K E N A L D E R

ASocialHistoryofUntruth: Lie Detection and Trustin Twentieth-CenturyAmerica

Inthewinterof2000 , shortly beforethe release ofthe nuclear scientistW en HoLee— the LosAlamos employee accusedof havi ng soldatom bombsecrets to the Chinese government—itemerged that agents ofthe Federal Bureauof Investigation had lied to Lee when they toldhim he hadfailed a poly- graph test.1 Thisarticle suggests how such lies have becomestandard procedure in the useof the polygraphmachine, an instrumentthat measures four basic physio- logicalparameters ( bloodpressure, galvanic ski nresistance,heart rate, and breath- ingdepth) whi le the subjectis interrogatedabout his or her activities. By the time Lee regainedhis freedom —hisprosecutors, it turned out, had misrepresented many otherfacts in the chargesagainst him— thousands of otherAmerican scien- tistsin the nationalweapons labs were being systematically subjectedto exams forlie detection. 2 Notlong after ,inthe wake ofthe RobertHanssen spycase, the FBIbeganto test itsown agents witha polygraph,even thoughyears ofannual polygraphexaminations had failed to catch the doubleagent AldridgeAmes at the CIA.3 Mostrecently ,the variousagencies thatwi llsooncomprise the Department ofHomelandSecurity have begunto polygraphdetai nees held inconnectionw ith terrorismon U .S.soi l.The useof the polygraphin these circumstancesdoes not seem tostrike most A mericansas eithersurprising or objectionable.Nor is itnew . What issurprisingis the historyof how we gotto thispoint: why Americanslook toa lie detectormachine to ferret out the truth,even thoughthere isabundant evidence thatthe machineitself depends on lies. In ASocialHistor yofT ruth Steven Shapinfocuses on the conditionsthat promptedseventeenth-century English gentlemen tospeak the truthon behalfof recalcitrantnature. 4 Thispaperw illfocuson the e Vortsof twentieth-century Amer- icanexperts tooblige recalcitrant men andwomen to tell the truthaboutthemselves. How diVerent arethese twoenterprises? When AlbertEinstein inscribed above his Žreplacethe motto,‘ ‘Nature’s Godis subtle,but He isnotmalicious,’ ’ he surely ac- knowledgedas acorollarythe possibility that people mightbe malicious,if alsosome- timessubtle. It is thislatter corollary that has inspiredthe proponentsof an Ameri-

Representations 80 · Fall 2002 q theregentsoftheuniversityofcalifornia is sn 0734-6018 pages1– 33. Allrights reserved.Send requests for permission toreprint toRights andPermissions, University of CaliforniaPress, Journals Division, 2000 CenterSt., Ste.303, Berkeley,CA94704-1223. 1 canscience oflie detection.Their premise is thatwhile a humanbeing may tell a consciouslie, that person’ s bodywi ll‘‘honestly’’ betrayhis or herawareness of this falsehood.By the middleof the twentiethcentury ,sometwo mi llionlie detector tests werebeing administered each year tocriminal suspects, members of the na- tionalsecurity apparatus, and ordinary citizens as aroutinepart of employment. Thisproliferation of lie detectortests intwentieth-centuryAmerica could not have occurred,of course,had not their expert proponentspersuaded their co-citi- zens thatthe tests served somepurpose. No novel technologycansucceed unless someonebelieves the claimsmade on itsbehalf. But, in the case ofthe lie detector, somethingadditional was required.There, the claimsmade on behalfof the tech- nologywerethemselves integralto the operationof the technology.Asseveral of itsproponentsacknowledged, the lie detectorwould not ‘ ‘work’’ (thatis, deter mine the fatesof itshuman subjects) unless itssubjects believed it‘ ‘worked’’ (thatis, distin- guishedtrue utterances from false ones).In otherwords, the machinecould not catchliars unless they believed they mightbe caught.T othatextent, the historyof the lie detectoro Versa dramaticexample ofthe degreeto whichthe transformative powerof technologymay residein whatmedical science has dismissively termed the ‘‘placeboe Vect’’ :the residualpotency produced by the ‘‘merely social’’ conŽ- dence thatmedical technology inspires in itslay subjects—and in itspur veyors too.5 Themachinery for catching liars, then, isanilluminatingexample oftechnol- ogy’s dependenceon the socialimaginary .Assuch,it may serve asan idealprobe intothe Americanpopular i magination. Indeed,as aninstrumentdesigned to assess the conŽdence thatone citizen may placein the utterancesof another ,the lie detectordirectly engages the problemof trustand mistrust that governs dai ly life inalarge,anonymous society .Duringthe courseof the nineteenth centuryAmericans became increasingly immersed in the exchanges ofmarketplacecommerce; yet they stilllargely encounteredone another inface-to-facei nteractions.As KarenHalttunen has shown,the abilityto read appearanceswas one of the acquiredski lls inthe repertoireof Victorian sociabil- ity,enablingcitizens to disti nguishthe conman from the legitimatesalesman. 6 Twentieth-centuryA mericans,by contrast, increasingly found themselves operat- ingwithinlarge hierarchical organizations— both withincorporate capitalism and stateinstitutions— organizations whose main rationale was the substitutionof bu- reaucraticpredictability for the (expensive) uncertaintiesof the marketplace.But couldthe managersof these newbureaucratic hierarchies trust their subordinates any morethan Victorianshad trusted traveling salesmen? Thispaper argues that the lie detectorwas oneof the principaltools by which twentieth-century A merican societytried to solve the problemof trust.Designed to drawa sharpline between lawfuland unlaw fulbehavior ,toprivilege expert insightover lay assessment, and toregulatelife withinhierarchicalinstitutions, the historyof the lie detectoris part ofthe historyof how America coped with the riseof a masspublic, on the one hand,and the riseof new large-scale organizationson the other.

2 Representations ABriefHistor yofDissembling

Ofcourse, all Yahoosocieties thrive byspeaking what Jonathan Swift called‘ ‘the thingwhich is not.’’ Despitephilosophical injunctions against falsehood fromSt. Augustine to Immanuel Kant,dissembling is asu Yciently advantageous practiceto befound among all peoplesat all times.There are Machiavellian lies disseminatedby the strongand defensive lies wovenby the weak.A ndof course there arethe many falsehoodswe individuallyand collectively tell ourselves—what onemight call ‘‘BasicLies.’ ’ 7 Perhapsfor that very reasonthere isagoodcase to bemade,as JosephBrodsky has done,that consciousness does not begi nuntilone tells one’s Žrst deliberate lie.8 Butif dissemblingis something ofahumanuniversal, the measurestaken to rootit outmay betreatedhistorically .Every Yahoosociety possesses its own individ- ualsandinstitutions whose authority depends on their presumed ability to unmask certainkindsof pettyfalsehoods— ifonly tobetter preser ve the bigones. One ven- erableapproach to this problem (sanctioned already i nclassicaltimes by physiog- nomy) has been toread morals by appearances: shifty eyes ora rosyblush may be signsof deceit. 9 ButconŽ dence men canmaster their faces, and women may paint. So,more probing tests have oftenbeen thoughtnecessary . The justicesystem has longwrestled with such tests becausecri minalactivity , almostby deŽ nition, cloaks itself in the sortof falsehoodthat society wishes to un- cover.Wecanidentify (roughly) three phasesin the developmentof suchtests in the West.In the medieval trialby ordeal, the alleged criminal’s innocencewas inter- rogatedby a physicalchallenge sothat God might deter mine the outcome.In one suchtest, presumed liars were asked tolick aburning hotpoker .IfGod wanted to commendtheir honesty ,theirtongues would not be burned. By the twelfthcentury ,asecondphase emerged on the EuropeanContinent: an inquisitorialsystem ofjustice. In the pursuitof certainty of judgment, magis- trateswere authorized to order the useof judicialtorture to obtai naconfession, then consideredthe ‘‘Queen ofProof.’’ Theproblem here (as the juristsunderstood fullwell) wasthat forced confessions might be unreliable.Hence, magistratescould only authorizetorture on the basisofstrongcircumstantial evidence; the examiners wereforbidden to ask overly suggestive questions;the confessorhad to supply cor- roborativeinformation; and the confessionhad to be repeatedafter the torturehad ceased.Still, jurists recognized how easily thissystem couldbe abused:that poten- tially innocentsuspects su Veredpain worse than any possiblesanction and that even reiteratedconfessions might be false. Althoughthe campaignto end torture wouldulti mately triumphunder the bannerof Enlightenment humanism,the practicealready was giving way in the seventeenth centuryto anewprobabi listic appraisalof the trustworthiness ofhuman testimony . 10 Itwasunder this probabilistic banner that the thirdphase took shape in early modernEurope. Increasingly ,statements ofwitnesses were probed in cross-exami-

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 3 nationby lawyers and judges, w ithultimate judgment on theirveracity— and the guiltof the accused—depending on the intime conviction ofthe magistrate(or in England,on the ‘‘moralcertainty’ ’ ofthe jury).At the same time,however,avariety ofexperts alsobegan to take aprominentrole in speakingon behalf of circumstan- tialevidence, evidence thatlay beyondthe ability ofthe laity (orthe magistrate)to assess and beyondthe powerof the accusedto dissemble— and that therefore could beusedto corroborate (or not) human testimony .Itis thisformof courtroominves- tigationthat, in all itsmani foldper mutations,has persisteddown to the present dayin Western Europeand the UnitedStates. 11 Yet,as the culturalauthority of science expandedin the nineteenth-century, many socialthinkers beganto hope that experts speakingon behalf of circumstan- tialevidence wouldat last cometo supplantentirely these merely probabilisticas- sessments ofhuman testimony .The problemhere (atleast injurisdictionssubject toAnglo-Americanlaw) was that the statehad increasi ngly delegatedthe gathering ofevidence tothe twoadversarial parties, and those parties had proved themselves adeptat Ž ndingadversarial experts tomake diametricallyopposite arguments. By the endof the nineteenth century,expert quarreling hadbecome a scandalin Americancourts. 12 Inspiteof this,there emergedi nearly twentieth-centuryA merica,a set of reform-mindedexperts whotried to inaugurate what they hopedwould be anew (fourth)phase in the investigationof accusedpersons and other witnesses: scienti Žc interrogationby means ofa‘‘polygraph’’ lie detector.Theirgoal was tocircumvent humandissembling by directly probing the thoughtsof their subjects. By measuring the subject’s basicphysiological parameters whi le he orshe wasunder interroga- tion,polygraph operators believed they couldtransform the subject’s bodyinto a pieceof counterfeit-proof circumstantial evidence thatcould corroborate (or not) the assertionsof the legal personhooked up to the apparatus.They hopedto return certaintyto the operationof modernA mericanjustice— much as ithadexisted in the daysof judicialtorture. 13

Why America?

By the middleof the twentiethcentury nearly twomillion polygraph tests werebeing administered each year inthe UnitedStates by Ž ve toten thousand operators.14 The polygraphwas usedin investigative policework, to screen business employees, fornational security checks, andas apublicitystunt. Its use continues, even thoughmany studieshave documentedthe machine’s fallaciesand limitations. Inthe mid-1980s,when the Reagan administrationtried to impose routine poly- graphexaminations on civilianfederal employees, the U.S.Congress ordered its OYce ofT echnologyAssessment (OTA)toassemble ameta-study.The resultsof the studyaccorded the methodan 80percentsuccess rate, an achievement signiŽ-

4 Representations cantly less impressivethan the 98percentsuccess rate regularly touted by polygraph examiners.15 Andeven the OTA’sstudywas considered optimistic by the noted psychologistDavid Lykken, whopointed out that in Želd studiesconducted under ‘‘truedouble-blind conditions,’ ’ the numberof false positivesjumped to 47 percent (thatis, the innocentwere called truthful only 53percentof the time). 16 In thispaper ,Iwilladdresstwo proximate historical questions about the lie detectorso as toget atone larger question about trust in Americanpublic li fe.First, Why wasthe polygraphdeveloped at the timeandplace and in the mannerit was? Andhow did it achieve sucha phenomenalsuccess? Here itis worthkeeping in mindthat no country outside the UnitedStates uses the technique. 17 Second, Why has the polygraphtest been consistentlybanned from U .S.courts? F ordespitethei r grandamibition to promotecertainty in Americanjustice, the reform-mindedex- pertsrepeatedly failed to introduce lie detectorevidence intocriminal trials. Ever sincethe Frye rulingof 1923—aruling that governed the admissibilityofallforms ofscientiŽc testimonyuntil the 1990s—American courts have excludedpolygraph evidence forhaving ‘‘failedto gain general acceptancei nthe particularŽ eld in whichit belongs.’ ’ 18 Thatis, the courtshave assertedin theirguise as sociologists ofscience thatthe relevant experts have rejectedthe techniqueas ‘‘badscience.’ ’ Thispaper shows that this characterization is notso much inaccurate as hopelessly insuYcient.Only inthe lastdecade— since the Daubertruling of 1993— have trial judgesbeen given abroaderset ofcriteriato allowscientiŽ c testimony,prompting somecourts to reconsidertheir ban on the polygraph. 19 When areconsiderationof the banon polygraphevidence recently camebefore the U.S.Supreme Court, sev- eral justicesconfessed their discomfort with this apparent contradiction between the law’s self-proscriptionand its toleration elsewhere ofthe practiceof lie detec- tion.20 Inoralarguments before the U.S.Supreme Court they posedthe question atthe heartof thispaper: Why doesthe U.S.government permit the useof the lie detectorin somany arenas,but forbid its admission into the courtroom? 21 Myargumenthere isthatthe answerto this question is fundamentallyhistor- ical.That is, I willarguethat the mixedreception of the polygraphin America turnedon the sort oflie detectionthat emerged over the courseof the twentieth century,andthat this particularpractice of lie detectionwas the outcomeof intense campaigning amongthe variousinterested parties. I willdemonstratethis byexam- iningthe careerstrategies of the fourcreators of the modernscience oflie detection between1900 and 1950: Hugo Mu ¨nsterberg,Wi lliam Marston,John Larson, and LeonardeKeeler .Theirshared assumptions point to what was particularlyAmeri- canabout the lie detector.Andtheir conscious disagreements tell usmuchabout the selective acceptanceof the lie detector.Indeed,it was the interactionof two distinctstrategies for validating expertise— one that sought to make itsreputation bypublicly disclosing knowledge, and another that sought to make aproŽt byhold- ingknowledge as proprietary—that together generated our current political econ- omyof lie detection.With thisepistemologicalgroundwork, we may then examine

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 5 the careerof the polygraphfor what it tells usaboutthe changingculture of trust inAmerica:about trust in science,trust in oursocial institutions, trust in ourfel- lowcitizens.

