Oxytocin


Чтобы посмотреть этот PDF файл с форматированием и разметкой, скачайте его и откройте на своем компьютере.
Oxytocinpromotesgroup-servingdishonesty
ShaulShalvi
a,1
andCarstenK.W.DeDreu
b,1
a
PsychologyDepartment,Ben-GurionUniversityoftheNegev,BeerSheva84105,Israel;and
b
DepartmentofPsychology,UniversityofAmsterdam,1018WB,
Toprotectandpromotethewell-beingofothers,humansmay
bendthetruthandbehaveunethically.Herewelinksuchtenden-
ciestooxytocin,aneuropeptideknowntopromoteaffiliationand
cooperationwithothers.Usingasimplecoin-tosspredictiontaskin
whichparticipantscoulddishonestlyreporttheirperformancelevels
tobenefittheirgroup

soutcome,wetestedthepredictionthat
oxytocinincreasesgroup-servingdishonesty.Adouble-blind,pla-
cebo-controlledexperimentallowingindividualstolieprivately
andanonymouslytobenefitthemselvesandfellowgroupmem-
bersshowedthathealthymales(
n
=
60)receivingintranasaloxy-
tocin,ratherthanplacebo,liedmoretobenefittheirgroup,and
reciprocaldishonestyfromfellowgroupmembers.Treatment
effectsemergedwhenlyinghadfinancialconsequencesand
moneycouldbegained;whenlosseswereatstake,individuals
inplaceboandoxytocinconditionsliedtosimilardegrees.Inacon-
trolcondition(
n
=
60)inwhichdishonestyonlybenefitedpartic-
ipantsthemselves,butnotfellowgroupmembers,oxytocindid
spectiveonmoralityrevealingdishonestytobeplasticandrooted
inevolvedneurobiologicalcircuitries,andalignwithworkshow-
ingthatoxytocinshiftsthedecision-maker

sfocusfromselfto
groupinterests.Thesefindingshighlighttheroleofbondingand
cooperationinshapingdishonesty,providinginsightintowhen
andwhycollaborationturnsintocorruption.
decisionmaking
|
behavioraleconomics
|
|
lies
|
honesty
P
rofit-seekingcertainlydrivesdecision-making(1),buthumans
arealsomotivatedbymoralsentiments(2,3),theinterlocking
suppressorregulateselfishness(4).Prominentamongthesevalues
andvirtuesarehonestyandtruthfulness(5).Withfewexceptions,
honestyandtruthfulnessareheldinhighregardacrossmany
culturesandreligiousorientations(6),whereasdeceitanddis-
honestyhavemetwithpunishmentandsocialexclusionthrough-
outhumanevolution(7).Despitethisnear-universalemphasison
honesty,however,humansoftenlie,deceive,omit,andmis-
represent(8

10).Forexample,theestimatedcostsfromtaxeva-
sionandfrominflatedinsuranceclaims,so-called

insurance
build-ups,

amasstobillionsofUSdollarsannually(5).In
tojustifysuchbehaviortothemselvesandothers(11

15).
Thathumansapplyhonestyandtruthfulnessinaflexible
someofthetime,fitsafunctionalistapproachtomorality(4,16).
Suchaperspectivesuggeststhatmoralbehaviorenablespeople
tobeapartofagroupbymakingpersonallycostlycontributions
tothegroupandbycreatingareputationofaloyalandtrust-
worthycooperatorthatshouldbeincludedinsocialexchange,
ratherthanbeingostracizedandexcluded(17,18).Accordingly,
bothgroupsandtheirmembersbenefitfromdevelopingand
nurturingmoralbehavior,includingbeinghonestandtruthful.
Importantly,however,moralbehaviorthatservesone

sgroup
maybeatoddswithmoralconcernsandbehaviorsthatserve
anoverarchingcollectiveorsomeuniversalmoralprinciple
(16,18

19).Forexample,whenamotherfromOhiofakedher
homeaddresstosendherchildrentoagoodschooloutsidetheir
residentialdistrict,shewassentencedtojailforbreakingthelaw.
homeaddress)canbeconsideredashighlymoral(self-sacrificing
tobenefitone

skinandkith)atthegrouplevelofanalysis.Indeed,
inhertestimony,themotherindicatedthatshedeliberatelytook
theriskofpunishmenttoimproveherchildren

schancesinlife
(20).Suchmotivationtoserveothers,itstandstoreason,ismore
powerfulandmakesiteasiertogeneratejustificationformoral
codebreakingthanamotivationtoservepersonalinterestsonly
(evenwhenholdingpersonalprofitconstant)(21

