. 7
( 8)


One might argue that the ideas of choice and consent make little
sense with regard to intimate relationships. We do not choose those
whom we love or even those to whom we are sexually attracted. In
marrying, then, we can often act against our best interests, subjecting
ourselves to an open-ended term of unhappiness, violence, and even
death. For this reason, Card thinks that no one should choose marriage
and, more importantly, that no one should struggle to open it up to
couples now barred from marrying.55 Nevertheless, the choice and
consent at issue in the right to marry are not choices to love or to be
sexually attracted to someone. They are rather choices about long-
term investments in, and involvements with, another person. We can
love someone and choose not to marry him or her. The right that the
meaning of civil marriage in the United States gives us is the right to
marry the person we choose.
Yet, if the identities entailed by the right to marry are identities
as free choosers and consenters, and if the right therefore fails to
involve either whites or non-whites, prison inmates or non-inmates
or indigent or non-indigent parents, it also fails to involve either men

Card, ˜˜Against Marriage and Motherhood,™™ p. 6.
M A R R I A G E , T H E M I L I T A R Y , A N D I D E N T I T Y 209

or women. If human beings are not interchangeable in the way that
railway cars are (or were supposed to be), then to deny someone his or
her choice of a spouse, however we understand that spouse, is to deny
him or her a ˜˜vital personal right.™™ Courts have allowed that these
choices can be restricted for purposes of health and particular states
have placed restrictions on marriages of people younger than what
those states consider to be the age of consent. Yet, they have also
repeated the claim in Zablocki v. Redhail that such gate-keeping
laws cannot ˜˜interfere directly and substantially with the right to
marry.™™56 Hence, neither understandings of participants in marriage
as blacks and whites nor understandings of participants as men and
women are intelligibly related to the meaning of that right as that
meaning has developed in the United States. This logic is the one
that courts in Massachusetts and New York City followed in deciding
that bans against marriage between same-sex partners violated
the Massachusetts and New York constitutions. In Goodridge v.
Department of Public Health, the Massachusetts Supreme Court
understood the right involved in marriage to be ˜˜at the core of indi-
vidual privacy and autonomy,™™ and it concluded that the right ˜˜would
be hollow if the Commonwealth could, without sufficient justifica-
tion, foreclose an individual from freely choosing the persons with
whom to share an exclusive commitment in the unique institution of
civil marriage.™™57 Although her decision was overturned by the higher
courts, the trial judge in Hernandez v. Robles agreed ruling that ˜˜The
˜liberty at stake™ is the freedom to choose one™s spouse. Thus, for the
State to deny that freedom to an individual who wishes to marry a
person of the same sex is to deny that individual the fundamental right
to marry.™™58
What does this understanding of the participants in civil marriage
as free choosers and consenters rather than as whites, non-whites,

Zablocki v. Redhail, p. 387.
Goodridge v. Department of Public Health 440 Mass. 309 (2003) pp. 328“329.
Hernandez v. Robles Supreme Court of New York, New York Country, 794 NYS 2d
579 (2005), p. 601.
210 A F T E R I D E N T I T Y

parents, men, or women mean for the potential ills to which those
opposed to a right to marry for same-sex couples point? If we extend
marriage rights to individuals irrespective of these identities, must we
also permit polygamy, bigamy, or marriages between parents and
children? Both bigamy and marriages between parents and children
would seem to be precluded by understanding the participants in
marriage as free choosers and consenters. In a bigamous relationship
one of the two partners does not know that the other is married to one
or more other people as well. Yet, if the partner has no knowledge of
these other relationships, it is hard to see how he or she could be
understood as a free chooser of or consenter to the arrangement. Nor
is it clear how children below the age of consent can be free choosers or
free consenters to another™s choice. Indeed, defenders of a ban on
marriage between same-sex partners emphasize children™s need for
nurturance, guidance, and parental authority. Hence, although these
defenders worry that lifting bans against marriage between same-sex
partners leaves no principled protection against incestuous marriages
between parents and children, the very capacities for choice and con-
sent that they imply children do not possess would seem to provide
just such a principle. We might disagree on when children reach the
age of consent. Yet, in detailing children™s need for protection as a
basis for civil marriage, defenders of a ban on marriage between same-
sex partners already provide the principle for prohibiting marriages
between parents and their young children. In any case, incestuous
marriages of any sort are ruled out by the legitimate gate-keeping
functions of marriage laws insofar as these are concerned with health
Of course, if we understand marriage as a way of off-loading
inquiries into private relationships onto a common coinage that com-
mands immediate respect and if we understand those with a right to
marriage as free choosers and consenters, little in this understanding
seems to preclude an extension of marriage to consensual groups or to
serially intimate relationships. Indeed, polygamous marriages are tra-
ditional in much of the world and serially intimate relationships are
M A R R I A G E , T H E M I L I T A R Y , A N D I D E N T I T Y 211

part of certain gay and lesbian communities. Why should the partic-
ipants in these relationships not have a right to the common coin of
marriage? Why should three or more individuals not have a right to
marry one another? One can also imagine a web of long-term relation-
ships that A has with B and C and that B has with A, D, and E. Why
should A not be able to marry B if B knows that A is also married to C
and why should B not be able to marry A if A knows that B is also
married to D and E? In this case, both parties can be said to consent
freely to the arrangement.
One might try to argue against a ban on polygamous marriages
on the basis of the history of the Mormon Church and the constitu-
tional guarantee of the free exercise of religion. In other words, one
might argue that in prohibiting polygamous marriages the law targets
a specific religion in a way that the state™s legitimate defense of the
health, safety, and welfare of it citizens cannot justify. This argument
is a difficult one to make, however, since the Mormon Church no
longer includes polygamy as one of its authorized practices. One
might also argue that while the development of marriage law in the
United States clarifies marriage™s contours, rights, and identities, that
law has yet to address fully the question of its binary character.
Instead, in the case of Utah, the federal government made accepting
this binary character a condition of admission into the Union. For this
reason, the move from a dyadic to a triadic or serial structure for
marriage would have to look to a different pedigree than the one to
which marriage between same-sex partners has access. Although this
latter pedigree is one gradually articulating the identities of those with
a right to marriage, it has left the question of the number of identities
To be sure, this analysis is insufficient as an argument for
excluding polygamy and serial forms of intimacy. If marriage is plau-
sibly understood as the right of free choosers to an immediately
respected form of legitimate intimacy, this understanding may well
encompass not only marriages between same-sex couples but other
forms of marriages as well. This concession, however, is unlikely to be
212 A F T E R I D E N T I T Y

satisfactory for many critics of marriage since civil marriage, even
under these conditions, would continue to legitimize certain intimate
relationships at the cost of delegitimizing others. Marriage, Michael
Warner says, ˜˜is the zone of privacy outside of which sex is unpro-
tected.™™59 The meaning of marriage and the logic of marriage rights
may dictate opening the institution to same-sex couples. At the same
time, the very legality of civil marriage gives the state license to
intrude on other relationships, to outlaw prostitution and bigamy,
and to condemn promiscuity, for example. Hence, Warner thinks we
should imagine ˜˜a world capacious enough in its recognition of house-
holds to be free from . . . invidious regulatory institutions.™™60 Yet,
presumably such a world would still need mechanisms for deciding
who should legitimately have mostly unquestioned access to whom
and under what conditions. Questions about inheritance, hospital
visitation rights, and the possibility of staying in a long-inhabited
home would still arise, as would questions about distributing com-
mon property if a household splits up. For these reasons, removing the
common coin of marriage would doubtless lead to more rather than
less intrusions by the state into private relationships. All relationships
would be subject to the scrutiny that the Erlangers received. If all
households are freed from state licensing, then none are free from
intrusive monitoring and regulation.
Nevertheless, we need not defend civil marriage as an institu-
tion, even one open to more forms of intimate relationship than it now
includes, in order to point out that nothing about it leads to an under-
standing of its participants as men and women. Instead, understanding
participants in the institution of marriage in sex and gender terms is as
distorting as understanding them in racial terms. If we return to the
hermeneutic circle of whole and part and take the history of marriage
and marriage rights to be the whole into which we must integrate the
parts, then, while we can integrate the identities of free choosers
and consenters, we cannot integrate the identities of races, inmates,

59 60
Warner, The Trouble with Normal, p. 96. Ibid., p. 105.
M A R R I A G E , T H E M I L I T A R Y , A N D I D E N T I T Y 213

non-custodial parents, or sexes and genders. For this reason, the strug-
gle to legalize marriage between same-sex partners may be more
important than its critics think. If successful, it will eliminate from
a socially influential public institution the plausibility of understand-
ings of its participants as men and women. What matters to marriage
is the choice to enter into it, not the economic or social status of those
that do so and not their political, racial, religious, sex, or gender
identities. The same holds for another important public institution,
the armed services.

