possible to think of any number of descriptions of acts that could consistently be willed to be universal laws but that we would not want to include as moral duties. Remember in Swifts Gulliver’s travels how the Lilliputians were involved in a war with their neighbours because one side broke their boiled eggs on the big end and the other side broke them on the sharp end? Swift sees this as a dispute over a trivial difference, but perhaps a Kantian would see it as a fundamental moral conflict. There are very clear maxims that one can apply to the act of breaking an egg, and one could, for example, act on the maxim: break the egg at the big end. It seems to be perfectly possible to make this into a universal law without leading us into inconsistencies or paradoxes, yet who would believe that this is a moral law?

Kantians may reply that this just shows that the Kantian test allows more actions than just duties. And that’s a reasonable response, I think, but what about the maxim ‘I shall kill Mormons’? That could be universalised without inconsistency, but who’d want to count it as a duty? It would be better to include the negation of the statement as a duty.

Value

Derivation

One of the curiosities of Kant’s moral theory is that he presents his fundamental moral principle in several forms, the consequence of viewing the problem of ethical actions from several different points of view, and claims that they are all equivalent. The second formulation, which is known as the Principle of Ends, he derives from considerations of the innate worth of rational persons. Just as the justification for the CI derived from an observation that obligation came in two forms, the justification for the PE derives from an observation that value comes in two forms. When we say that a thing is valuable, we often mean that the thing is valuable for a purpose, or more precisely, is valued by a person because it is a means to an end which they desire. If you want to get a good mark you will value your textbooks; if you want to win the championship you will value practise; to get to Pac Fair the number 6 bus is valuable. Thus: if one has an end that is desired, and one believes that a thing is necessary for the achievement of that end, then that thing is valued. A thing valued in this way has Conditional Value. Its value is derived entirely from the strength of the desire in the antecedent, and the value of the thing doing the desiring.

Quite different from this form of value is the form that Kant sees as being characteristic of moral agents. For them, we do not say that if such and such is desired then one values that person: we say simply that that person has value. These are not conditional statements, they are unconditional; and so the value that is attributed is Unconditional Value.  Being other than conditional, however, means that there is a problem about what gives them their value. Kant argues that whereas the conditional value of a thing is derived from the fact that that which values it has desires and is also valuable, this is a series that cannot go on forever. For what could give conditional value to the thing that is doing the valuing but another thing that has desires and is valuable. But if that is just another agent of the same kind then it looks as if we are getting into an infinite regress of value attributions. To avoid this (without invoking God), and because there is nothing that obviously relevantly distinguishes one valuable agent from another, the conclusion that suggests itself is that those moral agents all have unconditional value. We treat things according to their value for us, and any thing with unconditional value has value for us, and therefore we have the second statement of Kant’s fundamental moral principle (Kant, p. 429):

The Principle of Ends:

Act so that you treat humanity, whether in your own person or in that of another, always as an end and never as a means only.

Application

Consider again the case of the inquisitive murderer. In that case, if you lie to the intending murderer, Kant seems to think, you are making him a tool for your own purposes, you are using him for a means to an end rather than as end in himself.

Similarly, when you steal from another person you very clearly use them as a means to an end. You take their property to be your own because they have a bike, and you want a bike, and you reason that if you take their bike – and never mind what their desires are – then your end will be achieved by this means.

Can you construct a similar argument to show that you ought to keep your promises?

Problems with the Principle of Ends

As I mentioned, Kant thought that the CI and the PE were identical in meaning, and I invite you to closely compare the derivations that I have given for the two principles to see why he may have thought that that was the case. It is, nevertheless, very often claimed that the PE is not derivable from the CI. One of the reasons we might believe that rather than that Kant is correct is because the consequences of the PE seem to be (as far as we are able to tell) different from the consequences of the CI. For example, we do not find that the PE will allow us to adopt the maxim of action ‘Kill all Mormons’, nor will it have anything to say about breaking your eggs at the big end or the sharp end. Perhaps a better way to look at it is to say that the PE is complementary to the CI. This principle, however, has its own problems.

The Value of Rationality

Kant claims that what makes these moral agents valuable in themselves is not their humanity, which is not clearly definable, nor the fact that they belong to some group, but purely their rationality. Why does Kant think that rationality is sufficient to create value where it would not otherwise exist? Perhaps it has something to do with the fact that the ability to reason is required in order for a desire to create a conditional value in something that the rational agent values. Recall, that thing is valued only because the agent is able to reason that to achieve the end that he seeks the valued thing is a necessary means. But this just means that the unconditionally valuable could not give value to the conditionally valued unless it was rational. It does not seem to mean that the unconditionally valuable could not itself be non-rational.

In fact, Kant seems to think that the reasoning just given indicates that rationality itself is unconditionally valuable, but why he thinks that I don’t know (Kant 429). If he did think that those considerations established that conclusion then he would have been quite wrong, because it would be quite consistent with that reasoning for rationality itself only to be valuable as a means to an end. But perhaps, as a philosopher, Kant was not being entirely impartial in this estimation. After all, if we are valuable only in so far as we are rational, then surely the most rational amongst us are the most valuable. And since philosophers – and especially teachers of philosophy – are the most rational of all, it follows that philosophers are the best of people. I believe that both Plato and Aristotle held similar views.

Finally on the subject of rationality, the valuing of rationality to the exclusion of all other values has the consequence that there are no moral values for animals. You may or may not find that to be unacceptable. I, myself, am undecided.