T

spacioharbosharemanychildhStatue

A

The

s a setinternat

ousness,tor. But ths one feay seabornhood. TheofLiber

A

Intthe

S

tting fortional arrtheopenhe plaqueature witne anceste plaquertybearth

terne Sc

tephen

r an initrivalshallair,andte that greth the gretors, incle on Kenhesamei

nal carle

n Jay G

tial welcl atKennthesymboeets airboeat lady wluding alnedy's winscriptio

Braet W

Gould

come tonedyAirpoloffelloorne immwho gracll my grawall andon:Emma

andW

a newportpalesowshipinmigrantsced the aandparenthe pedeaLazarus'

d of

home, ts before tNewYorof our timarrival ofnts in thestal of t'spoemT

f

thetherk'smeso

heirtheThe

New Colossus—but with one crucial difference. The airport versionreads:

Give me your tired, your poor, Your huddled masses yearning tobreathefree...Sendthese,thehomeless,tempest‐tosttome:Iliftmylampbesidethegoldendoor.

Onemightbe excused for supposing that the elision representsalargeandnecessaryomissiontofittheessenceofalongerpoemontoasmallishplaque.Butonlyone line,easilyaccommodated,hasbeencut —and for a reason that can only reflect thoughtless (as opposed tomerely ugly) censorship, therefore inviting a double indictment onindependent chargesof stupidityandcowardice. (Asamemberof thelastpublicschoolgenerationtrainedbyforcedmemorizationofaholyhistoricalcanon,includingtheGettysburgAddress,thePreambletotheConstitution,Mr.Emersonontherudebridgethatarchedtheflood,andMs.Lazaruson thebig ladywith the lamp, I caught thedeletion rightaway and got sufficiently annoyed to write aNew York Times Op‐Edpieceacoupleofyearsago.Obviously,Iamstillseething,butatleastInowhavetheperversepleasureofusingthestoryformyownbenefittointroduce this essay.) I therefore restore themissing line (alongwithEmmaLazarus'srhymingschemeandsyntax):

Thewretchedrefuseofyourteemingshore.

Evidently, the transient wind of political correctness precludessuchaphraseas "wretchedrefuse," lestanyvisitor read the line tooliterally orpersonally.Did the authorities at ourPortAuthority everhear about metaphor and its prominence in poetry? Did they everconsiderthatLazarusmightbedescribingthedisdainofaforeignelitetowardimmigrantswhomwewouldwelcome,nurture,andvalue?

This story embodies a double irony that promptedmy retelling.We hide Emma Lazarus's line today because we misread its trueintention, and because contemporary culture has so confused (andoften even equated) inappropriate words with ugly deeds. But theauthorities of an earlier generation invoked the false and literalmeaning—theidentificationofmostimmigrantsaswretchedrefusetoaccomplish a deletion of persons rather than words. That is, thesupposedgeneticinferiorityofmostrefugees(aninnatewretchednessthatAmericanopportunitycouldneverovercome)becameaneffectiverallying cry for a movement that did succeed in imposing strongrestrictions on immigration beginning in the 1920s. These laws,strictly enforced despite pleas for timely exceptions, immuredthousands of Europeans who sought asylum because Hitler's raciallaws had marked them for death, while our quotas on immigrationprecluded any addition of their kind. These two stories of pastexclusionandtruncatedpresentwelcomesurelyillustratethefamiliarhistorical dictum that significant events tend to repeat themselves—thefirsttimeastragedy,thesecondasfarce.

In 1925, Charles B. Davenport, one of America's foremostgeneticists, wrote to his friendMadison Grant, the author of a best‐sellingbook,ThePassingoftheGreatRace,onthedilutionofAmerica'sold (read northern European, not Native American) blood by recentimmigration: "Our ancestors drove Baptists fromMassachusetts Bayinto Rhode Island, but we have no place to drive the Jews to."Davenport facedadilemma.Hesoughtageneticargument for innateJewish undesirability, but conventional stereotypes precluded theusualclaimforinherentstupidity.SoDavenportoptedforweaknessinmoral character rather than intellect. In his 1911 book, Heredity inRelation to Eugenics—not, by the way, a political tract but hisgeneration's leading textbook in thedevelopingscienceofgenetics—

hewrote:

In earning capacity both male and female Hebrew immigrantsrankhighandtheliteracyisabovethemeanofallimmigrants....Ontheotherhand,theyshowthegreatestproportionofoffensesagainstchastityandinconnectionwithprostitution,thelowestofcrimes . . . .The hordes of Jews that are now coming to us fromRussia and the extreme southeast of Europe, with their intenseindividualism and ideals of gain at the cost of any interest,represent the opposite extreme from the early English and themore recent Scandinavian immigration, with their ideals ofcommunitylifeintheopencountry,advancementbythesweatofthebrow,andtheuprearingoffamiliesinthefearofGodandloveofcountry.

The rediscovery and publication of Mendel's laws in 1900 hadinitiatedthemodernstudyofgenetics.Earliertheoriesofheredityhadenvisaged a "blending" or smooth mixture and dilution of traits byinterbreeding, whereas Mendelism featured a "particulate" theory ofinheritance, with traits coded by discrete and unchanging genes thatneednotbeexpressed inalloffspring independentandundiluted,butthatremaininthehereditaryconstitution,awaitingexpressioninsomefuturegeneration.

