A Role for Children in Hominid Evolution

8/7/2019 A Role for Children in Hominid Evolution

http://slidepdf.com/reader/full/a-role-for-children-in-hominid-evolution 1/31

A Role for Children in Hominid Evolution

Author(s): A. C. ZellerSource: Man, New Series, Vol. 22, No. 3 (Sep., 1987), pp. 528-557Published by: Royal Anthropological Institute of Great Britain and IrelandStable URL: http://www.jstor.org/stable/2802504 .

Accessed: 16/02/2011 19:41

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=rai. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

Royal Anthropological Institute of Great Britain and Ireland is collaborating with JSTOR to digitize, preserve

and extend access to Man.

http://www.jstor.org

http://www.jstor.org/action/showPublisher?publisherCode=rai

http://www.jstor.org/stable/2802504?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp



http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/stable/2802504?origin=JSTOR-pdf




A ROLE FOR WOMEN IN HOMINID EVOLUTION

A. C. ZELLER

UniversityofWaterloo

The proposalthatan increasein hominidfertilitycouldresultfromcontributionsmade byoffspringtotheirmother'ssubsistenceisexamined,usingthreelinesof evidence.The firstpoint

contraststhelow rateof reproductionin thecloselyrelatedgreatape taxawiththesituationexistingin modernhumansandwith theavailableevidenceforearlierhominids.The secondsourceof datais anexaminationof thefactorswhichaffectmodernhumanfertilityand howtheymighthavebeeninfluencedby the level of foodsupplyin thepast. The thirdapproachis anassessmentof the typesof contributionsactuallymade by childrenin hunter/gathererandhunter/horticulturalgroupsat thepresenttime.A surveyof fourteengroupsat thislevelrevealsthatintwelvecases childrencontributeto a widerangeofactivitiesthatreducesubsistencestresson theirmothers.In thetwogroupsinwhichthelevelofcontributionis low, thereproductiveratesarealso decreased.Evidencefromthesethreesourcessuggeststhatbothdirectand indirectcontributionsbychildrentotheirownandtheirmother'ssubsistencebasemayhavecontributedto thesuccessof humansinpopulatingtheworld.

Introduction

In the historyof researchon humanevolutiona numberof factorshave beenproposedas thecausal mechanismsunderlyingthehominisationprocess.Sincetool use was long consideredto be a 'marker'ofhumanity,themanufactureoftoolswas an earlycandidateas theinstigatingdevelopment(WashburnI959).This materialevidence was quicklylinkedwiththe patternof huntingas asubsistencebasewhichwould differentiatehominidsfromtheotherprimates

(Washburn& LancasterI968; LaughlinI968a; WashburnI98I). The develop-mentoflinguisticbehaviourwasseenas a corollaryoftheco-operationrequiredforhuntingin socialgroups(Hockett& AscherI964).

Morerecentthoughtsare based onthereinterpretationofdataobtainedfromhuntingandgatheringmodelswhichsuggestthatgatheringisthemajorsourceofdailyfoodinthissubsistencepatternat thepresentandmay easilyhave beenso duringearlystagesofthissubsistencestrategy(Lee I972; Kolata I974; TanakaI976; TannerI98I). This suggestionshiftsthefocusof attentionfrommalesastheprimesubjectsofevolutionarypressurestofemalesas at leastequal subjectsand sourcesofadaptivedevelopments(IronsI979). Thepotentialimportanceof

womenintheevolutionaryprocesswas recognisedbyDarwin (I872) butonlyrecentlyhas thisrecognitionbeen utilisedin constructingmodelsof hominiddevelopment(Lancaster& LancasterI98 3;Lancaster& King I985; TannerI98 I;Zihlman I98I).

Lovejoy's (I980) interpretationoftheavailabledatagoesfurtherinproposing

Mat, (N.S.) 22, 528-57



A. C. ZELLER 529

thattheproductionof childrenmayalso have hada roleto playin thecourseofhuman development.He argues (Johanson& Edey I98i) thatthe birthrate

increasedas earlyprehominidsshiftedtheirreproductivestrategyto a lessextremeversionof the 'K' selectionthatcharacterisespongids. In Lovejoy'spresentation,a morefrequentproductionofoffspringincreasedtherequirementfora concentrationof attentionandenergyon motheringbythefemales.Themales then developed a gatheringas opposed to foragingadaptationwithsubsequenteffectson matingpracticesand thedevelopmentof thebipedallocomotorpatternsincethemaleneededtocarrysufficientfoodforhismateandoffspring.Lovejoy'sfocus,however,isonthebehaviouraland structuraleffectson theparentsofbearingand nurturinganincreasednumberofoffspring,ratherthanon how largernumbersofchildrenmightactuallyaffectthe subsistencebase.

In order to begin the divergencefroman ape to a human patternofexploitation,some change,whichmayhave beenminorin itself,musthaveoccurred.Lovejoy's proposal of an increasedbirthrateis thetypeof minorchangewhichcould have majorramificationsintermsof thefuturesuccessofaspecies.An increasedbirthratewill reducetheage at whichtheyoung aresupplantedas a primaryfocusofmaternalcare.This differenceisemphasisedbya comparisonofapeand humanchildrearingpatterns.One notabledifferenceisthe scope of activityundertakenby youngoffspring.Ape infantsare totally

dependentphysicallyandemotionallyforfivetosixyears,whereasby age 3to4humanchildrenare beginningsome typesof usefulactivity(WhitingI963;

Munroeetal. I984). It isthepurposeofthisarticleto substantiatethisstatementand to arguethatthetypesof activitythathumanchildrenundertakecouldsignificantlyincreasetheirmother'sreproductivepotential.Thus, not onlycould anincreasednumberof childrenaffectthepatternsof adultbehaviourasLovejoy suggests,butitcouldalso actin apositivefeedbackcycleto increasethenumberof offspringper female,and eventuallytheoverall speciesnumbers.This wouldoccurifchildrenwere ableto contributeenough,eitherdirectlyorindirectly,to thesubsistencebasetoaffectmaternalfertilityratesby shortening

theinterbirthinterval,andincreasingoffspringsurvivability.The interactionofmanyfactorsissubsumedinthisexplanatoryfocuson oneaspectofthewhole. Evolutionworksinsuchawaythateachchangemustbe anintegrativephenomenonforsuccessfuladaptationto occur. Changestowardsfurlessness,bipedalism,language use, tool fabrication,and social groupingwereallpartofthepatternleadingto hominisationJohnsonI978). Neverthe-less it seems reasonableto propose that the continuedfeedbackinteractionbetweenbiologicalandculturalfactorswouldincreasinglypromotethesuccessof vitalhominidadaptations.Increasednumbersof offspring,coupledwithimprovingculturaladaptations,would allow hominidsto expandtheirranges

intonewregionsand eventuallypopulatetheworld.Three typesof datawill be utilisedto examinethisproposal.The firstis a

comparativestudyofreportedapeand humaninterbirthintervalsandreproduc-tivesuccess,in order to substantiatetheclaimthathumansreproducemoreeffectively.Extrapolationsto thesituationamong Australopithecinescan bemade from primatemodels (Tanner I98I; Teleki et al. I976) and from



530 A. C. ZELLER

archaeologicalandpalaeo-demographicevidence(Mann I975; Howell 1976a).The second pointto be discussedis a briefsurveyof the factorsinfluencing

currentnaturalhuman fertilityand how thesefactorsmighthave affectedfertilityratesin the past. The thirdsource of data concernsthe contributiveactivitiesof childrenin modernhunter/gathererand limitedhorticulturalsocieties.The publishedreferencesto childhood contributiveactivityhaveprimarilydealtwithpeasantand pastoralsocieties(AinsworthI985; DumondI975; Munroeet al. I984; Nag etal. I978; Hassan I980; WhiteI975; WhitingI963). However, publisheddata on economiccontributionsby childrenarerelativelyrare for horticulturaland hunting-gatheringsocieties. McDowell(I98I) compiled materialon the value of childrenin Papua New Guineahorticulturalsocieties,while childhoodactivitieshave been noted amongthe

hunter-gatherer!Kungalthoughtheircontributionsto thesubsistencebase arenot emphasised,forreasons which will be discussedbelow (Draper I976;

ShostakI976).

The quantityand scope ofthedatacollectedforthisstudywereonlyintendedto providean initialidea of thepotentialforthisresearchapproach.A muchlargerand moresystematicsurveywouldbe necessaryto confirmordenythetrendswhich are suggestedhere. I conductedinterviewswith eleven fieldresearchersstudyingthirteenhunter/gathererand limitedhorticulturalgroupsfromnorthAmerica,Africa,and Australia/PapuaNew Guinea. They wereasked to provideinformationaboutchildren'scontributionsto eightareas ofpotentialactivity,andtheagesat whichtheseactivitiesbegan.The materialonthe!Kungas a fourteenthgroupwas takenfrompublicationsby Draper (I976)

and Shostak(I976). Informationderivedfrombirthratedata, fertilityfactorsand childhoodactivitiesis used to examinetheargumentthattheactivitiesofAustralopithecinechildrencould have beena major contributingfactorto thesuccessof the hominidadaptationby increasingtheirmothers'reproductivepotential.This capacityforpopulationgrowthmay nothave beenofa scaletocausea rapidincreaseinnumbers(BirdsellI968; DeeveyI968; DumondI975;

Kennedy1975; WashburnI98I) butitwouldallowgroupsto recoverquickly

frompopulationbottleneckscaused byenvironmentalhazards,andeventuallyto increasetherangeofhominidoccupation(Lovejoy I980; Johanson& EdeyI98I). Thiswould contrastwithape populationgrowthwhichis estimatedtohaveremainedstablethroughoutthePleistocene(Telekietal. I976).

Ape reproductivepatternsDue to thelong reproductivespan in apes, longitudinalinformationon theinterbirthintervalandcompletedfertilityis stillbeing collected.Thereforethe

relativelengthsofhuman and ape interbirthintervalsare stillunderdispute.Lovejoy (I980) contendsthathumansin a stateof 'naturalfertility'producemoreoffspringand have a shorterinterbirthintervalthanapes. Short(I976;

I984) suggeststhathumansarecharacterisedbythelongestmammalianinter-birthintervals.Othersconsiderthattheinterbirthintervalinhumansisequival-enttothatinapes (Lancaster& LancasterI983; Lancaster& King I985; Sussman



A. C. ZELLER 531

I972; Tanner I98I; WashburnI98I). However, Teleki et al. (I976), Tutin(I980), Tutin & McGuinis (I98I), Harcourt et al (I980; i98ia; I98Ib) and

Galdikas(I978; I979; I98I), among others,havepresenteddatawhichindicatethatthebirthintervalinallthreeape speciesislongerthantheforty-fourmonthsfornomadic!Kung (Howell I979; Lee I979) who arereportedas havingthelongest natural human interbirthinterval(Howell I976a;b; I979; Huss-AshmoreI980; Lee I979). Partofthisdifferenceofopinionmaybe based on theuse of data fromcaptiveanimalswho havea markedlymodifiedreproductivepatternwhichis influencedby factorswhichwill be discussedbelow.

From her studiesofwild apes, Galdikas(I978; I979; I98I) commentsthatbirthintervalsprobablyexceed fiveyearsinorangutans(Pongo). Mackinnon,who also studiedfree-rangingorangutans,statesthatthemeanbirthintervalis5.5 yearsor 66 months(I979). These are minimumperiodswithsurvivinginfantsand maybelengthenedto sevenor eightyearsifecologicaldisruptionsornon-viablebirthsintervene(Galdikas,personalcommunication).Age at firstparturitionis about twelveyears,althoughfemalesoftenundergoa periodofadolescentsterilityforseveralyearsbeforetheygivebirth.The totalnumberofviableoffspringpossibleina forty-yearlifespanwithtwenty-eightfertileyearsand a 5.5 yearinterbirthintervalwould be about four with two possiblenon-viablebirths.This wouldbea veryhighrateofsuccessbutactualcompletefertilitiesarenotknownforfreerangingorangutans(tablei).

More extensiveinformationis availableon gorillassincetheyliveingroupsand can be studiedin largernumbers.Fossey (I979) has publishedknowninterbirthintervalsforeightcases,whichrangefromfortyto fiftymonthsinthemountaingorilla.However, the numbersof infantsperfemalesurvivingtothreeyearsinherfourstudygroupsrangedfrom.7 to I. 57infants/femaleperten-yearperiod (Fossey I983). Therefore,loss and replacementof unweanedinfantswould shortentheintervalwhen compared to contiguoussurvivingoffspring.

Anothersurvey,byHarcourtetal. (I98 Ia), reportedthatFossey'sstudysite,theKarisokeresearcharea of theVirungahabitat,was probablyone of thebest

areasforgorillasurvivalintermsoffoodandhuntingpressure,at leastupuntilI978. Theircollaborativestudyforthetwelve-yearperiodI967-I979 foundthattheaverageproductionofsurvivingoffspringover the wholeVirungaregionwas about one infantper eight yearsfora total of threeoffspringraised toreproductivematurityperfemale(Harcourtetal. I980). Femalegorillasbeginto breedat age io-ii, a littleearlierthan otherapes, but this is offsetby a

TABLE I. Reproductivepatternsofhigherprimates.

