§
Ens.

til. 3: l
Rug?
4..

a:
.héil

.. :3.
. .9. x..:.....!!:
g l‘lttf..51.
s I )2. i. l. v i
l f. ... (.0 .4. 21.! wit-.1...
. , . 1...! pro. vll

 

3.1.1....»0 . . .2711}: .0 .r. ,
. . . . . . V .'...a:u.l......:

 

W343}?

200?

LIBRARY
Michigan State
University

This is to certify that the
dissertation entitled

THE UNITY OF IDENTITY:
A DEFENSE OF AN IDEAL

presented by

JOHN ZILLMER

has been accepted towards fulﬁllment
of the requirements for the

 

 

 

 

Ph.D. degree in Philosophy
Jm,
Major Professor’s Signature
00 9
Date

MSU is an affirmative-action, equal-opportunity employer

-a-a-o-Q-o-o-n-b-o-—-—.—-——._-_. _. _--.—-o-—.-g-o—a--n-n-a-o-u-n-o-o-n—--n-o--a-c-o—-a-.--o—.‘-.-n-o-o-o—-

 

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

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K./Prolecc&Pres/ClRC/DateDue‘indd

THE UNITY OF IDENTITY:
A DEFENSE OF AN IDEAL

By

John C Zillmer

A DISSERTATION
Submitted to
Michigan State University

in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Philosophy

2008

ABSTRACT

THE UNITY OF IDENTITY:
THE DEFENSE OF AN IDEAL

By

John C Zillmer

The elements of one’s personal identity, including desires, projects, principles, group
membership, self-conceptions, and narratives are ideally uniﬁed. Uniﬁcation of identity
elements is characterized by a lack of conﬂict between these elements. Personal
identities are ideally uniﬁed since identity conﬂict undermines personal autonomy, moral
and other sorts of integrity, and both commitment and valuing within romantic

relationships.

TABLE OF CONTENTS

INTRODUCTION

DISCOVERING OR POSULATING THE ONE ......................................... 1
CHAPTER 1

UNITY, FRAGMENTATION, MULTIPLICITY ......................................... 8
CHAPTER 2

THE UNITY OF PERSONAL IDENTITY ............................................... 48
CHAPTER 3

AUTONOMY ................................................................................. 109
CHAPTER 4

INTEGRITY ................................................................................... 165
CHAPTER 5

LOVE: VALUE AND COMMITMENT ................................................. 207
CONCLUDING THOUGHTS ............................................................. 260
BIBLIOGRAPHY ............................................................................. 263

iii

— Introduction —

Discovering or Postulating the One

One way of dealing with the problem of diversity,
of the many, is by discovering, or postulating, the
one. (Why diversity is a problem is another
question, one I haven’t an answer to. What’s clear is
that for Western epistemology and politics it has
been a problem, perhaps the problem.)

- Naomi Scheman, Engenderings’

This work does not assume that diversity is the problem. The idea that there is
one, or even one central, type of problem is difﬁcult to accept; there are lots of problems,
and unfortunately for those who seek universal solutions, not all problems (and not all
answers) boil down to the one and the many. But in some ways diversity — the many —
can be a problem.

The speciﬁc sort of diversity that this work — one not in epistemology or politics
but in moral psychology — takes to be problematic is that of a disuniﬁed personal identity.
The overall aim in the present work is both to detail some of these ways in which certain
sorts of diversity within one’s personal identity can be a problem, and to attempt to
discover — rather than postulate — the solutions unity can offer.

Some introductory remarks on ‘unity’ and ‘personal identity’ are in order here.

One sort of unity that might characterize one’s personal identity is metaphysical. Derek

 

‘ Scheman (1993), p. 186.

Parﬁt has famously argued that there is no persisting subject of experiences such that
some future person is identical to the present person who is me.2 Such a view thus takes
‘personal identity’ to refer to just that thing the temporal persistence of which is
tantamount to the temporal persistence of the person. Parﬁt takes the subject of
experiences to fulﬁll this role; others — John Locke is typically thought to have held this
position — have taken memories to do so.

Parﬁt’s defense of the claim of diachronic disunity can be contrasted with an
argument denying the existence of synchronic metaphysical unity, such as that of Roland
Puccetti.3 Puccetti appeals to the behavioral split that occurs under certain conditions in
callosotomy patients — persons who have had the hemispheres of their brains surgically
separated — to argue that even neurally intact persons are composed of two
metaphysically independent beings. Thus, whatever it is that underlies behavior (Puccetti
is noncommittal as to just what constitutes this behavioral substratum), it is just this that
constitutes one’s identity.

The sort of unity of personal identity that the present work will take up is not a
metaphysical but rather a practical unity. For example, one’s gender identity has effects
on what she does, effects that are not a function of any physical characteristics that might
be thought to be associated with her biological sex. To illustrate: a woman has reason to
be cautious of accepting etiquette advice, since it is in virtue of her conceiving of herself
as a woman as well as 3m-person attributions of this identity to her that she is vulnerable

to exploitation through such channels as norms of feminine propriety.

 

2 Parfit (1984/1987).
3 Puccetti (1973).

Each person has more than one such practical identity: one may identify as, and
be identiﬁed as, a woman, and a Native American, and a mother, and a triathlete. These
identities might be practically uniﬁed - that is, they might have practical effects that are
concurrently realizable. More likely, though, some of a person’s identities are disuniﬁed.

What is it for an identity to be practically uniﬁed? This question has only an ideal
answer, since practical uniﬁcation is realizable only as a matter of degree — and, thus, the
claim that an identity is uniﬁed should throughout this work be taken as a relative, not
absolute, claim. Just as we call a bicyclist ‘fast’ or an artwork ‘beautiful’ without
thinking them the fastest or most beautiful, or perfectly fast or beautiful, we can call an
identity ‘uniﬁed’ and mean by that not perfect uniﬁcation (whatever that might be) but
instead mean that the identity has the feature ‘unity’ to such a degree or in such a way
that it is appropriate to note this feature.

One indication of a practically uniﬁed identity is that an action done for reasons
that ﬂow from one’s identity will tend not to be done ambivalently. A person who is
unambivalent about, say, eating organic food is wholeheartedly in favor of doing so, and
as such she disregards counter—considerations (such as the greater monetary cost and
relative unavailability of organic food). Talking to people who eat organic readily brings
to light two points about ambivalence: it is a relative notion, since virtually everyone has
some of it; and to be unambivalent is not to be single minded, as even those who are
wholehearted about a goal have many other goals as well. (Talking to such people
indicates that eating organic food is something that can be strongly identity—constituting,

as well.)

Another characteristic of a practically uniﬁed identity is that each of one’s actions
that ﬂow from reasons conferred by her uniﬁed identity will be unlikely to conﬂict with
other of her actions. A person whose practical identities include eating organic food as
well as frugality, for instance, will have regular practical conﬂicts, and as such her
identity is disuniﬁed. One’s actions will tend to cohere over time, as well, so a person
whose identity is diachronically practically uniﬁed will be less liable to undermine her
past projects with her present ones.4

This sort of practical unity of identity is one that, I’ll argue, is an ideal that is
desirable to hold. As an ideal, it is strictly speaking not perfectly realizable.
Nevertheless, the asymptotic approaching of the ideal can improve the lives of those who
do so. The argument for this claim will appeal to the ways in which the possessing of a
uniﬁed identity is a precondition both of acting autonomously and with integrity, and is
an important, if not strictly necessary, condition of romantic loving.

These aspects of human life are the focus here for one principle reason: each is an
important element of living well. Acting on one’s own principles and beliefs and
motivations, that is, autonomously, is a precondition of one’s leading a good life that it
often — in daily life, though not by philosophers — goes unexamined. Even Buddhist
writers — those who, having examined self— governance and found it lacking — imply that
the goal of no-self is to be attained only by one’s own wholehearted attention. Even
those who recommend an attitude of caring for others rather than one of preserving one’s

own rights and liberties endorse (one would hope) only self-motivated altruism; it would

 

4 It might be noted at this point that there are other reasons that one’s actions may conﬂict. One example is
the fact of value pluralism, which virtually guarantees conﬂict due to incommensurable grounds for
practical decisions. I will argue in Chapter 2 that the most important of such conﬂicts, and the values that
play a role in them, are ultimately grounded in practical identities.

be at the very least conceptually suspect to advocate forceful instigation of an ethic of
care. In any event, the importance of autonomy goes even deeper than such axiological
considerations — that being autonomous is, or contributes to, what makes a person’s life a
good one. While there may be some truth in that understanding, autonomy might more
clearly explain what makes someone’s life, good or bad, her own life at all.

Acting with integrity is perhaps a more ideal, rather than expected, characteristic
of a good life, since even those lives without excessive social and psychological barriers
to acting with integrity do not feature such actions unremittingly. Nevertheless, it is
axiomatic that, ceteris paribus, to act with integrity is better than not to; the term is often
used synonymously with a wide range of moral praise terms such as honesty, courage,
and reliability, and is sometimes used as a more general term of overarching personal
approbation. Though some certain situations might well demand that integrity be
forgone, the ability to act with integrity — and the regular appropriate exercise of this
ability — is desirable.

Loving is an especially important element of living well. The objects of our love
are sources of strength, meaning and pleasure, and without them life would be
impoverished. It tells a great deal about the importance of loving that a person is more
likely to die after the passing of her life’s partner, and suicide is much more common
among those who have recently ended a romantic relationship.5

Each of these aspects of a good life — autonomy, integrity and loving — is (and
likely has always been) undermined by forces both internal and external to particular

persons. Autonomy is thought to be undercut by ambivalence (arguably more prevalent

 

5 Kposowa (2000).

now than historically), coercion and oppression. To act with integrity, that is, from one’s
own moral principles in the face of pressure to do otherwise, is hampered by social norms
that selectively prescribe submission, the insurmountable resistance caused by various
oppressive forces, and the inability of some to act morally in certain social contexts.
Resistance to the ability to love is presented by (among many other things) social
pressure to value wrongly and the decline of commitment.

Each of these elements —- autonomy, integrity and loving — are in what follows
here the subject of a discussion that both outlines the philosophical terrain and defends a
certain conception of the notion. I then explain the ways in which the unity of identity
bears on the subject, showing how the possession of a uniﬁed personal identity improves
the ability to act with autonomy or integrity, and to love — that is to say, how a uniﬁed
identity is an ideal insofar as it underpins these elements of a good life.

Before these speciﬁc arguments are launched, two tasks are in order. One is a full
account of the notion of the unity of identity that will feature in the following arguments.
Though there is little in the way of explicit accounts of the uniﬁcation of identity in the
philosophical literature, there are a range of possible forms such unity might take. But
ﬁrst it is necessary to set the stage by considering the opposition to the ideal of unity. In
the eyes of many who are resistant to arguments for uniﬁed identities, any proffered
account of unity would be, I fear, a complete non-starter.

This resistance is held for understandable sorts of reasons, to be sure. Unity
ideals have been implicated in an oddly a priori sort of oppression of those whose
identities are multiplicitous or otherwise not amenable to reduction to a unity. The

Kantian uniﬁed will and the singular Cartesian cogito are taken as representatives and

even agents of patriarchy; contemporary accounts of autonomy and integrity are taken to
be further marginalizing of those who, because they are already marginalized, do not ﬁt
the criteria for these capacities; given that the social world is less than uniﬁed, to advise
that we inhabitants of it be uniﬁed seems to some an unattainable ideal.

These are serious charges. The account defended here will be as answerable to
these concerns as is possible. I believe that the notion of a uniﬁed identity can
accommodate many of the concerns of those who oppose it; I believe such an account can
accommodate all the most serious concerns. Nevertheless, the notion of personal unity
has been heavily criticized, and as an ideal, it must be admitted that the unity of identity

has a deep hole to climb out of before it can brush the dirt off and make itself presentable.

_1_

Unity, fragmentation, multiplicity

Do I contradict myself?
Very well then I contradict myself,
(I am large, I contain multitudes.)

-Walt Whitman, “Song of Myself”

Unlike Whitman, who in the above well-known passage admits his satisfaction
with a disuniﬁed self-conception, a number of prominent philosophers have argued — or
assumed — such a self-conception to be undesirable. Plato claimed that internal disunity,
like disunity within political and social organizations, entails injustice. After arguing that
injustice is inherent in an internally divided city, army, or family, he asks:

And is not injustice equally fatal when existing in a single person; in the

ﬁrst place rendering him incapable of action because he is not at unity

with himself, and in the second place making him an enemy to himself and

the just? Is that not true, Thrasymachus?l

Distinct from Plato, Kant thought it not (only) immoral but rather impossible that
an underlying disunity marked a person. The condition of having one uniﬁed
experiencing mind, the “transcendental unity of apperception”, is a condition he argued to

be a minimal condition of conceptualizing anything at all.2 Surprisingly to some, this is

intended to be even less of a metaphysical claim than that of Plato; our unity, for Kant, is

 

‘ Republic, 352.
2 Critique of Pure Reason, B 139-142.

a matter of our subjective self-conception, for “as far as inner intuition is concerned we
cognize our own subject only as appearance but not in accordance with what it is in
itself.”3

If, as is sometimes claimed, philosophy is a series of footnotes to Plato, and
modern philosophy footnotes to Kant, these two conceptions of unity will ﬁnd their way
into the present work whether intended or not. I will not explicitly address these views,
though - Kant’s being more cognitive than volitional, and Thrasymachus’ ‘yes, certainly’
being offered much too easily — in favor of addressing criticisms of contemporary views
of uniﬁed identity.

The Received Model

Of contemporary philosophers, two defend views that together draw the majority
of the criticism aimed at the ideal of unity. Often taken to defend a virtually identical
account, Gerald Dworkin and Harry Frankfurt hold views that are sometimes criticized as
one; John Christman characterizes them as such in terming them “the received model“.
Though sharing the same conceptual core, that of the hierarchical structure of the person,
the two accounts diverge both in ways that will be described immediately below, and
especially in ways that become apparent in responding to critics. These distinctions
between the two accounts will serve as response to some of the criticisms of uniﬁed
identity that this chapter will address, since certain criticisms of Dworkin’s account are

answerable by components of Frankfurt’s view. The remainder will be answered in the

spirit in which at least some have been intended: being corrective demands upon what is

 

3 Critique of Pure Reason, 8156.
" Christman (1991), p. 4.

(I will argue) an essentially defensible view, alterations to the inﬂuential accounts
described here will be suggested.

Dworkin’s and Frankfurt’s accounts share a notion of identiﬁcation. Both are
concerned with the effects of identiﬁcation on autonomy, as well, and both have a
hierarchical view of the self. Outside of these similarities, the accounts differ, as will
become clear.

What a person identiﬁes with are “the inﬂuences that motivate him”, on
Dworkin’s view.5 Such inﬂuences, termed ﬁrst-order motivations, can be desires, wants,
aims, motives, et cetera. For one’s ﬁrst—order motivations to be authentic her second-
order desires must “be congruent” with them.6 In this way — through wanting to be so
motivated — she identiﬁes with those ﬁrst-order motivations. One is not thus identiﬁed if
she is alienated from or resents her ﬁrst-order inﬂuences.7

Frankfurt’s notion of identiﬁcation is nearly identical. A desire that a person is
moved by is her own if she has a second-order volition by which she endorses her own
acting on the desire. Frankfurt admits some complexities to this simple core view,
though. Not all desires are ones that a person may wish to be moved by, even though she
might still identify with them and desire that she have them. For example, a person with
a desire for whiskey and also a desire for sobriety may identify with both, admitting that
both are desires are hers. But she may wish to act on the desire for sobriety - in light of a
second-order volition - and still identify with the desire for whiskey, though not want to
act on this desire. Thus the object of identiﬁcation need not be something that one

intends to be motivated by.

 

5 Dworkin ( 1988), p. 15.
6 Dworkin (1988), p. 15.
7 This latter point plays its strongest role in Dworkin (I970).

10

 

Usually, though, a desire that a person wants herself to have is a desire that she
wants herself to act on.8 It is also a usual occurrence that a person is ambivalent about
whether she wants to act on a certain desire. Though we often describe such a situation
in terms of conﬂicting reasons, the (compatible) description in terms of identiﬁcation is
this: some of one’s hi gher-order volitions are consistent with the desire, and some are
not.9 A person might be ambivalent regarding her desire, say, to marry Jorge; she may
love him, and want to legally join herself to him, but she might also desire to further the
aims of a certain sort of feminism which frowns upon heterosexual marriage.

Frankfurt’s intertwined notions of ‘resonance’ and ‘wholeheartedness’ are offered
as ideal-case characterizations of identiﬁcation. When a person identiﬁes with a desire,
(to continue the example) the desire to marry Jorge, wholeheartedly, she has no internal
conﬂict regarding the desire. In this way, her identiﬁcation with the desire resonates
throughout her entire volitional economy; like striking a tuning fork in the vicinity of
other tuning forks, the others return the reverberations at a harmonious pitch.

She may of course have external conﬂict regarding the desire: her best friend may
advise against her marrying Jorge, or Jorge may already be married to someone else.
This conﬂict makes for nuptial trouble, but not for identity trouble. Also a mode of
external conﬂict is a case in which her desire to marry Jorge conﬂicts with other desires
that move her but with which she does not identify — Jorge is shorter in stature than she

is, suppose, and while she does not want to be moved by her aversion to shorter partners

 

8 ‘Usually’, that is, as a matter of frequency. As a tangentially interesting matter, there is a conceptual core
to the notion of a ‘desire’, though, that makes for a certain amount of tension in claims about desires that
one does not desire to be moved by. In one sense, Ulysses desired to be lured by the sirens, but he did not
desire to be moved by their so luring him. Ulysses desired the phenomenological experience, but the
associated motivation. Thus one might truly say that Ulysses did not desire to be lured by the sirens.

9 A more complete defense of the desirability of understanding practical conﬂict in terms of identiﬁcation
rather than reasons is contained in Chapter 2, below. For the time being it should be sufﬁcient to note that I
do not deny that there are conﬂicting practical reasons.

11

(she realizes that her feeling is tied up in some complicated way to norms of feminine
submission to which she does not subscribe), she cannot ignore the feeling. Thus, her
desire to marry Jorge conﬂicts with a desire that is external to her, that is, a desire with
which she does not identify.

It is important to note the content—neutrality of both these accounts. The internal
relation of a person to a desire, not the content of the desire itself, is what makes the
desire part of the person’s identity. There may be some limits to the sorts of desires one
can take up; the desire to thwart all of one’s own goals may be one with which one
cannot identify, but this for structural reasons -— it would necessarily conﬂict with other
desires.lo But one could, on a structural account, identify with the desire to have no self-
respect, or the desire to be abused, or to be subordinate to others. Whether a desire is
properly said to be one with which a person identiﬁes is a matter of her relation to it, not
a matter of the desire’s appropriateness for a person (on some explicitly morally
normative conception of ‘person’) to possess.

Another feature of note is the conception of the self as constructed entirely out of
desires. While the locutions of Dworkin and Frankfurt include with the archetypal
‘desire’ such notions as ‘goals’, ‘aims’, and ‘projects’, it seems never to have been the
intent of these original accounts to accommodate categorically different identity—
constituents. This restriction seems at times to be an explicit one: Frankfurt claims, about
someone who doesn’t care which of his desires he is moved by, “he has no identity apart

from his ﬁrst-order desires”.“ More often, though, this restriction of non-connative

 

'0 And also, of course, with itself.
” Frankfurt (1988), p. 18.

12

elements from personal identity is made by exclusion; Frankfurt and Dworkin’s
arguments and examples simply do not reach outside the category of desires.

This limitation is misleading. The identities of persons, and thus the persons
themselves, are more plausibly described as comprising various characteristics (‘tall’,
‘playful’), group memberships (‘Ojibwa’, ‘feminist’, ‘union worker’), principles
(‘generous’) or abstracta (‘moral agent’).I2 But while I think that the idea of identity’s
being only a matter of desires a possibly misleading oversimpliﬁcation, structural
accounts of identity need not be grounded in such a reduction. Since the accounts of
Dworkin and Frankfurt are described in terms of desires I will talk about them in this
way; those of the following criticisms that are cast in terms other than ‘desires’ will
introduce the emendation of a basic structural view, to be completed in the following
chapter.

Criticisms of uniﬁed identity

Neither Dworkin’s nor Frankfurt’s is explicitly an account of identity. Both have
offered their account of identiﬁcation in the interests of grounding a certain conception of
autonomy. This fact has not prevented the accounts from being the subject of a great deal
of criticism regarding their claims about identity. Because these two accounts have been
so philosophically inﬂuential — positively as well as in the negative sense that will
become clear in the course of addressing criticisms - to show how resilient these
representative hierarchical accounts are is to begin to make plausible the ideal of unity.

Criticisms cluster around four distinct issues, which I will brieﬂy sketch here

before proceeding to a deeper analysis of each. At the most theoretic level, criticisms are

 

‘2 These categories are not mutually exclusive; ‘feminist’ ﬁts well into the ﬁrst three groupings. Nor are
they exhaustive.

13

spurred by the fact that Dworkin’s account (the speciﬁc target of John Christman, the
critic I will consider here) holds the structure of desires entailed by identiﬁcation to be of
a hierarchical nature. Christman argues that this hierarchical form subjects the account to
an unacceptable notion of ‘identiﬁcation’, in two ways. ‘To identify’ involves a state of
either evaluative approval or bare acceptance, neither of which characterize the range of
identiﬁcations. Secondly, the temporal process by which one comes to be so identiﬁed
are problematic, since endorsement on the grounds of a coercively acquired hi gher-order
desire is not, Christman claims, properly called ‘identiﬁcation’."

The second sort of criticism is that a structural account of identity cannot
accurately characterize the “adequacy of identity”‘4: adequate, that is, not only for mere
self-govemance, but for self— governance over actions that aim at one’s own best interests.
The quoted phrase and the language of ‘best interests’ belong to Susan Babbitt, but the
concern is shared by many other philosophers.l5 All sharing this concern agree that if a
person is properly said to be acting from her own identity this must entail that her actions
thus motivated aim (at least fallibly) at her own well-being, not just at the satisfaction of
desires she happens to have.

The third sort of criticism is that the unity of identity can entail ossiﬁcation or
rigidity that persons neither have nor ought to have. Diana Meyers claims, “Frankfurt’s
account shows how very difﬁcult it is to explicate integration without appearing to
recommend personal ossiﬁcation”.16 The internally uniﬁed are thought to be obdurate

and static, rather than resilient and open to change.

 

'3 Christman (1991).

‘4 Babbitt (1996), p. 104.

‘5 Benson (1991), Christman (2004), Meyers (1989).
'6 Meyers (2000), p. 169.

14

On the same subject, Amy Mullin notes, “it is important to avoid assuming that
effectively uniﬁed selves must be homogeneous or integrated to the point that harmony is
rarely threatened. Such a conﬂation of unity and homogeneity unduly limits our options
for coping with diversity and has damaging political consequences”.l7 These damaging
political consequences are the fourth point of criticism; these political consequences are
claimed to be due to a number of distinct features of the ideal of unity. One such feature
is the one that Mullin notes in the above passage: ‘unity’ is sometimes taken to be
‘perfect unity’, a pure state allowing of no conﬂict. Mullin uncovers another potentially
problematic feature of the ideal of unity in claiming an association between the supposed
unity of groups and the unity of identity: the fact that “. . .the appeal of unity is used to
eliminate the disadvantaged or marginal suggests another reason why we should be
uneasy about attempts to rid ourselves of inner differences or conﬂicts”l8

These criticisms — of hierarchical structure, inadequacy of identity, ossiﬁcation,
and damaging political consequences — come from different quarters: conceptual, such as
the concept of autonomy and the implications of structure; and practical, as the concerns
about oppression and resistance. These distinctions aside, each of the criticisms is culled
from a discussion that shares the practical focus of the conception of identity the present

work defends, and so the responses offered here will have this practical conception in

mind.

 

'7 Mullin (1995), p. 3.
‘8 Mullin (1995), p. 15..

15

0f hierarchical structure

The very basic conceptual feature of uniﬁed identity — structural unity of a
hierarchical form — has been questioned perhaps most thoroughly by John Christman.19
Christman’s project is the defense of a certain conception of a person’s volitional
economy — desires and beliefs that give one reason and motivation to act — on which she
is able to be autonomous. Though Christman’s ultimate concern is for autonomy, which
will be more closely analyzed in the third chapter of the present work, his criticisms focus
on the notion of identity that grounds the characterization of the autonomous person.
What it is for a person to identify with her desires, on the received model, is inadequate
to characterize the self—determination constitutive of autonomy. Christman takes
Dworkin’s hierarchical account as exemplary of the model, and in a supposition shared
with other philosophers, claims that the components of, and problems with, Frankfurt’s
account are the same. In light of the distinctions explained above, the criticisms
Christman offers have force not against hierarchical accounts generally but only against
Dworkin’s version. As I’ll go on to show, the criticisms therefore do not undermine the
ideal of unity in the way they are intended.

Christman argues that hierarchical models are subject to three problems.20 One is
a difﬁculty with the “condition of identiﬁcation.”21 Identiﬁcation with a desire occurs
when a person reﬂects on her desire, the desire to marry Jorge, suppose, and approves of
that desire - an approval grounded in one or more of her higher-order desires. Christman

offers two readings of this condition: (a) she acknowledges the fact of her possessing a

 

'9 Christman (1991).

m The tripartite taxonomy here is somewhat arbitrary; Christman’s explicit divisions between issues are not
as clear, and while this is not a problem in the original presentation, the divisions make a re-presentation
clearer.

2' Christman (1991), p. 5.

16

desire to marry Jorge, or (b) she both acknowledges and positively evaluates it. Both are
unacceptable, Christman argues: (a) because she could acknowledge that she has a desire
that she came to have with no control of her own and which she would prefer not to have,
though she admits she does in fact have it. Her desire to marry Jorge, or to marry at all,
might well be of this sort. This mere acknowledgment sense of ‘identiﬁcation’ is
problematic, for if this is all there is to identiﬁcation, then “I can just as readily ‘identify’
with those non—autonomous aspects of myself as the more ‘authentic’ parts.”22
Christman’s claim is that merely acknowledging that one has a desire is not sufﬁcient for
it to be her desire. The other reading, (b), is problematic since then “I would have to be
in some sense perfect (in my own eyes)” if my desires could be said to my own.23 Thus a
person who reﬂectively takes stock of who she is would, improbably, discover no ﬂaws,
for any possible identity element with which she ﬁnds fault would not be an element of
her identity.

The second related difﬁculty Christman points out regards the conditions under
which one comes to form her desires. The desire to marry Jorge might be one that she
approves of, even in the stronger evaluative sense, though this in virtue of hi gher-order
desires that are the result of illegitimate external forces. Again, the marriage example is
apt: suppose she came to have the desire to marry because of social and familial
expectation to do so, and further suppose that she positively evaluates her desire to marry
because she has internalized the hi gher—order desire to agree with social norms and the

expectations of her family. Thus, if a person acquires hi gher~order desires through some

 

22 Christman (1991), p. 5. Emphases in original.
1‘ Christman (1991), p. 5.

17

form of manipulation, the hierarchical view still takes the ﬁrst-order desires of which the
person approves to be her own desires.

This is unacceptable to Christman, who claims, “the acts of identiﬁcation must
themselves be autonomous”.24 This leads to a third difﬁculty, which takes as its form a
dilemma: either the acts of identiﬁcation don’t need to be autonomous, or they are in fact
autonomous. If the acts of identiﬁcation don’t need to be autonomous, that is, if
identiﬁcation with a desire is sufﬁciently done just by the brute fact of approval of the
desire, then identiﬁcation can occur ab initio, without foundations. Christman ﬁnds it
implausible that a lower-order desire can be one’s own without the process by which it
came to be one’s own also itself being relevantly grounded in oneself. The other horn of
the dilemma is the possibility that the lower-order desires are internalized not ab initio
but are in some way autonomously acquired. If they are, then this sort of autonomy
might be the same sort of autonomy that is supposed to characterize actions — that is,
being the object of a hi gher-order desire. This possible fashion of desire-acquisition
introduces an irdinite regress into the account, though, for every desire must then be the
subject of a still hi gher-order desire. If the desires with which one identiﬁes are acquired
by some other autonomous method, then the account is incomplete, since it needs to
characterize this other sort of autonomy.

Thus, Christman faults the received model for both failing to give a plausible
account of what it is to identify with a desire and for allowing this identiﬁcation to occur
via a process that must, but cannot, be autonomous. The ﬁrst concern, regarding the

condition of identiﬁcation, is admittedly problematic on Dworkin’s description, or as

 

2" Christman ( 1991), p. 7.

18

Christman implies, underdescription, of the notion. At least for purposes of clariﬁcation,
Frankfurt’s richly described notion of identiﬁcation can serve as a remedy. When a
person takes a positive volitional attitude toward a desire that motivates him, he thereby
identiﬁes with it: that he wants to have a desire that he is in fact motivated by is sufﬁcient
for his identifying with it. “Whether a person identiﬁes with these passions,” Frankfurt
writes, “or whether they occur as alien forces that remain outside the boundaries of his
volitional identity, depends on what he himself want his will to be.”25

Often enough, though, honesty will require a person to acknowledge that a desire
she would rather not have is indeed one for which she must take responsibility; here the
attitude toward the desire is less volitional than cognitive. Having unsuccessfully resisted
some desire for a long time, one might give up “this equation of the real with the ideal”26
in the understanding that her ideal self is not something she is willing, or able, to attain.
Thus Frankfurt claims that the evaluation criterion for identiﬁcation is sufﬁcient but not
necessary.

Christman objects to this sort of identiﬁcation — mere acknowledgment — calling
its objects inauthentic, as noted above. But this sort of identiﬁcation does not always
entail inauthenticity. A person can wish she didn’t desire to run, for example, since it
takes time and energy and dedication that she’d rather spend on other pursuits. But
admitting that she does have this desire, and is not willing to do whatever it would take to
divest herself of it, it is thus her own desire. Lots of runners are like this, and similar

examples can be offered ad nauseam.

 

25 Frankfurt (1999), p. 137.
2" Frankfurt (1988), p. 63.

19

Nor is the criticism concerning desires acquired or identiﬁed with ab initio
particularly troubling. We make all sorts of things our own in this way: that I take up the
hobby of backcountry camping on a whim, that I come to appreciate the dark humor of
Edward Gorey though it is grounded in none of my other literary values, that through
sheer happenstance 1 have acquired possession of this wholly unlikable cat, say nothing at
all against whether these things are properly mine in the various relevant senses.

But more troubling is a person’s identifying with some desire in virtue of a
hi gher-order desire that is not itself one with which she similarly identiﬁes. Christman’s
presentation of the problem is peculiar, however, since only actions, and not desires, are
said to be autonomous. One reason for this is that desires are only very rarely (if ever)
voluntaristically directly brought about, while it is a deﬁning feature of my action that I
brought it about thus. So, while I might govern the conditions for the formation of a
desire (eating a favorite food while performing an unpleasant task so as to come to desire
the task/treat complex), I do so though actions. The worry Christman brings out, though,
is real: one’s hi gher-order desires ought to be her own in the same way (or, if not by the
same method, at least with similar results) as are her ﬁrst-order desires, those on which
she acts.

The most forceful version of the concern is this: a person can be said (on the
received model’s notion of identiﬁcation) to identify with some desire that she endorses
with hi gher—order desires that have themselves been manipulated. This might occur
through hypnosis, brainwashing or oppressive socialization. To address this concern, I’ll

turn to the discussions of it by Paul Benson and Susan Babbitt.

20

The adequacy of identity

Paul Benson has inﬂuentially addressed the oppressive effect of socialization on
women’s self-conceptions. He writes,

In leading women to internalize gravely mistaken conceptions of

themselves from a very early age, such socialization systematically

prevents many women from recognizing more adequate views of their real

strength and value; it renders them unable to take seriously reasons there

are for them to regard their appearance differently.27
Benson is making two distinct points here, one psychological and one normative. The
normative point is that the socialization of women is often oppressive, this partly in virtue
of its encouraging the acceptance of incorrect values. Women are exhorted to ﬁnd their
personal value in their physical appearance (of a certain virtually unattainable type) or in
submission to men. These preferences — the preference to be submissive, or to ﬁnd one’s
value solely in one’s own appearance — are normatively incorrect, Benson rightly
maintains.

I’ll bracket the normative claim in favor of ﬁrst addressing the psychological
claim. In internalizing mistaken self-conceptions, women are endorsing the desires they
have, to take one example, to act in sexually submissive ways.28 Such a desire can easily
be understood to be the product of socialization: advertisements portray submissive

women as typical or admirable, and these images are reinforced by everyday interactions

with men (and women) who expect such submission as a matter of course.

 

27 Benson (1991) p. 396.

1’ Though the subjects of the discussion are oppressively socialized women, Benson realizes both that men
are subject to similar sorts of socialization, and also that non-oppressive socialization works in much the
same fashion. The focus on women is based on the particularly oppressive nature of the socialization of
women.

21

It is relevant that not only does such a woman have a desire to act in these ways.
Many of us have, as a matter of socialization, desires to act in certain self-harmful ways:
to eat unhealthy snack foods, or to forgo a college education are both supported by
certain social and cultural expectations. But the systematic inculcation of femininity
notably (though not uniquely) involves the manipulation not only of ﬁrst-order desires,
but also of those of the second order. Women are instructed not only to possess a mere
urge — as the urge to eat snack foods — to be submissive, but further to identify with
submission; that is, they not only desire to be submissive, but also want that desire itself.

In this way, “feminine socialization has insinuated its lessons into their most
stable views of what they are and ought to be as persons.”29 It is this argued shortcoming
of Frankfurt’s structural account of persons that grounds Benson’s critique, ultimately
aimed at the implications for autonomy.

The charges that structural accounts are unable to ground autonomy will be taken
up below, in chapter 3, and so for the time being I will bracket this speciﬁc concern. But
the indictment that in certain sorts of situations the structural account of identity yields
false self-conceptions is a serious one, which I will now address. The indictment might
hold, too, but not in the sorts of situation Benson describes, and not, I think, in any
situation that is not so pathological that our theory of identity, indeed our theory of
anything, need be held to answer.

The situations Benson is justly concerned about are those very common ones in
which social norms are given uptake and endorsed by women whose betterment those

norms do not promote. This socialization takes place from many quarters — from the

 

2" Benson (1991), p. 388.

22

mass media, from friends, from family members, from persons in positions of authority —
and from a very early age.30

But one relevant strength of a structural account of identity is that it accurately
characterizes the fact that, virtually always, not all of the grounds on which one might
endorse her lower level preferences are manipulated (at least, not all toward the same
end). For this to be the case, all of the sources from which one acquires her norms,
beliefs and values would have to conspire to give the same message. On certain
interpretations of various historical eras, such homogeneity of cultural pronouncements is
said to have existed. I do not know whether such uniﬁed societies have ever existed; if
perhaps they have, ours is not now one of them. Oppression is indeed systematic, but this
does not entail that it is perfectly pervasive. It is because oppression is not perfectly
pervasive that structural views of identity explain how it is that some (whether few or
many) people escape certain of its damaging effects, as I’ll explain.

Consider just one example, mentioned above, of a particularly injurious effect of
oppressive socialization: a self-image disorder such as anorexia. Sometimes claimed to
be the only potentially fatal psychological condition, anorexia is widely believed to have
strong ties to social norms advocating low body weight as the standard for feminine
beauty. The fashion model images of the 1970’s were extremely thin, especially
compared to the mid-century’s so-called sweater girl models. But — and this is the point
that has gone unnoticed -- at the same point in history, alternative social norms became
available, advocating the unimportance of the physical in favor of mental, social, and

psychic growth. If hippies were thin, it was because they refrained from eating meat, not

 

3° Indeed, even prenatally: see Scheman (1997) on being “perinatally pinked”.

23

because they emulated Twiggy. While it is undeniable that oppressive social norms play
a causal role in self-image disorders; it is similarly undeniable that not all culturally-
mediated statements advocate (then or now) extreme thinness as an ideal for women.
Indeed, one of the deﬁning elements in the development of the disorder of anorexia is a
stage of resistance — resistance, that is, to the perceived conspiracy against being thin.31

Contra Benson, then, a structural view of identity illustrates a way in which many
women in fact escape some of these more damaging effects of oppressive socialization.
Even if a woman has a hi gher—order preference such that she endorses her desire to be
extremely thin, she has other hi gher-order preferences as well. From a theoretical
standpoint, then, structural views have the beneﬁt of descriptive accuracy, at least in the
ways just mentioned. From a more practical standpoint, structural views show how a
number of hi gher-order desires in concert could tip the scales against a person’s
oppressive upbringing; this is what I take to happen in such situations as organized
consciousness raising. This also occurs when a person enters into a new social group and
comes to substantially reorganize her commitments as a function of the support she
garners from membership in that group. On a hierarchical account, there is no need to
think that one or another higher-order preference (or even some set of them) will deﬁne a
person’s life in such an encompassing manner as Benson suggests.

To be sure, there are cases in which a person’s life is controlled in a totalizing
fashion. Such Orwellian society has been reported to have characterized some of the
concentration camps of the Second World War, for example. In such circumstances, the

only self-conception possible (or, at least, perceived to be possible) is one to which there

 

3' Levenkron (2000).

24

 

is no alternative.32 Against such atrocities, though, we should not expect some certain
account of personal identity to defend.

Thus a structural account of identity does have resources to address the ways in
which harmful socialization exploits the process of identiﬁcation through hi gher-order
endorsement of desires. But still unanswered is the normative concern: the sort of
identity that would acceptably ground one’s actions must, the claim runs, be
characterized by an account which has an essential normative element. Thus any
structural account of identity, such as Frankfurt’s hierarchical account, is unsatisfactory
insofar as it is consistent with a person’s identifying with anything, no matter how
normatively undesirable this object of identiﬁcation.

Benson’s repeated references to ‘the reasons that there are’, to which a
satisfactory identity is supposed to lead its bearer, are echoed by the claims of Susan
Babbitt. Babbitt argues that structural accounts of identity are unable to capture the sense
of self that is adequate for autonomy, integrity and rationality: one that aims at one’s own
best interests, interests that may not be the goal of the desires that one happens to have.

She accuses as being “too simple” a view quite similar to Frankfurt’s, at any rate a
structural view: the “popular conception of individuals according to which the
autonomous individual is one who possesses the capacity to reﬂect on her preferences
and alter them in the light of information.”33 It is not just this relation that one has to her
preferences that constitutes identity, Babbitt argues, but rather it is the content of those
preferences that makes for an adequate identity. And though only explicitly demanding

that the adequacy of identity be an adequacy for the exercise of autonomy, it is clear

 

’2 Some survivors of such extreme violence report their self—conception having been reduced to that of a
mere physically existing being (Brison, 2002, chapter 3).
’3 Babbitt (1996). p. 103.

25

enough that Babbitt’s target is much more broad. She takes to be “equivalent” a number
of human capacities that seem, to this reader at least, to sum to personhood: personal
integrity, autonomy, rationality, identity, and even, minimally, a self.“ Further, without
an adequate identity, Babbitt argues that goods such as self—respect and dignity are
unimaginable and in fact impossible.”

Her proposed solution to these inabilities of certain persons is to require that one’s
sense of self be one that directs her to realize her own best interests. This is as much a
social as a theoretic demand. The social environment must be altered so that it allows
persons with inadequate identities the social resources needed to improve their self-
conceptions. Options for actions that make possible dignity, integrity, et a1, must be
available and be perceived to be available; minimally, such actions must be imaginable,
for it is the lack of possibilities not only in the world but especially in the range of
possible thought that makes for the capacities that concern Babbitt.

This much — that social changes should be promoted that open the range of
opportunities so that self-respect and the goods that accompany it are available to all — is
indisputable. The theoretic dimension of Babbitt’s demand, though, is less tenable. Part
of the problem here stems from the intended reconceptualizations of the central ideas:
‘identity’, ‘integrity’, ‘autonomy’, and ‘rationality’.36 The somewhat conﬂationary
redeﬁnitions make it difﬁcult, unfortunately, to get a clear grasp of the theoretically

intended suggestions. For example, “the term “personal integrity” is taken to refer to the

 

3“ Babbitt (1996), pp. 6 (on rationality), 8 (on autonomy), 43 (on the self), and 105 (on integrity and
identity).

35 Babbitt (1996), p. 8.

3" As pointed out in Walker (1998).

26

”’7; an adequate identity, in turn, just is that identity which leads one

adequacy of identity
to pursue her own real interests, not merely the ones she perceives herself to have. But
this is just autonomy: “a special kind of agency, the capacity to take control of one’s life
and to realize one’s best interests?”8 And each of these deﬁnitions is hardly
distinguishable from that of ‘rationality’, which is a matter of “whether a person’s
choices and actions constitute the best means to her most important ends.”39

There is a practical aim that comes through more clearly, though: Babbitt intends
a central normative element to be present in each of these concepts. This is quite
contrary to structural views, on which sufﬁciency is judged by the degree of uniﬁcation
of the constitutive elements whatever their content. Babbitt’s account thus leaves no
room for judging any of the central notions (identity, autonomy, integrity, rationality) on
structural grounds. For each of these notions, unity is just as preferable as disunity, so
long as one’s best interests are promoted.

Depending on just what we take the philosophical purpose of an account of
identity to be, though, it is not clear that Babbitt’s normatively-constituted view is
preferable. Babbitt’s motivation for making a certain content a requirement of a plausible
account of identity is that she wants to explain how identity leads a person to pursue her
own best interests. This is ﬁne, to a point; since the notion of identity at hand is one of

practical identities, if a person’s actions lead to her own well-being this will be, in part, a

matter of the content of her identity.

 

 

’7 Babbitt (1996), p. 104.

3" Babbitt (1996), p. 45. Interestingly, ‘control’, being “the realization of real interests” (ibid.), has the
same normative component shared by the other concepts at hand.

3" Babbitt (1996), p. 6.

27

But someone’s identity can be her identity without it grounding her own well-
being. Thomas Hill’s deferential wife, who through oppressive socialization has come to
completely identify with, and even gain pleasure from fulﬁlling, her subservient role in
relation to her husband, surely does not aim at her own well-being. She is worse off for
who she is, and would be happier and more self-respecting had she an identity that did
not cast her as essentially and completely subservient. But the (hypothetical) fact of the
matter is that this improved identity is not hers. Would that it were. She is subservient,
on her own sincere appraisal as well as judging from the subservient actions she regularly
engagesin.

Nor does the fact that one’s identity can fail to be self-beneﬁcial change in light
of the fact that identities are socially constructed, that ﬁrst-personal as well as third-
personal forces affect identity. Indeed, it is sometimes because one’s identity is not
purely self-constituted that a necessary normative element is implausible: it would be
inaccurate to deny that someone - a black slave in the antebellum South; an elderly
person assumed by her family to be incapable of making her own decisions because of
her age —- whose identity prevents her from promoting her own well-being is not in fact
the possessor of just this identity. Because of the social environment, the slave is a slave;
similarly the elderly woman is unable to make her own decisions. To say that these
identities do not belong to the people to whom they are attributed is to deny the
sometimes unjust power that others have in the creation of identity.

Attributing to Babbitt the claim that a person’s identity can fail to be her own is

not entirely easy to support textually. Babbitt’s repeated phrasing of the problem of

identity as one of the adequacy of identity can often be interpreted as either or both the

28

adequacy of accuracy or that of self-interest; her claim that an identity can be “wrong’”0
might be epistemic or moral. But one discussion of the Deferential Wife and why exactly
she is “falling short of her interests” involves a distinction between her actions being “a
result of mistaken beliefs about her self” and their being “the realization of her actual
self 1‘” In being subservient, the Deferential Wife acts on an inaccurate identity, one that
does not correspond to her actual self. Another similar discussion contains the claim that
the Deferential Wife is “improperly self-deﬁned”, suggesting, like other improper
deﬁnitions, not moral impropriety but epistemic inaccuracy. In perhaps the clearest claim
Babbitt makes on the matter, she discusses philosophers who “have not attempted to
answer the question. . .of what sort of person one should be — the more robust sense of
identity that is at issue when we aim at self-understanding.”42 If the aim of self-
understanding is not the aim of understanding who one in fact is, but rather who one
should be — if self-understanding takes as its object someone who one currently is not --
then it seems one’s identity can fail actually to be her own.

Now, it may be that, as in the case of the slave, the elderly woman and the
deferential wife, the identity that a person has is (simply) not adequate to lead to self-
beneﬁting actions. Even if we assume that this fully characterizes Babbitt’s point here,
this is not itself a fault of an account of identity so much as a problem with the speciﬁc
identities that persons possess. An account of identity ought accurately to characterize
the sort of identities that people do in fact have. With this even Babbitt agrees, arguing
that it is only her own normative account of identity that accurately accounts for a

person’s best interests (those she actually has), rather than her perceived interests (ones

4° Babbitt (1996). p. 30.
4' Babbitt (1986), p. 44. Emphasis added.
42 Babbitt (1986), p. 106. Emphasis added.

29

she only believes she has). And the fact of the matter — that some people (perhaps many
people; perhaps all of us) have identities that are inadequate — is not only captured by a
structural account but, as the present work has promised to take up, allows for an ideal
that promotes two of Babbitt’s concerns, autonomy and integrity.
OSSIﬁcation

For some, ‘unity’ and ‘rigidity’ are nearly synonymous, a uniﬁed thing being
inevitably rigid. A uniﬁed identity is assumed to be one that is inﬂexible, ossiﬁed into
one rigid unalterable structure. So rigid is such an identity, this assumption continues,
that it becomes homogeneous: lacking in internal diversity, each constitutive element has
become so wedded to its co-constituents that each is indistinguishable. Likely a visual
representation of such an identity would be an unbroken ﬁeld of dull gray; the
implication is that a life founded on such an identity would be the same dreary hue.“3

Two philosophers have taken up this charge against the ideal of unity. Diana
Meyers’s critique focuses on two elements of Frankfurt’s account: the notion of
wholehearted endorsement that is peculiar to his particular view, and the rank ordering
that is a fundamental component of a hierarchical view. Amy Mullin, offering more of a
warning than a criticism, is also concerned with the possible association of unity and
homogeneity. I will explain Mullin’s warnings and agree that they are apt, but ﬁrst I will
answer Meyers’s criticisms.

Meyers pits her reading of Frankfurt against the fact of intersectional identity,

ﬁnding Frankfurt’s account to inaccurately characterize the authentic self. Intersectional

identity, on Meyers’s usefully broad characterization, is “an identity drawn from diverse

 

‘8 As in Meyers’s (2000) description of “the autonomous individual as an isolated, plodding planner,
wholly lacking in panache and merriment...” (p. 168)

30

sources that may give rise to conﬂicting desires and rival allegiances?”4 Identity elements
from these diverse sources — race, gender, economic class, sexual identity, employment
roles, and so on — can intersect in distinctively harmful ways. For example, one harm of
racial oppression is that African-Americans are expected to pull themselves up by their
bootstraps and take the initiative to improve their collective situation, while people in
lower socioeconomic classes are expected to be humble in the face of those who hold
societal power; the conﬂicting identity of a poor black person should be apparent. It is
the authentic self of persons possessing this conﬂicting sort of identity that Meyers
claims Frankfurt’s view fails accurately to account for.

Meyers’s notion of the authentic self has been developed and somewhat altered
over the course of her writings, but nowhere does she give it an explicit and complete
deﬁnition. In a work earlier than the one critical of Frankfurt, she characterizes the
authentic self as a “repertory of skills” in conjunction with “a collocation of attributes”45
which are not merely a static set but are “dynamic“. Unsurprisingly synonymous with
possession of “the true self’“, to have an authentic self is incompatible with “regard[ing]
one’s self as alien’m’. In her later work, the repertory of skills is cast as a necessary
precondition of, but not a constitutive part of, the authentic self.49 Meyers’s earlier notion
of the integrated personality, argued to be a necessary facet of the autonomous person,

plays no role in her discussion of intersectional identity, consistent with expectations.

\

44 Meyers (2000), p. 154.
‘5 Meyers (1989), p. 92.
4" Neyers (1989), p. 61.
‘7 Neyers (1989), p. 20.
“8 Meyers (1989), p. 73.
“9 Meyers (2000), p. 166.

31

She gives a clearer deﬁnition of what she takes Frankfurt’s view on the authentic
self to be.50 Frankfurt argues that when a person wholeheartedly identiﬁes with a desire,
she identiﬁes with it in such a way that she cuts off any ambiguity or further internal
conﬂict regarding the desire. This identiﬁcation is done by excluding from one’s self the
desires she does not want to satisfy, and by establishing a rank ordering of the retained
desires that establishes one’s priorities. Meyers concludes that on Frankfurt’s view “to
have an authentic self is to be wholeheartedly committed to a rank ordering of the desires
one has decided to satisfy and to be wholeheartedly dissociated from those of one’s

”51 Thus the authentic self is an integrated

desires that one has decided against satisfying.
self.

Meyers points out that persons with intersectional identities do not ﬁt this model.
On a broad but plausible view, the set of possessors of intersectional identities includes
virtually everyone. We are each the possessor of multiple identities, and whether the
notion of intersectionality essentially includes the sort of notable conﬂict and interactive
harm as in Crenshaw’s crash metaphor of intersectionality, many pe0ple do exemplify
intersectionality.52

The lack of ﬁt is both descriptive and normative, Meyers contends. Frankfurt’s
view fails to accurately characterize the authentic self that intersectional subjects do

possess. It is in the interest of resistance to oppression that intersectional subjects do

Come to an understanding of their authentic self — that is, that they are indeed

\

5° The following paragraph is from Meyers (2000), p. 169.

5' Neyers (2000), p. 169.

52 The crash metaphor is a view of intersectionality that focuses on both the most harmful elements of
ide mity — race and gender insofar as persons are oppressed vis-a-vis them - and the “sinister synergy” that
characterizes their interaction: since intersectional persons are harmed in virtue of multiple elements acting
in concert, no elements are blamed individually (Meyers 2000, p. 154).

32

intersectionally identiﬁed, and to know this is to possess knowledge of one’s authentic
self. This knowledge, in turn, can motivate resistance to oppression: when a person
comes to realize that she is not only a woman (a subordinate identity) but also white (a
dominant identity that few white people are consciously aware of), she thereby gains
resources with which to resist her subordination.

But the view of the authentic self as an integrated self is in direct contradiction to
this fact, as made clear by considering such non-integrated identities. Normatively, a
view holding as ideal an integrated self is unattainable by many persons, and is in any
event not as ideal for multiply oppressed persons as it is for others; Meyers notes that for
someone to identify wholeheartedly with “her victimization” would be foolhardy, as
would wholeheartedly disidentifying with it, thus foreclosing the possibility of
resistance.” Ambivalence, in some cases, is the proper attitude to take up.

In summary, Meyers faults Frankfurt’s view for what she sees as an ideal of unity
that above all is unresponsive to the reality of intersectional subjects, who are in fact
internally divided and also stand to be harmed by taking wholehearted attitudes toward
their identities.

There are a number of elements of Meyers’s critique that put a great deal of
pressure on Frankfurt’s account. The observation that intersectional subjects are
motivated to resistance in virtue of self-knowledge of their own internal division should

give pause to the proponent of unity. But one might wonder if the two accounts are so
divergent on this matter; indeed, Meyers preempts the defense:

Now, it might seem that Frankfurt’s idea of identiﬁcation and integration
gives these individuals a technique for coping with this predicament.

 

53 Meyers (2000). P. 169.

33

They can disidentify with the harmful attributes they have internalized, for

example, self-doubt, servility, or submissiveness, and they can identify

with the empowering attributes they have internalized, for example, a

mordant sense of humor, persistence in the face of adversity, or a love of

convivial gatherings.54
This, though, would lead to the damage noted above: to wholly deny or to wholly
embrace, for example, one’s servility, is either to accept it and thus fail to resist it, or to
deny it with the same effect.