In anunderstandable initial enthusiasm for this great discovery,early geneticists committed their most common error in trying toidentify single genes as causes for nearly every feature of the humanorganism, from discrete bits of anatomy to complex facets ofpersonality. The search for single genetic determinants seemedreasonable for simple, discrete, and discontinuous characters andcontrasts (like blue versus brown eyes). But the notion that complexbehaviorsmightalsoemerge fromasimilarroot insimpleheredityofsingle genes never made much sense, for two major reasons: (1) acontinuity in expression that precludes any easy definition of traits

supposedlyunderanalysis(ImayknowblueeyeswhenIseethem,butwhere does a sanguine personality end andmelancholia take over?);and (2) a virtual certainty that environments can substantially moldsuchcharacters,whatever theirunderlyinggenetic influence(myeyesmaybecomebluewhateverIeat,butmyinherentlygoodbrainmayendupresidinginastupidadultifpoornutritionstarvedmyearlygrowthandpovertydeniedmeanyeducation).

Nonetheless, most early human geneticists searched for "unitcharacters"—supposed traits that couldbe interpretedas theproductof a single Mendelian factor—with abandon, even in complex,continuous, and environmentally labile features of personality oraccomplishment in life. (These early analyses proceeded primarily bythe tracing of pedigrees. I can envisage accurate data, and reliableresults,forafamilychartofeyecolor,buthowcouldanyonetracethealleged gene for "optimism," "feeble inhibition," or "wanderlust"—notto mention such largely situational phenomena as "pauperism" or"communality." Was great‐uncle George a jovial back‐slapper or areclusivecuss?)

Whatever thedubiousvalidityof suchoverextendedattempts toreduce complex human behaviors to effects of single genes, thisstrategycertainlyservedtheaimsandpurposesoftheearlytwentiethcentury's most influential social crusade with an allegedly scientificfoundation:theeugenicsmovement,withitsstatedaimof"improving"America's hereditary stock by preventing procreation among thesupposedly unfit (called "negative eugenics") and encouraging morebreeding among those deemed superior in bloodline ("positiveeugenics"). The abuses of this movement have been extensivelydocumented in many excellent books covering such subjects as thehereditariantheoryofmentaltestingandthepassageoflegislationforinvoluntary sterilization and restriction of immigration from nationsdeemedinferiorinhereditarystock.

Many early geneticists played an active role in the eugenicsmovement, butnonemore zealously than theaforementionedCharlesBenedict Davenport (1866‐1944), who received a Ph.D. in zoology atHarvardin1892,taughtattheUniversityofChicago,andthenbecameheadoftheCarnegieInstitution'sStationforExperimentalEvolutionatColdSpringHarbor,NewYork,wherehealsoestablishedanddirectedthe Eugenics Record Office, beginning in 1910. This office, with itsmixedaimsof supposedly scientificdocumentationandovertpoliticaladvocacy,existedprimarilytoestablishandcompiledetailedpedigreesinattemptstoidentifythehereditarybasisofhumantraits.Thehyper‐enthusiastic Davenport secured funding from several of America'sleading (and, in their own judgment, therefore eugenically blessed)families, particularly fromMrs.E.H.Harriman, theguardianangel andchiefmoneybagsfortheentiremovement.

Inhis1911textbook,dedicatedtoHarriman"inrecognitionofthegenerousassistanceshehasgiventoresearch ineugenics,"DavenportstressedthedependenceofeffectiveeugenicsuponthenewMendelian"knowledge" that complex behavioral traits may be caused by singlegenes.Writingofthe5,000immigrantswhopassedthroughEllisIslandeveryday,Davenportstates:

Everyoneofthesepeasants,eachitemofthat"riff‐raff"ofEurope,as it issometimescarelesslycalled,will, if fecund,playaroleforbetter or worse in the future history of this nation. Formerly,whenwebelievedthatfactorsblend,acharacteristicinthegermplasmofasingle individualamongthousandsseemednotworthconsidering:itwouldsoonbelostinthemeltingpot.Butnowweknow that unit characters do not blend; that after a sore ofgenerations the given characteristicmay still appear, unaffectedby repeated unions. . . . So the individual, as the bearer of apotentially immortal germ plasm with innumerable traits,becomesofthegreatestinterest.

That is, of our "greatest interest" to exclude by restrictingimmigration,lestAmericanhereditybeoverwhelmedwithadelugeofpermanentbadgenesfromthewretchedrefuseofforeignlands.

To illustrateDavenport'scharacteristicstyleofargument,andtoexemplify his easy slippage between supposed scientificdocumentation and overt political advocacy, we may turn to hisinfluential1915monographentitledTheFeeblyInherited(publicationnumber236ofhisbenefactor,theCarnegieInstitutionofWashington),especially topart1on "Nomadism,orTheWandering Impulse,WithSpecial Reference to Heredity." The preface makes no bones abouteither sponsorship or intent.With three ofAmerica'swealthiest andmost conservative families on board, one could hardly expectdisinterestedneutrality towardthe full rangeofpossibleresults.TheCarnegies had endowed the general show, while Davenport paidhomagetospecificpatrons:"Thecostoftrainingthefield‐workerswasmet by Mrs. E. H. Harriman, founder and principal patron of theEugenicsRecordOffice,andMr.JohnD.Rockefeller,whopaidalsothesalariesofmanyofthefield‐workers."