American Emotionology

The polygraphwas assembled out of variousphysiological instruments thatcame into use in Europeand the UnitedStates at the endof the nineteenth century.InFrance,in the 1860s,Etienne-Jules Marey Žrstbegan to usehis auto- maticapparatus to produce permanent, continuous, and graphical records of changes inbloodpressure, respi ration,and pulse rates while his patients experi- enced nausea,sharp noises, and ‘ ‘tension.’’ 22 Meanwhile, inAmericain the 1870s, the psychologistWilliam James usedtechniques of introspection to deŽne emotion asthosebodi ly changes thatoccur i nresponseto the cognitionof an exciting ‘‘fact.’’23 Butif James half retractedhis ownclaimin the years tofollow,hissucces- sorswere more brazen. Hugo Mu¨nsterberg—lured from Ger many toHarvardby James, and despised byhi minlateryears— founded the ŽrstmajorA mericanprogram of ‘‘brassinstru- ment’’ psychology,aswell asthe schoolsof appliedpsychology and industrial psy- chology.InhisHar vardlab, Mu ¨nsterbergand his students hooked their subjects uptoa physiologicalapparatus in anattemptto translatethe ephemeraof private, interior aVective experience ontoa public,universal grid. They then classiŽed, quantiŽed, compared, and aggregated these physiologicalrecords to make visible the normaland deviant psychological states of their volunteer subjects. The body, asan instrumentupon which ‘ ‘emotions’’ played,was then examined forsigns of adjustmentto the modernrhythms ofworkand play ,anxiety andease, stressand pleasure.Emotional bodies were black or white,male orfemale, andhonest or de- ceptive.24 ThisscientiŽ c programdirectly challenged the law’s venerable methodsfor as- sessing humanbeliefs and desires. Mu ¨nsterbergdenounced the courts’archaic pro- ceduresof adversarialcross-examination conducted by law yers unschooledi nthe newscience ofpsychology .Initsplace, he o Veredto introduceinto A mericanjuris- prudencea modern,mechanized version of the thrivingEuropean research pro- gramin the psychologyofcourtroomtestimony . 25 In 1907, Mu¨nsterbergtook a trainto Boise,Idaho, to examine HarryOrchard, whohad confessed to assassinating the governorof the state,but had laid the blame onaconspiracyof socialists, led byCharles Haywood, head of the radicalW estern MiningUnion. Haywood accused Orchard of lying, andthe Haywoodconspiracy trialquick ly becamea politicalcause ce ´le`bre,pitting organized labor against the corporatetrusts and the state.At the invitationof the prosecution,Mu ¨nsterberg subjectedOrchard to psychological tests andpublicly declared the mana truth-

6 Representations teller.By allowingthis opinion to be publishedbefore the juryhad rendered its judgment,he infuriatedthose who saw expertise asanattemptto usurp lay justice. 26 But Mu¨nsterbergwas not one to retreatfrom a good Kulturkampf .The next year,in hisbook entitled Onthe Witness Stand ,he accusedthe justicesystem ofwillfully spurningthe scientiŽc analysis oftestimonyand, hence, ofacallowdisregard for moderntechniques of sorting truth from falsehood. 27 And Mu¨nsterbergfound a sympatheticaudience for his views. A 1911editorial in the New YorkTimes pro- claimed,‘ ‘Soonthere will beno jury ,nohorde of detectives andw itnesses, no chargesand countercharges, and no attorneyfor the defense. These impedimenta ofourcourts wi llbeunnecessary.The Statewill merely submitall suspectsin a case tothe tests ofscientiŽ c instruments,and as these instrumentscannot be madeto make mistakes nortell lies, theirevidence wouldbe conclusive of gui lt orinnocence.’’ 28 William MoultonMarston was a Harvardlaw yer andstudent of Mu¨nsterberg whocarried on hisprogram after his death. In 1915,Marston continuously moni- toredchanges inawitness’s bloodpressure for signs of stresscaused by the guilty knowledgeof deception,thereby creatingthe Žrstmodern polygraph. He then went ontoimprove his instrumentunder the auspicesof the NationalResearch Council, testing iton soldiers at a Georgiaar my baseduring W orldW arI. 29 Marston is perhapsmore famous today as the creatorof the cartooncharacter of W onder Woman(which, as wewill see later,wasno coincidence). But long before he in- vented the feministAmazonian and her truth-lasso, Marston recognized that the polygraphdid not oVeran objectivemeasure of lying.Atbest,it measured whether the subjectwas distressed when she knowinglytold a falsehood,such that her physi- ologywas altered in waysshe couldnot suppress. A ndfurther more,that these changes couldbe distinguishedfrom the signsof otheremotions, such as the fear producedby the exam itself.That is, the test assumedthat while the lie wasa con- sciouschoice, the bodywas a slave tohabits shaped by an extended socialtraining (a conscience?)allowing interrogators to access the withheldknowledge. Marston admitted,for instance, that a pathologicalliar could never becaughtby the poly- graph.30 The challenge then, wastwofold: Ž rst,to design a standardizedpiece of ‘‘hardware’’ tomeasure the relevant physiologicalparameters, and second (and moreelusive), todesign the ‘‘software’’: aninterrogationtechnique that could cali- bratelie-producing stress against other forms of stress. In1922,Marston was invited to polygraphJames AlphonseF rye, an African Americanfrom W ashington,D.C., who had confessed to murder and then re- tractedhis confession.Marston’ s test clearedF rye, butthe judgestubbornly refused toallow Marston to testify tothis factin court,despite his ampleexpert credentials asa psychologist.According to the judge,Marston ’spolygraphexam seemed to invadethe provinceof the jury,whoseprerogative it wasto‘ ‘sizeup’ ’ the defendant. He acknowledgedthat the defendanthad had the badluck to appearbefore an old judgeunwi lling todiscard traditional methods for new fangledscience, but he stood

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 7 byhis refusal to admit evidence until‘ ‘itis establishedthat [a] scientiŽ c develop- ment has reachedsuch a pointas tobecome a matterof commonknowledge as toits results.’’ Apparently,the courtof appealsbench was alsodominated by old-timers, becausethe judge’s decisionwas sustainedin the famousFrye rulingof 1923,which rejectedthe liedetectorand advised judges henceforth to admit the scientiŽc testi- monyof only thoseexperts whosejudgments were derived from principles in line withthe consensusof the relevant scientiŽc community. 31 Forthe next Žfty years thisruling dictated the admissionof all formsof scientiŽc evidence toU.S.courts.

Sex, Lies, and

Paradoxically,the Frye rule’s banon the polygraphin courtcoincided witha vast expansionin the useof the lie detector.The secondphase of my story beginsi nthe wake ofthe Frye ruling,when twodisciples of AugustV ollmer,Police Chiefof Berkeley ,,adapted Marston ’smethodsfor use on criminalsus- pectsin policecustody ,ausethat lay outsidethe purview ofthe Frye ruling.One disciple,the self-righteous‘ ‘college-cop’’ JohnLarson, had a Ph.D.inphysiology fromBerkeley withan M.D.fromR ushto follow .The otherdisciple was the man- about-townand entrepreneur Leonarde Keeler ,namedafter Leonardo Da Vinci, thoughknown less grandlyas ‘‘Nard.’’ Inthe 1920sthese men workedcollabor- atively inBerkeley underV ollmer,butafter both moved to Chicago i nthe 1930s, they wenttheir separate ways, soon becoming rivals, and ending up something akin toenemies. Larsonjoined the Institutefor Juvenile Research andthence migrated intopsychiatry (with a degreefrom Johns Hopkins); Keeler briey gota jobwith NorthwesternUniversity’ s ScientiŽc CrimeLaboratory (the nation’s Žrstsuch lab), andthen wentto work as aprivateconsultant. The mainprogenitors of the lie detectorall sharedcertain features. T obegin with,all three men—Keeler ,Larson,and Marston— met theirwives throughthe lie detector.Marston’s wifewas his co-author and collaborator .Keeler met hisfu- turewife in the psychlab at Stanford University when he wasback on campus to Žnish hisB. A.AndLarson’ s Žrstsuccess with lie detectioncame in 1922,when he investigateda petty theftat a Berkeley sorority.Over the courseof several dayshe strappeddown Ž fty sororitysisters one by one in achair,hookedthem uptohis apparatus,and interrogated them aboutthe missing$500. The guilty party turned outto be the richestyoung woman in the house.Larson married one of the others. The recorddoes not indicate what questions he asked her. 32 Contemporarydescriptions of the lie detectoroften played o V genderstereo- types,with the interrogatingexaminer invariablycoded as male, andthe evasive subjectas female. Newspaperphotos typically showedtight-sweatered women strappedinto the machine. 33 Andthe technicalmanuals also printed striking pho- tosof womensubmitting to mock interrogations. Both the Americanlay andscien-

8 Representations tiŽc cultureshave longtyped women as emotional,secretive, anddeceitful, identi- fying them with‘ ‘nature’’ andas the subjectsof scientiŽc investigations.Those same lay andscientiŽ c worldviewshave, bycontrast, portrayed men asrational,forth- right,and frank, holding them upasthe idealobjective investigators. Hence, itis hardlysurprising that the pioneersof the polygraph(and newspaper editors quick tosense the ‘‘hot’’ angle toa story)used gender-typing to convey the ‘‘objective’’ natureof lie detectiontechniques and dramatize the ability ofpolygraphoperators toferret out hidden thoughts. 34 Onthe surfaceat least, William Marston’s texts forW onderW oman—aself- proclaimedfeminist hero —subvertedthese stereotypes.F orinstance,one of W on- derW oman’s archenemies (in thoseŽ rstnumbers of the early 1940s)is Dr.Psycho, ascientistwho has been spurnedby women all hislife becauseof his stuntedbody , butwho masters psychological powers that enable himto mesmerize vast crowds. Akindof Mario the Magician,he transformshimself Žrstinto Benito Mussolini, andthen, beforehis audience’ s eyes, intoGeorge W ashington,and then Žnally— toevade WonderW oman—into her square-jawedlover ,Steve. Yet WonderW oman ŽghtsDr .Psychow ithtactics that hardly di Verfrom the dissembler’s ownfascist propaganda.A lthoughshe espousesliberal rhetoric and is aŽerceadvocate of femi- nistequality ,when she tiesup Dr.Psychowith her truth lasso, he is obliged to tell the truth.Bound by her lasso, W onderW oman’s adversariesare ‘‘forcedto be free.’’ Revealingly,WonderW oman’s ownAmazonian powers are based on herwi ll- ingsubmission to a higherauthority: the powerof love. Herstrength derives from obedience.Almost every episodeshows her (orsome other woman) being bound, manacled,and enslaved— only tobeliberatedby submission to agreatergood. F or Marston,this popmythologizing was meant totap deep psychological truths about contemporaryrelations between men andwomen. His psychological theories pos- iteddominance and submission as the polaritiesthat underlie such garden-variety emotionsas ‘‘fear’’ or‘‘love.’’ Forhispart, Marston believed thatwomen, because oftheirsubmission, ought bethe dominantsex, andhe wantedto teach his adoles- cent male readersto respectfemale power. 35 Thelie detectorsimi larly produced truththrough submission. The device—featured in many ofthe early numbersof WonderW oman —extractedfrom its subjects their deepest desires, whatever their consciouswill. 36

True Crime

ForKeeler andLarson, i mmersedin the worldof crime-Žghting, the polygraph oVeredinsight into a di Verent sortof evasive individual.But the two reformersdid not just train their machine on criminals.Keeler andLarson also sharedan equalmistrust of old-timecops and municipal corruption and a corre- spondingrespect for August V ollmer,leaderof America’s interwarprogram of po- lice professionalization.

