Stanford Law Review

Sexual Orientation and the Politics of Biology: A Critique of the Argument from Immutability Author(s): Janet E. Halley Source: Stanford Law Review, Vol. 46, No. 3 (Feb., 1994), pp. 503-568 Published by: Stanford Law Review Stable URL: http://www.jstor.org/stable/1229101 . Accessed: 09/03/2011 01:19
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=slr. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Stanford Law Review is collaborating with JSTOR to digitize, preserve and extend access to Stanford Law Review.

http://www.jstor.org

Sexual Orientationand the Politics of Biology: A Critiqueof the Argument from Immutability
JanetE. Halley*
Threerecentscientificreportsthatpurportto show a biological basisfor have changedtheface of pro-gayequalprotectionlitigationby homosexuality more attractive. ProfessorJanet E. makingthe argument from immutability Halley critiquesthese studies and their receptionin legal culture. Because is for immutability not a requirement successfulpro-gay litigation,moreover, Professor Halley contendsthat pro-gay litigatorswho invoke the argument do from immutability so not onlyat theiroption,butat the riskof misrepresenting and dividingthe community hope to represent.She argues thatprothey gay legal argumentshouldfocus insteadon commongroundthat adequately representsthe self-conceptions bothpro-gay essentialistsand pro-gay conof structivists.And she suggestsjust such a common ground moreeffectively for articulating pro-gay equalprotectionarguments.
INTRODUCTION .................................................
I. II. THE ARGUMENT FROM IMMUTABILITY ......................... THE TOPOGRAPHY OF DISAGREEMENT .....5.... ...........

...

504
507 516

A. Choice .................................................. B. Genes ................................................... C. Autobiographies ........................................ D. Litigation ............................................... III. THESTUDIES ................................................ A. Recent Experiments ...................................... B. The Brain Study ......................................... C. The Twins Study ......................................... 1. The "bisexuals" .....................................

517 521 526 528 529 531 534 538 540

* Associate Professor of Law, StanfordUniversity. J.D., Yale Law School, 1988; Ph.D. (English), U.C.L.A., 1980; B.A., Princeton University, 1974. Thanks are gratefully extended to Paul R. Billings, Paul Brest, Jon W. Davidson, Kay Diaz, MarcusW. Feldman,ThomasC. Grey, Ruth Hubbard, Vicki Laden, Mark Kelman, Andrew Koppelman,Mary Newcomb, Daniel R. Ortiz, Judge RichardA. Posner, MargaretJane Radin, William H. Simon, EdwardStein, David A. Strauss,and CatherineWells for makingthe constructionof this argumentinto an extendedcross-disciplinary conversation;to the San Francisco Bay Area Group with Interestsin Genetics and Society, the New York University Gay and Lesbian Symposium, and the Feminist Legal Theory Seminarof the Universityof Chicago Law School for lively discussions; to Andy Eisenberg,Paul Lomio, and Iris Wildmanfor bibliographicalassistance; to Lisa Hayden and Suzanne Woods for clear-headedresearchassistance;and to the Dorothy Redwine Estate for researchsupport.

503

504

STANFORD LAWREVIEW 2. The nonresponsive brothers........................... 3. One autobiography ...................................

[Vol. 46:503 542 544

..... 546

IV. COMMON GROUND .....................................

A. Comingto Termswith Essentialism and Constructivism .... 1. Essentialisms ......................................... 2. Constructivisms ...................................... 3. Impureidentity....................................... B. Finding Common Ground ................................. 1. The lay of the land ................................... 2. Choosinga place to stand ............................ C. Arguingfor EqualProtectionon Common Ground........
INTRODUCTION

547 547 550 553 556 556 560 563

567

CONCLUSION .......................................................

The claim that sexual orientation biologically determined become is has that lesbiansand gay men comprisea increasinglysalient in legal arguments minoritypopulationwarranting meaningfulconstitutional protection.According to this argument, men andlesbiansconstitutea suspectclass underthe gay that Equal ProtectionClause in part because the characteristic differentiates them, and that constitutes the basis of discriminationagainst them, is immutable. A series of recent scientific reports,1 and the way that legal culturehas received them, have made biological causationtheories far more plausible. These studies have stimulated media and activist speculationthat,confronted with scientific proof of the immutability homosexuality, of judges will be forced to deem gay men and lesbiansa suspectclass.2
1. The most important studies are J. Michael Bailey & RichardC. Pillard,A GeneticStudyof Male Sexual Orientation,48 ARCHIVES PSYCHIATRY (1991); Dean H. Hamer,Stella Hu, VictoriaL. GEN. 1089 A Magnuson,Nan Hu & Angela M.L. Pattatucci, LinkageBetweenDNA Markerson the X Chromosome and Male Sexual Orientation,261 SCIENCE (1993); Simon LeVay, A Difference in Hypothalamic 321 StructureBetween Heterosexualand HomosexualMen, 253 SCIENCE (1991). For a discussion of 1034 related reports, see notes 102-105 infra. 2. RichardGreen,introducedon Ted Koppel's Nightline as an attorneyand psychiatrist,gave both legal and medical legitimacy to the claim thatbiological causationis the key to equal rights for gays and lesbians: Well, legally, I think, [the LeVay study] could make a very big difference. In American constitutionallaw, groups that are prejudicedagainst or stigmatizedare given special protection by the courts if the featurefor which they are discriminated what's called immutableor is innate or essentially unchangeable.... So if we can find, the scientists can find that a specific partof the brainis primarilyresponsiblefor sexual orientation,then the stigmatizationand the legal discriminationagainst gays and lesbians in this country should fall. Nightline (ABC television broadcast,Aug. 30, 1991) (available in LEXIS, Nexis library, ABCNEW File). Green has more recentlybeen quotedin a similarlysweeping statement. Natalie Angier, Studyon Sexual Orientation,N.Y. TIMES, July 18, 1993, ? 1, at 24, 24 ("'If sexual orientationwere demonstrated to be essentially inborn,' [Green]said, 'most laws that discriminateagainst gays and lesbians, including sodomy laws, housing and employment discriminationlaws, all would fall.'"). Similar statementshave appearedin majorprintmedia. E.g., Natalie Angier, The Biology of What It Means to be Gay, N.Y. TIMES, Sept. 1, 1991, ? 4, at 1, 4 ("If homosexualitywere viewed legally as a biological phenomenon,ratherthan a fuzzier matterof 'choice' or 'preference,'then gay people could no more rightfully be kept out of the military, a housing complex or a teaching job than could, say, blacks.");David Gelman, Donna Foote, Todd Barrett& MaryTalbot, Born or Bred?, NEWSWEEK, Feb.

February1994]

IMMUTABILITY CRITIQUING

505

At the same time, remarkable changes in identity politics over the past decade, most notablythe emergenceof queer identityand of an unrepentant movementof self-described bisexuals,have complicated and lesbiancomgay munities.3New voices areheard,offeringa sustained, attack community-based communities shouldendorsethe identitiesthrough on the idea thatsubordinated which superordinated groupssuppressthem. Manygay men, lesbians,bisexua andqueersrejectthe view thatthey constitute minority als, distinguished a by naturalidentity. stable, Two new theoreticaldevelopmentssupportthis attack. The postmodern of critiqueof liberalexplanations the self posits thatculture,not humannature, humanstheirsexualorientations.4 a distinctbutrelatedcriticalsetting, In gives of science arguethatscience is partof culture,not its opposite.5 culturalcritics
24, 1992, at 46, 48 (Proof that homosexualityis biologically caused "couldgain [the 'gay community'] the civil-rights protectionsaccordedany 'natural'minority,in which the legal linchpinis the question of an 'immutable' characteristic."). The scientists themselves have predictedthat their findings will change the equal protectionlandscape. Simon LeVay, authorof the brainstudy describedat text accompanyingnotes 117-136 infra has speculated as follows: "Is homosexualityimmutableor a chosen lifestyle? The last time the SupremeCourtruled on this matter,it was arguedthat it was biological, but nobody believed it. This work may show that sexual orientationis genetically determinedlike skin color, and may thereforehave implications for the civil rights of gays and lesbians." Jamie Talan, Study Shows HomosexualityIs Innate: Gay Scientist Is a Hero and a Villain, NEWSDAY, Dec. 9, 1991, at 41, 41 (quoting LeVay); see also Sharon Kingman,Nature, Not Nurture?, THEINOct. DEPENDENT, 4, 1992, at 56 ("'In the United States there is a law that protects people who have immutablecharacteristics,such as race, from discrimination. So if homosexualitywere proved also to be an immutablecharacteristic, then the law would have to be changed.'") (quotingLauraS. Allen, who coauthoreda brain study similar to LeVay's; see note 103 infra). 3. On bisexuals, see Bi ANY OTHER NAME: BISEXUAL PEOPLE SPEAKOUT(Loraine Hutchins & Lani Kaahumanu TO BISEXUALITY & FEMINISM eds., 1991); CLOSER HOME: (ElizabethReba Weise ed., 1992); Bisexualities: Theoryand Research, 11:1/2 J. HOMOSEXUALITY (Fritz Klein & Timothy J. Wolf eds., 1985) (special issue). On queer identity and politics, see FEAROFA QUEERPLANET:QUEERPOLITICS ANDSOCIAL THEORY FEAROFA QUEERPLANET];Lauren (Michael Warer ed., 1993) [hereinafter Berlant & Elizabeth Freeman,Queer Nationality, in FEAR OFA QUEERPLANET,supra, at 193; Steven Seidman, Identityand Politics in a 'Postmodern'Gay Culture:Some Historical and ConceptualNotes, in FEAR OF A QUEER PLANET,supra, at 105; Michael Warer, Introduction,in FEAR OF A QUEER REV. 11 (1992) PLANET,supra, at vii. See also Lisa Duggan, MakingIt Perfectly Queer, 22 SOCIALIST (tracinghistory and politics of queer movement);IdentityCrisis: Queer Politics in the Age of Possibilities, VILLAGE VOICE, June 30, 1992, at 27 (roundtablediscussion of queer identity with Holly Hughes, Gregg Bordowitz, MarcellusBlount, Eve Kosofsky Sedgwick, Alisa Soloman, and Jeff Nunokawa);Eve 1, Kosofsky Sedgwick, Queer and Now, in TENDENCIES 5-9 (Eve Kosofsky Sedgwick ed., 1993).
4. See, e.g., 1 MICHEL FOUCAULT, THE HISTORY OF SEXUALITY: AN INTRODUCTION 42-49 (Robert E. IDENTITY\DIFFERENCE: DEMOCRATIC NEGOTIAHurley trans., 1978); see also WILLIAM CONNOLLY, TIONS OF POLITICAL PARADOX 73-78 (1991); DAVID M. HALPERIN, ONE HUNDRED YEARS OF HOMOSEXUALITY AND OTHER ESSAYS ON GREEK LOVE 41-53 (1990). 5. See, e.g., LUDWIK FLECK, GENESIS AND DEVELOPMENT OF A SCIENTIFIC FACT (Thaddeus J. Trenn & Robert K. Merton eds. & Fred Bradley & ThaddeusJ. Trenn trans., 1979) (1935); DONNA J. HARAWAY, SIMIANS, CYBORGS, AND WOMEN: THE REINVENTION OF NATURE 7-20 (1991); KARIN D.
KNORR-CETINA, THE MANUFACTURE OF KNOWLEDGE: AN ESSAY ON THE CONSTRUCTIVIST AND CONTEXTUAL NATURE OF SCIENCE (1981); BRUNO LATOUR & STEVE WOOLGAR, LABORATORY LIFE: THE CONSTRUCTION OF SCIENTIFIC FACTS 43-90 (1979); 4 THE SOCIAL PROCESS OF SCIENTIFIC INVESTIGATION

(Karen D. Knorr,Roger Krohn & RichardWhitley eds., 1980); KathrynPyne Addelson, The Man of REALITY:FEMINIST PERSPECTIVES EPISTEMOLOGY, ON METAProfessional Wisdom, in DISCOVERING AND OF PHYSICS, METHODOLOGY, PHILOSOPHY SCIENCE (Sandra Harding & Merrill B. Hintikka eds., 1983); David Bloor, Durkheimand Mauss Revisited: Classificationand the Sociology of Knowledge, 13 STUD.HIST.& PHIL. SCI.267 (1982).

506

STANFORD LAWREVIEW

[Vol. 46:503

Together,these political and intellectualmovementshave producedpro-gay constituencies eager to deny the claim that homosexualityis biologically is caused or to assertthatthe biologicalstatusof sexual orientation an answer to the wrong question. The result is the celebratedessentialism/constructivism debate-or, more to This articleattempts get beyondthatimpasseby going accurately, impasse. shouldnot focus on positive aroundit. It arguesthat pro-gaylegal argument claims thathomosexuor claims of biologicalcausation, on pureconstructivist but shouldrepairto a commonmiddle is a historically artifact, contingent ality no ground. On that middle ground,sexual orientation, matterwhat causes it, the and social andpoliticalmeaningthrough material symbolicactiviacquires the ties of living people. This is the arenaof representation, arenain whichwe to signify to one anotherwho we are, negotiatethe normsattaching that, and hierarchy. arrangeand rearrange power along the sexual orientation and Part I examines the doctrinal, jurisprudential, strategichistory of the in fromimmutability generalandits use in cases challenging discrimargument inationagainstgay men, lesbians,andbisexuals. It concludesthatimmutability is not a requirement suspectclass statusand is unlikelyto become one, so for from immutability so at do that pro-gay litigatorswho invoke the argument from immutability a their option. Part II describesthe pro-gayargument as form of pro-gayessentialism,and notes thatpro-gayconstructivists vigorously of amongpro-gayadvooppose both. It mapsa topography this disagreement to in cates, and concludesthatwe shouldnot attempt resolvethis disagreement litigation. PartIII turnsto the scientific studiesthat have been cited in some orientation immutable. is gay-rightscases as conclusiveproofthathomosexual A close readingof two of those studiesargues,first, that they have been absorbedinto the broader cultureas scientificconfirmations homosexualorithat entationshouldbe understood essentialist as rather thanconstructivist, while in fact the studiesincorporate essentialistmodelsof sexualorientation untested as hypotheses;and second,thatthe more constructivist aspectsof sexual orientation identityexcludedfromthe studiesare nevertheless to important the social subordination gay men, lesbians,and bisexuals. Finally,PartIV seeks out of commongroundfromwhichto buildlegal arguments adequately that represent the self-conceptions pro-gayessentialists pro-gayconstructivists. proof and It ceeds by disaggregating essentialismand constructivism, the signifimapping cant territoryof their overlap in sexuality studies, and locating the optimal places in that conceptualzone for articulating equal protection arguments. This articlearguesthatpro-gaylegal arguments from biologicalcausation shouldbe abandoned.Instead,pro-gayessentialistsand constructivists should thatemphasizethe politicaldynamicsthatinevitablyatdesign legal strategies tend sexual orientation identity-no matterhow it is caused.

February 1994] I.

IMMUTABILITY CRITIQUING
THE ARGUMENT FROM IMMUTABILITY

507

before 1986, when the Gay rightsadvocateswritingaboutequalprotection Courtissued its banefuldecision in Bowers v. Hardwick,6 echoed a Supreme is refrain: Since homosexuality immutable, qualifiesas a suspect it reassuring classification,or at least meets one of severalcriteriafor suspectclass status underequalprotection on analysis. Most often this argument depended an emclaim that sexual orientationis either hardwired into us at birth or pirical brandedupon us so soon thereafter it cannotbe altered.7 that These argumentsacceptedan invitationapparently issued by a Supreme Courtpluralityin Frontierov. Richardson,8 which concludedthat discrimination based on sex warranted strict scrutiny.9The pluralityopinion stated a number discretereasonsfor its decisionthatsex discrimination constituof was tionally sensitive: "a long and unfortunate historyof sex discrimination" perpetuated through "stereotypeddistinctions between the sexes"; the "high ... visibilityof the sex characteristic" exposingwomento "pervasive discrimination";and the fact that "sex, like race and nationalorigin, is an immutable characteristic determined solely by the accidentof birth."10 Two featuresof the Court'ssubsequent treatment immutability of were acforeshadowed Frontiero. First, immutability not a requirement in is curately but a factor.11 Second,thatfactoris not immutability alone but immutabilityThe carelessreader Frontiero of well supposethatthe plurality had plus. might its analysis of immutability when it observedthat discrimination completed againstwomenbasedon theirsex "wouldseem to violate 'the basic conceptof our system that legal burdensshouldbear some relationship individualreto But thereis more. The Courtwent on to note thatmany imsponsibility.'"12 mutable characteristics-its examples were intelligence and physical decisionsthatare widely regarded disability-form the basis of discriminatory
6. 478 U.S. 186 (1986) (upholding state law criminalizingsodomy against due process and privacy challenges). 7. See Stacy Lynn Boyle, Note, Marital StatusClassifications:ProtectingHomosexualand Heterosexual Cohabitors, 14 HASTINGS CONST. L.Q. 111, 127-28 (1986); RichardDelgado, Fact, Norm, and Standardof Review-The Case of Homosexuality,10 U. DAYTON REV.575, 583-85 (1985); Kenneth L. L. Lasson, Civil Libertiesfor Homosexuals: The Law in Limbo, 10 U. DAYTON REV. 645, 656-57 (1985); Harris M. Miller II, Note, An Argument for the Application of Equal Protection Heightened Scrutinyto Classifications Based on Homosexuality,57 S. CAL.L. REV.797, 817-21 (1984); see also Note, The Constitutionality Laws ForbiddingPrivate HomosexualConduct,72 MICH. REV.1613, L. of 1625-26 (1974); Note, The Legality of HomosexualMarriage, 82 YALE L.J. 573, 576 (1973). 8. 411 U.S. 677 (1973) (pluralityopinion). 9. Id. at 688. The concurringJustices refused to apply strict scrutiny,id. at 691, and the determination that intermediatescrutiny should apply to classifications based on sex was not reached for another three years. Craig v. Boren, 429 U.S. 190 (1976). 10. Frontiero, 411 U.S. at 684-86. A fourth factor that the pluralityconsidered importantwas Congress' recent amendmentto Title VII of the Civil Rights Act of 1964 and the EqualPay Act of 1963, and its approvalof the Equal Rights Amendment,which togetherevinced a "conclu[sion]that classifications based upon sex are inherentlyinvidious." Id. at 687. 11. See, e.g., Bowen v. Gilliard, 483 U.S. 587, 602-03 (1987) (holding that relatives are not a suspect class); Lyng v. Castillo, 477 U.S. 635, 638 (1986) (same). I am unawareof any case in which the Supreme Court has gone beyond this moderatedemphasis on immutability. 12. Frontiero, 411 U.S. at 686 (quotingWeber v. Aetna Casualty& Sur. Co., 406 U.S. 164, 175 (1972)).

508

LAWREVIEW STANFORD

[Vol. 46:503

The as unproblematic.13 Court'sexamplesare a little dubious-one wonders and and whetherintelligenceis unequivocally certainlyimmutable, discrimination based on physical disabilityexemplifies not the class of unproblematic but discriminations the ones thatseem to requiremoreprobinganalysisthana Whatmakes characteristic. that meredetermination they turnon an immutable and basedon physicaldisabilityacceptable, othersunacsome discriminations to purposeat hand. Nearsightedness ceptable,is theirrelatedness the particular fighterpilots butnot potentiallaw profesmay legitimatelydisqualifypotential sors.14 Recognizingthese problemsat the level of doctrine,the Frontieropluthat heightenedscrutinywas needed in cases involving sex rality determined in because"thesex characteristic," additionto being immutable, discrimination to bearsno relationto abilityto performor contribute society."15 "frequently is Frontierothus expressesa conclusionthat,when a characteristic bothimmubasedon at to tableand unrelated the legitimate purposes hand,discriminations it may suggest unfairness.16 John HartEly has arguedthat this "relevance" criterion,once invokedto as a test for unacceptable cure the defects of immutability inequality, promptly factorwhole: swallows the immutability
based on physicaldisabilityand intelligenceare typicallyac[C]lassifications who ceptedas legitimate,even by judges andcommentators assertthatimmutability is relevant. The explanation,when one is given, is that those is characteristics (unlikethe one the commentator tryingto rendersuspect)are At thatpointthere'snot muchleft of the often relevantto legitimate purposes. theory,is there?17 immutability

immuWell, theremightbe, if thereis any reasonto supposethatan unrelated an unrelated basedon is table characteristic moreinvidiousthana classification mutableone. The boilerplate responseto this queryis to say (as the plurality did in Frontiero)that it is a "basicconceptof our system that legal burdens But to shouldbearsome relationship individual responsibility."18 thatresponse becauseit assumesexactlywhatis to be decided: thatthe state is unsatisfying
13. Id. at 688. the 14. This is not to underestimate difficulty of determiningrelatednessin the context of physical themselves mutable, may create relatedness where none disability, where pre-existing arrangements, need exist. The need to use a wheelchairis relatedto any job in a building with stairs and no rampsor elevators. See MarthaT. McCluskey,Note, RethinkingEqualityand Difference:Disability DiscriminaL.J. 863, 866-68 (1988); Rosalie K. Murphy,Note, Reasonable 97 tion in Public Transportation, YALE Accommodationand EmployerDiscriminationUnder Title I of the Americanswith Disabilities Act, 64 S. CAL.L. REV. 1607, 1613-14 (1991). 15. Frontiero,411 U.S. at 686. There are plenty of careless misreadersof Frontierowho construe it to state a freestandingimmutabilityfactor uninflectedby relatedness. See, e.g., Moss v. Clark, 886 F.2d 686, 690 (4th Cir. 1989) (holding that prisonersdo not constitutea suspect classification because the status of incarcerationis neither immutablenor an indicatorof invidiousness) (citing Frontiero on immutabilitywithout reference to relatedness). 16. For a repriseof the stipulationthat immutabilityshould triggersuspicion only when conjoined with irrelevance, see Mathews v. Lucas, 427 U.S. 495, 505-06 (1976) (conceding the immutabilityof illegitimate status, but noting that distinctionsbased on legitimacy may be rationalin some contexts). AND 150 HARTELY, DEMOCRACY DISTRUST (1980) (emphasis in original) (footnote 17. JOHN omitted). 18. Frontiero, 411 U.S. at 686 (quoting Weber v. Aetna Casualty& Sur. Co., 406 U.S. 164, 175 (1972)).

February1994]

IMMUTABILITY CRITIQUING

509

to actor has a rightfulpower to make it someone's responsibility conformto to state wishes that are (ex hypothesi)totally unrelated the job at hand. The immutabilityargumentthus rests on an unexpressedpremisethat the Equal well with a state-sponsored culturalconformity ProtectionClause harmonizes and assimilationism.19 Thereare otherproblemswith the immutability as argument well. Assumto Bruce Ackermanaring a CaroleneProductsapproach equal protection,20 gues that "anonymousand diffuse" groups are often more vulnerableto than "discreteand insular" invidious exclusion from politicaldecisionmaking and ones.21 The characteristics defineanonymous diffusegroupsare often that acutely mutable,especially when they can be hidden: Then they can be professed by and ascribedto a differentset of people every year, month,day, or even hour. Thatcertainly holds truefor gay men, lesbians,andbisexuals: Anonymityand diffusenessproduced the closet are our chief organizingchalby lenges, and they emergefrom elementsof our identitythatare often radically mutable-private andpublicidentity.22It's notjust thatthe discrete-and-insumodel, with its immutability lar-minority corollary,is a bad fit with the political realitiesof gay, lesbian,bisexual,and queerlife. The problemsgay men, lesbians,andbisexualshave entering fully into pluralist politicstendto confirm that,at least as it functionsin process-based analysis,the immutability theoryis simply wrong. The Courttentativelysuggestedin Frontierothat immutability a factor is that intensifiesthe invidiousness government-imposed of burdensunrelated to the job at hand. Academicassessments the theoryof suspectclassifications of have not encouraged Courtto give any further the emphasisto this factor. In19. The regulationof appearanceat work offers just one example of the way in which the argument from immutabilityexpressly facilitatesmonocultural impulses. See PeterBrandonBayer, Mutable Characteristics and the Definition of Discrimination Under Title VII, 20 U.C. DAVis L. REV. 769 (1987); PauletteM. Caldwell,A Hair Piece: Perspectiveson the Intersectionof Race and Gender, 1991 L.J. 365; Karl E. Klare, Power/Dressing:Regulationof EmployeeAppearance,26 NEW ENG. L. DUKE REV.1395 (1992); see also Cassista v. CommunityFoods, Inc., 5 Cal. 4th 1050, 1065, 856 P.2d 1143, 1153, 22 Cal. Rptr. 2d 287, 297 (1993) (holding that a worker denied employment because she was overweight could not obtain protectionunderstate disability law unless she could show that her weight was an immutable medical condition, "a physiological, systemic disorder,"over which she had no control). 20. United States v. CaroleneProds. Co., 304 U.S. 144, 152 n.4 (1938). This celebratedfootnote proposed a new role for federal judges following the demise of substantivedue process: Instead of policing legislatures'substantivedecisions,judges would monitortheirdecisional process. In particular, judges could use the Equal ProtectionClause to ensure that "discreteand insular minorities"are not unfairly boxed out of political decisionmaking. This process-basedapproachhas been most forcefully promotedby John Hart Ely. See ELY,supra note 17. But it has also been criticized for resting on an unstabledistinctionbetween substanceand process. See, e.g., C. Edwin Baker,Neutrality,Process, and Rationality:Flawed Interpretations Equal Protection, 58 TEX.L. REV.1029, 1058-61 (1980); Paul of ST. Brest, The Substanceof Process, 42 OHIO L.J. 131 (1981); Lea Brilmayer,Carolene, Conflicts,and the Fate of the "Inside-Outsider,"134 U. PA. L. REV. 1291 (1986); Lewis F. Powell, Jr., Carolene ProductsRevisited, 82 COLUM. REV.1087, 1090-91 (1982); LaurenceH. Tribe, The Puzzling PersisL. tence of Process-Based ConstitutionalTheories, 89 YALE L.J. 1063, 1064 (1980). 21. Bruce A. Ackerman,Beyond Carolene Products,98 HARV. REV.713, 723-24 (1985). L. 22. For a more detailedexplicationof this problem,see JanetE. Halley, The Politics of the Closet: Towards Equal Protection for Gay, Lesbian and Bisexual Identity, 36 UCLA L. REV.915, 930-63 (1989).

510

LAWREVIEW STANFORD

[Vol. 46:503

commentators deed, althoughacademic writingin the periodbetweenFrontiero the and Hardwickclearlyregarded justificationfor heightened judicialreview underthe EqualProtectionClause to be painfullyproblematic, they rarely,if as even alludedto immutability a conceptualsolution.23The Supreme ever, of Courtechoedthis lukewarm-to-cold jusreception the idea thatimmutability tifies heightenedscrutinyin its 1985 decision, Cleburnev. Clebure Living residentsof group homes Center.24Refusing to hold that mentallyretarded the or "quasi-suspect" constitutea suspect classification,25 Courtlooked ditheory. Layingout the completepasrectlyto Ely's attackon the immutability sage from Ely quoted above,26the Court emphasizedthat "those who are are immutablyso, in relevantrespects,"from mentally retarded" "different, the otherswhom the statetreateddifferently.27 Nevertheless, Courtrefusedto inferfromthis kindof differencea need for formallyheightened judicialsolicinot tude. Indeed,the CourtcitedFrontieroas precedent for relyingin whole or in part on immutability,but for inquiringinto relevance instead.28 After remainsa factor,but it is not clear thatthe Courtwill Cleburne,immutability to ever again make even an asymptoticapproach a claim that discrimination the basedon a characteristic bearercannotshed is intrinsically repellentto any
"'basic concept of our system.' 29

to since Clebure have contributed a startlingresorgiTwo developments mento of immutability-based amonggay-rightsadvocatesnotwitharguments

23. See, e.g., Baker, supra note 20, at 1090, 1094 (concluding that the Supreme Court's equal cases focused on either a "purpose[ [to] contribute[ to ] ] protectionpractice in the sex discrimination the subordinationof a sexual group"or a classification that "embodiesa negative or stereotypedjudgment about the capacities or qualities of either sex"); Brest, supra note 20, at 141-42 (concluding that laws "which treat people based on certainstereotypes,inflict a dignitaryharm,an insult, a stigma"that warrantssubstantivecondemnationas "anevil in itself," and recommendingthat equal protectiontheory stop trying to avoid such substantiveevaluation);RobertM. Cover, The Origins of Judicial Activism in L.J. 1287 (1982) (arguingthat heightened scrutinyproperlyapthe Protection of Majorities, 91 YALE plies to minorities subject to intermittentor occasional mistreatmentand especially to blacks whose disadvantagesare a productof the political process itself); Owen M. Fiss, Groupsand the Equal Protec& tion Clause, 5 PHIL. PUB.AFF.107, 148-55 (1976) (arguingthat the Equal ProtectionClause protects "specially disadvantagedgroups,"and arguingthat blacks constitutesuch a group because they form a and they social, not merely a legal group, they have been in a position of "perpetualsubordination," have been excluded from political power). None of these analyses look to immutabilityfor help in and justifying judicial interventionin political decisions. Even Fiss's distinctionbetween "natural" "artificial" classes actually distinguishesbetween groups that are socially generatedand those created by the majoritarian decision disadvantagingthem. Id. at 148, 156. 24. 473 U.S. 432 (1985). 25. Id. at 433. The plaintiffs in Clebure sought a zoning permit allowing them to establish a group home for mentally retardedresidents. Id. Though the Courtrefused heightenedscrutiny,it went on to hold that the permit denial failed rationalbasis scrutiny. Id. at 447-50. Its rationalbasis analysis makes no mention of immutability. Id. 26. Id. at 442 n.10 (quotingELY,supra note 17, at 150 (footnote omitted));see text accompanying note 17 supra. The Court also relied on Ely's position that immutabilityis not a good indicatorof process failure: "'Surely one has to feel sorry for a person disabled by something he or she can't do anythingabout, but I'm not aware of any reason to suppose that elected officials are unusuallyunlikely to share that feeling."' 473 U.S. at 442 n.10 (quoting ELY,supra note 17, at 150). 27. 473 U.S. at 442. 28. Id. at 440-41. 29. Frontierov. Richardson,411 U.S. 677, 686 (1973) (quoting Weber v. Aetna Casualty& Sur. Co., 406 U.S. 164, 175 (1972)).