Davenport's preface alsoboldly admits his political position andpurposes. Hewishes to establish "feeble inhibition" as a category oftemperament leading to inferior morality. Such a formulation willprovide a one‐twopunch for identification of the eugenically unfit—badintellectandbadmorals.Thegeneticbasisofstupidityhadalreadybeen documented in numerous studies of the feebleminded. Buteugenics now needed to codify the second reason for excludingimmigrantsanddiscouragingreproductiverightsofthenativeunfit—badmoralcharacter(asinDavenport'sfallbackposition,documentedearlierinthisessay,forrestrictingJewishimmigrationwhenhecouldnot invoke the usual charge of intellectual inferiority). Davenportwrites:

Awordmaybesaidastotheterm'feeblyinhibited"usedinthesestudies.Itwasselectedasafittermtostandasco‐ordinatewith"feeble‐minded" and as the result of a conviction that thephenomena with which it deals should properly be consideredapartfromthoseoffeeble‐mindedness.

Toallayanydoubtabouthismotivations,Davenportthenmakeshis political point: Feeble inhibition, leading to immorality, may beevenmoredetrimentalthanfeeblemindedness,leadingtostupidity.

I think ithelpstoconsiderseparately thehereditarybasisof theintellectandtheemotions.Itisinthisconvictionthatthesestudiesaresubmittedforthoughtfulconsideration.For,afterall,thechiefproblemin administering society is that of disordered conduct, conduct iscontrolled by emotions, and the quality of the emotions is stronglytingedbythehereditaryconstitution.

Davenport then selects nomadism as his primary example of aputatively simple Mendelian trait—the product of a single gene—based on feeble inhibition and leading almost inevitably to immoralbehavior.Heencountersaproblemofdefinitionat theveryoutsetofhiswork,asexpressedinanopeningsentencethatmustberankedasoneoftheleastprofoundstatementsintheentirehistoryofscience."Atendencytowanderinsomedegreeisanormalcharacteristicofman,asindeedofmostanimals,insharpcontrasttomostplants."

How, then, shall the "bad" form of wanderlust, defined as acompulsion to flee from responsibility, be distinguished from themeritorious sense of bravery and adventure—leading to "good"wanderlust—that motivated our early (largely northern European)immigrants to colonize and subdue the frontier. In his 1911 book,Davenporthadwarmlypraisedas"theenterprisingrestlessnessoftheearly settlers . . . the ambitious search for better conditions. Theabandoned farmsofNewEnglandpoint to the trait inourbloodthat

enticesustomoveontoreapapossibleadvantageelsewhere."

In a feeble attempt to put false labels on segments of complexcontinua,Davenportidentifiedthe"bad"formas"nomadism,"definedasan inability to inhibit theurgewealloccasionally feel to flee fromour duties, but that decent folks suppress. Nomads are society'stramps, bums, hoboes, and gypsies—"those who, while capable ofsteady and effectivework, atmore or less regular periods run awayfrom the place where their duties lie and travel considerabledistances."`

Having defined his quarry (albeit in a fatally subjective way),Davenport thenrequired two furtherarguments tomakehis favoredlinkofa"bad"traittoasinglegenethateugenicsmightlabortobreeddown and out: he needed to prove the hereditary basis and then tofindthe"gene"fornomadism.

Hisargumentsforageneticbasismustbejudgedasastonishinglyweak, even by the standards of his generation. He simply argued,based on four dubious analogies, that features akin to nomadismemerge whenever situations veer toward "raw" nature (wheregenetics must rule) and away from environmental refinements ofmodernhumansociety.Nomadismmustbegeneticbecauseanalogousfeatures appear as "the wandering instinct in great apes," among"primitive peoples," in children (then regarded as akin to primitivesunder the false view that ontogeny recapitulates phylogeny), and inadolescents (in whom raw instinct temporarily overwhelms socialinhibitionintheSturmundDrangofgrowingup).Theargumentabout"primitive" people seems particularly weak since a propensity forwandering might be regarded as well suited to a lifestyle based onhuntingmobile game, rather than identifiedas amarkof inadequategenetic constitution (or any kind of genetic constitution at all). ButDavenport,reversingtheprobablerouteofcauseandeffect,wouldnot

bedaunted:

IfweregardtheFuegians,Australians,BushmenandHottentotsasthemostprimitivemen,thenwemaysaythatprimitivemanis nomadic. . . . It is frequently assumed that they are nomadicbecause they hunt, but it is more probable that their nomadicinstincts force them to hunting rather than agriculture for alivelihood.

