. Luammxkfiym
.-r r .
3.0.3559
.1: _ ...

.fimfif: .. “mm

x

 

4. ,
2.

it:

I I.

Md. .. . .9...”
I.

.1 ......, . . Ln. . u
Wwawwfidmm
J4 ‘

.Jmfipfl.

6...? we:
is ,

‘ .nwurfi...

fig. V ‘ . 1%, _ , .2 ._ fiéfififiiy

 

, LIBRARY i
Michigan State
University

 

 

 

This is to certify that the
dissertation entitled

RECONSIDERING THE STATUS OF ANIMALS IN KANT‘S
ETHICS

presented by

HEATHER MARIE FIELDHOUSE

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Philosophy

 

 

MM

Major Professor’s Signature

flat/q 3/, 309’

Date

MSU is an Affirmative Action/Equal Opportunity Institution

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE
2b?1911

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm

RECONSIDERING THE STATUS OF ANIMALS IN KANT‘S ETHICS
BY

Heather Marie Fieldhouse

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Philosophy

2004

ABSTRACT
RECONSIDERING THE STATUS OF ANIMALS IN KANT'S ETHICS
BY

Heather Marie Fieldhouse

Kant claims that we have a duty to avoid inflicting unnecessary suffering
on animals because doing so damages our own humanity and hardens us to
human suffering. I argue that this theory of indirect duties to animals is weak
because it rests on an empirical claim that may not be true, and because it
encourages us to exploit what Kant sees as a mistaken analogy between
humans and animals. when it would be more consistent for Kant to advise us not
to make this mistake at all. Since Kant's indirect-duty theory fails, Kant's
followers must choose between accepting a theory which permits no duties with
regard to animals at all (which, though consistent, violates common intuition) or
constructing a new theory within the Kantian framework. I attempt to do the
latter. First, I analyze Kant's doctrine of animal minds, showing that although
Kant regards animals as possessing sensibility, he denies that they have

understanding or reason. I suggest that the evidence points to animals having at

least understanding and possibly some prudential reason, but that they do not
possess reason in the full sense required for moral agency. I argue that animals,
although they do not set their own ends and thus cannot be regarded as ends-in-
themselves, do have ends that are given to them by nature. As beings with
ends, they stand between mere things that have no ends, and rational beings
that are ends-in-themselves. I propose a broader version of Kant's kingdom of
ends, in which rational beings respect the ends of all other beings that have
them, including animals and other moral patients. The moral status of animals
would still be dependent on the existence of rational beings (without moral
agents, there would be no duties), but our duty to take their ends into account

would be a direct duty to them, rather than being a covert duty to human beings.

In memory of the General

iv

ACKNOWLEDGMENTS

I would like to thank Fred Flauscher for his support, assistance, and insight;
Martin Benjamin, Richard Peterson, and Winston Wilkinson for their helpful
suggestions; my husband, Christopher Fieldhouse, for his patience and
encouragement; and my friends in the philosophy department, for their good

cheer and companionship.

TABLE OF CONTENTS

CHAPTER 1

INTRODUCTION AND LITERATURE REVIEW .......................... 1

1. Overview ................................................. 1

2. Contemporary Discussion of the Indirect Duty View. . . . ........... 2

3. Other Kantian Views ....................................... 16

4. Looking Forward .......................................... 24
CHAPTER 2

KANT AND ANIMAL MINDS ........................................ 26

1. Introduction .............................................. 26

2. Animals, Sensibility, and Understanding ....................... 27

3. Animal Sensation and Consciousness ........................ 34

4. Reason in Kant .......................................... 40

5. Animals and Reason ...................................... 55

6. Expanded Kantian View .................................... 59

7. Conclusion .............................................. 77
CHAPTER 3

RATIONALITY AND MORAL AGENCY ................................ 78

1. Introduction .............................................. 78

2. Reason and Morality ....................................... 79

3. Reason and the Will ....................................... 81

4. Autonomy ............................................... 85

5. Rational Nature as an End in Itself ............................ 87

6. Animals, Reason, and Morality ............................... 91

7. Other Nonrational Beings ................................... 96

8. Conclusion ............................................. 102
CHAPTER 4

ANIMALS AS SUBJECTS IN THE KINGDOM OF ENDS ................. 103

1. Introduction ............................................. 103

2. Korsgaard on Animals ..................................... 105

vi

3. Members, Subjects, and Things in the Kingdom of Ends .......... 110

4. Animals Have Interests .................................... 117
5. Animals and Happiness ................................... 120
6. The Moral Status of Animals as Subjects ...................... 133
7. Objections and Replies .................................... 137
8. Future Directions and Conclusion ............................ 144
WORKS CITED ................................................. 148

vii

1. INTRODUCTION AND LITERATURE REVIEW

1. Overview

Kant explicitly denies that we have any direct duties to animals, so any
attempt to account for actual moral duties to animals (rather than indirect duties
that are actually to humans) will not be completely faithful to Kant. Kantians who
wish to incorporate animals more fully into their moral philosophy while still
remaining mostly faithful to the spirit of Kant can take two different general
approaches. First, and most strictly Kantian, they can try to defend Kant’s own
theory of indirect duties to animals, and perhaps show that it is applicable to
some of the current concerns about human treatment of animals. This approach
has the advantage that Kant's moral philosophy can be accepted more or less
intact, but the disadvantage that certain common intuitions are violated. Second,
they can attempt to show that animals do possess some qualifications which
admit them into the moral domain, possibly as moral patients but not moral
agents (whereas Kant held that only moral agents are moral patients). This

approach has the advantage of giving animals some intrinsic moral worth (and

thus better accounting for common intuitions), but the disadvantage that it will

require a greater revision of Kant’s position.

2. Contemporary Discussion of the Indirect Duty View

Kant holds that we have no duties to animals, but we have indirect duties
with regard to them. We ought not treat them cruelly, as it damages our natural
sympathies and thus can harden us in our dealings with other human beings.
Thus our duties with regard to animals are actually duties to human beings. We
may kill animals if it is done humanely, and put them to work if it does not strain
their capacities; in other words, we can use them as means to our ends, but we
must avoid being cruel as we do it (6:443, 272458-460).

A Kantian indirect-duty approach to the moral status of animals has
enjoyed a recent revival. Dan Egonsson, for instance, tries to show that it can go
beyond just the basic prohibition against wanton cruelty; Egonsson believes it
can be used to defend ethical vegetarianism.

According to Kant, being cruel to animals tends to make a person also
insensitive to his fellow man; that is why apparent duties to animals are actually

indirect duties, since ultimately they are duties to mankind. This does not seem

to apply to meat-eating, since it is possible to eat meat without being involved
oneself in the raising and slaughtering of the animal; in fact, most people are very
distanced from this process. Egonsson, however, writes that we can plausibly
extend Kant's remarks to also encompass accepting cruelty to animals (477).
Anyone who eats intensively-farmed meat is implicitly accepting cruelty to
animals. (Egonsson uses the example of intensively-farmed meat presumably
because humanely-raised animals which are killed painlessly would not be
suffering; since animals are only means, there would be nothing wrong with
killing them for food provided there is no cruelty involved with the raising and
slaughtering. Nevertheless, since most of the meat which is readily available
probably does not meet this ideal standard, Egonsson’s extension of Kant’s
position would tend to lead to vegetarianism.)

Egonsson decides that if accepting cruelty to animals does have a
desensitizing effect, then one cannot universalize the principle of eating meat.
“People might become less sensitive to your suffering - a consequence that is
probably undesired by you" (477). This is not a logical contradiction, but it does
seem to be a contradiction in will. Vegetarianism is thus supported by the

universal law formulation of the categorical imperative.

Egonsson then goes on to consider another formulation of the categorical
imperative, the formula of humanity. He considers a claim made by Tom Regan
that whereas the universal law formulation shows that vegetarianism is morally
required, the formula of humanity provides no guidance on the issue of
vegetarianism. Regan claims that this shows that the two formulations are not
equivalent. Egonsson argues that, given the desensitizing effect of allowing
animal suffering, both formulations equally promote vegetarianism: “Brutalization
is also undesired if we have the ambition to treat humanity as an end, and
therefore a moral case for vegetarianism will follow from both the formulations”
(48).

Allen Wood defends what he calls Kant's logocentrism - the grounding of
all duties in the value of rational nature - but takes issue with Kant's adoption of
what he calls the ”personification principle," which is the idea that to respect
rational nature is to respect it in someone's person (196). Wood's response is
that rational nature, including fragments of it and even preconditions of it, should
be respected wherever it is found, for instance, in the proto-rational or semi-
rational minds of animals (198). The problem with this position, as noted by

Onora O'Neill, is that it is not obvious that all or even most animals participate in

any kind of rational nature at all, fragmented or otherwise ("Kant and Duties"
224).

Kant himself regarded animals as possessing neither reason nor even
understanding. Even if some animals do have a degree of rationality (as I will
argue in Chapter 2), they do not have the kind of rationality that is worthy of
moral respect as an end in itself. Presumably the intelligence that animals have
is one of the preconditions of rationality, as Wood suggests, but it is not clear
why that gives it value. It is not as though the animal has the potential to develop
these capacities into full moral rationality (unlike children, about whom it makes
sense to say that we must respect their pre-rational nature).

Peter Carruthers has also defended a Kantian approach, although with a
change in emphasis. Whereas Kant claims that cruelty to animals tends to cause
people to become hard in their dealings with other human beings, Carmthers
claims that cruelty to animals reveals an existing flaw in the agent: a general
indifference to suffering, which will probably also express itself in the agent's
dealings with human beings (153-154). Carruthers’s approach is very much in
the spirit of Kant’s, but with one advantage and one disadvantage. The

advantage is its empirical plausibility. Animal welfare organizations often

emphasize a link between violence against animals (especially in youth) and
violence against humans. The implication is often that the former causes the
latter; much has been made of the fact that many infamous murderers were
previously caught abusing animals. It could just as easily be said, however, that
some underlying character flaw (indifference to, or even enjoyment of, others’
suffering) is responsible for both the animal and the human cruelty. Carruthers is
therefore not faced with the difficulty of showing how the one type of cruelty
causes the other; he only needs to show that there is a connection.

The disadvantage of Carruthers’s view is that although it may show why
we have a repugnance toward animal cruelty, it does not show that these actions
are immoral. If, as Kant holds, being cruel to animals causes us to be cruel to
humans, then we have a duty to refrain from being cruel to animals. If the animal
cruelty is only a symptom of a character flaw, rather than the cause of it, then it
would be deplorable but not evil. We would have no duty to refrain from it,
though we would be justified in passing unfavorable judgment upon the moral
character of those who engaged in it. By way of analogy, consider the mother
who suspects that her son’s style of dress indicates that he is involved with the

drug culture. This seems to be good reason for her to be concerned about his

character and lifestyle, and perhaps even to find his style of dress unpleasant.
However, it does not mean that his clothes are inherently harmful, and if she
responds by forbidding him to wear them, most would think her prohibition is
misguided. If the clothes caused the lifestyle, then the prohibition would make
sense. Hence an indirect duty view, to successfully establish that we ought to
refrain from being cruel to animals, must establish that such cruelty itself causes
the character flaw that leads to cruel treatment of humans, as Kant maintains.

Two main criticisms have been levied against the Kantian indirect duty
approach. The first is that it has certain consequences that are extremely
counter to intuition. James Skidmore identifies a counterfactual claim (clubbed
“CF") which is widely regarded as intuitively true:

Counterfactual (CF): Even if certain (e.g., cruel or brutal) treatment of
animals had no effect on an agent's attitudes or behavior towards human

beings, such treatment would 53m be morally objectionable. (8-9)1

 

Kant (and other supporters of the indirect duty view) must deny CF.
Furthermore, if it happened that somehow torturing animals made us kinder to

humans (for instance, by allowing us to release aggression), then we would be

 

1 The page references for Skidmore refer to a copy of his presentation provided to me by the
author.

obliged to do it (8).

Christina Hoff gives the example of a man who has always acted kindly
towards his family and towards human beings in general, but who is in the habit
of secretly burning stray dogs to death. According to Kant, he would not be
wronging the dogs, since we have no duties to dogs. Instead, he would be guilty
of wronging humanity, because such dealings with animals tend to make one
hard towards human beings. The terrible suffering of the dogs is in itself of no
importance. Indeed, before the arrival of mankind on the evolutionary scene, no
animal suffering or happiness had any value whatsoever; and upon the
awakening of rationality, it took on a merely indirect significance. Hoff regards
this as implausible and counter to our moral intuitions (67).

In Hoff’s view, this implausible claim belies a deep flaw in Kantian ethics.
“If there are any moral truths,” she writes, “this one is clearly among them:
suffering is an evil, and gratuitously and deliberately to inflict pain and suffering is
a moral evil. This needs qualification, but we must be wary of any moral
theory [. . .] that loses sight of it" (68). Furthermore, Kant is unable to
satisfactorily account for mentally impaired humans in his ethics: either they are

simply means to an end, like animals, or else some sort of leeway must be

introduced to allow them moral recognition - but any such leeway is likely to
make it even more difficult to exclude animals. “It is implausible that our duty to
feed a hungry retarded child would turn out to be indirect and, in this respect,
essentially distinct from our duty to feed a normal child" (68).

Hoff regards Kant’s limitation of the moral realm to humans to be merely
dogmatic, and says that in absence of any good argument for it, we should trust
our intuitions. It is certainly true that the indirect duty view is unintuitive, at least
for many people; however, it is uncharitable to simply discard Kant’s exclusion of
animals from the moral realm as dogmatic. He does indeed have an argument
for his position based on the inability of an irrational being to be autonomous (set
a law for itself), an argument not neatly excisable from his moral theory as a
whole. The fact that his position runs into difficulties with regard to irrational
human beings is, however, worth noting.

Are the counterintuitive consequences of Kant’s view a problem for it?
They certainly make it seem unpalatable. Alexander Broadie and Elizabeth
Pybus point out that, since Kant believed his moral system to accord with the
ordinary moral intuitions of the common person, it is legitimate to criticize his

moral system if it does not in fact accord with these intuitions. Aside from this,

however, there is another strong criticism of Kant’s theory which does not rest on
intuitions.

The other main problem cited against Kant's theory is that he cannot
successfully make the causal connection between cruel treatment of humans and
cruel treatment of animals. Broadie and Pybus show that Kant believes this
connection is founded on an analogy. Although animals are only things and not
persons, Kant claimed that they have some qualities which are analogous to
human qualities ("Kant's Treatment" 377). However, “His claim that animals are
analogous to persons appears to mean no more than that they behave as if they
have psychological states that we take to characterize people” ("Kant's
Treatment" 378). There is no further claim being made; Kant certainly does not
mean that they have anything like a faculty of reason.

Although they are things with features analogous to ours, they are
nevertheless still things, and therefore we do not have direct duties to them.

Kant refers to the mistaken notion that we have duties to beings other than men
as an amphiboly of the concepts of reflection ("Kant's Treatment" 379). Broadie
and Pybus note that an amphiboly is defined as “a confounding of an object of

pure understanding with appearance.” They interpret his remark as meaning that

10

the belief that we owe duties to beings other than men “is based on the false
assumption that beings which do not have a noumenal self, do” ("Kant's
Treatment" 379).

Although Kant holds that we do not have direct duties to animals, “L . .] he
holds that maltreatment of animals is wrong first because it leads us to be
unsympathetic to [. . .] other people. In other words it leads us to treat other
people merely as a means” ("Kant's Treatment" 382). Second, it is wrong
because it does violence to our own humanity, i.e., it leads us to treat ourselves
as a means.

Broadie and Pybus regard this position as inconsistent, because Kant is
claiming that in using certain things (animals) as means, by analogy we are led to
treat people as means. Kant cannot point to any morally relevant difference
between an animal and any other sort of mere thing, since the only possible
morally relevant difference would be the possession of rationality, which animals
do not have. Therefore, the authors claim, Kant is forced to say that nothing may
be used as a means; we have an indirect duty to any thing not to use it as a
means. “This is not merely absurd, but contrary to his imperative of skill” ("Kant's

Treatment" 382). They also claim that Kant cannot prove even an indirect duty to

11

animals, because Kant's position rests on a psychological claim about human
nature - that cruel dealings with animals make people hard towards other people
- which, even if he could prove it true, is irrelevant because it is “L . .] a
contingent matter of fact about human beings, and not a fact about Latigmal
beings” ("Kant‘s Treatment" 382).

The fact that it is a matter of fact about human beings and not about all
rational beings should not be a problem for Kant. Lying is morally wrong
according to the universal law formulation, but in order for there even to be such
thing as lying, we have to be the sort of beings who can communicate with each
other, and who can express themselves falsely, and so on. The specific fact in
this case involves the psychology of human beings, but the maxim could be
construed as “I will not perform actions which tend to harm my ability to behave
morally.” In order to apply this maxim, we do of course have to look at the
empirical facts about what does tend to harm this ability.

Tom Regan responds to the aforementioned article with the remark that,
although Kant’s position may go against intuition, it is not internally inconsistent.
Kant never claims, Regan points out, that we ought not use animals as means

(as beasts of burden, for example). He claims that we ought not maltreat them,

12

which is a narrower claim. “For we can, given Kant's views, use an animal as a
means without at the same time necessarily maltreating it, as when, for example,
a blind man uses a seeing eye dog but treats him with love and devotion” (471 ).

Regan’s response is brief, and does not address the central problem: what
does it mean to maltreat something? We cannot define it as “to use something in
such a way that goes against rationality (or morality)” because that begs the
question. Maltreating something cannot merely mean using it as a means, for
the reasons that Regan gives.

There is a hint in the Lagtgras gn Ethigs where Kant is reported as saying,
“Vivisectionists, who use living animals for their experiments, certainly act cruelly,
although their aim is praiseworthy, and they can justify their cruelty, since
animals must be regarded as man’s instruments; but any such cruelty for sport
cannot be justified” (272460). So, maltreating an animal for Kant is treating it with
unnecessary cruelty. Whether a given cruelty is necessary is probably
dependent on whether it is required for fulfilling a direct duty (or possibly even an
indirect duty) to human beings.

Broadie and Pybus's analysis of the belief that we have direct duties to

animals as an amphiboly, or mistaken analogy, raises an interesting point. If it is

13

a mistake that leads us to connect human and animal suffering, and if it is this
psychological connection that leads to the causal connection between the two
kinds of cruelty, wouldn’t the solution be to train ourselves not to make that
mistake? Kant’s theory seems to be aimed at damage control, rather than
prevention. Rather than accept that we will make that mistake, and then try to
make sure it doesn’t harm our sensibilities, it seems better to learn not to make
the mistake at all.

Skidmore makes a similar point. A weak or moderate connection between
cruelty to animals and inappropriate attitudes or behavior toward humans (i.e.,
that the former agaaaionally or asaafly leads to the latter) is not enough to
establish indirect duties to animals. If the weak or moderate connection were
established, it would not show that all agents have such duties; only certain
agents would be so obliged, and the others could treat animals however they
pleased. In order for the indirect duties to be universally applicable, it must be
true that cruel treatment of animals almost always, for almgst all agents, results
in inappropriate attitudes or behavior toward humans (Skidmore 6). There is a
lack of empirical evidence for this connection, and some evidence that suggests

it is false. Surely some cultures have existed in which animals were treated

14

brutally, without everyone in turn being brutal to each other. Consider Spain, for
instance; blood sports such as bullfighting are traditional and popular, yet there is
no evidence that the people of that country are any more brutal to each other
than in countries where such events are frowned upon.

The strong connection required for the indirect duty view may not be true;
but even if it is true, Skidmore argues, the connection would not be a necessary
one. If, as the indirect duty theorist claims, there is a clear moral difference
between humans and animals, then it should be possible for us to harm animals
without harming our sympathy for fellow human beings (11). In fact, since
sympathy for animals sometimes can distract our attention from our true duties, it
is not morally ideal. Therefore, we ought to try to shape our sympathy “L . .] to
reflect better the clear and crucial moral distinction (on Kant’s view) between
animals and persons” (13).