February1994]

IMMUTABILITY CRITIQUING

511

Courtthinking in the academic and standingtheiranemicconditionin Supreme literature.First, equal protectionlaw aboutsexual orientation began to focus on the identityor statusof homosexuals-an issue thatthe pro-gayargument seems tailoredto illuminate.This development from immutability beganwhen that the SupremeCourtheld in Bowers v. Hardwick30 states did not violate of federal constitutional guarantees due process and privacyby criminalizing same-sex sodomy. Hardwickwas soon followed by a series of federalcourt equalprotectionreview to discrimiholdingsthatrefusedto applyheightened due nationbasedon sexualorientation becauseof Hardwick's process holding. statesto criminalize same-sex They reasonedthat,becauseHardwick permitted that sodomy, and because same-sexsodomyis the "behavior defines the class the of of homosexuals,"31 Hardwick precluded application heightenedscrutiny "Afterall,"one courtreasoned, "there hardlybe can to anti-gaydiscrimination. more palpablediscrimination againsta class thanmakingthe conductthatdefines the class criminal."32 These cases challengedgay-rightsadvocatesto convince courtsthat sodomy alone does not define the class of gay men, lesbians,and bisexuals. LitiHardwick framing cases in which gatorsset out to constrain by equalprotection had been subjectedto unfavorable treatment becauseof any sexnot plaintiffs ual conductbut becauseof theirpublicandprivateidentitiesas gay, lesbian,or bisexual.33 This strategicchoice resultedin an emphasison militarycases
30. 478 U.S. 186 (1986). 31. Padula v. Webster, 822 F.2d 97, 103 (D.C. Cir. 1987) (holding that Hardwick "forecloses [plaintiff's] efforts to gain suspect class statusfor practicinghomosexuals"because, "[i]f the Courtwas unwilling to object to state laws thatcriminalizethe behaviorthatdefines the class, it is hardlyopen to a lower court to conclude that state sponsored discriminationagainst the class is invidious"); see also High Tech Gays v. Defense Indus. Sec. ClearanceOffice, 895 F.2d 563, 571 (9th Cir.) (reasoningthat "because homosexual conduct can . . . be criminalized,homosexuals cannot constitute a suspect or quasi-suspectclass"), reh'g denied, 909 F.2d 375 (9th Cir. 1990); Ben-Shalomv. Marsh,881 F.2d 454, 464-65 (7th Cir. 1989) (same), cert. denied, 494 U.S. 1004 (1990); Woodward v. United States, 871 F.2d 1068, 1076 (Fed. Cir. 1989) (holding that homosexuals cannot constitute a suspect class both because their defining characteristic not immutableand can be criminalized),cert. denied, 494 U.S. is 1003 (1990). 32. Padula, 822 F.2d at 103. 33. In Padula, the FBI refused to employ the plaintiff because a backgroundcheck revealed that she was a "practicinghomosexual." 822 F.2d at 99. Subsequent cases challenging discrimination against military personnel are distinguishablebecause they lack record evidence even that tenuous of homosexual conduct. In Steffan v. Cheney, 780 F. Supp. 1, 4-5 (D.D.C. 1991), anotherdistrict court similarly reasonedthat because the plaintiffwas dischargedfrom the Naval Academy based on his own statementsthat he was homosexual,and because his resignationfrom the Navy and subsequentrefusalto answer discovery questions foreclosed furtherinvestigationinto his sexual conduct, "this is primarilya case about the plaintiff's status as a homosexual." After an interestingskirmish in which the district court held that the Secretaryof Defense was entitled to discovery of Steffan's sexual conduct, 733 F. Supp. 121 (D.D.C. 1990), and the Courtof Appeals for the District of ColumbiaCircuitreversed,holding that the plaintiff s sexual conduct was irrelevantto a discriminationclaim based on sexual orientation status alone, 920 F.2d 74 (D.C. Cir. 1990), the districtcourt reintroduced conduct into the case by other means. Applying rationalbasis scrutiny,it held that the plaintiff could rationallybe dismissed, in parton groundsthat he "couldone day have acted on his preferences in violation of regulationsprohibiting such conduct." 780 F. Supp. at 13 (emphasisadded). Reversing,the court of appealsheld that the military policy under which the Naval Academy forced Steffan's resignation targeted status and not conduct; that it was irrationalto presumepast or futureconduct on the basis of Steffan's profession of homosexual orientation;and that, to the extent the policy was concernedwith a "propensity" engage to in homosexual conduct, it regulatedintentand thoughtsin violation of "variouscommon law and consti-

512

LAWREVIEW STANFORD

[Vol. 46:503

the deferencewith which courts typically treat notwithstanding unfavorable The military,after all, was in the business of discharging military policy. identity,and of generating potroops based solely on their sexual-orientation tential plaintiffswith recordsdevoid of any evidence of prohibited(or even conduct.34In manyof the resulting cases, however,courtsimputed disparaged) sodomiticalconducton the basis of identityand deniedheightenedscrutiny.35 thus What PatriciaA. Cain calls "litigatingaroundHardwick"36 became, in to of part,a campaign regainsome controlover the legal definition homosexual
identity.37

A seconddevelopment since Cleburne intensified has gay-rightsadvocates' interestin immutability theories: the cultural success of geneticsas a sourceof Beforethe HumanGenomeProject knowledgeaboutwho we are as humans.38 from imbecame a householdword, gay-rightsarticlesassertedthe argument The first serious even perfunctory, recitations.39 mutabilityin dispassionate, of for reevaluation equal protection gay men and lesbiansafterHardwickemtutionalprinciplesthat guardthe sanctityof a person's thoughtsagainstgovernmentcontrol." Steffan v. Aspin, 8 F.3d 57, 64-68 (D.C. Cir. 1993). Nor is the panel's decision likely to be the last word. At the time this article went to press, the Courtof Appeals for the D.C. Circuithas vacatedthe panel discussion pending en banc review. Steffan v. Aspin, 62 U.S.L.W. 2309 (D.C. Cir. Jan. 7, 1994). In Meinhold v. United States Dep't of Defense, 808 F. Supp. 1455, 1457 (C.D. Cal. 1993), a district court held that the Departmentof Defense's policy of banninggay men and lesbians "basedmerely on status, and not conduct" violated the Equal ProtectionClause. See also High Tech Gays, 895 F.2d at 571, 573 n.9 (holding that, after Hardwick, "it would be incongruous to expand the reach of equal protection to find a fundamentalright of homosexual conduct,"and rejectingplaintiffs argumentthat Hardwick should not apply to the heightened scrutiny issue because the challenged regulations "all relate to conduct"), reh'g denied, 909 F.2d 375, 380 (9th Cir. 1990) (Canby & Norris, JJ., dissenting) (arguing that "It is an errorof massive proportionsto define the entire class of homosexuals by sod[T]he Departmentof Defense is discriminatingagainst homosexuals for what they are, not omy.... what they do."); Watkins v. United States Army, 847 F.2d 1329, 1340, 1345 (9th Cir. 1988) (holding that Army's dischargeof soldier based on his homosexualidentityviolated the EqualProtectionClause and distinguishing Hardwick because plaintiffs discharge was based solely on his sexual orientation without referenceto conduct), vacated & order affd on other grounds, 875 F.2d 699 (en banc) (holding that plaintiffs discharge was improper on equitable estoppel grounds), cert. denied, 498 U.S. 957 (1990). Analysis of the conduct/identityproblem in post-Hardwicklitigation has been offered by Patricia A. Cain, Litigatingfor Lesbianand Gay Rights:A Legal History,79 VA. L. REV.1551, 1617-27 (1993); and Nan D. Hunter, Life after Hardwick, 27 HARV. C.R.-C.L. L. REV.531, 543-53 (1992); see also Halley, supra note 22, at 948-61. 34. For an implicit admission that the Justice Departmenthas found these pure identity cases difficult to defend, see Memorandum from JanetReno, AttorneyGeneral,to Bill Clinton,President(July 19, 1993) (assessing the Clinton Administration'sproposed new policy on homosexual conduct in the armed forces) (on file with the StanfordLaw Review). 35. See note 33 supra. 36. Cain, supra note 33, at 1617. 37. For a discussion of alternateroutes to legal protectionthat do not rely on the immutability argument,see text accompanyingnotes 244-253 infra. 38. WalterGilbert,A Visionof the Grail, in THE CODE CODES: OF SCIENTIFIC SOCIAL AND ISSUES INTHE HUMAN GENOME PROJECT 84 (Daniel J. Kevles & Leroy Hood eds., 1992) [hereinafter CODE 83, OFCODES] (predictingthat the Human Genome Project will help answer the questions: "[W]hatactually specifies the human organism? What makes us human?"). For a more detailed descriptionof the "new genetics," see text accompanyingnotes 72-89 infra. 39. See note 7 supra.

February1994]

IMMUTABILITY CRITIQUING

513

with readilyapparent reluctance.40 As from immutability bracedthe argument claims pointing to a genetic role emerged, that tone stronger biological immutaproposalemphasizing changed;indeed,the first strongpost-Hardwick of came fromthe intersection law andmedicine.41Bolsteredby citations bility to recentscientificexperiments claimingto show thathumansexualorientation fromimmutability becomethe has the rests on a biologicalsubstrate, argument on which many gay-rightsadvocatespreferto contestpost-Hardwick platform courts' equationof homosexualidentitywith criminalizable sodomy.42 fromimmutability Althoughpro-gayadvocatesoften advancethe argument with enthusiasm, is clearthatmanyjudges do not find it persuasive.In three it imnonbinding opinions,federaljudges have deemedhomosexualorientation mutablein the limitedsense thatattempting reverseit throughsocial policy to is as cruel and futile as attempting changeit through to medicalor psychological therapy.43 the actualholdingson recordrejectthe argument But outright.44 Strong biological evidence, however, might alter futurejudicial outcomes.
40. Elvia Rosales Arriola, Sexual Identityand the Constitution:Homosexual Persons as a Discrete and Insular Minority, 10 WOMEN'S RTS.L. REP. 143, 154-55 (1988) (arguing that "[m]aking discrete and insular status depend on ... immutability... fails on deeper analysis,"but nevertheless concluding that "at least for constitutionalpurposes,sexual orientationcan be treatedlike race or gender, since in the vast majorityof cases it is virtually an immutabletrait"). 41. RichardGreen, The Immutability (Homo)sexual Orientation:Behavioral Science Implicaof tionsfor a Constitutional(Legal) Analysis, 16 J. PSYCHIATRY 537 (1988). Green,who holds a J.D. & L. and an M.D., listed immutabilityas a necessaryrequirement suspect classification. Id. at 538-39. In of his more recent book, he has moderatedhis statementof the relevant law. RICHARD SEXUAL GREEN, AND THELAW SCIENCE 62-63 (1992) (arguingthatjudicial acknowledgementof new suspect classifications "often"depends on the immutabilityof a defining trait). But see Nightline, supra note 2, for his statementto the press that immutabilitynot only is a necessarybut may even be a sufficient groundfor invalidating official anti-gay laws. 42. For examples from popularlegal culture, see note 2 supra. 43. The successes include a dissent, a concurrence,and a districtcourt opinion reversedon other grounds. High Tech Gays v. Defense Indus. Sec. ClearanceOffice, 909 F.2d 375, 377 (9th Cir. 1990) (Canby, J., dissenting from denial of reh'g en banc) (relying on "the overwhelmingweight of respectable authority"to conclude that "[s]exual identity is establishedat a very early age; it is not a matterof conscious or controllablechoice"); Watkins v. United States Army, 875 F.2d 699, 726 (9th Cir. 1989) (en banc) (Norris, J., concurring)(concluding that even if therapymade sexual reorientationpossible, "the possibility of such a difficult and traumaticchange does not make sexual orientation'mutable'for equal protectionpurposes"and that "allowing the governmentto penalize the failure to change such a central aspect of individualand group identitywould be abhorrent the values animatingthe Constituto tional ideal of equal protectionof the laws"), cert. denied, 498 U.S. 957 (1990); Jantz v. Muci, 759 F. Supp. 1543, 1548 (D. Kan. 1991) (noting that "availablescientific evidence... strongly supportsthe view that sexual orientationis not easily mutable"),rev'd on other grounds, 976 F.2d 623 (10th Cir. 1992), cert. denied, 113 S. Ct. 2445 (1993). When the districtcourt's opinion in Jantz was reversedon other grounds,pro-gay advocates lost not only its finding that homosexualityis immutable,but also its for express holding that immutabilityis not a requirement strict scrutiny. Id. at 1548 (citing Supreme Courtcases that omit immutabilityfrom the list of suspect class requirements concludingthat "comand plete and absolute immutabilitysimply is not a prerequisitefor suspect classification"). For an account of therapeutic efforts to alter sexual orientation(always, as it happens,in the direction of heterosexuality),see GREEN, supra note 41, at 77-84. 44. High Tech Gays v. Defense Indus. Sec. Clearance Office, 895 F.2d 563, 573-74 (9th Cir. Woodwardv. United States, 1990) (finding that "[h]omosexualityis not an immutablecharacteristic"); 871 F.2d 1068, 1076 (Fed. Cir. 1989) (finding that "homosexualityis primarilybehavioral in nature" cert. denied, 494 U.S. 1003 compared to race and gender which "exhibit immutablecharacteristics"), (1990); see also Steffan v. Cheney, 780 F. Supp. 1, 6-7 (D.D.C. 1991) (finding that "homosexualorientation is neitherconclusively mutablenor immutable"), rev'd sub nom. Steffan v. Aspin, 8 F.3d 57 (D.C. Cir. 1993), vacatedfor reh'g en banc, 62 U.S.L.W. 2309 (D.C. Cir. Jan. 7, 1994). The Ninth Circuit's

514

LAWREVIEW STANFORD

[Vol. 46:503

over biological causationis a Several courts have noted that the controversy fromimmutability,45rationale mightcut the a that reasonto rejectthe argument were to reachconsensuson the etiology otherway if the scientificcommunity of homosexual orientation. Courts are increasinglyprepared,moreover,to commit the questionsraised when pro-gaylitigatorsrely on recent scientific reportsto the jury for factfinding. Not discouragedby this lacklustertrackrecord,gay-rightsplaintiffshave from immutability citing the new scientific begun to bolsterthe argument by In at least four recent cases challengingdiscrimination reports. against gay have cited the new scientific men, lesbians,andbisexuals,plaintiffs'attorneys that homosexualorientation a suspect is findings to supporttheir arguments classificationbecauseit is immutable.46
In Dahl v. Secretary of the United States Navy,47 challenging a sailor's

dischargefromthe Navy, in Steffanv. Aspin,48 challengingthe forcedresignation of a midshipmanfrom the United States Naval Academy, in Baehr v. in Lewin,49 which plaintiffshave alreadywon heightenedscrutinyof Hawaii's refusal to recognize same-sex marriages, and in Evans v. Romer,50 which in plaintiffsought and obtainedinjunctiverelief againstColorado'sAmendment have failed. The plaintiffin Dahl cited the studies of Two, these arguments Simon LeVay and of Bailey and Pillardto supporta claim that "'it has now been conclusivelyand authoritatively established sexualorientation biothat is logical, genetic and innate."'5' Plaintiffs offer to provethis claim at trialwas declined only because the courtdenied heightenedscrutinyon other grounds and proceededto grantthe plaintiff summary judgmentunderrationalbasis review.52 Lawyers asked for summaryjudgmentin Steffan in part on the
decision in High Tech Gays is particularlytroublingbecause the panel majority,though it described immutabilityas a factor, applied it as a requirement. 895 F.2d at 573. 45. See, e.g., Steffan,780 F. Supp. at 6 n.12 (statingthat "[w]ithouta definite answer at hand, yet confident that some people exercise some choice in their own sexual orientation,the Court does not Baehr v. Lewin, No. 91-1394-05, at 5 regard homosexuality as being an immutablecharacteristic"); (Haw. Ct. App. Sept. 3, 1991) (ordergrantingDefendant'smotion forjudgmenton the pleadings) (holding that homosexuals do not constitutea suspect class, in partbecause "[t]heissue of whetherhomosexuality constitutesan immutabletraithas generatedmuch dispute in the relevantscientific community"), rev'd on other grounds, 852 P.2d 44 (Haw. 1993). 46. Steffan v. Aspin, 8 F.3d 57, 64-68, (D.C. Cir. 1993), vacatedfor reh'g en banc, 62 U.S.L.W. 2309 (D.C. Cir. Jan 7, 1994); Evans v. Romer, No. 92CV7223, 1993 WL 518586 (Colo. Dist. Ct. Dec. 14, 1993); Dahl v. Secretaryof the United States Navy, 830 F. Supp. 1319 (M.D. Fla. 1993); Baehr, 852 P.2d at 44. 47. 830 F. Supp. at 1323. 48. 8 F.3d at 57. For plaintiffs immutabilityargument,see Plaintiffs Motion for SummaryJudgment and Affidavits I and II of Richard Green, published in slightly modified form in GAYSIN THE JOSEPH VERSUS UNITED THE MILITARY: STEFFAN STATES 17-19, 56-83, 171-73 (Marc Wolinsky & 3, Kenneth Sherrill eds., 1993) [hereinafterGAYSINTHE MILITARY]. 49. 852 P.2d 44 (Haw. 1993). 50. No. 92CV7223, 1993 WL 518586 (Colo. Dist. Ct. Dec. 14, 1993). 51. Dahl, 830 F. Supp. at 1323 (quoting Plaintiffs Memorandum Supportof Motion for Sumin mary Judgment). 52. The Dahl court held that plaintiff "may have submittedsufficient evidence to create a triable issue of materialfact as to whetherhomosexualityis an 'immutable'characteristic," refused heightbut ened scrutiny on other grounds. Id. at 1324. The fact that the court neverthelessgrantedthe plaintiff summaryjudgment under a rational basis review that carefully examined the Navy's discriminatory

February1994]

IMMUTABILITY CRITIQUING

515

groundsthatrecentscientificfindingsaboutthe etiology of homosexualorienin tation offered "conclusive"evidence of its immutability; findings on this reachedwithouta hearing,the districtcourtnoted that the immutability point to claim seemedirrelevant the "great'in between'"inhabited bisexuals,and by concludedthat scientific controversyon the issue left unchallengedits own intuition that "some people have some choice in their own sexual orientation."53 The appellatepanel reviewing this decision was able (like the trial court in Dahl) to rule for plaintiffon rationalbasis groundsand pretermitted considerationof plaintiff's argumentsfor heightened scrutiny.54Similarly claims were advanced Baehr,wherethey met an in strongbiologicalcausation intricate thatindicateshow liableto backfire argument the fromimmutabilfate ity, even when bolsteredby the new scientificfindings,really is.55
practices against Dahl, id. at 1335-37, suggests that the immutabilityargument,even if provable, remains unnecessary. 53. Steffan,780 F. Supp. at 6 n.12 (emphasisin original);see Memorandum Law In Supportof of Plaintiff's Cross-Motionfor SummaryJudgmentand in Oppositionto Defendant'sMotion for Summary IN MILITARY, Judgment,in GAYS THE supra note 50, at 3, 3-39 (plaintiff'sargumentfrom immutability). 54. Steffan v. Aspin, 8 F.3d 57 (D.C. Cir. 1993), vacatedfor reh'g en banc, 62 U.S.L.W. 2309 (D.C. Cir. Jan. 7 1994). 55. The plaintiffs in Baehr v. Lewin assertedthat Hawaii's refusal to acknowledgesame-sex marriage discriminatedagainst them on the basis of sexual orientationand violated the state's equal protection clause. 852 P.2d 44 (Haw. 1993). Their brief cited Simon LeVay's experimentalreport,examined in more detail below, see text accompanyingnotes 117-120 infra, which concludes that a sex-related brain structurediffers in size among homosexualand heterosexualmen and predictseventual identification of a causal link between brain structureand homosexual orientation;but the trial court held that in homosexualityis not an immutablecharacteristic, partbecause science was in controversy. See Baehr v. Lewin, No. 91-1395-05, at 5 (Haw. Ct. App. Sept. 3, 1991) (ordergrantingDefendant's motion for judgment on the pleadings), rev'd on other grounds, 852 P.2d 44. Plaintiffs challenged this finding as erroneousin their appeal,and relied on the study of Bailey and Pillard,examinedbelow, see text accompanying notes 138-163 infra. Opening Brief for Appellantsat 14-15, 15 n.5, Baehr, 852 P.2d 44 (No. 91-1395-05). Ruling for the plaintiffs, a pluralityof the Hawaii Supreme Court held that the state's refusal to recognize marriageof same-sex partnersdiscriminatednot on the basis of sexual orientation but on the basis of sex and, declaringsex to be a suspect classificationin state equal protectiondoctrine, remandedfor applicationof strict scrutiny. Baehr, 852 P.2d at 64-67. The pluralityvigorously denied that immutabilityhad anythingto do with the constitutionalclaim, but failed to convince Justice Bums, who concurredonly in the result. Id. at 53 n.14. The pluralitynoted, with admirableprecision, that a same-sex marriagedoes not necessarily involve a homoeroticallyrelated pair and refused to consider plaintiffs' professed homosexualityrelevantto the decision. Id. at 51 n.11, 53 n.14. Writing for himself, however, Justice Burs concluded that, to prevail on a sex discrimination claim, same-sex couples must prove that their homosexuality was "biologically fated" and thus a naturalcomponent of their "sex." Id. at 69-70 (Bums, J., concurring). At the time the plurality decision was handed down, Justice Bums' concurrenceappearedto be necessary to the judgmentin plaintiffs favor. (Even this is unclear: Of the five justices who heardoral argumentin the case, one dissentedoutrightfrom the judgment,and another-a temporary justice whose appointmentexpired before judgment was entered-sought to join in the dissent. Id. at 48 n.*, 70 n.1. His concurrencewas treated as legally ineffective by the plurality,but as part of the tally by the dissenter.) The plaintiffs filed a motion for clarificationseeking clear guidanceon the standard they would have to meet on remand. Justice Burns and the dissentercould have then blocked any clarificationof the immutabilityquestion, and forced a remandto a trialcourt which had alreadyfound "'that homosexuality was not an immutablecharacteristic,'"Baehr, No. 91-1394-05, at 5 (ordergrantingDefendant's motion for judgment on the pleadings). By the time the court ruled on the motion for clarification, however, the temporary justice concurringin the dissent had been replacedwith a new permanent justice who voted with the pluralityto issue an orderrequiringthe trial court to adjudicatethe case under the law as stated in the pluralityopinion.

516

LAWREVIEW STANFORD

[Vol. 46:503

Even when plaintiffsin Evans v. Romerframedmore moderate claims on the basis of new scientificreportsabouthomosexual orientation its causes, and the courtmisconstrued themas strongclaimsof the sortadvanced plaintiffs by in Dahl, Steffan,andBaehr. In Evans,plaintiffsarguedthat"sexualorientation is highly resistantto change, whateverits etiology,"and when they cited the work of Dean Hamerand SimonLeVay,they addedthe provisothat"it is not for trait necessaryfor a traitto be geneticallydetermined it to be an involuntary thatis highly resistantto change."56 a full trialon the merits,however,the At courthearda muchmoreabsoluteimmutability claim-"Plaintiffs stronglyarthathomosexuality inborn"-and rejected on the basisof the testimony is it gue offered by plaintiffs'own witness,Dean Hamer.57 The scientificevidenceavailablehas not madejudges morelikely to accept fromimmutability.As the foregoinglegal analypro-gayadvocates'argument sis suggests,gay-rightsadvocateswho base theirequalprotection cases on the from immutability so at theiroption. As I arguein the next Part, do argument they do so at the cost of dividingthe communities they serve. As I arguein PartIII, they do so despitethe failureof the existing science to supporttheir is empiricalclaim that homosexualorientation immutable.It is time to think fromimmutability anyjustihas carefullyaboutwhetherthe pro-gayargument fiable partto play in pro-gaylitigation.
II. THETOPOGRAPHY DISAGREEMENT OF

I arguein this Partthatthose who wish to premiselegal rightsof gay men, shouldcease and desist, and lesbians,and bisexualson a biologicalargument should seek instead a common litigationstrategywith those in the pro-gay fromimmutability.Some distinctionsare communitywho resist the argument in orderfirst. By anti-gay, I mean to describethose who believe that homosexualityis bad or harmfuland should be punished,hidden,or restrained; pro-gay, I by refer to those who believe that homosexualityis good or value neutraland shouldbe celebrated tolerated.Forpurposes this Part,an essentialistview or of of homosexualorientation claims that it is a deep-rooted, fixed, and intrinsic featureof individuals.This essentialistview assumesthathomosexual orientation is determined natureor nurture), chosen. (I will laterredesignate not (by this position "strongessentialism.")The constructivist view of homosexual orientation claims thatit is a contingent, malleabletraitthatarisesin a socially personas she managesher world,its meanings,and her desires. The pro-gay
56. Trial Memorandum Plaintiffs' Case in Chief at 35 & n.8, Evans v. Romer,No. 92CV7223, on 1993 WL 518586 (Colo. Dist. Ct. Dec. 14, 1993). 57. Evans, 1993 WL 518586, at *11 ('The preponderance credibleevidence suggests that there of is a biologic or genetic 'component'of sexual orientation,but even Dr. Hamer,the witness who testified that he is 99.5% sure there is some genetic influence in forming sexual orientation,admits that sexual orientationis not completely genetic. The ultimate decision on 'nature' vs 'nurture'is a decision for another forum, not this court, and the court makes no determinationon this issue."). For a critical review of plaintiffs' use of scientific testimony in the Evans v. Romertrial, see Donna Minkowitz,Trial by Science: In the Fight over Amendment Biology is Back-and Gay Allies Are ClaimingIt, VILLAGE 2, NOV.30, 1993. VOICE,

February 1994]

CRITIQUINGIMMUTABILITY

517

argument from immutability is, on these definitions, essentialist. When the pro-gay argument from immutability adds a reliance on biological causation theories, it merely locates the source of determination in nature. Neither essentialism nor constructivism is necessarily gay-affirmative. Anti-gay conservatives use both essentialist and constructivist justifications for their discriminatory policies, even as pro-gay advocates use essentialist and constructivist arguments to defeat them. Thus there are four, not two, opposed positions-pro-gay essentialism, pro-gay constructivism, anti-gay essentialism, and anti-gay constructivism-each incorporatingits causal theory into its social policy argument: * Pro-gay essentialismholds that becausehomosexualorientation fixed, is and it from discrimination. immutable, definitional, shouldbe protected * Pro-gayconstructivism holds thatall formsof sexualorientation mutaare momentof personal ble, eitheracross an individual'slife, at some important choice, or acrosshistorical periods,andthatsocial policy on sexualorientation shouldnot impedethese variations. * Anti-gayessentialismholds thathomosexualorientation fixed, immutais bad and ble, andnormatively or sick, eitherin itself or in its manifestation, that society should tailor discrimination againstgay men and lesbians to express normative of or judgments,determanifestations homosexualorientation, cure homosexualsof theirillness. * Anti-gayconstructivism either emphasizesthe mutabilityof heterosexual must be shoredup by anti-gaydisorientation, arguingthat heterosexuality or of that crimination, pointsto the mutability homosexual orientation, arguing discriminationshould be designed to convert gay men and lesbians to heterosexuality. Pro-gay activists usually limit their debate to the first two categories-progay essentialism and pro-gay constructivism. But without taking into account the anti-gay positions that these causal theories sometimes support, pro-gay analysis cannot adequately assess the relative merits of essentialism and constructivism. A. Choice

Anti-gay constructivists say that being gay, lesbian, or bisexual is a choice, and for that reason forms a proper target for a majority that thinks these ways of being are morally bad and seeks to deter people from adopting them. Former Vice President Dan Quayle became the most visible proponent of this position during the 1992 presidential campaign when he announced, "My viewpoint is that it's more of a choice than a biological situation.... I think it is a wrong choice."58
TIMES, Sept. 14, 1992, at A17. FormerVice PresidentQuayle furtherexplained his position: "It is a

58. Karen De Witt, Quayle ContendsHomosexualityIs a Matter of Choice, Not Biology, N.Y.

wrong; it is a wrong choice. I do believe in most cases it certainly is a choice." Id. Anti-gay political appealsecho Quayle's argumentwith increasingfrequency. The chief proponent of AmendmentTwo in Coloradotold the New YorkTimes: "'From up until the time I was in the eighth or ninth grade, I didn't really like girls, wasn't interestedin them.... If I had [had] a counselor who was predisposedin that way, he could easily have convinced me: "You'reone of us. We like guys."'"