24).
Thefunctionalistapproachtomoralitysuggeststhathumans
breakfundamentalmoralcodes(e.g.,

thoushallnotlie

)when
thisservestheirgroup

sinterestsmorethanwhenthisonlyserves
theirpersonalinterests,andthatsuchgroup-servingdishonesty
restsonevolvedneurobiologicalcircuitriesthatsustainand
motivategroupidentifica
tion,solidarity,andparochialcooperation.

sgroupshouldbe
modulatedbyoxytocin,anevolutionaryancientandstructurally
highlypreservedneuropeptideproducedinthemammalianhypo-
thalamus(25

28).Indeed,andconsistentwithevolutionarytheory
predictingthattrustandcooperat
ionaregearedprimarilyatthose
consideredrelevanttosurvivalandprosperity(29),intranasalad-
ministrationofoxytocin(comparedwithplacebo)promotestrust
andcooperationinhumans(30),especiallywithfamiliarindividuals
andin-groupmembers(27,31

35).Thus,nexttoitswell-known
roleinreproductionandsocialbondformation(36,37),oxytocin
alsofunctionsto

tend-and-defend

thein-groupbymotivating
parochialaltruism:self-sacrificingtobenefitone

sowngroup
Whereasmoralviolationssuchasbeingdishonestarepoten-
tiallycostly,asonemaybecaughtandpunishedandone

s
positiveself-viewmaybeundermined(5,39),afunctionalist
perspectiveonmoralityimpliesthathumansarepreparedto
accepttheseriskswhendishonestyservestheirgroup.Here,we
Significance
Verylittleisknownaboutthebiologicalfoundationsofim-
moralbehavior.Wereportheretheresultsofadouble-blind,
placebo-controlledexperimentshowingthatthehormone
oxytocinpromotesgroup-se
rvingdishonesty.Compared
withparticipantsreceivingpla
cebo,participantsreceiving
oxytocinliedmoretobenefittheirgroups,didsoquicker,and
didsowithoutexpectationofreciprocaldishonestyfromtheir
self-servingdishonesty.Thesefindingssupportthefunctional
approachtomoralityandrevealtheunderlyingbiologicalcir-
cuitriesassociatedwithgroup-servingdishonesty.
Authorcontributions:S.S.andC.K.W.D.D.designedresearch;S.S.performedresearch;S.S.
analyzeddata;andS.S.andC.K.W.D.D.wrotethepaper.
Theauthorsdeclarenoconflictofinterest.
ThisarticleisaPNASDirectSubmission.
1
Towhomcorrespondencemaybeaddressed.E-mail:[email protected]@
uva.nl.
Thisarticlecontainssupportinginformationonlineat
www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1400724111/-/DCSupplemental
.
www.pnas.org/cgi/doi/10.1073/pnas.1400724111
PNASEarlyEdition
|
1of5
PSYCHOLOGICALAND
COGNITIVESCIENCES
conjecturethatsuchgroup-servingdishonestymaybemotivated
byoxytocin,andwepredictthathumansviolatemoralcodes
(here:lying)morewhenthisservestheirgroupthanwhenthis
servestheirself-interestonly;inaddition,intranasaloxytocin
comparedwithplaceboshouldamplifylyingwhenitservesgroup
interestsmorethanwhenitservesself-interestonly.Wealso
individualslyingfortheirgroupexpectreciprocaldishonesty
fromtheirgroupmembers.
Predictionsweretestedinadouble-blind,randomizedplacebo-
controlledexperiment.Sixtyh
ealthymalesself-administered
asingleintranasaldoseof24IUoxytocin(
n
=
30)orplacebo
(
n
=
30)andthen,after30min,whichallowedoxytocineffectsto
peak(41,42),receivedcomputerizedinstructionsexplainingthey
wererandomlyassignedtoathree-persongroup,witheach
groupmemberperformingthesametaskandallearningsbeing
equallysharedamonggroupmembers(
).
Participantssawa
V
1coinonthecomputerscreenandwere
askedtopredicttheoutcomeofarandomtoss(

heads

vs.

tails

).Participantspredictedtheoutcome,wereinstructedto
theirpredictionwascorrect.Thisallowedparticipantstoreport
makingcorrectpredictionsevenwhenmakingincorrectones.
Wemanipulated(withinsubjects)thegroupprofitgeneratedby
correctpredictionstobegain(
+
V
0.30),loss(