The indirect duty theorist must then claim that shaping our natural
sympathy in this way is impossible. This claim is very implausible, given the
variation among people and cultures. Skidmore uses the example of abortion.
Some people have an acute sympathy even for embryos, whereas others see

them as nothing more than inconsequential tissue. “It seems rather obvious that

15

many people can and do shape their sympathies to reflect more adequately the
moral beliefs they come to hold” (14).

An additional point along these lines (though not raised by Skidmore) is
that at least some agents will, in the course of fulfilling their duties to humans,
have to inflict acute suffering on animals. The researcher who must injure,
poison, and inflict diseases upon animals for the benefit of humankind knows that
this is his duty. It seems that he has two choices: he can unlearn his natural
sympathy for animals because of his understanding that it is not morally
appropriate; or he can harm animals despite his sympathy, and therefore also
damage his sympathy for human beings. The latter would be immoral for the
indirect duty theorist, but if the former is possible then the strong connection does
not hold. Either the indirect duty theory must be abandoned, or else it must
become so strong that any use of animals which causes suffering - including

medical research - is forbidden.

3. Other Kantian Views
Besides the largely unsuccessful defense of Kant’s own position, the other

main Kantian strand in contemporary thought about the moral status of animals

16

has been to try to establish that direct duties to animals are possible within a
Kantian ethical framework. A first necessary step in any such project is to show
that, contrary to Kant’s stated position, it is possible for there to be moral patients
which are not moral agents. Kant believed that only moral agents can be moral
patients because the requirement for moral agency, and the criterion for moral
consideration, are the same: rationality.

One author who tries to separate the two is S. F. Sapontzis. Sapontzis
defines “moral being” as “any being which meets both of the following conditions:
it is capable of acting morally, and it merits moral rights" (45). First he considers
the issue of whether animals are capable of acting morally. Animals, he points
out, perform many actions which would readily be regarded as moral actions if
they were performed by human beings: dolphins risking their lives to save human
beings, guard dogs acting courageously in the defense of their owners, and so
on. Yet animals are usually regarded as incapable of moral action (by Kant and
others). Sapontzis reconstructs this argument as follows:

A-1 Only rational beings can be moral.

A-2 Animals are not rational.
A-3 Therefore, animals cannot be moral. (45)

17

In order to determine the soundness of this argument, Sapontzis discusses what
is meant by “rationality," why it is necessary for morality, and whether animals
indeed lack the required rationality.

According to Sapontzis, Kant and others who make rationality a
requirement for morality tend to use examples that indicate that “rationality”
refers to normal, adult, human intelligence. Animals, the insane, and the
retarded are usually thought of as amoral. On the other hand, morality is
supposed to be accessible to the common person, not just to geniuses. The
standard of rationality is considered to be “normal (adult, human) intelligence”
(45).

Why is reason considered necessary for morality? The argument as
Sapontzis sees it is as follows:

B-1 An action is moral only if the agent recognizes it is the moral thing to

do and does it because it is the moral thing to do.

82 Only rational beings are capable of such recognition and motivation.
B-3 Therefore, only rational beings can act morally. (46)

He differentiates between two senses of “moral." “Moraln" is the moral value that

an action has in itself, regardless of the motivations of the agent. “Ordinarily,
kindness and honesty are morally preferable to cruelty and dishonesty regardless
of the agent’s motives or understanding of the situation” (46). “Moral; is the

18

moral value of the agent’s motivations for performing the action and his -

understanding of the action. So, “An action is moralal only if the agent recognizes

its moraln value and does it because of its moraln value” (46). This premise is

based on the common sense notion that performing a moral action sincerely is
superior to either accidentally performing it, or performing it for some ulterior
motive.

Another type of argument for the necessity of reason for moral action is
identified by Sapontzis as follows:

C-1 Only beings which are free to choose what they will do can act

morallya.
C-2 Only rational beings are free to choose what they will do.

C-3 Therefore, only rational beings can act morallya.
A bird which feeds its hatchlings only because they have a certain smell is not
acting morally, because it is blindly reacting to a stimulus, one which is not even
morally relevant to the situation. It is like a change machine which dispenses
coins not out of a sense of dutiful exchange but simply because it is mechanically
constructed to do so.

Sapontzis argues that the moral" actions of animals do sometimes meet

the requirements of arguments B and C. “When a dog pulls a child from a fire, it

19

is not acting blindly, like an insect reacting to a chemical mating stimulus," thus
satisfying the requirement in argument C, “nor does it seem to have ulterior

motives,” thus satisfying the requirement in argument B (47). What morality
really requires is that the agent recognizes that a goal has some moralIn worth,

and that the agent makes a sincere effort to accomplish that goal. Sapontzis
backs this up with various examples of animals which act in ways that are
apparently not merely instinct-driven, such as a dolphin rescuing a human (a
member of a different species entirely).

Having given animals this foot in the door to morality, Sapontzis goes on

to qualify it. What animals lack, morally speaking, is the big picture. True
moralitya requires the commitment to a moral way of life:

D-1 An action is moral only if the agent does it because he believes it will
contribute to attaining an ideal way of life.

D-2 Only rational beings can project and dedicate themselves to ideal
ways of life.

D-3 Therefore, only rational beings can be moral.

He claims that animals do not perform moral acts, but do perform virtuous acts,
and introduces some new terminology:

[K]indness, courage, honesty, etc. are “virtues”; intentional, sincere acts of
kindness, courage, etc. are “virtuous acts”; and “moral acts" are virtuous
acts which are done as part of fulfilling an ideal way of life. Thus, many
animals are virtuous, but none are moral. (50)

20

Sapontzis then returns to the two conditions for being a moral being. The
first, as discussed, is the condition of being capable of moral action. The second
is the condition of meriting moral rights. A moral being “must be both a moral
agent and a moral recipient" (50). He notes that for Kant, these two conditions
are closely related; “[. . .] it is because rational beings are capable of acting
morally, which is an end in itself, that they must be treated as ends in
themselves” (50).

Animals can act virtuously, according to Sapontzis, but not morally. This
puts them in an odd position, because they are capable of doing good deeds but
they lack any sense of striving to do better. Sapontzis claims that this is similar
to Kant’s conception of the “holy will.” A holy will is a purely rational being which
always acts virtuously (in Sapontzls's terminology) out of its very nature. “God
always acts virtuously, but He is not striving to fulfill an ideal way of life and,
therefore, is not a moral being” (51 ). God, being unaffected by sensuous
impulses, does not have to “strive" at all. Sapontzis claims that if God is
considered worthy of respect, as a being who does good without being moral,
then it must be acknowledged that there are beings who are not moral that are

nevertheless worthy of respect. It is plausible that an animal, as another sort of

21

being which is capable of being virtuous but not moral, would also be worthy of
respect.

Sapontzis’s position is an interesting approach, although Kant would likely
not accept that an animal could act virtuously (here using the term as defined by
Sapontzis, not as defined by Kant), and would maintain that even a dog saving
someone from a burning building is merely responding automatically to the
situation. Granted, it is responding to a morally relevant feature of the situation
(saving a life) as opposed to the birds who only feed their offspring because they
smell right, who are responding to a morally irrelevant feature (the smell);
nevertheless, for Kant the dog is not acting freely, and hence its actions have no
moral worth. At best, such a dog might be like the person who does the right
thing out of sentiment instead of out of regard for the moral law.

Sapontzis’s attempt to show that the class of moral agents and the class
of moral patients do not have to be identical for Kant is especially interesting, but
may not be as Kantian as he claims. The conclusion that we have duties to God
(who is a moral patient but not strictly speaking a moral agent) and so we may
also have duties to animals, does not follow, because Kant does not believe us to

have duties to God. According to Kant, such a duty cannot exist because it

22

“L . .] would be a tragecmdant duty, that is, a duty for which no corresponding
external subject imposing the obligation can MIL/QB [. . .1” (6:241 ).

Broadie and Pybus have a different approach to establishing Kantian
duties. They believe we should entirely dispense with the very notion that a duty
must be a duty to an object instead of a duty to perform an action. “If we see a
man drowning we may judge that we have a duty to save him, without judging
that we have a duty to him [. . .]" ("Kant and Direct" 64). The trouble with this
strategy is that it requires us to discard the formula of humanity and its
associated arguments. However, the formula of universal law will still hold, which
is apparently what Broadie and Pybus consider more essential about Kant’s
doctrine.

Broadie and Pybus claim that Kant’s position is counter to our moral
intuitions because it denies that we have duties to animals, but reserves direct
duties for humans. Their solution is to abolish all direct duties. This seems to be
even less intuitive, not more. If it is counterintuitive to say that refraining from
beating a dog is not actually a duty ta the dog, it is just as counterintuitive to say

that refraining from beating a human is not actually a duty to the human.

Finally, Christine Korsgaard has recently offered a Kantian view of duties
to animals that makes use of the concept of passive citizenship in the Kingdom of
Ends. Since her approach has some important similarities to the interpretation

that I will offer in my final chapter, I will discuss it in detail there.

4. Looking Fonlvard

If Kant's own view is untenable, then a Kantian has two options: either
accept the counterintuitive result that we have no duties at all (indirect or direct)
to animals, or try to find some other way to establish duties to animals within
Kant’s system. My goal will be not only to find a better way to ground duties to
animals within Kant's framework, but also to show that these duties go beyond
mere avoidance of wanton cruelty. I will spend the next two chapters explaining
Kant's conception of rationality and the role it plays in his moral theory.

Chapter 2 is devoted to determining what mental capacities Kant believes
animals to possess and evaluating his position. I will consider whether animals
could be considered rational, but conclude that Kant's conception of rationality
prohibits this. I will argue, however, that animals do have minds, including some

of the same mental faculties possessed by humans. This will tie in to the

24

eventual claim, in Chapter 4, that animals should be accorded the status of moral
patients.

In Chapter 3 I discuss the role of reason, will, and autonomy in Kant's
ethics, showing how they come together in the concept of an end in itself. I will
suggest that although this concept is adequate to describe moral agency, it
cannot account for nonrational moral patients, including some human beings. I
acknowledge that animals are not ends in themselves, but leave open the
possibility that there may be another way to qualify as a moral patient.

Finally, the existence of non-agent moral patients will be defended in
Chapter 4. I will show that, despite Kant's own position on animals, Kantian
ethics can be expanded to include duties to them that are not merely indirect. In
doing so I will show that rationality can still be central to morality, and rational
beings can still be the center of the moral universe, even while we acknowledge

our duty to incorporate nonrational animals into our moral legislation.

25

2. KANT AND ANIMAL MINDS

1. Introduction

Kant did not regard animals as objects of direct duties. In order for
something to be an object of direct duties it must be an end in itself, and in order
to be an end in itself it must be rational. Animals, for Kant, cannot be objects of
direct duties because they are not rational. Thus, to fully understand the moral
status of animals in Kant's ethics, we must first determine why he does not
regard them as rational. It cannot simply be that they do not have minds at all,
because it is clear that Kant does not regard animals as Cartesian machines. He
specifically denies this in the Critique Qf tha Power at 3.lgggmant, stating that we
can infer from the similarity between animals' behavior and man's that animals
also act according to representations (5:464n). Since animals have
representations, and representations are not physical entities, animals must not
be entirely physical, as noted by Steven Naragon ("Kant on Descartes” 14). The
question, then, is not whether they have minds, but what kind of minds they
have. In this chapter I will address this question in two different ways. First I will

explicate Kant's discussion of animal minds in his written works, and thus

26

determine his stated doctrine on the subject. Second I will discuss whether
Kant's conception of the intellectual faculties necessarily leads to his position on
animal minds, or whether it is possible to hold Kant's view of reason without

coming to the same conclusions about animals.

2. Animals, Sensibility, and Understanding
First let us consider the cognitive faculties, sensibility and the
understanding. Kant describes the understanding as follows:
The understanding [. . .] [is] a nonsensible faculty of cognition. Now we
cannot partake of intuition independently of sensibility. The understanding
is therefore not a faculty of intuition. But besides intuition there is no other
kind of cognition than through concepts. Thus the cognition of every, at

least human, understanding is a cognition through concepts, not intuitive
but discursive. (A 67-68IB 92-93)

The understanding is a faculty of cognition through concepts. The
concepts of the understanding are what enable us to think of objects. Before the
objects can be thought of, they must be given through sensibility, which is our
capacity to be affected by objects. Sensibility gives rise to intuition through the
forms of intuition, space and time. Cognition, then, requires both the faculty of
sensibility in order for the objects to be given, as well as the faculty of

understanding in order for the objects to be thought. Kant writes that "without

27

sensibility no object would be given to us, and without understanding none would
be thought. Thoughts without content are empty, intuitions without concepts are
blind" (A 511B 75). The understanding, along with sensibility, makes it possible to
cognize something as an object.

The understanding is made up both of empirical concepts (conwms
derived from experience) and of pure concepts. Pure concepts are "strictly a
priori conditions for a possible experience, as that alone on which its objective
reality can rest" (A 95). They are the concepts of an object in general. For
something to be an object for us, we must already have the concepts that allow
us to cognize it as an object. We must, therefore, bring the pure concepts to
experience instead of deriving them from experience; otherwise there is no
comprehensible experience to draw from in the first place.

As in the case of sensibility, which requires the forms of intuition (space
and time) in order to give rise to intuitions, so the understanding presupposes
some kind of synthesis in order for the manifold of intuition to be transformed into
unified objects by the understanding. Kant writes that "All unification of
representations requires unity of consciousness in the synthesis of them.

Consequently the unity of consciousness is that which alone constitutes the

relation of representations to an object [. . .] and consequently is that which
makes them into cognitions and on which even the possibility of the
understanding rests" (B 137). This unity of consciousness, pure apperception, is
"that self-consciousness which [. . .] produced the representation Ltfligk" (B 132).
The simple representation "I think" must accompany the manifold of intuition in
order for cognition to be possible; that is, the representations that make up an
intuition are always thought of as my representations, and in this way are united
in one consciousness (B 132).

These, then, are the basic components of cognition: "first, the concept,
through which an object is thought at all (the category), and second, the intuition,
through which it is given" (B 146). All intuitions have space and time as their
form, whereas the understanding requires pure apperception (self—
consciousness) in order to unite the manifold of intuition into objects. These
cognitive processes are the same in all human beings; it is this commonality that
allows us all to recognize the same objects and to interact with each other as part
of the same world. Because these faculties are an inextricable part of our minds,
"The human understanding cannot even form for itself the least concept of

another possible understanding, either one that would intuit itself [i.e. would have

29

a non-sensible intuition] or one that, while possessing a sensible intuition, would
possess one of a different kind than one grounded in space and time" (B 139).
The fact that our understanding cannot form a concept of such an alien cognition,
however, does not mean that it is not possible: "We are acquainted with nothing
except our way of perceiving [objects], which is peculiar to us, and which
therefore does not necessarily pertain to every being, though to be sure it
pertains to every human being" (A 42).

Although a substantial part of the W is devoted to
the exploration of the cognitive faculties, it is mostly silent on the subject of
whether animals possess those faculties. Since its purpose is to analyze human
reason and understanding, it is unsurprising that animals are hardly mentioned.
One passage does imply that animals possess sensibility, since Kant describes
them as having an W, or sensible power of choice (A 534/8
562). In other works we find further support for the claim that Kant attributed
sensibility to animals. For instance, there is the above-mentioned passage from
the Critjgua of the Pgwar of Judgment, in which Kant asserts that animals act
according to representations (5:464n). Kant refers to sensibility as "the

motility of our mind to receive representations insofar as it is affected in some

30

way" (B 75/A 51). Animals receive representations, so they can be described as
having a faculty of sensibility. In the Lectures on Metaphysig, Kant states
explicitly that animals possess sensibility including "a faculty of sensation,
reproductive imagination, etc." (28:277).

Sensibility is a low-level faculty, comprised only of the capacity to be
affected by the world and receive representations. Animals do show every sign
of reacting to the world, which implies that they are affected by it. On the other
hand, it could be argued that if sensibility is no more than the ability to react to
the environment, then plants also possess sensibility. Plants react to sunlight by
growing toward it, and some plants (such as the Venus flytrap) can even make
quick movements based on stimuli. But to carry this line of reasoning further, if
any reaction to environmental change constitutes proof of sensibility, then litmus
paper possesses sensibility. Textual evidence does suggest that Kant regards
animals as having sensibility, but there is little evidence to suggest that he
regarded plants or nonliving things as possessing it. Extending the faculty of
sensibility to animals does not commit Kant to extending it further, because

sensibility is involved with representations, which plants and nonliving things

31

presumably lack. An explanation of the phototropic movements of plants can

readily be made without reference to any representations.2

The fact that animals have representations and, implicitly, a faculty of
sensibility does not mean that they have cognition. Cognition requires both
sensibility and understanding, so if animals do not possess understanding, they
can have only "blind" intuitions. Kant does not regard the understanding and
sensibility as inseparable; he writes, "T he categories of the understanding [. . .]
do not represent to us the conditions under which objects are given in intuition at
all, hence objects can indeed appear to us without necessarily having to be
related to functions of the understanding [. . .1" (A 89/B 122). Granting that
animals have sensibility, then, does not lead to the conclusion that they must
have concepts of the understanding; and, in fact, Kant states that they do not
have understanding (28:276. 29:879). Since they have sensibility and
representations, they do have minds; but since they lack cognition, their intuitions
are "blind."

What does it mean for a being's intuitions to be blind? Kant writes that

 

2 See also the Qousfiostumum (221418) in which Kant claims that animals have souls and plants
don't, because plants can be grafted, and therefore are not unified organisms.

32

"Without sensibility no object would be given to us, and without understanding
none would be thought" (A 511B 75). A creature that had sensibility but not
understanding would have objects given to it that could not be thought. In fact,
since the understanding contains the concepts that make it possible for
something to be an object for us, a being without understanding would not have
any awareness of objects as such. At most it would have the raw matter of
sense, completely unsorted and unsynthesized. It would not have an experience
anything like ours, and since we are conceptual beings we cannot even imagine
what its experience would be like.

Furthermore, the metaphor of blindness suggests the interpretation that
intuitions without concepts could not be experienced at all. Indeed, Kant equates
experience with empirical cognition (B 147), and cognition requires concepts of
the understanding. Therefore, a creature with sensibility but not understanding
would not have experience, for Kant.

Kant describes sensibility as "the capacity (receptivity) to acquire
representations through the way in which we are affected by objects" (A 191B
33). It seems strange to say that a creature has representations that it does not

experience, as would be implied for a creature that had sensibility but not

33

understanding. Stranger yet, since Kant defines sensation as "the effect of an
object on the capacity for representation" (A 201B 34), it would even be possible
for a creature to have sensation but not experience. However, the apparent
strangeness is merely a result of Kant's technical language. In common
parlance, "experience" and "sensation" are often taken to be synonyms, but Kant
assigns a more robust meaning to "experience." For Kant, experience is
necessarily an experience of objects, which is only possible through application

of the concepts of an object as such.

3. Animal sensation and consciousness

l have argued in the previous section that animals, for Kant, possess
sensibility, though not understanding. They have representations, but not
experience. Do they have any sensation or consciousness at all?

Kant repeatedly denies that animals are conscious. Many of these
passages appear in pre-critical works, and so they may not be representative of
his mature doctrine, but as late as in the Politz lecture notes (dating to sometime
within 1778-1780) Kant writes that animals have "a faculty of sensation,

imagination, etc., but all only sensible as lower faculties, and not connected with

34

consciousness" (282277). On the other hand, in a few passages Kant does
attribute consciousness to animals (Naragon, "Kant on Descartes" 11).

Naragon argues that when Kant denies consciousness to animals, he is
using the term in a specialized way that does not preclude the possibility that he
may have regarded animals as conscious in some sense:

If by "consciousness" is meant responsiveness to representations, then

Kant certainly allows that brutes are conscious. What he disallows is the

brute's ability to represent to itself some representation that it has, or to

bring several or all representations together, or to compare two separate
representations [. . .1. This use of "consciousness" appears similar to that

of 'inner sense.’ ("Kant on Descartes" 12)

One possible interpretation is that when Kant denied consciousness to animals
he really only meant to deny them self-consciousness. Since they lack self-
consciousness, they naturally also lack understanding (which requires the unity
of apperception). According to the Lagyres on Megphysics, "an animal has no
apperception, and therefore it is also incapable of any moral principles, of the use
of understanding and reason" (29:879). Lacking understanding, they would not
be able to judge by means of concepts, which could be described as "bring[ing]
several representations together" or "compar[ing] two separate representations."