InterFirst Life Fertile birth Potential Number

Genus birth span span (months) fertility raised

Pongo I2 40 28 66 5?Gorilla II 35 24 53 S? 3Pan I3 38 25 68 5 2Australopithecus

(reconstructed) I 5 40 25 48? 6? 4?Homo (Dobe!Kung) I9. 5 6o+ 25 44 7 5



532 A. C. ZELLER

shorterlifespanaveragingthirty-fiveyears, which provides approximatelytwenty-fouryearsofpotentialreproduction(tablei). Sincegorillasdo notshow

obvious sexual swellings,thelengthof gestation,whichaverages225 days,or81/2months(Tutin& McGuinisI98 I), mustbe usedto calculateprobabletimeofconception.It appearsthatabout 2-4 sexual cyclesoccur betweentheend oflactationalamenorrheaand refertilisation(Harcourtet al. I980). Theirrepro-ductivetimespanis maximisedbyrapidreplacementofinfantswho die and asomewhatshorterinterbirthintervalthanoccursinchimpanzees.The mediantimespansinvolvedaretwelvemonthsto replacea dead infantandfifty-threemonthsinterbirthintervalforviableoffspring(Harcourtet al. I980; i98 ib). Inorderforpopulationtomaintaina stablelevel atleast40 percent.of theanimalsshould be immatures(Harcourtetal. i9i8a). Male infanticideand deathsofmotherscontributetoinfantmortality,thuskeepingtheoverallnetpopulationincreasedown to .02 to .6 percent. per yearin thispopulation,in spiteof anadequatebirth-rate(see table2, Karisoke).

Whenthegorillaand wildchimpanzee(Pan)situationsarecomparedtherearevariationsin thetimespansinvolvedforreproductiveactivitiesbutverylittleinthe overall rateof populationgrowth.The twentyyear periodsurveyedatGombe (Goodall I979) indicatesthatinterbirthintervalsforsurvivingcon-tiguousoffspringrangefromfiveto tenyears.Tutin (I980) andTeleki et al.(I976) have examinedthereproductivepatternsof knownfemalesatGombein

an effortto assess total fertility.Wild chimpanzeesbegin to show sexualswellingsataboutages 8-9 withmenarcheoccurringI-3 yearslater.Aperiodofadolescentsterilityfollowsduringwhichyoungfemalesmatefrequentlybutdonot conceive.Thus mostfemalesdo not beginto reproduceuntilthey arethirteenyearsold, andnearlyhalflose theirfirstinfant(Tutin I980). The timerequiredto replacethelost infantis slightlylongerthanthetwelve monthsingorillas,averagingI2 to i6 months.BetweenI964 and I978 themeaninter-sibintervalforsurvivingyoungwas 68 monthswitha rangefrom53 to79months,althoughintwocasestherewereinterveningpregnancylosses. The numberofcyclesbetweenresumptionof sexual activityand thesubsequentpregnancy

rangesfromi to i i cycleswithameanof4.5. Itwasnoted,however,thatittooklonger(6.3 cyclesmean)forfemalestobecome pregnantthethirdtime.Tutinsuggeststhatprobablyearlypost partumcyclesareanovulatoryas is thecase inhumans(ShortI984). Therefore,witha probablelifespanof thirty-eightyears(Telekietal. I976; WashburnI98I) andmenarcheatageI3, chimpanzeefemaleswould have a reproductivelifespanof about twenty-fiveyears. With an

TABLE 2. InfantproductionperI000 femalesperyear.

no.of StatusofGenus Location infants population Source

Gorilla Calculated 125-148 Maintain Harcourt et al. Ig8iarequirement

Gorilla Karisoke 286.5 Increase Harcourtet al. I98 iaPan Gombe I64.3 Maintain Goodall I979Pan Kibale 123-I66 Maintain Goodall I979Homo Dobe Kung 128.o6 Increase Howell 1976b

(6 yearmean)



A. C. ZELLER 533

interbirthintervalof 68 months(Teleki et al. I976; Tutin ig80) this wouldprovidetimefor4-5 birthsto survivetoweaning(tablei). However,foreight

femalesobservedoverthelong term,theestimatednumberofoffspringrangedfromi to 5 withamediannumberof3 (TutinI980; Tutin& McGuinisI98 I). Ofthosefemaleswithknownoffspring,themaximumnumberto reachreproduc-tiveage is three(Tutin& McGuinisi98i), and themediannumberis 2 (Tutinig80), althoughit is possiblethatolderfemaleoffspringmayhavemigratedtootherpopulations.Tutin (ig80) concludesthat thereplacementratein theGombe populationonlymatchesthedeathrate,anddoesnotexceedit.

The situationis verysimilarin Nishida's chimpanzeepopulationin theMahale mountains.Over fourteenyearsofdatacollectiontheinterbirthintervaland speed of offspringreplacementare about thesameas thosedocumentedforGombe (Tutinig80). Two otherchimpanzeepopulationsstudiedby Ghiglieri(i984) in theKibale foresthad crudebirthratesequivalentto the estimatesofHarcourtetal. (i98 ia) forpopulationreplacement(table2).

Thus it isclearthatproductionofreproductivelyviable offspringby femaleapesisa veryslowundertaking(tablei). Althoughineachspeciestheinterbirthintervalssuggestthatabout fiveviable offspringshouldbe produced,inactualcountsthemaximumactuallyraisedisusuallythree,withamedianoftwo. Thislow rateimpedespopulationgrowthbothintrinsicallyand also becausetruerateofgrowthis basedontheproductionofdaughterstocontinuethespecies(gross

rateof reproduction,Howell I979). Producingtwo or threeviable sons willkeep up anindividual'srepresentationin thegene pool of thenextgeneration,but will not ensurethattheyreproduceor thatthepopulationwill continue.Anotherconstrainingfactorinape reproductionis theobservationthatinfantsbornwithinsixyearsoftheendofthemother'slifewillprobablynotsurviveher(Goodall I973;Teleki etal. I976). Orphanedapesmustbe atleastthreeyearsoldto survivetheloss of theirmother'smilk,butinmostcases,inspiteof carebyothers,themother'sdeathwillusuallydoomoffspringundersix todeathfromdepressionandneglect(Goodall I973; Teleki etal. I976). On a populationlevelalthoughsome individualfemalesmayhaveverysuccessfulreproductivelives,

someothersmaybesterileor nevermanageto raisean infant(Goodall I979).

ExtrapolationtoAustralopithecines

Whencomparingmodernapesto earlyhominidsitis clearthatsometypesofinformationwill be unobtainable,some will be based on inferencesand only asmallproportioncan be foundedon actual data. The lengthof the interbirthintervalforAustralopithecinesis probablyundiscoverablealthoughit is esti-mated by McKinley (in Dumond I975) as 3-5 years. The length of the

reproductivelifespancanonlybe assessedby drawingon modernhumansandapes as models.We do, however,have some data on overalllifespanwhichisanotherlimitingfactorin determiningoverallfertility.Howell (I976a) com-mentsthatprehistorichuntingandgatheringpopulationsprobablyfellintotherangeofa 20- to 30-yearlifespan.Data fromNeandertalburialssuggestthat5opercent.of thepopulationdied as infantsorjuveniles,butthatover 30 percent.



534 A. C. ZELLER

reachedthe age of3o; althoughonly a few(3 percent.) were consideredto beover 40 at death(Kennedy I976; WashburnI98I). Allan Mann arguesthat

Australopithecinelifespansmayhave been shorterthan this and suggestsanaveragelifeexpectancyofeighteenyears(Mann I975). This figurecan beputinperspective,however,byrealisingthattheaveragelifeexpectancyofGombechimpanzeesis I3.5 years(Telekietal. I976). Once theyhavesurvivedtoage I5bothchimpanzeesandAustralopithecinesareassesseda lifeexpectancyof27.7

years (Hassan I980; Teleki etal. I976). This is supportedby Mann's (I975)figuresfora combinedsample of thirtygracileAustralopithecinesin whichsevenwere underage i8, andtwenty-threeover age i8, withthreebeingoverage 35. He assesses the maximumlifespanas 40 years ?5, which is verycomparabletochimpanzees.

Although!Kungsubsistencepatternshaveoftenbeenusedasa modernhunterand gatherermodel on whichto baseinterpretationsofearlyhominidlifestyles,itis notthecasethattheyhaveadaptedto theircurrenthabitatoverthousandsofyears(Gordon I984; RipleyI980). Nonethelessextrapolationsare oftenmadefrom!KungtoAustralopithecinedata. The medianage atlastbirthfor!Kungwomenisabout32.5 yearsbuttheaverageageat firstbirthis I9.5 years(HowellI979). Thus, proportionately,the timeintervalbetweenestimatedage atfirstbirthinAustralopithecus(age I5) andprobablelifeexpectancy(aboutage 28) isquitesimilarto the fertileintervalof thirteenyearsthatoccursstatisticallyfor

!Kungfemales.Thesefiguresarebasedon therelativefertilelifeexpectancyafterthe attainmentofreproductiveage, as suggestedbyLancaster& King (I985).The nomadic!Kunghave an averageof4.7 to 5 livebirthsper womanwithamodal44monthinterbirthintervalmaintainedbylactationandpostpartumsextaboos(HarpendingI976; Howell I976b; I979; SaucierI972). Witha pre-reproductivemortalityrateof 34 per cent. (Howell I976b) to 40 per cent.(HarpendingI976) thisreproductivepatternprovidesa growthrateof0.5 percent./yearforthepopulation.

Anotherfactorwhichmay account forsome of thepopulationgrowthisincreasedsurvivability.Telekietal. (I976) estimatetheprobabilitiesofsurvi-

vorshipin chimpanzeesas .364 andinAustralopithecinesas .65o. Lancaster&Lancaster(I983) accountforthisdifferencebyfoodstressattheweaningstage.Only one in threechimpanzeessurvive thejuvenile stage,but the numberdoubles whenadequateprovisioningoccurs.Thus,insteadofraisinga mediannumberof two offspringto reproductivematurityout of 5-6 as chimpanzeemothersdo, Australopithecinesmayhave beenable to raise4out of6produced.

Mann (I975) and Washburn(I98I) statethattheAustralopithecinepatternoftootheruptionandenamelformationindicatea slowerdevelopmentalprocessthan occursforthe chimpanzee.Argumentsby Bromage and Dean (I985)suggest,however,thatAustralopithecusand earlyHomo demonstrate'bio-

logicalequivalenceto modernmanatroughlytwo-thirdsthechronologicalage'(Bromage& Dean I985). Mann'sargumentproposesthatsexualmaturitymayhavebeendelayeduntilage I5 inAustralopithecus,whereastheBromageandDean approachsuggestsa moreape-likepatternof I2 to I3 yearsfortheadventofreproductivebehaviourinfemales.Actualage at firstbirthwould havelessinfluenceon number of offspringproducedthanwould interbirthinterval



A. C. ZELLER 535

(Wood et al. I985). If Australopithecineshad the same birth intervalsaschimpanzees,approximatelythe same age at menarcheand an equivalent

lifespan,populationgrowthwould have beenat aboutthesame rateas occursinchimpanzees.If insteadof a 5-6 yearinterbirthinterval,however,thebirthspacing occurredon a moremodernhumanpatternof 3 to 4 yearsas Mann(I975) suggests,therateofoffspringproductionwould increaseby50to ioo percent.Even ifPleistocenepopulationsonly increasedat the rateof 0.0007 too.ooi5 per cent.per annum(Dumond I975) froma startingpopulationofIOO,OOO individuals,thenumberswoulddoublein IO,OOO years.Cohen(I98o)andHassan (I980) estimatetherateof Pleistocenepopulationgrowthat .OOI to.003 per cent. perannumwhichwould providean even fastergrowthrate. Itseems doubtfulthatearlyhominidpopulationsgrew even thatquicklybut itseemsclearthata constant,althoughslow, rateofgrowthdistinguishestheearlyhumandemographicpatternfromtheape pattern(BirdsellI968; Dumond I975;

Telekietal. I976). Thistrendtowardsgrowthis importantin allowinghominidpopulationsto recoverfromperiodicdisastersand to spreadinto unoccupiedareas.

Factorsaffectingfertility

Captivityandtheruralurbandichotomy.Captivity,with the resultantincreasedfoodsupplyandlessenergeticactivitypatterns,appearstodecreasetheinterbirthintervalquite markedlyin apes. Supportiveevidence for this includes thenumberof monthselapsingbetweenparturitionandresumptionof maximumsexual swelling.In a (small) captivesampleof chimpanzeesthis timeperiodrangedfrom3to26 months,whereasinwildchimpanzeesitrangedfromi i to8i monthswitha meanof42 months(TutinI980). Tutin commentsthatwildfemaleswhose infantsdie usuallyconceiveagainfromfourto eightmonthslater,which gives a strongindicationthatlactationand the mechanicsofsucklingaremajorsuppressivefactorsintheresumptionoffertilecycling(Tutin

I980). Anotherfactoraffectingthetotalnumbersofoffspringproducedisageatmenarche.Captive chimpanzeesreach menarcheat a mean age of 7 years,whereasforfreeranginganimalsfirstsmallswellingsappearedat ages 8 to 9yearsin femaleswith menarcheoccurringfromI-3 yearslater(Tutin I980).