There are at least three misleading inaccuracies in Meyers’s reading of Frankfurt.
The ﬁrst is remediable, but still worth noting. It is not at all apparent in what way
Frankfurt’s account would absorb the attributes, such as servility, that Meyers is
concerned with. On Frankfurt’s view the basic constituents of identity are desires.
Desires with which a person identiﬁes are associated with the things she cares about —
principles, ideals, political causes, other persons, her own projects — things she identiﬁes
with. Thus, unless the case involves a person who cares about being servile — that is, like
Hill’s Deferential Wife has a desire to be servile that she wants to persist — Frankfurt’s
account is simply silent regarding such attributes. Meyers’s misreading, then, is the
simple substitution of ‘attributes’ for ‘desires’ in Frankfurt’s account.

But on this point, Meyers’s characterization of identity, as including more than
just desires and objects of care, is more accurate than Frankfurt’s. While a person may
sometimes feel like nothing more than a bundle of desires, identity is more broadly

55

constituted, and includes, inter alia, attributes such as servility. While a person could

 

5‘ Meyers (2000), p. 170.

55 It could be argued that such attributes reduce to or ground in desires; ‘servility’ is just a term to
characterize a certain set of desires, and reference to these desires is ineliminable when explaining the
volitional, practical nature of personal identity.

34

have a desire to be servile, this is distinct from the notion of servility as an attribute.
What is the relation between the desire and the attribute?

The relation hinges on the second distinction. For Meyers, to identify with
something is to embrace it, to disidentify is to eschew.56 These suggestive terms (the only
characterization of ‘to identify’ that Meyers offers) obscure the difference that can be
drawn between desires and attributes.

To identify with a desire, Frankfurt has claimed, is sometimes to want to be
motivated by it.57 The simple (and simplistic, as this notion of identiﬁcation becomes
more rich in his later writings) distinction here is from those desires one does not want to
be motivated by. A person disidentiﬁes with her desire for a sports car if she does not
want to be motivated by the desire she has for such a car. She may realize that she has
the desire only because she saw an advertisement and does not want to be swayed by
advertising, or because her friend recently bought a sports car and she realizes her desire
is mere envy. She, in what amounts to a process of self-constitution, makes unwanted
desires external to her, and makes a desire with which she identiﬁes internal to her. The
desire with which she identiﬁes becomes not a force moving her, but she moving
herself.58

In later works the notion of identiﬁcation, and especially the radical freedom of
self—constitution the earlier notion included, Frankfurt admitted to be less a matter of

approval than of acceptance. One who identiﬁes with a desire not only wants it but also

“takes responsibility for the fact of having the desire”59 If our advertising victim ﬁnds

¥

:Meyers (2000), p. 170.
:Frankfurt(l988), p. 18; (1999), p. 105.
:,Frankfurt (1988), p. 170.
59,Frankfurt (1988), p. 170.

35

herself unwilling to distance herself from the desire for a sports car she may (again, if she
is honest with herself and others) admit that the desire is hers. The desire may be so
trivial that it does not merit the energy needed to expunge it (some of us have a strong
desire to put on our shoes either left-ﬁrst or right-ﬁrst, for example).

And about all of this, the person is satisﬁed: lacking the intention, desire, or
motivation to identify differently, one who is satisﬁed with her identity (or with certain
elements of it) is in “a state constituted just by the absence of any tendency or inclination
to alter its condition.”"0

On this view of the process of identiﬁcation, there appear to be more nuanced
options than either embracing or eschewing an attribute. A person who has internalized
the attribute of servility, on this view, does accept that she is motivated by this attribute;
she takes responsibility for her servility, not because she imposed it on herself and not
because she embraces it. She may identify with it as a matter of self-honesty, and with a
lack of resources to change this element of her identity.

On Meyers’s more hopeful narrative, the servile woman is imagined to possess
resources, other identity elements, in the light of which she might disidentify herself from
the attribute of servility. Because the woman identiﬁes with her persistence in the face of
adversity, she need not be satisﬁed with her servility. But this is not, as Meyers assumes,
to deny that the woman has the attribute of servility forced upon her. What Frankfurt’s
notion of identiﬁcation accurately shows about this sort of case is that a person can be
affected by something - the attribute of servility, which no matter how persistent the

woman, she will still be affected by to some degree - without her identifying with it.

‘

60 Frankfurt (1999), p. 104.

36

Thus a structural account need not lead to the sort of ossiﬁcation that Meyers
claims, for a number of reasons that we have just seen: it is not the case that Frankfurt’s
view entails that the authentic self is a uniﬁed self, nor are the options for a person’s
stance toward various elements of her identity exhausted by either complete acceptance
or total denial. Further, as I will now explain, any worry about the ossiﬁcation of identity
is not unique to a structural view.

While ‘ossiﬁcation’ is not a philosophical term of art (and so has no speciﬁcally
relevant deﬁnition), it does suggest an array of synonyms: ‘crystallization’,
‘immobilization’, ‘rigidity’, and ‘inﬂexibility’. Something that is ossiﬁed has all of its
constituent parts locked together in a certain unchanging, unchangeable pattern, an
assembly that might be thought to break under force, rather than bend. To become
ossiﬁed is to lose one’s resilience.

To be sure, a structurally integrated identity would lend its possessor a certain
degree of steadfastness in her actions and projects: she tends to know what she wants, and
how much she wants it in relation to other things she wants. She is not apt to be
ambivalent, and so will not be easily swayed from a decision to pursue a goal with which
she wholeheartedly identiﬁes.

Suppose a person identiﬁes with a desire to eat organic food. Suppose further that
the desire is one with which she identiﬁes wholeheartedly; this desire coheres with her
other desires — to attend to her health, to minimize the amount of chemicals produced, to
shop at that little co-op down the street that her friend is an owner of — and she

Wholeheartedly identiﬁes with no conﬂicting desires, such as the desire to be frugal. But

37

after she and her partner adopt a child, she ﬁnds (contrary to her earlier calculations) that
she cannot afford to feed herself and her child with organic food.

Now she may continue to identify with the desire to eat organic, even though she
cannot act on the desire. This would likely cause her a fair amount of distress, and so she
might well disidentify with the desire by a process of internal restructuring.“ Lower on
the rank-ordered scale would go her desire to attend to her health (she knew she’d get less
sleep and exercise while raising a child, anyway) to make room for the whole set of
desires that accompanies caring for a child.

This example is of a familiar sort. A person has all of her priorities organized,
and a change in the world — even a positive change — spurs a reorganization in response.
A structural account of identity, further, explains why a person presented with an external
change undertakes internal changes. Desires carefully ordered may be free of conﬂict,
but such a set is always subject the addition of desires caused by changes in the world.62

Desires may atrophy, too, for reasons external to their possessor, and this is well
characterized by the negative notion of satisfaction described earlier. To be satisﬁed with
one’s desires is not to experience some positive feeling (like the satisfaction one gets
from eating a favorite food). If this were the notion of satisfaction in play, it would be
unclear how a desire might fall out of one’s internal economy; the lack would be

noticeable, and even painful. But satisfaction deﬁned as the lack of internal conﬂict is

R

6' She could, and we sometimes do, continue to identify with desires that she knows cannot be satisﬁed. To
do this to a large extent, though, limits the practical nature of one’s identity; one who identiﬁed only with
impossible-to-pursue desires would be unable to act at all.

62 This is more consistent with Frankfurt’s actual claim regarding ‘rank ordering’. He explains that when
“the issue is which desire to satisfy ﬁrst [. ..| the competing desires are integrated into a single ordering”
(1988, p. 170, author’s italics), a set that would need to be reordered when a novel desire presents itself at
some point in the prioritization.

38

consistent with the loss of desires with which one was satisﬁed, so long as the loss of the
desire does not itself bring about a conﬂict.

Through the foregoing discussion it can be seen that the unity of identity can be
characterized in such a way that is attentive to the practical contingencies of the lives of
real persons. This is due in large part to the practical criteria on which structural unity is
based. The manner in which desires are moved around is not voluntaristic, at least not
primarily so; thus even to acquire, expunge or reorder desires involves a practical move,
not merely a mental one. The conﬂict on which unity itself is judged is often a practical
one as well.

Amy Mullin’s warning that “it is important to avoid assuming that effectively
uniﬁed selves must be homogeneous or integrated to the point that harmony is rarely
threatened,”"3 which she takes proponents of unity to have done, is thus a useful one.
That the ideal sort of uniﬁed self is an eﬁectively uniﬁed one, rather than absolutely

uniﬁed, is consistent with the present project.

Unity, oppression, resistance
Cui bono fuerit? (Whom did it beneﬁt?)

- Longinus Cassius

Tracing the steps the present chapter has taken shows that the concerns have been
primarily, though not entirely, theoretical. The notion of ‘identiﬁcation’ itself and the
related hierarchical process by which one comes to identify were questioned for their

conceptual plausibility. The suggestion that structural accounts support ossiﬁcation was

 

‘53 Mullin (1995), p. 3.

39

made in the interest of alternative philosophical theory. Paul Benson’s appeal to
oppressive socialization is best taken to be predominantly a counterexample used to
illustrate theoretical problems within hierarchical accounts. One with an interest in the
social effects of the ideal of unity and who agrees with Babbitt’s practical demands might
wish them to be grounded in less opaque conceptualizations.

That philosophical discussions of personal identity have, like the sorts outlined
above, been largely ignorant of the practical effects of our conceptualizations is clear.
But the matter of personal identity is much more than conceptual; the way in which
identity is characterized has real effects on the persons it purports to describe.

That the ideal of unity is just that — an ideal — straightforwardly suggests that
those who fail to attain it have fallen short. What is more, since the ideal of unity is an
ideal not so much about behavior (as, say, a utilitarian sort of moral ideal) as it is about
the sort of person one ought to be, those who fail to achieve or even to pursue the ideal
may be thought of as personally deﬁcient, even defective.

The ideal of the unity of identity does appear to be an ideology, in the pejorative
sense of the term — what Charles Mills characterizes as theory with a “reliance on
idealization to the exclusion, or at least marginalization, of the actual.”"4 Meyers’s
reference to intersectional identity was made in order to point out that real people have
disuniﬁed identities, and often face great practical barriers to altering that state. An
account advocating identity uniﬁcation must be responsive to these facts so as not to

marginalize those who do not attain the ideal advocated by the theory.

 

6“ Mills (2004). p. 166.

40

Even if the ideal of unity does not ignore the actual, it may further oppression by
contributing to the advantage of the dominant, in this way. If it is true that the ideal is
virtually unattainable by the socially subordinate, then the ideal will, to borrow another of
Mills’s characterizations of ‘ideology’, “reﬂect and contribute to perpetuate illicit group
privilege”"5 insofar as only the dominant will be able to attain the goods that are
associated with reaching the ideal. Who beneﬁts? The members of dominant groups.

Maria Lugones has, in advancing similar criticisms, provided a set of concepts
under which different sorts of unity can be understood. I will adopt her nomenclature,
and in describing it, draw out her associated criticisms and the responses that can be
offered.

Lugones argues that each human being — she uses the term ‘incarnate being’ — is
composed of multiple persons. This admittedly (and I think excusably) “ontologically
problematic”66 view ﬂows from a phenomenology peculiar to the oppressed. She appeals
to “examples of people who are very familiar with experiencing themselves as more than
one: having desires, character, and personality traits that are different in one reality than
in the other,”67 people who are not just of two minds about an issue, or who tend to be
more understanding with their children than with their colleagues (or vice versa, as the
case may be). The oppressed ﬁnd themselves in different realities, different
““worlds””"8. In one “world” such a person may be playful (to use Lugones’s example),
while in another it is impossible for her to be so. She is two persons: one playful, one

not.

 

65 Mills (2004), p. 164.

6" Lugones (2003), p. 89.

67 Lugones (2003), p. 57.

‘8 Lugones (2003), ch 4; internal quotation marks Lugones’s.

41

That these two situations are different realities and not merely different contexts
or environments is a feature not just of the altered subject who inhabits it. That the
characteristics, abilities and desires are different in different “worlds” is a function of the
different interpretations that are given, not only by the subject herself, in the various
“worlds”. “If one can remember the intentions of the person one is in the other world and
tries to enact them in the other, one can see that many times one cannot do so because the
action does not have any meaning or has a very different sort of meaning that the one it
has in the other reality.”"9

Regrettably familiar examples illustrate this claim well. A black man who intends
friendliness is interpreted as friendly in some realities, but in other realities his actions
(because of the interpretations given them by others) take on an aggressive meaning. A
woman’s ‘No’ in one “world” is deﬁant, while in another it is insubordinate, and in yet
another, coquettish.

An incarnate being leading a life (of a very common sort) marked by such shifts
in meanings and self—characterization is multiplicitous, on Lugones’s view. One of
Lugones’s central projects is to depict the ways in which such multiplicitous subjects can
resist the fragmentation that is alleged to characterize them. Thus Lugones’s target is the
ideal of unity: arguing that such internal discontinuities undermine one’s ability for
agentic action, or one’s bare personhood, or some other ability or status, is tantamount to
ﬁagmenting such persons. Such fragmentation — simply by making attributions of
fragmentation — that is the interpretation of multiplicity made by members of the

dominant group is itself oppressive. Thus, “the negotiations of the relation between

 

‘9 Lugones (2003), p. 57.

42

fragmentation and multiplicity by resistant subjects become central to an understanding
of resistant subjectivity.”70

Lugones’s recommendations for resistance introduce another valuable concept,
ambiguity. Actions can be intended ambiguously: instead of intending either a resistant
or an oppressed action — that is, acting from either a resistant or oppressed identity — one
can issue an action from both, intending that the action be interpretable as either resistant
or oppressed or both." For example, a black housekeeper might fail to arm the security
system, an action that she intends to be both oppressed (as when her employer attributes
the failure to her perceived ineptitude) and resistant (since it indicates disdain for her
employer’s possessions). A person can also (in a distinct, though still ambiguous, sense)
ambiguously inhabit an identity that has been externally imposed. To do so is to “block
identiﬁcation” with an oppressed identity, while still acknowledging it; Lugones does not
intimate that such an identity can be cast aside simply by refusing to acknowledge it, but
does offer this sort of ambiguity as a defense against bald uptake of oppressive
identities.72

As an example illustrating the notion in his second sense of ambiguously
inhabiting a certain identity, consider a lesbian couple who, on being publicly slurred for
their sexuality, engage each other in an openly passionate kiss.73 The act is oppressed:

after being accused of being lesbians (which is often what slurs partially entail), they

prove it, so to speak. It is also an act of resistance, openly acting in a way that is

 

7" Lugones (2003), p. 32.

7' Lugones (2003), p. 13.

72 Lugones (2003), p. 74.

73 This example is borrowed from Mason (2002), who discusses the conditions under which such acts of
resistance can be intended as such.

43

proscribed. In intending ambiguously, these women can inhabit an identity while not
giving it, and especially its oppressive aspects, full uptake.

It is the ﬁrst sense of ‘ambiguity’ that is especially interesting for the present
project. This sense of the concept requires a slightly different example, since it regards
ambiguity between, rather than within, identities. 74 A woman who is both a mother and a
wife might ambiguously intend her refusal to tolerate her son’s and husband’s collective
disrespectful sloth. As a mother, the act is authoritative; as a wife, it is insubordinate. In
intending ambiguously she intends the act to be interpreted in both ways, and the act
ﬂows from both of her identities.

Though this tactic would likely take substantial psychic resources, and likely
carry the same risks as any other resistant actions, it is an important observation that an
action need not be, to use another of Lugones’s terms, “monosensical”.75 The notion of
ambiguity explains how multiplicity can be retained without entailing fragmentation, the
fragmentation that is supposed to follow from internal diversity. Ambiguous intending
makes it possible for one to retain her moral and agentic standing in the face of
“mainstream, institutionalized morality” that presupposes the uniﬁcation of the self, since
she acts in a way that makes sense to such a morality, but also retains her integrity. 76 And
though resistant actions, even those ambiguously intended, open one up to substantial
risks, the motivation to perform such socially and personally valuable actions might well

be found in ambiguous intending.

 

7" Lugones’s own example of an act intended as both incompetence and sabotage (p. 13) is inadequate to
fully illustrate the notion of intent, since incompetence is more clearly a matter of ability rather than intent.
75 Lugones (2003), p. 7.

7" Lugones (2003), p. 14.

The notion of ambiguity, then, is valuable, as is its practice. Further, ambiguity is
not inconsistent with the ideal of unity. The notion of unity as a practical rather than
metaphysical unity is able to take up this element of multiplicity on which one brings her
internal economy into some sort of order through a harmony of action. Recall the lesbian
couple in the earlier example - each has conﬂicting reasons for action (to express herself
through a display of affection, as well as to refrain from doing so). Lugones advocates a
practical reconciliation of this conﬂict by intending ambiguously, a reconciliation that
entails a practical uniﬁcation: whatever the state of one’s identity, the actions that issue
from it are not in practical conﬂict.

The agreement between ambiguity and unity may be the result of a
misrepresentation or inaccurately narrow characterization of unity. At times (though
rarely) Lugones appears to take a speciﬁcally Kantian conception to task. “Reason,
including its normative aspect, is the uniﬁed subject. It is what characterizes the subject
as a unity.”77 But this ‘reason’, especially its normative aspect, is a practical notion:
reason — practical reason — is said to unify the subject so that she can act. This is the
same accomplishment as that of ambiguous intending: the practical (not metaphysical)
uniﬁcation of an agent. If practical uniﬁcation is problematic, ambiguous intending can
hardly be an acceptable alternative.

It is in the recurring claims about the fragmented subject’s supposed
“inconsequential or attenuated sense of agency”, without the appeal to ‘Reason’, that
Lugones suggests her target to be perhaps a more contemporary defender of unity, such

as Frankfurt, Dworkin, and even Benson and Chri stman, who each takes as central to the

 

7’ Lugones (2003), p. 129.

45

exercise of autonomy some variant of unity.78 Much the same criticism holds here: that
contemporary defenders of structurally-based accounts of autonomy advocate a unity that
is not necessarily based in rationality does not distinguish their projects from Lugones’s
as regards uniﬁcation in action.

The sort of uniﬁcation at which Lugones might ultimately be directed is the
supposed uniﬁcation of groups. “If the person is fragmented, it is because the society is
itself fragmented into groups that are pure, homogeneous.”79 Since group identity is an
important constituent of one’s overall identity, the sort of structure that characterizes
these groups might be thought to mirror the internal structure of those who gain their
identities from membership in these groups. As Amy Mullin writes, the fact that “. . .the
appeal of unity is used to eliminate the disadvantaged or marginal suggests another
reason why we should be uneasy about attempts to rid ourselves of inner differences or
conﬂicts.”80 The same sort of reasoning on which recalcitrant elements of a social group
are excluded, the worry runs, functions to exclude certain elements of a person from
herself.

Even if assumptions of group essentialism were the only or even the primary
cause of personal fragmentation, it is not at all clear how ambiguous intending could
constitute a remedy. To intend that one’s actions ﬂow from, and be interpretable in the
light of, more than one identity - identities that are each supposed to be homogeneous —
does nothing to undermine or even deny the purported homogeneity. The sharp

distinction between oppressive and resistant intendings that is a feature of ambiguity can

X

: Lugones (2003), p. 5.
so Lugones (2003), p. 141.
Mullin (1995). p. 15.

even be seen to reinforce one’s self-conception as a member of two essentially and
radically different groups.

It seems, then, that the weak point in Lu gones’s critique of the ideal of unity is the
imprecise characterization of ‘unity’. Though this may make the account of unity that is
being defended here perhaps too resilient, it is clear that at the heart of an account of an
identity that motivates action, enables one to act with integrity and to love — all practices
that require subtle and varied abilities, sensitivity to contingencies, and an appreciation of
the complexities both of the world and especially of the persons who inhabit it — one will
ﬁnd multiplicity without fragmentation.

If this is indeed the goal of both those who lack unity as well as those who are
better placed to realize it, that is, the goal of both the subordinate and the dominant, then
the answer to Cassius’s question is ‘everyone’, at least everyone with a direct and
immediate interest in these goals.8| In this way will the account of unity presented in the
next chapter and defended in those following be attentive to the oppressive possibilities

of marginalizing conceptualizations.

 

3‘ That is, I don’t take the ideal of unity to be relevant to the severely cognitively disabled, for example.
That this excludes a certain (perhaps large) set of people is true but unavoidable, since there are vanishingly
few substantial claims that are true of every human person.

47

_2_

The Unity of Personal Identity

The criticisms of the ideal of unity addressed in the last chapter were criticisms of
the unity of identity as conceived quite generally. That is to say, opponents of the ideal
of unity tend to take up arms against an underdescribed adversary. Having shown that
the criticisms of unity (conceived of generally) can be answered, the next task is to
outline the speciﬁc sort of unity of identity to be defended.

The unity of personal identity is a twofold notion, composed, of course, of
‘personal identity’ and ‘unity’. I will address these two concepts ﬁrst separately, and
then draw from these separate analyses in outlining the speciﬁc form of the ideal of the
unity of personal identity that I have in mind. I will explain — and defend as plausible —
various points of compatibility between views of identity including narrative, group
identiﬁcation, self-conception, and desire-endorsement views. That is, certain of the
issues on which competing views do not conﬂict (whether explicitly or implicitly) are, for
the present purposes at least, the most philosophically interesting issues (rather than, as is
usual, the points of contention). I will not here construct a theory of personal identity de
novo, though I do maintain that the views of personal identity on offer sum to a credible
if understandably incomplete account.

To look ahead: the main point of agreement between the accounts is as follows.
All the views of personal identity that will be discussed (which I take to be illustrative of

the range of major approaches), on a plausible reading of each, accommodate Harry

Frankfurt’s notion of endorsement as underlying the relation 3 person has to elements of

 

her identity. Just what counts as an identity element is what distinguishes the various
views that I will defend as compatible and complementary: narratives, desires, group
membership, practical self-conceptions are each elements of a person’s identity. Roughly,
for a person to endorse some element with which she identiﬁes is for her to be satisﬁed
with it, that is, for the element to be part of her psychic and practical life without conﬂict
or resistance from her. Following from this, I will describe a conception of unity as
freedom from conﬂict between one’s identity elements. I now turn to a critical
explication of major contemporary accounts of personal identity.
0n personal identity

Philosophical discussions of personal identity have primarily addressed one of
two only indirectly related questions, which Marya Schechtman has termed the
reidentiﬁcation question and the characterization question.1 The reidentiﬁcation
approach to personal identity, which this work is concerned with only indirectly, aims to
discover the criteria in virtue of which some person is identical with some temporally
prior person. Reidentiﬁcation accounts are thus concerned with the diachronic identity of
persons.

This discussion of reidentiﬁcation has dominated the philosophical literature,
spawning a wide variety of accounts of personal identity over time — memory‘,

consciousness’, psychological traits generally", character’ and other features of human

 

' Schechtman (1996).

2 Commonly attributed to John Locke.

3 What Locke more plausibly took to ground personal identity. He in fact denied the relevance of memory
explicitly: “being interrupted always by forgetfulness [. . .and] even the best memories losing sight of one
part whilst they are viewing another [...] concerns not personal identity at all.” (Locke 1690/1964, 2,
XXV II, 10.) See Schechtman (1996), pp. 107ff for an argued denial of Locke’s holding of a memory view.
4 Parﬁt (1984/ 1987).

5 In the somewhat peculiar sense of “projects and categorical desires with which the person is identiﬁed”
(Williams 1981, p. 14)

49

life have been adduced as the ground of personal identity. The most important reason for
noting this conception of personal identity — a conception that takes as central the
‘identical to’ sense of ‘identity’ — is to distinguish it and the philosophical literature that
addresses it from the characterization understanding of personal identity that is at issue
here.

Characterization accounts of personal identity, on the other hand, aim at
explicating the features in virtue of which a person is the person she is. The sense of
‘personal identity’ that is sought in the stereotypically adolescent or midlife question
‘who am I?’ is the sense such accounts address. (Preempting some misplaced criticism is
the reminder that personal identity is not automatically equivalent to biological identity,
historical identity, racial identity, and so on. That is to say, though ‘being a bipedal
mammal’ is, in everyday experience, an identifying criterion for persons, this
characteristic is not (typically) an element of personal identity.)

The question of what makes a person the person she is can seek two sorts of
answers. One regards the features of personal identity generally: of what elements is a
personal identity constituted? Call this ‘the general characterization question’. This
question assumes that identities are distinguishable entities — that they exist, have causes
and effects, and that true and false claims may be made about them. Similar assumptions
have been made - and questioned — in related discussions of the self, and in
reidentiﬁcation accounts of personal identity" I will not question the assumption here,
as I take it to be correct. Nor will I be centrally concerned with determining precisely

what features constitute a personal identity. I wish the arguments in defense of the ideal

 

6 Dennett (1989) questions the assumption that the self exists; Parﬁt ( 1984/ 1987) argues that whether in a
certain instance the relation of diachronic identity holds is indeterminate, and thus it is neither true nor false
that the relation holds.

50

of unity, the thesis I am supporting here, to be applicable to as wide a range of
conceptions of personal identity as possible.

The second question of characterization, which I will call the ‘speciﬁc
characterization question’, is, given some account of what features constitute identities
generally, whether speciﬁc instances of these features are to be attributed to a given
person. So, for example, imagining as Hume did that persons “are nothing but a bundle

or collection of different perceptions”7

, this second question of characterization asks,
regarding a given person, which perceptions belong to (or more in accordance with
Hume, are) her?

In defending the unity of identity, the speciﬁc characterization question will be
the most important characterization question. Since the unity of identity regards the unity
of the speciﬁc features of a given person’s identity, the question of which features
speciﬁcally characterize her will need to be answered prior to a determination of their
state of unity. But since, as noted above, I intend the following arguments to be
applicable to the range of answers to the general characterization question, I will brieﬂy
outline the most prominent established accounts of characterization. Given that I do not
expect agreement between all answers to the speciﬁc characterization question, I will
address speciﬁc views separately.

Narrative views
In order to illuminate the essential ideas of a narrative view of identity, 1 will

Concurrently discuss two recent accounts, those of Marya Schechtman and Hilde

Lindemann". I choose these two for a number of reasons. The approaches of both writers

\

1: Hume (1739/1978), Book 1, Part 4, section 6.
Then writing as Hilde Lindemann Nelson.

51

are more social than metaphysical — that is, the accounts aim to explain the ways in which
narrative and identity shows how people ﬁt into the social world, rather than defending a
certain view of what persons essentially are.9 Relatedly, the accounts share with the
present work an interest in the effects of identity on agency. Schechtman and Lindemann
also disagree in revealing ways, showing chief points of dispute between narrative views.
Schechtman defends what she calls a narrative self-constitution view of personal
identity. On this view, “a person creates his identity by forming an autobiographical
narrative — a story of his life.”10 By virtue of inclusion in such a narrative, actions,
experiences and personal traits come to belong to the person whose narrative it is. The
converse is said to hold as well; ”When a person is unable to explicate part of her
narrative, some set of her actions and experiences are incomprehensible to her and,
hence, not properly under her control.”ll This claim brings out at least two important
features of Schechtman’s narrative account. One is that a the elements of a person’s
identity must be comprehensible to her, in the way that narratives lend such
understanding — that is, by interpreting certain elements in the light of others, and in the
more overarching light of a meaningful, temporally unfolding progression of events.
Also suggested by the cited passage is the importance of self-understanding to agency, a
recurring topic in discussions of characterization, and one that I will return to shortly.
The core of the account — self-constructed narrative — carries with it seemingly
contradictory elements. One element is the plausible allowance that a person’s self-

narrative need not be entirely explicit (here I will follow Schechtman’s equivalent use of

 

9 The metaphysical-narrative foil 1 have in mind here is Dennett (1992), the central argument of which is
that selves, narratively constituted, are entirely abstract ﬁctions.

'0 Schechtman (1996), p. 93.

” Schechtman (1996), p. 118.

52

‘to make explicit’ and ‘to articulate’). Another, contradictory demand is that it must be a
story in a recognizable form. If the story must be in such a form — what Schechtman calls
standard (alternately, traditional) linear narrative — it is not clear how it could be
anything other than explicit, since that is just what narratives are.‘2

One way to relieve the tension caused by these claims is to note that
Schechtman’s standard linear narrative criterion, though given a fair amount of attention
by Lindemann, isn’t a demanding one. The notion of ‘standard narrative’ is described as
accommodatingly indeﬁnite and as such is an ineffective criterion for judgment. The
notion allows almost anything to be considered standard, even something of a completely
novel form. In addressing “the question of just how different a narrative must be before it
is removed from the family of narratives making up the standard form,” Schechtman
allows that both “brilliant moments in ﬁction” and analogously “brilliant lives” may
expand the notion.‘3 Thus what is standard is indeterminate, making it unclear just what
‘standard’ means if not ‘usual’ or ‘common’ or the like. Further, the criterion is not
ultimately a criterion for judgment at all. Since future cases determine just what is
standard, we can’t use that concept to judge a certain narrative; on Schechtman’s
description, narratives (“brilliant” ones, anyway) help us to ascertain what is standard,

not the other way around. '4

 

‘2 Even accepting that there is more to a narrative than just words (or even just images, as in narrative ﬁlm),
the other elements of narrative (characters, actions, meanings) are ultimately directly dependent on explicit
discourse (Barthes and Duisit, I975).

'3 Schechtman (1996), p. 105.

'4 And perhaps expand the concept to the point at which it does not do the work that Schechtman intends.
For example, the novel Life of Pi (Martel, 2001) is the story of a young boy, Pi, adrift on a lifeboat which
also happens to be carrying a Bengal tiger, a hyena, a zebra and an orang-utan. The story primarily
revolves around the plight of the castaways and Pi’s ultimate survival. What undermines the use of this
book’s speciﬁc sort of narrative in providing a model of a person’s life history is this: Pi offers not one but
two equally internally coherent, equally explanatory but radically distinct and incompatible accounts —
narratives — of his experience on the lifeboat.

53

It should be clear that the story of one’s life does not need to be fully articulated.
To make such a narrative completely explicit would be practically impossible.” A
narrative identity does, though, need to be locally articulable. On this issue, Schechtman
claims that a person must, for the most part, be able to explain her actions, thoughts,
feelings and desires with reference to the narrative that constitutes her identity.

But Schechtman also realizes that there are identity elements that an outside
observer can see are attributable to a person who ﬂatly denies their presence. She gives
an example of a man who claims only affection toward his brother but tends to ““forget”
his brother’s birthday, “unwittingly” serve his least favorite foods when inviting him to
dinner, “inadvertently” say things that humiliate him, and so on.”"’ A sharp observer can
see that the man has hostility toward his brother. On the narrative self-constitution view,
is this hostility part of the man’s self-conception — the story he tells about himself?

According to Schechtman, somewhat surprisingly, it is. The hostility is
motivating his behavior and affecting his emotions, and so is aptly ascribed to him. She
concludes that while “the hostility is not part of the story this person tells of himself,
then, it can still be part of his self-conception in a very real way.”'7 If this is true, though,
it is not clear that the l"-personal narrative is the root of personal identity. A plausible
criterion of local articulatibility, then, is that for some action, feeling, belief, or the like to

be attributable to a person requires that some advantageously placed observer — perhaps

an outside observer — be able to explain the way in which the trait ﬁts into its possessor’s

 

'5 As well as impossible in principle. In Jorge Luis Borges’s story “Funes, the Memorious”, Ireneo Funes
sustains a head injury that destroys his ability to forget. Ireneo thus sets out to catalogue the entirety of his
experiences, which the reader comes to realize is futile: he must, of course, catalogue his cataloguing, thus
condemning him - nightmarishly - to a life of inﬁnite regress.

‘6 Schechtman (1996), p. 116.

‘7 Schechtman (1996), p. 116.

54

life story. This might be more in line with Schechtman’s thought that whether or not
people ever do formulate an explicit narrative about certain elements of their identity,
they have “a general set of background assumptions about themselves and their lives to
guide the unfolding of experience.”'8 Given that unconscious self-assumptions ground
identity, though, casting the view as one of self-constructed narrative seems a bit
misleading.

Lindemann’s view does not attempt to defend such a ﬁrst-personal advantage in
the construction of identity. She maintains that an identity is a number of stories, both
ﬁrst- and third-personally given. As might be expected, stories from different sources
will not always agree. Lindemann claims, “some of the stories that constitute a person’s
identity may be deeply at odds with one another.”19 This is in direct contrast to
Schechtman’s demand that “in order for a narrative to be identity-constituting, it must
have a high degree of coherence.”20 While I maintain (and continue to argue) that there
are important reasons to think that a person’s identity ought to have a high degree of
coherence, it is another matter entirely to assert (as Schechtman does) that the lack of
such coherence is tantamount to the lack of an identity itself. There are many persons
who simply are unable to tell a coherent story about their lives. Have such persons no
identity?

A powerful example of such a failure of narrative ability and its effects on identity
is given by Susan Brison in her philosophical and personal exploration of her own

survival of sexual violence." Brison does not offer a distinct account of narrative

 

'8 Schechtman (1996), p. 116.
‘9 Nelson (2001), p. 76.

2° Schechtman (1996), p. 98.
2' Brison (2002).

55

identity; her project is an analysis of the destruction of the identity of victims of violence.
Telling a narrative, she explains, requires accurately descriptive words, concepts and
metaphors, as well as an audience willing to listen and able to understand; both of these
prerequisites, necessary for making coherent sense of one’s experiences, are often
unavailable to victims of violence. Brison argues that this failure of narrative ability
undermines the remaking of the self that is destroyed by violence.

Such an example might be taken in support of Schechtman’s implicit claim that
an incoherent narrative identity is no identity at all. Brison, though, fails to distinguish
between diachronic and synchronic identity, seemingly taking identity over time to
encompass the whole of personal identity. She claims, inaccurately, that the narrative
understanding of personal identity “is a version of the view that psychological continuity
constitutes personal identity.”22 What is true on narrative accounts of the sort Schechtman
defends is that a person is characterized with reference to elements that are interpreted
and understood with relation to temporally distant elements — thus, an important part of
the story of who I am is the story of my childhood. But neither a continuity theory nor a
narrative theory of personal identity would plausibly hold that a person is characterized
either by the connections she has with temporally distant elements nor by any speciﬁc
examples of those elements themselves; indeed, one ability of a narrative is to make
comprehensible the changes in a person over time.

Thus the most that can be concluded from Brison’s argument is the discontinuity
of identity over time, which is not equivalent to an identity destroyed. There may indeed

be effects of violence that involve the destruction of person’s identities, but these effects

 

2" Brison (2002), pp. 40-41.

56

are not due simply to a temporal discontinuity (a condition which is commonly caused,
for example, by certain insomnia medications which induce a short period of anterograde
amnesia). Brison’s claim that the victims of violence are in some sort of undesirable
identity state is certainly true, but the claim that they have no identity is not.

While I am not claiming that Schechtman is herself failing to distinguish
synchronic from diachronic identity (this is clearly not the case), it may be that
Schechtman’s single-story understanding of identity — rather than Lindemann’s multiple-
story conception — is committing her to an overly unitary account of what an identity has
to be for it to be an identity at all. If this is the case, then the multiple-story
understanding is much more acceptable.

The multiple-story understanding is more plausible for other reasons, as well.
Each person has at least two sorts of stories that constitute her identity — her own, and the
ones others tell about her. These are the two basic elements of Lindemann’s conception
of identity: “By ‘identity’ I mean the interaction of a person’s self-conception with how
others conceive her: identities are the understandings we have of ourselves and others.”23
Indeed, this is consistent with the earlier example of the man who harbored unconscious
hostility, which was part of his identity but only from the third-person point of view.
Lindemann explicitly addresses the crucial role that other’s narratives play in a person’s
identity.

One important avenue through which other’s understandings become part of a

person’s identity is the set of stories Lindemann terms master narratives. These stock

roles are culturally mediated, oppressive or uplifting, and sometimes caracaturish: ‘dirty

 

23 Lindemann (2001), p. 6.

57

old man’, ‘Renaissance man’, ‘absent-minded professor’, ‘princess bitch’, ‘thieving
nigger’, and so on. Master narratives are one important route through which one’s own
stories are affected by those stories others tell about her; master narratives form much of
the framework around which one’s own identity narratives are constructed. From a
practical standpoint, master narratives are not only of use in explaining and justifying
one’s own actions, but they function to oppress as well. The professor can appeal to a
master narrative to explain and justify her lack of concern for trivial matters, while the
‘thieving nigger’ master narrative explains and purportedly justiﬁes excessive police
presence in predominantly black neighborhoods.

Lindemann takes identity to have two basic elements — one’s self-conception and
the conception others have of her. I agree with Lindemann that a third—personal
attribution is part of one’s identity if she accepts it, through simple agreement or through
inﬁltrated consciousness (as in Hill’s description of the deferential wife who takes up
wholesale an inaccurate self-conception as a stable and value-conferring self-conception).
But what of narrative attributions that are both inaccurate and not taken up by the person
to whom they are attributed? Lindemann argues that such attributions are a part of the
identity of the person to whom they are attributed in virtue of the phenomenon of
deprivation of opportunity, damage that occurs when third-personal attributions unjustly
prevent certain persons “from occupying roles or entering into relationships that are
identity-constituting.”24 The damage, Lindemann argues, is caused to a person by others’
denial of her moral status in light of a negative identity attribution. So, in one of

Lindemann’s examples, an emotionally sensitive Latina identiﬁed by others as ‘an

 

2" Lindemann (2001) p. 20.

58

overexcitable Hispanic’ might be viewed as morally incompetent and denied the role of
caregiver.25

Another point of distinction between the two accounts is not unique to narrative
accounts. While Schechtman claims that one’s identity is who she is, Lindemann takes
an identity to be rather is a representation of who she is. Such claims — which I’ll show
to be less merely metaphysical than they ﬁrst appear - are usually made only implicitly.
Sometimes the assumption that a person’s identity is who she is appears in an equivalent
use of ‘identity’ and ‘self’ or ‘person’. Maria Lugones takes “the shift from being one
person to being a different person” the core of “an account of identity”.26 More often a
distinction between the self and identity is maintained, but the relation between the two is
not speciﬁed.

The likely motivation behind the notion of identity as representation is the thought
that a person can be incorrectly identiﬁed. If identity and the self were one and the same,
this disconnect between identiﬁer and identiﬁed would be impossible in principle; a
person cannot have an inaccurate self, though she be the subject of inaccurate identity-
attributions. A narrative view of identity clearly explains the common phenomenon of
misidentiﬁcation: the story that is told about a person (either by others or by herself)
simply fails to characterize the plain facts. If my earlier claim that inaccurate attributions
are not part of a person’s identity is true, though, the notion of identity as representation
is not bolstered by the phenomenon of misidentiﬁcation

Standing against such a notion of identity as representation — and in favor of a

tight conceptual link between identity and the self — is a recognition of the power that

 

2’ Lindemann (2001), p. 27.
‘6 Lugones (2003), p. 89.

59

identities have in our lives. First-personal intemalizations of incorrect identities - as
through inﬁltrated consciousness — can radically undermine agency. But if personal
identities are merely representations, such effects would be difﬁcult to explain in terms of
identity.

Some might see such claims as merely conﬂating or confusing ‘self’ and
‘identity’. The close conceptual relation I imply here, though, is not without reason. To
explain, it seems a helpful analogy to consider the relation between ‘art’ and two types of
works of art, as exempliﬁed by, say, Pollack’s Full Fathom Five, and the more familiar
and less abstract The Thinker by Rodin. Lindemann, for example, insists that the self is
that thing that is represented by one’s personal identity; thus the relation between one’s
personal identity and her self is the same as the relation between The Thinker and a
person engaged in cogitation.

There are two interesting problems with this line of thought. One is that not all art
(vis, Full Fathom Five, one of Pollack’s most striking pieces of so-called ‘drip art’) is
representational; I’ll return to this issue. Another problem is that while it is easy to
imagine that there was some (human) model to whom The Thinker stands as a
representation, it is less easy to conceive of the ‘self’ to which one’s personal identity
stands as a representation. This is in large part because identities create (affect, modify,
and constitute) selves.27 Indeed, whatever constitutes some self is either going to be
something with which I identify or something with which others identify me.

Now, an element unknown by me or by others — my as-yet unrealized interest in

promoting socialism, say — if it affects my life or my actions, is as a force moving me

 

27 The other reason why a self cannot be conceived of prior to personal identity is that the description of a
self (either the type or a token) will necessarily appeal to identifying elements, either general or speciﬁc.

60

from without. Lindemann’s claim that “how others identify me has a direct bearing on
how freely I can exercise my moral agency”28 is indubitable — if, say, others inaccurately
identify me as a transsexual based only on my androgynous outward appearance, my
actions might be restricted and I would then experience what Lindemann calls
deprivation of opportunity: I won’t be allowed to hold certain positions, or join certain
organizations, or live in certain neighborhoods. As Lindemann writes, “a personal
identity thus requires social recognition.”29

But all such considerations establish is that in some cases there are third-personal
effects on the agential efﬁcaciousness of personal identity. What these considerations do
not establish — and more, what is false — is that a personal identity requires social
recognition (or any other third-personal input) in order for it to be someone’s personal
identity.30 Sometimes (often?) others are wrong about or just ignorant of certain elements
of a person’s identity. Also, an element of a person’s identity need not necessarily be
causally efﬁcacious. Exemplifying both these claims is a person who neither acts on, nor
allows others to know of, his pedophilic desires (desires which, to simplify the example,
we can suppose he in fact wants to have).

Now, there is a clear disanalogy between art and personal identity, a critic might
notice: a piece of art, if it is a representation, is a static representation. Though the

objects that art represents change, the work itself (typically) does not; for instance,

though Rodin’s mistress Camille Claudel grew old, she is young in his F ugit Amor.

 

”8 Lindemann (2001), pp. 22-23.

2” Lindemann (2001), p. 81.

3" Except, it should be noted, in the somewhat trivial sense that any conceptualization requires a shared
language. In this sense, then, while it is true that George Washington could not have been a pervert — the
very concept post-dates Washington’s life — this is much less clearly a claim about Washington’s behavior
and self-conception than it is a claim about personal identity. The Marquis de Sade was paradigmatically a
sadist — that is, those characteristic behaviors lead to and were supported by his identiﬁcation with them —
even though the term post-dates his life.

61

 

Perhaps a better analogy, then, is one that takes the relation between the self and identity
to be like that between time and a clock — the clock tracks the change in time. The
serious difﬁculty in conceptualizing ‘time’ without ineliminable reference to those
material changes that track time, though, only lends support to my argument here."

Perhaps the critic might think a better analogy provided by the example of a
security video camera. The image provided by the camera represents the ongoing
activities of whatever or whomever is being ﬁlmed. As the subject changes, the
representation changes (that is the point of security cameras, of course — when the subject
changes in certain ways, the camera tracks those changes). To be sure, this analogy
supports Lindemann’s argument that third-personally attributed identity elements can
(through deprivation of opportunity) affect the lives of those to whom the elements are
attributed (another feature of the typical use of security cameras).

Besides the subject-relative nature of security camera images, this example
corresponds to the narrative theorist’s conception of identities as essentially
propositional, or at least similar to propositions in being truth-apt.32 This is not to say
that video is composed of verbal claims, but rather that video is often used as a claim
about some fact: the video can (in court, say) be used as a claim that she forcibly entered
the locked storeroom. As such, it can be true or false, just as, the claim runs, personal

identities can be true or false.

 

3' The difﬁculty is: assuming a very plausible sort of minimal physicalism with regards to epistemic beings,
the claim that there is this thing ‘time’ that can be in principle disconnected from the physical world (that
is, that there could be a period of time during which nothing happens) is an essentially empty claim, there
being no evidence for it in principle. The knower-relevance of time is at any rate essential to the regnant
scientiﬁc world-view (Einstein, 1961, ch 8, “On the Idea of Time in Physics”).

32 Lindemann, pers.comm..

62

This understanding is if not strictly incorrect at least deceptively limiting. This is
because, as will become clear through the investigations of the present chapter, personal
identities are not composed of only propositional elements (narratives). Personal
identities include as their various elements desires and group membership, to mention
two examples. But neither a desire, such as the one that Chavez win the election, nor
membership in a group such as ‘technogeek’, have truth—value. That such things as
desires serve as constituent parts of identity also undermines the idea that identity is a
representation, for desires are not representational.

To return to the art/artwork analogy, consider the remaining option: that a
personal identity is no more a representation than are many works of art.33 But this is not
to say that a particular artwork just hangs there (except, of course, in the most literal
sense), arbitrary and disconnected from anything else — artworks are about something,
namely art. In a related manner, personal identities are about selves — that is, about
persons. (What else would they be about?) Further, just as the only conceptual distance
between ‘a work of art’ and ‘art’ is the distinction between type and token (an irrelevant
distinction here), there need be thought no great conceptual distance between selves and
identities.

Though I’ve defended a particular stance on the matter, the question of the
relation between identities and selves is in large part terminological, yielding no more
practical knowledge, the sort with which I am primarily concerned here. With this
practical goal in mind, I would like to turn to a primarily phenomenological denial of the

plausibility of narrative understandings of identity mounted by Galen Strawson.

 

3’ Another striking example is given by DuChamp’s Fountain, which does not represent a urinal — it is a
urinal.

63

Strawson has criticized narrative views of personal identity as being both
factually and normatively incorrect. He argues that not only is it false that people
generally understand themselves narratively, but it is also not wholly desirable that they
do so. Since I take a narrative view of personal identity overall to be a plausible one, and
take my views in the present work to be compatible with such a view, a rebuttal of
Strawson’s arguments is in order here. I will ﬁrst explain how Strawson’s claim that he
is arguing against narrative accounts of personal identity is misleading in a way that
undermines the applicability of many of his arguments to narrative characterization
views. Second, I will address a few concerns about the overall plausibility of his own
view.

Strawson’s central distinction is that between persons who understand themselves
as persisting through time — such people are Diachronic — and those who do not —
Episodics. (It is important to keep in mind that this distinction regards the subjective
experiences of persons rather than some physical or metaphysical fact.) Strawson claims
to be Episodic himself, and as such does not experience a connection, in terms of stories
or otherwise, between himself and some past or future self. He writes,

“. . .I know perfectly well that I have a past [although] I have absolutely no

sense of my life as a narrative with form, or indeed as a narrative without

form. Absolutely none. [. . .] Another way is to say that it seems clear to

me, when I am experiencing or apprehending myself as a self, that the

remoter past or future in question is not my past or future, although it

certainly is the past or future of GS the human being.”34
Thus, even though understanding oneself as (inhabiting) a temporally persisting body, the

Episodic understands himself as not having a past and future (the same past and future

that the corporeal body does have).

 

3" Strawson (2004), p. 433.

Insofar as the point of contention is between Diachronicity and Episodicity,
Strawson is not clearly arguing against the truth or desirability of any distinctly narrative
conception of personal identity. What he is denying is the fact and ideal of a diachronic
self-conception, a view of oneself as importantly connected to some past and future
person in such a way that that past and future person is oneself. But, contrary to
Strawson’s claims, Diachronicity is not equivalent to Narrativity, nor is being Episodic
tantamount to being Nonnarrative.

One consideration in support of this claim is that Strawson is not arguing against a
narrative characterization view of personal identity, in spite of the fact that he somewhat
misleadingly offers Schechtman’s account as an example of a narrative view of the sort
he is discussing. Explaining how this is misleading will involve further elaboration on
Strawson’s claims, as follows.

Strawson explicitly claims that Narrativity is a “way for human beings to
experience their being in time.”35 That is, one might have the temporal experience of
herself as Diachronic, taking a special interest — self-interest — in the person she has been
and will become. Strawson plausibly claims that many people whose self experience is
of this sort, that is, people who are Diachronics, have a narrative view of the temporal
connections, their temporal continuance taking the form of a story.

In a related vein, Strawson offers a reply to the possible concern that Episodics
have a dysfunctional connection with their own past, a past that is part of the Episodic’s
life but not her own experience of it. This reply indicates the distance of his conception

of narrative identity from that of many others. He writes,

 

3’ Strawson (2004), p. 429.

65

“the past can be present of alive in the present without being present or

alive as the past. The past can be alive — arguably more genuinely alive -—

in the present simply in so far as it has helped to shape the way one is in

the present, just as musicians’ playing can incorporate and body forth their

past practice without being mediated by any explicit memory of it.”36
Thus the musician can, consistent with being Episodic, not identify explicitly with her
past self though that past self is in part an element of her present self.

Now, one implication of this is that identity-constituting narrative, at least in the
sense of it that Strawson is concerned to deny, is ﬁrst personally explicit - that is, the
connection with one’s past that Episodics are supposed not to have is a consciously
admitted connection. This leads to another reason to think that Strawson’s arguments put
little pressure on the claims of narrative identity theorists: neither Schechtman nor
Lindemann deny that a person can have narrative connections that fail to be “mediated by
any explicit memory” or thought. Now, these accounts might be thought to hold that at
least some narrative connections are explicit; what Strawson is ultimately denying is that
any such connections need be explicit.

Lindemann allows, though, that the narrative identity of, for example, infants, is
constructed of entirely third personal elements, and as such is not ﬁrst-personally
explicit.37 Infants are thus not Diachronic (very young infants, at least), though their
identity is a narrative one, constructed out of the stories of family and caregivers.
Invisibly to Strawson’s account (since it is entirely concerned with the person’s

subjective experience), these narrative elements are explicit. Again, Strawson’s claims

fail to ﬁt their target.

 

"6 Strawson (2004), p. 432.
37 Nelson (2002).

Aside from the inapplicability of the discussion of Synchronicity, some elements
of the notion beg for further defense. For example, the proffered case of the musician
does conform to a plausible intuition — that many elements of a person’s past are a part
of, and connected with, her present in perhaps more important ways that simply through
her explicit awareness of it. But the thought that a person’s past might be connected with
her present only in non—explicit ways is less plausible. It suggests that a person might
think of herself (as a person embodied or otherwise) and/or her personal characteristics as
coming into being very recently and ex nihilo (or at least, ex nihilo temporalis). While
this sort of self understanding seems psychologically possible, consistent with Strawson’s
denial that all persons do in fact conceive of themselves as diachronic, it would be at least
factually inaccurate since both physically and socially a person is causally dependent on
her past.38

It should be noted that even if it is the case that some persons experience
themselves as Episodic, and even if it is defensible to claim that the Diachronic life is no
morally better than the Episodic life, none of this undermines the claim that the Narrative
life is not or should not be the sort we experience. This is because a narrative identity
view need contain no distinctively temporal content. Much of Strawson’s argumentation
appeals, largely implicitly, to the intuition that narratives (those of identity as well as
literary fabulae) are essentially in the business of tracking events over time. In this
context, this intuition is misleading. While Strawson admits that “the Narrative outlook

clearly involves putting some sort of construction — a unifying or form-ﬁnding

 

3" This is not to say that all elements of a person’s causal past need be explicitly apprehended or even
epistemically available to her. My claim regards a person’s denial (or failure to apprehend) that she has a
causal past at all.

67

construction — on the events of one’s life, or parts of one’s life,”39 he takes the form to be
a necessarily temporal one.

But while many narratives draw connections between events on the basis of their
position in a chronology, such chronological progression is neither necessary nor
sufﬁcient for narrative. Narratives can contain ﬂashbacks (either in the minds of
characters or in the story lines themselves), be conveyed in reverse chronological order,40
or radically rewrite the past.4| Narratives, in their role of form-ﬁndin g, also appeal not
only to temporally distinct events as Strawson notes, but also to concurrent events (‘I
love her because she loves me’) or phenomena that do not, strictly speaking, occur at
some point in time (Lindemann’s master narratives neither happen at some certain time,
nor do they even need to be understood as historical products for them to shape one’s
identity). Thus the purported close link between subjectively-experienced Diachronicity
and Narrativity is not close at all, and Strawson’s arguments against the former have little
traction against the latter.

