PHI 385 UCL Cohen on Korsgaard and Moral Obligation Philosophy Question

1: rewrite and reword the introduction and delete the 400-500 words to go straight to the korsggard’s theory on moral obligation

2: the presentation of korsgaard’s view need to be complete , need to explain what problems she is trying to solve. explain why she thinks that reflective endorsement together with the idea of practical identity successfully answers the normative problem and establishes reasons.

3,explain why, according to her, we have an obligation to be moral

4. the criticism of Encho is not relevant to the question (delete)

5. Cohen has a few criticisms of korsgaard’s view , suggest that you simplify things and focus on the mafioso

6. asking yourself why does the mafioso show that is a problem with her view? the key to the answer, according to Cohen, is that the mafioso is an example of a person who has his reasons grounded exactly in the way in which korsggard rcommends and yet reaches an immoral conclusion

Cohen on Korsgaard and Moral Obligation Philosophy
PHI385 Sources of Normativity
Registration No. 170207618
Word Count: 3640
Introduction
Using the Korsgaard model, it is possible to analyze the potential influence of her
arguments in evaluating human behavior. On the same note, using the ideas
presented through these models, it is possible to create a high-level assessment of
the aspects that define the rationality of human decisions in different situations and
incidents. This project will explore the arguments presented in the Mafioso case
while critically responding to Cohen’s evaluation of Korsgaard’s theory. This analysis
will show the differences in perceptions and arguments raised by the two experts to
respond to normativity, constitutivism, and morality. While Korsgaard argues in
support of agents’ rationality, Cohen claims that this interjection fails to address
issues such as the shmagency and bad action problems (Pauer-Studer, 2018).
Korsgaard’s failure to respond to critiques by Cohen and individuals like Enoch shows
that her arguments lack the ideal connection with the ideas presented in her theory
(Ricciardi, 2019).
Korsgaard’s Argument
The first idea that this section will consider is the definition of the concepts of selfconception and normative reasons. According to Korsgaard, human beings are
different. Environment and education status make humans have varied views as
their mental capability is different. They influence how people view the world and
lead to the difference in perceptions and opinions. However, the primary distinctions
between the two are expressed in the variances in the thinking and that they are
controlled by their mental activities. Korsgaard argues that the distinct difference is
that while literate and civilized human beings are governed by normative ideals and
moral principles, others with lesser mental capacity are governed by emotion,
desire, and instinct. On the same note, Korsgaard argues that a problem emerges
from the tendency of humans to turn to the interpretations of their mental activities.
Korsgaard highlights the influence of reason in shaping the respective populations’
decisions about their situations. In the analysis of the influence of reason, Korsgaard
asserts that human beings can use their internal skills to determine the purpose or
benefits of acting on a given idea. In acting on a given issue, Korsgaard states that
human beings must have the intended desire. However, humans cannot find a
motive to act or respond to a given issue of concern without desire. Therefore,
desire is the foundation and drive for most human actions.
According to Korsgaard, human attitudes and ethical practices exhibit shortcomings
when compared to how humans treat other humans. Humans kill others for more
reasons than one. Whereas reasonable human beings could be provoked by others,
and they reason out together and solve their issue amicably, they can’t reason with
other individuals with less mental capacity. That might resort to punishing them,
including killing them. This view explores the difference in mental capacities and
directly relates this to the consequences of human being’s actions. The distinct
difference in this scenario is that human beings with a given literacy level can reason
together, while educated human beings and illiterate ones cannot. For instance,
there are unnecessary wars in less civilized nations as people lack the capability to
reason with each other. Each decision is made irrationally without weighing in on the
consequences. To Korsgaard, this difference explains why civilized humans can be
rational as they can evaluate the merits of their personal desires and decide against
them or go ahead with them, but others cannot because they lack the same ability.
In essence, Korsgaard suggests that human beings with lesser mental capacity lack
the capability of reasoning or do not have control when it comes to their desires.
Korsgaard says that civilized human beings can attach a reason to everything they do
not matter how right or wrong it is. In the case of killing others, human beings give
various reasons. One, they can say they kill them because they avoid danger or
protect others from death. Another reason is that humans can make peace
nationally and globally, for instance, the elimination of terrorists. Sometimes, the
reasons can be as trivial as other individuals destroying humans’ property, but the
bottom line is that humans can always attach a reason to the decisions they make or
the actions they take in the course of life. The stark difference is that learned human
beings are governed by normative ideals and moral principles while radical and
brainwashed act on instinct, desire, and emotion. The normative ideal can be wrong
morally, but it will still be used as a reason for actions. When humans kill others
morally, they may feel it is wrong. However, the reason for the killing action is to
base their actions on the normative ideals that govern them, whether or not they are
right notwithstanding.
Humans have a conviction that even though they use others for their desires and
purpose, they should be humane. Korsgaard argues that humans believe that while
punishing and killing others should continue, the killing should not be torturous but
humane. While the experimentation using other animals should not stop, humans
should ensure they minimize the pain they inflict on them to the lowest. Kant says
that this kind of treatment reveals human nature. Humans exhibit a moral concern
for others and treat them to show that human beings have a common nature of
consciousness and are sensate. Korsgaard says that this nature of human beings is
what makes them treat prisoners of war as enemies but, while at it, treat them as
people whose humanity they share. Korsgaard acknowledges, like Kant, that others
may be used as a usable resource as it is a fact during slavery or as mere means, but
in the process, the fact that human beings are sensate and have feelings influence
their treatment of others in a way that exhibits moral concern and humane action.
The emotions and desires of human beings are inherently controlled by their desire
to be moral, by their ethical code of conduct, and what they believe to be the
difference between right and wrong.
There is a moral asymmetry between how humans treat fellow humans. One theory
says that humans have a moral asymmetry. The second theory is based on humans’
distinctive relationship to doing what is right. Proponents of the first theory believe
in human superiority over others, and therefore, they have autonomy over what
happens. The superiority means that humans can do whatever they deem fit with
other animals. Human beings were gifted with the gift of intelligence, which has
been found lacking in most animals. This influences their actions; however, one
crucial factor to note is that the levels are different, and one human being can be
more intelligent and wiser than another. However, animals do not have the same
luxury because they are on the inferior side of the spectrum and can only be victims
of humans’ decisions. Proponents of the second theory say that humans return
goodness for goodness, obligated to reciprocate the actions of those they interact
with. The relationship is dependent on one’s treatment of the other. Humans
reciprocate good deeds with good deeds while they return the same when it is about
bad deeds. The relationship is purely symbiotic, where what one sows is what they
reap in the long run. Korsgaard argues that since animals are not moral, they cannot
have an obligation towards humans, and therefore, humans’ treatment of them is
less stringent and are not a direct debt to them. In this case, Korsgaard suggests that
human beings do not have a moral obligation to animals. This reasoning can be used
to explain human actions and activities. A human may feed his sibling today out of
his goodness or for the benefits he receives from it, but the sibling won’t provide the
parent tomorrow out of its own volition because he may not be moral. Therefore,
the relationship can only be one-sided, with the superhumans having the upper
hand.
From the study of the arguments raised above, it is possible to determine the steps
used to define the ideas that Korsgaard proposes. The first step in understanding her
claim is to evaluate the influence of a personal reflection of desires on actions
(Korsgaard & O’Neill, 1996). This claims that humans can affect their decisions to
either act or not based on reason and the influence of their mental states. The
reflective structure of the human mind helps humans evaluate the merits of their
desires. It equips them with the ability to assess the consequences of their needs
and wants. Humans have better faculties of reason, and their reflection hinges on
what reason designates as good or bad. It lies on the difference between what is
considered right and wrong. Reflection helps humans distinguish between good and
bad based on the moral code and, therefore, can go against desires that determine
tip the immorality scale. It helps human being identify their shortcomings and other
related factors that may influence them to become biased or prejudiced. After this is
done, a person is usually better positioned to make good decisions; in essence,
reflection acts as an eye-opener and helps human beings evaluate their desires and
decisions.
In the second step, the main focus is on the power of reflective distance away from
personal desires, which forces humans to decide against some actions because
reflection doesn’t find them fit according to reason. Korsgaard claims that reflective
distance is necessary for a human being’s functioning. The distance helps to reflect
and decide if any of the personal desires held are worth doing. In essence, they
decide whether some of their desires are worth actualizing; this requires an analysis
of the consequences of these desires.
Furthermore, Korsgaard proposes that each person’s normative reasons are defined
by the manifestation of their beliefs about themselves. She claims that selfconception is a product of the expression of each individual’s personality because
humans form various opinions about themselves, which most times influence how
they behave in particular situations. A person’s view of themselves influences how
they see the world; it impacts their relationship with themselves and, consequently,
other people as well. For instance, a person who perceives himself as a good person
might act wrongly while thinking they are doing good. In this case, a person’s
conviction of themselves influences their actions. These ideas have formed the
foundation for analyzing each person’s practical identities and the connection
between normative reasoning and self-conception. One of Korsgaard’s key concepts
is that both reason and desire influence the actions performed by any human being.
Reason exists along with the desire to act against or towards a given issue of
concern. However, this is significantly influenced by conviction and self-conception.
While a person might interpret a situation as wrong, another person might view it as
not wrong and essentially correct.
