Infanticide and the social status of children

InfanticideandthesocialstatusofchildrenAsreflectedinchildburials

MarliesKonijnenberg2103397BAThesisArchaeologyDr.J.P.CrielaardJune16th,2014Wordcount:9893

1

TableofcontentsIntroduction 2

Chapter1 4

Chapter2 11

Chapter3 16

Conclusion 20

Literature 22

Figuresandtables 24

Listoffiguresandtables 28

2

Introduction

“(...)Creusagavebirthinthehousetoachild,andbroughttheinfanttothesamecavewherethegodhadbeddedher,andthereexposedhimtodieintheroundcircleofahollowcradle,(...)gavehiminchargetothedaughtersof Aglauros; from which the Erechthidae have a custom to rear theirchildren in gold serpents.Ornamentswhich the girl had shehung aroundherson,andlefthimtodie.AndPhoebus,asmybrother,askedmethis:“Obrother,gotothenative‐bornpeopleofgloriousAthens,(...)takethenew‐bornbabyfromthehollowrock,(...)bringhimtomyshrineatDelphi,andplacehimattheveryentranceofmytemple;Therest—knowthatthechildismine—will bemy care.” (...) I tookup thewovenbasket andbrought ithere,andplacedtheboyatthebaseofthistemple,openingupthewreathedcradle,sothattheinfantmightbeseen.1”

A lot has been written about the subject of infanticide and the

abandonmentofnewborns.Quotedabove isapartof the tragedy ‘Ion’,writtenbytheGreekpoetEuripides.Theplaywasprobablyperformedaround413B.C.IttellsusaboutCreusa,daughterofanAthenianking,andhersonIon.ThegodApollo raped Creusa and after giving birth she brought her newborn son to acave,whereshelefthimtodie.AthisfathersrequestIonissavedandbroughttothe Apollo temple in Delphi. Here he grows up as a temple servant, and aftermanyyearsheisreunitedwithhismother.

Ancient tragedies,suchasEuripides’ Ion, tellstoriesofmothersdriventogruesomeacts,andthesestoriescouldgivetheimpressionthatinfanticidemighthaveoccurredonaregularbasisintheGreekandRomanworld.However,thereis a problemwith this assumption. Somemodern scholars are now convincedthatalthoughthethemeofinfanticideandexposurewasoftenusedasadramaticelement in ancient literary sources, exposure or infanticide in fact was veryexceptional.2Therearesomearchaeological sources that canbe interpretedassuch,butnotbyfarasmanyasthehistoricalsourcesletusbelieve.Eventhougha phalanx of scholars has tried to demonstrate that infanticide and exposureindeeddidoccur,itremainstobeatopicofdiscussion.

The invisibility of child graves and poorly preserved bones of childrenform a problem within the field of archaeology, and even more so for theresearchof infanticide.Whatcomplicatesmattersevenmore, is thatvictimsofinfanticideprobablydidnothavea formalburialatall(that is tosaya funeral,and/oraburial),andthismakesitextremelyhardtofindthemiftheyexistatall.

Ancient Greek and Roman authors often speak of female infanticide,suggestingthatgirlswereusuallyvictimsofinfanticide.Becausetherearesofewcasesofinfanticideknowninthearchaeologicalrecord,thereisnoevidencethatcorrespondswiththeancienttexts.Sinceyoungchildrenoftenwerenotseenasfullindividualsyet,theydidnotreceivesexspecificgravegoods.3Soeveniftheywereburied, the contextof gravesdoesnotalwaysgive informationabout thesexofthechild.1Eur.Ion15‐40:http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A1999.01.0110%3Acard%3D1,consultedonFebruary,32014.2L.R.vanHook1920,140‐141;Engels1980,112‐113.3Beaumont2012,89.

3

Historical and archaeological sources should both contribute to theresearch, tounderstandthepastascompletelyaspossible,andtoruleoutanydiscrepancies in texts that arewrittenwitha certainpolitical awarenessoranarchaeologicalexceptiontotherule.However,sinceinfantburialsoftenarenotrecognizableanymore,andvictimsofinfanticideprobablydidnotreceiveformalburial,archaeologicalevidence isveryhard to find in thisparticularcase.Nowscholars seek other ways to explore the possible historical correctness of theancientsources.

The goal of this paper is to create a general overview of the discussionamongst scholars about infanticide and child abandonment, and to discuss themethodsthatscholarshavecomeupwithtoproofwhetherinfanticidedidordidnot tookplace.Theexamples thatarementioned in thispaperwill range fromthe archaic period till the Late Roman period, when law forbade infanticide.SinceitwassupposedlyacustomoftheGreeksaswellastheRomans,sitesfromAthensandMessenewillbeaddressed,butalsoRomansitesinGreatBritainandinIsraelwillbedealtwithascasestudies.

In the following chapters an answer will be provided to the question ifinfanticideactuallydidhappen,andifso,iftheancientscaredfortheirchildren.Inotherwords,whatwas the social statusof children? In the first chapterwewillhaveacloser lookat thediscussionbetweenscholars,about thesubjectofinfanticide and exposure. The different methods that are being used todemonstrate whether infanticide indeed took place will be addressed. In thesecond chapter wewill see if archaeological sources can add anything to thisdiscussion.Archaeologicalsitesthatcanbeinterpretedassuchwillbediscussed.The thirdchapterwilldealwith thegeneralsocialstatusofchildren inancientAthens,andthedevelopmentsweseeinthegravematerial.Theirsocialstatusinthe ancient society of Athens, should give us an idea if infanticide could havebeen occurring at all. Thediscussion of status andburial customs for childrenwill be limited toAthens, becausemost sources reflectAthenian life. The finalchapterwillcontainalltheconclusionsofthispaper.Ananswertothequestionwhetherinfanticidehappenedinancientsociety,inrelationtothesocialstatusofchildren,willbesummarizedhere.

4

Chapter1

Inancientliteraturewefindalotofindicationsthatinfanticideandchildabandonment took place.What cannot be deduced out of literature is if thesecustomswerecommonpractice,orjustaliterarytopos.Primarysourcesdealingwith infanticideandabandonmentcanbedividedintotwocategories.Thefirstconsists of the many myths and legends in which children, mostly boys, areexposed,andNewComedy.Inthelattergenretheplotusuallyconcernsayoungmalecitizen,whofallsinlovewithanon‐citizengirl.Attheendoftheplaysheturnsouttobetheexposedandlong‐lostdaughterofcitizenparents,andall iswell in the world again. The second category consists of the stemmata andinscriptions.4 Aswith somany subjects in ancient history, sources are scanty.Thiscreatesasituationwhereevidence,takenoutofitscontext,canbethebasisforcontradictoryconclusions.

Inthischapterwewillhaveacloserlookattheacademicdiscussionaboutthe subject of infanticide. The differentmethods that they use to demonstratewhether infanticidedidordidnot takeplace, andwhy itmighthaveoccurred,willbeconsideredhere.

In thenineteenthcentury thediscussionabout infanticide,wascast inamoral framework contrasting the Judaeo‐Christian ethics with the antiquemorals. The twentieth century historians questioned whether infanticide hadoccurredatall,concentratingprimarilyon5thand4thcenturyAthens.Thedebatefocusedonthemisogynistconnotationsof femaleinfanticide.Scholarsattackedboththeconsensusthatfemaleinfantsmoreoftenfellvictim,andthebeliefthatthepowerofthefatherwasabsolute.Anothersignificantchangeintheattitudetowardsthissubjectwasthathistoriansgrewsuspiciousoftheancient literarytexts,andnowcametorelyondemographicargumentsinstead.5

ForGustaveGlotztheexistenceofinfanticideasaformoffamilyplanningwas simplya factof life inantiquity.Atheniancitizensonlyneededone son toguaranteethecontinuationofanoikos.Girlswereafinancialburden,sincetheyhad to be provided with a dowry, and marriage meant the end of parentalauthority, thus excluding the possibility of Greek fathers exercising influencethroughadvantageousmarriagealliances.Thisprovidedamotiveforinfanticide.Furthermore, because the Greeks refused to abolish slavery, many foreignershadtobeimportedasslaves.Sincelandandresourceswereavailableinalimitedamount, the resulting demographic pressure meant that many Greek nativenewbornshadtobeexposedonalargescale.6Insum,inGlotz’viewinfanticidewas caused by both economic and demographic pressure, and happened on alargescale.

