2002 Report

STANFORD UNIVERSITY EXCAVATIONS ON THE ACROPOLIS OF MONTE POLIZZO, SICILY, III: PRELIMINARY REPORT ON THE 2002 SEASON Ian Morris, Trinity Jackman, Emma Blake, Stanford University; Brien Garnand, Santa Clara University; and Sebastiano Tusa, Soprintendenza di Archeologia, Trapani1 with contributions from Tara Hnatiuk, Southampton University; Wendy Matthews, Reading University; and Hans-Peter Stika, Universitt Hohenheim In memory of Luke Bridgwater (1976-2002) 1.The projects intellectual context Research on Iron Age western Sicily has evolved rapidly. Thucydides comment2 that the ElymianpeoplehadmigratedherefromTroyhadlongfascinatedhistorians,3butuntil WorldWarIImostexcavatorsconcentratedonGreek,Phoenician,andRomanremains. TheirworkoftengeneratedindigenousIronAgematerialtoo,butthefirstmajor excavationaimedspecificallyattheIronAgewasVincenzoTusasatthecontrada MangositeatSegesta.Beginningin1953,heuncoveredahugesixth-centuryDoric temple.4 In 1970 he excavated at several other inland sites, including Monte Polizzo.5 He definedaseriesofcorequestions:DidElymianandSicanethnicityhavematerial markers?Ifso,didthesemarkersdefineterritories?Whendidtheseethnicgroupsform, and/orwheredidtheycomefrom?Andhowdidtheindigenouspopulationsbecome Greek?6 Not without reason did Massimo Ganci, introducing the first major conference on westSicilianindigenoussitesatPalermoin1989,callVincenzoTusailprimoElimo.7 Hisquestionsinspirednumerousfurtherexcavations,includingresearchintoIronAge IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 20022 deposits at Segesta.8 The three biggest inland projects begun before the 1990s were all at famousGreco-Romansiteswithmonumentsandinscriptions,whereIronAgedeposits were deeply buried or severely disturbed (the University of Zurich started work at Monte Iato in 1971, and the Scuola Normale di Pisa at Entella in 1983 and at Segesta in 1987);9 butallpaidmoreattentionthantheirpredecessorstorecoverIronAgematerial,andin 1991 Giuseppe Nenci of the Scuola Normale founded the Centro Studi e Documentazione sullAreaElima(CESDAE).ThismadetheIronAgeitsmajorfocus,10anditsGiornate internazionalidistudisullareaElima, held every third year since 1991, have transformed the study of this region.11 Asoftenhappens,thefieldworkthatVincenzoTusainspiredproduceddatathat raisedmorequestionsthantheyanswered.Archaeologistsdidnotfindcleardifferences betweenElymianandSicanmaterialcultures:insteadtheyobservedabroadzoneof similar pottery, settlement forms, and religious practices from the Salsa and Imera valleys tothewesterncoast,whilewithinthiszone,theyfoundconsiderablevariationbetween sites. And when looking at Hellenization they found that i modi, i tempi e le circostanze diquestoprolificoprocessodicambiamentifuronoassaimutevoli.12Newevidence calledfornewmethods:whereasarchaeologistsinthe1970s-80softendefinedElymian materialcultureandillustratedHellenizationbyjustshowingafewdecontextualized finds, in the 1990s they moved toward detailed accounts, often quantifying their data.13 Thisdata-drivenshiftcomplementednewideasthatancienthistoriansand prehistoricarchaeologistshadindependentlybeendeveloping.Theseideasledtotwo newsetsofquestions,whichwewillcallthesocioeconomicandpostcolonial frameworks.The first of these frameworks took shape in the 1960s-70s. Ancient historians had longsuspectedthatThucydidestoldusmoreaboutGreekethniccategoriesthanabout nativeself-identifications,14andtheSecondWorldWarmademanyscholars uncomfortablediscussingethnicityandrace.ParticularlyinItalyandBritain,historians moved away from ethnic questions and toward socioeconomic issues, influenced by Marx and Weber. In the first overall review of ancient Sicilian history produced in Britain since the1890s,MosesFinleyvirtuallyignoredethnicityinfavorof sociology; and in Italy the ScuolaNormaledevelopedadistinguishedschool,emphasizingdiet,demography,trade, IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 