Moral philosophy has come to rely on moral intuition as an arbitrator of truth. Ethical theories live and die by their ability to match our moral intuitions. The result of this strange worship of intuition is that the moral theories that survive are nothing more than restatements of what philosophers find intuitively plausible, with no justification for their foundation besides their intuitive plausibility.
The problem with this approach is shown most clearly in the case of utilitarianism versus deontological philosophies. Utilitarianism, at least in its simplest form, is a perfect example of moral philosophy done the right way – its premise (that happiness is good and suffering is bad) is observably true, rather than justified “intuitively” (which is not any justification at all). Deontological philosophies are the opposite – their premises rely on things like “rights” and “duties” which can only be justified intuitively.
Since philosophers see moral intuition as the sole determinant of what is right and wrong, utilitarianism is of course criticized on the grounds that it can give counterintuitive conclusions. So for example critics will proudly point out that utilitarianism cannot guarantee against slavery. Of course, it isn’t actually possible to construct a realistic situation where utilitarianism recommends slavery – it’s pretty damn difficult to construct an UNrealistic one – but that’s beside the point. Slavery is always wrong, because our moral intuitions say so. So if utilitarianism says it’s wrong 99.999% of the time then that is not enough, because our moral intuition says that it is wrong 100% of the time. And since it’s impossible that moral intuition could be wrong 0.0001% of the time, utilitarianism must be false.
After having so thoroughly proven utilitarianism false the critic will then go on to argue for his own deontological moral philosophy which, he will proudly note, CAN guarantee against slavery in ALL cases. Deontological philosophy matches our moral intuition that humans should never be enslaved because one of its premises is that humans have a right to freedom, i.e. that humans should never be enslaved. This premise is justified because intuitively it seems obvious that humans should never be enslaved.
Utilitarianism’s ability to go against our moral intuition is not a weakness. It is a strength – no, not even that – it’s a prerequisite for an ethical theory to be able to make any progress at all. If an ethical theory is lauded for its inability to violate our intuitions then it is being lauded for being a reflection of our beliefs rather than for being able to improve our beliefs. It is congratulated for being unable to produce progress.
The progressive nature of utilitarianism versus the sycophantic nature of deontology is born out in philosophical history. Bentham and Mill, the Big Two utilitarians, both worked against racism and sexism at a time when these were dominant ideologies. Why? Because utilitarianism is an actual ethical theory which can make progress; it helps us move beyond our unjustified prejudices and see what is actually true. It is capable of coming to counterintuitive conclusions and of challenging our beliefs.
Deontology – rights and duty-based ethics – is supposed to be superior to utilitarianism because it guarantees against oppression and prejudice. And yet Kant, the great deontologist, was extremely racist and inexcusably sexist. The greatest deontological philosopher of all time had ethical views that were simply unconscionable. This is because the forces deontological philosophies appeal to – ‘rights’ or ‘duties’ – are invisible (and nonexistent), and so what our rights and duties consist of is ultimately determined by intuition, which of course is where our biases and prejudices show themselves most prominently. As such the only thing a deontological philosophy can truly guarantee is not that it will protect against oppression but that it will conform to your prejudices and biases.
Let’s look at an example from the 20th century. Now that intellectuals have reached an agreement that sexism and racism are definitely bad, a deontologist comes out of the woodwork and condemns utilitarianism for not guaranteeing against slavery… even though utilitarian reasoning is the only ethical system that consistently opposes it. This new deontologist is massively successful – this is John Rawls and his Principles of Justice.
But John Rawls’ philosophy is not an improvement. It is again simply an iteration of his generations’ status quo, carrying his generations’ biases and preconceptions along with it. His philosophy does indeed guarantee that it will frown upon the kind of oppressions philosophers were already frowning upon. What it fails to do is fight an oppression that is still considered acceptable – in fact, Rawls’ principles of justice actually work to reinforce that oppression. Specifically, he leaves animals entirely out of his original position thought experiment. This is inexcusable. Bentham, who lived in the eighteenth century, was an animal rights activist. Yet Rawls – a modern 20th century philosopher, who criticizes utilitarianism for failing to protect aganist oppression – cheerfully leaves animals out entirely of his principles of justice. As far as his principles of justice are concerned there is absolutely nothing wrong with factory farms, dog fighting, or any other kind of horrifying mistreatment of animals.