La Rue van Hook, H. Bolkestein, and A.W. Gomme on the other hand,rejected theconsensus that infanticidewasaprevalentpracticeamong5thand4th century Athenians. They believed that scientific research could yieldempirical knowledge, and tried to isolate thephenomenonper timeandplace.Whentheytriedtoexcludethemultiplemeaningsofliterarysources,theygrew4Oldenziel1987,88.Forexample,Milete3(asdiscussedinPomeroy1993,209‐216)onwhichlistsofimmigrantsfromCreteinMiletusareinscribed.Theselistsshowalownumberoffemales,incomparisontothenumberofmaleslisted.5Oldenziel1987,88.6Glotz1892,930‐939.

5

suspiciousofitsliteraryfunction(inotherwords,wasinfanticideonlyaliterarytopos?). Philological arguments were replaced by demographical ones, andantiquity was divided up into specific times and places. The absence ofinfanticideintheliterarysourcestheystudiedwasaforegoneconclusion,simplybecause the narrow view taken of the isolated source material excluded anyrelevantsources.Asaresultofthisapproach,theabsenceofsourcesbecameanargumentfortheabsenceofinfanticideinthatspecifictimeandplace.7

VanHookradicallydismissedallofthesourcesthatwerecommonlyusedinthediscussionon infanticideasvalidevidence.Herejectedthereferences inmythicalnarratives,butalsotheonesintheworksofPlatoandAristotle.Inhisopinion they all referred to heroic measures for the limitation of population,“whichmightbe resorted to in their imaginarypolitics8”.NewComedydidnotdescribe Menander’s Athens, but rather reflected Plautus’ Rome. He thoughtliteraturecouldneverbeareflectionofhistoricalreality,becauseitonlyreflectswhatisrareandexceptional.9

Bolkestein also dismisses the possibility that infanticide occurred. Heargues that the Greekword enchutristriai referred to womenwho exposed ofchildren. The theory that newborn babieswere exposed in a jar (chutra), andthatthepracticewassoprevalent,thatagroupofwomenexistedsolelytointeryoung children in jars (and so received the title ‘bottlers’) rests on confusion,according to Bolkestein. He believes that the word enchutristriai, which hasalwaysbeenexplainedasaderivation from ‘ jar’ (chutra), actually is rooted in‘sacrificialpitduring funeral ceremony’ (chutros).Thewomen inAristophanes’Wasps 289, Frogs 1190 and Thesmophoriazousai 502ff, served during thechutrizein,wheretheysacrificedananimalandthrewpartsofthesacrificeintoapit.Itisonthedifferencebetweenchutraandchutrosthathebaseshisassertionthatexposuredidnothappenin5thand4thcenturyAthens.10AsadmirersoftheAthenian cultural and political achievements of these centuries, it seemedimpossible tovanHookandBolkesteinthat infanticideoccurred in thisperiod,and they argued that the practice could be restricted to the Hellenistic andRomanperiod.11

Gomme was the first to dissent from the consensus that if and wheninfanticideoccurred,itwouldhavebeenprimarilyfemaleinfanticide.Hearguedthat it would be inconceivable to assume that Athenian women were treatedcontemptuously,andheconsidereditimpropertosupposethatAthenianfathersthought of their daughters as financial burdens. Economicmotives alonewerenot enough reason for them to practice infanticide. He first presented acalculation inwhichhesupposed theexistenceofapopulation increase for5thcenturyAthens.Heassumedanormalmortalityrate,andthatahighbirthratewas likely. Because of this high birth rate in combination with a populationincrease,heassertedthatinfanticidecouldnothaveexistedin5thand4thcenturyAthens.12

Inaddition,hearguedthatevenifinfanticidedidoccur,itcouldnothave7Oldenziel1987,92‐93.8VanHook1920,143.9VanHook1920,139‐140.10Bolkestein1922,223‐225.11Oldenziel1987,97.12Gomme1933,75‐83.

6

been biased towards females, since female infanticide would have left theAthenianmenwithanimpossiblysmallnumberofmarriagepartners.13

Following Gomme’s reasoning, Donald Engels concluded that femaleinfanticide constituted a demographic impossibility. Engels argues that thepracticeofinfanticidewasofnegligibleimportanceinGreekandRomansociety,and a high rate of female infanticide was demographically impossible. Heexplainsthatosteologicalevidenceconcerningthesexratioandagestructureofancientpopulations is systematicallybiasedagainst femalesand infants.Bonesof females and infants survive less often, and excavators are more likely todiscardsmaller,brokenspecimens.Also,fewinfantswhodiedwithinafewdaysor weeks after birth receive standard burials in cemeteries. Since infantmortalitywasprobablyhighinantiquity,thisdistortstheage‐structurederivedfrombonematerialevenfurther.14

Besides all this, Engels mentions that unless the entire skeleton isavailableforanalysis,theprobabilityofcorrectsexclassificationmaybeaslowas20%.Hethenenumeratessomeageclassificationmethods(suchasanalysisoftooth wear and cranial sutures) before coming to the conclusion thatdetermination of sex ratios and age structures of ancient populations, derivedfromosteological evidence, are unreliable, nomatter howmany skeletons onelooksat.15However,hisargumentseemsincomplete,sincehedidnotdiscussanysex classification methods, such as examining the pubic bone. As a sexclassificationmethodthisisveryreliable,butonlyforadults.

Engels thendiscussessomedemographicprinciples, toshowthatahighrateoffemaleinfanticidewouldnothavebeenpossible.First,hestatesthatthesex ratio for newborn infants is about 1 female for every 1.05 males in allpopulations (and then uses a ratio of 1:1). Second, he argues that the rate ofnatural increase of any population (excluding immigration and emigration) isbased on the difference between its birth and death rate. The populationwillgrowwhentheformerishigherthanthelatter,butifthelatterishigher,itwilldecline.Therateofnaturalincreaseinantiquitywassmallorevennonexistent.Thehighestaveragepopulationgrowthrateinantiquitywasprobablylittlemorethan1per1,000peryear, andEngels assumesa stablepopulation for ancientsocieties,oneinwhichthedeathratewasinequilibriumwiththebirthrate.16

Furthermore, when a society’s average life expectancy is low, theproportion ofwomen surviving to themean age of reproduction is small. Theamount of women surviving to childbearing age can be calculated when thesociety’s average life expectancy at birth is known. The consequences that agivenratewouldhaveforthesizeofthepopulationcanbeestimated.If20%ofall female births per year were killed, this would have two effects on thepopulation: the population’s death rate would increase, and the population’sbirth rate would be reduced with one‐fifth. It would yield a negative rate ofnatural increase, and thepopulationwould suffer drastic declines in numbers.Forthesereasons,Engelsarguesthatevenarateofkilling10%ofallnewbornfemales per year, would be highly unlikely, and the rate would have never

13Oldenziel1987,98.14Engels1980,112‐113.15Engels1980,113.16Engels1980,114‐116.

7

exceededmorethanafewpercentoffemalebirthsinanyera.17To sum up, Engels points out that the preserved and excavated bone

material is biased towards males, and should be considered unreliable.Demographic principles show, according to Engels, that infanticide could nothavebeenpracticedatalargescalesincepopulationdeclinewouldoccur.