20023 andsocialorganization.15GiuseppeNenci,anepigrapheraswellasaleadingfigurein westSicilianarchaeology,broughtthehistoriansideastowestSicily,definingthearea elimatoincludeallSicilywestoftheImera-Salsarivers,andallperiodsthroughthe medieval.16 InthesameyearsEnglish-languageprehistoricarchaeologistsalsomovedaway fromculturehistory,withitsinterestinlabelingspatiallydefinedmaterialculture groupsaspeoples,towardsocialandeconomicproblems,particularlyhuman interactionswiththeenvironmentandtheevolutionofincreasinglycomplexsocial organization.17 Some scholars of the Greek and Italian Iron Ages found social evolution a usefulframework,andmadetheriseofthestateacentralresearchtopic.18InSicily, BronzeAgearchaeologistshavefoundthishelpful,19althoughmoststudiesofIronAge Sicily continue to use lo stato as a synonym for self-conscious ethnic groups.20 Inthe1980sresearchquestionsshiftedagain,fromsocietyandeconomyto identity.PostprocessualarchaeologistschallengedtheNewArchaeologists functionalism,arguingthatmaterialculturewasmeaningfullyconstituted,and manipulatedbyconsciousactors.Postprocessualarchaeologistsemphasizedthe construction,ratherthandiscovery,ofevidenceaboutthepast,andinsistedthatwesee archaeologyassocialandpoliticalactioninthepresent.21Newapproachestothewest MediterraneanhavecomeoutofthetheoreticalandmethodologicalturmoilinAnglo-Americanarchaeology,lookingattheconstructionofhybridculturesincolonial encountersthroughdifferentgroupsselectionandadaptationoftraits,ratherthanthe assimilationofindigenousculturesbydistinctandstrongerEastMediterranean cultures.22 Inthe1990sancienthistoriansalsochallengedessentialistmodelsofidentity. SomearguedthatGreeknesswasjustoneofseveralcompetingformsofcollective identity,whichevolvedfromanagglutinativemodel,inwhichnewgroupscouldclaim GreeknessandnewcriteriaforGreeknesswereconstantlyadded,toanoppositional model,inwhichaclosedgroupdefinedthemselvesasHellenesagainsttheoutside world.23MosthistorianssawthewarswithPersiaandCarthagein480astheturning pointinthisprocess.24Tosomehistorians,itmadelittlesensetospeakofHellenization priortothefifthcentury,becausetherewasnounitaryGreeknesstobetransmittedto IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 20024 barbarianOthers.ThismadedistinguishingGreekcoloniesfromindigenoussitesseem bothmoredifficultandlessnecessary.25InsteadofauniformGreekculture,these historianssawmyriadoverlappingandshiftingculturesinendlesscontestation, dissolvingmostoftheboundariesthatstructuredresearchinSicilianIronAge archaeology in the 1950s-80s.26 ThereisprobablymoreinteresttodaythaneverbeforeintheGreco-Phoenician expansion into the west Mediterranean, but less agreement about what actually happened, orevenwhattheimportantquestionsare.Thetraditionalframeworkemphasizesethnic identity,drawingclearboundariesbetweenpopulations,withtheGreekseventually swallowingupordrivingofftheothergroups.Itcontinuestogenerateimportantwork, althoughtheexplosionofnewdatainthepasttwentyyearshasraisedproblems.Avery differentsocioeconomicframeworkinspiredarchaeologistsactiveinmainlandItalyand ancienthistoriansinthe1980s-90s,buthadmoreimpactonBronzethanIronAge archaeologistsinSicily.Itemphasizesdemography,diet,andsocialhierarchy.Inthelate 1990sathirdapproach,thepostcolonialframework,wonpopularitywithAnglophone scholars.Likethetraditionalmodel,itstressescultureandidentity,butseesshifting identitiesinarchaictimes,hardeninginthefifthandfourthcenturiesintoa Hellene/barbaros dichotomy. OurresearchontheacropolisofMontePolizzoisdrivenmainlyby socioeconomic and postcolonial questions. We ask particularly how far we should see the archaicGreco-Phoenicianexpansionasapartofakindofancientglobalization,linking thewholeMediterraneanbasintogetherinnewwaysandchangingthecourseofsocial developmentinthewesternpartsoftheOldWorld.27Postcolonialargumentshave underminedthetraditionalframeworkscertaintiesaboutthereplacementofindigenous