Davenport then pursues his second claim—nomadism as theproductofasinglegene—bytracingpedigreesstoredinhisEugenicsRecordOffice.Onthesubjectivecriterionof impressionsrecordedbyfieldworkers, or written descriptions of amateur informants,Davenport marked all nomads in his table with a scarlet W forWanderlust,thecommonGermantermforanurgetoroam).HethenexaminedthedistributionofW’s throughfamiliesandgenerationstoreach one of the most peculiar and improbable conclusions everadvanced in a famous study: nomadism, he argued, is caused by asingle gene, a sex‐linked recessive located on what would later beidentifiedasthefemalechromosome.

Davenport reached this conclusion because he thought thatnomadism ran through family pedigrees in the same manner ashemophilia, color blindness, and the other truly sex‐linked recessivetraits. This status can be legitimately inferred from several definitepatternsofheredity.Forexample,fatherswiththetraitdonotpassitto their sons (since the relevant gene resides on the X‐chromosomeandmalespassonlyaY‐chromosomeonlytotheirsons).Motherswiththetraitpassittoalltheirsons,butnoneoftheirdaughters,whenthefatherlacksthetrait.(Sincethefeatureisrecessive,anafflictedmothermustcarrythegeneonbothX‐chromosomes.ShepassesasingleXtoher son, who must then express the trait, for he has no other X‐chromosome.ButadaughterwillreceiveoneafflictedX‐chromosome

fromhermotherandonenormalX‐chromosomefromherfather;shewill thereforenotexpressthattraitbecausethefather'snormalcopyof the gene is dominant.) Davenport knew these rules, so his studydidn'tfailonthisaccount.Rather,hiscriteriaforidentifyingnomadismasadiscreteandscorable"thing"weresosubjective,andsobiasedbyhis genetic assumptions, that his pedigree data turned out to beworthless.

Davenport's summary reached (and preached) a eugeniccrescendo: "The wandering instinct," he stated, "is a fundamentalhuman instinct, which is, however, typically inhibited in intelligentadultsofcivilizedpeoples."Unfortunately,peoplewhopossessthebadgeneW (thescarlet letterofwanderlust) cannotachieve thishealthyinhibition, and they become feckless nomads who run fromresponsibility by literal flight. The trait is genetic, racial, andundesirable.ImmigrantsmarkedbyWshouldbeexcluded(andmanyimmigrants must be shiftless wanderers rather than braveadventurers),whilenomadicnativesshouldbestronglyencouraged,ifnotcompelled,todesistfrombreeding.Davenportconcludes:

Thenew light brought by our studies is this: The nomadicimpulseis,inallthecases,oneandthesameunitcharacter.Nomads, of all kinds, have a special racial trait—are, in apropersense,membersofthenomadicrace.Thistraitistheabsence of the germinal determiner that makes forsedentariness,stability,domesticity.

Ofcourse,noonewouldnowdefendDavenport'sextremeviewofsinglegenesdeterminingnearlyeverycomplexhumanbehavior.Mostcolleagues eventually rejected Davenport's theory; he lived into the1940s,longpasttheearlyflushofMendelianenthusiasmandwellintothemodern era of understanding that complex traits usually record

theoperationofmanygenes,eachwithasmallandcumulativeeffect(not to mention a strong, and often predominant, influence fromnongenetic environmental contexts of growth and expression). Asinglegeneforanger,conviviality,contemplation,orwanderlustnowseemsasabsurdasaclaimthatoneassassin'sbullet,andnothingelse,caused World War I, or that Darwin discovered evolution all byhimself,andwewouldstillbecreationistsifhehadneverbeenborn.

Nonetheless,inourmodernageofrenewedpropensityforgeneticexplanations (a valid and genuine enthusiasm when properlypursued),Davenport's general styleof error resurfacesonanalmostdailybasis,albeit inmuchmoresubtle form,butwithall thevigorofhis putative old gene—yes, he did propose one—for stubbornlypersistentbehavior.

We are not questioning whether genes influence behavior; ofcoursetheydo.Wearenotarguingthatgeneticexplanationsshouldberesisted because they have negative political, social, or ethicalconnotations—achargethatmustberejectedfortwoprimaryreasons.First,nature's factsstandneutralbeforeourethicalusages.Wehave,tobe sure,oftenmadedubious, even tragic,decisionsbasedon falsegeneticclaims.But,inothercontexts,validargumentsabouttheinnateandhereditarybasisofhumanattributescanbeprofoundlyliberating.

Consider only the burden lifted from loving parents who raisebeautiful and promising children for twenty years and then "lose"them to the growing ravages of schizophrenia—almost surely ageneticallybaseddiseaseofthemind,justasmanycongenitaldiseasesofbodilyorgansalsoappear in thethirddecadeof life,oreven later.Generations of psychologists had subtly blamed parents forunintentionally inducing such a condition, then viewed as entirelyenvironmentalinorigin.Whatcouldbemorecruelthanafalseweightofblameaddedtosuchanultimatetragedy?Second,wewillneverget

veryfar,eitherinourmoraldeliberationsorourscientificinquiries,ifwe disregard genuine facts becausewe dislike their implications. Inthemostobviouscase, I cannot thinkofamoreunpleasant fact thanthe inevitablephysicaldeathofeachhumanbody,buta societybuilton the premise that King Prospero will reign in his personal fleshforeverwillnotflourishforlong.