One of the most common arguments that critiqued Korsgaard’s ideas comes from
Enoch and Cohen. Both experts have different interpretations of the ideas presented
in The Sources of Normativity (Korsgaard & O’Neill, 1996). In addition, Cohen (1996)
uses numerous examples, including the case of Mafioso and others, to illustrate that
Korsgaard’s ideas fail to provide the ideal platform for exploring the issues posed
through the normativity lens (Bagnoli, 2009). Understanding the role of these
critiques can help determine the definitive conclusions against Korsgaard and
normativity.
The defense of Korsgaard’s ideas comes from an understanding of the nature of the
agency problem. The primary justification is based on a response to Cohen’s bad
action problem where humans choose wrongdoing over right action because of
personal desires, particularly about the gaps within the constitutive theory.
According to Enoch, Korsgaard’s central ideas about the theory of constructivism are
not according to the ideal framework for exploring the connection between actions
and each human being’s reason and desire. Her theory has been critiqued for failing
to acknowledge the complexity of human nature and behavior, only considering the
influence of reason and desire to act instead of the multidisciplinary interactions
between diverse concepts. Critics claim that various factors are at play when
establishing a connection between reason and desire and should be considered.
Human nature is extremely complex and cannot be addressed across one factor; but
rather, all factors need to be considered. Furthermore, Cohen has critically analyzed
Korsgaard’s ideas, arguing against them because of the gaps that he claims she failed
to address. Instead of responding to Enoch’s critique, she gives a similar explanation
to Cohen’s bad action problem. In Cohen’s arguments, the theory above has failed to
specify the subjects’ ultimate reason to endorse either one thing or the other. Cohen
presented the associated concepts using moral theory in place of the normativity
concepts adopted by Korsgaard.
Cohen argues that Korsgaard’s impressions and ideas do not focus on defining the
agents’ independence in reasoning and responding to a given issue. Moreover, he
argues that Korsgaard fails to consider the differences that agents exhibit in their
constitution: subject to changes in their behavior and connection with the target
communities. He argues that Korsgaard’s argument should be based on concrete
factors rather than trying to make a connection between two factors. While the two
critiques form a foundation for exploring the challenges and gaps that Korsgaard fails
to resolve, she argues that her answer to Cohen’s wrong action problem is sufficient
to define the ultimate solution to Enoch’s shmagency challenge. By exploring the
shmagency problem, defining the ultimate solution that addresses Cohen’s criticism
through his bad action problem is possible. This argument can be used as a strategy
to respond to the weaknesses associated with the underlying issues and gaps in the
concepts of normativity and human nature constitution that Korsgaard fails to
address. By uncovering these concepts, the potential solution to understand human
nature could be revealed.
Korsgaard uses the concepts provided by the Kantian constitutive on the factors
influencing normativity exhibited by humans. In the argument, the primary source of
normativity is self-constitution (Lueck, 2009). She argues that normativity is the
product of reflective endorsement, which is associated with rational reflection.
Reflection endorsement can be defined as the ultimate decision-making process that
humans undertake to determine the best actions to implement when responding to
a given issue. According to Korsgaard, a reflective endorsement is a form of
commitment based on personal identity and can be used to define personal
authority when responding to an individual’s nature. Korsgaard added that being an
agent implies being able to solve problems through a given aim and goal. Each agent
intends to achieve a particular goal by solving diverse problems and issues.
According to this argument, the main conclusion is that reflective endorsement is
the fundamental determination of the laws that define the respective populations’
behavior when making decisions and acknowledging themselves as agents.
Reflection forms a huge part of the decision-making process and influences the
choices one makes. For instance, after reflecting on one’s desires and consequences,
an individual can make a good decision or rather moral decision.
Korsgaard proceeds to argue that the concept of self-constitution is the ultimate
activity that offers the elements associated with reflective reason. Acting for a cause
or performing duties based on what ought to be done is defined as constituting
oneself. According to the argument, the reason is defined as a set of fundamental
beliefs primarily endorsed by the person who believes or acts. The activity of
reflective endorsement has been linked to the development of the states of affairs in
reasoning. This statement implies that reflective endorsement activity helps people
make decisions according to the analysis of the options provided. Through reasoning,
it is possible to make decisions or act based on the perceived benefits and
drawbacks of the underlying issues. Korsgaard uses the self-reflection concepts as
the foundation for the derivation of normativity. The rationale for being
foundational is that there is no other form of reasoning that can be used to define
the concepts of normativity. As a result, using normative reasoning to establish a
connection between desires and one’s action is advantageous.
One of the conclusions that can be made about the concepts of normativity
discussed by Korsgaard is that she borrows ideas from the Kantian theory and
approaches; using these has created a reliable approach to defining the foundational
concepts associated with the elements of normativity and human nature. The
concepts of reason and desire, influenced by reflective endorsement, can be
explained through the Kantian philosophy model.
By understanding the ideas of the Kantian theory, Gerald Cohen (1996) critically
analyses Korsgaard’s model to determine the gaps that have not been fully explained
to reflect the accurate representation of normativity. According to Cohen (1996), the
ideas in Korsgaard’s model can be evaluated and summarized into eight primary
concepts. These concepts are essential since they influence the decisions made by
analyzing the gaps and weaknesses left out by Korsgaard. One of the main
summaries offered by Cohen is that since Korsgaard claims that humans are
reflective beings, they are driven to act by reason.
This idea has been supported by the concepts of human reflective endorsement
based on the foundations of action. However, if human beings do not have a
normative conception of their identity, it is impossible to have a reason for
performing any given action. In this analysis, Cohen (1996) claims that human beings
must have a normative understanding of their respective identities to become
rational agents in performing various actions. Therefore, human beings are supposed
to hold themselves valuable because of the nature associated with the rationality of
their actions. Moreover, Cohen concludes that Korsgaard argues that human beings
must be regarded as valuable. Therefore, this valuation leads to establishing a moral
obligation formed within the human agency nature.
In the first three steps, Cohen relies heavily on establishing morality, which is
achieved by appealing to the elements of the human constitution. Through this
summary, Cohen demonstrates that Korsgaard cannot derive conclusions from the
foundations offered above. Furthermore, Korsgaard’s theory only attempts to
analyze or explain the concept of morality and the foundations of normativity.
Instead of exploring the idea raised, the theory attempts to define the concepts of
normativity.
Cohen uses the Mafioso example to explain the problems associated with the
weaknesses of Korsgaard’s arguments. In his claim, Cohen relates moral laws and
human beings’ behavior governed by a particular code of conduct and honor. In this
example, Cohen relies on the potential weaknesses of living under an honor code
instead of the prevailing moral laws and regulations governing the community.
Through this weakness, the main conclusion offered in the process is that Korsgaard
fails to acknowledge the differences that her argument claims, especially when
considering the concepts of reflective endorsement.
Using the moral law and the honor code, it is not possible to convince Mafioso to
adopt the former instead of the latter. According to Cohen (1996), the inability to
offer a conclusive foundation for connecting reflective endorsement to the agents’
rationality as far as human beings are concerned makes it hard for Korsgaard to
provide valid arguments. The bad action problem’s concept argues against the
foundations of Korsgaard’s conclusions about normativity. Cohen asserts that
Korsgaard fails to show the distinction between reason and desire when performing
specific actions.
Furthermore, Cohen (1996) argues that reason alone is insufficient to provide the
ultimate connection between desire and performing an action. The relationship
between reason and desire in endorsing a given action has been argued from a
different perspective by Korsgaard. Cohen claims that Korsgaard does not explicitly
connect the desire to perform an action and normativity. Also, he claims that reason
solely is not sufficient to influence an individual’s behavior into performing a defined
action. Additionally, through the exploration of Korsgaard’s arguments, Cohen adds
that it is impossible to define that a reason or desire positively influences a given
person’s behavior.
Conclusion
One of the main issues raised by Cohen and Korsgaard is the analysis of the concepts
of morality and the definition of human behavior. Cohen has a different notion
regarding the principles offered by Korsgaard. While Korsgaard claims that human
nature’s normativity is based on the fundamental concepts associated with reflective
endorsement, Cohen argues against this conclusion by providing a bad action
problem. By analyzing the primary ideas based on Korsgaard’s principles, a person is
influenced by reason and a desire to perform a given action. The motivation to
perform a given action is dictated by a person’s will and provides the individual with
the ultimate reason for doing so. He tries to scrutinize and ascertain human action
by pointing the drive and motivation. It is evident that any conflict or war has a root
cause, and both parties have an outcome they desire, which is the motivation behind
their action. However, Cohen argues that it is not right for Korsgaard to associate
reason and desire with morality. He doubts that she could prove the principles as
indicated by the impossibility to associate reason and morality in the Mafioso case as
an example.
References
Bagnoli, C. (2009). The Mafioso case: autonomy and self-respect. Ethical Theory and
Moral Practice, 12(5), 477–493.
Bukoski, M. (2018). Korsgaard’s arguments for the value of humanity. Philosophical
Review, 127(2), 197–224.
Cohen, G. (1996). Reason, humanity, and the moral law. In C. Korsgaard & O. O’Neill
(Authors), The Sources of normativity (pp. 167-188). Cambridge: Cambridge
University Press. http://doi:10.1017/CBO9780511554476.00
Korsgaard, C.M, & O’Neill, O. (1996). The sources of normativity. Cambridge
University Press.
LeBar, M. (2001). Korsgaard, Wittgenstein, and the mafioso. The Southern Journal of
Philosophy, 39(2), 261–271.