It might seem that current law understands members of the armed
services in terms of sexual identities rather than in terms of sex and
gender ones: heterosexuals and closeted homosexuals are welcome
whereas open homosexuals are not. Under the ˜˜Don™t Ask, Don™t
Tell™™ policy, officials are not supposed to question either applicants
for military service or those currently serving in the military about
their sexual orientation. If applicants or service members reveal
homosexual orientations, however, or if they are discovered engaging
in homosexual conduct, they can be rejected from the armed services
or discharged. Yet, it is difficult to see how we can describe sexual
orientations or sexual conduct as heterosexual or homosexual without
thinking about those engaged in it or oriented toward engaging in it as
either two men, two women, or one of each. Sexual acts with a man are
normally not grounds for rejection or discharge if one is oneself a
woman and the same holds for sexual acts with a woman if one is a
man. They are grounds for rejection or dismissal if one engages in
sexual acts with a man and is a man or if one does so with a woman
and is a woman. To this extent, sex and gender remain the lens
through which the military understands service members and poten-
tial service members even though it addresses its policies to
To be sure, it is not quite clear what the military means by
homosexuals. The ˜˜Don™t Ask, Don™t Tell™™ policy defines a homosexual
214 A F T E R I D E N T I T Y

as ˜˜a person . . . who engages in, attempts to engage in, has a propensity
to engage in, or intends to engage in homosexual acts.™™61 It defines
˜˜homosexual acts,™™ in turn, as ˜˜any bodily contact, actively undertaken
or passively permitted, between members of the same sex for the
purpose of satisfying sexual desires and . . . any bodily contact between
service members of the same sex that a reasonable person would under-
stand to demonstrate a propensity or intent to engage in™™ such homo-
sexual acts.62 At the same time, a service member can rebut the charge
of homosexuality by, among other demonstrations, showing that his or
her homosexual conduct is ˜˜a departure from the member™s usual and
customary behavior,™™ that it is ˜˜unlikely to recur,™™ and that he or she
˜˜does not have a propensity or intent to engage in homosexual acts.™™63
Thus, not all homosexual acts constitute grounds for dismissal. Instead,
it is evidently only homosexual soldiers who cannot engage in homo-
sexual acts.64 The Ninth Circuit Court of Appeals made this clear in the
case of Watkins v. US Army by deciding that ˜˜If a straight soldier and a
gay soldier of the same sex engage in homosexual acts because they are
drunk, immature or curious, the straight soldier may remain in the
army while the gay soldier is automatically terminated.™™65 But, how,
then, are homosexual and non-homosexual soldiers to be distin-
guished? How many homosexual acts make one a homosexual? What
sorts of acts?
The military policy on homosexuality trades on the same sort of
ambiguity that courts employed in the nineteenth century in deciding
cases of racial identity. Just as courts deciding racial prerequisite cases

Title 10, Armed Forces, Subtitle A, General Military Law, Part II, Personnel, Chapter
37, General Service Requirements 10 USCS x654 (2005) section f(1).
10 USC x654, section f(3A) and (3B).
10 USC x654, section b(1A), (1B), and (1E).
The pre-1993 policy makes this delineation explicit by noting that the intent of
opportunity for rebutting the charge of homosexuality ˜˜is to permit retention only of
non-homosexual soldiers who, because of extenuating circumstances, engaged in,
attempted to engage in or solicited a homosexual act,™™ Watkins Cited in v. United
States Army 847 F.2d 1329 (1988), n. 11. Also see Martha Nussbaum, ˜˜A Defense of
Lesbian and Gay Rights,™™ in Martha Nussbaum, Sex and Social Justice (Oxford:
Oxford University Press, 1999), p. 188.
Watkins v. United States Army, p. 1339.
M A R R I A G E , T H E M I L I T A R Y , A N D I D E N T I T Y 215

defined white in whatever way would permit them to exclude those
they wanted to exclude, the military policy defines homosexuality in
ways that allow it to retain or eject service members at will. Miriam
Ben-Shalom asked in 1974 why she was not being discharged. The
answer, ˜˜We have no arguments with you, so don™t worry about it,™™
implies, as Card points out, ˜˜that if they did have ˜arguments™ that
were insufficient for a discharge, they could trot out the policy against
lesbians.™™66 We can also look at the case of Perry Watkins, a soldier
who served in the army during the Vietnam War and who repeatedly
told superiors that he was a homosexual. Although many soldiers
were dismissed on the basis of their sexual orientation, he was not.
After the war, however, when he tried to re-enlist, the Army tried to
discharge him on the basis of his homosexuality. Here the one ˜˜homo-
sexual act™™ the Army pointed to was that of squeezing another sol-
dier™s knee (although the person whose knee was touched could not
remember which ˜˜black™™ soldier had squeezed it).67 The Army thus
implied that declaring oneself to be a homosexual was not enough to
be one during combat, although squeezing someone™s knee was suffi-
cient to be a homosexual during peacetime.68 Watkins was not the
only target of this discrepancy. In general, the military is less con-
cerned with homosexuality during wartime while the number of dis-
charges on the basis of homosexuality increases during peacetime.69
The ambiguities involved in the military™s view of who and who
is not a homosexual would, as Martha Nussbaum writes, ˜˜be the stuff
of high comedy, or even farce™™70 except that, as in the case of ambi-
guities in racial understandings, they determine people™s lives.
Whatever understanding one has, however, of who is a homosexual,
this understanding also requires understanding individuals in sex and
gender terms. Of course, given that the military acknowledges that

Card, ˜˜The Military Ban and the ROTC,™™ p. 176.
Watkins v. United States Army, n. 2.
See Card, ˜˜The Military Ban and the ROTC,™™ pp. 175“176.
Randy Shilts, Conduct Unbecoming: Gays and Lesbians in the US Military (New
York: St. Martin™s Press, 1993), p. 6.
Nussbaum, ˜˜A Defense of Lesbian and Gay Rights,™™ p. 188.
216 A F T E R I D E N T I T Y

one can engage in homosexual acts without being a homosexual,
determining the sex and gender of a service member and his or her
sexual partner is not sufficient for determining that service™s mem-
ber™s ˜˜sexual identity.™™ Still, it is a start. Hence, as in the case of
registering for the draft, military service and the identities entailed
by it are still understood in sex and gender terms. The question then is
whether this understanding is a plausible one.
The primary ground on which federal policy defends its prohib-
ition of military service on the part of openly homosexual individuals
involves their potential threat to ˜˜high morale, good order and disci-
pline and unit cohesion,™™ where by ˜˜unit cohesion™™ it means ˜˜the
bonds of trust among individual service members.™™71 The argument
for the ˜˜Don™t Ask, Don™t Tell™™ policy is not that the presence of
homosexuals presents a threat to this group cohesion but rather that
knowledge of their presence does. The idea here is that military effec-
tiveness requires that military units operate as a cohesive force and
that introducing into a unit individuals whose acknowledged sexual
orientation differs from that of the majority will disrupt cohesion.
˜˜Open homosexuals would paralyze a unit, and degrade unit cohesion
and erode combat effectiveness.™™72 This argument is odd, however, for
at least two reasons. In the first place, one of the first tasks of military
training is to break down individuals™ primary group loyalties in order
to reform them into a new group with principal allegiances to one
another. Indeed, as Elizabeth Kier points out, groups composed of
individuals who are too similar to one another in attitudes and values
can be dangerous to the overall military mission. Desertions in the
Confederate Army, for example, were highest in companies composed
of individuals from the same general location.73 As a consequence,
˜˜few modern armies attempt to create homogeneous groups on the

