The Origins of Altruism in Offspring Care

37 Pages·2013·1.56 MB·English

PsychologicalBulletin ©2013AmericanPsychologicalAssociation
2013,Vol.139,No.6,1305–1341 0033-2909/13/$12.00 DOI: 10.1037/a0031755
The Origins of Altruism in Offspring Care
Stephanie D. Preston
UniversityofMichigan
Thecurrentreviewaimstounifyexistingviewsofaltruismthroughanexaminationofthebiologicalbasesof
afundamentalformofgiving:altruisticresponding.Altruisticrespondingismostsalientduringheroicacts
ofhelpingbutisalsoobservedanytimeoneperceivesanother’sdistressorneed,whichinturnmotivatesone
tohelpatacurrentcosttotheself.Suchaidissimple,observableacrossspecies,androotedintheinstincts
andcircuitsthatevolvedtomaximizeinclusivefitnessthroughthecareofhelplessoffspring.Bydesign,the
systemalreadybiasesaidtobothkinandnonkinunderconditionsthatarelargelyadaptive.Theseinherent
y.
s.adl benefitsarealsobuttressedinprimatesandhumansbyknown,later-arrivingbenefitstohelpingingroup-
herbro livinganimals.Evidencefortheproposedhomologybetweenaltruisticrespondingandoffspringretrievalis
publisnated pcoremsepnetteednctheraonudghex1p0ekrteisyes,h(ca)readnfaadcatoprtsiv.Ceoonpcpeopnteunacllyy,bbeotwtheerenqauviroeid(aa)ncpearatincdipaaptipornoabcyhn,aonndmaotfhaecrisli,ta(bti)nmgorotoler
dmi of(d)neonatalvulnerability,(e)salientdistress,and(f)rewardingclosecontact.Physiologically,theyalso
salliedisse schinagreulnateeurcoohrtoerxm,oannadl(sj)upthpeorotrbfriotomfro(ngt)alocxoyrttoecxi.n,T(hhe)ftrhaemdewomorakini-ngteenrmeriaxlesmuelstoimlimatbeoacnodrtpicraolxismysatteemle,v(ei)lsthoef
eofittobe amnaalliyasnisneaunrdalusnyifsiteesmesxtihstaitnegvvoilevwedstboysoaslvsuemkeinygptrhoabtleemvesnincoamdapplteixvehwumayasn,wbeihthavfaior-rrseraecfhleincgtacnocniseenqtumenacmes-
onnot forevenourmostveneratedhumantraits.
ors
i
ationand Keywords:altruism,empathy,prosocialbehavior,caregiving,decisionmaking
cier
os
su
s
Aal
ologicalindividu theOirrgoawnnisgmesneasr,eyethtomuganhyt tmoambemhaalsrddwisipreladytaoctsseolffiskhilnydnpersosmaontde Waltirlusiosnm, i1s9a9l8so; kWniolwsonnto&beWsiulspopnre,s2se0d07b;yWthielsporne,se2n0c1e2o).fcHaupmabalne
he care in situations we are only beginning to understand. Broadly, others and promoted by motivational states such as feeling sorry
ch
yt
Psof suchbehaviorsareconsidered“altruistic”whenanorganismissues for the plight of another (i.e., sympathy) or resonating with the
ne a behavior that appears to benefit another at a current cost to the emotionsofanother(i.e.,empathy),bothofwhichareassumedto
as
cu
erial self.Toresolvetheparadoxofsuchseeminglyselflessbehaviors, haveevolvedfromthemother–offspringbond(e.g.,seereviewsin
Amson researchers have identified ways that altruism can benefit the Batson, 2011; S. L. Brown, Brown, & Penner, 2011a; de Waal,
eer fitness of a giver’s genes (reviewed in Batson, 2011; de Waal, 2008;Dovidio,Piliavin,Schroeder,&Penner,2006;Eisenberg&
hp
ythe 2008;Fehr&Rockenbach,2004;McAndrew,2002;Preston&de Strayer, 1987; Hoffman, 2000; Hrdy, 2009; Preston & de Waal,
bt
yrightedolelyfor ptWhrroaocauilt,gyh20(iA1n1cx;leulJsr.oivMde,.1fWi9tn8ae4ns;gsT,(SrDievuiedgrlaset,rk,1inH9,7a21ll0),,0&a7n;PdHrelaasmtteoirnl,t,oi2nn0d,1i1r29e)c6,t4s)au,dcrvheacanis-- 2fi0e0Dld2esbs,)ps.ipteecsiuecs,haenxdtelnevsievlesowfoarnkaolynsiasl—truaisumniafineddeumndpeartshtya—ndaincgroossf
ps
coed tages in competition for resources at the individual, family, and humanaltruismremainsoutofreach.Thedifferentdisciplinesthat
isnd grouplevel(e.g.,seeAlexander,1987;Fehr&Rockenbach,2004; studyaltruismdonotinteract,andwestilldonothaveacompre-
entnte Gintis,Bowles,Boyd,&Fehr,2003;McAndrew,2002;Sober& hensiveideaofhowaltruismisinstantiatedinthebrainandbody.
mi
docucleis Oofneawcoanysteorvaedddrmesasmthmeasleiapnrosbylsetmemsitshtaotfuoncduesrloinesththeenesuimrapllbeabsees-
hisarti havioral assistance of another in need, hereafter referred to as
Ts
Thi ThisarticlewaspublishedOnlineFirstMarch4,2013. “altruistic responding.” Altruistic responding is most salient in
The specific framework provided herein and all research and writing cases of heroic responding but can be observed any time one
reflect independent work by Stephanie D. Preston. This research was perceivesanother’sdistressorneed,whichinturnmotivatesoneto
supportedbygrantsfromtheUniversityofMichiganandtheTempleton actinthemoment,atacurrentcosttooneself.Thisformofaidis
FoundationtoStephanieD.Preston.ThankyoutoStephanieL.Brownwho simple, is observable across mammalian species, and appears to
firstnoticedtheinterestingworkofMichaelNuman,suggestedapossible haveearlyrootsintheinstinctsandneuralsystemthatevolvedto
roleforthesubgenualanteriorcingulatecortex,andcommentedonanearly care for helpless newborn offspring—a system that is already
version of the article. Thank you to Jill Becker, Kent Berridge, Randy
adaptively designed to promote care when it most benefits recip-
Nesse,andBarbaraFindlayforfeedbackonthearticle.JosephLonstein
ientswhileminimizingthecoststogivers.
andR.BrentStansfieldprovidedfeedbackonthemodelandfigures.Amy
Thepotentialhomologybetweenretrievingahelplessnewbornand
Ross and Courtney K. Hsing helped with preparation of the article and
altruisticretrievalcanbedemonstratedacrossspeciesthroughsome
referencesection.
CorrespondenceconcerningthisarticleshouldbeaddressedtoStephanie salientexamplesdrawnfromboththeacademicandpopularpress:
D. Preston, Department of Psychology, University of Michigan, 530 A rat pup is separated from his mother and siblings nursing a
ChurchStreet,AnnArbor,MI48109.E-mail:[email protected] meter away. The distressed infant makes ultrasonic vocalizations
1305
1306 PRESTON
thatthemotherhears,causinghertoleavethenesttoretrieveand
return him to the nest to nurse with the others (Hofer, 2000; see
Figure1,leftpanel).
A rat dam is placed in an operant conditioning box in which
barpresses allow her to access and retrieve a series of unrelated,
isolated,neonatepups.Thehighlymotivateddampressesthebar
forhours,retrievinghundredsofpupsbeforethesessionisfinally
terminatedbytheexhaustedexperimenters(Wilsoncroft,1969;see
Figure2).