Lueck, B. (2009). Kant’s fact of reason as a source of normativity. Inquiry, 52(6), 596–
608.
Pauer-Studer, H. (2018). Korsgaard’s constitutive and the possibility of bad action.
Ethical Theory and Moral Practice, 21(1), 37–56.
Ricciardi, B. C. (2019). Hermann Cohen on compassion: A cognitivist take on the
peculiarly religious emotion. Journal of Jewish Ethics, 5(2), 207–227.
LECTURE 6
Morality and identity
Raymond Geuss
In her third lecture Professor Korsgaard distinguishes between
what she calls ‘the categorical imperative’ and ‘the moral law’
(3.2.4). The categorical imperative imposes a minimal condition on
free choice: such choice must be guided by a principle we have
given ourselves which has the form of a general rule or law (3.2.3).
A free will must choose a maxim it can regard as a law. What
Korsgaard calls ‘the moral law’, then, is a further specification of
what this law which I give myself must be: it must be the kind of
thing ‘all rational beings could agree to act on together in a workable cooperative system’ (3.2.4). As Professor Korsgaard quite
rightly points out, Kant doesn’t make this distinction and she suggests that awareness of the distinction will allow us to see a certain
incompleteness in Kant’s argument. He wants to show that any
free will is bound by the moral law but he, in fact, establishes only
the weaker claim that we are bound by the categorical imperative.
Korsgaard’s argument, then, has two parts. First, she defends
the bits of Kant where he got it right: normativity arises from the
structure of the free will and the free will must stand under the categorical imperative. Second, she completes Kant’s argument,
showing that autonomous human agents stand not only under the
categorical imperative but also under the moral law.
I’ll start with the first part, the part that is supposed to parallel
Kant’s own discussion. As humans the form our freedom takes is
that we are not forced to act on the desires we happen to find
present in ourselves. We have the capacity to take a step back from
them and decide whether or not we will endorse them as worthy
grounds for action. Furthermore, we have no choice but to see ourselves and our whole lives under the aspect of a potentially
189
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
190
RAYMOND GEUSS
continuous exercise of this capacity. We have no choice but to see
ourselves as constantly endorsing, or failing to endorse, the various
desires we encounter in ourselves. Note too that this rather odd
way of speaking as if my desires were simply something I
encounter in myself isn’t just an unsympathetic way of putting it
for the purposes of defaming the position, but rather it is something Kant’s defenders emphasize. As Professor Korsgaard puts it:
‘Anything outside of the will counts as an alien cause, including the
desires and inclinations of the person’ in question (3.2.3). There is
of course a long tradition of criticizing Kant’s ethics on this
account – namely that as an ethical agent I see my own desires and
inclinations as alien entities from which I must keep my distance starting with Friedrich Schiller’s well-known essay ‘Uber Anmut
und Wtirde’ (1793) and extending up to Professor Williams’s paper
here in Cambridge, in 1981.’
Standing back then from these alien entities that present themselves to me, my desires, I reflect on whether to endorse one or
another of them as reasons for action. Since all of my desires are
alien intruders I can’t use any of them to decide which of my other
desires to endorse. That is, it isn’t open to me to say: I have a desire
to listen to music, and now I’ll endorse that desire because I like listening to music and I like the desire to listen to music. If I were to
try this line I would have failed to reflect sufficiently, that is, I would
not have sufficiently abstracted myself from my desires. But if I
can’t use myfirst-or second- or any higher-order desires, and don’t
want to admit that it could be an exercise of my freedom just to pick
one existing desire for endorsement without following some antecedent principle of choice, then how do I come to endorse one of
my desires as a reason for action? Kant’s answer is that I must have
a principle of choice which isn’t derived from desire and which I
give myself. But if such a principle of choice is not dependent on
any desire, it can itself be nothing more than the principle: choose
what you can will as a universal law. That, however, is, for Kant,
just the formula of the categorical imperative (and for Kant,
although not, it seems for Korsgaard, the formula of the moral
law). For Korsgaard the situation is a bit more complicated because
1
Published under the title ‘Prasuppositionen der Moralitat’, in Bedingungen der Moglichkeit,
pp. 251-261.
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
Morality and identity
191
she seems to give two slightly different answers to the question. I
assume that she thinks that these two answers amount finally to the
same thing.
The first of Korsgaard’s two replies is very much like Kant’s:
The principle of endorsement is that I endorse only such desires as
are compatible with a maxim which I can will as a universal law.
Actually Korsgaard gives this line of argument a particular twist,
which allows her to connect it with the second of her two replies.
For Korsgaard I don’t just use a criterion of formal law-likeness as a
principle of endorsement of desires as reasons for action, I ‘identify’
with it. She writes: ‘The reflective structure of human consciousness requires that you identify yourself with some law or principle
which will govern your choices’ (3.3.3) and of course this law will be
the one prescribed in the categorical imperative. The question is
whether ‘using’ a principle is quite the same as ‘identifying’ with it.
Presumably Korsgaard holds that this is just a harmless linguistic
variation of usage. After all, if in reflection I see all my desires as
alien, then what else is there for me to identify with except the principle of endorsement (or rejection) I use?
I think it is striking, though, that Kant himself doesn’t talk about
‘identity’ in ethical contexts, and notoriously Kant thinks that
‘rational psychology’, the metaphysical discipline purportedly
studying the underlying bearer of personal identity, is a pseudoscience. Kant comes closest, I think, to discussing ‘identity’ in the
sense in which Professor Korsgaard uses the term in his discussion
of the ‘interests of reason’. In his Kritik der reinen Vernunft he claims
that the interest of reason is exhausted when one has given answers
to the three questions: ‘What can I know? What ought I to do?
What may I hope?’2 As Heidegger pointed out,3 in the introduction
to one of his unpublished lectures on logic Kant adds to these three
the fourth question: ‘What is the human being?’ The question:
‘Who am I?’ doesn’t appear, as if it were obvious that the correct
answer is: A human being’; that is, as if the questions ‘Who am I?’
and ‘What am I?’ were philosophically not properly distinct.4
2
4
Kant, Kritik, pp. 804-805. 3 Heidegger (Frankfurt/m, 1929), p. 187.
The failure to distinguish clearly between ‘Who am I?’ and ‘What am I?’, and thus to
address the former question at all, is one of Heidegger’s main criticisms of Kant, in his
early works Kant und das Problem der Metaphysik and Sein und £eit. Tom Baldwin has pointed
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
192
RAYMOND GEUSS
I make such heavy weather of this notion of ‘identity’ because
Korsgaard’s project of recentring Kantian ethics around notions of
identity seems to me to push toward a position in which it will be
difficult for the Kantian to reply to the line of argument developed
by Friedrich Schlegel in the 1790s.
In Korsgaard’s reconstruction, if I use my second-order desires as
my principle for endorsing or failing to endorse a given first-order
desire (as a reason for action) I have broken off reflection prematurely. I should continue until I reach the purely formal principle
embodied in the moral law. In a similar way Schlegel claims that
Kant broke off reflection prematurely: he stopped when he reached
the point at which he saw that we have the capacity to prescribe to
ourselves universal laws. But if what is at issue is my identity Kant
should have realized that no universal law or mere formal principle
can actually give me my identity. What T am will always go beyond
what can be given in any set of purely general laws, and to identify
myself with any such law or set of laws or with the mere capacity to
give myself such laws is to misunderstand and limit myself. What I
am is something that essentially cannot be identical with a law or the
capacity to give laws. If anything I am rather to be identified with a
specifically human capacity that is higher and more complex than
the mere capacity to prescribe universals laws, namely the capacity
to give myself a freely chosen formal law and then consciously
decide to violate it by making an exception of myself. Since I am not
and cannot be identical with any general law or principle, my proper
attitude toward any general law (even one I give myself) will be one of
keeping it at a distance from me, i.e. at best treating it ironically, and
precisely not identifying with it. What I should identify with is with
my continuing ability to distance myself in thought and action from
any general law. This is connected with an attempt on Schlegel’s part
to give a positive valuation of irony, frivolity, spontaneity, wilfulness
out to me that the place in his ethical writings where Kant comes closest to discussing
something like my ‘identity’ is in his discussion of my ‘Gesinmmg’ (and of such related concepts as m y ‘ Personlichkeif) in Die Religion innerhalb der Grenzen da Uofien Vernunfl (Konigsberg
1793, ‘Erstes Stuck’). The Religionschrift was a favourite among the Romantics, but Kant’s
notion of ‘Gesinnung’ seems to me still rather different from modern concepts of identity.
For Kant it seems as if I could have only one of two possible ‘Gesinnungen’: ‘good’ (if duty
is a sufficient motive for me to act without the need of any further motive), or ‘bad’ (if I
pervert the moral order and allow my inclinations to take precedence over the moral law).
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
Morality and identity
193
(and also laziness).5 My essential identity is a process of giving myself
laws and consciously deciding to treat them ironically, act frivolously
or wilfully, or consciously violate or change my self-given laws.