518

STANFORD LAWREVIEW

[Vol. 46:503

efforts to justify discrimination Most contemporary againstgay men and lesbiansrely on formerVice President Quayle'sfactualassumption.Forexamhis that,to the ple, JudgePosnerassumesthroughout analysisof homosexuality at extentthatsociety can preventor deterhomosexuality an acceptable cost, its repressionis justified to "makepersonswho would otherwisebecome or reAs main homosexualshappier."59 it happens,he concludesthatmany,but by no means all, forms of discrimination against homosexualsfunction inefficiently.60 But other conservativeanalystsjustify capaciousregimes of discriminationon the factual assumptionthat homosexualityis to some extent E.L. Pattullo,for example,posits that at commentator chosen. Conservative and choose theirsexualorientation, argueson thatbasis least some individuals on thatdiscrimination againstgay men andlesbians,includingconstraints their "waverers" from speech, should be tailoredto deter such sexual-orientation over heterosexuality.61 choosing homosexuality of reaPro-gayessentialismoffers a refutation this anti-gayconstructivist men and lesbiansas incaIt is an exoneration describing soning. strategy, gay and for pable of resistingtheirsexualorientation thus not "responsible" it. Its of claim to fairnesstaps a deep reservoir intuitiveplausibility:thatan individbecause he or she ual should not be criminallypunishedor civilly burdened essentialismis also a bears a disfavoredcharacteristic.62 Pro-gay helplessly is practicalstrategy,claimingthatpunishinghomosexuality useless because it
Dirk Johnson, "I Don't Hate Homosexuals,"N.Y. TIMES, Feb. 14, 1993, at A24 (quotingWill Perkins). Much of the revulsion expressed by young men in the militaryservices at the idea of being seen in the shower by a gay man is premised on anti-gay constructivism. Like Perkins, these men are concerned that their own heterosexuality,far from being immutable,is a tentative accomplishment. See Kendall Thomas, Shower/Closet,20 ASSEMBLAGE 83 (1993) (describingthis fear as the "shock of recogni80, tion that would follow from the revelation that the straightmale shower and the gay male closet are neither opposing nor even abuttingstructures"). 59. RICHARD POSNER, ANDREASON SEX A. SEX 308, 303-09 (1992) [hereinafterPOSNER, AND REASON]. Posner doubts, however, that homosexual orientationcan be chosen, id. at 297, and thus constructshis efficiency argumentson a premisehe is willing to have throwninto doubt. Since writing Sex and Reason, he has taken a more agnostic position on immutabilityand developed new arguments that do not rely on it. See RichardA. Posner,Economicsand the Social Construction Homosexuality, of in RICHARD POSNER, A. OVERCOMING (forthcoming1995) ("Itis of no importanceto my economic LAW analysis" whether homosexuality is immutableor not.). 60. In Sex and Reason, Judge Posner concluded that sodomy laws should be repealed, POSNER, SEX AND REASON,supra note 59, at 311, andthat the militaryshould exclude gay men and lesbians only for conduct that is also prohibited for heterosexuals, id. at 321. But Posner cautioned that antidiscriminationprotectionfor homosexualeducatorsof young people "maybe premature," at 322, 403id. 04, and concluded that society justifiably retains the ban on same-sex marriage,id. at 311-14. In his more recent examination of marriagepolicy, Posner concludes that principles of contract, not status, should govern marriage,and that such a regime removes the barrierto the legal recognitionof same-sex relationships. See Posner, supra note 59, at 16-17. 61. E.L. Pattullo, Straight Talk About Gays, COMMENTARY, 1992, at 22-23. Dec. 62. Surveys have shown that people who think homosexualityis immutabletend also to disapprove of discriminationagainst gay men and lesbians. Jeffrey Schmalz, Poll Finds an Even Split on Mar. 5, 1993, at A14 ("Americanswho say individuals cannot Homosexuality's Cause, N.Y. TIMES, change their homosexuality-43 percentof those surveyed-are more sympatheticto the gay view on these issues [gays in the military and gay lifestyles generally] than the 44 percent who see it as a choice."); see also Kurt E. Emulf, Sune M. Innala & FrederickL. Whitam, Biological Explanation, Psychological Explanation,and Tolerance of Homosexuals:A Cross-NationalAnalysis of Beliefs and REP. Attitudes,65 PSYCHOL. 1003, 1007-09 (1989) (finding similar results among Filipino and Swedish as well as American subjects).

February 1994]

IMMUTABILITY CRITIQUING

519

from immutability cannot be deterred. The argument may be the only "high in of discrimination the repertoire popular againstanti-gay concept"argument debatetoday. from immutability, when For a numberof reasons,the pro-gayargument think. First,as advancedas a legal claim, is not the silverbulletits proponents has we have seen, this "folk"formof antidiscrimination reasoning not survived the exactionsof constitutional analysis,whichseeks a principled way of distincharacteristics we that the many discriminations based on immutable guishing or of do not find normatively legally troubling. Second, as many proponents recordsuggestsnot that fromimmutability the argument concede,the empirical is changingsomeone's sexual orientation impossible,but thatit is so wrenchdifficult as to be cruel (and thus, in some versions,excessively costly ingly when talliedwith the scarcityof effectiveconversions).63 Stakingthe immutain on this particular is problematic severalways. The ground bility argument is of or in first resortof these arguments a description personality personhood, which the "traits" associatedwith homosexualorientation "areso centralto a for to person'sidentitythatit wouldbe abhorrent government penalizea person for refusingto changethem,regardless how easy thatchangemightbe physof from the ically."64 Personhoodarguments transposethe site of immutability to the personality. As Kendall Thomas argues, such a move risks body the "perpetuat[ing] psycho-medicalconceptionof the origins and natureof
sexual orientation .. .[,] leav[ing] the door open for effective regulation sim-

a ply by substituting medicalresponseto homosexualconduct... for a legal

63. It is also possible to cast a moderatedimmutabilityargumentin terms that emphasize the agency of gay men and lesbians in adoptingand affirmingtheir sexual orientations. Lesbianphilosopher Claudia Cardpoints out that one can have an option, exercise it, and thereafterbe unable to revisit that decision and "makeit again." ClaudiaCard,Lesbianismand Choice, 23 J. HOMOSEXUALITY 39, 41-42 (1992). Card argues that sexual orientationmight be labile, mutable,and subject to autonomousdecisionmaking at one point in a person's life, but that once a choice is made certainethical, psychic, and interpersonaldynamics make changing that choice impossible, or so costly as to be catastrophic. Id. or Similarly, William Connolly's accountof "branded entrenchedcontingencies"in identity-"obdurate contingencies"-allows him to "emphasize[] how [certain identities] are both contingent formations and resistantto modificationonce consolidated." CONNOLLY, supra note 4, at 176. Pro-gayessentialists shy away from this moderatedconception of immutabilitybecause it defeats their preferredpolemical point, that gays should not be punished because they cannot be deterred. The plaintiffs in Evans v. on Romer, for example, arguedthat discrimination the basis of homosexualorientationwarrantsheightened scrutinybecause sexual orientationis an "involuntary no characteristic," matterhow it is caused. Trial Memorandum Plaintiff's Case in Chief at 32-36, Evans v. Romer 1993 WL 19678 (Colo. Dist. on Ct. Jan. 15, 1993) (No. 92CV7223), aff'd, 854 P.2d 1270 (Colo.), cert. denied, 114 S. Ct. 419 (1993). 64. Watkins v. U.S. Army, 875 F.2d 699, 726 (9th Cir. 1989) (en banc) (Norris, J., concurring); see also Bowers v. Hardwick,478 U.S. 186, 203 n.2 (1986) (Blackmun,J., dissenting) (arguing that "[h]omosexualorientationmay well form part of the very fiber of an individual's personality");Trial Memorandum Plaintiffs' Case in Chief at 36, Evans (No. 92CV7223) ("Burdening on individualson the basis of sexual orientation... is also unjust because sexual orientationis a characteristiccentral to a H. AMERICAN LAW CONSTITUTIONAL 943 (1st ed. 1978) (stating person's identity.");LAURENCE TRIBE, that same-sex sodomy is "centralto the personalitiesof those singled out by" an antisodomy statute); Tribe, supra note 20, at 1075-77 (arguingthat anti-gay legislation should be rejected on the basis of a substantiveview of "whatit means to be a person"because it "denies those subject to it a meaningful opportunityto realize their humanity");Note, The ConstitutionalStatus of Sexual Orientation:HomoL. sexuality as a Suspect Classification, 98 HARV. REV. 1285, 1304-05 (1985) (asserting that "a gay person's sexuality is fundamentalto her personal identity"because "homosexualityis a determinative feature of personhood").

520

LAWREVIEW STANFORD

[Vol. 46:503

those one."65And this modifiedformof essentialism entirelyfails to represent that deny the centrality a particularized of homosexual pro-gayconstituencies to orientation theirpsychicmakeup,whetherbecausethey identifyas bisexual, of the becausethey seek to de-emphasize genderparameters sexuality,because aboutsexuality,or becausethey experiencesexualitynot areexperimental they but as serious self-expressiveness as play, drag,and ironic self-reflexivity. do Moreover,personhood arguments not establisha rationalefor delegitidecisions to sanctionvoluntaryconduct. As philosopher Edmatingpopular ward Stein has argued, pro-gay essentialism fails to address the anti-gay whether is mutable not, is expressedthrough it or that argument homosexuality, electedbehavior, rangingfromsame-sexeroticacts to practicesof self-identification.66Explaining conductimpingeon elementsof life why rulesburdening would requirenot a psychiatric psychologicaltheory or centralto personhood for of sexualitybuta politicalone. And thejustification thesepolicies neednot rest on an assumptionthat they tend to increaseor decreasethe amountof homosexualityand heterosexuality being expressedin a society: Anti-gayesthatto do otherwisewould sentialismmight espouse them on an assumption indicateapproval the conductof an immutably of definedclass. For example, in an argument aboutgay marriage that he has since modified,JudgePosner notedthat"[t]opermitpersonsof the same sex to marryis to declare,or more by precisely to be understood many people to be declaring,that homosexual is even a noble, conditionin whichto live."67Moreover, marriage a desirable, he offeredthisjustification prohibiting for same-sexmarriage an assumption on in that heterosexual orientation those who bearit is immutable.68 are Finally, anti-gayconstructivists often willing to concede that many, if not most, existinggay men andlesbianscannotswitchtheirsexualorientation. in Instead,they tailortheiranti-gaysocial policy to deternew enrollments the class of homosexuals. For these anti-gayconstructivists, sufferingof those the who have alreadyand irrevocably madethatchoice is unimportant-andso it doesn't matter,either, whetherthat sufferingarises from discrimination gay men and lesbiansare unableto duck becausethey cannotchangetheir sexual or that orientation, fromtransformative therapies cause themanguish. Pattullo, for instance,reasonsthateven if only some children sexualorientation are "waverers,"social policy must "give [them]clear,repeatedsignals as to society's that This preference" they elect heterosexuality.69 is the wise thingto do, Pattullo argues,even at the cost of "condemn[ing] youngsters,who from earliest
65. KendallThomas, Beyond the Privacy Principle, 92 COLUM. REV.1431, 1474 (1992). Such L. therapiesare currentlysuggested in ? 302.60 of the AmericanPsychiatricAssociation's Diagnostic and Statistical Manual (DSM-III) for "GenderIdentityDisorderof Childhood,"which legitimates therapies intendedto divert effeminacy in boys. See Eve Kosofsky Sedgwick, How to Bring YourKids Up Gay, in FEAR A QUEER OF PLANET, supra note 3, at 71-72. 66. EdwardStein, The Relevance of ScientificResearchAboutSexual Orientationto Lesbian and RIGHTS THE AND MEANING SCIENCE OF OUTING, CIVIL Gay Rights, in GAYETHICS: (forthcoming1994).
67. POSNER,SEX AND REASON, supra note 59, at 312.

68. Id. (declining to "suggestthat government'spronouncing homosexual marriagea beatific state would cause heterosexualsto rethinktheir sexual preference"). 69. Pattullo, supra note 61, at 24.

February 1994]

IMMUTABILITY CRITIQUING

521

memoryknow themselvesto be gay,"70 and a fortiori at the cost of condemning those who have chosen and now cannotrevise their choice. Indeed,the underthis rationale makespublicdisplays pedagogicaldesignof discrimination but of sufferingnot merelyacceptable valuable,as the clearestsignalof majority preferencesimaginable. An argumentfrom immutabilitythat relies on the futility of, and pain causedby, psychotherapeutic effortsat conversion does not refutea program of discrimination tailoredto preventpeople who can choose to become actually homosexualfromdoing so in the firstplace. To refuteanti-gayarguments takfrom immutability this form, a legal argument needs biologicalcausation. ing If pro-gayessentialistswant to stay in the game, they will have to claim that is becauseit is biologicallydetermined-causalprohomosexuality immutable essentialismtoutcourt. ForthatreasonI will assumein the followingParts gay that pro-gayessentialismassertsbiological causation. In fact, in three of the four recentcases relying on the new biologicalfindings,that is the form that the pro-gayessentialistargument taken.71 has
B. Genes

In the last thirtyyears, genetics has undergone astonishingascendancy an the life sciences. It has brokendisciplinary limits it adoptedin orderto among distanceitself from Nazi eugenics, and now forthrightly seeks to explain not the characteristics animalsandthe physiologicalfeaturesof humans, of merely buthumanbehavioral psychological and traits.72 ridesthe Pro-gayessentialism coattailsof moderngenetics' sweepingepistemological authority. Recent developmentsin science have been invoked to supportthe claim thathomosexuality now known,or will soon be known,to be a biologically is characteristic. two reasons,this sectionwill argue,proFor caused,immutable to gay essentialistsshouldhesitateto rely on these developments supportthe from immutability.First,as scientificprofessionals have amplyand argument studinsisted,behavioral repeatedly geneticsin general,andthe homosexuality ies in particular, not support claimthathomosexual do the orientation genetiis cally caused and therefore an immutablecharacteristic. Second, pro-gay of supportfor genetic explanations sexualorientation may boomerang valiby datingthe key premisesof anti-gayeugenics.
70. Id. 71. See notes 47-54 supra and accompanyingtext. 72. The most comprehensiveaccount of the transformation genetics is DANIEL KEVLES, of J. IN THE NAME EUGENICS: OF GENETICS THE AND USES HUMAN OF HEREDITY contri(1985). Otherimportant

STEVENROSE& LEONJ. KAMIN,NOT IN OUR GENES:BIOLOGY, AND phy is R.C. LEWONTIN, IDEOLOGY, HUMAN NATURE(1984).

OUSDIAGNOSTICS: SOCIAL THE POWER BIOLOGICAL OF INFORMATION (1989); Evelyn Fox Keller, NaOF ture, Nurture,and the HumanGenomeProject, in CODE CODES, supra note 38, at 281; and Daniel J. Kevles, Vital Essences and Human Wholeness:The Social Readings of Biological Information,65 S. CAL. L. REV.255, 273-77 (1991). An account that de-emphasizesthe retreatfrom eugenics after World War II and emphasizes instead broaderideological origins in Westernpolitics, economics, and philoso-

butions include RUTH HUBBARD& ELUAHWALD, EXPLODING THEGENE MYTH: HOW GENETIC INFORMATION IS PRODUCEDAND MANIPULATED BY SCIENTISTS,PHYSICIANS,EMPLOYERS,INSURANCE COMPANIES AND LAW ENFORCERS 13-38 (1993); DOROTHY NELKIN& LAURENCE DANGERTANCREDI,

522

STANFORD LAW REVIEW

[Vol. 46:503

Part III below examines the methods and claims of the most important recent studies of homosexual orientation. Even if pro-gay essentialists do not choose to heed the arguments offered there, they should beware of exaggerating the claims that human behavioral genetics can support as a general matter. Science professionals pursuing internal criticism of the behavioral genetics programme emphasize, first, that genetically caused characteristics are not necessarily immutable in the sense that they persist unchanged over time. For example, even if male pattern baldness is entirely genetically caused, it nevertheless emerges only in adulthood and then develops. And behavioral characteristics may exhibit even more complex developmental mutabilities than physical ones.73 Nor are genetically caused characteristics necessarily immutable in the sense that they are immune from environmental influence. Confusion on this point often arises from the popular assumption that characteristics shown to be "heritable" are, for that reason, immune from environmental influence. But population and behavioral geneticists use the term "heritable"only to describe "the level of predictability of passage of a biologically interesting phenotype from parent to offspring."74Heritability confounds phenotypic variance caused by heredity with phenotypic variance caused by environment.75 Moreover, even when geneticists are able to control environmental variance (as they seek to do, for instance, in experiments involving agriculturalplants and laboratory and farm animals), any finding that a certain phenotypic characteristic is likely generated through the action of a single gene not interacting with other genes ("narrow-sense heritability") assumes the particular environmental factors in which the characteristic emerged: A statistical estimatefor the narrow-sense heritability pertains only to a particular populationstudiedundera specific environmental regime. It cannotbe or regardedas valid for a differentpopulation underdifferentenvironments. Thereis nothingthatcan be inferred fromsuch an estimateaboutthe extentto which phenotypic differences between populations are due to genetic differences.76 Inasmuch as heritability studies do not even theoretically eliminate environmental factors, they do not eliminate the possibility that phenotypic variation
73. See L.J. EAVES, H.J. EYSENCK N.G. MARTIN, & CULTURE PERSONALITY: AND AN GENES, EMPIRICAL APPROACH 161-99 (1989) (developing a model for testing the relative contributionof genes and culture to behavioralcharacteristics that emerge in human development). 74. Marcus W. Feldman, Heritability: Some TheoreticalAmbiguities, in KEYWORDS EVOLUIN TIONARY BIOLOGY 151 (Evelyn Fox Keller & ElisabethA. Lloyd eds., 1992). 151, 75. Id. at 151-52; see also M.W. Feldman& R.C. Lewontin, The HeritabilityHang-Up, 190 SCIENCE 1163, 1164 (1975) ("The narrowheritabilitydoes not provide an index of the importanceof an individual's genotype in determiningthe phenotype. It is merely an index of the amenabilityto selective breeding and, as such, is of practicaluse in the constructionof breeding programs."). 76. Feldman, supra note 74, at 155 (emphasis added). Feldman concludes by observing that "normof reaction"studies, in which genotypically identical individualsare rearedin differing environments, producediffering measuresof heritabilityfor the same characteristic:"[E]ven if the heritability in one environmentwere high, it might not be in another." Id. at 157; see also DANIELL. HARTL& G. OF ANDREW CLARK, PRINCIPLES POPULATION 472 GENETICS (2d ed. 1989) (asserting that norm of reaction studies "underscorethe fact that heritabilityis a measure defined in one environment").

February 1994]

IMMUTABILITY CRITIQUING

523

arises througha process of interactionbetweengenotypeand environment.77 Nor do they eliminatethe possibilitythatenvironmental differences resultfrom and in turncause genetic variance.78 is difficultif not impossibleto design It that the experiments measureseparately wide varietyof ways in whichenvironment may contribute an observedregularity the transmission phenoto in of to typic traitsfromparents offspringwhenthe objectof studyis Drosophilaor and humanbehavioral corn. Humancharacteristics, particularly characteristics, are even more difficultto assess and even more inhospitable effortsto exto clude environmental cultural and factors.79Thus,according the internal to criticism offered by scientificprofessionals, experimental in human any program behavioral and geneticscan offeronly limitedclaimsof geneticcausation, even those claims do not necessarilydemonstrate immutability. At the sametime, pro-gayessentialism takessome unacceptable risks when that approach borrowsthe plausibilityof moder genetics. On a purelypragmatic level, pro-gayessentialismis just not differentenoughfrom anti-gayessentialism to mount an effective resistanceto the developmentof anti-gay fails to contestthe arguments crucialto the diseugenics. Pro-gayessentialism social policies of both anti-gayessentialism anti-gayconstrucand criminatory tivism: that homosexualityis bad eitherbecause it is immoralor because it the aims at the antidisrupts social order. Pro-gayessentialism parsimoniously constructivist that at least some homosexuality chosen; but it is gay premise is neglects the second premise,that homosexuality bad or harmful. Pro-gay constructivists thatthis failureto contestthe moralandpoliticalmeaning worry assigned to homosexualityleaves pro-gayessentialismvulnerableto cooptation. An anti-gayconstructivist could convertto essentialism, the maintaining that homosexualityis bad for moral or civic order, agreeing with premise the important and pointsexplicitlymadeby pro-gayessentialism, then,without undertake an eugenic program to eliminate breaking logical stride, homosexuals.

77.

Feldman, supra note 74, at 156; see also SUSAN L.

FARBER,

IDENTICAL TWINSREARED APART:

A REANALYSIS 22-31 (1980); HARTL& CLARK, supra note 76, at 458-63. Feldmanand Lewontin further argue that experimenterscannot partitiongene-environmentinteractionunless they also isolate gene-environmentcorrelation. Feldman & Lewontin, supra note 75, at 1164. 78. Feldman & Lewontin, supra note 75, at 1163 ("Thegenetic variancedepends on the distribution of environmentsand the environmentalvariancedepends on the distributionof genotypes."). 79. Id. at 1164 (concludingthatin experimentsinvolving humansubjects,"experimental controls" that allow the experimenterto partitionpurelygenetic and purelyenvironmental contributions well as as those involving gene/cultureinteractions"are either impossible or unethical,"rendering"statisticalinference about the heritabilityof traits that are phenotypicallyplastic . . . invalid"); see also Robert Plomin, The Role of Inheritancein Behavior, 248 SCIENCE 186 (1990) (concluding that "[g]enetic 183, influence on behavior appears to involve multiple genes rather than one or two major genes, and nongenetic sources of variance are at least as importantas genetic factors"). Twins studies have emerged as the most promising means of controlling for environmentalvariance in humangenetics. For pessimistic assessmentsof this theoreticalapproach,see HARTL CLARK, & supra note 76, at 480-81; FARBER,supra note 77. For an effort to design experimentalmodels adeET quately complex to accommodatethe problemsof researchin humans,see EAVES AL., supra note 73.

524

STANFORD LAWREVIEW

[Vol. 46:503

mainstream Anti-gayeugenicsis morethana theoretical danger. Important scientistshave praisedthe HumanGenomeProjectfor its eugenic potential,80 and researchers the causes of homosexuality on repeatedlyacknowledgethe of anti-gayeugenics.81Moreover,modem genetics aspiresnot to possibility improve or purify a nationalor racial gene pool-a task that would require levels of officialparticipation-butto enableindividual currently unimaginable to preventthe birthof less-than-optimal infantsthroughprivatelyobparents tainedprenatal and selective abortion.82 And of courseselective abortesting tion of fetuses deemedto be geneticallycommitted homosexuality not the to is conceivableway in which genetic information misinformation) could (or only be used. Current or imagined genetics technologies could be used to subordinate preventthe existence of gay men and lesbiansin a varietyof or could make involuntary identification ways: genetic fingerprinting possible; could provideparentswith reproductive cloning by optionsuntainted any "gay couldcurethe "condition" genotypichomoof gene";andgeneticmodification All of these approaches couldbe used whether targeted the sexuality.83 genetic
80. For example, Daniel Koshland,editorof Science, which has publisheda numberof the leading studies claiming to identify biological causes of homosexuality,has stated that genetic diseases "areat the root of many currentsocietal problems,"and has warnedagainst "the immoralityof omission-the failure to apply a great new technology to aid the poor, the infirm, and the underprivileged."Daniel E. Koshland, Jr., Sequences and Consequences of the Human Genome, 246 SCIENCE 189, 189 (1989). When asked whether funding proposed for the Human Genome Project might be better given to the homeless, Koshlandreplied: "'What these people don't realize is that the homeless are impaired.... Indeed, no group will benefit more from the applicationof humangenetics."' Keller, supra note 72, at 282 (quoting Koshland's address to the First Human Genome Conferencein October, 1989). 81. See, e.g., David J. Jefferson,Studyingthe Biology of Sexual OrientationHas Political Fallout, WALL J., Aug. 12, 1993, at Al, A4 (quotingDr. LauraAllen and Dr. Dean Hamer). Two researchers ST. in neuroendocrinology have indicatedthatthe principalobstacles to anti-gayeugenics are practicalones, making no mention of the potential moral problems that would be involved. Lee Ellis & M. Ashley A Ames, NeurohormonalFunctioningand Sexual Orientation: Theoryof Homosexuality-HeterosexualBULL.233, 252 (1987) (urging caution in attemptsto prevent homosexuality because ity, 101 PSYCH. "[s]everal decades of intense researchmay be requiredto adequatelytest the theory [that homosexuality and other "sexual inversions"are prenatallycaused], and ... to identify precisely where and when intervention might be feasible"). Elsewhere anti-gay eugenics are actively recommended. E.g., G. Duorner,B. Schenk, B. Schmiedel & L. Ahrens, StressfulEvents in Prenatal Life of Bi- and Homosexual Men, 81 EXPERIMENTAL CLINICAL ENDOCRINOLOGY (1983) ("These findings indicate that 83, 87 preventionof war and undesiredpregnanciesmay renderpossible a partialpreventionof the development of sexual deviations."). Often scientists raise a hortatorycry that prevention should not be attempted. See Hamer et al., supra note 1, at 326 (urgingreadersto regardhomosexual orientationas a "normalvariation in human behavior,"and stating that "[w]e believe that it would be fundamentally unethical to use [informationfrom the Human Genome Project] to try to assess or alter a person's currentor future sexual orientation,either heterosexualor homosexual"). Even those who do not share the normativebelief thathomosexualityis bad may endorsepreventionprogramsfor otherreasons. See, SEX REASON, e.g., POSNER, AND supra note 59, at 308 (wonderingwhether"science, which has worked so many wonders, may someday, perhaps someday soon, discover a 'cure' for homosexuality,"and opining that such a cure would be socially inefficient if applied to adults, who would suffer pain and dislocation in abandoninga social identity,but may be cost free in the case of fetuses, infants, and even children);Gelman et al., supra note 2, at 48 ("'No parentwould choose to have a child born with any factorthat would make life difficult for him or her.'") (quotingthe programdirectorof the Federationof Parents and Friends of Lesbians and Gays). 82. KEVLES, supra note 72, at 267-68; Keller, supra note 72, at 289, 295-97. 83. On genetic fingerprinting,see DNA ONTRIAL: GENETIC IDENTIFICATIONCRIMINAL AND JUSTICE (Paul R. Billings ed., 1992); Eric Lander,DNA Fingerprinting:Science, Law and the Ultimate OF Identifier,in CODE CODES, supra note 38, at 191. On cloning and genetic modification,see KEVLES, supra note 72, at 264-68.