Returning now to the earlier cited passage in which Strawson avers, “when I am
experiencing or apprehending myself as a self, that the remoter past or future in question
is not my past or future,”"2 we can uncover a problem with Strawson’s defense of
Diachronicity itself. Sometimes the ‘form-ﬁnding construction’ is necessary for
appropriate (I dare say ‘accurate’) interpretations of events and actions. If a person
understands herself (as Strawson’s explicit deﬁnition of ‘Episodic’ has it) as one “who

does not ﬁgure oneself, considered as a self, as something that was there in the (further)

 

’9 Strawson (2004), p. 440.

4" As in the clearly narrative ﬁlms Memento (Nolan, 2000) and Irreversible (Noe, 2002).
4‘ As in Martel (2001).

”2 Strawson (2004), p.433.

68

past and will be there in the (further) future“, we cannot, for instance, correctly interpret
her timorous demeanor as a reaction to her abusive childhood; there is no self-understood
past self who was the subject of abuse. Though some may take this to be a liberating sort
of self-conception, avoiding a Sartrean bad faith assimilation to one’s facticity, it seems
more troubling than emancipatory — at the very least, it closes off an avenue of
explanations of one’s less than desirable characteristics, putting the burden of
responsibility entirely on oneself.

Perhaps even more striking is the fact that, for an Episodic, certain common and
familiar feelings and actions would be in principle impossible. While an Episodic might
feel, for example, whatever is phenomenologically distinct about regret (since there is
nothing psychologically impossible in doing so), whatever she feels wouldn’t be regret —
there is (for her) no past self for whose behavior she feels bad. In fact, most if not all of
the morally interesting of what we might call the ‘again-feelings’ (resentment, regret,
remorse) and the ‘again-actions’ (react, revenge, recant, reconcile, redress) make
ineliminable reference to past events. Remorse that does not refer to some past event is
simply pathological guilt; it is wrong, of course, to take revenge for something that did
not happen, and incoherent to do so for something that is not even believed by the
avenger to have happened; and so on.44

Non-retrospective but still narrative constructions are relevant here as well.
Looking forward in time is necessary for a goal to be a goal, that is, a goal aims at some
future state of affairs. Non-temporally, there is a narrative on which the distinction

between, to take just one example, the stealthy murder of an enemy and the merciful

 

‘3 Strawson (2004), p. 430.
4" Strawson promises to address these sorts of criticisms in his upcoming book Life in Time.

69

euthanasia of a loved one can only be made — the raw data (two people, one sleeping and
the other injecting the ﬁrst with morphine) are inadequate to the distinction.

Narrative gives the tools for the understanding of and, indeed, the existence of
such actions and feelings, though these tools need not be temporally grounded. Thus
Strawson’s criticisms do not apply to distinctly narrative views of personal identity, nor
to they seem to offer a plausible alternative view.

Though we’ve seen no reason to think a narrative view incorrect, and many
reasons to think such a view useful, the goal of characterizing all the major views of
personal identity with which I take the ideal of unity to be compatible recommends that
we consider non-narrative accounts of identity. The ﬁrst of these that I will discuss is
that of Christine Korsgaard.

The importance of self-conception

A physician of my very close acquaintance recently related to me a hospital case
she was involved in, and quite troubled by. The absolute horror (her description) of the
parents of the patient caused her no small amount of agitation for days after the incident.
The patient was a lS-month old boy. As his father was pulling out of the drive in his
truck, to leave for work, or to do the shopping, or whatever it is that parents so often must
quickly and distractedly do, the child ran out to happily wave goodbye to his father. The
truck, backing up, struck the boy, and ran him over; he died at the hospital.

Though I was not witness to the emotional and personal disintegration the parents
experienced, I doubt that those present, or even the parents themselves, could ﬁnd words
adequate to characterize it. For it is a terrible thing for anyone to cause harm to a child,

even accidentally; fatally to harm one’s own child is quite beyond terrible. It is in a very

70

real sense impossible. It is impossible not because such things do not in fact happen
(they do). It is impossible because being a parent just is being a person who protects her
child. In failing to do so, this child’s parents failed to be parents; they failed, that is, to be
who they were.

‘Parent’ is an example of what Christine Korsgaard calls a practical identity.
Each of your practical identities - each person has many — is “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?” ‘Philosopher’ is a practical identity, as is
‘socialist’, ‘Jennifer’s husband’, ‘American’, ‘mountaineer’, and so on.

Much of Korsgaard’s argumentation is aimed at defending a certain (Kantian
constructivist) conception of norrnativity against other (more straightforwardly realist)
conceptions. Since this is a project that is tangential (though not irrelevant) to the one in
the present work, I won’t reconstruct all of the arguments supporting Korsgaard’s more
contentious conclusions here. Starting, rather, with a fairly uncontentious observation
(being one of simple moral psychology rather than complex moral metaphysics), a
person’s self-conceptions ground the grasp that practical reasons have on her. This claim
is perhaps most clearly supported by a negative sort of example: when you cease to think
of yourself as her friend, certain reasons that used to move you no longer do; you no
longer are motivated to call her on her birthday, for example.“5

The speciﬁc ways in which such identities are of practical signiﬁcance are

virtually unmentioned by Korsgaard, but are well worth noting. For one, identities

 

‘5 Korsgaard (1996), p. 101. I will follow Korsgaard in taking ‘value’ to refer only to positive value (as in ‘I
value your life’); I will use the term ‘evaluation’ in contexts where negative and positive degrees of value
are at issue (as in ‘Having evaluated the painting, she found it lacking’).

4" This point about motivation is made at p. 120 (Korsgaard 1996).

71

structure reasons and put them on offer in a way that enables them to be a routine,
accessible, helpful part of real lives. A mother does not have to identify, out of the
innumerable reasons that there are, the reason that should motivate her, say, to take time
off work to care for her ill child. As a mother, she simply does this; at the very least, she
has reason to do this.“7 Her practical identity gives her a usable view of herself and how
she ﬁts into the world of possible actions.

Identities provide interpretive schema, as well, so that a person can see a reason
ascribed to her from an outside source as a reason, rather than seeing it merely as
moralized browbeatin g. That is to say, for a reason to be my reason, for it to have
authority over me, I have to conceive of myself as being a person to whom that reason
applies. For example, though it may be true that any suitably placed person has an
obligation to replace her car with a bicycle, only if she conceives of herself as an
environmentalist (or a bicyclist, or a responsible citizen — the relation between identities
and reasons is not one-to-one) will this obligation be seen as an authoritative reason for
her.

The intractability of the moral skeptic (the philosophical problem, that is; the
skeptical person is intractable in a different sort of way) might usefully be understood in
terms of these reason-giving features of personal identity.48 The skeptic, demanding a
reason to be moral, insinuates that she does not identify as a moral person, or in

Korsgaard’s Kantian phrasing, as a Citizen of The Kingdom of Ends. This is why no

 

‘7 ‘Mother’ in a certain sense; ‘female biological parent’ is compatible with ‘a mother does not abandon her
child’, a statement I take to be false.

‘8 This situation might be usefully understood in a number of incompatible ways, of course. The point here
is not to establish some thesis about what is really going on with the skeptic (one reason for this is that
there are different sorts of moral skeptics), but that Korsgaard’s account has something interesting to say
regarding this ancient issue.

72

answer will satisfy the skeptic: for her to take anything as a reason to be moral, she needs
already to identify as a moral person, one who values herself as moral — or, if the answer
involves connecting morality essentially to, say, religion, then the skeptic needs already
to identify as religious. In this way the moral skeptic can be seen as exemplifying a
global dependence of practical reasons (at least moral reasons) on identity.“9

Korsgaard argues that the reasons that ﬂow from practical identities express
obligations. To be sure, the way in which obligatory reasons are expressed sometimes
implies that they are a function of one’s identity: A soldier never lays down his riﬂe; a
Christian always leaves vengeance for God. This source of obligation is sometimes
explicitly exploited, as in the Department of Transportation’s ad campaign using the
slogan ‘friends don’t let friends drive drunk’.

This is a very binding — categorically binding -- conception of obligation. For if
practical identities are (the) sources of self-value, and if the committing of certain actions
is tantamount to not possessing a certain practical identity, then certain actions entail a
loss, sometimes substantial, of self value. That this is sometimes the case should be clear
from the story of the parents that began this section. To violate your practical identity is
to lose the self-conception on which your life is worth living. Thus reasons that ﬂow
from practical identities — at least one’s most important practical identities -- are
unconditional in the sense that acting on them is the condition on which a person

continues on at all,

 

‘9 There is a stronger claim about reasons that Korsgaard makes: the very existence of a reason depends on
one’s practical identity; for example, in order for‘ it is the moral thing to do’ even to be a reason, the person
for whom it is a reason has to conceive of herself as a morally-bound agent I don’t address this sort of
claim here in part because it would take another book-length work even fully to describe the current state of
the debate, and also because (as 1 see it) none of the claims I make here depend on the truth of the reasons-
claim.

73

for to violate them is to lose [. . .] your identity, and to no longer be who

you are. That is, it is to no longer be able to think of yourself under the

description under which you value yourself and ﬁnd your life to be worth

living and your actions to be worth undertaking. It is to be for all practical

purposes dead or worse than dead.50

This is certainly an extreme claim. It is not unique to Korsgaard, though. S.I.
Benn presents a discussion of tragic dilemmas, in which no available option is acceptable
without substantial and intolerable moral damage, but in which the agent nonetheless
must act. “The kind of person one is,” he writes, “will be reﬂected in — or perhaps
constituted by — the orderings manifested in decisions of this kind.”51 But the future tense
— ‘will be reﬂected’ — is not necessarily apt; “Suicide may be an option, too, and could be
chosen, rationally and deliberately, as an act expressing one’s uncompromising
commitment to both [of two incompatible options] and one’s indignation at a world that
offers only such intolerable alternatives.”52 To be forced into an intolerable ordering that
shows “the kind of person one is” might recommend the self-termination of being a
person at all.

Examples supporting of the strong connection between identity and reasons to
continue one’s life are available as well. Consider Kostas Georgakis, who in 1970
responded to the rising political power in Greece of a military junta by immolating
himself. His hometown of Corfu has erected a monument to him, inscribed with his own

words, ‘I cannot but think and act as a free individual.’ Georgakis’s conception of himself

as a free individual gave him what he took to be a decisive reason for suicide.

 

5" Korsgaard ( 1996), p. 102.
5‘ Benn (1984), p. 35.
52 Benn (1984), p. 34.

74

Philosophy has its own paradigm instance of the value of a life determined by
one’s practical identity. Socrates, to Crito’s pleading offer of escape from his pending
execution, replied in what he claimed to be the words of his conscience as it spoke to
him:

...if you go to one of the nearby cities — Thebes or Megara, both are well

governed — you will arrive as an enemy to the government. [. ..] Or will

you avoid cities that are well governed and men who are civilized? If you

do this, will your life be worth living? Will you have social intercourse

with them and not be ashamed to talk to them? And what will you say?

The same as you did here, that virtue and justice are man’s most precious

possessions, along with lawful behavior and the laws?”

Socrates submitted to an execution that he, as a practical matter, could easily have
avoided. He submitted in the realization that since he would no longer be able to

propound his beliefs, no longer be able to be a philosopher, his life was not worth
living.

While accepting the implications of these and other similar examples, as well as
accepting the tenor of Korsgaard’s arguments, I will dissent on two interrelated points.
One point focuses on the categorical nature of identity-conferred reasons, the other on the
claim that one’s identities always bring value to their possessor. I am, that is, going to
accept (in light of the discussion immediately above) Korsgaard’s account as plausible
and thus subject to the unity of identity ideal, but I am going to dispute two points within
it.

If a person conceives of herself as, say, a mother, and this conception contributes

to the value that she takes herself to have, it is not necessarily the case that acting against

the obligations derived from this identity leads to a loss in self-value. One way in which

 

5’ Plato, Crito, 53c.

75

persons maintain perception of their own value even when they act against their self-
perceived identities is common enough: hypocrisy. The hypocrite continues to conceive
of herself as living under a valuable description, but fails to act in accord with that
description. A straightforward way to understand hypocrisy is as a cognitive failure to
see the dissonance between what one is doing and what those acts show about who she is.
This invisibility might be caused by erroneous interpretation of actions (‘That wasn’t a
lie, just a harmless ﬁb’) or of the requirements of certain identities (‘Yes, I’m a peace
ofﬁcer, but I don’t have to protect those people’). Some hypocrites, then, are simply the
victims of their own interpretive incapacity. Others deserve at least some of the moral
scorn heaped on them -— those whose hypocrisy is rooted in their narcissism, explicit
double standards, or disregard for living up to anyone’s expectations, including their
own. But whether hypocrisy is a cognitive or moral ﬂaw, the hypocrite can continue to
view herself as valuable even under the description that in fact fails to characterize her.
The second point of dissent looks to the opposite end of the evaluative scale;
while sometimes an identity can provide value no matter what actions accompany it, it
might seem that some identities fail ever to provide value. Consider ﬁrst an example in
which a person’s identity does fairly clearly contribute to her self-valuing: a person one
of whose practical identities is ‘philosopher’. Because she values herself as a
philosopher, she is pleased when she does philosophical things — like publishing papers,
giving lectures, talking to others about philosophical issues — and because she is thus
pleased she is motivated to continue to do such things. She sometimes justiﬁes her
actions with reference to her identifying as a philosopher (‘You spent the whole weekend

at the library?’ ‘Well, that’s what a philosopher does!’).

76

It is readily observable, though, that not all identities contribute to self-valuing.
Some identities are (typically) unwanted: being homeless, for example, is an undesirable
and undesired situation. Judging from the number of ‘I’m Homeless’ websites, forums
and weblogs emerging on the Internet, it is a situation that leads a substantial number of
people to identify as homeless (contrast this with other undesirable traits that only
relatively rarely do people self-identify with: hypocrisy, narrow-mindedness, or
miserliness). I speculate, though, that very few homeless people take ‘homeless’ to be a
description under which they value themselves (though some do). Other identities are
similar: ‘obese’, ‘lonely’, ‘member of an oppressed minority group’. In spite of this
disconnect of identity from self-value, these identities are still practical in ways similar to
the paradigm examples of practical identities. Even unwanted identities that the
possessor admits she possesses (like ‘homeless’) motivate behavior: though the mere fact
of homelessness can motivate behavior such as the seeking out of a shelter, the undesired
identiﬁcation as homeless can motivate behavior as well — perhaps behavior that serves to
cover up the fact of being homeless.

These observations undermine the necessity of a practical identity contributing to
self-value. Turning away from explicitly philosophical lines of argument for a moment,
certain observations of sociological theorists are relevant in explicating the connection
between undesired identities and self-value. The terminology is a bit different, but the
concepts are very similar. What Arnold Birenbaum calls a ‘social identity’ is constituted
of roles, sets of expectations that constrain and obligate agents. 5“ Birenbaum takes as

exemplary of roles ‘mother’, ‘senator’, ‘president’, ‘ex-convict’, and ‘scholar in

 

5‘ Birenbaum (1984). Here the focus is on the source of identity, which Korsgaard agrees is
overwhelmingly social.

77

residence’ — the same sorts of categories that are practical identities. In an interesting
merging of practical identities and (a certain conception of) narrative identities, he notes,
“roles are associated with orderly accounts of the events that produced or were connected
with [them]”55
In studying the disabled, Birenbaum reports that identifying as disabled, like

other undesired or stigmatized roles, does contribute to self-valuing, though in a more
roundabout way than Korsgaard suggests. Since disabled persons are stigmatized, taking
up an identity that offers a narrative of how it is the person came to be so identiﬁed can
reduce self-blame. The identity, as an account of events, explains and excuses
negatively-perceived attributes. Thus identifying as disabled does contribute to a
person’s sense of self-worth. Likely this phenomenon is generalizable; in a discussion of
the nature of the identiﬁcations that play a role in hate crime, Michael Blake draws
similar conclusions about the socially ostracized who take up their negatively-perceived
identities a matter of pride.56 Similar mechanisms may well be involved, though at a
more local level, in the process of terminological reclamation.57

Thus, there is empirical evidence to suggest that not all the elements of a person’s
identity need contribute to her self-value in any direct or even indirect way. A person can
gain value from her identity (hopefully this is the most common situation), but she may
merely admit to possessing a certain identity, and only be minimally satisﬁed with it.
This is an important point to bear in mind as the present discussion continues, since this

fact of identiﬁcation is central to the notion of endorsement. We will continue to see this

 

55 Birenbaum (1984), p. 322.
5" Blake (2001), p. 130.
57 As, famously, in Ensler (1998), “Reclaiming Cunt”.

78

importance in a look at another view of the source of personal identity, group
identiﬁcation
Group identiﬁcation

Korsgaard explicitly claims the elements of identity to be descriptions. They are,
though, descriptions of a broad sort. Virtually every example of practical identities she
mentions refers to membership in some group — psychologists, parents, Christians. Only
rarely does Korsgaard characterize practical identity in terms of anything like a speciﬁc
description — e. g. ‘the ﬁrst woman to walk solo to the North Pole’, or ‘Jennifer’s
husband’ (the most speciﬁc example she gives is “someone’s lover or friend”). Though
the notion of ‘a description under which you value yourself’ seems to imply relative
speciﬁcity (I take ‘description’ to suggest more rather than less speciﬁcity), the claim that
practical identities give practical reasons is most strongly supported with reference to
descriptions of group membership. While the description ‘the ﬁrst woman to walk solo
to the North Pole’59 is most certainly one under which a person could value herself, it is
not clear how it would confer obligations — since she has already done the thing that
characterizes the identity. A description that is more useful in understanding the practical
aspect of identity, such as ‘woman adventurer’, describes many - that is, a group — of
women."0

That claim that identities are grounded at least to a signiﬁcant extent in group
membership is made explicitly in some characterization accounts. Diana Meyers reminds

that “who we are — what we are like and how we think and act — is signiﬁcantly

 

’8 Korsgaard (1996), p. 101.

59 Helen Thayer did this in 1988 at age 50, solo so long as one does not count the assistance of her canine
companion, Charlie.

6" Such a large group that a number of companies make a business out of outfitting its members, and
Women’s Adventure magazine provides a journalistic and advertising venue.

79

inﬂuenced by social systems of domination and subordination.”"l Personal identity — who
one is - is thus signiﬁcantly inﬂuenced by one’s memberships (or ascribed membership)
in social groups, since social groups are the avenues for domination and subordination, at
least in those forms of domination and subordination that are properly considered
oppressive."2

The abovementioned article of Michael Blake’s makes connections between
oppression and identity as well. He notes that for “membership [to] give rise to feelings
of identiﬁcation [. . .] it is not just the case that we recognize our membership in such
groups - we also understand such membership as an essential part of our place in the
world.”63 This claim is made in an analysis of hate crime — a category of offense that
depends not merely on the victim’s group membership, but on a certain sort of
membership: being a member of a group with which she is identiﬁed. Many of us are
members of the group of people with brown eyes, but this is not a (type of) group with
which we tend to identify.

For our purposes, ‘feelings of identiﬁcation’ will not help in explicating what it is
for someone to identify with something, since feeling as if you are identiﬁed with some
group is neither necessary nor sufﬁcient for being identiﬁed with it. Nor is it clear why
felt identiﬁcation with a group must be essentially connected to our self-understood
social position. It is odd to think that someone would consider themselves essentially

homeless. Being homeless is plausibly thought of as an accidental characteristic of a

 

6‘ Meyers (2000), p. 153.

62 “If an individual is oppressed, it is in virtue of being a member of a group or category of people that is
systematically reduced, molded, immobilized.” (Frye, 1983, p. 8.)

(’3 Blake (2001), p. 129

80

person, and is nonetheless able — as many other accidental characteristics -- to contribute
to our identities.

The point of bringing up Blake’s understanding of group identity is not to point
out the different uses of the term for his project and for mine. In fact, the differences are
less notable than the connections his discussion makes between the various accounts of
identity that have been outlined here thus far. One such connection is that in appealing to
one’s group membership in her own social self-understanding, a person acquires a self-
conception of the sort Korsgaard takes to be the root of reasons.

Blake also notices what Lindemann did in constructing her account of
counterstories: some of the elements of our identities are grounded in conscious
resistance to oppression.64 Blake focuses on the accomplishment of reappropriation, the
taking up of disparaging labels as a matter of pride; what Lindemann describes is a
changing of the associations that the labels conjure up. What is important here is the
common element of alterability: group-based elements of personal identity, whether
characterized narratively or not, do not necessarily make inexorable demands of their
bearers.

Thus the practical impact of identities is not linked to a certain view of just what
constitutes a personal identity. This, if you will recall, is entirely in the spirit of this
chapter, which is importantly concerned with accommodating as well as conjoining
various accounts of identity. In the next section, I will explicate Harry Frankfurt’s

account of personal identity, which is centrally concerned with the practical efﬁcacy of

 

6‘ Blake (2001), p. 133.

81

personal identity. I will be supporting Frankfurt’s view, as well as his notion of
endorsement that will play a large role in characterizing the unity of identity.
Desire endorsement views

Addressing criticisms of the ideal of unity in the previous chapter gave much
opportunity to outline Frankfurt’s conception of personal identity, but only in a negative
context. Here it will be illuminating to review Frankfurt’s account in the interest of
showing the connections between it and other conceptions of personal identity just
discussed.

Frankfurt takes personal identity to be comprised of desires. Now, as an early
party to the philosophical discussion of personal identity, Frankfurt doesn’t use the
phrase ‘personal identity’ as many more recent writers do. Rather, he speaks of the
processes of identiﬁcation by which one acquires and divests herself of the desires that
comprise her as a person. But it is clear from later works that personal identity is just
what he takes to be at issue."5 The object of the process of identiﬁcation, Frankfurt

”6" and “the development of a distinctive and

repeatedly claims, is “individual identity
robust sense of personal identity.”67 Frankfurt takes persons most characteristically to be
volitional beings; of any person he argues “the boundaries of his will deﬁne his shape as
a person.”"8 The most stable of these boundaries, he argues, “comprise the necessary

conditions of his identity.”69 I don’t intend to defend the all implications of these claims,

but offer the passages only as illustrations of the core of such a view of personal identity.

 

‘55 And it is clear judging from the personal identity related claims of various of his critics, as discussed in
Chapter 1, above.

6" Frankfurt (1999), p. 110.

67 Frankfurt (1999), p. 108.

68 Frankfurt ( 1999), p. 114.

69 Frankfurt (1999), p. 115.

82

According to Frankfurt a person identiﬁes with desires through her endorsing
them. Endorsement takes a number of forms, all grounded in a structural relationship
between ﬁrst-order desires — desires simpliciter, motivating desires — and second-order
desires — those that take other desires as their object. Thus my desire to eat chocolate is a
ﬁrst—order desire, and my desire that my desire to eat chocolate cease is a desire of the
second order. If I, alternately, desire that I desire to eat chocolate, I thereby endorse the
desire to eat chocolate. If I, further, desire that my desire to eat chocolate is one that I
want effectively to be motivated by (since I could desire to have but not act on a desire to
eat chocolate), the desires are called volitions.70

Some desires, though they are endorsed, are fairly transient (my own desire to eat
chocolate is of this sort). Others are desires that their possessor desires to continue — for
example, were my desire to do philosophy wane, I would be concerned, and perhaps
upset, since I desire that my desire to do philosophy continue. Desires of this sort play a
role in caring. Those things that a person cares about are things that she wants her
desires for to persist. Intuitively, things that a person cares about are most central to her
identity, and we sometimes speak as if the person and the things she cares about are one
and the same: if her project is thwarted, she is thwarted; if her lover is displeased, she is
displeased.71 Beneﬁts to the things cared about are beneﬁts to she who cares about them,
as well.

Not every desire and characteristic that is properly attributable to a person is one

that she actively and consciously cares about, or even is one that she actively desires to

 

7" The general practice in this context is to call all desires, whether volitions or not, ‘desires’ unless there is
a speciﬁc point to be made with the distinction; the default is, understandably, to assume that an endorsed
desire is one the person wants to be moved by.

7' These two examples might be taken to illustrate distinct phenomena: thwarting her project is tantamount
to thwarting her, but there is a causal link between her displeasure and that of her lover.

83

have. Some undesired characteristics belong to a person if she simply accepts them - if
she has a characteristic she prefers not to have but is not willing to expunge it; harmless
vices that a person simply lives with are of this sort. Others a person wants to eradicate
but is unable to, such as some unalterable behavioral trait, or an intractable chemical
dependency. A point not often noted by critics is that Frankfurt does acknowledge (if
only in passing) this feature of actual lives. Regarding the possession of desires just as in
other aspects of life, he writes, “people often settle for less than what they think it would
be possible for them to get.”72

A person’s identity, on Frankfurt’s model, is implied to be the whole set of ﬁrst
order desires with which she is satisﬁed, plus all the higher-order desires she has.
Certainly a person identiﬁes with more than just her ﬁrst-order desires, her desires to do
something or other. Frankfurt acknowledges that second-order desires can be subject to
still hi gher-order desires; I might wish, contrary to fact, that I wasn’t satisﬁed with my
desire to sleep through the morning’s lecture. A theoretically possible inﬁnite regress
into ever higher levels of endorsement is open to practitioners of extreme self-reﬂection,
but for most of us, if we don’t ﬁnd internal conﬂict rearing its head of its own accord, we
don’t go looking for it.

A critic of Frankfurt’s view might question what makes the a persons highest
level desires (of whatever level the process terminates) hers.” The critic’s worry is that
if her ﬁrst-order desires are hers on the basis of her second—order desires, and those are

hers on the basis of her third-order desires. . .then the merely theoretical possibility of

inﬁnite regress seems, to the critic, more than merely theoretical.

 

72 Frankfurt (1999), p. 103.
73 This is a version of Christman’s concern, discussed in Chapter 1, above, that an autonomous person’s
higher-order desires themselves be autonomous.

84

One response to this worry is to reemphasize the point that endorsement is
primarily a matter of satisfaction, or lack of conﬂict.” The worry, that is, takes its
strongest form if ‘identiﬁcation’ is taken to be an active connection between lower- and
hi gher-order desires, since on this understanding in order to determine whether a lower-
order desire belongs to a person there must be an appeal to some speciﬁc higher-order
desire (the one that is to actively endorse the ﬁrst), a speciﬁc hi gher—order desire about
which it may be wondered whether it itself is hers. But since endorsement is ultimately a
matter of a lack of conﬂict, there is no speciﬁc hi gher-order desire the status of which
might be questioned.

Now, this negative use of ‘endorsement’ — negative insofar as it denotes a state of
lacking something, namely, conﬂict — might seem to stretch the notion of endorsement.
The origin of the term is grounded in the practice of endorsing checks by writing one’s
name on their back (‘dorsal’), certainly an active sense of the term. But consider this sort
of use:

In the fullness of time the ship was ready to receive her passengers. I was

introduced to the young gentleman who was to be my roommate, and

found him to be intelligent, cheerful of spirit, unselﬁsh, full of generous

impulses, patient, considerate, and wonderfully good-natured. Not any

passenger that sailed in the Quaker City will withhold his endorsement of

what I have just said.75
The claim is not that none of the passengers will fail actively to indicate their
belief in the positive qualities of the roommate, but rather that no passenger would

disagree. In this usage, ‘endorse’ is related to ‘approve’, a sometimes passive

exercise, as when, say, ninety Senators approved (of) the bill — certainly not all

 

7“ Other responses to it can be found in Chapter 1, above, where Christman’s regress objection to
Frankfurt’s notion of autonomy was addressed.
75 Twain (1869), ch. 2.

85

ninety sang the praises of the bill, but they were effectively satisﬁed with it, that
is, did not ultimately dispute its passing.

To return to the worry about inﬁnite regress in identifying with higher-order
desires: if it still remains, there is another response to the critic. The response is that a
person’s hi ghost-order desires aren’t hers, strictly speaking — they are, rather, what make
certain lower-order desires hers. If this is true, then it is inappropriate to expect that
whatever makes the lower-order desires hers must also make the hi gher-order desires
hers. An analogy to support this interpretation is found in the philosophical study of law.
Consider the law prohibiting burglary - it is a law in virtue of its being a bit of written
legislation.” More speciﬁcally, it is our law in virtue of its being written legislation. But
whatever makes bits of written legislation law cannot itself be a bit of written legislation,
of course, since if it were, the connection between written legislation and law would
never get off the ground. Whatever makes legislation law cannot be something of ours in
the same way that the law prohibiting burglary is ours. In the same way, we need not
expect (contra the claim of the critic) that a person’s highest-order desires be hers in
virtue of their being subject to some still-higher-order desire. But those hi ghost-order
desires are indeed hers, in the same way as whatever makes legislation law is part of our
legal system (on a suitably broad understanding of ‘legal system’).

The critic might reply that the analogy is a false one, since in the case of personal
identity, whatever makes my desires mine is different (in content, at least) from what
makes your desires yours. Not so in law, since whatever it is that makes a piece of

legislation law for us is the same thing that makes a different piece of legislation law for

 

7" Well, a number of bits, since I assume there to be pieces of legislation relevant to burglary on both state
and federal levels.

86

them; whatever grounds the local authority of legislation is the same in the United States
as it is in Canada. This reply, though, ignores the many differing grounds that have,
historically and currently, underwritten the authority of written legislation: the authority
of the monarch, religious authority, and general considerations of justice, to name just a
few.77

Other analogous cases can be outlined, the most obvious of which, like the case of
identity, involve a system comprising one level of components that are subject to the
evaluation of components of another level.78 Many such examples support the claim that
the grounds on which certain desires are a certain person’s desires can be a part of her
identity without those grounds being (impossibly) subject to the self-same grounds.
Thus, in a number of different ways, the inﬁnite regress concern is answerable.

A more applicable concern regards the constituent parts of identity generally.
One might suppose that the elements of personal identity are by no means exhaustively
categorized by reference solely to desires. Certainly, desires are explanatory, perhaps
essentially so, of other identity constituents. A person who identiﬁes as a socialist desires
that political structures be arranged in a certain characteristic manner; someone who cares
about the natural environment identiﬁes with a range of desires such as (for example) the
desire to use less fuel, and the desire that urban sprawl be reduced. But a person who

admits that she is homeless, and that being homeless is thereby part of her identity, is not

clearly standing in some internal relation to some set of desires.

 

77 This latter consideration has been invoked in trials of war criminals, who have been argued to be subject
to laws that are nowhere written down, and required to ignore legislation that is contrary to certain basic
principles of justice.

73 The rules of games, as a further set of examples, are the rules of certain games, though the determining
criteria are not themselves the rules of any of the covered games.

87

Indeed, each of these accounts has some degree of incompleteness; narratives,
group membership, and the things we care for are, as has become apparent in the
foregoing, are all elements of the identities of persons, but none fully characterize the
range of identity-constituents. Even a fairly modest understanding of general
characterization will likely need to take up components of each of these accounts. This
could take place coherently at a number of levels. Narratives are likely full of reference
to Frankfurtian desires, for if the point of a narrative self-understanding is to make sense
of one’s actions, often only by reference to desires can certain actions be understood.
The self-descriptions of practical identities will at least sometimes be a narrative
description, as for example understanding oneself as ‘Jennifer’s husband’ -— which, if it
has any content beyond that common to any husband, this content will likely take the
form of a historical narrative.

Having examined a number of accounts of personal identity -— answers to the
general characterization question - the speciﬁc characterization question still remains.
This, recall, is the question of which speciﬁc element — which story, which desire —
belongs to a given person. Although in the course of addressing the general
characterization question, illustrations appealing to speciﬁc identities were made, a more
focused look at the varied answers to the speciﬁc characterization question will show
there to be far less agreement on this question than the previous. One point of agreement,
though, will become clear: that endorsement plays an important role in each of the
accounts’ answer to the speciﬁc characterization question. Since the notion of

endorsement is especially difﬁcult to ascribe to Schechtman’s account, the following

88

section will directly consider Schechtman’s version of the speciﬁc characterization
question before I explain the role of endorsement in other accounts of personal identity.
Schechtman and the specific characterization question

Some approaches to what I am calling the speciﬁc characterization question serve
immediately to narrow the range of answers. Consider Schechtman’s claim that “the
most familiar examples of the characterization question are more speciﬁcally questions of
which characteristics are truly those of some person (as opposed, say, to those which are
his as result of hypnosis, brainwashing, or some other form of coercion).79 This claim
presumes that a characteristic a person is coerced into having is not truly hers - a
common thought, as was seen in the previous chapter.

Discussing the way in which Frankfurt’s account of personal identity answers the
speciﬁc characterization question — and how this answer ﬁgures into the broader
landscape of identity we now have before us — is a fruitful way in which to examine
Schechtman’s assumption.

A structural view of identity agrees in many instances with Schechtman’s
assumption that a coerced desire does not belong to the person so coerced. One can be
given group membership against her will — be inveigled or even bullied into joining a
labor union she does not want to be a part of, for instance. Desires can be instilled in her
that she prefers not to have — if you take a bite of this chocolate, you might come to
desire more, though prefer that you didn’t have such an unhealthy desire and so resist it.
On the structural view, these characteristics are not ones that the person identiﬁes with.

Since to identify with a desire, say the desire to eat that chocolate, entails that the person

 

7" Schechtman (1996), p. 73.

89

endorses the desire - she wants to have the desire — the structural view holds that the
unwanted desire to eat the chocolate is not hers. She doesn’t care about labor unions — in
fact, just the opposite — and so her membership is not a constituent element of her true
identity.

The more problematic cases involve a coercive and typically wholesale
manipulation of the second-order desires in virtue of which other characteristics are
identiﬁed with. Such cases range from the rare to the ubiquitous: conversion
experiences, hypnosis, brainwashing, socialization both oppressive and benign, and
childhood education and indoctrination. On the structural view of identity, so long as the
person endorses the desires on which she acts, they are her desires.

The pretheoretic interpretation of paradigm cases tends, I think, to be clear
enough. Consider a socially well-placed, emotionally and mentally competent adult who
acquires a desire to make a charitable donation to an organization that she sought out of
her own accord and put time into researching the goals and successes of the organization.
The desire corresponds to her other desires and commitments, and is not out of character
(she tends to be a cautious but regular charitable donor, and she is also giving of her time
to certain causes). There is little dispute over this desire properly being attributed to her.
At the other end of the spectrum, an impoverished, uneducated homeless adolescent who
is forcibly taken into the compound of a suicide cult and brainwashed through threats and
chemical manipulation into endorsing the aims of the organization is not typically
thought to truly identify with those aims. The desires she possesses because of

brainwashing are, we tend to think, not her desires.

9O

Cases lying somewhere between these extremes, though, are more difﬁcult to
judge. Imagine a not so well-off woman who has never made a charitable donation. One
day she is deeply moved by a mainstream news media story about the plight of Sudanese
women, and makes a large ﬁnancial contribution to a hi gh-proﬁle charity organization.
Or consider the various methods of indoctrinatin g children and adolescents that we
typically take to be legitimate: enforced education, the system of punishment and reward
that parents as well as the larger society employ to encourage positive behaviors, beliefs
and attitudes.

It can be very difﬁcult to separate our intuitions about characteristics, desires and
beliefs that are good for a person from our intuitions about those characteristics, desires
and beliefs that are her own. Those elements of a person which come about as the result
of required, usually public, education; those that benevolent parents instill through
various conditioning methods; the attitudes that accompany a positive change of heart
motivated by bare emotion can all beneﬁt the person who possesses them.

But when the content at issue is different, intuitions tend to change. An
elementary school child who has been taught that the forms and functions of biological
organisms are a function of evolution by natural selection is thought (ceteris paribus) to
have a belief that is truly hers. But a similar child who attends school across the state line
might be taught - within the context of a distinct world-view — that the forms and
functions of biological organisms are a function of the creator’s intent and active
construction. This belief, many would tend to think, is not her own.

But, as I have argued, the self-beneﬁt of a person’s identity elements is no more

relevant to whether they are properly attributed to that person than is the beneﬁt of a

91

person’s quotidian possessions — that bottle of whiskey might be bad for me, but it is
nonetheless mine. The same argument applies to elements that in some way deviate from
reality — that a belief is false (even dangerously so) has by itself no bearing on whether it
is mine.80

The introduction of Lindemann’s account of identity has given the opportunity for
another response, one that satisﬁes more intuitions than those grounded in property
metaphors. Explaining much of the damage of oppression requires that we hold such
harmful identity elements, even those acquired through non-self—directed means, to be
properly attributed to the persons these elements damage. Recall the notion of inﬁltrated
consciousness. Inﬁltrated consciousness is the internalization, by members of oppressed
minority groups, of the hateful, dismissive or otherwise disparaging characterizations that
dominant groups attribute to them.81 Much the same phenomenon has been argued to be
at work in the internalization of self-detrimental desires.82 The primary explanatory
virtue of the notion of inﬁltrated consciousness is its role in making sense of harm to
oppressed persons in cases where there is no outside force directly and actively harming
them. Such external forces are relatively visible: legislation that denies rights (e. g.
regarding property or marriage) to members of minorities; violent hate crimes
(sometimes more descriptively called ‘bias crimes’); the various mistreatments that stem
from stereotypes (as when police refuse to take seriously the accusations of a Latina

against her purported attacker, since she is just an overexcitable Hispanic).

 

8" A distinct sort of case is one in which the belief in question is one which is impossible

to hold — for instance, a manifestly self-contradictory one. Philosophers are, of course, in the business of
talking people out of such beliefs, but since many people seem to be in the business of believing just such
beliefs, I take this sort of case to be unresolved.

8‘ Nelson (2001), p. 21.

82 Benson (1991).

92

Not so visible is the method of other of oppression’s damages, those that follow
from other than externally visible mechanisms. The substantial damage caused by self-
image disorders such as anorexia and body dismorphic disorder, for example, which are
at least in part encouraged by unattainable social norms regarding physical appearance,
cannot be explained as a function of external force.83 External forces operate against the
desires of those they harm — gays who want their unions to be recognized by marriage
rights are denied these rights against their desires, for example. But the anorexic wants to
be dangerously thin, and identiﬁes with this desire. The notion of inﬁltrated
consciousness explains how it is that social norms damage the anorexic even though her
situation accords with her desires.

To deny that the anorexic is truly identiﬁed with her desire to be dangerously thin
is to recast her situation as one of (merely) external forces. But as the example of
anorexia vividly shows, this is to misunderstand the phenomenon entirely. Replacing the
external forces, the norms prescribing thinness, with more reasonable ones is entirely
unhelpful to anorexics: as noted earlier, anorexia is in part characterized by a stage of
resistance to recommendations of different body image.“ Thus, the motive force is
internal.

As suggested above, the determination of whether a desire acquired through non-
self-directed processes is to be attributed to a person cannot be made on the basis of a
universally applicable criterion. To deny wholesale that at least some coercively

acquired desires are truly the desires of the person in whom they reside and whose

 

8" ‘External force’, that is, in the sense that anorexics are not being withheld food. The original source of
the desire to be thin might be external, but the point here is that it is (and works as it does only because it
is) internalized.

8“ Levenkron (2000).

93

actions they motivate can entail ignoring the magnitude and even the very nature of the
damage that oppression inﬂicts on them.85 The structural view of identity, as I have
argued, usefully illuminates the way in which coercively-acquired desires are sometimes
truly attributed to their possessors.

On the matter of coercively-acquired desires, Schechtman’s narrative self-
conception view has a troubling implication. Her recommended “constraints on an
identity-constitutin g narrative” include the requirement that one’s self—conception “must
cohere with what might be called the “objective” account of her life - roughly the story
that those around her would tell.”8" As a criterion intended to deny manifestly false self-
conceptions, such as my conceiving of myself as Napoleon Bonaparte, it seems
warranted. But the constraint will also serve to undermine, for example, my self-
perception as a morally worthy being if those around me deny its accuracy.87

Furthermore, this constraint is in tension with the essential core of Lindemann’s
project — to defend the notion of counterstories with which a person can deny the truth of
the stories that those around her tell. If a person’s self-narrative must be in at least
general correspondence with the narrative others tell about her, it seems as if the degree
to which a person can construct her own identity, and react against parts of it she rejects,
is severely limited.

In explaining precisely what is wrong with Schechtman’s constraint, it is useful to

note that its aim is somewhat similar to that of counterstories. This, what she calls ‘the

 

85 The closely related question of whether actions that ﬂow from such desires are properly considered
autonomous actions will be taken up in Chapter 3, below.

86 Schechtman (1996), p. 95.

87 The ﬁlm Gaslight (Cukor, 1944) illustrates the loss of self-perceived moral worth. Paula is convinced by
her new, devious husband Gregory that she is insane; subsequently, she comes to believe herself unable to
make basic decisions about her own actions.

94

reality constraint’, is a practical requirement on the content of identity so that identities
might be socially useful tools. “Living the life of a person requires living in the same
world as other persons”, and to live in the same world, Schechtman claims, is largely to
agree on two fronts: facts and interpretations.88

There is indeed a certain modicum of factual agreement necessary for social
interaction and for the recognition of others’ personhood that is the precondition of such
interaction. But the reality constraint regarding interpretation is not so straightforward.
Schechtman uses as an example a paranoiac who imagines that, contrary to fact, everyone
around her is an agent of the CIA and is watching her. Schechtman says, “the narrative
self-constitution view does not take the alleged persecution to be part of the person’s
history, despite the role it plays in her narrative.”89 Schechtman seems here to be thinking
that persecution couldn’t be a part of this person’s identity since the persecution doesn’t
exist, just as it can’t be the case that Santa Claus brought the gifts, since Santa Claus
doesn’t exist. But this is a matter of the reality constraint regarding facts: the fact of
persecution. The fact of persecution cannot be a part of this person’s identity because
there is no such fact. This claim should be relatively uncontroversial.

But contrary to Schechtman’s assertion, there are important senses in which the
interpretation is a part of her identity. This could be the case for a number of reasons. As
Schechtman admits, “the line between observation and interpretation is notoriously
difﬁcult to draw”,”0 and so we might dispute whether the paranoiac’s imaginings have at

least a grain of truth to them. Interpretations are also quite observer-dependent; a

 

8" Schechtman (1996). P. 119.
8” Schechtman (1996), p. 126.
9° Schechtman (1996), p. 127.

95

chronically sexually-harassed woman is more attuned than most people to noticing
disrespectful behavior toward her and interpreting it as such.

There is an even more straightforward way in which one’s interpretations feature
in her identity, whether or not these interpretations are accurate in some sense. The
existence of the content of the interpretation (i.e. that there are CIA agents watching) is
not what is relevant to her identity, but the story that constitutes the interpretation itself
is. That the interpretation is part of her identity is indicated in multiple ways: by its place
in her (narrative) self-concept, its role in explaining her actions, and (a point Schechtman
doesn’t address) the probable agreement by others that she is making just such an
(incorrect) interpretation. And unlike the case of Santa Claus, there does actually exist
such an interpretation to take its place in her identity: her own interpretation.

Another constraint Schechtman defends is the articulation constraint, which
“acknowledges that the elements of a person’s narrative he cannot articulate are his, but
says that they are only partially his — attributable to him to a lesser degree than those
aspects of the narrative he can articulate.”91 Thus someone whose fear of ﬂying, for
instance, prevents her even from entering an airport but does not feature into her self-
narrative — she cannot explain how this fear relates to her other fears (it doesn’t), nor tell
a historical narrative of its genesis (she might not know how she came to fear ﬂying) — is
part of her identity, but less a part than if she were able to tell how the fear features into
the whole of her life’s narrative. It is often thought that such fears as phobias are

somewhat alien to their possessors (or, revealingly, ‘victims’).

 

9' Scehchtman ( 1996), p. 117.

Other cases are dissimilar. Imagine an extreme narcissist. She considers herself
perfect in every way. She cannot articulate her narcissism (since she takes herself to be
perfect, while narcissism is a fault), but it is central to her identity: virtually every one of
her beliefs, desires and actions are best interpretable as a belief, desire or action of a
narcissist. So unlike other examples in which an identity element is not in fact, but could
in principle be, articulated - consider a person aptly identiﬁed as a creep who takes
himself to be a perfect gentleman, or psychotics, who tend not to self-identify as
psychotics -- at least some identity elements exclude even the possibility.

In the light of these sorts of examples it seems that the articulability of the identity
element is not relevant. The notion of endorsement, in the sense of lack of dissatisfaction
with the element, is more explanatory. The phobic cannot narratively articulate her fear
(she can admit that it moves her, but she cannot make sense of it in the story of her life),
and is dissatisﬁed with it.92 The narcissist shares with the phobic the inability of
articulation, but the narcissist is not dissatisﬁed with this element of her identity. The
inarticulable narcissism is attributable to the narcissist, but not the phobia to the phobic —
a distinction tracked not by articulability but endorsement.

There are other shortcomings regarding the articulation constraint. Schechtman
claims that the point is not merely that inarticulable elements are less one’s own, but that
the relationship between these elements and a person’s life is importantly different than
that shared with articulable elements. “Those features of one’s life which cannot be

articulated do play a role, but they play a role different from that played by those which

 

92 According to the DSM IV (the most recent edition of the Diagnostic and Statistical Manual of the
American Psychological Association), adult phobics recognize that their fear is unreasonable.

97

can be” she says; “their inﬂuence is rigid and automatic”.93 But simply because
something is inarticulable, because it is unconscious, does not mean that it is rigid. A
person’s unconscious hatred of her mother can manifest itself in quite varied and
situation—sensitive ways. An unconscious feature is indeed automatic, in the sense that it
operates without one’s conscious awareness (like the processes of digestion, or most
sleeping dreams) but then Schechtman’s claim is the trivial one that unconscious desires
operate unconsciously.”
The heart of identity

Throughout the examination of views of personal identity, the notion of
endorsement has repeatedly come to the fore. Endorsement of an identity element, the
satisfaction entailed either by approving of the element or admitting to possession of it, is
sufﬁcient for identiﬁcation. The important role of endorsement is seen especially clearly
in considering externally-attributed identity elements: inﬁltrated consciousness is a
person’s endorsement of a harmful or oppressive conception that others have of her. The
distinction between inﬁltrated consciousness and (mere) oppressive attributions rests on
endorsement: a person who takes up as her own the oppressive attributions of others
experiences no inner conﬂict; she is satisﬁed with the characterization of her identity; she
endorses it.

Whatever elements we take to constitute the identities of persons, we are now in a
position to see that endorsement is the means by which speciﬁc elements come to be

those with which a person is identiﬁed. Endorsement of desires by hi gher-order volitions

 

93 Schechtman (1996), p. 118.

9“ Which may in fact be the claim she intends to defend. In a footnote she afﬁrms, “[me claim that
unconscious material is rigid should [...] be taken to indicate [. . .] that it cannot be directly affected by
what is part of consciousness.” (1996, p. 141)

98

just is the core of the structural account defended above, and the notion explains how
other sorts of identity elements are identiﬁed with.

Korsgaard’s notion of a practical identity, a conception under which a person
values herself, involves endorsement in the ‘approval’ sense. Recall also Blake’s
discussion of the reappropriation of disparaging labels. The group memberships that
come about as a response to oppression, such as his example of taking up the ‘geek’
epithet as a matter of pride illustrates endorsement in something much more like the
‘acceptance’ sense; reappropriation involves creating the self-respecting interpretation on
which a person can admit a not entirely benign attribution into her identity. The ‘geek’
label is attributed to persons, certain of whom likely do not endorse it, but those who do
endorse that identity can do so through reappropriation — reinterpreting the label so that it
does not cause identity conﬂict.

The telling of counterstories involves endorsement insofar as the process makes
the absence of endorsement manifest. Faced with dissatisfaction with some (narrative)
identity element, a person dissociates herself from it through constructing an alternative
and incompatible narrative. The dissociation that is a function of counterstories is
practical in nature; counterstories are told in the interest of making some real change in
the world. But inseparable from this practical alteration is an internal disafﬁliation from
the attributed identity element. The exact relation between the internal and external
disidentiﬁcation will vary from case to case — internal repudiation may provide the
motivation and justiﬁcation for the practical telling of the counterstory; internal

dissociation might, alternately, be dependent on the recognition of others that can be

99

gained only after the counterstory is given uptake. Whatever the mechanism at play in
speciﬁc instances, the counterstory is a practical mechanism of endorsement.
Uniﬁcation

I will argue in the following chapters that uniﬁcation of one’s identity is desirable.
A reminder to the reader of the general content of the arguments of the following
chapters is in order at this point. The unity of identity will be defended on three fronts.
First, on the structural view of autonomy that I will defend, possessing a uniﬁed identity
is a precondition of acting autonomously. Second, acting with integrity has often been
considered (by philosophers, at any rate) conceptually equivalent to being internally
uniﬁed; though I will deny this thesis, I will argue that the unity of identity ﬁgures into
acting with integrity as a practical precondition as well as an explanatory feature of
integrity. Third, romantic loving is underpinned by the unity of identity at two points: the
valuing of one’s beloved and romantic commitment. The remainder of the present
chapter will outline the notion of unity that will feature in the subsequent discussions.

Though the aim here is to support the ideal of the unity of identity, since the
approach to identity is practical rather than metaphysical, and since the arguments
outlined above all appeal to certain activities, an important sort of unity will be the unity
of action and motivation. Action need not be thought of as uniﬁed in the sense Alasdair
MacIntyre advocates, in the sense of having a telos, either in the sense of “the good of a
whole human life, the good of a human life conceived of as a unity”, or the more global
sense of “the good life for man”.95 Having more than one goal does not entail disunity;

analogously, many artifacts are uniﬁed, but have more than a single goal (or, without

 

95 MacIntyre (1981), p. 189. p. 172.

100

appearing to attribute intentionality to artifacts, ‘a single function’). An automobile, that
uniﬁed mass of protective, motive and comforting elements, has more than one goal; an
art museum is often expected both to house artworks and itself serve some aesthetic
function. But what uniﬁcation of action does entail is relative absence of conﬂict between
various actions. If an art museum was itself beautiful but failed to protect the artworks
within (perhaps because of a magniﬁcent, sublime hole in the roof), it would fail to attain
the goals expected, and be characterized by a functional disunity.

In the analogical examples as well as in regards to personal identity, ‘absence’ of
conﬂict is often a fuzzy concept, and as such not only demands judgment rather than
algorithmic application, but also suggests that there will be some boundary cases about
which there might be reasonable disagreement. For while it is clear that a museum of
charcoal drawings with no roof would fail to be uniﬁed in its functions (since all the
works within would be ruined by the ﬁrst rainfall), it is not as clear (and perhaps
indeterminate) whether an art museum whose windows let in daylight sufﬁcient to view
artworks under the best lighting conditions but also sufﬁcient to degrade the pigments of
the paintings over time is uniﬁed in its functions. Likewise, the internal conﬂict between
my caring about the environment and my desire for an ambulance ride, rather than a
walk, to the hospital may be too slight to be relevant.

In order for actions to be uniﬁed, the motives that ground them are, exceptional
cases aside, best uniﬁed. This sort of uniﬁcation is straightforwardly characterizable
much in the same fashion as unity of action — the absence of conﬂict. This is the level of
unity most relevant to desire-endorsement views, as discussed above. If a person has

both a desire to smoke and a desire not to smoke, her desires are as such not uniﬁed. If

101

she has a desire to write the paper this afternoon and a desire to sleep the whole
afternoon, her desires are not uniﬁed. When these conﬂicting desires are ones the person
identiﬁes with, a disuniﬁed set of desires is tantamount to a disuniﬁed identity. In the
next chapter I will address those sorts of cases in which the disuniﬁed desires are each a
motive.96 Such a unity of motives is necessary for autonomous action, according to the
hierarchical account of autonomy I will defend.

The notion of ‘uniﬁed around’ -- as in ‘the painting is uniﬁed around the beatiﬁc
Madonna’ — is relevant to the unity of identity. In a painting about which this sort of
claim is made, that which is the unifying element of the painting - the Madonna, continue
to suppose — is the most important element. This notion of ‘important’ can be cashed out
in various ways, many illuminating of personal identity. Were the image of the Madonna
erased from the canvas, the painting would be considered destroyed, though if other
elements were erased, the work would be merely damaged. We might well understand
the other elements of the painting by reference to the Madonna (e. g. the child in the
painting is understood differently than is the child in a painting uniﬁed around plump
dancing peasants, a la Bruegel). The presence of the Madonna puts constraints (aesthetic
ones, in this analogy) on the inclusion of other elements; a Fuselian imp, for instance, is
precluded.