Most people find this position unappealing. Hegel with
uncharacteristic lack of charity cites Schlegel’s views as the main
instance of a category he calls ‘Evil’.6 The issue is not, however,
whether or not we think Schlegel is wrong, but rather whether a
Kantian position has the conceptual means to give us a reason to
reject Schlegel. Hegel thinks he has grounds to reject Schlegel’s
position. The ideal Schlegelian life, after all, is a ‘constant succession of self-creation and self-destruction’7 and thus not a life
devoted to the cultivation of continuing habits of socially responsible action. As Hegel points out, to keep frivolously making and
unmaking laws or to treat given laws ‘ironically’ would be no way
to run a state, but perhaps that just shows the difference between
principles we might use in giving stability and decency to our social
life and issues of our identity. The more the Kantian focuses on
issues of identity the harder a time he or she is going to have in
dealing with Schlegel.8
That then is the first of Korsgaard’s two replies to the question
how, in reflection, I can come to endorse some desires as reasons
for action and reject others: I give myself a general rule, identify
with this rule, and use it as a standard. Korsgaard’s second reply to
this question is the claim that ‘the reflective structure of the mind’
‘forces us to have a conception of ourselves’ (3.3.1) and this conception functions as our standard. Such a conception gives me, she
writes, ‘A description under which you value yourself, a description
under which you find your life to be worth living and your actions
to be worth undertaking’ (3.3.1). On this second view my identity
(in this sense) gives rise to normativity both positively in that this
identity allows me to turn desires into reasons and negatively in
that ‘obligations spring from what [my] identity forbids’ (3.3.1).
This reconstruing of the Kantian project in terms of values
seems to me to weaken the appeal of the position considerably. In
5
7
8
Cf. Schlegel’s Lucinde (1799). 6 Hegel (Berlin, 1821), section 140.
Schlegel (1798), Fragment 51.
Although Hegel thought Schlegel’s position dangerously wrong, he also thought it the
correct dialectical successor to Kantian ethics.
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
194
RAYMOND GEUSS
her lecture 3 Korsgaard considers the hair-raising claim that you
might be able to obligate me to value what you value. It is a significant strength of the Kantian view that I stand under no obligation
whatever to value what you empirically value – and a good thing
that is, too, given that you might value rock music, sports cars, the
ballet, geraniums, small porcelain figurines, or various other inherently worthless things. For Kant I am at best under certain obligations not to interfere in various ways with you in your misguided
pursuit of worthless rubbish, not to value it myself.
Korsgaard seems to shift back and forth uneasily between a very
strong sense of ‘obligation1 and an exceptionally weak one.
Sometimes ‘obligations’ are unconditional demands I should die
rather than violate, something the violation of which is a threat to
my very identity. There might be such things, but what I normally
call my ‘obligations’ aren’t like this at all. I have, and acknowledge,
an obligation to pay my legitimate debts and perhaps this is even an
‘unconditional’ obligation but one would have to have a pathologically fragile sense of self to feel one’s identity threatened if one
defaulted on a few debts. I would not even think of risking my life
in order to be able to repay a small bank loan.
Failing to discharge minor debts doesn’t threaten my identity,
but I also don’t think defaulting a desirable thing. I would prefer
not to do it and will go to some lengths (although not to any length)
to avoid it. But one could easily imagine cases in which doing something which was a violation of some of my obligations &/form part
of my identity, i.e. followed from my conception of what made my
life worth living and my actions worth undertaking (3.3.1).
To use a very anachronistic example, Filipo Argenti is in the
Inferno rather than in the Purgatorio not because he very frequently
committed the sin of anger, but because anger formed a part of his
identity. Part of what made him the person he was and wanted to
be, of what made life worth living for him, was the angry ‘violation
of his obligations’.9
Korsgaard would presumably reply to this that although as a
9
This example is anachronistic in at least two ways. First of all because I put it in terms of
‘identity’ rather than state of the soul, will, or form of ‘love’. Second, a character trait
such as ‘anger’ is something quite different from the violation of an ‘obligation’. I hope
the point can be seen through the smudge I have made of this.
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
Morality and identity
195
matter of fact Argenti may have had a practical identity which
gave undue prominence and inappropriately positive valuation to
being angry, he ought not to have had this identity, and thus perhaps
rightly ended up stewing in the dismal swamp he inhabits down
there. He ought not to have identified himself as an essentially irate
person but as a potential member of the party of universal humanity. So the issue then becomes in what sense it can be claimed that
anyone ‘ought’ to acquire this particular identity (as a member of a
Kingdom of Ends). I will come back to this later.
There is then in Korsgaard’s lectures a very strong sense of
‘obligation’ which connects it with my very identity and with
unconditional demands I should die rather than violate.
Sometimes, however, the notion of an ‘obligation’ is so weak and
thin that you can put me under an obligation just by calling my
name. ‘By calling out your name I have obligated you. I have given
you a reason to stop’ (4.2.7). I think it isn’t just accidental that ‘I
have obligated you’ and ‘I have given you a reason’ seem to be used
interchangeably here.
I’m also somewhat unclear about the concept of ‘reflection’
Korsgaard uses. In lecture 3 she envisages two possible outcomes
to a process of reflection:10 Either I come to be able reflectively to
endorse my initial set of desires (or a modified one that arises in
the process of reflection), then I have ‘reason’ to act on them, or I
reject them in reflection as incompatible with my identity. Then I
have an obligation not to act on them (3.3.1). But it seems to me
that earlier in her lectures Korsgaard countenanced a third possible outcome of reflection. She spoke there (for instance in the discussion of Hume) of the possibility that reflection might
undermine the claims of certain considerations to be reasons for
action. Reflection on morality might have a sceptical outcome,
undermining the claims of morality. This would presumably not
mean that we had an obligation to refrain from being moral, just
that we saw we didn’t have the kinds of reasons we once thought
we had. I don’t really know what to do with this observation
about the concept of ‘reflection’ but I do think it important to dis10
One possible outcome of reflection is that I am unable to decide whether to endorse
certain desires or not, but let’s leave that aside for the moment and consider just the case
in which reflection does lead to a determinate decision.
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
196
RAYMOND GEUSS
tinguish carefully between reflective undermining of reasons and
reflective generation of an obligation.
I said that the first part of Korsgaard’s argument was a defence
of the validity of the categorical imperative. I’ve nowfinishedmy
discussion of that and would like to turn to the second part, the
argument for the moral law in the version Korsgaard prefers. The
conclusion is that my identity is really and essentially my relation
to humanity as a whole or as an equal self-legislating member of
the Kingdom of Ends, and thus reflection consists in endorsing
those desires that are compatible with this identity and rejecting
those incompatible with it.
I must confess that I don’t understand the argument Korsgaard
gives for this claim. Sometimes it seems as if for Korsgaard we
just are able to see a real intrinsically moral fact: ‘. . . you are a
human being and so if you believe my argument you can now see
that that is your identity . . . And that is not merely a contingent
conception of your identity, which you have constructed or chosen
for yourself, or could conceivably reject. It is simply the truth’
(3.4.9). So it is just a non-contingent fact about me that my essential identity is that of citizen of the Kingdom of Ends and that
this identity trumps all my contingent identities (as a resident of a
certain city, native speaker of a certain language, member of a
certain association, etc.). If I just see that this is true of me, presumably we have a form of realism. Surprisingly, realism emerges
at the end of the lectures as one of the positions that is ‘also true’.
This is rather puzzling given the vigorous criticism of realism in
lecture 1.1 assume that in lecture 1 Korsgaard was criticizing what
she took to be a crude version of substantive realism, and that the
realism which emerges as part of the winning team at the end is a
more sophisticated version; indeed, in the passage just cited
Korsgaard doesn’t say we just see that we are essentially citizens of
the Kingdom of Ends (and only contingently Muslims, atheists,
Serbs, etc.) but rather that ‘if you believe my argument’ you will
come to see this.
There should then be some kind of argument to the conclusion
that I am essentially a member of the Kingdom of Ends (and
only contingently an American). Or rather there need to be two
arguments: one to convince me that / a m necessarily a member of
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
Morality and identity
197
a Kingdom of Ends and then another to show the necessary universal extension for this Kingdom of Ends (for me) to all other
humans.
Korsgaard’s argument seems to run: if I reflect rightly I will see
that no other feature of myself but only my mere humanity is the
source of reasons and values for me. Thus I must see the mere
humanity of any other human as equally a source of value and
reason to act for me.
Korsgaard is well aware of the standard objection to this argument, namely I may well come to see my mere humanity as a
source of value for me, your mere humanity as a source of value
for you; how does it follow from that thatjyoar humanity must be a
source of value for me? The Serbs have what I can see are quite
good reasonsyor them to act as they do, reasons which (if you will) I
can see as arising from their ‘mere humanity’, but it doesn’t follow
that these reasons have any standings/or me; they aren’t reasons for
me to act the way they do or even to refrain from condemning
their actions. It seems to me an elementary fact of life of the late
twentieth century that we are constantly encountering people
whose reasons for action we understand perfectly well and which
we see are genuinely good reasons for them, without in the least
endorsing these reasons or sharing their values.
I think it is a grave mistake to run together questions of the
understanding of motives, reasons, and values and questions of
endorsement. We understand perfectly well why certain groups of
Muslims might want to kill Rushdie – he is a threat to their identity — and we can fully appreciate that the considerations that
move them are quite good reasons for them without in the least
thinking that they, or anything like them, are or would be reasons
for us (and also without thinking that we stand under any obligation whatever to fail to try to protect Rushdie from their acting on
their good reasons). We also assume, quite rightly I think, that the
only way to change their minds would not be to present them
with some new argument — they will have heard those that will
occur to us and are not impressed – but to engage in some much
more complicated process of restructuring their way of life.