W.V.HarrisandM.GoldenbothofferedareplytoEngelsbypointingoutthat one could well imagine alternative marriage, birth and death patternsperfectly consistent with population reproduction, while simultaneouslyallowing for the presence of infanticide. In addition, Golden suggested whyfemaleinfanticidedidnotcauseashortageofbrides(asGommeclaimedbefore).Heevenarguesthatwithout female infanticidetheremighthavebeenanover‐supplyofbrides,whichwillbediscussedlater.18

W. V. Harris argues against Engels that the exposure of infants, oftenresultingindeath,wascommoninmanypartsoftheRomanEmpire.Further,hestatesthatthereprobablywasnocontinuouspopulationstability,sothenaturalpopulation growthwas not a stable figure.19 According toHarris, peopleweretolerating fatal childexposurewhen they thought that toomanychildrenwerebeing born, or too many of one sex. Some, if not all, of these periods willcorrespondwithperiodswithgenuinelyhighnatural increase. In suchperiodstheoveralleffectofextensiveinfanticidemayhavebeengreatorsmall.Great,ifitreduced swift population growth to stability, or changed stability into swiftdecline(inthelattercase,whichistheoneespousedbyEngels,wewouldexpectinfanticidetodecreaseafteracertainperiod),orsmall,ifthefactorsdeterminingfertility reacted to the act of child exposure (a decrease in the number ofabortionsandearlierageoffemalesatmarriage).20

Astothefactorsdeterminingfertilityinasociety,theyincludetheageatwhichitsfemalemembersbegintopracticecoitus,coitalfrequency,fertility,useof contraception, and foetal mortality. These factors are quite immeasurable,exceptforone:intheRomanperiodtheageofgirlsatmarriagewaslow,namelyfavourable to high fertility. Harris’ opinion is that this was caused by theexposureofinfants.21 AccordingtoHarris itmaywellhavehappenedthatsomeregionsoftheRoman Empire witnessed rates of infanticide notably higher than thepercentages that are discussed by Engels, and that this happened withoutsufferingabruptpopulationdecline.Also,EngelswritesthatearlierstudiesfromCameronandBolkesteinhaveestablishedthatthepracticeofinfanticidewasofnegligibleimportanceinGreekandRomansociety.22However,Harrisdismissesthis,sinceEngelsbasedhisopiniononworksfromthetwentiesandthirties,butvery conveniently left their latestwork out, inwhich Cameron andBolkesteinbothconcludedthat infanticidewas in factcommon.23AdemographicstudybyDickeman points out that infanticide occurred at rates of 5‐50% in hunter‐gatherers, horticulturalists, and stratified agrarian societies. Following the17Engels1980,118‐120.18Oldenziel1987,99.19Harris1982,114‐116;Golden1981,318.20Harris1982,115.21Harris1982,115.22Engels1980,112.23Harris1982,116;Bolkestein1922,238;Cameron1932,105.

8

argumentsoftheseauthors,thereisnotheoreticalreasonwhyitcouldnothaveoccurredamonggirlsintheGraeco‐Romanworld.24 Mark Golden also disagrees with Engels’ arguments. Engels states thatfemaleinfanticidecouldnothavetakenplaceinanypopulationwithalowrateofnaturalincrease,whichisbasicallyeveryprimitiveorpre‐industrialpopulation.Golden looks at a study of ethnographical reports on 393 widely scatteredcultures, and comes to the opposite conclusion: infanticide, especially femaleinfanticideisprevalentinsuchpopulations.179oftheseculturescommonlyand29occasionallypracticedinfanticide,predominantlyofgirls.25 Golden also stresses that Engels did not take into account social andcultural measures, when explaining the slow rate of population growth inantiquity. Engels focused on matters such as lack of food and disease, whichcould cause this low rate of increase. Yet cultural measures such as abortion,contraception,segregationofwomenafterchildbirth,prolongedbreast‐feeding,concubinage, postponed marriage and infanticide, are not even mentioned. InGolden’s opinion, the low rate of increase in the ancient population is at leastpartly a result of conscious efforts in population control, including infanticide.The equilibrium between birth and death rate (of which Engels thinks thatinfanticidewouldoutbalancethiswithdisastrousconsequences)isitselfpartlyaproductofinfanticide.26 One specifically important cultural variable that should be taken intoaccount,accordingtoGolden,ismarriagepatterns.Goldenclaims,againstEngels,that exposure did happen in Athens on a larger scale. He notices that thereprobablyweremorewidows thanwidowers, and that thesewomen often gotremarried,becauseotherwisetheylackedakyrios(andbecausetheyhadfewerrightsthanmales).Becauseofthis,therewasasurplusofyoungunmarriedgirls,what he calls a ‘marriage‐squeeze’. The solution for this problemwould havebeeninfanticide.Goldenthinks10%ormorewasverywellpossible.27 TakingthelifeexpectancyrangeatAthensintoaccount,Goldenassumesthat one in every five or six Athenian girls would find no husband of thepreferredagewhenshereached15.Theseunmarriedwomenwouldrepresentaneconomicdrainontheindividualoikosandasocialanomalyforthepolisasawhole. The longer a girl stayed at home, the more costly it would be for herfamily, and the less likelihoodof hermarrying at all. Unmarriedwomen couldchoose prostitution, or life as a concubine. Both situationswere obviously notdesirable. Also, the Athenians charged their political opponents by preventinggirlsfromgettingmarried.Theyprovideddowriesfortheirowndaughters,andfor those of their relatives. Law required that poor women be supplied withdowries in certain circumstances, and this indicates that the community as awholefeltsomeurgencytohelpgirls findhusbands.28Toconclude, inGolden’sopinion,marriagepatterns, dowries, andunmarriedwomenwhowere seenas

24Harris1982,116.25Golden1981,319‐320.26Golden1981,320.27Golden1981,321‐330.Goldendoesadmitthatheisignoringotherdemographicfactorsinhisdiscussionofthesubject.ForthefifthcenturyinAthenstheseincludebattlecasualtiesofthe460sandthePeloponnesianwar,thefamineofthe450sandtheplague.28Golden1981,325.

9

an ‘economicdrain’ forbothoikosandpolis, couldhavebeenplausiblemotivesforinfanticide.

SarahPomeroyalsobelievesthattheburdenofprovidingdowrieswasaprobable motive for the practice of female infanticide. Pomeroy believes thatwomenoftenlivedshorterthanmales(theoppositeofwhatGoldenclaims),andamongstothercausesshementions infanticideandexposure,butalsoa ‘subtlebutpervasiveattitudethatgavepreferentialtreatmenttoboys’29.HerargumentcomesforthoutofalawattributedbyRomulusthatrequiredafathertoraiseallmale children, but only the first‐born female. A greater number of men thanwomenexisted in all social strata throughout antiquity, according toPomeroy.Shearguesthisisremarkablesincemen,ratherthanwomen,faceddeathincivilwars,whichthereforeshouldhaveresultedinagreaternumberofwomen.Thepracticeofwidespreadinfanticideshouldbethecauseforthisphenomenon.30

However, many of Pomeroy’s arguments, such as theundercommemoration of women in tombstones and bone material, and theconnection with disproportionate sex ratios, are yet refuted. Literature aboutwinningwarsandmaintaininganempire,butalsothefactthatwomen’sboneslessoftenremainedpreservedbecauseofthethicknessofthebones,biasedthebone sample towards males.31 Nevertheless, her remark about dowries oftencomesbackintothediscussionaboutfemaleinfanticide.

Ingalls counters her ideas about dowries. In his essay Wayne Ingallsarguesthatinfanticideingeneral,andofdaughtersinparticular,wasunlikelytohavehappenedatafrequentrate.Heexaminesthedowryinanefforttouncoverevidencethatmightsuggestthatitwasafactorindecisionstoexposedaughters,and claims that girls were not at greater risk of exposure than boys. Ingallsargues that Greek families were small, not because of parental choice, butbecauseofhighratesof infantandchildmortality.Heclaimsthatonly50%ofnewbornswasstillalivebyadulthoodinancientGreece.Giventhishighrateofchild mortality, he thinks parents were unlikely to have practiced infanticide(except under the most exceptional circumstances such as deformity,illegitimacy, or extreme poverty).32 Furthermore, he claims the Athenianscertainlywantedtheirchildren.Childrenwereseenasguarantorsofprosperityandhappiness, theywere toprovidenourishmentanddefence inoldage, theywere the instruments for the transmission of property, the performers ofreligiousritesassociatedwithancestorcult,butalsotocarrythehouseholdandfamily,andasamechanismforsupplyingthestatewithfuturecitizens.Aboveall,theygrievedthelossoftheirchildrenwhentheydied.Thus,heclaims,sincetheydesiredandneededchildren,andconsideringthelosstheyfeltattheirdeaths,itwould seem highly unlikely that the practice of infanticide would bewidespread.33 AgainsttheargumentsfromPomeroyandGoldenthatmostlygirlswerevictims of infanticide, Ingalls argues that Athenians wanted girls as much asboys. Even if a man had sons, there was no guarantee that they would notpredecease him. When this happened, his daughter, being the epikleros (the29Pomeroy1975,164.30Pomeroy1975,164.31Engels1980,112‐113.32Ingalls2002,246.33Ingalls2002,247‐248.