Less metaphorically, it can be noted that some of a person’s narratives are more
central to her self-understanding than others, some of a person’s self-conceptions are
more motivating. One or a small subset of a person’s self-understandings often serve to

anchor the rest. These anchors — central and important concerns and self-understandings

 

9" As it was noted above, a desire can be other than a motive or volition.

102

— are a person’s principles. Such principles can be the identity elements in virtue of
which a person stands in a relationship of endorsement to her volitions, desires, and other
identity elements; so long as her ﬁrst-order desire does not conﬂict with her principles,
she is said to endorse that desire, and her identity is in this respect uniﬁed. Thus, another
relevant mode of uniﬁcation is grouping one’s motives, etc. under the auspices of one’s
principles. I will address the connection of this sort of unity with the notion of integrity
in Chapter 4.

Since self-conception underlies motives, a number of relevant sorts of unity
regard one’s self-conception. One’s own self-conception can be internally consistent,
that is, relatively free from internal conﬂict between its various elements. This might be
in virtue of the content of one’s practical identities: there seems no conﬂict in the sorts of
reasons associated with the practical identities ‘mother’ and ‘gardener’ (thought there
may of course be conﬂicts of other sorts, most obviously temporal conﬂicts: there is
simply not enough time in a week to meet all the demands of each). All the possible and
desirable aspects and forms of uniﬁcation are not exhausted by the avoidance of conﬂict:
simply the conceiving of oneself as a whole, that is, as uniﬁed, is itself a mode of unity.
It is a desirable one, as well, as I will argue in Chapter 5.

A sort of resolution of conﬂict can be attained by compartmentalizin g one’s
various self-conceptions. A person who conceives of herself as a feminist in her personal
life but cannot while she is in her corporate work environment is engaged in
compartmentalization, creating insular identities that, because they are not allowed to
interact, do not conﬂict. Calling this lack of conﬂict ‘unity’ might be misleading,

though, since what is interpreted as reconciling conﬂict within a person is equally

103

interpretable as avoiding conﬂict by creating two persons. Since I don’t take the
metaphysical outcome of such a debate to be important, I will argue instead, also in
Chapter 5, that compartmentalization is quite undesirable in light of certain effects it has
on our practices of loving.

A uniﬁed identity, then, is one which is marked by endorsement - that is, a lack
of conﬂict between and among elements. But considering the extensive literature
criticizing the nature of the structure in structural accounts of autonomy, it is surprising
that there is little written that attempts to characterize the possible structures a uniﬁed
identity might take.97 Opponents of the ideal of unity generally take there to be a limited
range of possibilities, often simply equating ‘unity’ with ‘homogeneity’ or
‘ossiﬁcation’.98 Seeing no reason to defend one certain form of unity, but desiring to say
something so to initiate the philosophical investigation into unity, we might simply lay
out a few of the more obvious forms that such unity might take.

To equate ’uniﬁcation’ with ‘homogeneity’ is both practically and theoretically
implausible. Practically, one would have a very narrow range of reasons for action
(perhaps so narrow as to make much more than the basic necessities of life impossible to
acquire). Theoretically, the notion of ‘homogeneity’ excludes the possibility of a number
of distinct practical identities; a homogeneous entity is the same throughout, having no
distinguishable elements. That a highly uniﬁed identity might have a certain degree of

ossiﬁcation is certainly possible, but the likelihood of such an identity being undesirably

 

97 Even Mullin (1995), explicitly claiming (p. 3) “we need to speak with more clarity when we refer to
selves as uniﬁed”, fails to offer clariﬁcation. Her notions of ‘assimilation’, ‘compartmentalization’, and
‘negotiation’ are notions of how selves get uniﬁed; as to what unity is, her discussions, like those of many
others, repeatedly assume it to be ‘homogeneity’.

9" The ‘ossiﬁcation’ charge was defended against in the previous chapter; homogeneity is discussed
immediately below.

104

rigid is low; change in one’s identity can be the function of many factors, only one of
which is the relation between practical identities. Internal to the person but external to
her identity are many potential causes for identity alteration: change can be motivated by
the realization of error, either of the grounds of one’s endorsements99 or the
understanding of just what one’s endorsement of a certain identity entails; change in
identity can be the result of some sort of internal emotional disturbance such as
depression. External to the person entirely are a multitude of forces — social, political,
physical — that will affect the constitution of an identity that is even minimally responsive
to the world.

The possible structural variations of a uniﬁed identity fall into two broad
categories: hierarchical and non-hierarchical uniﬁcation. A hierarchically—uniﬁed
identity might be organized as a pyramid: one primary identity that trumps all others by
having authority over some few which in turn have authority over a larger set, and so on.
One identity ‘having authority’ over another means that the authoritative identity is the
one the reasons of which are typically authoritative in cases of identity conﬂict. Often
one’s moral identity occupies the top of this sort of hierarchy in a Kantian fashion, and so
a test of moral permissibility might resolve lower-level conﬂict.loo

A similar but non-pyramidal structure is an autocratic one: reasons of one identity
typically trump those of all others which themselves have relatively equal authority.
Though an identity of this sort could be excessively rigid, it is usually a caricature;
dramatic characters and the public personae of famous and infamous persons seem to

overlay such an identity, but it is the nature of ﬁctional and popular characters to portray

 

9” Here, as at other points where I neglect the cognitive in favor of discussing the connative, I owe the
noting of this important point to James Lindemann Nelson.
“’0 The most well-known critique of such a model is Wolf (1982).

105

just a segment of a full identity. Real persons possessing such an identity might be
overwhelmingly pitiable: survivors of extreme violence, such as rape or genocide, often
identify only with their being minimally physically alive, and have no other reason-
giving self-description.’°'

A parliamentary structure would be one in which more than one identity element
share authoritative force. As a contingent matter, some identities tend not to conﬂict, at
least in very problematic ways - avid novel-reader and environmentalist seem a fairly
nonconﬂicting pair — but likely these have not enough content on which alone to base a
meaningful life. Identities that in principle don’t conﬂict are more promising: the
conﬂict, both philosophical and quotidian, between one’s relational identities and her

‘02 and the desirability of rendering these somehow

moral identity is signiﬁcant,
conceptually unconﬂicting is indicated by philosophical attempts to do so.103

An aversion to talk of internal authority might lead one to consideration of non-
hierarchical structures, on which one’s various practical identities are relatively equal.104
Many of our identity conﬂict-reducing tactics aim at such an egalitarian internal
economy; in the conﬂict between sexual orientation and religious beliefs, for example, a

sort of reconciliation is often attempted by altering one of the identities (e. g. change what

it is to be a Catholic) so that the two neither conﬂict nor rule over one another.105

 

'0‘ See for example Brison (2002), chapter 3.

'02 The philosophical conﬂict I have in mind here is the one begun by Williams (1981).

'03 E. g. Velleman (1999), which characterizes love as a moral emotion; if love is a source of moral reasons,
then though it might conﬂict with other sources of moral reasons (duty to impartial beneﬁcence, say) it
can’t conﬂict with morality itself.

'04 Marilyn Friedman’s (1986) account of this sort is discussed, albeit brieﬂy, in Chapter 3, below.

'05 Dubowski and Korda’s (2001) ﬁlm Trembling Before G—d, a documentary look at the lives of
homosexual Hasidic and orthodox Jews, is a powerful illustration of the conﬂict that motivates such a
reconciliation. Lindemann (2001) is an account of how such l"-personal identity change might take place.

106

Similarly for identities that conﬂict not in principle but contingently, we attempt
to alter the world such that conﬂicts are minimized. Making it more practically possible
to be both a mother and a careerperson is one way we have restructured not the identities
themselves but the ways in which they are able to be acted upon. When this alteration is
not so much of the world but of our organization of it, persons and their lives can get
compartmentalized; the parent who reduces identity conﬂicts by (merely) temporally
delimiting the authority of his identities (being a businessperson from 9 to 5, and a parent
nights and weekends) can“ indeed reduce conﬂicts but not only risks not having a
desirably coherent life but also must guard against the failure of the compartmentalization
(for example, being a parent gives reasons to act even during the time a parent is a
businessperson).

Another undesirable sort of uniﬁcation can be achieved by arbitrariness: conﬂicts
between identity- granted reasons are resolved arbitrarily each time they arise. This is
probably common enough, but is self-undermining: practical identities don’t so much
give reasons for speciﬁable, distinct actions as they give reasons for sets of actions that
constitute projects. If ‘environmentalist’ is one of a person’s practical identities she has
reason not so much to ride her bike to work today but to ride her bike to work generally.
The less often she acts on this reason, the more her ‘environmentalist’ identity is
undermined.

The present work is intentionally noncommittal as to the speciﬁc structure the
uniﬁed identity might take. It is possible given any of these structures both to realize the
(relatively) conﬂict-free coexistence of varying practical identities and to have a rich

capacity for autonomy, integrity, and the ability romantically to love. A uniﬁed identity,

107

one marked by endorsement, is of central importance in supporting these capacities. This

claim, and its defense, is what I will take up below.

108

_3_

Autonomy
Saw the mirrors in two.
I’m a divided man.
The little I try to do
Is less than I can.
- Theodore Roethkel

In the previous chapter, I argued that a person’s identity, by giving her practical
reasons, underpins her actions. More accurately, each of her practical identities (which
together constitute her overall identity) gives her such reasons. This connection between
who one is (however characterized) and her actions is central to the notion of autonomy.

It might be thought that a person is autonomous just in case she acts in virtue of
reasons conferred by her practical identities. This thought is implied, anyway, by the
foregoing discussion of practical identity. But even if this were the only contender for an
account of autonomy, the matter cannot be so simple. Numerous practical identities give
each of us numerous reasons for numerous actions. Such exponential multiplication of
practical commands makes it unlikely that they will all congeal into a coherent directive.
And, as the poet Theodore Roethke notes (in a fragment; we do not know the context of
this observation of his), we don’t need to appeal to the mathematics of exponential

growth to know that the connection between identity and action is not always clear and

 

' Roethke (1974).

109

direct. We look in the mirror, look at who we are, and see that we are divided. This
division, in turn, limits the actions that one may properly count as her own — or so
Roethke suggests, and so I will argue here.

In philosophical discussions the foregoing observations about a person’s actions
and her relation to them typically are made in terms of autonomy.2 Speciﬁcally, the
question of whether internal division is centrally relevant to the judgment of a person’s
autonomy is a question given an afﬁrmative answer by those who subscribe to a
structural view of autonomy, and a negative answer by those who hold either a content-
speciﬁc or a procedural view.

Such accounts share many features, but differ as to their central requirement on
autonomous action. Content-neutral accounts hold autonomy to be judged only by the
structural relation of elements of the agent’s volitional economy. Content-speciﬁc
accounts take the correspondence of action to speciﬁed moral facts to be a deﬁning
criterion of autonomy. Procedural accounts hold that the content and structure of an
autonomous agent’s internal economy must either have come about by, or be modiﬁable
by, processes that do not themselves depend on the actual economy itself. In this chapter
I will defend a structural account of autonomy against content-speciﬁc accounts and
procedural accounts.

I will support a structural account of autonomy though three routes: primarily, by
presenting reasons in light of which procedural and content-speciﬁc accounts are
untenable as they stand (and thus supporting a structural account by default); by

answering criticisms of structural accounts; and also by showing that one desirable

 

2 Typically, that is, in the analytic philosophical tradition.

110

feature (the ideal self) of procedural and content-speciﬁc accounts can be satisfactorily
accommodated by a structural account.

Having defended a structural notion of autonomy, an explication of this sort of
autonomy will sufﬁce to show how unity of identity is necessary for autonomy. That is,
the defense of a structural account of autonomy is tantamount to the defense of the ideal
of unity. As will be explained more fully below, such an account takes the arrangement
of the autonomous person’s internal economy to be a practically uniﬁed one; the
elements of one’s personal identity, through their practical role in agency, are ideally
uniﬁed on a structural account of autonomy.. The last brief section of this chapter will
explain why autonomy is desirable, and also what its limits are in underpinning a good
life.

An outline of conceptions of autonomy

The range of what has been argued to count as ‘autonomy’ is very wide, and so a
brief taxonomy is in order.3 The most general distinction might be made between
content-speciﬁc and content-neutral views.

A contemporary content-speciﬁc view (one that will be discussed in more detail
in what follows) is that of Susan Babbitt.4 Babbitt argues that a person is autonomous
only if her actions aim at her own real interests in human ﬂourishing. Since oppression
can lead a person to possess an inadequate self-concept — convincing her, incorrectly, that
she is subordinate, or merits no self-respect, to mention just two elements of such a self-
concept — her real interests may not be those interests she perceives herself to have. Thus

she will not necessarily be autonomous in virtue of acting from her perceived interests;

 

3 Addressing both political and personal autonomy, and the relations between them, would broaden the
discussion even more. Here I address only personal autonomy.
4 Babbitt (1996).

111

she will be autonomous only if she acts so to further her real self interests. Some actions
— those that fail to promote one’s own ﬂourishing — are not autonomous regardless of
who performs them.

This sort of view is in contrast with a content-neutral view, on which whatever
criteria there are for autonomy, the criteria will not be grounded in a certain content.
Rather than the content of the actions or their motives, the structure of the agent’s internal
economy — her desires, commitments, various identities, elements of her self — is the
determinant of autonomy. The most prominent structural view, sometimes called the
‘Dworkin/Frankfurt model’ is, not surprisingly, an amalgam of the accounts of Gerald
Dworkin and Harry Frankfurt.’ Frankfurt’s version of the account was distinguished
from, and shown to be superior to, Dworkin’s in Chapter 1, above, and in what follows I
will continue to focus on Frankfurt’s version. On this structural view, an autonomous
action is one that is the product of a desire that the agent endorses via a higher order
desire. The speciﬁc structure that characterizes autonomous actions, then, is a hierarchy.
On a structural view, an action with any content whatsoever might count as autonomous,
so long as it issues from a person with the requisite sort of internal constitution.

On procedural accounts, whether an action is autonomous is judged on the
procedure by which the agent acquires or maintains the desires that move her. Some
accounts (such as those of Paul Benson and John Christman) focus on the legitimacy of
the process by which the desire was acquired, while others (that of Diana Meyers)
include a host of interrelated, ongoing processes. Complicating the construction of a

helpful taxonomy, procedural views share features with both content-speciﬁc accounts

 

5 The moniker is used in e.g. Christman (1988).

112

and content-neutral accounts: Meyers’s account, as we will see in the next section, has
an explicit moral requirement, and the following explanations of other procedural views
will show that they implicitly deny the autonomy of actions with certain moral content"

What all these accounts — content-neutral, content—speciﬁc, and procedural -—
accept that the conceptual core of autonomous action is self-govemance. That is, all
agree that only elements of a person herself can motivate an action if that action is to be
considered an autonomous one. But which elements of a person are truly her own?

Frankfurt ’s structural account

Harry Frankfurt, as outlined in the previous chapter, takes personal identity to be
essentially a matter of volition, and so his accounts of identity and autonomy are closely
intermingled. This can be seen most clearly by looking at the most basic elements of his
view: desires. The desires a person identiﬁes with are those she endorses, that is, desires
which are in no conﬂict with any of her hi gher-order desires. When both of the desires at
issue are volitions — that is, when both are desires she wants to be effective in her own
motivation - they are the grounds of autonomous action. Simply put, an action is
autonomous when it is motivated by a self—endorsed volition.

This might seem a bare, somewhat austere view of autonomous action.7 As the
account has developed, both conceptually and historically, it shows a more human face in
its notion of caring. The objects of some desires are the things a person cares about, such
as moral principles, political ideals, other persons, and the like. If a person cares about,

say, socialism, she identiﬁes with her desire for socialism: she desires that political

 

6 Of course, structural views are subject to an apparently similar claim: there is an implicit moral aim
underpinning the account. The ‘moral content’ that I refer to here regards the content of the (purportedly)
autonomous acts themselves, not to the moral goals (in the pejorative sense of ‘agenda’) furthered by the
promotion of the account.

7 And, of course, of personal identity as well; this criticism was allayed in the previous chapter.

113

structures take a socialist form, and she is satisﬁed that she has that desire. She also,
insofar as she cares about socialism, desires that her desire persist. Simply desiring
socialism would lead to no particular response if the desire ended, but if she cares about
socialism she would be upset if her desire for it waned.8 To care, then, is not merely (as
we sometimes imply) to desire strongly, as we might say in a teenager who we say cares
about being popular. It is rather, in Frankfurt’s usage, to want the desire to continue, and
relatedly to be disposed to experience a loss were the desire to disappear -- unlike most
teenagers who I suspect are later relieved to ﬁnd that adulthood has divested them of such
desires.9 Because the conceptual core of caring resides in the notion of endorsement, that
which a person cares about can motivate autonomous actions as well.

Grounding autonomy in the things we care about is intuitively satisfying, at least
more so than the bare ‘desire-endorsement’ model. Many of us sometimes (and some of
us often) feel as if we are doing things against our will when the things we do are things
we do not care about. The Frankfurtian model helps to explain and even to legitimate
this feeling. It must be kept in mind that a person has many (autonomy-grounding)
desires for things she does not care about, but when she realizes distinctly that she does
not care about the object of her actions, this tends to be a matter of her acting on desires
with which she does not identify. Career disillusionment, or falling out of love, or even
broader existential crises (as they are sometimes called) can herald a failure to care about
one’s employment, partner, or life (respectively); these things can then take the form of a

goad — moving us non-autonomously.

 

8 That is, if only the desire for socialism, but not her hi gher-order desire that her desire persist, waned. It is
common enough to cease to care about something (or someone) and experience no dissatisfaction thereby.

In these sorts of cases, though, one’s hi gher-order desires have withered as well as those of the ﬁrst-order.

9 As teenagers, we tend to think we care about a wide range of things that we come to realize (for better or

for worse) we merely had powerful desires for.

114

The basic desire-endorsement structure of the model itself is sound, as well, and
withstands a range of criticisms — or so I will continue to argue throughout the current
chapter. Taking up a certain one of these criticisms, one lodged by Marilyn Friedman,
will serve to support this claim of soundness as well as to explicate some further details
of the view.IO Friedman’s criticism is aimed at the notion of what she calls the ‘split-level
self’, which Frankfurt might be thought to advocate. On his understanding, the desires
that motivate autonomous action have a distinctly hierarchical structure. Lower-order
desires are endorsed (literally, ‘on the back of’) by higher-order ones. Not only are there
two unequal sets of desires, but the self is split (at least conceptually) into two distinct
elements.

Friedman’s proposed ‘bottom-up’ structural approach, on which ﬁrst-level desires
play a role in evaluating higher level “principles”, as she calls them, could be taken as
opposed to Frankfurt’s structural account; indeed, it is intended as such. Instead of a
split-level self, the two elements could be seen to work in a sort of harmony. Not the
least virtue of describing persons’ internal economy in this way is its accuracy —
sometimes our (mere) desires do play an important role in the modiﬁcation of principles
(and commitments, allegiances, and the like).

But Friedman criticizes the accounts of no less than seven philosophers together
under the heading of ‘split level self’. The split self that she seems to be wary of is a
conception she attributes in large part to Dworkin, who understands the reﬂective self to
be the ‘true self’; thus within a person there can be two selves, the true self and the

merely apparent self. Friedman derives some elements of the conception from the view

 

‘° Friedman (1986).

115

of S.I.Benn, on which the true self is linked to one’s values, ideals, and principles. The
idea that there are two selves, only one of which is the true self, is distinct from
Frankfurt’s view on which a person identiﬁes with some desires and disidentiﬁes from
others; in order to cohere with Friedman’s conception, the disidentiﬁed desires would
have to coalesce to form another resident self, an impossibility if being a person requires
possessing hi gher-order desires from which reﬂection proceeds.”

Another point on which Friedman’s bottom-up view fails to contravene
Frankfurt’s speciﬁc formulation regards the characterization of desires vis-a—vis their
structural role. Since on Frankfurt’s view hi gher-order desires are distinguished from
lower-order desires simpliciter by their taking other desires as their object, it is incoherent
to suggest that lower-order desires take higher order elements (whether characterized as
desires or principles) as their objects.‘2 If, as an example of a common sort of
occurrence, a person’s ﬁrst-order desire to smoke plays some role in the reformulation of
her hi gher-order desire that she not smoke (i.e. the desire not to have the desire to
smoke), the desire to smoke must be accompanied by a distinct hi gher-order desire — the
desire to desire to smoke.13 There is no reason to think this incompatible with Frankfurt’s
view; the view lays no restrictions on the sources of novel second-order desires, and so
they could well come about in response to certain ﬁrst-order desires that are rewarding to
satisfy, or quite difﬁcult to divest oneself of. Otherwise, this is simply a case of urges

overpowering principles, which is not at all what Friedman is proposing.

 

“ The other self, the one that is not the true self, would according to Frankfurt’s view be a paradigm
example of a wanton.

‘2 This point, as well as other responses to Friedman’s criticisms, is made in Christman (1987).

‘3 A possible reply could involve a desire that is by its nature, so to speak, both a ﬁrst- and second -order
desire, but I am unable to think of an example of such a desire.

116

Friedman’s proposal that all the things with which a person identiﬁes, whether
bare desires or deeply-held principles, play a role in evaluating each other, shares an
important feature with Frankfurt’s structural account. That feature is the uniﬁcation of
identity.

If an autonomous action is one that is motivated by an endorsed desire, the
agent’s identity (constituted of the desires with which she identiﬁes), vis-a-vis that action,
is uniﬁed. For example, acting on a desire to smoke (to use the above example) can be
autonomous or non-autonomous. Someone acting in the grip of her desire to smoke (as it
is sometimes put) is not autonomous when she smokes if being ‘in the grip of’ implies a
resistance, however futile, to the desire. Such a resistance entails an internal conﬂict; her
desires are not uniﬁed.

The example of a smoker is an apparently problematic but actually illuminating
one with which to continue to explicate the structural view. If the desire to smoke is one
the smoker identiﬁes with, and as such has no conﬂicting desires, she acts autonomously.
The desires (relevant to smoking) with which she identiﬁes are uniﬁed, and so her action
is a self-directed one.

There are a number of points at which this sort of claim about an autonomous
smoker meets resistance. Certainly she has some conﬂicting desires — the desire not to
smell like smoke, or the desire to not spend money on cigarettes, or the like. This is one
form disunity takes: conﬂict between ﬁrst-order desires.l4 In autonomous action these
dissenting desires might well exist (they often do), but are silenced; these are desires a

person (if honest, and sufﬁciently self-aware) admits herself to have, but are not desires

 

‘4 Frankfurt (1988), pp. 164-165.

117

by which she wants to be moved. Conﬂict is absent since the desires that are not in
accord with the volitions on which she acts are just that: desires, and not volitions.15 She
possesses them, but since she does not want them to motivate her action, there is no
conﬂict.

It might be objected that smoking is a peculiar example to use to describe
autonomy since people are hooked — smoking is (to limit the range of types of smokers
we are considering) an addiction. There are at least two possible issues here, one relating
to freedom of action, and the other to the status of the desire that motivates the action.
Regarding freedom of action: to the extent that a person is addicted, she is not free to act
otherwise — the phrase ‘freedom from addiction’ is almost a mantra in some circles. But
freedom (i.e. freedom of action, not metaphysical freedom — the freedom of an unmoved
mover) isn’t the issue — you need autonomy to be free, but not vice versa. Autonomy is a
precondition of freedom of action since if your act is not autonomous, it is heteronomous,
controlled by someone or something other than yourself. To be controlled by another is
one way in which a person can be unfree.

But the sort of freedom that a satisﬁed smoker lacks is not necessary for
autonomy. When Martin Luther, asked to recant the claims contained in his writings,
declared, “Hier stehe ich; ich kann nicht anders” (‘Here I stand; I can do no other’), he
was not refusing to recant, but stating the impossibility (not metaphysical, but agential,
impossibility, of course) of his recanting. He was unfree, but not thereby non-

autonomous. '6 He was expressing his most deeply-held convictions - convictions which,

 

'5 See, above, the distinction between desires and volitions.

‘6 Not unlike the smoker in the grips of her desires, Luther admitted to being “captive to the Word of God.”
The distinction here is that Luther was assumedly completely satisﬁed with his lack of freedom to do
otherwise.

118

because they are his, guide his autonomous actions, but convictions that were so deeply
rooted that to act against them was impossible. Likewise, then, the fact (if it is one) that a
smoker is not free not to smoke does not itself entail that she is non-autonomous. So long
as the desires with which she identiﬁes guide her actions, she acts autonomously.

The second abovementioned issue regarding addiction, one that helps even further
to explicate Frankfurt’s view, is that an addict has the desire whether she wants it or not.
The foregoing discussion entails only that some (unspeciﬁed subset of) addicts act
autonomously when they are moved by the desires characteristic of their addiction. Now,
when addiction takes the form of an inner conﬂict, the conﬂicted person is not fully
identiﬁed with either (or any, where there are more than two) of the desires that are in
conflict.[7 This is why, consistent with widely-held intuition, this sort of addict - the one
who does not identify with the urges that move her — is not autonomous. But when an
addict is satisﬁed with her desire, she acts autonomously when she is moved by it, even
though, like Luther, she can do no other. This sort of addict, like Luther, is not free to do
otherwise, though she is not moved against her will; her own will moves her, and so she
is autonomous.

On Frankfurt’s view, then, whether an action is autonomous is judged strictly on
the grounds of the structural relation between the ﬁrst-order desires by which she is
motivated and relevant second-order desires that may conﬂict with those of the ﬁrst
order. The necessary structural relation is one marked by a lack of conﬂict — that is, as

outlined in Chapter 2 above, a uniﬁed one.

 

'7 ‘Fully’ might seem to imply that ‘identiﬁcation’ is a scalar notion. On Frankfurt’s view, ‘fully’ alludes
to the degree to which conﬂict must be absent — namely, fully absent - regarding some desire in order for
that desire to be one with which the person is identiﬁed. There are, on this view, degrees of conﬂict, but
identiﬁcation (and hence autonomy) requires that this precisely speciﬁed bar be reached (“Identiﬁcation
and Wholeheartedness”, Frankfurt, 1988, pp. 159-176).

119

C ontent-speciﬁc accounts
To lead a bit obliquely into the discussion of content-speciﬁc accounts of
autonomy, the locus classicus of what is sometimes called ‘inﬁltrated consciousness’ is

an example from Thomas Hill.‘8

His example of the deferential wife is of a person who
has fully accepted social norms of feminine subordination. Fully and wholeheartedly
devoted to her husband, she actively subordinates herself to him, gaining her happiness
and her self-identity from her servility, and all with her own full endorsement. She forms
no independent desires of her own, desiring only to please her husband, and when she
satisﬁes desires of his, she feels happy and fulﬁlled. Her consciousness is inﬁltrated
(though Hill does not use these terms, he is certainly referring to the same phenomenon)
in virtue of her not only being subject to the norms of servility, but having integrated
them into her own stable, conﬂict-free internal economy.

Hill’s example appeals to the intuition that even though the deferential wife
considers herself happy and fulﬁlled, acts only on her own desires and is not forced
against her will into subservience, something is amiss. The ex hypothesi claims that she
is free, happy, and in control of her own life, it is argued, cannot be true. Not often noted
by secondary authors, but relevant here, is the fact that Hill himself used the example to
further argue that the deferential wife violates her duty of self-respect. Her fault is a
moral one (though this is not to say that she is blameworthy), stemming from her having

a certain morally wrong “attitude concerning one’s rightful place in a moral

community.”'9 The deferential wife fails to fulﬁll her duty to maintain self-respect.

 

‘8 Hill (1973).
'9 Hill (1973). p. 90.

120

Such a duty-based interpretation has apparently failed to appeal to a number of
philosophers offering alternative readings. The ﬁrst to be discussed here is Susan
Babbitt, who argues that oppressed persons, like the deferential wife, who accept and
endorse inaccurate claims of their subordinate status fail to possess an adequate self-
conception.20 It is because of this adequate self-conception that persons in similar
positions are not autonomous, in spite of certain understandings of autonomy on which
they may be seen to be so.

The argued lack of autonomy the deferential wife exhibits is not a matter of a
failure of self-govemance. The “liberal view’”' of autonomy that Babbitt positions
herself against has trouble making sense of the lack of autonomy in cases of inﬁltrated
consciousness. Babbitt presents the liberal view of autonomy as a view of practical
rationality, on which someone acts rationally (i.e. autonomously) “when she weighs her
options in light of relevant information and does or chooses that which is most likely to
advance more of her aims than other options.”22 Further, on the liberal view, an agent’s
interests are determined by this same process — a person’s interests lie in satisfying her
aims, fulﬁlling her desires, pursuing her goals. On the liberal view, Babbitt claims, an
autonomous act is one that leads to the satisfaction of one’s aims, and so it is in a

person’s interests to perform actions that best satisfy those aims.

 

2" Though similar to Hill’s interpretation of the issue as being one of self-regarding attitude, contemporary
moral-psychological and moral-sociological interpretations of the deferential wife diverge sharply from the
original deontological focus.

2' Babbitt (1996), p. 44. Babbitt notes that this view is not aptly so named in virtue of its being advanced
by liberals, but rather because it takes the ﬁrst—person perspective as the favored one. Though her target is
ostensibly John Rawls, it is not clear that this is Rawls’s conception of autonomy as “acting from principles
that we would consent to as free and equal rational beings” (Rawls (1971), p. 516). Rawls’ conception
seems to imply an ideal self view not entirely unlike Babbitt’s, though I won’t pursue this criticism of
Babbitt’s argument here.

22 Babbitt (1996), p. 38.

121

The important feature of such a view of autonomous action is its instrumentalism.
In being instrumental, the process is evaluated by the results relative to the initial aims of
the agent; whatever aims the agent perceives herself to have. The instrumental focus also
implicitly sidesteps the procedure of determining just what a person’s interests are;
interests seem simply self-evident. The process of autonomous action is one of
information processing, of gathering relevant information and determining, largely
empirically, what action is most likely to help satisfy one’s aims. Thus the view implies
that the primary barrier to autonomy is a lack of information, a deﬁciency of factual
knowledge.

Babbitt points out that the deferential wife is, on the liberal view, autonomous on
all counts. Her desires are indeed well satisﬁed by actions that are directed toward the
satisfaction of her husband’s desires, and so the instrumental feature of autonomy is
fulﬁlled. Nor would any factual information need be thought to change the actions or
desires of the deferential wife. Hill puts this point well: “She readily responds to appeals
from Women’s Liberation that she agrees that women are mentally and physically equal,
if not superior, to men. She just believes that the proper role for a woman is to serve her
family.”23

The intuition that something is wrong with the situation of the deferential wife is
perhaps best brought out by Babbitt’s appeal to ‘interests’. The deferential wife is just a

person who is supposed not to choose actions that lead to the fulﬁllment of any interests

relevant to a good human life. Whether she knows it or not, Babbitt points out, the

 

23 Hill (1973). p. 89.

122

deferential wife has an interest in ﬂourishing, in self-respect and dignity.M But pursuing
these interests and satisfying the desires she actually has are not coterrninous

Thus the main shortcoming of the liberal view of autonomy, as Babbitt sees it, is
that a person’s interests are not always self-known, because she may have desires that
even under conditions of full factual knowledge and capacity for instrumental reasoning
might not lead her to actions that are in her real interests. A person’s real interests may
not follow from her current desires by any path.

Autonomous action is to be judged by what “would constitute autonomous action
if conditions were such that it were possible for her to deliberate as a full human being.”25
These conditions, for the deferential wife, are a change in her social environment, and -
following from this — a change in herself. Her oppressive social environment is the root
cause of her inadequate self-understanding: she has been inculcated with norms of
subordination and with the abilities (how to manage her appearance, how to walk, how to
talk, how to think) that help her to obey these norms. But even more importantly for her
to be autonomous, Babbitt argues, is a change in who she is, in who she takes herself to
be. The deferential wife and those in similar situations need to conceive of themselves as
someone whose proper role is to pursue their own ﬂourishing, not that of others to whom
they are submissive.

Thus, on this account of autonomy, some actions are non-autonomous no matter
who performs them. Servile actions, self-harmful actions, actions that undermine or are
inconsistent with one’s own self-respect are, by deﬁnition, not autonomous actions. It is

this implication that makes Babbitt’s a content-speciﬁc account.

 

2‘ Babbitt (1996), p. 14 and throughout.
25 Babbitt ( 1996), p. 6. Italics in original.

123

There are two main concerns that show Babbitt’s conception of autonomy and her
defense of it to be untenable. One is that whatever the merits and accuracy of the appeal
to cases of oppression in addressing issues of autonomy, to deﬁne autonomy in terms of
resistance to oppressive forces, forces that prevent certain persons’ ﬂourishing, has
yielded an excessively narrow conception of autonomy. This can be seen through a
number of familiar counterexamples that undermine Babbitt’s view and show it to fail to
account for autonomous action generally, as I will go on to explain. The second concern
is perhaps even more troubling. The claim that a person’s interests might well be better
determined by someone other than herself is not necessarily itself problematic, but when
this claim is taken to support the notion that a person could be more autonomous when
acting from interests only others realize, the possibility of real personal damage is close
at hand. I will address these issues in turn.

Babbitt’s analysis of the damage caused by oppression to the capacity of persons
to act in ways that promote their own overall ﬂourishing is comprehensive. She
establishes about those persons in situations similar to that of the deferential wife that
“desiring the kind of self-determination we might think characteristic of a good life
would depend on their undergoing a change to their actual selves — a change to their
sense of themselves as well as a change to the actual interests and commitments they
possess.”26 That is, many victims of oppression would live a good life only if they acquire
different goals, if the (self-)direction of their lives were to change in fundamental ways.

But the self-interest criterion for autonomy is too narrow. It may well be that one

reason that some Oppressed persons are not able to act autonomously is that they are

 

2° Babbitt (1996). p. 45.

124

prevented from acting in ways that promote their own self-interest. But this is not to
agree that autonomy generally is a matter of acting thus. Consider actions that are
altruistic in the strong sense of the term. While we do call altruistic actions that beneﬁt
persons other than the agent, altruism in the strongest sense characterizes an action that
beneﬁts another person while inﬂicting costs on the agent. Such self-sacriﬁce would
necessarily be non-autonomous on any account of autonomy that requires the promotion
of self-interest (though actions that are altruistic in the weaker sense could be
autonomous). Acts of straightforward self-sacriﬁce — parents who give up their own
goals and instead promote those of their children, people who help strangers with no
expectation of reciprocation, acting morally without regard to one’s own desires -— would
be judged to be non-autonomous following Babbitt’s demand that autonomous action
beneﬁts oneself.27

Many of the paradigm examples of autonomy that Babbitt offers support the
foregoing claims. She makes repeated references to the character Sethe in Toni
Morrison’s novel Beloved as an exemplar of autonomous action; Sethe kills her children
rather than allow them to be returned to slavery. What Babbitt ﬁnds important (important
enough to quote twice, and refer to numerous times) is not that the children’s death was
merciful, but that Sethe took it to be a self-interested act: “It was a kind of selﬁshness I
never knew nothing about before. It felt good. Good and right.”28 Indeed, most of the

proffered examples of autonomous action involve harm to others: Ridley Scott’s Thelma

 

27 Though there is no indication that this response is one Babbitt might offer, it can be noted that these
examples are not, as they are in other contexts, subject to a egoist interpretation on which an apparently
self-hannful action can in fact beneﬁt the agent through her desire e. g. to save the stranger; there may be,
on the strongest version of this view, no altruistic actions at all. But this response is inconsistent with
Babbitt’s viewpoint since it embraces the so-called liberal view that Babbitt has positioned herself against;
the egoist response, that is, appeals explicitly to desire-maximization in its interpretation of supposedly
altruistic actions.

2” Babbitt (1996), p. 22 and again on p. 27 .

125

and Louise is taken to be about “the struggle for autonomy”29; the central scene in the

ﬁlm involves Louise’s shooting to death Thelma’s rapist (out of a sort of vicarious
vengeance, not because Thelma needed to be defended). Marleen Gorris’s A Question of
Silence is “the story of three women — a secretary, a waitress, and a housewife — who
come together in a dress shop and (gruesomely) murder the dress-shop owner, a man they
have never seen before.”30 Babbitt considers their action “reasonable’”' and takes this
situation to be “signiﬁcant as regards the individuals’ moral responsibilities to maintain
their own sense of self-respect and dignity”, that is, the objects of autonomous action.32
My attribution of a self-interest view of autonomy to Babbitt is not simply a
matter of interpreting her paradigm cases of autonomy thusly. Babbitt explicitly deﬁnes
‘autonomy’ in terms of self-interest. “Autonomy is a special kind of agency, the capacity
to take control of one’s life and to realize one ’s best interests.”33 She writes of the
preconditions required “to act autonomously, to be in control of their lives, where control
is the realization of best interests.”34
But surely, contrary to both Babbitt’s strongly suggestive examples as well as her
explicit demand that autonomy essentially promote one’s own interests, actions that
beneﬁt others while bringing harm to oneself can be autonomous. Indeed, a common and
plausible assumption is that actions that proceed in spite of contrary self-beneﬁting

desires, desires which are suspected of being the source of much heteronomous action,

are likely candidates for the label of autonomy.

 

2" Babbitt (1996), p. 88.

3° Babbitt (1996), p. 112.

3' Babbitt (1996), p. 121.

32 Babbitt (1996), p. 112.

33 Babbitt (1996), p. 45. Italics added.
3" Babbitt (1996), p. 105. Italics added.

126

There are two responses that could be offered in defense of Babbitt’s arguments
regarding the nature of autonomy. One is that Babbitt is not attempting to defend a
conception of autonomy generally, but rather is arguing that there is a need in cases
where person ’s moral self-conceptions are damaged by oppression for just these persons
to attend to their self-interest in order to regain autonomy.

To be sure, all of the examples that Babbitt offers of autonomous action are
examples of distinctively oppressed persons in oppressive circumstances. But at least in
claims such as “control [of one’s life] is the realization of best interests”, and others
quoted above, she offers her arguments as applying to questions of autonomy generally.
Further, Babbitt offers no reason to think that a person who because of oppression fails to
realize her own interests differs from a person who for some other reason — moral
saintliness, everyday codependency, habitual self-deprecation, or some other non-
systematic cause — fails to aim her actions at her own self-interests. Indeed, in Hill’s
original essay, the Deferential Wife was not different in this way from a person he termed
the Self-Deprecator; each sort of person, Hill presumed, were in the same situation,
morally and agentially speaking. Babbitt offers no reason to think otherwise.

The second possible response is that it might be that what Babbitt has in mind is
not simply an action’s outcome as determining whether it is an autonomous one, but
rather the distinctly moral quality of the action — whether this moral quality is judged by
the effect of the action or by its grounds. Though such an argument need not take a

Kantian line, this is arguably what Kant had in mind regarding autonomy - that only

127

morally worthy acts are autonomous.35 Unfortunately, Babbitt here offers no explicit
argument as to why we ought to think that autonomy includes a moral element.
Diana Meyers (whose account will be examined more fully in the
following section) does offer such an argument in her account of procedural
autonomy. Appealing to the objective point of view that, it is often argued, we
take toward ourselves in quiet moments of self-evaluation as well as in defending
our choices, commitments and actions publicly, she writes,
People’s intense concern with their personal decisions, the complexity of
the processes whereby they ponder their options, and the ways in which
they sometimes criticize the choices they have made would be
inexplicable if personal autonomy were reducible to uncoerced choice.36
If what Babbitt calls ‘the liberal view’ were correct, then, our prospective
deliberation regarding, as well as our retrospective evaluation of, our choices
would be impossible to comprehend.
What Meyers seems to ignore is the multitude of evaluative criteria that do

”37, that is, on objectively

not, as she claims, depend on “genuinely good things
valuable considerations. We criticize our own choices, for example, when they
were guided by desires that were less strong or less subjectively important than
others that could have guided our choice. The complexity of decision processes
can be quite pronounced in clearly non-objective contexts - choosing a new pet
that is to be compatible with one’s own personality and tastes, or deciding on a

gift that will be appreciated by its recipient are both based essentially on nothing

“genuinely good” but rather on what can be just as elusive: the determination of

 

’5 This line of thought regarding autonomy is not pursued here, as I take it that the Kantian model has been
sufﬁciently undermined to preclude its viability.

3" Meyers (1989), p. 18.

’7 Meyers (1989), p. 18.

128

what is subjectively good. Lastly, people’s concern for their own decisions may
be due in part to their striving for moral rectitude, but such concern is surely at
least sometimes explicable in other terms: narcissism or even pedantry explain
such concern even if they do not justify it. Indeed, there seems little reason to
suspect that moral rigorism is at all common in daily life.38 There is much more
to decision and action than bare (instrumental) autonomy, certainly, but this does
not require that autonomy have a moral element.

Without a more satisfying argument for the inclusion of moral elements in the
notion of autonomy, the instrumental use of ‘autonomy’ (the concept) in attempting to
promote best interests remains the most plausible interpretation of Babbitt’s arguments.
This focus on the effects of one’s actions (the effect being promoting one’s best interests)
plays a role in the second, more troubling, reason to think Babbitt’s account untenable.
Valuable in identifying the details of this criticism is a passage from Babbitt, partially
cited above. “But to the extent that some people are in fact deprived of dignity and self-
respect in their actual lives, desiring the kind of self-determination we might think
characteristic of a good life would depend on their undergoing a change to their actual
selves — a change to their sense of themselves as well as a change to the actual interests
and commitments they possess.”9 Babbitt’s point here is that no amount of factual
information will be of help to certain oppressively socialized persons, since one thing that
is restricting their autonomy is “their actual selves”.

It is hard to doubt that the lives of many persons — all of us, likely — would not be

improved with certain changes to our actual selves. Many of us would be happier and

 

’8 And even less reason to think it desirable.
3° Babbitt (1996). p. 45.

129

morally improved, ultimately, were we coerced into certain changes in our preferences
and the associated actions. Some of these improvements may be so important as to
justify paternalistic intervention. To be sure, Babbitt devotes an entire chapter to the
denial of Will Kymlicka’s implausible claim that “no life goes better by being led from
the outside according to values the person doesn’t endorse”.4O

But on Babbitt’s view it is crucial here to recall that autonomy is essentially a
matter of a person’s pursuing her own best interests, and that one of the primary checks
on autonomous action by oppressed persons is an inadequate self-conception, one that
fails to represent their real value as persons. This conception of autonomy along with the
plausible claim that a person is not always (and some people are never) in a privileged
position to identify their own best interests as well as their personal value sums to the
peculiar claim that certain persons might be more autonomous were their lives led by
values they themselves don’t endorse.

Babbitt’s response is to argue that without active and possibly coercive
intervention into societal structures, it is impossible for certain persons even to make real
choices about their interests. Furthermore, such intervention is already ubiquitous: the
social structures already in place force certain understandings on persons, often
oppressively.

This response, though, does not go far enough to defend against the possibility of
(further) illegitimate intervention in other’s lives. On the liberal picture that Babbitt is
positioned against, the beneﬁts of paternalistic intervention in the life of a person is at

least countervailed against by the value of autonomy, of her living her own life in spite of

 

4° Babbitt (1996). p. 84.

130

the well-being she might garner by external intervention. On a view that holds personal
well-being to be conceptually equivalent to autonomy, there is no such guard against
intrusive intervention.

Thus Babbitt’s account of autonomy leaves open the possibility of political abuse
to an even greater degree than does the so-called liberal view. This is a real worry since
just what ‘human ﬂourishing’ entails is an open question. Historically, widely accepted
conceptions of ﬂourishing in the Western world have included adherence to a certain
elitist set of virtues, obedience to the church, commitment to scientiﬁc reasoning, and
membership in a master/slave hierarchy supposedly for the beneﬁt of members of both
groups. It is historically still the case (within mainstream American political and social
culture) that the dominant conception of human ﬂourishing is tied up with assumptions of
rugged individualism, heteronorrnativity and an economic model of human value that
would be even more undesirable were it enforced in the name of autonomy. (Nor even
need we restrict conceptions of ﬂourishing to those of the mainstream to ﬁnd such
examples — consider Michigan Congressperson Barbara-Rose Collins’s early 1990S
support of the Unremunerated Work Act, legislation that would have required the
ﬁnancial value of the unpaid work done primarily by women to be ﬁgured into the US
Gross Domestic Product — yet another economic model of human value.) Without a
substantive conception of ‘ﬂourishing’, which Babbitt does not offer (instead taking “the

”41

concept “human ﬂourishing” to require empirical investigation ), the recommendation

that ﬂourishing be imposed on persons is a dangerous one.

 

4' Babbitt (1996), p. 9.

131.

Meyers ’s procedural account

On procedural accounts, speciﬁable processes and the abilities required for one to
undertake these process are essential to autonomy. Such accounts vary according to the
nature of the deﬁning process — the most general distinction within procedural accounts is
between views that focus on the procedure by which a person comes to acquire the
desires in which her autonomous actions are grounded, and those that take autonomy
itself to be a procedure.

Diana Meyers argues most explicitly for the latter, that autonomy itself is best
understood through an analysis of the various procedures that constitute its practice, and
as such she defends a procedural account.“2 She takes autonomy to be a tripartite
phenomenon, including self-discovery, self-deﬁnition, and self-direction. Thus an
autonomous person knows who she is, is able to effect changes in who she is, and is able
to direct her life in accordance with who she is. Meyers understands each of these three
elements of autonomy essentially as processes underpinned by “autonomy competency”,
which refers to “the repertory of coordinated skills that make self-discovery, self-
deﬁnition, and self-direction possible.”43

The skills of autonomy competency are numerous and varied. Autonomous
persons are able to be resistant to external pressures, and, relatedly, have resolve — the
determination to act on their own judgments.44 Autonomy involves the evaluation of

one’s own desires, including the ranking of desires, the envisaging of alternative desires,

 

‘2 Though Meyers does in fact characterize autonomy as being inseparable from the processes she outlines,
and she does label her account as a procedural one, I will go on to show that there are central elements of
her account that overshadow the procedural criteria.

‘3 Meyers (1989), p. 76.

‘4 Meyers (1989), p. 83.

132

and the ability to imagine the outcomes of acting on various desires.“5 The autonomous
person is engaged in ongoing adjustments to her self-image as her actions dictate, and as
she fashions and adjusts her life plan."6

This last characterization, of a person involved in a dynamic relationship with a
self-constructed life plan — which Meyers explains to be “a largely schematic, partially
articulated vision of a worthwhile life that is suitable for a particular individual”,47 and,
somewhat differently, “a comprehensive projection of intent”48 — is also clearly one of
process. Working out such a plan, determining how to best act in accordance with it and
modifying it in the fact of change in desire, ability and external conditions are ongoing
tasks.

The notion of a life plan is a central element of Meyers’s account. For people
with life plans, the plans are also, Meyers argues, a “mirror [of] their authentic selves”,
since they are partially systematic and partially explicit articulations of one’s desires.49
Thus to act in accordance with one’s life plan is just to act in accordance with one’s
authentic self (alternately, “true self ’).5° As to the centrality of the authentic self in
understanding autonomy, Meyers claims, “personal autonomy requires a touchstone.
That touchstone is the unique authentic self.”SI

There are a fair number of demands Meyers makes on the autonomous person that
one can take issue with. One is the seemingly ad hoc demand that the authentic self be

unique. In a fairly trivial sense, every entity is unique — numerically unique — but Meyers

 

‘5 Meyers (1989), p. 81.
4" Meyers (1989), p. 84.
‘7 Meyers (1989), p. 51.
‘8 Meyers (1989), p. 49.
‘9 Meyers (1989), p. 50.
5° Meyers (1989), p. 75 and elsewhere.
5' Meyers (1989), p. 19.

has something more substantial in mind. “The leading phenomenon of personal
autonomy,” she writes, “is human diversity. Unless one is impressed by the uniqueness of
individuals, and unless one delights in this variegation, one will not suppose that these
unique personalities should somehow be manifest in people’s lives, and one will never
wonder how personal autonomy is possible?”2

This is likely a reaction to two concerns. One cause of this reactive demand is the
dominant Kantian understanding of autonomy as rational action grounded in universal
principles. In taking issue with such an understanding Meyers claims, “as rational
beings, people are all alike.”53 Dissatisfaction with such an absurd characterization of
autonomous persons motivates the uniqueness criterion. Another cause for the demand is
likely the existence of the sort of socialization that aims to instill certain desires
uniformly; oppressive socialization might be thought to be of this sort”4 But it is too
strong a demand to put on autonomous persons that they be unique. The degree to which
I share your desires, whether accidentally or even causally (some cases of empathy, for
example, involve shared desires), is not clearly related to the degree to which I am
autonomous.

As noted above, the demand that one’s authentic self be unique is not apparently

an integral part of the conception of autonomy Meyers defends. The traits and abilities

that characterize autonomy competency, though, are the constitutive elements of the

 

52 Meyers (1989), p. 82.

5’ Meyers (1989), p. 13.

5“ A fact easily overlooked is that the uniformity of desires is not itself importantly relevant to oppression,
but certain patterns of instilled desires certainly are. The fact that persons are oppressed in virtue of their
group membership entails that oppressive socialization will tend to aim at unifomrity of desires within
groups (so that, for example, all women will desire to be domestically skilled), but the oppression is a
function of the possession of the desire by women and not men.

134

account — and many of these traits and abilities are either not characteristic of people
generally, or not part of the authentic selves of many actual persons.

Those with autonomy competency are said to engage in candid communication
about their own concerns, have lively recall, vivid imagining, openmindedness, and
alertness to the reactions of others. ”These are skills with which everyone is acquainted
and with respect to which each of us knows roughly how proﬁcient she is.”55 That
autonomous persons must possess these qualities is an unrealistic demand, for two
reasons. The ﬁrst is that the claim that people generally are aware of the degree to which
they themselves possess these qualities is simply false. Even without empirical data it is
not difﬁcult to determine that most people overestimate some of their own abilities at
least occasionally. But data are not lacking. Numerous studies have shown that such
overestimation of one’s own abilities is ubiquitous.56

The second reason that such characteristics are not legitimate criteria for judging
autonomy — especially on an account that holds autonomous action to be that action
grounded in the authentic self - is that such characteristics simply are not part of
everyone’s authentic self. Meyers claims that

chronic obliviousness to self-referential responses, awkwardness or

rigidity in envisaging and appraising options, uncommunicativeness about

one’s needs, desires, values, and so forth, imperviousness to others’

feedback, timidity about acting on the basis of one’s own deliberations,

and obstinate inﬂexibility in executing a chosen plan [. . .] signal the
individual’s inability to fathom his or her authentic self”.57

 

5" Meyers (1989), pp. 87-88.

5" Though the most thorough attempt to determine the extent of personal overestimation is in medical
contexts (Ward, et al., 2002), social psychologists are indeed interested in the phenomenon in the general
population (DeAngelis, 2003).

57 Meyers (1989), p. 88.

135

Thus to take oneself to be uncommunicative, impervious, or timid in these ways is to
misunderstand one’s own true self. But this demand entails that there is no one who in
fact unimaginative, uncommunicative and unobservant, a claim that excludes many
persons.

This may seem to be a misguided interpretation of Meyers’s claims. It would
seem unlikely that that she is denying anyone an authentic self. The plainness with which
this claim is stated, though, makes other interpretations difﬁcult: “An authentic self is a
self that has autonomy competency and that emerges through the exercise of this
competency.”58 If an authentic self has autonomy competency, then any self without
autonomy is not an authentic self. Since it is agreed by all parties to the debate over
autonomy that there are some persons who lack it, on Meyers’s account there are some
persons who lack an authentic self. That suggests that there is no person who they truly
are.

What Meyers is more plausibly interpreted as seeking in the notion of the
authentic self is better termed the ‘ideal self’. A bit of textual exegesis is required to
substantiate this claim, and so I will end the discussion of Meyers’s account by doing just
that.