culturesbyGreekcultureinSicily,andfollowinginthispath,wefocusonthe constructionandnegotiationof identities. But we also emphasize the material forces that drovethedemographicexpansion,andthematerialconsequencesthatfollowedit.We hopetoexplorenotonlymaterialstandardsofliving(e.g.,nutritionandhousing)and socialhierarchy(e.g.,thecentralizationofpowerinresistancetointruders),butalsothe constructionofnewsubjectivities(e.g.,classandgendercategories)andnewformsof communalexpression(e.g.,inreligiousrituals).Wehighlightdifferentialresponses(i.e., IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 20025 whetherallgroupsandcommunitiesbenefitedorsufferedequally,whetherthesixthand fifthcenturiesBCwidenedsocialdifferencesorevencreatedentirelynewones,whether different groups embraced the widening of their world or resisted it, etc.).ThelogicalwaytoanswerthesequestionsisnottoexcavateanotherPhoenician or Greek site, but to examine an indigenous site in close contact with east Mediterranean settlers. Monte Polizzo, located between Segesta (the major Iron Age indigenous center), Motya (the main Phoenician settlement in Sicily), and Selinous (one of the most powerful Greek cities) seems an ideal choice. 2.The site Figure 1Major sites in western Sicily mentioned in the text IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 20026 Monte Polizzo lies 6 km northwest of Salemi, in Trapani province (37 56 N, 12 46 E: figure1),andconsistsofaninterconnectedgroupofridges(figure2).The highest point is725.9mabovesealevel.AntiquariansfromSalemiandCorleonehavelongknown aboutthesite,althoughtherehasbeennopermanentsettlementherein800years. Tradition holds that the neve near the top of the hill (figure 3) was used for ice storage in theeighteenthandnineteenthcenturies.Inthe1950sthehillwasdeclaredaforest preserve, plowed, and planted with live oak and pine. The Forestry Service maintains dirt roadsandfirebreaksonMontePolizzo,scrapingthelatterwithbulldozers each summer. Morerecently,awater-pumpingstationwasinstallednearthefootofMontePolizzo, supplyingSalemi,andthehillhasbeenafavoritespotforrabbithunters.28Thelower slopes are used for pasture, and shepherds bring flocks to the upper slopes. Figure 2Aerial photograph of Monte Polizzo, with excavation areas marked (cf. figure 3. Prepared by C. Sevara) MontePolizzowasoneofthemainsitesin Vincenzo Tusas 1970 campaign, but thenewquestionsthathaveemergedsincethencallfornewfieldmethods.Sincethe IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 20027 1990sithasbecomenormalinSicilytorelateallfindstopublishedstratigraphic matrices,toprovidefaunalandgeologicalreports(thoughfloral,palynological,and chemicalstudiesremainrare),toanalyzeawiderangeofartifacts,andtoquantifydata. Systematicsurfacesurveyshavealsoproliferated.Anewprojectmustmeetthese standards and go beyond them. In particular, we need (a) excavations that explore several partsofaparticularsettlement,recoveringnumerouscompletehouseplans,entireritual andadministrativeareas,andsamplingwidelyenoughtohaveagoodsenseoftown planning,specialistquarters,fortifications,etc;and(b)evenmoredetail,recordingand quantifyingtheentireassemblagestratigraphically,andestablishingagreedstandardsfor data collection and publication, so we can make valid comparisons between sites. Mostarchaeologistswouldagree;butinpracticethereisrarelyenoughtimeand moneytomeetthesegoals.Inresponsetothesechallenges,SebastianoTusa,Direttore dellaSezioneArcheologica,SoprintendenzaaiBB.CC.eAA.diTrapaniandProfessor ofArchaeologyattheUniversityofNaples,andKristianKristiansen,Professorof ArchaeologyattheUniversityofGothenburg,createdtheSicilian-Scandinavian ArchaeologicalProjectin1996.TheyselectedMontePolizzoasthebestsitefornew work,withinitialgoalsof(i)understandinginteractionsbetweentheindigenous populationandGreekandPhoeniciancolonists;(ii)clarifyingtheformationofElymian ethnicidentity;and(iii)producingstratigraphiccorrelationsbetweendatableGreek