However,ifweoftenfollowerroneousbutdeeplyrootedhabitsofthinking to generate false conclusions about the role of heredity inhumanbehavior,thenthesehabitsshouldbeexposedandcorrected—all the more vigorously if such arguments usually lead torecommendations for action that most people would also regard asethicallywrong(involuntarysterilizationofthementallyretarded,forexample). I believe thatwe face such a situation today and that thegeneticfallaciesunderlyingourmisusagesbearastrikingsimilarityinstyleandlogictoDavenport'serrors,howevermuchwehavegainedinsubtletyofargumentandfactualaccuracy.

Throughout the history of genetics, the most common politicalmisuses have rested on claims for "biological determinism"—theargument that a given behavior or social situation can't be helpedbecausepeopleare"madethatway"bytheirgenes.Onceweattributesomethingwedon'tliketogenes,wetendeithertomakeexcusesortomakelesseffortforchange.Forexample,manypeoplestillarguethatwe should deny educational benefits and social services to groupsfalselyjudgedasgeneticallyinferior.Theirpovertyandmisfortuneliein their own heredity, the argument goes, and therefore theircondition cannot be significantly ameliorated by social intervention.Thus,historyshowsaconsistentlinkagebetweengeneticclaimsinthismold and conservative political arguments for maintenance of anunjuststatusquoofgreatbenefittopeoplecurrentlyinpower.

Ofcourse,noseriousstudentofeithergeneticsorpoliticswould

nowadvancesuchanargumentinDavenport'sstyleof"onegene,onecomplex behavior." That is, no one today talks about the gene forstupidity,promiscuity,orlackofambition.Butaseriesofthreesubtle—and extremely common—errors leads all too often to the sameeugenicalstyleofconclusion.Somehow,weremainfascinatedwiththeideathatcomplexsocialbehaviorsmightbeexplained,atleastinlargepart,byinherited"atoms"ofbehavioralpropensity lyingdeepwithinindividuals.Weseemsomuchmoresatisfied,somuchmoreintrigued,by the claim that a definite gene, rather than a complex andinextricable mix of heredity and social circumstances, causes aparticularphenomenon.Wefeelthatwehavecomesomuchnearertoa real or essential cause when we implicate a particle within anindividual, rather than a social circumstance built of multiplecomponents,asthereasonbehindapuzzlingbehavior.Wewillavidlyreadafront‐pageheadlineentitled"GayGeneFound,"butnewspaperswill not even bother to report an equallywell‐documented story onother components of homosexual preference with a primary socialrootandnocorrelatedgeneticdifference.

The common source of these errors lies much deeper than anycorrelationtoapoliticalutilitymostofusdonotevenrecognize—andwould disavow if we did. I suspect that the source lies in a generalview about causality that has either been beaten into us by a falsephilosophyaboutscienceandthenaturalworldormayevenrecordanunfortunate foible inourbrain's evolvedmodeof operation.We likesimple kinds of explanations that flow in one direction from small,independent, constituent atoms of being too complex and messyinteractionsamonglargebodiesororganizations.Tousethetechnicalterm, we prefer to be "reductionists" in our causal schemes—toexplain the physical behavior of large objects as consequences ofatomsinmotion,orthesocialbehaviorof largeanimalsbybiologicalatomscalledgenes.

But the world rarely matches our simplistic hopes, and theadmittedly powerful methods of reductionism don't always apply.Wholes can be bigger than the sums of their parts, and interactionsamongobjectscannotalwaysbedisaggregatedintorulesofactionforeach object considered separately. The rules and randomness of asituationmustoftenbeinferredfromdirectstudyoflargeobjectsandtheir interactions, not by reduction to constituent "atoms" and theirfundamental properties. The three common errors of geneticexplanation all share the same basic fallacy of reductionistassumptions.

1.Wethinkwehavebecomeohsosophisticatedinacknowledgingthat both genes and environment produce a given outcome, but wethenerrinassumingthatwecanbestexpressthiscorrectprinciplebyassigningpercentagesandstating, forexample, thatbehaviorA is40percent genetic and 60 percent environmental.Wemust understandwhy such reductionist expressions have no meaning. Genetics andenvironment do interact to build a totality, but resultantwholes areunbreakableandirreducibletoseparatecomponents.Watercannotbeexplainedastwo‐thirdstheseparatepropertiesofhydrogengasmixedwith one‐third oxygen's independent traits—just as wanderlustcannotbeanalyzedas30percentofageneforfeebleinhibitionmixedwith 70 percent of social circumstances that abet an urge to hit theroad.

2.Wealsothinkthatwehavebecomesophisticatedinsayingthatmanygenes,notjustaDavenportianunity,setthehereditarybasisofcomplex behaviors. But we then take this correct statement andimpose the reductionist error of asserting that if behavior A isinfluencedbytengenesandis50percentgenetic(bythefirsterror),theneachgenemustcontributeroughly5percenttothetotalityofthebehavior. But complex interactions are not built as the sum ofindependentpartsconsideredseparately.Iamnotone‐eighthofeach

ofmygreat‐greatgrandparents(althoughmygeneticcompositionmaybe roughly so determined); I am a unique product of my owninteractingcircumstancesofsocialsetting,heredity,andalltheslingsandarrowsofindividualandoutrageousnaturalfortune.