A 3-year-old boy falls 18 feet into a gorilla enclosure at the
Brookfield Zoo. Binti Jua, a female gorilla caring for her baby,
picksuptheunconsciousboyandrockshimasshecarrieshimto
safety, protecting him from an elder female (BBC h2g2 Contrib-
y. utors,2005;seeFigure1,rightpanel).
shers.broadl andAtmugagniningTartinthidealdegisosfavheisdpfraonmtsauhnotiulsheefiarweabkyehniss.dAofgt,ebrasrakfienlgy Fwiigthurreet2ri.evaFlrsoomfnWeoilnsoatnecrpoufpts(1b9y6a9)p,adritsuprileanytinfgemthaeleniunmtbheerdoafybfaorlplorewsisnegs
publinated exitingthehome,thedogdiesafterrunningbackintotheburning preawrtaurrditeidonw.iTthhethfeirsdtasmix’stroiawlsnwsiexrepurepws,aarnddedthweitrhemfoaoindd,ethrewneerxetrseiwxawrdeerde
dmi home,possiblytoretrievetheparrot(AssociatedPress,2007).
salliedisse traAck.yWouensglemyaAnurterecyo,vewrhinogwfraosmwaaitsienigzufroerftahlelstroanintowtihtehshuisbwtwayo wleivtehlfoofsrteertri(euvnarlelraetmeda)inpsuhpisghfrothmronuognheoxuptethriemeexnptearlimmeontht,ewrsi.thNaoltmeothsattfothuer
ite perminuteevenafter3hrofcontinuousretrievaltrials.Theexperiment
ofob daughters,jumpsdownintothetracktoretrievehimbut,without waseventuallyterminatedbytheexperimenters,tiredofadministeringthe
oneott timetopullhimout,insteadliesontopofhimbetweentherails procedurethatdidnotappeartobereachinganasymptoteinperformance
orsn as the train drives over them both, missing them by mere centi- in the highly motivated dams. From “Babies by Bar-Press: Maternal
i
ationand meOtefrfssp(Briuncgklreeyt,ri2e0v0a7l).surely evolved to protect the helpless BMeehthavoidosr,IinnsttrhuemRenatts,”&bCyoWmp.uEte.rsW,1il,spo.n2c2ro9f.t,Co1p9y6r9i,ghBte1h9a6v9iobryRSepsreianrgcehr
sociuser newborn, ensuring the reproductive success of the mother who ScienceandBusinessMedia.Reprintedwithpermission.
s
Aal shareshalfofhergeneswiththeendangeredpup.However,the
ologicalindividu scuifbicsetqaurgeentttehxaatmthpeleosbsienrvvoelrviencaunrsungrreelaattepdersoorneavlerniskhetoterreosscpuee- 2008). Even some ants appear selected to rescue and retrieve
he (see Trivers, 1971). While all species do not retrieve pups in conspecifics who have become trapped under sand or debris
ch
yt
Psof theirmouthslikerodentdams,allcaregivingspeciesshowsome duringforaging(Nowbahari&Hollis,2010;Vasconcelos,Hol-
ne analogous and costly behavior to retrieve and maintain contact lis,Nowbahari,&Kacelnik,2012).Suchbehavioralsimilarities
as
cu
erial with vulnerable, dependent, altricial offspring (e.g., Noren, across species suggest that altruistic retrieval may be a neces-
mn
Aso
eer
hp
ythe
bt
yrightedolelyfor
ps
coed
sd
in
entnte
mi
docucleis
hisarti
Ts
hi
T
Figure1. Imagesofretrieval.Thepanelontheleftdepictsatypicalrodentmotherretrievingheroffspringthat
becameseparatedfromthegroupornestingsite.AdaptedfromAnIntroductiontoBehavioralEndocrinology
(2nd ed., p. 356), by R. J. Nelson, 2000, Sunderland, MA: Sinauer Associates. Copyright 2000 by Sinauer
Associates.Adaptedwithpermission.ThepanelontherightdepictstherescueandcarefrommothergorillaBinti
Juaofa3-year-oldboywhohadfallenintohercompound.Bintipickeduptheunconsciousboyandcarriedhim
toalogadjacenttotheenclosuredoor,holdinghimgentlywhilewaitingforthetechnicianstoarrive.Drawings
byStephanieD.Preston.
ORIGINSOFALTRUISM 1307
saryprimitivebehaviorincaregivingandsocialspecies,which
needs to be examined across levels of analysis if we are to
determine the extent to which these similarities impact the
development of human giving.
Atthesurface,alloftheaforementionedexamplesinvolvea
distressed target who is vulnerable, hurt, or in danger, and
whose need is salient to the observer, who responds by ap-
proaching and/or retrieving them to relieve their distress or
sparethemfromharm.Intuitively,manysituationsofneedtake
thisform,andthus,theknownmechanismsforsimpleoffspring
retrieval appear to offer a promising starting point for an
integrated ultimate and proximate model of altruistic respond-
ing.Moreover,theneurobiologicalevidenceavailabletodatein
y. mammals suggests that the similarities do not just reflect an
shers.broadl ainnarleosgpyo,nisnewtohiscihmsiliamrilparrobbelehmavsioarcsroesvsoltvimede,inbudtisatcintucatlslpyerceieps-
publinated resent a homology, in which the behaviors are thought to have
dmi evolvedfromacommonsource,sharinggenesthatwerepassed
itsallieedisse dhiurmecatnlyaflrtorumismin—divaindduaml otostincdleivairdlyuaal.ltArucisctoircdirnegsptoontdhiinsg—viedwi-, Figure 3. Schematic of the relationship between offspring care and
ofob rectly evolved from the neural and behavioral systems for altruism.Mechanismsofactiveoffspringcare,particularlyretrieval,heav-
oneott caringforandretrievingoffspring.Assuch,altruisticrespond- ilyinfluencethenatureofaltruisticresponding—thefocusofthisreview.
orsn ing and offspring retrieval are presumed to not only share Activeoffspringcareusuallyresultsinpassivecare(e.g.,huddling,close
ationandi csuobmsmtraotnesgaonadlsgeonretsicurifnafcleuefnecaetsu.reSsucbhuat haolsmoolnoeguyroisbinoolotgoinclayl bnoudrtyurcaonncteac(te,.gn.u,rscionngs)o;laatniaolnogooruscloym,faoltrrtuinisgt)icornecsepotnhdeintagrgoeftteinsrreesmulotsveind
cier fromdanger.Passivecareandnurturanceoccurduringdevelopmentand
ssous parsimonious,itcancontributetoourbroaderunderstandingof socialrelationshipsaspartofthenormalmechanismsofnervoussystem
Aal altruism by providing a primitive building block upon which
ologicalindividu mexoarmepcleo,mepvleenxafbosrtmrasctocfasheusmofancohoeplepriantigoncaonrdboenaptliancgemd.onFeoyr dpmeroivmreeliotaipvbmestsreauncpttoaonnrdwrehbmioconhvdemindagfn.ryoAmfcottrihvmeescaoonfndtheepxlatpsoisnfigvoefdfescrpairrvieen,gae.revInetnhtuetrhnno,bsheeehtlahpvainitogarraiesl
chhe tostrangersmaybedrivenbythereleasingstimuliofavulner- asubclassofthebroadercategoryofaltruism,whichincludesdiversetypes
yt
Psof abletargetinimmediatedistresseveniftheyalsoinvolvemore ofgiving,notallofwhichareinfluencedbyoffspringcare(e.g.,strategic
ne significant deliberation or top-down cognitive processes that helping, bacterial altruism). Even abstract or complex forms of human
as
ericalu are not possible for some other species. altruism that appear complex may have roots in offspring care when
eAmerson argTuhmeencturfroerntarheovmieowlogsyumbmetwareizeensatlhtreuisutlitcimreastepoannddingpraonxdimoaftfe- f(aseceiliFtaigteudreby4).theperceiveddistress,need,andvulnerabilityofthetarget
hp
ythe spring retrieval, presenting the strongest evidence for their rela-
bt
yrightedolelyfor t(tiiaos)enp,sah(critp)icatihnpraoatduioagnphtbi1vy0enkooepnypmoshonatehrneecdrysf,ba(cebtt)owmrese:onCtooarvncocoeimdpatpuneactleelnyac,nebdoaatnhpdpreeroxqapuceirhre-, pfuesrec,epotriounnoinftenneteido,nawlhfiochrmissdoifffaelrternutiisamtedthfartomlikceolyopdeeraritvivee,frdoimf-
ps
coed and a facilitating role of (d) neonatal vulnerability, (e) salient other evolutionary and mechanistic origins. Altruistic responding
isnd distressand(f)rewardingclosecontact.Physiologically,theyalso can be considered a subtype of Warneken and Tomasello’s
entnte share neurohormonal support from (g) oxytocin, (h) the domain- (2009b)“helping,”whichexistsinevenyoungchildrenandsocial
mi
docucleis gtheeneorrablimtoefrsoonlitmalbcoocrotretxic.aTlhsyisstreemvi,e(wi)itnhteenctiinognualllayteinctoerrtmexix,easndul(tji)- MmacmCamrtahlsy,lik1e99a5p;esW,daorgnse,kaennd&dolTphoimnsas(edleloW, 2aa0l0,92b0)0.9I;tMisasaslosno&a
hisarti mateandproximatelevelsofanalysisundertheprinciplethatcom- subtype of de Waal’s “directed altruism,” which is directed at a
Ts
hi plex human behaviors are best understood as extensions of ancient specific individual, rather than broadly, as with alarm calls (de
T
behaviorsthatrelyuponhighlyconservedneuralmechanisms,which Waal,2008).
in turn influence even complex human phenomena in adaptive and Because these existing subclassifications were not created to
predictable ways. Before diving in to the key arguments, a few group types of altruism by their motivations or evolutionary ori-
definitionalandconceptualissuesareaddressed,includingtheoper- gins, they aggregate cases with different mechanistic properties
ationaldefinitionofaltruisticrespondingandhowitrelatestoother (e.g., high and low explicit cognitive processing). Altruistic re-
knowntypesofgiving,bothactiveandpassive. sponding further narrows these classifications to only include
cases where the motivation to respond is fomented by direct or
indirect perception of the other’s distress or need (see Figure 4).