potteryandlocalwares.Theyenvisagedatwo-wayexchange,bringinginternational perspectivestobearonwesternSicily,andusingwestSicilianmaterialto clarify debates takingplaceinarchaeologiesofotherpartsoftheworld.Theybeganfieldworkin1998 withapreliminarysurveyofthesite,suggestingthatitcovered15-20hectares. ChristopherPrescottoftheUniversityofOsloexcavatedmid-sixth-centuryBCHouseI in 1998-2001, a deeply stratified Iron Age deposit (the Profile) in 1998-99, and a series of soundings on the northwest slope of the acropolis in 1999 (see figure 3 for all locations withinMontePolizzo).29In2002KristianKristiansenandChristianMhlenbockbegan new excavations with students from the University of Gothenburg outside House I and on DanishHill,200-300mfurtherwest,wheretheyexposedfurtherIronAgeremains. SebastianoTusaexcavatedsixth-centurybuildingsatthePortellaSantAnnain1999-2000, and Antonello Rizzo opened several graves in the Iron Age cemetery in 2000-2001.IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 20028 Figure 3Excavation areas, 1998-2002. Contour interval 20 m NorthAmericanteamsworkedalongsidetheSicilian-Scandinavian ArchaeologicalProjectfromthebeginning.MichaelKolbofNorthernIllinoisUniversity surveyedaroundMontePolizzoandtheBronzeAgesiteofMokartain1998-2000, excavatedaBronzeAgetombatPitrazzionMontagnaGrandein2000-2002,andin IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 20029 2001-2002excavatedfourtrenchesinSalemi,findingmedievalandfourth-centuryBC deposits.30 In 2000 Tom Boving of the University of Rhode Island dug a small trial trench at the neve, probably the settlements major water source.Stanford University joined the project in 1999, and has excavated on the acropolis since2000.31In2002ateamfromtheUniversityofCalgaryandgraduatestudents from threeItalianuniversitiesandtheAmericanAcademyinRomessummerprogramjoined theStanfordexcavation.TheStanfordteamhad84membersin2002,andinallmore than 120 archaeologists conducted research at Monte Polizzo that season. Eachteamispursuinganindependentresearchproject,buttheprojectsinterlock into a larger whole Previous work in western Sicily had shown that the highest point of a settlementwasoftenitsreligiouscenter,andtheStanfordteamchosetofocusonthe acropolistogetabetterunderstandingofindigenousreligion.However,wecanonlydo this through systematic comparisons with other parts of the town, and by putting our data inthecontextoftheregionalsettlementhistory.Thevariousteamsproceduresare thereforeastightlylinkedasispractical.Allcloseddepositsaredry-sievedthrougha5 mmmesh,andtheteamssharethesamemacrofossilanalysis,GISdigitalrecording system,andfindsdatabases.TheGISandfindsdatabases(inMSAccess)arelinkedby theTotalStationGISpointnumbersgeneratedforfindsandstructures.Thepottery, whichmakesupthebulkofthefinds,undergoesatwo-stepsortingprocess.Allsherds aresortedbyfabrictype,countedandweighed.Diagnosticsherdsarethenrecordedin moredetail,andreceiveanartifactnumber.Wealsocoordinatefloralandfaunal techniques with excavators at Selinunte. We hope to establish a baseline that can be used for comparisons by future field projects in western Sicily. In2000-2001theacropolisexcavationconcentratedonzonesA,B,andC.This revealedpartsoffivestructurescurvilinearA1andrectilinearA3,A4,B1, and C1as well as altar A2. We concluded that A1 was a sixth-century shrine of a type known from contemporary sites. It collapsed around 500-475 and was replaced around 350 BC by the small structure A3. This was abandoned around 300 BC, and zone A was not reused until the1970s,whenrabbithuntersusedshelterA4.StructureB1/2andsurroundingwalls werebuiltinthesixthcenturyBCthenrenovatedinthetenthcenturyAD,whenroom B1/1wasadded.Onlyasmallpartofsixth-century structure C1 was excavated. In 2001 IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200210 wealsoclearedapartofaprobablysixth-centurystaircase,andProfessorJennifer TrimbleofStanfordUniversitycarriedoutaprotonmagnetometersurveyaroundthe acropolis. In 2002, we expanded the excavation in zones A, B, and C; opened zone D, on a terrace immediately north of structure A1; and excavated a 60-meter-long Great Trench connectingzoneBwiththeTusaHouse,oneofthestructuresexcavatedbyVincenzo Tusa in 1970 (figure 4). 