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 9 Municipalpolice forces had grown in the early-twentieth-centurytobecomea paramilitarypresencein largeA mericancities like Chicago.With the expansion ofurbancenters and their growing diversity and inequalities, the lawcourts in- creasinglyhad per mittedcity police to conduct their investigations and interro- gationsw ithoutaccording suspects the fullrange of Constituti onalprotections grantedby the Bill ofRights. In doingso, the magistrateshad freed the policeto operatein atwilightzone of quasilegality . 37 Theresult? By the timeofProhibition, abroadsegment ofthe Americanpublic had begun to despairof clearly distinguish- ingbetweenlaw-abiding citizens and sco Zaws,as well asbetweencops and crimi- nals.38 Vollmer’s programof police professionalization was intended to restorepub- licorder and public respect for the lawby making the policethemselves law-abiding. On the basisof thisprogram,V ollmerwas broughtdown from Berkeley in1924 tobe chief of police in Los A ngeles andclean upcorruption there. He triedto institutevarious refor ms.He implemented standardsfor police recruitme nt(in- cludingIQ tests), promotionon the basisof exams (withcivil ser vice protectionfor oYcers),and specialization of policetasks (includinga largeforensic science unit). Butthese reformsconjured up a passionateresistance among rank-and-Ž le cops andtheir political patrons, and V ollmerdid not remain in LosAngeles long.Back inBerkeley he continuedhis campaign for police reform on apublicstage. In his chapteron the ChicagoPolice for the inuential IllinoisCrime Survey ,he wrotethat ‘‘the fundamentalcause of the demoralizationof the policedepartment is corrupt politicalin uence.’ ’ 39 Vollmer’s programbelonged to a largerprogressive move- ment thatappealed to public respect for the ruleof law by attacking the graft,pa- tronage,and strong-ar mtacticsthat dominated municipal ‘ ‘machine’’ politics. 40 Atthe heartof this‘ ‘oldsystem’ ’ wasa patternof brutal interrogation. Early- twentieth-centurypolice were notorious for thei rbrutalitytoward suspects. When Larsonarrived in Chicagoin the late 1920s,he wroteto V ollmerthat the local methodof ‘ ‘elicitingevidence iswithrubber hose, black jack, and boot, and I have seen someŽ rst-handexamples.’ ’ Atalecturefor police recruits, he hearda chief detective justifythe beatingof suspects. 41 Vollmerand his fellowreformers viewed suchpractices as bothine Vectualand li kely toerode the public’s trustin lawen- forcementpersonnel. The national Wickersham reportof 1931—co-authored by Vollmer—condemnedthis quasi-o Ycial‘ ‘thirddegree,’ ’ andproposed that the judicialbranch reassert its authority over pretrial i nterrogation. 42 Larson’s study ofthe polygraph,published that same year,likewise builta case againstthe useof ‘‘thirddegree’ ’ methodsusing the classicEnlightenment argumentsagainst judicial tortureas apunishmentworse than any o Ycialsanction and aslikely toelicit unreli- ableconfessions. F orthese abuses,however ,Larsono Vered a diVerent solution.The useof the lie detector,Larsonsuggested, would put an endto such abuses and place the gatheringof testimony on ascientiŽc basis. 43 Itdidnotgo unnoticed that, should the policethemselves learn tomaster the newmachine, they wouldmaintain their monopolyon the interrogationof suspects.

10 Representations Keeler sharedLarson’ s disdainfor the nexus ofpoliticsand police work. Ac- cordingto Keeler ,the link betweenpolice appointments and the spoilssystem of municipalgovernment explained why the ‘‘humanmaterial’ ’ onthe forcewas cor- rupt‘ ‘upanddown the line.’’ 44 Intothis world of corruption,the lie detectorshone the brightlight of truth.Noting the underhandeddoings at the IllinoisState Peni- tentiary,Keeler boastedto his father that ‘ ‘[thanksto the lie detector],all thisis aboutto change. I amthe Žrstshot from the gunof destructionof political graft andthe constructionof orderlyscientiŽ c management.More and more of the ad- ministrationof this penitentiary w ill befrom this o Yce.’’45 Nowonder,then, thatmost police resisted the lie detector—at least initially. First,police resented outsidersstealing theirrole as guardiansof the publicorder . Even inBerkeley,old-stylecops resented Larson’ s collegiatedo-gooder attitude and his‘ ‘infallible’’ machine. 46 Andwhen NorthwesternUniversity soldits scientiŽ c crimelab to the ChicagoPolice Department, the policerefused to let Keeler take the helm, lest hismania for publicity drive him to steal all the creditfor solvi ng crimes.47 Second,police wielded authority on the streetsand in the station-house throughtheir discretionary power over the bodiesof subjects.Where policehad formerlycovered for one another and for their patrons, they nowfaced a threatto theirautonomy .Hence, even FBIdirectorJ. Edgar Hoover was skeptical about the lie detector.Ashe putit: ‘ ‘Ipersonallywould not want to accept solely whatthe operatorof alie detectorsays the instrumentshows in provingthat a manwas or wasnot a sex deviate.’’ 48 (Andgiven whatwe nowknow about Hoover’ s sexual predilections,no wonder .)Andthird, cops were the Žrstgroup of Americansto be routinely subjected tothe test.In the famous‘ ‘canarymurder case,’ ’ Keeler usedhis polygraphto extract a confessionfrom a copwho had stolen a$100‘‘trickcanary’ ’ froman estate he wassupposedto be guarding .Thepresiding judge Henry Horner predictedthat routine polygraph examination of policeo Ycerswould soon follow . 49 Indeed,in townssuch as Evanstonand Wichita, whenever disciplesof Vollmerwere namedpolice chief, they immediatelysubjected their subordinate o Ycersto routine polygraphtesting . 50 By 1933,Keeler wassubjecting Evanston’ s mayorand police chiefto lie detectortests onchargesof municipalcorruption. 51 Inshort,the lie detectorbelongs to that particular American strain of the En- lightenment projectwhich seeks toreplace personal discretion with objective mea- sures,and politics with science. This is nota projectthat appeals to politicians who wantto make patronageappointments or police o Ycerswho seek toselectively enforcetheir authority on the streets.By contrast,V ollmer,Larson,and Keeler campaignedfor the polygraphby appealing to apublicsentiment thatjustice de- pendedon adispassionatesearch for truth conducted by impersonal rules. Thislogic places the liedetectorsquarely in the currentof the early-twentieth- centuryA mericanpush for intelligence testingand post-T ayloristindustrial man- agement—techniques o Veredby the newly emergent disciplineof professionalpsy- chologistseager to sell theirser vices totheir patrons in the stateand corporate

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 11 administration.T otake the formerparallel Ž rst:the democraticappeal of these multiple-choiceintelligence tests (fromthe IQtest tothe SATs) residesin the way they ostensiblytreat all subjectsalike. 52 The ‘‘mechanical’’ waysuch tests canbe gradedis particularlyappealing in acountrylike the UnitedStates that does not trustits elites todistinguishamong its citizens and, therefore, insists on publicly veriŽable and quantiŽ able standards for judgment— even ifthismeans thatthese tests areshockingly remotefrom any functionaltest ofrealability .Taylorismtoo followsa similarlogic, in whicha scientiŽcally calculated‘ ‘onebest way’ ’ suppos- edly precludessubjective judgments about the value oflabor —while, ofcourse, standardizinga moreproŽ table degree of exertion.A llofwhichexplains oneof the centralappeals of the lie detectorin the UnitedStates: the charadethat it is the polygraph machine andnot the examiner thatassesses the veracityof the subject. Yet itis importantto understand that neither Vollmer’s programof policeprofes- sionalization nor the lie detector necessarily restrictedthe discretionof examiners.

TwoStrategies forExpertise

Andit was here thatLarson and Keeler partedcompany .Larsonpur- suedthe strategyof ‘ ‘openscience,’ ’ andKeeler ,the strategyof proprietaryknowl- edge.But I wishto emphasize that each strategydependedon the other,andeach waswracked by internal tensions not easi ly overcome. 53 The strategyof openscience assertsthat objective knowledge is producedwhen the scientist’s ‘‘disinterestedness’’ isguaranteedby a set ofinterlaced social mecha- nisms:(1) norms that denounce venality andreward priority of discovery,(2)the publicdissemination of thosediscoveries in journalsvetted byexpert peers,and (3) meritocraticinstitutions that translate those assessments intoa livelihoodand the resourcesto conti nueresearch. Under such a system, ascientist’s reputationis his orhermost prized possession. The question is: why wouldany societysponsor such knowledge?Princely states or privateuniversities might do soto enhance theirpres- tige,but this hardly accounts for the ratioof fundingawarded to the NationalSci- ence Foundationversus the NationalEndowment for the Arts.Infact,the di Verence infundingis largely duetothe additionalclaim (often advanced by scientists them- selves) thatscientiŽ c knowledgeis useful—ifnot immediately ,then overthe long haul.This, of course,begs the questionof usefulto whom and i nwhatway .The answerto these questionshas longimplicated scientists in pointingthei rresearch indirectionsthat ser ve theirpolitical and economic sponsors. W emay call thisthe ‘‘utility’’ dilemmaof objectiveknowledge. The strategyfor proprietary knowledge-maki ng takes thissocial uti lity asits startingpoint. Here, the aimis toextractrents (or coercive power) from knowledge asitissubstantiatedin productsor services,and this means holdingthat knowledge asprivate,so as notto dilute its marketvalue. One wayto do so istokeep the knowl-

12 Representations edge secret,like the early moderngui lds,the Coca-ColaCorporation, or the Man- hattanProject. The problemhere isthatthe holderof privateknowledge knows it isnoteasy tokeep asecret.And, for its part, society worries that valuable secrets willdie w iththeir possessor ,andwi llnever beusedto generate additionaluseful knowledge.That is why modernstates have createdpatent systems fortechnical knowledge.F orthe holderof private knowledge, the challenge then becomesdecid- ingwhen tokeep the informationsecret, when toapply for a patent(which requi res publication),and when torely onthe licensingof expertise.Behind thisproblem oftiminglies the problemof showing that this knowledge(or its substantiatedtech- nologies)can be applied by strangers, but this may mean givingmuch of the secret game away.Moreover,the licensingof proprietaryknowledge, such as the fees that expert witnessesdemand, give auditorsgood reason to doubt the impartialityof theirtestimony ,leadingexperts todemonstrate the extent towhichtheir knowledge iswidelyassented to.W emay call thisthe ‘‘publicity’’ dilemmaof usefulknowledge. The pointof thisall-too-bri ef analysis is not thatthese types ofknowledge- making exist inthese idealfor ms.Rather ,itis the uneasy hybridizationof these two strategies—and the resultingregime of intellectual property—that has produced ourcontradictory political economy of expertise.So let usnowsee howKeeler and Larsonpursued their respective strategies, Ž rstwith regard to the hardwareof the polygraph,and second with regard to the softwareof interrogation.

Hardware: ProŽts and Publicity, Priority and Reputation

Keeler chosethe strategyofproprietary knowledge. F orhi m,success meant seeing hislie detectorwidely employed —andcounting the remunerationin hispocket. That is why he soughta patent.During his Ž ve-year struggleto secure apatentfor his machine, he oscillated betweenproviding Larson with reports on hisprogress(under V ollmer’s scolding),and jealously guarding his methods. 54 After beingrepeatedly forced by the PatentO Ycetomoderate his claims ( hismachine incorporatedfew new principles), he Žnally secureda patentin January1931. 55 He then enlisted Western Electro-MechanicalCompany to manufacture the instru- ment.But Keeler insistedon retainingveto rightson every sale. He realizedthat the machineitself could not guarantee reliable results and that the reputationof hismachine (and hence itslong-ter msales prospects)might be damaged i fhe ‘‘turn[ed]outmachines promiscuously to untrained individuals.’ ’56 Forits part, the manufacturercomplained the lie detectionwould never beacceptedwidely untilKeeler solda standardizedinstrument— and standardization was di Ycult to achieve insmall productionruns. Behind Keeler’s moves wasa delicatebalancing act. He needed tohave areli- ablemachine and have itaccepted as the industrystandard. But he understoodthat

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 13 the bestway for him to make money fromthe device was,as he putit, to ‘‘control the instrumentand lease hisservices.’’ When the Walgreens department-storecom- panywanted to buy several ofhismachines and set uptheirownin-house security team,he refusedto sell them any detectorsand o Veredto consultfor them instead. Ashe conŽded to a closecolleague, he madeonly $125from each machinesold, andeach onehe soldcreated a competitor.Worstof all, each ofthese inadequately trainedoperators would damage the ‘‘reputationof the Želd’’ andthereby hurthis ownabi lity tosell hisser vices. 57 Then,when hispatent expired in the postwarperiod, Keeler switchedstrate- gies.Now he abandonedthe restrictionson sales andser vices,and told his manufac- turingcompany to ‘ ‘goahead and sell toanybody .’’ 58 Heshiftedthe focusof his businessfrom services totraining .Over the next fewyears he trainedlarge numbers ofoperatorsin shorttwo-week courses to runastandardizedmachine. 59 His school, Keeler Associates,was the Žrstto turn out polygraph operators i nbulk,and though Keeler provedan inadequatebusinessman, his younger associates, like JohnReid, successfullyexpanded on Keeler’s formulaafter he diedsuddenly in 1951. Throughoutthis period, however ,Keeler hadgreat success in publicizinghis services. Hescoredwrite-ups in Readers’Digest ,rodecelebrity cases, and fed the presstitillating stories about the machine’s abilityto ferret out marital inŽ delities. Andone of Keeler’ s early students—Chester Gould —wenton toinvent the comic stripcharacter Dick Tracy ,the personiŽcation of scientiŽc crime-Žghting and a tire- less advocatefor the polygraph.But if popularnotoriety was crucial to the poly- graph’s eVectiveness inthe marketplace(as wewi llsee), italsoproved its undoing inthe courtroom. By contrast,Larson took the routeof open science. He publishedhis resultsin journalsof criminologyand psycholog y.Ashe noted,it was priority‘ ‘whichmatters inscience.’’60 Andhe pridedhimself onhaving refusedthe ‘‘unethical’’ routeof patentinghis own (earlier) device. 61 Heworriedthat Keeler’ s policyof selling ma- chines andtrai ningoperators ‘ ‘superŽcially’ ’ would‘ ‘mess things up,’’ andthat Keeler’s maniafor publicity would give the newscience a‘‘badreputation.’ ’62 Yet Larsonalso recognized that Keeler hadproduced a standardpolygraph instrument withoutwhich researchers like himself couldnot hope for a science oflie detection. Thatis why he wroteto Keeler in1927,saying ‘‘Iwantedyou to handle the appara- tusend of it and derive whatever compensation might be. I couldthen devotemy time toclinical experimentation.’ ’63 Thispattern of mutual reinforcement and subversion emerged even more starkly when itcameto the ‘‘software’’ ofinterrogation.