This disparityin age may be closelyrelatedto the greaterweightsforage incaptiveanimals(Telekietal. I976). In thissamplethemeanweightforcaptivenon-obeseadultfemaleswas 56.8 ? 2 kg.,whereaswild females(usuallywithclingingoffspring)rangedfrom32.3 to 37 kg. (Tutin I980). Since the meanweightat menarcheforcaptivefemaleswas 29. I ? 4.2 kg. it is clearthatwildfemalesdo notgainsubstantiallymoreweightthanthelevelrequiredtoachieve

fertility.It is quite possiblethatwild chimpanzeesmustweigh the same ascaptiveonesto reachmenarche,buttakelongertogaintheweight(Teleki etal.I976). Allthesefactorssuggestthatsituationsprovidingimproveddietintermsofquantityandconsistency,possiblycoupledwithlessactivelives,canhaveaneffecton thepotentialnumbersofoffspringborn.

Among humansthereis a dichotomyin the timingof reproductiveactivity



536 A. C. ZELLER

betweenruraland urbanwomen even when nutritionlevels are presumedadequateinbothcases. Urbangirlsbegintocycleearlierthanrurallivingones

(ShortI976). A Polish studycomparingtheageatmenarcheof ruralgirlswiththosewho livedinWarsawindicatedthatthecitygirlsbegantocycleI.03 yearsearlieron averagethanthose fromruralregions(Laska-MierzejewskaI975).Age at menarchehas also declinedduringthiscentury.In a Frenchstudythemeanage ofvillagegirlswas I5.7 yearsin i900, andI3.4 yearsin I946. Theseageswere stillinexcessofthemeanagesatmenarcheinPariswhichfellfromI3.5 to I2.9 yearsoverthesametimespan(LeridonI977).

Among Rwandan women, urbanlivingis correlatedwith a twelve-monthreductionin theinterbirthintervalfornursingmothers,when comparedwithrurallivingwomen(Bonteetal. I974; seealsoBongaartsI980; BuchananI975;Van GinnekenI977; Huss-AshmoreI980; ShortI976). Thisdifferencemayhave manyunderlyinginfluencessuch as socioeconomicfactors,activitylevelsandbreastfeedingpatterns.BothLee(I979; I980) andShort(I976) notethatinthecasesof!Kungwhohavebecomemoresedentaryandhaveaccessto milkandgrainas dietarysupplements,periodsof amenorrheaand birthintervalsarereduced30 to 50 per cent. Modern westernwomen have mean interbirthintervalsof 24 months,while ruralPunjab women average 30.5 monthsinterbirthinterval(LeridonI977). Thus thereis a slight,thoughmeasurable,differencebetweentheproportionsof theirlivesthatwomeninruralandurban

situationsareat riskofbecomingpregnant.For casesinwhichdietarydiffer-encesarenegligible,atleastpartofthisdifferenceprobablyarisesfromincreasedlevels of manual labour in ruralwomen (Laska-MierzejewskaI975). Thesedifferencesofreducedageat menarcheand shorterbirthintervalsareinthesamedirectionas thedifferencesbetweenwild andcaptiveapes.Itseemsreasonabletosuggestthatbothdietandactivitypatternsarefactorswhichcould affecttherateofoffspringproduction.

Diet.IftheinterbirthintervalsinAustralopithecineswerefiveor moreyearsthereproductivespanwould only allow the same numberof birthsas occur in

moderngorillasandchimpanzees.This is barelysufficientto maintaina stablepopulation,as indicatedabove. The questionthenbecomes,whatfactorscouldaffectthefertilityratesofearlyhominids,andpossiblyalso reducemortalitylevels.GenerallyspeakingBronson'sstatement'Diet is thecore environmentalconstrainton a mammal'sreproduction'(I980: i i6) seemsto be thefoundationoffertilityeffects.The caloriesconsumedmustmeettherequirementsforbodymaintenance,thermoregulation,foraginglocomotion,growthand reproduc-tion.Intheabsenceofsufficientcalories,growthandreproductiveeffectivenessare thefirstprocessesadverselyaffected(BronsonI980).

The abilitya woman has to buildup thenutritionalreservesnecessaryto

conceive,carrya pregnancyto termandlactateuntilthechildcan eat sufficientamounts of food to maintainitselfis based on threefactors.These are theavailabilityoffood,theenergya femalespendsacquiringitandcaringforheroffspring,and thespeedwith which theyoungcan transfertheirdependenceawayfromthemotheras theirmainsourceof food.

The availabilityof foodintermsof calorielevelis largelya factorof external



A. C. ZELLER 537

conditionsandsubsistencetechnique.The typeoffoodmayalso influencetheoutcomeofa pregnancy,sincelack ofproteinand mineralsinthedietmaydraw

upon themother'sbodyintheformationofthefoetuswithresultingpoorhealthformothersandnewborns(BuchananI975; CohenI980; Huss-AshmoreI980;Katz I972; ScrimshawI983). Ifthe level ofproteinintakeis low theage ofpubertymay be retarded(ShortI976). Body weightdoes notappearto affectfertilitysignificantlyaslongas thecriticalmass(seebelow)isattained,butlowermaternalweightincreasesboththelengthof thelactationperiodand thelengthoflactationalamenorrhea(Premaetal. I979 in Knauer I984; BongaartsI980;FrischI977; vanGinnekenI972; Huss-AshmoreI980; WrayI977). Frisch(I977) also commentsthatlow maternalweighttendsto resultin a low birthweightfortheoffspringwhichreducesitssurvivability.

The numberofavailablekcalrequiredtosustainpregnancyand threemonthslactationareestimatedbyFrischandMcAndrew(I974; FrischI975; I977) asI44,000. Ofthisamount,Buchanansuggests(I975) that55,oookcalarerequiredto maintainthepregnancy.Althoughlactatingmammalsareextrememodelsofefficiencyin the use of nutrients(Misra I979), thelevelof caloricsupportforhumanlactationis estimatedat I,000 kcal/dayorgo,oookcalforthreemonths(CohenI980; FrischI975). Insituationsofrestricteddiet,calorieaccumulationcan be slow andfertilityratesareaffected.Lee's (I980) input-outputanalysisof!Kungdietindicatesthatadultsconsume2I40 kcal/daywhileexpendingI,975kcalto acquirefoodand maintainthemselves.At thislevel,calorieaccumulationtothelevelthatFrischsuggestswould takeseveralyears.SteinandSusser(I977)investigatedthe reproductiveresultsof a five-monthfamineduringwhichcaloricintakedroppedfromI 500 to700 kcal/dayanddiscoveredthatoffspringproductionwas severelycurtailed.Undertheseconditionsit wasmainlyyoungmothersbearingtheirfirstchildwho had thenecessaryreservesto continueprocreation.In light of the calorielevels considerednecessaryBongaarts's(I980) studyon thedistinctionbetween'low' (I300 kcal/day)and 'high' (I630kcal/day)levelsofmaternalcalorieintakeprobablyonlysampledwomen atthelow endofthescale.Thushisresultthatcalorieintakeonlyaffectedthelengthof

lactationby one month,maynotbe valid forhigherlevels ofnutritionsinceShort(I976) statedthateven 2I00 kcal/dayand 42 gmsofproteinwere notsufficienttomaintaina positivenitrogenbalanceinthemother.

Frisch(I975; I977; I978; I982) appearsto have established(Cohen I980) thatthereis a criticalweightof44to48kilos,andproportionofbodyfat(I7-22 percent.)which is necessaryto allow a developingfemaleto undergomenarche.She hasalsoarguedthatthisstagecoincideswitha changein metabolicratecostfrom35 cal/kg/dayto 28 cal/kg/dayanda changeinproportionof leanbodyweightto fatfrom5:I to3: I (I97 5). Althoughcriticalweightseemsestablished,thereis someargumentconcerningherassertionthatthelean:fatratiomustbe

regainedbeforereconceptioncanoccur(Bongaartsig80;Jain& SunI972;Jain&BongaartsI98I; RipleyI980; ShortI984; Lee I972; I980). Frisch(I975; I978)

has suggestedthatthe mechanismbywhichfataffectsmenstrualcyclicityandtheabilitytoconceivemaylieintheinteractionofgonadalactivityandestrogenlevelswiththelevel of fatinthebody.Protein,calorieandvitamindeficitsarecorrelatedwithgonadal hypofunction(Katz I972). Estrogenmetabolismmay



538 A. C. ZELLER

be influencedby factorsinvolvedin fatmetabolismand thus moreabundantwhena highfatdietis consumed(FrishI975; I978). Thereforeratherthana

simple storage of calories,fatmay providethe basis forvaryinglevels ofreproductivehormones.

Cohen (I98o) hassuggestedthatthelean:fatratiomaybe irrelevantin moderndietsbecausethelevelofcaloricintakeand thusthelean:fatratiomaynolongerprovidea reliableguideto thebody, indicatingan environmentsuitableforreproduction.He arguesthatearlyinhominiddevelopmentthelevelof caloricintake,andthusthebody'slevel of storedfat,was also an indicatorofproteinandothernecessarynutrientsinthediet.Consumingcurrentadultdietsloadedwith non-nutritivecaloriesmayallow modernwomen to conceiveand bearyounginenvironmentswhichdo notreallyhaveenough proteinresourcesto

sustain them.One example of thisis the increasein the fertilityrate to 8offspring(Neel I983) in the Yanomamo sincetheybeganthe cultivationandconsumptionof bananaswhich have only I.2 gms of proteinper ioo gms(Cohen I980). Thuscaloriclevelsmayallowreproductionto occurinsituationsinwhichproteinresourcesareinsufficientto maintainthehealthofthechild.Ontheotherhand,therearealso somepopulationsinwhichthewomenweigh35 to40 kg. and are able tomaintainpregnancyandlactationon a dietof I500-I 6oocalories/day,with35 to40 gm.ofprotein.Adaptationmechanismsinpregnantwomen seem toallow anextremelyefficientutilisationofprotein,calciumand

iron (BeatonI96I in Misra I979).

During pregnancysome nutrients

appearto

be storedinthebodyand releasedduringlactation(Misra I979). These factorsmay contributeto theoccurrenceof pregnanciesand continuedlactationinwomenwho do notappearto have anadequatecalorieintake.However,underthese conditionsstillbirthsand low birthweightchildrenoccur 'with muchgreaterfrequency'(p. I83) thaninwell-nourishedwomen(Misra I979).

If infantsare born to a poorlynourishedmothermany proteinand fatresourcesare drawnfromher body. In situationsin which maternalcaloricintakeis below 2,000 cal/day,thevolume ofmilkdrops tolevelsof400 to600c. c. perdayratherthantheusual8ooC.c. (WrayI977). Inonestudy,Edozian (in

WrayI977) supplementedmaternaldietswithI00 gramsofproteinwhichwascontrolledso as not tochangethecalorieintake,and milkproductionincreasedby upto 200 c. c. perday.Thusproteinlevelsinbothmother'sandinfant'sdietarea majorfactorindetermininghow well theinfantwillthrive.

Diet also influenceslevel ofprolactionwhichcaninhibitmenstrualcycling.This effectwas demonstrated(Lunnet al. I984) in a studyofpregnantandlactatingGambianvillagewomen. They gavea highenergyproteinandvitaminsupplementthatraised calorielevelsto 229I kcal. eitherduringlactationorduringpregnancyandlactation.As a pregnancysupplement,itimprovedthebirthweightofchildren,andinbothcasesitloweredprolactinlevelsduringthe

lactationperiod.The women supplementedduringpregnancyand lactationresumed cyclingand ovulatingin one-halfthe time that unsupplementedwomen did (43 V. 79 weeks forresumptionin 50 per cent. of thewomen).Improvingthe dietduringpregnancyand lactationratherthanmerelyduringlactationwas more effectivein shorteningtheinterbirthinterval(37 V. I9 percent.)whencomparedto theunsupplementedcondition.



A. C. ZELLER 539

Activitylevelsand nursingpatterns.The effectof maternalactivitypatternsmayoccurthroughseveral channels.In a predominantlygatheringsituation,the

quantityoffood may be adequatebut thecaloricenergyexpendedin amassingand processingit may be a significantlylimitatingfactoron the woman'scapacityto buildup thenecessaryfatlevelsto conceiveandbringthepregnancyto term(BuchananI975; Dumond I972; Howell I976b; I979; LaughlinI968b;Lee I972; I979; I980). Lee statesthat'any changein thesubsistenceeconomythatallows reducedmobilitymay be sufficientto increasefertility'(I979: 3I9).

Thisobservationappearsto besubstantiatedby thedeclineininterbirthintervalsfrom44 to36 monthsamong!Kungwomen who havebecomemoresedentary(Lee I979). Even in industrialisedwesternsocietywomenwho engagein highlevels of activity,such as balletdancersor marathonrunners,may undergo

primaryorsecondaryamenorrhea(WarrenI979 inHuss-AshmoreI980). Frischstatesthatthe loss of 30 percent. of body fatreserveswill triggersecondaryamenorrhea(I977).