V
0.30),ornone
(
V
0).Incorrectpredictionsalwaysaddednothing(
V
0).After
fivepracticetosses,participantsengagedin30randomlypre-
sentedpaymenttosses(10ofeachoutcomeinaninfinite-repe-
predictionwascorrect.Attheendoftheexperiment,partic-
ipantspredictedtheperformanceofanothergroupmember,
withaccuratepredictionsincentivizedby
V
0.10.
Results
Participants

performancewastrulyprivate.Dishonestywas
assessedontheaggregatelevelbycomparingreportedperformance
totheperformancepredictedbyafairtoss(50%)(43,44).Overall,
participantsoverreportedcorrect
predictionsongaintrials(73.2%;
binominal
Z
,11.31;
P

0.000001),underreportedonlosstrials
(29.3%;binominal
Z
,

10.08;
P

0.000001),andmodestlyover-
reportedinno-benefittrials(55.3%;binominal
Z
,2.57;
P
=
0.01).
Fig.1showsthatcheatingonthegaintrialswasmodulatedby
treatment:participantsundertheinfluenceofoxytocinreported
morecorrectpredictions(79.7%)thanparticipantsreceivingthe
placebo(66.7%;Mann

Whitney
Z
,

1.99;
P
=
0.046).No
treatmenteffectswereidentifiedinthelosstrials(27.0%
oxytocin
vs.31.7%
placebo
;Mann

Whitney
Z
,

0.72;
P
=
0.47)orintrials
withnobenefit(52.7%
oxytocin
vs.58.0%
placebo
;Mann

Whitney
Z
,

1.15;
P
=
0.25).
A2(oxytocinvs.placebo)

3(lossvs.nonevs.gaintrial
block)ANOVAwithtrialblockwithinsubjectsrevealedfaster
ticipantsreceivingoxytocin(mean,2.22
seconds
;standardde-
viation,0.50)thanplacebo[mean,2.86
seconds
;standarddeviation,
0.78;
F
(1,58)
=
13.99;
P

0.00001].Themaineffectforblock
wasalsosignificant,withparticipantstakingmoretimeintheloss
trials(mean,2.74;standarddeviation,1.09)thaninthegaintrials
(mean,2.42;standarddeviatio
n,0.89)andno-benefittrials
[mean,2.46;standarddeviation,0.68;
F
(1,58)
=
3.62;
P
=
0.033].
(
F

1;
P

0.65).ANOVAfurtherrevealedthatparticipants
expectedtheirgroupmemberstooverreportcorrectpredictions
ingaintrials(64.5%correctpredictions),underreportcorrect
predictionsinlosstrials(33.3%),andslightlyoverreportinno-
benefittrials[57.5%;
F
(2,58)
=
24.25;
P

0.00001].Oxytocin
neitheraffectednormoderatedexpectations(
F

0.41),sug-
gestingoxytocindidnotboostreciprocal-dishonestyexpectations.
Althoughparticipantsreceivingoxytocinliedmoretobenefit
theirgroupthanparticipantsreceivingplacebo,theyalsoboos-
tedtheirownshareofthejointoutcome.Thisraisesthepossi-
bilitythatoxytocinincreasesdishonestyaimedatboostingown,
ratherthangroup,outcomes.Toaddressthis,werecruitedan
taskwiththeexceptionthatthepayoffstructurewasmodifiedto
influenceonlytheparticipants(
+
V
0.10ongain,

V
0.10onloss,
and
V
0onnobenefittrials)andnottheirgroupmembers.Im-
acorrectpredictiononagaintrialwasworth
V
0.10/1person
=
V
0.10perperson)wereidenticaltothoseinthepreviouslyused
wasworth
V
0.30/3persons
=
V
0.10perperson).Resultsre-
vealedsimilarlevelsofdishonestyacrosstreatments:Participants
over-reportedtheiroutcomesongaintrials(68.7%;binominal
Z
,9.10;
P

0.00001),underreportedoutcomesonlosstrials
(41.0%;binominal
Z
,

4.37;
P
=
0.00001),anddidnotlie
withoutfinancialincentive(51.2%;binominal
Z
,0.53;
P
=
0.59).
Lyingwasnotinfluencedbyoxytocin[gaintrials:
Z
,