The DiscretionaryExpert

Keeler usedthe relevant-irrelevant technique.Its goal was to calibrate the polygraphmachine for the individualbody by comparing a presumablyhonest

14 Representations answer(‘ ‘Y es, Ididhave acigarettethis morning’ ’) witha possiblelie (‘‘No,I did notcommit the murderlast Thursday’’). But,of course,a subject’s strongreaction toa questionabout murder ,say,mightsimply indicate that this topic was more stressfulthan the topicof morningcigarettes. One partialsolution was to calibrate the machineby obliging the subjectto tell alie. One ofKeeler’ s favoriteways to dothiswas the ‘‘cardtrick.’ ’ In thisruse, Keeler asked the monitoredsubject to select aplayingcard from a deck and,without naming the card,deny thatevery cardwas the correctone as the subjectreviewed the entiredeck cardby card— includingthe correctone. Keeler then examined the polygraphrecord and by read- ingthe squigglytraces that measured the reactionsof the subject’s body,identiŽed the selected card.This game hadtwo purposes: (1) it induced the subjectto tell a lie (andthereby establisha base-line ofmendacityto comparewith other possible lies), and(2) it convinced the subjectthat the operatorcould catch himorher telling alie (andthereby heighten the fear ofbeingcaught, and hence the chances of being caught).In actuality,Keeler wasableto pull o V the cardtrick only byhimself deceiv- ingthe subject—usually by marking the deck.The moregeneral versionof this same technique(still inusetoday as the ‘‘controlquestion test’ ’) istotrick the subject intotelling afalsehoodby asking adeliberatelyambiguous question like ‘‘Have you ever committeda crime?’’ aboutwhich the subjectwill presumably lie toapolice oYcer.64 Inshort,Keeler’ s polygraphtechnique depended on adisingenuousform of interrogationdesigned to create stressand an atmosphereof intimidation —andthis witha deŽnite purposein mind.F orstarters,many subjectsfeel compelledto com- ply witha requestfor a polygraphtest even thoughthe agreement tosubmit must bylaw be ‘ ‘voluntary.’’ Forinstance,prosecutors promise to release pretrialdetain- ees fromjail i fthey take andpass the exam. Andmany jobapplicants and employees understandablyfear retribution if they refuseto take the test—even when stateor federallaw formally prohibits employers from requiring one. 65 Andthen, afterthe exam isover,the examiner confrontsthe subjectwith the inky,graphicaltraces of whatthe examiner assertsto be the subject’s body’s supposedbetrayal (which, of course,the subjectis unableto read),andadvises the subjectto confess. Under the circumstances,many subjectsdo self-incriminate. In 1939Keeler privatelysurveyed thirteenmunicipal and state police units usinghis polygraph machine across the country.Thissur vey remainsthe largest andmost thorough study we have ofthe resultsof polygraphexami nationsof crimi- nal suspects.And asthe respondentsnever expected theirdata would be made pub- lic—indeed, they remainsafely storedin the archivesto date—the resultshave a certaincredibi lity,especially since the respondentsseemed unawarethat their methodsmight be consideredobjectionable. Of the nearly nine thousandsubjects examined, the policereported, 97 percenthad ‘ ‘voluntarily’’ agreedto take the test, only 1percenthad refused, and 2 percenthad confessed before even beingstrapped down.About one-third of the subjectswere labeled deceptive,of whoma stunning

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 15 average of60percentwere then persuadedto confess to the crime—with the dra- maticexception of amere 6percentat the handsof the IndianaState Police (an exceptionI explain later).A mongall these confessions,only 1percentturned out tobefalse—at least according tothe police.As for those labeled deceptivewho did not confess,roughly half wereconvicted and half hadtheir cases dismissed. Contrariwise,of the two-thirdsfound not deceptive, only atiny percentage(0.3 percent)went on tobe found gui lty.Ofcourse, such statistics are not meant betaken atface value. Butthese privatelygathered and unpublished data are far and away the bestportrait we areever likely tohave ofhow the policeused the polygraph. Andthey dotell atale. Forinstance, the examiner inMadison,Wisconsin, counted foursubjects who had confessed to acrimesimply upon being threatened w iththe machine—or Žve, he notedin the margin,if youcounted the onewho commit- tedsuicide. 66 So,despite having issueda formalprohibition on the useof the lie detectorin courtroomtrials, the judiciaryallowed police to usethe techniqueto screen sus- pects,determine their suitability for trial, and extract confessions. In thissense, Keeler devisedthe lie detectorto operateaccording to the same logicas ancienre ´- gimejudicialtorture. August V ollmercandidly called the lie detector‘ ‘amodiŽed, simpliŽed and humane third degree.’ ’67 Andthat is why the policehave ultimately welcomedthe technique.Keeler notonly madethe lie detectorinto an instrument almostanyone couldoperate, even aminimally trainedpolice o Ycer,butbecause ofthe wayhe conceivedof its operation,he alsoactually enhanced the discretionary powerof the examiner,whowas less interestedi nthe polygraphrecord per se than inusingit to intimidate the subjectinto confessing— which was ( please recall) the only sortof lie detectorevidence acceptablein courtduring the post-Fryeera. Thisis the reasonthe hardwareof the polygraphmachine has changedso little since the developmentof Keeler’ s Žrstdevice inthe 1930s,despite the tremendous progressof physiologicaland psychological knowledge since then. Given the nature ofthe ruse,the internalworking of the machinerywas almost beside the point. Thisis an example ofopeningthe technologicalblack box and Ž ndingit empty . Indeed,police examiners have sometimes goneso far as towringconfessions from suspectsby having them placetheir hands on aphotocopymachine, which they Žll withpaper printed with the word‘ ‘LIAR!’’68 Andthe onemajor technical innova- tionsince the 1930sisthe exceptionthat proves the rule.In the 1990snew computer algorithmswere developed that could mechanically amalgamateand assess the sub- ject’s physiologicalresponses, and data processors able to dojust that have been incorporatedinto the newest generationof machines.But, because these algorithms mightpreclude operators from accusing subjects of lying (whetherthe machine says they arebeing truthful or not),the topexaminers atthe Defense Department PolygraphInstitute tell me they usuallyturn the computero V.69 The continuedhold of whatMichel Foucaultcalled the culturede l’aveu —the cultureof confession— shows the extent towhich our justice system isstillinthrall

16 Representations tothe socialresolution provided by the dramaof confession,long after the demise ofthe ancienre ´gimeand the abolitionof judicial torture. In part,this can inter- pretedas partof the moderndrive for e Yciency:confessions are said by law-and- economyrationalists to save the police,the prosecution,and judiciary considerable timeandexpense. Indeed,as some90 percentof criminalconvictions in the United Statesare not won in aformal trial,but with confessions or pleabargains extracted priorto courtroom trial ( bargains thatare themselves, almostby deŽ nition, false confessionsby accused individuals to lesser crimes inthe interestsof mutualexpedi- ency), the lie detectorplays acrucialsorting role in Americanjustice even though ithas been bannedfrom the courtroom.At the same time,however ,the factthat the polygraphoperates beyond strict legal scrutiny(and outside of many ofthe pro- tectionsof the Constitution)allows the legal system towillfully ignorethe social andi nstitutionalcontext i nwhichthese confessionstake place.Suspects face tre- mendouspressure to take apolygraphtest andhave little controlover who adminis- tersthe test orhow.Indeed,this wi llful blindnesson the partof the courtsis part ofalargerlicense thatallows the policeto employ dissembling to ferret out the truth.70 Awayfrom the publicscrutiny a Vordedby the strictapplication of the me- chanicalrules of justice,then, the machineryof American law thrives onthe (poten- tially) coerciveexercise ofdiscretionarypower .

Trustin America

AsKeeler soondiscovered, these same principlescould be appliedin contemporaryrelations between employers and employees. During the GreatDe- pression,Keeler inaugurateda newphase in the useof the polygraphwhen he sold hisinterrogation techniques to corporatemanagers, a vast newmarket for decep- tiontesting that he wasthe Žrstto cultivate. In 1931,the Chicagorepresentative ofLloyd’ s InsuranceCompany approached Keeler inthe hopeof cuttingback on the $337million that American busi nesses losteach year dueto employee pilfering. The Lloyd’s representativeo Veredto reduce the insurancerates for those banks thatallowed Keeler toroutinelytest employees. There,in the supposedpursuit of aspeciŽc bankfraud, Keeler wenton Žshingexpeditions that typically revealed that20 to 30percentof bank tellers hadtaken pettysums of money atsome point intheircareers. Bank managerswanted to Žrethese employees—criminalizing whathad been, in e Vect,the customarypractice of pocketingthe occasionaltwenty- dollarbill— but Keeler urgedthat they beretai nedand retested every year.He assuredthe managersthey wouldhenceforth be the mosttrustworthy employees the bankhad. This also brought Keeler backfor another round of remunerative testing.71 Using the same sales pitch,Keeler alsogot the U.S.government to use the exam toscreen securityrisks in the 1940s,another market he wasthe Žrstto exploit.In

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 17 1945,he examined Germanprisoners of warbeingtrained in RhodeIsland to ser ve aspolicemenfor occupied Ger many.Keeler foundthat 36 percentharbored Nazi sympathies—or moreto the point,found it stressfulto be asked abouttheir politics. The administratorsof the programdenied that asking POWs torat out hidden Nazis(and Communists) among their comrades would give these futureGer man policemena false impressionof Americandemocratic values. 72 Keeler’s largeste Vortbegan in 1946at the Oak Ridgenuclear faci lities,where he initiateda programthat would eventually leadto the routineexamination of somesix thousandscientists, engineers, andlaborers— all employees ofthe Carbon CarbideCorporation, a subcontractorto the AtomicEnerg yCommission(AEC). Again,the goalwas less touncover speci Žcinstancesof fraudthan to enforce a new formof employee behavior.Outof the 690exams Keeler conductedin February 1946,he foundthat nine persons(or 1 percent)admitted to having ‘‘stolen product material.’’ Ineach case,however ,the infractionproved minor: in onecase atiny chipof uraniumhad been removedfrom the siteintentionally ,andin another,some workershad planted a tiny chipin the pocketof oneof theircomrades as ajoke. Farmore prevalent werethe customarypractices that always surround work sites; Keeler foundthat 10 percent of employees hadlied on theirjob application, 12 percenthad stolen tools, 3 percenthad used an aliasat some point in theirlife, and soon. F romthere itwas a shortjump to astillbroadertest ofreliability.Aset of 6,058exams taken duringa six-monthperiod in 1952(under Russell Chatham, Keeler’s successor)showed that the focushad expanded to includepolitical loyalty . Ofthosesubjects whose results were labeled ‘‘derogatory,’’ one-thirdwere found to have ‘‘friendsor relatives associatedw ithorganizations considered un-American.’ ’ Yet oncloserexamination, those who ‘ ‘sympathizedwith the Communistmove- ment’’ turnedout to include individuals who supported federally subsidizedhous- ingprojects and the Valley Authority. 73 Buteven asthisAEC programended in 1953amid charges of coercionand pseudoscience,the same McCarthyiteperiod saw a tremendousexpansion in the useof lie detectors.It was at this time that the StateDepartment began to use the machineto screen civilservants suspectedof homosexuality .Ofseventy-six poly- graphcases investigatedthere inthe early 1950s,seventy-four were ‘ ‘morals’’ cases, nearly half ofwhichresulted in confession,followed by either dismissal or resigna- tion.As DavidJohnson has pointedout, contemporary anxiety aboutthe collusion betweeninternal and external enemies led the stateto try to regulatedeviance of thoughtand deed in the entwinedways— sexual andpolitical— inwhich Cold W ar AmericadeŽ ned ‘ ‘normalcy.’’74 Alreadyat the timeEdward Shils had o Veredthe classicanalysis ofthis period when the state’s publicmechanisms of regulation over whelmedthe citizen’s private autonomy—all inthe name ofrootingout secret conspiracies and protecting all- importantstate secrets. 75 ButShi lsconsideredthis period an exceptionalbreak- downin the balancebetween legitimate state power and private i nterests,whereas,

18 Representations by1950 there wasno longer anything exceptionalabout the useof the lie detector. By the middleof the twentiethcentury ,the polygraphhad become a routinepart ofAmericanlife, deŽ ning a newnational normalcy .The polygraph’s very ubiquity suggeststhat we mustsearch for a moresystematic source of this pervasive new emphasison surveillance inAmericaninstitutional life. The Žrsthalf ofthe twentiethcentury saw the riseof vast newbureaucracies inmanagerialcapitalism and the national-securitystate. As economic and institu- tionalhistorians since R. H.Coaseand Alfred Chandler have noted,these newhi- erarchicalstructures emerged as asubstitutefor the informationcosts associated withrisky marketrelations. 76 Butcould managers really trusttheir salaried subordi- nates any betterthan they trustedsellers onthe openmarket? Here iswherethe lie detectorstepped in. Nolonger need managersrely onthe sortof sentimental avow- alsthatpassed for sincerity in Victoriantimes. 77 The twomillion annual polygraph exams beingconducted by midcentury suggests that managers expected the liede- tector—or itsintimidati ngshadow—toensure the reliabilityof thisnew white col- larwork force. Indeed,the lie detectorwas yet anotherar mament inthe newindustrial psy- chologydeveloped by Mu ¨nsterbergand his followers.Mu ¨nsterbergwas amongthe progenitorsof thisexplicit post-T ayloristattempt to suppressdeviance inthe work- place,especially socialcon ict and public displays of anger. 78 Aspartofthisbroader program,the liedetectorbelongs to are´gimeof emotionalsuppression, character- izedby Peter Stearns as the new‘ ‘Americancool,’ ’ andepitomized by the stereotypi- cal‘ ‘organizationman’ ’ ofmid-twentieth-centuryAmerica,trained in emotional opacity,yet alwayseager to exhibithis command of ‘ ‘positivethinking .’’ 79 The questionwas notjust one of enforcingemployee loyalty,butof the manage- ment andownership of knowledge. The Keeler polygraph,which thrived on the exercise ofaprivateand proprietary ski ll, proveda valuable toolto remind techni- calworkers that the knowledgethey createdwas the propertyof theircorporate or stateemployers. As its champions acknowledged, the mainvalue ofthe lie detector wasthatit ‘‘kept securityuppermost in aman’smind’’ andacted as a‘‘psychological deterrentto the disclosureof classiŽed infor mation.’’80 Thisform of employee se- crecywas the same intellectual propertyregime that fueled both the SecondIndus- trialRevolution and the Manhattan Project.