This interpretation is consistent with Kant's letter to Herz (May 26, 1789) in which

he attributes consciousness, but not self-consciousness, to animals (11:52).

35

lf Kant meant only to deny animals self-consciousness, why did he so
often deny them consciousness? Naragon's interpretation is that consciousness
and self-consciousness are separate concepts, so that when Kant denies
animals consciousness he is not simply denying them self-consciousness. He is
instead denying them inner sense, meaning "me ability to have representations
of 'inner things“ (namely, other representations)" ("Kant on Descartes" 12). Inner
sense is contrasted with outer sense, the ability to have representations of outer
things. Kant writes that "by means of outer sense we represent to ourselves
objects as outside us"; inner sense, on the other hand, is the means by which
"the mind intuits itself, or its inner state" (A 221B 37). Naragon's interpretation of
"consciousness" in certain passages as meaning inner sense is plausible, and is
strengly supported by the following passage in the W:
Accordingly we ascribe to these beings a faculty of sensation,
reproductive imagination, etc., but all only sensible as a lower faculty, and
not connected with consciousness. We can explain all phenomena of

animals from this outer sensibility and from mechanical grounds of their
bodies, without assuming consciousness or inner sense. (282277)

Naragon concludes that Kant is denying animals inner sense while leaving
them with outer sense. The former is a higher level of consciousness than the
latter; we could think of them as a rich consciousness (inner sense, or the ability

to have representations of representations) compared with a bare consciousness

36

(outer sense only). According to Naragon, Kant regards animals as having a
bare consciousness, but lacking a rich consciousness; and since self-
consciousness is an even higher level of consciousness, animals lack that also.
Kant is clear that inner sense and the faculty of apperception should be
distinguished from each other (B 154-155); hence, if Kant uses "consciousness"
in some passages to denote the inner sense, it is consistent with Naragon's
interpretation that consciousness and self-consciousness are not the same for
Kant.

Karl Ameriks makes the claim that when Kant denies that animals are
conscious, he is simply denying that they have the ability to judge (Ameriks
242n). It is clear that Kant does deny them this ability (Naragon, "Kant on
Descartes" 13); however, it seems unlikely that Kant would use the term
"consciousness" simply to mean judgment. Naragon suggests that the
connection between animals' inability to judge and Kant's denial of their
consciousness is less direct; according to Naragon, "a certain reflective ability
such as we find in inner sense... seems to underlie the ability to judge" ("Kant on
Descartes" 13). Naragon thus ties the ability to judge with consciousness (in the

rich sense); since animals do not have rich consciousness, they cannot judge.

37

There is, however, a simpler explanation for why animals cannot judge; judgment
is defined as "the way to bring given cognitions to the We unity of
apperception" (B 141 ). Judgment thus requires self-consciousness, not merely
consciousness, and furthermore it involves the application of concepts of the
understanding, which animals (for Kant) do not have.

Naragon makes a good case that Kant recognizes different levels of
consciousness, which accounts for his inconsistent remarks about animal
consciousness. Further evidence that animals must be considered conscious in
some sense can be found in Kant's hierarchy of representation:

The genus is representation in general (repraasantatig). Under it stands

the representation with consciousness (perceptio). A perception that

refers to the subject as a modification of its state is a sensatlon

Wig); an objective perception is a cognltlon (cggaitiq). (A 320B
376).

Since sensation is there described as a kind of perception, and perception
as being representation with consciousness, it follows that a creature with
sensations must be conscious. Several passages show that Kant does attribute
sensation to animals. In a passage from the Leggras an Mataphysigs Kant
specifically attributes "a faculty of sensation" to animals (28:277). Indirect
evidence can also be found in passages where he attributes feelings such as

enjoyment or misery to them. In asserting that we have indirect duties with

38

regard to animals, he cites the "shared feeling of their suffering" as the sentiment
that is damaged by violating these duties; "suffering" would seem to describe a
variety of sensations (6:443). In a passage from the Metaphfiics Qf Mgrals Kant
discusses the impulses of nature that human beings have in common with the
animals, and he includes "the preservation of his capacity to enjoy life" (6:420). It
would be absurd to attribute enjoyment to a being without sensation;
unconscious representations surely cannot constitute enjoyment. Another
example is in the 59mm; there Kant remarks that "Malebranche chose to deny
to non-rational animals a soul, and therefore feelings, rather than to admit that
horses had to endure so much misery..." (6:74n), with the implication that
Malebranche's claim is absurd. Finally, in a letter to Marcus Herz (May 26, 1789)
Kant suggests that animals have feelings and desire (11:52).

If Kant's denials of consciousness are a denial of rich consciousness
(inner sense), then they also imply that animals do not have the further quality of

self-consciousness. Self consciousness is required for understanding, so this

39

reinforces Kant's contention that animals do not have understanding.3 According
to Mrongovius's notes:
Now [self-consciousness] belongs to the human soul alone, and
distinguishes it from all animals. An animal has no apperception, and
therefore it is also incapable of any moral principles, of the use of
understanding and reason [. . .] (29:878-879)
Animals are aware of their surroundings, but do not have a conception of
self - they do not distinguish between "I" and "other" or even between "inner" and
”outer." They have sensation, but do not distinguish the world as objects. They

have minds of a sort, but so rudimentary as to be incomprehensible to us. They

are more than Cartesian machines, but much less than persons.

4. Reason in Kant

First let us turn our attention to a matter of terminology. Kant uses
"reason" (yarngnft) in both a specific and a general sense. The general sense
refers both to reason proper and to the understanding, and as such is a general
description of a person's intellectual workings. Although Kant makes no formal

announcement of these different uses, they can be seen in various passages. In

 

3 This is entailed even if Kant's denials of animal consciousness can be interpreted as simply a
denial of animal self-consciousness.

the Critigga 9f Pure Reason Kant usually uses "reason" or "understanding" as
appropriate, but the general use of "reason" is evident in places: for instance, the
preface to the second edition contains various observations about the relation
between reason and nature, stating that "reason has insight only into what it itself
produces according to its own design" (B xiii) and so "it has to seek in [nature]

[. . .] in accordance with what reason itself puts into nature" (B xiv). It is true that
reason in the specific sense does contribute guiding principles to the organization
of the natural world, but in speaking of what reason "produces" and "puts into"
nature, Kant is probably also referring to the contributions that the understanding
makes to the world of appearances through pure concepts such as causality.
Another use of the general sense can be found in the Prglaggmena (in which
Kant refers to the mathematical concepts associated with the understanding as
products of reason (280-2811).

The Transcendental Logic, which is the section that makes up the largest
part of the Critique of Pure Reason, is divided into the Transcendental Analytic
and the Transcendental Dialectic. The Analytic deals with the understanding,
and the Dialectic with reason proper. Why, in a work titled Critique 9f Pare

Reagn, is so much of the discussion devoted to the understanding, rather than

41

reason? Part of the answer is that the "reason" in the title refers to the general
sense of the term explained above. The purpose of the critique, Kant writes, is to
"supply the touchstone of the worth or worthlessness of all cognitions a_p_r_io;i" (A
121B 26). Since the understanding is a cognitive faculty, it must surely be
included in such a critique. A parallel passage in the W
Juggmam makes the inclusion even more explicit (5:288).

Another part of the answer is that Kant is setting up the understanding in
contrast to reason. The understanding plays the cognitive role in Kant's system
that reason plays in many of his predececessors'. Instead, reason has a higher
calling in the practical realm.

The idea that reason is a cognitive faculty is associated with dogmatism,
"the presumption of getting on solely with pure cognition from (philosophical)
concepts according to principles" (B xxxv). A typical dogmatist is Leibniz, who
held that all truths are discoverable by reason. This is because the universe is
ordered rationally; it was created by a rational God in accordance with universal
truths (glam 414-415). Reason for Leibniz, therefore, is a cognitive faculty

by which we acquire knowledge of all things, both sensible and non-sensible

(Mommies! 459-460).

42

Kant holds that cognition can arise only from the interaction of the
understanding and intuition. Although the concepts of the understanding are a
priori, those concepts must always relate to intuition to be meaningful. The only
kind of intuition that human beings have is sensible, so all cognition is related to
experience. Leibniz and the other dogmatists have overstepped the bounds of
human cognition by claiming knowledge of ideas such as God. Kant agrees,
however, that human reason poses questions greater than those answered by
the mere cognition of sensible objects. Kant breaks with the dogmatists by
concluding that reason is not a cognitive faculty.

Leibniz's concept of reason is that of deterministic process. Susan
Neiman writes that "Leibniz holds that truth determines reason. That is, once we
understand the truth, we cannot help but accept it [. . .1. Truth itself commands
reason's assent" (31). Furthermore, deductive reason is mechanical: Neiman
suggests that "The correct use of reason involves a simple procedure that can be
followed without choice or embellishment to reach the truth that is reason's goal"
(31 ). This model of reason led Leibniz to believe that all disagreement could be
resolved by calculation, if only the correct procedure, the universal calculus,

could be discovered. Neiman describes the process of reasoning as having an

43

"automatic character," since "reason has no choice in assenting to the truths that
the universal calculus uncovers."

The understanding, for Kant, has a mechanical quality similar to what
Leibniz attributes to reason. Although the understanding is spontaneous rather
than merely receptive, Neiman points out that spontaneity should not be
confused with autonomy: "the categories of the understanding are spontaneous
in that they are applied to, rather than derived from, experience" (49). This
distinguishes them from the entirely passive association of ideas posited by
Hume, but their mode of application is still mechanical in nature, engaged in
”merely spelling out appearances according to a synthetic unity in order to be
able to read them as experience" (A 314/B 370-371).

Leibniz gives reason a cognitive role, but the price is that reason is turned
Into a mere instrument to be used in processing truths. Kant's separation of
reason from the understanding liberates it from this routine process, but also
acknowledges that its role is not a cognitive one. Whereas Leibniz saw reason as
a passive discoverer of the world‘s truths, Kant sees reason as an active creator
of ideas. This will prove significant because even if evidence shows that some

animals are able to use conceptual tools to arrive at conclusions about reality,

such evidence will not be enough to prove that they possess reason in the
Kantian sense.

Having shown the limits of the understanding (and hence of cognition) in
the Transcendental Analytic, Kant turns to the role of reason in the
Transcendental Dialectic. There he returns to the claim made in the preface, that
human reason "is burdened by questions which it cannot dismiss, since they are
given to it as problems by the nature of reason itself, but which it also cannot
answer, since they transcend every capacity of human reason" (A vii). The
understanding, since it is limited to objects given in (sensible) intuition, cannot
provide answers to questions about ultimate reality, but it is the nature of reason
to pose such questions.

Here again we can see a contrast between Kant and Leibniz. Richard L.
Velkley paints a somewhat less antagonistic picture of their disagreement than
does Neiman. He points out that both Kant and Leibniz acknowledge the
"metaphysical demands of reason" and attempt to account for reason's striving
toward the goal of complete knowledge (21 ). The contrast between Kant's and
Leibniz's reason, according to Velkley, is in their relationship to this goal. Leibniz

makes reason an inert instrument that can be set to work on achieving this goal,

45

which is external to it. But Kant, Vekley asserts, sees a contradiction between
the Enlightenment project to emancipate reason, and its use as such an
instrument. Instead, Kant has made reason the originator of its own goal: "In
some way, reason prescribes its own object; it spontaneously originates the goal
of wholeness" (22). Velkley's interpretation differs from Neiman's in that he
emphasizes the difference in what reason is striving for, whereas Neiman
emphasizes the difference in how reason functions. Both agree that Kant‘s view
of reason empowers it, as it is now free to create its own ends.

The term "reason" often brings to mind logic, and indeed Kant regards
logic, epitomized by the syllogism, to be a "formal [. . .] procedure of reason" (A
30618 363). He is careful to point out that “the syllogism does not deal with
intuitions [. . .] but rather deals with concepts and judgments" (A 306/B 363).
Since logic does not deal with intuitions, it does not result in cognitions. Reason
in its logical use serves to give order to the cognitions of the understanding by
comparing its concepts and subordinating them to each other.

Reason, however, is more than just the housekeeper of the
understanding. It also seeks "to find the unconditioned for the conditioned

cognitions of the understanding, with which its unity will be completed" (A

307/8364). This is the supreme principle of reason, but "L . .] no adequate
empirical use can ever be made of that principle" (A 308/8365). Reason, in other
words, always seeks an explanation for the things it encounters in nature (the
realm of the understanding), but there is always a further explanation for the
explanation — that is, there is always another condition behind each condition.
Ultimately, reason seeks the unconditioned, the final explanation that would unify
all observations. Leibniz and Kant are in agreement that reason is not satisfied
by a merely intermediate explanation. The dogmatists' mistake, according to
Kant, is that they assumed reason will be able to find the unconditioned that it
seeks.

Reason has its own concepts, but unlike the concepts of the
understanding, they are not limited to experience, nor can they be satisfied by
experience: "If they contain the unconditioned, then they deal with something
under which all experience belongs, but that is never itself an object of
experience" (A 311/8 368). Kant identifies three primary concepts of reason,
which he dubs the transcendental ideas: the soul, the world, and God. Each one

is the idea of an unconditioned unity. The soul is the unity of the thinking subject,

47

the world is the unity of the series of conditions of appearance, and God is the
unity of the condition of all objects of thought in general (A 334/8 391 ).

Kant devotes more space in the first C_ri_ti_q;@ to showing what reason
cannot do, than to discussing what it can do. When philosophers fail to correctly
distinguish between concepts of the understanding (cognitive concepts) and
ideas of reason, error ensues. All cognition is related to possible intuition, but the
unconditioned is not a possible intuition, so the ideas are not cognitive concepts;
they cannot be used speculatively (to gain knowledge). What, then, is the role of
reason?

As mentioned previously, one role of reason is a merely logical one, in
which it discovers the relations between the various concepts of the
understanding. The ideas of reason, however, are its own contribution entirely
apart from the understanding. They have two valid modes of employment:
regulative, and practical.

The regulative purpose of the cosmological ideas (those pertaining to the
world) is to "L . .] prescribe a rule to the regressive synthesis in the series, a rule
in accordance with which it proceeds from the conditioned [. . .] to the

unconditioned, even though the latter will never be reached" (A 510/8 538). It

provides a guide for science by commanding us to forever seek the conditions
behind everything in nature, but the journey will never be complete. The
regulative use of the ideas is the greatest extent to which they can be used in a
speculative (knowledge-gathering) context. They provide a program for scientific
inquiry, but they cannot be used to extend the concept of the world beyond
possible experience.

The ideas also have a practical employment:

Pure reason has a presentiment of objects of great interest to it. It takes

the path of mere speculation in order to come closer to these; but they flee

before it. Presumably it may hope for better luck on the only path that still
remains to it, namely that of its DIEM use. (A 796/8 824).

Reason prescribes pure practical laws, known as the moral laws. They are in
answer to the question of "I. . .] WW the will is free, if there is a
God, and if there is a future world" (A 800/8 828). These are questions about the
world, God, and the soul, respectively. The "better luck" Kant refers to when
discussing the practical employment of reason, is the fact that the ideas are more
robust in the practical sphere than in their regulative use; no longer limited to
mere principles of empirical regress, they acquire more content in response to

the demands of practical reason.

49

Kant emphasizes the primacy of reason's practical employment over its
theoretical. What exactly is meant by that is a subject of some debate. Onora
O'Neill takes Kant's claim about the primacy of practical reason, combines it with
his claim that the Categorical Imperative is the supreme principle of practical
reason, and concludes that Kant is committed to the claim that the Categorical
Imperative is the supreme principle of reason. Her interpretation of the primacy
of practical reason is that it is a temporal primacy: we must be able to cooperate
if we are to engage in reasoning, and thus we must live by moral rules that make
cooperation possible before we can begin to reason. O'Neill stresses Kant's
references to the "tribunal" of reason in the first C_rifigu_e_. The purpose of the
tribunal is to pass judgment on reason (and thus to vindicate it), but, O'Neill
writes, "tribunals can pass judgment only when they have been constituted. The
task of constituting a 'trlbunal' must come earlier on the agenda" Was
9). The metaphor of a tribunal also implies a community of people, in contrast to
the Cartesian project of an individual reasoning in isolation. The categorical
imperative makes such a community possible: "L . .] we must, if we are to evade

the threat of Babel, act only on that maxim through which we can at the same

time will that it be a universal law" (W 20). It is only through practice
that we are able to recognize the importance of the categorical imperative: "it is
[. . .] in thinking, communicating and acting that we discover that we must
discipline tendencies to rely on strategies of acting in ways that make others'
adoption of like strategies impossible" (Constatctions 27). O'Neill concludes that
the self-critique of reason is possible only if we understand reason itself to be a
construction of the tribunal that is critiquing it, and that the construction of reason
is constrained only by the principle that makes construction possible - the
categorical imperative (Canstructions 27).

O'Neill's interpretation is that the primacy of practical reason means that it
is logically, and perhaps even temporally, prior to the theoretical use of reason.
The textual evidence suggests a more modest interpretation. Kant's claim about
the primacy of practical reason can be found in several places, including a
dedicated section of the second Cfltiqga. There he writes, "8y primacy among
two or more things connected by reason I understand the prerogative of one to
be the first determining ground of the connection with all the rest. In a narrower

practical sense it signifies the prerogative of the interest of one insofar as the

51

interest of the others is subordinated to it (and it cannot be inferior to any other)"
(5:1 20).

Reason in its two employments has two corresponding interests. First, it
has a speculative interest, which is an interest in the cognition of objects.
Although the understanding and not reason provides cognition of objects, reason
provides regulative principles to guide the understanding, and this is what Kant
calls the theoretical employment of reason. Second, reason has a practical
interest, which is an interest in the determination of the will. In claiming that
practical reason has primacy over speculative or theoretical reason, Kant is
claiming that speculative reason must be subordinate to practical reason. This
does not mean that, in the face of conflict between the two, practical reason is
declared the winner; instead, this relationship of primacy and subordination
ensures that the two employments will not conflict, '1. . .] since if [pure
speculative and pure practical reason] were merely juxtaposed (coordinate), the
first would of itself close its boundaries strictly and admit nothing from the latter
into its domain, while the latter would extend its boundaries over everything and,
when its need required, would try to include the former within them" (5:121). A

proposition that cannot be affirmed by speculative reason may be affirmed by

52

practical reason, as long as the practical interest is necessary (arising from
reason itself) and not merely contingent on pathological conditions such as
happiness . After all, Kant reminds us, "L . .] it is still only one and the same
reason [. . .1" (5:121).

Kant argues that it must be speculative reason that is subordinate, and it
is impossible for practical reason to instead be subordinated, because "all
interest is ultimately practical and even that of speculative reason is only
conditional and is complete in practical use alone" (5:121 ). In stating that the
interest of speculative reason is complete in practical use alone, Kant is pointing
out that reason's striving toward the Leibnizian complete cognition cannot be
fulfilled by its theoretical employment. The theoretical ideas of reason provide a
direction for speculation, not an actual cognition. In its practical employment,
however, reason can assume those ideas that are forever out of cognition's
reach.

Kant's statement that "all interest is ultimately practical" suggests an
additional point: that although reason has a theoretical employment, this
employment would be impossible without practice. When reason provides

theoretical ideas, what it is doing is providing a direction for speculation and

53

research. It is commanding a certain scientific endeavor. Thus, even in its
theoretical employment, reason is ultimately a practical faculty.

Kant's discussion of the primacy of practical reason has to do with the
relationship of subordination between the two interests of reason (which, we are
reminded, stem from one and the same reason), rather than the relationship of
logical priority suggested by O'Neill. He does not paint a picture of one part of
reason constructing the other, but of one reason possessing two potentially
conflicting interests. The primacy of practical reason is not that it comes first, but
that it is supreme.