As noted above, the concern surrounding the authentic self is grounded in the
worry that a person can fail to know who she really is, and thus fail to act as she really
wants to. Meyers makes this claim in many different ways, and in varied contexts and
arguments. At one point, she puts it this way:

Autonomous people must be disposed to consult their selves, and
they must be equipped to do so. More speciﬁcally, they must be able to

 

5" Meyers (1989), p. 53.

136

pose and answer the question “What do I really want, need, care about,

value, etcetera?”; they must be able to act on the answer; and they must be

able to correct themselves when they get the answer wrong.”

This is a familiar sounding passage, but focusing on the last portion of it will
likely reveal something novel. What is it to get the answer — to the question ‘what do I
really want, need, care about, value’ — wrong? It could be that a wrong answer is one that
is personally mistaken — as when I have a feeling that makes me think I need sleep, when
in fact I need to drink some water, or (somewhat differently) when a person buys a new
car thinking she wants more creature comfort in her life, when in fact she wants envious
attention from others. In these sorts of cases the error regards determining one’s actual
internal state (thirst, vanity). The mistake in these sorts of cases is epistemic.

Another way in which a person might be wrong about what she really wants,
values, and so on, is that she could in fact want and value the wrong things. This is the
sense at hand in the judgment of the amateur psychologist who claims that what you
really need is love, and thus your need for money (though descriptively true of you) is
wrong, morally. This is the sense in which valuing the sorts of things that Klan members
characteristically value is morally wrong.

This bipartite interpretation of Meyers’s view is supported by claims that imply
both an element of self-govemance and a moral element to autonomy. “Any ostensible
case of autonomy could be debated. Was the crucial choice the right one to make, or not?
Was the agent really in control of his or her life, or not?”60

As we saw in the previous section, Meyers offers an argument in support of the

claim that there are moral criteria involved in the judgment of whether an act was

 

5” Meyers (1989), p. 52.
6° Meyers (1989), p. 9.

137

autonomous. If autonomy is grounded in the authentic self — the “touchstone” of
autonomy — then it would seem there is a self that a person ought to have, namely, the
self from which autonomous actions ﬂow. Meyers does argue, in a line of thought quite
similar to one pursued by Lindemann (discussed in chapter 2, above), that persons
possess identity elements that they would rather not — disparaging, oppressive, and
simply undesired third-personal attributions. She describes the ways in which persons
can and do go about reconﬁguring their own identities — by attempting to make one’s
actions cohere with her aspirations, reconciling her speciﬁc goals with her group
memberships, and so on - so as to engage in the process of self-conﬁguration of authentic
selves.

But here, too, not only is the authenticity necessary for autonomy a matter of
discovering (or ‘deﬁning’; Meyers uses both terms) one’s self by acting on the desires
and values that one has, it is also a matter of determining which “desires, personal traits,
values, interests and aims shouldl seek to enact”."l That is to say, Meyers’s notion of
autonomy involves not acting on the desires I have so much as acting on the desires I
ought to have. In this way, the desires with which a person in fact identiﬁes might well
be less germane to her own autonomy than desires with which she (morally) ought to
identify. The authentic self is thus better understood as an ideal self — ideal, that is, not
merely in some normative sense or other, but rather ideal regarding the possibility of
acting autonomously.

Though Meyers has named her own account ‘procedural’, it should be apparent

from the foregoing explication that the most prominent features of the account have less

 

6' Meyers (2000), p. 166; emphasis added.

138

to do with its procedural features than with the notion of the authentic self that plays so
central a role in autonomy. Although the next part of the discussion afﬁxes the
‘procedural’ label to accounts that were not so named by their original authors, the
accounts of autonomy from Paul Benson and John Christman illustrate more centrally
important procedural requirements on autonomous action.

Procedural responses to the structural model

Both of the accounts of autonomy, from Paul Benson and John Christman, to be
discussed in what follows are largely responses to the structural model of autonomy, as
defended by Harry Frankfurt. (in what follows, ‘the structural model’)."2 On this model,
a person is autonomous just in case she identiﬁes with her motivating desire. What it is
to identify with a desire is less simple, as discussed above, in Chapter 2, but recall the
basic outline of ‘identiﬁcation’: to identify with a desire is just to endorse it, where to
endorse a desire entails the possession of no second-order desire that conﬂicts with it. So
someone who acts on a desire she wants to have (more speciﬁcally but less
straightforwardly, since the criterion is a negative one, ‘a desire about which it is false
that she wants not to have it’) has thereby acted autonomously.

Both Benson and Christman are unsatisﬁed with the structural model for similar
reasons as those which motivated criticisms from Babbitt and Meyers. Each argues that
Frankfurt’s hierarchical account of identity is unacceptable since it does not account for
the importance of the manner in which one comes to possess her hi gher-order desires.

On the structural account, a person who comes to possess desires through coercive or

 

62 In this section, like in the discussion of the structural model in Chapter 1, I will appeal most often to
Frankfurt’s claims, rather than those of Gerald Dworkin, primarily because of Frankfurt’s more extensive
writings on the subject. Since Benson criticizes only Frankfurt’s version of the model, and Christman takes
the two versions to be “essentially the same” (1991, p. 5), this focus on Frankfurt is legitimate.

139

manipulative processes and endorses them on the basis of similarly-acquired second-
order desires is autonomous when she is motivated by these desires. This is so since
autonomy is judged simply on whether the endorsement relationship holds, not on the
basis of the way in which this relationship came about.

Christman recommends adding to the hierarchical model the requirement that a
person must have the self-reﬂective ability to approve of the process by which she
acquired her desire. This ability must be self-transparent, non-self—deceptive and
minimally rational. Benson argues that hi gher—order endorsement — the core of the
structural model — is alone neither necessary nor sufﬁcient for autonomy. He proposes
adding to the model a condition of critical competence regarding action — that a person
need to be able to discern the reasons that there are for her to act. I will discuss each of
these proposals in turn.

C hristman ’3 historical criteria

A clear formulation of Christman’s principal claim is “that the central focus of
autonomy must make particular reference to the processes of preference formation, in
particular what makes them ‘manipulative’ in a way crucially different from ‘normal’
processes of self-development.”63 Thus, Christman realizes that which was discussed at
some length above: simply because a desire is manipulatively or coercively instilled in a
person is not sufﬁcient to ground a charge of illegitimacy; such inculcations are
ubiquitous and permissible in early education, for instance. The illegitimate sorts of
manipulations of desires might come about as the result of brainwashing, or oppressive

socialization of the sort involved in Hill’s example of the deferential wife.

 

6" Christman (1991), p. 10.

140

Consistent with Frankfurt’s account is the possibility “that a thoroughly
manipulated individual could be considered autonomous.” Christman asks us to
“[i]magine someone who has been secretly hypnotized to want strawberries. . .But such a
person is surely not autonomous relative to this desire for strawberries”64 This last claim,
employing the vague phrase ‘relative 10’, could mean either or both of two things. It
could mean that her acting on the desire for strawberries (by, e. g. eating one) is not an
autonomous action. It could also mean that there is something non-autonomous about her
very having of the desire.

The former interpretation is just the sort of claim at issue in the present chapter,
and as such it calls for no direct response here. The latter reading, involving desires that
are themselves autonomous, is a claim that was discussed above, in Chapter 1. There I
argued that only actions, and not desires, are said to be autonomous. Given this further
explication through an example of the phenomenon at issue, we might more clearly
understand Christman’s point. Rather than (as ‘autonomous’ suggests) understanding
desires ideally to be self-governed, or self-motivated, we might see the point about
hypnotically-acquired desires to be that they are inaccessible to self-alteration. Once a
person submits to (or is otherwise subjected to) hypnosis, she can neither eliminate nor
alter the desires she comes thus to possess. I assume (having no experience with
hypnosis) that the post-hypnotic subject feels no internal conﬂict regarding the implanted
desires.

But, as Christman acknowledges, this inalterability is true of many of our desires,

including some of those that are most central to our self-conceptions. Such desires are

 

6" Christman (1991), pp. 8-9.

141

often “too much entrenched”"5 into one’s self-conception to be subject to alteration by
any but the most radical (and, notably, non-self-directed) upheavals. In agreement with
many of the this work’s readers, Christman admits that “I would not say that my desire to
study philosophical questions, which for me has this character, is not autonomous.”66
Thus, the mere ability to alter one’s desires is not at the heart of Christman’s concern
regarding autonomy.
Christman’s view aims to “eliminate the need for the condition of identiﬁcation
(as such) altogether.” 67 Irrelevant to determining autonomy is agent’s evaluation of her
desire at a particular time, Christman claims. What matters is what the agent thinks
about the process of coming to have the desire. Thus, a person is autonomous just in case
she acts on a desire that she acquired through some process of which she approves.
Speciﬁcally, the criteria Christman proposes are these:
(i) A person P is autonomous relative to some desire D if it is the case that P
did not resist the development of D when attending to this process of
development, or P would not have resisted that development had P
attended to the process;
(ii) The lack of resistance to the development of D did not take place (or
would not have) under the inﬂuence of factors that inhibit self-reﬂection;

and

(iii) The self-reﬂection involved in condition (i) is (minimally) rational and
involves no self-deception?"

Among the problems with such a view is its silence on cases in which a person
acts (intuitively, autonomously) on some desire she possesses in virtue of some

unapproved-of process. Consider a philosopher who realizes that argument from

 

‘55 Christman (1991), p. 6.
6" Christman (1991), p. 6.
‘7 Christman (1991), p. 10.
‘8 Christman (1991), p.11.

142

authority is insufﬁcient grounds on which to endorse a desire. Nonetheless, at one time
she acquired a desire to drink Diet Pepsi because her then—idol, Ray Charles, promoted
the beverage in advertisements. Does she now act autonomously when she drinks Pepsi-
Cola to satisfy her desire for a sweet-tasting, caffeinated, carbonated beverage? I think
she does. Christman’s view might allow for the intuitive interpretation of this case if we
stipulate that the desire she now has to drink Diet Pepsi is a different desire than the (non-
autonomy-grounding) desire she had to drink Diet Pepsi when she was watching cola
advertisements. Other than for the support it would lend to Christman’s View, though,
there seems no reason to think the two desires are distinct.

It might be replied that the imagined case violates none of the criteria. It does not
violate the most obvious, (iii) the notion of ‘minimal rationality’, since Christman takes
this merely to be reasoning that “does not contain manifest contradictions?"9 It does
violate (i) if it is allowed that the person as she is now is the one who does not
retrospectively resist the development of the desire; the now-philosopher now knows that
celebrity endorsements are poor grounds for desires. This reading of (i), though, cannot
be the reading at issue, since then it fails to accommodate the sort of cases Christman has
in mind, e. g. brainwashing. The now brainwashed person does not retrospectively resist
the process of her brainwashing, just as Paul does not resist the process of his conversion
(though Saul might have)?" If (i) refers only to the person that she was at the time of the

desire-acquisition, then this precludes many cases of retrospective disapproval — once a

 

6” Christman (1991), p. 13.

7° Acts 22. “I could not see for the glory of that light” (22:11) Paul recounts at least partially
metaphorically, and he admits that “even while I prayed in the temple, 1 was in a trance” (22: 17). Paul
became blind and unaware, and thus is in a situation similar to that of the hypnotized or oppressively
socialized person.

143

person has approved of the process, she identiﬁes with the desires that it gives her ad
inﬁnitum. To disallow future reevaluations of one’s identity is certainly incorrect.

There are further problems involving the endorsement of a desire’s formative
process. Whether the self-evaluation of the process is made by the person as she is
before she has acquired the desire in question or after, the evaluation itself is inextricably
dependent on more than rationality, reﬂective ability and freedom from self-deception.
Self-evaluation is essentially dependent on the content of one’s desires (and also her
beliefs, and fears, and so on).

The example of the deferential wife shows that rationality, reﬂection and freedom
from self-deception are not sufﬁcient to ground the sort of self-evaluation that would
enable her to repudiate her putatively illegitimate desires." She is minimally rational, in
Christman’s sense of being free from manifest contradiction. There is no reason to doubt
that she is self-reﬂective; “No one is trampling on her rights, she says,” Hill imagines,
“for she is quite glad, and proud, to serve her husband as she does?72 That is, the
deferential wife responds to the claims of others that her rights are being violated; she
looks inward to see whether she feels this way, and ﬁnds she does not - in fact she ﬁnds a
feeling of pride. (Indeed, the force of the example depends on her reﬂecting and, having
done so, ﬁnding nothing at all amiss.) She is, though not as obviously, free from self-
deception as well. She is clearly deceived, but not by herself. She herself is indeed in

error (regarding her moral beliefs, or her empirical beliefs, or the source of her personal

 

7' Two points bear mention regarding this claim. The deprecating tone of ‘putatively illegitimate’ here
refers not to the undesirability of the deferential wife’s desires (with which I completely agree), but rather
to the sense of ‘illegitimate’ at issue here - namely, ‘not truly her own.’ Also, the claim consciously
brackets consideration of those clearly external forces (e. g. the threat of physical harm) that prevent persons
from repudiating certain of their desires; that is to say, rationality, reﬂection and freedom from self-
deception are never, strictly speaking, sufﬁcient (without certain external facts obtaining) to self—
evaluation.

72 Hill (1973), p. 89.

144

value, depending on the commentator), but deception essentially involves a connection
with an accepted truth from which other beliefs or actions deviate.73 Thus the deferential
wife meets the proposed criteria for autonomy.

The example of the deferential wife cannot be generalized to all cases in which
we aim to determine autonomy, of course. For example, many cases of less-
comprehensive oppressive socialization clearly involve a degree of self-deception; a
victim of intimate abuse might deceive herself into believing that she deserves the abuse,
in order to construct some explanation, however implausible, of her experience. But
Christman’s account entails other implausible judgments. On his view there could be a
person who acts autonomously when she is motivated by some desire she acquired
through a process of which she approves but which in virtue of its content she disdains
and repudiates. What Christman is appealing to is not the general sort of process by
which a person comes to have a desire (e. g. my parent’s benevolent paternalism) but
rather speciﬁc instances of processes (e. g. the instance of benevolent parental paternalism
in virtue of which I acquired my desire to study science in the interest of landing a good-
paying job upon graduation). I may, indeed, come to think that this instance of my
parents’ guidance undesirable; I may think that studying with future pecuniary reward in
mind is tantamount to squandering an education. But in this case, then, the content of the
desire is doing all the work of determining which processes yield desires with which I

will identify. Thus, the subjective judging of desires on the grounds of their genesis is

 

73 This is why ‘lying’ is not in principle inconsistent with ‘truth-telling’ - a person who is in error regarding
the truth value of some proposition, and who also lies in regards to it, ultimately conveys the truth.
Knowing that you are being lied to — but being unaware of the state of the liar’s knowledge — alone tells
you nothing about the truth value of the deceptively-intended claims; thus deception is an intentional
concept, not an epistemic one.

145

either patently false (‘1 got this benighted desire from my benevolent parents, so it must
be mine’) or superﬂuous, since it depends essentially on the content of the desire.

What Christman seems to be aiming for is a view on which desires are subject to
more evaluation than mere endorsement in order for them to ground autonomous actions.
Of course, ‘endorsement’ does not (contrary to Christman’s implication) exclude
‘evaluation.’ In either approving of or accepting the various elements of one’s identity
there is (in some circumstances more than others) a critical consideration of the value of
the element, the degree to which it coheres to other elements with which one identiﬁes,
and so on.74 But any such evaluation is limited, Christman realizes, to judgment on the
basis of the desires, beliefs, and the like, that one already endorses. Thus while
attempting to elucidate a manner of self-evaluation that is not essentially dependent on
that which is being evaluated is an understandable sort of project, Christman’s suggestion
does not helpfully further that project, for the reasons laid out above.

Benson ’5 critical competence condition

Paul Benson has much the same goal as does Christman, and a proposal that
shares elements with both Christman’s and Susan Babbitt’s View.” He aims to account
for cases of oppressive socialization like that of the deferential wife. Women who are
socially coerced into shaping their appearances so as to be visually pleasing to men often
come to internalize the judgments that failing to maintain their appearances for men’s

pleasure is a personal failing. This judgment, of course, is untrue, and so he claims such

 

7" Even in situations of what we might call ‘endorsement by silence’ (endorsement by refusing or failing
critically to evaluate), the decision to refuse to evaluate is itself a critical evaluation of a higher order (since
in order to make such a decision one must have considered whether further evaluation would be
worthwhile). Cases in which one simply fails to evaluate — in which a person accepts a desire of hers out of
disinterest in or ignorance of her possession of it — involves her not considering the desire worthy of her
attention, which is itself an evaluative stance.

7’ In Benson (1991).

146

socialization “systematically lead[s] women to internalize false construals of their
personal value and, in consequence, to misconstrue many of the reasons there are for
them to act?“

Like Christman, Benson takes autonomous action to be grounded in desires and
other motives that are subject to more than hi gher-order endorsement. Like Babbitt, he
argues for the inclusion of a content-speciﬁc epistemic element in the notion of
autonomy. Babbitt’s requirement, recall, is that autonomous actions aim at the actor’s
best interests, essentially involving the autonomous person’s adequate self-conception
that leads her to knowledge of what her best interests are.77 Benson describes
oppressively socialized women as lacking correct beliefs about what constitutes their own
personal worth. This epistemic error leads, as noted in the cited passage above, to failure
to act on appropriate reasons.

Most basically, Benson’s dissatisfaction with structural accounts lies in their
restricted characterization of ‘autonomy’. “These views assume that critical reﬂection
facilitates autonomy mainly in virtue of its efﬁcacy in the motivation of an agent’s action,
regardless of whether or not the agent is a sufﬁciently competent criticizer to be able to
recognize and appreciate for herself the reasons there are for her to act in various ways.”78
Thus structural views, the claim runs, allow all self-directed actions to count as
autonomous, whether or not they are the actions of a sufﬁciently competent criticizer.

It is worth noting that simple epistemic error does not itself entail a loss of

autonomy. If it did, then no one, or radically fewer people than we tend to think, has

 

7" Benson (1991), p. 389.

77 This is a central element of Babbitt’s notion of “personal transformations” by which one can acquire
“beliefs that make possible an alterative interpretation of his society.” (p. 3)

7'3 Benson (1991), p. 396.

147

been autonomous: we just didn't know (and probably still don't know) just what our own
individual or more broadly human ﬂourishing entails. Suppose we come to realize, not
entirely implausibly, that human ﬂourishing requires maintaining a strong connection to
the natural world, through working to preserve a pre-human environment and seeking to
understand ourselves primarily as part of the natural (as opposed to, say, the more
broadly moral) world, and the like. It would be quite odd to think that this fact would
entail that all of us who had failed to live in this natural fashion — though they might have
been failing to ﬂourish, or may have even been living morally badly in some broader
sense — were not autonomous.

Benson’s claim is, fortunately, not so categorical. “Persons do not suffer reduced
autonomy merely because they have false beliefs about what they should do, or because
they act contrary to what there are the best reasons for them to do?” Autonomy is
preserved, Benson argues, in at least two sorts of situations: (1) in which the person
“autonomously decides to curtail rational consideration of certain of his motives or
available lines of action”, and (2) in which the person, though having exercised critical
competence, simply makes a mistaken judgment.80 By way of agreement, I will pass over
(1) and focus on (2), especially on the notion of ‘critical competence’. This notion, I’ll
go on to show, is problematic in at least two ways:

The critical competence necessary for autonomy is an ability to reﬂect not merely
on one’s higher-order preferences, though Benson’s view requires this ability as well; at
least at many points in the discussion, he poses his recommendations as additions to

Frankfurt’s view. The critical reﬂection that is the function of hi gher-order endorsement

 

7'9 Benson (1991), p. 397.
8° Benson (1991), p.397.

148

of ﬁrst-order volitions must be sensitive to “the reasons there are”.8| There are at least
two readings of the quoted phrase, each of which indicates problems with the object of
the critical reﬂection criterion, namely practical reasons (I’ll discuss the process of
critical reﬂection itself in a moment). On the weak reading, on which the reasons at issue
are (merely) “the reasons that their hi gher-order motives afford”82, no addition has been
made to a standard structural account. This is because Frankfurt-style endorsement
already provides the agent with reasons grounded in her hi gher-order desires. The
apostle Paul, for example, has at least one more reason to serve God than does Saul — that
reason is that Paul endorses (whether actively or by mere lack of internal conﬂict) his
serving God.83 On the other hand, a strong reading, on which ‘the reasons there are’
refers to reasons external to any desires (lower- or hi gher-order) of the agent, might seem
to cast autonomy’s net problematically wide, in ways discussed in earlier sections of the
present chapter.

Benson is apparently aware of the problematic nature of content requirements on
autonomous action, for, as noted above, he allows for the possibility of autonomous
actions grounded in mistaken judgments, so long as critical competence features in
deliberation. There are two points to be discussed regarding this reasons requirement.
The ﬁrst is grounded in the nature of practical reasons, and the second in Benson’s
fallibilist characterization of the critical competence that ideally leads to the reasons there

are.

 

8' Benson (1991), p. 396.

82 Benson (1991), p. 399.

33 There may, of course, be disagreement as to whether Saul has a reason to serve God or not; the point is
that however many such reasons Saul has, Paul has (at least) one more.

149

The nature of practical reasons is a concern here in light of a contemporary
dispute over whether a certain reason for a certain person must, to be considered a reason
for her, appeal to some desire, aim, or motive that the person already has. Benson
realizes that an account of autonomy that includes a responsiveness-to-reasons
requirement might seem to separate the grounds of action from the agent herself, since it
appeals to something conceptually outside the agent’s will, too far from the “self-
regulation, or self-determination, that seems to stand at the conceptual center of the
notion of autonomy.”84. This is a worry primarily in light of a conception of reasons as
things that “serve to justify actions; their status as reasons does not depend on their
actually belonging to explanations of the agent’s actions.”85 This external conception of
reasons, as it has come to be called, denies that a reason need be in some way connected
to some desire, aim or motive of the agent to whom it is a reason.

Benson denies the accuracy of such a conception, at least in the context of
autonomy. “Now, when I act autonomously,” he claims, “. . .these reasons must be
capable of being the grounds on which I form my decision and subsequently undertake to
act?“ The distinctly autonomous agent is one for whom “the motivational capacity of

”87 than for the nonautonomous person who, as exempliﬁed by

reasons is more robust
oppressively socialized women, is blind to certain of the reasons there are. So the

autonomous agent, at least, is a person to whom the relevant reasons are internal - that is,

the reasons do have connections with her motives.

 

8“ Benson (1991), p. 398.
85 Benson (1991), p. 401.
8" Benson (1991), p. 402.
87 Benson (1991), p. 408, fn 23.

150

If Benson is making a claim about the nature of reasons, then, it might seem a
limited one. Reasons, on his account, are by nature justiﬁcatory claims, but in the special
case of autonomous actions they serve a motivational purpose in virtue of the
autonomous agent’s appreciation of them. Thus the connection between a person and her
actions is preserved, since whatever role “the reasons there are” play in her actions, it is
only because she sees them as reasons that they are connected to her actions.

The criterion of autonomy is now shifted from some feature of the relation
between an agent and her actions to the agent herself - more speciﬁcally, to that part of
her self-identity concerned with her conception of her own value. Broadly speaking, the
critical competence condition puts restraints not so much on what reasons a speciﬁc
autonomous agent can act on as it puts restraints on what she is to identify with at all.88
This entails that certain persons fail to act autonomously not because they fail to govern
themselves, but because they have an identity that could not, given any connection with
the person’s actions, be involved in self— governance. The oddity of such a view — on
which a person with certain identity elements cannot be autonomous with respect to them
regardless of the actions they motivate — has been repeatedly noted in the foregoing
discussions.

Speciﬁcally, Benson’s view denies autonomy to anyone who is in error, to some
substantial degree, regarding her own personal value. This seems at ﬁrst to be a speciﬁc
subtype of the now-familiar ideal self. Benson’s ideal self is one who self-identiﬁes as a
person with the degree and sort of personal value that an objectively accurate

identiﬁcation of her would reveal. But in fact, Benson’s version of this view involves a

 

8” It also assumes that successful critical reﬂection will lead to ethical convergence on ‘the reasons there
are’, an assumption incompatible with Williams’s brand of intemalism.

151

troubling dilemma.89 If the oppressively socialized person acts in accordance with her
misvaluation, she fails to respond to the reasons there are, and is thus not autonomous. If
she acts contrary to this valuation - that is, acts as a person who knows her real value, in
accordance with the reasons there are — she would fail to act autonomously, not being
self—regulating.90 Given that “feminine socialization has insinuated its lessons into their

”91

most stable views of what they are and ought to be as persons , the only possibility for
autonomy for such persons will often be radical change in who they are —- perhaps,
practically, to become someone else.

The second concern mentioned above, regarding Benson’s characterization of
critical competence, shows that the critical competence condition is little help here, at
least insofar as one might support an addendum to the structural view. The critical
competence condition appears not to be a part of the structural view in cases in which a
person endorses her desires in the negative, or passive, sense of ‘endorsement’ — that is,
where she experiences no internal conﬂict regarding a ﬁrst-order desire. If she fails to
experience such conﬂict because she has simply been blinded (through oppressive
socialization) to the possibility of negatively evaluating (and thus experiencing internal

conﬂict regarding) her desire, critical competence is supposed to give such her a tool with

which to come to this negative evaluation.

 

8” This dilemma is likely present certain other views under discussion here. It is especially clear in
Benson’s account, though, since this account weights equally heavily the notion of self-regulation (which,
for example, Susan Babbitt’s account was relatively unconcerned with) and acting with regard to objective
reasons.

9° A great deal depends here on the details of exemplary cases. If she is coerced into acting in a self-
valuing way -- say she is ﬁned unless she acts in such a way -- she is not autonomous vis-a-vis her
understanding of her own value, though it might be debated whether she acts autonomously in relation to
the ﬁnancial motive.

9' Benson (1991), p. 388.

152

It is not clear, though, what this process would amount to that is over and above
endorsement. It certainly cannot be a process on which every one of a person’s ﬁrst-
order desires are actively evaluated before (or even after, in retrospect) they can ground
autonomous action — the range of autonomous actions is much greater than the range of
actions performed on the basis of motives that have been subjected to explicit critical
evaluation, unless we think there are vanishingly few autonomous actions performed.
More plausibly, autonomous actions are grounded in motives that could be the subject of
such evaluation just in case the agent saw some reason to evaluate them. Think here, for
example, of the process by which most people are able critically to evaluate the safety of
our immediate surroundings — we (again, most of us) do not always check that no one is
lurking behind the door, that the iron is off, that the carbon monoxide detector isn’t going
off, and so on. Rather, if we notice something a bit amiss, we have the ability to look
behind the door, check the iron, and so on. Now this ability is one the autonomous agent,
on the structural view, does indeed have. If she senses something (internally) amiss, she
can either probe for the root of the conﬂict, or remain in conﬂict without knowing why;
in either case, the desire in question is not one she is autonomous in regards to, just as
Benson demands.

If, rather, we interpret the critical competence condition in the light of the stronger
reading of “the reasons there are” - that is, the reasons there are regardless of the
person’s subjective motives — the situation is little better. Not that Benson explicitly
recommends such a reading of the critical competence condition. Indeed, critical
competence as a fallible ability — one, that is, that ideally leads to the reasons there are,

but is a necessary component of autonomous action even when it goes awry — might

153

appear to offer a reply to the abovementioned dilemma of a person’s being autonomous
only if she becomes someone else. Given that critical competence in evaluating one’s
motives is a fallible tool, the problem that faces oppressively socialized and hence
nonautonomous persons is not centrally their failure to apprehend reasons, but is rather
that they lack the ability to do so (whether or not that ability leads to success). If this is
the case — that the radical alteration of one’s values, aims, social commitments is not so
necessary for autonomy as is simply the ability to apprehend reasons - then the worry
that a person might have to in effect become someone else is misplaced.

But although Benson explicitly claims that critical competence is a fallible ability,
and that “one can exercise the requisite critical competence and still arrive at a mistaken
assessment of what there is reason for one to do””, he describes it solely in terms of its
success. He gives an example of a woman who “had sufﬁcient competence at critical
reﬂection to be able to appreciate the unreasonableness of some features of her gender
training.”93 Benson ﬂatly states that “autonomy requires having sufﬁcient competence at
critical reﬂection to be able to detect and appreciate the reasons there are to act in various
ways?” In both the example and the statement, sufﬁciency of competence is
characterized in terms of what goal a person is able to reach. A bit later, Benson tells of a
situation in which “inasmuch as I can apprehend that my desire is not irrational and that I
have no good reason at the moment to do anything else, I can possess all the critical

9 ’95

ability that autonomy demands And since Benson is explicitly discussing the

autonomy of speciﬁable actions, not “that sort of autonomy which attaches only to

 

92 Benson (1991), p.397.
9" Benson (1991), p.397.
9" Benson (1991), p. 397.
95 Benson (1991), p. 400.

154

”9" it is not the case that ‘ability’ means something like

extended periods of agents’ lives
‘tendency’. The emphasis, it is clear, remains still on connection with the reasons there
are, and thus ultimately on the values, desires, goals, and commitments with which the
autonomous person identiﬁes?7

The reasons of the ideal self

The foregoing proposed accounts of autonomy each suffer ﬂaws that make them
inadequate replacements for the structural account. The reasons for this inadequacy, as
we saw, vary: Babbitt’s narrow range of concerns leads to a narrow understanding of
autonomy; Meyers’s account is grounded in an implausible notion of the authentic self,
and contains an inadequately-supported moral element; Christman’s criteria for
preference formation ultimately do not support judgments in the intended cases.

Beneath these distinctions, these accounts share a conception of the central
problem of autonomy. I agree with the claim implied by all the philosophers under
discussion that if there is something about a person’s volitional economy that is outside
her control, there will be a degree of autonomy lacking in actions that issue from that
portion of her volitional economy. This just follows from the ‘self-control’ element of
autonomy — since, for example, my motive to hand over my wallet to the mugger is not in

my control (it is, rather, in the control of the mugger), I am thereby not autonomous when

I hand over the wallet.” This should be uncontentious.

 

9" Benson (1991), p. 390.

97 There may, of course, be a way to characterize the ability without reference to its success. Meyers’s
similar notion of ‘autonomy competency’ was characterized by the various abilities that constitute it, but as
I argued there, many of these abilities are ones that apparently autonomous persons lack.

9‘"That the example depends speciﬁcally on coercion might make it contentious regarding autonomy, but the
point it illustrates should be clear.

155

But the absence of such gun-to-the-head coercion is not sufﬁcient for the sort of
control at issue in the non-structural accounts. What these accounts require for autonomy
is that the motivational elements with which a person identiﬁes are under the auspices of
some further elements with which she identiﬁes but that are not themselves part of the set
of motives they govern. An internal economy that supports autonomous action must have
some portion insulated from the rest, at least when ‘the rest’ fails to meet the criteria
peculiar to the different accounts. For example, the closed-minded, socially unaware
person is autonomous on Meyer’s account only if she acts openmindedly and with
alertness to the reactions of others.

The shared intuition, then, is that there is some sort of ideal self that the person
could autonomously act from within, if only she could become (or act as) that self:
Babbitt claims that any person, whatever her perceived interests, is really interested in
ﬂourishing; Christman implies that the self I become is really only me if I have control
over that self-becoming process; Benson argues that only by identifying with speciﬁc
sorts of morally desirable reasons can a person be autonomous; Meyers explicitly argues
that central to autonomy is the acting not on apparent desires but rather on the real desires
of what she calls the authentic self, which I argued to be an ideal self.

Ideal-self views are by no means assuredly benign. Consider the marginalizing
effects of denying the autonomy of persons who are not able to critically evaluate the
makeup of their internal states. On the strongest (by no means an implausibly strong)
interpretation of ‘critical evaluation’, this includes all persons, given the ‘view from
nowhere’ requirement of such self-evaluation. In a certain sense, every person is like the

paranoid who denies attributions of paranoia to her on the grounds that they are certainly

156

lies constructed by those who are persecuting her. This observation should at least give
us pause:

There can be no well-ordered inquiry into the question of how one has

reason to live, because the prior question of how to identify and to

evaluate the reasons that are pertinent in deciding how one should live

cannot be settled until it has ﬁrst been settled how one should live.99

There is, alternately, a more localized interpretation of the self-reﬂection demand,
on which a person evaluates not her entire internal economy but rather portions of it. The
thought here might be (as, for example, John Christman has suggested) that certain
legitimate elements of one’s identity are to support reﬂection on those whose legitimacy
is questioned. But as we saw, the criteria Christman recommended were insufﬁcient to
the task (as they failed to explain exemplary cases), and other more substantial elements
would likely fall to an inﬁnite regress as the above quoted passage suggests.

Aside from these problems, the relative rarity with which persons tend to engage
in even this localized sort of self-reﬂection would entail that there are very few
autonomous actions performed. When I look at my own life, at least, the ratio of
autonomous actions to self-reﬂective actions is abundantly weighted toward the former.

Finally, as I’ve noted earlier, the danger of misguided paternalism can be
increased by holding an ideal-self requirement on autonomy as well. On some ideal-self
views (for example, that of Susan Babbitt) a person’s action may in certain fairly
common cases be said, paradoxically, to be autonomous (over and above merely
beneﬁcial) only when it is guided by the wishes of someone other than herself.

None of the foregoing should be taken to mean that in all situations one’s ideal

self is irrelevant to autonomy. Some actions are coincident with a desire that one wants,

 

’9 Frankfurt (2004), p. 26.

157

but does not (yet) possess. A parent who disdains organized education, but who sees the
happiness school brings to her child, might want to act as if she valued education. When
she acts as if she had a positive opinion of education, she acts autonomously, for this is
the manner in which she desires to act.

As a practical matter these situations lead often enough to the eventual realization
of one’s ideal self. This is relevant not only to the case of the parent - who in acting out
of love for her child and as if she valued education, comes to value education — but also
to situations in which a person intentionally acquires a wanted desire by acting as if she
already has it. A person who wants to desire to watch art ﬁlms would perhaps best go
about gaining this desire by watching some art ﬁlms, and hopefully coming to appreciate
them and thus to desire to watch them.‘00 Persons who want to act on desires the objects
of which we sometimes call ‘acquired tastes’ fall under this description. Examples of
these sorts suggest that there is room for an ideal-self element in an account of autonomy
that is neither implausible nor conducive to political exploitation. In what follows, I will
outline such a notion.

Korsgaard, as we saw, argues that reasons follow simply from one’s existing
identity: ‘An X doesn’t Y.’ I have supported this view as accurate; I have argued, though,
that it is too narrow. A person gains reasons not only from her identiﬁcation with the self

she is but also from the one she desires to be. Since a person is not typically entirely

 

1"" On the structural desire-endorsement model, the explanation of such a situation is interestingly complex.
Ex hypothesi, the person has a second-order desire to want to watch art ﬁlms, but does not yet have the
ﬁrst-order desire that is the object of the second-order desire. This might seem to entail that the second-
order desire is actually a desire of the ﬁrst order, since some desire of hers must be motivating her to watch
the ﬁlms, and it was stipulated that she does not have the requisite ﬁrst-order desire. What she does
possess (if she is practically rational) is a ﬁrst-order instrumental desire to do whatever she determines
necessary and sufﬁcient for her to acquire the desire to watch art ﬁlms. The italicized passage, of course,
happens also to refer to ‘the desire to watch art ﬁlms’, a desire which she does not, by stipulation, possess.
Since this in an intensional context, though, there is no incoherence in claiming her to desire ‘to do
whatever she determines. . . ’ and not desire ‘to watch art ﬁlms’.

158

identiﬁed with each of her practical identities, but rather is such to a degree, and typically
either waxing or waning in relation to the identity, she also gets reasons from her stance
toward the identity-element. If she endorses, say, being an environmentalist (which no
one is perfectly all the time), and so she desires to be more closely identiﬁed with
‘environmentalist’, she gets a reason not since an environmentalist simply does buy a
fuel-economical car (they all don’t), but rather she gets a reason to buy a fuel-economical
car in virtue of her wanting to be (more of) an environmentalist. Her becoming more
closely identiﬁed, so to speak, with a certain practical identity can take a number of
(concurrently realizable) forms. She might gain a relatively wider range of reasons from
the identity element; the reasons that stem from this identity might tend to trump those of
other of her identities; the identity might take a hi gher-order role, being the basis on
which other of her practical identities are judged.

The structural account of autonomy that is at the core of this understanding
explicitly contains an ideal-self element, in fact. Frankfurt gives an example:

Someone who loves justice, for instance, necessarily wants to be a person

who serves the interests of justice. He necessarily regards serving its

interests not only as contributing to the realization of a desirable social

condition, but also as integral to the realization of his ideal for himself.

His love deﬁnes for him, at least in part, the motives and preferences of

his ideal self.101
Described in this way, acting in accord with one’s practical identities -— where these

practical identities are afﬁliated with the things the person cares about — is to promote

one’s ideal self.

 

'0‘ Frankfurt (1999), p. 139.

159

It is consistent with the notion of the ideal self that appears in the quoted passage
above that a person who loves justice not only be a person who wants to serve the
interests of justice but also does serve the interests of justice. One’s ideal self need not be
entirely distinct from one’s actual self; many of the desires, beliefs, values, and the like
that a person wants (ideally) to identify with are ones with which she already does
identify. This is consistent as well with the notion of the ideal self as it featured into
views discussed above. Babbitt’s account maintains only that oppressed persons fail
actually to realize the interests that they have in ﬂourishing, and so nonoppressed persons
might well realize their own best interests and as such act from their ideal selves. When
Meyers asks which “desires, personal traits, values, interests and aims shouldl seek to
enact”, the implication is that there is a correct answer to this question, an answer that
some persons get right. "’2 Christman certainly would not deny that some persons have
the ability to self-reﬂectively evaluate their desires, nor would Benson claim that no one
succeeds in appreciating the reasons there are. Thus, the ideal self of the structural view
is consistent in this aspect — that sometimes the ideal and the actual coincide — with the
implicit ideal self demand of the previously discussed accounts.

This is, though, still a weaker notion of the ideal self than those earlier discussed.
The ideal self must be within the possible present imaginative or conative state of the
person. This was not necessarily the case regarding other views. Babbitt, for example,
explicitly argued that some oppressed persons who fail to have a self-conception
adequate to their own ﬂourishing lack the capacity even to imagine a world in which such

a self-conception might ﬁgure into their identity.

 

‘02 Meyers (2000), p. 166; emphasis added.

160

The ideal self that is constituted by desires, goals, values, et al, with which a
person wishes to but does not yet identify is, I have suggested and continue to maintain,
more plausibly the source of autonomous action than some objectively ideal self. The
ideal of this ideal self refers to the person whose actions are under scrutiny, thus
maintaining the self— governance that is the core of the notion of autonomy, as is claimed
by all parties to the debate.

To summarize the argument of this chapter: I have argued that identity disunity
undermines autonomy because the most acceptable account of autonomy is a hierarchical
account. The reasons that alternative accounts are less acceptable are as follows. Susan
Babbitt’s ‘human ﬂourishing’ account of autonomy is not only too narrowly focused on
oppression’s effects on the possibility of autonomy to be taken as an account of
autonomy proper, but it is also even more open to political abuse as the mainstream
liberal view that motivates Babbitt’s own account. Diana Meyers’ procedural view of
autonomy is unacceptable insofar as the characteristics that the autonomous authentic self
is said to have are characteristics that the authentic selves of many real persons simply do
not have. John Christman’s requirement that an autonomous person approve of the
process of her desire-acquisition is implausible since such approval is either irrelevant (in
cases where a person gets a desire she actually wants, though through avenues she
disapproves of) or, since it is grounded in desires the person already has, superﬂuous to
the Frankfurtian account. Paul Benson’s critical competence condition, which requires
not only that autonomous persons be self-regulating but also that they be motivated by
objective reasons, implies that those who are non-autonomous due to oppressive

socialization cannot be self-regulating.

I61

The hierarchical account of autonomy holds that a person acts autonomously just
in case she acts on a desire that she endorses, that is, a desire with regards to which she
has no internal conﬂict. These desires just are the ones with which she identiﬁes, and so
disunity (conﬂict) of identity will negatively impact autonomy. And as I have argued in
the previous chapter, the notion of endorsement is applicable not just to desires but also
to a wide range of possible identity elements, elements which (like desires) motivate and
give reasons for action. Whatever characterization we give of the identities from which
we act, they are from the point of view of autonomy ideally uniﬁed.

The value of autonomy and the unity of identity

A few more general comments are in order at this point, not the least reason for
which is to indicate the direction of the subsequent discussion. An observation not yet
noted is that some amount of the resistance to the thought that a uniﬁed identity is
necessary for autonomy is grounded in a certain belief about the value of autonomy.
Since some persons have distinctly disuniﬁed identities (and most persons do, to a
degree), and since the ability to act autonomously is of the highest value, to rest
autonomy on unity entails denying the value of the lives of many persons. Maria Lugones
alerts us to the multiplicitous subject’s “invisibility and worthlessness in the eyes of those
who attempt to control multiplicity” by (inter alia) supporting unity requirements on
autonomy. ’03

But a lack of autonomy makes a person metaphorically invisible only if boundless

autonomy is a sine qua non of human life, and makes a person worthless only if

 

‘03 Lugones (2003), p. 130.

162

autonomy is the only, or perhaps primary, source of human value. Both these
assumptions, while perhaps commonly implied, need not be accepted.

No person is autonomous in every action; neither is anyone (victims of terrible
psychiatric or social pathologies notwithstanding) fully bereft of autonomy; the truth of
these claims should be obvious. Less obvious is the idea that autonomy is not a trait of
the highest value. Autonomy can be valuable in itself (many of us often act as if it is, at
least, and so it is at least valuable to many of us), though some will value it more than
will others. But as I’ve continuously stressed, the fact that an action is autonomous does
not entail that the action is a good one, morally or otherwise. Further tempering any
thoughts of the overarching value of autonomy, the pursuit of many other valuable things
— many personal relations such as romantic love and parenting, as well as the undertaking
of certain sorts of long-range projects’o“ -- often requires restricting one’s (episodic)
autonomy. If there is some sort of hierarchy of values, autonomy may not (always)
occupy the top position.

This is not to say that autonomy is not a good to be pursued. Some valuable
things are reduced in value partially or completely if not attained autonomously; things
that are valuable mainly in virtue of their having been acquired (or created, or pursued)
by the person to whom they are valuable are of this sort. Food is valuable whether she
who consumes it acquired it autonomously or not, but the Grand Prize-winning pumpkin
is (most) valuable to she who grew it; there is much value in satisfying one’s own desire
to run a marathon, but virtually none at all in a 26.2-mile forced march. It often matters

that one’s actions are her own.

 

'0‘ Those projects, at least, that are not embraced wholeheartedly. Training for competition in endurance
events sometimes involves Ulysses tactics, for example running a route on which at the point one is likely
to want to turn back, completing the course is actually no greater a distance.

163

Even to give up or minimize the self-guided nature of one’s actions — as certain
sorts of Buddhists aim at, and is suggested by those ethical views that are grounded in
altruism rather than personal duty — is, on these views, to be done by a person, not to her.
As Odysseus’s famous choice shows, there may be good reason to give up one’s
autonomy. There is less reason, though, to have it taken away.

Beyond the value of one’s actions being her own, it does of course matter that

"’5 This is the intuition (possibly connected with the

one’s actions are morally good ones.
inﬂated valuation of autonomy) behind much criticism of structural accounts of
autonomy, as has been discussed above. While I maintain that autonomous actions can
be immoral or imprudent, there is a notion closely allied to autonomy about which this
claim does not hold. This is the notion of integrity.

The conceptual and practical relations between autonomy and integrity take a
multitude of diverse forms in the philosophical literature. Susan Babbitt, for example,
seems to take integrity to be conceptually identical to autonomy.'°" On some views,
integrity is at odds with the demands of morality - a commitment to morality can be and
often is incompatible with acting with integrity.‘07 Sorting out some of these connections

is the task of the next chapter, which takes up the discussion of integrity and its relation

to autonomy, principles and commitments, morality, and the unity of identity.

 

"’5 It doesn’t merely matter morally that one’s actions are morally good; this claim is not tautologous. It
matters to many of us personally (to those persons, that is, who take morality up as a personal project), and
the morally worthy life might be seen to have an additional aesthetic value, as well.

‘06 Babbitt (1996).

“’7 Williams (1973), pp. 108 ff., and (1981).

164

_4_

Integrity

The discussion of the present chapter will focus on the connection between
integrity and the unity of identity. The notion of integrity has been the subject of a fair
amount of philosophical attention since Bernard Williams’s revival of interest in the topic
over three decades ago', and since then a number of strongly disputed accounts have been
presented to support the claim, often implicit, that integrity just is a form of personal
integration or unity maintained in the face of disunifying forces. Beyond this shared
feature, there has been little in the way of conceptual convergence among these accounts.

The present contribution to the philosophical discussion of integrity has four aims.
First, to shift the discussion from the character trait or personal virtue of integrity to
integrity as a characteristic of a speciﬁc act. My second aim is to present a view of
integrity, which I will call the taxonomic view, which identiﬁes distinct sorts of integrity
and thus will help to make sense of much of the stalemate in the current literature. Third,
this view of integrity addresses and attempts to accommodate an intuition that surfaced
repeatedly in the last chapter’s discussion of autonomy — the intuition that the moral
content of an action is relevant in determining whether that action is one’s own. The
fourth and main goal here is to show that the unity of identity is a crucial element of, but
not simply equivalent to, the ability to act with integrity.

Certain character traits are commonly exempliﬁed by public ﬁgures or literary

characters: Othello ’s Iago is an exemplar of treachery, Mother Teresa a paradigm of

 

‘ Williams (1973).

165

altruism, Oscar Wilde of wit. Integrity has a number of supposed representatives: Martin
Luther King, J r., Mohandas Gandhi, and Abraham Lincoln are representatives of the
character trait of integrity to the extent that their names can be taken as synonymous with
the term. “My father never backed down,” I might say. “Why, he was a virtual Gandhi
for his whole life.” At least in a standard context, the intent of such a statement is easy to
grasp: I am claiming that my father was a man of integrity. (It should not be thought that
all of the work in characterizing my father is done by the phrase ‘never backed down’. I
could use that phrase to describe him followed by a reference to Attila the Hun and the
character trait I am attributing to him would be understood very differently.)

But (of course) the reality of such commonly lauded public ﬁgures is more
complex. If we point out that Gandhi tested his commitment to brahmacharya, total
control of one’s own desires, by enlisting various (often young) women to lie naked
beside him, the attribution of integrity might seem less apt.2 Or consider King: as the
man who fought so long and so bravely for the rights of black Americans, he is clearly a
person of integrity. But as a man who plagiarized portions of his PhD dissertation’, or as
a womanizer, we might not know whether to award him the honoriﬁc label of ‘integrity’.

One important philosophical concern regarding integrity seems to be the
construction of an account of integrity that will help us reliably to make such judgments,
as seems especially difﬁcult when judging the integrity of a person who seems for some
reasons to deserve the attribution but for other reasons not to deserve it.

I don’t assume that integrity is the only character trait that is complicated in this

way. I have chosen to investigate integrity since (among other reasons) we do tend to

 

2 As reported by Mehta (1993), p. 43.
3 A habit that began with the presenting as original his ﬁrst sermon, which was actually one “lifted” from
Rev. Harry Emerson Fosdick (Frady (2002), p. 19).

166

weight it very heavily in certain important situations — we often aim to choose our
friends, our lovers, our business associates and our political leaders on the grounds of
their integrity. Integrity is also often taken to be a laudable quality of one’s own self.
“Whatever happens,” a person may tell herself, “at least I have my integrity.” (I myself
happen to have kept a message contained in a fortune cookie, which in a ﬁt of whimsy I
afﬁxed to the cover of my checkbook: “You value your principles higher than any money
or wealth?)

The currently available philosophical conceptions of integrity do not help us to
perform this important task of judging integrity. This is because each of the main
accounts are subject to quite simple counterexamples, cases of persons who we take not
to have integrity but who are characterized by the account as possessing integrity, as I
will now show.

Contemporary accounts of integrity

As mentioned above, Bernard Williams is often credited with sparking the current
philosophical interest in integrity. Williams has argued that acting from moral concerns
(he restricts his discussion to strictly utilitarian sorts of moral concerns) can be at odds
with acting with integrity. This is because, as Williams claims, “each of us is specially
responsible for what he does, rather than for what other people do”, an idea “closely
connected to the value of integrity?" The demands of utilitarianism, though, make no
distinction between what I do and what other people do, and so utilitarianism cannot

make sense of the special connection I have with my own actions.

 

" Williams (1973), p. 99.

167

One way in which Williams is understood is as giving a clean hands view of
integrity; this is how Cheshire Calhoun, for one, has characterized his view.5 On a clean
hands view, a person acts with integrity when she refuses to be the agent of some speciﬁc
moral harm, a refusal that may itself entail some greater moral harm coming about.
Williams asks us to imagine George, a biochemist in need of a job. He is offered one, but
it is ajob that involves research aimed at developing chemical and biological weapons.
George objects to chemical and biological weapons. Although George and his family
need the money, and his taking the job will prevent some more zealous researcher from
taking the job, and George’s family have no qualms about his taking the position, he
turns it down. George knows it to be likely that his turning down the job will not prevent
any such weapons from being created, and that he could have even slowed the
manufacture of some of these weapons by accepting the job and then being a bit of a
slacker, he shows integrity, on the clean-hands picture, by refusing to participate in what
he takes to be evil. It is not that he prevents evil, but he keeps his hands clean of it.

The attribution of integrity to George seems apt. After all, even though George
could have had a good-paying job, and his family would have thought none the worse of
him, he respected his commitments to himself, to his beliefs that if there are going to be
chemical and biological weapons in the world, at least he himself will not be the maker of
any of them

It should be borne in mind that though this view of integrity is commonly
attributed to Williams, he does not offer much in the way of a defense of the view. His

argument, rather, is that utilitarianism carries with it a notion of negative responsibility

 

5 Calhoun (1995).

168

that sometimes (as in George’s case) demands that one dirty her hands to prevent great
evil; such a moral requirement can (again, in certain cases) undermine the possibility of
integrity. Thus, Williams’s explicit position is that the possibility of integrity depends on
(as Calhoun characterizes the matter) being able to keep one’s hands clean, but he does
not claim that integrity just is doing so.

Calhoun points out that the clean hands view is at least incomplete, for
sometimes dirtying one’s hands shows integrity. A soldier in a just war might be thought
to have integrity in virtue of her dirtying her hands for a greater moral good. A person
who dirties her hands by intervening in the marriage of a friend who is having an affair
might similarly be judged to have integrity. Suppose she knows that her interference will
offend her friend with its moralism, perhaps even losing her this friend, and further
suppose she stands nothing to gain from the cuckolded husband, whom she has never
thought much of to begin with. She simply thinks (however distasteful the means to
thwarting it) that such a wrong ought not to be occurring 6 So, the clean-hands picture of
integrity is incomplete. It accurately accounts for some, but by no means all, cases of
integrity.

Perhaps one complicating factor is the inclusion of the moral element.
Considering the cases of both George and the friend of the unfaithful wife shows that the

attribution of integrity does not follow on the relation of their action to the greatest moral

 

6 Depending on the back—story, these two examples can do different work than the example of George.
George’s situation is one in which there is only one moral harm at issue: the production of weapons. The
case of the soldier might be of this sort, if we suppose that the just war is one in which some persons will

be killed, but with the goal of preventing the killing of a greater number of persons which will occur if the
war is not fought. But a just war may be fought in which one moral harm — the killing of a number of
persons — is perpetrated in the interest of preventing a distinct (and greater) moral harm — genocide, for
example. The case of the friend has similar interpretive options. The one moral harm at issue might be that
of failing to be trustworthy; two distinct harms - marital inﬁdelity and invasion of privacy, perhaps — might
both be relevant.