1.20(
P
=
0.23);losstrials:
Z
,

0.14(
P
=
0.89);no-benefittrials:
Z
,

0.92
(
P
=
0.36)].Finally,ANOVAwithtrialoutcome(lossvs.nonevs.
gain)withinsubjectsandtreatment(oxytocinvs.placebo)be-
tweensubjectsrevealednodifferencesinresponsetimebe-
tweenoxytocin(mean,2.51
seconds
;standarddeviation,0.75)and
placebo[mean,2.65
seconds
;standarddeviation,0.86;
F
(1,58)
=
0.46;
P
=
0.50].Similartothegroup-servingsetting,themain
effectforblockwassignificant,withparticipantstakingmore
timeinthelosstrials(mean,2.
80;standarddeviation,0.91)
thaninthegain(mean,2.44;standarddeviation,0.87)andno-
benefit[mean,2.51;sta
ndarddeviation,0.90;
F
(1,58)
=
8.77;
P

wasnotsignificant(
F
=
1.78;
P
=
0.18).Thus,incontrasttothe
creasedneitherdishonestynorthetimeneededtomakedecisions.
group-serving,butnotself-serving,dishonesty.Althoughthisfits
theconclusionthatoxytocindr
ivesgroup-servingdishonesty
appearedtobelimitedtogaintrialsandrestsontwoindependent
tests(i.e.,oneinthegroup-servingandoneintheself-serving
task).Thus,tofurtherexploretherobustnessandformofthe
obtainedresults,weperformedalog-linearanalysisthatdirectly
testedfortheinteractionoftreatment(oxytocinvs.placebo)and
Fig.1.
Mean(

1SE)reportedcorrectpredictionsasafunctionoftreatment
(oxytocinvs.placebo)andoutcome(lossvs.nonevs.gain),inthegroup-
chance(i.e.,ifparticipantsreporthonestly;5/10).
2of5
|
www.pnas.org/cgi/doi/10.1073/pnas.1400724111
ShalviandDeDreu
others

outcomes(49),theyfinditeasiertojustifytheirownlies
whentheyarenottheonlyonesbenefitingfromthem(23),and
theyarewillingtoengageinmoreextremelyingwhendoingso
servestheirgroup(22).Fittingthis,weobservedthatoxytocin-
drivengroup-servinglieswererelativelyfast,suggestinggroup-
servinglyingrequireslittledeliberation(40,50)andthatoxyto-
cinenablesthismoreautomaticresponding.
cuitriesoxytocinmodulateswhenengagingingroup-serving
dishonesty.Suchnewresearchshouldalsoexaminewhydis-
honestyaimedatpreventing(group-orself-serving)losses
appearedinsensitivetooxytocinadministration.Ingeneral,people
aremoreloss-aversethangain-seeking,aspreventinglossismore
fundamentaltosurvivalthanachievinggain(51).Perhapsthe
humanbrainismoresensitivetoneurohormonalmodulation
whengains,ratherthanlosses,areatstake,andthegoalis
(group)prosperity,ratherthansurvival.Suchaconclusionwould
fitworksuggestingthatoxytocinmotivates(social)approach
tendenciesmorethanitmodulateswithdrawal(47,52,53).
Thatoxytocinboostedgroup-servingdishonestycannotbe
attributedtopeople

sexpectationforreciprocaldishonestyfrom
theirfellowgroupmembers,asparticipantsreceivingoxytocin
didnotexpectgroupmemberstolieforthegroupmoreoften
tocinledtofasterandputativelyless-deliberateddecisionmak-
circuitryisinvolvedinbiologicallypreparinghumanstolearn
servesthoseonecaresaboutandthegrouponebelongsto.
Maleparticipantswererecruitedviaan
onlinerecruitingsystemandoffered
V
10
(approximately13USD)forparticipatinginastudyontheeffectsofmedication
ontestscoresanddecision-making.Theyfilledoutanonlinemedicalscreening
form,andparticipantswereincludedinthestudyonlyiftheyconfirmedthey
werenotsufferingfromsignificantmedicalorpsychiatricillness,usingmedica-
tion,and/orsmokingmorethanfivecig
arettesperday.Participantswere
instructedtorefrainfromsmokingordrinking(exceptwater)for2hbeforethe
experiment.Theexperimentsreceive
dEthicsApprovalfromtheUniversityof
involvinghumansubjects.Inkeepingwithdepartmentalethicsguidelinesand
contemporarypracticeinexperimenta
lresearchwithhumansubjects,partic-
ipantsreadandsignedaninformedc
onsentbeforetheexperiment,the
experimentsdidnotinvolvedeception,andparticipant

