So far, since, as I mentioned at the outset, this short course has a broadly Kantian orientation, I’ve done my level best to present Kant’s ethics as charitably and effectively as possible.

But now we must also consider three classical worries about it:

(i) a worry about the intelligibility and workability of Kant’s moral criterion of universalizability,

(ii) a worry about the over-strictness and excessive universality of Kantian moral principles, aka Kant’s rigorism, and

(iii) a worry about contradictory conflicts between so-called “strict” or “perfect” duties, aka moral dilemmas.

Let’s look at these worries one-by-one.

1. The intelligibility and workability of Kant’s moral criterion of universalizability.

As I mentioned above, The Formula of Universal Law says that nothing will count as a morally permissible maxim (i.e., a first-person-centered principle of volition, or act-intention) unless it consistently generalizes.

A prima facie problem in this connection is that there are troublesome maxims, of at least three different kinds.

First, consider the maxim,

I ought to play tennis on Sundays when everyone else is in church,

and then consider its generalization,

Everyone ought always to play tennis on Sundays when everyone else is in church.

This seems morally perfectly acceptable, yet it’s not consistently generalizable because then everyone would be on the tennis courts, instead of in church, which would make it impossible for everyone to play tennis.

So not every morally permissible maxim is consistently generalizable.

Second, consider the maxim,

I ought to be the last person off a sinking ship,

and then consider its generalization,

Everyone ought always to be the last person off a sinking ship.

This again seems not only morally perfectly acceptable but also even in some contexts morally obligatory — e.g., if you’re the captain of the ship, or even if you’re not the captain but everyone else on the ship is a sick person or otherwise especially vulnerable — yet it’s not consistently generalizable, because then no one could ever be the last person if everyone were always trying to be the last person.

So again not every morally permissible maxim is consistently generalizable, but even worse, some maxims that in some contexts are morally obligatory, aren’t consistently generalizable.

Third, consider the maxim,

I ought to be an egoistic liar in a world populated exclusively by egoistic liars,

and then consider its generalization,

Everyone ought to be an egoistic liar in a world populated exclusively by egoistic liars.

Strictly speaking, this is consistently generalizable — we’d simply all be (un)happy little egoistic liars together — but obviously it’s not morally permissible.

So some consistently generalizable maxims are morally impermissible.

How could Kant, or anyhow a Kantian, reply to the troublesome maxim problem?

The obvious problem with the troublesome maxims is that the formulation of the worry presupposes something that’s false, namely, that for every maxim there is one and only one relevant generalization.

In fact, on the contrary, for every maxim there are at least two substantively different — and in particular, logically non-equivalent — ways of generalizing them.

For convenience, let’s call that The Law of Multiple Generalizations.

So, holding The Law of Multiple Generalizations before us, let’s look again at the first troublesome maxim,

I ought to play tennis on Sundays when everyone else is in church,

and now consider a non-equivalent alternative generalization of it,

Anyone, in a world populated by lots of other people who are Sunday church-goers, and also by a few non-church-going tennis enthusiasts, ought always to play tennis on Sundays with some other non-church going tennis enthusiasts, on one or more of the tennis courts that’s left empty by all the church-goers,

which is perfectly logically and conceptually consistent.

So now all we need to do, in view of The Law of Multiple Generalizations, is to assert the general thesis that a maxim is consistently generalizable if and only if at least one of its relevant generalizations is logically and conceptually consistent, and that neatly avoids the prima facie difficulty raised by that maxim.

Now let’s look again at the second troublesome maxim, which says

I ought to be the last person off a sinking ship,

and consider the following non-equivalent alternative generalization of it,

Anyone, in a world populated by a sinking ship and other people (including crew, passengers, stowaways, etc.) on the ship who all desire not to drown and therefore want to get off that ship as quickly as possible, ought always to be the last person off that ship,

which is also perfectly logically and conceptually consistent.