169

harm; George has integrity in the case where a greater harm is allowed, whereas the
friend has it in the case where she refuses to allow (what she takes to be) the greater
harm. What the attribution of integrity seems to depend on, rather, is what the action
shows about the agent. George’s refusal to take the weapons manufacturing job shows
that he is not the sort of person who makes such weapons. George is not committed to a
moral ideal, at least not a primarily consequentialist one — that is, that there be fewer
chemical and biological weapons in the world — since if this were the case he would have
taken the job and done slovenly work, undermining the efﬁcient production of such
weapons. George is, rather, committed to a certain conception of himself, a conception
as a person who does not make such weapons. Likewise, in trying to stop her friend’s
inﬁdelity, the woman shows something about herself — her conception of herself as
primarily a moral agent, and only secondarily a friend to be trusted come hell or high
water.7

These considerations recommend that the proper account of integrity be one on
which attributions of integrity respond to the connection between a person’s actions and
her self-conception. We might call this an identity view of integrity.8 On this view, a
person acts with integrity when she acts from those beliefs, commitments and principles
that are central to her own self-understanding. Calhoun attributes this sort of view to
Williams as well, an attribution that has at least some suggestive textual support. About
the demand that we might make on George to take the weapons research job so to prevent

a more zealous person from doing so, Williams claims that this would be “to alienate him

 

7 Of course, a trustworthy friend is not a defacto outcast from moral community. But on a certain ethical
view made plausible by considerations in Williams (1981) and Wolf (1982), the demands of personal
relationships and the demands of morality can and do conﬂict. If they do, there is a choice to be made as to
which will be more authoritative, that is, a choice as to how to conceive oneself.

8 Again, following Calhoun (1995).

170

in a real sense from his actions and the source of his actions in his own convictions [. . .]
in the most literal sense, an attack on his integrity.”9 If to alienate a person from her own
convictions is an attack on her integrity, preserving the correspondence between one’s
own convictions and actions would seem to support her integrity. A commonsense sort
of view of integrity — emphasizing the etymological connection with ‘integration’ and
‘integral’ —that holds that integrity just is a certain sort of wholeness is consistent with
this view as well.

The identity view does have a number of strengths. It accounts for the case of
George and of the adulterer’s friend, as just noted. The identity view also explains the
damage of oppressive social forces to persons who are prevented from acting with
integrity.10 The members of oppressed social groups - for example, women — are often
forced into certain patterns of actions, and thus can be prevented from acting in
accordance with their own principles.ll Their ability to act with integrity is in this way
impeded. (Is it any wonder why the paradigm representatives of integrity - King, Gandhi
and Abraham Lincoln already mentioned, and we might add Winston Churchill and John
F. Kennedy — are all men?”) On the identity view, the mechanism by which the
oppressed are prevented from acting with integrity is clear. Integrity just is acting in

accord with your self-conception, an ability often denied the oppressed.

 

9 Williams (1973), pp. 116-117.

'0 As argued by, e.g. Babbitt (1996) and Lugones (2003).

“ And are sometimes prevented even from coming to have their own principles at all.

'2 Mother Teresa and Rosa Parks come to mind as women of integrity. Mother Teresa’s integrity, though,
is in virtue of her extreme subordination of any self-directed desires, a peculiar (though still appropriate)
example of acting from principle. Parks tends to be publicly misunderstood as a woman who had an
isolated if bold idea (not to yield her bus seat), but few take that action to be grounded in long—standing
principle and association with the NAACP (in a notice of her death (October 24, 2005), ABC News wrote
of her “public conception as a quiet, domestic woman who was just too tired from a hard day’s work to get
up from her seat”).

171

Alas, there are intuitive examples of integrity that the identity view fails to
capture. Recall Othello’s Iago, who out of wanton jealousy caused multiple murders and
a suicide. Shakespeare does not offer much as to Iago’s own self-conception, but it
certainly was not a positive one; in one of his rare moments of self-reﬂection Iago admits
“I am nothing if not critical.”13 He also importantly takes himself to be dishonest: when
Othello mentions personal discord, Iago sarcastically replies with a reference to musical
discord in saying,

0, you are well tun’d now!

But I’ll set down the pegs that make this music,

As honest as I am."

Further, Iago is acting in accord with his self-conception, as he implies that he is not at all
at the mercy of some force outside himself:

’Tis in ourselves that we are thus or thus. Our bodies are gardens to which

our wills are gardeners; so that if we will plant nettles or sow

lettuce. . .either to have it sterile with idleness or manured with industry;

why, the power and corrigible authority of this lies in our wills. If the

balance of our lives had not one scale of reason to poise another of

sensuality, the blood and baseness of our natures would conduct us to the

most preposterous conclusions... '5

And so Iago connives to effect the murders of Desdemona, Emilia and Roderick,
and Othello’s suicide. Now, the morally demanded action would have been (as Iago
seemingly should have realized, at least judging from his understanding of reason’s
trumping of the baseness of his nature) for him to take arms against his own unjustiﬁed
hatred and jealousy of Othello, to resist the powerful internal forces moving him to

murderous deception. (Indeed, oftentimes the most powerful force to resist in acting with

integrity is a force emanating from within oneself.)

 

'3 Othello, Act 11, scene 1.
'4 Othello, Act 11, scene 1.
'5 Othello, Act 1, scene 3.

172

A far different play than Shakespeare’s Othello might have Iago resisting the
“green-eyed monster” of jealousy that drives him. This Iago could have realized that
jealousy was driving him; this Iago would have taken up his own counsel of Othello to
beware jealousy; this Iago would have taken up noble reason, rather than base passion, as
the core of the will he speaks so highly of; this Iago would take up cool (presumably
moral) reason as his principle. This Iago would have acted with integrity in standing up
to desires that moved him against (what we are now hypothetically supposing to be) his
reasonable will.

That is, this Iago would act with integrity if he would act from moral
considerations. Morally speaking, Iago ought to put aside his self-conception of a person
who is critical and dishonest, and refuse - by resisting his powerful feelings of jealousy —
to manipulate those of whom he is jealous.

The inclusion of a moral element in judgments of the character trait of integrity
has intuitive appeal. If we deny the attribution of integrity to Martin Luther King, the
denial likely rests on a judgment that no matter what brave and selﬂess things he did to
motivate racial desegregation, he is not a man of integrity because of the innumerable
liaisons he engaged in to the moral harm of his family, his political cause, and in many
cases the women he misled. Here, our judgment depends on the thought that integrity is
attributed to a person in view of their whole life, or at least a substantial portion of it, and
that this life must have been a moral one (overall, or on balance, or at least not marked by
egregiously immoral content).

The hypothetical (and admittedly contentious) ‘Nazi of integrity’ might be

interpreted in the same way. For example, it is entirely consistent with the tenets of the

173

National Socialist Movement that a member of the party largely dedicates her life to the
preservation of the natural environment. But if we are less comfortable attributing
integrity to her than we are to John Muir or Rachel Carson, it is likely because of the
conﬂict between her other beliefs and actions with the attribution of integrity. The
immorality of one aspect of the Nazi’s life (it might be thought) taints the totality of her
life, at least as regards integrity. So on a moral View, integrity might be said to
characterize a person only if she leads her life at least largely in correspondence with
morally permissible principles.

But recall our judgment of George. There was no hesitation in the evaluation of
George’s integrity in spite of our ignorance of whether he is a member of the Nazi party,
or encourages young girls to sleep next to him, or plagiarized his dissertation. That is, we
were content to judge the integrity of George’s action (or, perhaps, the integrity of
George qua agent of a speciﬁed action), rather than George himself.

Judging the integrity of a person’s actions rather than judging the person herself
can change the polarity of our judgment. Even if Martin Luther King is not a person of
integrity (because, for example, of his sexual habits), some of his actions show great
integrity. His refusal to leave jail after his arrest in Albany, Georgia, until negotiations
between the city and black leaders had been settled was an act of integrity. It was an act
done out of selﬂess commitment to a lofty moral principle, an act that he feared the
personal consequences of (believing, understandably, that the police might kill him).
Implicit in this judgment is a view of integrity on which an action is one of integrity only

if it is motivated by a moral principle.

174

Considering actions rather than character can cause our judgments to alter in the
other direction as well. It is the belief of many people that Abraham Lincoln was a
person of integrity, though some of his most important actions — such as defending the
emancipation of slaves as a political tactic aimed at the preservation of the union rather
than aimed at the freeing of the slaves per se — arguably lacked integrity.“ In considering
the person, we ﬁnd integrity; in considering the action, we do not. On which ought
judgments of integrity to focus?

One response involves the observation than even persons of integrity need not to
be perfect in their performing of actions of integrity. Thus, we can continue the quotidian
practice of attributing integrity to persons, not to their actions. But this response is not
entirely satisfactory, since there can be imagined a person who performs all acts of
integrity save one, and this one so egregious that we are moved to refuse the attribution
of integrity to her. The Nazi of integrity is again a possible example here.

More plausible is a view on which integrity is ﬁrst considered as a characteristic
of actions, a characteristic that becomes attributable to a person who engages in some
certain number of acts of integrity, or ratio of integrity-featuring actions to other actions,
or something of the sort (though I won’t be offering any helpful formula here, since I do
not expect that there is such a formula). This is a more plausible view since attributions
of integrity must ground in the integrity of actions in some fashion; it is absurd to speak

of some person of integrity who has never performed an act of integrity, or to attribute

 

'6 Of course, most actions lack integrity in the sense that judgments of integrity are simply not relevant -
most quotidian acts are of this sort. But an action can also lack integrity in the sense that judgments of
integrity do apply, and the action is found lacking. This is the sense in which one might take Lincoln to
have lacked integrity in using slaves as mere means, means to the preservation of the Union.

175

integrity to a person and claim their actions are not at all relevant to the judgment.’7 An
attribution of the character trait of integrity presupposes a decision as to which of the
person’s actions motivate the attribution. This is why a dispute over whether, e.g., Martin
Luther King was a person of integrity will (I suspect) devolve into a discussion of which
of his actions were central to his character, or most relevant in some other way to the
attribution. Once that is decided, the judgment of integrity is a straightforward matter.
These relevant actions must be actions with the quality of integrity, either in themselves
(and thus they lend the quality to the person who performs them) or in the sense that they
are actions such that to engage in them either makes the agent, or shows her to be, a
person of integrity.18

This conclusion — that integrity is most basically a characteristic of acts rather
than of persons — is certainly contrary to traditional views. Philosophers have tended to
take integrity to be a trait of character, or a disposition, or a characteristic of a whole
human life.

But to characterize integrity as essentially a characteristic of persons is mistaken.
As I’ve argued above, each of such views fails to reﬂect our judgments of integrity in
certain cases. As it seems from the foregoing discussion, taking integrity to be a
characteristic of persons leaves the question of whether a certain person is a person of

integrity open. Consider whether George is a person of integrity. Williams claims he is,

 

‘7 The former might not quite be absurd; I suppose that a dispositional account of integrity might be
defended on which certain persons (the oppressed, perhaps, or, for quite different reasons, persons in
persistent vegetative states) who are disposed to act with integrity but are effectively prevented from doing
so still warrant the attribution of integrity. It is not clear how well such an account would hold up, but it
seems at least coherent. But at any rate (and this is my point), we would ﬁrst need an account of what
makes the acts to which the person is disposed acts of integrity.

'8 That is, actions can be either the perceptual apparatus through which we see that a person has the
character trait of integrity (‘1 see from her actions that she has integrity’), or the actions can be the
constituent parts of that trait (‘Doing that will give her integrity’).

176

since George refused to accept the weapons-research job. But just as one swallow does
not a summer make, one act of integrity does not (I assume) a person of integrity make.
Certainly other of George’s acts suggest (If George is at all like you and me) that he is
not a person of integrity (after all, he’s forcing his wife ﬁnancially to support their family
because of his ultimately ineffective paciﬁst hang-up). Even if integrity is acting in
accord with one’s personal commitments in refusing to be an agent of harm, or being
disposed to do so in relevant circumstances, or habitually doing so for the right reason at
the right time, reasonable disagreement will arise regarding many cases of supposed
integrity.

There is a further reason to take integrity to be primarily a characteristic of acts.
In judging whether an act is one of integrity, we are of course judging the act of a person
(so there is no fear that attributing integrity to acts involves conceiving of agentless acts).
But in judging the integrity of an act we do not need to appeal to whether the agent is a
person of integrity. For example, in evaluating the integrity of Martin Luther King’s
unfaithful sexual behavior, the fact (if it is one) that King was a man of integrity has no
bearing. Thus the judgment of an act ’s integrity can be made independent of knowledge
of the agent’s (supposed character trait of) integrity.

Not so regarding judgments of the character trait. For example, in judging
whether King was a person of integrity, there will be ineliminable reference required to
his acts not simply qua acts, but qua acts of integrity. In order, that is, to judge whether
King was a person of integrity, the question of whether (and which of) his acts had the

quality of integrity has to have already been addressed.

177

In short, the primary philosophical motivation for here investigating act-integrity
rather than character—integrity is epistemic. Investigating the character trait of integrity
has so far yielded only a handful of inconclusive and conﬂicting accounts, as outlined
above. Further, as I have just argued, even though it makes perfect sense to attribute
integrity to persons, this attribution is made in the light of certain actions, actions which
are actions of integrity, and for which we need an account. I do not suppose that there is
any deeper relevance of acts to the nature of integrity.

Whatever the precise connection between acts of integrity and persons of
integrity, then, we can say two things: a person of integrity must perform at least some
acts of integrity at least sometimes; and following from this, if we search for a account of
persons of integrity, we need ﬁrst an account of acts of integrity.

Even though we may have made headway by reducing the problem of integrity to
that of acts of integrity, it is no more clear what the desired account might look like. The
views described above, the clean hands view, the identity view, and a moral view each
have problems corresponding to the range of our everyday judgments. Most notably,
seemingly contrary judgments seem equally plausible. George’s refusal to take the
weapons research job is an act of integrity, but if George were instead to take the moral
high ground — which is, I take it, to reduce the production of weapons whatever the
personal cost — and take the job in spite of his distaste, this would be plausibly interpreted
as an act of integrity. Lincoln’s dedication to the preservation of the Union was a central
element of his self-identity, and so pursuing that goal to the point of civil war was an act

of integrity on an identity view, but I take it that it would have showed integrity had

178

Lincoln allowed the Confederacy to secede rather than feigning concern for the slaves in
pursuing ulterior political motives.19

The solution I will suggest to remedy these odd conclusions may itself sound odd.
The suggestion is that our judgments are correct: George shows integrity in refusing the
job, and he shows integrity in taking the job. Lincoln showed integrity in pursuing the
cohesion of the Union, and he would have showed integrity by refusing to pursue that
goal in light of his having to dissemble to do so. What alleviates the oddness of this
suggestion is the provision of a taxonomy of integrity, as follows.

What George shows in refusing the weapons research job is personal integrity.
This, recall, was the crux of Williams’s point in this example — that personal integrity is
often in direct conﬂict with utilitarian sorts of moral demands (such as, in this case, the
utilitarian demand to minimize harm). What has gone unexamined in the philosophical
discussions of integrity is the distinction between integrity of this sort and moral
integrity. Moral integrity and personal integrity are distinct potential evaluations of an
action. As the case of George shows, an action that shows personal integrity need not be

an action of moral integrity.20

 

‘9 This judgment depends in part on the facts that slavery in the American South (at that time, the
Confederate States of America) would have ended of its own internal decay in a relatively short time, and
that Lincoln privately professed to have absolutely no interest in whether slaves were held or not “My
paramount object in this struggle is to save the Union, and is not either to save or to destroy slavery. If I
could save the Union without freeing any slave I would do it, and if I could save it by freeing all the slaves
I would do it; and if I could save it by freeing some and leaving others alone I would also do that. What I
do about slavery, and the colored race, I do because I believe it helps to save the Union; and what I forbear,
I forbear because I do not believe it would help to save the Union.” (Lincoln, letter to Horace Greely,
August 22, 1862.)

2" George’s situation needs to be explained a bit differently for an attribution of integrity upon his accepting
the job to be appropriate. Suppose that George is very opposed to weapons research and the thought of
taking such ajob turns his stomach — he does not think he could respect himself if he were to take the job.
George might even wow that his family will come to think less of him were he to take any part in such an
industry. But he realizes that he can perhaps save just one life, or perhaps even many lives, by taking the
job and being less efﬁcient at it than some other more zealous researcher; he overcomes his distaste out of

179

There are other sorts of integrity that, while virtually unexamined by
philosophers, are straightforwardly uncovered. Consider whether Abraham showed
integrity in his intention to sacriﬁce Isaac.2| To answer afﬁrmatively, to agree that
Abraham’s is indeed a “story of integrity”22 is to appeal to religious integrity. Consider
Stanley Mil gram’s (in)famous experiments on obedience to authority; if Mil gram showed
integrity in engaging in this research, it is intellectual integrity. William Styron’s
character Sophie exhibited interpersonal integrity in choosing to die with Nathan rather
than live with Stingo.23

We might call a view that recognizes these distinct types of integrity (types that
are severally but not yet jointly described) and the relation between them ‘the taxonomic
view of integrity’. This is the view I am defending here, and is a view that takes up many
elements of established accounts, for example the observation that personal integrity (as
acting from one’s principles or commitments) and moral integrity (as doing what
morality demands whatever the personal cost) can and often do conﬂict. What sets the
taxonomic view apart from others is the observation that personal integrity and moral
integrity (and many other sorts) are all parts of a useful account of integrity. To claim
that, say, only personal integrity is integrity (as Williams has strongly implied) is
incorrect on the taxonomic view.

The suggested labels of these sorts of integrity that feature into the taxonomic

view may not be familiar. For example, though we do sometimes speak of intellectual

integrity, ‘interpersonal integrity’ is (to me, at least) an unfamiliar term. Nonetheless, the

 

the realization that he has a moral obligation to minimize harm in the world (or to respect the ends of
others, or to act bravely), and to conquer his own irrational and ultimately self-centered emotions.

2‘ Genesis 22.

’2 Mooney (1986), p. 23.

23 Styron (1979).

180

 

judgments that they track should be recognizable. We do evaluate these sorts of acts as
acts of integrity. But we also get confused, as I’ve argued, when we fail to distinguish
the distinct types of integrity.

A further feature of the taxonomic view (which I’ll use an extended example to
illustrate below) is that it takes all sorts of integrity except moral integrity to be qualiﬁed
sorts of integrity. Moral integrity is integrity unqualiﬁed. In claiming, for example, that
Stanley Mil gram acted with intellectual integrity we limit our claim. To claim, simply,
‘Mil gram acted with integrity’ would invite explanation (and puzzled looks). This is
because he did not act with moral integrity, but rather with a subtype of integrity, as I will
explain.24

Consider the following “general principle” described in the American
Psychological Association’s Code of Ethics:

Principle C: Integrity

Psychologists seek to promote accuracy, honesty, and truthfulness in the

science, teaching, and practice of psychology. In these activities

psychologists do not steal, cheat, or engage in fraud, subterfuge, or

intentional misrepresentation of fact. Psychologists strive to keep their

promises and to avoid unwise or unclear commitments. In situations in

which deception may be ethically justiﬁable to maximize beneﬁts and

minimize harm, psychologists have a serious obligation to consider the
need for, the possible consequences of, and their responsibility to correct

 

7" For those readers unfamiliar with Mil gram’s series of experiments: Mil gram enlisted subjects on the
pretence of testing the effects of punishment on learning. The actual aim was to determine whether
subjects would cause (what they thought to be) substantial harm, by electric shock, to other persons when
this harm was demanded by a person in authority. Many test subjects (that is, those who were
administering the shocks) experienced extreme emotional distress during the experiment. In the ﬁrst run of
the experiment, 26 out of 40 subjects completed the experiment - that is, they continued to administer
shocks gradually increasing in intensity to the maximum 450 volts in spite of the extreme (though feigned)
pain to the recipient. The famous ‘Mil gram’s 37’ comes from a later iteration of the experiment, in which
the subject did not directly administer the supposed shocks, but directed a peer to do so; 37 out of 40
subjects in this situation completed the experiment. (This synopsis and other details I mention below are
drawn from Milgram 1963, 1964 and 1977 and Baumrind 1964.)

181

any resulting mistrust or other harmful effects that arise from the use of

such techniques.25
Mil gram seems to have failed to maintain this principle on nearly all counts. He deceived
his test subjects, by misrepresenting both the object of the study (as punishment rather
than obedience ), the role of the participants (who thought themselves helpers rather than
test subjects), and the effects of their actions (subjects thought themselves to be
administering real electrical shocks). It is not clear that he minimized harm to the
subjects, some of whom were “reduced to a twitching, stuttering wreck [. . .] rapidly

””26

approaching a point of nervous collapse , and though three subjects experienced “full-

blown, uncontrollable seizures”27 “

only one instance of the experiment was terminated
because of this. Mil gram was accused of neglecting to correct for harm to subjects,
exhibiting a “posture of indifference” toward the emotional state of his subjects.28

The APA integrity principle is clearly a morally characterized one, speaking as it
does of responsibility and obligation to honesty, fairness, and minimization of harm.
Mil gram thus, in violating on many points the commonsense understanding of integrity
that it codiﬁes, acted without moral integrity. But consider some of his responses to such
criticisms of his actions. Though Mil gram did reply to a number of the moral concerns,
much of his defense appealed to the intellectual and scientiﬁc value of the experiment.
“['T]he problem of destructive obedience,” he writes, “ ...because it is the most

perplexing, merits intensive study.”29 Nor did he take this stance to be his alone, for he

claims that “many [of the subjects] felt gratiﬁed to have taken part in scientiﬁc research

 

25 American Psychological Association (2002), p. 3.

2" Mil gram (1963), p. 377, attributed to an unnamed observer.
27 Milgram (1963), p. 375.

2" Baumrind (1964), p. 422.

29 Milgram (1964), p. 848; italics added.

182

they considered to be of signiﬁcance”30

, and a subsequent survey of the test subjects
revealed that 4 out of 5 of them “felt that more experiments of this sort should be carried
out”.31 Even a very disapproving editorial writer for the St. Louis Post-Dispatch noted
that “the rationale offered as justiﬁcation is that it was in the interests of science.”32

Indeed, the polemical, somewhat despairing tone of even Milgram’s articles
published in psychology journals indicates that he took the knowledge gained through the
experiments to be of utmost importance. The entire ﬁrst page of his original article
describing the experiment (in the Journal of Abnormal and Social Psychology) speaks of
the horrors of obedience to the Third Reich. (Mil gram’s experiments took place shortly
after, and were almost certainly given impetus by, the trial of Adolf Eichmann; surely the
cotemporaneous escalation of the (so-called) Vietnam Conﬂict played a similar role.)
The intellectual question to which Mil gram desperately wanted an answer was this: “if X
tells Y to hurt Z, under what conditions will Y carry out the command of X, and under
what conditions will he refuse?”’3

This extended discussion of Stanley Mil gram is intended to support the claim, and
the underlying theoretic account, that a person might act without moral integrity but with
e.g. intellectual integrity.’4 Note that while different types of integrity are
distinguishable, many will not be categorically distinct. Recalling William Clifford’s

argument that belief has a central ethical element suggests that Mil gram’s pursuit of

knowledge was inextricably moral; similarly, Abraham certainly took ethics and religion

 

3" Milgram (1964), p. 850.

3' Milgram (1964), p. 849.

32 St Louis Post-Dispatch, editorial, November 2, 1963.

’3 Milgram (1964), p. 848.

3" There might, I suppose, be thought to be distinct categories of moral integrity (Kantian and Utilitarian,
say) which could conﬂict.

183

to be complementary if not identical. Nonetheless, distinctions can be made — Mil gram
took his inquiry to be fundamentally an intellectual one, into a “most perplexing”
question, and if we don’t take there to be a distinction between moral and religious
demands, then it is unclear why Abraham’s willingness to sacriﬁce Isaac is so
noteworthy.35 Mil gram was a scientist, not an insurrectionist (especially if these are
success-terms; Mil gram largely succeeded in answering the psychosocial question, but
we can say without unreasonable cynicism that the answer has had no noticeable practical
effect on human behavior).

Returning now to the heart of the taxonomic view, it has been implied by Bernard
Williams that personal integrity, not moral integrity, is integrity proper. Utilitarianism’s
problem with understanding integrity, he claims, is grounded in a failure to “coherently
describe the relations between a man’s projects and his actions?” Harry Frankfurt
suggests this sort of ranking as well. He imagines someone who “has no personal
boundaries whose inviolability he might set himself to protect,” and so “there is no such
thing for [him] as genuine integrity.”37 But if personal integrity is true, unqualiﬁed
integrity, then attributing integrity to Mil gram’s actions wouldn’t raise eyebrows and call
for an explanation (which it does).

There are certainly cases that might seem to suggest that personal integrity is

integrity proper. Ridley Scott’s ﬁlm Thelma and Louise tells the story of two women

 

35 Perhaps there are other interpretations of the situation — supposing both that religion and morality are one
and the same and that Abraham did (or at least should have) willed Isaac’s sacriﬁce only in the face of a
great inner struggle suggests that Abraham acted out of moral integrity rather than the conﬂicting
interpersonal integrity. 1 do not expect that the provision of a taxonomy of integrity will make attributions
of integrity a matter of applying an algorithm; there are a number of ways to interpret a situation, but the
point is that the taxonomic account is relevant to all of them.

3" Williams (1973), p. 100.

37 Frankfurt (1999), p. 115. It is worth noting that Frankfurt does not offer quite an account of integrity; but
makes suggestive remarks that, like those of Williams, are aimed only at describing certain conditions
under which acting with integrity is impossible.

184

who act immorally but at the same time act with integrity. To take a speciﬁc scene as an
example, Louise prevents Thelma’s rapist from continuing the act, but even after the
immediate threat is past, Louise (motivated by her own remembered rape), shoots the
man to death. This could be taken as an act of integrity; clearly Louise acts for reasons
that form, for her, a principle — a set of beliefs, self-conceptions and motives that are
central to who she is.

Now, the attribution of integrity to Louise’s act is disputable in a way that, say,
Martin Luther King’s refusal to leave the Albany jail is not. That is, one might argue that
no act of murder could be also an act of integrity; holding a ‘clean-hands’ view of
integrity such as the one Calhoun attributes to Williams would clearly entail such a stance
(Louise importantly dirtied her hands). But if there is something to the thought that
Louise did show integrity in committing an illegal, arguably immoral, and ultimately self-
destructive act, the taxonomic view makes sense of that thought — Louise showed
personal, though not moral, integrity. And it is because Louise’s act is debatable in a
way King’s act (of moral integrity), is not — that is, because a demand for qualiﬁcation of
the attribution of integrity to Louise’s act is perfectly understandable — that I claim moral
integrity to be integrity unqualiﬁed.

Personal integrity is, further, a precondition of moral integrity — an act of moral
integrity is the act of someone whose personal principles are moral ones.38 This is why,
for example, Huckleberry Finn’s self-perceived wanton act of helping the slave Jim was a

morally positive act, but did not involve integrity on Huck’s part; it was morally good

 

3” McFall (1987), p. 16. A point not made by McFall, but worth noting here regards cases in which a
person acts with moral integrity but out of character. Here, the degree to which we attribute integrity to
such an act is the degree to which the person is, through the act itself, taking the moral principle as her
own.

185

that Huck sheltered Jim, but whatever moral principle is relevant here, it wasn’t Huck’s
moral principle.39 Personal integrity is thus distinct in its relation to moral integrity, since
neither intellectual, nor religious, nor the other subtypes of integrity are preconditions of
moral integrity.’0

There are further elements of integrity that are not unique to the taxonomic
account. For one, as has been implied throughout the foregoing discussion, acts of
integrity are resistant — to general social forces, to features of speciﬁc situations, or to
powerful contrary inclinations. A person can show integrity by resisting the norms of
male privilege, resisting her father’s speciﬁc demands for male privilege, or (for a man)
resisting one’s own desire to’exploit the powers of male privilege. So, for example, a
man might show integrity by turning down an offer for a desired job, an offer made to
him at the expense of a more deserving woman who was passed over for the offer.
Absent the forces he resisted (which in this case are two: male privilege, and his own
desire for the job), integrity would not be an issue here.

This element of integrity is implicit in many views. In Williams’s original
discussion, for example, acts of integrity are importantly resistant to (utilitarian) moral
demands. Similarly for Frankfurt, integrity is posed against demands to violate the limits
of a person’s own volitions, “limits that might serve at once to anchor his judgment and
to specify the requirements of his integrity?"I Lynn McFall, elements of whose view will

be discussed below, maintains that “integrity requires “sticking to one’s principles”,

 

’9 Denying integrity to Huckleberry Finn does not preclude the possibility of other honoriﬁc evaluations
pertaining to him and his conduct; he may, for example, be virtuous (Arpaly, 2002).

4" Interestingly distinct is academic integrity, which seems just to be a certain sort of moral integrity,
characterized by the practical context in which it is displayed. Thus it cannot be said of a person that she
showed academic but not moral integrity.

4' Frankfurt (1988), p. 179.

186

moral or otherwise, in the face of temptation”."2 There seems no reason to doubt the
correspondence of this shared element of philosophical views to our everyday attributions
of integrity. Certainly acts of integrity come neither effortlessly nor without some degree
of risk."3

A resistance feature has been taken to be the very deﬁning feature of integrity in
Cheshire Calhoun’s account of integrity as ‘standing for something’."4 “Integrity,” she
argues, “calls us simultaneously to stand behind our convictions and to take seriously
others’ doubts about them.”45 The person of integrity adopts “the co—deliberative
perspective?“ taking her own deliberative conclusions to be important, both to herself
and to others, while also respecting the claims and criticisms of others. Failure to stand
behind one’s judgments in front of others explains judgments of failures of integrity —
one can lack integrity through making a considered judgment but then abandoning it
cognitively or practically; neither (according to Calhoun) are fanatics, those who push
forward in denial of the advice and criticism of others, as such candidates for integrity.

Standing for something in the sense Calhoun defends, though, is neither integrity
proper nor integrity qualiﬁed, but rather is one aspect of integrity considered generally.
Standing for something is neither sufﬁcient nor even necessary for integrity, and cannot
stand on its own as an account of integrity. At the very least, it suffers from the same
shortcomings as do other contemporary accounts: it is susceptible to straightforward

counterexamples. Many of these illustrate that respecting the claims of others is not

 

‘2 McFall (1987), p. 7.

‘3 How much resistance has to be overcome in order for an act to be one of integrity is not a static matter, of
course; for Iago to act honestly toward a person of whom he is jealous is a contender for the title of act of
integrity, though for me it would likely not be, since I don’t have Iago’s powerful inclinations to subdue in
myself.

4“ Calhoun (1995).

‘5 Calhoun (1995), p. 260.

‘6 Calhoun (1995), p. 258.

187

necessary for integrity. Socrates, in submitting to execution, refused to accept the
considered judgments of those other members of his deliberative community. To be sure,
he did engage his peers in (characteristically one-sided) discussion on the matter, but
there seems to be no reason to doubt that he would remain unconvinced by their
arguments and entreaties. Martin Luther, at the Diet of Worms, certainly acted with
integrity in refusing to recant (“Here I stand; I can do no other”), though he made it clear
that there was nothing anyone but God could say to change his mind (“my conscience is
captive to the word of God”) — a clear instance of a person’s placing himself outside of
the deliberative community.

Fanatics can thus sometimes exhibit integrity in the very acts that suggest their
fanaticism, and examples can be provided at length: Socrates and Luther just mentioned,
as well as both Albert Einstein’s and, earlier and more powerfully, Galileo Galilei’s,
intellectual integrity/fanaticism in supporting an (at the time) absurd view of the physical
world; at least one prominent writer has argued that Mother Teresa is a fanatic at least in
part in light of acts the taxonomic account would call religious integrity."7 Thus it is not
at all clear that Cal houn’s denial of integrity to so-called fanatics is tenable.

It is not even clear that there needs to be a deliberative community, in the sense of
a group of relative equals who listen to, consider and respect the reasoned opinions of
each other. Imagine (as some people do) a black woman named Rosa Parks who had a
bold, singular and completely novel idea, perhaps born of frustration with the very lack of

a deliberative community, to refuse to give up her seat on the bus. This Rosa Parks (like

 

‘7 Hitchens (2003).

188

the actual Rosa Parks) acts with integrity, though Calhoun’s account cannot explain this
fact.

Other situations suggest that there may be tension within Calhoun’s view itself,
between the central characteristic of integrity (standing for one’s deliberative
conclusions) and the requirement that a person must stand for something only as a
member of a deliberative community. When, imagine, standing for one’s deliberative
conclusion that nonwhite persons have and deserve exactly the same moral
considerations as do white persons against members of the National Socialist Movement,
it is odd to think that respecting the deliberative capacity of the Neo-Nazis, taking
seriously their doubts about the basic rights of non-Christians, homosexuals, and blacks,
could possibly contribute to one’s integrity. In the well-worn example of the Nazis at the
door of the house in which the Jews are being sheltered, certainly the provider of shelter
acts with integrity. It is quite implausible to think that having a reasonable conversation
(actual or even hypothetical) with the Nazis, carefully considering their reasons for their
project of extermination and their underlying racial hatred, would make the attribution of
integrity any more strong or appropriate.“ In such a case, refusing to give any uptake
whatsoever to arguments supporting racially-grounded political violence might be just
what integrity demands.49

Calhoun’s view also fails to account for cases where a person such as George can
stand for either of two things: his personal commitments or a moral principle; the

distinction in George’s options is invisible to Calhoun’s account. This observation is

 

‘8 It may, of course, be prudent to feign such interest in the Nazi’s opinions.

‘9 It is likely that this intuition rests (with a bit of circularity, admittedly) on the assumption that moral
demands ground true integrity, and further that the relevant moral consideration here is respecting the
personhood and moral self-esteem of nonwhite people by excluding Neo-Nazis from discourse entirely.

189

especially relevant here since distinctions between the sorts of integrity outlined in the
taxonomic view hinge on distinctions in the source of resistance: acts of personal
integrity are grounded in one’s personal principles; religious integrity stems from
religious commitments; acts of interpersonal integrity ﬁnd their strength in personal
relationships; and so on.

A ﬁnal point to note about the resistance requirement is that it is the most
apparent characteristic that distinguishes moral integrity from the closely related notion
of moral rigorism (the trait of always actively doing what one thinks is morally best),
since one can be a moral rigorist even if she never experiences any resistance to her
views or actions.50

Lynne McFall has offered a view of integrity that shares some characteristics with
the taxonomic view, but I disagree with some of the details of her account, as well as
with her ultimate conclusion. For McFaIl, integrity is ultimately a matter of acting from
one’s identity-conferring commitments, “a set of principles or commitments that makes us
who we are.”5| Defeasible commitments, on the other hand, do not ground integrity -
being defeasible, one can leave them behind with no remorse or loss to one’s sense of
self-value — but neither do just any identity-conferring commitments. The grounds of
personal integrity must be important. The grounds of moral integrity must,
unsurprisingly, be moral.

That the grounds of integrity must be important (in some sense than more than
merely subjectively important) should be apparent; trivial contexts are not the places we

look for integrity (this is another feature of integrity that is shared at least implicitly by all

 

5" Another distinguishing feature is the diachronic nature of rigorism; a speciﬁc act can be one of integrity,
but no one act is an act of moral rigorism.
5' McFall (1987), p. 13.

190

contemporary philosophical views). But that acts of integrity must be moral is only clear,
I am arguing, on the taxonomic view, and McFall’s argument for the moral grounds of
integrity is unconvincing. Her defense involves a distinction between personal morality
and social morality. Personal morality is “that set of moral principles and commitments
that I adhere to that I do not expect everyone to adhere to and that need not be
characterized by impartiality.”52 Social morality is, conversely, both universal and
impartial. This distinction is necessary to sidestep the Williamsian observation that
integrity can be and often is at odds with (impartial) morality; McFall’s move, then, is to
conceive of the partial commitments that are opposed to what Williams considers moral
demands as themselves moral — personally moral.

McFall maintains that personal integrity is a precondition of moral integrity
insofar as for a person to act with moral integrity, the moral principles that ground her
action must be her own; about this there should be no dispute. McFall’s ultimate point is
that “moral integrity is as much a threat to social morality as [is] personal integrity.”53 By
this she means that personal moral integrity (the sort of integrity grounded in non-
universal and non-impartial principles) is a threat to morality of the universal and
impartial sort just as much as (per Williams) integrity qua acting from one’s own
possibly amoral principles is such a threat.

There are two problems with McFall’s conclusion. One is that the distinction on
which it depends, that between personal and social morality, isn’t anywhere near a
compete categorization. There are moral stances one can take that are universal and

partial (‘honor thy father and mother’; ‘everyone ought to look out for number one’);

 

52 McFall (1987), p. 17.
5* McFall (1987), p. 20.

191

there are moral stances that are non-universal and impartial (‘my parents raised me to
consider everyone’s interests equally, but you might justiﬁably disagree’), and her view
gives no account of cases in which these ﬁgure.

Second, it is not clear that in the end her distance from Williams’s view is
anything but nominal. Calling certain non—universal and non-impartial stances ‘moral’
may be legitimate and even illuminating in a number of ways, but does not undermine
Williams’s point that if universal and impartial considerations are the ﬁlter through which
other values must pass, there is a real loss to leading the sort of lives in which partial and
subjective commitments feature as prominently as they often do.

To summarize a bit, the taxonomic view of integrity has a number of strengths.
Most importantly, it explains a wide range of our judgments, a feature I’ve appealed to in
defending the plausibility of the view. It also accounts for cases in which a person can
show integrity through performing either of two conﬂicting options — George’s taking the
job on moral grounds, or refusing it on personal principle — so long as the action
performed is grounded in principle against resistance.

Noting another substantial strength of the View of integrity proper as moral
integrity returns us to an issue left unresolved in the previous chapter’s discussion of
autonomy. Recall that a number of accounts of autonomy emphasized the relevance of
the moral characteristics of action in evaluations of actors; though I attempted to show
this implausible, the thought that an autonomous action must be in some sense a moral
action might well have an intuitive attraction. The view of integrity on offer here can be

seen as at least a partial vindication and explanation of that intuition.

192

Recall that on the taxonomy into which moral integrity ﬁts, personal integrity —
the acting on one’s own principles in the face of pressure to do otherwise — is a necessary
but not sufﬁcient condition for integrity proper. This condition along with the condition
that these principles are moral ones are conjointly sufﬁcient for integrity. That is, true
integrity is possible only for the person whose practically effective principles are moral
ones. The structure of this account resembles, then, the structure of content-speciﬁc
accounts of autonomy but, as I’ve argued, offers a more plausible locus of morality in the
life of the agent.

Integrity and unity

It may yet be unclear how the moral view of integrity supports the ideal of unity
thesis. That integrity rests essentially on an action’s being grounded in a moral principle
(rather than in endorsement, say) implies that even a radically disuniﬁed person could act
with integrity, so long as the right principle is in play.

Consider just what we might mean when talking of principles. We speak
sometimes of basic principles, the core concepts of some area of investigation or
practice. As examples, consider that Freud held the basic principles governing human
behavior to be eros and thanatos (the life and death instincts, respectively); the four
principles of visual art are proportion, scale, balance, and unity within variety; equality is
a basic democratic principle. The focus here is on the overarching nature of the speciﬁc
principle — so, for example, equality is to ground or at least feature into all (or most of, or
the most important of) the speciﬁcs of democratic practice.

‘Principle’ in the moral context can refer to an unwaveringly held rule: ‘Honesty

is her highest principle.’ In this sense, the degree of strength, importance, or scope seems

193

to be emphasized by calling something a principle rather than simply a rule or guideline;
the occasional mendacity of someone for whom honesty is (merely) a guideline should
not be surprising. ‘Moral principle’ is distinct not only from ‘guideline’ but also from
‘moral belief’, in at least the following way. A thief might hold the moral belief that it is
wrong to steal (i.e. that stealing is a wrong-making consideration); she asserts the belief
to her conﬁdants, holds an epistemic pro-attitude toward the belief, can explain what
makes stealing wrong, and so on — but she nonetheless has to steal to feed herself and her
family. We wouldn’t say, though, that she has a moral principle regarding the wrongness
of theft.54 Indeed, it is one of the harms of certain sets of social conditions that some
persons are prevented from possessing such principles, though they might possess such a
belief. To attribute a moral principle to someone is thus often to suggest a stronger and
more reliable connection with action than is suggested by an attribution of moral belief.55
To call someone, more broadly, a principled person implies something further
about the content of her principles, for while ‘proﬁt regardless of the human cost’ might
be someone’s highest principle, we wouldn’t call this person ‘principled.’ A principled
person tends to be so considered if she is one for whom the positively evaluated
principles of honesty, respect, charity, and so on are overarching concerns, and so guide

her actions.

 

5" It may be (I think it is) that the wrong-makingness of theft is the case, and that this is necessarily,
conceptually true; if there is nothing wrong with my appropriating that object that you are in possession of,
then it has to be the case that the object wasn’t in fact your property (since if it was, there is at least the
slight wrong of violating your right to property). If I am correct about the nature of theft, then the example
above is more broad: everyone (on pain of conceptual incoherence) believes that theft is wrong, though
fewer hold this as a principle.

55 It should be clear that this analysis of ‘principle’ and ‘belief’ and so on is an analysis of quotidian usage
of and distinctions between the terms. The powerful philosophical arguments regarding the motivational
power of practical beliefs are, given this context, not to the point here.

194

Not all philosophical theories, though, hold morality and principles to be
associated in this way. Moral particularism is one such theory, and an explanation of just
what particularism takes a principle to be will help us to understand how principles ﬁgure
into the lives of persons who act with integrity.56

The philosophical debate between moral particularists and moral generalists
employs the notion of a principle quite heavily.57 ‘Moral principle’ - which, by the way,
is something that generalists claim we need in order to engage in moral practice, and
particularists deny that we so need — has been understood in two distinct ways that
feature into the particularist debate. One understanding carries the sense of an absolute
guideline, as in Kant’s argument in favor of the principle of honesty. Jonathan Dancy
calls this conception, which he takes to be “largely discredited”’8, the subsumptive option.
A principle in this sense is a rule that admits of no exceptions or mitigating conditions,
and is decisive when it applies. For instance, to take honesty to be a moral principle in
this sense is to hold that when the situation is such that one can be honest or dishonest,
one must be honest, full stop. This conception agrees fairly well with some common
usage; someone who claims to subscribe to the principle of benevolence but then neglects
to help an injured person might expected to explain her action in a certain way: why the
action wouldn’t have been one of benevolence.

Some claim that this absolutist conception, the subsumptive option, doesn’t

exhaustively characterize the ways in which principles can be understood, and argue that

 

5" Another moral view that relies heavily on the notion of a principle, as we saw earlier in the discussion of
Bernard Williams’s concerns about integrity, is Utilitarianism. The main reason I choose to illustrate the
matter by appeal to Particularism is that I take the Particularist position to pose a problem for my account
of integrity, which I take up in the following section.

57 I mention this debate not only as a means to introduce some useful conceptualizations, but also since, as I
will point out, there is a conﬂict between the notion of integrity I support and the desirability of leading a
distinctly particularist sort of life.

5" Dancy (2004), p. 3.

195

there are (many) defeasible moral principles. Benevolence, taken as a defeasible rather
than absolute principle, might be understandably foregone when the cost is too high. The
existence of conﬂict between the reasons given by various principles also recommends
understanding them as defeasible. Honesty might (more plausibly) be such a defeasible
principle, and charitableness another; these two principles each give reasons for acting in
a certain manner, reasons which may in some easily imaginable cases conﬂict. A moral
principle, on this reading, points to non-trumping considerations for action, and at best
plays into an all-things-considered judgment of what one ought to do.

It is important to keep in mind that these last two conceptions of principles are
more metaphysical than practical conceptions, as they are concerned with “theories about
reasons rather than with life”;59 they are certainly more metaphysical than moral-
psychological. But there are illuminating and relevant points of contact between moral
principles (of the sort ‘lying is wrong’) and what is at issue in the notion of integrity that
is supported here: that is, moral principles taken up as personal principles — as in a person
who takes up as her own principle ‘lying is wrong’.

Even Jonathan Dancy, who denies that moral practice best be grounded in
principles, admits that as a matter of fact “people order their lives (and the lives of others)
according to their principles”."0 What is it for a person to order her life according to her
principles? The answer to this question uncovers the connection between integrity and
the ideal of the unity of identity.

We might understand a person’s principles by analogy to principles themselves.

For example, much as equality is a democratic principle insofar as it is supposed to

 

59 Dancy (2004), p. 2.
6° Dancy (2004), p. 2.

196

ground or at least feature into all the speciﬁcs of democratic practice, a person’s
principles ground or at least ﬁgure into the speciﬁcs of her life. Someone whose
principle is honesty is largely honest; in the life of a person among whose principles is
fairness, her actions will be marked by fairness, whatever their content."1

Whether a speciﬁc one of a person’s principles is absolute or defeasible depends
on the possibility of acting on the principle, conceptual features of the principle itself, and
a person’s relationship to the principle, facets which I will discuss brieﬂy in turn. First.
many of us take honesty not to be a good choice for an absolute principle — there are
simply too many situations in which variously, the amount of harm to be avoided by
lying is too great to recommend honesty, other moral demands preclude honesty, or
personal (or some other sort) of integrity demands dissemblance. Such a principle,
because of real-world practical demands, is best not taken to be an absolute principle.

Second, the degree of defeasibility of such principles might also be determined
primarily with reference to the principles themselves — the concepts rather than the
practice. It might plausibly be argued, say, that the principle of freedom of speech ought
to be subordinate to the principle of harm. Such an argument might appeal primarily (or
initially) to just what counts as ‘speech’ and what counts as ‘harm’ — that is, we will
determine just what the principles themselves are.”

But most importantly to the present discussion, when considering principles as

characteristics of persons — ‘benevolence is her principle’ — the matter is different. For

 

6’ I am not claiming that every person who holds onto some so-called principle, e.g. the principle of
freedom of speech, holds it as her principle; freedom of speech ﬁgures into speciﬁc human lives in
different ways than it ﬁgures into the behavior of the state. That is, a person can subscribe to something
that in some other context takes the form of a principle without it taking the same role in her own behavior.
62 This is not to deny that adjudicating between the relative values of principles does not sometimes or even
best involve consideration of the practical effects of holding up one or the other of these principles as more
important.

197

while a person may well consider her principle of honesty to override her principle of
kindness, this ranking need not be so in virtue of some objective facts about the value of
honesty and the value of kindness (perhaps she doesn’t know these facts). This ranking
can be grounded in the relative degrees to which she identiﬁes with these principles.

Recall the various modes of identiﬁcation that were discussed and defended
above. Identiﬁcation can be described in narrative terms — as the construction and
attribution of stories on the basis of which we can comprehend our own and others’
actions. Identiﬁcation might be thought of as a self-description on the basis of which a
person values herself. Identities are in part structures of desires, or are composed of
group afﬁliations. At the root of these conceptions, I argued, is the notion of
endorsement — the positive evaluation or acceptance that is marked by a lack of internal
conﬂict regarding the identity element.

It is not true, of course, that a person with principles cannot experience conﬂict
regarding them. Sometimes quite the opposite is the case. The example of the conﬂict
between honesty and benevolence is a famous one, of course.63 A person whose principle
is healthy living might experience an overwhelming urge to eat an entire plate of cookies,
or to smoke a cigarette. The uniﬁcation that characterizes the identity of a person who
has principles is, in fact, often enough initiated by internal conﬂict.

To attempt to structure one’s life in such a way as to be able consistently to
endorse a speciﬁc principle - in order, that is, for the principle to be one’s own — involves
unifying one’s identity. This uniﬁcation can be described with the terms of any of the

accounts of personal identity that have been outlined above. A woman whose narrative

 

6" As evidenced by the common resistance to Kant’s “On a Supposed Right to Lie Because of Philanthropic
Concerns.”

198

identity elements include ‘good Catholic’ and ‘radical feminist’ will (very) likely
experience internal conﬂict regarding them. It is not implausible to think that such
conﬂict might be so damaging as to motivate conciliation. Suppose she takes up
feminism as her principle, or comes to realize about herself that she does, unbeknownst to
herself, hold feminism as a principle."4 Feminist demands then take precedence in her
life — precedence over, among other things, religious claims. She adopts feminist
principles, endorses feminism, and uniﬁes her identity around it. She thereby reduces
(internal) conﬂict with religious aspects of her identity; when faced with a decision
whether to support Planned Parenthood, conﬂict is lessened by the fact that though she
may still be a Catholic, her principles are feminist ones.

Identities grounded in group membership are similar. Consider the identity
‘African-American’, doubly interesting here because this supposedly unitary identity is
actually composed of a historical/racial element as well as a national element, a pair
which can easily be imagined to ground conﬂict."5 One with such conﬂict might take up
racial (that is, rather than national) ﬁdelity as a principle.

Desires can be at the core of principles as well."6 One who desires that political
society have a socialist sort of organization might desire that her desire persist; that is,
she cares about socialism. When she takes up socialism as her principle, she organizes

her life around the principle, by voting for certain political candidates, joining appropriate

 

6" Remember that since endorsement is characterized by an absence of active internal conﬂict, a person can
be said to endorse something she is not aware she identiﬁes with. Thus, one or more of a person’s
principles can be unknown to her, and that a person holds X as a principle is often enough not a decision of
hers but rather a discovery about herself.

65 I write “supposedly unitary”, because, for example, of the common sorts of race/gender/disability
information forms that conﬂate racial and national identity with the unitary category ‘African-American’ in
the same way such forms conﬂate sexual and gender identity by allowing only two choices (ignoring, say,
biological males who gender identify as other than male).

6" Neither this nor any of the foregoing claims entail that principles do not ineliminably contain other
elements such as beliefs, for example.

199

political organizations, and disavowing certain of her other desires, such as the desire to
own a business.

Though conﬂict has a role in the adoption of principles (and thus in identity
uniﬁcation), it is less clear that a person can be ambivalent regarding her principles. If
she is unsure which principle to adopt, neither of the principles are yet hers (having not
yet been adopted). A person’s principles, when they themselves conﬂict, are not her
principles at all. A person cannot concurrently hold as her principles both feminism and
ﬁdelity to the sexual status quo. Both of these principles cannot, it should be apparent, be
core organizing features of her practical life. She may espouse feminist beliefs, even
sincerely so, but be internally compelled to remain relatively subservient; in this case
neither one is her principle. She may be (sincerely) faithful to the status quo, but desire
that she be otherwise, that she be a feminist; in this case she has a principle, though it is
one that is on it’s way out, so to speak. Conﬂicting principles are thus either not
principles, or are at best potential (though not potentially coexisting) principles.

In summary, then, to act with integrity is to act on a principle in the face of
resistance. When that principle is a moral one, the act is one of moral integrity — what is
often considered integrity proper. Because, as I’ve argued, principles are composed of
just those elements that constitute personal identities, and because principles guide action
in a structured, coherent, holistic way, acting with integrity is at odds with identity
conﬂict.

Integrity and particularism
Mentioning, as I did above, the debate over moral particularism uncovers a

possibly unacceptable entailment of the view of integrity as moral integrity. In this

200

section I will address this potential problem for the account of integrity I’ve defended,
ﬁrst outlining brieﬂy Jonathan Dancy’s moral particularism and how he ﬁgures integrity
into it, then showing at least two ways in which integrity as I characterize it appears to be
in conﬂict with particularism. I will proceed to explain how particularism does not
undermine the moralized view of integrity defended here, and then note why Dancy’s and
my sort of principles are not as different (as some readers may already have noticed) as
they seem.

Particularism’s two main claims are these: the practice of morality does not
require the application of general moral principles to speciﬁc cases, and doing so will
often enough fail to give morally correct results. Both these claims depend on the theory
of contributory reasons: what is a reason for thing in one case may in another case be a
reason not to (1), or even be no reason at all. For example, ‘because that man possesses it’
is sometimes a reason against Robin’s taking it, though if the man is a rich man and
Robin is Robin Hood, then the reason is a reason in favor of Robin’s taking it. If Robin
Hood acts on the principle ‘taking someone else’s possessions is wrong’ he will in many
cases fail to act correctly.

