It is difficult to have any misconceptions about Immanuel Kant’s position on certain moral issues. His thoughts on whether it is permissible to lie are perhaps the most striking example of his moral absolutism. If the title of his essay “On a Supposed Right to Tell Lies from Benevolent Motives” doesn’t make the absolute impermissibility of lying clear in itself, the notorious case of the enquiring murderer certainly will.
But first, how did Kant get to absolute moral principles – ones which are not relativistic, and apply to us all, no matter what our preferences or circumstances might be? In a very simplified form, the argument goes like this: While moral theories like utilitarianism speak of happiness as the goal of morality, Kant instead focussed on what we need to do to be worthy of happiness at all. In terms of morality, this involved doing what was right, regardless of whether it made one happy or not.
How do we know what is right? Compare what Kant referred to as hypothetical imperatives and the stronger categorical imperative. Hypothetical imperatives derive their force from relevant desires – for example, my desire to not get wet results in the imperative to carry an umbrella. However, if I lacked that particular desire, the imperative would have no force. By contrast, categorical imperatives derive their force from reason and are binding on us all, because they are products of a principle which every rational person must accept (at least, according to Kant).
That principle is the Categorical Imperative, which (in one formulation) says: “Act only according to that maxim by which you can at the same time will that it should become a universal law”. This summarises the procedure for deciding whether something is morally permissible. We should ask ourselves what rule (maxim) we would be following by performing the action in question, then ask whether we could will (in other words, make it so, rather than simply imagine) that everybody followed that rule.
In other words, what we are being prompted to consider is whether the rule in question could be universalised. If it can be, then the rule may be followed, and the act is permissible. If I chose to borrow money from you that I had no intention of repaying, I could ask whether the rule “it is permissible to borrow money without intending to repay it” could be universalised. For Kant, the answer is obviously “no”, in that this rule is self-defeating – it would effectively eliminate commerce.
So reason itself is meant to lead us to categorical imperatives involving repaying debts, and also not lying, in that a rule permitting lying would, according to Kant, make communication impossible, or at least futile (given that much communication is premised on the sharing of useful and true information). An early challenge to this prohibition on lying came from Benjamin Constant, who pointed out that according to Kant, one is morally obliged to tell a known murderer the location of his intended victim, if he were to request that information.
A lie intended to divert the murderer is impermissible, because it’s possible that unbeknownst to you, the intended victim has in fact changed location, and is now in the place you identify in your lie. This thought experiment can easily be modified to make an answer such as “no comment” equally revelatory – and therefore leaving us responsible for the death of the victim in both these cases. Or not, at least according to Kant, who said that so long as you tell the truth, it’s the murderer who has done wrong, not you. You’ve done what’s morally obliged by reason, consistency, and the categorical imperative – you have acted from a “good” will.
Some may want to bite this particular bullet and agree with Kant that the only way we can avoid moral rules becoming the victims of subjective preference or other forces is to treat them as absolute. However, I’d be willing to wager most of you would thinking there’s something absurd about not allowing us to lie to an enquiring murderer, simply because doing so would introduce cost-benefit analysis into our choice. Of course, we’re not all equally good at such analyses and, of course, some situations don’t lend themselves to these calculations as well as others.
Regardless, it’s reasonable to suspect that most people would agree it’s morally permissible to lie when doing so appears guaranteed to save an innocent person from death. And if you do agree, then you admit that at least some moral principles are not absolute, but are instead ideals to be followed as closely as possible. Or, you might say that it’s unreasonable for Kant to formulate the maxim or rule in such absolute terms: What’s wrong with a maxim like “don’t lie, except when you can save an innocent life by doing so”?
For Kant, what’s wrong is that this sort of maxim appeals to consequences, and thus offers us no absolute, rather than context-specific guidance. It opens the door to a potentially infinite number of revisions and subtle qualifiers, and leaves us in exactly the moral mess that he thought he was clarifying with his deontological (duty-based) ethics. But an inflexible set of rules that don’t appear to account for what can be known (like when you know that a lie can save a life), or that don’t allow for any ranking of rules (if all rules are absolute, what would one do when a rule to tell the truth conflicts with a rule to save lives) is an unsatisfactory attempt to clean up that mess.