3.Summary of results We have identified five phases of activity on the acropolis, spanning 3500 years: I.BronzeAge(c.1500-800BC).Wehavefoundfragmentsofmorethanadozen BronzeAgevessels,spanningtheMiddlethroughFinalBronzeAge.Allcame from Iron Age contexts. We have found no intact Bronze Age deposits. II.Sixth century BC (figure 5). Initial study suggests that we can break phase II into three sub-phases based on the types of Greek imports:!PhaseII.a,characterizedbyCorinthianimports,withsmalleramountsofEast Greek, probably c. 575-550 BC; !PhaseII.b,characterizedbyEastGreekimports,withsmalleramountsof Corinthian, c. 550-525; !PhaseII.c,characterizedbybothAtticblackglazeandEastGreekimports,c. 525-475. Sofar,themainoccupationbelongstophaseII.b.Inzone A(seefigure8)small areaspredatingbuilding A1 have been exposed, but none has yet been excavated. A1 is a round building, 6.4 m in diameter, built around 550, and (either as part of orsoonafteritsinitialconstruction)subdividedintothreesmallchambers. Outside we have found a stele, at least one altar, enclosure walls, a partially paved area, and rectilinear structure A5. A5 may also belong in phase II, but we have no securedatingevidenceyet.A1saltar(s)was/wereusedforburnedsacrifices, particularly of deer, but the meat was consumed elsewhere. The jaw and teeth IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200211 Figure 4Excavation zones on the acropolis IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200212 fromanadolescenthuman,aged12-14years,werefoundscatteredaroundA1. Theiroriginalcontextremainsunclear.A1wasabandonedaround500-475,but not in haste; before leaving, a pit filled with ash was sealed with clay, and a round clay hearth heaped with ash may also have been sealed. The mudbrick upper parts of A1s walls decayed gradually after 475. Zone B saw complex activity in phases II.b and II.c (see figure 17). Wall h and associated layers in trench L108 may date tophaseII.a.Thegroundsurfacenortheastofwallh(trenchM108)wasthen raisedwithdumpedearth,andadrainwasinstalled,emptyingoverwallh.A largedumpofash,storagepottery,andbones(particularlyreddeerantlers) covered much of zone B in phase II.b. Around 525, structure B1/2 was built over this dump. B1/2 went out of use by 500. Walls from an earlier structure (B2) have beenfoundundermedievalroomB1/1.B2probablydatestophaseII,butwe need to dig more to confirm that. In zone C, a rich destruction deposit on a paved floorintrenchN/O113/114datestophaseII.a.PartofstructureC1walla belongswiththisdeposit;theotherwalls post-date it. The Great Trench between zone B and the Tusa House revealed parts of another rectilinear building, C2, and surfaceexplorationssuggestthattheacropoliswasringedonthewestsideby an enclosurewall,withthinnerwallsrunninginfromittowardthetopoftheridge. Trench J/K 112/113 dated this enclosure to phase II. In zone D we exposed parts of two rectilinear structures and dense concentrations of storage vessels, dating to phase II.c. III.LatefourthcenturyBC.32StructureA3,perhapsasmallshelter,wasbuiltonthe ruinsofA1.FindsincludeaPunicstele,9bronzePunic-Siciliancoins,and5 stone dice. Zone A is the only part of the acropolis with phase III remains. IV.TenththroughtwelfthcenturiesAD(figure6).Theby-then1,500-year-oldruins ofB1/2wererebuiltandreused,roomB1/1addedtothenortheast,andseveral wallsbuilttodivideexternalspace.IronAgestructureC2wasalsorebuilt,and further medieval walls (poorly preserved) were found near structure C1. Medieval debris was found under the main phase IV floors in zone B, suggesting that thereIAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200213 Figure 5Phase II structuresIAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200214 Figure 6Phase IV structures IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200215 areoldermedievalstructures in the area. Zones B and C are the only parts of the acropolis with phase IV remains. V.LatetwentiethcenturyAD.33Inthe1950sfurrowsfortree-plantingweredeep-plowedaroundtheentirehill,andinthe1970sastoneshelter(A4)wasbuiltat the summit, severely damaging phase III structure A3. We may need to modify these phases in the light of further work.Inthemainpartofthisreport,wepresentthefindsfromzonesAthroughDin turn, in each case proceeding from the earliest deposits to the latest. 4.Zone A (figure 7) We continued work in zone A to clarify our picture of sixth-century religious behavior. Phase II TheearliestactivitydetectedinzoneAisclaylevelingfillsintheuneven sandstone/pebblenaturalsurfaceofthehill.RoundbuildingA1(diameter6.4m)was builtonthissurface,probablyaround550BC.In2001wesuggestedthatA1was semicircular,34butin2002foundthatitwasafterallround,althoughmostofthestones from the northern part of the circular wall c had been robbed in antiquity (figure 8). Walls aandbdivideA1intothreesmallcompartments.Asfigure8shows,wallbisonthe same line as wall e and the possible altar west of A1; and a stone block visible in the balk betweentrenchesM98andM99seemstobeonthesamelineaswalla,suggestingthat this also continued outside building A1. Three interpretations are possible: 1.Wallsa,b,e,thealtar,andthestonesintheM98/99balkbelongtoanearly rectilinear building, and circular structure A1 was built over this. 2.All the walls are contemporary. 3.StructureA1wasbuiltfirst,thenreplacedbyarectilinearstructurecomprising walls a, b, e, the altar, and the stones in the M98/99 balk. IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200216 Figure 7Structures in zone A, phases II, III, V IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200217 ThereisnosequenceofstratifiedfloorsinsideoroutsideA1tosettletheissue definitively(seefigure11),butthestratigraphicevidencewedohavesuggeststhat interpretation 2 or 3 is the most likely. In favor of theory 1, we should note that the west end of the altar is very close to the edge of the excavated area, and there are some small stones between the structure and the balk. It remains possible that this structure is in fact justtheendpartofawallcontinuingtothewest,inwhichcasethegapbetweenthe altarandA1wallccouldoriginallyhavebeenadoorwayintoarectilinearbuilding. Furtherexcavationwillresolvethis;butoverall,theory1seemsleastlikely,forfour reasons. First, while A1 walls b and c are not bonded, the east end of wall b (the west end ishiddenunderA3)wasneatlybuilttofitagainstthecurvinginnerfaceofwallc.Itis possiblethatacontinuouswallb-ewastherefirst,thenwaspartlydemolishedtomake roomforc,andfinallycarefullyrebuilttofitagainstc;butthisseemsveryunlikely. Second,whilewallbsurvivessevencourseshigh,wallehasonlyonecourse,witha steleinsituontopofit(seefigure11).Thissuggeststhatbandewerenevera continuous wall in a rectilinear building. Third, as figure 9 shows, A1 walls b and c stood 50cmabovethefloorofbuildingA1,makingithighlyunlikelythatA1replacedan earlierrectilinearbuilding.Finally,thereisclearstratigraphicevidencethatthealtar (whether part of an longer wall or not) postdates A1 wall c. Theory 3 can accommodate the junction of A1 walls b and c, and also provides an economical explanation for both the robbing of the north part of A1 wall c and our failure in2001todetectarobbertrenchcuttingthroughtheclaydeposit (layer 8 in figure 11b) created by the collapse of the mudbrick walls: there was no such trench, because wall cs stoneswerereusedinthenewrectilinearbuildingbeforethemudbrickcollapsedeposit wasformed.However,figure11bshowsthatthiswasnotthecase.Layer9,theashy abandonment deposit, postdates A1 walls a, b, and c, and