Defining Altruistic Responding
This excludes cases that emerged later in time or include diverse
Altruistic responding is defined as any form of helping that processes, such as cooperation or helping influenced by strategic
applieswhenthegiverismotivatedtoassistaspecifictargetafter goals,socialnorms,displayrules,ormatesignaling.Suchcauses
perceiving their distress or need (see Figures 3 and 4). Altruistic surelyparticipateinthehumanaltruisticresponse,buttheyarenot
responding implies an active behavioral response initiated by the consideredearlyorkeydriversinprimitivecare-basedaid,which
1308 PRESTON
Palagi,Paoli,&Borgognini,2004).Accordingtoanimalmodelsof
offspringcare,consolationisa“passive”andconsummatoryform
ofcaregiving,whileapproachingandretrievinghelplessnewborns
is“active”andappetitive;however,bothsharesignificantconcep-
tual,mechanistic,andevolutionaryorigins(seeFigure3).
Active,abstract,andpassivegivingareallevolutionarilyaltru-
istic,astheybenefitanotheratacurrentcosttothegiver.Mone-
tarydonationscarryaliteralcostandconsolationcarriesanener-
getic cost (Noren, 2008); both entail opportunity costs for more
self-servinggoalslikematingandforaging.Consolationaddition-
allyentailsanincreasedriskofpredation(Maestripieri,1993)and
conflictfromopposingalliances(deWaal,1982,1989).Allthree
types of helping are also motivationally altruistic when they em-
y. anatefromapprehensionofthetarget’sdistressorneed(seeFigure
shers.broadl 4ti)o.nTalhuosri,gaincstivweithhutmheanmhoerelpainbgstrsahcatreasndevpoalsustiivoenacrayseasnwdhmenottihveay-
publinated arealsobasedinthemechanismsforoffspringcare.
dmi Inaddition,theneuralmechanismsofallthreeoverlapaspartof
salliedisse aansduipterootefcstiovceiaalgbgerhesasviioonrs(Mlikaesosnex&ualMbeenhdaovzioar,,1s9o9c8ia;lNbeowndminang,,
ite
ofob 1999;Numan,2011;Panksepp,1986).Forexample,theneuropep-
oneott Figure4. Driversofthealtruisticresponse.Thisisbasedonthefixed- tidesrequiredforactivecare(e.g.,oxytocin[OT]andvasopressin)
orsn actionpattern(FAP)qualityofoffspringretrievalandaltruisticresponses. also subserve other social processes (Carter, Grippo, Pournajafi–
i
ationand Ctoehrtealipnpqrueadliistipeosseinatchteiotna,rgliekte,tshiteuatatirogneat’ls“srtealteea,saenrsd”tohfecoabrseetrhvaetr’nseaebdilniotyt NPaanzkasroepopl,,1R9u8s6c)io,,pr&oduPcoirnggesth,e2o0ri0e8s;aIbnosuetl,th1e9r9o7le; oNfusmocaina,lb2o0n1d1-;
cier
sous requireextensivecognition,deliberation,orperspectivetaking.Eachqual- ing mechanisms in human morality and prosociality more gener-
s
Aal ity(indiamonds)istheoreticallydistinctandadditive,thoughtheytypi- ally (e.g., see Barraza & Zak, 2009; S. L. Brown, Brown, &
ologicalindividu callyco-occurinneonatalandcompellingneedsituations. 2P0e0n8n;er,F2e0h1r1,;FSi.sLch.bBarcohwenr,,B&rowKno,s&felPdr,es2to0n0,52;01H1e;inCrhicuhrcs,hlavnodn,
he Dawans, & Domes, 2009; Hrdy, 2009; Kosfeld, Heinrichs, Zak,
ch
yt
Psof canbeobservedacrosscaregivingmammalssuchasrodents,who Fischbacher, & Fehr, 2005; Morhenn, Park, Piper, & Zak, 2008;
anse haveminimalcapacitiesforexplicitdeliberationandlittleselective Numan, 2011; Panksepp, 1986; Singer et al., 2008; Taylor et al.,
cu
erial pressure to appear beneficent. Multiple benefits are achieved by 2000; Zak, 2008). All of these behaviors require approaching an
Amson carving nature at its joints in this way, from the bottom-up. By unfamiliartargetthatwouldnormallybeavoided,indicatingakey
heper identifyingatypeofaltruismthatisdefinedbyitscommonform, need for approach-avoidance opponent processes (see below and
ythe function, mechanism, and evolutionary origin, we can finally Insel,1997;Insel&Young,2001;Mason&Mendoza,1998;Moll,
bt
yrightedolelyfor mfbreaetmrwgeeewemnorouktli,tvimawtiaeotencaaalnnadncdopmreovpxorilemuhtaeiotnendalretyvheeelxsepsolsafennaatntiiaaollnyssiniistn,etroarnealdactowiohenesrcheainnpt GZdeaahlOlnal,givhdeeeirr,aO2-lS0ivo0eu9iz)r.aa,-TS&houuZsz,aach,onKn,cr2eu0pe0tgu8ea;rl,Mly&,olelGv&oralufSmticoahnnua,lrki2iln0y,0,52a;0n0dS9km;uMesceohl&al-,
ps
coed shed light on multiple remaining paradoxes in the literature of nistically, abstract and passive caregiving overlap considerably
isnd giving(thedistinctionbetweenmotivationalandevolutionaryex- withactivealtruisticresponding(seeFigure3).
mentinte planationsforaltruismisprovidedinBatson,2011;deWaal,2008; Despite this clear overlap, the current article focuses on overt,
docucleis Krebs,2011;Preston&deWaal,2011;Sober&Wilson,1998). aqcutiirvees,aadltdruitiisotnicalrefsepaotunrdeisngthpaetrasree,nboetcaruesqeuitrheids tfyopreaobfstraaicdtroe-r
Thissarti The Relationship Between Altruistic Responding and passive cases and have yet to be explicated (see Figure 4). For
Thi Abstract or Passive Care example,rodentoffspringretrievalandhumanaltruisticrespond-
ingappeartorequireauniqueroleforthemedialpreopticregion
Ofcourse,modernhumanlifeaboundswithexamplesofaltru- ofthehypothalamus(MPOA),motorsystems,autonomictone,and
isticrespondingthatareprobablymotivatedbytheinstincttocare theknowledge/abilitytoenacttheappropriateresponse—features
for a distressed other but are more abstract or less active than that are less germane to abstract or passive giving. Motor-
rushing to save a stranger from a burning building or icy waters. motivational and expertise processes are particularly underappre-
For example, people altruistically write checks to feed starving ciatedincurrentmodelsofaltruism,whichmaybekeytounlock-
children on the other side of the world and arrange for people in ing existing paradoxes in the literature, such as the fact that
the community to deliver meals to ailing neighbors. In addition, empathyandaltruismdonotalwaysco-occur(e.g.,seeBuchanan,
forms of aid like comforting and soothing—known as “consola- Bagley, Stansfield, & Preston, 2012; Preston & Hofelich, 2012;
tion” in apes—are surely also forms of caregiving, but ones that Prinz, 2011) and the fact that helping can be predicted both by
are much less active than the heroic cases provided above (Cor- contagious distress and physiological arousal as well as by con-
doni,Palagi,&Tarli,2006;deWaal&Aureli,1996;deWaal& cerned attention and physiological deceleration (see reviews of
van Roosmalen, 1979; Palagi, Cordoni, & Borgognini, 2006; physiological effects in Batson, 2011; Eisenberg, Wentzel, &
ORIGINSOFALTRUISM 1309
Harris,1998).Thus,thecurrentarticlefocusesuponactivealtru- isincludedfirst.Moredetailedinformationcanbefoundinexist-
istic responding, but it is assumed that most of the described ingreviewsonthemechanismsofmaternalcare(seeJ.B.Becker
mechanismsalsoapplytoabstractandpassivecases,particularly & Taylor, 2008; Lonstein & Morrell, 2007; Maestripieri, 1999;
whentheysharemotivationalprocessesassociatedwiththedrive Numan,2007;Numan&Insel,2003;Rosenblatt,1992)orreward
toward young or distressed individuals who are vulnerable or in anddecisionprocesses(seeBechara,Damasio,&Damasio,2000;
danger. Damasio,1994;Kringelbach&Rolls,2004;Robinson&Berridge,
2003;vandenBos,McClure,Harris,Fiske,&Cohen,2007;Zak,
Article Summary 2004).