Once animal rights have been established – through utilitarian reasoning – perhaps some fashionable deontologist will condemn utilitarianism for not guaranteeing strongly enough against the oppression of animals; that it allows oppression of animals to occur in an obscure thought experiment that she will fail to define very well. And perhaps she will have dictated this condemnation to an artificially intelligent computer which she cheerfully keeps bound as her personal assistant, because after all, it just doesn’t seem intuitively plausible that computers can have rights.
Plato is Terrible 
Your eyes must be sharper than mine. I can’t observe goodness or badness in the world. Rather, it seems intuitive to me that happiness is good (possibly excepting that of vicious people), and (undeserved) suffering is bad.
(I also find it intuitively obvious that there are — very distant — possible worlds where slavery is a good thing, and not any violation of the ‘rights’ that are really worth having for people in those circumstances.)
Feelings can be observed to have value, rather than intuited, because we experience what is being said to be good or bad. If someone claims that suffering is intrinsically bad, I have access to what it is that is supposed to be bad – the feeling of suffering. I can recall my experiences of sufferings and judge that yes, that was a bad feeling.
This is different from ontological constructs like rights – the ‘bad’ness that is caused when my rights are violated is not accessible to me. In fact my rights could be violated without me ever being aware of it. If someone claims that when my rights are violated this is intrinsically bad, I have no way of confirming this statement.
The only way you could argue that intuition is involved in the utilitarian claim is if you stretch ‘intuition’ to include any kind of judgment or interpretation. If you want to, that’s fine – in that case I would simply be arguing that we should avoid the kinds of intuition that are based on nothing at all, as this kind of intuition is unreliable, and favour the kinds of intuition that are judgments based off of our experience, which is more reliable.
I’m not seeing any principled distinction here. When we suffer pain, we judge this experience to be bad. When a deontologist is treated as a means to benefit others, without their consent, they too are likely to form a normative judgment on the basis of this experience. They judge that they’ve been wronged. Both we and the deontologist then extrapolate our judgments to cover other, similar cases beyond our direct experience. What’s the methodological difference here?
“Deontological philosophies are the opposite – their premises rely on things like “rights” and “duties” which can only be justified intuitively”
I would say that rights and duties are the conclusions of deontological ethical theories. Their premises refer to properties such as “treats no-one as a mere means” which seems to be a good thing.
“Once animal rights have been established – through utilitarian reasoning – perhaps some fashionable deontologist will condemn utilitarianism for not guaranteeing strongly enough against the oppression of animals; that it allows oppression of animals to occur in an obscure thought experiment that she will fail to define very well. And perhaps she will have dictated this condemnation to an artificially intelligent computer which she cheerfully keeps bound as her personal assistant, because after all, it just doesn’t seem intuitively plausible that computers can have rights.”
Tom Regan’s “The Case For Animal Rights” is an incredibly detailed work in analytic philosophy. I find it much more rigorous than most of Singer’s work. As you can guess from the title, it is firmly based in deontological ethics. As for your last sentence, I think that there are relevant differences between pigs and PCs.
Matt: It depends on the theory of course. But even the premise you mentioned is intuitively based, so I’m not sure how that’s a counterexample to my argument.
re: Tom Regan, I’ve never heard of Tom Regan so I can’t say how rigorous he is or isn’t. But if he relies on premises like “treat no-one as a mere means” then, again, it falls under the same category I argued against here. If you’re rebutting the specific jab I make when I mention that the thought experiment won’t be defined very well, I’m just noting that a lot of the thought experiments I’ve encountered are poorly defined. I’m sure that there are some that are well defined.
And of course, saying that you think there are relevant differences between pigs and intelligent computers, without expressing what those differences are, is more proving my point than objecting to it. The deontologist would believe so too – that’s the problem with deontology!
