Kant's Theory of Ethics

click here for ordinary reading
click here for a high contrast version

Agenda for Part 1 & 2

Agenda for Part 2
Discussion site I've set up a discussion site for continuation of discussion started in lectures and seminars. It's a Yahoo site - you have to register and put up with ads but it seems the easiest way just now. If you know of a better, let me know please.



Kant reacts to the threat of determinism

When we last met I was trying to worry you about determinism.

This week I want to have another go.

Determinism is a worry because it seems to imply we have no control over events, even events that are very close to us, like what we are going to do next. Determinism claims it is all out of our hands, fixed before we had any hands, fixed perhaps for ever if the universe has always existed, or from its start, whenever that was.

But I'm silly to call it a worry. In a way this idea spells the end of all worry. If worrying is something we choose to do, we had better stop it if determinism is true, since whatever will happen will happen. But if determinism is true, we won't be able to stop ourselves worrying, of course...


Kant reacts to the threat of determinism: summary

Kant led a reaction to the deterministic thinking that gathered force from the early triumphs of science. The reaction became the 'romantic movement' which developed at the end of the 18th Century.

Kant argued that there were two aspects to reality: one was the world as it appeared to observation and entered into thought, the other was the world as it really was, independently of any conceptual framework of an observer. In observing and thinking about the world, human beings apply categorization.

Anyway, let's not worry that we might be worrying unnecessarily.

The greatest worrier about determinism was one of the greatest philosophers. He lived at a time when science, the mother of determinism, was enjoying terrific success. His name is Immanuel Kant and he lived or at least flourished towards the end of the 18th Century.


Newton the great giant of early modern science, was regarded as having more or less done physics and the view of the universe and everything in it as more or less complicated bits of machinery was widely persuasive. Laplace caught the prevailing outlook when he wrote of the vision science in the 18th Century held before them: that if only you knew where all the bodies in the universe were at a particular time, then by applying the laws established by science you could work out where they would be at any time in the future. And by bodies he meant atoms particularly, so it was a way y of saying that absolutely everything was in principle predictable

Kant, the person I want us to think about today, was part of the reaction to this vision and the threat it posed to human beings' image of themselves. These revolutionaries woke up to the horror of what was proposed, and began shouting that it was completely unacceptable. It was a mix of saying: you mustn't think of human beings as

Some web stuff


mechanical because that would lead to dreadful things, and human beings are actually not like that, and science must be wrong if it thinks we are.

I'm talking here of what became the Romantic Movement, a reaction to science running let's say, from 1775 to 1830.

Causality belongs to the world of appearance only

Causality belongs to the world of appearance only: summary

Kant argued that causality was a concept that thinking beings brought to their experience of the world - it wasn't part of the world as it was in itself.

The world as it was in itself was therefore, in Kant's view, not subject to causality.

One argument he had for this was that the causal principle seems a priori.


Kant 's solution was the dramatic one of saying that causality was a kind of illusion.

Kant thought we looked out on the world through a framework of concepts, a conceptual apparatus, and what we saw, and then what we thought about, was structured by these concepts. Human experience was partly made up of input from the world, but partly made up of the concepts human beings bring to it. Think of wearing rose-coloured spectacles. What you see, a rose-coloured Alexander Square for example, is made up of two components, the square itself making some input, but also the glasses which contribute the rosy glow.

Kant's stupendous idea was that causality belonged to our conceptual apparatus and not to the world itself.

Sometimes as you get old, you get spots before the eyes - little specks of black which flit across your vision. To begin with you keep thinking it's flies or whatever. Later you tumble to that the specs are not in the world at all - they are on your retina.

It's like that, sort of, thinks Kant, with causality... We may think of it as a feature of the world, but in fact it is something that comes from us.

One argument that you may think supports this is that the causal principle seems a priori. That is, we don't seem to need to pile up much evidence before we are prepared to say 'every event has a cause'. In fact, it can seem that we don't need any evidence at all. You just have to think about ti and you will come up with the conclusion that every event MUST have a cause. If we are sympathetic to this line of thought we might well see it as supporting Kant's dramatic claim that the principle of causality is not something we get from observing nature, it's something we insist on ourselves - a principle we use in interpreting nature.