February 1994]

IMMUTABILITY CRITIQUING

525

traitunilaterally causedhomosexual an orientation, merelypredisposed individual to such an orientation, hadno relationship humansexualdevelopment or to at all. Pro-gayessentialismis not rhetorically materiallyprepared deal or to with the furtivecultural dispersalof anti-gayessentialism takingall, or any, of these forms.84 also embodiesideologicaldangers couldengender that Pro-gayessentialism concretepoliticalones. Geneticsdoes not merelydescribereality;it also incorporatesculturalnormsinto its premisesand diffuses into the wider cultureits own implicitnormsandepistemological commitments.SciencescholarEvelyn Fox Keller uses the example of alcoholismto illustratehow the ideological of new meaningsto political assumptions geneticsmightcontribute disturbing life. Kelleremphasizesthatthe medicalappropriation modem genetics, foof cused as it is on the conceptof genetic disease, makes genetic normalitythe fundamental basis of analysisand yet pervasivelyfails to define it:
"Take alcoholism. Alcoholismgets defined as a genetic disease and we look for genes thatcharacterize of to populations peoplewho areaddicted alcohol. We don'tlook at the genes thatcharacterize populations can drink the that withoutbecomingaddicted.... . . . [L]et's supposea durable,robustclaim for a genetic contribution to certain physiological responses that are correlatedwith alcoholism is obtained.... At thatpointit becomesnecessary ask: 'Howdid the categoryof to alcoholismget definedin the first place?'"85

Genetics thus raises the question of "how the authorityfor prescribingthe meaningof 'normal'is distributed"86-aquestionwhose answerwill emerge as scientificclaims are interpreted used in culture. One dangerof shaping and pro-gay legal strategyon genetics is that this patterndeveloped in genetic medicinewill be reiterated the contextof sexuality: Definitionsof the norin mal thatexclude homosexuality alreadybe embedded the science when will in it entersculture. Nor is the dangersimplythatheterosexuals get to definethe normalas will identicalwith themselves. Heterosexual identityis a complex,indeedunstable as is suggestedby studiesshowingthatmen affirmtheiridentities phenomenon, as heterosexualeven when they acknowledgehaving recent same-sex contacts.87 The instabilityof genetic normalityis mirrored the instabilityof in conheterosexualityas a social-representational practice. Social-descriptive
84. I do not think that the dangerof appropriation justifies abandoningscientific studies that hypothesize a biological cause of homosexuality. If I did, I could not consistently write and publish this article, which is ripe for misappropriation anti-gayconstructivistswilling to distortits centralpoints. by Nor do I think that gay men and lesbians who find themselves affirmed or reflected in narrativesof biological causation should cease to say so. I argue only that they should not use litigation to obtain official approbationfor their sexual orientationidentity. See Part II.D infra. 85. LarryCasalino, Decoding the Human Genome Project:An Interviewwith Evelyn Fox Keller, SOCIALIST REV.,Apr.-June 1991, at 111, 121-22 (quoting Keller); see also Keller, supra note 72, at 296-97. 86. Keller, supra note 72, at 299. 87. See, e.g., RobertE. Fay, CharlesF. Turner,AlbertD. Klassen & John H. Gagnon,Prevalence and Patterns of Same-GenderSexual ContactAmongMen, 243 SCIENCE 338 (1989); Janet Lever, 338, David E. Kanouse, William H. Rogers, Sally Carson & Rosanna Hertz, Behavior Patterns and Sexual

526

STANFORD LAW REVIEW

[Vol. 46:503

cepts that are this manipulable invite, and often receive, opportunistic redefinition in the political sphere. The power to define and redefine who and what the "normal"is, and who and what "heterosexuals"are, can become itself a method of doing politics to secure some aspects of social superordination. The overlap of genetic normality with heterosexual identity doubles the danger. Despite the danger of the genetics model, however, it would be imprudent for pro-gay constructivists to insist that legal strategy exclude and contradict pro-gay essentialism. As Keller warns, robust scientific claims that homosexuality is genetically caused may be made any day now: "Thefirst reactionof leftiststo the [Human] GenomeProject... [is to argue] thatbehaviorandpersonality not products nature, of nurture.I think are of but that'sa positionthatis being radically undermined the GenomeProjectand by will be impossibleto maintainover the courseof the next years. ... I don't have any doubtthat [geneticsresearchers] going to be able are to makecredibleclaimsfor multi-gene "determination" personality of traits."88 Legal strategy predicated exclusively on pro-gay constructivism would be severely injured if and when researchers show, within the standardsof proof that pertain in reputable modem science, that homosexuality as it is currently manifested in our culture is genetically determined. In that event, pro-gay constructivism could maintain its program of legal reform only by successfully attacking the standards of proof that pertain in reputable modem science. Though pro-gay constructivists are entitled to, and should continue to, mount such critiques,89 it may not be pragmatic to predicate legal strategy on their success. C. Autobiographies Tom McNaught stated an autobiographyin three sentences: "It's not a matter of choice. It's who I am.... It's genetic."90 Pro-gay essentialism makes autobiographical sense to a significant number of gay men and to many, though perhaps fewer, lesbians. But other people who experience anti-gay discrimination tell quite different stories. Some understandthemselves to have chosen the form of their desire or the ways in which it structures their lives.91 Others
Identity of Bisexual Males, 29 J. SEXRES. 141, 151-53 (1992); see also Janet E. Halley, Reasoning About Sodomy:Act and IdentityIn and After Bowers v. Hardwick,79 VA. L. REV.1721 (1993). 88. Casalino, supra note 85, at 115-16 (quoting Keller). 89. For a study of ways in which African-Americans Jews combattedthe "scientific racism" and ascendantbetween 1870 and 1920, which claimed that they were biologically inferior,see Nancy Leys the Stepan & SanderL. Gilman,Appropriating Idiomsof Science: The Rejectionof ScientificRacism, in THEBOUNDS RACE: OF PERSPECTIVESHEGEMONY RESISTANCE (Dominick LaCapraed., ON AND 72 1991). 90. Tony Rogers, 'Why' of Homosexuality, CHI. TRIB., June 2, 1993, at C2 (quoting Tom McNaught). 91. See, e.g., Card, supra note 63; Carla Golden, Diversity and Variabilityin Women's Sexual
Identities, in LESBIAN PSYCHOLOGIES: EXPLORATIONS AND CHALLENGES 19 (Boston Lesbian Psycholo-

gies Collective ed., 1987).

February1994]

IMMUTABILITY CRITIQUING

527

identityis a prodgroundthathomosexual occupythe hotly contestedhistorical uct of modernity,not natureor humannature.92 Still othersworrythatthe designations constrain the at "gay"and"lesbian" very momentof their application.Among these are self-identifiedbisexuals, who repeatedly reportthe difficultythey confrontin fittingtheirlives and experiences into the simple narrativeform propounded Tom McNaught.93 by Otherstrongcurrentsin the pro-gaymovementscritiquethe very impulse to a organizearoundgay and lesbianidentity,eitherbecausedoing so suppresses from homosexuality,94 because it or sexuality distinct and semi-autonomous obscuresthe historical,institutional, politicalprocessesthatproduceidenand are tity.95 To the extent that these self-articulations anticategorical, they are underthe rubric"queer."96 increasinglyperformed
92. THE SEXUALCOMMUNITIES: MAKING A HOMOOF See, e.g., JOHND'EMILI, SEXUALPOLITICS,
SEXUAL MINORITY IN THE UNITED STATES, 1940-1970, at 10-13 (1983); FOUCAULT, supra note 4, at 43; HALPERIN, supra note 4, at 41-53;EVE KOSOFSKY SEDGWICK, BETWEEN MEN: ENGLISH LITERATURE AND MALE HOMOSOCIAL DESIRE (1985); EVE KOSOFSKY SEDGWICK, EPISTEMOLOGY OF THE CLOSET

THE OF MODERN HOMOSEXUAL (1990) [hereinafterSEDGWICK, EPISTEMOLOGY]; MAKING THE (Kenneth Plummer ed., 1981); JEFFREY COMING IN FROM THE POLITICS BRITAIN, OUT:HOMOSEXUAL WEEKS, NINETEENTH CENTURYTOTHE PRESENT11-44 (1977) (tracing a shift in the definition of homosexual "from sin to crime," thence to "the medical model," and at last to "a way of life"). John Boswell has voiced the most articulateoppositionto this view. John Boswell, Categories,Experienceand Sexuality, in FORMS DESIRE: OF SEXUAL ORIENTATION THESOCIAL AND CONSTRUCTIONIST CONTROVERSY 133 (EdwardStein ed., 1990); John Boswell, Revolutions, Universals, and Sexual Categories, reprintedin
HIDDEN FROM HISTORY:RECLAIMING THE GAY AND LESBIANPAST 17 (Martin Duberman, Martha

A. Boone & Michael Cadden eds., 1990)). 95. See, e.g., Joan W. Scott, The Evidence of Experience, 17 CRITICAL INQUIRY773 (1991); Steven Seidman, Identityand Politics in a "Postmodern"Gay Culture,in FEAR A QUEERPLANET, OF supra note 3, at 105. 96. Sedgwick, supra note 3, at 8 (describing"one of the things that 'queer' can refer to: the open mesh of possibilities, gaps, overlaps,dissonancesand resonances,lapses and excesses of meaning when the constituentelements of anyone's gender, of anyone's sexuality aren't made (or can't be made) to

PLEASURE AND DANGER:EXPLORING FEMALESEXUALITY 267 (Carole S. Vance ed., 1984); see also Judith Butler, Imitation and Gender Insubordination, in INSIDE/OUT: LESBIAN GAY THEORIES THEORIES, 13 (Diana Fuss ed., 1991); Andrew Parker, Sensitive New Age Guys, LESBIAN& GAY STUD. NEWSL., Mar. 1993, at 31 (reviewing ENGENDERING MEN: THE QUESTION MALE FEMINIST OF CRITICISM (Joseph

Bauml Vicinus & George Chauncey,Jr. eds., 1989). 93. For example, CLOSER HOME: BISEXUALITY & FEMINISM, TO supra note 3, collects a numberof narratives that vary widely in their responseto this problem. Perself-descriptiveand autobiographical haps the most salient difference among these essays is their implicit response to the question whethera unitary biography is possible for people whose erotic lives are invested in men and women. Some contributorsaspire to lifestories of "wholeness." See, e.g., MargaretMihee Choe, Our Selves, Growing TO Whole, in CLOSER HOME, supra, at 17, 24 ("I am now at the point where I have grown tired of chopping myself up to order into tiny, less-than-human pieces. I'm going to let the cuts heal and let myself grow whole."); Ruth Gibian,Refusing Certainty:Towarda Bisexualityof Wholeness,in CLOSER TOHOME,supra, at 3, 14 ("Inclusion. Discoveringoneness where we believed therewas polarity. Healing false dichotomies."). Others,however, conclude that their lives are too "fluid"for such integration. TO See, e.g., Nina Silver, Coming Out as Heterosexual,in CLOSER HOME,supra, at 35, 46 ("My life is a continual process of coming out. Ultimately, I cannot identify myself even as bisexual. As a human being striving to reach my fullest potential,I can only relate as the lover.");Dvora Zipkin, WhyBi?, in id. at 55, 72 ("[The label] bisexual ... comes closest to describingwho I am, who I have been and who I may become. Might that change? Of course-that's what this essay has been about.");see also Kathleen Bennett, Feminist Bisexuality: A Both/And Optionfor an Either/Or World, in CLOSER TO HOME, supra, at 205, 228 (reviewing a selection of bisexual lifestories and concluding that bisexuals, feminists, and "gay liberationists"should "bas[e] our identity not only on fluidity, but on fluidity of sexuality"). 94. See Gayle Rubin, ThinkingSex: Notes for a Radical Theory of the Politics of Sexuality, in

528

STANFORD LAWREVIEW

[Vol. 46:503

differ aboutwhether As long as people who sufferanti-gaydiscrimination bornor becamegay-indeed, aboutwhetherthey are gay-neither a they were can approach adequately ground purelyessentialistnor a purelyconstructivist notwithpro-gaylegal theory. And differwe do, mediareportsto the contrary offers no theoreticalfoundation legal protection for standing.97Immutability of those gay men and lesbianswho experiencetheirsexual orientation conas tingent,mutable,chosen. This exclusion will only get worse as a distinctive movementof bisexualstakes shape: The fairnesstheoryof pro-gayessentialism does not explainwhy bisexuals-by hypothesis sexcapableof satisfactory ual encounterswith membersof the so-called "opposite" sex-should not be or for encouraged forcedto do so. But buildinga new foundation legal protection on the contrary is and assumption-that sexualorientation constructed not biologically determined-would risk the same exclusion in reverse. An adequate legal theory shouldprotectthe entiresocial class on whose behalf it is articulated.
D. Litigation

Pro-gaylitigationhas a numberof important objectives. It seeks concrete remediesfor plaintiffswho have been materially harmed anti-gaydiscrimiby nation. It also seeks the more symbolicattributes justice for plaintiffsperof sonally, by restoring to them some measure of their dignity and civic engagement. More broadly,it seeks to establishrules of law that will benefit gay men, lesbians, and bisexuals (and, it is to be hoped, other subordinated the and groupsas well), eitherin subsequent litigationor through gradual mysteriousprocessesby whichlegal rulesshapepublicandprivatenorms. Finally, pro-gaylitigationinvitescourtsto use theirprivilegedpositionas social producers of meaning-a kindof epistemological rivalledtodayby science, authority but still formidable-to arrest,and hopefullyto replace,the meaningsgenerated by anti-gaydiscrimination.Of course, pro-gay litigationcan fail, even to spectacularly, obtain these goals. But win or lose, the effects of pro-gay litigationfar exceed those felt by the individualplaintiff. Unlike otherpoliticalactivities,litigationis specialbecauseit activatesthe possibility of success or defeat on any or all of these very differentfronts. When pro-gayadvocatesturnto litigation,then, they face difficultpragmatic
OF signify monolithically");Michael Warner,Introduction,in FEAR A QUEER PLANET, supra note 3, at xii, xxvi ("The preferencefor 'queer' represents,among otherthings, an aggressive impulse of generalization; it rejects a minoritizinglogic of tolerationor simple political interest-representation favor of a in more thoroughresistanceto regimes of the normal.... 'Queer'... suggests the difficulty in defining the population whose interests are at stake in queer politics."). 97. Media reportson the scientific studies examinedbelow tend to exaggeratethe extent to which members of gay/lesbian/bisexual/queer communities supportthe argumentfrom immutability. For a discussion of how this exaggerationreflects the mainstreammedia's relationshipsto gay, lesbian, and bisexual community sources, see Kay Diaz, Are Gay Men Born That Way?,Z MAGAZINE, Dec. 1992, at 42, 46; KathrynE. Diaz, The Cultural(Mis)appropriation a Brain Cell Study? The Media, the Moof tive, and the Hypothalamiof Gay Men, GAYCOMMUNITY Oct. 6-12, 1991, at 9, 9-11. For a NEWS, discussion of pressures within gay, lesbian, and bisexual communitiesto suppress differences on this point, see Marc A. Fajer, Can Two Real Men Eat Quiche Together? Storytelling,Gender-RoleStereoL. types, and Legal Protectionfor Lesbians and Gay Men, 46 U. MIAMI REV.511, 522-27 (1992).

February1994]

IMMUTABILITY CRITIQUING

529

andnormative questionsthatshouldnot inhibitthemwhenthey engagein other forms of politicalaction. Outside of litigation,gay men, lesbians, bisexuals, and queers-in fact, anybodywho bringsan openmindto debatesaboutsexuality-should continue to disagreeas vociferouslyas we wantaboutthe degreeto whichour identities are essentialor constructed.But the distinctiveand broadeffects of litigation, in in particular powerto normalize law andcultureany definitionsof homoits sexuality and homosexualsit adopts,must be measuredagainstthe exiguous that need for a doctrinal argument defineswho we are in ways thatsome of us object to and cannot,and will not, conformto. Pro-gayessentialistsand proshould stop treatingtheir conflict over legal strategyas a gay constructivists winner-take-all contest,and seek commonground.
III. THE STUDIES

Recent studies in neuroanatomy, endocrinology,and behavioralgenetics fromimmutability.These the have strengthened allureof the pro-gayargument have enteredpopularculturesufficientlyto influencethe thinkingof reports in as litigatorsandjudges andhave actuallyappeared pro-gaylitigationstrategy for the argument from immutability. proudsupports recentstudiesreportThis Partundertakes critiqueof the most important a ing on biological correlatesto homosexualorientation.Three considerations limit the scope of this critique. First,I will reporton but will not pretendto as the develop further critiqueofferedby scientificprofessionals they evaluate the experiments light of the standards modem science and delineatethe in of (quite narrow)reach of the studies' well-justifiedconclusions.98Second, I
98. William Byne and Bruce Parsons provide an intelligent and intelligible description of how experimentaldesign limits the inferences that can be drawn from published researchreports. William Byne & Bruce Parsons,Human Sexual Orientation:The Biologic Theories Reappraised,50 ARCHIVES GEN.PSYCHIATRY (1993). Among the questions debated in that literatureare whether biological 228 causes of humansexual orientationcan ever be deducedfrom neuroanatomic findingsor studiesof twins raised apart. A series of studies issued in the mid-1980s demonstratedthat experience could cause changes in braindevelopment. Id. at 229, 237 nn.6-8;see also Russell D. Fernald,Cichlids in Love, THE SCIENCES, July/Aug. 1993, at 27; RichardC. Francis,KiranSoma & Russell D. Fernald,Social Regulation of the Brain-Pituitary-Gonadal NAT'LACAD.SCI. 7794 (1993) (reporting on Axis, 90 PROC. changes in the size of cichlid fish brain structureafter manipulationof fishes' social structureand gonadal state). A well-conducted twins study cannot prove more than a correlation. Because Bailey and Pillard studied twins raised together, they were unable to do more than wish away possible environmental differences in experience. See TheodoreLidz, Reply to [Bailey and Pillard's] "AGeneticStudyof Male Sexual Orientation,"50 ARCHIVES PSYCHIATRY (1993) (letterto the editor);J. Michael Bailey GEN. 240 & RichardPillard, in id. (responseto Lidz defendingmethodologyof twins study). Even twins studies using twins raised apartcan produce correlationsthat can be described as genetically caused only on highly speculativegrounds. One famous study of twins rearedapartfound thatdifferenttwin pairs liked the same brandof cigarettes,had marriedwomen with the same name, and had chosen the same names for their children and dogs. Donald Dale Jackson,Reunion of Identical Twins, Raised Apart, Reveals Some AstonishingSimilarities,SMrrHSONIAN, 1980, at 48; see also ThomasJ. Bouchard,Jr., David Oct. T. Lykken, Matthew McGue, Nancy L. Segal & Auke Tellegen, Sources of HumanPsychological Dif223 ferences: The MinnesotaStudyof TwinsRearedApart, 250 SCIENCE (1990). For an examinationof the many ways in which environmentalinfluences can contributeto the development of similar traits even in twins raised apart, see FARBER, supra note 77.

530

LAWREVIEW STANFORD

[Vol. 46:503

on leave to other critics the internaland externalcriticismof this literature aboutgender.99 that it incorporates assumptions questionable grounds here-to find common ground from And third, the project undertaken can which pro-gay essentialistsand pro-gay constructivists launch litigation strategiesthat include ratherthan divide pro-gayconstituencies-can hardly mustcertainlybe fromimmutability that begin on an assumption the argument and wrong. For pragmatic ethicalreasonsarisingfromthe specialdemandsof falls shortof a claim the litigationstrategy, followingcritiquequitedeliberately are that the relevantexperiments bad science. To be sure, this Partoffers a aboutsexual orientation criticismof the experiments' categories, assumptions here is not thatthe studiesare flawed becausethey but the argument pursued definimake such assumptions.Science must always begin with hypothetical it of the phenomena seeks to study. drawnfrom conventional tions, language, A realist would posit that science eventuallyrefines those conventionalaswhile a pragof correctdescriptions nature, to sumptions arriveat transparently matist would merely requirethat good science graduallyaccommodateits categoriesclosely enoughto the emergingdata that it can offer more or less serviceabletechnologies. The rereliablepredictionsand, where appropriate, thanconcluderesearch rather orientation cent studiesof homosexual inaugurate that realists grantto in this field: They are owed the forbearance programs to all science, even if theircategorical and that pragmatists grant early work, reflect social biases.10? assumptions They are not owed more than that, however. Specifically, they are not is owed any deferenceat all on the questionwhetherhumansexualorientation bimodal or continuumized:whetherit appearsin the forms of homo- and betweenthese two. When an exor heterosexuality in more minutegradients to decisionwhether assumehumansexualorienfaces the threshold perimenter tation to be bimodal or continuouslyvariable, she faces a question that over the pro-gay underliesthe political disputeamongpro-gayconstituencies favor a bimodalmodel; opponents from immutability:Proponents argument model. Thus, science must choose betweenpolitically favor a continuumized of loadeddefinitionsfor its hypotheses. Becausethe ultimatestrength statistitraitsthat recal findings dependsin parton the model of sexual orientation and the available searchers adopt as their hypothetical apparatus,101
99. Neuroanatomicaland hormonalreportsrepeatedlyassume that homosexuals are sex inverts, such that gay men resemble women, and lesbians men, along some dimensionthat has been found to be dimorphicfor sex. This assumptionhas led hormoneresearchersto classify the male rat who presents but himself to be penetratedas "homosexual," the male rat who penetrateshim as "heterosexual."Byne a & Parsons,supra note 98, at 231. This inversionhypothesisrichly warrants thick culturalexplanation, but I do not attempt to fashion one here. For a brief restatementof the rationale for distinguishing sexual orientationfrom gender, and pursuinganalysis of the former on the assumptionthat it operates semi-autonomouslyfrom the latter, see Halley, supra note 87, at 1724-26. 100. For an argumentthat science may incorporatesocial biases and neverthelessbe "good sciAND 132, ence," see Elizabeth Potter,Modeling the Gender Politics in Science, in FEMINISM SCIENCE 141-44 (Nancy Tuana ed., 1989). ET 101. See EAVES AL.,supra note 73, at 45 ("Estimatesof genetic and environmental parameters will be biased if the model is wrong.").

February 1994]

IMMUTABILITY CRITIQUING

531

are apparatuses politicallycontested,science may neverfind itself hypothetical in a position to adjudicate essentialism/constructivism the dispute. The troubling featureof recentscientificdevelopments the areaof human in sexual orientation not experimenters' is definitions simple use of conventional about homosexualand heterosexualorientation (they could hardlydo otherhave been wise). The problemarisesin the way the resultingscientificreports assimilated the broader into cultureandinto legal culturein particular. Oftenat the urgingof the scientiststhemselves,this science has been understood the in broader cultureas a series of empirical thatthose definitionsare confirmations naturaland fixed. But the studieshypothesizedefinitionsof homosexualand heterosexual orientation are subjectto questionpolitically;the definitions that the scientistsincorporate shouldremaintopics for politicaldebate. Thus, the following critiquefocuses not on what science can prove,but on how science worksat its commonborderwith culture. At thatborder,cultural and practicesbecome scientificassumptions, scientificfindingsare offered to confirm and disconfirmculturalintuitions. As the following analysis shows, the politicalcharacter sexualorientation of categorieslurksat the outeredge of the experimenters'method. Those of us who work politically and legally shouldnot allow the brightlight thrownby scientificfindingsto dazzle us so thatwe cannotsee the politicalproblems the experimenters that assumedaway. defiIndeed,as I arguein PartIV, the excludeddynamicsof sexual-orientation nitionarepreciselywherecommongroundbetweenpro-gayconstructivists and essentialistsis to be found. pro-gay A. RecentExperiments Scientists have acclaimed three recent research reports-one from and in neuroanatomy two from behavioralgenetics-as majorbreakthroughs the search for biological causes of homosexualityin humans.102Simon structure launchedthe currentavid scientific LeVay's study of hypothalamic andmediainterest,perhaps becauseit was the firstneuroanatomic studyclaimbetweensexualorientation a brainstructure has and that ing to find correlations been shown to have directcontrolover sexual behaviorin rats.103
102. See Bailey & Pillard,supra note 1; Hameret al., supra note 1; LeVay, supra note 1. Recent work on hormonal patternshas not achieved similar success. William Byne and Bruce Parsons have concluded that hormone research currentlycomprises two separate literatures-studies of nonhuman mammals and studies of humans-that are not as yet mutually illuminating. Byne & Parsons, supra note 98, at 230-34. Research in nonhumanmammals has sought examples of gender nonconformity exhibited in sexual behavior,for example, male rats who exhibit "lordosis,"or readinessto be mounted, and female rats who mount. Id. at 231. But this model has not been successful in predictingthe bewilbetween dering variety of gender nonconformityfound in researchon humans,where the interrelations hormonebalances, genital configuration,and social gender attribution complex; where sexual orienare tation identity and erotic gesture are relatively autonomous;and where sexual fantasy often entertains differentobjects than does sexual behavior. Id.; see also Louis Gooren, Biomedical Theories of Sexual Orientation: A Critical Examination, in HOMOSEXUALITY/HETEROSEXUALITY: CONCEPTS SEXUAL OF 71 ORIENTATION(David P. McWhirter,StephanieA. Sanders & June Machover Reinisch eds., 1990). 103. LeVay, supra note 1. For earlier studies of brain structuresnot linked to sexual behavior in nonhuman animals, see Laura S. Allen & Roger A. Gorski, Sexual Orientationand the Size of the Anterior Commissurein the Human Brain, 89 PROC. NAT'LACAD.Sci. 7199 (1992); D.F. Swaab &

532

LAWREVIEW STANFORD

[Vol. 46:503

have emergedfromwhatI will call Geneticclaims aboutsexualorientation "twinsstudies"and,most recently,in a pedigreeand geneticlinkagestudy. In seek evidenceof geneticrelationships examining a twins study,researchers by whethera traitappearsin identicaltwins, fraternal twins, nontwinsiblings, or greateror less thanone would expect from the adoptedsiblings in proportions C. at in trait'sdistribution the population large. J. MichaelBailey and Richard Pillard's widely acclaimedstudy of male twins and adoptedbrothersis the largestand most carefullydesignedtwins study to date.104At the time of its the publicationit constituted strongestevidenceyet that any biological differmen cause ratherthanresultfrom theirsexual ences betweengay and straight
orientations.105

Since publicationof Bailey and Pillard'stwins study, a study using pedigree and DNA-linkageanalysis has producedan even bolder genetic claim. examinedgay men's family tree Dean H. Hamerand a team of researchers in or pedigrees,and observedhigherratesof homosexuality maternal patterns, thanpaternal rather relatives,suggestingthatmale homosexuality mightbe geIn throughthe mother.106 the next phaseof theirinvestineticallytransmitted materialfrom a linkage study, Hamer'steam examinedchromosomal gation, withno morethanone lesbianrelativeandno homosexual brothers homosexual found that 64 percentof the sibling pairs fathers or sons. The researchers sharedan identifiable Accordingto geneticsequenceon the X chromosome.107 is the authors,thatresultwas "evidencethat one form of male homosexuality side the maternal andis geneticallylinkedto transmitted through preferentially chromosomalregion Xq28."'08 They ultimately concluded, however, that in to to while this gene sequenceappeared contribute sexual orientation some it could not be a determiningfactor for homosexual orientation males,
generally.109
RES. 141, 146 M.A. Hofman, An Enlarged SuprachiasmaticNucleus in Homosexual Men, 537 BRAIN (1990). Both studies used methods similar to LeVay's. 104. See Bailey & Pillard, supra note 1. 105. See id. I dub Bailey and Pillard's study of male twins "the twins study" merely for shorthand, though (1) it focuses on twins, nontwin siblings, and adopted siblings; and (2) other important studies have used its methods. Bailey and Pillardhave conducteda more recent study of sexual orientationin women. It followed almost exactly the methods used in their study of men and drew virtually the same conclusion: that homosexuality in women is significantly heritable. J. Michael Bailey, RichardC. Pillard, Michael C. GEN. Neale & Yvonne Agyei, Heritable Factors InfluenceSexual Orientationin Women,50 ARCHIVES PSYCHIATRY (1993). Because of these similarities,the twins study of women is subjectto the same 217 criticisms I bring to the twins study of men, and I give it separateattentiononly when it departsfrom the male study in ways importantfor my analysis. See, e.g., note 144 infra. A more recent twins study focused on a source of variance ignored in the Bailey and Pillard studies: differences in the subjects' sexual practices. FrederickL. Whitam, Milton Diamond & James Martin,Homosexual Orientationin 187 SEXUAL BEHAV. (1993). For a Twins:A Reporton 61 Pairs and ThreeTripletSets, 22:3 ARCHIVES ANDHOMOSEXUALITY: A collection of articles on twins studies and sexual orientation, see TWINS CASEBOOK (Geoff Puterbaughed., 1990). 106. Hamer et al., supra note 1. 107. Id. at 325-26. 108. Id. at 325. 109. Id.