Therefore, the second maxim also falls directly under our general thesis that a maxim is consistently generalizable if and only if at least one of its relevant generalizations is logically and conceptually consistent, and the prima facie difficulty with it is also neatly avoided.

Finally, let’s look again at the third troublesome maxim,

I ought to be an egoistic liar in a world populated exclusively by egoistic liars,

and consider the following non-equivalent alternative generalization of it,

Anyone who’s really a non-egoistic truth-teller, even though he may sometimes be strongly tempted to act like an egoistic liar, ought to live in a world populated exclusively by people who really are egoistic liars,

which is clearly logically and conceptually inconsistent, since by hypothesis he’s not an egoistic liar but on the contrary a non-egoistic truth-teller, yet the world he would supposedly live in contains nothing but people who really are egoistic liars.

Correspondingly, again in view of The Law of Multiple Generalizations, all we need to do is to hold the general thesis that a maxim is not consistently generalizable if and only if at least one of its relevant generalizations is logically and conceptually inconsistent, and that also neatly avoids the prima facie problem raised by that maxim.

Another prima facie problem about Kant’s moral criterion of universalizability concerns its epistemic indeterminacy.

The epistemic indeterminacy problem has two sub-problems.

First, Kant says explicitly that some maxims that turn out not to be consistently generalizable, do so by virtue of the fact that their generalization leads to a “contradiction in conception,” which means that they’re logically or conceptually inconsistent, whereas other maxims that turn out not to be consistently generalizable do so by virtue of the fact that their generalization leads to a “contradiction in willing,” which aren’t logical or conceptual contradictions.

But what’s the difference between the two kinds of contradiction?

If we can’t spell out the difference clearly and distinctly, then the moral criterion of universalizability is epistemically indeterminate, i.e., then we can’t know how to apply the criterion.

Let’s look at some examples.

If we consider the maxim,

I ought to be an egoistic liar in a world populated exclusively by non-egoistic truth-tellers,

and then consider its generalization,

Everyone ought always to be an egoistic liar in a world populated exclusively by non-egoistic truth-tellers,

then that’s obviously logically or conceptually inconsistent, aka it’s a contradiction in conception: for if anyone were an egoistic liar in that world, then there wouldn’t be nothing but non-egoistic truth-tellers in that world, but that’s contrary to hypothesis, which says that the world has to be filled exclusively with non-egoistic truth-tellers.

Now let’s look at the maxim,

I ought to spend my life refusing charity to others in a world populated by many other people who are needy and vulnerable, including being mortally threatened,[i] even though I’m comparatively well-off,

and then consider its generalization,

Everyone ought always to refuse charity to others in a world populated by many other people who are needy and vulnerable, including being mortally threatened, even though they (i.e., the charity-refusers) are comparatively well-off,

then there’s clearly no logical or conceptual contradiction in that: but what does it mean to say that it’s a “contradiction in willing?”

Here’s what.

Given the generalization of that maxim, it’s obvious that I’m laying a fatal trap for myself, practically speaking: for insofar as I myself ever become needy and/or vulnerable, including being mortally threatened, and want someone else’s charity, then I’m up the creek without a paddle.

In other words, I’m saying,

Help me, I’m starving! (or whatever), but no one should ever provide charity.

Therefore, Kant’s contradictions in willing are simply what’s nowadays called pragmatic contradictions, which are not logical or conceptual inconsistencies within the explicit content of what’s said, but instead are inconsistencies between what’s implicit in my act of utterance and the explicit content of what’s said by me — in this case my plea for charity for myself, together with an explicit ban on all charity.

Second, another epistemic indeterminacy problem arises when we realize that there are different descriptions of the same human body-movement with categorically different moral values, such that under one description, that body-movement is morally permissible, but under the other description, it’s morally impermissible.

E.g., suppose that I suddenly move my arm in such a way that my clenched hand makes forcible contact with another physical object.

So, corresponding to that, consider the maxim,

I ought to move my arm suddenly in such a way that my clenched hand makes forcible contact with another physical object.