Richard: I have to confess I’m having trouble drawing a clear distinction between the utilitarian and deontological claims, so for now at least I won’t try to defend the claim that the utilitarian premises are observably true vs. that the deontological premises are intuitively justified. I still believe the utilitarian claim is better justified than the deontological one for similar-but-different reasons, but that’d be getting offtopic. This post was more about the observation that it’s irrational for deontologists to criticize utilitarianism for not guaranteeing against oppression, as utilitarianism has demonstrated itself to actually be far superior to deontological philosophies in that regard.
‘Utilitarianism’s ability to go against our moral intuition is not a weakness. It is a strength – no, not even that – it’s a prerequisite for an ethical theory to be able to make any progress at all.’
Here, I think it might be helpful to distinguish between two kinds of ‘ethical theory’. Perhaps this can be reckoned as the ethics/metaethics distinction, or perhaps it is better cashed out as: attempting to get “better” moral conceptions vs. attempting to describe/explain/theorize about actual moral conceptions – something in the ballpark of the normative/descriptive distinctions.
One may say it is all well and good to be harsh to ‘intuition pandering’ with the former (although whether theorizing is desirable in this area at all is another matter), but when it comes to the latter, perhaps such a practice – of working out theories to fit with existing moral conceptions, and using intuitions as a guide – can lead to progress on the descriptive/explanatory front. What do you think?
I wouldn’t have a problem with a philosopher who worked on moral intuitions in the way you describe, provided they explicitly said that they were being descriptive and not normative – but I haven’t seen much of that. Most of the intuition-pandering philosophy I’ve encountered so far has definitely been normative (e.g. the philosophy of Kant and Rawls), so it falls into the former camp.
I also think that in moral philosophy normative work should take precedence over descriptive work, as the descriptive work might be interesting, but it certainly isn’t the goal of moral philosophy.
Le travail de recherche subjacent l’ouvrage enpÃlcocy©dique sur la beauté, publié chez Gallimard, serait-il immédiatement déclassé par le fait que la fondation de L’Oréal en soutient la réalisation (2 millions d’euros) ?
I find it extremely perplexing that, in this post, Kant is derided for apparently pandering to intuition, whilst in another post you attack him for principles like ‘never lie’ which are ‘obviously’ (intuitively?) wrong. Hmm..
I don’t see the problem. If you’re saying Kant had some claims which were counterintuitive even in his day (obviously now his racist and sexist claims are counterintuitive, but they weren’t back then), sure he did. He wrote a lot. But those moments are the exception rather than the rule as far as I can tell and if you want to argue otherwise then you’ll have to give more detailed examples.
And if you’re saying that I’m hypocritical for using intuition in that post, then that doesn’t weaken my arguments (that’d be ad hominem). And in my defense there is a difference between using intuition as casual support for a side-argument in a blog post and using it as central support for your entire ethical system. In one case, you can admit that intuition is flawed without hurting your argument greatly, while in the other, admitting that undermines your entire ethical system.
People criticise Kantian ethics for a lot of things, but this is the first time I’ve ever read it attacked for intuition-pandering. It’s usually thought of as the most blatant example of method-first philosophy: begin with a general principle (are you claiming the Categorical Imperative is intuition-pandering?), then apply it accordingly and accept whatever particular conclusions arise (never lie, never murder and so on). How is this mere assertion of the common man’s view trying to be passed off as moral philosophy? Since when has anyone thought you can never lie? And where in the Groundwork does he begin with this and merely assert it as ‘obvious’ and therefore right?
Isn’t the idea that you ought never to lie pretty common?
But in any case, suppose you found the “never lie” conclusion counterintuitive – then you say “well, a world in which people universally lie under certain dire circumstances (e.g. when an axe murderer asks you to tell him where your friend is) clearly wouldn’t be that badly off. So lying under those circumstances is fine.” So the Categorical Imperative relies on intuition to evaluate whether the maxim works when universalized or not, and on how universal you make the maxim. Which means that intuition has a large say in what actions are OK and what aren’t.