Lying onthe Couch

Oppositionto the lie detectorcame from several quarters.Libertarians like DwightMacdonald and the AmericanCivi lLibertiesUnion spoke out against the coercivepowers of the test.A ndthe AFL-CIOandother unions protested againstemployers’ right to polygraph their workers at wi ll. Butdecades of congres- sionalhearings and complaints made no headwayunti lthe late 1980s,when Presi-

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 19 dentReagan, furiousat regular leaks ofgovernmentinfor mationto the press,tried toexpandthe systematicuse of the lie detectorso that it might be used,at w ill, onall civilservants. Acongressionalbacklash led tothe Employee PolygraphPro- tectionAct of 1988.This law has somewhatcurbed the numberof tests performed inrecentyears, although signi Žcantexceptions per mitthe testingof suspects in policecustody ,all federalemployees, andany corporateemployee suspectedof malfeasance. 81 Butfor now ,let usreturnto the 1930sto examine the responseof JohnLarson. Larsonwas infuriated by Keeler’ s trickery.Inletters toVollmer,he bitterlyde- nouncedKeeler asthe ‘‘highschool boy in shortpants’ ’ he hadtrained, but who hadnever goneto medical school and had instead been luredby money . 82 He di- rectly wroteto Keeler toinsistthat he cease an exam the momentthe subject‘ ‘had any objectionto the test.’’83 HeadmonishedKeeler toalways ask speciŽc questions, suchas: ‘ ‘Didyou take $160,000?’’ not‘ ‘Have youever stolen anything fromthis bank?’’84 Andhe condemnedthe ‘‘unethicalcommercialization’ ’ Keeler hadculti- vated andexploited. 85 Foratimein the 1920s,Larson had himself hopedthat the polygraphcould besuccessfullysubjected to a ‘‘standardizedscientiŽ c controlledobjective evalua- tionof validity .’’ 86 ForLarson this meant Žrstof all thata conclusivepolygraph exam couldbe veriŽ ed by a genuine double-blindanalysis ofthe graphicalrecord, withoutmaking useof unrecordedclues obtained during the examination.But in 1927he wrotesourly to V ollmerthat under those conditions ‘ ‘noŽ ve ofuswho have donethe mostwork in thisŽ eld canagree on the interpretationof a record.’’87 Andsecond it meant securingindependent corroborative proof of whetheror not each suspectin aŽeld studyhad been lying, informationthat could be obtained only forcases thathad been ‘‘positivelycleared up,’ ’ usuallyby an independent confession.Y et Larsonrecognized the rarityof suchcases andthat even indepen- dentconfessions might be false. 88 By the early 1930she hadacknowledged that ‘‘all scientistsbecome suspicious of the techniqueand method of investigation if [thetechnique] shows up 100percentwhen dealingwith such factors as human emotions.’’89 Thisdid not mean, however ,thatLarson considered the polygraphuseless. Af- terall, nodisease had a sure-Žre diagnosis,yet physicianshad a profoundunder- standingof many illnesses. Inthe early 1930sat Chicago’ s Institutefor Juveni le Research,Larson inaugurated a noncoerciveclinical approach to the useof the polygraph,using a teamthat consisted of apolygraphexpert, a psychologist,a phy- sician,and a lawyer. 90 Intests ofjuvenile delinquentswho had been referredto the Institutefor apparently w ithholdinginfor mation,Larson noted that his team- basedpolygraph exams hadcleared 20 percent of cases andelicited confessions from33 percentof the youths.These confessions,Larson assured his readers,were securedwithout accusing the subjectof lying, norof misrepresenting the e Vec- tiveness ofthe polygraph.As forthe remaining47 percent of tests, Larson marked

20 Representations them ‘‘disturbed,’’ andsent them backto the juvenile courtswith a noticethat the polygraphtest notbe usedas evidence inthe case. 91 AndLarson persisted with this team-clinicalapproach in mental andcorrectional faci litiesthrough the 1940sand 1950s.92 He insistedthat the polygraphremain a toolof psychiatric diagnosis. A fter all, inmany psychiatrictraditions, a lie isnoless revelatorythan a trueutterance— tothe extent thatdistinguish ingtruth from falsity matters at all insuchcases. 93 Afterhis rebuke in the Frye case,Marston, too, primarily used the device aspart ofatherapeutic,psychological practice: to promote ‘ ‘healthy love adjustment’’ by uncoveringthe lies thatmen andwomen told themselves, thereby freeingthem from ‘‘twists,repression and emotional con icts.’ ’94 Butaside from the occasionaluse of physiologicalinstruments in biofeedbacktherapy ,thisis notthe sortof liedetector techniquethat has ourishedin America. Inany case,Larson insisted the polygraphmust not become a green lightfor criminalprosecution, nor a toolto force a confession.Indeed, the anomalouslow confessionrate among the IndianaState Police (found in Keeler’s survey of1939) canbe directly attributedto Larson,who had himself trainedthe examiner there. 95 Throughouthis career ,Larsonboth publicly and privately opposed Keeler’ s ongo- ing eVortsto introduce lie detectorevidence inthe courtroom. 96

Selective Acceptance

Thisbrings us to America’ s selective acceptanceof the lie detector— andto what this selective acceptancetells usaboutthe politicaleconomy of trustin the twentieth-centuryUnited States. The 1923Frye rulingdeclared that scientiŽ c evidence ‘‘mustbe su Yciently establishedto have gainedgeneral acceptancein the particularŽ eld inwhichit belongs.’’97 This,of course,begs the questionof who can claimto bethe relevant experts.Polygraph examiners have longclaimed that man- tle, andhave arguedvehemently thatthe polygraphis su Yciently reliablefor use incourt:98 percentreliable according to some of theirstudies. But the courtshave generally lookedto academic psychologists as the relevant experts tojudge poly- graphtechniques. 98 Andinsurveys ofthe Želd conductedat both ends ofthe Keeler- Larsonera (in 1926and agai nin1952),these academicpsychologists pronounced themselves skepticalabout the reliabilityof the technique. 99 Inthe 1980sand 1990s, these surveys themselves becamecontroversial as the twosides jockeyed for the right tospeak inthe name ofthe ‘‘relevant’’ experts. 100 The factis, however ,thatthe courtshave acceptedmany dubiousforensic sci- ences—such as handwritinganalysis, ballisticidenti Žcation,and forensic psychol- ogy—all ofwhichare treated with considerable skepticism outside the immediate circleof practitioners.Y et the polygraphalone hasbeen deniedadmission into the courtroom.This conti nueseven thoughthe Frye rulehas ceasedto hold sway .The FederalR ules ofEvidence (1977)and the recentDaubert decision (1993) indicate

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 21 the courts’wi llingness togrant trial judges a gatekeeping rolei nadmittingtesti- monythat might provide probative evidence forthe trierof fact.But the polygraph expert isstill generally barred,and the Sche Vercase, heard by the SupremeCourt in1999,fai led toclarify this issue. 101 Somecourts and commentators have hintedthat the ‘‘real reason’’ the liedetec- torhas been rejectedby the courtsis notits failings, but its power .Because the lie detectorgoes to the heartof the assessing adefendant’s innocenceor guilt,they argue,the polygraphexpert— were he orshe tobe believed—wouldbecome a super-expert,overshadowi ng allothertestiŽ ers, and unduly in uencing —oreven supplanting—the jury. 102 Keeler,ofcourse,hoped that juries would beconvinced bythe polygraph.Indeed ( like the NewYorkTimes of1911),he hopedto see the jury system abolished.He wantedcriminal cases triedby expert criminologistswielding apolygraph,w itha judgeto ruleon legal technicalities. 103 Inshort,he subscribed tothatstrain in the Western system ofjustice that sought to achieve certaintyof judgmentby setting aside human testimony as inherently untrustworthy(especially asassessed bylay citizens)in favorof reliablecircumstantial evidence (as itismade tospeak bydisinterested experts). This,even thoughthe instrument—even inthe handsof the bestoperator— was far from infallible. Perhapswe should not Ž ndit sosurprising, then, thatAmerican courts have continuedto reject the polygraphexam andto call it‘‘unreliable.’’ Remember,how- ever,thatthis repudiationfollowed not simply from Larson ’sassertionthat lie detec- tionwas‘‘unscientiŽc,’ ’ butfrom the success ofKeeler’s brandof salesmanship. That is,this paperhas madea historical argument.It does not deny thatjudges may have behaved rationallyin decidingthat the probativevalue ofpolygraph results has been outweighedby the possibility thatsuch results might mislead the jury.Itdoes suggest,however ,thatone cannot understand the basisfor this repudiationwithout comingto terms with the sort ofliedetector—and examiner —thatemerged in the UnitedStates. First,everyone agrees thatthe mainobstacle to crediblepolygraph tests isthe largenumber of incompetent examiners: 80percent of them,according to advo- catesof the polygraph. 104 Butit wasKeeler andhis studentswho initiated the quick commercialtraining of polygraphoperators and cultivated a vast marketplacefor the kindof expertise thatthrives onenhancing the discretionof the examiner (and hisorher employer). 105 Thatis, the polygraphonly succeedsat its principle task of extractingconfessions and intimidati ng subjectsbecause polygraph operators have deliberatelyshied away from even the mostbasic self-regulating norms and stan- dards.They arethe consummateantiprofessionals. A ndsecond, the factthat the general publicgives credenceto the lie detectormeans thatjudges wish to shield impressionablejuries from lie detectorevidence. Butit was Keeler andhis fellow pioneerswho had deliberately cultivated the publicmyth ofthe liedetector’s e Vec- tiveness, notonly toi ncreasethe demandfor their ser vices, butalso to make the lie detectorthat much more e Vective byenhanci ng itspower to intimidate subjects—

22 Representations andthereby toextract confessions, deter petty crimes, and enforce political loyalty . Hence, juristsbanned the lie detectorfrom the courtroomprecisely because of the mannerin whichit has thrivedin the commercialmarketplace. Indeed, the liede- tectoris thatparadoxical techno-science that works to the extent the popularcul- turehas been convincedthat it works— even thoughit works only becauseits opera- tors lie.

Conclusion

Thispaper has provideda historicalexplanation for a uniquelyA meri- cansocial practice: the polygraphtechnique for lie detection.It has alsotried to showhow this practice exempliŽ es the waymodern expertise emerges fromthe uneasy hybridizationof twostrategies for producing and validati ng knowledge.It shouldbe notedthat the institutionsof A nglo-Americanjustice are themselves the historicaloutcome of a tensionbetween this hybridized‘ ‘republicof expertise’ ’ and othersocial practices. The jurysystem isoneof the mostvisible democraticprac- ticesthat still values citizenparticipation as agoodin itself,as well asanextension ofthe claimthat justice depends on popular assent. The judgments of citizen-jurors areunderstood to be subjective both about matters of factand culpability; indeed, the system acknowledgesthat their decision-making is collective,consensual, unar- ticulated,unjustiŽ ed, and shielded from public view . Tobesure,the supposedvulnerability of lay judgmentsto demagogic appeals (as Keeler andmany othersfeared )has meant thatA nglo-Americanlaw has gener- atedelaborate rules of evidence toŽ lterwhat jurors may hearat trial, including expert testimony.Norshould we overlook the waysin whichthe demographic‘ ‘rep- resentativeness’’ ofthe juryand the forensicstrategies of advocateshave been in- creasinglyshaped by lawyers wieldi ng psychologicaland social-scientiŽ c theories assembledby the same disciplines thatgave usthe lie detector.AndŽ nally,we shouldnot forget that the lie detector—though banned from the courtroom—itself permitteda vast expansionin the numberof cases thatcould be settled without trial,substituting the supposedcertainty of aconfessionextracted by techno-science forthe lengthy,expensive, andunpredictable judgment of alay jury. Isitpossible that the eradescribed in thisarticle is comingto a Žtfulclose? On the onehand, the courtsare feeling agrowingpressure to allowpolygraph evidence intothe courtroom,and, on the other,the useof the lie detectorin the corporate setting wouldseem tobe diminishing.The Daubertruling of 1993has been read bysome lower courts to mean thatdefendants may usethe polygraphto exonerate themselves (an appealingstrategy since the polygraphmachine is generally more proneto false positives,meaning that the machine’s billofinnocence can be touted as‘‘extra reliable’’ ).Somebelieve thatprosecutors should be ableto introduce incul- patingpolygraph evidence aswell. In the process,some advocates of admitting