Anotherpossibleeffectofmaternalactivitypatternmayresultfromtheeffectofdailyscheduleson thefrequencyandpatternsofnursing.Whenmaximumnursingintervalsoccurbeforesixmonthshavepassed,orwhennightnursingisdiscontinued,cyclingis resumed more quickly even thoughlactationmaycontinueup untilI8-20 months(Knauer I984). Womenwho nurseless thanfour timesa day after6 monthsshow a drop in prolactinlevels which willusuallyresultintheresumptionofcycling(Jain& BongaartsI98I). The use ofwaterandsupplementaryfoodsbefore6 monthswill also reducetheperiodoflactationalamenorrhea.Fullbreastfeedingondemandwithoutsupplementaryfoodorliquidsforsix monthsisstronglycorrelatedwithlengthenedperiodsofamenorrhearangingfromI I to 36months(KnauerI984). The resumptionofcyclingmaynot be evidenceoffertilitysincethefirstfewreturningcyclesmaybe anovulatory(ShortI984). On theotherhand,ovulationcan occurbeforecyclingis evidentand severalstudiesindicatethata smallpercentageof womenbecomepregnantagainwithoutmenstruating(5.4 percent.Bonte& Van BalenI969; 6 percent.Jain& Sun I972). Thus in spiteof folkwisdom,menstrual

cyclingcan resumewhile lactationis stillin progress(LeridonI977). This isrelevantbecause long lactationperiodshave frequentlybeen suggestedas amajor cause of long interbirthintervalsin non-contraceptingpopulations.Infantswho geta smallervolume ofmilkmaynursemorefrequentlyandwithgreatervigour.Bothfrequencyandintensityofsucklingarereportedto havemarkedeffectson thehormonecyclewhich controlsthe onsetof ovulation(BongaartsI980; Frisch1975; Knauer I984; Konner& WorthmanI980; LeeI972; I979; ShortI976; I984).

Thuswomenwho leave theirinfantswitha caretakerforsubstantialpartsoftheday,eventhoughtheyare stillnursing,will tendto resumecyclingmore

quicklythanthosewhokeeptheinfantwiththemand feeditfrequently.Lee inparticular(I972; I979; I980) statesthatcontinueddemandnursingup

toage 3maybe a majorcause of thelongbirthspacingamongthe!Kung.Thisargumentis basedmoreon thestimulusofsucklingon the mother'sneuroen-docrinesystemthanontheeffectofcontinuednursingonthemother'sfatlevels(Konner& WorthmanI980; Lee I980). Theimportanceofthesucklingstimulus



540 A. C. ZELLER

is confirmedbyexperimentaldenervationof mammaryglandsinanimals,whothenresumedreproductivecyclingwhilelactating(ShortI976).

Weaningfoods.In additionto diet,maternalactivityand patternsofnursing,cycleresumptionis also affectedby thetypesofweaningfoodsavailable.Mostsocietiesbegin supplementaryfeedingbetween6 monthsand i yearof theinfant'slife.Until 6 months,lactationcan supply all of the infant'sdietaryrequirementsbutpastthatpointitisnot sufficient,althoughfrequentlyitis theinfant'sonlysourceof high qualityprotein(Frischi982). Between6 and I2monthsanadequatelynourishedmothercanprovideaboutthree-quartersoftheinfant'sproteinrequirements(BuchananI975). Scrimshaw(i983) suggeststhat.75 gm/kghigh qualitymilk,meat or egg proteinis requiredto establishand

maintainadequategrowthratesandprotectionfrominfection.Ifthemotherispoorlynourishedmanyoftheseproteinand fatresourceswillbedrawnfromherbodytosupplytheinfant.

Continuednursingamong the !Kung is considerednecessaryto provideadequateproteinforinfantsbecauseofthelackofsuitableweaningfoodsinthenomadichuntingandgatheringdiet.Thisproblemof proteinsupplycontinuesintothesecondand thirdyearoflifeamongthe Kungandisemphasisedbytheseveregrowthretardationobservedin post-weaningchildren(Truswell &Hansen I976). !Kung infantsgrowatratescomparableto westernchildrenforthefirstsixmonths,nourishedonbreastmilk,butafterthisdroptobelowthe3rd percentileof theheight/weightstandardforage of American children(Truswell& Hansen I976; Lee ig80). Between6 and 12 yearsof age theselowweight tendenciesare more pronounced,with !Kung childrenattaininganaverage weightof only 63 per cent. of the soth percentileaccordingto theHarvard standard(Truswell& Hansen I976). This resultsin adultswho areshortand light,with male and femalefatlevelsonly43 and 6i per cent.ofWesternstandardsrespectively(Lee I979). The adultweightoffemalesaveragesabout 42 kg. (Lee I979) which is close to thelow end of therangeFrischpostulatesas necessaryto achievereproductivestatus(FrischI975). Children

who liveat thecattlestationswithpotentialweaningsupplementsare tallerandheavierfortheirage to adulthood (Lee I979). Thereforeit seems clear thatperiodsofproteinorproteincalorieshortagearepresentduringthe!Kunglifecycleandmayhave aneffecton theexceptionallylongbirthintervalnotedforthisgroup.

Overall, then,the literatureon modernhumanssuggeststhatinterbirthintervalsaregovernedby a complexsetoffactors,butseldomextendbeyond36months.The exceptionsnotedincludejainandBongaarts(i98i) studyindicat-ing38.2 monthsinIndia, 38.4monthsinIndonesia,37.9 monthsinSriLanka,and Lee's (I972) (Howell I976b; I979) reportsuggesting44 monthsfor the

!Kungintheirnomadicphase.These maximuminterbirthperiodsare stillonlytwothirdsthelengthofthosereportedforwildapes.

To summarise,thefactorsaffectingthelengthof the interbirthintervalinhumansincludelevelofbody fat,maternaldiet,maternalactivity(subsistencepatterns),lengthof lactation,patternof nursing,intensityof sucklingandpresenceofweaningfoodsintheenvironment.There are otherfactorssuch as



A. C. ZELLER 54'

maternalagewhichaccountforintra-individualvariance,butarenotconsideredhere.Levelof foodsupplyanditseffecton maternalactivityarestartingpoints

forexaminingthevalue of children'scontributionsto theirparents.

Contributionsofchildren

Pastoralandpeasantcultures.Data on theeconomiccontributionof childrenareincludedin somestudiesof peasantand pastoralsocieties.Care of infantsandlivestockis a frequentformofcontributiveeffort.Althoughtheseactivitiesmaynot providedirectsubsistencegain theyfreeadultsforsuchactivityand canbe

classedas a valuableindirectcontribution(WhiteI975). Borgerhoff-Mulder&Milton(I985) examinedpatternsofinfantcare,and thereforepatternsof siblinginvolvementin caretakingamong theAfricanKipsigis.They notedthatfairlyyoungchildren(medianage6.3 years)assumedhighlevelsof responsibilityforinfants.In Munroe etal.'s studyof fourcultures(I984), childrenspent23 percent. of theirtimeworking.This figurewould have been only 7 per cent.ifinfantcareand domesticchoreshad notbeenincludedas contributions(Munroeet al. I984). Indeed,childcare is mentionedspecificallyas a contributionbychildrenamongtheGusiiofKenya(LeVine& LeVine I966), theRajputsofIndia(Minturn& HitchcockI966), the MixtecansofMexico (Romney & Romney

I966) and theTarongofthePhilippines(Nydegger& NydeggerI966) whichareall peasantlevel societies. Other typesof helpfulactivitiessuch as runningerrands,caringfor livestockand domestic chores are also mentionedforchildrenfrom3 to IO yearsofage. Child care and animalcare are particularlynotedas economiccontributionsbyJavaneseand NepalesechildrenstudiedbyNag etal. (I978). Six- toeight-yearold childrenspentanaverageof3.5hrs.perday in work activitiesinJavanesevillagesand 4.3 hrs. per day in Nepalesevillages.Nine- to eleven-year-oldJavaneseandNepalese childrenspentfrom3.I to6.5hrs.perdayincontributiveeffort.The hoursspentperday continuedtoriseas theagesincreasedup to adulthood(Nag etal. I978). These studiesall

indicatethat childrenin agriculturallyorientedsocietieshave an economicimportanceto theirparents,whetheror nottheparentsperceivetheactivitiesasbeingveryuseful(McDowell 198I).

Hunter-gathererandhorticulturalsocieties.The materialfora surveyofchildren'scontributionsinhunter-gathererandhorticulturalsocietiesis notreportedintheliteratureforverymanysocieties.Irons(I983) commentsthatchildreninthesesubsistencepatternsareaneconomicburdenuntilaboutage I2 to I 5 whentheybeginto becomeeconomicproducers.Since I agreewith White(I975) (also

Munroeetal. I985; Nag etal. I978) thatthe contributiveactivitiesofchildrencouldbothdirectlyandindirectlyreducetheeconomicdemandsmade on theirparents,I interviewedelevenfieldresearcherswhoprovidedfirsthandinforma-tiononthirteencultureswithhunting-gatheringand limitedhorticulturalbases(table3). The !Kungareoftenconsidereda prototypicalhunter-gatherergroupand no first-handinterviewswereavailable,so I used2 publishedsources,one



542 A. C. ZELLER

fromthe fieldworker'sperspective(Draper I976) and the othera reportofchildhoodmemories(Shostak I976), bringingthetotal to I4 cultures.The

surveydealtwitheightmajorclassificationsofbehaviour,each subdividedintovariousactivitiesconstitutingeitherdirectorindirectcontributions.The activi-tieslistedarethosebywhichchildrencould contributeto theirown maintenance(direct)or assisttheirmothers(indirect)towork moreefficiently.Table 4 liststheactivitieswithparticipationdividedby ageintothreestages,3-5 years(stagei), 6-Io years(stage2) and I I-I 5years(stage3). The materialis not dividedbygendersinceinmostcases bothsexes contributeassistance,althoughinseveralcasesitwas notedthatgirlshelpedtheirmothersmorethanboys.Theageoffirstcontributionis indicatedbut activitiesnoted at one age are usuallyfoundinsubsequentage categories.

Childrenoftheyoungeststage (age 3-5) made theirmajorcontributionsattheindirectlevelofchildcare,carryingobjectsandrunningerrands(table5). Inmanycasestheyused tools, althoughat a fairlylow level; nevertheless,theywerebeing givena veryearlyopportunityto familiarisethemselveswith thetechnologicalaspectsof theirculture.Knives, axes, fire,and carrynets wereusedbychildrenatageswhichprovokedcommentbytheinformants.'Whenamotherneededherknifeshetookitfromtheyoungchild,who wasplayingwithit,used it,and gaveitback tohim'(Rodman,personalcommunication).Snaresand diggingstickswere put to use by four-yearolds, who rapidlygained

proficiencywiththem.Inmanycasestheyare consideredreasonablycompetentaroundfire,andamongtheCree aresometimesdetailedto watchtheroastingmeatand turnitasnecessary.Contributionstogardeningwerenotveryhighatthisstage,probablydue to lack ofstrength,but theconceptsofplantingandharvestingwere introducedat an early age in those cultureswhichutilisedgardens.

The majorityof contributionsin both directand indirectactivitieswere

TABLE 3. Contributorstosurveyofchildhoodactivitypatterns.

Informant Location Tribename Subsistence

ElspethYoung P.N.G. Highland GardenElspethYoung CentralAustralia Wabwi H & GNancyMcDowell P.N.G. Bun Fish/SagoPatTownsend P.N.G. Saniyo-Hiyowe H & G, SagoJaneGoodale N. Australia Tlwi H& GJaneGoodale S.W. New Britain Kaulong H & G, GardenMargaretRodman Vanuatu Vanuatu GardenNaomi Scaletta W. New Britain Kabana Garden,FishDorothyCounts W. New Britain Kaliai Garden,H & GLeslieMarshall Truk Namaluk Garden,FishSara Preston N. Ontario Cree H& GJackieSolway Kalahari Bakgalagadi Pastoral,H & G

MikeLambeck Comoro Islands Mayotte Swidden,FishP. Draper* Kalahari !Kung H& GM. Shostak** Kalahari !Kung H&G

*Draper,I976** Shostak,1976Iwouldliketoextendmythankstotheeleveninformantswhoprovidedthematerialonwhichthissurveywas

based.



A. C. ZELLER 543

TABLE 4. Childhood activitypatterns-subsistence relatednumber of groups.