10 USC, x654, section a(6), a(7).
Air Force Chief of Staff, General Merrill McPeak, in testimony before US Senate
Committee on Armed Services, July 20, 1993, in Policy Concerning Homosexuality
in the Armed Forces (Washington, DC: Government Printing Office, 1995), p. 710.
See Elizabeth Kier, ˜˜Homosexuals in the US Military: Open Integration and Combat
Effectiveness,™™ International Security, 23 (2), 1998, p. 16.
M A R R I A G E , T H E M I L I T A R Y , A N D I D E N T I T Y 217

basis of common ethnicity, race, class, regional origin, age, personality
traits, or upbringing.™™74 In the second place, even where socially cohe-
sive groups are not a direct threat to military effectiveness, they are
less efficient than groups whose cohesion rests on other grounds. Kier
takes the sort of cohesion crucial to military effectiveness to be ˜˜task
cohesion,™™75 or cohesion constructed out of the need to accomplish
specific purposes. Groups that are cohesive in this way are more
efficient than socially cohesive groups because they do not devote
any of their attention to maintaining personal relations or social
communication and because, since they have no personal investment
in the interpersonal relationships, they do not hesitate to correct any
actions on the part of their members that may be counter-productive.
The question, then, is why federal policy should suppose that
the integration of openly gay or lesbian individuals into the military
would be any different than the integration of different ethnicities,
races, classes, or those of different ages, regional backgrounds,
upbringings, and attitudes. Indeed, if unit cohesion is enhanced by
diversity, one would suppose that the military would be interested in
whatever diversity it could find: not only the diversity of homosexual
and heterosexual identities, but those with identities as Northerners
and Southerners, Red Sox fans and Yankees fans, intellectuals and
deep-sea divers. Despite widespread concerns that the racial integra-
tion of the American military would diminish cohesion and effective-
ness, it did not. As one enlisted person explained when he was
interviewed in 1951, ˜˜when it comes to life or death, race does not
mean any difference.™™76 In another 1951 interview, a service member
said ˜˜Concerning combat, what I™ve seen an American is an American.
When we have to do something we™re all the same.™™77 In the early 1990s
service members made the same point about women. One said ˜˜we
don™t see it as male and female, we see it as a team™™78 and another

74 75
Ibid., p. 22. Ibid., p. 17.
Cited in Kier, ˜˜Homosexuals in the US Military,™™ p. 26.
Cited in Kier, ˜˜Homosexuals in the US Military,™™ p. 26.
Cited in Kier, ˜˜Homosexuals in the US Military,™™ p. 27.
218 A F T E R I D E N T I T Y

in basic training said, ˜˜there was some initial flirtation between the sexes,
but that was quickly moved to the back burner as the trainees realized
that teamwork was essential if everybody wanted to graduate.™™79 In
the context of basic training or combat, then, individuals™ primary
understanding of one another is not as blacks and whites or men and
women but as Americans and team members. Yet, if race and sex
identities as blacks, whites, men, or women do ˜˜not mean any differ-
ence™™ why should identities as homosexuals or heterosexuals? Why
would the integration of African Americans, Irish Americans, Latinos,
Latinas, Asians, Northerners, Southerners, and Westerners increase task
cohesion, and the integration of gays and lesbians diminish it?
Tarak Barkawi and Christopher Dandeker argue that neither the
integration of African Americans nor the integration of women into
the armed services serves as a good indication for the harm to unit
cohesion that the integration of openly homosexual service members
can cause. Like Kier, they point to the need to break down primary
group loyalties in order to reform recruits into a cohesive unit with
principal allegiances to one another. Yet, they stress that the identities
that are reformed in this way are specifically masculine. Recruits
must be transformed into soldiers who are competitive, aggressive,
and willing to kill. Such a transformation, Barkawi and Dandeker
contend, produces what they call ˜˜warrior masculinity.™™ This mascu-
linity is not undermined by the recruitment of women because they
can undergo the same transformation. ˜˜Indeed,™™ Barkawi and
Dandeker write, ˜˜the contemporary image of the service woman is
precisely not that of ˜traditional femininity™ but of someone who is
just as tough and capable as male soldiers given the limitations of
physical strength.™™ The same, they think, is not true for homosexual-
ity, for it is not ˜˜compatible with socially derived constructions of
warrior masculinity.™™80

Cited in Kier, ˜˜Homosexuals in the US Military,™™ p. 28.
Tarak Barkawi and Christopher Dandeker, ˜˜Rights and Fights: Sexual Orientation
and Military Effectiveness,™™ International Security, 24 (1), 1999, p. 185.
M A R R I A G E , T H E M I L I T A R Y , A N D I D E N T I T Y 219

Yet why not? Barkawi and Dandeker criticize Kier for not mak-
ing a distinction between ˜˜sex™™ and ˜˜gender.™™ Women™s biological
˜˜sex™™ does not, in their view, mean that they cannot be made mascu-
line in the ˜˜gender.™™ Since they acquire a feminine gender through
acculturation and socialization in the first place, re-acculturation and
re-socialization can transform it. But if women can be made masculine
why cannot homosexuals? By the idea of masculinity, Barkawi and
Dandeker designate traits of aggressiveness, competitiveness, and
willingness to kill. We can certainly question whether this designa-
tion is a fair one, given the many other traits we might prefer to label
masculine. Still, if we accept it and if we also agree with Barkawi and
Dandeker™s refusal to link masculinity with what they refer to as male
or female ˜˜sexes,™™ it becomes entirely unclear why it should be linked
to sexual preferences. Barkawi and Dandeker try to make this con-
nection between masculinity and heterosexuality by noting both the
prevalence of prostitution around military bases and the marching
songs that refer to women as whores.81 By masculinity they therefore
mean not only aggressiveness, competitiveness, and a willingness to
kill but also a willingness to pay for sex and to sing songs demeaning to
women. Of course, it still remains unclear why these traits should be
labeled masculine ones since many men do not view themselves and
are not viewed by those who know them as aggressive, competitive or
willing to kill, pay for sex, or sing songs demeaning to women. Other
people might possess some of these traits and preferences and not
others. Indeed, we might mix and match all of these characteristics:
we can understand ourselves and be understood by others as mascu-
line, unwilling to fight, competitive, not aggressive, and interested
sexually in just about anyone who comes along.
In any case, if prostitution and marching songs make no differ-
ence to the integration of female service members, why suppose that
they would make a difference to the integration of open homosexuals?
Barkawi and Dandeker refer to the heterosexuality of a warrior

Barkawi and Dandeker, ˜˜Rights and Fights,™™ p. 184.
220 A F T E R I D E N T I T Y

identity, but what is this? How is the question of whom one would be
willing to sleep with connected to the question of whom one would be
willing to kill? Even those who are against allowing openly homo-
sexual individuals into the armed services concede that they have
been and continue to be effective soldiers. The Army consistently
commended Perry Watkins from the time of his enlistment in 1967
until it sought to discharge him in 1982. Moreover, Kier points out
that ˜˜discharge proceedings against homosexuals are filled with testi-
mony of many of these individuals™ outstanding records, dependability
and dedication to their jobs.™™82 In these cases, the discharged homo-
sexuals had apparently adopted warrior masculinities although doing
so did not save their careers. By defining masculinity in terms of a set
of heterogenous traits and a specific sexual orientation, Barkawi and
Dandeker embroil themselves in the same morass of perplexing iden-
tity determinations that plague the way the armed services try to
define homosexuality.
Suppose we were to rethink the military policy by looking at the
point of the armed services, just as we looked at the meaning of
marriage and marriage rights. In its ˜˜Policy concerning homosexuality
in the Armed Services™™ the military insists that there is no constitu-
tional right to serve.83 At the same time, it states that ˜˜The primary
purpose of the armed forces is to prepare for and to prevail in combat
should the need arise.™™84 The tasks of preparing for and prevailing in
combat require a number of skills, assets, and tools. During the
Vietnam War, the military developed smaller and lighter weapons
that could be used by Vietnamese soldiers who were smaller than
their American counterparts. Given the availability of these sorts of
weapons there is no longer any rationale for distinguishing the possi-
ble combat roles of men and women. There is also no rationale for
distinguishing homosexuals and heterosexuals in their potential ˜˜to
prevail.™™ The military tries to distinguish between the valor of

Kier, ˜˜Homosexuals in the US Military,™™ p. 6.
83 84
10 USCS x654, section a(2). 10 USCS x654, section a(4).
M A R R I A G E , T H E M I L I T A R Y , A N D I D E N T I T Y 221

individual homosexuals and the effect of open homosexuals on unit
cohesion. It need only look at the testimony in Perry v. United States
Army to see that this rationale will not wash. In 1975, Watkins™ fellow
service members testified that ˜˜Watkins™s homosexuality was well-
known but caused no problems and generated no complaints.™™85