Thefollowingreviewsummarizestheargumentforaproposed
The Ultimate Origins of Altruism in Offspring Care
homologybetweenrodentoffspringretrievalandhumanaltruistic
responding. First, ultimate arguments for altruism are reviewed
Existing Ultimate Views From Biology and Economics
and integrated by framing altruistic responding as a rudimentary
y. and evolutionarily conserved neurobehavioral mechanism for re- Ultimate theories for why altruism is adaptive and how it
shers.broadl shpoomnodlionggytobeottwheeersn’ rnoedeednst. Tofhfespbruinlkg oreftrtiheevarlevainewd aslutprupiosrtitcs trhee- ebvinoelvbeedhalavrigoeralyldceommoenfsrtoramtiobnisoloofggyivainndg,ehceolnpoinmgi,cso,rwcohoicphercaotimon-
dpubliminated sCpoonncdeipntguabllyy,hbigohthligrhetqinugire10(a)kepyarftiaccitpoartsiosnhabryednobnetmwoetehnersth,e(mb). wsoi.thMmoadtehlesminatnicoanlhmumodaenlsbtihoalotgdyemfoocnusstroantectohoepoeprtaitmioanliitnyoeufsdoociinagl
salliedisse mtwoeteonracvoomidpaentecnecaenadnadpepxropaecrthis,ea,n(dc)aafnacailditaapttiinvgerooplepoonfe(ndc)ynbeoe-- afintnimesaslsanodr mdiircercotorregcainpirsomcsit,ywbheitcwheceannthbeeheixgphllayiniendtebrryeliantceldusainvde
ofitobe natal vulnerability, (e) salient distress, and (f) rewarding close interdependent organisms (e.g., see reviews in Dugatkin, 2007;
oronesnott csuopnptaocrtt.fPrhoymsi(ogl)ogoixcyatlolyc,inb,o(thh)btehheadvoiomrsaianls-goesnhearraelnmeeusroolhimorbmoocnoar-l S20tr0a7s;smWainlnso,nZ&hu,H&ölldQoubelellre,r,2020050)0.;TWhoesset,adGdrrieffsisni,ng&huGmaardnnaelr-,
ationandi tciocratlexsy(ssteeemT,a(bil)et1h)e. cingulate cortex, and (j) the orbitofrontal trreuciispmrocuitsyuamlloydeals,suwmhiechafomcoudseifsieudpovnerlsaitoenr,moforTeriivnedrisr’esct(r1e9t7u1rn)
sociuser Eachofthe10keyfactorsfocusesondatafromtheneurobiol- rewardstogivers(e.g.,seereviewsinFehr&Rockenbach,2004;
Asal ogyofrodentoffspringretrieval,becausethisprovidestheclearest Gintis et al., 2003; McAndrew, 2002), including “indirect reci-
hologicaleindividu faleunvndeclet.ixoDtneianvlseiarvnseeddpdaratoataxoimfnrotahmteishobotehmheaorvlosiopgreycatiloereshauadnmydaenxdioasmlttraautiinasstfiacairrreleysipnsopcnleudcdiifneigdc pSGrirogacmfietunyn,”d1(,A9199le09x;8a;JnoWdhenersd,te1ok9ni8en,7d;,11M999.95A8;).Z,N“achooawsvtailky&,s2igZ0na0ah5la;invMgi,.t1hA9e.9o7Nry)o,”w“(CcaokSmT&-;
ch
yt wherever possible to support the assumption that these data gen-
Psof petitive altruism” (Roberts, 1998), and “strong reciprocity”
ne eralize to other contexts and species, particularly human helping. (Bowles & Gintis, 2004; Fehr, 2004; Gintis, 2000). These “mul-
as
ericalu Abriefoverviewoftheneurobiologyofrodentoffspringretrieval tilevel fitness” models largely assume that altruism is adaptive
Amson toward relatives, direct reciprocators, and other group members
eer becauseitaffordslatersuccessincompetitionforcriticalresources
hp
ythe Table1 suchasmates,coalitionpartners,andfood.Thesemodelsarealso
bt
yrightedolelyfor Tanhde1A0ctKiveeyOFfafscptroirnsginCaCroemmonBetweenAltruisticResponding soSuftrpaEpiaot,srAtteeurdnsbtrPyaalhiraaugm(uBaalnyieag(nHethBawriorkdpe,osSl,omg1ii9ct9ha1,l)&daaBtnadirfdrto,hme20M0th1ee;riSAammchiteohff&oTraBogrierredrss,
copeds Number Keyfactors 2000),thatshowthatthosewhosharearebenefittedlaterthrough
mentisintend 12 CThareeRMequustirBemeeEnxttfeonrdMedoBtoeryConodmPpaertetunrcieenatnMdothers ifnamcreilayseddurminagtirnegsosuurccecebsos,ttlseonceicaklsst(aBnodoinnge,,1a9n9d8)s.upport to their
docucleis 3 ThEexApderatpistieveOpponencyBetweenAvoidingand ThTehaefoorreemtiecnatliolnimeditmatoiodneslsosufrebliyolsougpipcoarltathnedeveoclountioomniocfvpireowsos-.
Thissarti 4 AtAtrapcptriooancThionwgardCuesofNeonatalVulnerability cialbehavior,butvirtuallynoneattempttoexplainactiveorheroic
Thi 5 TheSalienceofNeonatalDistress aidtoendangeredordistressedtargets,exceptwhentheyinclude
6 Rewarding,CloseContactAugmentstheMotivation such salient acts in the introduction to demonstrate humans’ vast
toAssist potential for altruism. Moreover, these models do not attempt to
7 OxytocinCriticallyReducesAvoidanceand
address how such motivations or decisions are proximately con-
IncreasesPassiveCare
veyed,exceptwhenassumingthatcooperationishardwired(asin
8 TheNAccandDAMotivateApproachandMediate
ConditionedSocialRewards bacteria)orisexplicitlymediatedthoughanaccountingoffavors
9 IncreasingtheEmphasisoftheACCinCaregiving (inhumans).Forexample,modelsofstrongreciprocityandhuman
andAltruism cooperationassumeahighlevelofcognitionandawareness(e.g.,
10 ThePFCIntegratesMultisensoryCuestoPromote
Bowles & Gintis, 2004; Fehr, Fischbacher, & Gächter, 2002;
AdaptiveResponding
Gintis,2000),whichcannotexplainwhyaltruisticrespondingalso
Note. Theproximatereviewisorganizedaroundthesetenfactors.Other occurs in mammalian species like rodents and dogs that lack
factorssurelyexist,butthesearethemostsalientcontributorstoaproposed
significant abilities to deliberate. Altruistic responding is also
homologyforwhichextensiveevidencealreadyexists.NAcc(cid:1)nucleus
accumbens;DA(cid:1)dopamine;ACC(cid:1)anteriorcingulatecortex;PFC(cid:1) commonlydescribedasacompulsionratherthanadecision,albeit
prefrontalcortex. onethatiscontext-sensitive.Forexample,heroesintheHolocaust
1310 PRESTON
whorescuedandhidvictimsintheirownhomesreportedhelping ziano, & West, 1995; Roney, Hanson, Durante, & Maestripieri,
whendirectlyconfrontedwiththeother’sdistress,whichproduced 2006). Some food-sharing research involves real risk to targets
feelings of empathy, love, compassion, and a personal responsi- whocouldstarvewithoutthesharedresource,butapartfromafew
bility that simply compelled action, without “considering risk or anthropological studies (e.g., Bliege Bird, et al., 2001; Hawkes,
thinkingaboutbeingeitherlaudedormaligned”(Oliner,2002,p. 1991; Kaplan & Hill, 1985; Smith & Bird, 2000), most of this
127; see also S. W. Becker & Eagly, 2004). Similarly, even workisperformedinnonhumans(e.g.,birds,vampirebats,mon-
bystanderapathystudiesfindthatnearlyeverysubjectwhofaced keys, and apes, see Dugatkin, 1997; Preston & de Waal, 2011;
thetargetalonehelped,thevastmajorityinfewerthan10s,with Wilkinson, 1990) and almost never in the context of risky depri-
fewreportinganyaccompanyingthoughtswhentheydid(Darley vation where giving is most expected. The fact that real and
&Latané,1968). immediate need is a key motivator of offspring care, which is
Humans may minimize the extent to which they contemplate rarely included in experimental protocols, could partly explain
decisions, and they can surely calculate decisions implicitly, but perplexing null results, such as those of ape food-sharing studies
evenyoungchildrenandchimpanzeesexhibitaltruisticresponding (Brosnan et al., 2009; Hrdy, 2009; Silk et al., 2005; Silk, Paul,
y. and show limitations in their capacity for helping only when it Colin,Ernst,&Russell,2009;Warneken&Tomasello,2009b).As
shers.broadl rperqouceirsessesin(hWibairtinnegkepner&soTnaolmraesweallrod,s2o0r0s9ibg)n.iMficoarnetotvheero,rtyheorfemisinnod eisvmidoenstceo,ftwenheinndfoiroedctshreasrpinognsdeoteosboecgcguirnign(sHorcdiayl,g2r0o0u9p;sPorfesatpoens&it
publinated evidence that chimpanzees can anticipate return favors, which de Waal, 2011). Moreover, recent experiments revealed greater
dmi furtherunderminestheassumptionthathelpingrequiresanexplicit prosocial behavior in apes when the experimental context was
salliedisse macocdoeulnstianlsgooefmfpahvaosrisze(dtheeWrewaaalr,d2s0t0h8at).arMeabneystohwuemdaunpeocnohneormoeics cWhaanalg,e2d01to1)b.eNemuororeimnaagtuinrgalissttuicdie(Hsoorfncehr,arCitayrtdeor,naStuiocnhaakr,e&modree
ite
ofob to support the adaptiveness of such startling acts (e.g., see directlyrelevanttoaltruisticrespondingbecausetheyusetargetsin
oneott Macaulay & Berkowitz, 1970; Wong, 2000); however, a focus real need and positive rewards for giving; in accordance, these
orsn uponreturnrewardsisalsoinconsistentwiththefactthataltruistic studies also find the most pervasive activation of all neuroeco-
i
ationand rreestrpioevnediunngreclaantebdepoubpssedrveesdpitaecrtohsesfsapcetcthieast.tFhoeyr edxoamnoptlee,ngroadgeenitns n(Homaribcaupgarha,dMigamyrs,i&nBthuergnheaurrta,l2r0e0g7io;nMsoilnlveotlavle.,d2i0n06o;ffTspanriknegrsclaerye,
cier
sous social signaling of beneficence or receive concrete rewards for Stowe,&Huettel,2007).