Now, Dancy takes integrity to have the status of a contributory reason.“ That
means that ‘because the act is one of integrity’ can be a reason for doing it, or be a reason
against doing it, or be no reason at all. On a certain conception of integrity, this is
initially plausible. If we take integrity to be commitment to one’s deepest personal ideals

(a‘ la Bernard Williams), that an act is one of integrity is a reason for the biochemist

 

‘7 Dancy (2004), p. 121.

201

George to do it, though it is a reason for a mobster beginning a process of self-reform not
to do it.

There are a number of problems with taking integrity itself to be a contributory
reason, though. For one, the assumption that the integrity of an act yields some further
reason, over and above that act’s being, as on the Williamsian view for example, an
expression of one’s deepest personal ideals, is questionable. Recall George’s decision to
refrain from accepting the weapons research job. George’s reason (one of them, anyway)
for acting as he did is that investigation into new and better weapons technology is
morally wrong, at least as George sees the moral situation. If George acted as he did
because weapons research is wrong, the fact that the wrongness of weapons research
ﬁgures as one of George’s personal ideals cannot give George yet another reason to act as
he did, since this is not another further fact about the situation. To make the point with a
Wittgensteinian ﬂavor: there is nothing left after we subtract from George’s acting on his
personal ideal that weapons research is wrong George’s acting on the wrongness of
weapons research?"

Nor is it clear that the motivational aspect of such a reason would be coherent. A
person of integrity does not, for example, give her last dollar to a local food bank because
it is an act of integrity — she does so because it will provide food for the hungry,
something she thinks she ought to help remedy. Insofar as an act is done because it is an
act of integrity, it isn’t. This is true both on William’s view (which the present discussion

has so far been implying) as well as on the taxonomic view. If integrity is just personal

 

6" To be sure, George’s personal ideal — that is, his principle — regarding the wrongness of weapons
research implies other further reasons: because his principle is that weapons research is wrong it is likely
that he also is against the employing of such weapons, a moral stance which provides him an additional
reason. The point, though, is that even where the propositional content of a reason is also (part of) the
propositional content of a principle, this does not itself constitute a further reason.

202

integrity, that is, acting from one’s personal ideals, a person might well take ‘acting with
integrity’ as a personal ideal. But this itself has no content — there is no action that is just
acting with integrity, although refusing to promote weapons research might be an act of
integrity, as might be drinking the hemlock or nonviolently resisting the colonial
authorities. Thus, ‘because it is an act of integrity’ is a poor candidate for a reason,
contributory or otherwise. From the ﬁrst-personal standpoint, the reason undermines
itself.

But integrity isn’t just acting from one’s deepest personal ideals — integrity is
acting against resistance and on personal principles that are moral ones. Thus ‘act with
integrity’ would itself have to be a moral principle in order for it to ground an act of
integrity. Below, I will argue that ‘act with integrity’ is not itself a moral principle, and
thus can’t itself ground moral integrity.

Remembering that integrity is acting on distinctly moral principles leads to the
other way, alluded to in at the beginning of the current section, in which integrity seems
to be at odds with particularism, a relatively more practical than theoretical conﬂict.
Recall that acting on moral principles is just what the particularist says we ought not to
do, since doing so will cause us to get things wrong, practically speaking. Acting with
integrity is precisely the wrong way to go about being a moral person on the particularist
view.

This is indeed a conﬂict between particularism and the taxonomic view of
integrity. In addressing the conﬂict, I will ﬁrst concede that it is not always advisable,
prudent, or all-things-considered- good to act from moral principle. Consider the moral

principle ‘always give a person what she deserves’ (possibly the strongest contender for

203

the status of universally applicable principle), and a situation in which an armed maniac
threateningly waves her gun around a crowded room and demands moral praise from you.
It would certainly be ill-advised — and wrong — to act with moral integrity in giving the
gunman what she deserves (which is, I take it, moral reprobation).

Such cases suggest that to act with integrity is not always to get things right,
morally speaking. This seems accurate; even denying the claims of particularism, there
are many ways in which an act grounded in moral principle can go wrong, from bad
moral luck to social prohibition against certain persons acting from moral principle.

As an example of the latter, recall the earlier discussion (in Chapter 2, above) of
Ridley Scott’s ﬁlm Thelma and Louise, and the act of self-respect — the killing of
Thelma’s rapist — central to the ﬁlm. Arguably, that act of integrity went badly wrong:
with one exception (the actions of Slocum, the detective who only halfheartedly attempts
to capture Thelma and Louise), their act of resistance is misunderstood as pointless
violence; much physical and psychological harm and property damage follows from the
act; Thelma and Louise take their own lives two days later. There are strong reasons to
think that Thelma and Louise had no morally agreeable options open to them; there seem
no such reasons, though, to deny that they acted with integrity. Once the score is taken,
so to speak, to act out of moral principle can turn out to have been the wrong choice.

But insofar as it is desirable and praiseworthy to act with integrity, the distance
particularism makes between correct acting and principled acting is problematic.
Because particularism incorrectly takes integrity to be a contributory reason, it
misunderstands the value of acting from principle in situations where to do so is to act

with integrity. So rather than particularism posing a problem for the taxonomic view of

204

integrity, the taxonomic view in fact, to some degree, undermines the plausibility of
particularism.

In closing, I am aware that some readers may have noticed an apparent category
mistake in the foregoing discussion. When Dancy speaks of a principles, what he is
referring to is a moral-metaphysical entity that many people think stands in favor of (or
opposed to, as the case may be) a certain action in all cases where it applies. What I am
calling a principle is a certain set of identity elements as they ﬁgure into some certain
person’s identity — a moral-psychological entity. Thus, it might be thought that Dancy
and I have nothing to say to each other regarding integrity, since we are not using
‘principle’ in at all the same way.

But this thought is incorrect. What links the two conceptions of ‘principle’ is this:
a moral principle — ‘charity is always right’, say — qua moral-metaphysical entity is
something that applies to all relevant acts of all persons. Notably here, it applies to all
the relevant acts of each speciﬁc person; looking at the way in which principles of the
sort Dancy denies shows that a person’s adopting of this sort of principle has the effect of
unifying a person’s identity in the way I have described. Dancy himself notes this in
mentioning persons who “act ‘consistently’, where this means making their decisions in
such a way that they can all be ﬁtted in under the same set of principles?"9 The relation
between metaphysical principles and moral-psychological principles can be described in
the opposite direction as well: a person whose personal principle is charity believes it to

be true that charity is always right.70 Moral-metaphysical principles and moral-

 

"9 Dancy (2004), p. 2.
7" Always right for her, at least; a person who adopts a principle, even a moral one, need not think that the
principle applies to other persons.

205

psychological principles are not only closely linked, but they might be thought to be
different ways of talking about the same normative regularities.
In summary

Integrity, that feature of actions by which they are performed in the face of
resistance and on the grounds of one’s personal moral principles, is grounded in the unity
of identity insofar as principles serve to guide and focus large sets of one’s actions and
large segments of her identity.

The taxonomic view of integrity defended in support of this claim was, as we saw,
also one that explains and supports the intuition regarding the content relevance of
autonomy, which I argued to apply rather to integrity. Because true integrity is a
characteristic of actions grounded in one’s personal moral principle, acting with integrity
is not merely a matter of acting out of a uniﬁed identity, but instead requires an identity

uniﬁed around a moral principle or principles.

206

_5_

Love: value and commitment

Our minds are porous and forgetfulness seeps in; I
myself am distorting and losing, under the wearing
away of the years, the face of Beatriz.

- Jorge Luis Borges, “The Aleph”

In Jorge Luis Borges’s ostensibly metaphysical story ”The Aleph”, he writes of a
place in space and time that contains in itself all other places in space and time; all
events, all people and all moments are contained in this Aleph. The reader comes to
realize, though, the Aleph is a ﬁction, for Borges weaves into this story one of a poet,
Carlos Argentino Daneri, who attempts an epic poem that aims to lay out in verse the
entirety of history; the parallel Borges implies between Daneri and the Aleph suggests
that anything all-encompassing is futile, ridiculous, or impossible.

Borges’s story is, in the ﬁnal analysis, about love: the tale’s ultimate parallel is
between the impossible Aleph, the ridiculous Daneri, and the futility of the narrator’s
fully characterizing (or even completely remembering) his love, Beatriz. Indeed, just
below the story’s surface is the won'y not only that the memory of his beloved Beatriz,
but also love itself, is fading; I share this worry, and it is the motivating thought of this

chapter.

207

I also agree with Borges’s implication that love is, like the whole of the universe
or even the whole of human history, not fully characterizable. Thus 1 will not here give
an account of love; though my claims are to be held to standards of consistency, they are
not intended to be comprehensive. There is little to be gained (rather more accurately,
there is something to be lost) by spelling out necessary and sufﬁcient conditions for the
concept and practice of love. Instead, I will focus on two elements that are relevant and
perhaps central to a plausible conception of love: the valuing of, and commitment to,
one’s beloved. I will explain why I take each of these to be important.

To determine the nature of the value in loving is theoretically important since
loving involves — or according to some just isI — a certain sort of valuing. As I will argue,
the nature of the value that is relevant to the practice of loving is practically important as
well: the sorts of characteristics of persons that we are socially prescribed to value are not
the same characteristics that we ought (on other criteria) to value. This discrepancy is
grounded at least in part in a misconception of just where the value of one’s beloved
derives from.

Related to this last point is one about commitment. To commit oneself to another
in a loving relationship can be as fulﬁlling as fairy tales, ﬁlm and other narratives often
portray.2 Romantic commitment is supposed to facilitate an increased ability to further
one’s own projects and desires, for at least a wide range of projects and desires — the
emotional support of a committed partner aids the other’s career pursuits and parenting
goals, to mention just two obvious examples. And at least one recent sociological study

has found that women in marriages marked by commitment on the part of both partners

 

' E.g. Velleman (1999).
2 Pace Trollope, who wrote in Barchester Towers, “There is no happiness in love, except at the end of an
English novel.”

208

report being happier than do other married women (even happier than women in less
committed but more egalitarian marriages).3 But this sort of commitment is seemingly
rare in contemporary western society, and we are, I think, the less fortunate for this rarity.
Commitment, of course, is not always desirable; to pledge oneself to another person
(which is what commitment is commonly taken in no small part to involve) can lead to
psychological, social and even legal entrapment in a life that is unpleasant and even
harmful.

In this chapter, I will argue the following. The mode of valuing in which we
engage in romantic relationships involves bestowed rather than appraised value. This
process is distinctively undermined by identity disunity since romantic love both takes a
singular object and extensively permeates a lover’s identity. Though the focus here will
for largely idiosyncratic reasons be on romantic love, some of the considerations can be
applied mutatis mutandis to paternal, ﬁlial and agapic love.4 In the chapter’s ﬁnal
section, I will defend a conception of ‘commitment as giving oneself’ as a more plausible
alternative to a common conception of ‘commitment as a promise.’ I will then show how
commitment to a romantic partner — like other sorts of commitments that circumscribe
the boundaries of a person’s identity — is tantamount to unifying large areas of the

landscape of one’s identity.

 

3 Wilcox and Nock (2006).

" Most clearly, the ways in which commitment to the object of our love serves as an identity-unifying force
are relevant to love whatever its object. Other features, such as the singularity of the object of love, are not
relevant to other sorts of love, while a bestowal view of value is central to agapic love.

209

Valuing
A claim to objectivity is a claim to speak for all

those whose voices count; it is a license not to
listen.

- Naomi Scheman, Engenderings

It may not be immediately clear how objectivity is relevant to a discussion of
value and loving. It should be more clear that value is an important element of the notion
of love. While love may be a matter of more than valuing, I take it as indisputable that if
Jane loves Jill, Jane values Jill. What is eminently disputable is the nature of this
valuing.

To say that Jane values Jill might be to say that there are certain qualities that Jill
possesses that are inherently — and so objectively — valuable, and Jane’s love for Jill is
(partially) constituted by an apprehension of those qualities. This is what has been called
an appraisal view of love. Altemately, Jane’s loving of Jill might involve Jane’s
bestowal of value onto Jill. On such a bestowal view, the value of Jill as a beloved is
entirely a matter of Jane’s placing of value on her. Some bestowal views allow that in the
initial stages of falling in love, the nature of the value of the beloved is of the appraised
sort, since otherwise it is difﬁcult to understand what would motivate the lover to choose
to bestow value on one possible beloved among many who are equally devoid of the
relevant sort of inherent value.

Other views of value are certainly possible, but it is not apparent that there will be
views not subsumable under the ‘appraisal/bestowal’ distinction. For example, it might

be thought that the value of a beloved person (read ‘value qua beloved person’ here and

210

in what follows) is only apparent to her lover — an appraisal view. Further, none of my
claims here should be taken to imply that Jane and Jill’s relationship cannot have both
appraised and bestowed sorts of value; Jane and Jill’s relationship is (supposing it is a
realistic relationship) not solely a romantic one. They are likely involved simultaneously
in a relation of friendship, and perhaps numerous more instrumental relationships, such as
that between rock-climbing partners. If differing modes of valuing feature in different
sorts of relationships (as I think they might), then differing modes of valuing will feature
in the set of Jane and Jill’s relationship. The discussion here artiﬁcially separates out their
romantic relationship.

This might all be thought to interest only the academic. Whether my beloved is
inherently valuable or is valuable only in virtue of my bestowal certainly cannot matter to
her or to me. While the philosophers go on disputing, she and I will go on loving, none
the worse for the theoretical uncertainty. But this is not entirely true. The nature of
love’s value is of more than theoretical importance. Whether we love in virtue of value
that is thought to be inherent in our beloved, or in virtue of value that is bestowed by the
lover, makes a great deal of practical difference. Hence, I will argue in support of one of
the contending accounts.

The appraisal view

A number of prominent writers take love to be a matter of appraisal of the value
of the beloved object, idea, or person. I don’t take one or another of such accounts to be
authoritative; indeed, not all the important philosophical claims about love are made in
the interest speciﬁcally of defending some certain account over another. Nor to I take

one or another writer’s claims particularly susceptible to criticisms. I will appeal to a few

211

such accounts — that of David Velleman, Iris Murdoch, and Plato — in order to sketch the
outline and a few illustrative details of the sorts of claims that make an appraisal view
distinct.

On appraisal views, love is, to take David Velleman’s characterization as one
example, “an arresting awareness” of “the value inherin g in its object”.5 Within Iris
Murdoch’s suggestive discussion of morality as “attention to reality inspired by,
consisting of, love?" she argues that “the ability to direct attention is love,” a
precondition of the “intellectual ability” of “accurate vision.”7

Claims of love essentially involving an awareness of some valuable quality or
other — usually accompanied by visual metaphors — go back at least to Plato’s
Symposium. There, after considering a number of views of love as a process of making up
for some deﬁciency in the lover, Socrates recounts the arguments of Diotima, who said,

The man who has been thus educated in matters of Love, who has beheld

beautiful things in the right order and correctly, is coming now to the goal

of Loving: all of a sudden he will catch sight of something wonderfully

beautiful in its nature.8
Not only is love thus a matter of ‘beholding’ or ‘catching sight of’ some quality —
namely, beauty -- of the beloved. This quality is in the beloved’s nature — an objectively
apprehendable quality, as Diotima had explained:

It is not beautiful ...in relation to one thing and ugly in relation to another;

nor is it beautiful here but ugly there, as it would be if it were beautiful for
some people and ugly for others.9

 

5 Velleman (1999), p. 360.

6 Murdoch (1970), p. 67.

7 Murdoch (1970), p. 66.

9 Cited in Solomon and Higgins (1991), p 26.
9 Cited in Solomon and Higgins (1991), p 26.

212

Though a number of views can be thus interpreted as appraisal views, only the
source of the value of the beloved is agreed upon; the supposedly valuable qualities
themselves vary. Diotima took beauty to be the only relevant valuable characteristic of
love’s object. Velleman’s account, distinctly, takes the valuable characteristic of the
beloved to be her “capacity to care about things in that reﬂective way which is distinctive
of self-conscious creatures like us”, that is, her moral personhood grounded in her
rational will.” Murdoch implies more simply that love seeks to attend to “the idea of
perfection” which moves us “because it inspires love?” It might seem that Murdoch’s is
not a view about the love of persons (imperfect as we are); if this is true, it is not true
because of the nature of love, but rather because of the way in which we in fact tend to
love: “human love,” she writes, ‘is normally too profoundly possessive and also too
‘mechanical’ to be a place of vision.”12

A more homely appraisal view might take the valuable characteristics inhering in
a person to be such things as her sensitivity, her love of nature, even (following Yeats)
her yellow hair.l3 Whatever the speciﬁc object of the lover’s awareness, it is just that: an
awareness, a perception, an apprehending of some value that exists in the beloved and
independently of the lover.”

There are a number of difﬁculties with appraisal views taken generally. One is

what we might call the problem of substitutability. The problem is this: if loving involves

 

” Velleman (1999), p. 365.

“ Murdoch (1970), p. 62.

'2 Murdoch (1970), p. 75.

’3 "Never shall a young man,/ Thrown into despair/ By those great honey-coloured/ Ramparts at your ear,/

Love you for yourself alone/ And not your yellow hair. "

'4 ‘Independently’ in a metaphysical, but not necessarily causal, sense. Her love for nature might exist only
because it was nourished by her history with her lover, but the point is that the value of this attitude of hers
is not dependent on her lover or their history.

213

an appraisal of objective value, the object of love ought to be substitutable for an object
with the same or, possibly, different but equally valuable qualities.” It is apparent
(perhaps analytically so) that equally valuable things ought to be valued equally, at least
by a person who is not axiologically disabled in some way. We do not tend to think,
though, that a lover presented with a person who is value-identical to her beloved ought
romantically to value this new person. While in other contexts it is legitimate to expect
all to value certain characteristics of persons relatively equally — we justiﬁably expect all
academic analytic philosophers to value any person’s clarity of thought, for instance — it
is not legitimate to expect a person who values the tenderness of her beloved to value that
same quality in some other person.”

There are various replies to this criticism. One defense against this oddity of an
appraisal view is that objects of love are not substitutable because the reason for loving a
person is given by the relationship the lover has to the beloved.17 On such a view
whatever valued qualities of a beloved might be duplicated in another person, but these
two persons are not substitutable as beloveds since the lover shares a unique history and
trajectory of love with only one of them. This might be taken as no more than an
amendment to the list of valuable qualities of the beloved, though; one of my beloved’s
valuable qualities might be that she shared some experience with me, or shares some plan

or project with me.18

 

'5 Only possibly, since distinct sorts of value are not (always) commensurable.

'6 This, of course, assumes a set of practices of loving that does not include such things as forced marriages.
Since I aim to explicate facts about our current practices this assumption is justiﬁed.

'7 Kolodny (2003). By ‘relationship’ Kolodny is referring to “my present concern” together with “our
history of shared concern and activity and [...] my friend or lover’s present concern for me.” (p. 162).

'8 On a slightly different note, the claim that one’s relationship to the beloved gives special (positive) value
to her is not always the case: if a person whom I love has an affair, I might well disvalue her less for this
action if she is someone with whom I have little history or concern; the longer our history, and more
prominent our shared concern, the more 1 might despise her.

214

Velleman’s Kantian account, on which what is valued in loving is the beloved’s
rational will, is able to reply to the problem of substitutability with a distinction between
‘dignity’ and ‘price’. Since the value of quotidian objects is commensurable - the value
of one pair of shoes can be compared to that of another pair of shoes, or to that of a
coffee maker — objects have a price. But if something “is exalted above all price and so
admits of no equivalent, then it has a dignity?” Perhaps this avoids the substitutability
worry, but at a cost: the cost of being unable to judge between the value of you and the
value of her, exalting both above all price, and being compelled to love you both. Call
this, then, the problem of specialness.

The problem of specialness, which Velleman details quite extensively, is the
worry of a beloved that if everyone is distinctly lovable for some invaluable quality, there
is nothing special about oneself, romantically speaking. Velleman takes the worry to be
very deeply rooted, at least from a developmental point of View: “We are told by adults
who love us, and want us to feel loved, that we are special and irreplaceable [...] and we
wonder: If everyone is so special, what’s so special about anyone?”20 Nor is the thought
that one is unique (qualitatively or numerically) a help, since “it is easily defeated by the
very same standard. How valuable can our uniqueness make us if everyone is unique?”21

Given Velleman’s concern for the problem, his solution to it is unsatisfying. He
introduces a distinction between judging value and appreciating value. We properly
judge the value of objects, and in this way rate, rank and compare their value relative to

each other; we properly (to recall Velleman’s abovementioned distinction) consider

objects to have a price. But because persons have dignity rather than price, the proper

 

'9 Kant, Groundwork, 4:434, quoted in Velleman (1999), p. 364.
2° Velleman (1999), p. 363.
2‘ Velleman (1999). P. 363

215

way to apprehend the value of persons is not to judge it but to appreciate it. Because we
appreciate but not judge the value of our beloved, there is “a ban on subjecting him to

9922

comparisons , such as the demanded comparison implicit in the question ‘what makes

me more special than her?’. The question ought not to be asked, even, for being special
“entails being seen to have a value that forbids comparisons.”23

But if everyone has this sort of incomparable value, then an explanation of why
we love only certain persons is necessary. Velleman supposes that this is a question of
why we love a person “for his observable features” such as physical traits, mannerisms,
and (presumably, though unmentioned) character traits such as courage and honesty. At
least regarding mannerisms such as ‘the way he walks’, Velleman claims that we love
people for them because such love is “a response to his gait as an expression or reminder
of his value as a person?” This much seems true, and a valuable observation - that our
tendency to say (and believe) that we value, for instance, our beloved’s tenderness is
better described as a valuing of our beloved for (in virtue of) her tenderness. But such a
description returns us to where we began: if any relevant mannerism or trait of body or
character is an expression or reminder of one’s value as a person, then the question ‘why
love me, rather than him?’ stands unanswered.

The problem of the best beloved is this: even on an appraisal view that does not

include the notion of value as ‘a dignity’, it would seem to follow that the accurate

appraiser of value would be compelled to love certain persons.” Indeed, if there were a

 

’2 Velleman (1999), p. 367.

” Velleman (1999), p. 370.

2" Velleman (1999), p. 371.

” The problem, of course, is worse on an account that does include the notion of a dignity: I am
normatively compelled to love every person. Velleman notes that such love is “constitutionally
impossible”, and our choosing one of all the obligatory beloveds is assisted by the fact that “the human

216

most valuable person, everyone who was an accurate perceiver of value should love her,
thus making the current relatively egalitarian distribution of beloveds somewhat odd.”
The bestowal view (and a hybrid account)

On a bestowal view, the value of the beloved does not exist in her inherently. Her
value is solely a function of her lover’s valuing her — ‘valuing’ here being an active
verb.” The lover does not perceive the value of her beloved, she rather gives her beloved
value, by valuing her. “When we love,” writes Irving Singer, “we create value”.”

In order to illustrate what I take to be the central distinction between bestowal and
appraisal views, it will be valuable to consider the ways in which such views agree. On
both views, a speciﬁc feature of one’s beloved -- that set of behaviors of hers that we
might consider as constituting her sharp wit, for instance — does exist in her. That is, she
really does have a tendency spontaneously to interject intelligent remarks characterized
by subtle punning or inventive metaphor, all with a quick if slightly dark humor. IfI
value my beloved in part for this particular set of behaviors, I perceive that she in fact
exhibits these behaviors.

What distinguishes a bestowal view of value from an appraisal view is the

supposed source or locus of whatever value this set of behaviors has. On an appraisal

 

body and human behavior are imperfect expressions of personhood [. . .] hence the value that makes
someone eligible to be loved does not necessarily make him lovable in our eyes.” (p. 372) Not only is the
shift from moral personhood as a normatively and rationally necessary object of respect to personhood as
making one eligible for loving respect illegitimate and unexplained, but this explanation of Velleman’s is
just that — an explanation, speciﬁcally of why we love selectively. But given Velleman’s project of
reconciling loving with Kantian - that is, impartial and inexorable — moral demands, a justiﬁcation of
selective love seems required.

2" And, it might be noted, making for an odd interpretation of the old saw ‘There is someone for everyone.’
That is, instead of Vx3y(ny), 3ny(ny).

” ‘Value of the beloved’ should be taken throughout to mean ‘value of the beloved to the lover, and this
only in her role as beloved.’ That is, ‘value’ does not here include human value, self-value, value to others,
value to God, and all the other sorts of value (inherent or bestowed) properly attributed to her.

” Singer (1987), p. 392.

217

view, since valuing is perceiving the value of one’s beloved, I perceive the valuable
characteristic of ‘wit’ in her. This claim is both phenomenological and metaphysical; not
only does it seem like I perceive value that she in fact has, but also the phenomenology
could well be accurate — if the value of her behaviors is anywhere, it is (so to speak) in
her. A bestowal view is — if indeed I perceive the value in my beloved — an error theory,
for although the features in virtue of which value is bestowed do exist objectively, the
value is not perceived, but is rather conferred by the lover.29

Thus (as further illustration), the reason why a lover might value the wit of his
beloved, a person who others might rather take to be cynical, ﬂippant or facetious, a
bestowal view explains as follows. Since cynicism, ﬂippancy and facetiousness are each
both descriptive and evaluative terms, who I call witty and you consider ﬂippant are
being described the same by you and by me — where we differ is in our bestowal of value
on the (physical) behaviors we both perceive.”

Because the value of a beloved is not inherent in the beloved but rather bestowed
by the lover, the bestowal view neatly sidesteps the problems of substitutability and
specialness that make appraisal views untenable. The problem of substitutability — that
someone with my same value (or degree of value) should be an adequate romantic

replacement for me — is not an issue on a bestowal view since there is no other person

 

29 That the bestowal view is an error theory will not need to be accepted universally. The phenomenology
of the overall experience of love differs widely between persons, as do certain elements of it such as the
valuing of one’s beloved, and so for some lovers the bestowal view is phenomenologically accurate.

30 This point might be clearer to the reader with analogical examples. A somewhat complex one: suppose I
take a certain man to be a ‘thieving nigger’, while you take him to be a ‘poverty-stricken African American
man’. The point here is that you and I can agree on all the (non-normative) facts - e. g., that he stole
another person’s property, he is homeless, he is unemployed and uneducated, he lives in a society which
systematically denies to him opportunities it allows to white people, and so on - but at the same time
disagree on our evaluative judgments, as evidenced by the difference in terms we use to characterize him.
A very simple illustration: you and I taste the same beer, a heavily-hopped India Pale Ale. You blurt out
‘Eww...bitter!’ and 1 say ‘Yum...bitter!’

218

who has the same (degree of) value as me." Similarly for the problem of specialness:
where on the appraisal view the lover judges (or as Velleman points out, appreciates) the
beloved’s uniqueness, a quality everyone has, the bestowal view holds that the lover
takes only her beloved to be unique or special. Thus, the bestowal view is subject to
neither of these problems.

A bestowal view, like an appraisal view, can be seen to entail a number of odd
notions that demand a bit of discussion. One is the description of valuing that is
inconsistent with (some people’s) phenomenal experience. There are at least two ways in
which one’s experience might fail to accord with the account defended here. One is in
the difference between love’s feeling like a passive happening — ‘falling in love’;
‘overtaken by love’ - as against the apparent active valuing that is supposed to be
undertaken by the lover. One response is that the account of valuing defended here is
not argued to be a comprehensive account of the whole of love (as I explained at the
beginning of the chapter); the feeling of falling is explainable by some other feature of
love.32 Another response is to point out that the phenomenological experience of loving
takes widely varying forms between persons, and so perhaps an appeal to subjective
experience can be illuminating but nowhere near conclusive.

The other phenomenological discrepancy regards the source of the value of one’s

beloved: a person might well feel that his beloved’s wit really is in her, and that he

 

3‘ Except, that is, from the lover’s point of view. But the bestowal view’s recognition of the not uncommon
phenomenon of coming to love someone else — through bestowing value on some other, new, beloved -
further lends credence to such an account.

32 Possible explanations range from ‘falling in love is no more loving itself than falling in the shower is
showering itself’ (that is, just a small part of the total practice, as well as an accident there are good reasons
to try to avoid) to the observation that the way in which we conceive of a practice such as loving can
radically affect our subjective experience of that practice (and so our experience reﬂects and is caused by
an incorrect understanding of what we are in fact doing), an observation that carries some substantial
argumentative weight in the present chapter.

219

himself sees that valuable quality of hers. This, I suspect, is the experience that motivates
the construction of appraisal views (as well as the ubiquitous visual metaphors), and
understandably so. But to motivate the construction of an account is not the same as
supporting it; this, in the face of the deep conceptual problems with the appraisal view, is
something that the phenomenal experience of some set of persons is not sufﬁcient to do.
Another, non-phenomenological, issue might be raised against the implication of
the bestowal view that any personal characteristics or any person (however horrible)
might romantically be valued. The thought that love might be given to anyone,
regardless of their worth, is admittedly consistent with observation, for even more
bafﬂing than the extraordinary nastiness that characterizes certain people is the fact that,
often, they are loved by someone (consider Henry VIII’s six wives, at least some of
whom loved him in some substantial sense, or the high value Bonnie placed on Clyde).
In this regard, then, the bestowal view shows a distinct phenomenological strength. Nor
is there any real weakness entailed by this implication, for a person who values her
beloved for, say, his domineering, violent treatment of her can (or better, should) still be
strongly criticized. She (we can legitimately think or say) should not value that in him
since if she did not she would be happier, she would have more personal freedom, others
would respect her more, she would have the sort of self-respect required of a Kantian
moral agent, and so on. Further, these considerations might be expected to have some
traction on her. If she indeed wants happiness (or freedom, et al) and the connection
between these of her desires and a change in her valuing is well-drawn she will ceteris

paribus be drawn toward revising her objects of value.33 Some people will, of course,

 

’3 Ceteris paribus, that is, since she must be minimally practically rational, and there are not stronger (to
her) considerations moving her to maintain her current valuing practices, and so on.

220

continue to value such things. But (to make the speciﬁc point at hand) what the bestowal
view denies is that what is wrong with her valuing what she does is that what she values
simply isn’t valuable. Such an appraisal view is, as I continue to argue, implausible and
unnecessary to make sense of (and improve) our practices of valuing.

To return to addressing possible criticisms, it might also appear that a bestowal
view has the lover bestowing value gratuitously. If a lover’s evaluations are not a
response to some real value, then they are not a response to anything. Perhaps one’s
beloved ought not to feel so special after all, if she was chosen as an object of love
arbitrarily, and even at that is given more value than she deserves. Singer attempts to
respond to these sorts of entailments of a bestowal view; I will outline his responses
below, and then follow with what I will argue to be more plausible answers to these
concerns.

Singer claims to defend a hybrid account on which love is marked by a “complex
network of bestowals and appraisals”.”4 Though he admits that bestowal is necessary for
love, he distinguishes two sorts of appraisals that occur as well. What he calls objective
appraisals are made using the value scale of one’s community. Singer offers as
illustration the evaluations of judges at a beauty contest. Individual appraisals are those
made with special reference to one’s own interests and preferences.

It seems that the straits between the Scylla of appraisal and the Charybdis of
bestowal are perhaps too narrow to navigate successfully, and Singer scrapes the rocks at
many points. In an elaboration of bestowal he claims, “the lover accepts the beloved,

responds afﬁnnatively to her attributes even though he knows that from an appraisive

 

3" Singer (1987), p. 390.

221

point of view they could be evaluated in a less positive way”.35 But to respond, and to
evaluate in the sense of ‘judging’ used here, is just what happens in appraisal. If this
claim of Singer’s is true, then value is appraised. But at another point he claims, “all
appraisals must ultimately depend on bestowal since they presuppose that human beings
give importance to the satisfying of their needs and desires. Without such bestowal
nothing could take on value of any sort”.” If this is true, all value is bestowed. In any
case, it is ﬂatly contradictory with his claim that “bestowal would not occur in human
beings without appraisal”.37 Finally, in an argument aimed at showing the incoherence of
the notion of ‘inﬁnite worth’, he writes, “in themselves [...] people have neither “inﬁnite
worth” nor inﬁnite worthlessness. They acquire value and the lack of it only as they are
valued or value themselves in their relations with one another”.38 Such acquiring is the
deﬁning feature of a bestowal view.

To summarize a bit: I have argued that there are substantial problems with
appraisal views, especially with but not limited to Velleman’s Kantian account; I have
also shown how at least one attempted hybrid view39 -— Singer’s — contradictorily
conﬂates two distinct sorts of valuing.

Thus the abovementioned apprehensions toward accepting a bestowal view still
remain. One, recall, is that there is a certain arbitrary ﬂavor to bestowal; in addressing
this concern here I will make it more explicit by casting it in two closely related ways.

One way to characterize the concern is that bestowal seems to allow a limitless, willy-

 

” Singer (1987), p. 392.

3" Singer (1987), p. 393.

97 Singer (1994), p. 141. In Singer’s defense, this claim might be a purely psychological one, and the
former (that bestowal is the foundation of all value) an axiological one. I am not aware, though, that Singer
explains how the practice and theory of valuing could be so fundamentally opposed.

3" Singer (1987), p. 403.

39 The only one of which I am aware; the literature on valuing in love is not large. I do not deny that a
hybrid view of some sort is not possible.

222

nilly casting about of value. Velleman raises this sort of issue in motivating his own
appraisal view against views which allow that “we needn’t see or be moved to see [one’s
beloved] as he really is?” He claims that lacking accurate perception of the real value of
one’s beloved often entails “overvaluation.”"l Singer, more richly, claims that bestowal
without appraisal entails

love [to be] a ﬂight of fancy, or an emotional explosion that destroys the

humdrum facts of ordinary life, or a blind allegiance, or an afﬁrmation

(however sublime) that ascribes objective perfection despite what

everyone knows about the limits in mere humanity.“2
In these claims, Velleman and Singer confuse the nature of value with its excess. Simply
because something is bestowed gratis rather than in response to some prior state of affairs
does not entail that such bestowal will be unjustiﬁably gratuitous. Consider here gift-
giving versus bill-paying, and the fact that while gifts are bestowed, only rarely are they
bestowed excessively.

The second way to characterize the concern about arbitrariness comes from a
more ﬁrst-personal point of view. If value is bestowed, rather than judged, I may wonder
whether I warrant her love —- another version of the problem of specialness. After all, it is
not in virtue of some objectively valuable quality of mine that she loves me. There seems
no reason for her to love me rather than him.

While it is indeed not by the light of some inherent value that a lover sees her

beloved as she does, all is not darkness. This woman who I call bold, you call brazen. I

bestow value on her, while you do not."3 In my loving her, I provide the light by which

 

4° Velleman (1999), p. 351.

4‘ Velleman (1999), p. 351.

42 Singer (1987), p. 394.

43 “Some say thy fault is youth, some wantonness;/ Some say thy grace is youth and gentle sport;/ Both
grace and faults are lov’d of more and less:/ Thou mak’st faults graces that to thee resort./ As on the ﬁnger

223

she may be seen, and thus the reason for loving her exists (satisfying those who expect
that love is to be responsive to reasons). Indeed, in bestowing value on her, I may
provide the light by which she can see herself as valuable and thus see the reason she is
loved legitimately and non-arbitrarily.

This thought — that only through being loved can a person see herself as a proper
object of love — is not mere romanticism. Certainly the value that others bestow on us
contributes to our self-conceptions as valuable persons -— this is a not trivial feature of
awards and honors, for instance. But what is more, valuing oneself may be practially
possible only insofar as one is valued by others. Recall the deferential wife, who
believably was supposed to lack a sense of self-value because of her husband’s failure to
value her. Reconsider also Lindemann’s notion of master narratives, those often third-
personally constructed character-narratives. A master narrative such as ‘single mother’
only now (not, say, as the character-type was understood in the 1950’s) carries with it a
bestowal of value; now (much more than 50 years ago) is a single mother as such able to
value herself. Through being valued, a person can come to see herself as valuable;
through being loved, a person can come to see herself warranting that love.

Even if the beloved is not valued in virtue of his inherently valuable qualities,
there is something in virtue of which he is valued; there is something, though inherently
valueless, to which the attributions of boldness and brazenness refer. Admittedly,
though, bestowal refers in large part back to the lover herself. She is the creator of value,
the artist to the canvas that is her beloved. Value is bestowed on the beloved in virtue of

facets of the lover as well as what might be thought to be mere contingencies — the

 

of a throned queen/ The basest jewel will be well esteem’d;/ So are those errors that in thee are seen/ To
truth translated, and for true things deem’d. . .” Shakespeare, Sonnet XCVI.

224

 

11 rant

i.l|!l¢l|ll.ir 1

 

‘1 it «will

incident luck of a chance meeting, the unlikely state of interpersonal compatibility, that
obscure thing called ‘chemistry’. Love is fortuitous, a fact the bestowal view does not
attempt to hide.

The bestowal view is thus the more plausible alternative since it answers the two
most troubling problems to which appraisal views are subject: the problems of
substitutability and specialness. But a bestowal view, as mentioned earlier, is taken by
some to be an error theory, failing to correspond to some people’s experience of loving.
There seem to be three ways to respond to an error theory. One is to defend an
alternative theory, though I’ve argued above that the two alternatives defended in the
literature — appraisal and a hybrid account — are unsatisfactory. Another is to accept the
theory, and accept that our phenomenal experience is simply nonveridical. The third
option, as I ’11 recommend in what follows, is to reconceive our own understanding of our
practices, thus bringing phenomenology in line. The reason we ought to reconcieve our
practices of valuing is because, as I now explain, such reconception will help to remedy a
set of oppressive effects of loving. Many of the instances of romantic love that are
damaging are distinctively damaging to women, and this in light of socially-prescribed
patterns of valuing that direct women to value certain characteristics in their romantic
partners.

Value and oppression

The oppressive effects of loving ﬁnd their roots in whom we love, or more
accurately, with whom we ought to love. The question of who we ought to love is based
in no small part on the question of what personal characteristics ought to be valued in a

beloved. Men ought to seek out women who are physically beautiful, domestically

225

talented, conﬁdent yet appropriately submissive. These are the valuable qualities of a
wife, unless my cultural radar is substantially out of tune.

Who ought she to love? A man, ﬁrst and foremost. Beyond this, a man who
allows her her career, and compliments her on her good hair days. She should value a
man who takes control of the family’s affairs (except for the monotonous elements of
childcare, of course) and allows her some bare minimum of moral personhood.

The point of these nearly accurate caricatures is this: men are socially guided to
value qualities of women that are beneﬁcial to men (sexual attractiveness, concern for
male comfort, and so on). Women are socially guided to value qualities of men that are
beneﬁcial to men (thinly restrained dominance, freedom from emotional burden)."" That
is, the socially recommended values of desirable beloveds are values that are
overwhelmingly beneﬁcial not to women but to men.

This set of conditions, whereby the system of norms is structured such that they
ensure the beneﬁt of the members of one group over the members of other groups, is a
commonly accepted deﬁning example of oppression. Thus, insofar as love is partially
characterized by a certain sort of valuing, and insofar as the sort of valuing that occurs in
our own practices of loving is a sort that tends to beneﬁt men as it harms women, love is
oppressive.

There are of course many ways in which to pursue the amelioration of this state of
affairs. Most commonly, the content of the purportedly valuable characteristics is
recommended for change. For example, the idea that the male partner might be the

primary domestic caretaker is gaining adherents, as is the virtually mainstream idea that

 

4“ The debt here to Marilyn Frye should be apparent. See Frye (1983).

226

 

 

romantic partnerships can be non-heterosexual. Because, as I’ve argued, the social
inﬂuences of identity are neither monolithic nor homogeneous, much social reevaluation
of this sort can have favorable and substantial effects. This is not to say that none of
these sorts of efforts at changing the content of norms are misguided; for example,
assertiveness training programs for women, involving both interpersonal communication
as well as assault prevention, can be seen as unfairly shifting responsibility for change. 45
In these speciﬁc cases, the most practically feasible option has the unfortunate effect of
perpetuating the unequal burden of responsibility.

But it is not only the speciﬁc content of these norms that makes for this damaging
oppression. The supposed nature of the value plays an important role in the systematicity
of the oppressive norms, as I’ll argue next.

The bestowal view reconsidered

On a bestowal view of value, the beloved’s value comes from her lover. On such
a view, the idea that there are values inhering in the beloved, that the lover is to judge, is
incorrect, if not incoherent. To accept a bestowal view is thus to deny the force of
socially mediated prescriptions of objective value. In other words, if the value of one’s
beloved is bestowed, no one can be expected to judge value in some certain way, since
such value isn’t judged at all. Thus a woman who is exhorted to perceive a controlling if
tolerant man as appropriately valued can come to see that whether she bestows value on
him is rather a matter of the sort of value she herself is to bring into existence. She can
understand her valuing him to be an act of creation that need be responsive neither to

external expectations nor even to internal needs.

 

‘5 The error here is that the content of the norm prescribing feminine reserve is assumed to need alteration,
rather than the one prescribing masculine dominance.

227

The bestowal view thus portrays value as relative to the lover. The value of the
beloved is more relative to the lover in a less instrumental sense than Singer’s ‘individual
appraisal’. Recall that individual appraisal has the lover judging the value of her beloved
on criteria applicable to how well the beloved and his characteristics fulﬁll some needs or
desires of hers. Thus value is relative to the lover in the sense that it is not relative to
some prior community.” There is, though, something a bit distasteful about a conception
on which the value of the beloved is to be understood as satisfying some need of the lover
— as a lonely person might value a dog, or a hungry person food.’7

Visceral reactions to instrumental conceptions of love aside, predominant
understandings of the needs of men and women are skewed in such a way that also
contribute to making the individual appraisal understanding undesirable. The slogan of
resistance ‘A woman needs a man like a ﬁsh needs a bicycle’ has yet to be fully accepted;
women are assumed to need men for personal protection (the most visible element of
assault-prevention organizations on college campuses tends to be male walking escorts),
and women are still, I think, taken to need men for partnership (consider the negative
connotation of ‘spinster’, an unequal analogue to the much desired state of

‘bachelorhood’)."8 Men may be thought to need women, too, but for much more limited

 

4" Nothing in this distinction need be taken to deny that the desires and needs of speciﬁc persons are a
function of society through socialization. Whatever one thinks about the degree to which a person’s own
standards can diverge from society’s standards, it is at least phenomenologically apt to distinguish between
a person’s valuing something that society has deemed valuable and valuing something in spite of societal
expectations to the contrary.

‘7 These metaphors are distinct insofar as the former at least implies that there might be some return of
utility to the beloved (since the dog hopefully gains from playing this role, as well), but the latter implies
that the value is expended in satisfying the desire in virtue of which it is valued. If I appraise you
positively in virtue of your ability to satisfy my desire for political power (a quality you have but I lack),
when I have acquired what I seek, do I thus cease to value you? One can see how an appraisal view can be
construed as an overly cynical view of love.

‘8 Thus even though non-heterosexual relations are socially accepted to a degree unknown in recent history,
it is my impression that they are still taken by many to be a lesser sort of partnership.

 

228

 

purposes; for example, I take the still-popular sentiment ‘why buy the cow when you can
get the milk for free’ to apply primarily to men (I have yet to hear a popular adaptation
involving a bull). So a ‘need-satisfaction’ understanding of value — that is, Singer’s
‘individual appraisal’ model — plays a role in much of the same social damage to women
as does his ‘objective appraisal’ understanding.

On a bestowal view the value of the beloved is ‘relative to’ the lover in a different
way. The value of the beloved is not taken to be a solution to the problem of how to
satisfy desires. On a bestowal view, valuing is not a solution at all, but is, rather, an
expression. The expression is an expression of who she, the lover, is. Her medium is her
beloved. Through making him valuable, she expresses herself.

The artistic creation metaphor suggested here is not mere rhetoric, for bestowal,
like many sorts of expression, can be experimental in a way that appraisal cannot.
‘Experimental’, that is, in the sense that convention can be ignored (or at least bracketed)
for the sake of exploring new ways of living.49 Because of this, a person who sees herself
as bestowing value can feel free to ignore social expectations in the way an artist can but
an art critic cannot (on some normative reading of these roles — a good art critic). What
the critic asserts must be accurate, or at least useful; what an artist creates need have no

such epistemic or utilitarian goal, and perhaps no preconceived goal at all.”

 

‘9 More precisely, certain speciﬁc conventions can be ignored or bracketed. There is no way to step outside
the whole of what I here call ‘convention’ - socially-mediated factual and evaluative understandings. There
are, though, ways to pick and chose which aspects of conventional thought one will accept or dispute, and
this choice is made more free by conceiving of ourselves as lovers as notjudging by the criteria of socially-
mediated standards.

5° 1 am thinking here of the artist as someone like the later Jackson Pollack (originator of what is sometimes
called ‘drip art’), or at least a certain vulgar understanding of him as someone who created with neither
content nor aesthetic assertion in mind.

229

 

Valuing and the ideal of unity

Having defended a subjectivist account of value in romantic relationships has not
yet made the connection with the unity of identity any more transparent. In the following
explanation of this connection I will argue that the valuing in love is ideally supported by
a uniﬁed identity, in two distinct ways: ﬁrst, while appraising value is a process that has
relatively little connection to the unity of identity, bestowing value is in fact undermined
by identity disunity. Thus, I will show how the subjective valuing that characterizes
romantic relationships is specially related to the unity of identity. Secondly, valuing
one’s romantically beloved — as contrasted with, say, valuing a friend, or a work of art, or
a favored piece of clothing — is distinct insofar as we tend to believe (as I think we ought
to) that each of us can have only one romantic partner. This is a fact about romantic
loving that is undermined by a certain sort of fragmentation that characterizes at least
some disuniﬁed identities. That is, I will argue that the distinct object of romantic
valuation makes the unity of identity specially important. I will discuss each of these two
issues in turn.

The ability to value is undermined by disunity much more on the sort of bestowal
view of value that is defended here than on an appraisal view. To make this point, I will
make use of the distinction between telic states (or processes, or relations), and thetic
states (or processes, or relations)?l Empirical belief, for example, is a thetic state - that
is, the aim of belief, or what it is for belief to go well, is that a person’s belief about a
thing actually track the truth that is out there in the world. When this process goes

wrong, the error is with the person’s mental state (speciﬁcally, her belief). On thetic

 

5' The telic/thetic terminology was ﬁrst presented in Humberstone (1992).

230

 

processes, the mind is supposed to track the world To correct such an error, we attempt
to inﬂuence a change in the person’s mind by, among other things, presenting her with
more or better information, or giving her tools of reasoning.

Consider, now, the contrasting telic process. A desire — my desire to eat
chocolate, say — is an element in a telic relation, in which when things go wrong, the
problem is that the world has not altered to satisfy my desire; in a telic process, the world
is supposed to alter to ﬁt the mental state. To correct this sort of error, we attempt to
effect a change in the world by, in this example, ﬁnding me some chocolate.

Now, appraisal of value is a thetic process. If something has value that is inherent
in it, the aim of valuing is for the valuer’s belief to ﬁt the truth about the valuable thing.
Money is like this — a ﬁve dollar bill has inherent value (ﬁve dollars’ worth, to be exact).
If a person values a ﬁve dollar bill at one dollar, she has erred, and the error is in her
belief.” Her belief could be corrected: she could get better eyeglasses so she can read the
numbers on the bill, or she could be corrected in a (very peculiar) mathematical
misunderstanding that 1:5. There are likely other possibilities, all involving information
or improved tools for acquiring it. Notably, none of the corrections involve her personal
identity in any important sense. The ﬁve dollar bill is worth ﬁve dollars, no matter what
the content or structure of one’s identity.53

But bestowing value, on the other hand, does importantly involve the elements of

one’s identity. Explanations of why a person attributes value to X rather than Y that

 

52 There are, of course, cases in which, for example, three quarters are worth more to a person than a dollar
bill (as when she needs exact fare for a bus). These cases involve distinct kinds of value, though, and so it
would be fallacious (involving equivocation) to conclude that a dollar is worth less than three quarters.

53 Neither is there is a good analogue to a practical identity in the realm of theoretical reasons (we don’t, for
better or worse, value ourselves as persons who act in accordance with the law of noncontradiction), to
disarm a possible counterargument on this matter.

231

 

 

appeal to her identity often go unrequested, since they are obvious. Why I value Socrates
more than Babe Ruth (in the realm of famous ﬁgures), coffeehouse conversation over
poetry slams (for a pastime), or Jorge Luis Borges over Elizabeth Barrett Browning (for
recreational reading) is clear: I am a philosopher.54 What makes a person who identiﬁes
as a Christian value humility is likewise apparent; similarly for a ﬂower-child and world
peace. Thus group identities importantly ﬁgure into a person’s value bestowals."5
Elements in a person’s complex of desires play a role in bestowing value, as well
— unsurprisingly, since desires and, by extension, the things we care about are constituent
elements of personal identity. The role of desires in valuing is a more direct role than
that of group identities, though, for while group identity can explain a person’s valuing
certain things, her desires just are her valuations. If a person desires something, then she
values it. This is not to say that a person cannot desire something revolting, say — some
things (a substantial number of very desirable ﬁlms, for instance”) are desirable in large
part because of the revulsion they cause. Nor does a person desire something she takes to
be valueless, since at the very least she values what she desires instrumentally (being, as
it is, the satisﬁer of her desire)” But taking an instrumental stance toward what one
values is not necessary, nor as I argued above is it always appropriate. This stance is

most clearly inappropriate in the context of interpersonal and more narrowly romantic

 

5‘ It cannot be countered that what makes me a philosopher is just (in part) that I have these preferences,
since that assumes that I had the highly unlikely concatenation of evaluations (Socrates, conversation,
Borges, and all the rest) prior to my being a philosopher.

55 None of this is intended to imply that group identities are sufﬁcient for value bestowals. Identity
elements in toto may be sufﬁcient, but without an understanding of the range of identity-constituents (an
understanding that we have yet to gain, as I argued in Chapter 2 above), such an argument would be
difﬁcult to construct.

5" Kubrick’s A Clockwork Orange, Noe’s Irreversible, and Coppola’s Apocalypse Now would be far less
desirable were they not revolting.

” This instrumental connection between desire and value is not unique, as it holds for appraisative value as
well.

232

relationships — as when, for example, I value my friend’s recovery from illness because it
frees up my afternoons from visits to the hospital. Such an instrumental understanding,
further, is not even necessary. I may desire that my beloved succeeds at her project, in
spite of the fact that her success precludes the success of my project, simply because I
value her success.

Conceiving of personal identity as a set of practical identities — self-descriptions
under which one values herself, her actions and her life, on Korsgaard’s description -
suggests many examples supporting the tight conceptual connection between one’s
personal identity and the patterns of her bestowals of value. Earlier (in Chapter 2),
Socrates’ response to Crito’s offer of ﬂight was taken as exemplary of the high degree of
value placed on certain things in virtue of one’s identity. Recall the words of Socrates’
conscience:

...if you go to one of the nearby cities - Thebes or Megara, both

are well governed — you will arrive as an enemy to the government. [. . .]

Or will you avoid cities that are well governed and men who are civilized?

If you do this, will your life be worth living? Will you have social

intercourse with them and not be ashamed to talk to them? And what will

you say? The same as you did here, that virtue and justice are man’s most

precious possessions, along with lawful behavior and the laws?58
Socrates, in his defense of his peculiar (from the point of view of Crito, his friends, and
“the majority”) set of values, explicitly appeals to relevant elements of his identity in
saying that “not only now but at all times I am the kind of man who listens only to the
argument that on reﬂection seems best to me.”59 That is, because he is a philosopher, he

values — above his own life — social intercourse with civilized men, virtue, justice and

lawfulness.

 

5” Plato, Crito, 53c.
59 Plato, Crito, 46b; italics added.

233

Similarly, Hill’s example of the Deferential Wife is illuminating here. She,
because of her self-identifying most centrally as my husband ’3 wife came to bestow value
exclusively on her husband’s projects. Given a different self-understanding — ‘Kantian
moral agent’, as Hill recommended, or ‘feminist’ as others have implied — she would
have bestowed value differently.