Now clearly, any number of generalizations of that would be logically or conceptually consistent, and therefore morally permissible.

But what about the maxim,

I ought to sucker-punch some innocent person,

which also correctly describes the very same human body-movement?

Clearly, that’s morally impermissible, and it’s easy to formulate a generalization of it that’s either logically or conceptually inconsistent, or a pragmatic contradiction.

So how are we to know which is the relevant maxim?

Here I think that we can follow the same overall strategy as in the case of the troublesome maxims, and note that the problem arises if and only if we make the false presupposition that for every human body movement there is one and only one relevant maxim.

In fact, on the contrary, for every human body-movement there are at least two substantively different — and in particular, logically non-equivalent — ways of describing them in maxims.

For convenience, let’s call that The Law of Multiple Maxims.

And now, in view of The Law of Multiple Maxims, all we need to do is to assert the general thesis that a given human body-movement is morally impermissible if and only if at least one of its relevant maxims is such that at least one its relevant generalization is either logically and conceptually inconsistent, or pragmatically inconsistent, and that also neatly avoids the second prima facie problem about epistemic indeterminacy.

2. The over-strictness and excessive universality of Kantian moral principles, aka Kant’s rigorism.

Quite apart from Kant’s or Kantian moral theory, looking at the various facts and phenomena of human moral life, we’ll naturally wonder whether it’s always wrong to lie?

And in all likelihood, we’ll answer: no, it’s not always wrong to lie, citing, e.g., so-called “white” or trivial lies that harm no one, and also “polite” lies that spare people’s feelings.

Indeed — as we’ll see shortly — there are also some actual or really possible situations in which, arguably, it’s morally obligatory to lie.

Similarly, we’ll naturally wonder whether it’s always wrong to commit suicide?

And correspondingly, in all likelihood, we’ll answer: no, it’s not always wrong to commit suicide, perhaps citing cases like that of Stephan and Edith Körner.

Indeed, there may well also be some actual or really possible situations in which it’s morally obligatory to commit suicide, perhaps citing a minor variant on Bernard Williams’s case of Jim the South American Traveller, mentioned in section V.2.2, which I’ll call Jim the South American Traveller Redux.

In Jim the South American Traveller Redux, again for reasons of life-defining personal integrity, Jim chooses to commit suicide (perhaps by putting himself forward to be shot by the cruel police officer, hence a case of suicide-by-police), instead of being forced to kill one Indian, even in order to save nineteen other Indians.

In view of such and similar answers and examples, this is sometimes also called the problem of Kant’s rigorism.

One simple response that Kant, or at least a Kantian, can make to this prima facie difficulty is, as I’ve noted above, that Kant’s terminology is extremely misleading and that Kant’s so-called “strict” or “perfect” duties are in fact only blah blah, period duties, namely, moral obligations for us to consider them in every situation in which it’s really possible for human persons to choose and act.

And that’s perfectly consistent with its being at least sometimes morally permissible, or even morally obligatory, to lie or commit suicide.

Indeed, now looking back at our discussion of moral relativism in section II.1, the Kantian can also quite consistently say that for every blah blah, period duty, there is a fairly universal and objective moral principle corresponding it, as per the following two principles:

(i) other things being equal (i.e., ceteris paribus), it’s always wrong to lie, but if other things aren’t equal in some historical context, then it’s really possible for it to be either morally permissible or even morally obligatory to lie, and

(ii) other things being equal (i.e., ceteris paribus), it’s always wrong to commit suicide, but if other things aren’t equal in some historical context, then it’s really possible for it to be either morally permissible or even morally obligatory to commit suicide.

So the prima facie worry about Kant’s rigorism can be avoided.

3. Contradictory conflicts between so-called “strict” or “perfect” duties, aka moral dilemmas.

So-called “strict” or “perfect” duties, in historical contexts, sometimes lead to moral dilemmas, aka moral contradictions, and (so goes the worry) any moral theory that leads to moral dilemmas or contradictions must be rejected, therefore Kant’s moral theory must be rejected.