Korsgaard’s response to this, if I have it right (and I probably
don’t) is that the only reason you could have for denying that
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
198
RAYMOND GEUSS
you stood under an obligation to accept what you acknowledge
are good reasons for them as also good reasons for you would be
that you thought of reasons and reasoning as inherently private
entities and processes, but reasoning is public and we can sometimes share reasons. Not only can we share reasons, but we
usually in fact do exist in contexts of shared reasons. For me to
reject the project of obligatory common reasoning with you ‘I
would have to hear your words as mere noise, not as intelligible
speech’ (4.2.10) and that is impossible (if you are speaking a language I know).
Note the shift here from: ‘We can share reasons (because they are
not in principle private entities)’ and ‘We usually do live in contexts
of shared reasons’ to ‘We must engage in common reasoning with
all other humans’. But from the fact that the sharing of reasons is
possible (or even, highly desirable) it only follows that I could get
together with you in a Kingdom of Ends (and that in many circumstances, when it is possible, it is also a very good idea), but not that I
must.
Equally I don’t at all see why I should be thought to have the
choice only of hearing your words as mere noise, or being committed to joint citizenship with you in the Party of Humanity. I can
quite easily understand you perfectly well and ignore you and
what you say. That might not be very nice, but it is certainly possible, and why is it even immoral? Even if it were to turn out to be
immoral just to ignore people in some contexts, that would be the
result of bringing to bear further moral argumentation on this
situation; the fact that I ignore you is something to be evaluated
morally (perhaps), not itself an automatic origin of a moral claim.
I’m surely not required to listen to everyone’s reasoning just
because he or she succeeds in producing some in my presence,
and I’m not even required to give any account of why I ignore
you unless there is some special reason for me to take you into
account.
Maybe it just is a Moral Fact (or a Divine Command) that we
stand under an obligation to take account of others in certain
ways. Even if it isn’t a moral fact, it might be a good thing to do,
but that would be for further reasons which we would have to try
to specify and it isn’t at all clear to me how one could get uniDownloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
Morality and identity
199
versal, strictly binding reasons to take account of everyone in all
circumstances. What one would be left with would be a highly
context-dependent, non-Kantian form of reflective endorsement.
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:35:58, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.008
LECTURE 8
History, morality, and the test of reflection
Bernard Williams
I THE NORMATIVE QUESTION
Korsgaard says that the normative question (which I shall label
[N]) is necessarily formulable in different ways. It may be helpful to
her argument to spell out more fully the relations between some of
the formulations. For instance, there are significant differences
between [Ni] ‘What justifies the claims that morality makes on us?’
(1.4.3), a n d [N2] ‘Is there anything we must do?’ (1.1.1, 1.3). [N2] is
at least broader than fNi], since there are non-moral forms of
normativity. Korsgaard accepts this, and indeed uses the notion of
means-end normativity to elucidate (via the idea of the will’s relation to itself) the moral sort of normativity. But this does not seem
to allow enough for non-moral forms of normativity (prudential,
aesthetic, etc.) which, like the moral sort, can equally give trouble
with inclination. It is not entirely clear to me whether Korsgaard
thinks that there is a problem about the nature of normativity
before we ever get to the specifics of morality. (Perhaps there is a
Kantian preconception hovering here, in the idea that it is only the
opposition of morality to inclination that really puts the nature of
normativity on the line.)
I take it that the reflective question [R] ‘Can morality survive
reflection?’ provides a way of approaching [Ni] ‘What justifies
morality’s claims on us?’ But fNi] and [R] line up neatly with each
other only if two things are granted. We have to assume (a) that the
reflective question about morality is concerned overwhelmingly
with its obligatory aspect, its ‘claims on us’: Aristotelians and others
might be more impressed by morality’s role as an enabling device
210
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:39, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.010
History, morality, and the test of reflection
211
for the agent’s own life, or by other considerations distinct from
those of obligation. The second assumption is (b) that the justification of morality is presented to the agent in the agent’s own reflection. Morality might ‘survive reflection’ in the sense that we could
recognize it as something that it is necessary to have around. Such
reflection would show, in effect, why enough moral claims had to
be recognized by enough people. I take it that on Korsgaard’s view
this kind of consideration would not in fact provide a positive
answer to [R] in the sense in which it is meant. More needs to be
said, then, to justify the requirements on [R] which Korsgaard
assumes (a Socratic requirement, in fact) that the answer to [R]
should be one that is given to each agent – or at least/or each agent,
as I put it in Ethics and the Limits of Philosophy.’
These questions are closely related to Korsgaard’s idea that [N]
and [R], as opposed to such issues as scepticism about the external
world, are both very continuous with practice. I think that here,
too, Korsgaard puts a lot of weight on the force of the conflicts
with inclination: each agent, if minimally reflective (in a fairly
uncontentious, not particularly philosophical, not culturally local,
sense) must confront, in her terms, the status of the moral claim
when it is so uncomfortably going against ‘her heart’s desire’. But
can so much be taken for granted in getting the discussion started?
Korsgaard is already confronting a motley crew of romantics,
Nietzscheans, Lawrentians, and merely Epicuri de grege porcos, who
will say that if the claim in question can and should have that much
power against the heart’s desire, it had better have a footing in the
heart’s desire. I come back to their concerns below, in sections 5
and 6.
2 REFLECTION AND EXPLANATION
The idea of morality’s failing the reflective test is that the true
explanation of why we have moral beliefs may not sustain those
beliefs. But why should the explanation of morality have to sustain
it in the sense of providing a normative reason for it? Why should
morality not be sustained, rather, by what is mentioned in the explana1
See pp. 39!!.
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:39, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.010
212
BERNARD WILLIAMS
tion? We should consider here what Thomas Nagel in his remarks
says about self-conceptions, and indeed these matters tie up with
Nagel’s much earlier work, about the absurd. Consider, for
instance, the true explanation of (the other bits of) her heart’s
desire: what is revealed in psychological accounts of the origins of
her passions may not normatively endorse them, but this does not
mean that it renders them meaningless when they are considered
‘from inside’.
How external to the agent’s existing concerns is the explanation
to be? If it is very external indeed (for example, very reductive),
maybe all human interests will lose their significance under such
explanation. If the explanation, on the other hand, uses much of
the material of the agent’s world, then it can make reflective sense
of morality, and much else beside (such as her heart’s desire). If the
demand is that reflection should endorse morality in virtue of the
way that it explains it, we need to be told the appropriate level or
type of explanation, and what its materials are, for this to be a reasonable or even a determinate demand.
We can be misled in this respect by the case of factual knowledge.
Here, not only are explanation and justification closely related to
one another, but the level of explanation for which this holds is
already understood in terms of the subject matter and (relatedly) in
terms of what count as claims to knowledge in the area. Suppose I
am disposed to believe that P. Then I can ask about the origin of
this state: and I answer, Q, which is another belief of mine (about
the origin of my belief that P). Moreover, CMS the kind of belief
that could tend to falsify P, or, again, support it; or if it does not
falsify P, it could falsify my claim to know that P. Granted this, it is
desirable that the particular Q_I come up with not be of either falsifying kind, and standardly it will be better if it actually supports P.
All this applies, however, just because of the relations between,
and the origins of, factual beliefs, and because the explanations that
we give of our own beliefs are of the kinds that can play a role in
validating claims to knowledge. (Edward Craig’s admirable discussion of the relations between knowledge and the origins of belief is
relevant here: see his Knowledge and the State of Mature.) But the relation of a normative attitude to its explanation (which latter is a
factual belief) surely could not, ex hypothesi,fitjust this pattern, on
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:39, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.010
History, morality, and the test of reflection
213
anyone’s view of the matter.
David Wiggins in some recent work (‘Moral Cognitivism, Moral
Relativism and Motivating Beliefs’) has tried to apply a notion of
‘justificatory (or vindicatory) explanation’ to certain moral beliefs,
in a way that is designed to parallel its application to some factual
and simple arithmetical beliefs. It is significant in relation to
Korsgaard’s strategy that Wiggins takes such an argument to work
only for very ‘thick’ ethical concepts. It is also significant that
Wiggins’s argument does not (in my view) actually work: basically,
because it begs the question of the respondent’s using the thick
concept in question. This is just the reason, I take it, that Korsgaard
would not want to follow Wiggins’s path: she needs a thin concept,
such as ought, that everyone must use.
3 A REMARK ABOUT MY OWN VIEW
Korsgaard perhaps suggests, at the end of lecture 2, that I think
reflective questions are only of philosophical, not practical, interest. I do not think this.
I agree otherwise with her account of my views, except that —
and this is only a matter of emphasis – she perhaps represents me
as rather more neo-Aristotelian than I am. I am more sceptical
than perhaps she suggests about the project of grounding die
ethical life in something like psychic health or a state of flourishing.
I have wanted to claim only that this project at least makes sense;
that it operates, so to speak, in the right corner of the field. I agree
with Korsgaard that the realist is in a weak position because he or
she raises a question that he or she cannot answer. Korsgaard raises
it and hopes to answer it. I myself think that if it is raised it cannot
be answered, but I am less clear than she is about what counts as
raising it, as comes out in my previous remarks about levels of
explanation, and about the status of [N], ‘the’ normative question.
My basic doubt can be put like this: is there a question which at
once
1 is about ultimate justification,
2 is rationally inescapable,
3 is practically relevant,
and 4 the answer to which justifies by explaining?