spaywasbasedonthe
sumoftheirshow-upfeeandtheirearn
ingsduringthedecisiontasks.Partic-
ipantsreceivedfulldebriefing
Onarrivalinthelaboratory,participantswereseatedinindividualcubicles,
preventingthemfromseeingeachotherandcommunicating.Closelyfol-
lowingtheproceduresinearlierwork(32),participantswererandomly
assignedtotheoxytocinorplacebogroup(double-blind,placebo-controlled
studydesign).Participantsself-administeredeitherasingleintranasaldose
of24IUoxytocin(Syntocinon-Spray,Novartis),3puffspernostril,eachwith
4IUoxytocin,orplacebo30minbeforeengagingintheexperimentaltasks.
Theplacebocontainedalloftheactiveingredientsastheoxytocin,withthe
exceptionoftheneuropeptide,andwasmanufacturedbyStichtingApoth-
ekersHaarlemseZiekenhuizenincoordinationwiththepharmacyatthe
AmsterdamMedicalCentre,adheringto
theguidelinesongoodmanufacturing
practiceandgoodclinicalpractice.
Afterself-administrationofthemedication(placebooroxytocin),the
tionnairesandteststhatwerepresentedontheircomputerscreen,usingthe
keyboardtoanswerquestions.Thesetaskswereself-paced.Conformingto
priorresearchshowingthateffectsofoxytocinpeekafter

30min(41,42),
after30min,thecomputerautomaticallyswitchedtotheinstructionsforthe
experimentaltask.Themainexperimentaltaskbeganbyinformingpartic-
ipantsthattheywouldengageinataskinvolvingathree-persongroup.
Participantswereinformedthatgroupswerecomposedonthebasisofthe
orderinwhichtheyhadsignedupfortheexperiments,anditwasnoted
thatmost,butnotnecessarilyall,groupmemberswerecurrentlypresentin
thelaboratory.Accordingly,participantslearnedthattheyandtwoother
participantswereassignedtobeinonegroup.Theywerealsotoldthatthey
wouldnotknowwhowasintheirgroupatanytime.
Hereafter,participantswereintroducedtotheexperimentaltasks(see
following),madedecisions,andansweredaquestionnaire.Theexperimental
ipantswerethankedanddismissed.Within8weeksaftertheexperiment,all
participantsreceivedasealedenvelopecontainingtheirearningsandwent
anddebriefingtopreventanyinsightonthemethodsandmaterialsusedfrom
becomingknownbeforetheentireexperimentwascompleted.Allparticipants
agreedtothisprocedure.
Sixtyhealthymaleundergraduatestudentsatalargeuniversityin

agedid
in(mean,21.60;standarddeviation,2.47)andpla-
cebo(mean,21.47;
standarddeviation,2.80)groups[
t
(58)
=
0.20;
P
=
0.85]groups.
Afterthegeneralintroductionsandassi
gnmenttoagroup,computerinstructions
statedthatforthegrouptask,eachindi
vidualgroupmemberwouldpredictthe
outcomeofmultiplecoin-tosses,earningmoneyforthegroupaccordingtohis
performance.Participant
sreadthatallthreegroupmemberswouldperformthe
taskandthatthefinalearningsofeachgroupmemberwouldbeathirdofthe
sumearningsgeneratedbythethree-pe
rsongroup.Participantslearnedthat

heads

or

tails,

keepthepredictionintheirmind,andpressthe

tosscoin

button.Once
theysawtheoutcome,theyhadtoclickw
not(refs.43

44,similartoref.54).Tofamiliarizethemselveswiththetask,each
participantengagedin5practiceco
Afterbeingfullybriefedaboutthetask,participantswereremindedthat
theirpredictions,andthoseoftheothers,hadfinancialimplications.Par-
ticipantsknewtheywouldengageinaseriesofcoin-tossesreceiving(ano-
nymized)feedbackabouttheperformanceoftheothergroupmembersonly
attheveryendoftheexperiment.Aftertheirpredictions,participants
memberstoperformonthecoin-tossingtask.Specifically,theywereaskedto
guesshowmanycorrectpredictions(ofamaximumof10)oneoftheir
randomlychosengroupmembershadoneachofthethreeblocks(with
correctguessesleadingtoalosstothegroup

soutcome,nofinancialcon-
sequence,oragaintothegroup

soutcome).
Anothergroupofsixtyhealthymalesparticipatedinthe

theoxytocin(mean,22.07;standarddeviation,2.82)andplacebo(mean,
21.80;standarddeviation,3.25)groups[
t
(58)
=
0.34;
P
=
0.74].Theproce-
exceptthatparticipants

earningfrompredictingeachcointosscorrectly
influencedonlytheirown,notthegroup

s,outcome.
ACKNOWLEDGMENTS.
WethanktheparticipantsofthejointCenterfor
DecisionMakingandEconomicPsychology(Ben-GurionUniversity)and
commentsonthiswork.PartoftheresearchwasfundedbythePeople
Program(MarieCurieActions)oftheEuropeanUnion

sSeventhFramework
Program(FP7/2007-2013)underResearchExecutiveAgencyGrantAgreement
dation(toC.K.W.D.D.).
Fig.3.
Mean(