Card asks the same question about the ˜˜Don™t Ask, Don™t Tell™™ policy
that she asks about the definition of marriage as the union of one man
and one woman: namely why anyone, particularly gays or lesbians,
should care given that both the military and marriage are suspect
institutions. However, while she does not think we should fight to
open up marriage to same-sex couples, she thinks that we should fight
to make the military accessible to them. Why? Her reasoning holds for
both institutions. ˜˜What is at stake is one™s dignity in communities in
which one lives daily.™™86
Reserving marriage for opposite-sex couples and military oppor-
tunities for heterosexuals and closeted homosexuals betrays an unten-
able conception of identity as monolithic. While we can understand
those who want to marry each other as free choosers and consenters
nothing about either marriage or the right to marry makes it plausible
to understand them as African Americans, Irish Americans, baseball
fans, prison inmates, men, or women. We cannot view civil marriage
as a union for the purposes of procreation if we want to make sense out
of the legality of marriages in which children are not possible,
intended, or forthcoming. Nor can we view civil marriage as a two-
in-one-flesh communion if we want to make sense out of the legality
of a host of other marriages. If, instead, we understand civil marriage
as the publicly recognized zone of privacy for an intimate relationship
there is no reason to limit its legitimate participants by their prefer-
ences, sexual or otherwise. It is equally odd to suppose that armed

Watkins v. United States Army, p. 1331.
Card, ˜˜The Military Ban and the ROTC,™™ p. 191.
222 A F T E R I D E N T I T Y

service members are men and women. There is no right at issue in this
case, as the military makes clear. Nevertheless, if the point of the
military is ˜˜to prepare for and to prevail in combat should the need
arise,™™ it is at least worth asking whether the participants the military
should try to recruit are not those able to prepare and prevail. The
identification of service members and potential service members as
homosexuals or heterosexuals and, hence, men or women, reflects a
misunderstanding of who or what these service members and poten-
tial service members are. They are not men and women and hence not
homosexuals and heterosexuals any more than they are baseball fans
and Barbie-doll collectors, chess players, and race car drivers.
Imposing sex and gender identities on the military imposes identities
that make no sense in the context and forces out identities that do
make sense: those of willing warriors.
Hermeneutics and the politics
of identity

The idea of a hermeneutic circle of whole and part might seem to be
an odd idea to bring to issues of identity. Indeed, one might argue
that David Reimer™s troubles began because of his parents™ and physi-
cians™ presumptions about the need to integrate parts and wholes.
Hermeneutic premises project unity on texts and look to a standard
of coherence as a criterion for revising interpretive projections of
meaning that cannot be integrated with one another. If one™s under-
standing of a part of the text cannot be integrated with the meaning
one has projected for the whole, one has to revise either one™s under-
standing of the part or one™s understanding of the whole. In David™s
case, the loss of part of his body suggested to his parents and physi-
cians that they revise the whole of his sex and gender identity. This
same need for revision in the name of coherence explains surgeries on
the genital parts of intersexuals so that the whole of their bodies can
coherently mean one sex and gender. It also explains sex-reassignment
surgeries on the part of individuals who think that their inner and
outer selves do not cohere with one another. Such appeals to coher-
ence might even try to justify attempts to ˜˜cure™™ homosexuals on the
theory that their sexual desires are at odds with their sexes and genders
and need to be revised to be consistent with them.
Nevertheless, were we to apply the hermeneutic circle in such a
narrow and dogmatic way, we would have to ask whether we should
bleach the skins of ethnicities who act white or Anglo or bring up
infants who lose their thumbs as animals. Instead, the hermeneutic
circle supports a radically situational account of identity. The
assumptions behind David Reimer™s operation and upbringing were,
first, that identity as a boy undergirds or is a part of all life-contexts
and, second, that it requires a penis. Yet the second assumption
224 A F T E R I D E N T I T Y

overlooks the variety of ways we can understand individuals as boys or
men. There are no necessary and sufficient conditions that exhaust
what it is to be a baseball fan. Rather, the identity is elastic, open to
differences in degree, and subject to variations in the habits, incomes,
and life-conditions of different individuals. In contrast to this elastic
approach to identity, David Reimer™s sex identity was meant to
depend upon only and absolutely the absence of a penis when it was
determined that he should be raised as a girl and his gender identity
was meant to depend only and absolutely on the absence of certain
behaviors when it was determined that he was really a boy. To be sure,
the medical profession can stipulate a strict definition of diseases such
as anorexia nervosa, reserving the designation of anorexic for those
who eat less than a certain number of calories a day. Yet this definition
is consciously stipulative and tailored to helping doctors treat a dis-
ease. Furthermore, its justification depends entirely on whether it
does help in treating the disease and the medical profession can
redefine the contours of the identity if it does not. In contrast, doctors
and psychiatrists do not agree on a stipulative definition of men and
women. Instead, they simply assume that strict definitions exist and
disagree on what they are. Nor does the medical profession suffi-
ciently reflect on its own agenda in attempting to define men and
women or boys and girls. Do doctors serve patients by surgically cor-
recting their bodies so that they cohere with standard male and female
models or by educating both patients and the public on the range of
variation? Must individuals always endure painful surgeries in order to
wear a dress coherently or might we use the possibilities opened up by
the structure of understanding to stretch the boundaries of sex and
gender identities so that they are more flexible and less dogmatic?
Nor need we buy into the first assumption behind David
Reimer™s operation and upbringing, that identity as a boy is a coherent
element of all life-contexts. In whatever way individuals are boys and
men, girls and women they are also more and other than boys and
men, girls and women. They may be baseball fans, philosophers,
poker-players, gourmets, and countless other identities as well. Seen
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 225

hermeneutically, lives are lived as a series of different contexts of
which different identities are parts. Moreover, lives loop. In thinking
about our future, we think about who we are, which identities we have
that we want to maintain, and which we want to revise or discard. At
the same time, in trying to understand who we are, we do so in terms
of a continuum that contains both our past and our future. Our task as
individuals is to develop and organize our identities in ways that give
our lives the meaning we want for them. Of course, some individuals
may find that being men is the most important identity they possess or
desire. In this way, being a man may count as one of David Copp™s self-
esteem identities1 or, in other words, as the central identity around
which these individuals organize their lives. For others, identities as
baseball fans or philosophers might be a more significant aspect of
their self-esteem identity and provide a better key than their sexes or
genders to who they most importantly are. Our question, however, has
not been so much about the place of identities in our moral psychology
as it has been about what identities are, whatever place they have in
our moral psychology. And the answer to that question is that identi-
ties, themselves, are answers to questions “ questions about who or
what we are and, crucially, questions that are always asked in partic-
ular contexts to which only certain possible answers make sense.
To be sure, it may be difficult to see what content an answer
such as ˜˜I am a man™™ can have if we strip ˜˜manhood™™ of its associa-
tions with traits such as aggression and a lack of interest in children. In
this regard, it is easier to see the sense of answers such as ˜˜I am a
philosopher™™ or ˜˜an Irish or African American,™™ since these answers
link individuals to traditions, disciplines, and ancestral histories
whereas the former seems to link individuals only to a set of disput-
able stereotypes. Nevertheless, for some individuals being a man is
the most important identity in their moral psychology and for some
of these, because of they way they understand what being a man is, the

David Copp, ˜˜Social Unity and the Identity of Persons,™™ Journal of Political Philosophy,
10 (4), p. 369.
226 A F T E R I D E N T I T Y

possession of a penis will be crucial. Hence, they will undergo pain-
ful operations in order to acquire a penis if they do not have one or
consider it a special feature of their bodies if they do. For others, a penis
will be less crucial to their identity and they may not undergo an
operation or give that particular body part pride of place at all.
The point about these assessments, however, is that we develop
them ourselves in the course of living our lives, in trying to figure out
what we care about. David Reimer™s tragedy was that other people
viewed it as their prerogative to make these assessments for him
and that they made them in not only dogmatic but also imperialist
ways. They deemed the possession of a particular appendage to be
necessary to the coherence of the whole of identity as a man; nor
could they conceive for David any identity other than a sex or gender
one. Accordingly, they intervened in his life in a disastrous way. As
David himself put their problem, ˜˜It™s like your whole personality,
everything about you is all directed “ all pinpointed “ toward what™s
between the legs. And to me, that™s ignorant. I don™t have the educa-
tion that these scientists and doctors and psychologists have, but to
me it™s very ignorant.™™2