3. We think that we are being sophisticated in qualifyingstatementsabout"genesfor"traitsbyadmittingtheironlypartial,andoftensmall, contributiontoan interactive totality.Thus,wethinkwemay legitimately talkof a "gaygene" so longasweadd thatonly15percent of sexual preference records its operation. We need tounderstandwhysuchstatementsaretrulymeaninglessandthereforeworse than merely false. Many genes interact with several otherfactors to influence sexual preference, but no separable "gay gene"exists.Even to talkabouta "gene for"10percentofbehaviorA is tocommit the old Davenportian fallacy on the "little bit pregnant"analogy.

To give a concrete example of how a good and important studycanbe saddledwithall theseerrors inpublic reporting (andalsobycarelessstatementsofsomeparticipatingresearchers), theNewYorkTimesgreeted1996withaheadlineonthe frontpageof its issue forJanuary 2: "Variant Gene Tied to a Love of New Thrills." The articlereportedontwostudiesintheJanuary1996issueofNatureGenetics.Two independent groups of researchers, one working with 124Ashkenazi and Sephardic Jews from Israel, the other with a largelymale sample of 315 ethnically diverse Americans, found a clearlysignificant, if weak, association between propensity for "novelty‐seeking behavior" (as ascertained from standard surveyquestionnaires) and possession of a variant of a gene called the D4dopaminereceptor,locatedontheeleventhchromosomeandactingasoneofatleastfivereceptorsknowntoinfluencethebrain'sresponsetodopamine.

This gene exists in several variant forms, defined by differinglengthsrecordingthenumber(anywherefromtwototen)ofrepeatedcopiesofaparticularDNAsubunitwithinthegene.Individualswithahighnumberofrepeatedcopies(thatis,withalongergene)tendedtomanifest a greater tendency for novelty‐seeking behavior—perhapsbecause the longer form of the gene somehow acts to enhance thebrain'sresponsetodopamine.

Sofar,sogood—andveryinteresting.Wecanscarcelydoubtthatheredityinfluencesbroadandbasicaspectsoftemperament—abitoffolkwisdomthat surely falls into thecategoryof "whateveryparentwithmorethanonechildknows."Nooneshouldbeatalloffendedorthreatenedbytheobviousfactthatwearenotbornentirelyblankorentirelythesameinourmixtureofthebroadbehavioralpropensitiesdefining what we call "temperament." Certain genes evidentlyinfluence particular aspects of brain chemistry, and brain chemistrysurely affects our moods and behaviors. We know that basic andpowerful neurotransmitters, such as dopamine, strongly impact ourmoods and feelings (particularly, for dopamine, our sensations ofpleasure).Differing formsofgenesthataffect thebrain'sresponsetodopaminemayinfluenceourbehaviors,andaformthatenhancestheresponsemaywellinclineapersontowardnovelty‐seekingactivities.

But the long formof theD4 receptordoesnot thereforebecomethe(orevena)novelty‐seekinggene,andthesestudiesdonotidentifythisbehaviorassuch‐and‐suchapercent"genetic"inorigin—althoughstatements in this form dominated popular reports of thesediscoveries. Even the primary sources—the two original reports inNature Genetics and the accompanying "News and Views" featureentitled "Mapping genes for human personality"—and the excellentTimes story (representing the best of our serious press) managed,amid their generally careful and accurate accounts, to propagate allthreeerrorsdetailedabove.

The Times reporter cited the first error of assigning separablepercentagesbywritingthat"abouthalfofnovelty‐seekingbehaviorisattributabletogenes,theotherhalftoasyetill‐definedenvironmentalcircumstances."Dr.R.P.Ebstein,principalauthorofonereport, thenstated the second error of adding up effects without consideringinteractions when he argued that the long form of the D4 geneaccountsforonlyabout10percentofnovelty‐seekingbehavior.If,bythe first error, all of novelty seeking can be viewed as 50 percentgenetic,and ifD4accounts for10percentof thetotality, thenwecaninferthataboutfourothergenesmustbeinvolved(eachcontributing10percentforthegrandtotalof50percentgeneticinfluence).EbsteintoldtheTimesreporter:"Ifweassumethatthereareothergenesouttherethatwehaven'tlookedatyet,andthateachgeneexertsmoreorless the same influence as the D4 receptor, then we would expectmaybefourorfivegenesareinvolved."

But the most significant. errors, as always, fall into the thirdcategory ofmisproclaiming "genes for" specific behaviors—as in theNature Genetics title previously cited: "Mapping genes for humanpersonality."(Ifourprofessionaljournalssoindulge,imaginewhatthepopularpressmakes of "gay genes," "thrill genes," "stupidity genes,"and so on.) First of all, the D4 gene, by itself, exerts only a weakpotential influence on novelty‐seeking behavior. How can a geneaccountingforonly10percentofthevarianceinatraitbeproclaimedasa"genefor"thetrait?IfIdecidethat10percentofmyweightgaincamefromthecaloriesintofu(becauseIlovethestuffandeatitbytheton), this item, generally regarded as nutritionally benign, does notbecomea"fatnessfood."