BUZZ: Causality for Kant is part of our conceptual apparatus and not part of reality as it is in itself.

Would you want to put forward any other ideas or concepts which might be similar in this regard?

Discussion site


Any how, Kant's theory is there are two aspects to reality. One aspect is reality as we see it through our conceptual structure.

The other is reality as it is when we're not looking.

The world as we experience it is subject to the principle of causality, but the world as it really is not.

When this is applied to the human being you get two aspects as well. There is the human being as they appear to other human beings, and there is the human being as they are when nobody is looking.

The real human self is autonomous

The real human self is autonomous:

This meant that the human being, in his or her self, was not subject to causality - in other words, human beings have free will


The human being as others experience him or her is subject to causality. But the human being as he or she really is not.

Human beings as they really are are free. They have a free will.

Kant's theory of ethics

Agenda for Part 2

Kant's theory of ethics: summary:

If the will is free, what is to guide it?

Nothing? - the existentialist position.

Others argue that as an autonomous being it is important for the human being to act 'well'. What then is there to lay down what is right or good for human beings to do, in the exercise of their freedom?

Kant argued that as free beings we were obligated to do what our reason dictated


But if the will is free, what is to guide it?

Or is it enough simply for the will to act? Existentialists have argued that this is all that is necessary, or possible. Human beings are free, they can act.

Others argue that as an autonomous being it is important for the human being to act 'well'.

What then is there to lay down what is right or good for human beings to do, in the exercise of their freedom?


Kant's notion of practical rationality

Kant's notion of practical rationality: summary

Kant meant by rational that which involved no inconsistency of will.

He says if when you articulate the principle you are about to act on and you find it would involve you in willing both p and not p, the act would be irrational.

Acting on the principle that you may break your promise should it be in your interest to do so would involve an inconsistency. You can't will that you should follow this principle unless you willed everyone to follow it. But if everyone followed it 'promise' would lose it's meaning. Therefore in following this principle you would be willing to break a promise and that there wouldn't be a promise to break, which is inconsistent.

So Kant's thesis is that pure rationality rules some actions out.

The idea that reason on its own might guide your behaviour may be tempting: is it defensible?


Kant's theory was that what guided us was 'rationality'.

As free beings we were obligated to do what was 'reasonable'.

Today we often use 'reasonable' very broadly to mean anything that those around us would find unobjectionable. Kant meant it though in a very strict sense. You are behaving rationally when you are behaving in a way that doesn't involve any inconsistencies.

You know what an inconsistency is in argument. You say at one point that the only thing nice about television is 24 and later on let slip that you really enjoy Bad Girls as well. You may be charged with inconsistency because flat-footedly you seem to be committed to two things which can't both be true: that there is only one good thing on tele just now, and that there are two good things on tele just now.

It is very difficult to deal with someone who doesn't mind about this. If a person seems perfectly happy to go on defending two inconsistent propositions reasoning seems to be at an end. If it comes to that we can only use a sharp stick.

The grip reason has on us in that sense is pretty firm.

But Kant is talking about action, not arguing. He is saying a free person has to act rationally - has to act without inconsistency.

How does our point about argument translate into a point about action?

Kant says this. When you are considering an action, articulate what principle you will be following if you do it. If that principle involves a contradiction you have identified an act that would involve a contradiction.

Buzz: Can you think of any principles like this?

Discussion site

- I'll completely smash this window with a hammer, and then I'll shatter it to smithereens with a baseball bat.

Is that a principle with an inbuilt inconsistency?

- You want a bowling green built in your front garden. Certainly. I will make it perfectly round, with just four straight sides.

These examples seem to fit the bill, but they are of course completely useless. They would only be guides to action if they ruled out some real possibilities.

What sort of examples does Kant have in mind?

I suppose it is a special sort of inconsistency. He says there are some principles which involve inconsistencies of the will.

They involve you in willing that a thing should happen, and willing that it should not happen: both at once.

He says if when you articulate the principle you are about to act on and you find it involved you in willing both p and not-p, the act would be irrational.

Supposing you get into a jam, from which the easiest way out is to break a promise you have given. Should you break the promise?

To find the correct answer, Kant says, first formulate the principle you would be acting on. Would it be like this?:

'When it suits your interest to do so, break your promise.'