February 1994]

IMMUTABILITY CRITIQUING

533

Hamer'slinkage study thus forges out beyond Bailey and Pillard'stwins can studyby proposingthatthe causationof male homosexuality be studiedat the level of a specific, identifiedgeneticsequence. It is important, however,to the understand limits of Hamer's study, as the study itself describes them. Hamer'slinkagestudy focused on a groupof homosexualmen selectedto inwas subjects'homosexuality tensify the likelihoodof findingthatthe primary the to transmitted themby theirmothers. Accordingly, "oneformof male holinkedto the geneticsequenceis the formthataffects mosexuality" purportedly no gay men with homosexualbrothers, lesbianrelatives,and no gay fathersor aboutthe causes of sons.l10 The study providedno basis for any statements in gay homosexuality familieswith lesbiansor paternally-related men, no basis for inferring thatthe genetic markers identifiedin the studywould reappear as of significantin a study of the generalpopulation gay men, and no basis for of supposingthat non-gaybrothers men in the studydo not also have the genetic markers identifiedin the study."' Moreover,as the authorsare quick to point out, the study did not show any genetic concordancein seven of the study's forty brotherpairs.112Thus, even if one assumes that the identified the genetic sequenceexpressesitself somewherein its bearers, findingsdo not commandthat theirhomosexuality thatplace.113 is contestedunderstandings the categories,homosexualand hetof Culturally erosexual, became part of the experimental hypothesesof all three studies. Thatin itself is not troubling.It is troubling, of however,thatcultural reception these studieshas led nonscientificreadersto supposethat the categorieshave themselvesbeen foundto exist in nature.Forthreereasonsit is most appropriate to examine this problemin LeVay's brainstudy and Bailey and Pillard's twins study, and to pretermit of consideration Hamer'spedigreeand linkage Hameremployscategorical study. First, assumptions very similarto thoseused and Pillard,but providesless detailabouthow he deployedthem."4 by Bailey Second,LeVay,Bailey, andPillardhave all activelyandpersonally engagedin mediacoverageof theirstudies,fosteringthe misunderstanding this body of of science, while Hamercuts a farless prominent figurein the mediadatabases.15 And third,Hamerdrawsconclusionsfrom his data that affordan opportunity

110. See Mary-ClaireKing, Sexual Orientationand the X, 364 NATURE 288 (1993). 288, 111. Hamer et al., supra note 1, at 325; see also Anne Fausto-Sterling& Evan Balaban,Letters: Genetics and Male Sexual Orientation,261 SCIENCE (1993) (describinglack of a control group of 1257 nonhomosexual brothersas "[t]he most obvious weakness[ ]" of the Hamer study). 112. Hamer et al., supra note 1, at 325. 113. Biologists Anne Fausto-Sterling and Evan Balaban point out additional weaknesses in Hamer's study, in particularits reliance on certainassumptionsthat, if disproven,might rob their findings of statistical significance. Fausto-Sterling& Balaban,supra note 111; see also Neil Risch, Elizabeth Squires-Wheeler BronyaJ.B. Keats, TechnicalComments:Male Sexual Orientationand Genetic & 2063 (1993); Dean H. Hamer, Stella Hu, Victoria Magnuson, Nancy Hu & Evidence, 262 SCIENCE 2065 Angela M.L. Pattatucci,Reply to Male Sexual Orientationand Genetic Evidence, 262 SCIENCE (1993). 114. Hamer et al., supra note 1, at 321-22. 115. See texts accompanyingnotes 121, 123, 145-148, & 238 infra.

534

LAWREVIEW STANFORD

[Vol. 46:503

to for pro-gay essentialistsand pro-gayconstructivists meet on the common I identifyin PartIV below.116 ground B. The Brain Study In the most celebrated to studylinkinghomosexuality biologicaltraits,Simon LeVay found that a groupof cells in a certainportionof the brain-the nucleusof the anterior thirdinterstitial or hypothalamus INAH 3-was larger in men he classified as heterosexual than in men he classified as homosexual.117His experiment used postmortem samplesfrom the brainsof forty-one men and six women.118He drewfromhis findingsessensubjects,thirty-five tialist conclusionsthat increasedin confidenceand breadthas he moved beyond "highscience"fora to more popularculturevenues. In the research in reporthe published Science,LeVayclaimedonly thathis has and findings "suggest[] that sexual orientation a biologicalsubstrate" "ilin is to lustrate[] thatsexualorientation humans amenable studyat the biological level."119He concededhis inabilityto determinewhetherthe size of an individual'sINAH 3 "is the cause or consequenceof that individual'ssexual but for of orientation," indicateda preference the formerexplanation his findsimilarbrain ings by pointingout studiesshowingthatthe size of an apparently in structure maleratsremainsstableafterbirthandstronglycorrelates with "the amountof male-typicalsexual behaviorshown by the animals."120 In discussing his study with the popularpress, LeVay has consistently insistedthathis studywill pointedout these limits, andyet has simultaneously for be foundational determining whethernatureor nurture causes sexual orientation-indeed, that his researchmakes this the questionupon which future inquirymust focus.121 "We can't say on the basis of [the brainstudy] what
116. See text accompanyingnotes 236-238 infra. 117. LeVay, supra note 1, at 1035. 118. Id. LeVay's study adoptedthe assumption,prevalentin hormonalresearch,that male homosexuals' brains would be more like female heterosexuals'brains than like male heterosexuals' brains. Id. The underlyingparadigmis one of male homosexual feminization. Because LeVay was unable to identify any of his female subjects as homosexual,he drew no conclusions about the etiology of homosexual orientationin women. Id. I do not examine this interestingassumptionabout the intersectionof gender and sexual orientation. 119. Id. at 1034, 1036. 120. Id. at 1036. 121. See also David Perlman,Brain Cell Study Finds Link to Homosexuality:Tissue Differs Between Gay and StraightMen, S.F. CHRON., Aug. 30, 1991, at Al, A12 ("'What the study means is that I've observed structural differencesin the brainsof gay and straightmen, but it doesn't indicate whether you're born with those differences ... or whetherthere's some aspect of sexual behavior-perhaps even in AIDS itself-that leads to structural changes.... [W]hat's most importantis that homosexuality is now a topic that can be studied in the laboratoryas a problemin neurobiology,and not something that must be left to the psychiatristsor the psychoanalysts.' (quotingLeVay); MacNeil/Lehrer NewsHour: ") Sex and the Brain (PBS television broadcast,Aug. 30, 1991) (transcript No. 4150 at 4-6, available in LEXIS, Nexis Library,CurrentFile) (conceding that his study did not resolve whetherthe difference in brain structurewas the cause or the effect of homosexuality,but insisting that his study opens up avenues for discovering "what it is that causes us to be homosexual or heterosexual"). In these statementsLeVay proposes that his study frames the question for future investigatorsto answer. The goal is to find the cause of humansexual orientation,the options are natureor nurture,and the forms in which sexual orientationappearsnaturallyare mappedby the homo/heterodichotomy. To be sure, he has occasionally made more equivocal claims about his work. See Simon LeVay, Replica-

February 1994]

IMMUTABILITY CRITIQUING

535

makespeople gay or straight," explains,"[b]utit opens the doorto find the he As answerto thatquestion."122 he raisesthis question,he also suggeststhatthe will ultimatelybe found to match designationshomosexualand heterosexual theirbiologicalbases, makingno secretof the answerhe expects to find: "'If thereare environmental influences,'he says, 'theyoperatevery earlyin life, at the fetal or early-infancy stage, when the brainis still puttingitself together. I'm very much skeptical of the idea that sexual orientationis a cultural thing.'"123In anotherinterview,he indicatedthe depthof this convictionand the strengthwith which he believes thathis findingsdemonstrate truth: its was when of LeVay alonein his fifth-floor laboratory his moment discovthe blindto ery came."I was measuring totalvolumein the brainsamples, wheretheycamefrom,andit wasrightthere." Thescientist looked downat thetissuesamples, blurred tears."I now by wasveryemotional about I hada lotinvested mywork.... I havealways it. in felt thatI wasborngay,"he said.124 LeVay's claim to the mediais thatthe brainsof gay men differfromthose of heterosexual men in a way thatcauses theirhomosexuality, that essenand tialistassumptions aboutsexual-orientation will be proventrue. One categories need not decide whethersuch a claim is empirically plausibleor implausible, or bad, or pragmatically useful or harmful,to see that in normativelygood reachingit LeVay begs the questionof essentialismvel non. LeVay assumes essentialismin framingthe findinguponwhichhis conclusionis basedas well, and thus begs the questionas to even the less controversial componentof his claim that INAH 3 size differsbetweengay and heterosexual men. could be promoted the firsttentativeproofthathoas LeVay's experiment mosexualorientationis a biologicalfact, only becauseconstructivist elements in homosexualidentitywere excludedfrom its method. A somewhatdetailed review of LeVay's experimental indicatesthatprofoundly procedure complex social andpoliticalelementsof sexualorientation identityfall outsideits frame. medicalrecords LeVayobtainedthe braintissuesamplesandcorresponding "at directlyfromhospitalpathologists routineautopsiesof personswho died at seven metropolitan hospitalsin New Yorkand California."125 LeVay's report does not suggest that he soughtor obtainedpermissionsfrom the individuals beforetheirdeaths. LeVaythushadno access to his subjects'own assessments of their sexual orientations to the historyof theirsame-sexor heterosexual or
tion Will Tell, N.Y. TIMES, Oct. 7, 1991, at A16 (letter to the editor) ("The ultimatesignificance of my differences reportwill depend on whetheror not it can be replicatedand on determiningif the structural I have seen are actually related to a person's sexual drive.");see also Angier, supra note 2, at 1 ("Dr. LeVay in no way claims to have discovered the-or even a-cause of male sexual preference, but merely suggests he has detected something worth furtherinvestigation."). But the modesty of these statements only brings into high relief his more typical claim to have framed the agenda for future research. 122. Gelman et al., supra note 2, at 50 (quoting LeVay). 123. Id. at 50-52. Gelmanfurtherreportsthat"LeVay... thinksa small numberof sex genes may be isolated, perhapswithin five years: 'And that's going to blow society's mind."' Id. at 48 (quoting LeVay). 124. Talan, supra note 2, at 41. 125. LeVay, supra note 1, at 1035.

536

STANFORD LAWREVIEW

[Vol. 46:503

contacts. Instead,LeVay dependedwholly on the subjects'medical records. "at The recordsof the twenty-sixsubjectswho haddied of AIDS indicated least one AIDS risk groupto which [each] patientbelonged(homosexual,intravethe nous drug abuser,or recipientof blood transfusions)";126 recordsof the indicatedsexual orientation only by fortuity. remainingsubjects or on LeVay classified his subjectsas "homosexual" "heterosexual" these scant records. If a subjecthad died of AIDS and his recordsindicatedthathe was in the risk group"homosexual," LeVay classifiedhim as homosexual.127 classifiedas heterosexual AIDS patientswho, accordtwo Conversely,LeVay to theirrecords,had deniedengagingin homosexualactivity.128At least ing one subject'srecordsindicatedthathe had been identifiedas bisexual;LeVay If classified him as homosexual.129 the recorddid not indicatepast sexual classifiedthe subjectas heterosexual the basis of the nu"on practices,LeVay merical preponderance heterosexualmen in the population."'30 of This last includedall the women.131 group In each instance,LeVay treateda representation sexual identityas the of identitiesto the thing itself. The entireprocessof assigningsexualorientation brainsassumedthatthe artifactual document a patient'shospitalrecordwas of a transparent window into his or her sexual history. But to take at face value denialsof homosexualactivityor identityis to ignore possibly self-protective the social historyof AIDS as it has shapedthe meaningof gay male identity. The AIDS/HIVepidemichas been conducted a way thatrequires in everyHIVabouthomosexual positivepersonto do something identity:disavowit, confess the it, embraceit, ally with it. LeVay's methodexcludes from consideration of the complexsocial patterns identityprofessionandascription, refracting layers of representation which the image of sexual orientationis managed, in groomed,appropriated, negotiated,and captured.He reducedthis complexity to a single characteristic: essentialsexualorientation lodged neatlywithinthe atomizedindividualwho has died.132 This exclusionof the social and representational aspectsof sexual orientation identity makes LeVay's handlingof his bipolarcategories,homosexual and heterosexual, almoststartlingly crude. If a subjecthad died of AIDS and had left recordsclassifyinghim as belongingto the risk group"homosexual," that LeVay implicitlydetermined thatsubjecthad engagedin homosexualconductandwas therefore A properly given a homosexual identity.133 single act of homosexualconductthatcould transmit HIV is treated irrevocably as defining:
126. Id. at 1037 n.9. 127. Id. at 1036 n.7. 128. Id. 129. Id. at 1035. 130. Id. at 1036 n.7. 131. Id. at 1035. 132. For a more satisfactorybut necessarily more complex account of the relationshipbetween homosexual identificationand HIV or AIDS diagnosis, see CINDY INVENTING AIDS 120, 131 PATTON, (1990) ("The relationshipbetween identity-especially gay ... identity-and HIV is highly unstableat that present.... [T]he AIDS narrativeexists as a technology of social repression;it is a representation attemptsto silence not only the claims of identity politics, but the people marginalizedby AIDS."). 133. See LeVay, supra note 1, at 1036 & n.7.

February 1994]

IMMUTABILITY CRITIQUING

537

It was of itself constitutiveof a subject'shomosexuality.Conversely,LeVay that "presumed" any subjectnot identifiedas homosexualwas heterosexual.134 defendedhis research critiquethat againstthe internal LeVay has subsequently as he hadno way of assuring his renders dataunreliable, this assumption that,if subbetterinformed,he would not have designatedmany of his heterosexual Whatever ultimateresolutionof those debatesabout the jects homosexual.135 of technique, LeVay's deployment his categoriesremainsopen to experimental culturalcriticismfor ignoringthe complexityof his subject. By makinghis class a universal heterosexual default,LeVayinsiststhatall personsare indeed of locatedin one of his two classifications.As a matter theoretical assumption, he eliminatesthe possibilityof a personwith a sexualityneitherheterosexual nor homosexual. as LeVay himself defined sexual orientation "thedirectionof sexualfeelof or behaviortowardmembers one's own or the oppositesex."136Clearly ings fails to capturethis nuance. LeVay never knew his categorical apparatus and to what extent,male subjectswhose recordsindicatedno homowhether, had sexual contactsnevertheless had conscioushomoeroticfeelings on which of they had not acted, either because they disapproved those feelings themsocial disapproval them. of selves or were not willing to contradict widespread Nor did he know to what extent those subjectshad had homoeroticfantasies as how complexly whichthey hadfailedor refusedto construe such. No matter they might have engagedin homosexualdesires,LeVay's methodmade them heterosexual. Conversely,naming a man gay because he has been infected with HIV throughanalintercourse with another manis to deny the complexity of fantasy,desire,andinternal publicidentitythatgive subjectiveandpuband lic heterosexual identityits layeredinstabilities. errorwas more grave than simply misclassifyingsome subjects LeVay's who are "really" heterosexual homosexual,or vice versa. He has changed as the natureof his categoriesfrom the merely lexical to the ontological. His method forced these categoriesto describeand conclude the entire range of humanpossibility-to constituteus, no matterwho we are and what we do or feel. That'sessentialism. Manypeoplecommitted increasing to legal protection for lesbians and gay men are deeply averseto this view of sexual orientation categories;others arejust as staunchlydevotedto it. But both partiesto the debateshouldbe able to agreethatan expro-gayessentialism/constructivism on an essentialistassumptioncannot prove an essentialist perimentresting conclusion.
134. Id. at 1035. 135. Joseph M. Carrier& George Gellert, Letters, 254 SCIENCE (1991) (letter to the editor) 630 (noting "misclassificationbias" in LeVay's use of "bipolarcategories of 'heterosexual'and 'homosexual' men"). But see Simon LeVay, Letters, 254 SCIENCE 630 (1991) (letter to the editor) ("I may 630, well have oversimplified the problemin my study, but sometimes such oversimplificationis necessary to make progress in a novel field."). 136. LeVay, supra note 1, at 1034 (emphasisadded). My thanksto Lisa Haydenfor pointing this out.

538

LAWREVIEW STANFORD

[Vol. 46:503

C. The TwinsStudy As LeVay acknowledged his research in report,even a conclusiveshowing that a particular brain structure largerin heterosexual is than in homosexual men cannotexclude the possibilitythatthe differenceis the effect ratherthan the cause of sexual orientation.137 Thus, when J. MichaelBailey and Richard C. Pillardannounced carefullyconstructed administered a and studycomparing the degreeto which congenitalmale twins, fraternal twins, and adoptedbrothers exhibitedthe sameor differentsexualorientation,138 hopes were raisedthat genetic effects would be isolated. Bailey and Pillardfoundthatidenticaltwins were morelikely to be consonant for homosexualitythan fraternaltwins or adopted brothers.139Like LeVay, Bailey and Pillardmake conservativeclaims for their study in their researchreport. In the Archivesof GeneralPsychiatry, they reachthe guarded conclusionsthat their statistics"suggestthat genetic factorsare important in individualdifferencesin sexual orientation," are "consistent and determining with some genetic influence."140 of They estimatethatheritability homosexuin ality is between31 and74 percent, depending parton the incidenceof homoIn sexualityin the population.141 commentsconsistentwith otherresearchers' predictionsof the strengthof heritabilityfindings in behavioralgenetics,142 does not indicatethe presenceof they also explainthata findingof heritability a genetic on-off switch, concedingthat hereditymay cooperatewith environmentalfactorsin producing individuals willing to live as gay men.143Indeed, they pointout thatfindingthatanyidenticaltwinspairsdifferin sexualorientation proves that sexual orientation not entirelygenetic.144 is like LeVay, however,Bailey and Pillardmake more heroic claims Again for theirdatain othervenues. They begana New YorkTimeseditorialessay by that proclaiming "[s]cienceis rapidlyconvergingon the conclusionthatsexual orientation innate."145 an interviewwith Science, Bailey announced is In the
137. Id. at 1035-36. 138. Bailey & Pillard, supra note 1. 139. Id. at 1093. An anomaly in Bailey and Pillard's male study was that fraternaltwins were only slightly more likely to be consonantfor homosexualitythan adoptivebrothers. Id. at 1092 ("[T]he difference was only marginallysignificant."). Their study of lesbians resultedin similar findings. Bailey et al., supra note 105, at 219 ("Rates for [fraternaltwins] and adoptive sisters did not differ significantly."). 140. Bailey & Pillard, supra note 1, at 1093. 141. Id. 142. See, e.g., Plomin, supra note 79. 143. See id. at 1095. "[G]iven any heritabilityestimate, there are a variety of possible developmental mechanisms. For instance, these data are consistent with heritable variationin prenatalbrain development or in some aspect of physical appearancethat, by way of differentialparentaltreatment, leads to differences in sexual orientation."Id. 144. Bailey, Pillard, and their coauthors make this point explicitly in the lesbian twins study. Bailey et al., supra note 105, at 222 (explainingthatidenticaltwins "who differ in sexual orientationcan do so only because relevant environmentalfactors differ"). LeVay makes the same point in a book addressedto a popularaudience. SIMON BRAIN (1993) (concluding from the 138 LEVAY, THESEXUAL variability of sexual orientationbetween identical twins that "naturealone, or nurturealone, cannot provide an adequateexplanationfor our sexual individuality"). 145. J. Michael Bailey & RichardPillard,Are Some People Born Gay?, N.Y. TIMES, Dec. 17, 1991, at A21.

February 1994]

IMMUTABILITY CRITIQUING

539

team's belief that their twins study suppliesthe elementof causationmissing from LeVay's brainstudy.146He statedthathe and Pillardmaintaina "workthat the genetic substrate ing hypothesis" they identified"affect[s]the partof If the brainthat [LeVay]studied."147 confirmed, hypothesis this would suggest thathypothalamus causes homosexualorientation.In the same interview, size Bailey also threwcold wateron theoriessuggestingthatsocial experiencecontributesto homosexuality:"'No one has ever founda postnatal social environmental influence for homosexualorientation-and they have looked plenty
..
"'148

instance,promptlyexpressedthe view that,thoughearlierstudiessuggesteda to the genetic contribution homosexuality, Bailey andPillardstudy"reallysort

of clinches it."149

Others seem to agree. Psychology professor Gregory Carey, for

As Bailey and Pillardusheredtheirstudyinto the mainstream media,they consistentlyimplied that they had shown that homosexualitywas a discrete trait,not a continuousone. TheirNew YorkTimeseditorial,for instance,describes homosexualityand heterosexuality as conventionalcategoriesof not often act analysis, but as the labels for distinctpopulations: "Homosexuals in brainstructure differentlyfrom heterosexuals early childhood"; may differ "between homosexual and heterosexualmen"; "a biological explanationis But of good news for homosexuals."150 a close examination theirmethodsindicates that, even more than LeVay, they have simply assumedthis bipolar model of sexual orientation, have systematically and excludedthe social and political dynamicsof sexual orientation identity. in 161 Advertising urban publications, gay Bailey andPillardrecruited selfidentifiedgay and bisexual men (I will call them the "volunteers") their and male twins or adoptivebrothers(the "brothers").151 test their hypothesis, To and Pillardhad to assign sexualorientation identitiesto both groups.'52 Bailey of They used threeindicators sexualorientation identity. First,when possible, asked subjectswhetherthey identifiedas "homosexual/gay," they "bisexual," or "heterosexual." Next, they asked subjectsto ratethemselvesfor adultfantasy and behavioron the Kinsey scale.153When they were unable to ask a brother these questions(eitherbecausehis volunteer siblingrefusedpermission
146. Constance Holden, Twin StudyLinks Genes to Homosexuality,255 SCIENCE (1992). 33 147. Id. In a recent interview,however, Pillardexpressedcautionaboutrelying on LeVay's study, indicating that it should be replicatedbefore such inferences are drawn. Edward Stein, Evidencefor & Queer Genes: An Interviewwith RichardPillard, 1 GLQ:J. LESBIAN GAYSTUD. 103-04 (1993). 94, 148. Holden, supra note 146, at 33. Bailey does, however, accept the possibility of environmental influences that are "biological"ratherthan social. Id. Nevertheless, Pillardmakes sweeping claims for the exoneratingpower of even a genetic contributionto sexual orientation,implicitly denying that that would be any differentfrom an unmediatedgenetic determination a genetic contributionthat cooperor ates only with prenatalfactors: "'A genetic component in sexual orientationsays, "This is not a fault, and it's not your fault ...."'" Gelman et al., supra note 2, at 48 (quoting Pillard). 149. Gay Men in TwinStudy,N.Y. TIMES, Dec. 17, 1991, at C5 (quotingGregoryCarey, Assistant Professor of Psychology at the University of Colorado). 150. Bailey & Pillard, supra note 145. 151. Bailey & Pillard, supra note 1, at 1090. 152. For a detailed discussion of their methodology, see id. at 1090-91. 153. See ALFRED KINSEY, C. WARDELL POMEROYCLYDE MARTIN, B. & E. SEXUAL IN BEHAVIOR THE HUMAN MALE 638 (1948).

540

STANFORD LAWREVIEW

[Vol. 46:503

to contacthim, or becausehe declinedto participate the studyor to answer in abouthis sexual orientation154), and Pillardaskedthe volunquestions Bailey teer sibling to designatehis brother'ssexual orientation.Once again, the options were "homosexual/gay," If "bisexual,"or "heterosexual." the brother could make such an identification with a high degreeof certainty,Bailey and Pillardused that identification the brother.155 for Unlike LeVay's subjects,Bailey and Pillard'swere alive when the experimentersmet them. They chattilyascribedsexualorientation identitiesto themselves, estimated the proportionof their desires and conduct that were homosexualor heterosexual, sexualorientations, in and guessedtheirbrothers' some cases activelyrefusedto answerquestionsabouttheirsexualorientations. In short, Bailey and Pillardset up a living theaterof sexual orientation selfand in representation allo-interpretation. Particularly its handlingof bisexual the brothers, twins study reachesout more identityand of the nonresponsive decisively thanLeVay's to the borderbetweenscience and culture,creatinga more detailedphotographic that negativeof the culturalassumptions it silently
incorporates. 1. The "bisexuals."

all Bailey and Pillardclassified as homosexuals self-designated bisexuals, an unspecifiednumber subjectswhose Kinseyresponsesindicatedthattheir of sexual practicesand fantasieswerejust as often heterosexual homosexual, as andthreesubjectswhose sexualpractices fantasiesweremoreoftenheteroand sexual than homosexualbut who "ratedthe idea of having sex with men as 'very sexually exciting.'"156 This methodof classifyingambiguoussubjectsis entirelyconsistentwith similaressentialist LeVay's, andincorporates assumptions.Onceagain,human sexual orientation assumedto be bimodaland polar(homosexual/heterosexis identitiesand are imposedon ual). These categoriesswallow all intermediate indication homosexual of any every subject; anything (self-description, fantasy, or conduct)conclusivelyclassifiesa subjectas "ahomosexual," while the mere absenceof those indicators leaves the subjectin a defaultclassification, of that heterosexuals.By forgoingthe opportunity recognizea bisexualcategoryor to to allow for a continuumof sexual orientation categories,Bailey and Pillard embracean essentialisthypothetical apparatus.Theiruse of the Kinsey scale not only frustrates heuristic the purposefor whichthe scale was designedin the first place, but also ignoresimportant refinements sexual-orientation in models that have arisento overcomethe scale's defects. The social science of sexual orientationhad alreadymoved beyond the crudedichotomous systemhypothesized Bailey and Pillardwhen AlfredC. by
154. Bailey and Pillardwere denied permissionto contact21% of the brothers; those whom the of experimenterswere permittedto contact, almost 6% did not returnquestionnaires. Bailey & Pillard, but supra note 1, at 1091. One brotherfilled out the questionnaire, left the questionsabout his sexuality blank. Id. 155. Id. 156. Id.

February 1994]

IMMUTABILITY CRITIQUING

541

the Kinseyscale reflects Kinseyintroduced very scale they use. The multipoint the proportion a person'ssexualbehavior fantasythatis homosexual and and of heterosexual.157 the scale in this way to defeat the use of Kinsey designed Far "homosexual" "heterosexual" personaldesignations.158 fromrealizand as used to flout it, as Bailey and ing this intention,the Kinsey scale is repeatedly Pillard'sunapologetic illustrates.And Bailey and Pillardare misappropriation not alone: People familiarwith the scale readily invoke it as a catalog of humantypes (e.g., "he's a Kinsey 5" or "she's a Kinsey 1"). To stanchthese sex have soughtto amendthe Kinsey scale tendencies,subsequent researchers in ways thatemphasizepossiblerefinements the hypothetical in construction of sexual orientation categories. An important alternative measure by proposed MichaelStormsfocuses on a pointthatBailey andPillardfailedto notice: the way in whichthe Kinseyscale and as represents homosexuality heterosexuality mutuallyexclusive opposites, such that one must retreatif the otheris to advance. Of coursea personwho scores as a Kinsey3 (equallyhomosexual heterosexual) and couldhave heterosexual eroticismas intense as any Kinsey 0 and homosexualeroticismas intense as any Kinsey 6. Stormstherefore arguesthat the intensityof homo- or heterosexualeroticismshould be registeredas a variableindependent the of eroticism.159 heteroeroticism proportionof homo-or heterosexual Arraying by intensityon one axis and homoeroticism intensityon anotheryields a rich by field of variation readilysubsumed not into the identitieshomosexualand heterosexualexcept at its extremes. Anotheroften-noted defect of the Kinseyscale is its aggregation behavof ior and fantasy.160 This featureof the scale, thoughit properlyacknowledges the relevanceof two quiteincommensurate elementsof eroticlife, at once confounds them and omits many other factorsthat enter into sexual orientation identity. Sexualfantasyandbehavior may entertain divergent objectsandmay, in turn,differ from desire, personalself-designation, public identity. To and addressthis problem,FritzKleinhas developeda multifactorial to register grid the possibility that, in a single person, sexual attraction, behavior,fantasies, emotionalpreference,social preference,self-identification, "hetero/homo and life style" mightnot "match"-and furthermore distinguish to betweenthe re-

157. Kinsey's seven points on the continuumwere: (0) exclusively heterosexual;(1) largely heterosexual but with incidentalhomosexualhistory;(2) largely heterosexualbut with a distinct homosexual history; (3) equally heterosexual and homosexual; (4) largely homosexual but with a distinct heterosexual history; (5) largely homosexual but with incidental heterosexualhistory; (6) exclusively homosexual. KINSEY AL.,supra note 153, at 638-39. ET 158. Explainingthe findings he producedon the basis of this continuum,Kinsey stated:"Malesdo not representtwo discrete populations,heterosexualand homosexual. The world is not to be divided into sheep and goats. Not all things are black nor all things white." Id. at 639. 159. Michael D. Storms, Theories of Sexual Orientation,38 J. PERSONALITY Soc. PSYCHOL. &
783, 784-86 (1980); see also ALFRED C. KINSEY, WARDELL B. POMEROY & CLYDE E. MARTIN, SEXUAL BEHAVIOR THE HUMANFEMALE IN 470-72 (1953).