Yet, parents inevitably attempt to mold their children™s identities.
In imparting their values they try to make them liberal Democrats
or conservative Republicans. They take them to baseball games to
try to develop a love of the sport; they give them ballet lessons and
they put them on soccer teams. These efforts may fail; the identity
a parent wants to centralize in the children™s moral psychology
may become inconsequential for them or even non-existent.
Nevertheless, much of the literature on identity focuses on the ques-
tion of the obligations of social and political institutions to help
parents in these endeavors and to support centrally identity-forming

Cited in John Colapinto, As Nature Made Him: The Boy Who Was Raised as a Girl
(New York: HarperCollins, Perennial Books Edn., 2001), p. 262.
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 227

communities.3 Should liberal democracies allow for collective rights
that permit individuals with particular American Indian tribal iden-
tities to engage in activities such as fishing or whale-hunting while
individuals with other identities cannot? Should liberal democracies
allow members of particular religious groups to exempt their children
from mandatory schooling if such schooling puts their religious iden-
tity at risk? If we support the prerogative of parents to try to mold
their children so that they take certain identities to be central to
them, do we not need to support and work to preserve the commun-
ities that sustain these identities? If we admit that someone can
make being Amish the central identity of his or her life and if we
also admit that this person can try to make being Amish the central
identity of his or her children™s lives, then we would seem also to
have to support governmental policies that work to sustain the
Amish way of life, for one cannot be centrally Amish without the
existence of an Amish community of which one is a part. Do we not
also need to allow whatever dispensations are necessary to enable
American Indian cultures to survive? Furthermore, why should trying
to develop our children™s identities as boys and girls or men and
women be any different from trying to develop their identities as
Navajo or Amish? If we need to support collective rights on the part
of the Amish, do we not need to support the efforts of doctors and
parents to preserve the sex and gender communities we currently
possess? Must we not make the same claim for racial communities?
Indeed, if the argument of this book makes sense, then all of our
identities have the same hermeneutic status as ways of understanding
others and ourselves. Hence, if the government supports the Amish
community as a way of supporting those that make their Amish

See Kwame Anthony Appiah, The Ethics of Identity (Princeton, NJ: Princeton
University Press, 2005); Amy Gutmann, ed., Multiculturalism: Examining the
Politics of Recognition (Princeton, NJ: Princeton University Press, 1994) and Amy
Gutmann, Identity in Democracy (Princeton, NJ: Princeton University Press, 2003);
Will Kymlicka, Multicultural Citizenship: A Liberal Theory of Minority Rights
(Oxford: Oxford University Press, 1995); Susan Moller Okin et al., Is
Multiculturalism Bad for Women? (Princeton, NJ: Princeton University Press, 1999).
228 A F T E R I D E N T I T Y

identities their central ones, it would appear that it should support
all of the communities that sustain all of the identities different
people might make central to their lives, including the communities
of major league baseball and neo-fascism. Conversely, we could
decide that given the status of all identities as understandings and
self-understandings of parts within contextual wholes, the govern-
ment should not be in the business of supporting any of them. Which
way should we go?
Calling the second option ˜˜benign neglect,™™ Will Kymlicka
denies that we can adopt it.4 To begin with, governments decide on
the language of courts, legislatures, and schools. In doing so, they
cannot help but favor one identity over others “ say, Anglophone
identities in the United States over Latino or Latina ones. Kymlicka
points out that English is not simply a natural outgrowth of the
language the majority of the US population uses. If it were, we
would have to change the language with shifts in immigration and
majority populations. Moreover, English was never simply a natural
outgrowth. In adding states to the union the federal government delib-
erately declined to accept territories unless or until English speakers
outnumbered non-English ones in the territories in question. It drew
state boundaries in ways that guaranteed a majority of English speak-
ers, as in the case of Florida. It delayed statehood until sufficient
numbers of Anglophones moved into the territory in question, as in
the case of Hawaii. And, where English speaking was not likely to
become dominant, it established a different sort of political unit, as in
the case of Puerto Rico.5 In these ways, then, the government worked
actively to support English-speaking communities over non-English-
speaking ones.
Kymlicka also insists that governmental support for certain
identities over others goes beyond language. Governments favor cer-
tain identities when they decide on public holidays such as Christmas
and when they decide on the contours of the work-week, selecting one

4 5
Kymlicka, Multicultural Citizenship, p. 108. Ibid., pp. 28“29.
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 229

that works for Christians, for example, instead of one that works for
Muslims. Governments also decide on public uniforms for the police
and military that can be more restrictive for certain expressions of
identity than they are for others and they design state symbols that
reflect certain identities, not others.6 Indeed, one might say that cer-
tain state symbols “ for example, state flags involving Confederate
symbols “ are not simply instances of a benign neglect of other iden-
tities but, instead, instances of an outright disrespect for them.
If, in even the best instances, governments cannot help but favor
certain identities over others, how are they to decide which to favor?
Kymlicka argues that because decisions on language, internal bounda-
ries, public holidays, work-weeks, and state symbols inevitably sup-
port the majority culture, what is necessary is ˜˜similar support for
minority groups through self-government and polyethnic rights.™™7
Groups that are entitled to self-government, he thinks, are those
such as American Indian groups whose cultures and territories were
invaded by what has become the majority culture. He also thinks that
ethnic and religious groups such as Jews, Muslims, and those with
different heritages deserve some accommodations: consideration in
designing state symbols, perhaps, as well as exemptions from holiday
closings and uniform codes.
Yet, even if we restrict our examination to their effect on reli-
gious and ethnic identities, such measures raise problems. A state flag
that looks neutral to African Americans in the southern United States
may appear to other groups to be a complete erasure of their culture.
To some French citizens, French laws that prohibit girls from wearing
headscarves to school may appear necessary as a way of sustaining
France™s secular culture, while to others they may appear to be an
attempt to undermine Muslim traditions and a Muslim identity.
Moreover, some communities, and perhaps all communities, con-
strain their members in certain ways. David Reimer suffered a partic-
ularly drastic form of cultural constraint when surgeons cut into his

6 7
Ibid., p. 115. Ibid., p. 115.
230 A F T E R I D E N T I T Y

body to make it conform to a standard sex and gender community.
Women in some countries suffer the same sort of invasive cultural
practice when they must undergo clitordectomies, ˜˜marriage by cap-
ture,™™ and even murder if they are accused of adultery.8 But there are
less drastic versions of cultural constraint as well. According to a 1939
ordinance of the Santa Clara Pueblo, children born of unions between
male members of the Pueblo and female non-members are themselves
full members of the Pueblo. In contrast, children born to female
members and male non-members are not. If a Pueblo woman knows
that having children with a Navajo man will deny her children the
right to live at the Pueblo, hunt or fish on the land, use irrigation
water, and share in economic benefits, is she really free of a cultural
constraint to marry within the tribe?9 On one reading, Santa Clara
women are denied at least some of the social and political rights that
the Pueblo grants to men. Should liberal societies grant these sorts of
cultures the sort of accommodations that Kymlicka advocates?
Kymlicka tries to deal with such cases by distinguishing
between ˜˜external protections,™™ which attempt to reduce a minority
culture™s vulnerability to majority decisions, and ˜˜internal restric-
tions,™™ by which minority cultures curtail the basic civil or political
rights of some of their members.10 He thinks that liberal societies
should support the former, establishing those collective rights neces-
sary to protect minority cultures and, hence, minority identities
against the encroachments of the majority culture. But he thinks
that liberal societies cannot support cultures that impose restrictions
on their members. Kymlicka does not think denying support need
lead to forcible intervention into the culture. Nevertheless, he thinks
that liberal societies can use incentives meant to encourage a liberal-
ization of illiberal communities. Still, as many commentators have
pointed out, the distinction between external protections and internal