10

daughterofamanwhohadnomaleheirs),wouldprovideanheir.Whensonswere lacking, the oikos could continue through the adoption of the daughter’shusband or through the epikleros. However, since Athenian women were notallowed to hold property in their own name, she was required to marry herfather’snearestmalerelativeinordertokeepherfather’spropertyinthefamily.Ifshewasalreadymarried,andthehusbandwasnotadoptedintotheoikos,shewas required to divorce her spouse to marry that relative. In other words,daughterscouldprovideawaytoavoidhavingone’sestatepasstomoredistantrelatives.34 Summing up, according to Ingalls, infanticidewould not have occurred,because childrenwere important for anoikos to survive. About the subject ofdowriesasamotiveforfemaleinfanticide,hestressesthatthereisnoevidenceof any unwillingness of fathers to dower their daughters. After all, it was fareasier fora fathertodoweradaughter, thantosharehisestateamongseveralsons.Besidesthis,adowryalsoprovidedanopportunitytodisplaywealthandstatus.ThisiswhyIngallsthinksthatfatherstendedtobegenerous.Sogenerouseven, that Solon is creditedwith an attempt to limit their size, and also Platowould have banned dowries and limited expenses related to weddings, in hisidealstate.InIngallsopinion,thesuppositionthatfathersworriedaboutthecostwhen deciding whether or not to rear a daughter, rests on no firm evidence.Infanticidewouldnothavebeenencouragedbecauseofdowries.35

So far it looks plausible that infanticide did occur, however, in whatquantities remains the question. Primitive societies, aswell as the Greeks andRomans,mighthavesufferedahighchild‐mortalityrate,sothatatleast40%ofthenewbornchildrendiedwithinthefirstyear.Consideringthisextremelyhighdeathratioamongstchildren,thehighlevelofmortalityofwomengivingbirth,and fertilityproblemsbecauseof diseasesduring childhood, it is veryunlikelythat Greeks or Romans indeed secured their economic or social position bylimiting their offspring through extensive infanticide.36 On the other hand,argumentssuchastheoneputforwardbyHarris,stressingthatpeopletoleratedfatalchildexposurewhentheythoughtthattoomanychildrenwerebeingborn,ortoomanyofonesex,duringtimesofhighpopulationincrease,soundplausibleaswell.Alsoeconomicreasonsmighthavebeenanimportantmotivationforthisphenomenontooccur.

Despitetheappearanceofscientificaccuracyofthedemographicdebates,scholars have yet to provide a conclusive consideration on the infanticidequestionwhetheritoccurredornot.Statisticaldatafromantiquityisweak,andnotalottrustworthierthanliterarysources.Also,theaprioriargumentsaboutthedemographical feasibility of infanticide should consider anumberof socialfactors, such as marriage and fertility, birth rates, and death rates, beforedrawing conclusions about the subject.37 Archaeological sites could perhapsthrowmorelightonthissubject.Inthenextchapterwewillhaveacloserlookatthepossibleevidence.

34Ingalls2002,248‐249.35Ingalls2002,251.36Haentjens2000,264.37Oldenziel1987,100.

11

Chapter2 In thepreviouschapter thequestionwasdiscussed,whether infanticideand child abandonment inancient textswasonly a literary topos,or anactualhistoricalphenomenon.However,scholarshavebasedtheirconclusionsmainlyon literarysources, and therefore it remainsdifficult topointout if infanticideindeed occurred. Archaeological sites could add information towhat is knownaboutthissubject. Youngchildrenusuallyreceivedadifferentburialtreatmentthanadults,andincaseofabandonmenttheyprobablydidnotreceiveaformalburialatall.In some cases infants who were excluded from formal burial, were to bedeposited inside wells. At the Academy of Plato a deep pit was found, whichcontainedaround40smallamphoraewiththeburialsofsmallchildreninninesuccessivelayersdatedtothesecondhalfofthe8thandearly7thcenturyB.C.38Inthe area of the Agora of ancientMessene, P. Themelis excavated awellwhichcontained a number of infant bones, along with burial amphorae and dogbones.39 A Hellenistic well from the Athenian Agora also contained bones ofanimals,mostlydogs,andofapproximately447infants.40Someofthesechildrenwho were deposited in wells, might have been victims of infanticide. Thereasoningtowhythismighthavebeenthecase,willnowbediscussed. InMessenesomeinfantsburiedinpots,accompaniedbyfourdogburials,havebeenuncovered inside theprecinct ofHellenistic funerarymonumentK3(see figure 1) in the Gymnasium area. Along its west wall twenty‐five childburials were found. These were infant inhumations in cooking pots, calledenchytrismos(seefigure2).41Awelldepositofthe3rd‐2ndcenturyB.C.wasalsobrought to light in the agora of the city. It contained the commingled skeletalremainsof262non‐adult individuals,andbonesofseveraldogs,aswellaspotfragments and pointed amphorae. Estimation of age at death was possible bymeasuring intact bones. Long bones’ measurements included the clavicle,humerus, radius, ulna, femur and tibia. Cranial bones’measurements includedthe zygomatic and pars basilaris, and pelvic bones included the ilium, ischiumand pubis. The Minimum Number of Individuals (MNI) was calculated by therepresentationoftheleftfemur.42 Basedonthemeasurements,thenon‐adultsfallwithinthefollowingagecategories:pre‐term(lessthan37weeksofgestation),full‐term(37to42weeksofgestation),post‐term(morethan42weeksofgestation),neonatal(frombirthto 28 days), stillborn (infant born after a gestational period of 24weeks,whoshowsnosignoflife).43 Themacroscopicandradiologicalexaminationof the immature remainsrevealedno specificpathological conditions that couldhave contributed to thechildren’s early deaths. Therefore, the risks tat come with pregnancy andchildbirth, which could affect both mother and child, are considered as aplausible explanation (since there are no visiblemarks thatmight explain the38Vlachou2012,383.39Bourbou/Themelis2010,111‐128.40Rotroff/Little/Snyder1999,284‐285;Papadopoulos2000,96‐118.41Bourbou/Themelis2010,111.42Bourbou/Themelis2010,112and123.43Bourbou/Themelis2010,112.

12

cause of death). A number of factors could have affected mother and baby,namelymaternal and fetal problems. Poor health and nutritional status of themother, inadequatecareduringpregnancyanddelivery,poorhygiene,butalsoprematurebirth,andageneral lackofnewborncaremayresult indeathoftheinfant.44 Itissuggestedthatthechildburialsfoundinthewellwereinitiallyburiedinpotsandamphoraeataspecificburialground,andlaterre‐depositedintothewell.Thiswellprobablycontinued toreceivedead infantsasaknownplaceofdisposal of infants who died at or after birth.45 The re‐depositing of theseremainsdoesraisethequestioniftheancientscaredfortheirchildren,andalsoifthesechildrenalldiedbecauseofhealthproblems,orperhapsbecausesomeofthem were abandoned by their parents after birth. On the other hand, theprimary formal burials suggest theywere not simply disposed of,without anycare. Dogs were essential to the comfort and dignity of the dead, escortingthemtotheunderworld,orasguardsofHadesandsymbolsofdeath.Theyarealso reported as sacrificial animals for easing childbirth, or in funeral/ritualcontexts.Itisthoughtthatthemajorityofdogburialsinvolvedsacrificeandwereassociated with human burials.46 The large quantity of dog remains suggestschthonicritualstookplacehere,whentheinfantswereburied.Itseemspossiblethattheuseofthisprimaryburialgroundchangedovertime,andthepotburialswerere‐depositedinthewell.Thelargenumberofnon‐adultremainsmightalsosuggest that the well continued to receive dead infants as a known place ofdisposal. The latter, seemingly cruel, disposals of infantsmight have been theresult from exposure, but perhaps it could also be explained by their socialstatus. If early deaths were so common, emotional responsemight have beenminimal.47Thestatusofchildreninantiquitywillbediscussedinthefollowingchapter. TheAthenianAgorawellalsocontainedtheremainsofmanyinfantsanddogs. In 1945 J.L. Angel published a short description of the human skeletalremainsrecoveredfromthewell.Inadditiontoanadultmaleandan11‐year‐oldchild, he reported that thewell contained the remains of circa 175 infants. In1995 itwas discovered that not all the bonematerial from thewell had beensorted.Duringare‐sortingofthecollection,theestimatedminimumnumberofinfants increased dramatically to 450. The presence of the male adult in thesample, stresses the unusual character of the deposition compared to formalburials. Severe osteoarthritis throughout the joints of his body may suggestextensivedegenerativejointdisease,oralifetimeofhabitualheavylabour.48Thiscould mean that this man was a social outcast, because he suffered externaldeformities, or perhaps that he was a slave, and therefore did not receive aformalburial. Theestimationofageatdeathofeachinfantspecimenhasdemonstratedthat,whilesomecanbeclassifiedasfetusesandolderinfants,themajorityoftheindividualswereneonates.AccordingtoL.M.Little,thisfact,andthesizeofthe44Bourbou/Themelis2010,113‐114.45Bourbou/Themelis2010,115‐116.46Bourbou/Themelis2010,116‐117and124.47Bourbou/Themelis2010,115.48Rotroff/Little/Snyder1999,284.