Whether providing regulative or practical principles, reason's defining
activity is the setting of ends. Kant defines "end" as "what serves the will as the
objective ground of its self-determination"; it is what the will points its actions
toward. Kant describes this as self-determination because these ends are
generated by a rational being's own reason. The understanding does not set
ends; it just processes the world and discovers facts. Only reason can set goals
and prescribe laws of action (including, in its theoretical employment, setting the
goals of scientific programs). In giving this role to reason, Kant deprives it of

cognitive authority but transforms it into an active faculty, as opposed to the

somewhat passive faculty of the understanding. It is not the mechanical,

automatic reason of Leibniz, nor is it limited to the narrow scope of formal logic.

5. Animals and Reason

Since animals do not have understanding, they certainly do not have
reason (in the specific sense and, therefore, in the general sense) (see 29:879).
Kant's position on animals and reason is made clear in several passages. He
claims that when a human being uses reason merely for the furtherance of his
own happiness, "reason. .. serve[s] him only for the sake of what instinct
accomplishes for animals; reason would in that case be only a particular mode
nature had used to equip the human being for the same end to which it has
destined animals [. . .1" (5:61). Kant also states that

[. . .] the ground of the artistic capacity in animals, designated as instinct,

is in fact specifically different from reason, but yet has a similar relation to

the effect (comparing, say, construction by beavers with that by humans).

- Yet from the fact that the human uses reason in order to build, I cannot

infer that the beaver must have the same sort of thing [. . .]. (5:464n)

Both passages indicate that although animals accomplish things that are the

result of reason in human beings (construction and engineering, survival and the

55

fulfillment of desires), we must not infer that animals also use reason to achieve
these results. Animals have instinct to guide their actions, instead of reason.
Instinct can account for a wide variety of animal behaviors, including such
complex procedures as the construction of dams by beavers, or the cutting and
curing of hay by pikas. It would be a stretch, however, to attribute all animal
behaviors to instinct. Most animals are able to incorporate novel elements into
their patterns of behavior — in other words, they exhibit learning. A dog will not
instinctively know its friends from foes; nature cannot implant the knowledge of
which humans are going to kick and throw stones at it, and which humans are
going to give it food. Kant's model of animal minds can account for such basic
learning by association. In a letter to Marcus Herz (May 26, 1789) he claims that
if he had no understanding, his sense data
[. . .] could still (if I imagine myself to be an animal) carry on their play in
an orderly fashion, as representations connected according to empirical
laws of association, and thus even have an influence on my feeling and
desire [. . .1. (11 :52)
So, an animal is essentially a Humean learner. It does not have reason, but it
has empirical laws of association that allow it to connect certain sense data (the

satisfaction of being fed) with other sense data (the appearance of a certain

person) through repeated conjunction.

Aside from the merely logical employment of reason, the primary function
of reason is to set ends (as discussed in Section 4). Kant writes that only
through our humanity, not our animality, are we capable of setting ends for
ourselves (6:387). Animals do not set ends for themselves, but act according to
the ends given to them by nature." Human beings, because they are rational,
have the ability to choose ends instead of being moved by impulses. Both
animals and humans have a sensible power of choice (meaning a power of
choice that is pathologically affected), but the human power of choice is free,
whereas the animal power of choice is pathologically necessitated (A 534/8 562).
It seems strange to refer to a power of m as necessitated. The word is
mung, formerly often translated as "will," now commonly translated as "power of
choice" to distinguish it from Wfle. Animals have no will (Wfle), as that term
refers to the autonomous power of reason. mm is translated as "power of
choice" because it deals with a lower-order capacity: the simple freedom to
choose one action over another (hence the paradox of "choosing to be unfree,"

that is, of choosing the path of heteronomous impulses instead of the path of

 

4 Strictly speaking, Kant denies that animals have ends at all, since he defines "end" as "an
object of the choice (of a rational being)" (6:381). I will argue in Section 5 that Kant should not
have limited the term to rational beings.

57

autonomous reason). Animals are not free in any sense, because they do not
have a faculty of reason to present them with any alternative to the
heteronomous path. Their power of choice will always "choose" that path; hence
they have a pathologically necessitated power of choice. In the case of animals,
it might be better to think of it as a "power of decision," the capacity of an animal
to commit to an action based on its desires. The animal may be unsure which
course of action will best satisfy its desires (such as a dog that sees its favorite
toy on one hand, and its favorite person on the other, might momentarily stand
between them looking confused), but once one impulse or the other emerges as
the stronger, the animal is committed to that action. A human being, on the other
hand, can choose to satisfy his or her desires, to delay them, or to defy them
entirely; hence the human W is truly a power of choice.

Because of Kant's rich conception of reason, denying it to animals goes
beyond the simple claim that animals lack logic. An irrational animal by necessity
is unable to set ends for itself, and therefore unfree. Furthermore, since reason
is the source of morality, animals cannot act morally. The link between reason
and morality is deep and important, and will be explored further in the following

chapter.

6. Expanded Kantian View

Before concluding this chapter, I will examine the question of whether it is
possible to attribute greater mental faculties to animals than Kant did, while still
remaining faithful to his general mental framework. Given contemporary
understanding of animal abilities, we may arrive at different conclusions
regarding animals' possession of understanding or reason. Thus even if we
retain Kant's theoretical position that rationality is the criterion of moral
considerability, we may differ from Kant in the empirical evaluation of whether
some animals would qualify as rational.

First, let us consider the understanding. Evidence that Kant denies
understanding to the animals has been presented in section 5. The reason for
this denial is twofold: first, animals do not use concepts, which is the essence of
the understanding; second, animals lack self-consciousness, which is a
prerequisite for understanding. If evidence suggests that these claims are false,
then understanding can be attributed to animals.

The belief that animals do not use concepts is common still. Since

animals do not use language, an inherently conceptual activity, it is assumed that

59

they are unable to use concepts. (This belief actually has two different forms.
Some hold the Cartesian belief that the inability to use language is symptomatic
of the inability to use concepts; others regard the possession of language as a
prerequisite for having concepts.) Recent attempts to teach languages to
animals may call this belief into question. The most famous cases are those of
apes that have been taught to use nonvocal symbols such as hand signs or
pointing to glyphs. Whether these activities constitute language use is a matter
of much controversy. Many of the most interesting uses of language attributed to
such "talking apes" are only anecdotal. Perhaps the most famous among them is
Koko the gorilla, but a transcript of Koko's "live chat" on the Internet (April 27,
1998) appears to show Koko making mostly random signs, with her trainer
offering an ad hoc explanation for each non sequitur.

A striking instance of a talking animal is Alex, an African gray parrot who,
according to researcher Irene Pepperberg, is able to understand some human
language and to use language meaningfully. Pepperberg claims that Alex can,
for instance, respond correctly to requests such as "give me the orange square"

and other phrases containing familiar words in novel combinations, and can

correctly answer "what's same?" with high accuracy when presented with a
variety of items with a particular repeated characteristic (such as color).

If Alex can indeed respond correctly to such commands and questions,
then it would imply that he is able to recognize discrete concepts and combine
those concepts in various ways. Pepperberg's research has even greater
implications for our conception of animal minds than the previous experiments
with apes, because rather than just acknowledging that our closest evolutionary
relatives — those generally regarded as the "highest" of the nonhuman animals -
might have some ability to use language, it suggests that such abilities are
present in other areas of the animal kingdom.

A thorough analysis of Alex and other cases of "talking" animals is beyond
my scope here. However, I believe that the phenomenon, though fascinating and
worthy of further study, does not have much impact on my analysis of whether
animals have understanding in Kant's sense. The claims made by Pepperberg
and others have not yet received wide scientific acceptance, nor have they been
adequately replicated by other laboratories; that aside, even assuming that a few
animals have demonstrated the ability to use language, it would not prove that

animals in general, or even a significant percentage of animals, are able to do so.

61

The use of language is evidence of the ability to use concepts. On the
other hand, possessing concepts does not necessarily entail the ability to use
language, in Kant's view. Kant distinguishes between empirical concepts and
pure concepts. The former are "borrowed from" experience, and the latter are
the concepts of the understanding that constitute "the form of the empirical
cognition of objects" (A 220/8 267). Empirical concepts are those such as
"orange" or "square" - the kinds of concepts that Alex the parrot is supposed to
be able to use. The use of empirical concepts presupposes the existence of the
pure concepts, because without the pure concepts nothing would be an object for
us at all. We must first be able to identify something as an object before we can
determine if it is red or a block. Using concepts in our language is a sure sign
that we possess understanding. On the other hand, concepts are not inherently
linguistic. The concepts of the understanding, for Kant, shape not just our
language but our very world. Their role is very deep, in that they serve to
constitute our experience, yet very basic, in that they merely provide the concept
of an object as such. It is possible to imagine a creature with understanding but
not language. It would experience a world of objects much as we do, but would

not have the ability to describe the relations between the various objects. This

62

could be because it has difficulty forming empirical concepts from its experience;
it is not contradictory, at least, for a creature to experience the world objectively
through pure concepts, but lack the tools to create further concepts from its
experience. Such a creature probably would not be able to use language, since
most meaningful statements involve empirical concepts. Or, a creature might
have pure and empirical concepts, but lack the logical skill to describe the
relations between these concepts. The role of the understanding is to bring
concepts to intuition and to make the experience of objects possible.
Discovering the relations between concepts, however, is left to reason in its
logical employment. It is possible to imagine a creature that has a conceptual
experience but completely lacks the ability to relate those concepts to each other
logically. Such a creature would have a limited use of language at best because
much of our language is devoted to logical relations.

The use of language is not, therefore, required for understanding,
although it may be evidence of it. In any case, if the reports of language use by
animals are accepted, it only means that a few exceptionally intelligent species of
animals can use language; and perhaps only a few exceptionally intelligent

members of those species, at that. A more pertinent question is whether the

63

 

 

 

mall

the

tug

bes

UIIC

the

W0

Kai

she

Ihrc

wht

UP.

SUCI

Ioca

Sam

majority of animals (or perhaps some selected large group of animals such as
the vertebrates, or the warm-blooded animals) could have understanding. I have
argued that lack of language does not entail lack of understanding. What
besides language, however, might provide evidence that animals do possess
understanding?

Since our experience is constituted by, and hence saturated with, the
categories of our understanding, it is fair to say that a being that did not have
these categories would not really live in the same world as we do. Their world
would not be made up of objects, so they would not perceive the objects that we
perceived. If we assume that animals have sensibility but not understanding (as
Kant asserts), then they do not join us in experiencing our world of objects. They
share the same raw intuitions, but they are not processed as cognitions. When I
throw a stick to a dog, we both watch the stick fly through the air and land. I see
where the stick has landed, and so, apparently, has the dog; he runs to it, picks it
up, and brings it back to me. If Kant is right, the dog does not perceive a stick as
such. Yet somehow it still follows the same movement I follow, and is able to
locate the stick just as I do. Because animals seem to recognize roughly the

same set of objects in the world that we do, we may be inclined to assume that

they must have a similar understanding. Kant, however, cautions against this
kind of argument: we cannot assume that because beaver dams are reminiscent
of human constructions, beavers too must use reason in creating them. Rather,
nature provides animals with instincts that achieves some of the same results
that reason achieves in us. Kant might have extended this argument to the
understanding. Our faculty of understanding allows us to process the world as
objects, and so it makes our recognition of the stick possible. The dog behaves
as though it recognizes the same object, but instead this parallel behavior is
rooted only in the empirical laws of association that Kant attributes to beings with
sensibility but not understanding (letter to M. Herz 11:52).

One might respond by pointing to Kant's claim that "in the natural
constitution of an organized being... we assume as a principle that there will be
found in It no instrument for some and other than what is also most appropriate to
that end and best adapted to it" (4:395). Since the self-preservation of animals is
always one of the ends that nature has given them, and since the ability to
recognize features of the world (prey animals, water, places of shelter, and so
on) is instrumental for self-preservation, one could argue that Kant should have

recognized that animals must have a faculty of understanding because it is much

65

better suited for the recognition of such objects than is sensibility alone.
However, this line of argument once again fails to acknowledge that the same
and could be equally well achieved by more than one faculty. Sensibility and me
laws of association might serve animals just as well for recognizing and
responding to relevant features of the world, as understanding serves us.
Examples of such parallel adaptation certainly exist in nature. The eyes of a
squid serve roughly the same purpose as the eyes of a human being. Squid
eyes are extremely complex for an invertebrate and they are able to focus, see
colors, and generally perform most of the same functions as human eyes. But
they are constructed completely differently; the function is the same, but the
manner in which they function is completely different. Perhaps the same might
be said for the human understanding and the animal laws of association.

Still, there is something suspect about this conclusion. First, although
many animal behaviors can be adequately explained by laws of association,
some behaviors are hard to fit into that mold. Consider the recent case of Betty,
a crow at Oxford University that figured out how to bend a wire Into a hook in
order to retrieve a piece of food that was out of reach (Weir et al. 981 ). The bird

had not been trained to do so, and had not observed anyone else using a piece

of wire in this way. It is difficult to imagine how this insight could have been
accomplished by an animal that had no ability to conceive of the hook as an
object. Such glimmers of insight are not well-explained by laws of association,
and seem to require the ability to understand objects and how objects can relate
to each other.

Furthermore, if an animal such as a dog can, through sensibility alone,
respond to roughly the same features that a human being responds to, and if it
behaves as though it is reacting to the same objects and events that we react to,
then why assume that humans have understanding at all? The Transcendental
Deduction rests on the claim that the categories of the understanding are
necessary for the possibility of experience. If a dog has something very like our
experience but does not have understanding, then the Deduction loses some of
its power.

I conclude that, unfortunately, without a first-person perspective - or even
a firsthand report - on what an animal's consciousness is like, we have no way of
concluding either that an animal has understanding, or lacks it. Why does Kant
conclude that the animal lacks understanding? His main argument, according to

Naragon, is that "Since it is possible to think of the animal behavior as occurring

67

111119.11! the use of judgments, we will assume that it does. (With humans, on the
other hand, we are directly aware of our own acts of judging, and so cannot
proceed similarly with our own behavior)" (_R_e_asga 104). Such an approach
makes good sense from the standpoint of reason. Reason demands that we
always strive to find unity of principles in the natural world. Since the evidence
strongly suggests that animals have sensibility, and since it is possible to
attribute all of their behaviors to that faculty, then it is best to do so. Then
science can obtain a more systematic understanding of animal behavior than if
we suppose that perhaps some animals also have understanding.

Two considerations should be weighed against this argument, however.
First, as I have suggested previously, as more complex animal behaviors are
uncovered by science, it becomes harder to account for them adequately through
more association of representations. Rational inquiry may be better served by
leaving open the possibility that animals possess understanding, even though it
means sacrificing economy of principles. Second, in the context of ethics (rather

than science), serious transgressions could be committed by making

unwarranted assumptions about the sort of mind that a creature has.5 Kant, for
reasons that will be discussed in the following chapter, places animals outside
the moral sphere, and so would not regard them as beings that can be
transgressed against. He does, however, regard kindness toward them as an
indirect duty. What constitutes cruelty and kindness to a creature may depend
on what sort of mind it has, so the question of animal minds could still hold some
moral importance in Kant's doctrine. For these reasons I think it is more
appropriate to reserve judgment on the question of whether animals, or some
animals, have understanding.

Note that if Kant had acknowledged that animals possess understanding,
it would have entailed that they are self-conscious. However, the meaning of
"self-conscious" is not the common usage which often includes a robust self-
conception. The self-consciousness, or apperception, that Kant regards as a
condition for understanding is a relatively bare concept, devoid of any content

apart from the simple representation "I." It is the ability to recognize that all of the

 

5 Note that Kant does not dispute that animals have some kind of mind, and I have argued that
he even attributes consciousness to them. And so, when I argue that we must be careful when
assuming what faculties animals have or do not have, this does not entail also giving tables and
Chairs the benefit of the doubt. Once we have evidence, however, that a being is conscious and

can therefore be affected by what is done to it, then we must temper our assumptions.

69

representations in the manifold of intuition belong to one subject, ma. It does not
require any greater self-understanding than that, and is therefore not implausible
to attribute to animals (even though Kant specifically denied that they possess it).

Whether or not animals have understanding is not such a concern for Kant
as whether or not they have reason, since reason is required for moral worth. It
is important to recall what reason means for Kant. It is not a merely cognitive
faculty like the understanding, and it also goes beyond logic. Therefore, showing
that animals have conceptual abilities or even some logical abilities is not enough
to establish that they have reason in Kant's sense.

Consider the story of Chrysippus's dog, which comes to us through Sextus
Empiricus. Supposedly, the dog comes to a three-way fork in the road while
chasing his prey. He sniffs down the first and second paths, and does not small
the game, so he immediately runs down the third without checking for the scent.
Apparently, the dog has performed a disjunctive syllogism.

Does this anecdote have any truth? Perhaps not. But if it did, it would at
most indicate that some animals have a weak faculty of reason that permits a

logical employment. It would not show that animals possess reason in the full

70

sense. That would require animals to be capable of setting ends for themselves,
rather than merely using reason as a tool to achieve externally-imposed goals.

It is unlikely that animals employ reason in the regulative sense, since
they are not engaged in the business of scientific systematization. A more
interesting question is whether animals show any signs of employing practical
reason. Practical reason has two modes, for Kant: prudential, and moral.
Prudential reason is the use of reason to generate hypothetical imperatives.
These imperatives prescribe the best course of action to take to satisfy a desire:
"If you desire x, then you should do y." Reason in this prudential mode is not
determining the will directly, but only mediately through the feelings of pleasure
and displeasure (5: 24-25).

It could be argued that some animals do show signs of using prudential
reason. Betty the crow figured out how to bend a wire to retrieve a piece of food.
This could be the result of some low-level prudential reasoning: she wanted that
piece of food, and determined that bending the wire was one way to get it. My
remarks about understanding also apply here: many animal behaviors are not

well explained by either instinct or laws of association.

71

If some animals, however, do have some small faculty of reason that they
can employ prudentially, this does not show that they are able to use reason in
the full sense. The most important and highest use of reason is to set ends. In
creating a tool to retrieve food, Betty the crow may show admirable insight, but
the end she obtains is not one that she set for herself. Nature has already
chosen for her. It is the destiny of crows to scavenge food. It may at times be up
to them how to scavenge food, but they cannot choose not to do it at all. Rather

than "if you desire x, then you should do y," Betty's imperative is "you desire x, so
do y." Her ends are not autonomous, but imposed by her nature.6

Applying the language of "ends" to animals even in this limited sense is
partly at variance with Kant's stated position. In the Grganflork, after defining
the will (Mia) as "a capacity to determine itself to acting in conformity with the
W," Kant defines "end" as "what serves the will as the
objective ground of its self-determination" (4:427). This would exclude animals

from having ends, since animals do not have a will of that kind. However, he

 

6 Kant does hold that there is one end that is imposed upon human beings by their nature: their
own happiness (5:25). He points out, however, that happiness is an end only in a very general
sense; what constitutes happiness varies from person to person, and so do the particular ends
that each person adopts in furtherance of his or her happiness (5:25). In any case, the important
thing is that we can adopt ends besides our own happiness, moral ends, and that is the important
difference between us and the animals.

72

then goes on to distinguish between "subjective ends, which rest on incentives"
from "objective ends, which depend on motives" (4:428). Incentives, in Kant, are
sensuous, whereas motives are rational. Nonrational beings could not have
motives, and hence objective ends, but there does not seem to be any reason to
deny that nonrational beings have subjective ends. This is especially true in light
of the different definition of "end" given in the Metaphfiias Qf Morals. There Kant
defines it as "an object of the choice (of a rational being), through the
representation of which choice is determined to an action to bring this object
about" (6:381 ). It is not clear whether the parenthetical restriction to rational
beings is meant to imply that only rational beings have ends, or whether he is just
limiting the current discussion to the ends of rational beings (since the rest of the
passage deals with matters of moral obligation that would not apply to
nonrational beings). The interesting difference between the Gmuadwgfls
passage and this one is that the former relates ends to the will (Me), and the
latter to choice (Willk'ui). Animals do not have a will in the true sense, but they
do have a power of choice; despite Kant's apparent denial, it would make sense
to regard the objects of the animal power of choice as "ends" in analogy with the

objects of the human power of choice. The animal power of choice impels

animals toward some goal, and the animal is determined to do what it can to
achieve that goal. The only difference between that and how it works in rational
beings, is that our power of choice is free and our ends are self-selected: "I can
never be constrained by others to_l'ial_/_e_an_e_rlg: only I myself can maka
something my end" (6:381). Animals, on the other hand, are constrained by
nature to have certain ends, and their power of choice is always directed at those
ends.