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 23 polygraphevidence, seeking tomollify the courts’anxieties thatthe lie detector willsupplantthe jury,have foundthemselves inthe paradoxicalposition of arguing thatthe general public’s growing skepticismabout results dressed up in the patina ofscience means thatlay juriesare less likely totake the wordof the polygraph operatoras gospeland, hence, thatsuch expert testimonywi llnotoverly persuade ajury’s decision.It would be ironici ndeedif judgesŽ nally admittedthe polygraph intotheir courtroom because they wereconvinced that no onebelieved itsresults anymore. Atthe same time,there has been adecline inthe numberof examinationscon- ductedby private corporations. In part,this may bedue tothe same general skepti- cismabout scientiŽ c claims.More directly ,itis due to the 1988Polygraph Act, whichgenerally restrictedthe ability ofprivateemployers to polygraph thei rem- ployees withoutcause (although they may stillinsiston apolygraphas acondition ofhiring,and the justiŽcation for a roundof polygraph testing is still fairly easy to trumpup). This double-switch —agrowingacceptance of the polygraphin the courtsand a decreasein itsuse in the privatesector —doesnot invalidate the gen- eral thesis ofthis article, as itmight Ž rstappear .Rather,itunderlines the article’s mostgeneral claim:that the riseof the lie detectorin twentieth-centuryAmerica wasahistoricalresponse to aspeciŽc set ofimperatives.These includedthe popular demandthat the statedraw a sharpline betweenlawful and unlaw fulbehavior (amongboth cops and criminals), the ongoingnegotiations that produced a partic- ularregime of intellectual property,andthe need forreliability in newhierarchical institutions.Binding these imperativestogether has been the penchantfor A meri- caninstitutions to preferto presentto the publica faceof objectivityand rule-bound fairness,even asthe distributionof power and social rewards are handled, o Vstage, throughinformal and unregulated arrangements. The resultingculture of (mis)- trustis increasingly under new pressures. The rulesof intellectual propertyare beingrewritten. The hierarchicalinstitutions of the twentiethcentury are giving wayto corporateorganizations that allow for quasi-market relations within the Ž rm, sothat some employees— especially thosewith the ability towithholdor resell their skills orintellectual property—are entitled to actwith relative autonomy.Insuch auniverse,the lie detectorseems aterriblycrude instrument with whichto enforce loyalty.Indeed,the very notionof loyalty has somethinganachronistic about it. Only inthe realmof law enforcement does the useof the lie detectorseem to begrowing.Publico Ycials(including prosecutors, spy-catchers, and antiterrorism experts) rushto reassure the publicthat no e Vortwi llbesparedto sift truth from lies. Andas proofof their diligence, they announcethat the lie detectorw ill be copiouslyadministered. Despite highly publicizedexpose ´softhe polygraph’s fail- ures(i nthe Wen HoLee case,and in the AldridgeAmes case), the instrumentis still trottedout as the goldstandard in high-proŽle criminalcases, as awayto plug securityleaks, andas an instrumentto extractthe truthfrom those suspected of threateningA merica’s safety.The historyof untruth is anever-ending tale.

24 Representations No t e s

Anearlierversion ofthis paperappeared in printas Ken Alder,‘‘The Honest Body: The PolygraphExam andthe Marketingof American Expertise,’’ Historical Reec- tions/Re´exions Historiques 24(1998):487– 525. Permission toreprint previously pub- lished materialfrom that article has beengranted by the journal.Research forthis paperhas alsobeen generously supportedby the NationalScience Foundationand by avisitingfellowship atthe American BarF oundation. 1. Washington Post ,8January 2000,A2. 2.Protests fromLivermore scientists have somewhat limitedthe scope ofthe polygraph testing there; see Department ofEnerg yPublic Hearing onPolygraph Examination Proposed Rule,Livermore,Californ ia(14 September 1999). Available http://www.spse.org/ Polygraph comments Livermo.html . 3.Eric Schmitt, ‘‘Security Moves Means Lie Test for50 0atF .B.I.,’’ NewYork Times , 25 March 2001,A19. Aldridge Ames has taken tospeaking out from prison against the eYcacy ofthe liedetector test. 4.Steven Shapin, ASocial History ofT ruth: Civility and Science inSeventeenth-CenturyEn- gland (Chicago,1994). 5.F orthe literatureon the moralcomplications of the administrationof placebos,see John Forrester, Truth Games:L ies, Money,and Psychoanalysis (Cambridge,Mass., 1997). 6.Karen Halttunen, ConŽdence Men and Painted Ladies: AStudy ofMiddle-Class Culture in America,1830 –1870 (New Haven, 1982). 7.Saint Augustine, ‘ ‘Lying,’’ trans. Mary SarahMuldowney ,and‘ ‘Against Lying,’’ trans. Harold B. JaVee, in Treatiseson Various Subjects ,ed.Roy J .Deferrari( New York,1952), 47–179. Immanuel Kant, TheGroundwork oftheMetaphysics ofMorals ,trans. anded., MaryGregor(Cambridge, 1998). F orasurvey ofcurrent moralphilosophizing about lying,see Sissela Bok, Lying: Moral Choicein Public and Private Life (New York,1989). Forananthropologicalaccount oflying, see F.G.Bailey, ThePrevalenceof Deceit (Ith- aca,N.Y .,1991).F orasociologicalaccount, see J.A.Barnes, APackof L ies: Toward a Sociology ofL ying (Cambridge,1994). 8.Joseph Brodsky, LessT han One: SelectedEssays (New York,1986), 7. 9.Caspar Lavater , Re`gles physiognomiques,ou observations sur quelques traits caracte´ristiques (The Hague,1803). On the history ofphysiognomy,see PhillipProger ,‘‘Illustration asStrategyinCharles Darwin’s‘The Expression ofEmotionsin Man andAnimals,’ ’’ in Inscribing Science:ScientiŽ c Texts and theMateriality ofCommunication ,ed.Timothy Lenoir(Stanford, 1998), 140– 81. 10.On the theory ofjudicialtorture in the ancien re´gime,see John H.Langbein, Torture and theL awofProof: Europe and England in theAncien Re´gime (Chicago,1977); on its practice ineighteenth-century France, see Richard Mowery Andrews, Law, Magis- tracy,and CrimeinOldRegime Paris, 1735–1789 (Cambridge,1994). On the rise ofprob- abilistic thinking andits impacton legalanalysis, see Ian Hacking, TheEmergenceof Probability: APhilosophicalStudy ofEarly IdeasAbout Probability,Induction, and Statistical Inference (Cambridge,1975), and Lorraine J .Daston, Classical Probability intheEnlight- enment (Princeton, N.J.,1988). 11.Barbara J .Shapiro, BeyondReasonable Doubt and ProbableCause: Historical Perspectiveson theAnglo-Americ anLawofEvidence (Berkeley,1991). 12.T alGolan,‘ ‘ScientiŽc Expert Testimony inAnglo-American Courts,1782– 1923’ ’ (Ph.D.diss.,University ofCalifornia,Berkeley ,1997).

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 25 13.On the relationshipbetween the legalperson and his/ her body,see Alan Hyde, Bodies of Law (Princeton, N.J.,1997). 14.Anthony Gale, ‘‘Introduction:The PolygraphT est, Morethan ScientiŽc Investiga- tion,’’ in ThePolygraph Test: Lies, Truth, and Science ,ed.Anthony Gale (London,1988), 7.The American PolygraphAssociation boastsroughly three thousandmembers, trainedi nsome thirty accredited,privately owned schools o Vering fourteen-week courses. GordonH. Barland,‘ ‘The PolygraphT est inthe USAandElsewhere,’ ’ in Polygraph Test , 75. 15. OYce ofT echnology Assessment (OTA), U.S.Congress, ScientiŽc Validity ofPolygraph Testing: AResearchRevie wandEvaluation—AT echnical Memorandum (Washington,D.C.: GPO,1983).The backgroundto this OTAstudycan befoundin Jack Brooks,‘ ‘Poly- graphT esting:Thoughts ofa SkepticalLegislator ,’’ andLeonard Saxe, Denise Dou- gherty,andTheodore Cross, ‘ ‘The Validityof Polygraph T esting: ScientiŽc Analysis andPublic Controversy ,’’ AmericanPsychologist 40(1985):348– 66. 16.David Thoreson Lykken, ‘‘The CaseAgainst PolygraphT esting,’’ in Polygraph Test , 117.See also David Thoreson Lykken, ATremorin theBlood: Usesand Abusesof theL ie Detector (New York,1981). 17.The few countries outsidethe that make limiteduse ofpolygraphtests have adisproportionatelysmall numberof examiners, have intimatesecurity ties with the United States,and have only taken an interest inthe technique inthe last decade. Barlandestimates that there areapproximately 110– 120 examiners inCanada,90– 120in Japan,90– 110 in Turkey,40–50 inSouthKorea, and 40– 45 inIsrael; Barland, ‘‘PolygraphT est inthe USA,’’ 77.Areview bythe British government in1985repudi- atedthe polygraphtest; Gale, introductionto Polygraph Test , 1. 18. Fryev. United States ,293F ed.1013 (1923). The Frye rulingcame froma district court andso did not legally bindstate ormost federaljurisdictions. However ,many courts citedF rye intheir decisions onscientiŽc testimony andespecially onpolygraphevi- dence. The sole exception tothe general banon polygraphevidence (underthe Frye rule) iswhen prosecutionand defense stipulatein advance that they willbeboundby atest andspecify carefully the terms underwhich itwillbecarriedout; see Lawrence Taylor, ScientiŽc Interrogation:Hypnosis, Polygraphy, Narcoanalysis, VoiceStress, and Pupillo- metrics (Charlottesville, Va.,1984), 247– 92. 19. Daubert v. Merrell Dow Pharmaceuticals, Inc. ,509U.S.579 (1993). The interpretationof Daubertis still beingworked out. Some recent federalcircuit courtdecisions have citedDaubert in orderingtrial judges to set asidethe perse exclusion ofpolygraph evidence; see United Statesv. Posado ,57F .3d428 (5th Cir .1995); United Statesv. Galbreth , 908F .Supp.877 ( D.N.M. 1995);and United Statesv. Crumby ,895F .Supp.1354 ( D. Ariz. 1995).However ,otherjurisdictions have repudiatedthis stand anda Yrmed the perse exclusion; see United Statesv. Kwong ,69F.3d336 (2d Cir .1995). 20.The case of United Statesv. Edward G.Sche Ver beganwhen amilitary o Ycer (ScheVer) triedto introduce a polygraphexam— conducted at the military’s own insistence— toexonerate himself inacourtmartial. Whi le the courtmartial was willing tohear evidence that Sche Ver’s urinehad tested positivefor drug use, it refused to allow testi- mony bythe polygraphexaminer indicatingthat Sche Ver hadnot been deceptive when he deniedusing drugs. On appeal,the topmilitary court called thisanunconsti- tutionalinfringement onthe rightof the accused topresent adefense,and cited Daub- ert; see United Statesv .Sche Ver,44M.J.442(1996). 21. United Statesv. Sche Ver,96U .S.1133 (1998), see Stevens’s dissent. Fororalarguments, see JoanBiskupic, ‘ ‘Justices VoiceConcerns AboutPolygraphs,’ ’ Washington Post , 4