This is why most ethical frameworks – the ones that seem to summarise our behaviour, rather than the ones discussed by moral philosophers – allow for some sort of hierarchy of principles, or exceptions to rules such as the “white lie”. We know that life is more messy than the ideal existence imagined by a very anal-retentive Prussian, whose daytime walks were so regular that (according to legend, at least) locals used to set their clocks by him as he passed by their houses.
While all but a handful of existing Kantians are probably behind a desk at a philosophy department somewhere, we still have a large number of deontologists in our midst. This is because the other historically popular route to absolute moral rules is through religion, and the Abrahamic religions in particular. I’ve perhaps said enough about the absurdity of basing contemporary moral reasoning on the superstitions and lack of knowledge of our ancestors in previous columns, so will make only a few simple points here.
First, as Plato argued in Euthyphro’s dilemma (around 399 BCE), propositions can’t be right simply because they are commanded by God, because this could make anything potentially “right”, if God had that particular whim. It would strike most believers as implausible that God commands them to cut off their ears – and rightly so, seeing as the immediacy of the harm should make any non-immediate (and untestable) rewards for doing so less attractive. Believers might respond that God would not command this. But this is precisely the point: We expect some conformity between the commands of God and what we can understand as rational or reasonable, and this shows that instead of what God commanding being right (by virtue of the command), God instead commands what is independently right.
Second, the Golden Rule – which strikes us as an absolute principle, and one common to many religious outlooks – is also far from absolute. It’s also far from being derived via religion, given that reciprocal altruism can be observed in many non-human animals. The idea of “doing as you would be done by” is, however, of little use if we were to apply it in all situations, without any recognition of context. As Simon Blackburn points out, imagine the criminal asking the judge “how would you like to be sentenced?” In other words, there are cases in which we think applying the Golden Rule to be inappropriate, because it’s been superseded by some other moral judgement – again showing that we do rank these things in practice, rather than treat them as absolute.
Both of these versions of a deontological morality fail, often because they provide easy answers to questions that should be understood as difficult. It is often politically desirable to give people answers that correspond to an imagined black-and-white world. But that is not the world we live in anymore, now that more of us have educations, and now that we find no compelling reasons to subjugate ourselves to edicts and forces that are not intelligible.
They fail also by offering moral confusion where there should be none. Consider policy decisions like whether to allow euthanasia or gay marriage. Both are cases in which one has to work very hard (with little reward) to understand how they can ever be considered immoral practices in themselves. For surely morality has to have something to do with making life better or worse? If I prefer for my life to end, not allowing euthanasia makes my life worse than it could otherwise be, and it seems immoral to deny me that right. And even if you want to object that having myself euthanased could make life worse for some, by (for example) causing my wife distress, consider instead the case in which I’m a widower, and have no other family? Again, the context changes the analysis – or at least it should.
Most importantly, defining morality as necessarily absolute and objective is an illegitimate way to privilege religious morality, even as it continues to become less and less useful to people living in a modern world. Not only less useful, but also potentially harmful, in that our moral sensibilities atrophy further every time we think a dilemma should not be debated – or cannot be debated – because of some absolute principle.
And just as religious morality should be considered part of our intellectual history, rather than inform our actions in the present, so too for much of what secularists call “Enlightenment reason”. While the Enlightenment project did the species an enormous favour at the time, our knowledge of cognition in the 18th century should not be considered a permanent guide to how to live. Reason, for Kant, was absolute and transcendent, but in reality it is probably no such thing.
There is no unified locus of rationality in the brain, and no disembodied faculty of reason – our will is tied to established habits, and also seems to follow after our emotional moral judgements, rather than generate them. As Jonathan Haidt suggests, the emotional dog wags the rational tail – we form emotive reactions, and then construct post-hoc rationalisations to justify them. John Dewey’s characterisation of the self as an “interpenetration of habits” should remind us that some of the very concepts of morality (like “person”) are metaphorical – they have no necessary and sufficient conditions.
For moral theory to make sense in light of what we now know about how we reason, it must take the work of people like Haidt, De Waal, Harris and Binmore into account. Of course, the answers are going to be messy, at least for now. But so are we, and so is the world we live in. If you don’t want to help with cleaning it up, so be it – but standing on the sidelines reciting dogma (religious or otherwise) is simply getting in the way of important work. And that’s surely not morally good, by any definition. DM