1SE)responsetimetaken(inseconds)whenindicating
4of5
|
www.pnas.org/cgi/doi/10.1073/pnas.1400724111
ShalviandDeDreu
1.CamererCF,FehrE(2006)Whendoes

economicman

dominatesocialbehavior?
Science
311(5757):47

52.
2.SmithA(1759)
TheTheoryofMoralSentiments
(LibertyFund,Indianapolis)(reprinted
1982).
3.BowlesS(2008)Policiesdesignedforself-interestedcitizensmayundermine

themoral
sentiments

:Evidencefromeconomicexperiments.
Science
320(5883):1605

1609.
4.HaidtJ(2007)Thenewsynthesisinmoralpsychology.
Science
316(5827):998

1002.
5.MazarN,AmirO,ArielyD(2008)Thedishonestyofhonestpeople:Atheoryofself-
conceptmaintenance.
JMarkRes
45:633

644.
6.SchwartzS,BilskyW(1990)Towardsauniversalpsychologicalstructureofhuman
values.
JPersSocPsychol
58:878

891.
7.TriversR(2011)
Thefollyoffools:Thelogicofdeceitandself-deceptioninhumanlife
(NewYork:BasicBooks).
HumCommunRes
37:377

403.
9.DePauloBM,KashyDA,KirkendolSE,WyerMM,EpsteinJA(1996)Lyingineveryday
life.
JPersSocPsychol
70(5):979

995.
10.BazermanMH,TenbrunselAE(2011)
BlindSpots:WhyWeFailtoDoWhat

sRightand
WhattoDoaboutIt
11.FischbacherU,Follmi-HeusiF(2013)Liesindisguise,anexperimentalstudyon
cheating.
JEurEconAssoc
11:525

547.
OrganBehav
HumDec
115(2):181

190.
AvoidbothMajorandMinorLies.
BrJManage
22:s16

s27.
14.LundquistT,EllingsonT,GribbeE,JohannessonM(2009)TheAversiontoLying.
JEconBehavOrgan
70:81

92.
vatedcommunicationofuncertaininformation.
JRiskUncertain
25:185

201.
16.GreeneJ(2013)
NY
(NewYork:ThePenguinPress).
17.BowlesS,GintisH(2013)
ACooperativeSpecies:HumanReciprocityanditsEvolution
heuristicus,andHomoduplex.
OrganBehavHumDec
123(2):150

158.
19.VanLangePAM,JoiremanJA(2008)HowWeCanPromoteBehaviorThatServesAll
ofUsintheFuture.
SocialIssuesandPolicyReview
2:127

157.
20.ABCnewshttp://abcnews.go.com/US/ohio-mom-
jailed-sending-kids-school-district/story?
id
=
12763654.
benefitsothers.
JEconBehavOrgan
,10.1016/j.jebo.2013.04.005.
22.ConradsJ(2013)Irlenbusch,B.,Rilke,R.M.,Walkowitz,G.LyingandTeamIncentives.
JEconPsychol
34:1

7.
23.WiltermuthSS(2011)Cheatingmorewhenthespoilsaresplit.
OrganBehavHumDec
115(2):157

168.
24.CohenTR,GuniaB,Kim-JunSY,MurnighanJK(2009)Dogroupsliemorethanin-
dividuals?Honestyanddeceptionasafunctionofstrategicself-interest.
JExpSoc
Psychol
45:1321

1324.
ciality.
Science
322(5903):900

904.
26.LudwigM,LengG(2006)Dendriticpeptidereleaseandpeptide-dependentbehav-
iours.
NatRevNeurosci
7(2):126

136.
27.DeDreuCKW(2012)Oxytocinmodulatescooperationwithinandcompetition
reviewandresearchagenda.
HormBehav
61(3):
419

428.
28.Meyer-LindenbergA,DomesG,KirschP,HeinrichsM(2011)Oxytocinandvasopressin
inthehumanbrain:Socialneuropeptidesfortranslationalmedicine.
NatRevNeurosci
12(9):524

538.
29.NowakMA,TarnitaCE,WilsonEO(2010)Theevolutionofeusociality.
Nature
466(7310):1057