160. See, e.g., Mary McIntosh,The HomosexualRole, reprintedin THEMAKINGOFTHE MODERN HOMOSEXUAL 39 (Kenneth Plummered., 1981) (originally published in 1968). 30,

542

STANFORD LAW REVIEW

[Vol. 46:503

suiting patterns in a subject's present, past, and "ideal" life.161 These factors not only complicate-perhaps the better word is multiply-the identity "bisexual," they also acknowledge that the social designation "heterosexual" may be assigned to a person whose sexuality is quite complex. Bailey and Pillard claim that they merely simplified their analysis when they reduced sexual orientation to a homo/hetero dichotomy.162 But they did much more than that, particularlywhen they introduced their study to the general public with no proviso that their categories "homosexual" and "heterosexual" were experimental devices, not human types soon to be mapped unerringly onto nature. In a subsequent interview with philosopher Edward Stein, Richard Pillard expressed the only misgivings I have found in the public record about the misprision he and Bailey have invited. Commenting on the study of lesbians, Pillard said: I thinkwomen are muchmoreflexible in theirsexualorientation-they don't as often label themselvesas gay or straight.Usuallywhen you ask thatquestion of men, at least men over the age of thirtyor thirty-five,a few will say they're bisexualbut most will say they'regay or straight; they dichotomize. Women often will say, "Well, it dependson who I'm with, on what sort of I'm that relationship having...." Andthey'll oftenhavehadrelationships are lesbian and relationships are heterosexual that .... You might want to call morewomenbisexual,whichwe endup doing,butI thinkwomen'sorientation is really much more complicated thanmen's sexual orientation.163 If the struggle to define sexual-orientation categories is a political one, the imposition of Bailey and Pillard's model on experimental subjects who actively resisted it-even though it occurred within the framework of the experiment itself-is also political. By presenting their results to the press without explaining that this imposition was merely an experimental convenience operating at the level of an untested hypothesis, Bailey and Pillard effectively transferred some power to make decisions about sexual categories out of the political and social spheres into the scientific laboratory. 2. The nonresponsive brothers.

In one of their most tendentious moves, Bailey and Pillard assigned nonresponsive brothers the sexual orientations attributedto them by their volunteer siblings if the volunteers were willing to say that they made the attribution with complete or virtual certainty.164 Bailey and Pillard based this
161. Fritz Klein, The Need to View Sexual Orientationas a MultivariableDynamic Process: A Theoretical Perspective, in HOMOSEXUALITY/HETEROSEXUALITY, note 102, at 277, 280; see also supra Braden Robert Berkey, Terri Perelman-Hall& Lawrence A. Kurdek, The MultidimensionalScale of 67 (1990); Richard E. Whalen, David C. Geary & Frank Johnson, Sexuality, 19 J. HOMOSEXUALrrY Models of Sexuality, in HOMOSEXUALrTY/HETEROSEXUALrrY, note 102, at 60. supra 162. Bailey & Pillard,supra note 1, at 1091 ("Becausethe broadercategoriesare fewer in number and more readily understood,we have used them herein."). 163. Stein, supra note 147, at 99 (first ellipsis in original). A gender analysis would ask whether the difference Pillard comments on here results in an account of homosexuality less appropriateto women than to men. 164. Bailey & Pillard, supra note 1, at 1091.

February 1994]

IMMUTABILITY CRITIQUING

543

inferenceon theirown andotherstudiesshowingthatsiblingpairswith responsive brothersalmost always agreed on a descriptionof the brothers'sexual This practiceexemplifies the experimenters' orientation.165 exclusion of the in which one brother'ssemaphore his sexual oriof dynamics representational entationto his brother possiblyotherfamily membersbecomespartof the and other brother'ssexual orientation identity. Once again,social-representational are excludedonly by hypothesis,andcannotbe deemedto be proven dynamics unimportant a studythat ignoresthem. by First,there's the problemof why Bailey and Pillardhad a problemin the firstplace. A volunteer siblingmightrefuseto grantthe researchers permission to contacthis brotherfor a greatnumberof reasons,many of which directly touchthe questionof how well a brother knowshis sibling'ssexualorientation aboutlettingtheirbrothers (andposidentity. Gay men areoften apprehensive sibly, throughthem, other family members)know they are gay. Thus, a gay man might well hesitatebefore doing anythingto indicatean interestin his brother'ssexual orientation. particular, openly gay man who thoughthis In an brother was gay but deeplyclosetedmightpausebeforemakingsuch inquiries. of in Conversely,the brother an openly gay man mightrefuse to participate a behavioralgenetics study that probessexual orientation because the precisely genetics studyitself revives fearsor hopes thatthe brothers' sexuality,or their styles of self-disclosure,are tightly linked. instructedeach volunteerto consider Bailey and Pillard's questionnaire himself "completelycertain"about his brother's sexual orientationif his brother told the volunteer had how he identified, to considerhimself"virtuand certain"if he felt he could decide on the basis of his brother's"behavior ally alone."166It is not clear why the experimenters reposedsuch confidencein a brother's decontextualized verbaldisclosureof his own sexualorientation.Nor is it clear why declarations heterosexualorientation of were deemed as unas problematic professionsof homosexuality.It is not clearwhy behaviorwas considereda less reliableindicator thanspeech,or what kindsof behaviorthe volunteerswere supposedto recall, or what inferencesthey were supposedto drawfromdifferent with stronglyheld typesof behavior.Of course,volunteers ideas aboutthe semiologyof sexualorientation inconsistent with these instructions might ignorethem: Why shouldtwo brothers' aboutultimate agreement designationerase this rich sourceof misprision? In relying on a volunteer'sassessmentof his brother'ssexual orientation, sexualorientation Bailey and Pillardtreatedthe brother's categoryand his sibconfidencein ascertaining categoryas independent.But they can't that ling's be, because you're less likely to feel certainabout which categoryto assign when applyinga rigid either/or categoricalsystemto someonewho sends ambiguous signals. Indeed, Bailey and Pillardnote that the volunteerswhose brothers to "werenot accuresponded questionsabouttheirsexual orientation ratein predicting whethera nonheterosexual relativewouldlabel himself 'gay/
165. Id. 166. Id.

544

LAWREVIEW STANFORD

[Vol. 46:503

homosexual' or 'bisexual.'"167The volunteerstended not to mislabel selfor identifiedheterosexualrelativesas bisexual,168 to mislabel self-identified Whereerrorsoccurred, homosexualrelativesas bisexual.169 they involvedthe who identifiedthemselvesas bisexual. Of of volunteers'descriptions brothers the the six such brothers, volunteers only one as bisexual;they desdesignated and six two of the remaining as heterosexual, threeas homosexual.170 ignated The siblings tendedto confirmidentitiesin the polarcategorieswhen describrelaand ing theirbrothers to imposethe polarcategorieson more ambivalent did tives. (There'sno reasonto thinkthe volunteers not do somethingsimilar in describingthemselves.) These complex epistemologiesof the closet171 suggest not only that two tolerancefor disclosurein light of whateach thinks brothers mightadjusttheir their the otherthinksis good, but also thatthey mightactuallyreconceptualize howeverglimidentitiesin light of theirunderstanding, own sexualorientation identities.They suggestthatsexual mering,of theirsiblings'sexualorientation in orientation each brothermay not be an atomized,essential state, but a reunderthese circertainty sponsiveactivity. To rely directlyon the volunteers' the cumstancesis to deeply underestimate complexityof sexual knowledge. was conducted,however,withoutmuch attentionto these The experiment of difficulties. Bailey and Pillardconcludethattheirroutinemiscategorization bisexualsposes no problemto theirmethodbecause"[its]majordistinction... and is betweenheterosexual nonheterosexual relatives,which [the volunteers] and made quite well."172In otherwords,the researchers theirsubjectsshared that the same essentialistassumptions led themboth to exclude the classificaone in tion "bisexual" the firstplace. But the knowingness has aboutsomeone on how that person enacts it and on one's else's sexual orientation depends to of interpretation the resultingperformances.The twins study purports examine sexual orientation; instead,it covertlyand complexlyenacts sexual orientationidentity. 3. One autobiography. autobiOne of Bailey and Pillard'sown subjectsgave a sexual-orientation that suggestshow deeply theirexperiment reflects,ratherthanexamography ines, social practicesof sexual-orientation categorization.Accordingto this Barnett" until he was confidentin his heterosexuality "Doug autobiography, was twenty-eightyears old, when his identical twin brother"came out" to is him.173Barnett,who believes that sexual orientation genetic, becamecuriwith sex with men and was ous abouthis own orientation.He experimented
167. Id. 168. They did so for only 1 out of 80 such relatives. Id. 169. They did so for only 4 out of 35 such relatives. Id. 170. Id. 171. I am pluralizingthe term invented by Eve Kosofsky Sedgwick. SEDGWICK, EPISTEMOLOGY, supra note 92, at 67-68. 172. Id. 173. Gelman et al., supra note 2, at 46. Newsweekcarefullypoints out that"Doug Barett" is "not his real name." Id.

February 1994]

IMMUTABILITY CRITIQUING

545

"A to "morefulfilling."174 yearlaterbothtwins surprised find suchencounters in told theirparentsthey were gay."175They laterparticipated Bailey and Pillard's twins study. Barnett'sautobiography Bailey andPillard'sstudyincorporating imand it asked the same question,"WasBarnettever straight?"Barnett'sown plicitly is reportedbelief that homosexuality genetic implicitlyanswersthat his substantialheterosexual was delusional-a long, unwillingconscription of history a naturallyhomosexualbody into heterosexual culture. Bailey and Pillard's conclusionthat Barnett'shomosexuality probablygeneticallycaused transis lates that belief into the authoritative of vocabulary science. But once we decidethatBarnett be wrongabouthis sexualorientation, can why not say he is wrongaboutit now? Of course,passingas straight captures moresocial rewards thanpassingas gay, a circumstance createssome prefthat erence for believing Barnettnow and not then. But thatexplanation does not accountfor Barett's sweet surpriseat his homosexualengagementor its imhad plicit suggestionthat his sojournin heterosexuality not felt like a forced marchthrough alien territory.To save appearances this case as it is reported in to us, we need to acknowledge Barnett'sstablebelief thathis twin brother that was heterosexual, his laterloss of thatbelief, may have been causes of his and sexual orientationat least as important their common genes. We cannot as believe anything aboutBarnett's unlesswe also believe thathis self-description to some extenton his perception his brother. of self-description depends Barnett'sperception his brother of depends,in turn,on his brother's representation himself. Thatrepresentation apparently of was plastic,at least in that the brotherhad some power over when to tell Barnettthat he was gay. It is almost unimaginable the brother'sreasonsfor delay were unrelated a that to desireto avoid the stigmaof homosexual the knew identity. Moreover, brother that as long as he failed to "come out" to Barnett,Barnettwould think him heterosexual.Thus, the social and representational of superordination heteroand the supporting of sexuality over homosexuality, ideologicalconstruction as deviates,acted twice to heterosexuality a normfrom which homosexuality of shape Barnett'sperception his brother'ssexual orientation.
Causa causae est causa causati: "The cause of the cause is to be consid-

ered as the cause of the effect also."176If perceivinghis brotheras straight causedBarnett thinkof himselfas straight, to thenthe social superordination of over homosexuality, the interpretive and that supports heterosexuality system to himselfas straight.Perceivinghis brother it, also causedBarnett experience as gay activatedthe reversecausalchain. In both phasesof Barnett'slife, the intervened his sexual-orientain representational system of sexual orientation tion identity. Researchers between homosexuality using twins to study the relationship and geneticsworryaboutascertainment but not enough. They worrythat bias,
174. Id. 175. Id. 176. BLACK'S LAWDICTIONARY (6th ed. 1990). 220

546

STANFORD LAWREVIEW

[Vol. 46:503

twins studies,unlessbasedon a random sample,will attract subjectswith some of commitments characteristics will distorttheir findor that atypicalpattern researchers worrythatpeople who have a personalinvestings. In particular, ment in promoting of biologicalexplanations homosexuality-perhapspeople like Barett-may volunteerfor twins studiesmore often thanpeople who do not.177But a moreprimary formof ascertainment occurswhen studysubbias sharethe experimenters' thattwo types of sexualorientation jects assumptions is exist, that one (homosexuality) a markeddeviationfrom the norm represented by the other (heterosexuality), that any homosexualbehaviorexand cludes one permanently fromthe class of heterosexuals.Whensubjectsdo not sharethese assumptions, in the cases of Bailey and Pillard'sself-described as heterosexualswith high scores for homoeroticfantasy,the bisexuals,and the decisionto imposethose assumplesbians,the experimenters' identity-resistant tions anywayimplicatesthe politicalproblemsdescribed above. But when exand subjectsagreeon these cultural perimenters assumptions, they slip into the conclusionscovertly.
IV. COMMON GROUND

As a matterof logical coherenceand simple honesty,pro-gayessentialists need not capitulate pro-gayconstructivists, vice versa,to reacha common to or essentialistsand pro-gay constructivists litigation strategy. Pro-gay already share commonground. It, ratherthantheirdifferences,can and should form the basis of theircommonlitigationstrategy.178 This Partbegins the task of definingthat commongroundby setting out ranges of conceptuallyavailableessentialistand constructivist positions. It then spells out how historians anthropologists sexualityhave differedin and of the degreeto whichthey pursuepureessentialist, or pureconstructivist, merged and offers the optimal approaches, arguesthatone particular mergedapproach commonpositionfromwhichpro-gayadvocatesshoulddeveloplitigationstratthat egy. Finally, it concludesthat equal protectionarguments can be articulated from that position offer betterconceptualand descriptiveresourcesfor thanmore essentialistalternatives. pro-gayequal protectionarguments

177. A recent twins study invited such a possibility. The researchersadvertisedin the gay press for study subjects using the following inducement:"Twinresearchis an importanttechniquefor understanding the nature vs. nurtureproblem in the social sciences ascertainingwhich aspects of human sexuality are learnedand which are biologically determined."Whitamet al., supra note 105, at 203 app. A. Even without such a bald invitation,it is hardto imagine gay, lesbian, or bisexual twin adultswho do not know why experimentersmight want their time and attention. 178. The differences will remain,but will, I hope, be irrelevantto the argumentsand descriptive claims pro-gay litigants will need to make in court. For the many importantissues on which pro-gay constructivistsand pro-gay essentialists disagree, fora other than litigation remain fully available. Indeed, any proposal that such a strategicdivision of audiences is unethical probablyrests on the easily challenged assumption that the state is entitled to a clear, unmediatedview of all one's meanings at once.

February 1994]

IMMUTABILITY CRITIQUING

547

and Constructivism A. Comingto Termswith Essentialism Up to this point, I have assumedthat an essentialistview of homosexual orientation claims thathomosexuality a deep-rooted, is fixed, andintrinsicfeature of individuals,determined not chosen. I have assumedthat the conand structivist view of homosexual orientation holdsprettymuchthe opposite: that it is a contingent, malleabletraitthatarisesin a personas she manages socially her world, its meanings,and her desires.Essentialismand constructivism are a good deal morecomplexthanthat. The purposeof this sectionis to actually describethe range of possible conceptualpositionsthat could fall underthe rubricsessentialistandconstructivist.Distinguishing amongessentialismsand constructivisms particularly is becausehigh feelings in essentialism/ important constructivism controversies tend to obscureunderlying definitional problems, and to foreclose the possibilityof agreement strategiccooperation. and 1. Essentialisms. In her influentialbook on the culturaland literaryissues raised by postmodem antiessentialism, Diana Fuss defines essentialismas "a belief in true essence-that whichis most irreducible, and constitutive unchanging, therefore of a given personor thing."179 particular, notes thatnatureis the paraIn she digmaticsource of true essence: "[F]orthe essentialist,the naturalprovides the raw materialand determinative starting point for the practicesand laws of the social."180 Fuss' definitionselide threeaxes along which essentialistpracticesdiffer. First,she recognizesonly one sourceof essences: nature. Second,she uses an ambiguousterm-"unchanging"-as though it had a single meaning. And third, she treats all essentialismas a matterof individualratherthan group characteristics.In the following paragraphs, explore the consequencesof I I of makingthe first two of these distinctions; defer consideration the thirdto the conclusion of this section because it implicatesthe relationship between essentialismand constructivism practice. in Distinguishingessentialismfrom biological causation. Essentialismassumes at minimum a pureandperfectdefinitionof a particular that thingcan be found. An essentialistview of a treeis that,even if the humanmindis not now can it, capableof articulating some definitionof "tree" be framedthat is irreduciblein the sense thatit has all the necessarydescriptors no unnecessary and in the ones, and thatis constitutive the sense that,wherever qualitiesdescribed by those terms appearin a single thing, that thing will be a tree. If this is what an essence is, clearly naturecannotbe the only source of essences. Any attribution an irreducible, of constitutive characteristic a perto son or thingattributes it an essence. An artifact very plausiblybe saidto to can

179. DIANA SPEAKING: NATURE DIFFERENCE & 2 (1989). FUSS,ESSENTIALLY FEMINISM, 180. Id. at 2-3.

548

LAWREVIEW STANFORD

[Vol. 46:503

have an essence, as Plato himself acknowledgedin the Seventh Letter.181 Thinkof the kitchentable caughtin the hedges of To the Lighthouse.182 Sorulesestablishessences: Naturedid not createfelons or cially descriptive legal and but tortfeasors, law nonethelessimposes on them irreducible constitutive characteristics. of Attribution a naturalessence, then,is but one kindof essentialism. It is "essentialism-plus": generic essentialismthat specifies a certainsource-nature-for the essence it detects. I propose,therefore,to distinguishbetween of and thereforeconstitutive "weak essentialism"(attribution any irreducible of essentialism" of characteristic, (attribution regardless its source)and "strong such a characteristic is also natural biological). that or Detaching essentialismfrom naturepermitsone to see that biologically causedcharacteristics neednot also be essentialones. Freckles,tonsils,the size of "moons"at the base of one's fingernails, abilityto roll one's tonguethe these examples suggest that biologicalcauses determinemany featuresof an individual arerarely,if ever,considered defineher.183Even a biological that to characteristic needs some otherrationaleto explainwhy it is essential. In all but a few cases, humancharacteristics socially and culturally are rationale designatedas essential. To be sure,fromtime to time the additional that explains why a characteristic constitutiveand defining is so predomiis between esnantly naturalthat it virtuallyrestoresthe univocal relationship sence and nature posited by Fuss. Imagine a genetically caused physical conditionso harmful healththatno infantbornwith it lives morethana few to days. Such a condition(along with any genes thatcause or preventit) distinguishes with grim certaintythose who must die from those who may. The decisiveness of this distinctionmightjustify our deeming the characteristics that producesuch outcomes essential featuresof individuals,particularly of those who must die becauseof them. Thatis, nature,appearing its guise of in imminentand inescapable physicalmortality, mightproperlybe consideredto
181. Plato, Letter VII, in THECOLLECTED DIALOGUES PLATO OF INCLUDING LETTERS THE 1574, 1589-90 (Edith Hamilton & HuntingtonCairns eds. & L. A. Post trans., 1963). Plato wrote: For everything that exists there are three classes of objects through which knowledge about it must come; the knowledge itself is a fourth, and we must put as a fifth entity the actual object of knowledge which is the true reality. We have then, first, a name, second, a case if description,third, an image, and fourth,a knowledge of the object. Take a particular you want to understandthe meaning of what I have just said; then apply the theory to every object in the same way.... The same doctrineholds good in regardto shapes and surfacesboth straightand curved, in regard to the good and the beautiful and the just, in regard to all bodies artificial and natural, in regardto fire and waterand the like, and in regardto every animal,and in regardto every quality of character,and in respect to all states active and passive. Id. (emphasis added). 182. When Andrew Ramsay tried to explain his father's philosophicalwork to Lily Briscoe, "she said Heavens, she had no notion what that meant. 'Think of a kitchen table then,' he told her, 'when if you're not there."' Lily went on to think, "Naturally, one's days were passed in this seeing of angular essences, this reducingof lovely evenings, with all their flamingo clouds and blue and silver to a white deal four-legged table (and it was a markof the finest minds so to do), naturallyone could not be judged
like an ordinary person." VIRGINIA WOOLF,To THE LIGHTHOUSE 38 (1927).

183. Of course, many differences that are not currentlyconsideredto be biologically determined, such as preferencefor bright colors or spicy foods, are similarly deemed nonessential.

February 1994]

IMMUTABILITY CRITIQUING

549

essential.184Short of provide the rationalefor deeming these characteristics such cases, the importance even of a biologically caused characteristic-the decisivenessthatinducesus to considerthatits possessorsare irrevocably and suffusedin their very being with that characteristic-will almost ceralways tainly be social. That means that a commitment essentialism,even strongessentialism, to cannoteliminatethe obligation examinethe social andcultural to processesthat to designatecharacteristics be so important definewho or whatpeopleare. they Distinguishingkinds of change. Under an Aristotelianconceptionof esthat be means sence, the requirement an essentialcharacteristic "unchanging" not that it cannotbe alteredor effaced,but thatalteringor effacingit changes the natureof the thing underdefinition.185 Thatthe letters"S-T-O-P" esare sentialto a stop sign does not imply thatthey are impervious change;if that to series of lettersis removed,the sign may still be a sign, but it is no longer a stop sign. But thatis not to answerquestionsaboutotherkindsof change. Is it a stop but sign if only the "S"is missing? If it says "S-T-O-P" is paintedgreen,or is fails round,or is in Beijing? Fuss' use of the term"unchanging" to anticipate these varioustypes of change,all of whichmay be crucialto decidingwhether a given sign remainsa stop sign. Similarly,the claim that a sexual-orientation categoryis immutablemay have a rangeof meanings. If the Aristotelian sense of immutability appliesto claims of identity,then the essentialistmerely claims that to be a lesbian, a woman must have whatevercharacteristic defines a lesbian, and if she loses thatcharacteristic ceases to be a lesbian. Claimsthathomosexualorientashe tion is immutableusuallyrely on a much strongerdefinitionof change-that whateverit is thatconstitutes essence of homosexual the identitycannotbe lost or removed from a persononce it exists, whetherthat occurs at conception, before birth,in infancy, at a wild high school party,or in an agony of early adult self-re-creation. None of these distinctions whatthe essence of sexual orienpredetermines tationis. Some peopledefinea "homosexual" a personwho entertains as desire for erotic contacts with a person of the same sex. Othersrequiresame-sex fantasy,as well or instead. Still othersrequireactualeroticcontacts,andthere is a lot of disagreement about whetherit takes one contact or many, recent contactsor any, to makea homosexual.Finally,the problemof identityvexes as any effort to establishall important aspectsof homosexualorientation unchanging. If a person has a lot of same-sex contacts, desires them and fantasizesaboutthem, but does not imaginehimself to be gay, or a homosexual, or bisexual, is he wrong? And public identity has its own range of
184. For an extrapolationfrom this point to a more inclusive "Aristotelianessentialist" list of see capabilitiesthat define "humanness," MarthaC. Nussbaum,HumanFunctioningand Social Justice: In Defense of Aristotelian Essentialism, 20 POL.THEORY 214-23 (1992). 202, 185. Essence and Existence, in 3 THEENCYCLOPEDIA OFPHILOSOPHY 59 (Paul Edwardsed., 1967) ("ForAristotle, the essence of an object. . . was what finds expression in the concept which the object embodies, the concept under which it must be identified if it is to be identified as what it is.").

550

STANFORD LAWREVIEW

[Vol. 46:503

mutabilities by largelyregulated the closet and the rule thatanyonenot designatedhomosexualis by defaultconstrued heterosexual. as It is not at all settled,insidepro-gaycommunities outsidethem,whichof or these elementsis essentialto the identityhomosexual,and which is merelyan accidentthat can alter withoutproducinga definitionalchange. Weak essentialismis committedto characterizing or moreof these elementsas essenone tial; strongessentialismadds that the essence(s) are producedby nature;and from immutability most often say strongessentialistsespousingthe argument thatthe essence(s) are installedat conception, beforebirth,or not long thereafter. But the constructivist more interestedin the special sort of mutability is in engendered the simultaneous by operation a single cultureof incommensurate selections from this meta-definitional menu. If you think someone is a homosexualbecausehe routinely same-sexeroticcontacts,andI thinkhe is has not becausehe fantasizesaboutwomen,on a constructivist analysisour struggle to define him is a struggleto changewhat he is.
2. Constructivisms.