Consider, e.g., the famous fictional case of The Murderer at the Door, that Kant himself considered in a philosophically unfortunate 1797 essay called “On a Supposed Right to Lie from Philanthropy,” which discusses whether

it would be [morally impermissible] to lie to a murderer who asked whether a friend of ours whom he is pursuing has taken refuge in our house,[ii]

but in more recent versions, the moral thought-experiment concerns whether it’s morally impermissible to lie to a Nazi murderer who shows up at your door and demands to know whether Anne Frank is hiding in your house (and she is), hence it’s sometimes called The Nazi at the Door.

In any case, the prima facie moral contradiction between so-called “strict” or “perfect” duties is between, on the one hand, the duty to tell the truth, and on the other, the duty to protect rational human life — in this case, someone else’s life.

Another example to consider is real-world, and concerns Cassie Bernall, an 18 year old born-again Christian high-schooler, who, during The Columbine High School Massacre in 1999 (reportedly — twenty years later, this remains highly controversial), was asked by the mass murderers Eric Harris and Dylan Kliebold whether she believed in God, and then was shot and killed by them when she said “yes.”

In 2000, Bernall’s mother published a book called She Said Yes: The Unlikely Martyrdom of Cassie Bernall — so I’ll call this case The High School Massacre Martyr.

Like the case of Stephan and Edith Körner, the case of The High School Massacre Martyr has special emotional force and poignancy for me.

When I was teaching one version of this course in Fall 1999 and discussing The High School Massacre Martyr in class, one of my students suddenly burst into tears, completely distraught and clearly traumatized: then, sobbing uncontrollably, she told us all that she’d been a classmate of Bernall’s at Columbine, and a friend of hers, and had only narrowly escaped being killed herself that day.

For me and the other students, it was one of those real-world, Thomas-Whitaker-like, epiphanies about morality and the human condition, and, to put it mildly, a class-stopper.

In the case of The High School Massacre Martyr, again, the prima facie contradiction between so-called “strict” or perfect duties is between, on the one hand, the duty to tell the truth, and on the other, the duty to protect rational human life.

But in this particular case, even leaving aside its emotional force and poignancy,

not only (i) does the issue of life-defining personal integrity come into play, as in Jim the South American Traveler Redux, hence also the real possibility of Bernall’s death being a case of suicide-by-mass-killers,

but also, (ii) because in this case the duty to protect rational human life obviously applies directly to Bernall’s own life, the issue of what I’ll call a moral right to lie in self-defense also comes into play.

What could a Kantian (other than the 76 year old Kant, that is, who already unfortunately had his say in “On A Supposed Right to Lie from Philanthropy”) say in response to this worry in general, and to these very hard cases in particular?

Common sense moral judgment clearly and distinctly says that in The Murderer at the Door/The Nazi at the Door, it’s not only morally permissible, but perhaps even also morally obligatory to lie to the murderer/Nazi murderer.

And common sense moral judgment also says, although perhaps slightly less clearly and distinctly, that in The High School Massacre Martyr, it’s not only morally permissible for Bernall to lie in self-defense, but also morally permissible for her to tell the truth for reasons of life-defining personal integrity, and perhaps even also morally obligatory for her to do so, given her born-again Christianity.

I myself am with common sense moral judgment on these claims.

But the crucial thing for us to see, is that all of those common sense moral judgments are fully consistent with the Kantian responses to the moral rigorism worry that I sketched above, as extended to the moral dilemmas worry, according to the following argument.

1. Kant’s terminology is extremely misleading and Kant’s so-called “strict” or “perfect” duties are in fact only blah blah, period duties, namely, moral obligations for us to consider them in every situation in which it’s really possible for human persons to choose and act.

2. That’s perfectly consistent with its being at least sometimes morally permissible, or even morally obligatory, to lie or commit suicide.