Moreimportantly,genesmakeenzymes,andenzymescontroltheratesofchemicalprocesses.Genesdonotmakenovelty‐seekingoranyother complexandovertbehavior. Predispositionvia a long chainofcomplexchemicalreactions,mediatedthroughanevenmoreintricate

series of life's circumstances, does not equal identification, or evencausation.Atmost,thelongformofD4inducesachemicalreactionthatcan, among other possible effects, generate a mood leading somepeople to greater openness toward behavior defined by somequestionnairesas"noveltyseeking."

Infact,afurtherstudy,publishedin1997,illustratedthiserrorina dramatic way by linking the same long form of D4 to greaterpropensityforheroinaddiction.TheoriginalTimesarticleof1996hadexultedinthe"firstknownreportofalinkbetweenaspecificgeneandaspecificnormalpersonalitytrait."Butnowthesamegene—perhapsviathesamerouteofenhanceddopamineresponse—correlateswithasevere pathology in other personalities. So what is D4—a "noveltyseeking" gene in normal folk, or an "addiction" gene in troubledpeople?We need to reform both our terminology and our concepts.ThelongformofD4isagenethatproducesachemicalresponse.Thisresponsemaycorrelatewithdifferentovertbehaviors inpeoplewithwidelyvaryinghistoriesandgeneticconstitutions.

The deepest error of this third category lies in the reductionistand really rather silly notion that we can even create rigorousdefinitions for discrete, separable, specific traits within the complexcontinua of human behaviors. We have enough trouble specifyingcharacterswithclearlinkstoparticulargenesinthemuchclearerandsimpler features of human anatomy. I may be able to specify genes"for"eyecolorbutnotforleglengthorfatness.HowthenshallIparsethe continuous and necessarily subjective categories of labilepersonalities?Isnoveltyseekingreallya"thing"atall?CanIeventalkin any meaningful way about "genes for" such nebulous categories?Have Inot fallenrightback into theerrorsofDavenport's search fortheinternalscarletletterWofwanderlust?

I finallyrealizedwhathadbeentroublingmesomuchabout the

literatureon"genesfor"behaviorwhenIreadtheTimes'saccountofC. R. Cloninger's theory of personality (Cloninger is the principalauthoroftheNatureGenetics"NewsandViews"feature):

NoveltyseekingisoneoffouraspectsthatDr.Cloningerandmanyother psychologists propose as the basic bricks of normaltemperament, the other three being avoidance of harm, rewarddependence and persistence. All four humors are thought to beattributableingoodparttoone'sgeneticmakeup.

The line about "humors" crystallized my distress, for I realizedwhy the canny reporter (or the scientist himself) had used this oldword.Considerthetheoryinoutline:fourindependentcomponentsoftemperament, properly in balance in "normal" folks, but with eachindividual displaying subtly different proportions, thus determiningour personal temperaments and building our distinct personalities.But if one humor gets out of whack by substantial over‐ or under‐representation,thenapathologymayresult.

But why four, and why these four? Why not five, or six, or sixhundred?Whyany specificnumber?Why try toparse suchcontinuainto definite independent "things" at all? I do understand themathematicaltheoriesandproceduresthatleadtosuchidentifications(see my book The Mismeasure of Man), but I regard the entireenterpriseasamajorphilosophicalerrorofourtime(whileIviewthemathematicaltechniques,whichIuseextensivelyinmyownresearch,ashighlyvaluablewhenproperlyapplied).Numericalclumpsarenotphysicalrealities.Afour‐componentmodeloftemperamentmayactasa useful heuristic device, but I don't believe for amoment that fourhomunculi labeled "novelty seeking," "avoidance of harm," "rewarddependence," and "persistence" reside in my brain, either wing fordominanceorcooperating.

The logic of such a theory runs in uncanny parallel—hence thecleverchoiceof"humor"asadescriptivetermforproposedmodulesof temperament—with the oldest and most venerable of gloriouslywrongtheories in thehistoryofmedicine.Formore thana thousandyears, from Galen to the dawn of modern medicine, prevailingconcepts regarded the human personality as a balance among fourhumors—blood, phlegm, choler, and melancholy. Humor, from theLatinwordforliquid(preservedinourdesignationofthefluidsofthehumaneyeastheaqueousandvitreoushumors),referredtothefourliquids that supposedly formed the chyle, or digested food in theintestine just before it entered the body for nourishment. Since thechylearose,ononehand, froma rangeof choices in the foodweeatand,ontheotherhand,fromconstitutionaldifferencesinhowvariousbodiesdigestthisfood,thetotalityrecordedbothinnateandexternalfactors—anexactequivalenttothemodernclaimthatbothgenesandenvironmentinfluenceourbehavior.

The four humors of the chyle correspond to the four possiblecategories of a double dichotomy—that is, two axes of distinctionbasedonwarm‐coldandwet‐dry.Thewarmandwethumorisblood;cold and wet generates phlegm; warm and drymakes choler; whilecold and dry forms melancholy. I regard such a logically abstractscheme as a heuristic organizing device, much like Cloninger'squadripartite theoryofpersonality.Butwemakeamajorerror ifweelevatesuchaschemetoclaimsforrealanddistinctphysicalentitiesinsidethebody.

