The humanity formula by hcj



The humanity formula. Notes for Philosophy 13 Lecture Dick Arneson
In section II Kant introduces three formulations of the Categorical Imperative principle (universal law,
humanity, and autonomy). He asserts the three formulations are equivalent in their implications for conduct
and may be viewed as alternate expressions of a single underlying principle. This is a striking claim
because the three formulations appear to be different in content.

Recall that a categorical imperative is a requirement of reason that binds unconditionally. It binds one
independently of any aim that one might or might not adopt. One route to the Categorical Imperative is by
arguing that just in virtue of being a rational agent, hence committed to acting on reasons that are general,
one is committed to the categorical imperative. This is supposed to yield the universal law formula. Another
route is to investigate whether there is an end that as a rational being one cannot disavow--an objective end,
in Kant's jargon. If there is, it could be the basis of a categorical imperative. On pages 36-37, Kant presents
an argument (which I rehearse below under the “regress argument” heading) to the conclusion that there is
indeed such an end. If we make a choice, we have made an evaluation, and evaluating something as good
requires an exercise of rational agency. If we must value what is necessary for our valuing, we must value
our rational agency. Rational agency is an end we cannot then disavow. This is supposed to yield a version
of the Categorical Imperative, the humanity formula: "So act that you use humanity, whether in your own
person or in the person of any other, always at the same time as an end, never merely as a means" (p. 38).
In other words, act always in such a way that you treat rational agency capacity, yours or that of any other
person, never merely as a means.

Kant's humanity formula has seemed to many a ringing declaration that we must always treat persons with
respect in virtue of their rational nature. What does this involve? I briefly indicate three alternative
construals of the meaning of the humanity formula.

1. Treating rational nature not merely as a means but also as an end is treating persons always
according to principles that qua rational (if they were fully rational) they would accept. On this
construal, I treat persons with respect just in case I treat them according to the principles that they would
accept if they were fully rational. If someone is confused, ignorant, or malicious, but I treat him according to
principles he could accept if he were rational, I do not treat him with disrespect.

Example: If the only way I can startle a small child in danger and save his life is by shattering your expensive
vase, I do not treat you as a mere means if I am treating you according to principles you could if fully rational
accept. The question then is whether "accept minor inconvenience to save a life" is a rational principle.
Kant might hold that this issue is settled by seeing whether the principle is universalizable according to the
universal law test. On this construal, it is easy to see why Kant believes that the two principles are aspects
of the same basic idea.
The humanity formula on this interpretation is unobjectionable but also unhelpful. It is unobjectionable
because it merely says “Act in conformity with correct moral principles.” It is unhelpful because it gives us
no help in figuring out what the correct moral principles are (if there are any).

2. Treating rational nature not merely as a means is refraining from treating persons in ways to
which they could not possibly give consent. This line is suggested by Kant’s comment, applying the
humanity formula to the false promising example, that “he whom I want to use for my purposes by such a
promise cannot possibly agree to my way of behaving toward him, and so himself contain the end of this
action” (p. 38). Some interpreters suggest that when judged by this standard, deception and coercion will be
revealed to be always impermissible. When I deceive someone, I treat that person in a way to which she
could not possibly consent, because she does not know what is occurring, and so it would be logically
impossible for her to consent to it. In a similar way, so it is claimed, if I am coerced, I cannot voluntarily
consent to what is occurring.

Quibble: Does the test on this construal really rule out deception and coercion? It seems I can possibly
consent to being deceived, as when I tell my doctor to keep me from learning the truth for as long as
possible if he diagnoses me with cancer. I think I can also imagine a case of consenting now to be deceived
now and then actually being deceived. Similar examples suggest one can consent to being coerced. If a
burly person challenges me to a fight in a bar and my wife holds me down, preventing me from fighting, it
can be true simultaneously that (a) I an coerced not to fight and (b) I consent to this coercion, I am glad it is
happening. --This quibble does not challenge the suggested test, but only disputes what would be the
results of applying it.
Objection to the test: The could-not-possibly-consent test seems too weak. I could possibly consent to
being assaulted even when I do not actually do so and have no good reason to consent. The mere fact that
it is possible that I could consent seems insufficient to render the act of assaulting me morally permissible.

Objection: the could-not-possibly-consent test also seems too strong. Maybe in given circumstances you
could not possibly consent to the deception that is necessary to save your life, but the deception is OK.