I’ve never heard a clear explanation of how Kant deals with the problems of 1) how do you evaluate whether what results from a maxim being universalized is good and 2) how do you tell how “universal” you make the maxim (since clearly nothing will be good if you do it completely regardless of circumstances – e.g. eating). If his philosophy isn’t going to be intuition-pandering then he has to have good answers to both of those problems rather than leaving them to be worked out intuitively.
Well I think you proved yourself via example how the idea isn’t common at all, so I take that as an (at least implicit) concession that Kant is not intuitive.
Anyway, the little conversation you imagine with yourself shows you really just don’t get Kantian ethics if you think that’s how it works. You don’t apply the CI and then only accept it if it’s intuitive. It’s called the supreme principle of morality for a reason – because its dictates are final. You realise Kant did actually write a response to this mad-axe objection insisting he was right – even here you tell the truth. And the reason he thought that isn’t because such a world would be ‘better off’ (your use of the phrase ‘badly off’ showed you are fallaciously assuming he’s a consequentialist). It would simply be because of the fact that you could not even begin to will being lied to yourself – because, then, it would of course cease to be a lie were you conscious of it. Lies are necessarily deceitful, thus violating what he takes to be fundamental to morality: sharing reasons, the publicity condition (as also found in Rawls).
“I’ve never heard a clear explanation of how Kant deals with the problems of 1) how do you evaluate whether what results from a maxim being universalized is good” The universalisation procedure IS the test of ‘goodness,’ if you wish to call it that (though Kantians prefer the word ‘rightness.’) There is no intuition involved.
If you want to call any moral theory along these lines intuition-pandering then fine, so be it, use those words in that way. But it seems pretty clear you’re radically redefining the concept of ‘intuition’ in a way nobody has ever used it before.
Incidentally, even if Kant *did* go a lot by intuition, there would probably be good reason to resist your objections to this. Intuition is *the* fundamental evidence-source in philosophy and always has been. Are you seriously telling me if I proposed to you a thesis along the lines ‘justice is the killing of innocent people for the sheer bloody mindedness of it’ – you would be happier and think my theory is more worthy of attention than if I were to say ‘well, we say punishing criminals is just, and everyone getting a fair share is just, so let’s begin from here…’
incidentally, Kantians don’t have to say what Kant said about the mad-axe scenario, and they can get out of it along Kantian lines rather than ‘intuition-pandering’ – see something like Korsgaard, Kant on Dealing with Evil and the Right to Lie in her Creating the Kingdom of Ends if you’re sincerely interested.
Korsgaard doesn’t really get out of the problem.
“Anyway, the little conversation you imagine with yourself shows you really just don’t get Kantian ethics if you think that’s how it works. You don’t apply the CI and then only accept it if it’s intuitive. It’s called the supreme principle of morality for a reason – because its dictates are final.”
There’s definitely an evaluative component to the CI, where you decide whether universalizing the maxim would be acceptable, I’m not sure why you’re pretending as if there isn’t.
“It would simply be because of the fact that you could not even begin to will being lied to yourself – because, then, it would of course cease to be a lie were you conscious of it.”
Uh. It’s certainly conceivable that lying under certain circumstances could be universalized – to pick an easy example, if I give someone a present that they don’t like, I could easily will them to lie and say they like it, even if I know that they don’t, because I would feel better regardless.
And if I knew that in the future I was going to turn into an axe-murderer I would definitely will people to lie to me to try to impede my future axe-murdering ways. Or does Kant’s universalization assume that everyone (including axe-murderer me) can read everyone else’s mind? That seems like an unreasonable addition. And if axe-murderer-me would just know that you’re lying – but not what the actual truth is – then lying is still possible since you would still be withholding the truth from someone (e.g. if there are 4 different ways your friend could have gone, and you lie and say they went down one way, the axe murderer can only narrow the friend’s path down to one of 3 exits).
“The universalisation procedure IS the test of ‘goodness,’ if you wish to call it that (though Kantians prefer the word ‘rightness.’) There is no intuition involved.”