If constructivism an essence, it is the claim that humanactivities of has or or perception, conceptualization, (or description, workproduce maintain produced or maintained) some partof the worldor the world itself.186 Constructivisms vary along two axes: who or what does the constructing, and who or whatgets constructed.Labellingtheorydevelopedin sociology in the 1950s and 1960s, for instance,predominantly concerneditself with the ways in which individuals designated othersas, for example,deviant,come by to occupy the meaningsand institutionsassignedto them, and thus come to in themselvesas deviantand in "managing" identhat participate constructing In labellingtheory'sstudiesof deviance,the constructing doneby the is tity.187
186. Constructivismis not the same as the "antiessentialism" that has come under fire at the intersectionof critical race theory with feminist legal theory in articles critiquingfeminism's use of the category "woman." See, e.g., KimberleCrenshaw,Demarginalizingthe Intersectionof Race and Sex: A Black Feminist Critiqueof Antidiscrimination Doctrine, Feminist Theoryand Antiracist Politics, 1989 U. CHI.LEGAL 139, 152-57; Trina Grillo & Stephanie M. Wildman, Obscuringthe Importanceof F. Race: The Implication of Making ComparisonsBetween Racism and Sexism (Or Other -Isms), 1991 DUKE 397, 404-05; Angela P. Harris,Race and Essentialismin FeministLegal Theory,42 STAN. L.J. L. REV. 581, 586-87 (1990). First, the intersectionality literatureis not primarilyconcernedwith replacing the essentialism it critiques with constructivism. Second, when intersectionalityscholars seek to keep white feminists from covertly defining feminism's constituency as "white women" ratherthan "women," the actual targetof their analysis may not always be feminist essentialismin the sense in which I use that term here. On that sense of the term, feminist legal theory is essentialist only if it involves a claim that being a woman is conceptually inconsistent with being a person of color, but most of the examples cited in the intersectionality analysesprobablyinvolve white women'sforgetting or not caring about women of color-conceptual and normativeerrorsthat are more accuratelydescribednot as essentialist but as racist. cussing deviance by reference to those inside or outside a particularset of social rules); ERVING
GOFFMAN, ASYLUMS: ESSAYS ON THE SOCIAL SITUATION OF MENTAL PATIENTS AND OTHER INMATES

187. See HOWARD BECKER,OUTSIDERS: S. STUDIESIN THESOCIOLOGY DEVIANCE OF (1963) (dis-

(1961) (examining the relationshipof an inmate's world view to his role in the institutionalculture); ERVING THE OFTHE IN GOFFMAN, PRESENTATION SELF EVERYDAY (1959) (examining the role an LIFE individual plays in shaping others' impressions of her); ERVING STIGMA: NOTES THE ON GOFFMAN,

February 1994]

IMMUTABILITY CRITIQUING

551

and and particular playersinvolvedin deviancerelationships institutions, what constructare specific deviantand, implicitly,normalpersons. they social studies of science, leave Sociology of knowledge,and particularly that largelyin place labellingtheory'sassumption individuals actingaloneor in but groups within institutionsare the source of epistemicconstructions, they have considerably claim aboutwhatthose individudeepenedthe constructivist als construct.Sociologistsof knowledgeinsistedthatall perceptions reality of areproduced But through conceptual systems.188 they did not attendsystematically to the social locationor interestsof those who produceknowledge,and have been criticizedfor giving little attention "theextent to which socially to constructed meaningsmay reflect and reinforcethe disparate power of ruling elites."'89 To some extent,recentworkin sociology of science moves beyond this limit. Social studies of science pursuethe hypothesisthat professional forms of organization, and communication, activity (and not an objectivelyIn perceivednature)are whatproducescientific"facts."190 doing so, they focus sustainedattentionon the constraints academicdiscipline,on the pracof tices of group work, and on the local cultureof careercompetition.19lThis inquirythus takes into accountthe possibilitythat the activitiesof producing knowledgeproduce,in turn,the producers. This double extension of social constructivism reachedperhapsits most extent in the work of Michel Foucault. Foucaultpositedthathuman thorough facts aboutlanguageor humanbeknowledgearises "notfromtranscendental but from the systematicqualityof knowledgeitself, organizedinto an ings," that set "episteme,which is the total set of relationsor discursiveregularities out the possibilitiesof meaningandtruth."192 specificdiscourseswhichan The functionto produceknowledgenot by theirreferepistememight incorporate ence to things outside of discourse but by their internal regularities.193 discourse "[T]herecan be no question,"Foucaultbelieved, "of interpreting with a view to writinga historyof the referent."194 formsof human Moreover, experienceand subjectivitywere themselvesproducedthroughthe historical
MANAGEMENT SPOILED OF IDENTITY (1963) (formulatingconcepts about how stigmatized individuals convey self-image). 188. PETER BERGER THOMAS L. & THE CONSTRUCTION OFREALITY: TREAA LUCKMANN, SOCIAL TISE IN THE SOCIOLOGY KNOWLEDGE OF 97-98 (1966) (describingthe "conceptualmachineriesof universe-maintenance" that construct "the social phenomena");Ernst von Glaserfeld,An Introductionto Radical Constructivism,in THE INVENTED REALITY: How Do WE KNOW WHAT WE BELIEVE WE KNOW?CONTRIBUTIONS TO CONSTRUCTIVISM 17, 19 (Paul Watzlawick ed., 1984) (explaining that "for are constructivists,all communicationand all understanding a matterof interpretiveconstructionon the part of the experiencing subject"). ch. 2, p. 141, on file with the StanfordLaw Review). 190. See, e.g., LATOUR & WOOLGAR, supra note 5, at 105. For other works in this school, see note 5 supra. Ernest Sosa eds., 1992). 193. Id.
191. LATOUR & WOOLGAR, supra note 5. 192. Linda Alcoff, Michel Foucault, in A COMPANION TO EPISTEMOLOGY 143 (Jonathan Dancy & 189. ANDREW KOPPELMAN, THE ANTIDISCRIMINATION PROJECT (forthcoming 1995) (manuscript
at

194. MICHEL FOUCAULT, THE ARCHAELOGY OF KNOWLEDGE 47 (A.M. Sheridan Smith trans., 1972), quoted in HUBERT L. DREYFUS & PAULRABINOW, MICHEL FOUCAULT: BEYOND STRUCTURALISM AND HERMENEUTICS ed. 1983). 61 (2d

552

LAWREVIEW STANFORD

[Vol. 46:503

operationof discourses. Knowledge producesknowers, not the other way

around.195

In all these forms,and in the manyotherswhich I do not distinguishhere, is constructivism a causal theory,in which the causal agent is always some version of "culture."It thus opposes strongessentialism,which identifiesnature as the causal agent,but not necessarilyweak essentialism,which merely identifiesthe existenceof essences.196Some very strongversionsof constructivism identify categoriesthemselvesas the instruments unjustifiedpower of and sources of oppressiveconstraint. This hypothesisis (again) most frethat quentlyassociatedwith Foucault,who proposed knowledgeandpowerare interlinked:"'Truth'is linkedin a circularrelationwith systems inextricably of powerwhichproduceandsustainit, andto effects of powerwhichit induces of offers an attack tably associatedwith the application force, constructivism even on weak essentialism. Shortof strongethicalconstructivism this sort, of and are however,essentialism constructivism opposedonly whenbothreachan exclusive theoryof causation. The possibilityof commongroundbetweenessentialismandconstructivism, conexploredbelow, arisesfromthis incomplete tradiction the formerby the latter. of Nor does constructivism refuteclaimsof immutability.I argued necessarily above that essentialismsdiffer a greatdeal in the sense of "unchangingness" thatthey posit. Correspondingly, constructivism claimsthatimportant features of humanbeings and our world are contingent,historicalproductsof human but activity and interaction, it does not claim that anythingit describesas soconstructed would be easy, or even possible, to change. Eve Kosofsky cially Sedgwickwryly comments:
I remember buoyantenthusiasm the with whichfeministscholarsused to greet the findingthatone or another brutal formof oppression not biologicalbut was "only"cultural! I have often wonderedwhat the basis was for our optimism aboutthe malleability cultureby any one groupor program.198 of and which extend it."197 If the constructivist generation of categories is inevi-

involve a converse ethical problem: Indeed, some forms of constructivism whetherthey leave any roomat all for humanagencyanddecision.199 Thus,it is possible for a constructivist claim thatsexual-orientation to is mutaidentity
195. DREYFUS RABINOW, & supra note 194, at 168-83 ("TheGenealogy of the Modem Individual as Subject"). 196. For a discussion of this incomplete disagreementof essentialism and constructivism, see Daniel R. Ortiz, Creating Controversy: Essentislism/Constructivism the Politics of Gay Identity,79 and VA. L. REV.1833, 1842-43 (1993). 197. Michel Foucault, Truthand Power, in THEFOUCAULT READER 74 (Paul Rabinow ed., 51, 1984). 198. SEDGWICK, EPISTEMOLOGY, note 92, at 41. supra 199. One such critiqueis offered by Bryan D. Palmer,who arguesthat the promotionof language as the sole constitutive activity in some discourse theories tends "to reify language, objectifying it as unmediated discourse, placing it beyond social, economic, and political relations, and in the process
displacing essential structures and formations to the historical sidelines." BRIAN D. PALMER, DESCENT
INTO DISCOURSE: THE REIFICATION OF LANGUAGE AND THE WRITING OF SOCIAL HISTORY 5

(1990).

February 1994]

IMMUTABILITY CRITIQUING

553

ble across the range of humanpossibility,withoutmakingthe distinctclaim that it is mutablein a given person,or even in a given society or era.20 3. Impureidentity. It is a cherishedpremiseof the essentialism/constructivism impasse that essentialismand constructivism mutually are exclusive. On this assumption, if in the formation description essentialismand constructivism or appear together of a social category,logical incoherence hypocrisymustbe at work. Probing or a thirddistinctionpassed over by Fuss indicatesthat this premiseshould be rejected. Fuss focuses exclusivelyon the attribution a fixed, determinative of characteristicto a particular or person. Of course,one can also attribute such a thing characteristic an entirecategory,makingthe entiregroupessentialist. Nothto orderthroughthe step of forming requiresthatone proceedin any particular a categoryon one hand, and the step of placing individualsin it on the ing other. One mightproceeddeductively, establisha categorybeforeconsultand the qualitiesof proposedindividualmembers. One might pursuea more ing inductivecourse,andaggregate similarindividuals beforeforminga apparently in whichto encompass them. Essentialism thesetwo differin generalcategory ent kinds of definitionalpracticetakes correspondingly differentforms: Deductive essentialism would posit that a group shares a certain constitutive while inductiveessentialismwould incrementally note thatindicharacteristic, viduals are stronglymarkedby a characteristc increasinglyseems importhat tant enoughto be consideredconstitutive them. A researcher of might decide before makingany experimental for observations, instance,thata humancatefor enthusiasm havingsex on airplanes gory is definedby its members' (deductive essentialism); or she might find upon interviewing a number of experimental subjectsthatsome of themare salientlydifferentiated such an by enthusiasmand consequently decide to classify them using it as an index (inductiveessentialism). Moreover, each of these essentialistpracticescan take a weak or a strongform: The researcher mightsupposethatnoticingsome subjects' enthusiasmfor having sex on airplanesis a fleetingly useful heuristic (weak essentialism)or that it is a deeply embedded,gravely differentiating characteristic un(strongessentialism).The possibilityof these discontinuities derlies the rifts dividingLeVay and Bailey and Pillard'srationalesfor setting for up crude,conventional hypothetical categories"homosexual/heterosexual" their experiments(weak deductive essentialism) purposes of administering from theiractualpracticeof categorizing individualsubjects(stronginductive essentialism),and again from their misleadingstatementsto the press sugconfirmedthe natural gesting that theirexperiments originof those categories (strongdeductiveessentialism).
200. David Halperin,for example, argues that "[p]articular cultures are contingent, but the personal identities and forms of erotic life that take shape within the horizons of those cultures are not." HALPERIN, supra note 4, at 51-52.

554

LAWREVIEW STANFORD

[Vol. 46:503

Additionalcomplexityarises,first,from the fact thatdefinitionsof groups and of individualscan proceedon constructivist ratherthanessentialistpremises; and second, from the fact that a given definitionalpracticecan exhibit both inductiveand deductivesequences. Indeed,one mightwell find it impossible to proceedin any way otherthanby combining induction deduction and in a process that gives neitherdecisive priority.201 Both forms of complexityinconsistentrelianceon essentialistand constructivist modes of definitionand deductionto inductionto deduction-can be detectedin the cycling through armedservices'practiceof applyingan essentialist defintionof "ahomosexual" as any personwho evinces homosexualdesire202 women who have resisted to sexual harassment men.203However essentialistthe regulatorycategory, by these women are includedin it on the basis of adventitious, contingent,even attributed characteristics: are constructed lesbians. And this infalsely they ductivephaseof the military'sarticulation the categoryof homosexuals of produces a new, deductive essentialism: homosexualityin women has as its essence not the positive demonstration sexual desire for women, but the of failureto evince sexual desire for men. In an important recentanalysisof the essentialism/constructivism impasse in gay and lesbianidentitypolitics,law professor DanielR. Ortizsuggeststhat the impassebe resolvedby makinga strongformaldistinction betweenits antipodes. Ortiz argues that "[t]he questionsof how a person comes to have same-sex desire and how that personis viewed are completelyindependent"
such that "the constructivist debate ....

into a particular identitycategorybut ratherhow the identitycategoryitself is formed."204 This utterdivorceof the processof defininggroupsfrom that of in frominduction, placingindividuals them,of deduction may be conceptually
201. Kant distinguishedbetween "intuition" (which "can never be otherthan sensible"),and "unOF REA("the faculty . . . which enables us to think"). IMMANUEL derstanding" KANT, CRITIQUE PURE SON 93 (Norman Kemp Smith trans., 1968). He concluded that they are mutually informing: "To neither of these powers may a preferencebe given over the other. Withoutsensibility no object would be given to us, without understanding object would be thought. Thoughtswithoutcontentare empty, no intuitions without concepts are blind." Id. The claim that empirical observationis "theoryladen" and thus not prior to theory has been advanced by MARY HESSE,REVOLUTIONS RECONSTRUCTIONS PHILOSOPHYSCIENCE AND INTHE OF 63-109 THE OF INFERENCE S. HESSE, STRUCTURESCIENTIFIC THESTRUC(1980); MARY (1974); THOMAS KUHN, TURE SCIENTIFIC OF REVOLUTIONS (2d ed. 1970). 10-51 202. See, e.g., 32 C.F.R. pt. 41, app. A ?? H.l.a, H.l.c(2) (1993). 203. Militaryregulationsprovidingfor the dischargeof homosexualshave been disproportionately applied to women. MarkThompson,Lesbian TroopsMore Often Get Boot: Insecurity,Vindictiveness FREE Feb. 16, 1993, at 4A (reportAmong Men May Be WhyWomenFace Allegations, DETROIT PRESS, ing that women are less than 10% of militarypersonnel,but 23% of those dischargedfor homosexuality). Many commentatorsexplain this disparityby noting that a woman's noncompliantresponse to sexual overturesby male soldiers can lead to investigationand discharge,whetheror not she is lesbian. Id.; Judy Mann, ConductUnbecoming,WASH. POST, July 7, 1993, at E13 (reportingthat an investigation of a female lieutenant"involved with a woman in her battalion"was triggeredwhen a "manwhom she had rebuffedturnedher in");Randy Shilts, Dismissed: ThePurging of Gay and Lesbian Troopsfrom the ArmedForces, L.A. TIMES MAG.,Apr. 25, 1993, at 10, 14 (speculatingthat an officer was investigated for being a lesbian "becauseshe rejectedthe passes of male crew members,"and reportingthat, unlike other women dischargedin the course of the same investigation,she was exoneratedby testimonyof her boyfriend that she was "a good sexual performer"). 204. Ortiz, supra note 196, at 1835, 1838 (emphases added).

asks not how a person comes to fall

February1994]

IMMUTABILITY CRITIQUING

555

imaginable,but it offers a parchedanalysisof social categoriesas laden with and meaning as sexual-orientation race groups.205For instance,Ortiz states that skin color is geneticbut that"race... is socially constructed," thatthe so constructedcategory "race"independently construesa naturalgiven, "skin in color."206But the culturalsystemof race interferes the very designationof some skin colors as "white" of "pinko-grey," (instead etc.) "orange," andothers as "black" of "mocha," (instead "brown," "yellow,"etc.);differentcultural systems of race notice differentgradations color.207Culturethus interferes of in whatthe geneticendowment skincolor is. And conversely,genetic of deciding race adjudicates cultural as race,inasmuch one cannotknow whose cultureAfrican-American cultureis, withoutresorting (secretly,if one wantsto keep faith with a commitmentto strong constructivism) a biological conceptionof to
race.208

of models of Precisely such an imbrication essentialistwith constructivist race appears in Gloria Anzaldua's invocations of mestiza identity.209 Anzaldua'smestizaheis the radicallyunstableidentityof womenliving on the borderbetweenthe United Statesand Mexico;betweenEnglish,Spanish,and Indianlanguages; betweensexualitiesandgenders; betweenracesso complexly mixed thatthey cannotbe disentangled.210 Anzalduia celebratesthe rich range of possibilitycreatedby this instabilityand porosityin an implicitinsistence thatmestizaidentityis not reducible any essence. But she also designatesits to essence precisely where pro-gay biological essentialismwould locate it: in historicized genes.211This is to establisha provisional, negotiable,contingent, and then insertindividualsinto it on the basis of strongessentialist category, characterizations. Ortiz's proposalthat the formation an identitycategoryproceedsindeof us pendentlyof the processof locatingindividual personsin it wouldrequire to of and ignore the complex interdependence constructivist essentialistelements in Anzaldua'smestizahe. Genetic race is part of the social meaningof that debatesaboutthe geneticelementsof sexualorientation identity,just as current
205. A similar reduction of descriptive complexity to serve analytic clarity occurs in Richard Mohr's resolution of pro-gay definitional debates in favor of essentialism. RICHARD MoHR,GAY D. IDEAS: ANDOTHER OUTING CONTROVERSIES 221-26, 234 (1992). Mohr proceeds from the empirical claim that "most social history is simply irrelevantto the question whetherthere is an innate biological drive in some people to have erotic encounterswith membersof their biological sex." Id. at 234. This claim describes a possible world, but not our world. Because he ignores the ways in which cultural definitions produce the need for biological ones, Mohr is inattentiveto the extent to which his own proposed "minimalor core" definition-"the desire for sexual relations with members of one's own biological sex," id. at 240-41-is politically loaded, adjudicatingthe boundariesbetween homosexual, in bisexual, and heterosexualin a way that many participants currentsexual politics disagree with. 206. Ortiz, supra note 196, at 1839. 207. See generally F. JAMES WHOIS BLACK? ONENATION'S DEFINITION DAVIS, (1991). 208. Walter Benn Michaels, Race into Culture:A Critical Genealogy of Cultural Identity, 18 CRITICAL INQUIRY 675-85 (1992). 655,
209. GLORIA ANZALDUA, BORDERLANDS/LA FRONTERA: THE NEW MESTIZA (1987).

210. Id. at 1-13, 53-64, 77-91. 211. Id. at 77 (describing mestiza consciousness both as the "consciousnessof the Borderlands" and as definitionallygenetic, a "confluenceof two or more genetic streams"producinga "moremalleable species with a rich gene pool").

556

STANFORD LAW REVIEW

[Vol. 46:503

identity have become important parts of gay, lesbian, and bisexual identity. If inductive and deductive processes of category formation can be interlocked, and if either, neither, or both processes can be (weak or strong) essentialist or constructivist, then the essentialism/constructivism debate has to deal with a myriad ways in which essentialism and constructivism can be interlocked in the articulation of complex social categories. B. Finding Common Ground

Essentialisms differ in the depth or "thickness" of their causality: Weak essentialism merely claims that some entity (here, homosexuality) is an "irreducible, unchanging [in the Aristotelian sense] and therefore constitutive" while strong essentialism gives that definitional core not only characteristic,212 a cause but a cause in nature. I propose that pro-gay essentialists litigate from a position of weak, not strong, essentialism. In all but its strongest forms, constructivism already coheres with such weak essentialism. 1. The lay of the land.

Different versions of constructivism differ in the depth or thickness of the thing they claim is caused by socially and historically contingent circumstances and activities. In a wonderfully clear essay differentiating constructivisms in the study of human sexuality, Carole S. Vance argues that constructivists "differ in their willingness to imagine what was constructed":213 At minimum,all social construction [to approaches sexuality]adopt the view thatphysicallyidenticalsexualacts may have varyingsocial significance and subjectivemeaningdepending how they are definedandunderstood on in differentculturesand historicalperiods.... A further of step in social construction theorypositsthateven the direction sexual desireitself, for example,objectchoice or hetero/homosexuality, not is in intrinsicor inherent the individual is constructed.Not all constructionbut ists takethis step;for some, the direction desireanderoticinterestarefixed, of this althoughthe behavioralform interesttakeswill be constructed prevailby and ing culturalframes,as will the subjectiveexperienceof the individual the social significanceattached it by others. to The most radicalform of constructionist the theoryis willing to entertain idea that thereis no essential,undifferentiated sexual impulse, 'sex drive' or 'lust', which resides in the body due to physiologicalfunctioningand sensation.... This position,of course,contrasts sharplywith more middle-ground constructionist which implicitly accepts an inherentsexual impulse theory which is then constructed terms of acts, identity,community,and object in
choice.214
212. Fuss, supra note 179, at 2. 213. Carole S. Vance, Social ConstructionTheory:Problems in the History of Sexuality, in HoWHICH MOSEXUALITY, HOMOSEXUALITY? (Dennis Altman, Carole Vance, MarthaVicinus & Jef13, 21 frey Weeks eds., 1989) (emphasis in original). 214. Id. at 18-19 (first emphasis added) (footnote omitted).

February 1994]

IMMUTABILITY CRITIQUING

557

each with a more Vance discernsfive differentforms of constructivism, claim as to what is constructedthan the last. The following expansive furtherexplicateVance's essentialism/constructivism spectrum. paragraphs Social meanings constructivism. This approachregards sexual object and are choice as fixed. The categorieshomosexual heterosexual good descripand Sexual object tions of all humanbeings,cross-culturally transhistorically. of choice in turndictatescertainpatterns behavior,which also, therefore,apin pear in all humansocieties withoutalteration theirform. But differentcultures and historicalcontexts give these differencesin object choice and the attendant meanings. Activirangeof activitiesdifferentsocial andexperiential in ties thatare thoughtnormalin one settingaretabooor criminalized another. affectionfor a motherto kiss Thus,in one cultureit mightbe a sign of maternal her infant son's penis, while in anotherthe same act would be deemed child abuse.215In the formerculture,this act wouldmakea womanfeel like a good This shift in levels of social approval mother;in the latter,like a transgressor. mightmeanthatin the first culturea lot of womenfellate theirsons, but some to in the second,who reallyrelishthis ritual,wouldbe tempted do it even at the cost of sanctions. The behavioralraw materialunderlyingthat differencein valuationdoes not vary: A personwho wouldenjoy this act in the firstculture for would enjoy it in the second because she comes hardwired female/male fellatio (of a particular kind). Behavioralconstructivism.This approach regardssexual object choice as fixed across culturesand historicaleras. In the most commonversionof this some people are predominantly disposedto have sex with people of approach, theirown sex; some arepredominantly disposedto have sex with people of the so-calledoppositesex. Thatmuchis determined nature.But the behavioral by patternsavailable for satisfying these predispositions vary across time and to place. A fortiori, the meaningsattributed sexual varietydiffer, as do the in feelings they arouse. John Boswell has describedthis position as "realist" the sense thatit regards sexualorientation "'real'apart as fromsocial structures
bearing on it,"216and distinguishes it from a nominalist view, on which "cate-

gories are only the names (Latin: nomina) of things agreed upon by Boswell himselfadoptsthe realistview that"'gay persons' [are] humans."217 . . . those whose erotic interestis predominantly directedtowardtheir own gender (i.e., regardlessof how conscious they are of this as a distinguishing Boswell's own work is a testimonyto the vast range of characteristic)."218 social constructive allows a realist,or a behavioral activitythathis assumption to existenceof constructivist, describe: Thoughhe assumesthe transhistorical whathe calls "gaypersons," workcontinually his noticesthe astounding range of social forms availableto themandthe volatilityof the social and normative
215. Cf William Davenport,Sexual Patterns and TheirRegulation in a Society of the Southwest BEHAVIOR 166-69 (FrankA. Beach ed., 1965); MarvinK. Opler,Anthropologi164, Pacific, in SEXAND cal and Cross-CulturalAspects of Homosexuality,in SEXUAL INVERSION (JuddMarmored., 1965). 108 216. Boswell, Revolutions, Universals, and Sexual Categories, supra note 92, at 17, 35. 217. Id. at 18-19. 218. Id. at 35.

558

STANFORD LAW REVIEW

[Vol. 46:503

meanings attributed to their erotic lives.219 His approach is even calibrated enough to register differing proportions of realist and nominalist views of "gay persons" in different ancient and medieval cultures.220 Behavioral constructivism as Vance configures it is rather specific about what aspect of sexuality remains essential: gender-of-object-choice. A weaker form of behavioral constructivism is distinguishable, however, in which some object is intrinsically and inherently preferred,but it is not necessarily congruent with sex or gender or even a particularkind of person.221 Weak behavioral constructivism acknowledges the powerful reality of sexual-orientation categories as we know them, but posits that some other form or forms of human variance are primary. It thus challenges us to imagine beyond the sexual-orientation categories homo- and heterosexual. Eve Kosofsky Sedgwick provides a suggestive list of object choices that might definitively distinguish people, but that our culture either ignores or sorts only roughly into categories differentiated by gender-of-object-choice: * Some people spenda lot of time thinkingaboutsex, otherslittle. * Some people like to have a lot of sex, otherslittle or none. * Many people have theirrichestmental/emotional involvementwith sexual acts thatthey don't do, or even don't wantto do. * For some people,it is important sex be embedded contextsresonant that in with meaning,narrative, connectedness and with otheraspectsof theirlife; for otherpeople,it is important theynot be; to othersit doesn'toccurthatthey that might be.222 It is difficult (though not impossible) to imagine that an array of individuals essentially committed to one of these object choices might tend to sort predominantly into the currently available diacritical categories, homosexual and heterosexual. Other sexual object choices-a taste for anal penetration or for sexual daringness or safety, or for reviving or avoiding certain childhood memories in sexual encounters-might be more likely to produce regular correlations, without installing gender-of-object-choice as the essential substrate of behavioral constructivism. And it is not difficult to imagine genetic causes of these differences, for instance, by genes causing development of more neurons in a particularzone of the body or, as biologists Anne Fausto-Sterling and Evan Balaban suggest for the repeatedly studied group of gay men with gay male
219. See, e.g., JOHN BOSWELL,CHRISTIANITY, SOCIALTOLERANCE AND HOMOSEXUALITY: GAY PEOPLEIN WESTERNEUROPE FROMTHEBEGINNING THECHRISTIAN OF ERA TO THEFOURTEENTH CENTURY(1980).

220. Boswell, Revolutions, Universals, and Sexual Categories, supra note 92, at 24-29. William N. Eskridge, Jr. provides anotherexample of behavioralconstructivismof the sort that animates Boswell's work. William N. Eskridge,Jr., A History of Same-SexMarriage, 79 VA. L. REV.1419 (1993). Eskridgeassumes that the same-sex-ness of same-sex marriagehas a stable cross-cultural unity, providing an unmoving platformfrom which to admirethe spectaculararrayof its behavioraland ideological manifestationsin various times and places. Id. 221. Edward Stein articulatesthis model of sexual orientationas "sophisticatedessentialism." EdwardStein, Conclusion: The Essentials of Constructionism and the Constructionof Essentialism,in
AND THESOCIALCONSTRUCTIONIST FORMSOF DESIRE:SEXUALORIENTATION CONTROVERSY 325, 332-

38 (Edward Stein ed., 1990).

222.

SEDGWICK, EPISTEMOLOGY, supra note 92, at 25.

February 1994]

IMMUTABILITY CRITIQUING

559

to siblings, "anygene thatmightincreasethe tendencyof brothers psychologiwith one another."223 Even in our culture,it is not clearwhether cally identify the identityhomosexualor heterosexual wouldoffer a personessentiallydiffera entiatedby such characteristics more suitablemeans of exercising it. But neitheris it clear that a patternwould not emerge.
Gender-of-object-choice constructivism. In this approach, sexuality-the

of patterned organization sex driveor sexualappetitecapacityfor a coherent, is predetermined invariant, sexualobjectchoice, the associatedbehavand but ioral forms, and their social and experientialmeaningsremainavailablefor social adaptation.At this point constructivism that abandonsthe assumption the terms"homosexual" "heterosexual" and necessarilydescribehumansexual of open, most historians predispositions.Thoughthis categoryis conceptually of kind sexualitywho pursueconstructivism a morethoroughgoing thanbehavioral constructivism constructivist developthe stronger hypothesiswhichI dub sexualityconstructivism. as of Sexuality,Vol. I, but decliningto extendconstructivism far as he did in thatwork,distinguishes betweena rawphysicalcapacityfor eroticpleasureand the organization thatcapacityinto a coherent,patterned of elementof the self. The formeris part of being human,but the latteris the productof complex historicalcontingencies as which,once in place,areas inescapable the capacity for orgasm. Objectchoice, behavioral social meaning,and sexual repertoire, emotionall fall withinthe domainof sexuality,and areproduced its means. by David Halperinprovidesan exemplary, even pyrotechnic, displayof whatthis which proclaimsthe view thathomosexuality quite a recentinvention.224 is Halperinstates that, "[u]nlikesex, which is a naturalfact, sexuality is a culturalproduction:It represents appropriation the humanbody and of the of its erogenouszones by an ideologicaldiscourse."225 Sexualityis relativelyautonomousfrom gender,and "[t]hatis preciselywhat makes sexualityalien to the spiritof ancientMediterranean where"sexualtypologiesgenercultures," ally derivedtheir criteriafor categorizing people not from sex but from gender."226 Halperin does not question that "there really are, nowadays, homosexualand heterosexual people .. .: they reallydo desirewhatthey do, andthatis afact aboutthem."227 this is a weakessentialistformproduced But causes. by social ratherthannatural Sex constructivism. This may be the most thorough view of constructivism in the areaof sexualityyet articulated.It assumesthatthe sheerrecognition of certainbodily sensationsas sexualis constructed. This is not merelyto say, for "sex drive"would be a differinstance,thatliving in a culturethat"implants"
223. Fausto-Sterling& Balaban,supra note 111. 224. David M. Halperin,One HundredYearsof Homosexuality,in ONE HUNDRED YEARS OF HOMOSEXUALITY, supra note 4, at 15-18, 24-27. 225. Id. at 25 (footnote omitted). 226. Id. 227. Id. at 43 (emphasis in original).

Sexuality constructivism. This approach,inauguratedby Foucault's History

approach can do in his book One Hundred Years of Homosexuality, the title of

560

STANFORD LAW REVIEW

[Vol. 46:503

ent thing thanliving in a culturein which sex originatesin an "appetite."It is to insist thatculturesuppliesthe very termsfor understanding bodily sex, in or between persons,as distinctfrom other modes of physical configuration, achis tion, or sensation. Halperincrisply distinguishes more moderateposition from Foucault'ssex constructivism:
I should point out that Foucaulthimself decisively abandoned distinction the between sex and sexuality,as I have drawnit. Not only is Foucault'sfinal deniesthat"sex"is conceptionof "sex"muchless positivistic(he categorically a biologicalfact), but his own understanding the distinction of between"sex" and "sexuality" reversesthe sequencepostulated here [in OneHundredYears on of Homosexuality]:"[S]exuality," his view, arises in the eighteenthcentury and eventuallyproduces"sex,"as an idea internalto its own apparatus, only in the nineteenth century.228

The equivalentof "sex constructivism" the study of gender is, perhaps, in to the questionwhethershe had a vagina,that she MoniqueWittig's response did not.229
2. Choosing a place to stand.