3-5 years 6-1oyears 11-15 years Total

IndirectcontributionChildcare-start6-io months

carry I 9 IO

watch 5 9 I4protect 4 8 I2

feed I 4 I 6cook for I 5 I 7

Carrytools 6 2 I 9water I 7 I 9firewood 2 7 I IO

messages(fire) 5 I 6

runerrands 6 I 7food(produce) 3 4 7buildingmaterial 2 2 4walklongdistance(miles) 3 3

Householdhelpsweep 6 6wash 8 8

Gardenweed I 3 2 6fixfences 2 2

plant I 3 4harvest I 3 4prepareground 2 2

own I 2 3Careforgrandparent 5 I 6newmother 2 2

livestock 4 4chickenls I 2 3

DirectcontributionUse tools

fire 3 5 8knives 4 4 8axes 3 4 7diggingsticks 2 3 5

carrynets 3 2 5slingshots I 2 3gun,bow, spear I 3 4netfish 2 2 4snare I 3 4canoe 2 3 5

Catchandgatherbirdsandeggs 2 7 9smallanimals 3 5 8grubs 2 5 7shellfish 4 4plants 3 6 9fish(spear,hook) 2 2 4

hunt-spear, gun I 3 4netfish 3 3ownprotein I 4 5

Preparefoodhelpa little 3 2 I 6wash,grate,pound, peel I 3 2 6cookforfamily 3 2 5huskandgratecoconut I 3 2 6



544 A. C. ZELLER

initiatedbetween6 and io yearsofage (Stage2, table4). On a proportionalbasis,child carewas themostimportantactivitycarriedoutby children,and

rangedfromplayingwiththebabytocaringforhim orheroverthecourseofadayorevenovernight(LambeckI985). Thiswouldrelievethemothersofmuchofthestrainofcontinuouslycaringfora one-tothree-year-oldsinceshecouldforageorgarden,undistractedbya clinginginfantwhomshewouldnothave tocarryandnurseall day. Otherindirectcontributionsincluderunningerrandsandcarryingfirewood,water,produceandmessages.Veryyoungchildren(age3)maystartwithoneortwo sticksofwood, oryamsina carrynet,butby age8theyarecarryingfivelitresofwaterwhiletheirmotherscarrytwenty(Solway).Constantpracticefroma youngagebuildsupboththestrengthandtechniquesnecessaryto transportvarioustypesof loads. At thispoint,carryingone- totwo-year-oldsiblingsmayoccurforsubstantialportionsoftheday. AmongCentralAustralians,femalesunderage I 5 performed38 percent. of all infanttransport(Denham I974). In addition,helpwithhousehold choressuch aswashingclothesanddishes,sweepingandmending,is frequentlyexpected,ascanbeseenfromtable4.

One specialisedaspectofdomesticwork,which canbe regardedas bothadirectandindirectcontribution,is thetimeandenergyconsumingpreparationofcoconuts,taro,seedsandnuts.Bysixto tenyearsofagechildrenwash,grate,pound, peeland cook food on a moreorlessregularbasis. Insomecasesyoung

childrenare lefthomein thechargeof a ten-yearold who caresforthemandpreparesthefamilymeal.Inthissituationtherearealmostalwaysotheradultsinthevicinityincase ofemergency.Evenunweanedchildrenareleftsincetheycanbenursedmorningandnight,andfedmashedbananasor theequivalentduringtheday.

The abilityto undertakehouseholdmaintenanceactivitiesalsomadechildrenusefultoadultsotherthantheirparents.Infactitwas specificallymentionedbyinformantsthatchildrenarefrequentlyadoptedbyindividualswho donothavethem,such as grandparents,forboth theirlabourand companionship.Thesechildren,whorangeinagefrom5to IO,contributetotheproductiveprocessby

gardening,carryingwater and wood, preparingfood and washingclothes.LambecknotesthatamongtheMayotteone oftheprimaryreasonsfortakinginchildrenisto have a helpinghand(LambeckI985: 9). Youngmothersmaytake

TABLE5. Categoriesindescendingorderofproportionalcontribution.

Categories Stage1 Stage2 Stage3 Total *Proportion

Childcare 12 35 2 49 9.8Carry 23 25 7 55 6.8Householdhelp 0 14 o 14 7.0

Catchandgather I I 34 8 53 5.8Preparefood 5 II 7 23 5.7Use tools 17 28 5 50 5.0Garden 4 15 2 21 4. 2Carefor I 13 I 15 4.2

*Proportionreferstothetotalnumberofcontributionsinacategorydividedbythenumberofactivitiesinthatcategory.



A. C. ZELLER 545

in a nieceoryoungsiblingto assistthemwiththeworkinvolvedincaringforanewbaby. Among theCree, olderpeoplewho stillwish tolivealoneareable to

managetheirhouseholdchoreswiththehelpof a five-orsix-yearold,whocancook, carrywood and clean fish(Preston).It is quite possible thathavingchildrenwhoarecapableofundertakingtheseactivitiesforolderkinwill relievemothersof taskswhich theywould otherwisebe socially constrainedtoperform.Animalcareis includedinthiscategorybecause itinvolvesa certainamountofresponsibility,althoughof a differenttype.Since,however,mostoftheseculturesweresurveyedon thebasis of a hunter/gathereror horticulturalsubsistencebase,onlya fewgroupshad livestockorpoultrytobe caredfor.Asmentionedabove animal care is oftena major contributionof childreninpastoral(AinsworthI985; Hassan I980) orfarmingcommunities(White1975).

Thisage categoryhas alsobegundirectsubsistencecontributionby collectingbirdsandeggs,smallanimals,grubs,shellfish,plantfoodsandby nettingfishwhereavailable.They maynotaccompanytheirmothersto do this,butmayformpartiesofpeersandforageon theirown.YoungAustralianAboriginegirlsbegindiggingforlizardsat aboutage 4 andareroutinelysuccessfulatthisskilledandenergeticactivityby age8to io (weightoflizard:2 to4 lb, ElspethYoung).Childrendo notalwayssharefood withadults.Frequentlytheymake firesandcook it in the bushduringtheday (Rodman). In some cases thisprovidesasubstantialportionof theirproteinintakefor the day (Chowning I972).

LaughlincommentsthatAlaskan Eskimo '(c)hildren. . . actuallyproduceagreatpercentageoftheirown foodsupplyin thesubarcticzone-including,especially,shellfish'(i968b: 241-2). There are a numberof casualreportsonchildrencatchingbirds,rats,grubsandinsects,whichtheyeithereatthemselvesor sharewiththeirmothersandsiblings(Dwyer I974). Thenutritionalvalue ofthispatternof searchingforand eatingfoodduringthe courseof theday isconfirmedbyRobsonandYen's (I980) reportfortheTasaday. Theydiscoveredthat,based on the amountof food broughtback to the home cave, dailyon-the-spotconsumptionofsmallitemsintheforestmustprovidealmost2/3oftheTasaday's dailyintakeofenergyandprotein.Childrenmayalso undertake

time-consumingtaskssuch as berrypicking,grassseedcollecting,or diggingforgrubs-which theydo bringback to thefamilyunit.Girlsmay accompanytheirmothersonto the reefto gathershellfishandboyswilloftenaccompanyteenagerson a spearfishingexpedition(Scaletta,personalcommunication).

Thereareonlya few culturesinwhich contributiveeffortis initiatedat theeleven-tofifteen-year-agecategory.There are sometypesofassistancesuch ascookingforyoungersiblingsorhuntingfairlydangerousanimalswithspearsandfirearmswhichare morecommonlyadded to thebehaviouralrepertoireatthisage. Heavy gardenworksuchasfencing,preparinggroundandharvestingsago maybe leftuntilthisage also, althoughweedingandharvestingsmaller

foodmay beginearlier.By age I2 someHuntingandGatheringchildrenarepotentiallyself-supportingin a subsistencesense (Tiwi and Cree) whereasothers,suchas the!Kung,arenotexpectedto be. GoodalementionedthatsomeTiwi childrensubsistedforseveralmonthsinthebushafterthedeathof all theadultsintheirgroup(personalcommunication).SolwaysaysoftheBakgalagadithatby the timea girlis I2 she can run a household,and beforethis she is



546 A. C. ZELLER

expectedtoputinafulldayofwork.AmongtheKaliaigirlsage I 3canmanageagardenand are capableofmakingcopraforsale. Thus it is not so muchthe

differencesinsubsistencebase thataccountforvariationsin childproductivity,as differencesin whatis expectedin theculturalpatternand whatisexpectedofchildren.ThispointissupportedbyBorgerhoff-Mulder& Milton'sresearchonthesocialand economicfactorswhichappearto be correlatedwithhighlevelsofsiblinginvolvementinchildcare.Theyfoundthatfamilystructure,patternofresidenceand divisionoflabour,contributeto differencesbetweenculturesinthepatternsofinfantcaretaking(i985).

Levelsofcontribution.The fourteengroupssurveyedshowgreatvariationinthenumbersof activitiesbywhich childrencontributeto theirmaintenance.Theeightcategoriesof contributionare subdividedintoforty-eightareasinwhichthereispotentialforchildhoodinvolvement.Intwelveofthegroupsassistanceoccursin fromten to thirty-eightactivities,whilein the othertwo culturesminimalamountsofhelpinthreeareasarenoted(table6). IntheI2 cultureswithmoderatetohighlevelsof childhoodcontributions,a totalof270 activitiesareundertakenwitha meanof22.5 activitiespercultureoutofapossibletotalof48.The averageparticipationis actuallymuchhigherthanitappearsbecausesomeof these activitiesare not presentin all cultures.For example, theCentralAustralians,KungandBakgalagadido nothavefishingtools oractivityaspart

oftheirculture,andinsomegroupsgardeningisnotpartofthesubsistencebase(e.g. Tiwi,CentralAustralia,Cree).Thus ofthe48 potentialcategories,thefourgroupsinwhichchildrenundertake30ormoreactivitiesshow a veryhighlevelof childhoodinvolvementintheoverallculturalpattern.

At theotherendofthescaleare thetwoculturesinthissurveywhichshownegligiblelevelsofinputfromchildren.These are theSaniyo-HiyoweandtheKaulong, bothofPapuaNew Guinea.Inbothcases theinformantscommentedthatlevels ofcontributionby childrenwereverylow andthiscorrelatedwithlow levelsofcompletedfertility.Among theSaniyo-Hiyowe,sagoprovides85percentofthe caloriesutilisedwhilemeatsauces, madefromferalpig, small

game,fishand insectlarvaaccountforio percent.The remaining5per cent.is

TABLE6. Contributionlevelsofchildrenindescendingorder.

Stagei Stage2 Stage3 Total

Kabana 5 22 II 38Kaliai I0 22 2 34Vanuatu 5 26 I 32Cree 10 17 3 30Mayotte I 19 I 21

Numluk 8 8 3 19Bakgalagadi 2 15 2 I9

CentralAustralian I 5 3 0 I 8PNG 13 4 0 17Tiwi 0 15 2 17!Kung o 8 7 15Bun 2 7 0 9Sanlyo-Hiyowe 2 I 0 3Kaulong 0 I 2 3



A. C. ZELLER 547

composedofotherplantfoods.Enough sagocan be gatheredquiteefficientlybyone woman to lastforseveralweeks;itis,however,a veryhighstarchdiet,and

peoplemustsupplementit withtheotherdietarynecessitieswhicharegatheredfrom the forest.This is especiallyimportantfor young childrenwho areundergoingthetransitionfromnursingto adultdiet. Sago isnot easilydigestedand a numberof one-to three-yearold childrenshow symptomssuchas soft,discolouredhairand listlessness,which are associatedwithproteinshortage(Townsend I980). Thisproblemof feedingyoungweanlingchildrentendstoprolong the nursingperiod, and as a result post partumsex taboos andinfanticideare used tomaintainbirthspacingatabout threeyears.Townsend's(I980) studyindicatesthatthemean numberoflive birthswas 5.3, withinfantand earlychildhoodmortalityof43.2 percent.Itisan interestingpointthat57ofI32 childrendied beforeweaning,and only fivedied after.This may relatetodietaryproblemsintheprolongednursingperiodas well as thehighincidenceofdiseaseslethalto thesechildren.Inmostcultureswithhighinfantmortalitythegreateststressoccursin thepost weaningperiod(Lancaster& LancasterI983;Lancaster& King I985; Lee I972; I980; Ripley I980). Eleven per cent.of allchildrenbornaresubjecttoinfanticidewhichisthecauseof23 percent.ofinfantand earlychildhood deaths.In almost all cases the explicitreason givenforinfanticideis birthspacing.It seems to weighmoreheavilyon girlsthanboys,sinceeightofelevenvictimswerefemale(Townsend I980). The demographic

consequencesofthis

practiceis a

populationwhichis barelyreplacingitselfin

termsoffemales(grossreproductiverate,Howell I979). Theforty-ninewomenofcompletedfertilityinthisgroupproducedforty-ninedaughterswhomaynotallsurviveandreproduce(TownsendI985). Thus, althoughthesewomenhad amean birthrateof5.3, onlytwo or threesurvivedtoreproductiveage,withabiasinfavourof males.This is about thesame level of successas apereproduc-tion.The correlationbetweenlevelsof contributionbychildrenandreproduc-tive success aremaintainedin thisculturesinceoffspringengagein very fewcontributiveactivities.Girlsdo nothelpwithchildcarebeyondplayingwiththeinfantfora few hours,althoughtheydo helptheirmothersgatherfirewood.

These activitiesare the only ones mentionedduring the data collectionsurvey.Anothergroup with low levels of populationreplacement(about four

offspring)andminimalhelpfromchildrenaretheKaulongofSW New Britain(Goodale). Thesepeopleacquireapproximately6opercent.of theirfoodfromthebush and about40 percent.fromgardening.Birthspacingof fouryearsismaintainedbyprolongednursingand sexualabstinence.Fouryearsofnursingisconsideredessentialforthecontinuedlifeand healthofyoungchildrenandpoorhealthis amajorcauseofinfantmortality.The oldersiblingmustbeableto walkindependentlyto thegardenarea beforea new infantwill be allowed to live.