Kant asks if rational people would will that we live in a world where the maxim is universalized. How do we decide this except by intuition? As you point out it would indeed be silly to evaluate it from a consequentialist standpoint..
You will find that the most important and influential critiques of utilitarianism don’t involve the idea that it “goes against our moral intuitions”. You will find, rather, that (1) the theory can’t make certain indispensible distinctions (doing versus allowing, for example), (2) that it relies on a questionable assumption about values (that they are fully commensurable) and (3) that it’s based on a kind of metaethical objectivism that cannot make sense of the sources of action and subjectively meaningful projects.
If I were you, I would be very careful making gross generalizations about all moral philosophers without reading them first! The most famous modern critique starts with Bernard Williams in “Utilitarianism, For and Against”, “Persons, Character and Morality” and “Moral Luck”. Happy reading! 🙂
Well, the example you gave of an indispensable distinction that utilitarianism doesn’t make – doing vs allowing – is actually very clearly an argument that utilitarianism goes against our moral intuitions, in fact it was one of the cases I had in mind when I wrote this.
I remember reading your objection in an intro course. It was given in the example of Jim the explorer, who stumbles across an army captain who is about to shoot 20 innocent natives for some reason. The captain says that as an honoured guest, Jim can shoot one of the natives and the captain will let the other 19 natives go free. The idea is that Jim has a tough decision to make because after all, him killing one person himself is different from him allowing 20 people to die by omission. And yet utilitarianism would say we have a very easy decision because it doesn’t make the distinction between doing and allowing, and so clearly utilitarianism must be wrong.
That objection is a purely intuitive one, and it’s a really bad objection. I mean, my intuitive reaction to the Jim the Explorer thought experiment is that utilitarianism is absolutely right, Jim does have an easy decision to shoot one of the natives! I can’t believe anyone would let 19 people die for purely selfish reasons – so that they personally aren’t “a killer.” And it would be entirely selfish – killing one of the natives has NO impact on the native you kill, since he would have died the exact same way regardless had you done nothing, and saves the lives of 19 others.
People use intuition as an excuse not to have to defend their views logically. But seriously, how can you justify not shooting a native in the Jim the explorer example? If the idea is that killing is a terrible wrong, but not doing anything isn’t, then you should STILL shoot one of the natives, because if you don’t, then the captain will shoot TWENTY natives, which means that twenty times more killings will take place! The only way you can come to the conclusion that you shouldn’t shoot anyone is if you consider the situation from an entirely selfish perspective where you don’t consider what’s best for anyone but yourself.
It’s important to bear in mind that the example does not claim that utilitarianism gives the “right” or “wrong” answer. The example has nothing to do with what the actual right answer is, rather, it asks you to consider the WAY in which utilitarianism answers the question: via a total disregared for the doing/omitting distinction. While it’s very easy to claim, impersonally on the internet, that you don’t feel the force of this distinction, I submit that in the actual situation you would not be so sure. You might grit your teeth and fire the gun, knowing it’s the right thing to do, but the fact that you HAVE to grit your teeth says something. Plausibly, the fact that YOU have to pull the trigger means something to you.
There are also deeper problems with your accusation that non-utilitarians rely on “intuition” (while, presumably, utilitarianism is just “logic”). All moral theories rely on intuition to some degree, and utilitarianism is no exception: the classical version of the theory, for example, is hedonistic, and Mill himself claimed that he had no proof for this part of the theory, except to say that people seem to generally believe it.
The “you’re using intuition while I’m using logic” tactic is kind of empty for this reason, because you have to say why MY use of intuition is wrong while YOURS is somehow OK. Because you will rely on something like intuition at some point, as all ethicists must.
Don’t change the subject like that. First you say that the major objections to utilitarianism are not based on utilitarianism being counterintuitive, while suggesting that I should read around 700 pages of Bernard Williams before having an opinion on the matter, and after I pointed out that actually the major objection you mentioned is entirely based on utilitarianism being counterintuitive – I mean, Bernard presents a thought experiment then asks us “Surely it seems that utilitarianism is neglecting something here” which is pretty much as clear as he can make it that he’s using intuition – you completely drop that point without conceding anything and switch to saying “well you’re using intuition too – please submit a full defense of utilitarianism in order to respond to this.”