See Susan Moller Okin, ˜˜Is Multiculturalism Bad for Women?,™™ in Okin, Is
Multiculturalism Bad for Women?, p. 18.
See Santa Clara Pueblo v. Martinez 436 US 49 (1978).
Kymlicka, Multicultural Citizenship, p. 152.
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 231

restrictions is often far from clear. Take the example of Quebecois
language laws that prohibit French-speakers and immigrants from
sending their children to English-speaking schools.11 Is this example
one of external protections, allowing for the survival of a French-
speaking culture in the English-speaking Canadian nation? Or is it
instead an instance of internal restrictions that forbid French-speaking
parents a right to send their children to the schools to which they want
to send them? The membership laws of the Santa Clara Pueblo raise
the same question. If the cost of a woman™s marrying a non-Santa
Claran man is the disenfranchisement of their mutual children is
this cost a way of preserving and protecting the Pueblo or an internal
constraint on its inhabitants?12
If we cannot use a distinction between external protections and
internal restrictions to separate identities and cultures worthy of sup-
port from identities and cultures that are not, what can we use? And if
we cannot find any criterion does the impossibility of ˜˜benign neglect™™
mean that governments must support all and every identity that any
given individual happens to find central to his or her life? Much of the
debate on issues of multiculturalism and collective rights assumes
that we must pick between two alternatives: either we pursue a
politics of difference that allows for the recognition of select identities
such as the Amish and the Pueblo or we pursue a politics of benign
neglect that leaves all minority identities to flounder equally within a
majority culture. Yet, if the account of identity and identification
I have argued for in this book makes sense, then selective recognition
and benign neglect are not the only options open to us, for the minority
and majority status of our identity shifts depending on the particular
context in which we are involved at a given time and the identities
that are part of that context. Both selective recognition and benign
neglect misconceive identity. We can be understood in many ways
and therefore possess many identities. We may be Amish in certain

See Charles Taylor, ˜˜The Politics of Recognition,™™ in Gutmann, ed., Multiculturalism,
pp. 52“53.
Also see Appiah, The Ethics of Identity, pp. 79“80.
232 A F T E R I D E N T I T Y

situations and hence in the minority; but we may be European
Americans in other situations and hence in the majority. The question
of which identities are to be recognized and which neglected depends
on which are intelligible parts of the context of concern and interpre-
tive framework at hand.
None of our identities penetrates every aspect of our lives. If I am
with my child I may be, for that period of time, a mother and if I
am playing in a chess tournament I am not a mother, unless, perhaps,
I am playing chess with my child. Similarly, if I am pregnant, I am a
pregnant person, again for a limited period. Indeed, given how limited
this period is, it is a least somewhat bizarre that the capacity to
become pregnant should be so hegemonic with respect to identity.
How often, we might ask, do contemporary human beings reproduce
or, at least, want to reproduce? And how much of a life does it really
compose? Individuals in advanced Western countries have some con-
trol over their reproductive lives and where individuals in developing
countries do not, they nonetheless often aspire to it. Hence, it seems
quite odd that we continue to define individuals in terms of their
reproductive role. Indeed, if the norm in the West is approximately
two children per family then we are defining certain people as females
or women on the basis of eighteen months of their lives. Even if we
identify individuals as females and women for somewhat longer
because we suppose that they have a greater responsibility for child
care and we equate femaleness with child-rearing, we are still defining
them in terms of a continually interrupted and, indeed, limited span of
time. And yet this identity is meant to be who and what one is at every
moment and in every sphere. Surely, if we go by length of time alone,
we would be better defined as workers or sleepers.
The same holds for being Amish, a Quebecois, or a Santa Clara
Pueblo Indian. Neither is all that one ever is. Indeed, it is conceivable
that one™s Amish, French, or Santa Claran identity is more prominent
to outsiders than it is to oneself. For neither outsiders nor insiders,
however, can an individual possess only one identity, any more than a
text can possess only one meaning. Instead, just as we approach texts
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 233

from within different interpretive wholes and therefore can under-
stand their meaning in different ways, we approach individuals from
within different wholes and therefore can understand their identities
in different ways. If the whole is the context of marriage, then those
individuals will possess identities “ put otherwise, they will be intel-
ligible as “ certain sorts of people different from those they are in the
context of the military, asthma-research, or childbearing. If the whole
is the context of religion, then they will be intelligible as Amish and
Quakers, Catholics and Wicca. Given this account of identity, how-
ever, we need to rethink the conclusions of two central texts in the
literature on the politics of identity: Charles Taylor™s defense of
French language laws in Quebec in his ˜˜The Politics of Recognition™™
and the 1970s Supreme Court ruling in Wisconsin v. Yoder,13 dealing
with the exemption of Amish children from schooling.

Taylor argues that certain restrictions on inhabitants of the Canadian
province of Quebec are justified as means of ensuring the survival of
French-speaking culture in Canada.14 Neither Francophones nor
immigrants are to send their children to English speaking-schools;
all signs are to indicate what they are about in the French language
and all businesses with more than fifty employees must be run in
French. In reviewing the restrictions, Taylor argues that ˜˜One has to
distinguish the fundamental liberties, those that should never be
infringed and therefore ought to be unassailably entrenched, on
the one hand, from privileges and immunities that are important but
that can be revoked or restricted for reasons of public policy . . . on the
other.™™15 What justifies Quebec™s public policy in restricting certain
language and schooling privileges is the goal of ensuring not only that
the French language remains a resource available to those who want to
make use of it but also that in the future there will be a community of

Santa Clara Pueblo v. Martinez 436 US 49 (1978); Wisconsin v. Yoder 406 US 205.
Taylor, ˜˜The Politics of Recognition,™™ p. 52.
Taylor, ˜˜The Politics of Recognition,™™ p. 59.
234 A F T E R I D E N T I T Y

people who do want to make use of it. In Wisconsin v. Yoder, the US
Supreme Court appealed to a similar logic of cultural preservation in
exempting Amish adolescents from formal schooling after the eighth
grade. ˜˜Compulsory school attendance to age 16 for Amish children,™™
it said, ˜˜carries with it a very real threat of undermining the Amish
community and religious practice as they exist today.™™16
Such policies and exemptions understand identity in monolithic
and absolutist ways, however. Whatever other reasons support Quebec
language laws or Amish exemptions from schooling, reasons that try
to preserve a cultural or religious identity do not. In the first place,
they allow for only one kind of a Francophone or Amish identity and,
in the second place, they allow individuals to be only Francophones or
Amish. Surely one can have a Francophone identity in different ways,
whether because one speaks French exclusively, or because one speaks
it at home, or because speaking French contributes to one™s life in
some other way. Just as it was odd to reduce David Reimer™s male
identity to the presence or absence of his penis, it is odd to reduce the
question of a French-speaking identity to conducting one™s business in
French or sending one™s children to a French-speaking school. Nor
does either necessarily contribute to enhancing a French-speaking
identity. Parents might send their children to French-speaking schools
not because they have identities as Francophones and want to ensure
that their children have identities as Francophones but simply because
they have identities as snobs. One can also possess identities in addi-
tion to a Francophone one, including, for example, a parental identity
concerned with the capacity of one™s children to flourish in an English-
speaking nation.
One can also surely be Amish in more than one way. It would be as
odd to assume that all Amish share every belief as it would be to assume
that all Catholics do. Moreover, one can be Amish and a student. The
Supreme Court agreed with the Amish that missing two additional years
of schooling was not likely to damage the psychological or physical

Wisconsin v. Yoder, p. 218.
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 235

health of the adolescents, burden the surrounding society, or render
them unable to support themselves. Yet, why not reverse the question?
How would an additional two years of schooling undermine the Amish
community? Moreover, if it would, why pick an adolescent™s Amish
identity as the identity to try to preserve over other identities he or she
might possess or, indeed, come to possess as a result of staying in school?
Suppose a greater amount of knowledge and exposure to other forms of
life renders an Amish teenager a skeptic about the Amish way of life.
Will some skeptic society of America now demand help to preserve that
culture? Individuals, even if they are Amish are no more Amish than
they are schoolchildren and as schoolchildren they should be treated
in the same way as other schoolchildren under the jurisdiction of the
state. The context of education is different from the contexts of religion
and life-style, and if the individuals in question are Amish in the latter
contexts they need not be in the former one.
Francophones, too, are no more Francophones than they are
businessmen or parents. If and only if sufficient numbers of individu-
als want identity as a French-speaker to be either one part or a central
part of their lives or their children™s lives does the identity have a
˜˜right™™ to survive. And even if it survives it will be neither monolithic
nor constant. People will be intelligible as French-speakers in different
ways and they will be intelligible as other identities in other contexts.
Moreover, support for identities outside of the contexts in which they
have their meaning is a misapprehension of the hermeneutic condi-
tions of identities “ or, in other words, understandings “ of who we are.
Such support is no less dogmatic than discriminatory laws that require
individuals to be races in the context of citizenship or than practices
that assume individuals are blacks in the context of driving.
These conclusions suggest a different way of thinking about
the politics of identity. For the important question now is not whether
the state has an obligation either to support us in our identities as
Francophones or Amish or to treat them with benign neglect. Rather,
the question is how governments can help to accomplish two different
tasks: to curtail the public determination of identities to the particular
236 A F T E R I D E N T I T Y