13

collection,mightraisethequestionofhowthebodiesoftheseinfantshavecometorest inthiswell.Theremainsmayrepresentindividualswhodiedofnaturalcauses,andwhosebodiesweredisposedofinthewell,ratherthangivenformalburial.However,thepracticeofinfanticidemayalsoprovideanexplanationforthedeposit.49 Stampedamphoraehandles indicate that thedepositwasclosedaround150 B.C. By then the neighbourhoodwas abandoned, and other nearbywatersourceswere filled inat thesame time.Theuniformityof thepotterysuggeststhat thedepositaccumulatedoverashortperiodof time.Mostartefactsareoftypes familiar from domestic dumps elsewhere around the Agora, but somemight have special reference to the newborns: for example, a baby feeder, afemale herm, and a largenumber of sizeable basins. Little, Snyder andRotroffargue that the deposit was formed under normal circumstances, rather thanunderabnormalconditions,suchasepidemic,famineandsiege.Theystressthepossibilitythatchildrenwhowereexposed,couldhavebeenburiedhere.50

Theabandonedindustrialneighbourhood,inwhichthewellwaslocated,offeredasecludedspotforexposureordisposal.ItsproximitytotheAgoraalsooffered the comforting possibility that the exposed child might be found andrearedbyothers.51

Thepaleopathologicalinvestigationdemonstratedcasesofcleftpalate(acondition inwhich the two palatines in themouth are not completely joined)among a number of the infants thatwere deposited in thewell, and thus it ishighlysuspectedthatthesechildrenwerethevictimsofinfanticide.Infantsbornwithcleftpalatehadlesschanceofsurviving,sinceitwasimpossibleforthemtobe breastfed, or even to receive a feeding bottle, and there were alsosuperstitiousattributions tomalformed individuals inantiquity.52The fact thattheywere buriedwith the remains of an adultmale thatmight have sufferedfrom degenerative joint disease, could mean that these children were socialoutcastsaswell.

ThetwositesdescribedabovecanbothbedatedinHellenistictimes,butinfanticide also occurred in Roman times. Two Roman sites will be discussednow.ThefirstsiteisYewdenVillainHambledon,England.ThesecondsiteistheLateRomansiteofAshkeloninIsrael.

At Yewden Villa the remains of 97 infantswere found. The burials aremostlylocatedinanareatothenorthofthemainvillabuildings.Thisarealiesoutside the main perimeter wall, but within an outer wall. There are a goodnumberofburial locationsacrosstheyard, includingsomeinrubbishpits,andone skeletonwas placed in awall cavity. In a different location to the others,there are twobabies foundunder the floor in the secondhouse, anda furtherthreeskeletonswerefoundinapitadjoiningthethirdhouse.Overall,theinfantswereburiedinsmall,unmarkedgraves.Theburialsaremarkedwithcrossesinfigure3.Somecrossesaremultipleburials.53

OsteoarchaeologistSimonMaysisintheprocessofanalyzingthebonesof49Rotroff/Little/Snyder1999,284.50Rotroff/Little/Snyder1999,284‐285.51Rotroff/Little/Snyder1999,285.52Bourbou/Themelis2010,113.53Mays/Eyers2011,1933,enhttp://www.chilternarchaeology.com/infant_deaths.htm,consultedonMay6,2014.

14

theseinfantswho,inhisopinion,wereallkilleddirectlyafterbirth.Studyofthebones suggests that the infantswere victims of infanticide. This assumption isbasedonthemeasurementsof the longbonesofthearmsandlegs,which leadarchaeologiststobelievethatallthebabiesdiedatthesameage,rightatthetimeof birth. If the deaths had been natural, it would be expected to see someprematureinfants,somewhodiedaroundthetimeofbirth,andotherswhodiedin the weeks after birth. However, distribution was dissimilar to this naturalpattern,andwasinterpretedasevidenceforinfanticide.54

As towhy the infantswere killed directly after birth, there are severalexplanationspossible.IthasbeenarguedthattherewasabrothelontheRomansite, and theprostitutesweredisposingof theirunwantedchildren. Ithasalsobeensuggested,thattheinfantsweretheoffspringofillicitliaisonswithslaves.55Althoughotherexplanationscannotbedefinitivelyexcluded,themostplausibleexplanation for the patterns of long‐bone lengths and estimated ages at deathamongtheinfants,appearstobeinfanticide.

At Ashkelon in Israel the skeletal remains of nearly a hundred infantswere found in a Late Roman‐Early Byzantine sewer beneath a 4th centurybathhouse.Thelackofcarefulburialmaysuggesthastydisposal,resultingfromadisruptioninsociallife(epidemic,war),orcouldindicateinfanticide.Infanticideisusuallycarriedoutimmediatelyafterbirth,sovictimsofinfanticideshouldallfallwithinanarrowage‐range.56

As for the identificationof thecauseofdeath, suffocationanddrowningwere themost commonmethods used for infanticide, according to Smith andKahila.Neithermethodwouldhave leftdiagnosticsignsonthebones,althoughtheymight cause discoloration of the teeth (from blood being forced into thedentinetubules).Therefore,theosteoarchaeologistsconsidereddeterminingtheage range of the skeletal remains the best approach to use in attempting todifferentiatebetween infanticideandothercausesof infantdeath. If all infantswereneonatesitcouldbeassumedthattheywerethevictimsofinfanticide.Thearchaeologistsdismissedthepossibilitythattheywerestillborn,becauseofthelarge number of individuals represented. On the other hand, if they were ofdifferent ages, then it could be assumed that they either died from naturalcauses,orthattheyweremassacred.57

Thebathhouseunderwhichthe infantremainswerefound,wasbuilt inthe4thcenturyAD,anduseduntilthe6thcentury.Theinfantbonesweremixedwith animal bones, potsherds and isolated coins. No signs of careful burial orassociated grave goods were observed. All parts of the skeletons wererepresented,althoughthejuxtapositionofsomanybonesprecludedseparationofindividuals.Theresearchersinterpretedthegoodconditionofthebonesasanindicatorthattheirdisposalinthesewertookplacesoonafterdeath,whenthesofttissueswerestillpresent.58

To determine the age of the infants, three independent methods wereused.Bone length, dental development, and examination of ground sections ofthe teeth for neonatal lines were examined. Since the bones were mixed54Mays/Eyers2011,1931‐1932.55Mays/Eyers2011,1932‐1933.56Smith/Kahila1992,668.57Smith/Kahila1992,668‐669.58Smith/Kahila1992,669.