You should then reflect rationally on this principle, says Kant.

When you do, you will realize that, if everyone followed this principle, 'promise' would lose its meaning.

The whole point of a promise is that you stick to it even when it doesn't suit your interest, so if everyone followed the principle 'break your promise if you need to', promising would be destroyed.

So Kant's thesis is that rationality rules some actions out.

He thinks of rationality then as the guide to behaviour.

Kant's picture is that there are laws of causality which rule the world of appearance, and there are rules of reason which rule the world as it really is.

Buzz: what answer then does Kant give to the question: why should I behave rationally?

I think the answer he must give is: because otherwise you will be willing inconsistently.

If you say So what? All he can do perhaps is point to rationality applied to arguing. When you have been convicted of inconsistency in your argument, it is also quite open to you to say 'So what?' But if you do say 'So what?' - if you do claim not to be bothered if you contradict yourself - you are putting yourself in a very curious position ...

The example of promising is the easiest one to take to make the point that some principles seem to involve a contradiction.

Easiest and perhaps the only.

Buzz: Do you agree that breaking a promise involves you in inconsistency?

Can you think of any other examples where this claim might with any plausibility be made?

Discussion site


To me, the idea that reason on its own might guide your behaviour is a tremendously attractive one, but it is one that I find very difficult to defend.



Hypothetical and categorical imperatives

Hypothetical and categorical imperatives: summary:

It is much easier to defend the idea that if you know what you want, reason will help you secure it. This kind of guidance Kant calls 'hypothetical imperatives'.

Kant thinks that reason also issues categorical imperative. This is what he calls reason's guidance when it tells us that we should not act in a way that involves an inconsistency of will.

That is one formulation of the categorical imperative. Kant has other formulations:

Of course much easier to defend is the idea that if you know what you want, reason will help you secure it. It will tell you if you want x, then you should do y. If you want to help, keep your mouth shut.

Where reason just helps you work out what to do when you know what you want produces what he calls 'hypothetical imperatives'. It tells you what you must do IF you want a certain outcome.

For Kant, the key role of reason in morality is not this, is not its issuing hypothetical imperatives.

Kant puts his main thesis by saying that reason also issues a categorical imperative. This is the thesis I have just set out: that as rational free beings we should not act in a way that involves an inconsistency of will, that we should not act in any way that involves the willing of two things that are inconsistent with each other.

That is one formulation of the categorical imperative.

Kant has other formulations - 'formulations' which don't seem to be quite variations of the same thing, in spite of the fact that Kant seems to put them forward as such.

One of his formulations is:

The Golden Rule

Act only on that maxim that you can at the same time will to be a universal law.

He thinks of this as a version of the Golden Rule:'Do as you would be done by'.

John Stuart Mill comments here.

The Kingdom of Ends

The Kingdom of Ends: summary

Treat the other as an end and not as a means.

But what is there in reason (in reason alone) that gives us the obligation to respect other rational beings?

Another is this:

Treat another person as an 'end' and not as a means to an end.

With many things in nature it seems acceptable to make them useful to us. If we come across a convenient flat-topped boulder around lunchtime as we trudge up Old Man Coniston, it seems intuitively acceptable to use it as a seat, or as a table for our picnic.

But suppose it wasn't a boulder but a person - a man - sitting there: sitting a little awkwardly with his legs arranged so as to offer us a flatish surface for our bottle of wine and glasses. Would it be OK to go ahead

Dali: Ghost of Vermeer of Delft Which Can Be Used as A Table (1934)

Courtesy the Dali Museum

and simply lay out our spread on this happy configuration of limbs?

We might worry that he might move. We might think it wise to instruct him not to move.

But if we took that precaution would we be acting reasonably?

Kant says we wouldn't.

A person is different from a boulder. We can use a boulder for our own purposes without further ado, but we can't use a person, or his legs, in the same way. There is something more to consider.

What more? Kant suggests the more to consider is that the person is like us, someone who has a free will. There are just some special entities in the Universe which have wills, wills possessed of autonomy. They deserve from each other a respect that is not due to things which do not have independent wills.

His way of expressing this is to say that other people are to be treated as ends and not means.

Buzz: What if it were a conveniently lugubrious sheep? Might you picnic with a clear conscience on that?