layer 6, the trench produced by the removal of stones from wall c, clearly postdates 9. Themostplausibleinterpretationiseitheraversionoftheory2,thatA1wallsa, b,andcarecontemporary,withthealtaraddedlater(alongwiththepossiblewallin theM98/99balk),oracombination of theories 2 and 3, that A1 was originally consisted ofwallconly,andwassubsequentlysplitintothreechambersbywalls a and b, perhaps at the same time that wall d, pavement e, the altar, and the possible wall in the M98/99 IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200218 balkwerebuilt.Theory(b)postulatesageneralremodelingofthearea,inwhichthe old roundbuildingA1remainedinuseaspartofalargercomplex.Thecombinationof theories2and3seemsmorelikelyatthispoint,butextendingtheexcavatedareamay produce further evidence. TheentrancetoA1wasprobablyontheeastside,intochamberA1/1.The bedrockwassmoothedtomake a floor in the south part of chamber A1/1;35 in the north part,wheretherockslopesaway,claywasaddedtomakealevelsurface.Ashallow circularclayhearth(diam.64cm,exactlyone-tenthofA1sdiameter)wassetintothe floor (figure 10). Theashlayers(figure11alayer7and11blayer9)containedbronzebeadsand manysherdsfromamphorasandcups.Thesedepositshavenotyetbeenquantified,but seem to contain a higher proportion of Greek material than assemblages from other parts ofthesite.MostfinewaresherdscomefromIonicB2cups,buttherewerealsoafew Corinthianfragmentsand several Attic black glaze sherds, which date A1s abandonment c.500-475BC.Fewornovesselscanbereconstructedfromthesherds,suggesting that thepotsactuallyusedinA1/1wereremovedwhenthebuildingwasabandoned,leaving only fragments from vases that had already been broken and discarded. ChamberA1/2hadaroughlyleveledrockfloor.Apit(diam.80-90cm; maximumdepthapproximately20cm)hadbeendugintothis.Likethecircularclay hearthinA1/1,thepitwasleftheapedwithmulticoloreddepositsofash;butthepitin A1/2 was also sealed with two layers of very hard baked clay, each about 2 cm thick. The upper layer was red and the lower one white. The hole dug for the modern shelter A4 had damagedtheashmoundanditsprotectivelayers(figure11a),butpartsofthedeposit were trapped intact under A4 wall c. The basin in A1/1 may have been sealed in a similar way,butifso,thesealinglayerswerepoorlypreserved.TheA1/2abandonmentdeposit contained fewer artifacts than that in A1/1, but otherwise the finds in the two rooms were similar. Fourth-centuryA3wallccoversmostofA1/3,andwecouldonlyexcavatethe eastendofthechamber.Itsleveled-bedrockfloorwasveryclean,withnotracesofthe ash that characterized A1/1 and A1/2. IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200219 ImmediatelyeastofA1,asmallsteleremained in situ, standing on pavement e, a singlecourseofflatslabs(figure12).Thebottomofthestelewassetintothebedrock. Immediately north of the stele, we found ash, charcoal, and animal bones in 2001; east of it,alarge piece of red deer antler lay on pavement e (figure 13). We return to this antler in section 7 below. InareaA1/4westofA1,wefoundtwophasesofactivity.WhenA1was originallybuilt,thenaturalrockservedasthesurfaceinareaA1/4,withpatchesofclay to level it. Intense burning went on, and a deposit of scorched red clay gradually built up againstthesouthwestfaceofwallc,withpocketsofashandcharcoal.36Again,only fragmentarypotterywasfound.Theseincludedmoreindigenousgraywarecupsthan inside A1, and local and Greek amphora sherds, dating to the later-sixth century. In the second phase, probably close to 500 BC, wall d was built, along with what mightbeanaltar.Thiswasarectangularstonestructure,1.3x0.55 m, on the same line aswallbinsideA1(seefigure8).Weinterpretthisstructureasanopen-airaltar,since Burningandanimalsacrificewentoninthisspotbeforethisstonestructurewasbuilt, andcarriedonafteritsconstruction,accumulatingtoadepthof10cmagainstitsfaces. Because of this, we suggest that the structure was probably an open-air altar. In 2001, we suggested