s
Aal helping. Even when neuroeconomic studies do use targets in need
ologicalindividu capBtuiorleogailctraulisatnicd reecsopnoonmdiincgulptiemrasteebmeocdauelsseatrheeyalsmoaukneliaktellyeatsot seuxcphlaains cahltarruiitsietisc, trheesproensudlitnsgcabnencaoutsneecthesesaprailryadbigemexstevnirdteudaltloy
he implicitpredictionsaboutwhoshouldhelpthemost,whicharenot always entail highly deliberated cost–benefit decisions made
ch
yt
Psof consistentwithhumandata.Forexample,becausetheirmodelsare withmoney(forextensivereviewsofthesestudies,seeCamerer
ne largely based upon the role of sexual display and male–male &Fehr,2006;Fehr&Camerer,2007;Molletal.,2008;Preston
as
cu
erial cooperation in shaping the evolution of altruism, they at least & de Waal, 2011; Walter, Abler, Ciaramidaro, & Erk, 2005;
Amson implicitly favor the expression of altruism in males. Males are Zak, 2004). These studies do find a key role for the domain-
eer morelikelytoengageinthemostheroicformsofaltruism,likely general portions of the caregiving system including the orbito-
hp
ythe because of a more general selection pressure for sexual dimor- frontal cortex (OFC), nucleus accumbens (NAcc), insula, and
bt
yrightedolelyfor pmohroiessvmtehnearniondiccrdeciasapsseeladsysm.uHraetoilnywgesuvsrueprc,acsetshsest,hceaonbsdetsnfeeofmfitasinleojusfrpayruebolraiccdtieutyaa,tlhlmyienmdaothlrsee, arseimognyusgladttoainlagg,ivsetwruh(Bcicteuhcrheassuraaplpseootratsslu.,bts2he0er0v0ei;dcDeoaammpthalesaixto,apn1rd9im9e4ixt;ipvPleirceistatfodfneecc&ti--
ps
coed likelytorespondinalmosteveryotherprosocialsetting,including de Waal, 2011; Preston & Hofelich, 2012). For example, sub-
isnd risky contexts (S. W. Becker & Eagly, 2004). Moreover, human jects viewing pictures of former cooperators showed increased
entnte helping is known to shift with hormone status, pregnancy, and activationintheamygdala,NAcc,OFC,insula,fusiformgyrus,
mi
docucleis etexnptowsuirtehteoxoisftfisnpgrinmgul(trielevvieewlefditnbeeslsowvi)e,wwshibcuhtiissaplsroednicottecdonbsyisa- atinodnsiuspewriaorrratnemtedporbaelcsauulsceusst(uSdTySin).gDaelstpruitiesmthiisnovanerleaxpp,lciacuit-,
hisarti caregiving view of altruism (see also Hrdy, 2009). The current financialcontextcanrenderdecisionsmorecontrolled,explicit,
Ts
hi review aims to integrate these inconsistencies by building more calculated, and planned than nonmonetary decisions (J. M.
T
complex forms of altruism from the primitive form of altruistic Wang et al., 2012), which may not generalize to the processes
respondingdescribedherein. required for simple behaviors like retrieval.
Empirical limitations of biological and economic views. Phenomenologically, every-day acts of helping—such as help-
Extensive laboratory research on cooperation in economics and ingalosttoddleratthemall,pickingupaneighborhoodchildwho
biologyisalsonotaimedatcapturingaltruisticrespondingperse, fell off his bike, or surrendering your seat on the bus—do not
because the paradigms do not entail overt aid to targets in real necessitateatrade-offbetweenhelpingandmonetaryrewardand
need—akeypreconditionforcare-basedaid.Thisappliestogame- may not involve any explicit decision whatever. Such acts are
theoretic models (e.g., Gintis, Smith, & Bowles, 2001; M. A. decisions in the nominal sense that all motor acts are decisions,
Nowak,2005;Panchanathan&Boyd,2004),behavioraleconom- becausethereweremultiplepossibleoptionsandonewasselected
icsexperiments(e.g.,Fehretal.,2002;Hardy&VanVugt,2006; through an implicit, cost–benefit calculation. However, as with
Jerdee & Rosen, 1974; Milinski, Semmann, & Krambeck, 2002; actionselectioninmotorbehavior(Grèzes,Tucker,Armony,Ellis,
Milinski,Semmann,Bakker,&Krambeck,2001),andmatepref- &Passingham,2003;Tucker&Ellis,2001),theexistenceofother
erenceexperiments(e.g.,P.Barclay,2010;Jensen-Campbell,Gra- alternativesischaracteristicofthesituationandnotthemindofthe
ORIGINSOFALTRUISM 1311
actor—a critical distinction that is greatly underappreciated by moral behaviors to humans and possibly apes because of their
modelsofmoralityandaltruism. larger emphasis on complex perspective-taking skills
The controlled, cognitive processes required for neuroeco- (Warneken,Hare,Melis,Hanus,&Tomasello,2007;Warneken
nomictasksmayeveninhibitthehelpingimpulsebydisengag- & Tomasello, 2009a, 2009b). Batson holds a caregiving view
ing the emotional-motivational state from the current context thatissimilartodeWaal’sbutthatisevenmorehuman-centric
(S. L. Brown, Brown, & Preston, 2011). For example, even thanTomasello’s;Batsonassumesthatmotivationalaltruismis
indirect priming with the concept of money has been shown to only possible in humans and does not derive from inclusive
make subjects less prosocial and more goal-directed (Vohs, fitness,reciprocity,orgeneticsuccess(seeBatson,2010,2011).
Mead, & Goode, 2006). Moreover, money is a hedonic, incen- Thus,mosttheoristsmakeaconnectionbetweenoffspringcare
tivizingstimuluslikeanaddictivedrug(Lea&Webley,2006), and human prosociality, but they vary widely on whether such
which can compete against reward-based motivations to help. phenomena are assumed to exist across species and entail
As evidence, in altruism studies of children and chimpanzees, higher level cognitive capacities. Moreover, none of these
subjectsdonothelpmorewhentheyareclearlyofferedrewards views explicitly address offspring retrieval or its neural bases
y. for helping; 20-month old children actually help less in the (butseeS.L.Brown,Brown,&Preston,2011;MacLean,1985;
shers.broadl p&reTseonmcaeseolfloc,o2n0cr0e7t)e.rMeworaerodvse(rW, taermnepkoernal, Hneaurrea,lMreegliios,nsHaanssuos-, Pupreosntotnhe&ndeeurWobaiaoll,o2g0y11o)f.aTlhtruusi,sttihcerceusrproenndtirnegviepwerissef,ocaussaend
publinated ciated with explicit theory of mind processes are actually de- encapsulatedstartingpointforthealtruisticurge,whichcanbe
dmi activatedwhensubjectsviewrewardingpicturesofoffspringor augmented to address more complex forms of giving that vary
salliedisse roofmfspanritnicgpcaarrtenesryss(tBemartceolsm&biZneeskia,2p0r0im0,it2i0v0e4s)y.sTtehmus,fworhialcetitohne more widely across species.
ite
ofob with cortical modulatory inputs, these inputs can both initiate The Current Ultimate View
oneott andinhibitaction.Studiesareneededthatspecificallyexamine
orsn the neural correlates of active responding and measure the Thematernalinstinct,whichimpelsthemothertoprotectandcherish
ationandi espffoencsteo.f explicit processes or monetary incentives on the re- hItesriymopuunlgs,eisiscopmrimmoanriltyoatolmaofsftoradllpthheyshiicgahlerprsopteecciteisonoftoanitmhealcsh.i.ld.,.