Animals are consequently incapable of adopting moral ends. They cannot
act according to the moral law, and thus cannot use reason in its highest
practical employment. Even the highest animals show little sign of true morality:
they may have social restrictions on their behavior; they may sympathize with
those who are hurt, or protect those that they have affection for. All of these
behaviors are explainable by instinct; and even if they weren't, they are still
based on desires, which makes them nonmoral for Kant. The kindness that one

animal has for another, as Kant says in a different (but related context), "however

amiable it may be, has nevertheless no true moral worth" (4:398).7 And, of

E

7

Kant is describing persons who are "so sympathetically attuned that I. . .] they find an inner
satisfaction in spreading joy around them [. . .1". Such a person would be similar to an animal that
is kind to its peers because it gets pleasure out of such behavior.

74

course, many animals are capable of inflicting great suffering on those of other
species and even of their own species. We do not blame them for this because
we recognize that it is not an end that they adopted for themselves, but is rather
the result of the ends that nature has given them. It is not a moral matter at all,
however unpleasant it may be.

What would it mean for animals to have some ability to use logical and
prudential reason, but no ability to use moral reason? These are not separate
faculties, for Kant, but different ways that we employ the faculty of reason.
Furthermore, as noted by O'Neill, the Categorical Imperative is the supreme
principle of practical reason, and practical reason has primacy over theoretical
reason. If animals do not act according to the Categorical Imperative, does it
follow that they must not have reason? On O'Neill's interpretation, it seems that
it would, since she holds that the primacy of practical reason (and the Categorical
Imperative) means that a species would have to be able to act according to the
moral law before it could begin to develop any other aspects of reason. On my
own, more conservative interpretation of the text, a deficiency in practical
employment does not rule out the possibility of theoretical or prudential

employment. The primacy of practical reason means that if a being has the

75

ability to make moral judgments then its theoretical or prudential judgments must
be subordinate to them, but it is neutral on the subject of whether a being could
in fact be limited to theoretical or prudential judgments. Perhaps animals could
be said to have an immature or deficient reason that admits of only some limited
employment. However, the most important component of human reason, and
what makes moral reasoning possible, is the ability to set ends for oneself, which
animals lack.

I conclude that it is possible, and largely consistent with Kant's theory of
mind, to ascribe understanding to animals, and furthermore that it is consistent
with current data on animal behavior. Furthermore, I conclude that it is at least
possible that animals may possess some ability to employ reason, but only in a
lower sense. Since Kant does not regard reason as the cognitive or merely
logical faculty that his rationalist predecessors did, no matter how many amazing
and intelligent things animals are capable of, and no matter what logical or
conceptual abilities they may have, it will not be enough to show that they have

reason in the full sense.

76

7. Conclusion

Kant did not regard animals as being self-conscious or possessing
understanding, and therefore also did not regard them as rational. l have
argued, with Naragon, that he does regard them as conscious and therefore as
possessing minds. Although I believe that he could have ascribed understanding
and even some low-level reason to them, and given current information about
animal behavior it would be more plausible to do so, I agree with the most
important point, that animals are not capable of setting ends for themselves. In
the next chapter, I will explain why Kant regards this capability as being a

requirement for moral consideration.

77

3. RATIONALITY AND MORAL AGENCY

1. Introduction

In the previous chapter, I discussed Kant's position on animal minds:
animals, according to Kant, have sensibility but lack both understanding and
reason. I argued that Kant was too quick to deny these faculties to animals, as
the evidence suggests that many animals possess understanding and even a
limited prudential reason. However, even if animals can reason to some degree
(as l have claimed), it would not be enough to make them moral agents for Kant.
Moral agency in Kant's philosophy requires a specific kind of rationality that
animals lack. Thus while disputing some of Kant's empirical findings about
animal minds, l ultimately concluded that he is correct to deny that they are fully
rational.

In this chapter I turn from the empirical question of whether Kant was right
about the nonrationality of animals, to the theoretical question of whether he
should have excluded nonrational beings from moral consideration. Since Kant
links moral agency with moral considerability (i.e., moral patienthood), he denies

that any direct duties are owed to animals. In this chapter I will explore this link

78

 

3:-
I1

‘Tfi

by discussing the relation between rationality, moral agency, and being an end in
oneself. I will show that animals, since they lack the ability to reason morally,
cannot be ends in themselves, but I will question whether being an end in oneself
should be required for moral consideration. Kant's strict correlation between
moral agency and moral patienthood would exclude even some human beings

from moral consideration, a result that most would find unacceptable.

2. Reason and Morality

When one thinks of Kant's ethics, one thinks of the categorical imperative,
and rightly so: all our duties are derived from it. Kant argues for it as follows in
the Mk. The only thing that is good in itself is a good will (W_illa).
Anything else that is considered good - for instance, talents of mind, qualities of
temperament, and happiness - are only judged to be good when they
correspond with a good will. Mental talents can be evil and harmful if employed
wickedly, and happiness itself seems like an abomination when it is being
enjoyed by an evil person (4:393). The good will is thus the highest good.
Further, "A good will is not good because of what it effects or accomplishes [. . .]

but only because of its volition" (4:394).

79

A will that is not subject to selfish temptations would be a holy will, such as
God would presumably have; but as human beings we have a strong natural
instinct to promote our own happiness. The human will must be commanded to
do good; this command is called "duty." (In contrast, a holy will is not subject to
duty, because it always does the good automatically and does not need to be
commanded.) (4:397; 4:414) Kant defines duty as "the necessity of an action
from respect for law" (4:400). Since only the will is good in itself, the moral worth
of an action cannot lie in the expected effects of an action, but only in the will
behind an action. The law that duty commands us to follow, therefore, cannot be
derived from empirical effects (such as happiness). For principle to have the
binding force of law, it must be necessary and not contingent, but all empirical
conditions are contingent. If we strip away all the empirical effects, the only thing
left that could determine the good will is "the conformity of actions as such with
universal law, which alone is to serve the will as its principle" (4:402). In other
words, it is not the (empirical) matter of the law that determines us, but its form;
we judge our maxims by their correspondence with this form.. Thus, "Act only in

accordance with that maxim through which you can at the same time will that it

become a universal law"; this, along with its various formulae, is what Kant calls

"the categorical imperative," and all particular duties are derived from it (4:421 ).

According to Kant, this "L . .] W in itself [. . .] QQLLQI
nl in r “on I i " (4:401). This is because "Only a rational
being has the capacity to act WW laws. that

is, in accordance with principles, or has a will" (4:412). Rationality and morality

 

are thus linked in Kant's ethics. Not only is rationality necessary for morality, but
to be immoral is to be irrational, since violating the categorical imperative
involves committing a contradiction. This can be either a contradiction in thought
(universalizing the maxim is impossible), or a contradiction in will (universalizing
the maxim results in a situation in which one's own ends are frustrated) (4:424).
Thus, one must be rational to be moral, but one must also be moral to be fully

rational.

3. Reason and the Will
In light of the link between morality and rationality in Kant, it seems natural
to refer to reason as the source of the categorical imperative and thus of morality.

Kant does do this at times (for instance, 5:31 ), but more often he refers to the will

81

as giving the moral law to itself (for instance, 4:431-433). This is no
inconsistency, but a product of the relationship between reason and the will. The
will (Willa) turns out to be a guise of reason: "Only a rational being has the
capacity to act W291) of laws, that is, in
accordance with principles, or has a will. Since Leaaga is required for the
derivation of actions from laws, the will is nothing other than practical reason"
(4:412). Kant also defines will as "a kind of causality of living beings insofar as
they are rational" (4:446). Essentially, "will" is what Kant calls reason in its role
as the source of actions.

If will is identical with one kind of reason, and reason is linked with
morality, then we are left with a puzzle. What is the source of evil? It cannot be
the will, since the will is practical reason and should therefore always do what is
moral. A possible solution to this puzzle is found in Kant's distinction between
Mia and _W_illfllr. Both can be translated as "will," but more recently the former
is usually rendered as "will," and the latter as "choice" or "power of choice." Mla
can be seen as the Iegislating will, whereas M can be interpreted as the
choosing or deciding will. This interpretation is appealing because it fits well with

Kant‘s claim that the Willa is identical with practical reason. It is also well

82

supported by passages in the W. There Kant contrasts the

MM"! as the faculty of desire considered in relation to action, with the [Vila as
the faculty of desire considered in relation to the determining grounds of choice.
"The will (yl/flle) itself, strictly speaking, has no determining ground; insofar as it
can determine choice (M), it is instead practical reason itself" (6:213). The
picture he gives there is one in which the V_V_ill_e, through the categorical
imperative, legislates moral action, and thus can (and should) serve as the
determining ground of the MIKE of a rational being. Unfortunately, the human
Wi_lllg.1r can also be affected by nonmoral sensible impulses. Thus, although the
WM can choose evil, the V_l/ll_la is the voice of reason and so inherently good.
Separating will from choice in this way does make it more difficult to understand
the claim that the Willa gives the law Mt, since in this model it instead gives
the law to a separate faculty, the MM. It may be better to interpret Kant's
remarks about the will giving the law to itself as being a shorthand for the idea
that rational beings give the law to themselves (since both faculties are united in
the mind of one individual) through the legislation of the Ma.

Although this doctrine is explicit in the Metaphysics of Morals, it is not

clearly in place in earlier works. Sometimes Kant seems to use Wille to refer to

what he later termed MIME. while also using it in the Iegislating sense at other
times. In the Gragngwark, for instance, the entire discussion of the worth of a
good WMa seems to presuppose a contrast with an evil, or bad, Willa. The
contrast is even explicit in some places: for instance, Kant states that certain
talents of mind "are undoubtedly good and desirable for many purposes, but they
can also be extremely evil or harmful if the will [\_/V_illa] which is to make use of
those gifts of nature [. . .] is not good" (4:393). Yet in the same work Kant
identifies the Wma with practical reason (4:412), implying that it is inherently
rational and moral. Since the Wifla/Wiflkfll distinction resolves this tension, and
since it is present explicitly in a later (and thus more mature) work, I believe it
should be regarded as Kant's true doctrine.

The will (Wflla), then, is practical reason; the will gives the law of morality
to itself; therefore, reason is the legislator of morality. The ability of the rational
will to govern itself through the legislation of morality gives rise to the concept of

autonomy.

4. Autonomy

Autonomy is a very important concept in Kant's ethics, and once again it is
linked to reason, since it is linked to the Iegislating will. Kant writes, "autonomy of
the will is the property of the will by which it is a law to itself (independently of any
property of the objects of volition)" (4:440). As discussed in Section 2, reason
(as Willa) is the source of the categorical imperative. Maxims are judged to be
good if they have the correct form according to the categorical imperative,
abstracted from their content. Since our own will is giving the laws, we are
considered autonomous when we adhere to those laws. On the other hand,
when our choices are determined by empirical matter - i.e., sensible impulses -
we are heteronomous. Kant regards empirical matter as being external to us, so
in allowing ourselves to be determined by inclination (which always has some

empirical matter as its object), we are not autonomous.

Kant has two conceptions of freedom, negative and positive. The

negative conception is that of the freedom of the Willkiir. The animal Willkiir

 

 

(amilriam brgtgm) is always determined by sensible inclination. The human
Willki’lr has the potential to be determined either by inclination or by reason

(through the Wille) (6:213-214; cf. 41446-447). Therefore, the animal power of

85

choice is unfree. and the human power of choice is (negatively) free. It is a
negative conception because it only refers to the fact that the power of choice is

not determined.

The positive conception is that of autonomy, which is expressed when a
rational being's choice is determined by the Iegislating will rather than by sensible
inclinations. In choosing the dictates of the will, the agent has chosen to be
governed by his or her own will. By choosing autonomy, the agent becomes truly
free through self-determination. Note that this view is compatible with the idea
that an agent can freely choose unfreedom, since the freedom would be that of

choice-freedom, but the unfreedom would refer to autonomy-freedom.

Autonomy is only possible through the categorical imperative, which is
given by reason: "A free will and a will under moral laws are one and the same"
(4:447). Reason, morality, and autonomy are linked. The set of moral agents
and the set of rational beings are the same. Only rational beings can have duties
to others; so far, however, it is still unclear why Kant believed that no duties can
be owed to nonrational beings. The answer lies in Kant's conception of an end in

itself, a status that is linked to autonomy.

5. Rational Nature as an End in Itself

Kant writes that "The human being and in general every rational being
aLisla as an end in itself" (4:428). He defines an end in itself as "something the
axialanm at whiah in il§alf has an absolute worth" (4:428). This is contrasted
with the ordinary kind of ends based on inclination. Those kinds of ends cannot
furnish moral principles, since they are subjective and contingent. On the other
hand, rational beings, as ends in themselves, are necessarily ends for everyone
and thus are objective ends. Beings without reason "have only a relative worth,
as means, and are therefore called things" (4:428). The rational being's quality of
being an end in itself is related to the legislative will: "It is [. . .] this fitness of his
maxims for giving universal law that marks him out as an end in itself" (4:438).
Kant links this with the concept of autonomy, implying that it is the autonomous

quality of our will that makes us ends in ourselves:

[. . .] a human being alone, and with him every rational creature, is an end
in itaalf: by virtue of the autonomy of his freedom he is the subject of the
moral law, which is holy. Just because of this every will, even every
person's own will direct to himself, is restricted to the condition of
agreement with the amonamy of the rational being [. . .] hence this subject
is to be used never as a means but as at the same time an end. (5:87)

We are, as rational beings, the source of the moral law, and we each possess a

will. The good will is the only thing that is good in itself rather than as a means

87

for something else. Therefore, insofar as we are beings with wills, we are ends
in ourselves, with an infinite, rather than relative, worth. The term "dignity" is
used to describe this infinite worth, as it is contrasted with price (i.e., relative

worth).

Kant relates the idea of an end in itself to the categorical imperative. The
basic form of the categorical imperative is "Act only in accordance with that
maxim through which you can at the same time will that it become a universal
law" (4:421). Kant derives several formulae from this, which he claims are all
equivalent. One of these is what I will call "the formula of humanity": "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" (4:429). (The fact
that he formulates it in terms of humanity rather than rational nature in general is
puzzling, but I will not deal with that issue here. It will suffice to point out that
Kant does not limit the quality of being an end in itself to human beings; see

4:428, for example.)

Although Kant regards the formula of humanity as equivalent to the
formula of universal law, it is my belief that it does add something that is missing

from the formula of universal law: it provides guidance in the proper description

of maxims. It is not my intention here to delve too deeply into the problem of
maxim description, although it would be a fascinatingly perilous endeavor. The
problem to which I refer is this: how the maxim of one's action is described
affects whether it is judged permissible according to the formula of universal law.
The same action can be described in various ways, and there is no defined
procedure for determining which is the correct description to use when judging
the maxim. For instance, if I lie to a potential murderer to save my friend's life, is
my maxim "I will lie," or "I will lie when necessary to save a life"? (Kant,
infamously, chooses the first, when intuition seems to recommend the second.)
The first is not universalizable, but the second probably is. Overly general
maxims tend to forbid too much, overly specific maxims are usually too
permissive (e.g., "I will kill people with red hair" - which I can will to be a

universal law because my hair is brown).

One particular area of ambiguity is the question of which beings to take
into account when formulating one's maxim. If I go into the woods and shoot a
deer, have I violated a stricture against killing? Kant would say no, but the
formula of universal law does not readily provide this response. My maxim could

be, "I will kill other living things," which I cannot will to be universal law, since I

89

too am a living thing and would be killed. Kant must bring in a concept of
personhood in order to show why that maxim is badly framed. Kant equates
personhood with being an end in itself. Our maxims should make reference to
persons as such. This explains both why "I will kill others" is a too-general
description of hunting and "I will kill people with red hair" is a too-specific
description of a murder (since it makes a mere empirical attribute the focus,
rather than the victim's personhood). Once the concept of personhood is used to
guide what counts as well-formed maxim, then the formula of universal law

comes into line with the formula of humanity.

Through the concept of an end in itself, it becomes clear that not only are
rational beings the only possible moral agents in Kant's philosophy, they are also
the only beings worthy of moral consideration. Without will and thus without
autonomy, nonrational beings can only have a relative value. Rational beings, as
ends in themselves whose value is absolute, are permitted to use nonrational

beings as means.

6. Animals, Reason, and Morality

Kant claims that animals lack reason and even understanding. Therefore,
it is obvious that he believed animals could not be moral agents. Without reason,
they cannot act according to principle or have a will (4:412). Thus they are
things, not persons; they have no moral worth and deserve no consideration in
our moral judgments, except insofar as our treatment of them indirectly harms
rational beings. Kant's official position on this subject is cut and dried, and
explication seems unnecessary. Instead, I will turn to considering how animals
would fare in Kant's moral theory given the expanded view of their mental

abilities that l have suggested in the previous chapter.

I have argued that the evidence favors animals having understanding at
least, and some amount of reasoning (less than what is possessed by human
beings) at most. Since Kant makes rationality the center of the moral universe,
the possession of mere understanding does not gain animals any further ground
than if they only had sensibility (as in Kant's official position). But what if animals
do have some reasoning powers? I have claimed that, insofar as some animals
display reasoning ability, it is of the problem-solving, means-ends variety. In

other words, given some desire that an animal has, it can use its limited

91

reasoning powers to discover how to satisfy this desire. New Caledonian crows,
for instance, have been known to make and use tools in the wild. During a
laboratory experiment at Oxford University, Betty the crow spontaneously bent a
piece of wire to use in retrieving a food item (Weir et al. 987). She had not used
or been shown how to use wire before. Her behavior is strong evidence of
problem-solving abilities. This is a type of practical reason that Kant calls

"prudential reason."

Prudential reason cannot be the basis of morality, since it incorporates
empirical matter in the form of sensible impulses. It can provide hypothetical
imperatives (if you want X, then do Y) but not a categorical imperative. Its
imperatives are based on contingent desires, and so they can never have the
necessity required for a categorical imperative. Betty may be able to learn rules

for the best ways to satisfy her desires, but she cannot act on principle.

The limited reason that animals have cannot be called a will (Willa), since
they cannot formulate principles and are not in the business of Iegislating
morality. There is a possible problem here. Kant equates the will with practical
reason. If animals such as Betty have prudential reason, which is a kind of

practical reason, then how is it possible that they don't have a will? This no

greater (or less) a problem than the existence of a similar state of affairs in
children. A young child might be able to figure out how to open the cupboard in
which there are cookies, but not how to determine whether or not stealing a
cookie is wrong. (She might fear punishment, but that is not the same as
understanding why something is wrong.) Although the child will someday have a
will, it is not yet developed. Animals can be thought of as evolutionary children.
Their reason has reached only a certain stage of development; it is immature
reason. The difference is that theirs is fixed at that point; a child and a crow
might be at the same stage of rational development, but the child will soon leave
the crow behind. The point, however, is that such a state of partial rational

development is possible.