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:39, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.010
214
BERNARD WILLIAMS
4 A POINT ABOUT HUME
Hume did not think that there was a question that satisfied all these
conditions, even leaving aside 4. It is not clear to me that Hume’s
method answers the same question as Korsgaard wants answered.
In any case, there is a problem. Korsgaard says that for Hume,
the principles of the understanding fail the reflexivity test, but the
moral sentiments pass it. This seemingly depends on the principle
that if the understanding cannot justify itself, then it is not justified.
(Though it is less than clear, of course, what this last claim would
mean for Hume: the interpretation of his scepticism, its irrelevance to practical life, and so forth are relevant here.) But the
operation of the moral sentiments requires the principles of the
understanding, and the explanation of the moral sentiments
invokes the principles of the understanding (as Hume remarked).
So if ‘. . . cannot justify themselves’ implies ‘. . . are not justified’;
and if ‘.. . depends on what is not justified’ implies ‘. .. is not justified’; then it looks as though the moral sentiments are not justified
either.
5 FROM HUME TO BENTHAM
The lawyer in lecture 2, when she reflects, supposedly finds that
her disapproval seems ‘poorly grounded, and therefore in a sense
irrational’ (2.5.2); this can lead to utilitarianism, in parallel to an
historical development from Hume to Bentham. But this brings us
back once more to the level and type of explanation (which in this
case is self-applied). Does the argument not underestimate the
lawyer’s normative resources – particularly her ‘practical identity’
as a lawyer, to use a phrase that occurs in lecture 3? The lawyer can,
of course, have the thought: ‘I just happen to have been brought up
as a lawyer with these rules, etc . . .’, and then this thought about
her identity as a lawyer may go dead on her. But there is a real
question of what resources do or can go dead on a given person,
and of what this means (the question ties up with matters discussed
earlier, in section 2).
Korsgaard’s requirement seems indeed to be Kant’s, diat nothing
will serve as an adequate normative resource in such reflections
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:39, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.010
History, morality, and the test of reflection
215
unless ‘I just happen to . . .’ cannot even intelligibly be applied to it
(though, presumably, at a later stage we can go on to include
considerations that can be legitimated by being based on considerations that pass this test). But is this a reasonable requirement? People
say such things as ‘I just happen to love him …’, ‘she just happens to
be my daughter …’, or ‘I happen to be a vegetarian . . . ‘ (It is interesting that such formulae as ‘I happen to be a Catholic . . . ‘ have an
apologetic use in modern anti-dogmatic circles: no-one says ‘I
happen to be a Protestant . . .’ in places where such things really
matter, such as Belfast.) Indeed, might not someone say ‘I happen to
be someone who thinks in terms of principles’? And suppose
nothing (relevant) passes the test? As Nietzsche was disposed to say,
what then? Or rather (see next section): what now?
The immediate point, however, is that Korsgaard does not seem
to me to have located Bentham’s destructive (as I agree it to be)
‘advance’ from Hume securely in the process of reflection itself. I
think that what is weak in Hume is his conception of self-interest
and its relation to the moral sentiments, which themselves are conr
stituted by various extensions, projections, and universalizations of
‘self-interest.
I come back to the historical dimension itself in section 7 below.
6 PRACTICAL IDENTITY: OTHER PEOPLE
I see a Platonic inheritance here, in Korsgaard’s suggestion that die
person who is not (sufficiently) open to the claims of morality is in
some way cognitively defective in relation to others. Plato, using
this line, can plausibly be charged with having ducked the full force
of Callicles’ challenge, by equating the bad with the addicted or
brutal, whereas Callicles’ suggestion (his first suggestion, at any
rate) was that they did not need to be like that. A fortiori, they do not
need to be like that if diey are simply insufficiendy open to the
claims of morality. But let us exaggerate, and call the figure in question the ‘amoralist’. One cannot, without further argument, claim
that the amoralist has to be defective in these obvious, unlovely, and
unenviable ways. One also cannot say, without further argument,
that the amoralist regards other people as worthless; he may just
regard them as others (for example, as enemies). (If the argument is
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:39, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.010
2l6
BERNARD WILLIAMS
going to turn on what is involved in the recognition of ‘the Other’,
Korsgaard may want to consider someone who has tried to work
out similar ideas in a different style, Levinas.)
My main question, however, is not whether there are answers to
be found in this direction, but rather the following: even if there
are, how do they help to answer the questions [N]? How, for
instance, are these considerations supposed to work in the lawyer’s
reflection? What will they do to strengthen moral claims against an
agent’s heart’s desire? If acknowledgment of others is implicit in
one’s practical identity, then it is already so – already, when the
morally normative, and such things as her identity as a conscientious lawyer, supposedly go dead on her. Why is reflection on these
considerations about other people going to make the required elements come alive again? If obligation is ‘calling’, it is already so: if
others have the power to tell us what to do and make us do it (by
telling us, by existing, by being there), then what happened when
she did not hear them? (When I say that she did not hear them, I
mean that she did not hear the supposedly morally compelling
voice among them; her heart’s desire, after all, may well have been
listening to some one among them.)
I think that Korsgaard needs to provide some more argument on
two things, which are in the end the same thing, (i) How are the
considerations about others relevantly activated in reflection? (2)
How do they mesh into ‘practical identity’ in such a way as to
satisfy Kant’s requirement that one speaks to oneself, and under an
identity which one does not ‘just happen’ to have? (2) relates
directly to what Korsgaard says about others telling us and making
us do things. It relates, that is to say, to the Categorical Imperative
(perhaps, in this way, it relates more directly to the idea of a
Categorical Imperative of morality than Korsgaard suggests when
she says that Kant’s argument needs some supplementation to help
in the direction of the Kingdom of Ends; but I am not sure about
this). The idea that (some) people can make us do (some) things by
telling us to do them is quite helpful, I think, in understanding how
it is for the agent who is alive to obligation and its claims, but as of
course Kant saw, it cannot explain the force of obligation from the
ground up unless there is an account of how, from the ground up,
the agent accepts the force of what other people say. This is the
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:39, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.010
History, morality, and the test of reflection
217
familiar point that ‘Categorical’ is not a grammatical category:
unwanted, bullying, intrusive agents can make their imperatives to
me as unconditional in form as they like, but it does not make their
instructions Kantianly unconditional in the relevant dimension of
my having a reason to obey them.
Hence (1) and (2) come to the same thing. To repeat: even if there
is (which of course I doubt) a consideration linking practical identity in a sense that is inescapable with acknowledgment of others in
a sense that is morally sufficient, how can this link be mobilized
normatively in reflection, so as to answer Korsgaard’s own very
radical question?
7 HISTORICAL FOOTNOTE
Or, rather, a footnote about the very idea of the history. In comparing possible processes of individual rational reflection to historical
developments from Hobbes to Hume, and from Hume to Bentham
(or, better, from Hume to Kant), Korsgaard raises the question of
how these historical developments are to be understood. There are
several related questions. Does she accept a ‘Whiggish’ view to the
effect that the historical emergence of universalistic morality is, so
to speak, self-propelled (as might be implied, for instance, by
certain remarks made in the discussion about the ‘discovery of
equality’)? (I suspect that Nagel really does accept this, or at least is
so uninterested in any further explanations that it comes to the
same thing.)
If one accepts that historical and social developments were necessary to the emergence of universalistic morality – which is hard
to deny – one is faced with some notorious Hegelian problems.
First, does one accept that among the conditions of the emergence
of universalistic morality were many historical activities that
depended on the non-acceptance of universalistic morality? As
Hegel himself (and of course in a nastier, less redemptive, sense,
Nietzsche) asked, does the Kantian really wish that Kantian morality had prevailed?
Second, why should the history be supposed to stop at (roughly)
Kant? After all, the history in fact went on from Kant to Friedrich
Schlegel (as Geuss reminds us), and to Hegel (as Hegel reminds us),
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:39, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.010
2l8
BERNARD WILLIAMS
and might not as good a story be told about why it ‘had to’ develop
in these later ways as can be told about the earlier developments?
Indeed, cannot one also see why it should have gone on from Hegel
to (some version of) Nietzsche?
This is not meant to imply, in the spirit of traditional historical
materialism for instance, that we should be so impressed by the
onward march of the historical process that we feel we must accept
its latest stage. (That line of argument is anyway open to the
unanswerable objection — a valid version of the ‘naturalistic
fallacy’ argument – that if we have doubts about the ‘latest stage’,
then it cannot be that stage, but rather our doubts, that constitute
the latest stage.) The point is rather that if Korsgaard takes historical categories seriously (as she does, unlike many of her fellow
Kantians) she has to explain why the Kantian moment in that
history is privileged. She also, relatedly, has to take account of the
fact that history having got to the present point, it is not only
impossible to ignore that question, but it is also hard to take seriously most of the answers that have been given to it at earlier stages
of that history, for instance in the Kantian moment itself.
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:39, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.010
LECTURE 7
Universality and the reflective self
Thomas Nagel
Christine Korsgaard has provided us with an illuminating analysis
of the problem of the normativity of ethics. She observes that it is
the reflective character of human consciousness that gives us the
problem of the normative – the fact that unlike other animals, we
can fix our attention on ourselves and become aware of our intentions, desires, beliefs, and attitudes, and of how they were formed.