In the medical theory of humors, good health results from aproperbalanceamongthefour,whiledistinctivepersonalitiesemergefromdifferentproportionswithin thenormalrange.But toomuchofany one humor may lead to oddness or pathology. As a fascinatinglinguistic remnant, we still use the names of all four humors asadjectives for types of personality: sanguine—dominance of the hot‐

wet blood humor—for cheerful people; phlegmatic, for stolid folksdominated by the cold‐wet humor of phlegm; choleric, for angryindividualssaddledwithtoomuchhot‐drycholer;andmelancholic,forsad people overdosed with black bile, the cold‐dry humor ofmelancholia. Is themodernquadripartite theoryofpersonality reallyany different from this older view in basic concepts of number,balance,andthecausesofbothnormalpersonalityandpathology?

In conclusion, we might imagine two possible reasons for suchuncannysimilaritybetweenamodernconceptionoffourcomponentsto temperament and the oldmedical theory of humors. Perhaps thesimilarity exists because the apparatus is true, with the modernversion representing a great refinement of a central fact that ourancestors could only glimpse through a glass darkly. Butalternatively—and ever so much more likely in my judgment—thestunningsimilaritiesexistbecausethehumanmindisaconstantthing,despite all our growth of learning and the historical changes inWestern culture. We therefore remain sorely tempted by the sameeasyfallaciesofreasoning.

Isuspectthatweoncechosefourhumors,andnowdesignatefourbasic constituents of temperament, because something deep in thehuman psyche leads us to impose simple taxonomic schemes ofdistinctcategoriesupontheworld'strulycomplexcontinua.Afterall,our forebears didn't invoke the number four only for humors. Weparsedmany other phenomena into schemeswith four fundamentalproperties—thefourcompasspoints,theagesofman,thefourGreekelements of air, earth, fire, and water. Could these similarities becoincidental, or does something about the workings of the humanbrainfavorsuchartificialdivisions?CarlJung,forreasonsthatIregardasdubious, felt strongly thatdivisionby four representedsomethingdeepandarchetypalinhumanproclivities.Hearguedthatdivisionsbythree lack balance and lead onward (for one triad presupposes

another for contrast), whereas divisions by four stand in optimalharmony.Hewrote:"[B]etweenthethreeandthefourthereexiststheprimary opposition of male and female, but whereas fourness is asymbolofwholeness,threenessisnot."

I think Jungcorrectlydiscernedan inherentmentalattraction todivisions by four. I suspect the basis for this propensity lies in ourclear (andprobablyuniversal)preference fordichotomousdivisions.Four represents a kind of ultimate dichotomization—adichotomy ofdichotomies:twoaxes(eachwithtwofundamentalproperties)atrightangles toeachother.Wemayexperience fourasanultimatebalancebecause such schemes fill ourmental spacewith twodichotomies inperfectandoppositecoordination.

Inanycase, ifthissecondreasonexplainswhyweinventedsucheerily similar theories as four bodily humors and four basicconstituentsoftemperament,thensuchquadripartitedivisionsreflectbiases of the mind's organization, not "real things" out there in thephysicalworld.Wecanhardlytalkabout"genesfor"thecomponentsof such artificial and prejudicial parsings of a much more complexreality. Interestingly, the greatest literary work ever written on thetheory of humors, the early‐seventeenth‐century Anatomy ofMelancholy,bytheEnglishdivineandscholarRobertBurton,properlyrecognized the four humors as just one manifestation of a largerpropensity to divide by four. This greatman, who used the balm ofliterature to assuage his own lifelong depression, wrote of hiscondition:"Melancholy,coldanddrie,thicke,blacke,andsowre...isabridle to the other two hot humors, bloode and choler, preservingthemintheblood,andnourishingthebones:Thesefourehumorshavesomeanalogiewiththefoureelements,andtothefoureagesinman."

Iwouldthereforeend—andhowcouldanessayistpossiblyfindamore appropriate culmination—with some wise words from

Montaigne, the sixteenth‐century founder of the essay as a literarygenre. Insteadof trying to identifyapropensity forwanderingor fornovelty seeking (perhaps a spur to wandering) in a specific, innatesequenceofgeneticcoding,perhapsweshouldpaymoreattentiontothewondrouswanderingsofourmind.Foruntilwegraspthebiasesandpropensities of ourown thinking,wewill never see through thehumorsofourvision into theworkingsofnaturebeyond.Montaignewrote:

Itisathornyundertaking,andmoresothanitseems,tofollowamovement so wandering as that of our mind, to penetrate theopaquedepthsof its innermostfolds, topickoutandimmobilizetheinnumerableflutteringsthatagitateit.1

                                                            

StephenJayGouldteachesbiology,geology,andthehistoryofscienceatHarvardUniversity.HeisalsotheFrederickP.RoseHonoraryCuratorinInvertebratesattheAmericanMuseumofNaturalHistory. 

NaturalHistory107(March):22‐25,70‐78.