Some animals might also seem to have a partial morality. Capuchin
monkeys, for instance, have exhibited a sense of fairness in laboratory studies: if
a monkey knows that it has been given a lesser reward for the same work as
another monkey, it will react with anger, even if the reward is one it would
normally be happy to accept (Brosnan and de Waal 297-299). Monkeys who
saw their neighbor getting a grape (a favorite treat) when they got only a

cucumber would sometimes even throw the cucumber out of their cage in

protest. Cucumbers are a food they normally like, although not as much as
grapes; so their behavior seems counter to their own happiness, since a small
reward is better than none. It is important to note, however, that the monkey that
got the better reward did not also protest. The monkeys' evaluation of fairness
shows some degree of rationality, but not morality. Ultimately their motivation is
self-interest: if I am going to be cheated, it is not in my interest to play along. In
the short term, they lose out, but presumably their intention is to win out in the
long run by showing they won't tolerate being shorted. They consider the
unfairness only as it relates to themselves, not in terms of whether it is suitable

for universal law.

Although animals do not have a will (Mlla), they do have a power of
choice (mum). Kant writes, "a faculty of choice [. . .] is merely animal (am
mm) which cannot be determined other than through sensible impulses, i.e.
pathologlcally" (A 801 /B 830). Whereas the human power of choice can
choose either morality or self-interest, the animal power of choice will always
follow the perceived self-interest. (As I noted earlier, the term choice is
somewhat misleading when applied to animals, since it is usually thought of as

selecting between options. The animal power of choice - perhaps better termed

a "power of decision" - is what enables animals to take action based on their
perception of what will bring them the most satisfaction, but they will always
choose whichever action they think that is.) Since Kant believed that animals
respond only to representations and not concepts or rational calculations, he
presumably envisioned animals as always attracted to whatever representation
appears to be the most desirable at that moment. I do not believe that it is quite
so automatic, since animals seem to be able to take delayed effects into account,
and to employ problem-solving abilities. I concur with Kant, however, that their
power of choice will always follow what either mere inclination or prudence

(which is based on inclination) recommends.

Animals are unfree in both senses. In the negative sense, they are
unfree because their power of choice has no alternative but to choose inclination.
In the positive sense, they are unfree because they are governed by forces that
are not self-legislated. They are never autonomous, but always heteronomous.
They can, in some cases, use prudential reasoning to achieve their ends, but
those ends are not freely assumed; they are dictated by nature. Kant writes,
"Rational nature is distinguished from the rest of nature by this, that it sets itself

an end" (4:437). This is the essential way in which even the most advanced

95

animals fall short of full rationality. They do not set their own ends, but only do
their best to achieve those that are assigned to them. Because are not free, we
also do not hold them accountable for their actions. They cannot do good, but

neither can they do evil. They are what they are.

Animals, as heteronomous, therefore cannot be regarded as ends in
themselves. They lack dignity, instead having only a price. Hence rational
beings may use them as means to fulfill our ends (as food or beasts of burden,
for instance). They have ends, but are not ends in themselves. Kant concludes

that we can have no duties directly to them.

7. Other Nonrational Beings

Although Kant believes that his system merely codifies what most people
know by intuition, the trouble is that intuition is notoriously variable. Many believe
that some minimal duties, ones that are not merely duties to human beings In
disguise, are owed to animals. Others have no such intuition. Animals, however,
are not the only group that is excluded in Kant's moral theory. A much more
challenging case is that of nonrational human beings. The intuition that duties

are owed to such human beings is much more widely held, to the point that the

96

case could be (and has been) used as a counterexample against the claim that
rationality or moral agency is required for moral considerability. There are
several cases in which a being could be human but lack the necessary level of
rationality for moral agency. The human being could be nonrational currently, but
with the potential to be rational in the future (a child); it could be formerly rational,
but now nonrational (someone with permanent mental debilitation); or it could be
permanently nonrational and have always been nonrational (someone with
developmental disabilities). All are cases in which moral agency is not present,
but most believe we still have some duties, if not exactly the same ones. Usually
such human beings are thought of as moral patients. But Kant does not

recognize the possibility of moral patients that are not also agents.

One way that they could be admitted into the moral circle is to say that
being a member of the human spades is sufficient to gain moral status, that
although rationality is what makes the human species in general valuable, once
we recognize that value we also extend it to nonrational members of the species.
This seems more like a description of our moral intuitions than a justification. If
rationality is the essential element, why do we extend our respect for humanity to

nonrational individuals? Isn't it just a mistaken analogy (of the kind that Kant

thinks we make when we think we have duties to animals)? Admittedly, Kant
seems at times to hold the position that being human is sufficient for moral worth
(e.g., "Humanity itself is a dignity; for a human being cannot be used merely as a
means by any human being" [61462]). Yet this does not seem consistent with his
claim that it is the rational will that makes us ends in ourselves. Species
membership is merely an accident of birth, not chosen by us; it does not have a
necessary relation to autonomy. Kant is clear that non-human rational beings
can be autonomous, but he seems reluctant to admit that nonrational human
beings cannot be autonomous, although it is clearly implied by his theory. In any
case, the moral status of nonrational humans is problem that must be dealt with,

either by abandoning intuition or by explicating Kant's doctrine.

First let us look at the case of not-yet-rational beings, specifically children.
One plausible way to deal with this case is by pointing out that rational nature is
infinitely valuable and necessarily exists as an end for all other rational beings.
Although a young child is not yet rational, we are sensitive to the fact that he or
she is developing that way. Although our specific duties to children will be
different from our duties to adults, we still must have duties to treat them well on

the grounds that doing othenlvise may interfere with the development of reason.

Our actions toward children must always be aiming toward the end of helping
them develop rationality. This is consistent with Kant's remarks on the education
of children. The remaining problem with this is the same as with any argument
based on potentiality: how early on must we begin to regard something as a
potential reasoner? At birth, before birth, as sperm and egg? Taking a cue from

Kant, I will leave such details for others to solve.

A harder case is that of the formerly-rational being. Someone who was
once rational has sustained an illness or injury that renders him or her
permanently incapable of moral reasoning. Is there any grounds for continuing to
regard such beings as persons rather than things? Unlike children, they are not
going to become rational in the future, so their moral status cannot be based on
potentiality. A possible solution is to take a metaphysical turn. Kant claims that
the will of a rational being is outside time, which is how it can be free even in a
deterministic world. He also claims we must assume that the rational being
survives after death. So, it would seem that the rational part of a person cannot
pass out of existence. For Kant, although the phenomenal aspect of a person is
all we can know and interact with, it is really the noumenal reason that our moral

respect is directed toward. In a sense, then, the body is just serving as the

phenomenal sign of the real person. Since the rational part of the now-disabled
person is eternal and must still exist, we should continue treating the
phenomenal aspect as its temporal sign and pay it moral respect accordingly,
even though the rational part is no longer being expressed through it. This is an
interesting approach, although it emphasizes the metaphysical aspects of Kant in

a way that some would find unpalatable.

Although formerly-rational beings and children can get a foot in the door
because of what they were or will be, permanently nonrational beings lack such
credentials. Such a being would never be autonomous, and so would not be an
end in itself. A human being born with defective or missing reason would seem
to have the same status as an animal in Kant's ethical theory. Would Kant or his
followers be willing to bite the bullet and say that we may kill mentally disabled
humans as long as it is done painlessly (to avoid damaging the killer's
sensitivities), or experiment on them as long as it benefits rational beings? Is our
repugnance at this idea based only on a mistaken analogy with rational human

beings?

Of course, Kant can take the same route he took with animals: he can

claim that harming such human beings damages our humanity by hardening us

100

to human beings that are rational. Since the similarity between a rational human
being and a nonrational human being is even closer than the similarity between a
human being and an animal, harming nonrational human beings is probably even
more likely to result in the hardening of our sympathy. However, the problems
that exist in the theory of indirect duties to animals (discussed in my first chapter)
also exist in a theory of indirect duties to nonrational humans. For instance, it
seems to permit us to do whatever we want to them to further our ends, as long

as it is done carefully and out of sight.

Kant's failure to make room for non-agent moral patients in his philosophy
results in the exclusion of at least some human beings from moral consideration.
Although my primary concern is with the moral status of animals, I believe I have
shown that some human beings are affected by this aspect of Kant's moral
theory. Unless we are willing to accept some unintuitive and disturbing
conclusions, we must find some elbow room for moral patients in Kant's ethics, if

we are not to discard it entirely.

101

8. Conclusion

In this chapter I have shown how reason, will, and autonomy are the
central concepts in Kant's ethical theory, and how they come together in the
concept of an end in itself. This is an interesting and fruitful concept, but Kant
uses it to equate moral agency with moral considerability. The result is that not
only animals, but even some human beings are excluded from moral
consideration, and thus must be regarded as things. In my next chapter, I will
propose a way for Kant to better account for animals and other moral patients by
suggesting that being an end in itself should be regarded as sufficient, but not

necessary, for moral considerability.

102

4. ANIMALS AS SUBJECTS IN THE KINGDOM OF ENDS

1. Introduction

In Chapter 1, I argued that Kant's position on the moral standing of
animals is weak and not consistent with his own ethical theory. First, it does not
accord with common intuitions (as Kant claimed was a strength of his ethical
theory), since it claims that the duty not to inflict cruelty on animals is not a duty
to the animals, but rather a duty to humanity in oneself and others. This entails
the very unintuitive claim that no wrong is done to the animals themselves when
we mistreat them. Second, the indirect duties to animals proposed by Kant rest
on a contingent claim about psychology - that cruelty to animals damages our
sensitivities and makes us more likely to be cruel to humans - that may not
actually be true at all, or may be highly variable from person to person. Third,
since this psychological phenomenon is, on Kant's view, based on a mistaken
analogy between humans and animals, it would seem that it would be more
consistent for him to recommend that we learn not to make this mistake, instead
of accommodating it. For these reasons I rejected Kant's theory as

unsatisfactory, both in general and in terms of internal coherence.

103

In Chapters 2 and 3 l analyzed why it is that Kant holds animals to be
undeserving of true moral consideration by looking at Kant's theory of rationality
and its role in his moral theory. In Chapter 2 I reviewed Kant's doctrine regarding
animal minds, determining that he regards them as having sensibility but not
understanding or reason. I argued that contemporary knowledge of animal
behavior makes it more likely that they have at least understanding and possibly
some reasoning abilities, but that they still must lack the power of moral
legislation that is the hallmark of full rationality in Kant's system. In Chapter 3, I
showed that Kant regards moral agency as the requirement for moral
consideration. Moral agency for Kant is tied to the ability to set ends for oneself,
which is what makes a rational being an end in itself rather than a means.
Animals are not ends in themselves and thus they are mere things.

In this chapter I will propose a replacement for Kant's indirect duties theory
that will allow human beings (and other as-yet undiscovered rational agents) to
retain their unique status as ends in themselves and the special consideration
that attaches to this status, while also showing that we have duties to animals
that are not merely covert duties to human beings. My theory will make use of

Kant's idea of the Kingdom of Ends.

104

2. Korsgaard on Animals

The obvious approach to forming a Kantian theory of duties to animals
would be to claim that Kant should have recognized animals as ends in
themselves. This is essentially the route taken by Tom Regan in his broadly
Kantian approach to animal ethics, but insofar as he was successful it was by
reconstructing the end in itself into a new concept of moral personhood that he
calls "subject-of-a-Iife." Whatever the merits of this position, it makes no claim of
being a truly Kantian position. Tampering with Kant's conception of the end in
itself would require extensive revision of his ethical theory as a whole.

On the other hand, Christine Korsgaard claims that "end in itself" as
described by Kant would include animals. She argues that the aspect of
ourselves that we value, when we regard ourselves as ends in ourselves, is not
only our rational nature, but also our animal nature. What differentiates the good
for an animal from the good for an object such as a car, is that an animal has a
natural good that matters to it, just as a human-qua-animal has a natural good
that also matters to him or her. Korsgaard holds that a good "matters" to a being
when that being "welcomes, desires, enjoys, and pursues its good" (30). Thus

she implies (as I will also claim later) that for a being to have a good that matters

105

to it, that being must be conscious.‘3 As Iegislating moral agents, we legislate for
that natural good by promoting things that preserve our bodily lives, and
forbidding things that harm, terrify, or destroy us; thus as moral arbiters of the
universe we have declared our animal nature to be valuable (Korsgaard 32-33).
"In taking ourselves to be ends in ourselves we laglslata that the natural good of
a creature who matters to itself is the source of normative claims. Animal nature
is an end in itself, because our own legislation makes it so" (Korsgaard 33).
Having attempted to establish that animals are ends in themselves,
Korsgaard then uses the idea Kingdom of Ends to show how animals fit into a
Kantian moral universe. She holds that animals are "passive citizens" of the
Kingdom, and that the other members can, and ought to, legislate on their behalf
(21-22). Korsgaard claims that because of the shared animality of human beings

and animals, when we legislate for our own protection from such things as bodily

 

8 In a footnote, Korsgaard considers the argument that plants also have a "natural good" that is
"good for" them. It, then, we endorse our own natural good, why does that not include endorsing
the natural good of any functional entity? She acknowledges and accepts the possibility that her
theory would entail this (33h). My own interpretation will emphasize happiness (as the state
resulting from the fulfillment of ends) rather than natural good, and will claim that consciousness
is required for this; thus I will not include plants in the scope of moral concern, except insofar as
we have a duty to appreciate and preserve nature (see 6:443).

106

harm and terror (things that also matter to animals), we legislate these
protections for animals, too.

Although I am in agreement with the general theme of Korsgaard's
interpretation, I believe she has erred in elevating animals to the status of ends in
themselves. She equates the quality of being an end in itself with the status that
rational agents accord themselves through their moral legislation. This, however,
is only part of the story. Being an end in itself has two facets, which are closely
related. The first is, as Korsgaard notes, that an end in itself is
never a mere means, and cannot be used like a tool to achieve other ends
(4:428). The second, which Korsgaard does not here note, is that an end in itself
is a itself source of ends, an autonomous legislator (4:435). These are not
independent ways of being an end in itself; the first is dependent on the second,
as it is our status as moral legislators that makes it impossible for us to ever be
merely a means and thus forbids others from treating us as such: "morality is the
condition under which alone a rational being can be an end in itself, since only
through this is it possible to be a lawgiving member in the kingdom of ends.
Hence morality, and humanity insofar as it is capable of morality, is that which

alone has dignity" (4:435). The law determines the worth of various things,

107

relative to each other, but the lawgiving itself has infinite worth; thus it is our
moral dimension that makes us ends in ourselves (42435-436; cf. 42437-438).
Kant regards autonomous legislation as both necessary and sufficient for a being
to have a dignity rather than a price, that is, to be an end in itself rather than a
means. Kant's intention is not to show that autonomy is one possible criterion,
but that it is the only possible criterion; this is evidenced when Kant claims that "l.
. .] morality is the condition under which alone a rational being can be an end in
itself [. . .]" and that "L . .] humanity insofar as it is capable of morality is that
which alone has dignity [. . .]" (4:435). Everything else can be used merely as a
means (5:87).

Korsgaard's claim that animals are ends in themselves does not,
therefore, sit well with Kant's theory. If we confine ourselves to just the first
sense of "end in itself" (something that is not a mere means) then we might
include animals in it, but we would need a whole new explanation of what justifies
declaring anything, human or animal, an end in itself. The fact that we both have
a natural good that matters to us isn't enough to show that this forbids our use as
means. More work would need to be done there and it would entail a wide

divergence from Kant (probably ending somewhere near Regan's "subject-of-a-

108

life" criterion). Kant has an internally coherent justification for according humans
this status, but it is one that cannot be extended to animals. In the interest of
showing that Kant can accommodate a more robust theory of duties to animals
while doing as little violence to the rest of his ethical theory as possible, I must
reject Korsgaard's claim that animals are ends in themselves.

My argument will also make use of the Kingdom of Ends; the idea of the
"subject," as explicated in the next section, is essentially identical with
Korsgaard's concept of the "passive citizen," although I think my terminology
more accurately reflects the subordinate status of animals. Like Korsgaard, I will
argue that the assignment of value by rational beings is what gives rise to our
obligation to take nonrational beings into account in our moral legislation. My
starting place will not be our shared animality, however. I find two weaknesses in
that approach. First, as Korsgaard herself acknowledges (27h), when Kant
discusses our duties to ourselves with respect to our animal nature, all the
examples ultimately demonstrate that abusing our animal nature is bad because
in doing so we also abuse our rational nature. Thus Kant is showing that it is our
rational nature, not our animal nature, that exists as an end in itself. Korsgaard

is willing to diverge from Kant on this point. I would be, too, but her argument

109

has not provided a compelling reason to do so. Second, Korsgaard has not
successfully shown that in Iegislating for the protection of our animal needs, that
it is me animality we are valuing. Instead, one could argue that what is being
valued is the autonomy of rational beings, first by preserving our lives (a
necessary precondition for preserving our autonomy), and second by protecting
our right to pursue private ends related to our animal nature, as long as they do
not interfere with others' right to do the same. In Section 5, I will try a different
approach - one that is also based on the creation of value by rational beings, but
focuses on the value of happiness rather than animal nature. Sections 3 and 4
will argue, first, for the addition of the category of "subjects" to the Kingdom of

Ends, and second, for the claim that animals would fall into that category.

3. Members, Subjects, and Things in the Kingdom of Ends

In rejecting the claim that animals are ends in themselves, it may seem
that I have eliminated any possibility of anything more than indirect duties with
respect to animals. The end in itself is the center of Kant's ethics and the source

of all value. How can there be any room left for animals? Since I believe that the

110

answer lies in Kant's idea of the Kingdom of Ends, I will first explicate this idea
before showing where I think animals should fit into it.

Kant formulates the Categorical Imperative in at least three distinct ways.
The first is the familiar criterion of universalizability, known as the formula of
universal law: "I ought never to act except in such a way that I could also will that
my maxim should become universal law" (4:402). Kant describes this as "mere
conformity to law as such," as it involves the defining characteristic of laws — their
universality - abstracted from any particular content (4:402). The second main
formulation rests on the concept of humanity as an end in itself: "So act that you
use humanity [. . .] always at the same time as an end, never merely as a means"
(4:429). As discussed in the previous section, it is the power of moral legislation
that justifies rational beings' status as ends in themselves.

The third main formulation makes use of the idea of a Kingdom of Ends,
which Kant describes as "a whole of all ends in systematic connection (a whole
both of rational beings as ends in themselves and of the ends of his own that
each may set himself)" (4:433). In other words, it is the idea of a world in which
all the ends of rational beings are respected and none are in conflict. In such a

world, all rational beings are both legislators and subjects of the law. We should

111

judge our own actions by whether they would accord with this harmonious
kingdom of ends. Kant formulates this as "act in accordance with the maxims of
a member giving universal laws for a merely possible kingdom of ends" (4:439).

Although it is not entirely clear that the three formulations are equivalent
as Kant claims, they are certainly interrelated. The formula of universalizability
uses the form of the law as a measuring stick and the formula of humanity uses
the matter of the law. The Kingdom of Ends synthesizes the two into one
concept in which form and matter unite, resulting in a world of universal laws that
respect and protect the ends of all rational beings.

It is not hard to see why the Kingdom of Ends has appealed to many
recent commentators. Its incorporation of the previous two formulations makes it
the richest conception of the categorical imperative that Kant offers. Although
Kant assures us that all the formulations express a single categorical imperative,
the formula of the Kingdom of Ends makes it easier to see how the formula of
universal law and the formula of humanity work together.

In accordance with Kant's remarks in the lectures. the Groundwork implies
that animals have no place in the Kingdom of Ends apart from their usefulness to

rational beings. Beings without reason "have only a relative worth, as means,

112

and are therefore called things" (4:428). Animals, then, must be things. (This
would hold even if Kant were to acknowledge my claim that some animals have
degrees of rationality, because the rational capacity that concerns Kant in these
passages is the ability to act according to principle, a capacity which even the
most intelligent animals lack.) Kant sharply divides the Kingdom of Ends into
persons and things; the former have dignity, the latter merely a price. Because
animals cannot participate in the Kingdom as legislators, they are not members
and their ends are not taken into account in the systematic unity. In dividing the
world into persons and things, Kant places animals into the latter category. I
believe that dividing the moral universe in this way is a mistake, because it does
not recognize an important distinction between animals and other nonrational
"things."