But it is not awareness alone that does it; a further aspect of our
reflective consciousness is involved, which can appropriately be
called freedom. Here is what she says:
Our capacity to turn our attention on to our own mental activities is also a
capacity to distance ourselves from them, and to call them into question
. .. Shall I believe? Is this perception really a reason to believe? . . . Shall I
act? Is this desire really a reason to act? (3.2.1)
The new data provided by reflection always face us, in other words,
with a new decision.
The normative problem does not arise with regard to everything
we observe about ourselves: we cannot decide whether or not to be
mortal, for example (though we may have to decide how to feel
about it). It is only beliefs, and acts, or intentions that face us with
the problem of choice, and it is in our response to this problem that
values and reasons reveal themselves.
Korsgaard’s account of the normativity of ethics, and her criticisms of rival accounts, appear within the framework of this
conception of the human mind and the human will. I should like to
explore her view by considering how she would answer the following three questions, which the reflective conception naturally
poses:
200
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
Universality and the reflective self
201
1 Why does the reflective self have to decide anything at all, either
granting or withholding endorsement, instead of remaining a
passive observer of the beliefs and actions of the non-reflective
self?
2 Why, when it decides, must it try to do so on the basis of reasons,
which imply generality, or something law-like?
3 How does it determine what those reasons are?
The third question of course comprehends all of moral theory
and epistemology, but I plan to discuss only the general character
of the answer.
Korsgaard’s answer to the first question, with which I would
agree, is that the reflective self cannot be a mere bystander because
it is not someone else; it is the very person who may have begun
with a certain unreflective perception, or desire, or intention, but
who is now in possession of additional information of a special,
self-conscious kind. Whatever that person now concludes, or
chooses, or does, even if it is exactly what he was about to do
anyway, will either have or lack the endorsement of the reflective
view — even, as Korsgaard observes, if he behaves like a wanton
and allows his first-order desires to move him without interference.
Given that the person can either try to resist or not, and that he is
now self-conscious, anything he does will imply endorsement, permission, or disapproval from the reflective standpoint.
The second question is more difficult. What is it about this
reflective individual that leads him to make the new choices with
which he is faced in a way that has universal implications – even
when he merely endorses or refrains from interfering with his firstorder beliefs or desires? As I understand her position, Korsgaard
believes, in my view righdy, that we cannot avoid giving an answer
which is in some way general. But why is this? Why isn’t the reflective individual just someone with more information, who can
therefore make choices which may be different but need be no less
purely personal – or even temporally fragmented? How do
reasons, law, and universality get a foothold here – one that cannot
be dislodged? Presumably it has something to do with the difference between reflective and unreflective consciousness, but why
should awareness of self bring with it this further regularizing
effect?
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
202
THOMAS NAGEL
Korsgaard’s final answer will be one which attempts to transcend
the opposition between the rational and the personal, and my
doubts about it have to do with that ambition. But she begins by
accepting an argument of Kant’s that I have never understood: the
argument from free will to the categorical imperative. The conclusion she draws from it is weaker, because she distinguishes between
the categorical imperative and the moral law – meaning by the categorical imperative only the principle that we must act according
to some law – but the form of the argument is the same. ‘The free
will must be entirely self-determining’, she says, paraphrasing
Kant. ‘Yet, because the will is a causality, it must act according to
some law or other’ (3.2.3).
But why is the second sentence of this argument true? If the will
is self-determining, why can’t it determine itself in individual, disconnected choices as well as according to some consistent law or
system of reasons? A neo-Humean regularity theory of causation
seems an inappropriate model for free self-determination. If the
idea makes sense at all, the free choice of actions which conform to
a law is no more nor less a form of causality than the free choice of
actions which do not. (And the same could be said of the free adoption of beliefs.) So far as I can see, choosing freely in a law-like
pattern is merely a way of mimicking causality; if I always put on
my left sock before my right, that does nothing to establish the
causality of my will, so why does the categorical imperative do any
better? There has to be something more compelling about the
demand for universality than this.
I think the true explanation is quite different. We are drawn irresistibly to the search for general reasons and justifications in the
endorsement of our actions and beliefs, not because of the requirements of causality but because of the externality of the reflective
view. This does have something to do with freedom, but Kant’s
argument obscures the connection. The freedom in question is
freedom from direct control by our impulses and perceptions. The
capacity for self-consciousness changes the nature of the being
who is making the choices — whether they are decisions to act or
decisions to believe – by introducing irrevocably the distinction
between appearance and reality, between how things seem from
our personal point of view and how they really are – and facing us
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
Universality and the reflective self
203
with the need and perhaps providing us with the capacity to arrive
at an answer that can be seen to be right not just from our individual point of view but from the reflective standpoint that takes that
view as its object.
The reflective self is in its nature more universal than the original, unreflective self, because it achieves its self-conscous awareness
by detaching from the individual perspective. The reason we can
no longer decide from the purely local perspective within which the
original appearances or impulses are found, is that once we observe
ourselves from outside, and achieve the distance of which
Korsgaard speaks, our choice becomes not just what to believe or
do, but what this person should believe or do. And that has to be a
decision about what any person so situated should believe or do,
since the external view does not give any consideration to the fact
that the person is me – it describes me in terms which would be just
as available to someone else sufficiently well informed about me.
Even before we reach this impersonal level, there is the perspective of one’s life as a whole, which leads to the search for principles to govern choice which will apply at any time — not just,
what shall I do now? but, what should I do in circumstances like
this whenever they occur? Given the reflective perspective, every
individual choice automatically becomes a general choice – even if
it is the general choice not to strive for law-like consistency, and to
act always on the impulse of the moment. (That is not easy, and its
analogue for belief — staying with your impressions — is even
harder.)
Suppose, then, that for whatever reason, it is granted that a
reflective being will inevitably be led to seek general or law-like
answers to the questions of belief and action which face him. The
next question, and the largest, is the third on our list. What if anything will determine his reflective response when faced with the
question what a person in his position should do or believe, in
general?
Though she accepts the Kantian argument that freedom implies
conformity to law, Korsgaard departs from Kant in holding that
the content of the law depends on something else, namely our
conception of our practical identity. This distinctly unKantian,
rather existentialist idea is the heart of her position. It introduces a
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
204
THOMAS NAGEL
strong element of contingency and therefore of relativism, because
depending on how we conceive of ourselves as reflective beings,
the law may be egoistic, nationalistic, truly universal, or just plain
wanton. In other words, occupation of the reflective standpoint,
though it implies law-like determination, can yield different results
for different individuals because each person has his own reflective
point of view. The law selected by the categorical imperative will
be the moral law only if that is the principle you regard as expressive of yourself, because you identify yourself as a member of the
Kingdom of Ends — rather than just as someone with certain interests, for example. This is Korsgaard’s answer to the third question.
We give consent to the law by identifying with a certain self-conception, and that also explains the law’s hold on us. Going against
such a law flagrantly enough is like destroying yourself.
A natural question about this practical self-conception is, first,
whether we are supposed to have some choice in the matter, and
second, whether there is any right or wrong about it if we do have a
choice. The way we think of ourselves seems in Korsgaard’s
conception to be an empirical matter of fact. Even if we could
change our stripes, it would seem that there could be no reasons
one way or the other for changing in a particular direction, since all
reasons have to depend on a pre-existing self-conception.
According to this picture, the final say goes to whatever determines
our identity. She says:
The test for determining whether an impulse is a reason is whether we can
will acting on that impulse as a law. So the test is a test of endorsement.
(3-3-5)But then:
When an impulse – say a desire – presents itself to us – we ask whether it
could be a reason. We answer that question by seeing whether the maxim
of acting on it can be willed as a law by a being with the identity in question. (3.3.7)
where ‘the identity in question’ is just whatever way I think of
myself:
Different laws hold for wantons, and egoists, lovers, and citizens of the
Kingdom of Ends. (3.37)
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
Universality and the reflective self
205
She discusses some of the pressures which push us toward wider
and wider communities of identification, including those which
derive from participation with other reflective beings in a public
language. But at the end of the line, the explanation of the content
of rationality that derives from these identifications is first-personal. It is a matter of who you think you are.
This is an anti-realist position in ethics, and I should like to
connect my objections to it with what I think is the true issue about
– realism. In lecture 1 (1.4.4), Korsgaard introduces a false opposition
between procedural moral realism, which she says is trivial, and
substantive moral realism, by identifying substantive realism with a
metaphysical belief in the existence of moral ‘entities’, ‘facts’, or
‘truths’ (though she turns Mackie’s sceptical use of the term ‘entities’ against him to wonderful effect at the close of lecture 4). But a
substantive realism need not (and in my view should not) have any
metaphysical content whatever. It need only hold that there are
answers to moral questions and that they are not reducible to anything else. Procedural realism, by contrast, is compatible with all
kinds of reductionism. The issue is, what does the truth or falsity of
statements about what we have reasons to do or believe, or what we
should do or believe, depend on? Does it depend on what we think
or what we choose, more or less, or not?
Korsgaard believes that in ethics, at least, it does. She does not
say whether she believes the standards of reasoning with regard to
theoretical and factual questions also derive from our self-conceptions. That too is a possible anti-realist view, but the two need not
go together.
I think that giving the last word to the first person is a mistake in
both domains. It is an example of the perennially tempting mistake
of seeking to explain an entire domain of thought in terms of
something outside that domain, which is simply less fundamental
than what is inside. In the end, the explanation of why a belief or
action is justified must be completable, if it is completable at all,
within the domain of the relevant reasons themselves.