15

together,itwasnotpossibletodistinguishbetweenbonesbelongingtodifferentindividuals.TheMNIwasthereforeassessedfromthemaximumnumberofthelongbones(withleftandrightsidesrecordedseparately).59

The diaphyses of intact long bones were measured, and the lengthrecordedwasusedtoestimateagebycomparisonwithlongbonemeasurementsof modern fetuses and neonates of known age, and archaeological samples ofinfantremainsrecoveredfromdifferentsites.Dentaldevelopmentwassimilarlyassessed by comparison of crown height of the deciduous incisor tooth budsfromthesample.Thesewereexaminedunderalightmicroscopetodetermineifany signs of a neonatal line were present (this line results from enameldepositionintheimmediatepostnatalperiod,andisseenasevidenceofsurvivalforatleastafewdaysfollowingbirth).60

Many of the tooth buds were stained in a dark red‐brown colour,suggesting impregnationof thehard tissueswith ironoxide. Sincenoneof thebonesshowedanysimilarstains,groundsectionsof twoof thesestainedteethwere further examined, to identify the nature of the stains. Microprobe withmass spectography showed that iron oxide was present in relatively highquantities inthestainedareasoftheteeth.Theironoxidewasconcentratedinthe external layer of developing immature enamel and may representdegradationproductsofhaemoglobin,releasedbybleedingintotheoraltissuesatthetimeofdeath.Inforensiccases,breakdownproductsofhaemoglobinhavebeen identified in the dentine of adult teeth, especially after suffocation. It isascribedtobloodbeingforcedintothedentinetubuli.Fromtheseobservationscanbeconcludedthattheinfantsweremostlikelysuffocated.61

Theintactlongbonesareshownintable1,andthedistributionoflengthmeasured for complete long bone shafts is shown in figure 4. The extremelynarrow range in bone length suggests that all bones belong to infants ofapproximatelythesameage.Figure5showsthedistributionoflongbonelengthin the Ashkelon sample, in comparison with the bone length in the smallestinfants recovered from the lateOttoman cemetery atDor. This figure shows amuch greater variation in long bone length for theDor infants, indicating thatherewe are dealingwith infants of different ages, an representingdeath fromnatural causes.Whereas in the sample from Ashkelon the size range is muchnarrower,whichindicatesoneagegrouponly.62

The lengthof the bones in theAshkelon sample is given in table 2. Forcomparison,meanvalues forbone length inmodernneonates,andfor thoseofnewborns from various archaeological sites, are given in table 3. Length of allbonesmeasured from the Ashkelon infants, corresponds to that of recent andarchaeological samples of neonates. Examination of the teeth showed that noneonatallineswereseenonthegroundsectionsoftheincisors,whichisanotherindicatorthattheinfantsdiedsoonafterbirth.63

ThenarrowsizerangeofthebonesoftheAshkeloninfants,demonstratesthatallwerethesameagewhentheydied.Theinfantsfallwithinthesizerangeofmodernneonatesforlimblengthandtoothdevelopment,andgroundsections59Smith/Kahila1992,671.60Smith/Kahila1992,672.61Smith/Kahila1992,673.62Smith/Kahila1992,673.63Smith/Kahila1992,673.

16

of the teeth show no neonatal line, so they died almost directly after birth.Infanticide appears to be themost likely explanation for the disposal of thesenewbornsinthegutter.AllarchaeologicalreportsaboutinfantremainsinIsraeldescribe careful burial treatment, either with adults, or separate in pots orsheds.64Disposaloftheinfantsinthegutterseemsanomalous,andpresumablyreflects either indifference, or an emergency measure resulting from acatastrophe, such as an epidemic or massacre. However, catastrophes wouldaffectchildrenofdifferentages.Sinceonlyneonateswerefoundhere,infanticideseemstobethemostlikelyexplanationinthisparticularinstance.

In this chapter two Greek and two Roman sites have been discussed.EspeciallytheRomanevidenceseemsquiteconvincing.TheHellenisticevidence,ontheotherhand,stilllacksdecisiveproof,sincenoforceofphysicaltraumacanbediscovered in thebonematerial. In these cases infanticidewould indeedbepossible,butitmightjustconcerninfantswhodiedoffnaturalcauses,andwhodid not receive formal burial. The social status and burial habits for youngchildren, might perhaps explain why these infants were exposed of in such amanner.

Chapter3In thischapterthesocialstatusofnewbornsandyoungchildrenwillbe

discussed.ThesurvivingevidencerefersmainlytoancientAthens,andthereforethischapterwillfocusonthisarea.

InClassicalAthens,beingbornwasinsufficienttomakeachildamemberofanoikos.Eventhosewithtwocitizenparentshadnoautomaticrightofentry,but had to be accepted by the kyrios first. Two ceremonies marked thisacceptance:theamphidromia,whichtookplaceonthefifthorseventhdayafterbirth,andforthosewhocouldaffordasecondcelebration,thedekate,whichwason the tenth day after birth. At the amphidromia the father carried the childaround the household hearth, an offerwasmade to the gods, and friends andrelatives sent traditional gifts, such as octopuses and cuttlefish. It waspresumably at this time that the family decorated the doorway of its home toannouncethebirth.Aboywasannouncedwithawreathofolive,andagirlwithwool.Theinfantwasalsonamedatthisrite.However,ifafamilycouldbeartheextra expense, the baby received its name at thedekate. This celebration alsoinvolvedanofferingtothegods,andwasmorefestive.Thewomenweredancing,andtheritewasopentooutsiders.65

Until they reached the age of maturity, boys could not vote in theassembly,serveinthearmedforces,orrepresentthemselvesinlawcourts.Girlswerelifelongminors.However,minorswerenotcompletelyexcludedfromciviclife. Religion provided the main activity for children in the polis life. TheAnthesteriaiscertainlyconnectedwithchildren,especiallyitssecondday,Choes.AritualcalledAiorawasthenperformed.Thiswasarite inwhichthechildrensatinswingstiedtotreebranches,andmighthavebeenafertilityrite.AnotherfestivalwithaspecialsignificanceforchildrenwastheOschophoria.Thisrelated

64Smith/Kahila1992,674.65Golden1948,21‐22.

17

tothecomingofageofAthenianboys.66On a linguistic level, the Greek conception of childhood is inextricably

intertwinedwithplay.Pais,‘child’,sharestherootofafamilyofwordsincludingpaizo,‘Iplay’,andtopaigma,‘play,sport’.Playcharacterizeschildrenincontrasttoadults,whoareinvolvedinmoreseriousactivities.67However,thewordtokosmeans both ‘child’ and ‘interest on a loan or investment’,which indicates thatchildrenwerenecessaryforthefutureofanoikos.68

The fact that infantswere not citizens yet directly after birth, and thattheywere party excluded from civic life shows that children did not have thesame social status as adults. That girls remained to be treated asminors, andtheycouldnotvote,alsoindicatesthattherewasadifferenceinstatusbetweengenders. Depictions of boys and girls some years after infancy on funerarymonuments, show different characteristic attributes, such aswagons for boys,andducksordollsforgirls.69

Atheniangirlswerenotallowedtocompeteinathleticfestivalsduringthearchaicandclassicalperiod.However,Spartangirlswereallowedtoparticipate.Not only there were social differences between boys and girls, children andadults,butalsobetweenpoleis.70

Aswehave seen in theprevious chapters, childmortalitywasprobablyhigh.Inaworldinwhichearlydeathswereroutine,theintensityanddurationofemotional responsesmight have been limited, especially if the social status ofchildren was indeed lower than the status of adults. It is difficult to find outwhethertheancientscaredfortheirchildren.Perhapsparentslimitedthedegreeoftheirpsychologicalinvolvementwiththeirinfantchildren,sothatiftheydiedatayoungage,theymighthavefeltmoreorlessindifferenttothis.71

However, infants and young children were among the most carefullyburied individuals inancientAthens.Cemeteriesdevotedprimarily to childrenextendedoverlargeareasatthecitygates.ThedevelopmentofdefinitionsofagethroughburialpracticedemonstratesremarkablecontinuitythroughoutthefirstmillenniumBC.Yet,theemphasisplacedonformalchildburialvariedovertime,particularlybetweentheperiodcoveringtheformationandpeakoftheAthenianpolis(720‐400BC),andthecenturiesbeforeandafter(1100‐720BCand400–0BC, resp.). Far from being individuals in whom Athenians showed no officialinterest, infants held special significance in burial customs at the height ofAthens’powers.72

Thecarecharacterisingthearrangementofgravegoods,andthedisposalofthesmallbody,isstriking.Alsonoteworthyisthewayinwhichtheageofthedeceasedchildsignificantlyinfluencedthechoiceofburialtypeandgravegoods.From 1100‐400 BC, those who buried children found it necessary to expressthroughburialcustoms,thefactthatachild(asopposedtoanadult)haddied.Itis likely that female relativeswere responsible for burying the infant. Archaic

66Golden1948,38‐41.67Golden1948,53.68Golden1948,93.69Golden1948,72.70Golden1948,72.71Golden1948,82‐83.72Houby‐Nielsen2000,151.

18

andClassical vasepaintingsdepictwomenpayingvisits to tombs.73Also, childcemeteries lay near, or en route, to the sanctuaries of goddesses, such asDemeter or Artemis, where female cults concerned with fertility werecelebrated.