Mothershavetotalcareoftheinfants,whichincludesholdingthemat alltimes,even when theinfantsare asleep. Afterage 3 infantcare is sharedby oldersiblings,buttheyarenotleftin a highlyresponsiblesituation.Whenthefamilygoesto thegardenchildrenpickupsnailsandnibbleonvariousfoodsduringthecourseoftheday. Girlsbegin helpingin thegardenearlierthanboys but allchildrenare stillquitedependentuntiltheyareintheirmid-teens.Mothersdo



548 A. C. ZELLER

mostofthewood, waterand foodcarryingas wellas cookingforchildren,butmenandteenagersmustcook forthemselves.

These two cases oflow levelsof populationreplacementcoupledwithlowchildhood productivityare in contrastto the other culturessurveyed.The!Kung appearto be an anomalouscaseinthatfieldresearchersreporta verylowlevel of child involvementin contributiveactivitiesalthoughfifteenareas ofactivityarementionedintheirreports.Theseoftenoccur,however,underadultsupervisionand atanolderstageof childhood(stage3) thanis thecasein manyothergroups. Another factoraccountingfor thisdiscrepancymay be theamountof timeinvolved. Solway (personalcommunication)commentsthatBakgalagadichildrenarecontributingseveralhoursperday by thetimetheyare6 to 8 yearsold and mostof theday by age io. Draper's (I976) studyof !Kung

childrenfindsthattheycontributedforonlya few (I-5) minutesin a samplehour. Lee (I979) mentionsthat!Kung childrenare not pressedintothesubsist-ence quest untiltheyare teenagers,althoughtheyrunoccasionalerrandsat anearlierage. The correlationbetweenlow levelsofcontributionbychildrenandlow birthrateshowever,agreeswiththehypothesisproposedin thisarticle.The!Kung haveanaverageof4.7 to5 childrenwitha loss rateof about 34per cent.(Howell I976b). Thus theirrateofpopulationgrowthin nomadic groupsisquite low. However,the !Kung livingsituationis undergoingrapidchangeastheybecomesedentaryandthebirthintervaldecreases(LeeI972; I979; I980).

ContributionsbyapeandAustralopithecinechildren

Ape pattern.The majornonhumanmodel of Australopithecinebehaviourisbased on the informationderived fromstudyingchimpanzees,as a highlyintelligent,tool using, social, closelyrelatedprimate(Tanner I98I; ZihlmanI98I). Chimpanzeeinfantcare differsfrom the human patternin severalsignificantways. In contrastto thehumansituation,ape mother-offspring

relationsaremuchmoreheavilyslantedtowardsthemotheras provider.Afterchimpanzeefemalesbear theiryoung, theynurseand carrythemextensivelyuntiltheirthirdyear.Betweenage 4 and 5, finalweaningoccurs,butmothersstillcarrytheiryoungiftheytravela longdistance.Theyounglearnfoodhabitsby watchingtheirmotherspick fruit,or fishfortermites,and occasionallymotherssharefoodwiththem,or allow themto takeit(Goodall I979).The nextsiblingisusuallybornwhentheyoungsterisin itssixthyear,butin chimpanzeesandgorillasitis stilldependenton themotherforemotionalsecurityandsomecare.Juvenilesfrequentlysharethe mother'snightnestwithher and the newinfant.Not untilthenewinfantis about i yearold and thejuvenile about7will

the motherallow thejuvenileto carryit, and eventhenonlyforshorttimeperiodsand undersupervision.Apesdo not seem toutilisethemonkeypatternoffairlyextensiveallomotheringandbabytendingwhich can serve togive amonkeymothera respitefromher infants.Usuallyjuvenile apes begin towanderfromtheirmothersforincreasedlengthsof timeat aboutage 7 or8, astheymatureintoadolescents.It is not untiltheyhave achievedthislevel of



A. C. ZELLER 549

emotionalindependenceand knowledge of the environmentthatthey arecapableofsuccessfullysurvivingon theirownoradoptinganorphanedsibling

(Telekietal. 1976). Youngeranimalshavebeenobservedattemptingto serveassurrogatemothers,buttheyarerarelysuccessfulinkeepingtheorphanedinfantalive.

Mother chimpanzeesmustnursetheiryoung duringthetime thattheyarebecomingfamiliarwith,and capableofexploiting,the naturalfood sources.The shiftto solid foodtakes longer forchimpanzeeyoungthan forhumansbecausetheymustbe able to climbtreestogetatthefruit,crackthenuts,pullupvegetation,andoperatea termitestickthemselvesbeforetheyarenutritionallyindependent.All theseoperationsrequireboth learningand some level ofstrengthor dexteritytoachieve.Theirmotherswill sharesome foodwiththem,

butnotenoughto nourishthemiflactationwereto cease when theinfantwasstillquiteyoung.Sincejuvenilesdo notcontributetotheirmother'soryoungersibling'sdiet,and do notrelievethemotherofanysubstantialchildcareburden,theyareofverylittlehelpinraisingthenextsibling.Infact,theystilldependonthemotherfora considerableamountof socialandpsychologicalsupport,andsometimesfortransportation.

Australopithecinepatternanditsinfluenceon hominisation.Clearly, manyof thedifferencesbetweenAustralopithecineandchimpanzeelifestyleareresultsoftheprocessofhominisation,butwhenandwherethesebehaviourswereinitiatedinhominiddevelopmentis notyetevident.Fourmajorchangesinlifestylewouldcontributeto thedevelopmentofthedifferencesin childcarepatterns.The firstdevelopmentconsideredis the transitionfromforagingto gathering,coupledwiththedevelopmentofsimpletoolsandcarryingaides.Secondlytheestablish-ment of group life and the division of labour would be fosteredby thedevelopmentoflanguageout ofa moregeneralisedcommunicationsystem.It isnotknownwhenculturalpreparationof fooddeveloped,butthischangemighthave markedeffectson the survival rate of theyoung, especiallyin crisissituations.Fourthly,a new locusof resourceexploitationutilisedbyhominids

mayhaveprovidedan advantageover thepreviousape-likepattern.There areroleswhichimmatureoffspringcouldfillin all of thesedevelopments.

Gatheringvforaging.The habitofgatheringfoodas opposed tomerelyforagingforitis amajormilestoneinhumandevelopment.Itishighlycorrelatedwiththeearlierweaningpracticesof humans(RipleyI980). All thegroupssurveyedinthisstudyhadbegunthepatternofsupplementalfeedingatbetween6 and i 8

months,whenhumaninfantsarenotable toforageforthemselves.Young apesbegintosupplementtheirmilkdietatabout I2 to i 8monthsbutmustcontinuenursingmuchlongerbecausetheymustrelyon theirown resourcesto achieve

thissolidfoodintake.The additionaladvantageforhumangatherersis thattheycan leave theiroffspringin someone else's care,and returnto feed them (apatternseen in birds and carnivores,but not in Anthropoidea).The use ofchildrenage 6-Io tosupplementthecareofyoungerchildrenwould undoubt-edlyrequirethepresenceof otheradultsin thenearvicinity,and thusa groupliving pattern.Young hominidscould also contributeby performingsimple



550 A. C. ZELLER

repetitivetaskssuch as berryor seed gathering.Small edibles at lake or rivcredge such as crayfish,snailsorfrogscouldeasilybe gatheredbytheyoungand

consumedon thespotorcarriedback to thegroupin a container.Data on theeffectsof proteinshortagespresentedearlierin this articleemphasisethatalthoughnursingchildrencan surviveon milklow inprotein,theyaremuchmore viable whenproteinresourcesareadequate.Thus smallcontributionsofproteinaceousfoods made by childrento theirmothersmay have a nutritionalimpactfarexceedingtheirsize and caloriccontent.These 'snacks'would also beverybeneficialiftheywereconsumedonthespotbythechildren,becausetheywould relievetheparentsofpartof theirforagingresponsibilities.

Major advantageswould accruetoyounghominidsfromtheuseof tools andcontainers.The earlyinitiationof thisdevelopmentseemsprobable,based on

analogous ways whichotherprimatesextend theirreach, open tough foodsourcesand dealwithproblemsofprocurement(Goodall 1973; Tanner I98I;

Zihlman I98I). If theearlieststone tools date from2.6 millionyears ago, itseemspossiblethatwood, shelland skinmightalso have been usedat least asearly.The developmentoftools wouldallow childrentoaugmenttheirnaturalstrengthandcapacityby carryingmaterialinskinbagsor ostrichegg shells.Adiggingstick would enhancetheirabilityto gathergrubs, small burrowinganimalsor tubers.These tools aretechnologicallyverysimpleandwould notleave muchtracein thefossilrecord,but it seemsquite possible based on achimpanzeemodel of tool use (TannerI98I) thattheymayhave beenavailabletoearlyhominids.Theyarecertainlywithinthecapacityof modernchildrentomanipulate.

Languageandgroupliving.The developmentof languageskills sufficienttoenable humansto plan and co-ordinateactivities,instructthe young, andarrangea rendezvousis one ofthemorevaluabledevelopmentsbecauseitallowstheorganisationofgroupactivities.Evidenceforthisdevelopmentisenigmatic,sincethetracesofitspresenceleftinthefossilrecordareso slight.Thereissomesuggestionin endocranialcasts ofAustralopithsthatBrocca's area on the left

surfaceof the brainis beginningto show increaseddevelopment(Holloway1976). Nevertheless,the developmentof languageand its role in enablingmotherandoffspringto communicateis undoubtedlya majorcontributiontohominidlife.Language skills would enable a motherto instructand directoffspringin thevarioustechniquesofgatheringratherthanrelyingon thechild'sinteresttoprofitfromobservationallearning.The motherwouldalso be able todirectthechild'sactivityat a distance,andtogivethechildinstructionstofollowinherabsence.Languagemightalso have fosteredchangethroughtheperiodofinnovationandplaycharacteristicofyoungprimates.Itcould have resultedinanexpansionofthediet,ifthechildgatheredsomethingthemotherwould not

have, and offeredto share it. Language would allow mothersto teach theiryounghow to careforinfants,ratherthanrelyingonobservationallearning,andtoorganisethebaby-sittingprocess.

Grouplifehasmanyadvantagesin termsofpredatorprotectionandforagingpractices.A situationsuchas thatoccurringamong theAustralianAboriginesmayhavedeveloped.Here mothersandchildrentraveltogethertoa protected



A. C. ZELLER 55I

placeneara goodgatheringarea.Motherswithbabiesstayhereand watchovertheyoung children(5-7 yearolds)who are caringforthetoddlers.By thisage

young girlsarefrequentlycarryingone- and two-year-olds(Denham 1974).

The five-to seven-year-oldsplaywiththetoddlers,keepthemout ofthefireanddo a littlegrub,seedorsmallanimalcollectingas well. Mothersdo notliketotaketoddlerscollectingwiththembecauseof thenecessityofcarryingwaterforthem(Denham 1974). With thisarrangementmotherscan leave unweanedchildrenforseveralhourswhilegatheringfood andthenreturnto nurse.Thusone ortwoadultscan overseeeightortenchildrenaltogetherwithoutbecomingexhausted,or actuallyfeeding,caringforor playingwith thechildren.Themotherswhoareoutcollectingreturnto this'dinnercamp' and preparefoodforthemselvesand thechildren.Onlya smallportionofwhatiscollectedduringthe

dayis actuallycarriedbackto themaincampatnightfallto be sharedwiththeothers.

Preparationoffood.A thirdhumanadvantageistheculturalcapacityto respondtolactationalfailureby cooking,poundingand pre-chewingfood forinfantstodigestmoreeasily.Humans have a culturalpatternwhichencouragesmotherstoprovidefood fortheiryoung,bothin timesof milkshortage,and duringtheweaning period. Among the Saniyo-Hiyowewho sometimesexperienceashortageof suitableinfantfood, non-motherswho findeasilydigestedgrubs,

etc., may bringthem as a

giftto the infant

(TownsendI980). The human

patternsofprocessingfoodmayaid theyoungin achievinga morecompletediet.Cooking, pounding,choppingandshellingnutsare allbehaviourswhichchildrencouldundertaketo liberatecalories(Cohen1980) andthuscontributetotheirown and theirsiblings,dietarysuccess.Chimpanzees(andotherapes) donotundertakethesetypesofbehaviourontheiroffspring'sbehalf.The bestthattheydo istoallow themto sharein foodthattheyhavepreparedforthemselves.

Resourcelocation.A fourthdifferentiatingfactoris that most human foodresourcesare locatedat orneargroundlevel. Motherswho are carryinginfants

donotusuallyhavetoclimbtreestogatherresourcessuchas nutsand fruit.Thedivisionof labour is suchthatchildren,men or non-motheringwomenwillusuallyobtain and share theseresources.This means that neithernursingmothersnoryoungoffspringhavetonegotiatetreesand runtheriskoffalling,whichiseverpresentinthearborealprimates.