In deciding, for example, whether to accept a perceptual
appearance or to substitute for it some other belief, the only thing
to do, once one has adopted the reflective view, is to think about
what the world probably has to be like, in order to explain why it
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
206
THOMAS NAGEL
appears as it does. In other words, you have to think about the
world, of which you are a part, rather than about yourself and who
you feel yourself to be.
Formally, it is the same witii morality and other practical issues.
To decide from the reflective standpoint what to do you have eventually to stop thinking about yourself and think instead about the
question at issue – not in this case about what entities the world
contains, but about whether what has made you want to do something is really a reason to do it. The answers to such questions may
partly determine your identity, but they don’t derive from it.
The temptation to offer an egoistic answer to egoism has been
a weakness of etiiical theory since the dawn of the subject. Korsgaard’s grounding of morality in a self-conception which you
would rather die than violate seems to me close to being an
example of this strategy. But it is not the only possibility. Here is
another one:
If someone accepts death rather than betraying a number of
other people to the killers, it might be unappreciative to explain
this in terms of the conception he had of himself. Of course if he
cares about the survival of the others, and is unwilling to save his
own life by betraying those others, then that is in fact an important
aspect of his conception of himself. But to explain the grip on him
of those reasons in terms of the self-conception would be to get
things backwards, and incidentally to cheapen the motive. The
explanation in its natural form can stop with the lives of the others
versus his own — not with romantic thoughts of what it would annihilate his personality to do, however useful these may be in stiffening his resolve. Even if he can get motivational help from
thinking that he couldn’t live with himself if he saved his life by this
method, that is not the final explanation – indeed it couldn’t be.
The real explanation is whatever would make it impossible for him
to live with himself, and that is the non-first-personal reason
against the betrayal. These remarks are not about what is on the
conscious surface of the mind, but about the person’s real reasons
for acting.
This alternative story sounds rather high-minded, I realize, and
to be frank, on the psychological level I have my doubts: perhaps
only romantic egoists can make sacrifices of this extreme sort — on
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
Universality and the reflective self
207
the model of religious martyrs who expect eternal bliss. It even
helps to explain why some genuinely heroic personalities are so
unbearable. But if that kind of answer is the only available solution
to the problem of normativity, then morality is an illusion, in my
opinion, and the sceptics are right.
I think moral truth requires another type of answer – one with
the universal applicability that Kant sought. The reflective attitude abstracts from the present moment and the point of view of
the particular individual, because it takes those perspectives as
objects of its attention. Suppose it faces the choice between
endorsing egoism or the moral law as the general principle of
conduct for someone occupying such a perspective. Since the
reflective attitude is being taken up by an ordinary human being,
he is of course thinking about himself as well as everyone else.
The issue is an evaluative one and it can only be addressed evaluatively.
I believe that the crucial question he has to answer is whether he
is prepared to regard that individual, reflectively considered, as
worthless. If so, then the reflective standpoint will offer no evaluative constraints to the carrying out of the individual’s personal
desires – will keep its hands off, so to speak. The reflective standpoint will bring no further reasons to bear on the individual’s
choices, in addition to those which appear from within his individual point of view, because it will then regard all reasons as existing
onlyTor individuals, in virtue of their aims or interests. This is itself
a general attitude, but one which supports egoism (or wantonness,
if even the egoistic value of present satisfactions is denied). Egoism
as a general principle is equivalent to regarding myself as valueless
from a reflective point of view, because it says that my interests, like
those of every other person, provide others with no reason for
action except in so far as they can be linked to the other person’s
prior motives.
If, on the other hand, from a reflective standpoint we do not
regard ourselves as worthless, then we must accord a more general
weight to at least some of our reasons for acting. And because of
the character of the reflective attitude, this weight will automatically be accorded to similar reasons arising in the lives of others,
and these will in turn constrain what we are justified in doing in the
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
208
THOMAS NAGEL
pursuit of our own lives. None of this can be explained in terms of
our practical self-conception, though it might well be described as
determining our practical identity.
I may be overestimating my disagreement with Korsgaard. She
too believes that morality arises from the value we find it irresistible
to grant ourselves as sources of reasons once we take up the reflective attitude towards our own actions. And she too wishes to identify her argument as a version of Kant’s argument that, if we are to
take our desires as providing reasons for action, we must on reflection regard ourselves, and hence humanity itself, as intrinsically
valuable. The trouble is that she seems to hold that the kind of
reflective value that comes out of this argument is consistent with
egoism (4.2.1). But so far as I can see, the same kind of reflective
judgment that yields reflective endorsement of rational self-interest can be carried further to yield reflective endorsement of values
which obligate others. In each case, it is a matter of being faced
with the alternatives, and having to decide which is more credible.
We do not make these things true by taking some kind of leap, or
even by taking a cautious collective step. The invocation of
Wittgenstein doesn’t help, because egoism doesn’t violate publicity.
I don’t deny that some values are adopted or created, but morality,
in its basic oudines, is not among them. Our practical identity is its
product, not its source.
Because I don’t know where else to put it, let me close with an
unnconnected exegetical point. I think that in lecture 1,
Korsgaard misinterprets Hobbes. He was not, in her sense, a
voluntarist, because he did not believe the command of the sovereign was the source of obligation. Rather, the sovereign’s commands, and his monopolistic capacity to enforce them, remove
the excusing condition of insecurity which makes the laws of
nature oblige only inforo interno when we are in the state of nature.
Even the command of God is not the source of moral obligation.
I read the passage from Leviathan she quotes in section 1.3.2 as a
purely linguistic point – that we can’t literally call these moral
principles ‘laws’ except in so far as they are commanded – but
they oblige us nonetheless, since they are rational dictates of selfpreservation, which is our first aim.
Korsgaard has put before us a characteristically rich, ambitious,
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
Universality and the reflective self
209
and original system of philosophical ideas developed to the highest
intellectual standard. My comments, largely in opposition, constitute a very selective and inadequate response to the whole; her
project is of the first importance.
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:36:22, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.009
LECTURE 9
Reply
Christine Korsgaard
My commentators have raised many important questions about,
and made some very forceful objections to, what I have had to say
in these lectures. I can only wish that I could give satisfactory
responses to them all. In what follows, I address just a few of the
points which they have raised. Specifically, in section 11 address the
question why we must will in accordance with a universal law. In
section 2 I discuss some ways in which, according to my commentators, my account of obligation departs from Kant’s, to its detriment, and I try to defend myself both against the claims of
departure and of detriment. In section 3 I discuss the status of
desire, both in Kant’s account and in my own. In section 41 take up
the question whether my focus on the idea of identity makes my
account of moral motivation unattractively egoistic. Finally in
section 5 I consider some issues about the relationship between
giving a psychological explanation of the sense of obligation, and
giving a justification of obligation itself.
I AN OLD QUESTION RAISED AGAIN! THE
UNIVERSALIZABILITY REQUIREMENT
Near the beginning of lecture 3 (3.2.3), I cited the argument by
which Kant undertakes to establish that we must submit our
maxims to a test of universalizability. Kant argues: (1) that we must
act ‘under the idea’ that we have free will, where a free will is one
which is not determined in accordance with any law external to
itself; (2) that a free will, if it is to be a will at all, must nevertheless
be determined in accordance with some law or other; (3) that it
must therefore be determined in accordance with its own law – that
219
Downloaded from https://www.cambridge.org/core. University of Sheffield Library, on 20 Jan 2021 at 02:37:01, subject to the
Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511554476.011
220
CHRISTINE KORSGAARD
is, be autonomous; and (4) that this shows that the categorical
imperative is the law of a free will. For by this point in the text (the
opening of Grundlegung, chapter in), Kant has already shown that
the categorical imperative is the law of autonomy. And indeed this
is clear in any case, since the categorical imperative tells us to
choose a maxim which has the form of a law, and that is what an
autonomous will by its very nature must do – it must choose a law
for itself. The categorical imperative, in fact, simply tells us to be
autonomous. In so far as we must act autonomously, we must of
course conform to it.
Confronted with this astonishingly simple argument, it is impossible not to feel that some sort of sleight of hand has taken place;
and, accordingly, Kant’s readers have protested at almost every
point. Let me review the objections here.
The first objection is that Kant has not shown that a free will’s
dictates must be universal, even in a purely formal sense. Cohen,
for instance, who agrees with me that the ability to reflect puts the
will in a position of self-command, asks why the will cannot give
itself singular commands, edicts, or orders, rather than deriving its
reasons from general principles.’ A yet more radical version of this
objection emerges in Geuss’s invocation of Schlegel, who thought
that true freedom consists in violating one’s laws, proving that one
is something above and beyond any law.2 Why, then, is what the will
requires a universallaw?
Assuming that this objection can be surmounted, we then get
the series of related objections traditionally comprehended under
the complaint ’empty formalism’. The first of these is the one I
myself deployed in the lectures: that until we settle the domain over
which die law universalizes, the requirement of universalizability
does not yield any particular content. We must argue that the law
ranges over human beings or rational beings in order to get what I
called ‘the moral law’, and that, according to the objection, cannot
be done. So universalizability does not get us to morality. Next
comes the version of the objection made most familiar to us in a
long tradition stretching from Hegel and his followers and John
Stuart Mill down t…