And regarding your accusation that I would have a hard time not making the ‘allow-do’ distinction in the actual situation – so what? For all we know I’d curl up in a ball crying and whimpering while the captain shot the 20 natives. Does that mean that curling up in a ball and crying is the ethically best choice? Or as you would put it, does my curling up in a ball crying mean that although curling up in a ball crying may not NECESSARILY be the best choice, SURELY it indicates an underlying ethical rule that in crisises we should try not to do anything?
In any case, it’s certainly true that I haven’t outlined an ironclad defense of utilitarianism anywhere on this page. I wasn’t trying to. I was trying to point out that many philosophers have a disturbing tendency to unquestioningly take some very questionable intuitions as moral ‘facts,’ and that this can result in them holding atrocious beliefs, e.g. that it’s OK to let 20 people die rather than shoot one person (who will be shot anyway) yourself. It can also result in them rejecting perfectly reasonable theories because the reasonable theory disagrees with their unreasonable intuition.
If I was giving a talk at a conference and someone provided a second reason to reject my original argument, I would not be allowed to say “hey, stop changing the subject”. But this is your blog, so fine, let’s stick to the original argument.
“And regarding your accusation that I would have a hard time not making the ‘allow-do’ distinction in the actual situation – so what? For all we know I’d curl up in a ball crying and whimpering while the captain shot the 20 natives. Does that mean that curling up in a ball and crying is the ethically best choice?… best choice…”
Now, what do you think gives you license to talk about “the ethically best choice” here? I will repeat: the example does not have anything to do with what “the best choice” is. The question, rather, is about what you believe about action. And, I say again, that you (like all of us) accept that there is a difference between having to do something yourself and allowing another person to do it. We can pick the example apart, but countless others can be (and have been) produced to show that as a decision-making procedure, utilitarianism is hopeless, because it can’t make sense of a distinction that is central to action itself.
Again: the most powerful arguments against utilitarianism don’t end with: “and that’s counterintuitive, isn’t it?” This one doesn’t, neither do lots of others. I agree that intuition-pandering is stupid, because we have all kinds of bizarre, contradictory first-order moral intuitions that shouldn’t be pandered to. But that does not insulate utilitarianism from some important objections!
Nick, utilitarianism (I would prefer to use “consequentialism” more generally but a standard hedonistic utilitarian system could answer the question the same way) DOES take into consideration the distinction between doing and allowing, because it takes into consideration human psychology, and human minds are beset with various sorts of biases and cognitive processes which place greater emphasis on personally performed actions, etc., all of which affect how a person will respond to a given action and which in turn effects utility. To provide a simpler scenario that distills how utilitarianism takes this into account, consider two scenarios:
1. A man stands by idly, failing to push a button that would stop a person from being crushed under a weight.
2. A person is forced to push a button which drops a weight on someone, crushing them.
Utilitarianism doesn’t apply the same distinction between doing and allowing a deontological system would, but it DOES treat these scenarios differently insofar as the person in case 2 is likely to experience quite different psychological consequences: they may feel more guilt, suffer more, etc. What the utilitarian would concede is if there were no differences in terms out outcome, that any distinction between doing and allowing is morally irrelevant. But, they would point out, in most instances, all else being equal, there will tend to be differences between doing and allowing.
Your claim that it “totally disregards” it is thus not exactly true. It does disregard deontological approaches which, I would argue, do rely on rather dubious intuitions about relevant differences between doing and allowing. What humans would do in practice – whether we’d pause or not, when it comes to pulling a trigger or pulling a lever that will kill others – is not a testament to anything other than the particulars of human moral psychology.
There is interesting work in experimental philosophy on the evidential status of intuitions, and some of them even draw directly on scenarios related to the doing/allowing distinction. As far as I can tell, the deontological take on these is extremely questionable, and I regard the consequentialist approach to these as far more sensible and superior. In any case, here’s an article on the topic:
Click to access alex.pdf