26 Representations November1997, A5; and Linda Greenhouse, ‘‘Justices Grapplewith Merits ofPoly- graphsat T rials,’’ NewYork Times ,4November1997, A14. 22.Etienne-Jules Marey,‘‘Etudes graphiquesdes movements respiratoire,’’ Journal del’a- natomie etdelaphysiologie 2(1865):276– 301, 425– 53; 3 (1866):225– 42, 403– 16. F or the sphygmograph,see RobertG .Frank Jr.,‘‘The Telltale Heart:Physiological Instru- ments, Graphic Methods,and Clinical Hopes, 1854– 1914,’ ’ in TheInvestigative Enter- prise: ExperimentalPhysiologyinNineteenth-CenturyMedicine ,ed.Wi lliam Colemanand Fredric L.Holmes (Berkeley,1988),211– 90. American physicians only beganto tenta- tively employquantitative measures ofblood pressure inthe early twentieth century, andthe valueand interpretation of this new informationwas still controversial;see Hughes Evans, ‘‘LosingT ouch:The Controversyover the Introductionof Blood Pres- sure Instruments intoMedicine,’ ’ Technology &Culture 34(1993):784– 807. 23.William James, ‘‘What Is an Emotion?’’ Mind 9(1884):188– 205; William James, Prin- ciplesof Psycholog y (1890;reprint, New York,1950), 2:4 42–87. 24.On HugoMu ¨nsterberg’s laboratory,see the outputof his Harvard Psychological Studies , 5vols.(Lancaster ,Pa.,1903– 22). On Mu¨nsterberg’s program,see DeborahCoon, ‘‘Standardizingthe Subject:Experi mental Psychologists, Introspection, andthe Quest fora TechnoscientiŽcIdeal,’’ Technology&Culture 34(1993): 757– 83; and JuttaSpi llman andLothar Spi llman, ‘‘The Rise andFall ofHugo Mu ¨ nsterberg,’’ Journal ofthe Histor yofthe Behavio ral Sciences 29(1993):329 –30.On the science of emotionologyinthe early twentieth century,see Otner E.Dror,‘‘Creating the EmotionalBody: Confusion, Possibi lities,and Knowledge,’ ’ in An Emotional His- tory ofthe United States ,ed.Peter N.Stearns andJan Lewis (New York,1998), 173–96. 25.F oroneover view ofthe applicationof psychological science tothe credibility oftesti- mony inEurope,see MattK. Matsuda, TheMemoryof theModer n (New York,1996). Inthe 1890sand early 1900s,some prominentEuropean criminologists and psycholo- gists—among them CesareLombroso and C. G.Jung—deployed physiological in- struments toverify the veracity ofhuman testimony,with aneye onintroducingthe results asevidence intothe courtroom.Not only were their e Vorts rebuVed by the Europeancourts (much aswere the e Vortsof their American counterparts),but their techniques alsofai ledto Ž ndacceptance amongpolice administrators and govern- mental andindustrial elites (quiteunlike the case inAmerica). Foruse ofthe blood pressure gauge,see Dot.Cougnet and Cesare Lombroso, ‘ ‘SŽgmograŽ a didelinquenti edalienati,’ ’ Archivio dipsichiatria,scienze penali edantropologia criminale 2(1881):234– 35,472.Vittorio Benussi later followedup onthe work ofhismentor,Lombroso,and madesimilar observationsabout the e Vect ofdeceitfulness onbreathing patterns; see VittorioBenussi, ‘ ‘DieAtmungsymptome derLuge,’ ’ Archiv fu¨rdiegesampte Psychologie 31(1914):24 4–73. Finally ,the methodof skin resistance was usedby C. G. Jung;see Frederick Peterson andC. G.Jung,‘ ‘Psychological Investigations withthe Galvanom- eter andPneumograph in Normaland Insane Individuals,’’ Brain: AJournal ofNeurol- ogy 30(1907): 153 –218. 26. Hugo Mu¨nsterberg,‘ ‘Experiments with Harry Orchard,’’ inHugoMu ¨nsterberg Pa- pers,Boston Public Library ,Boston,Mass. (hereafter HMP),2450. On the publicity that followedMu ¨nsterberg’s declaration,see Boston Herald ,3July 1907,1. New York Times,5July 1907,6. Foranaccount ofthe Orchard/Haywoodtrial, see J.Anthony Lukas, Big Trouble: AMurder inaSmall WesternTown SetsO V aStruggle forthe Soul of America (New York,1997). 27. Hugo Mu¨nsterberg, On theWitness Stand: Essays onPsychologyand Crime (New York,

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 27 1908).See also John Henry Wigmore’s witty defense ofjudicialcaution in the face of new science; John Henry Wigmore,‘ ‘ProfessorMuensterberg and the Psychologyof Testimony,’’ NorthwesternUniversity LawReview 3(1909):399 –445. 28.‘ ‘Electric Machines,’’ NewYork Times ,10September1911, 6. 29.Wi lliam M.Marston,‘ ‘Systolic BloodPressure Symptomsof Deception,’ ’ Journal of Experimental Psychology 2(1917):117– 63. F orMarston’s workunder the Army’s Na- tionalResearch Councilduring W orldWar I, see WilliamM. Marston,‘ ‘Reaction- Time Symptomsof Deception,’ ’ Journal ofExperiment alPsychology 3(1920):72– 87. 30.Marston, ‘ ‘Symptomsof Deception,’ ’ 162. 31.National A rchives andRecords Administration,W ashington,D.C.( hereafter NARA), RG21CaseŽ le 38325: United Statesv .Frye ,U.S.District Courtfor the District ofColumbia, 17– 20 July 1922,11– 18. See also NARA RG276U .S.Court of Appeals forthe District ofColumbia,Case Ž le 3968;and RG 204U .S.Pardon Attorney Case Žle 56–386. F orasummary ofthe Frye case, see J.E. Starrs,‘ ‘‘AStill-Life Watercolor’: Frye v.United States,’’ JournalofForensic Evidence 27(1982):684– 94; and Golan, ‘ ‘Sci- entiŽc Expert Testimony,’’ 358–82. 32.F orLarson’ s retrospective account,see John AugustusLarson Papers,Bancroft Li- brary,University ofCalifornia,Berkeley (hereafter JLP), c. 2,f. misc.: Larson,‘ ‘Taken bythe DipMob,’ ’ n.d.A cleaned-upversion ofthe case wasreportedin John A.Lar- son,‘ ‘Cardio-Pneumo-Pyschogramin Deception,’’ Journal ofExperime ntal Psychology 6(1923):436– 40. 33.Alva Johnson, ‘‘The MagicLie Detector,’’ Saturday Evening Post ,15April194 4,9– 11, 72;22 Apri l1944,26–27,63; 29 April194 4,20,101– 2. 34. See GeoV Bunn,‘ ‘Constructingthe Suspect:A BriefHistory ofthe Lie Detector,’’ BorderLines 40(1996):5– 9. F orahistorical treatment ofthe gendereddivision between natureand its investigators,see L.J.Jordanova, Sexual Visions: Images ofGenderin Science and Between theEighteenth and theTwentieth Centuries (Madison,Wisc., 1989). Thoughsuch genderdualities are largely social conventions, many modernscientists, such associobiologists,still invokethem. 35.Wi lliam Marston, Wonder Woman 5(April/May 1943);reproduced in WilliamMoul- tonMarston, Wonder Woman ,intro.Gloria Steinem (New York,1972). F orMarston’s psychology,see William MoultonMarston, Emotions ofNor mal People (New York, 1928);and , C. Daly King,and Elizabeth H.Marston, Inte- grative Psychology: AStudy ofUnit Response (New York,1931). F oran analysis of Wonder Woman andMarston, see Geo V Bunn,‘ ‘The Lie Detector, Wonder Woman ,andLiberty: The Lifeand W orkof William Moulton Marston,’ ’ Historyofthe Human Sciences 10 (1997):91– 119. 36.Wi lliam MoultonMarston, TheLieDetectorT est (New York,1938). 37.Akhil ReedAmar , TheConstitution and Criminal Procedure:First Principles (New Ha- ven, 1997). 38.Claire Bond Potter , Waron Crime: Bandits, G-Men, and thePolitics ofMass Culture (New Brunswick, N.J.,1998). 39.Berkeley PoliceDepartment Papers, Bancroft Library ,University ofCaliforn ia, Berkeley (hereafter BPP),c. 8,f. Police T raining School:[ Vollmer],‘ ‘Outline of Coursesof Instruction fora Schoolfor Police O Ycers, 1916–1917,’ ’ December 1916. ForV ollmer’s blueprintfor Los Angeles, see AugustV ollmer, LawEnforcementin Los Angeles,intro.Joseph G. Woods(1924; repri nt, New York,1974). August V ollmer, ‘‘The ChicagoPolice,’ ’ in TheIllinois Crime Survey ,ed.John Henry Wigmore(Chicago, 1929),372. Over the years, Vollmer’s many students tookup managerial positions in

28 Representations many ofthe leadingpolice forces inthe nation;e.g., O .W.Wilson, whoserved aspolice chief inChicago1960 –69. 40.Gene E. Carteand Elaine H.Carte, PoliceRefor min theUnited States: TheEraof August Vollmer,1905–1932 (Berkeley,1975).Samuel Walker , ACritical HistoryofPolice Refor m: TheEmergenceof Professionali sm (Lexington, Mass., 1977).Eric H.Monkkonen, Police inUrban America,1860 –1920 (Cambridge,1981). 41.BPP b.10, f. Larson: Larson toV ollmer,17October 1924. 42.Wickersham Commission, Report onPolice ,FourteenthReport of the NationalCom- mission onLaw Observance andEnforcement ...Under the Directionof August Vollmer (Washington,D.C.: GPO, 1931).See also the popularcondemnation in Emanuel H.Lavine, TheThird Degree: ADetailedand Appalling Expose´ ofPolice Brutality (Garden City,N.Y.,1930). 43.John AugustusLarson, Lying and Its Detection: AStudy ofDeception and Deception Tests , withGeorgeW .Haney andLeonarde Keeler; intro.by August V ollmar[ sic] (Chicago, 1932),65 –121. 44.Leonarde Keeler Papers,Bancroft Library ,University ofCalifornia, Berkeley (here- afterLKP), c. 2,f. Original: Keeler ,‘‘PoliceSystems ShouldBe Divorced from Poli- tics,’’ [1930].This undergraduatepaper was written inthe 1930swhen Keeler was alreadyemployed by Northwestern’ s ScientiŽcCrimeLab, was anationalŽ gurefor his lie detectorwork, and had returned to Stanford to pick upa few last credits for his degree. 45.Charles Keeler Papers,Bancroft Library ,University ofCalifornia,Berkeley (hereafter CKP),b. 7, f. L. Keeler: L.Keeler toC. Keeler,18July 1929. 46.JLP c. 2,f. misc.: Larson,‘ ‘Taken bythe DipMob,’ ’ n.d. 47.Author’ s interview with FredInbau, 5 September1996. See also the corroborating letter ofInbau to Charles M.Wilson,9 June 1938(from the personalŽ les ofInbau, copyin author’s possession). 48.Quoted in DwightMacdonald, ‘ ‘The Lie Detector Era,’’ TheRe porter ,8June 1954, 10–18; 22 June 1954,22– 29, here 26. 49.Leonarde Keeler ,‘‘The CanaryMurder Case,’ ’ AmericanJournal ofPolice Scie nce 4 (1930):381– 86; reprinted in Polygraph 23(1994): 145– 48. ‘ ‘Science Solves Canary Death,’’ Chicago Tribune ,24May 1930.BPP b.10, f. L. Keeler: Keeler toV ollmer,17 August1930. 50.Leonarde Keeler ,‘‘The Lie-Detector Proves Its Usefulness,’’ Public Management 22 (1940):163– 66; reprinted in Polygraph 23(1994): 185. BPP b.16, f. Wi ltberger:Wilt- bergerto V ollmer,17June 1924;26 October 1926;13 December 1926. 51.CKP b.7, f.L. Keeler: L.Keeler toC. Keeler,20April 1933; L. Keeler tohis mother, 11 May 1933. 52.F orintelligence testing andapplied psychology inthis period,see F.Allan Hanson, Testing Testing: SocialConsequences ofthe Examined Life (Berkeley,1993);John Carson, ‘‘ArmyAlpha,Army Brass, andthe Search forArmy Intelligence,’’ Isis 84(1993):278– 309;Michael M.Sokal,ed. Psychological Testing and AmericanSociety ,1890–1930 (New Brunswick, N.J.,1987);Nicolas Lemann, TheBigTest: TheSecretHistor yoftheAmerican Meritocracy (New York,1999). The democraticappeal of such atest does not preclude the possibilitythat the test itself favorscertain kinds ofsubjects overothers, nor that the test doesnot measure any relevant qualityor talent. SeeKen Alder,‘‘Engineers Become Professionals,or ,HowMeritocracy MadeKnowledge Objective,’ ’ in The Sci- encesin Enlightened Europe ,ed.William Clark, Jan Golinski, andSimon Scha Ver (Chi- cago,1999), 94– 125.

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 29 53.The followingdiscussion represents my ownblend of recent scholarship inthe history ofobjectivity .SeeTheodore Porter , Trust in Numbers: QuantiŽcation in Scienceand Public Life (Princeton, N.J.,1995);Lorraine Daston, ‘ ‘Objectivity andthe Escape fromPer- spective,’’ Social Studies ofScience 22(1992):597– 618’ ’; Ken Alder, Engineering theRevo- lution: Armsand Enlightenmentin France,1765 –1815 (Princeton, N.J.,1997);and P .A. David,‘ ‘CommonAgent Contractingand the Emergence ofan ‘Open Science’Insti- tution,’’ AmericanEconomic Review 88(1998): 15– 21. 54.JLP c. 7,f.corr .misc.: Keeler toLarson, 1 May1927; 19 April1928. Keeler cut out his coworker,Dr.Edwards ofUCLA,when he learnedthat Edwards wantedto go into partnershipto manufacture the machine; JLPc. 7:Keeler toLarson, 29 April1924. 55.Leonarde Keeler ,‘‘Apparatus,’’ U.S.Patent, No. 1,788,434; Ž led30 July 1925;granted 31January 1931. 56.LKP c. 1:Keeler toW .J.Foster( Western Electro-Mechanical Company),23 Apri l 1931.Keeler aggressively blockedAssociated Research (his second supplier)from manufacturingpolygraph machines designedby his rivalinventors; LKPc. 2,f.busi- ness: Inman toKeeler ,with notes byKeeler ,13August1948; Inman toReid,17 Au- gust1948; Inman toKeeler ,25August1948. 57.LKP c. 2,f. technique: Keeler toWi lson, 20July 1931. 58.LKP c. 2,f. business: Inman toKeeler,with notes byKeeler ,13August1948. 59.Between 1947and his deathin 1950,Keeler’ s income fromcasework droppedfrom $32,000/year to$20,00 0/year,while training fees rosefrom a negligiblesum to $45,000/year;LKP c. 2,f. busi ness: LeonardeKeeler ,Inc., ‘‘Balance Sheet,’’ 31 March 1950. 60.JLP c. 7,f.corr .misc.: Larson toCharles DeLacy,5February1941. 61. Ibid. 62.JLP c. 7,f.corr .misc.: Larson toKeeler ,21December 1931. 63.JLP c. 7,f.corr .misc.: Larson toKeeler ,23March 1927. 64.The ControlledQuestion T echnique (CQT)was developedby John Reidin the mid- 1940s.John E.Reid,‘ ‘ARevised QuestioningT echnique inLie-Detection Tests,’’ Journal ofCriminal Law, Criminology, and PoliceScience 44(1947):542– 47. 65.These restrictions have increased signiŽcantly since the federal1988 Employee Poly- graphProtection Act. 66.Keeler Papers,Department of Defense PolygraphInstitute, Ft. Jackson, S.C.(here- afterDoDLKP). Keeler survey of1938–40,responses fromthe policedepartments in East Cleveland,T oledo,Indiana State, Indianapolis, Kansas City,Bu Valo,Honolulu, Madison,Michigan State,St. Louis, North DakotaState, Cincinnati, Wichita. 67.‘ ‘PoliceT ryMachine toDetect Lies,’’ LosAngeles Times ,25January 1924. 68.David Simon, Homicide: AYearon theKilling Streets (Boston,1991). 69.Personal communication,DoD polygraph operator ,June 1998. 70.This continues despitethe famousMi randaruling of the mid-1960s.F redE. Inbau, John E. Reid,and Joseph P.Buckley, Criminal Interrogationand Confessions ,3ded. ( Balti- more,1986). 71.LKP c. 2,f. technique: Keeler toNewman Baker,21July 1933.LKP c. 1,f.research: Keeler toHenry ScarboroughJr .(Lloyd’s ofLondon), 10 June 1931,10 F ebruary 1934;Keeler toNagle, n.d. F oranadvocateof the use ofthe polygraphin business, see the popularwrite-up of Keeler’s work inJ.McEvoy,‘‘Lie Detector Goes intoBusi- ness,’’ Readers’Digest 38(1941):69– 72. F oran article opposingthe use ofthe polygraph inbusiness, butwritten in1962,see Richard A.Sternbach, Lawrence A.Gustafson,