Houby‐Nielsen argues thatwomenoftenburied their children in such awaythatitispossibletodistinguishbetweenthreeagegroups:infants(0‐1yearold),smallchildren(1to4yearsold),andtheolderchild(3‐4to8‐10yearsold).Over time, these age groups were increasingly formalised through choice ofburialtype,andthenatureofgravegoods.Inaddition,childgravesmayalsobeidentified through the length of the grave, pit or coffin. This categorisationculminated during 720‐400 BC. However, at no time were adolescentsdistinguished from adults in burial rites.74 In this paper I will follow Houby‐Nielsen’sreconstructionofagegroupsandburialcustoms.

In theSubmycenaeanperiod (1100‐1050BC)sexhad little influenceongraveforms,andchildrenweregenerallyburiedincistsorinsimplepitgraves,withoutstonewalls.75Thisperiodsawthefirstattemptstodistinguishinfantsasanagegroup.76

TheProtogeometricperiod(1050‐900BC)sawaradicalshiftintheburialofadults,with thevastmajoritybeingcremated.This tradition lasteduntil the6thcentury,andremainedinuseeveninthe5thand4thcenturies.However,thecategory of infant burials remained the same, since infants and small childrencontinued to be inhumed in pits, or (from this moment on) in medium‐sizedstoragejars.Theseinfantburialurnswereoftenclosedwithastone,ora largepotsherd,andthegravepitsintowhichtheywereinsertedcoveredwithslabs.77

InhumationforadultsreturnedintheLateGeometricperiod(2ndquarterofthe8thcentury).Theusualgravetypewasinhumationinapitgrave,orafairlysimpleshaftgraveforadultsandolderchildren,andinhumationinsideavasefortheveryyoung.However, someadultmaleswere still cremated, andburied inamphorae.AtdifferentsitesinAttica,suchasAnavyssosandtheAcademy,somechildrenwerecremated,andtheashesplacedinamphorae,sothereweresomeexceptions.However,themajorityofinfantswereinhumed.Atmanysites,infantvaseburialswerecutintotheshaftsofadultgraves.Somegraveswerequiterich,particularly in pottery, but before the end of the Late Geometric periodquantitiesofgravegoodsweredeclining.78

Between720and400BC,thedefinitionofagethroughmortuarypracticebecameevenmoredistinct.,andformalisedburialritesforsmallchildren,agedbetween1and3‐4years,emergedaround500BC,throughtheuseofterracottabasins.Thesebasinsmeasuredbetween80and100cm.This typeofburial forsmallchildrenwasverycommonthroughoutthe5thandmostofthe4thcentury.Previously, this age group had been inhumed in pits, cist graves, or woodencoffins. In comparison to the urn and basin burials given to infants and smallchildren, older children (aged between 3‐4 and 8‐10) remained a moreindistinct, although recognisable, group in terms of grave type. The main73Houby‐Nielsen2000,152.74Houby‐Nielsen2000,152.75Morris1987,18.76Houby‐Nielsen2000,152.77Houby‐Nielsen2000,152.78Morris1987,20‐21.

19

distinguishingcriteriaaretheuseof inhumationastheacceptedritual,andthesizeofthepitorcoffin,whichwasfittedtocontainthebody.79

After700,primarycremationbecamethenormagainforadults,andtheadult graves seemed to have been marked by mounds. Grave goods stoppedrather abruptly around this time. In the late 7th century funerary sculpturesbegan to appear. By the mid‐6th century, inhumation was becoming morecommon for adults, and grave markers reached a peak of magnificence (buthereafter declined sharply). By 500 BC, child inhumations in clay tubs and ingravesmadefromrooftileswereappearing.Generally,burialformsweregettingsimpler.80

Intheperiod300‐0BC,infantsandsmallchildrencontinuedtobeburiedinvesselsandhouseholdbasins,butitalsoseemsthatcremationwasnowmorefrequentlyusedforchildrenthanbefore.81

Identification of the most common categories of grave goods found inrelationtochildburials,indicatesthatanambiguousawarenessofagepervadedthechoiceofgravegoodsforalmostamillennium.Thispatternwassomewhatmoreconsistentbetween720and400BC, incomparison to theprecedingandsucceedingperiods. Ingeneral, infantsdidnotreceivesexspecificgravegoods,and the number of pieces of jewellery, terracotta, toys, and items that weregenderspecificgraduallyincreasedwiththeageofthechild.ThisindicatesthatAtheniansocietyperceivedtheinfantstageasessentiallygenderneutral.Genderspecific grave goods, such as male‐associated athletic equipment or female‐relatedtoiletitems,includedpyxides,swords,belly‐andneck‐handledamphorae,aryballoi, mirrors, spindle‐whorls, needles, arrowheads and cauldrons. On theotherhand,theprovisionofinfantswithvasesfordrinkingandeatingtendedtodecreasewithage.82

Theenchytrismosburials,fortheyoungestinfants,receivedmostlyeatinganddrinkingvessels,suchasthebambylios(orbabyfeeder).Thismightreflectthepreoccupationwith thenourishmentof thechild in theearliest stageof itslife.Fromabout500BConwards,theolderinfants,agedbetween1‐3,werealsomarkedoutby their interment in terracottabasins(larnakes),asyetdescribedabove. For them the most common grave offering, which took the form of aterracottaanimalfigurine,relatednottheconceptofphysicalnourishment,butprobablyrathertothedevelopingchild’sdesireforplay.83

IntheClassicalperiod,infantsandsmallchildrenwerealsoprovidedwithmany more small and miniature jugs than older children and adults. Thus, itseems that artefacts in child graves were selected with a view to theirappropriatenessforparticularagegroupsofchildren.84

IntheClassicalandHellenisticperiods,morechildrenwereburiedcloseto,ortogetherwith,adults.Anthropologicaldeterminationsandassociatedgravegoods suggest that all theseadults arewomen.The importanceof themother‐child relation could indicate the ideal of legitimate children, since citizenshiplawsweregainingground.This isatrendthatcontinuesintotheRomanburial79Houby‐Nielsen2000,153.80Morris1987,21‐22.81Houby‐Nielsen2000,153.82Houby‐Nielsen2000,153.83Beaumont2012,88‐89.84Houby‐Nielsen2000,153.

20

habits.85To sum up, young children were among the most carefully buried

individuals in ancient Athens. Cemeteries devoted primarily to childrenextended over large areas at the city gates, and near temples associated withwomen and fertility. Infantswere party excluded from civic life, and probablyhadadifferentsocialstatus thanadults.However, theywere important for thefuture of both oikos and polis. This is reflected in their burial: the carecharacterisingthearrangementofgravegoods,andthewayinwhichtheageofthe deceased child significantly influenced the choice of burial type and gravegoods,isstriking.Itseemsthen,thattheinfantsdisposedinwellsorinsewers,as described in the previous chapter, were disposed of in a very nonchalantmanner. The enormous difference between these burials, and the carefullyburied infants in cemeteries, could indeed very well indicate that the infantsburiedinwellswerevictimsofinfanticide.

Conclusion

In the academic discussion about infanticide manymethods have beenused, to establish whether it was a literary topos or if it reflected ancientcustoms. Linguistic, economic anddemographic arguments arementioned. Forexample,oneofthelinguisticargumentsisthatenchytristriaiinfactdidnotreferto the chutrai in which the infants were buried, but to chutros, which was asacrificialpit.Aneconomicargumentthatismentionedofteninthediscussion,isthat dowries could be an important motive for infanticide. Demographicargumentscouldbethatinfanticidecouldcauseashortageofmarriageablegirls,or the opposite, that without infanticide, there would have been too manymarriageablefemales.Themanyargumentsputforwardinthediscussion,makeit seem plausible that infanticide might have occurred. However, conclusiveevidencecannotbedistilledfromtheseargumentsalone. Fourarchaeologicalsiteshavebeendiscussedabove.ThetwoHellenisticsites cannot be described as full‐proof evidence for infanticide, although it isstriking that theMessenian infantswere re‐deposited into awell, and that theinfantsfromtheAthenianAgorawellwereallneonates.Someofthemhadacleftpalate, so they had less chance of surviving, and their deformitiesmight havecausedthemtoberegardedassocialoutcasts.