contexts in which particular identities make sense and to secure the
public framework of rights within which we can sort and shape the
identities we take to be important to our private flourishing. On
this view, the questions we should be asking are not whether Amish
adolescents should go to school or what sort of schools Francophone
children should attend. Rather, we should also ask how a government
can guarantee to individuals that they can be both Amish and school-
children or both Francophones and members of an English-speaking
nation. Moreover, if we ask this question, certain obvious answers
present themselves: the Amish might establish private schools that
meet state guidelines without posing a threat to an Amish identity
and the Qu´
ebec government might certify the right of parents to choose
their children™s schools while helping the Francophone community to
set up Saturday or Sunday schools for French language and culture on
the model of Hebrew schools or the Chinese and Japanese schools to
which some Chinese Americans and Japanese Americans send their
children on weekends.
In addition to a monolithic view of an Amish identity, however,
the ruling in Wisconsin v. Yoder betrays a monolithic view of the
identity that it thinks education is meant to help to develop. The
court reasoned that the Amish do not participate in the welfare
state, do not make use of social security funds, and do not become
burdens on the state. Hence, their continuing education is not neces-
sary in the way that is for a child whose future is not so assured.17
Yet this analysis supposes that the sole point of education is to make
sure that individuals can support themselves as workers. Others,
however, understand the identities that education is meant to
develop differently “ as democratic citizens, for instance “ and they
therefore deny that an eighth-grade education is sufficient.18 Instead,
children arguably need to develop an understanding of their rights,

See Wisconsin v. Yoder, p. 225.
See Richard Arneson and Ian Shapiro, ˜˜Democratic Autonomy and Religious
Freedom: A Critique of Wisconsin v. Yoder™™, in Ian Shapiro, Democracy™s Place
(Ithaca, NY: Cornell University Press, 1996), p. 146.
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 237

responsibilities, and opportunities as citizens as well as a capacity for
critical reasoning that allows them to assess opinions and views
either similar to or different from their own.19 In addition, children
arguably need some understanding of science, if only to be able to
evaluate environmental threats to their way of life or to assess the
implications of certain governmental policies. They could also use
an understanding of world history, the history of the United States,
and the history of democratic institutions, if only to recognize their
own position within these histories.20 Of course, many children
complete high school without acquiring these understandings and
skills. Nevertheless, as Richard Arneson and Ian Shapiro point out,
˜˜The failure of citizens . . . to provide education adequate for prepar-
ing youth for future citizenship does not justify a decision . . . to
cease upholding and enforcing these norms.™™21 Rather, if part of
the goal of education is the development of individuals who can be
competent members of a democracy, then the two years that Amish
children miss may well be crucial.
How are we to decide between these accounts of the identities
that an education is meant to help to develop? Are American schools
meant to produce workers or citizens? Obviously the answer here is
that education can surely produce both, but this answer confirms the
dogmatic character of the court™s decision on the Amish. Not only are
Amish adolescents both Amish and schoolchildren; as schoolchildren
they are more than future self-supporters. Instead, education in the
United States is meant to serve at least two goals: that of preparing
students to take up identities as workers in a global economy and that
of preparing students to take up identities as citizens in a multicul-
tural society and democratic polity. If, for the individuals in question,
their present identities as Amish preclude their future identities
as workers in a global economy, their Amish identities do not preclude
their future identities as citizens. In deciding as it did, then, the court

See Amy Gutmann and Dennis Thompson, Democracy and Disagreement
(Cambridge, MA: Harvard University Press, 1996), p. 65.
20 21
Ibid., p. 147. Ibid., p. 148.
238 A F T E R I D E N T I T Y

imposed a dogmatic and monolithic understanding on both the
adolescents in question, which it saw only as Amish, and on school-
children in general, which it saw only as future workers. Recognizing
that the adolescents must be understood to be more than Amish and
that schoolchildren must be conceived of as more than future workers
does not prohibit the Amish from setting up their own private high
schools, ones that they could presumably tailor to fulfill the goals of
democratic education and their own religious and cultural identity
needs. Recognizing multiple identities does prohibit a US court from
imprisoning the Amish or schoolchildren in any one of their multiple

It is not difficult to see how education can incorporate the goals of
creating reliable workers, competent citizens, and, in the case of
religious schools, possible believers. But can recognizing a multiplic-
ity of identities not often overburden institutions and practices? Can
differing interpretations of who and what we are not sometimes pre-
empt one another? Take the identities of being both a Christian
Scientist and a parent with a very ill child. For Christian Scientists,
illness is the result of spiritual alienation and imperfect understanding
so that, for them, prayer is a valued form of medical intervention.22 For
most Western doctors, medical care involves more scientifically
informed forms of intervention. Since their religion does not allow
Christian Scientists to receive conventional medical treatment, the
issue arises as to what state authorities are to do when Christian
Scientist parents withhold medical care from their gravely ill minor
In this instance, we cannot decide the question by tailoring the
identity to the framework of interpretation within which it is an
identity, for part of the problem is how to understand that framework.

See Anne D. Lederman, ˜˜When Religious Parents Decline Conventional Medical
Treatment for Their Children,™™ Case Western Reserve Law Review, 45, 1995, p. 918.
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 239

Nor can we allow for both frameworks of interpretation. When we
recognize the multiple ways in which we can understand Sense and
Sensibility, we come to admire the novel all the more and to marvel at
its countless interpretive meanings. When we recognize the different
ways we can understand a dying child, we are caught in a relativist
nightmare. Is the context for understanding the child that of health
or religion? Which contextual interpretation of the identity of the
child should be decisive: that of medicine within which the child is
a diseased corporeal body or that of Christian Science within which
the child is a soul alienated from God? Moreover, what is the proper
context for understanding the potential death of the child? For
Western medicine it is an avoidable event, looming only because of
the parents™ irrationality. For Christian Scientists, ˜˜What appears to be
an ending is merely a passing, ascending to a realm of higher under-
standing.™™23 Consequently, the focus of Western medicine on the body
alone is misdirected.
It may be that taking seriously the different interpretations of
a text that stem from different contexts and different textual relations
serves to deepen our understanding and appreciation of the text.
Yet, texts do not require us to act, whereas deciding how to proceed
in the context of medical care does. What, then, should doctors do
when parents refuse to permit them to care for their children? Shapiro
offers a possible way out of the problem. First, he distinguishes
between a child™s basic interests which include his or her needs for
food, shelter, education, and the like and the child™s best interests,
which involve interests that the family thinks are important to his or
her religious, ethical, or spiritual development, or to his or her partic-
ular talents and special needs.24 Second, like John Locke, Shapiro
argues that responsibilities for children™s interests are fiduciary ones.