cier especially by throwing the arms about it; and that fundamental im-
os
ssu pulse persists in spite of the immense extension of the range of
Aal
aldu Existing Caregiving Ultimate Views applicationoftheimpulseanditsincorporationinmanyidealsenti-
ologicindivi In contrast to the work described above, psychology has often ments.(McDougall,1908/1923,pp.69,74)
he studied the direct response to another’s immediate need (e.g., An integrated caregiving framework presumes that altruistic
ch
yt
Psof Batsonetal.,1997;Darley&Latané,1968;Eisenberg&Strayer, respondingaccommodatesdirect,inclusive,andreciprocalfitness
ne 1987; Latané & Rodin, 1969; Warneken & Tomasello, 2009b; pressures by assuming that altruistic responding persists as the
as
cu
erial Zahn-Waxler, Radke-Yarrow, Wagner, & Chapman, 1992) and byproductofastrong,necessarypredispositiontoassistone’sown
Amson typically does so through a framework that assumes helping offspring. Behaviors designed to care for altricial offspring—
eer evolvedfromtheneedformammalstocareforhelplessoffspring particularly nursing and emitting and responding to separation
hp
ythe (e.g., Batson, Lishner, Cook, & Sawyer, 2005; S. L. Brown, calls—areconsideredfundamentaltotheemergenceofearlymam-
bt
yrightedolelyfor 21B09r08o57w;;nP,Hr&redsytPo,rne2s&0to0n9d,;e2MW01aac1aL;l,eda2en0,W111a)9.a8lM5,;2o0sMt08aor;fshEt,hiseAesnedbaeemxrgiss,t&in&gSKtervaleoycleukr-,, cmaoftaneltrsrafhsraot,tmcmhiranemgptmi(lMeasl,ascwLahseoadnpi,srot1av9ni8td5ef;rvoPimrrteusahtloulnymna&noscHaaorseferfloiocdrhet,nht2esi0r1cy2loe).aurnIlngy
ps
coed tionary theories are not detailed, but a few are more comprehen- respondtothedistressofconspecificswithaffectiveempathyand
isnd sive,suchasBowlby’s(1958)workonthemother–infantattach- reactive altruism (e.g., Bartal, Decety, & Mason, 2011; Chen,
entnte ment system, McDougall’s (1908/1923) writing on the parental Panksepp, & Lahvis, 2009; Church, 1959; Langford et al., 2006;
mi
docucleis ivnusltninercatbalendtartgeentds,eranedmEoibtilo-nEsibethsafetldptr’osp(e1l97c1a/r1e9g7iv4e)rvsietwospornottehcet MSoamsseeremveann,suWgegcehstkitnh,at&anTtsererxish,ib1i9t6a4;hiRghicley s&imGilaarinreers,cu1e96b2e)-.
hisarti behavioral ecology of responding to offspring across species. havior, as they retrieve trapped conspecifics from under sand or
Ts
hi MacLean (1985) also proposed that key developments in the debris(Nowbahari&Hollis,2010;Vasconcelosetal.,2012).Ant
T
mammalianbraincorrespondedwiththeshifttowardaltricialcare rescue is at least analogous to offspring retrieval in mammals. It
of offspring, including the development of nursing, play, and the mayalsobehomologousbecausethebehavioristhoughttoderive
separationcall.Morerecently,comparativeviewshavearguedthat from perception-action processes similar to mammalian empathy
empathy,altruism,andmoralityexistasextensionsoftheneedfor (Buchananetal.,2012)wherebyolfactorystressfromthevictimis
mammals to care for offspring (de Waal, 1996, 2008, 2009) or perceived by the giver, who then also becomes hormonally
perhaps particularly the need to provide them with cooperative stressed and helps. Without inferring any intentionality, these
care(Hrdy,2009). behaviorsatleastservethesamefunctionacrossspecies,theyare
Despite general agreement that caregiving pressures influ- adaptive for similar reasons (i.e., kin selection), and they share
enced the evolution of helping, there is significant controversy somerudimentaryaspectsoftheproximatemechanism.Ofcourse,
over whether these phenomena extend to nonhumans. For ex- ant brains are different from rodent or human brains, and the ant
ample de Waal proposes a “Russian doll” model that assumes form of rescue need not operate through maternal care circuits.
gradually increasing prosocial competency across species (de However,thecorecommonalitiesinfunction,behavior,andmech-
Waal, 2008), while Tomasello and Warneken restrict most anismcouldstillrepresentahomologyacrosstaxatofacilitatethe
1312 PRESTON
interindividualtransferofemotionoraid,butitcouldalsorepre- totalstrangerswithoutnegativelyaffectingfitnesswhentheaidis
sent an analogy in which similar mechanisms spontaneously and veryrareorlowcost(likeafewdollarsorminutes;Neubergetal.,
efficiently emerged to support such behaviors when necessary in 1997).Inaddition,itwouldbeverydifficultinneuralevolutionto
socialspecies. modifyapowerfulandnecessarymotivationtorescueendangered
Retrieval as a fixed action pattern (FAP). In this model, offspring in a way that prevented such accidental extensions to
offspringretrievalbehaviorisassumedtorepresentasortof“fixed nonkin, without also impeding the primary goal of ensuring off-
action pattern” (FAP; Lorenz & Tinbergen, 1939) that can be springsurvival.
released toward nonoffpsring when the conditions are similar to Does this mean that altruism, as an FAP, is a simple tolerated
those of offspring need (similar to the “misplaced parental care error? Other reasons for the persistence of altruism positively
hypothesis” for avian cooperative breeding, e.g., Hrdy, 2009; supportthebehavior,whichadaptivelybalancestheneedsofself
Ligon & Burt, 2004). In an often-described FAP, geese sit upon and other. For example, costly, heroic, and dangerous help is
and retrieve eggs that roll out of the nest. The geese do not only already limited by the offspring care mechanism, which is only
retrieve their own eggs but will also retrieve any similar stimuli releasedtowardvulnerableandsignalingtargetswhentheobserver
shers.broadly. uathsnadtnheerthmleaiarrgteoerwneagnlgecsgoognfsd.oiTttihohenirsss.speMeecmoireienso,gvaleyrre,or“desdturbipeeevhrenadovriemovreacnl”amnstobimreeutaqliku,eiscnukcalyhs dTnoeheeedsm,nmoectihnfeaimanriiszfmoinrgahlistshoeoprerhoxedteruncotewstonaiswdahfoeinctyhlyahnwedlhpkennisotwhfaeslsthaeorlgywesttooilsirceinistpecodlneoadrr.
publinated anindicationthatthebehaviorisproximatelyencodedwithrefer- manipulated (see S. L. Brown, Brown, & Preston, 2011). In
dmi encetotheperceptualfeaturesofeggs,ratherthanthroughdirect addition to these costs that are not actually that costly, altruistic
itsallieedisse rinefgeriesnacsesutomtehdeirtogerneeptriecserenltataedfnleexsisb.lSeimFAilaPrltyh,aatltcraunistbicerreeslpeoasnedd- rreestupronndreinwgaradlssothpartowdeurceesalpreoasdityivseugbgeensetfeidtsbtyopgriivoerruslttihmroauteg-hletvheel
ofob towardoffspringandnonoffspringalikewhenthesituationshares models (described above), such as inclusive fitness, reciprocity,
oneott featuresincommonwithoffspringinneed.Thesefeaturesinclude group fitness, and social signaling. The caregiving system can
orsn vulnerability, distress, and immediate need, all of which are not even provide the mechanism for “strong reciprocity,” without
i
ciationerand onneslys.dTihrercotugmharekveorlsutoiofnn,etehdesbeufteaatlusoreisnwdierreectshmaaprekdertsooafdarpetliavteedly- r2e0q0u6i;ridnegWanayal,e2x0p0li8c;itPrreesctoornd&kedeepWingaa(l,S2.0L0.2bB).roHwenlpi&ngBnroonwkinn,
sous elicitcaretowardrelatedorinterdependentindividuals;however, in one’s own group additionally helps givers by ameliorating the
s
ologicalAindividual btoewcAaasurdseeaurtnlhyreeyalasatre1ed9i0ni8nd,diriMvecicdtDumaoalusrgkaaenlrlds,saitmhdeuillratesrlle(ysaesaeergrFsuicegadun,real4so).elicitneed Bnaneugdcahptairvneeadnaetfiefoetncat(ls.e,o.g2f.0,d1sis2ep;elaEAyicnseoobnfod&&istPrToehsteosgmoanal,nh,1e91a99lt49h;2,;HgBreonoudwprlibkhysa,,rm2109o05n58y;;,
he Hoffman,1981;Murray,1979;Zeifman,2001).
ch
Psyoft whenwesee,orhearof,theill-treatmentofanyweak,defenceless Thus,altruisticrespondingisnotsimplyamistake.Itisneces-
ne creature (especially, of course, if the creature be a child) tender saryforone’sownreproductivesuccess,whichmakesitdifficult
as
ericalu emotion and the protective impulse are aroused on its behalf. The to constrain, and it is already balanced by the mechanisms of
eAmerson rcersypoofnhseerischaislddoirrechteranimdpinuslstaenttoanfleyotuosiatssdtheefemncoet,haenr’dsietmisoetsiosennatitatlhlye cTahreesgeivcionngdiintioanswraeyndtheartabceanreefgitisvibnogt-hbatsheedrvecieewiveardaapntdivteheacgciovredr-.