Discussion site

The question arises:

What is there in reason that gives us the obligation to respect other rational beings?

Kant's test is inconsistency. Is there some kind of inconsistency in one rational being treating another rational being as simply a means to an end?

I know it would be nice if they didn't, but is there an actual strictly logical inconsistency there?

I hope you will have a chance to discuss this in the seminars.

Further exploration of Kant's theory


Duty: summary

People often think of Kant as a champion not so much of reason, as of 'duty'.

His view contrasts with any kind of utilitarianism, which thinks that in deciding what to do it is the consequences of your action (or inaction) which count.

('Consequences' for Kant belong to the world of appearance and so can have no bearing on how the real self should act.)

Philosophers often feature Kant as a champion not so much of reason as of 'duty'.

He is known for maintaining that you should never lie, no matter what the circumstances. So that for example if by telling a lie you could save the planet you still shouldn't lie. The usual example is this. You are standing at a junction in the road when a whey-faced person flies by, looking terrified over his shoulder. He disappears, turning left. Shortly afterwards his pursuer comes into view, screaming blue murder and brandishing a correspondingly murderous instrument. 'Did he go left?' he bellows.

Kant thinks that in these circumstances the wrong thing to do is to say 'No, mate, he went up there!' pointing to the right. This would be to utter a lie, and lying is always wrong.

Here Kant takes a view which contrasts very sharply with the view you were thinking about last week, utilitarianism. For utilitarianism, the consequences of an act are crucial. If you are wanting to decide whether it would be right to do a certain thing, what you should do is to assess its likely consequences, and the likely consequences of not doing it and then you should make up your mind in the light of this assessment. Of course that is not all there is to Utilitarianism, because it has to take a view as to what consequences should be sought. Should you maximize happiness, or feelings of pleasure, or money or culture, or what? But the common point is that it is the consequence of a proposed act (or principle) that matter.

Again the question arises, what is the connection between the thesis that you should be guided by reason - i.e. you should avoid inconsistency - and the principle that you should follow simple 'Thou shalt not' type rules without taking the consequence of following them into account?

Again maybe one can see dimly the relevance of consistency. Can one imagine a person saying it is simply inconsistent to lie on one occasion and not lie on another? But I have to confess I find this terrifically simple-minded. I have to leave the question with you.

I can have a go at explaining though why utilitarianism for Kant must be wrong. It goes back to his two aspect theory, the person as he or she appears, and the person as he or she is for real. For Kant what you want, and what you enjoy belong to yourself at it appears (in this case, to you yourself). They don't belong to the self as it really is. The only thing that belongs to your self as it really is is: rationality.

But moral action belongs to the self as it really is. It is only the self in this aspect that has the freedom to act at all. So for Kant the self cannot be right in deciding what to do in the light of wants or benefits. It doesn't have any wants. Benefits all belong to appearance. So moral action cannot be governed by the weighing of consequences.

The good will the only thing that is good

Buzz: Consider the following two scenarios:

A driver falls asleep at the wheel, careers off the motorway and
(A) lands on a sheep.
(B) lands on the main railway line, derails a train loaded with nuclear waste heading for Sellafield, causes a spillage which causes, over the years, the decimation of the North West.

Which is the morally more serious act?

Discussion site

'Nothing can possibly be conceived in the world, or even out of it, which can be called good without qualification, except a good will.' Kant, Foundations of the Metaphysics of Morals, opening sentence.

Kant's theory gives us an understanding of a very striking feature of our moral judgements. We feel when we are judging somebody's behaviour that what matters morally is what they meant to do, not what they did, and not the consequences of what they did. If someone does their very best to save someone's teddy bear, but in doing so destroys a whole city, though the story is a shocker, we don't think the person is a moral fiend for doing so much harm. If he or she was genuinely trying to do the right thing, most of us judge him or her to be morally in the clear, despite the devastation.

Evidence for a world in which rationality guides

Evidence for a world in which rationality guides: summary

The special experience we human beings know which is the feeling that we 'ought' to do such and such.

You may think Kant's theory is worth thinking seriously about.

If it was right, it would confirm a number of intuitions we seem to share about morality. For example, it would explain to us why we feel when we are judging somebody's behaviour that what matters morally is what they meant to do, not what they did, and not the consequences of what they did.