Thevalue ofthesecontributionsoccursat botha directand anindirectlevelofinfluenceon thesubsistencebase. Ifchildrencanhelpto wean themselvesearlyby eatingsolid food thefrequencyand intensityof sucklingwill diminish,allowingthemother'shormonalcyclingtobeginagain. Slightlyolderchildrencanhelpto feedthemselvesand also theiryoungersiblings.They can gather

foodfordirectconsumptionandiftheymakeworthwhilecontributionsto theirmother'sdiet,hervolumeofmilkmaybe increasedforanursinginfant,and herbody storesof nutrientsnot so severelydepleted.A bettermilksupplywillpromotethesuccessof thenursinginfant,whilemaintainingher reserveswillallow herto becomepregnantagainmorequickly.Indirectcontributionssuchas childcare will allow mothersto gathermoreefficiently,and spendfewer



552 A. C. ZELLER

caloriescarryinginfantson theirgatheringexpeditions.Leavinginfantsin aprotectedcampwould also reducetheirriskofexposureto predation.In the

eventofapredatorencounterwhilegathering,motherswouldonlyhavetosavethemselvesratherthanbe burdenedbya heavyhelplesschild. Assistancewithfoodpreparationwouldallow a greatervolume of food tobe processedwithagreaterprobabilitythattherewould be enoughforall theyoung.The majorinfluenceof all thesecontributionswouldbe anincreaseinthemother'senergylevels,so thatshecouldconceivemoreinfants,andmoresuccessfullyraisetheones shebore.

Australopithecinefertility.The interestin Australopithecinefertilitypatternsisbased on theirstatusas ancestralhominidforms.Presentday Homo is un-

doubtedlyverydifferentfromearlyhominids,but theevolutionarylinkageshould allow some extrapolationto thepast. Chimpanzeesare also used asmodels sincetheirbodysize andforagingpatternsmaymorecloselyresembletheAustralopithecinecondition,andtheirancestorswerecloselyrelated.

Thus fertilitypatternsof bothchimpanzee(Pan) and modernHomo canbeexaminedforanyindicationstheymay giveconcerningtheAustralopithecinepattern.AlthoughAustralopithecinelevels of fatstorageandlactationalener-getics are unknown,the similarityin pregnancylengthsbetween Pan andmodernHomo supportan argumentthattheAustralopithecineconditionwasnot radicallydifferent.They probablyresembledwild, ratherthancaptivechimpanzeesin not gainingmuch more weightthanthe level requiredtobecome fertile.Accordingto theirskeletalremainsthislevel wouldhave beenabout 30 kg. which is just a littleless thanchimpanzees.The lengthof birthintervalsis basedpartlyon theamountof storedenergywhichcan be used forlactation.Wildchimpanzeesrequirea meanof42 months(TutinI980) to regainfertilityaftera previousbirth.Modernwesternwomenrequireonlyabout12 to

14 monthsto resumecyclingevenwhenbreastfeedinginfantstotally(Kippley1974). Post-birthamenorrheadoesnotusuallyexceedI 8 monthsin anyhumanpopulation, althoughfertilitymay be delayedfor severalmonths(Leridon

1977). Cycle resumptionin Australopithsprobablyfellbetweenthe Pan andHomo levels,althoughtherangeislikelyto be closerto thechimpanzeeend duetobodysize andactivitypattern.However, dietarysupplementationofmothersby kin,mate and possiblychildrencould operateto reduce thelengthof theinfertileperiodthroughthemechanismsmentionedin thisarticle.This change,coupledwitha greaterchanceof survivalforpost-weanlingchildren,couldincreasethenumberofoffspringsuccessfullyrearedby a mother.

Several authorshave investigatedthepossibilitythatfemalescould act toincreasetheirreproductivesuccessbychosingmateswhowould help themreartheiroffspring(Hrdy& WilliamsI983; Irons1979; I983; TriversI972). Others

have arguedthatAustralopithecusmay have had a more matrifocallivingsituationandutilisedaidfrommaternalkinwho would bemorecloselyrelatedto themthanweretheirmates(FischerI982; TannerI98I). This articlesuggeststhatoffspringwho areequallyrelatedtothemother,no matterwho she mateswith, mayhavecontributedtheirefforts,bothatmaintainingthemselvesandtheirsiblingsand increasingtheirmother'spotentialforadditionaloffspring.



A. C. ZELLER 553

Ironscomments(I983) thata largersibshipmayhave been advantageoustoearlyhominidsfortheformationofalliancesas well as thespreadofpopulation,

as long as thespacingwas not so closeas tojeopardise individuals.SuccessfulfemaleAustralopithecinescould benefitthemselvesby choosinga matewhowould helprearoffspring,as wellas byhavingchildrenwho wouldhelpboththemselvesandher.Iftheoffspringweresocialisedto assisteachother,theywould bein anadvantageouspositionwithrespecttoothersibshipswho didnotoperatein that fashion.If this behaviouraltraditionbecame common to amajorityofthepopulationitcouldincreasethenumbersofAustralopithecinesandpermitthemtoexpandfromtheiroriginalhabitat.

Conclusions

Thedatapresentedin thisarticlesupportanargumentthatsubsistenceactivitiesofhominidchildrencandirectlyorindirectlysave sufficientenergycostsforthemotherto influenceherreproductiveratebyshorteningtheinterbirthinterval.Studiesof greatape reproductionindicatethattheirapproximatelysix-yearbirthintervalisbarelysufficientto maintaintheirpopulationlevel.Fertilitydatasuggestthatnutritionalfactorsarea majorconstrainton humanreproduction,especiallyintermsofenergyexpendedforcaloriesgained.This is particularlyrelevantafterthebabyhas beenbornand themotherismaintainingbothherselfandthechild.Thepresenceofweaningfoodsofsuitablequalityandconsistencyarea majorfactoraffectingboth survivalofchildrenand thefrequencywithwhich theyare produced.Lengthof nursing,frequencyof infanticideandvoluntarychildspacingareallinterrelateddeterminantsofoverallfertility.

The actual contributionsmade by childrento the subsistencebase arediscussed in fourteenhunter/gathererand partiallyhorticulturalsubsistencesystems.Direct contributionsincludeprovidingsomeof theirown foodfromanearlyagewhichwillreducetheload on theirmother.Thiscontributionwillbe even more effectiveifolderchildrencontributeto theiryoungersiblings'foodsupply.Childrenmaysharesmallbutdietarilysignificanttypesof foodwhichtheyhave timetocollect,such asberries,insectsandinvertebrates,with

theirmothersandsiblings.Movingto a moreindirectlevelchildrenmayhelptheirmothersgatherwood andwater,workin thegardensandperhapsdo somehousework.Theycanalso relievetheirmothersofmuchof theburdenofdailychildcarefortoddlersandyoungchildren.Alltheseactivitiesallow themothertoconductherown subsistenceworkmoreefficientlyin termsoffood collectedand caloriesexpended.

In his surveyof theeconomicvalue of childreninJavanesecultureWhite(1975) pointsoutthatindirectcontributionsarejustas valuableasdirectones infreeingtheparentstoperformvitaladultactivities.Themostvitalbehaviourinanevolutionarysenseis theproductionandrearingofoffspring.As thisarticle

indicatesthereare a greatmanyfactorsinvolvedinsuccessfullyachievingthisundertaking,even afterthemaleandfemalehave choseneachotherandbegunthematingprocess.In additionto themale-femalebond are levels ofmaternaldiet,maternalactivitypatterns,lengthof lactationalperiod,infantdiet,andsuccessindefendingtheyoung,whichareallmajorvariablesindeterminingthenumbersofoffspringwho canbe raisedto maturity.These factorsaffectnot



554 A. C. ZELLER

onlypresent-daypopulationsbut alsoearlierhominidswho haddifferenttypesofresourcesbut thesametypesofproblemsto face.The datapresentedin this

article support the suggestionthat one resource was their offspringwholabouredwith theirparentsto maximisethe size oftheirsibshipandestablishhominidsas themajorprimatetaxon.

NOTE

This is a revisedandexpandedversionof a paperpresentedto theCanadianEthnologicalSocietyMeetingsinToronto,May i985. Thepaperis dedicatedto thememoryofmyfriendandcolleague,Dr MelissaJ.Knauer,who diedina tragicaccidentonherwayto conductfieldworkwiththe!KunginJulyi985.

REFERENCES

Ainsworth,H. G. i985. Biologicalanthropologyandnutrition.J.hum.Evol. 14, 335-46.Birdsell,J. B. I968. Some predictionsforthe Pleistocenebased on equilibriumsystemsamong

recenthuntergatherers.InMan thehunter(eds) R. B. Lee& I. Devore. Chicago: Aldine.Bongaarts,J.1980. Does malnutritionaffectfecundity?Asummaryofevidence.Science208, 564-9.Bonte,M. & H. van Balen I969. ProlongedlactationandfamilyspacinginRwanda.J.biosocialSci.

I, 97-I00.

E. Akingeneye,M. Bashakauba,E. Nabarutso & M. Nolans I974. Influenceof thesocioeconomiclevel on theconceptionrateduringlactation.Int.J.FertilityI9, 97-I02.

Borgerhoff-Mulder,M. & M. Milton i985. FactorsaffectinginfantcareinKipsigis.J.anthrop.Res.4Ir 23 I-62.

Bromage,T. G. & M. C. Dean i985. Re-evaluationoftheageat deathofimmaturefossilhominids.Nature,Lond.317, 525-7.

Bronson,F. H. I980. Theadaptabilityofthehousemouse.Sci.Am.250, I i6-25.

Buchanan,R. 1975. Breastfeeding:aid to infanthealthand fertilitycontrol.Popul. Rep. J. 4,49-68.

Chowning,A. I972. Childrearingand socialization.InEncyclopediaofPapuaNew Guinea(ed.) P.Ryan.Melbourne:Univ. Press.

Cohen, M. N. I980. Speculationson the evolution of densitymeasurementand populationregulationinHomo sapiens.In Biosocialmechanismsofpopulationregulation(eds) M. W. Cohenet al. New Haven:Yale Univ. Press.

Denham,W. D. I974. InfanttransportamongtheAlyawaratribe,CentralAustralia.Oceania44,253-77-

Deevey, E. I968. Pleistocenefamilyplanning.In Man thehunter(eds) R. B. Lee & I. DeVore.Chicago:Aldine.Draper,P. I976. Socialandeconomicconstraintson childlifeamongthe!Kung.InKalaharihunters

andgatherers(eds)R. B. Lee& I. Devore. Cambridge,London:HarvardUniv. Press.Dumond, D. E. I972. Prehistoricpopulationgrowthand subsistencechangeinEskimo Alaska. In

Poptulationgrowth:anthropologicalimplications(ed.) B. Spooner. Cambridge,Mass.: M.I.T.Press.

I975. The limitationof humanpopulation:a naturalhistory.Sciencei87, 7 I3-20.

Dwyer, P. D. I974. The priceof protein:five hundredhours of huntingin theNew Guineahighlands.Oceania55,278-93.

Fischer,H. E. I982. Thesexcontract:theevolutionofhumanbehavior.New York:W. Morrow.Fossey,D. I979. Developmentof themountaingorilla(Gorillagorillaberingei).In Thegreatapes

(eds)D. Hamburg& E. McCown. Menlo Park:Cummings.I983. Gorillasinthemist.Boston:HoughtonMifflin.

Frisch,R. E. I975. Criticalweights,a criticalbody composition,menarche,and themaintenanceofmenstrualcycles.In Biosocialinterrelationsinpopulationadaptation(eds) E. S. Wattset al. TheHague:Mouton.

I977. The effectoffood intakeon reproductiveability.In Nutritionandhumanreproduction(ed.) W. H. Mosely.London: PlenumPress.



A. C. ZELLER 555

I978. Population,foodintakeand fertility.Science199, 22-30.

1982. Malnutritionand fertility.Science215, I272-3.

Frisch,R. E. & J. W. McArthurI974. Menstrualcycles: fatnessas a determinantof minimumweightforheightnecessaryfortheirmaintenanceoronset.Sciencei85, 949-51.

Galdikas, B. I978. Orangutan adaptationat Tanjung Puting Reserve,centralBorneo. Thesis,UniversityofCalifornia,Los Angeles.

I979. Orangutanadaptationat Tanjung PutingReserve:matingand ecology. In Thegreatapes(eds)D. Hamburg& E. McCown. Menlo Park:Cummings.

I98I. Orang utanreproductionin thewild. In Reproductivebiologyofthegreatapes(ed.) C. E.Graham.New York:AcademicPress.

Ghiglieri,M. P. I984. ChimpanzeesoftheKhibaleforest.New York: ColumbiaUniv. Press.Van Ginneken,J. K. I977. The impactof prolongedbreastfeedingon birthintervalsand on

post-partumamenorrhea.In Nutritionandhumanreproduction(ed.) W. H. Mosely. London:PlenumPress.

Goodall,J.van Lawick I973. Culturalelementina chimpanzeecommunity.In Preculturalprimatebehavior(ed.) E. W. Menzel. Basel:Karger.

I979. Lifeanddeathat Gombe.Nat.Geogr.May 1979, 59i-620.

Gordon,R.J. I984. The !KungintheKalahariexchange:anethnohistoricalperspective.InPastandpresentinhuntergatherersocieties(ed.) C. Schrive.London: AcademicPress.