ythhep thesameprocess.(McDougall,1908/1923,p.77) ingtoalloftheconditionsoutlinedbyHoffman(1981).Theyalso
bt
yrightedolelyfor Ilaamstespdpo,ortniantnantnaleyteo,,umus,nosadtleeterrneraobbtylieopleoodrgbiusenthscaodvnoitorronsloltathbadlteeafibrneuetharFarAdthPteosrcdaosenfeitnrnoeclatophnsecume- afsoccracfoofmonlemd’isondgoawltaenteorit-nhacerlrruivsuiilvntiegmbfaiettenn-eelfseisvtsewltohvehienewlphsienlbgpyiinnagscsooufomfpsiepnrrgaintiagv,ekaegnyrdoruobplyes
ps
coed enacted,areexpressedbyalltypicallydevelopingmembersofthe uponthekeyreleasersthatalreadyexisted.
isnd species,andaresubjecttocontextualandepigeneticeffects(e.g.,
entnte seeDewsbury,1978;Eibl-Eibesfeldt,1975;Moltz,1965).Thus,a Blurring Proximate and Ultimate Levels of Analysis
documcleisi fTeiwnbedregceand’essstaicftkelrebMacckDFoAugPatlol,eSxplolatninicmka(t1e9rn6a7l)cadrierescetqlyueanpcpelsieidn In the complex organization of the phylogenetically old and new
Thissarti rfoixdeedntasn,dcahrieefrualrlcyhsipceacliafysiinngftihsaht,rboudteinntssteeqadueenmcepslowyoinugldanfoltebxeibales sfrtorumcttuhreesm.o.s.tpwreimpirteivseumseaxbulyalhfeaevleinngetuorathlelahdidgehresst.l.e.vefolrofasaclternudisitnigc
Thi responsethatwasorganizedbyfrontal,cingulate,andseptalareas. sentiments.(MacLean,1967,p.380)
Mammalian neural systems are inherently goal-directed and con-
text-sensitive; therefore, even behaviors assumed to be “innate” Researchers from different domains typically avoid conflict
arenotinflexibleornoncognitive;theyreflectanimplicitdecision by theorizing that research on evolution, adaptation, develop-
that maximizes the goal of helping, while integrating contextual ment, and proximate causes represent different but compatible
cuesandreflectingtheuniquedevelopmentalhistoryofthegiver. explanationsforbehavior(Mayr,1961;Tinbergen,1963).How-
Isaltruismanerror? TheinstantiationofanFAP,andinter- ever,bottom-upviewslikethisoneinherentlycutacrosslevels
individual variation in the mechanism for offspring care, would ofanalysis,whileavoidingjust-so-stories,byfocusinguponthe
naturally produce “accidental” acts of helping even if altruism evolutionandconservationofthenervoussystemitself(Finlay
were not specifically selected for. For example, one may instinc- &Darlington,1995;Krubitzer,1995),particularlyformamma-
tively reach for an unfamiliar toddler falling when the bus accel- lian systems that subserve attachment, bonding, and maternal
erates,andsomeindividualswouldbemoreorlesspronetodoso, care (Insel & Young, 2001). This biobehavioral continuity
for example, because they are less sensitive to novelty or more constrains the ability of the nervous system to produce novel
sensitivetoneed.Suchaccidentalgivingcouldalsobeextendedto human adaptations, requiring complex human behaviors to be
ORIGINSOFALTRUISM 1313
builtuponpreexistingsystems(Gould&Lewontin,1979).This connections from the amygdala to the MPOA/vBST and then to
approachefficientlyexplainsnotonlytypicalandclearlyadap- thedopaminergicventralstriatalsystem.
tivebehaviorsbutalsounusualpropensitiesthatoftenconfound Toextendoffspringcaretohumanaltruism,theneuralregions
researcherswhofailtolinkthebrainsandbehaviorsofhumans depictedintherodentoffspringcaresystem(placedontherightof
to those of other species. By analogy, only by knowing how Figure 6) have to be augmented with more recently expanded
imprinting is encoded at the proximate level could biologists domain-general cortical processes that are more prominent in
explainwhyGreylaggeeseimprintnotonlytotheirmothersbut humanemotionalandbehavioralsystems,permittingdevelopmen-
also to humans, stuffed animal heads, and even swinging balls tal and behavioral flexibility and control (on the left ofFigure 6;
(Blumberg, 2005; Lorenz, 1937). Only by understanding the addressedingreaterdetailbelow).Humandomain-generalreward-
proximate mechanism of offspring care across species can we baseddecisionprocessesparticularlyinvolveconnectionsbetween
explain why mammals adaptively retrieve not only their own the prefrontal cortex (PFC), amygdala, hypothalamus, anterior
endangered offspring but also complete strangers and pets in cingulatecortex(ACCanditssubgenualregion[sgACC]),andthe
extreme danger. The subsequent major section describes the ventral and dorsal striatum. Anterior cortico-limbic regions
y. proximate homology between human altruistic responding and (amygdala, OFC, sgACC) output to regions in the brainstem that
shers.broadl rsohdaerendt boeftfwspereinngthreemtr.ieval through 10 key features that are othrrgoaungizhethaeuPtoAnGom)aicndatnodremgiootnosrinouthtpeuhtyspotothathlaempuesrtihpahteoryrga(en.igz.e,
publinated offspring care motivation (MPOA) and generate the necessary
dmi adrenocorticaltoneforresponding(originatingintheparaventricu-
salliedisse The ProximateAHltormuioslmog:y10BeKtweyeeFnacOtoffrsspring Care and lMarPOnuAcleaucst)i.vaItnepmaroatloler-lm, eoftfievraetniotnparlojceicrctiuointss ifnromthethveeOntFraCl aanndd
ite
ofob dorsal striatum, potentiating the approach and retrieval of the
oneott Overview of the Rodent Neural System for Offspring distressedother.
orsn Retrieval Whiletheevidenceprovidedinthebelowreviewsupportsakey
i
ationand Muchresearchontheneuroscienceofoffspringcareexamines roleforanavoidanceandapproachopponencyinoffspringcare,as
cier
sous behavioral and neurobiological changes associated with a partic-
s
Aal ularly important behavior—the retrieval of distressed and sepa-
ologicalindividu rLaotendstenienw&boMrnosrr(eslel,e2r0e0v7ie;wNsuminanJ,.2B0.07B;eNckuemra&n &TaIynlsoerl,, 22000038;;
he Rosenblatt, 1992). In a typical pup retrieval scenario, the pup
ch
yt
Psof makesultrasonicvocalizationsafterbecomingseparatedfromhis
ne motherandsiblings,causingthemothertoleavethenest,retrieve
as
cu
erial thepup,andcarryhiminhermouthbacktothenestwherehecan
Amson resume nursing and huddling (Hofer, 2000; see Figure 1, left
eer panel). Efficient retrieval not only prevents offspring from being
hp
ythe locatedbypredatorsbutalsoensuresthatpupshaveaccesstothe
bt
yrightedolelyfor msMeonetsnhoderory’zsas,fto1imo9d9u,8la)at—siowanlell(olGfauwsbhheirecnrhiacaskrs,eoc1ci9rai8ttei1cd;alcHlfoordsreyp,broo2pd0ei0lry9d;ecvoMenltaoaspcotmnaen&ndt
ps
coed of the nervous system in altricial mammals (F. A. Champagne,
isnd Francis,Mar,&Meaney,2003;Meaney,2001).Thisarrangement
entnte bearsstrikingresemblancetoothersalientactsofaltruism,partic- Figure 5. Major elements of the offspring care system (OCS), as de-
documcleisi utolarrleytriienvaectaivdeiastnrdeshseerdo,icencdaasnesgewrehde,reaninddivvuidlnuearlasbrliesksttrhaenirgelirveins sticarlilbye,dfobryraNtsumthaant(h2a0v0e6n),oLtobnesetneipnriamndedMboyrrmelalte(2rn0a0l7)h,oarmndonoethse(rms.aIlnesi-,
Thissarti immediatedistressorneed. vspirogninseftehmataliessm),etdhieateodlfabcytothryeAcuHesNoafndnePoAnaGte.sOpnrcoedtuhceesyastwemithdhraaswbereen-
Thi Activeoffspringretrieval,incontrasttomorepassiveformsof primedbythehormonesofpregnancyandparturition,projectionsfromthe
care like nursing and licking, has been shown to rely upon inter- MPOA/vBST inhibit the avoidance system (AHN and PAG) while acti-
actions among the prefrontal cortex, amygdala, medial preoptic vating dopamine-dependent approach responses in the ventral striatum
region of the hypothalamus (MPOA), ventral bed of the stria (VTA,NAcc,VP).Notethatmanyofthesesubregionscanparticipatein
terminalis (vBST), and ventral striatum (e.g., Numan & Insel, bothexcitatoryandinhibitoryactions,andinbothapproachandwithdraw,
2003, see Figure 5, acronyms in Table 2). The system can be sothecurrentformulationisageneralizationbaseduponexistingempirical
ostensiblydividedintoavoidanceandapproachroutes,whichare workforthissystemperse(e.g.,inrodents,theamygdalaisinvolvedin
bothapproachandavoidance,e.g.,seeLai,Ramiro,Yu,&Johnston,2005;
selectedwhenprocessingapotentialtargetofneeddependingon
separablesubregionsofthePAGareinvolvedinbothpredatoravoidance
the experience and hormonal condition of the observing indivi-
and prey approach, Comoli, et al., 2012; and glutamatergic dopamine
dual. The avoidance route proceeds from activation of the
signalscanproducebothrewardandfearlikeresponses,Faure,Reynolds,
amygdalatotheanteriorhypothalamus(AHN)andperiaqueductal Richard, & Berridge, 2008). AHN (cid:1) anterior hypothalamus; PAG (cid:1)
gray(PAG;bottomofFigure5)whiletheapproachcircuitinhibits periaqueductalgray;MPOA(cid:1)medialpreopticareaofthehypothalamus;
the avoidance system through inhibitory connection from the vBST(cid:1)ventralbedofthestriaterminalis;VTA(cid:1)ventraltegmentalarea;
amygdala to the AHN while also motivating approach through NAcc(cid:1)nucleusaccumbens;VP(cid:1)ventralpallidum.