30 Representations andRonald L. Colier,‘‘Don’t Trust the Lie Detector,’’ Harvard Business Review 40 (1962):127– 34. 72.NARA RG389(1941– 58), box 40, Ž le Lie Detector Tests: ‘‘Memorandumto Col. Franklin W.Reese,’’ 30August 1945. F orKeeler’ s account ofthese state-sponsored examinations,see LKPc. 1,f. training: Keeler ,‘‘Training,’’ 1949. 73.NARA RG326,box 48 631,4/ 8/8:USAEC,‘‘Use ofLie Detector atAEC Installa- tions,’’ 24March 1953,Keeler data,2; 1952data, 70– 77. F oranexplanationfor the ‘‘thefts,’’ see NARARG326,box 149 631 4/ 8/4:USAEC,In the Matterof ‘ ‘Lie- Detector’’ Panel Meeting,24 January 1952:Hardie, 42. F orthe deeperinvestigation ofthe politicalviews, see SouthEast RegionalO Yce ofthe NationalArchives and Records Administration,East Point,Ga. ( hereafter SE-NARA), RG326:F .P.Cal- laghan toJ .S.Denton,‘ ‘Interview ofJoseph E.Deye Concerning His Polygraph Exam,’’ 28May 1951.F oranexpose´ ofthe OakRidgeprogram, see Anthony Leviero, ‘‘U.S.T ests Sta VsbyLie Detectors,’’ NewYork Times ,20December 1951,1, 20.F or the decisionto end the program,see USAEC Doc.#708937 ‘ ‘Advisory Councilfor BiologyandMedicine,’ ’ 4–5 April1952. F ora partialhistory ofOakRidge,see John G. Linehan, ‘‘The Oak RidgePolygraph Program, 1946– 53,’ ’ Polygraph 19 (1990): 131–37. Keeler initiatedthe program,but he was forcedaside by a rivalpolygraph operator,Russell Chatham,in the late 1940s. 74.NARA RG59(1953–60), box 12, Ž le Bureauof Security: Thurston Morton(Asst. Sec. ofState) to Sen. Olin Johnston, 30June 1955,‘ ‘Polygraph,’’ [1955].David K. Johnson, ‘‘The Lavender Scare:Gays andLesbians inthe Federal CivilSer vice, 1945–1975’’ (Ph.D.diss., Northwestern University,2000). 75.Edward A. Shils, TheTorment ofSecrecy: T heBackground and Consequences ofAmerican Security Policies ,intro.Daniel P .Moynihan (1956;reprint, Chicago, 1996). See also DanielP .Moynihan, Secrecy:T heAmericanExperienc e ,intro.Richard GidPowers (New Haven, 1998). 76.R. H.Coase,‘ ‘The Natureof the Firm,’’ in TheFirm,theMarket, and theL aw (1937; reprint,Chicago, 1988). Alfred Chandler, TheVisibleHand: TheManagerial Revolution inAmericanBusiness (Cambridge,Mass., 1977). 77.Halttunen, ConŽde nce Men , 198–210. 78. Hugo Mu¨nsterberg, Psychologyand IndustrialE Yciency (1913;repri nt, Easton, Pa., 1973). 79.Peter N.Stearns andCarol Z. Stearns, ‘‘Emotionology:Clarifyi ng the History of Emotionsand Emotional Standards,’ ’ AmericanHistorical Review 90(1985): 813– 36. Seealso Christopher Lasch, TheCulture ofNarcissism: AmericanL ifein anAgeofDimin- ishing Expectations (New York,1978). 80.NARA RG326Box 48 6314/ 8/8:USAEC,‘‘Use ofLie Detector atAEC Installa- tions’’ ;‘‘deterrent’’ quote:19. 81.Civi llibertariansbegan to object to the test inthe McCarthy erawhen itwas decried as an eVorttosubvert Fifth Amendment guaranteesagainst self-incrimination; Mac- donald,‘ ‘Lie Detector Era.’’ Morerecently ,laborlawyers have challenged the coercive andintimidating atmosphere created bypolygraph exams inthe workplace;Edgar A.Jones Jr.,‘‘American IndividualRights andan AbusiveT echnology:The Torts ofPolygraphing,’ ’ in Polygraph Test ,159–87. Unti lthe federalEmployee Polygraph ProtectionAct was enacted in1988,thirty states stillallowedthe use ofpolygraph testing ‘‘atwi ll’’ inbusiness settings (forpreemployment screening, investigationsinto allegedwrongdoing, and routine checks).

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 31 82.BPP b.10, f. Larson: Larson toV ollmer,28April 1931. August V ollmer Papers,Ban- croftLibrary ,University ofCalifornia,Berkeley (hereafter AVP), b.18, f.Larson: Lar- son toVollmer,2June 1951. 83.JLP c. 7,f.corr .misc.: Larson toKeeler ,31October 1930. 84. Ibid. 85.JLP c. 1:Larson toDouglas Kelley ,26December 1950. 86. Ibid. 87.BPP b.10, f. Larson: Larson toV ollmer,28April1931. 88.BPP b.10, f. Larson: Larson toV ollmer,20April1927. 89.BPP b.10, f. Larson: Larson toV ollmer,9October 1931. 90.John A.LarsonandG .W.Haney,‘‘Cardio-RespiratoryV ariationsin Personality Stud- ies,’’ TheAmericanJour nal ofPsychiatry 11(1932): 1035– 81. Larson hadalready cooper- atedwith the institute’s director,Herman Adler,see Herman M.Adler andJohn A. Larson,‘ ‘Deceptionand Self-Decept ion,’’ Journal ofAbnormal Psychology 22 (1928): 364–71. 91.V erne W.Lyon,‘ ‘DeceptionT ests with Juvenile Delinquents,’’ Journal ofGenetic Psy- chology 48(1936):494 –97.Previousstudies at the institutehad examined lying inchil- drenwithout specifying how they couldtell whether the children were lying ornot; LutonAckerson, Children’sBehaviorProblems: A Statistical Study Basedupon 5000Child- ren Examined Consecutivelyat theIllinois Institute forJuvenile Research ,2vols.(Chicago, 1931–), 1:57,175– 77; 2:357– 65. 92.John A.Larson,Alan Canty,andClaude Broom, ‘ ‘Lie Detectors Rashly Operated andOversold,’ ’ PoliceJour nal 25(Nov.–Dec. 1939):13, 24–25;26(Jan.–Feb. 1940): 4, 20.R. F.Borksteinand John A.Larson,‘ ‘The ClinicalT eamApproach,’ ’ in Academy Lectures onLieDetection, ed.V .A.Leonard(SpringŽ eld, Ill., 1957),1:11– 20. 93.F orrester, Truth Games. 94.See Marston, LieDetector , 119, 138. 95.DoDLKP Don L. Kooken( IndianaState Police) toKeeler ,24August1939. 96.BPP b.10, f. Larson: Larson toV ollmer,9October 1931.Larson, Canty ,and Bloom,‘ ‘Oversold.’’ 97. Fryev .United States 293F ed.1013 (1923): 46– 47, decided by the Courtof Appealsof the District ofColumbia,opinion by F .Van Orsdel; annotatedin AmericanL awReports 34(1923): 147– 48. 98.F orthe judicialdebate over who are the relevant experts, see PaulG. Giannelli andEdward J .Imwinkelried, ScientiŽc Evidence ,2ded. (Charlottesville, Va.,1993), 1:232–35. 99.Jurists have conductedvarious surveys ofpsychologists. In1926a lawyer surveyed 88members ofthe American Psychological Association,selected fortheir presumed interest inthe Želdof lie detection,and found that ofthe 38who responded,18 thoughtthe methodwas ‘‘ofsu Ycient accuracy astowarrant considerationby judges andjurors’ ’ (thoughmost expressed some qualiŽcations); 13answered no;and 7 an- swers were toomixed to classify .Tests conductedfor McCormick byJ .F.Dasheill (University ofNorth Carolina),reported in C.T.McCormick,‘ ‘DeceptionT ests and the Law ofEvidence,’ ’ California LawReview 15(1926–27): 495– 98. In 1952,when the University ofT ennessee polled1,682 criminologists, polygraphers, and psycholo- gists, itfound that ofthe 719responses, twice asmany polygraphers(75 percent) as psychologists (36percent) believedthat the mainreactions tothe instrument were causedby deception or the attemptto conceal it;see EdwardE. Cureton,‘ ‘AConsen- sus astothe Validityof Polygraph Procedures,’ ’ TennesseeL awReview 22(1953):728 –

32 Representations 42.This survey was proposedby the ChathamCompany ,then conductingits system- atic polygraphexaminations of employees atOak Ridge,and the companyhelped select the respondents.F orthe backgroundto the examination,see PaulV .Trovillo, ‘‘ScientiŽc Proofof Credibility ,’’ TennesseeL awReview 22(1953):760– 61. 100.At the behest ofthe Societyfor Psychophysiolog yResearch (SPR),the GallupOrgani- zationconducted a pollof members in1984and found very few willing tosee poly- graphresults usedin court,though 62 percent believedthe polygraphwas auseful diagnostictool. Gallup Organization, ‘ ‘Survey ofthe Membershipof the Societyfor Psychological [ sic]Research.’’ Polygraph (1984):153– 65. A rephrasedquestionnaire sent outin 1994to members ofthe SPRbypropolygraph advocates found that 61 percent saidthe methodwas ‘‘usefulfor legal proceedings’ ’; see SusanAmato and Charles Honts, ‘‘What DoPsychophysiologists Think AboutPolygraph T ests? ASur- vey ofthe Membershipof S.P .R.,’’ Psychophysiology 31(1994):S22. 101.The majorityopinion in Sche Ver byJustice Clarence Thomas turnedon the narrower questionof whether the president’s executive powerwas used‘ ‘arbitrarily’’ when he imposeda perse exclusion ofthe polygraph.It explicitly puto V any judgmentabout whether inthe absence ofa clear rulingby the legislatureor executive powerthe courts oughtto admitor deny polygraphevidence, takingnote instead of the distinct rulings emergingin the variousdistrict courts underthe new Daubertrule. United Statesv. ScheVer 96US1133(1998). 102.A district attorneyin Appleton,Wisconsin expressed this view in1930.LKP Scrap- book: AppletonPost ,8April1930.The courts expressed this fearmost eloquentlyin United Statesv .Alexander 526F .2d161 (8th Cir .1975),which excluded polygraphevi- dence because itwas ‘‘shroudedwith an auraof near infallibility,akin tothe ancient oracle ofDelphi.’ ’ 103.LKP c. 2,f.original: Keeler ,‘‘The Jury System ShouldBe Abolished,’’ [1930];also c. 1:Keeler,‘‘TalktoSigma [ Xi],’’ MayoClinic, Rochester ,Minn.,18 January 1934. 104.Fred E.Inbau,‘ ‘CaseAgainst the Polygraph,’’ American BarAssociati on Journal 51 (1965):857. 105.In lightof this problem,some have suggestedobliging all operatorsto pass through arigoroustraining program and meet strict licensing requirements.Keeler himself advocatedstate licensing forexaminers, thoughhe alsobelieved that ‘‘inthe mean- timeit rests with the honorand integrity of each member ofaprofessionto keep the professionpurged of incompetency anddishonesty .’’ On the few andfeeble attempts bypolygraph operators to policethemselves, see Charles R.Honts andMary V.Perry, ‘‘PolygraphAdmissibi lity:Changes and Challenges,’ ’ Lawand Human Behavior 16 (1992):369– 73. Currently ,the American PolygraphAssociation accredits schools, and it,along with the American Association ofPolice Polygraphers, sets standardsfor the use ofthe polygraph.But the standardsare voluntary and there isnoenforcement mechanism. Moreover,some twothousand operators are not members ofeither orga- nization;Giannelli andImwinkelried, ScientiŽc Evidence , 218–19.

ASocialHistory of Untruth: Lie Detection andT rust inTwentieth-Century America 33