3. The third construal is offered by Thomas Hill in his essay “Humanity as an End-in-Itself.”
According to Hill, treating rational nature not merely as a means is treating people always in ways
that respect the unconditional and incomparable dignity of their rational nature.

Interpreters who follow this third line take their cue from Kant's distinction between dignity and price. Kant
distinguishes what has price from what has dignity. What has dignity is above price—is priceless, as we
might say. What has price can reasonably be traded for any other good that commands a price, as a certain
quantity of hay might be traded for a smaller pile of diamonds. What has dignity is beyond price. Kant says
that what has dignity has “an unconditional and incomparable worth.” On this basis Kant is interpreted as
maintaining both that what has dignity is incomparably more valuable than anything with price, so that one
should not accept the tiniest loss in dignity value in exchange for the greatest gain in what has price, and
also that what has dignity has a special nonaggregative value, so one cannot reasonably quantify dignity

Consider the maxim of committing suicide to obtain relief from pain. According to the incomparable-dignity
construal of the humanity formula, it forbids suicide done for this reason, which amounts to giving up rational
nature (continued life in which rational agency capacity can be exercised) for greater pleasure or lesser
displeasure. But this is in effect to exchange what has dignity for what has merely price value. Mutilating
oneself in a way that lessens one’s rational agency capacity to gain some conditional value is also forbidden
by the humanity formula in the same way. Hill gives other examples of the application of the humanity
formula. He writes, “ since the exercise of rationality is something to be cherished, in trying to influence
others one should appeal to their reason rather than try to manipulate them by nonrational techniques.”

The exchange argument. The general argument seems to go as follows:
1. One ought never to treat rational agency capacity in oneself or in others merely as a means (but rather
always also as an end).
2. If and only if one acts in a way that exchanges rational agency capacity for things that have value just
because one wants them, one exchanges dignity for what has mere price.
3. If and only if one exchanges what has dignity for what has mere price, one treats rational agency capacity
in oneself or in others merely as a means.
4. One ought never to exchange rational agency capacity for things that have value just because one wants

Comment: The idea that the categorical imperative commands each of us always to respect persons in
virtue of their possession of rational capacity and never to act in ways that express wrongfully low valuation
of rational agency is attractive and appealing. But does rational agency have a value incomparably greater
than the values labelled goods that have only price? Why think this? Does Kant have any argument for this
claim? An alternative view is that rational agency has large but not incomparably greater value. Consider a
choice to engage in a risky sport for fun even though the activity increases one’s chances of suffering an
injury that destroys or damages one’s rational agency capacity. If this is reasonable, then having rational
agency capacity is not incomparably more valuable than having fun. Or consider going across town,
accepting a slightly elevated risk of suffering death in a car crash and so slightly increasing one’s risk of
premature loss of rational agency capacity, in order to get tastier bread than one can obtain in one’s home
neighborhood. Or consider becoming drunk at a wedding, to be festive, in safe surroundings. One
sacrifices some amount of time in the possession of full rational capacity in exchange for conviviality. If
these activities can be permissible, then the humanity formula as interpreted by Hill is delivering implausible
verdicts. One might respond to the exchange argument by making two objections: (1) Some things that
have value just because one wants them may, in sufficient quantity, outweigh a given quantity of dignity
value. (2) Anyway some things other than rational agency capacity have value independently one whether
or not one happens to want them—accomplishment, friendship, knowledge, and so on. Exchanging or
trading off rational agency capacity for other values is not necessarily immoral.

Consider also the idea that dignity value cannot be quantified. Suppose one has to choose between
preventing one person from being injured in a way that damages rational agency capacity and preventing a
thousand persons from suffering the same injury. If dignity values cannot be aggregated, it seems one
cannot say 1000 persons losing rational agency is worse than one. To some, the notion of an incomparable
and nonquantifiable value, a value to be respected not promoted, sounds fishy.