Pam Robbins and Robley Whitson, ˜˜Mary Baker Eddy™s Christian Science,™™ in
Christian Science: A Sourcebook of Contemporary Materials (Boston, MA:
Christian Science Publishing Company, 1990), cited in Lederman, ˜˜When
Religious Parents Decline Conventional Medical Treatment,™™ p. 918.
Ian Shapiro, Democratic Justice (New Haven, CT: Yale University Press, 1999), p. 86.
240 A F T E R I D E N T I T Y

Parents are to represent their children™s interests until the children are
able to represent their own and they are to exercise authority over
their children only in their children™s interests.25 For some time,
Western governments have also had a fiduciary responsibility towards
children™s interests, not only providing for their education but also
looking out for their physical safety and working to protect their
health and nutrition.
In Shapiro™s scheme, state and parental responsibilities comple-
ment one another. Parents are the primary custodians of their own
children™s best interests and have ultimate authority over them. They
are the secondary custodians of their basic interests. Hence, where the
state fails to protect these basic interests parents can and must legit-
imately intervene. For its part, the state is the primary custodian for
children™s basic interests and secondary custodian for their best inter-
ests. In the case of some basic interests, such as health and nutrition,
the state usually gives up day-to-day control to the parents, subject to
the proviso that the state has ultimate authority in this area and can
intervene if the parents fail to discharge their tasks. The same holds
for children™s best interests. Where parents neglect these interests or
fight over what the child™s best interest involves, the state must
become the judge.
Using this scheme of fiduciary responsibility for best and basic
interests, Shapiro declares that ˜˜we should not be troubled when the
preferences of Christian Scientists to withhold essential medical care
from children are overridden by courts.™™ Rather, ˜˜these are instances
where parents™ conceptions of a child™s best interests lead to a viola-
tion of the child™s basic interests.™™ Since the state has ultimate fidu-
ciary responsibility for basic interests, it can and should intervene.26
Shapiro does not deny that Christian Scientist parents™ actions are
directed at their children™s best interests as they understand them “
in this case, their interests in spiritual salvation. Nevertheless, he
thinks that their concerns are properly overridden by the state since

25 26
Ibid., pp. 73“75. Ibid., pp. 93“94.
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 241

it is responsible for the children™s basic interests “ in this case, their
interests in physical survival.
Nevertheless, this division of duties raises more complicated
interpretive issues than Shapiro acknowledges. For, from a Christian
Scientist perspective, in rejecting medical care for their children,
Christian Scientist parents are securing their children™s basic interests
in compliance with their secondary fiduciary duty to be taken up
when the state cannot or will not fulfill its primary fiduciary role in
this area. The Christian Scientist parents, in other words, may under-
stand their duty as parents in the same way as Shapiro understands it:
namely, as a duty that directs them to protect their children™s best
interests in all circumstances and to protect their basic interests when
the state fails to do so. Yet, in withholding medical care, they take
themselves to be doing just that: looking out for their children™s best
interests in spiritual salvation and for their basic interests in spiritual
survival, precisely because the state will not.
Our understanding of education can encompass different dimen-
sions as an institution multiply geared to developing well-qualified
workers, democratic citizens, and, in some cases, religious believers.
In contrast, medical care cannot attend to the body without damaging
the soul according to Christian Scientists and cannot attend to the
Christian Scientist soul without damaging the body according to
Western medicine. The two different understandings of the child
as ill patient and alienated soul thus lead to two different but ulti-
mately inadequate responses: non-action in the face of imminent
harm or failure in properly respecting a minority identity. We can
ask why individuals should be Amish in the context of education
and we can understand them to be both Amish and high-school stu-
dents either at public schools, if their elders will allow it, or in private,
Amish schools, if they will not. Yet, it is more difficult to se how
children can be both gravely ill children and Christian Scientists
since in the eyes of their parents their being Christian Scientists
precludes their being gravely ill and in the eyes of the medical profes-
sion their being gravely ill precludes their being Christian Scientists.
242 A F T E R I D E N T I T Y

The contextual solution to the problem “ namely that the children are
patients in the hospital and Christian Scientists at church “ also fails.
For a devout Christian Scientist parent, the very fact that his or her
child is ill “ or, in their view, alienated from God “ indicates that the
relevant identity in the hospital is a religious one.
In 1996, Congress passed legislation that requires states to pro-
vide medical treatment for dangerously ill minor children but also
permits the states to allow for religious exemptions to findings of
parental abuse and neglect in instances in which the parents objected
to or failed to seek medical help.27 Perhaps we can take this law as an
example of the sort of compromise that might be necessary in such
cases. To be sure, ultimately the law favors what Kymlicka might
see as a majority identity. In the end, the children receive treatments
as medical patients, not Christian Scientists. Nevertheless, it is pos-
sible to view the law as trying to go as far as it can in recognizing and
respecting a minority identity and the hermeneutic perspective it
frames on who its children are. Perhaps more importantly, the law
asks that minority culture to recognize and respect the different iden-
tities its members have. They are not only Christian Scientists and not
only parents with their own understandings of the basic and best
interests of their children. In addition, they are members of a techno-
logically advanced Western society, just as the Amish are also citizens
of a democracy and Canadian Francophones are members of an
English-speaking nation. We are all required to balance the different
identities we possess and to bear the consequences of whatever incom-
patibilities they involve. Hence, the 1996 law may be the best accom-
modation Christian Scientists can expect. At the same time, we
should recognize that there are different ways of being members of a
technologically advanced Western society and that the religious way
that Christian Scientists adopt may not be an unimportant one. In
fact, in taking it seriously, non-believers might deepen their own

See Janna C. Merrick, ˜˜Spiritual Healing, Sick Kids and the Law: Inequities in the
American Healthcare System,™™ American Journal of Law & Medicine, 29, 2003,
pp. 269“299.
H E R M E N E U T I C S A N D T H E P O L I T I C S O F I D E N T I T Y 243

thinking about what human life is and they might use religious
views to work out their own views on a number of issues including
physician-assisted suicide, artificial means for extending life, and so
on. Intervening to save the gravely ill children of Christian Scientist
parents does not mean that we cannot respect and even learn from
their perspective on who their children are.
The same holds for the possible insights of frameworks and
contexts of which other identities are a part. There are also different
ways of being members of a democracy and we can try to learn from
those who understand the identity differently than we do.28 We need
not tolerate identities that encourage violence as part of who they are,
if for no other reason that identities that require violence cut short the
possibility of learning from alternative understandings. Still, the idea
of alternatives in understanding is as important to our thinking about
our identities and our lives as it is to our thinking about our texts.
Recognizing the multiplicity of ways of understanding who and what
we are opens us to multiple allegiances and tells against our encasing
ourselves in one identity, no matter how important that identity is to
us or to the politics of difference. In addition, our multiple identities,
allegiances, and differences allow us to try accommodation in public
policy and to refuse to see it as simply the product of defeat.
To be sure accommodation smacks of appeasement. We cer-
tainly should think more than once about appeasing certain sorts of
identities including dogmatic or fundamentalist ones, neo-Nazi ones,
or ones for which violence is a given. But we cannot start our political
thinking from the ground up, deciding in advance which identities we
want the world to contain. Rather, we already have identities and we
are already parts of different practices and institutions. In this world
into which we are thrown, the virtues of recognizing the multiple
ways we can understand who and what we are reflect democratic
virtues. They allow us to acknowledge the equal status of our different

See Georgia Warnke, Legitimate Differences: Interpretation in the Abortion
Controversy and other Public Debates (Berkeley, CA: University of California Press,
244 A F T E R I D E N T I T Y

identities and to be sensitive to the different contexts in which they
have their meaning. In addition, these virtues allow us to listen and
learn from identities we do not possess. Governments and laws may
not always be able to accommodate all the understandings that issue
from the perspective of different identities. Yet, if we refuse to
entrench ourselves in only one of our identities and if we take seri-
ously their interpretive status, we can at least listen to others. In the
end, this point may be the one Butler is making in asking whether we
have ˜˜ever yet known the human.™™29 I would say that we have, but also
that there is always more to know.

Judith Butler, ˜˜The Question of Social Transformation,™™ in Judith Butler, Undoing
Gender (New York, Routledge, 2004), p. 222.

The claim I have tried to make is that identities are parts of contexts
and make sense only within the contexts of which they are a part. Just
as the question of who Elinor Dashwood is makes no sense outside of
the context of Sense and Sensibility, the question of who someone
is or whether someone is black or white, male or female, Amish
or student makes no sense unless we know with regard to what.
Moreover, depending upon how I understand the whole of the text of
which Elinor Dashwood is a part, I will understand who she is differ-
ently. If I place the novel in the context of onanism, I may understand
her as an incestuous lover. If I understand her in the context of democ-
racy I may understand her as a model of independence. Likewise, if
I understand an ill child in the context of Western medicine, I will
understand him or her as a medical patient. If I understand him or her
in the context of Christian Science, I will understand him or her as an
alienated soul. In concluding this book, I want to expand on two
remaining issues. First, if one of the points of the book is to emphasize
the different ways both identities and the contexts of which they are a
part can be interpreted, why accept my interpretations of such insti-
tutions as marriage, education, and the military? Second, if one of the
points of the book is to support liberal goals of non-discrimination,
comparisons between literary interpretation and identity might
seem to be total overkill. For what actually is the difference between
the traditional liberal thesis that a person™s race, sex, and gender are
irrelevant in public life and my claim that race, sex, and gender are
unintelligible except within limited contexts? Put otherwise, what is
the difference between restricting all identities to the contexts in


. 7
( 8)