Each constructivist theoryexcept the last renderssome aspectof sexuality essential. The rangeof potential commongroundis nevertheless rather limited. Most pro-gayessentialistsare deeply committedto denyinggender-of-objectchoice constructivism, the quitelogical reasonthatgender-of-object-choice for constructivismis intrinsicallyinconsistentwith social meaningsconstructivism.230Neitherside shouldseek to legitimateits claim on these points in the context of litigation. But there is an alternative:behavioralconstructivism. The view that objectchoice may be essentialand indeed,biological,and that the relatedbehaviorandits meanings belongto the domainof cultureis already essentialist. Behavioral constructivism comes in two versions,each distinguished the by with which it describesits most extensiveessentialistelement,obspecificity constructivism is ject choice. Both strongbehavioral (gender-of-object-choice into is essential,butits translation behavioral patterns not) andweakbehavioral constructivism (object choice is essential but is not necessarilyinflected by offer some common groundaccessible to most pro-gay essentialists gender) and constructivists.Two principles selectionindicatequitedecisively,howof constructivism moresuitableto prois ever, thatthe weakerformof behavioral and alike. First,it offersless offense to gay constructivists pro-gayessentialists the principle that neither pro-gay essentialists nor pro-gay constructivists
228. Id. at 162 n.47 (construingFOUCAULT, supra note 4, at 152-57). 229. JUDITH GENDER TROUBLE: FEMINISM THE AND SUBVERSION IDENTITY n.54 OF BUTLER, 157 (1990). 230. Halperinconcedes the mutualantagonismof gender-of-object-choice constructivismand behavioral constructivism. In an interview in which he seems most concernedto assert gender-of-objectchoice constructivism,and not the broaderclaim that sexuality is constructed,he quite rightly states that any conclusive and unassailablescientific proof thathomosexualityas such is biologically caused would defeat his project entirely. David M. Halperin, "Homosexuality": CulturalConstruct(An Exchange A
with Richard Schneider), in ONE HUNDREDYEARS OF HOMOSEXUALITY, supra note 4, at 41-49.

February 1994]

IMMUTABILITY CRITIQUING

561

shouldbe required enterinto litigationwith an argument requiresthem to that to say somethinginconsistent with theircommitments.And second, it is simply more consistentwith the rich varietyof humansexualitiescross-culturally and with the evidence of modem science itself. A compromisestoppingat strongbehavioral constructivism would require to pro-gay constructivists retreatto the position that, while an essential and possibly naturaldifferenceamong humansdirects some of us to seek erotic contactswith personsof the same sex, and othersto select personsof the socalled oppositesex, cultureprovidesthe humancharacterof these essentially differentsexualities,and the social organization theirrealization. Settling for on strong behavioralconstructivism would not be a total rout for construcconstructivism be given quitea pronounced can tivists, inasmuchas behavioral constructivist to spin. Thosepre-committed same-sexcontactsmightbe pederasts, sodomites, mollies, berdache,inverts,homosexuals,gay men, lesbians, queers,and so on. People's subjectiveexperienceof sexuality,andthe behavior they undertake supportit, would be radicallycontingenton the identity to offered by their culturefor personsof their object choice and on their own for construcopportunities alteringor shapingthe optionson offer. Behavioral tivism is alreadya respectable JohnBoswell is pro-gayconstructivist position. largely (thoughnot entirely)rightwhen he denies being an essentialist.231 It is precisely Boswell's maintenance transhistorical of humancategories "homosexual" "heterosexual" manypro-gayconstructivists objecand that find however. Ourdifficultyon this pointis not trivial: It is that such an tionable, a invokedto explain assumption, naturalizing humancharacteristic by currently sometimesto justify) a social hierarchy make the role of power in (and may that invisible or obscure. Pro-gaycongeneratingand maintaining hierarchy structivists who take this positionshouldnot be required sharein litigation to with pro-gayessentialists the cost of positivelycontradicting point at a strategy this important theirdistinctiveanalysis. to Weakbehavioral constructivism a poses constructivists muchless strenuous case of conscience. To subscribeto this position,which implicitlyrecognizes the possibilitythatdifferentindividuals into homo-andheterosexual sort orientationswith varyingdegreesof natural we aptitude, would merelybe askedto agree with William E. Connolly,that "[t]o treatidentity,as I do, as a site at which entrenched definitionsis not dispositionsencounter socially constituted to insist that any such definitionwill fit every humanbeing equally well or
badly."232

Strongessentialistswho seek to rely on scientificfindingsshouldlike the weak versionbettertoo, becauseit almostexactlymatchesthe findingsto date and scientists'predictions aboutwhatthey will find in the future. Indeed,scientificresearch datesupports stronger to no essentialist claimthanweakbehav231. Boswell, Revolutions,Universals,and Sexual Categories,supra note 92, at 35 (assertingthat "no currenthistoriansconsciously defend an essentialist point of view"). Boswell is not entirely accurate in denying all essentialism here, inasmuchas he maintainsan essentialist view of gender-of-objectchoice. 232. CONNOLLY, supra note 4, at 163 (emphasis added).

562

STANFORD LAWREVIEW

[Vol. 46:503

ioral constructivism-a fact thatputs pro-gayessentialistswho invoke science on a virtualestoppel. As PartIII abovesuggests,pro-gayessentialists have not and "heterosexual" the are demonstrated why the categories "homosexual" ones that will ultimatelycorrelatebest with any biological causes of sexual object choice that science mighteventuallyidentify. Pro-gayessentialismhas to date been silent when faced with the question,"Whatreason have we to and think that the categories'homosexual,''heterosexual,' perhaps'bisexual' will be the categoriesexplainedby an advanced[essentialist] theoryof sexual orientation?"233 Indeed, RichardPillardhas gone on record agreeing with weak behavioralconstructivism.234 It is no surprisethat a prominentpro-gay essentialistresearcherwould reachthis conclusionwithoutan apparent sense of self-contradiction. their In of lesbians,Bailey andPillard(workingwith a new teamof researchers) study the acknowledge"the power of genetic studies such as ours to demonstrate and illuminatethe natureof environmental influences."235 This importance power is confirmedin the Hamerstudy,in which seven out of forty gay male subjectsdid not have the specific gene sequenceassociatedwith homosexual in orientation men.236Hamerand his team concludethat "[g]iventhe overall that complexityof humansexuality,it is not surprising a single genetic locus does not accountfor all of the observedvariability," recommendfurther and researchto identify "additional or environmental, genes experiential,or cultural factors(or some combination these) thatinfluencethe development of of male sexual orientation."237 a rarecommentto the media,Hamerhas given In a similarcaution: "[W]hat have foundis a linkagebetweena small region we on one chromosomeand sexual orientation men, and what that suggests is in thatpart of whethera personis gay or heterosexual influenced the genes is by that they inherit. Howeverwe have not foundthe gene, whichwe don't think exists, for sexual orientation."238 These are compellingreasons for pro-gayessentialistsand pro-gay constructiviststo conductcommon projects-particularlylitigationon behalf of
233. Stein argues that the answer is "none": "[A] viable version of essentialism should not assume our standardcategories of sexual orientation." Stein, supra note 221, at 337. 234. Stein, supra note 147, at 103. In response to Stein's claim that "homosexual"and "heterosexual" are merely "folk categories"and that biological researchcan at best be hoped to confirm weak behavioralconstructivism,Pillardsaid: 'That puts us a step closer than I thoughtwe were. You suggest an interestinglocution: 'folk categories.' That's exactly what they are, because we don't really know in any specific way what behavior genes code for." Id. (quoting Pillard) (emphasis added). 235. Bailey et al., supra note 105, at 222. 236. Hamer et al., supra note 1, at 325. 237. Id. at 325-26 (emphasisadded);see also Fausto-Sterling Balaban,supra note 111, at 1257. & and Balabanstatedthat a "gene affecting sexual orienCommentingon Hamer's study, Fausto-Sterling tation in some segment of the male populationmight do so very indirectly. For instance, any gene that might increase the tendency of brothersto psychologically identify with one anothermight influence their similarity in such mattersand would be picked up in the present study." Id. 238. Tim Radford, Your Mother Should Know, THEGUARDIAN, 17, 1993, at 23 (quoting July Hamer's comments broadcast on BBC radio) (emphasis added) (paragraph break omitted); see also Hamer et al., supra note 113, at 2065 ("We did not say that Xq28 'underlies' sexuality, only that it contributesto it in some families. Nor have we said that Xq28 representsa 'major'gene, only that its influence is statistically detectable in the selected populationthat we studied.").

February 1994]

IMMUTABILITY CRITIQUING

563

as gay men, lesbians,bisexuals,andqueers-from the positionI have described weak behavioral constructivism.The exactionsof equalprotection analysisdo not appearto requirethis level of specificity,however. As the following general review of equalprotection arguments suggests,these projectsmay require to constructivism simonly an agreement litigatefroma positionof behavioral pliciter (and thus to abandon all strong essentialist argumentsbased on biology). C. Arguingfor EqualProtectionon Common Ground The recenthistoryof gay equalprotection plaintiffsincludesmale inmates who "appear[ weak, smallor effeminate" the sheriffsat the ShelbyCounty to ] Jail and are therefore in the segregated special"gay"cells;239 shamedandconfused JamesMiller,confessingto militaryinquisitors he had committeda that "homosexual act" and renouncing the it;240 discreteMajorieRowland,quietly and indicatingto a coworkerthat she was bisexual;241 the bold Miriambenand Shalom,consistently comingout as a lesbianto hermilitarysuperiors suing to invalidateher dischargefor more thanten years.242These plaintiffsdo not even have a sexual orientation common. Whatthey shareis a life-altering in rejectionimposedbecausethey enteredthe zone in whichthe social representation of same-sexdesirebecamea meansof institutional ordering.The essence of the class thatneeds protection not, then,in its members:It is in the interis actionsamongclass membersandbetweenthemandothersaboutthe meaning and value of same-sexerotic desire. Underthese circumstances, strongessentialistmodels of sexual orientation are likely to exhausttheirresourceslong beforea fully adequate groups analysis of anti-gaydiscrimination be built. And conversely,at each stagein the can explication of equal protection analysis-from the determinationwhether whether heightenedscrutinyis possible afterHardwick,to the determination is due, to the assessment(underheightened heightenedscrutiny scrutinyor rationalbasis review) of the ends and meansof anti-gaydiscrimination-behavioral constructivismenriches the conceptual,analytic, and representational resourcesof pro-gaylitigators. of Indeed,at the threshold suspectclass analysis,strongessentialistmodels of sexual-orientation identityarea positiveimpediment.Severalfederalcourts have held thatHardwick foreclosesheightened becausecriminalizable scrutiny is "sodomydefinesthe class"for whichprotection sought.243 Thoughlitigators
239. Gay Inmates of Shelby County Jail v. Barksdale,No. 84-5666, 1987 WL 37565, at *1 (6th Cir. June 1, 1987). 240. Beller v. Middendorf,632 F.2d 788, 802 (9th Cir. 1980), cert. denied, 452 U.S. 905 (1981). 241. Rowland v. Mad River Local Sch. Dist., 730 F.2d 444 (6th Cir. 1984), cert. denied, 470 U.S. 1009 (1985). 242. See benShalom v. Secretaryof the Army, 489 F. Supp. 964 (E.D. Wis. 1980) (challenging benShalom's discharge for publicly acknowledging her homosexuality);BenShalom v. Marsh, 703 F. Supp. 1372 (E.D. Wis. 1989) (challengingArmy's refusalto reinstatebenShalomon the same grounds), rev'd, 881 F.2d 454 (7th Cir. 1989), cert. denied, 494 U.S. 1004 (1990). 243. Padulav. Webster, 822 F.2d 97, 103 (D.C. Cir. 1990) (emphasis added);see note 31 supra.

564

LAWREVIEW STANFORD

[Vol. 46:503

it have the option of attackingHardwickdirectly,244 is most likely that they it.245 Litigatorscan will preferin equal protectioncases to "litigatearound" its refute the extensionof Hardwick attacking legal premise,that due proby cess precedentsapply to equal protectioncases.246 And they can attackits definitionof the class. constructivists essentialistsalike should and To make the latterargument, the shed unnecessarydefinitionalclaims that undermine distinctionbetween which andthe class of homosexuals.Social meaningsconstructivism, sodomy in its naturalcategories"homobehavioral installs a predetermined repertoire sexual" and "heterosexual," actuallysupportsthe extension of Hardwick. A fortiori, unmediated strongessentialismdoes the same: It posits the same unthat brokenequationbetweenhomosexualsand theirbehavioral repertoire diswill work ables social meanings constructivism.Behavioralconstructivism only if it acknowledgesthat,in our culture,many elementsof sexual-orientato of with andindependently behavior constitute tion identityoperatein tandem and to populatethe classificationhomosexual. is The constructivist view that sexual orientation mutablebecauseof slipof pages and rearrangements desire, fantasy, behavior,private identity, and refutation a definitionof homosexuof publicidentityis possiblythe strongest shouldbe able to agreewith ality thatmakessodomyits essence. Essentialists this view as long as it falls shortof its outermost reach,thatno one's subjective entertained sexualorientation reflectsthe gender-of-object-choice by accurately his or her desire and fantasy. Unlikepro-gayessentialistuses of the argument fromimmutability, whichat leasttheoretically excludesome gay men,lesbians, zone of heightenedscrutiny,a pro-gayarguand bisexualsfrom the protected the ment that distinguishes Hardwick emphasizing varietyof gay, lesbian, by class membertoo. the bisexual,andqueeridentitiesembraces most unwavering resources for Behavioral constructivismis also rich in representational that heightenedscrutinyis warranted.On a probuildingpositive arguments allows advocatesto focus on cess-based analysis, behavioralconstructivism wherethey pinch: at the closet door.247The process-based identitydynamics argument emphasizesthatthe volatilitiesandfixities of publicidentitymakeit exceedinglydifficultfor gay men, lesbians,bisexuals,andtheirfriendsto enter
244. Gay rights litigatorswho can file their cases in states with sodomy laws and well-developed of privacy rightshave been engaged in a state-by-staterepudiation Hardwick. See, e.g., Commonwealth v. Wasson, 842 S.W.2d 487 (Ky. 1992) (holding that criminalstatuteproscribingconsensual, homosexual sodomy violates privacy and equal protectionguaranteesof KentuckyConstitution);Michigan Org. for HumanRights v. Kelly, No. 88-815820 CZ (Mich. Ct. App. July 9, 1990) (holding that any application of state sodomy and gross indecency statutesto consensual, private conduct violates state privacy guarantee). Federallitigants who do not have real clients, or whose clients do not mind knowing at the outset that they'll lose, might also consider seeking a reversal of Hardwick. These are now the only routes open to litigators who want to avoid making an implicit concession that Hardwick is in some sense legitimate. 245. Cain, supra note 33. 246. See Cass R. Sunstein, Sexual Orientationand the Constitution: Note on the Relationship A Between Due Process and Equal Protection, 55 U. Cm. L. REV.1161, 1167-68 (1988) (arguingthat an unfavorabledue process holding should have "no consequences for an attackbroughtunder the Equal ProtectionClause"). 247. Halley, supra note 22.

February 1994]

IMMUTABILITY CRITIQUING

565

fully into politicaldebate. Too manygay men, lesbians,and bisexualsremain silent so that they can keep theirjobs; too many sympatheticheterosexuals remainsilent so that no one will supposethey are queer.248Social-meanings becauseit assumesthatpublicidentityunivocallyandtransparconstructivism, lacks the conceptual to apparatus deently reflectsintrinsicsexualorientation, because it is scribe these political impediments. Behavioralconstructivism, hospitableto the claim thatpublichomosexualidentityis affectedby cultural forces, can supportthe furtherclaim that the resultingsocial patterns deprive access to the politicalprocess. pro-gayadvocatesof unimpeded Process-basedtheorieshave been subjectedto fairly stiff theoreticalcrievalutique,however,on the groundthatthey do not actuallyavoidsubstantive ation.249In light of this critique,behavioralconstructivism possesses a key It paintsa richandvariedpictureof the ways in whichofficial antiadvantage: the enliststhe statein reproducing social and lexical suborgay discrimination dinationof gay men, lesbians,and bisexuals. Many of the slippagesthat become visible on this model are extremelypainfulto behold,and may formthe of assessment anti-gaydiscrimination. basis for a coherent,negativenormative And judges who thinkthey do not know any homosexualsneed to be given a "thick"picturenot of homosexualsbut of homosexualidentityif they are to the understand social and politicaltextureof anti-gaydiscrimination.250 All of this assumes that the goal of equal protectionlitigationis to get heightenedscrutiny. Of course, litigatorscan and usually must settle for rational basis review, but that is no longer equivalentto losing. Three federal judges have held that the military'sformeranti-gaypolicy fails rationalbasis review.251Moreover,both heightenedscrutinyand rationalbasis review re248. Id. at 944-46. 249. See note 20 supra. 250. Psychologists GregoryM. Herekand Eric K. Glunt have found that heterosexualswho know that they know gay men are more likely to entertainfavorable attitudes towards them. Gregory M. Herek & Eric K. Glunt, InterpersonalContactand Heterosexuals'AttitudesTowardGay Men: Results from a National Survey, 30 J. SEXRES.239 (1993). These findings take on an alarmingcast when one acknowledges, with Judge Posner, that "judges know next to nothing about the subject [of sexuality] beyond their own personalexperience, which is limited, perhapsmore so than average, because people with irregularsex lives are prettymuch (not entirely,of course) screenedout of the judiciary." POSNER, SEX AND REASON, supra note 59, at 1. One solution to the problemof judicial ignoranceof the realities of anti-gay discriminationmight be to reverse the screen, and ensure that some gay men, lesbians, and bisexuals become federaljudges. StephenReinhardt, KeepingJustice in the Closet: Judge Argues Time Is Right to BreakDown Barriersfor Gays, DAILY Dec. 6, 1993, at 4. But that approach,if it works at J., all, will work very slowly. Another is to use litigation as a setting for the first-personnarrativesof lesbians and gay men. Fajer, supra note 97. But that method presupposes the speaker's own unmediated access to the truth of her "experience"-a presuppositionthat has been convincingly challenged. See Daniel A. Farber& Suzanna Sherry, Telling Stories Out of School: An Essay on Legal Narratives, 45 STAN.L. REV.807 (1993); Scott, supra note 95. The approachsuggested here is that litigatorsgive judges not the equivalentof but a substitute knowing someone gay: a nuanceddepicfor tion of the social, political, and representational dynamics in which sexual-orientationidentities are negotiated and ranked. 251. See Steffan v. Aspin, 8 F.3d 57 (D.C. Cir. 1993) (rejecting as irrationalSecretary of Defense's justifications for formermilitaryban that it eliminatedmemberswith a propensityto engage in illegal conduct; that it preservedthe morale or privacy of heterosexualsoldiers; that it preventedtransmission of the HIV virus; or that it reducedthe incidence of blackmail), vacatedfor reh'g en banc 62 U.S.L.W. 2309 (D.C. Cir. Jan. 7, 1994); Dahl v. Secretaryof the United States Navy, 830 F. Supp. 1319

566

LAWREVIEW STANFORD

[Vol. 46:503

ends and means-that is, of the purposes quire examinationof governmental To andmethodsof anti-gaydiscrimination. the extentthatanti-gaydiscrimination is sometimesjustifiedon an essentialistand sometimeson a constructivist model of sexualorientations, is impossibleto determine priori whether it a progay advocates will encounterany temptationto invoke strong essentialist claims. In recentmilitarycases, the government soughtto justify its antihas models: gay policy by an unstableamalgamof essentialistand constructivist The military,it is said, shoulddefer to the strongfeeling amongheterosexual men areessentiallyunlikethem,and to the same male soldiersthathomosexual soldiers'terribleanxietythattheirproximityto gay men erodestheirown heterosexuality.252 If the military cases are any guide, the temptation to refute

will modelsof sexualorientation oftenbe mutedby the using strongessentialist of more constructivist ones. In most cases, the focus countervailingutility of shouldfall not on the nature the peoplediscriminated againstbut on the idea of themformedin the policiesandprograms disadvantage that them,andon the of all anti-gaylegal policy, that the state should serve as a implicit premise thoseideasinto social reality.253 facilityfor constructing Strongpro-gayessentialism draws attentionto the former;behavioral constructivism providesthe for understanding latter. the conceptualapparatus

(M.D. Fla. 1993) (rejectingdefendant'srationalesfor formermilitarypolicy as motivatedexclusively by prejudice, and granting plaintiff summaryjudgment on equal protectionclaim after rational basis review); Meinhold v. Dep't of Defense, 808 F. Supp. 1455, 1458 (C.D. Cal. 1992) (holding that "[t]he Departmentof Defense's justificationsfor its policy banninggays and lesbians from militaryservice are based on culturalmyths and false stereotypes" have no rationalbasis). These decisions indicate that and failure to obtain heightened scrutinyis not necessarily fatal to a gay or lesbian plaintiffs case. 252. See Dahl, 830 F. Supp. at 1319 (summarizingDepartmentof Defense rationalesfor anti-gay policy); Steffan v. Cheney, 780 F. Supp. 1, 11-13 (D.D.C. 1991) (same). 253. Several courtshave held thatthe Department Defense's invocationof heterosexualtroops' of recoil from homosexual proximity gives official form to private prejudice in violation of Palmore v. Sidoti, 466 U.S. 429, 433 (1984) (holding that "privatebiases may be outside the reach of the law, but the law cannot, directly or indirectly,give them effect"). Dahl, 830 F. Supp. at 1331-32; Steffan,780 F. Supp. at 12-13. There remains an avenue aroundHardwick and to heightened scrutiny that does not necessarily depend on an identity claim: The argumentthat discriminationon the basis of sexual orientationis really discriminationon the basis of sex. That was the basis of the Hawaii Supreme Court plurality opinion in Baehr v. Lewin, 852 P.2d 44 (Haw. 1993), summarizedin PartI above. The Baehr plurality understoodthe argumentas a purelyformalone: Hawaii would not let a woman do what a man was free to do, namely, marrya woman. The justices explicitly stated that plaintiffs' professed homosexuality was irrelevantto their logic. Id. at 53 n.14, 58 n.17. In separatearticles,Andrew Koppelmanand Cass Sunstein have offered more socially-descriptiverationales for holding that sodomy laws violate the constitutionalbar on sex discrimination. Koppelmanemphasizes that "theirfunction is to maintainthe polarities of gender on which the subordinationof women depends,"Andrew Koppelman,Note, The L.J. 145, 147 (1988), while SunMiscegenationAnalogy: SodomyLaw as Sex Discrimination,98 YALE stein emphasizes the rejectionof caste implicit in Loving v. Virginia. Cass R. Sunstein,Homosexuality and the Constitution,in LAWS NATURE: & SHAPING PREFERENCE THE AND FAMILY SEX, (David Estlund & MarthaNussbaum eds., forthcoming) (manuscriptat 1-3, on file with the Stanford Law Review). Neither the formal nor the social-descriptiveargumentis constrainedby essentialist or constructivist versions of sexual-orientationidentity, though the richerclaims advancedby Koppelmanand Sunstein call for a separateassessment of various theories of how sexual orientationand gender intersect.

February 1994]

IMMUTABILITY CRITIQUING
CONCLUSION

567

The argument from immutability respondsto a particularly contemptuous and dismissive form of anti-gayanimuswith elegant simplicityand plangent availableto appeal. It also works. Indeed,it oftenis the onlyeffectiveresource theirparents,coworkers, gay men, lesbians,andbisexualsseekingto persuade and neighborsthat they can love someone of the same sex and remainfully human. Moreover,for most of the gay children,workers,and neighborswho use the argumentfrom immutability these settings, it is absolutelytrue: in They can't changetheir sexual orientation. Whenthe argument fromimmutability leaves those settingsandbecomesa and however,the termsby whichwe shouldjudge its plausibility legal strategy, effectivenessshift. While it may be entirelyresponsiveto the particular form of personalcriticismfaced by many gay men, lesbians,and bisexuals("Why don't you just change?"),it is not fully or even coherentlyresponsiveto the forms of anti-gay argumentused to justify state-sponsored discrimination againstall of us. Threenew elementsneed to be takeninto account. First,anti-gaypublicpolicy is complexandflexible, andfinds ways to justhat orientation many,most,or in tify itself even on the assumption homosexual all its bearersis immutable.Second,the reasonswhy the state shouldnot discriminateagainstgay men, lesbians,and bisexualsare differentin important ways from the reasonswhy parentsshouldnot thinkill of theirgay children. Suspectclass analysis(whengiven its best reading)asks whetherthe resources of the state are being used to enforce,confirm,and validatesocial hierarchies. The argument fromimmutability neverattained preeminence suspect has the in class analysisthat some pro-gayadvocatesattribute it becauseit carriesso to little waterin thatanalysis. And third,the argument fromimmutability, when advancedon behalf of a complex movement,many of whose memberscan changesome aspectof theirsexualitythatis targeted anti-gaypolicy, is less by becomes directlyresponsiveto the problemwe face. Moreover,the argument burdened with an ethicalproblemit does not have when used privately:When fromimmutability beforea courton behalf pro-gayadvocatesuse the argument of gay men, lesbians,and bisexuals,they misrepresent us. Even worse, when the pro-gayargument fromimmutability annexesrecent scientificfindingsto bolsterits empiricalclaim thathomosexualorientation is it immutable, becomes simply incoherent.As I have indicatedin my summaries of the internalcriticismof behavioral sciengeneticsand of the particular tific articlesthat advocatesof the pro-gayargument from immutability have considered most useful,biologicallycausedtraitscan change;thereis as yet no has proof that humansexual orientation a biologicalcause; and even if a biowere eventuallyidentified,the conlogical cause of humansexual orientation and ceptuallydistinctquestionwhetherit causeshomosexuality heterosexuality would remainoutstanding. Failureto emphasizethatlast pointis perhaps mostdisturbing the featureof the way in which pro-gayadvocatesof the argument from immutability have introduced scientificfindingsinto the broader the culture. Of course,one day

568

STANFORD LAWREVIEW

[Vol. 46:503

science may find thatthe sexual-orientation categoriesof our cultureare natuis buttodaytheiradequacy a politicalquestion-in fact the very samepolitral, fromimmutability ical questionthatdividespro-gayadvocatesof the argument and many of its pro-gaydetractors.Proponents the pro-gayargument of from a have covertly withdrawn political questionfrom the political immutability sphereby falsely implyingthat science has answeredit. Retrievingthat political questionfor social and political debatealso provides a betterbasis for antidiscrimination for analysis. A prerequisite the latter is a representation homosexualsand heterosexuals, homosexualityand of of that (or heterosexuality, does not divide pro-gaycommunities divides them as little as possible). The controversy over the pro-gayargument fromimmutabilbut ity is a controversybetweenessentialismand constructivism, it has been conductedwith a dangerously thin notionof the complexityof the logical entailmentsof those theories. It has assumedthat an essentialistview of sexual orientation committed to a claimthathomosexuality, such,is the product one as of nature;and thata constructivist view committedone to a claim thathomosocial artifact, sexualityis an entirelycontingent subjectto changeat whim. It has assumedthatessentialism constructivism logical opposites. A more and are carefulexplanation whatis involvedin essentialism constructivism of and indicates thata weakformof essentialism-holding thata given entityis subjectto consistentconventional definition-actually subtends manyformsof construcvaries quite widely in the depthof its contintivism, and that constructivism gency claims. the thus Disaggregating variousforms of essentialismand constructivism indicatesthatthey are actuallyintertwined all but the most extremeends of in their own ranges, and offers the possibilityof finding a conceptuallocation from which pro-gayessentialistsand pro-gayconstructivists frame legal can that avoid the argument from immutability while not contradicting arguments its empiricalpredicate.Recent sexualitystudiesin history,anthropology, and culturalstudiesvary more or less continuously the depthof theirclaim that in sexual-orientation categoriesare sociallycontingent.All but the most extreme formsof constructivism retaina weak essentialistview of homosexualorientation. And all but the most extremeforms of essentialismpermitattentionto focus on the social meaningsof sexual desire, behavior,and self-description. My proposalis thatpro-gayadvocatescan formlitigationstrategyat the resultof and Therearedistinctlegal ading intersection essentialism constructivism. to this compromise, moreover:Strongessentialismactuallysupports vantages courts in holding that Hardwickforecloses heightenedscrutiny;the justifications for heightenedscrutinyarise from preciselythe social and politicalelements of sexual-orientation identitythat constructivist analysisidentifies;and rationalbasis analysis,becauseit involves examination the discriminator's of reasonsfor discriminating, looks not at what gay men, lesbians,and bisexuals really are, but at what we are thoughtto be. Litigatingon commongroundis thus not only the rightthing to do-it is also more likely to work.