Even if the idea that rational agency capacity has incomparable and unconditional value has counterintuitive
consequences, we should recognize that Kant argues for the claim. If the argument is sound, ”we” (that is,
those of us who go along with the objections so far advanced) would have to revise our judgments. The
argument appears in the Groundwork, Section 2, on page 37. Kant writes, “All objects of the inclinations
have only a conditional worth; for, if there were not inclinations and the needs based on them, their object
would be without worth. But the inclinations themselves, as sources of needs, are so far from having an
absolute worth. . .Thus the worth of any object to be acquired by our action is always conditional. . . .
Rational beings, therefore, are not merely subjective ends, the existence of which as an effect of our action
has a worth for us, but rather objective ends, that is, beings the existence of which is in itself an end, and
indeed one such that no other end, to which they would serve merely as means, can be put in its place,
since without it nothing of absolute worth would be found anywhere; but if all worth were conditional and
therefore contingent, then no supreme practical principle for reason could be found anywhere.”

This passage is hard to interpret. It seems to say that, on the assumption that there is a supreme principle of
morality, then it must be the case that rational nature (rational agency capacity) has unconditional and
incomparable worth. This leaves open the possibility that there is no supreme principle of morality—for all
we know, morality might be bunk, a big hot air balloon that skeptical reason can puncture and deflate.

The argument might alternatively, or additionally, claim that when we act, we must assume that what we
pursue is objectively worthwhile, but if this is assumption is true, that rational nature (rational agency
capacity) has unconditional and incomparable worth. Here’s an attempt to state the argument; the Kant
commentator Christine Korsgaard calls this the “regress argument”:

1. Whenever one (a rational agent) acts, one presupposes that what one seeks is valuable.
2. What one seeks is only conditionally valuable.
3. What is conditionally valuable is so only if its conditions are met.
4. If those conditions are only conditionally valuable, then ultimately we must reach unconditionally valuable
conditions, or what one seeks is not valuable.
5. One’s inclinations are not unconditionally valuable.
6. Nothing but one’s inclinations or one’s rational nature (the source of one’s valuations and inclinations)
could be unconditionally valuable.
7. So, whenever one acts, one presupposes one’s rational nature is unconditionally valuable.
8. What one proposes to be unconditionally valuable, one must presuppose to be incomparably valuable.
9. So, whenever one (a rational agent) acts, one presupposes that one’s rational nature is unconditionally
and incomparably valuable.
10. One’s rational nature can be unconditionally and incomparably valuable if and only if anyone’s rational
nature is unconditionally and incomparably valuable.
11. So whenever one ( a rational agent) acts, one presupposes that one’s rational nature is unconditionally
and incomparably valuable and that anyone’s rational nature is unconditionally and incomparably valuable.

So interpreted, the argument is not compelling. First, I don’t see why one must presuppose when one acts
that what one seeks is valuable. One might believe nothing is objectively valuable, one just wants the thing
and the action chosen will get one what one wants. Second, the conditions for something’s being valuable
need not themselves be valuable at all, much less incomparably valuable. This is so whether the conditions
in question are causal or conceptual. A causal condition for your winning the tennis game is that tennis balls
are available but there need not be anything objectively valuable per se about tennis balls being available.
A conceptual condition of your being virtuously courageous is that you are capable of having character traits
but there might be nothing objectively valuable per se about your being capable of having character traits.
Third, premise 6 looks to be an arbitrary assertion. Why are possibilities mentioned the only possibilities?
The regress argument is not convincing, and the conclusion Kant draws from it is counterintuitive, so we lack
reason to accept the conclusion—that rational nature is unconditionally and incomparably valuable (or that, if
we are rational, we have to assume this is so).

Further comment: Some commentators have suggested that if we interpret Kant’s Categorical Imperative
principle as a test for deciding what maxims it is morally OK to act on, we run into problems, some of which
have been canvassed in these notes. If we interpret Kant’s Categorical Imperative principle as giving us a
procedure to determine what our moral duties are, e.g. that lying is wrong or that we morally ought not to
borrow money on the basis of a false promise to repay, we run into problems. These problems are serious.
Perhaps we should think of the point of the CI procedure in a different way. One proposal is that the
supreme principle of morality is that we should act only in ways that express proper respect for the
unconditional and incomparable value of rational agency capacity, but what this requires in practice cannot
be set forth in any determinate principles. We just have to figure out in each particular context what respect

for humanity (rational nature) requires in that context. The norm of respecting rational nature then is
functioning somewhat like W. D. Ross’s list of prima facie duties—there is no formula according to Ross that
specifies what, given these duties, we ought to do in any particular circumstances. {{Don’t worry about the
references to the author Ross—he is not a course author.}} In some ways this way of reading the CI is close
to construal #1 of the humanity formula plus the additional claim that no other formula of the CI supplies
determinate content as to what we morally ought to do. On this latter view, the CI is a way of conceiving of
morality and not a substantive principle that specifies or determines what we ought to do. On the “express
respect for humanity” reading, in contrast, the CI is substantive but interpreting its requirements always
requires additional evaluation case by case.