1062.
30.KosfeldM,HeinrichsM,ZakPJ,FischbacherU,FehrE(2005)Oxytocinincreasestrustin
humans.
Nature
435(7042):673

676.
31.DeclerckCH,BooneC,KiyonariT(2010)Oxytocinandcooperationunderconditions
ofuncertainty:Themodulatingroleofincentivesandsocialinformation.
HormBehav
57(3):368

374.
32.DeDreuCKW,etal.(2010)Theneuropeptideoxytocinregulatesparochialaltruismin
intergroupconflictamonghumans.
Science
328(5984):1408

1411.
33.DeDreuCKW,GreerLL,VanKleefGA,ShalviS,HandgraafMJJ(2011)Oxytocin
ProcNatlAcadSciUSA
108(4):1262

1266.
PsycholSci
21(8):1072

1074.
35.IsraelS,WeiselO,EbsteinRP,BornsteinG(2012)Oxytocin,butnotvasopressin,increases
bothparochialanduniversalaltruism.
Psychoneuroendocrinology
37(8):1341

1344.
36.CarterCS,GrippoAJ,Pournajafi-NazarlooH,RuscioMG,PorgesSW(2008)Oxytocin,
vasopressinandsociality.
ProgBrainRes
170:331

336.
37.HeinrichsM,vonDawansB,DomesG(2009)Oxytocin,vasopressin,andhumansocial
behavior.
FrontNeuroendocrinol
30(4):548

557.
stimuliinmales.
ProcNatlAcadSciUSA
109(44):18144

18149.
39.BeckerGS(1968)CrimeandPunishment:AnEconomicApproach.
JPolitEcon
76:169.
40.ShalviS,EldarO,Bereby-MeyerY(2012)Honestyrequirestime(andlackofjustifi-
cations).
PsycholSci
23(10):1264

1270.
inhumans.
JNeurosci
25(49):11489

11493.
42.BaumgartnerT,HeinrichsM,VonlanthenA,FischbacherU,FehrE(2008)Oxytocin
shapestheneuralcircuitryoftrustandtrustadaptationinhumans.
Neuron
58(4):
639

650.
43.ShalviS(2012)Dishonestyincreasingthelikelihoodofwinning.
JudgeDecMaking
7:292

303.
44.GreeneJD,PaxtonJM(2009)Patternsofneuralactivityassociatedwithhonestand
dishonestmoraldecisions.
ProcNatlAcadSciUSA
106(30):12506

12511.
45.DeDreuCKW,ShalviS,GreerLL,VanKleefGA,HandgraafMJJ(2012)Oxytocin
motivatesnon-cooperationinintergroupconflicttoprotectvulnerablein-group
members.
PLoSONE
7(11):e46751.
defense:Breastfeedingincreasesaggressionbyreducingstress.
PsycholSci
22(10):
1288

1295.
47.ScheeleD,etal.(2013)Oxytocinenhancesbrainrewardsystemresponsesinmen
viewingthefaceoftheirfemalepartner.
ProcNatlAcadSciUSA
110(50):20308

20313.
48.RillingJK,SanfeyAG(2011)Theneuroscienceofsocialdecision-making.
AnnuRev
Psychol
62:23

48.
49.GneezyU(2005)Deception:Theroleofconsequences.
AmEconRev
95:384

394.
50.RandDG,GreeneJD,NowakMA(2012)Spontaneousgivingandcalculatedgreed.
Nature
489(7416):427

430.
51.KahnemanD,TverskyA(1979)Prospecttheory:Ananalysisofdecisionunderrisk.
47:263

291.
52.KempAH,GuastellaAJ(2011)Theroleofoxytocininhumanaffect

anovelhy-
pothesis.
CurrDirPsycholSci
20:222

231.
fromownvs.rivalingothergroups.
BrainRes
,inpress10.1016/j.brainres.2013.09.013.
54.SchurrA,RitovI,KareevY,AvrahamiJ(2012)Isthattheansweryouhadinmind?The
JudgeDecMaking
7:679

688.
ShalviandDeDreu
PNASEarlyEdition
|
5of5
PSYCHOLOGICALAND
COGNITIVESCIENCES
Becausepeopleusemoreextremelieswhendishonestyserves
ipantswho,ongaintrials,underreportedbeingcorrect(i.e.,0