Bestowing value, then, is a telic process. Given that I am a philosopher, I value
Borges over Browning, or law over life; given that I care about your happiness, I value
the success of your project. In each sort of case, if there is a discrepancy - if, say, I care
about your happiness but fail to value the success of your project — the fault lies with the
world: speciﬁcally, with the value that has not been bestowed on it.

Now, thetic processes — like those involved in belief — are subject to a simple cure
for personal disunity. A person who believes both, say, that it is raining here right now
and it is not raining here right now can remedy the conﬂicting beliefs by a basic ﬁx to her
rationality (namely, an understanding of the principle of noncontradiction) and some
empirical information (seeing whether she gets wet when standing outside).

But consider an analogical situation involving value bestowal. A person who is a
political philosopher (of a certain sort) may value her beloved’s sense of justice, fairness
and concern for the less privileged. This selfsame philosopher is also, imagine, a modern
Western capitalist, and so values the competitive spirit, the almost effortless ability to
extract money from seemingly unproﬁtable transactions, and the drive to create vast
economic empires. While not strictly contradictory, these are certainly conﬂicting value
bestowals (indeed, if the reader sees oddness in the example of a committed capitalist

with a sense of justice, the point about conﬂict has been made). The conﬂict here is

234

remedied by a change to her personal identity: by resolving the conﬂict that lies there.
Because value bestowal is not a thetic process, the other sorts of solutions — corrections to
rationality, improved information — are not relevant here (at least not centrally relevant; a
person might bestow value in conﬂicting ways simply because she is in error about some
relevant fact). She cannot value both (that is the conﬂict), and she values both in virtue
of conﬂicting elements of her identity. To reduce or eliminate the conﬂict between the
relevant identity elements would be to reduce or eliminate the conﬂicting value
bestowals. Thus, because value bestowal is a telic process, it is undermined by the
disuniﬁcation of personal identity in a way that thetic processes are not. This is then, one
way in which the valuing that is a part of romantic relationships is undermined by
disunity.

People do often have a practical solution to the value discrepancies that are the
product of a disuniﬁed identity: we compartmentalize ourselves and our lives. Some
compartmentalization is temporal; since being a parent and being a career-seeker often
conﬂict, a practical sort of resolution is attempted by delimiting certain times during
which each identity is to be in force. “It is Friday,” a person might say to her child, “a
work day. I can’t go to the park with you today; but tomorrow I have off from being
CEO.” Other compartmentalization is social; 1 might value justice in one person,
competitiveness in another. Likely a large set of distinct sorts of compartmentalization
could be characterized: cultural, similar to that Lugones has described (discussed above
in Chapter 2), and as Kipling famously suggests in noting that “East is East, and West is
West, and never the twain shall meet”; or gendered, in which value bestowals are made

differentially according to the gender or sex of the recipient.

235

 

I have argued above that such compartmentalization is undesirable. Such
compartmentalization is an especially unsatisfactory solution to the value conﬂicts that
can arise relevant to the value bestowal of romantic love. As a mother, I might value
your gentleness and concern (since I might suppose these qualities to feature in your
interactions with my children), while as an ExxonMobil executive, I might value your
estimation of money above all else, and your squashing anything and anybody that gets in
the way of proﬁt. Likely, my valuing in these discrepant ways will need to be focused on
more than one person: as a mother, I will value one person, and as an ExxonMobil
executive, I will value another person. This is how we sometimes manage friendships —
one (set of) friend(s) is valued in virtue of certain aspects of our identities, another (set)
by another aspect. I might have my rock-climbing friends and also my philosophy
friends.

Such a conﬂict-alleviating tactic is not available to the romantic lover, since the
typical structure of romantic relationships is quite different from that of friendships.
Herein lies the second abovementioned way in which romantic valuation is specially
related to identity, insofar as the object of valuation — a romantic partner — is very unlike
other objects of valuation. Importantly, romantic relationships are typically both
monogamous and encompassing. In being monogamous, the option to value more than
one person as a romantic lover is checked. Conﬂicting valuations cannot be remedied by
valuing different things in different beloveds.

The encompassing nature of romantic relationships does not directly preclude
compartmentalized valuing. This aspect of love is a scalar consideration, for while a

person’s relationship to her beloved either is monogamous or it isn’t, the range of aspects

236

with regard to which a person bestows value on her beloved is variable. He might value
her for her wit and creativity but not for her sloth."0 There is a limit, though, to how
many aspects of one’s beloved go unvalued —- under this (subjectively determined) limit,
the individual respect that is part of love diminishes, or disappears. Thus, the more (and
more important) aspects of a lover’s identity in virtue of which she cannot value her
beloved, and the degree to which she cannot value them, the closer she approaches to not
loving at all.

A lover, then, is to value her singular beloved across diverse circumstances
(circumstances sometimes akin to Lugones’ “worlds”). Were our ideal romantic
relationship less than exclusive (most romantic relationships aim at monogamy in the
bed, the heart, the home, the nursery, and the checkbook), identity disunity might not be
problematic; I value you in the home setting and someone else in the bed — one’s sexual
and domestic identities could be unproblematically disuniﬁed. But given that our partners
are central in many of these as well as other realms, such discrepant valuing is not
practically desirable. It is because of this exclusivity characteristic of romantic love —
which I have no reason to argue against — that makes identity disuniﬁcation especially
relevant to romantic objects of valuing.

An objector might attempt to turn the overall line of argument I’ve presented here
on its head: since uniﬁed valuing of the sort I’ve defended is enormously practically
demanding (and at the limit, perhaps impossible), this implies that we ought to question
the desirability of monogamy. Certainly other arguments both in and outside of the

philosophical literature have been raised against it. For example, in an article arguing

 

6" Though as I’ve mentioned, he may value her for her sloth.

237

 

that monogamous sexual relationships are damaging to individual women’s identities,
Christine Overall relates this reasoning:

Women, I suggest, are generally expected to incorporate the sexual partner

into their own identity [. . .]. A woman who freely takes on a second sexual

relationship in addition to the one she already has is likely to feel [an]

expansion of her identity, or even a claiming or reclaiming of self."1
Further, women are argued to be differentially affected by the possessive aspect
of monogamy — the term literally means ‘one wife’, and polygamy ‘many wives’;
the thought that multiple sexual partners are more permissible for men than for
women is certainly common.

But monogamy need not be thought of as a sexual arrangement, just as —
unsurprisingly - marriage need not be thought of as sexual possession. Overall
usefully notes the example of Simon de Beauvoir and Jean-Paul Sartre, committed
life-partners who were somewhat famously non-monogamous in the sexual sense.
Their arrangement “suggests that it is possible and useful to distinguish between
sexual exclusivity and possession, on the one hand, and on the other, the creation
of a long-term, even life-long project with another person?"2 Beauvoir and Sartre
were romantically, though not sexually, monogamous.

Not only, then, is romantic monogamy clearly the subject of the bulk of criticisms
aimed at monogamy, romantic nonmonogamy is not a live possibility, practically

speaking. Monogamy (in the sense of a long-term project, less so in the relatively

unimportant sexual sense) is insinuated to an extraordinary degree into religious”,

 

6’ Overall (1998), p. 9.

62 Overall (1998), p. 6.

‘3 Virtually all contemporary Judeo—Christian sects condemn polygamy (in the sense of multiple
contemporaneous marriage), though the Bible clearly does not.

238

I”; psychological” and historical”

moral”, social”, political and legal”, biologica
understandings of who we are both as speciﬁc persons as well as humans generally. I do
not think that one or more of these understandings could not be (or will not be, or has not
been, or should not be) overturned. I do think, though, that sufﬁcient of these
understandings to ensure the continuance of monogamy as the only viable option for
workable romantic relationships will continue into the foreseeable future; the magnitude
of social revolution that would be necessary to effect the viability of nonmonogamy is
enormous. Since, then, the appeal to monogamy in the above argument turns on the
practical implications of it on identity and valuing, the claims regarding the fact and
practice of romantic monogamy here are axiomatic. In a different world (better in some

ways, perhaps, but very different), my arguments, as all arguments, will have less traction

than they do here.

 

"4 For example, Bernard Williams’ famous and powerful argument against impartial moral demands appeals
to the practical reason given a person by the thought that love “is something which has enough power and
even authority to conﬂict badly with morality” (Williams, 1981, p. 17). Williams’ argument takes as its
most forceful example monogamous love and the sometimes morally conclusive fact, simply, “that it is his
wife” (p. 18).

‘55 Consider even the multitude of ways in which the physical aspects of the social world are arranged, e. g
the fact that automobiles have two front seats (me’nages a trois would be counterproductively
uncomfortable even on the drive to dinner).

66 Perhaps the easiest to unravel of all the threads of monogamy, somewhat oddly.

67 Imagine, for example, the vast gulf (more vast than any so far) that would emerge between humans and
other animals were we to come to think of ourselves as not importantly pair-forming creatures.

6" For example, while psychological understandings of parent-child relationships are fairly easy modiﬁed to
accommodate gay or lesbian couples as parents, it is another matter entirely to adapt them to, say, a ‘three
fathers and two mothers’ models. Even the literature on relationships between children and their remarried
parents is not clearly applicable, nor do 1 see abundant reason to think these sorts of family relationships
especially normatively good models.

69 Two thousand years of narrative on love would be outmoded, for instance; imagine the themes of Romeo
and Juliet as laughable as the inhabitants of Huxley’s Brave New World found them.

239

Commitment
It leaves you baby, if you don’t care for it...

- U2, “One”

Loving relationships fail for many reasons, and sometimes it is indeed better that
the relationship fail. Legal changes that make it more possible to end problematic
marriages have undeniably led to many improvements in people’s material lives and in
their capacities for self-respect and self-direction.

The other side of the story, though, is less uplifting; examples of the damage to
persons -— both women and men, and both partners and their families —- by the ending of
romantic partnerships can be cited at length. For example, shorter education,
unemployment, and more risky health behavior are more common among people,
especially among women, with a background of parental divorce”; widowhood is
strongly correlated with increased risk of suicide, at least among heterosexual men";
suicide rates are twice as high among divorced or separated than among married
persons”. I will offer a conception of commitment that is both consistent with the
continuation of love that the partners to the relationship feel no pressing reason to end,
and is also instructive as to how to perpetuate a loving relationship that one would on
balance desire that it continue.

What is it to commit oneself to a partner in a romantic relationship? Such a
commitment, which I take to be a variety of the sorts of commitment persons have to

aims, projects, and principles, is sometimes taken to share a conceptual core with a

 

7" Huurre at al (2006).
7’ Luoma and Pearson (2002).
7" Kposowa (2000).

240

promise. On such an interpretation a commitment is a type of self-initiated act of putting
oneself under a temporally persisting obligation. A commitment of this sort might be
seen as valuable in romantic partnerships, those ongoing associations that are inevitably
marked to some degree by doubt, changes of heart and even changes of identity. That is,
sometimes the person we love is a different person, although the same incarnate being,
than the one we initially fell in love with. Our commitment to that being, or perhaps our
commitment to the relationship itself, can help to perpetuate the partnership though such
changes, just as a promise serves to perpetuate an obligation through similar changes.73

Understanding commitment between changeable persons as essentially a promise
is not without its problems, though. As Derek Parﬁt writes,

I might say, ‘I, and all of my later selves, shall help you.’ But it could be

objected that I can bind or commit only my present self. This objection has

some force, since it resembles the plausible claim that I can bind or

commit only myself. In contrast, no one denies that I can promise you that

I shall help other people, such as your children. It is therefore clear that I

can promise you that I shall help your later selves.74
Thus it is difﬁcult to understand how commitment, as promising or self-binding, could
fulﬁll its role of helping to perpetuate a relationship through changes within each of the
partners. It may even be more difﬁcult to understand than Parﬁt lets on; in a speciﬁcally
romantic commitment, the idea that one might promise to love some future person,
presently unknown, is implausible.75

This problem may be even more common, too, than Parﬁt’s view of it might seem

to imply. For even if we take such radical changes — Saul-to—Paul conversions —- to be

 

73 On the role of the valuing of the relationship itself, see Kolodny (2003).
7“ Parﬁt (1984/1987). p. 327.
75 This suggests that many instances of long-term promises may be in some sense incoherent. I think this is

true, but this because many instances of long-term promises involve commitment of the sort explained
below.

241

rare, substantive changes occur in within the same person over time. Few promises are
categorical — my promise to go to a ball game with you next summer holds unless you
cease to enjoy ball games, or the team dissolves, or the like - and so even relatively
minor changes in persons can affect the content and perhaps even bindingness of
commitments conceived of as promises.

It may be that we (as lovers) don’t need commitment to play this role: sometimes
one or both partners change so much that it would be best if the partnership dissolved.
Furthermore, if the changes of self are not so radical as to recommend the dissolution of
the old partnership, there is no obvious reason why one cannot love this new person — the
partner who has changed — anew. Although lovers may not, strictly speaking, need
commitment“, I take it that within our common practices it is often commitment that
perpetuates many romantic partnerships through changes in the partners."

Whether holding commitment to be necessary for romantic partnerships or not,
philosophical accounts of love often take commitment to be a promise. In arguing that
“commitment is precisely what love is not” Robert Solomon claims that “a commitment
is an obligation sustained whether or not one has the emotion that originally motivated
it.” 78 This is indeed central to the practice of promising: the creation of an obligation and
motivation that persists in spite of the failure of the original motivation. Thus I promise
to help you move next Saturday so that you can rely on my help even if I wake up that

morning with no desire whatsoever to help you move. The promise also helps me by

 

7" Commitment as promising, that is. As we shall see, the sort of commitment that is a part of love shares
structural and functional aspects with the sort of encompassing valuations that I have argued to mark
romantic love.

77 The implied claim should not be read too strongly. I do not deny that many relationships persist not
because of commitment but rather because of lack of exit freedom due to scarcity of alternative partners,
ﬁnancial pressures, social expectation, and the like. I do not wish to imply that commitment is more
prevalent than it actually is.

78 Solomon, p. 27 .

242

giving me a motive that I might otherwise be able to produce in myself only through the

psychically difﬁcult process of voluntaristic desire formation. The obligation created by
my promise will motivate me whatever my conative stance toward lifting boxes of books
is come moving day.

Solomon notes that this sort of obligation is a poor basis for motivating the
persistence of a romantic partnership.79 Motioning toward the Kantian notion of ‘moral
worth’ as being dependent on a good will, Solomon suggests “we might say that the
emotional worth of an action is determined precisely by the degree to which we want to
do it, regardless of our obligations. And that means that an act of love —- and ultimately
love itself — is the very opposite of a commitment of any kind.”80

For the present purposes, I will pass on questioning whether every act associated
with love must have emotional worth. We might also ignore the implication that love
itself is reducible to acts of love.81 What does stand in need of investigation, though, is
whether the thought that love ought to go on only because it is desired is inconsistent
with a commitment of any kind. Is there a kind of commitment that is distinct enough
from promising that might play an important role in romantic relationships?

Richard White, in an investigation of friendship, notes that a person can commit
herself implicitly.

[W]hen I promise that I will repay the money that you lent me, I create an

expectation that is equivalent to a commitment, given the institution of

promise-making and all that it entails. Likewise, when I spend time with

someone, accept their help, and make myself available to that person, by
sharing the more intimate aspects of myself, I am also creating an

 

7’ Perhaps a stronger indictment of the thought that love might persist as a function of an (externally
imposed) obligation is that since loving is not voluntaristic, obligation is not pertinent to it.

so Solomon, p. 224.

8‘ And, of course, the categorical cleft between desires and obligations is highly contentious.

243

expectation that is equivalent to a commitment, given the institution of
friendship and all that it commonly entails.82

Now, a promise to repay money is explicit. If I borrow money from you without
explicitly promising to repay, but under circumstances in which it is reasonably
understood that the transaction was a loan and not a gift, I have allowed an expectation to
come about in you to the effect that I will repay the money. This is an expectation that I
am obligated to fulﬁll. But it is not a promise — if I fail to repay the money, you may
legitimately demand repayment, but only in virtue of my having borrowed it, not in virtue
of some promise that is over and above the mere fact of borrowing.

The foregoing distinction between promising and commitment — that promising is
explicit while commitment need not be — is insufﬁcient to relieve Solomon’s worries
about obligatory loving. For even an implicit commitment can create such an obligation.

A perhaps more vivid distinction between a promise and a commitment is that
while a promise is explicit and thus self—evident, my being committed is something that I
can come to ﬁnd out about myself. Retrospective explanations of my behavior may
accurately appeal to commitments of which I was unaware. When one day I wonder why
I keep ﬁnding myself defending a friend against all those who criticize her, I may come
to realize that I have been committed to her. A person can discover herself committed to
a cause or principle in a similar manner.83

That commitments can be unintentional is most vividly seen when one intends not
to be committed, but ﬁnds himself committed nonetheless. A person might join a

strongly feminist philosophy department, intending to remain open-minded but skeptical

 

82 White (1999), p. 82.

83 A person can forget that she promised, of course. The distinction from commitment is that she does not
(in non-pathological cases, anyway) later discover that she promised, although she may remember - or fail
to remember — that she did.

244

about such an intellectual and political position. After time with his new colleagues,
during which he is exposed to many ideas about which he tries to be considerate but
resistant, he may come to realize that he has become committed to feminist principles.

Another distinction between commitment and promising is that I can un-commit
myself, but not un-promise myself. To act contrary to a promise one has made is just to
break the promise, to fail to fulﬁll the obligation the promise entails. The promise has not
been unmade, but rather unfulﬁlled. But a person can undo a commitment without
breaking it, by acting contrary to it. This last claim is perhaps less a fact about
commitment than it is a point of clariﬁcation regarding the details of the notion of
commitment being defended here, as I will explain.

There do seem to be cases of commitment in which the person who has
committed has an obligation to act in accordance with the content of the commitment.
Suppose you are halfway up a shaky ladder and 1, walking past, stop unbidden and hold
the ladder. Once you continue climbing the now-secure ladder, it might seem clear that I
have given you good reason to believe that I have committed (but not promised, of
course, since whatever binds me to continue to hold the ladder is an implicit
arrangement) to continue to assist you by holding the ladder. To offer a second sort of
example, imagine a person who has worked tirelessly for years for some cause —
Greenpeace, say — being actively engaged in the ongoing projects of this cause, and even
starting many new constituent projects which she and the other members of the
organization pursue. A last sort of case: implied romantic interest can be taken to
indicate commitment — in a different day and age, a commitment to marry, though now

more likely a commitment to focused romantic attention.

245

There are differing responses that show each of the cases not to be cases of
commitment in which the person so committed is culpably bound to fulﬁll the supposed
obligation. In the case of the helpful ladder-holder, it is hard to deny that were he to
cease to hold the ladder while the need for such assistance remains, he would not be
culpable for harm or even for the psychological security he led the ladder-climber to
expect. But to suppose that this is a case of commitment implies that analogous cases are
too. The expectation that I will not wantonly injure you with my automobile; the
obligation a grocer has not to sell food she knows to be spoiled; the resentment you
would feel on my holding the door for you only to let it close when you are halfway over
the threshold - all these are real and legitimate (the expectation, the obligation, the
resentment), but none of these are cases of commitment. To think they are is
unrecognizably to alter the term.

The second sort of commitment, commitment to a cause, is indeed an example of
the sort of commitment outlined here. Further, a member of a cause such as Greenpeace
could be expected to answer for her discontinuing her involvement in the project of, for
instance, protecting the Minke whales. But this is because the project is a morally
laudable one, not simply because she has committed to it. To criticize her abandonment
of the whales then reduces to an appeal the sort of moral omission described in the ﬁrst
sort of case, above.

The third sort of case — of one who is perceived to have committed to marry or at
least romantically to attend to some person — is often enough not a case in which there
exists culpability for failure to carry through the so-called commitment. A person who

allows an expectation of marriage or continued romantic affection in another has ‘led her

246

on’, to use a common phrase, and is certainly liable to cause resentment for that; he is
culpable for whatever harm follows from allowing an illegitimate expectation; he is
guilty of deception (intended or not) insofar as he has encouraged or permitted her
incorrect interpretation of his behavior and intentions. But all these are cases of
something other than failure to carry through on one’s commitment, by a sometimes
blameworthy action or set of actions contrary to the commitment. I will return to address
this issue shortly, after a bit more necessary discussion of the notion of commitment.

To act contrary to a commitment is either to ﬁnd out that one has become
uncommitted or, more actively, to engage in the very process of uncommitting oneself.
Suppose I were committed to you and found myself acting contrary to that commitment.
If my so acting were somewhat unthinking, that is, if it indeed was a discovery I made
about myself that I acted in this way, my acting contrary to my commitment is a sign that
I am not committed, either not at all or not in the way I took myself to be.

To consciously act contrary to a commitment can be a way of ending the
commitment. If I come to desire not to be committed to you any longer, I might end my
commitment by acting so to end the conditions that supported the commitment.84 If my
commitment to you involved my accompanying you to Botswana for ﬁve years while you
do your PhD research, I can go about ending my commitment by declining to accompany
you to Botswana. To come to desire not to be committed, though, suggests that the
commitment has already begun to end, since commitment entails a desire that the

commitment continue.85 Thus while a person might actively go about ending the

 

8“ If I simply come not to desire to be committed (that is, I lose the desire without gaining the opposed
desire not to be committed), the commitment may persist of its own inertia for a long time.

85 I will return to further discuss this important feature of commitment. Interestingly, this self-perpetuating
nature is common to a number of phenomena of both psyche and character as diverse as persistence and

247

commitment by wrapping up the loose ends, so to speak, the important part of the
phenomenon is the discovery — the discovery that one is no longer committed. A person
can discover that she herself is committed, and she can also discover that she is not
committed.

So a commitment need not be conceived of as a sort of promise. Another, more
narrow, conception of commitment that understandably makes its way into philosophical
meditations on love is that of commitment as exclusivity. Robert Nozick writes of
commitment as a special sort of promise, a promise of exclusivity, especially sexual
exclusivity. Probing the limits of an economic metaphor for love, he suggests that the
energy expended in building a romantic partnership calls for a “guarantee that the other
party will continue to trade with you”.86 Indeed, his account of love is aimed at answering
the question “why would we want actually to give such a commitment to a particular
person, shunning all other partners?“

While commitment may and often does entail exclusivity, this is largely
accidental; the entailment is not a conceptual one.88 Consider commitment to ideals. A
person can be committed to more than one ideal, as when someone is committed to both
vegetarianism and environmental preservation. It is important to notice, though, that she
cannot be committed to both environmental preservation and leisure travel of the world
by private jet. So while some co-commitments are impossible, being committed to more

than one ideal is not automatically ruled out.

 

apathy. Both essentially include intended or expected continuance, since e. g. a person who is feeling
globally unmotivated now but self-consciously and currently possesses a motive to act in some way in the
future is not apathetic, but rather merely resting.

8" Nozick (1989), p. 78.

87 Nozick ( 1989), p. 77.

88 As mentioned in the previous section, the entailment is practical though very deeply rooted.

248

Regarding romantic commitment to another person the matter is different, given
the pervasiveness of monogamous romantic partnerships. Romantic commitment is to
one person (and even in what is probably the most common variation on monogamy,
what occurs is not simultaneous commitments but rather a violation of the rules of
commitment altogether: cheating). Sometimes (often?) one’s commitment to a person is
even brought to an end by means of committing to another person. But this is not a basic
fact about commitment per se. It is a fact about how romantic relationships tend to ﬁgure
into our lives.

A look at basic locutions might indicate that we have been on the wrong track all
along. A person ‘commits herself’ and even ‘commits her life to’ some ideal or some
person. These phrases suggest that what is being committed is inseparable from she who
is doing the committing. ‘Co-mitto’ is, literally, ‘together, I am handing over’. That
which is given is not distinct from the giver — in other words, I give myself. A person
committed to an ideal has given herself, not merely her actions, to the ideal. She
formulates her life around the ideal, and comes to think of herself as being someone
committed to the ideal. Likewise, a person who commits herself to another person in a
romantic partnership identiﬁes with the object of her commitment. This is why I do not
merely act in husbandly ways - rather, I am a husband. One who commits gives herself
by means of a change in her identity.

This conception of commitment as giving oneself has to meet the criteria
discussed above. It has to explain its own role in perpetuating romantic relationships in
spite of the forces of personal change. It must be immune from Solomon’s worry about

obligation’s trumping of desire. It must be consistent with the observation that while a

249

person can actively undertake a commitment, it is sometimes a matter of self-discovery
that one is committed. It might also illuminate the way in which we do tend to commit
only to one person, while we commit to many ideals and projects. The conception of
commitment as giving oneself can meet these criteria.

There seems initially to be a bit of tension between the notion of giving oneself
and the aim of explaining some fact about essentially changing persons. If the selfI am
giving in my commitment is different enough from that self that is supposed to be
engaged in a persisting relationship, it doesn’t seem that this notion of commitment is
going to get us very far. In (nearly) everyday language, the concern is this: your
commitment to continue to love me isn’t very helpful if you will soon be someone other
than the person who so committed. Put more technically, the commitment C by person X
to person Y at time T is not the same as C’ by X' at T2, and so can’t explain the role it is
supposed to play in perpetuating relationships. Interestingly, the problem is distinct from
the one that undermines commitments as promises in the earlier passage from Derek
Parﬁt. To promise on behalf of my future self, recall, is illegitimate because promises I
make bind only myself. Thus, if I change enough (enough to have a different self, that
is), no prior promise will be binding on me, since the self that I obligated no longer
exists.89 But commitment as giving oneself has a worse possible problem: the self that I
gave no longer exists.

If this appears to be a problem for the notion of commitment as giving oneself, it

is only because of a snapshot view of the self that is being assumed. Such a view

 

8” There is a different conclusion that could be drawn here: because promises are binding on me, change
though I might, I will not become a different person. That is, persisting obligations serve to bind together
the parts of my temporal life, effecting a diachronic uniﬁcation. Whatever the situation regarding
promises, I claim that such temporal cohesion is one of the results of commitment.

250

assumes that not only does a person exist as given in a snapshot - that is, at a moment —
but also that there might be two portions of a person’s extended life, metaphorically
represented by two snapshots, which are in important ways independent of one another.
The snapshot of me at birth bears no resemblance whatsoever with that of me receiving
my college degree. This is a misleading view of persons. The life of a person is
connected, moment to moment, day to day and year to year. These connections are
effected by, inter alia, the commitments she makes and discovers.90 Thus, what is being
given in commitment is a temporally persisting self, at least that portion of it that remains
to unfold (it is not clear how a person might commit herself as she used to be, except by
becoming that person again).

So not only is commitment a giving of a temporal self, but commitment itself
supports diachronic continuity. A person who is committed shapes her desires and plans
around the object of her commitment. These desires and plans are themselves temporally
extended: plans are just sets of intentions that look into the somewhat distant future
(while a person can intend to turn the page of this essay, that is not the proper object of a
plan, since it is too near and too simple a goal). Since plans extend forward in time, and
commitment is an arrangement of these plans, a person’s commitments extend her
forward in time.91

Commitment also has a retrospective element. It is when a person comes to
realize that she has unconsciously been forming her own goals around those of another

person that she discovers her commitment. We can know our own goals and desires

 

9° This continuity is also effected by memories, the less-momentous elements of commitments such as
goals, desires and projects, and those elements of identity such as group membership and character traits
that are attributed both lSt and 3'‘1 personally. To think of diachronic continuity as dependent only on some
subset of these items would be a misleading view as well.

9‘ This observation was forcefully made in Williams (198]), Chapter 1.

251

without knowing their structure, the way in which they as a group correspond to some
external point of reference. Such ignorance of the collective focus of our desires might in
fact be the norm. It is when that focal point comes to light that the fact of one’s own
commitment — the fact that one has given herself, in the form of her desires and plans, to
a certain object or aim — is discovered.

That commitment is a condition of one’s desires, rather than fundamentally
opposed to desire, as Solomon assumed, explains two features of commitment: its self-
perpetuating nature, and its inherent motivational force that is independent of external
obligation.

A committed person needs neither stick nor carrot to motivate her to perform
actions commensurate with the commitment. This is, simply, because she is committed.
Many of her desires collectively aim at the object of her commitment, and so she has the
natural motivation of desire satisfaction to motivate her. Not every one of her desires is
captured in this arrangement, but insofar as commitment entails the structuring of a
substantial subset of one’s desires toward some object, the motivational weight of this
subset will tend to win out over isolated recalcitrant urges to the contrary.

This view of commitment might seem to allow for a commitment to be unwilling.
Someone willingly committed to a certain person can come to knowledge of some fact
(often a fact about the object of her commitment) that makes her present commitment
unwilling. She might, alternately, discover herself to be committed to someone who she
would never have consciously consented to commit to.92 If this is possible, then the

motivational picture is not as clear. Someone unwillingly committed would have an

 

92 I suspect this phenomenon is fairly common, since while love may or may not be blind, falling in love
almost certainly is.

252

arrangement of desires with a certain object, but she herself would not approve of this
arrangement. Her commitment becomes a stick, moving her against her will.

This does not correspond well to common use of the term ‘commitment’, and so
recommends an addendum to the account. We simply tend to call something a
commitment only if it is willing. This is the case especially in romantic commitments;
familiar is the thought of Agathon in Plato’s Symposium: “And the effects [love] has on

’93 Even

others are not forced, for every service we give to love we give willingly.’
commitments that come to self-awareness only after their formation would be, I suspect,
retroactively reinterpreted as something other than a commitment if its possessor failed to
endorse it. Hill’s deferential wife was, ex hypothesi, committed to her husband, though
were she to be the subject of either feminist consciousness-raising or a morally positive
conversion experience she may (for a time, at least) behave in the same way as before,
but she would no longer properly be called ‘committed’. Finding oneself with unwanted
(and unshakeable) large-scale arrangements of desires may be one element of ‘identity
crises’, but not of commitment.

The claim that a person might herself be up against her commitment is plausible
only in an inverse proportion to the scope of the commitment. This is because
commitments are signiﬁcant constituents of persons, and so the degree to which a certain
commitment constitutes a person is the degree to which she cannot be positioned against
it.

It is apparent that certain commitments are more self—constitutive than others.94

That is, we give our selves more fully to some objects of commitment. ‘More fully’ is

 

9" Cited in Solomon and Higgins (1991), p. 21.
9“ Although it is often less obvious which of a person’s commitments are more self-constitutive.

253

apt here, since the commitments that are more important to a person tend to be the ones
which are formed from a weightier subset of her identity elements, whether these
elements are desires, self-conceptions, group identiﬁcations, or narratives. Romantic
commitment can be (and is ideally, I would argue) structurally similar to what S.I Benn
calls a "core commitment", which is a principle "such that not to act on it would be to
disregard, downgrade, or repudiate some central part of the web of one's moral beliefs,
which supplies a unity and coherence, perhaps, to all the rest.”95

It should be clear by this point that romantic commitment is not usefully
characterized as having the form or function of a promise. If the alternative account of
romantic commitment I have presented here is plausible, then it should further be clear
that the unity of identity plays an important role in such commitment. A person often
identiﬁes with the object of his commitment; I do not act like a husband, but rather
become and identify as a husband, and so commitment is not so much an act a person
performs as it is a characterization of an important, central part of his identity.

I noted that when discovering oneself committed to some person, what is
discovered is a particular clustering of desires, narratives, self-conceptions, goals, or
other identity elements. Thus, to be committed is in large part just to be uniﬁed around,
in the case of romantic commitment, another person and that person’s desires and goals.

The motivational considerations that serve to distinguish committing from
promising also serve to illustrate the internal uniﬁcation that commitment entails. A

person who is committed does not require the sort of external pressure to act that a person

who has (merely) promised might, since the committed person’s desires aim at the object

 

95 Benn (1984), p. 43; emphasis added.

254

 

of her commitment. Even when an isolated desire is absent, the collective desires that
constitute a commitment can motivate. For example, though I may not desire to go to the
opera with you per se, I may still be motivated to go if I am committed to you — that is, I
desire that you experience the pleasure you get (and I do not get) from attending the
opera with me, I desire more generally that your projects not be thwarted, and so on.
Important to notice here is that the motivating force of commitment is the motivation of
an interconnected cluster of desires; the less cohesive this cluster, the more conﬂict
present in the set of desires, the less apt is the attribution of commitment. Thus
commitment, understood properly -— that is, not as a promise but rather as a clustering of
identity elements - is itself a form of unity of identity.

The issue of conﬂict in commitment is not limited to conﬂict within a single
commitment. The conception of commitment defended here makes good sense of the
conﬂict between commitments, as in cases of romantic partners with different religious
commitments, for example. Because each partner is committed to her religion and to her
partner (who is in turn committed to a different and possibly incompatible religion),
practical and ideological conﬂict sometimes arises.

An even more direct connection to the precise notion of unity under discussion in
the present work is illuminated by the fact that a committed person endorses the identity
elements that make up her commitment. I had suggested above that a person who
undertakes to end her commitment isn’t doing just that, since the motivation to end a
commitment to something occurs only in the relative absence of a commitment to that
thing. This is not the case regarding promises, of course — one can consider breaking her

promise while still fully in the grips of the obligation the promise entails. But

255

 

considering a few examples will illuminate both the nature of commitment’s terminus
(and the connection with endorsement that this point is made in support of) as well as
lending just a bit more support to the unity—based conception of commitment I am
defending here.

Imagine Karl, who is committed to socialism. He votes socialist, pumps Citgo
gasoline into his car”, and shops at co-operatively owned markets and grocers. Karl self-
identiﬁes as a member of the Socialist party, and gains a great deal of personal self-worth
from this identiﬁcation; he sometimes thinks that were he a capitalist he could not look
himself in the face. Karl desires that society be organized under socialist principles.
When these desires falter or are opposed, as might happen after being psychologically
manipulated by commercial advertisements, or when affected by people who he respects
but who do not share his political views, he distances himself from these desires (‘I feel
as if I want that new shiny Proctor and Gamble product, but I really don’t want such a
thing’, he tells himself). That is, Karl endorses his desires that society as well as he
himself be socialist.

Because Karl is committed, he has a great many internal resources with which to
perpetuate his commitment. He has many desires that correspond with socialism that
trump isolated contrary desires; he has a great many self-conceptions (socialist, co-op
member, nationalist oil supporter, and the like) in which to ﬁnd self-value, and so can
afford to forego other possible (and popular) self-conceptions (successful investor, self-

made millionaire).

 

9" Citgo is a subsidiary of Petroleos de Venezuela, the Venezuelan national petroleum company.

256

It is because of the presence of this body of interconnected identity elements that
it would be incomprehensible were Karl to end his commitment to socialism while all
these elements remain in place.97 Certainly Karl might be motivated to end or alter his
commitment by acquiring a powerful desire to control a vast ﬁnancial empire, for
instance. But if we suppose that Karl has this desire (and thus accepts it as his own,
rather than distancing himself from it) then we are supposing that Karl’s commitment is
already in the process of ending — disintegrating from the rift struck by this incompatible
desire.

As an illustration of romantic commitment’s similar structure, imagine Jenny,
who is committed to Karl. She desires that Karl is happy, and that his projects, by and
large, are successful. (Certainly she need not be committed to all the projects, ideals, and
desires to which Karl is committed, and likely she is not; if she is committed to none of
them, though, it is obscure in what way she is committed to Karl himself.) She conceives
of herself as a person who backs Karl and his projects in this way. She gains self-value
both from supporting Karl and from the success of his projects, and is thus committed to
these projects as well.

There are, of course, many ways in which Jenny’s commitment to Karl could be
undermined. These ways are diverse and virtually innumerable: Jenny might come to
identify with projects aligned with capitalist economic relations, and thus be unable to
identify with socialist aims; were Karl’s sexual inﬁdelity to move Jenny to end her
commitment to Karl she might continue to endorse socialism but cease to desire that

socialism qua Karl’s project be realized; Karl may become afﬂicted with some sort of

 

97 lncomprehensible, at any rate, in the normal course of events. Saul-to-Paul sorts of conversions involve a
wholesale shifting of desires of both the ﬁrst and second order, and so are explicable in terms of a change
into a different person.

257

 

dementia or psychosis that destroys his ability even to have such projects to which Jenny
could be committed.98 But for Jenny to continue to endorse socialism and Karl’s projects
as originally described and at the same time to seek to end her commitment to Karl is
inexplicable.

The story of Jenny and Karl serves well as a point of return to the unanswered
question regarding the culpability of one who has ended her commitment. Surely in
some of the hypothetical paths of Jenny and Karl, certain ways in which Jenny’s
commitment comes to an end are occasions for resentment from Karl. Were Jenny to end
her commitment in virtue of her identifying rather with capitalist goals, or with the
projects of some other person, or come to realize that she cannot identify with any male
romantic partner’s projects when she discovers that she is a lesbian, Karl’s resentment
seems warranted (even when Jenny abandons an Alzheimer’s stricken Karl who cannot
experience resentment, it would seem warranted).

Depending on the back-story, Jenny is certainly the proper subject of
disapprobation if, for instance, she deserts the socialist project when she realizes that as a
capitalist she could have a servant, a BMW and an increased power to direct other
people’s lives at will. Such disapprobation need not be essentially moral; she might be

legitimately charged with a failure of personal integrity for giving up her personal

 

principles under pressure. Other cases will be more plainly moral: completely to abandon
one’s romantic partner to dementia exhibits a profound lack of respect, a merely

instrumental use of other persons that we tend to think morally heinous.

 

9" The account of commitment defended here denies that a person can, strictly speaking, be committed to
someone who is dead, comatose, or severely demented, though he may be committed to that someone’s
memory, her imagined or remembered projects (as Benigno was committed to Alicia in Almodovar’s
(2002) Talk to Her), or to her care.

258

Commitment is thus one mode by which we organize our identities, and thus our
lives. It is the content of our commitments -- the goals to which and the persons to whom
we commit - that has moral standing. But although there is no culpability to ending a
commitment simpliciter (again, commitments not being promises), there is at least one
thought motivating their continuance. This thought is the barrenness of a life without
commitment -— to projects, to causes personal and political, to principles and especially to

other persons.

259

 

—6—

Concluding thoughts

In oneself lies the whole world...

-- J iddu Krishnamurti

The wide range of concerns in the foregoing chapters have taken us time and 1'
again outside the typical ambit of philosophical discussions of personal identity; if there
is truth in Krishnamurti’s observation, then the breadth of any substantial discussion of
the person is expected, even necessary. A very brief reﬂection on where the arguments
of this work were aimed will, I hope, make my central conclusions individually and
jointly perspicuous.

The one guiding thought throughout has been this: there are reasons thus far

unmentioned in the philosophical literature on personal identity both to believe beneﬁcial

 

and to strive for a uniﬁed personal identity — and so the uniﬁcation of one’s identity is an
ideal toward which, for the reasons I offer, one might well strive. The sense of ‘uniﬁed
personal identity’ I have defended is the relative absence of subjectively perceived
internal conﬂict among the elements of one ’s personal identity, elements which include
but are not limited to desires, narratives, principles, group identiﬁcations and self-
conceptions.

More speciﬁcally, I have argued that the most prevalent criticisms of uniﬁed

identity views can be answered. Speciﬁcally, worries about the coherence of hierarchical

260

accounts of personal identity, the inadequacy of such identity to ground a person’s real
interests, the alleged rigidity of a uniﬁed identity, and the damaging political
consequences of promoting such a view were all molliﬁed, and some I suggested are
quite possibly groundless.

I defended the ideal of identity unity for the role it plays in supporting personal
autonomy primarily by showing how a structural account of autonomy — on which a
person acts autonomously when she acts on desires she endorses — is a more plausible
account than are either procedural or content-speciﬁc views.

I explained how a better philosophical understanding of the notion of integrity (an
understanding that I take to have value not merely within the context of the current
project) enables us to see that when a person acts with integrity, she acts on principles —
‘principles’ being a sort of shorthand term for those clusters of things with which a
person identiﬁes that are uniﬁed into coherent, conﬂict-free clusters. Thus when a person
acts with integrity she acts as a person whose identity is, at least locally, uniﬁed.

Finally, I argued that the romantic lover has two pressing reasons to avoid a
disuniﬁed identity. Since the mode of valuing of one’s beloved that is most theoretically
plausible and socially beneﬁcial is the telic process of value bestowal, identity disunity
undermines such valuing in a way that does not affect the practice of appraising value.
Lastly, I argued that commitment between lovers — that is, that state of organizing large
measures of one’s life around her beloved — is itself tantamount to the unity of identity.
Thus to value one’s beloved, a person best have a uniﬁed identity; insofar as she is

committed, she does have such an identity.

261

I have defended the views here against what I take to be the strongest criticisms
the philosophical literature offers. I do not suspect that my arguments here have been
universally convincing, nor do I suppose that I have not missed substantive criticisms
both extant and yet to come. It may come to be demonstrated that there is a feature of
good lives, something more important than acting with autonomy, integrity, and love, that
is possible only in the presence of substantial conﬂict between the various elements with
which a person identiﬁes. But insofar as that has yet to be demonstrated, and as
autonomy, integrity and love do feature prominently into satisfying, good lives, we ought

best to hold up the unity of personal identity as an idea].

262

Bibliography

American Psychological Association (2002). Ethical Principles of Psychologists and
Code of Conduct.

Arpaly, Nomy (2002). Unprincipled Virtue: an Inquiry into Human Agency. Oxford UP.

Babbitt, Susan E. (1996). Impossible Dreams: Rationality, Integrity, and Moral
Imagination: Westview Press.

Barthes, Roland and Lionel Duisit (1975). “An Introduction to the Structural Analysis of
Narrative”, New Literary History, Vol. 6, No. 2.

Baumrind, Diana (1964). ”Some Thoughts on Ethics of Research: After Reading
Milgram’s “Behavioral Study of Obedience”.” American Psychologist, 19, pp.
421-423.

Benn, SJ. (1984) “Persons and Values: Reasons in Conﬂict and Moral Disagreement”,
Ethics, Vol. 95, no. 1, pp. 20-37.

Benson, Paul (1991). “Autonomy and Oppressive Socialization”, Social Theory and
Practice, Vol. 17, No. 3, 385-408.

Birenbaum, Arnold (1984). “Toward a Theory of Role Acquisition”, Sociological
Theory, Vol. 2, pp. 315-328.

Blake, Micheal (2001). “Geeks and Monsters: Bias Crimes and Social Identity”, Law and
Philosophy, Vol. 20, no. 2, pp. 121-139.

Borges, Jorge Luis (1970). “The Aleph”, The Aleph and Other Stories. E.P. Dutton and
Co.

Brison, Susan J. (2002). Aftermath: Violence and the Remaking of a Self. Princeton UP.

Calhoun, Cheshire (1995). “Standing for Something”, Journal of Philosophy, May
1995: 235-260.

Christman, John (1987). “Autonomy: A Defense of the Split-Level Self”, Southern
Journal of Philosophy, Vol 25, pp. 281-294.

Christman, John (1988). “Constructing the Inner Citadel: Recent Work on the Concept of
Autonomy”, Ethics, Vol. 99, No. 1, 109-124.

Christman, John (1991). “Autonomy and Personal History”, Canadian Journal of
Philosophy, Vol. 21, no. 1, pp. 1-24.

263

 

Christman (2004). “Relational Autonomy, Liberal Individualism, and the Social
Constitution of Selves”, Philosophical Studies, Vol. 117, No. 1-2, 143-164.

Cukor, George (1944). Gaslight (ﬁlm), Metro-Goldwyn-Mayer.
Dancy, Jonathan (2004). Ethics Without Principles, Oxford, Clarendon Press.

DeAngelis, Tori (2003). “Why We Overestimate Our Competence”, Monitor on
Psychology, Vol. 34, No. 2.

Dennett, Daniel (1989). “The Origins of Selves”, Cogito, 3, pp. 163-173.

Dennett, Daniel (1992). “The Self as a Center of Narrative Gravity”, in F. Kessel, P. Cole
and D. Johnson, eds, Self and Consciousness: Multiple Perspectives, Hillsdale,
NJ: Erlbaum.

Dubowski, Sandi Sirncha and Susan Korda, dirs. (2001). Trembling Before G—d. New
Yorker Films.

Dworkin, Gerald (1970). “Acting Freely”, Nous, Vol. 4, No. 4, pp. 367-383.
Dworkin, Gerald (1988). The Theory and Practice of A utonomy, Cambridge UP.

Einstein, Albert (1961). Relativity: the Special and the General Theory, Crown
Publishers.

Ensler, Eve (1998). The Vagina Monologues. Random House.

Frady, Marshall (2002). Martin Luther King Jr. Penguin.

Frankfurt, Harry (1988). The Importance of What We Care About. Cambridge UP, 1988.
Frankfurt, Harry (1999). Necessity, Volition, and Love. Cambridge UP.

Frankfurt, Harry (2004). The Reasons of Love. Princeton University Press.

Friedman, Marilyn ( 1986). “Autonomy and the Split-level Self”, Southern Journal of
Philosophy, Vol. 24, pp. 19-35.

Frye, Marilyn (1983). The Politics of Reality. The Crossing Press.
Hill, Thomas (1973). “Servility and Self-Respect”, The Monist, 57, 87-104.

Hitchens, Christopher (2003). “Mommie Dearest: The Pope Beatiﬁes Mother Teresa, a
Fanatic, a Fundamentalist, and a Fraud.” Slate, October 20, 2003.

Humberstone, LL. (1992). “Direction of Fit”, Mind, vol. 101, no. 401, pp. 59-83.

264

Hume, David (1739/1978). A Treatise of Human Nature. L.A. Selby-Bigge and
P.H.Niddich, eds. Oxford UP.

Huurre, T., H. Junkkari and H. Aro (2006). “Long-term Psychosocial effects of parental
divorce: A follow-up study from adolescence to adulthood”, European Archives
of Psychiatry and Clinical Neuroscience, 256, 4, pp. 256-263.

Kolodny, Niko (2003). “Love as Valuing a Relationship”, Philosophical Review, 112, 2,
pp. 135-190.

Korsgaard, Christine (1996), The Sources of Normativity, Cambridge UP.

Kposowa, Augustine J. (2000). “Marital status and suicide in the National Longitudinal
Mortality Study”, Journal of Epidemiology and Community Health, 54, pp. 254-
261.

Levenkron, Steven (2000). Anatomy of Anorexia, Norton.

Locke, John (1690/1964). An Essay Concerning Human Understanding, A.D. Woozley,
ed. Meridian Publishing.

Lugones, Maria (2003). Pilgrimages/Peregrinajes: Theorizing Coalitions Against
Multiple Oppressions. Rowman and Littleﬁeld.

Luoma. Jason B. and Jane L. Pearson (2002). “Suicide and Marital Status in the United
States, 1991-1996: Is Widowhood a Risk Factor?” American Journal of Public
Health, 92, 9, pp. 1518-1522.

MacIntyre, Alasdair (1981). After Virtue, University of Notre Dame Press.

Martel, Yann (2001). Life of Pi. Harcourt.

Mason, Gail (2002). The Spectacle of Violence: Homophobia, gender, and knowledge.
Routledge.

McF all, Lynne (1987). “Integrity”, Ethics, 98; 5-20.
Mehta, Ved (1993). Mahatma Gandhi and His Apostles. Yale UP.

Meyers, Diana Tietjens (1989). Self, Society and Personal Choice, Columbia University
Press.

Meyers, Diana Tietjens (2000). “Intersectional Identity and the Authentic Self” in
Relational Autonomy, Mackenzie and Stoljar, eds. Oxford UP: 151-180.

Milgram, Stanley (1963). “Behavioral Study of Obedience.” Journal of Abnormal Social
Psychology, vol. 67, no. 4, pp. 371-378.

265

 

 

Milgram, Stanley (1964). “Issues in the Study of Obedience: Reply to Baumrind.”
American Psychologist, 19, pp. 848-852.

Milgram, Stanley (1977). “Subject Reaction: The Neglected Factor in the Ethics of
Experimentation.” The Hastings Center Report, vol. 7, no. 5, pp. 19-23.

Mills, Charles (2004). “Ideal Theory as Ideology” in Moral Psychology: Feminist Ethics

and Social Theory, Peggy DesAutels and Margaret Urban Walker, eds. Rowman
and Littleﬁeld.

Mooney, Edward F. (1986). “Abraham and dilemma: Kierkegaard’s teleological
suspension revisited”, International Journal for Philosophy of Religion, 19, pp.
23-41.

Mullin, Amy (1995). “Selves, diverse and divided: Can feminists have diversity
without multiplicity?” Hypatia, Vol. 10, No. 4, pp. 1-31.

Murdoch, Iris (1970) The Sovereignty of Good. Routledge.
Nelson, Hilde Lindemann (2001). Damaged Identities, Narrative Repair, Cornell UP.

Nelson, Hilde Lindemann (2002). “What Child is This?” The Hastings Center Report,
vol. 32, no. 6, pp. 29-40.

Nolan, Christopher, dir. (2000) Memento [motion picture]. Columbia Tri Star.

Nozick, Robert (1989). The Examined Life: Philosophical Meditations. Simon and
Schuster.

Overall, Christine (1998). “Monogamy, Nonmonogamy, and Identity”, Hypatia, vol. 13,
no. 4, pp. 1-17.

Parﬁt, Derek (1984/1987). Reasons and Persons (1987 reprint with corrections), Oxford
University Press.

Puccetti, Roland (1973). “Brain Bisection and Personal Identity”, The British Journal for
the Philosophy of Science, Vol. 24, No. 4, pp. 339-355.

Rawls, John (1971). A Theory of Justice, Harvard UP.

Roethke, Theodore (1974). Straw for the Fire. Selected and arranged by David Wagoner,
Anchor Books.

Schechtman, Marya (1996). The Constitution of Selves, Cornell UP..
Scheman, Naomi (1993). Engenderings: Constructions of Knowledge, Authority, and

Privilege. Routledge; Ch 16: “The Body Politic/T he Impolitic Body/Body
Politics”, 185-192.

266

 

Scheman, Naomi (1997). “Queering the Center by Centering the Queer: Reﬂections on T
Transsexuals and Secular Jews” in Feminists Rethink the Self Diana Tietjens
Meyers, ed. Westview Press.

Singer, Irving (1987). The Nature of Love: Volume 3, The Modern World, University of
Chicago Press.

Singer, Irving (1994). The Pursuit of Love, Johns Hopkins UP.

Solomon, Robert C. and Kathleen Higgins, eds (1991). The Philosophy of (Erotic) Love.
UP of Kansas.

Strawson, Galen (2004). “Against Narrativity”, Ratio, Vol. 17, no. 4. pp. 428-452.
Styron, William (1979). Sophie ’s Choice, Random House.

Twain, Mark (1869). The Innocents Abroad.

Velleman, J. David (1999). “Love as a Moral Emotion”, Ethics, 109: 338-3 74.

Walker, Margaret Urban (1998). Review of Susan Babbitt’s Impossible Dreams, in
Hypatia, Vol. 13, No. 3, pp. 168-173.

Ward, Mylene, Larry Gruppen and Glenn Regehr (2002). “Measuring Self- Assessment:
Current State of the Art”, Advances in Health Science Education, 7, pp. 63-80.

White, Richard (1999). “Friendship and Commitment”, Journal of Value Inquiry, 33,1,
pp. 79-88.

Wilcox, W. Bradford and Steven L. Nock (2006). “What’s Love Got to Do With It?
Equality, Equity, Commitment and Women’s Marital Quality”, Social Forces,
Vol.84, no. 3, pp. 1321-1345.

Williams, Bernard (1973) “A Critique of Utilitarianism”, Utilitarianism for and Against,
J .J .C. Smart and Bernard Williams: Cambridge UP.

Williams, Bernard (1981). Moral Luck, Cambridge UP.

Wolf, Susan (1982). “Moral Saints”, The Journal of Philosophy, 79, 8: 419-439.

267

 

EEEEEEEEEEEE

llljlllllﬂlﬂllllllllllllllllllllllllllllllllll

2956 6