The big picture. Hill presents a sympathetic interpretation of Kant’s humanity formula. In the end he
rejects Kant’s position as he interprets it. He says Kant’s position “reflects an extreme moral stand that few
of us, I suspect, could accept without modification.” If what Kant asserts, unmodified, is unacceptable, what
modifications are needed? Hill suggests that rational agency capacity is a great value but not incomparably
greater than other values. Another possibility is that rational agency capacity by itself is not a value but
rather a potential for value and maybe an enhancer of value. Having rational agency capacity, and a life to
live, you have an opportunity to gain lots of goods for self and others. Moreover, that a being has rational
agency capacity perhaps enhances the value of the goods it gets: maybe your getting one unit of pleasure is
morally more valuable than a dog or a goldfish getting one unit of similar pleasure. But the mere existence
of rational agency capacity in a being does not per se generate any value or good; just staying alive then is
morally neutral, neither good not bad. On this view, suicide for pain relief might be morally acceptable or
morally unacceptable, depending on what possibilities of life cutting short one’s own life right now eliminates.
A pouting young man committing suicide to relieve his upset at a temporary disappointment squanders the
valuable opportunity for good life as a rational agent and is doing what is morally wrong whereas a person
with terminal cancer, untreatable severe pain, and no chance of getting any significant goods by staying
alive is arguably acting in a way that is morally permissible and even admirable. Another example: Consider
a mother suffering from severe depression that antidepressants cannot relieve. Each day as she
experiences it is a horrible bleak gray blank. Here perhaps suicide for pain relief would be excusable, and
morally permissible, even though the death of the mother is bad for her child.

I have criticized some details of Kant’s Categorical Imperative principle. But in a broader sense, Kant could
be right about the nature of morality for all that I have said. He claims that if morality is not a sham, then we
are subject to categorical imperatives—commands of reason that bind us independently of what we happen
to want or choose. These commands ultimately come down to principles—general claims about what we
ought to do—that are universal (they hold without exception at all times and places), necessary (their
authoritative force does not depend on contingent features of our circumstances, features that might or
might not be the case), and supremely authoritative (moral reasons take priority, they trump any other
reasons there may be). Just in virtue of being rational agents, we have the power to think about what to do
from an impartial perspective that does not give any more weight to one person’s interests than to anyone
else’s. What principles emerge as categorical imperatives from this perspective? Common rules such as tell
the truth, keep your promises, respect private property, don’t kill the innocent, and the like cannot be
(contrary to Kant’s own opinion) categorical imperatives, because these rules all admit of exceptions.
Sometimes one ought to lie, break one’s promise, steal, or kill the innocent. On the other hand,
consequentialist principles such as “One ought always to do whatever would bring about the best outcome
impartially assessed” seem extremely demanding. Kant’s thought seems to be that moral principles while
supremely authoritative need not be imperialistic or all-encompassing. From an impartial perspective that
treats all the same, maybe there are good impartial reasons to accept principles that allow each person to
be partial sometimes to herself and to those near and dear to her. So maybe the supreme principle of
morality is some relaxed version of consequentialism, such as, “One ought always to act in whatever way
would bring about an outcome that is close enough to the best outcome one could have brought about
instead.” So long as the shortfall between what one brings about and the best one might have brought
about is not excessive, one is permitted to put a thumb on the scale in favor of one’s own interests and
concerns in deciding what to do. Another possibility, following hints given by Amartya Sen in “Rights and
Agency,” is that the supreme principle of morality is a balancing rights principle, such as “Always act in such
a way that properly balances respecting all people’s moral rights and promoting the fulfillment of all people’s
moral rights, with rights weighted by their importance and the proper balance allowing some favoring of
oneself.” To get a specific proposal from these suggestions, one would have to specify the allowable
shortfall range in relaxed consequentialism or the correct weights to be assigned to different moral rights
and the limits of favoring oneself in the Senian principle. These are just examples of how one might pursue
the Kantian project broadly construed.

To top