Kant identifies two roles for persons in the Kingdom of Ends. Every
rational being is either a member or a sovereign. Members both legislate, and
are subject to universal laws; this is the category that we as human beings fall
into. A sovereign, on the other hand, would be a rational being who legislates
without being subjected to anyone else's will; this is possible "only in case he is a

completely independent being, without needs and with unlimited resources

113

adequate to his will" (4:434). God would presumably be such a being; elsewhere
Kant explains that the concept of duty does not apply to a holy will, a claim that
he also makes for the sovereign of the Kingdom of Ends (4:434). Everything that
is not a member or a sovereign is merely a thing, to be used as a means for
rational beings to achieve their ends.

Kant's political analogy lacks an obvious group that can exist in a natural
kingdom - that of the subject, who is governed (and, ideally, protected) by laws
but cannot legislate them. Kant may have seen this role as being filled by things,

but I hold that subjects are distinct from things. Subjects are beings whose

interests9 are taken account of in the laws that members and sovereigns enact,
even though they do not participate in Iegislating those laws; typically this would
include children, aliens, and so on. Laws can be enforced on their behalf. This
is the legal equivalent of moral patienthood. Things, on the other hand, have no
interests of their own (although we may have an interest in them) so it would not
make sense to legislate on their behalf. In accordance with the political

metaphor that Kant uses in discussing the Kingdom of Ends, I will use the term

 

9 See section 4 for a discussion of what it means to have interests.

114

"subject" to describe those who would have the status of moral patients in the
Kingdom of Ends.

Kant does not acknowledge the existence of subjects in the Kingdom of
Ends. He does, however, make reference to the idea of a subject when
discussing why human beings must be subject only to laws that we give
ourselves. Human beings are autonomous, and thus in the Kingdom of Ends
cannot be mere subjects. If we were subject to the law without being its
originators, then the law must compel us by some external means, that is,
heteronomously (4:433). Thus the law must be one that we give ourselves.
Being both subject to and legislator of the law is what gives us the status of
member in the Kingdom of Ends.

Animals cannot be moral legislators because they are incapable of acting
according to self-originated principles; thus they cannot be members in the
Kingdom of Ends. They also are not subject to the law. We no longer put
animals on trial or execute them for crimes; we may kill a dog that has injured
peOple, but we do so for reasons of safety, not justice. We consider it
appropriate to constrain an animal's behavior according to relevant laws, but it is

the animal's guardian that is truly subject to the law. A dog does not understand

115

that it is not permitted to foul the sidewalk, but we still have laws against that
behavior and we expect the dog's guardian to compel it to act in a certain way.
This can be carried over by analogy into the ideal Kingdom of Ends. Since
animals are not internally motivated to follow the law by reason, they may be
externally compelled in certain ways by the members of the Kingdom. The law
would be addressed to the moral agents responsible for them, not to the animals
themselves. In a way animals are subject to the law, in that we can legislate
certain restraints on their behavior. But this is not really the same sense in which
a moral agent is subject to the law, because the agent can understand the law
and act accordingly. An animal can be compelled physically or psychologically to
act a certain way, but will be responding to something unrelated to the law. The
agent, on the other hand, can be motivated by the law itself. Animals are a
subject at the law, rather than being subject la the law.

Animals, I argue, should be regarded as subjects in the Kingdom of Ends,
and we must legislate on behalf of them and in their best interests. In order to
make this argument, I will need to establish two main claims. First, that animals
have interests in a way thatthings do not; and second, that those interests

should be considered by the legislators of the Kingdom of Ends.

116

4. Animals Have Interests

A being with interests, as I use the term, is one that has ends that matter

to it.‘° I argued in Chapter 2 that animals have ends, although those ends are
not freely chosen like ours, and are dictated by their nature rather than given by
reason. Having ends is a quality that they share with human beings (and by
extension other rational beings), and that they do not share with anything else.
This is because having ends implies at least a power of choice, if not a will.
Things like tables and cars have ends only metaphorically or by extrapolation of
the human ends involved with them. Plants are a harder case, because they
would seem to have ends implanted by nature the way animals do, that compel
them to grow toward a light source and other basic movements. But Kant does
not attribute a power of choice to plants, and I see no reason to disagree. The
power of choice is a mental faculty, and it is implausible to attribute mind to

plants as we do to animals.

 

1° Kant defines "interest" as "the dependence of a contingently determinable will on principles of
reason" (4:413n). There are two kinds of interest, practical interest in the action itself, and
pathological interest in the object of the action. The latter refers to the use of reason to obtain an
object of inclination — in other words, the prudential use of reason. I have previously argued that
some animals may employ prudential reason, so such animals could have "interests" in Kant's
sense; most, however, would not. However, I am diverging from Kant's technical use of "Interest"
in order to use the term in the more familiar sense of "acting in someone's interests." As I am
using the term, any creature with ends that matter to it will have interests.

117

It is possible that one could argue that plants possess a low-level
calculating function that governs their heliotropic behavior, and which is different
only in (albeit large) degree from the basic weighing of inclinations that is
involved in the animal power of choice. Or, one could argue that it is possible to
build a machine that mimics the animal power of choice in terms of working
toward externally-imposed goals (which we would then perhaps be justified in
calling "ends"). If machines and plants also have ends, then the apparently
special status of animals and the importance of ends is diminished.

Even if I did acknowledge such similarities between animals, plants, and
machines, animals' ends would still be of a different kind, because their ends
matter to them - that is, it matters to them whether those ends are achieved or
frustrated. An animal is, as Korsgaard puts it, "an organic system to whom its
own good matters" (30). Things can go well or badly for it in a way that things
cannot go well or badly for a table or a car, or even a tree.

Of course, there is a sense in which things are "good for" a table or car. It
is good for a wooden table to be oiled periodically, and bad for it to be sat upon
by a heavy person; it is good for a car to have proper oil pressure, and bad for it

to be exposed to a lot of salt. In a biological sense, various things are good or

118

bad for trees. But these things do not capture what I mean for something to "go
well or badly" for a being.

Kant's argument for indirect duties to animals itself rests on an intuition
that animals' ends matter to them. Why is it that he identifies mistreatment of
animals as something that damages our humanity, but not mistreatment of
carriages or tables? (Likewise for plants; although he does argue that it is a duty
to avoid wanton destruction of natural beauty, his reasons for this are separate
from his reasons for avoiding cruelty to animals, and he regards the latter as "far
more intimately opposed" to our duty [62443].) Kicking the wheels of a carriage
or taking a hatchet to a table might be foolish, but it does not damage our
humanity as would kicking a horse's legs or taking a hatchet to a dog. Why not?
It cannot be just that the objects do not resemble human beings enough to
produce the amphiboly, because on the surface an animatronic pirate at
Disneyland resembles a human being a lot more than a dog does, yet no one
would suggest we have duties with respect to animatronics.

The answer lies in Kant's own remark about why we have duties with
respect to animals: because cruelty "dulls his shared feeling of their suffering"

(6:443). Tables and animatronics do not suffer; that is, they do not emerience

119

suffering. There is no feeling there for us to share. But when one kicks a dog,
the dog suffers; kicking it is not just bad for it, it is bad for it in a way that matters
to it. Thus, for a being's ends to matter to it, that being must be conscious. Our
intuition that animals are conscious is implicitly acknowledged by Kant in his
theory of indirect duties. Furthermore, I have argued, in Chapter 2, that Kant's
theory of animal minds implies that he believed animals to be conscious. Thus,
animals have ends that matter to them, a characteristic they share with human
beings but that sets them apart from things; I term this "having interests."
Showing that animals have interests at least gives a possible criterion that
allows them to be separated from mere things in the Kingdom of Ends. However, '
it is not sufficient to establish that these interests must be taken into account by

the members of the Kingdom.

5. Animals and Happiness

In the second Crin‘gue, Kant defines "happiness" as "the state of a rational

being in the world in the whole of whose existence W

hia wish and wil " (5:124). By limiting this to rational beings, he seems to exclude

the possibility that animals can be happy at all.

120

Why would rationality be required for happiness? It might simply be due
to context, since he is discussing rational beings in the passage. Or, it could be
that somehow happiness is dependent on the possession of a will (Willa). It is
true that animals do not have a will (Wflla) as this is limited to fully rational
beings; however, animals do have a power of choice (Milt). Insofar as an
animal is able to obtain the objects of its power of choice (which might be
described as having things "go according to its wishes"), this would seem to fit
Kant's definition of happiness. Happiness for a human being would be different
from happiness for an animal, since the former would involve the rational will and
the latter a merely animal power of choice, but both could fall under the general
term "happiness."

Perhaps it is not the fulfillment of every individual desire that constitutes
happiness, but the satisfaction of knowing that all one's desires have been
fulfilled; this would require a higher-order ability of reflection than most, if any,
animals possess. It would seem to require reason, although not necessarily the
moral component of reason. As I argued in Chapter 2, it is moral reasoning that
animals conspicuously lack, whereas some other aspects of reason are

apparently present. So it is possible that some of the most intelligent animals

121

have the ability to conceptualize that their desires have been satisfied — or not,
as in the case of Brosnan and de Waal's slighted monkeys. Still, the requirement
that a creature be able to reflect on its own satisfaction would exclude almost all,
if not all, animals from being capable of happiness.

The more general definitions of happiness that Kant gives in the
QLQLIDSMQIK - "that complete well-being and satisfaction with one's condition" -
(4:393), and "the entire satisfaction of [needs and inclinations]" (4:405) — do not
make reference to rationality as a requirement for happiness, although they do
not rule out that claim either. If Kant does incorporate a requirement for rational
reflection into his concept of happiness at times, he does not do so consistently:
when discussing the use of practical reason to further one's happiness, Kant
states that a person's reason "L . .] does not at all raise him in worth above mere
animality if reason is to serve him only for the sake of what instinct accomplishes
for animals [. . .]" (5:51-62). Given the context, it is plausible to take "what
instinct accomplishes in animals" to be the attainment of happiness.

The solution, I believe, is in the following passage:

[. . .] as far as our nature as sensible beings is concerned, all that counts

is our hapainasa if this is appraised, as reason especially requires, not in

terms of transitory feeling but of the influence this contingency has on our
whole existence and our satisfaction with it [. . .] (5:61)

122

Two important points can be gleaned from this passage. First, that Kant
acknowledged a broader definition of happiness under which it is possible to
evaluate it in terms of transitory feeling (which animals certainly possess).
Second, that rational beings always evaluate their own happiness by seeing how
the satisfaction of particular desires fits into their well being as a whole. I take
this to mean mat for beings that are capable of reflecting on the fact that their
needs have been met, this reflection is part of their happiness, but that this ability
to reflect is not required for beings that do not reflect in this way. This would
explain why it is more difficult for humans to be happy than for animals; our
happiness has an additional requirement that animals' does not.

According to Kant, happiness is not in itself good, but only in combination
with a good will, which "seems to constitute the indispensable condition even of
worthiness to be happy" (4:393). Kant thus defines virtue as worthiness to be
happy, and states that the highest good of a possible world is for happiness to be
distributed in proportion to virtue (52110). This is similar to the idea of the
Kingdom of Ends, with the addition that we must postulate God in order to ensure
that happiness actually is distributed in proportion to virtue, which does not

necessarily happen in nature.

123

Since happiness in proportion to virtue is the highest good, we must make
the happiness of others our end. This would apparently not include animals,
because animals, as moral agents, lack virtue. They cannot, therefore, be
worthy to be happy, as Kant has described it. On the other hand, they are not
deserving of unhappiness, either. Their happiness is simply irrelevant, exwpt
insofar as our indirect duties require us to preserve it for our own ends.

Kant argues that happiness is not good in itself because, without the
influence of a good will, it can give rise to boldness and arrogance (4:393), and
because we dislike seeing immoral people enjoying happiness, and only approve
of it when it is had by virtuous people (4:393). Thus only when paired with the
one thing that is good in itself - a good will, which animals cannot have - does it
have value.

Kant's argument suffers from a flaw: if happiness cannot be judged on its
own merits, apart from the presence of a good will, then why is it that we take
pleasure in a virtuous person being miserable, and displeasure in an evil person
being happy? We base this evaluation on the belief that a good person deserves
happiness, and a bad person does not. How do we know what a good person

deserves? The concept of desert involves some sort of equivalence between the

124

character of the agent and the quality of the thing that is deserved. Good people
deserve good things. So in order to declare that good people deserve
happiness, we must already have judged happiness to be a good (not
necessarily the highest good, but a good nonetheless). There must be some
sense in which happiness is a good independent of the presence of a good will.

If happiness is a good, then why is it that in some situations, such as
those Kant mentions, we don't judge it to be good? Doesn't the fact that we
dislike seeing evil people happy suggest that it achieves its goodness only when
combined with virtue? I admit that there is a relationship between virtue and
worthiness to be happy, but it is complex. If our intuitions reveal that there is
something wrong with virtuous people being unhappy or evil people being happy,
they also reveal further wrinkles. A small child happy at play is a sight that meets
with almost universal approval; conversely, nothing is more upsetting than the
sight of a child in distress. Yet a child, prior to the age of reason, is incapable of
morality. Our judgment that children should be happy is not based on the child's
moral character.

Whenever children are raised as an example in moral philosophy, the first

inclination is to treat them as de facto moral agents because of their potential to

125

become moral agents. One could argue that children have the potential to
become virtuous, and thus are worthy to be happy. Potentiality, however, does

not solve the current problem. Children have the same potential to become evil

as to become good.11 This would seem to imply either that the two cancel either
other out, and children should be neither happy nor unhappy, or else that both
are entailed, and the child should be made both happy and unhappy (perhaps
alternately). We have no justification to assume that a child is going to become
good (nor, of course, that it is going to become evil).

What is it, then, that is behind our intuitions about the happiness of
children? Why is the happiness of a child so cherished, and the misery of a child
so disdained? The answer, I believe, lies in the concept of innocence. As pre-
moral beings, young children can do no evil. It is the lack of evil that we regard
as the prerequisite for happiness, not the presence of moral goodness. In a

rational adult, the lack of evil entails the presence of moral goodness; in a

nonrational or pre-rational being, neither is present.

 

1‘ Fred Rauscher points out that, given Kant's belief in the radical innate evil of humankind as
expressed in the Raligian, we have reason to expect that all children are destined to become evill
They may also manage to fight against the innate evil and become good, but that is not
guaranteed. Extrapolating from the Rallgjan, it is merely possible that a given child will become
good, but cartaih that he or she will become evil.

126

This also applies to our evaluation of happiness in animals. We take
pleasure in animals' happiness, just as we have a shared sense of their suffering.
The sight of happy animals at play does not produce as strong a reaction as the
sight of happy children (no doubt for natural reasons related to the propagation of
the species), but we do approve of it. We have a sense that animals are
innocent, that they do nothing (morally) wrong, and although we may resent their
presence at times, when we are thoughtful and objective we do not blame them
when their actions have a negative impact on us. ‘

I believe that it is not virtue that makes one worthy of happiness, but the
lack of vice. Rather than the presence of virtue making happiness good, it is the
presence of evil that ruins it, robbing it of its prima facie goodness. But what
grounds this goodness? As moral agents, we necessarily confer value upon it.
Kant states that happiness is the "L . .] aha end that can be presupposed as
actual in the case of all rational beings [. . .] and therefore one purpose that they
not merely malt: have but that we can safely presuppose that they all actually gla
have by a natural necessity" (42415; of. 5225). As finite, rational beings - is.
human beings - we necessarily regard happiness as valuable. Although it is our

own happiness that nature instructs us to cherish, as rational beings and

127

members of the Kingdom of Ends, our adoption of happiness as an and confers
value upon it and thus we legislate for its general promotion. Korsgaard points
out that Kant is not a value realist. That is, he does not believe there are
intrinsically valuable things or qualities, which we then discover and arrange our
morality around. Instead, as moral legislators we are the source of value
(Korsgaard 18).

Since happiness is a natural and of all finite, rational beings, it has a value
that we confer upon it. We see it as a prima facie good, and thus we seek to
promote it wherever it appears, except when it is tainted by evil. It is not, for the
reasons Kant shows, absolutely good, and it is not the source of value (rather,
rationality is the source of value, but we place this value on happiness).

But when we take happiness as a value, why do we make an exception for
immoral persons? All human beings have their own happiness as an end. As
rational beings, we are supposed to evaluate all our actions as though we were
legislators in the Kingdom of Ends, in which everyone's ends are in harmony.
Thus, our own actions must promote, and not interfere with, the happiness of
others. When rational beings interfere with others' ends, they are trying to make

an exception of themselves. They will for their own ends to be met while others'

128

are frustrated, but no consistent law can be made from this in the Kingdom of
Ends. Thus in the real world, when we see such people being happy, we
recognize the inconsistency between their willing unhappiness for others but
receiving happiness for themselves, and this offends our rationality. In the case
of children and animals (as well as those with a good will), happiness entails no
inconsistency in will, and thus the value we confer upon it is not forfeited.

As rational but finite beings, we confer value upon happiness. Although
the origin of this value is our own desire for happiness, as rational beings we
bestow a value upon it that goes beyond its humble origins. The value of
happiness is not absolute, since the presence of an evil will can override our
approval of it. Its value, however, is not entirely dependent on being linked with a
good will. This argument diverges from Kant, but not as severely as it may seem
on first impression. It leaves intact Kant's definition of virtue as "worthiness to be
happy," if it is understood that this definition is limited to rational beings. Since
rational beings are the only ones that can participate in the moral realm, this
limitation is not a serious hindrance for the definition. Worthiness to be happy, in
my theory, entails not being evil; in moral beings, this is the same as being good.

Hence within the scope of rational beings, worthiness to be happy would still

129

correspond with morality. Children, animals, and other nonrational beings, on the
other hand, would be naturally worthy of happiness due to their innocence.
Further, this argument does not contradict Kant's claim that the only thing good in
itself is a good will. I have argued that happiness can be evaluated as a prima
facie good independent of whether it is attached to a good will, but its value still
comes ultimately from its endorsement by the human will, and so its good is still
derivative rather than inherent.
Kant's own remarks about the good will fit well with my claim about the
prima facie value of happiness:
Understanding, wit, judgment and the like [. . .] are undoubtedly good and
desirable for many purposes, but they can also be extremely evil and
harmful if the will which is to make use of these gifts of nature [. . .] is not
good. It is the same with gjflaafloflgha. Power, riches, honor, even
health and that complete well-being and satisfaction with one's condition
called happflasa, produce boldness and thereby often arrogance as well

unless a good will is present which corrects the influence of these on the
mind [. . .] (42393)

Note that Kant claims that it is the presence of an evil will that renders some
talents of mind and qualities of temperament harmful, and then remarks that the
same is true of happiness. Furthermore, he says that the reason a good will
must be present in conjunction with happiness is that it will correct its influence

on the mind. In an animal, no evil will can be present, and thus no correction is

130

required. Kant's remarks in this passage support my claim that it is the absence
of an evil will, rather than the presence of a good will, that is the important factor
in judging worthiness to be happy.

It may be objected that in claiming that being good and not being evil are
the same for a rational being, I have obscured the range of goodness that can
exist in human beings. Not being evil seems to require only the fulfillment of
perfect duties, which does make a person good to some extent. There is quite a
range, however, in the fulfillment of imperfect duties; a person who gives a small
amount of money every month to charity is following an imperfect duty, but not to
the extent that a person who puts in several hours a week of volunteer charity
work, and so on up to a person who devotes his or her life to serving humanity.
That Kant recognized such a range is implicit in the idea of happiness in
DLQQQEILOJJ to virtue.

This is a valid point, but it can be incorporated without abandoning the
claim that good, for a rational being, is entailed by not being evil. First, although
imperfect duties are wide in their application and thus we are allowed to exercise
judgment in when and how we fulfill them, making it one's principle never to fulfill

them at all would be a conflict of will and therefore immoral. Thus acknowledging

131

and discharging imperfect duties is a requirement for not being, to some extent,
evil. The question remains how to account for the difference between the merely
good and the saintly. This can be accomplished by regarding the absence of evil
as a minimum requirement for us to approve of happiness at all, while allowing
that greater degrees of goodness are worthy of even greater degrees of
happiness. I will return to this in the next section.