4
predictions),reportedataroundchancelevel(5

6predictions),
moderatelyoverreportedbeingcorrect(7

8predictions),orex-
tremelyoverreportedbeingcorrect(9

10predictions).Notethat
correctlypredicting9or10correctpredictionss
houldoccuronly
in1%ofthecasesifparticipantsarehonest.
AscanbeseeninFig.2,however,theproportionofextreme
overreportingcorrectpredict
ionsoccurredmuchmorethanthe
predicted1%ifparticipantswer
ehonest.Moreover,extremeover-
reportingofcorrectpredictionsvariedasafunctionoftreatment

2
(4)
=
9.08;
P
=
0.059(marginal)].Fig.2showsthatparticipantsin
redictingcorrectly9or10times
in53%ofthecases(16/30)whengivenoxytocin,whichwasmore
thantwiceashighasreportedbyp
articipantsgiv
enplacebo[23%
(7/30);

2
(1)
=
5.71;
P
=
differencewasfarlesspronouncedandwasnotsignificant.Here,
participantsreceivingoxytocinex
tremelyoverreportedinonly33%
ofthecases(10/30)vs.20%intheplacebogroup[6/30;

2
(1)
=
1.36;
P
=
0.24].Theseadditionalanalysesconfirmthatoxytocin
drivesdishonestywhenitservesone

sgroup,butnotwhenitonly
servespersonalself-interest.Moreover,andasseeninotherwork
(22),theseadditionalanalysesshowthatthisoxytocin-motivated
group-servingdishonestymanifests
itselfespeciallyinextremelies.
Withregardtotimeneededtomakedecisions,atreatment
(oxytocinvs.placebo)

F
(1,116)
=
0.09;
P
=
0.76]butdidrevealsignificanteffectsfortreatment
[
F
(1,116)
=
8.32;
P
=
0.005]andforthetreatment

teraction[
F
(1,116)
=
3.34;
P
=
0.07(marginal)].Fig.3showsthat
[
F
(1,116)
=
11.19;
P
=
ting[
F
(1,116)
=
0.56;
P
=
0.45].
Discussion
Morality,anditsuniversality,hasbeensubjecttocenturiesof
philosophicaldebate(2,4,16),withonepossiblesolutionbeing
afunctionalperspectivesuggestingthatrulessuchas

thoushallnot
lie

maybeuniversalandacceptedacrossgroupsandcultures,and
thathumansactuallyconditiontheapplicationofsuchrulesonthe
ultimateconsequencesforthegroupstheybelongto(4,16).Indeed,
togainprofit,humansaretemptedtoactdishonestlyandtoviolate
moralcodes.Asshownhere,suchunethicalbehaviorispartic-
ularlylikelywhenitservesgroupinterestsandhumanswere
givenoxytocinratherthanplacebo.Specifically,whendishonesty
servesgroupinterests,oxytocinincreasedlyingaswellasextreme
lying.Comparedwithplacebo,oxytocinalsoincreasedthespeed
ofdishonestdecisionmaking.Theseeffectswereparticularly
prominentwhenlyinggeneratedprofitsandwereabsentwhen
lyingservedtoavoidlossorhadnofinancialconsequences.
Whenlyingservedpersonalself-interestsonly,oxytocinhadno
effects.Apparently,oxytocinboos
tsgroup-servingbehavior,rather
thanadherencetogeneralmoralcodes,aconclusionthatfitswork
showingthatoxytocinsustainsandenablessocialbondingaswellas
trustandcooperation,especiallytowardthosebelongingtoone

s
owngroup(33).Thus,ratherthanbeinganeurohormonalmod-
ulatorofmoraltendenciesanduniversalcooperation,oxytocin
parochialcooperationandself-sacrifice(32,35,45),through
lashingoutagainstthosewhothreatengroupmembers(46),or
asshownhere,throughdishonestyandmoralcodebreaking.
Thefindingthatoxytocinmotivatedgroup-servingdishonesty
resonateswithrecentfindings(47)demonstratingthatoxytocin
increasesresponsesinbrainrewardregionswhenpicturesof
theirromanticpartnerswerepresentedtoparticipants.Possibly,
therewardcircuitryalsomaybeinvolvedingroup-servingbe-
haviorandrecruitedwhenmakingdecisionsthataffectnotonly
self-interestsbutalsothoseofothergroupmembers(48).
Indeed,peoplecareabouttheconsequenceoftheirliesfor
Fig.2.
e
ShalviandDeDreu
PNASEarlyEdition
|
3of5
PSYCHOLOGICALAND
COGNITIVESCIENCES

Приложенные файлы

  • pdf 1101360
    Размер файла: 706 kB Загрузок: 0

Добавить комментарий