But it's a lot to swallow!

We have to believe that causality is a concept that doesn't apply in the world as it really is at all. And we have to believe we ourselves have these two different aspects. There is myself as I appear to myself. Under this aspect I do appear to be subject to causality. I appear to belong to a world which has all kinds of impacts on me, and I on it. It seems to me for example that I sometimes play squash, and that when I am doing so it appears to me that I sometimes manage to hit the ball, and that I sometimes crash into the wall, and that this sometimes hurts, sometimes bruises. That is myself as it appears to me.

But these appearances for Kant are partly generated by the system of categories that I bring to experience. I can't have experiences without bringing my categories of understanding to bear. But I can wonder, Kant appears to be saying, what I am really like, what the self is on its own, when it is not being subjected to categorization. And if I ask this I arrive at the view that I am a being who is subject not to causality but to rationality. I am free, autonomous, but able to be guided by reason.

If all this is not a tall story, then being asked to believe it is certainly a tall order.

Is there any evidence for it? I have argued that if this Kantian account of the two-aspected self were true it would explain some features of our intuitions about morality.

Is there anything else?

Maybe just one thing.

The thing perhaps we keep coming back to in thinking about causality and determinism. Just this: we don't actually feel as though we belong to the world of causality. We actually feel that we are free to choose, free to act.

We find ourselves agonizing over decisions. Quite often we feel we ought to do this, or we ought not to do that.

And these oughts that we feel seem to be completely different from our awareness of facts.

We know that carrots have not always been orange and we know that teashops are plentiful in the Lune Valley. But knowing that we ought not to walk by on the other side - that seems to be quite different from these factual sorts of knowledge. If we come across someone - probably at the end of a trail of corpses - who really doesn't seem to have these moral feelings, these tugs inside us telling us that we ought to do x, or that y is completely wrong, we are extremely concerned. We think of them as very seriously defective as human beings, and we apply warning labels like 'psychotic'.

Kantians point to these features of human experience and say: it is at the point when we are considering what to do that our real selves make themselves known to us. It is our real self that at these points is feeling the tug of rationality. When we see another human being in trouble, we feel we ought to help.

At these points we are feeling not the pulls and pushes that might be felt by the cogs in a machine, but, says the Kantian, our obligations as free citizens of the moral order.

Marx on Kant


Mary Shelley's Frankenstein

I want to show you Frankenstein, creating what came to be called his 'monster', an expression of the Romantic Movement, the turn against science at the end of the 18th Century.

Dr Frankenstein, a 'brilliant young scientist', pursues the project of creating a human being, using bits from morgues and electricity. He succeeds, the creature is made and lives. But nevertheless the author has the project go horribly wrong.

I think it is difficult to make out exactly what the problem is, but what happens is the creature turns murderous.

What appears to me to happen is that he turns murderous as a consequence of being rejected by human beings. (I think Branagh's film departs crucially from its source on this crucial point.)

If I am right, the point Shelley is making is that science and human beings don't make a team. Science is successful in its own terms, but as humans try and relate to it and its successes they ruin everything.

It is a popular view that human beings, not scientists, are to blame for the misuse of science. This goes further - decisively further - in adding that they can't help it. The implication is that we should give it up.

Review questions


1. Which best expresses Kant's view:

A. Human beings are free when they are alone and can think for themselves B. Human beings are free even though all their decisions are causally determined    
C. The human being as experienced and thought about by others is subject to causality D. The mind of the human being is free but the body can be quite expensive  
Ask a friend

2. Which of the following does Kant maintain tells us what we absolutely MUST do?

A. The Imperial Category B. The Categorical Imperative    

C. The Hypothetical Imperative

D. The Empirical Imperative  
Ask a friend

3. What does Kant think qualifies a thing to be treated other than as a means to an end?

A. Being an evolutionarily advanced animal B. Being morally upright    
C. The possession of reason D. The capacity for feeling pain  
Ask a friend

4. Which best expresses Kant's view?

A. What makes an action right is the extent to which you have thought about it carefully beforehand

B. The end justifies the means

C. What makes an action right is the likelihood that it will bring benefit to other people D. There are some actions that are absolutely never justified   Ask a friend