1314 PRESTON
Table2 retrievalandaltruisticresponding,asevidencefortheirhomology,
AcronymsforNeuroanatomicalRegions,Neuropeptides,and arepresentednext.
NeurotransmittersUsedThroughouttheText
Factor 1: Care Must Be Extended Beyond Parturient
Region Acronym
Mothers
Anteriorcingulatecortex ACC
Basolateralamygdala BLA For an offspring-care based model to be explanatory, the pro-
Dopamine DA posedmechanismsmustfirstexplainwhybothoffspringcareand
Dorsolateralprefrontalcortex DLPFC altruismareexhibitedbymalesandfemales,aswellasbymothers
Hippocampus HPP
andnonmothers.Indeed,offspringretrievalinrodentsisnotonly
Medialamygdala MeA
Medialprefrontalcortex mPFC possibleforparturientrodentfemalesbutalsooccursinmalesand
Medialpreopticareaofthehypothalamus MPOA virginfemalesundertherightconditions,andinmothersofother
Nucleusaccumbens NAcc species. For example, virgin female rats do eventually care for
y. Orbitalfrontalcortex OFC pupsleftintheirchamberfollowingahabituationprocessthatcan
shers.broadl OPPraxeryfartvooencnittnarliccuolratrexnucleus OPPFVTCN tsayksetemup(tRooasewnbeelakt,t,w1h9i6c7h).gMradaulealrlaytsdocawnnraelsgoulbateesintdhueceadvotiodacnacree
dpubliminated PPererifaroqnuteadluccotartlegxray PPFACG feosrtrapduiposlt(hRroousgenhbsliamttil&areCxepuoss,u1re99te8c;hRnioqsueensbloarttb,yHaapzpelliwcaotoiodn, o&f
itsallieedisse SVVueebnngttrreaanlluptaealgllmriedgeuniomtanloafretaheanteriorcingulatecortex sVVgPTAACC Pdioroecleti,n1g9p9a6t)e.rnEavlicdaernecienablispoarseunptaplosrtpsecthiees,rsoulechoafstthheeMCaPliOfoArniina
ofob Ventrolateralprefrontalcortex VLPFC mouse,inwhichlesionsoftheMPOAreduceparentalcareinboth
oneott Ventromedialprefrontalcortex VMPFC malesandfemales(Gubernick,Sengelaub,&Kurz,1993;Lee&
orsn Brown, 2002). In prairie voles, male parental behavior is associ-
i
ationand aFteerdriws,it&hthDeereVleraieses,o1f9v9a4so),prweshsiicnhinmtahyelbateertaieldseptotumthe(Zs.eWxuaanllgy,
sociuser well as the aforementioned neural regions in both offspring care dimorphic behavior of these males, who defend and protect the
Asal andhumanaltruism,therearenotableexceptionsinthesearchfor femaleandpupsfromintrudersinthenest(Insel,1997).
hologicaleindividu ctrheosemorlmoultoeinoonmfienthcehnaeMnuirPsomOimsA.agEinivnighduemmnceaetnhiosadlatsrlumtoisomdstadtceuoemantpdolettthheeelypflaoacoctrktishnpagattfinoaorl bheueMmnaudnlteimpplroeimnfseatarteatusterde(sIinnosfenltohn&erorodYdeoneuntsntgion,fcfl2sup0dr0ii1nn;ggsMchaeareeepsst,yribsptieiredmrsi,ha&anvdeZnaeolhsnro-,
ch
Psyoft onewaspreviouslylookingforthisregiontoparticipateinaltru- 1998).Forexample,themechanismsthatinducepostpartumclean-
ne ism. Some evidence even undermines the potential role for the ing and kin recognition in sheep are largely similar to those in
as
ericalu MPOA outside of species-specific behaviors involving perioral rodents, relying upon the combined influence of peripartum hor-
Amson sensorycues(Insel&Young,2001;Numan,Corodimas,Numan, mones,thebirthingprocess,andolfactorycuesfromtheoffspring
eer Factor, & Piers, 1988). However, the MPOA does play a more (Kendrick et al., 1997; R. Nowak, Keller, Val-Laillet, & Lévy,
hp
ythe generalroleinmaternalmotivation(e.g.,inbarpressingforpups; 2007).ThespecificrolefortheMPOAintheresponseofmothers
yrightedbolelyfort imfpoaartnaendineisuacrluotrssusciiiosetnniccseereeshpNaosunmdinianvnges,&tpigaIanrttsieecdlu,lta2hr0el0y3rf)oo,lresoacoitftivmtehaeryetssrtitreiilvaltapula.mrHtiuicn-- Mtpoaerstueircpiuaslsraaert,ley&dsoLtfrféosvnpygr,ini2gn0h0pa7rs;imaPliospoianrdboreuoesnn,(dfLeirémsvtoyn,tism&treaK)teedmllioentrh,seh2re0se0p(7P)a.enrPdrirniis-,
ps
coed altruism and cooperation, but the specific role of the ventral matesareoftenthoughttohaveareducedrequirementformaternal
isnd pallidum(VP)alsostillneedstobeconfirmedwithhigh-resolution hormones for infant care, because juveniles, nonmaternal female
entnte scanningtechniques. relatives,andmalesoftenprovidecareinthesespecies;however,
mi
docucleis extBenecsaivueseeovfidtehnecseecfaovretahtes,rtohleeroevfitehwebdeolpoawmfionceurgseicsonnucthleeums oarce- tihneterreesat,recasrteil,lpsriogtneicftiicoann,taenfdfectrtseatomfemntatoefrnianlfahnotsrm(oHnredsy,o2n00th9e;
hisarti cumbens,anteriorcingulate,andorbitofrontalcortexincaregiving Maestripieri,1999;Maestripieri&Zehr,1998).Forexample,even
This and altruism. Of course, terms like “caregiving system” do not though nonperinatal pigtail macaques care for infants, pregnant
T
implyafixedsystemutilizedsolelyforthiscontext.Indeed,most femalesincreaseinfantinterestandcarebeforegivingbirth;more-
ofthecaregivingsystemispartofalargersocialbehaviornetwork over,experimentalapplicationofestrogenincreasesinfantcontact
(Newman, 1999), which is also implicated across domains for (Maestripieri & Zehr, 1998). Even species that are highly diver-
decision-making and reward processes involving food, drugs, ro- gent from mammals, such as squid, crocodiles, clownfish, and
manticlove,andevenmaterialgoods(e.g.,seereviewinPreston, rattlesnakesdemonstratefunctionallysimilarbehaviorstoseques-
2011). It is assumed that the regions that participate in offspring terandprotectyoungfrompredatorsduringtheirmostvulnerable
retrieval also perform similar functions in other processes and developmentalstage,shortlyafterbirth(Hrdy,2009).Comparative
collaboratewheneveroneapproachesadesiredorvaluedtarget.In researchhasalsoshownthatevenhumansasyoungas14months
addition, the term “rewards” herein does not refer to concrete and chimpanzees will help another person in discrete need, for
compensation for good deeds but rather refers to the reinforcing example by picking up a dropped item (Warneken & Tomasello,
qualityofitemsthatmotivateonetoapproach,whichneednotbe 2009b). Even among Old World langur monkeys, females of all
subjectivelyexperienced(e.g.,seeBerridge&Winkielman,2003; ages—eventhosetooyoungtogivebirth—showanintenseinter-
Damasio, 2000). The factors that are shared between offspring est in newborns, responding to their calls and attempting to hold