The two Roman sites, however, do seem quite convincing cases ofinfanticide.InboththeYewdenVillaandtheAshkelonsample,alargenumberofneonateswasdisposedof,andallofthemhaddieddirectlyafterbirth,accordingtotheosteoarchaeologists.Incaseofnaturaldeaths,differentage‐classeswouldhave been represented. At the Yewden Villa site, some infantswere buried inrubbish pits, and at Ashkelon they were found in a sewer, so not all of themreceivedaformalburial.Furthermore,theAshkelonsamplealsoshowedthatitisverylikelythattheinfantsdiedfromsuffocation,orthattheyweredrowned.Theirtoothbudswerediscoloured,duetoironoxide.

Regardingthesocialstatusofchildren,itisstressedbyscholars,thattheyhad a lower status than adults, and that girls probably enjoyed a lower status

85Houby‐Nielsen2000,162.

21

thanboys.Nevertheless, throughout thewhole firstmillenniumBC, infantsareamongthemostcarefullyburied individuals, soeventhoughtheirsocialstatusmighthavedifferedfromanadults’status,timeandcarewasspentonburyingthem.

The question whether the ancients cared for their children cannot beanswered easily, since the formal burials do reflect that most children wereburiedwithcare.Thegravegoodsindicatethattherewasanawarenessofage,andthattheageoftheinfantinfluencedthechoiceofgravegoods.

Even though they were partly excluded from civic life at a young age,childrenwereimportantforthesurvivaloftheoikosandpoleis,andtheircarefulburialsreflectthissenseofhopeforcontinuity.Inthislightitissurprisingthatinfanticidewould have occurred. The infants from the Greek sitesmight havebeendisposedof,becausetheirchancesofsurvivingwerealreadylow.Childrenwith cleft palate could not be fed, and therefore might have been exposed.Perhapstheirparentsweresousedtothehighchildmortality,thattheydidnotputanyeffortinweakerinfants.Emotionalresponsesmighthavebeenminimalwhena childwithadeficiencywasborn, andespeciallypoorer familieswouldneedall theresources theyhadto feedtheirhealthychildren,so that theyhadsomesortofinsurancefortheirownfuture. However, in the Roman samples there seems to be no indicators tobelieve that these infants were already weak. These children were probablyunwanted. Toconclude,infanticidedidoccur.However,inwhatquantitiescannotbedetermined without more archaeological evidence. It seems likely that thedecisiontorear,ornottorear,achild,wasamorepracticalonethanisacceptedinourmodernworldviews.Childrenwereimportantfortheancients,buttakingtheirsocialstatusintoaccount,weakerorunwantedinfantscouldbelefttothewillofthegods.

22

Literature

Beaumont,L.A.2012,ChildhoodinancientAthens:iconographyandsocialhistory,

London/NewYork.

Bolkenstein,H.1922,TheexposureofchildrenatAthensandtheἐγχυτρίστριαι,

CPh17.3,222‐239.

Bourbou,C./P.Themelis2010,Childburials at ancientMessene, inA.Guimier‐

Sorbets/Y. Morizot (eds.), L’Enfant et la mort dans l’antiquité I. Nouvelle

recherchesdestombesd’enfants(actesdelatablerondeinternationaleorganisée

àAthènes,Écolefrançaised’Athènes,29‐30mai2008),Paris,111‐128.

Cameron,A. 1932, The exposureof children andGreek ethics,ClassicalReview

46,105‐114.

Engels,D. 1980, Theproblemof female infanticide in theGreco‐Romanworld,

CPh75,112‐120.

Golden,M.1948,ChildrenandchildhoodinClassicalAthens,Baltimore/Londen.

Golden,M.1981,DemographyandtheexposureofgirlsatAthens,Phoenix35.4,

316‐331.

Gomme,A.W.1933,ThepopulationofAthensinthefifthandfourthcenturiesB.C.,

Oxford.

Glotz, G. 1892, Expositio, in Ch. Daremberg/E. Saglio (eds.), Dictionnaire des

antiquitésgrecquesetromainesd’aprèslestextsetlesmonumentsII,930‐939.

Haentjens, A.M.E. 2000, Reflections on female infanticide in the Greco‐Roman

world,L’antiquiteClassique69,261‐264.

Harris, W.V. 1982, The theoretical possibility of extensive infanticide in the

Graeco‐Romanworld,CQ32,114‐116.

vanHook,L.R.1920,TheexposureofinfantsatAthens,TAPhA51,134‐145.

Houby‐Nielsen,S.2000,ChildburialsinAncientAthens,inJ.S.Derevenski(ed.),

Childrenandmaterialculture,London,151‐166.

Ingalls,W.2002,Demographyanddowries:perspectivesonfemaleinfanticidein

ClassicalGreece,Phoenix56.3,246‐254.

Mays/Eyers2011,Perinatal infantdeathat theRomanvilla siteatHambleden,

Buckinghamshire,England,JArchSc38,1931‐1938.

23

Morris, I. 1987, Burial and ancient society: the rise of the Greek city‐state,

Cambridge.

Oldenziel, R. 1987, The historiography on infanticide in antiquity. A literal

stillborn,inJ.Blok/P.Mason(eds.),Sexualasymmetry,Amsterdam,87‐107.

Papadopoulos, J.K. 2000, Skeletons in wells: towards an archaeology of social

exclusion in theAncientGreekworld, in J.Hubert (ed.),Madness,disabilityand

social exclusion: the archaeology and anthropology of difference (One world

archaeology40),MiltonPark/NewYork,96‐118.

Pomeroy, S.B. 1975, Goddesses, whores, wives and slaves: women in Classical

antiquity,NewYork.

Pomeroy, S.B. 1993, Infanticide in Hellenistic Greece, in A. Cameron/A. Kuhrt

(eds.)Imagesofwomeninantiquity(rev.edition),Londen/Detroit,207‐222.

Rotroff, S.I./L.M. Little/L.M. Snyder 1999, Colloquium: The reanalysis of awell

depositfromthesecondcenturyBCintheAthenianAgora:animalsacrificeand

infanticideinLateHellenisticAthens?,AJA103,284‐285.

Smith, P./G. Kahila 1992, Identification of infanticide in archaeological sites: a

case study from the Late Roman‐Early Byzantine periods at Ashkelon, Israel,

JArchSc19,667‐675.

Vlachou,V.2012,Addendumtovol.VI,DeathandBurialintheGreekworld,inA.

Hermary/B.Jaeger(eds.),Thesauruscultusetrituumantiquorum(ThesCRA)VIII,

LosAngelos/Basel,363‐384.

Websites:

Eur.Ion15‐40:

http://www.perseus.tufts.edu/hopper/text?doc=Perseus%3Atext%3A19

99.01.0110%3Acard%3D1,consultedonFebruary,32014.

http://www.chilternarchaeology.com/infant_deaths.htm,consultedon

May6,2014.

24

Figuresandtables

Figure1:GroundplanofGraveMonumentK3anditsprecinct

25

Figure2:Childburial(enchytrismos)

Figure3:Distributionofinfantburials.ViewsfacingWandE

26

Figure4:RadiographsoffemursfromdifferentindividualsatAshkelon.Notethesimilarityinsize

27

Figure5:SizerangeofmeasurementsoflongbonesatAshkelonandDor.Ashkeloninblack

Table1:NumberofinfantlongbonesfromAshkelondrain

28

Table2:Longbonemeasurements(mm)fromAshkelon

Table3:Meanlengthofbonesininfantsfromdifferentsitesandperiods

29

Listoffiguresandtables:

Figure 1: Ground plan of Grave Monument K3 and its precinct, from:

Bourbou/Themelis2010,126fig.1.

Figure2:Childburial(enchytrismos),from:Bourbou/Themelis2010,127

fig.2.

Figure3:Distributionofinfantburials,from:

http://www.chilternarchaeology.com/infant_deaths.htm,consultedon

May6,2014.

Figure4:Radiographs of femurs fromdifferent individuals atAshkelon,

from:Smith/Kahila1992,669figure1.

Figure5:SizerangeofmeasurementsoflongbonesatAshkelonandDor,

from:Smith/Kahila1992,670figure2.

Table 1: Number of infant long bones from Ashkelon drain, from:

Smith/Kahila1992,668table1.

Table 2: Long bone measurements (mm) from Ashkelon, from:

Smith/Kahila1992,671table2.

Table3:Meanlengthofbonesininfantsfromdifferentsitesandperiods,

from:Smith/Kahila1992,671table3.