Harcourt,A. H., D. Fossey,K.J.StewartandD. P. WattsI980. Reproductioninwild gorillasandsomecomparisonswithchimpanzees.J. reprod.Fert.Suppl. 28, 59-70.

D. Fossey&J. Sabater-PiIg8Ia. DemographyofGorillagorilla.J.Zool. i95, 2I5-33.

K.J. Stewart& D. FosseyI98 ib. Gorillareproductioninthewild.In Reproductivebiologyofthegreatapes(ed.) C. E. Graham.New York: AcademicPress.

Harris,D. R. I978. The originsofagriculturein thetropics.In Readingsinphysicalanthropology(eds)D. E. Hunter& P. Whitten.New York:Harper& Row.

Harpending,H. I976. Regionalvariationin Kung populations.InKalaharihuntersandgatherers(eds)R. B. Lee & I. DeVore. Cambridge,Mass., London: HarvardUniv. Press.

Hasson, F. A. I980. The growthand regulationof human populationin prehistorictimes.InBiosocialmechanismsofpopulationregulation(eds)M. W. Cohen et al. New Haven: Yale Univ.Press.

Hockett,C. F. & R. Ascheri965. The humanrevolution.Curr.Anthrop.5, I3 5-I52.

Holloway,R. L. I976. Paleoneurologicalevidenceforlanguageorigins.InOriginsandevolutionandlanguageandspeech(eds) S. R. Harnadet al. (Ann.N.Y. Acad. Sci. 280). New York:AcademyofScience.

Howell,N. I976a. Uniformitariantheory.In Thedemographicevolutionofhumanpopulations(eds.) R.Ward& K. Weiss.New York: AcademicPress.

I976b. ThepopulationoftheDobe area!Kung.In Kalaharihunter-gatherers(eds)R. B. Lee &I. DeVore. Cambridge:Univ. Press.1979. DemographyoftheDobe!Kung.New York:AcademicPress.

Hrdy,S. B. & G. C. WilliamsI983. Behaviouralbiologyand thedoublestandard.In Socialbehaviorinfemalevertebrates(ed.) S. K. Waser.New York: AcademicPress.

Huss-Ashmore,R. I980. Fat andfertility:demographicimplicationsof differentialfatstorage.YB.phys.Anthrop.

Irons,W. I979. Culturalandbiologicalsuccess.In Evolutionarybiologyandhumansocialbehaviour(eds)N. A. Chagnon& W. Irons.Massachusetts:DuxburyPress.

I983. Human femalereproductivestrategies.In Socialbehaviorinfemalevertebrates(ed.) S. K.Wasser.New York: AcademicPress.

Jain,A. K. & T. H. Sun I972. Theinterrelationshipsbetweensocio-demographicfactions,lactation

andpost-partumamenorrhea.Demogr.Ind.I, 3-I5.Jain,A. K. & T. BongaartsI98I. Breastfeeding:patterncorrelatesandfertilityeffects.Stud.Fam.

Plan. 12, 79-99.

Johnson,L. L. I978. Humanevolution:apostulatedsequence.New York: M. S. VassarCollege.Johanson,D. & M. Edey I98I. Lucy:thebeginningsofhumankind.New York: Simon & Schuster.Katz, S. H. I972. Biological factorsin populationcontrol. In Populationgrowth:anthropological

implications(ed.) B. Spooner. Cambridge,Mass.: M.I.T. Press.



556 A. C. ZELLER

Kennedy,K. A. R. I975. Neanderthalman.Minneapolis:Burgess.Kippley,S. I974. Breastfeeditnganidntaturalchildspacing.New York:Harper& Row.Knauer,M. J. I984. Breastfeedingpatternsandpost-partumfertilityinurbanCanadian women.

Thesis, Universityof Toronto.

Kolata, G. I974. Kung huntergatherers:feminism,dietand birthcontrol.ScienceI85, 932-4.Konner,M. & C. WorthmanI980. Nursingfrequency,gonadalfunctionand birthspacingamong

!Kung hunter-gatherers.Science,207, 788-9 I.

Laska-Mierzejewska,T. I975. Effectof ecological and socio-economic factorson the age atmenarche,body heightand weightof ruralgirlsinPoland. InMasnandtnatuire(ed.) F. Hulse.New York: RandomHouse.

Lambeck,M. I985. Motherhoodand othercareersinMayotte.In In herprime(eds)J. Brown & V.Kerns. Massachusetts:Bergin& Garvey.

Lancaster,J. & B. King. 1985 An evolutionaryperspectiveon menopause.In In herprimne(eds)J. Brown and V. Kerns.Berginand Garvey,Massachusetts.

& C. S. LancasterI983. Parentalinvestment:theHominidadaptation.In How humanadapt:abioculturalodyssey(ed.) D. Ortner.Washington:SmithsonianInstitution.

Laughlin,W S. I968a. Hunting:an integratingbiobehaviorsystemand itsevolutionaryimport-ance.In Mantthehunter(eds),R. B. Lee& I. DeVore. Chicago:Aldine.

I968b.The demographyofhunters:anEskimo example.In Man thehunter(eds) R. B. Lee &I. DeVore. Chicago:Aldine.

Lee, R. B. 1972. Population growth and the beginningsof sedentarylife among the !KungBushmen.InPopulationgrowth:anthropologicalimnplications(ed.) B. Spooner.Cambridge,Mass.:M.I.T. Press.

1979. The !Kung Sani: meni,womenanidworkin a foraginigsociety.Cambridge: Univ.Press.

1980. Lactation,ovulation,infanticideand women's work: a study of huntergathererpopulationregulation.In Biosocialmechanismsofpopulationregulation(eds) M. W. Cohen etal.New Haven, London: Yale Univ. Press.

Leridon,H. 1977. Humanfertility.Chicago:Univ. Press.LeVine, R. A. & B. B. LeVine I963. Nyahsongo:a GusilcommunityinKenya. In Six cultuires:

studiesofchildrearing(ed.) G. G. Whiting.New York:JohnWiley.Lovejoy,C. 0. I980. Hominidorigins:theroleofbipedalism.Am.J.phys.Anthrop.52, 250.

Lunn,P. L., S. Austin,A. M. Prentice& R. A. WhiteheadI984. The effectofimprovednutritiononplasmaprolactinconcentrationandpostpartuminfertilityinlactatingGambianwomen.Am.J.clin.Nutr.39,227-35.

Mackinnon,J.I979. Reproductivebehaviorinwildorangutanpopulations.InThegreatapes(eds)P.Hamburg& E. McCown. Menlo Park:Cummings.

Mann, A. E. 1975. Paleodemographicaspectsof the South AfricanAustralopithecine(Publ.Anthrop.i). Philadelphia:Univ. ofPennsylvaniaPress.

McDowell,N. I98I.

Reproductivedecision

makingand value of childrenin rural

PapuaNew

Guinea. Papua New Guinea Instituteof Applied Social and Economic Research.ReportpreparedforthePopulationProgramof theOfficeof EnvironmentandConservation.

Minturn,L. &J. T. HitchcockI963. The RajputsofKhalopur,India. In Six cultures:studiesofchildrearing(ed.) B. B. Whiting.New York:JohnWiley.

Misra,R. I979. Breastfeedingandweavingin two Indianvillages.InBreastfeedingandfoodpolicyinahungryworld(ed.) D. Raphael.New York: AcademicPress.

Munroe, R. H., R. L. Munroe & H. S. Shimmin I984. Children'swork in our cultures:determinantsand consequences.Am.Anthrop.86, 369-79.

Nag, M. G., N. F. Whitehead& R. C. Peet I978. Ananthropologicalapproachto thestudyoftheeconomicvalue of childreninJavaandNepal. Curr.Anthrop.19, 293-306.

Neel,J.U. I983 . Some baselinesforhumanevolutionandthegeneticimplicationsofrecentcultural

development.In How humansadapt:a bioculturalodyssey(ed.) D. Ortner.Washington:Smith-sonianInstitution.Nydegger,W. F. & C. NydeggerI963. Tarong:An Illocasbarrioin thePhilippines.InSix cultures:

studiesofchildrearing(ed.) B. B. Whiting.New York:JohnWiley.Ripley,S. I980. Infanticideinlangursandman:adaptiveadvantageor socialpathology?In Biosocial



A. C. ZELLER 557

mechanismsofpoptilationregulation(eds)M. W. Cohen etal. New Haven,London: Yale Univ.Press.

Robson,J. R. K. & D. E. Yen I980. Some nutritionalaspectsof the PhilippineTasaday diet. InEcologyandculture(ed.)J. R. K. Robson. New York: Gordon& Breach.Romney,K. & R. RomneyI963. The MixtecansofJuxtlahuaca,Mexico In Six cultures:studiesof

childrearinig(ed.) B. B. Whiting.New York:JohnWiley.Saucier,J. F. I972. Correlatesof thelong post-partumtaboo:a cross-culturalstudy.Curr.Anthrop.

13, 238-49.Scrimshaw,N. S. I983. Food, past present,and future.In Howhomninidsadapt:a bioculturalodyssey

(ed.) D. Ortner.Washington:SmithsonianInstitution.Short,R. V. I976. The evolutionof humanreproduction.Proc.R. Soc.Lond.B195, 3-24.

I984. Breastfeeding.Sci.Am.250(4), 35-4I.Shostak,M. I976. A !Kungwoman'smemoriesofchildhood.In Kalaharihuntersandgatherers(eds)

R. B. Lee& I. DeVore. Cambridge,Mass.: HarvardUniv. Press.

Silk,J.B. I978. Patternsof foodsharingamongmotherand infantchimpanzeesatGombe NationalPark,Tanzania.Foliaprimatol.29, I29-4I.

I979. Feeding,foragingand sharingbehaviorof immaturechimpanzees.Foliaprirnatol.31,

I25-42.

Stein,Z. & M. SusserI977. Famineand fertility.In Nutritionand humanreprodtuctiotn(ed.) W. H.Mosely. London:PlenumPress.

Sussman, R. I972. Child transport,familysize and increasein human population duringtheNeolithic.Curr.Anthrop.13, 258-69.

Tanaka,J. I976. Subsistenceecologyof centralKalahariSan. In Kalaharihuntergatherers(eds) R. B.Lee & T.DeVore. Cambridge,Mass.: HarvardUniv. Press.

Tanner,N. i98i. Onbecominghuman.Cambridge:Univ. Press.Teleki, G., E. E. HuntJr.&J. H. PfifferlingI976. DemographicobservationsI963-I973 on the

chimpanzeesofGombe NationalPark,Tanzania.J. hum.Evol. 5, 559-98.Townsend, P. K. I980. New Guinea sago gatherers:a study of demographyin relationto

subsistence.InFood,ecologyandculture(ed.)J. R. K. Robson.New York: Gordon& Breach.I985. Infantmortalityin theSaniyo-Hiyowepopulation,Walio Sio Census Division, East

SepikProvince.PapuaNew Guineamned.J.28(3), I77-82.

Trivers,R. L. I972. Parentalinvestmentand sexualselection.In Sexualselectionandthedescentofman1871-1971(ed.) B. Campbell. Chicago: Aldine.

Truswell,S. A. &J. D. L. HansenI976. Medical researchamong the Kung. In Kalaharihuntersandgatherers(eds)R. B. Lee & I. DeVore. Cambridge,Mass.: HarvardUniv. Press.

Tutin, C. E. G. I980. Reproductivebehaviourof wild chimpanzeesin theGombe National Park,Tanzania.J.Reprod.Fert.Suppl.28, 43-57.

Tutin,C. G. & P. R. McGuinisI98I. Chimpanzeereproductionin thewild. In Reproductivebiologyofthegreatapes(ed.) C. E. Graham.New York:AcademicPress.Washburn,S. L. I959. Speculationson theinterrelationsof thehistoryof tools and biological

evolution.In Theevolutionofman'scapacityforculture(ed.)J.N. Spuhler.Detroit:WayneStateUniv. Press.

I98I. Longevityin primates.In Aging,biologyandbehavior(eds) J. March &J. McGough.New York: AcademicPress.

& J. S. LancasterI968. The evolutionofhunting.In Man thehunter(eds) R. B. Lee & I.DeVore. Chicago:Aldine.

White,B. I975. The economicimportanceofchildreninaJavanesevillage.InPopulationandsocialorganization(ed.) M. Nag. TheHague:Mouton.

Whiting,B. B. I963. Six cultures:studiesofchildrearing.New York:JohnWiley.

&J. W. M. WhitingI975. Childrenofsix cultures.Cambridge,Mass.: HarvardUniv. Press.Wood, J. W., P. L.Johnson& K. L. CampbellI985. Demographicand endocrinologicalaspectsof

low naturalfertilityinhighlandNew Guinea.J.biosocialSci. 17, 57-79.Wray,J.D. 1977. Maternalnutrition,breastfeedingand infantsurvival.In Nutritionand human

reproduction(ed.) W. H. Mosely. London: PlenumPress.Zihlman,A. L. I98I. Women as shapersof thehumanadaptation.In Womanthegatherer(ed.) F.

Dahlberg.New Haven:Yale Univ. Press.

Documents

A Role for Children in Hominid Evolution