3. For every blah blah, period duty, there is a fairly universal and objective moral principle corresponding to it, and three of these fairly universal and objective moral principles are:

(3a) other things being equal (i.e., ceteris paribus), it’s always wrong to lie, but if other things aren’t equal in some historical context, then it’s really possible for it to be either morally permissible or even morally obligatory to lie,

(3b) other things being equal (i.e., ceteris paribus), it’s always wrong to commit suicide, but if other things aren’t equal in some historical context, then it’s really possible for it to be either morally permissible or even morally obligatory to commit suicide, and

(3c) other things being equal, it’s always morally permissible to act in self-defense, including telling a lie in order to save oneself from being murdered, but if other things aren’t equal in some historical context, even when acting in self-defense and lying in order to save oneself from being murdered are indeed morally permissible, then it’s also really possible for it to be either morally permissible or even morally obligatory to tell the truth and be a martyr.

4. Therefore, Kant’s or at least Kantian moral theory can smoothly accommodate all cases like The Murder at the Door/The Nazi at the Door, and The High School Massacre Martyr without leading to moral contradictions, and the worry about moral dilemmas is thereby avoided.[iii]

V.4 All-Things-Considered Conclusion of This Section

Looking back now at our critical discussions of Aristotelian virtue ethics, Millian utilitarianism, and Kant’s ethics of persons and principles, all things considered, I conclude that Kant’s ethics of persons and principles is on the whole superior to Aristotelian virtue ethics and Millian utilitarianism alike.

We’ve already gone through the basic arguments for and against the three classical moral theories, so I won’t reprise those.

The essential reason for the on-the-whole superiority of Kant’s ethics of persons and principles, which I think isn’t generally noticed, is this:

All the standard objections to Aristotelian virtue ethics and to Millian utilitarianism can be fully conceded, and at the same time the core truths of those views (i.e., the moral significance of virtue, happiness, benefits-based moral reasoning, and consequences) can all be accommodated and included, as fairly universal and objective moral principles, within an overall Kantian moral theory of persons and principles.

— Of course, you’re perfectly free to disagree!, and, as reasonable people, we can also freely agree to disagree, as a “normal condition” of philosophical discussion.

So let’s move on.

NOTES

[i] Here, of course, we can think back to Singer’s and Unger’s argument for famine relief, and our critical analysis of it, in section III.4.

[ii] I. Kant, “On a Supposed Right to Lie from Philanthropy,” trans. M. Gregor, in I. Kant, Immanuel Kant: Practical Philosophy, pp. 611–615, at p. 611 (SRL 8: 425). It’s philosophically unfortunate for two reasons. First, in that essay, Kant doesn’t in fact consider whether it would be morally permissible to tell a lie in that context per se, instead he actually considers whether it would be morally permissible to tell the lie for consequentialist (and in effect Utilitarian) reasons alone, which is an importantly different question. And second, in that essay, the 76 year old Kant (he died in 1801) clearly shows himself to be, at least in his old age, a moral fanatic about truth-telling.

[iii] For purposes of philosophical full disclosure, I should also mention that in Kantian Ethics and Human Existence: A Study in Moral Philosophy (THE RATIONAL HUMAN CONDITION, vol. 3) (New York: Nova Science, 2018), ch. 2, available online in preview HERE, I develop a somewhat different, five-step response to the worry about moral dilemmas, which is (i) to reject a presupposition of the worry, namely that any moral theory that leads to moral dilemmas or contradictions must be rejected, (ii) to accept the existence of “local” moral dilemmas or contradictions in some historical contexts, (iii) show how the moral dilemmas or contradictions that arise in such contexts don’t logically affect Kant’s absolutist moral generalism, which applies only to The Categorical Imperative, (iv) show in particular how the four (or five) formulations of The Categorical Imperative are all generally objectively valid and non-contradictory even when we allow for “local” moral dilemmas or contradictions, and finally (v) show how this version of the Kantian theory of moral principles captures certain profound existential insights about our “human, all-too-human” moral lives. But all that’s too longwinded, somewhat philosophical strenuous, and therefore, again, “too much like hard work,” for a short course.