Human beings can, and in fact must, have ends other than their own
happiness. Animals, as nonmoral, are always concerned ultimately with
satisfying their own needs, and thus with happiness. Taking account of animals'
interests in the Kingdom of Ends would mean providing for their happiness.
Since we have acknowledged happiness as valuable provided it is not
accompanied by a bad will, we should legislate on behalf of animals in the
Kingdom of'Ends. Their status would be that of subjects, protected by the laws
but not participating in their creation. This expanded view of the Kingdom of
Ends allows us to see where Kant should have placed moral patients in his

ethical theory.

132

6. The Moral Status of Animals as Subjects

Animals, as discussed previously, cannot be regarded as ends in
themselves because they are not a source of ends. Whereas Korsgaard has
termed them "passive citizens" of the Kingdom of Ends, I prefer the term
"subjects." "Citizen" implies an equality that they do not have with ends in
themselves; even with the modifier "passive," the implication is that they are
different from true citizens in degree, not kind. On the contrary, the fact that they
are not ends in themselves does put them in a subordinate moral position.

Rational beings are the creators of the moral universe. Without rational
beings to impose a moral order, everything that happens - in the words of a
cynical determinist acquaintance - is just "the universe doing stuff." Animals
would exist, and would still be happy at times and suffer intensely at others, but
this would not be right or wrong; it would simply be. This is difficult for us to
acknowledge. The temptation is to insist that pain and suffering are evil, and
happiness good, regardless of our presence. But that is only because we are
looking at it as rational beings. We cannot truly imagine ourselves out of the
picture, because our mind's eye is still there to judge what we envision. Thus we

assign value to these things through time, before and after our existence, and

133

even across possible worlds. Nevertheless, that value is dependent on us.
Animals' status in the moral realm is not just different or inferior to ours, it is in
fact dependent on ours.

Korsgaard holds that although animals cannot assign moral value, they
still have a natural good that matters to them; thus it would not be true to say that
without rational beings, nothing in the universe would matter (36). Things would
still matter to animals. Unhappiness (such as pain) would still be bad from their
perspective, and happiness good. I have also argued that animals have ends
that matter to them, and thus I agree that things would still matter to animals if
rational beings did not exist. There would, however, be no moral value attached
to their likes and dislikes. It is only rational beings who can say that animals
amulgl be happy and ahoalaaat suffer. The moral status of animals is
dependent on our existence as rational beings.

Since our own moral status is the precondition for animals to have any
status, the preservation of rational beings must take precedence over animals'
interests. This would include not just the preservation of the lives of rational
beings, but also the preservation of rationality in those beings. In a situation

where harming an animal is required to save the life of a rational being, there

134

would be conflicting grounds of obligation”, one pointing toward protecting the
animal's interests as a subject of the Kingdom of Ends, the other based on the
rational being's status as an end in itself. Since rationality is a prerequisite for an
animal's status as a subject, it is a stronger ground of obligation and thus in this
conflict its preservation must be upheld as a duty at the expense of the animal's
interests.

This account of animals' status has an advantage over Korsgaard's
because, in holding animals to be ends in themselves, she seriously complicates
the issue of conflicting obligations. It is true that conflicting grounds of obligation
exist even without considering animals, but the problem takes on a decidedly
expanded scOpe under her interpretation. Her response is to take the tragic view
that it is impossible for us to uphold all our duties because we do not live in the
Kingdom of Ends; we try to shape the actual world into the ideal Kingdom as
much as we can, but we cannot expect nature to behave in accordance with our
ends. We must simply do our best to treat all creatures as ends in themselves,

even if we must fail (34-35).

 

'2 For Kant's discussion on conflicting duties, see 62224-225, 272508, and 27:537. Briefly, he
holds that there are no conflicting duties. only conflicting grounds of obligation. The weaker
ground gives way before it reaches the status of a duty, and thus there is only one actual duty.

135

How acceptable one finds this argument in a Kantian framework will
depend on how one interprets Kant's claim that ought implies can (alluded to at
5:30) and whether one believes it is dispensable. Even so, the trouble with
Korsgaard's position is that it gives no guidance in how to choose among
conflicting obligations. "Do your best" is not a useful command without some
way to determine what the best thing is. Recognizing the subordinate position of
animals - based not on mere fiat, but on their dependent moral status - will help
arbitrate at least some of the moral difficulties that we will encounter upon
admitting them as subjects in our Kingdom.

The dependent status of animals does not mean that their interests will
always be overridden by our own ends in the case of a conflict. The moral status
of animals is dependent on our existence as rational beings, but not, for example,
on our personal comfort or desires. There will be times when various personal
ends will conflict with animals' happiness, just as our various personal ends can
conflict with other rational beings'. We will at times have to forgo our desires in
order to care for our subjects. This is not a problem, but simply the nature of our
obligations as members of the Kingdom of Ends. It is, after all, also true of our

relationship with other rational beings. We should pursue our private ends in

136

such a way that they can harmonize with the ends of the subjects of the Kingdom

of Ends, as well as with the other members.

7. Objections and Replies

There are still some objections that one might raise with regard to the view
that animals have the status of subjects in the Kingdom of Ends. In this section I
will address three lingering concerns: first, that my scheme poses a problem for
the allotment of happiness in proportion to virtue; second, that my theory offers
few real protections for animals; and third, that ultimately l have offered another
indirect-duty theory, with all the attendant difficulties.

If animals are worthy of happiness insofar as they are not unworthy of
happiness, as l have proposed, then in the ideal moral universe, animals and
virtuous human beings would be equally happy. Something about this does not
seem to fit with the ideal of happiness in proportion to virtue. We would want to
say that the greater attention a person pays to his or her duties, the worthier that
person is of happiness. Animals do not forfeit the value of their happiness
through evil action, but they also are not capable of the amazing heights of moral

goodness to which human beings can strive, through devoting themselves to the

137

fulfillment of imperfect as well as perfect duties. It bothers our sense of fairness
to think that a dog should be as happy as a hardworking philanthropist. Has the
latter not earned a greater degree of happiness than the human being who at
least avoids evil (if not strives to fulfill imperfect duties to quite the same degree),
and does that moral-but-not-saintly person not in turn deserve more credit than a
nonmoral animal? If we strive to keep animals happy, will they get more than
their fair share?

First, if this is a problem, it is one for Kant's stated view as well as for my
theory. After all, Kant believed that we should try to provide for animals'
happiness as much as possible; although his rationale was that sensitivity to
animals contributes to our own moral abilities, the result would still be that
animals would be made happy as much as possible.

Second, it is plausible that animals, due to their lesser intellectual
faculties, are naturally incapable of appreciating happiness to the extent that
human beings do. Admittedly, since their desires tend to be simple, it may be
easier for them to achieve happiness, but surely the kind of simple bodily
satisfaction that comprises animals' happiness cannot be compared with the

higher happiness felt by a fully rational being. The latter not only can appreciate

138

having his or her ends met, but also can reflect on the fact that those ends have
been met. It is still possible, then, that the proportion of happiness allotted to the
philanthropist will be greater than that allotted to the dog, even if the dog is as
happy as it can be.

The other two objections deal with how well I met my own goals in
undertaking this project. My first goal was to show that the reason Kant gives for
treating animals well is inadequate, and to suggest that a better reason can be
offered within Kant's framework. It might be objected that what I have proposed
is simply an indirect duty view under another guise, and thus that it is subject to
all the same problems that I criticized Kant's view for in Chapter 1. Since
animals' moral status is dependent on the value that rational beings place on
happiness, it might seem that our duties to animals are really only covert duties
to ourselves, to respect our own values." I objected to Kant's indirect duty view
for several reasons, one of which is that animals' moral status is entirely
dependent on human psychology; thus, if it turned out that we could learn to
separate our sympathy for animals from our sympathy for rational beings, there

would be nothing wrong with torturing animals. My own view would seem to be

139

equally contingent and thus equally problematic. We must provide for animals'
happiness because we happen to value happiness in ourselves; thus if we did
not value happiness, we could harm animals as we wished.

Although it is true that my view maintains Kant's claim that only rational
beings are ends in themselves, and thus the moral status of other beings is
dependent on ours, it is not contingent in the same way Kant's is. Kant claims
that happiness is "necessarily the demand of every rational but finite being"
(5:25). Although it is contingent in the sense that what constitutes happiness
varies from person to person (and presumably would vary even more if we knew
of any other rational but finite beings), it is subjectively necessary in the sense
that as finite beings we have needs and these needs are necessarily
incorporated into our desires (5:25). Furthermore, happiness is such an
important value for us as finite rational beings, that its presence (in proportion to
virtue) is necessary for the attainment of the supreme good. Linking animals'
status to the value we place on happiness does not put them in the precarious

position that linking their status to the promotion of sympathy does. My position

 

‘3 Note that this criticism would apply equally to Korsgaard, since she makes animals' status as
ends in themselves dependent on the value that rational beings place on their animal nature.

140

does not violate the intuitive claim identified by James Skidmore and which he

dubs "CF" ("counterfactual"): "Even if certain (e.g., cruel or brutal) treatment of

 

animals had no effect on an agent's attitudes or behavior towards human beings,
such treatment would stlll be morally objectionable"(8-9). My own position does
not make the wrongness of cruel treatment of animals dependent on any effect
that the treatment would have on the agent or on other human beings.

Is the view I am proposing still an indirect duty view? That is an
interesting question. Since animals are not ends in themselves, it would seem
that their ends, insofar as we are obliged to take them into account, are really
ends l9r_lls. The duties that we have to promote their ends are not, however,
merely covert duties to human beings. We must promote animals' happiness
because happiness is valuable; and it is the happiness olthatahlulal that we are
obliged to promote, not some indirect effect on the happiness of human beings.
Therefore my view is not an indirect-duty view as Kant's is. Animals, however,
do have a status that is contingent on our own, and so our duties to them are not
duties to them in quite the same way that our duties to human beings are. This is

because it is not the animals that bind us in a relation of obligation; the source of

141

the obligation is ourselves as rational beings. Our duties to other rational beings,
on the other hand, are duties that originate in them as well as in ourselves.

My second hope was that whatever replacement I found for Kant's theory
would entail a larger set of duties to animals than the one that he offers. The
problem is that in declaring that animals have a subordinate status as subjects, I
may have left them in nearly the same position that Kant did: the object of some
weak and easily-overridden duties that amount to little more than "use animals as
you wish, but avoid being cruel when possible." This is not, however, an
accurate characterization of my position. It is true that I regard the preservation
of rational beings as necessarily taking priority over the interests of animals. As
previously noted, however, this does not mean that all human interests take
priority; far from it. Only in the narrow case of preserving a rational being's life
(or preserving his or her rationality) do rational beings' interests automatically

take precedence. This would seemingly allow animal experimentation under

certain circumstances“, but would not permit eating animals (at least for

 

M Korsgaard rightly points out that what. exactly, counts as "saving a life" is a matter of some
ambiguity, especially in the realm of research (36n). Allowing animals to be sacrificed when
doing so would save a life might allow less research than one might initially assume.

142

members of first-world nations at this point in history),15 hunting, and other uses
of animals for mere gratification.

An important consideration is that as we strive to create the Kingdom of
Ends, we are attempting to set up a world in which everyone's ends are in
harmony. This entails not just specifically carrying out or refraining from certain
narrow actions, but trying to bring about various institutions and practices that will
bring us closer to the possibility of such a kingdom. Thus although we are
permitted to harm animals for the preservation of rational life, we have an
obligation to try to prevent such necessities from arising in the future. For
instance, we must not be content to use animals for scientific experiments on the
grounds that doing so preserves rational life; we must also be working as hard as
we can to discover ways to reach the same ends in the future without animal

experimentation.

 

15 It may be that killing animals does not in itself violate their interests. If true, this would
presumably make it permissible for us to kill them provided that it is done painlessly (this would
be true not just in the theory that I am proposing, but in many theories that involve obligations to
animals). The usual argument for this is that animals do not have plans for the future, and so we
are not frustrating their plans by ending their lives. We could raise them for food and thus use
them as means so long as we also made their happiness our end while they were alive. I am not
entirely convinced that survival is not an interest for animals, but even if the point is conceded, we
would have to radically reevaluate how we currently raise food animals, probably to the point at
which it would no longer be economically or practically viable. The constraints on our behavior
toward animals would still be much stronger than we currently acknowledge.

143

Those are the main objections to my theory as I see them, and I believe
my responses are adequate to meet the challenges. Some questions about my

theory remain unexplored, and I will discuss these in the next section.

8. Future Directions and Conclusion

My intention has been to carve out a space for animals in Kant's moral
theory by discarding his unsatisfactory account of indirect duties and replacing it
with a theory that provides for a more robust conception of animals' place in the
moral sphere. Although I have focused on animals, I believe that my conclusions
will also help account for permanently nonrational human beings, such as
severely mentally retarded persons. In striving toward the Kingdom of Ends, we
should take into account all beings with ends that matter to them. Although it
may be equally consistent with the rest of his moral theory to simply deny that
there is anything wrong with treating animals as cruelly as we wish, I believe that
this violates Kant's attempt to create a moral system that accords with common
intuition. In proposing the doctrine of indirect duties to animals, Kant seems to
be trying to get them in through the back door; it is as though he wants to

acknowledge the common intuition that we have duties to them, but has been

144

trapped by his belief (which I do not dispute) that moral agency is the source of
all moral worth. His mistake, I believe, is in not recognizing that we can locate
moral worth in things other than ourselves. Its source need not be the same as
its location.

My goal has been only to show that there is room for a better account of
duties to animals in Kant's framework, rather than to show what exactly those
duties would be. As it stands, I have offered only a small sketch of what would
be entailed by taking animals' interests into account. Various projects for future
study are immediately apparent, including how to resolve conflicting grounds of
obligation between animals and human beings in cases other than the narrow
one that l have described (saving a human life or preserving human rationality),
how to resolve conflicting grounds of duty that involve two animals, and to what
extent we are obliged to interfere or refrain from interfering in the affairs of wild
animals. Another important question is how to decide what creatures count as
"animals" under this theory. I have used the term loosely to refer not to the
whole biological category of animals, but only to those animals that count as
beings with ends that matter to them; which beings those are is a controversial

matter deserving of further consideration.

145

I will conclude by noting that the incorporation of animals into the Kingdom
of Ends has one further benefit for Kant's theory: it allows for a satisfying parallel
to Scripture. Kant was fond of alluding to Biblical passages, and endeavored to
bring his moral theory into line with a religious worldview, so I shall offer an
allusion of my own; I think Kant would have appreciated it.

In the Garden of Eden, Adam was given the task of ruling over the
animals, but there is no indication that he used them; at that time, all creatures
apparently ate plants (Naw Amarign Stamiartl, Gen. 1:26-30). Human beings
did not need to use animals, because all creatures' needs were provided for. But
the world we live in now is not the Garden of Eden; animals kill and eat each
other, and sometimes to survive we too have to act against their interests.
Someday, the Bible promises, we will once again live in harmony in God's
kingdom. When that day comes, all the animals will once again be in harmony
with each other and with us:

And the wolf will dwell with the lamb,

And the leopard will lie down with the young goat,

And the calf and the young lion and the fatling together;

And a little boy will lead them. (Is. 11 :6)

This is impossible in the natural world, of course. But our role is to strive to bring

about a moral world, a Kingdom of Ends, like the kingdom of God on earth. We

146

cannot make the wolf dwell with the lamb, but we can and should lead them all
toward the greatest fulfillment of ends that we can achieve within the cruel world

of nature.

147

WORKS CITED

Ameriks, Karl. Kantjs Thaary at Mim: An Analysis af tha Paralmisms af Para
Elaasoh. Oxford: Oxford UP, 1999.

Broadie, Alexander and Elizabeth M. Pybus. "Kant and Direct Duties." [Magus
20 (1981): 60-67.

---. "Kant's Treatment of Animals." Pit—ilasamll 49 (1974): 375-383.

Brosnan, S. F. and F. 8. M. de Waal. "Monkeys Reject Unequal Pay." Nature
425 (2003): 297-99.

Carruthers, Peter. Tm Animals lama. Cambridge: Cambridge UP, 1992.
Egonsson, Dan. "Kant's Vegetarianism." rn I f l l i 31.4 (1997):
473-83.

Hoff, Christina. "Kant's lnvidious Humanism." Environmental Ethias 5 (1983):
63-70.

Kant, Immanuel. W. Trans. and ed. Arnulf Zweig. Cambridge:
Cambridge UP, 1999.

_ri_Lque of Practical Reason. Practical Philosaphy. Trans. and ed. Mary J.
Gregor. Cambridge: Cambridge UP, 1996.

Critique of Pure Raasan. Trans. and ed. Paul Guyer and Allen
Wood. Cambridge: Cambridge UP, 1997.

148

---. ri‘ f P wer f m n. Trans. Eric Matthews. Ed. Paul Guyer.
Cambridge: Cambridge UP, 2000.

---. r f h i of l . Pragtlaal Philosoahy. Trans. and
ed. Mary J. Gregor. Cambridge: Cambridge UP, 1996.

---. Legglasahgthlca. Trans. Peter Health. Ed. Peter Heath and J. B.
Schneewind. Cambridge: Cambridge UP, 1997.

---. W. Trans. and ed. Karl Ameriks and Steve Naragon.
Cambridge: Cambridge UP, 1997.

Mataghysias at Marals. Pragtlgl Philasqphy. Trans. and ed. Mary J.
Gregor. Cambridge: Cambridge UP, 1996.

---. W. Trans. Eckard FOrster and Michael Rosen. Ed. Eckard
FOrster. Cambridge: Cambridge UP, 1993.

Korsgaard, Christine. "Fellow Creatures: Kantian Ethics and Our Duties to
Animals." Author's personal web site. 2004. Harvard University. 12 April
2004 < http://www.people.fas.harvard.edul~korsgaarl
CMK.FeIIowCreatures.pdf> Forthcoming in W
Human Valaas 25/26. Ed. Grethe 8. Peterson. Salt Lake City: Utah UP.

Leibniz, Gottfried Wilhelm. W. Trans. George
Montgomery. Th R ' Ii . New York: Doubleday, 1960.

Monaglalogy. Trans. George Montgomery. Tha Ratlanalisls. New York:
Doubleday, 1960.

Naragon, Steven. "Kant on Descartes and the Brutes." Kantstudien 81.1 (1990):
1-23.

149

--. "Reason and Animals: Descartes, Kant, and Mead on the Place of Humans
in Nature." Diss. University of Notre Dame, 1987.

Tha Naw Amarign Stanglarq Bibla. Anaheim, Calif.: Foundation Publications,
1997.

Neiman, Susan. Tha Unity at Raasqn: Rereading Kant. New York: Oxford UP,
1994.

O'Neill, Onora. W. Cambridge: Cambridge UP, 1990.

---. "Kant on Duties Regarding Nonrational Nature II." Pr in

91 me Arislglalian Sfiiefl Stlpglamant 72 (1998): 21 1-228.

Pepperberg, Irene M. "Studies to Determine the Intelligence of African Grey

Parrots." Pr i f h In r ' n lAvi l ri oi n
W. International Aviculturists Society. 11 April 2004

< http://www.mecca.orgl~rporter/PARROTS/grey_al.html>

Regan, Tom. "Broadie and Pybus on Kant." Philqmphy 51 (1976): 471-72.

Sapontzis, S. F. "Are Animals Moral Beings?" W
17.1 (1980): 45-52.

Skidmore, James. "The lncoherence of Indirect Duties Regarding Animals."
2001 Central Division Meeting. The American Philosophical Association.
Hyatt Regency, Minneapolis. 2-5 May 2001.

Velkley, Richard L. Fraflam anq tha Eng af Raasan. Chicago: University of
Chicago Press, 1989.

Weir, A. A. S., J. Chappell, and A. Kacelnik. "Shaping of Hooks in New
Caledonian Crows." Salarlca 297 (2002): 981.

150

 

Wood, Allen. "Kant on Duties Regarding Nonrational Nature l." Prqaaedings
at tha Aristotelian Saciety Supplemant 72 (1998): 189-210.

151

illiiiilljiijiijli[iii