that structure A2 was an altar contemporary with or earlier than A1, and that it went out of use while A1 was still functioning.37 If this is right, the structure in A1/4 may havereplacedA2whenwalldwasbuilt,markingoffA1/4asaspecialenclosurefor sacrifices. However (as noted above) the new structure may turn out to be merely the end of a longer east-west wall. Further excavation will resolve this.Walldcurvedaroundtheedgeofanoutcroppingofbedrock, which dropped off tothesouthwest.Againstthefaceofthepre-existingwallc,walldwasbuiltontothe scorchedredclaydepositsthathadaccumulatedinthefirstphaseofA1suse;further west, it was cut through these layers and into the bedrock. Southwest of d, a clay packing (trenchL100layer21)wasadded,extendingtheareaavailableforuse.Burning continuedinareaA1/4inthefinalphaseofA1suse,andmoredepositsofred clay and charcoal,includinganimalbonesandteeth(someburned),accumulatedagainstwallsc, d,andthealtar.Thefindsfromthelatestlayersassociatedwiththealtarincluded IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200220 fragmentarygrayware,Corinthian,EastGreek, and Attic black glaze cups, and the spout of a late-sixth or early-fifth century west Greek oil lamp. Figure 8Hut-shrine A1, phase II At some point, rectilinear structure A5 was dug into the clay southwest of A1 wall d.L100layer21providesaterminuspostquemforA5inthe late-sixth century, but the smallpartofthefoundationtrenchexcavatedin2002containedno datable material, and excavationinsideA5hasnotyetreachedlayersthatcanprovideafirmdate.Thefew datablefindsassociatedwithA5aresixth-century,soweassumethatthisbuilding belongswithA1inphaseII,butthisiscurrentlyonlyguesswork.A5 consists of at least two spaces, divided by wall b. Both are packed with several courses of flat stones. TheseIAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200221 Figure 9Structure A1, facing south Figure 10Chamber A1/1, showing the circular clay hearth (scale 50 cm) IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200222 seemtobeplatformsratherthanpavedfloors,butwehavenoevidenceyetfortheir functions. IntrenchesM-N98weuncoveredanunevenscatteringofstoneslabs,probably fromapoorlypreservedpavedsurface,datingtophaseII.c.Thepavingitselfandwhat seemstobealevelingfillbeneathitcontainedsmallAtticblackglazeaswellas CorinthianandEastGreeksherds.Underthedisturbedslabswaspartofthejawand somelooseteethofanadolescentaged12-14years(figure14).38Anotherhumantooth was found in 2001 in trench M100, in the layers of sacrifices pre-dating A1 wall d.39 We do know yet what context the bones originally came from. A1wasdeliberatelyabandoned,withoutdestructionbyfire.Asnotedabove,the pitinA1/2andperhapsthecircularclayhearthinA1/1werecarefullysealed,andapart fromanironcleaverfoundin2001(seefigure8),40virtuallynocompleteartifactswere leftbehind.Wehavefound19beads,4112otherfragmentsofbronze,27fragmentsof iron,and9fragmentsofworkedboneinA1suseandabandonmentlayers,butallwere probably discarded or accidentally lost, rather than being votive offerings. Afterc.500-475,A1smudbrickwallsgraduallydecayed.Thehutmayhavehad athatchedroof,asGiuseppeCastellanasuggestsforhutshrine1atMontagnoli;42buta largepieceofclayrooftilefromthefillunderthepavingslabsinM98andseveral smaller fragments in other trenches suggest that A1 had a tile roof, renewed at some point beforethepavingwaslaidinM/N98.Ifthatiscorrect,thentheroofwascarefully dismantled and removed around 475 BC. 2002 provided no new evidence for continuous occupation between phases II and III. Many amphora sherds could date between 475 and 350, but no objects have to date to theseyears.ItstillseemsthatafterA1wasabandonedaround475,theacropolisstood emptyformorethanacentury.Atsomepointafter475,mostofthestonesinthe northernhalfofA1wallcwererobbed out. The robber trench presumably filled up with thesameyellow-grayclay that characterized the layers either side of it, originally formed bythedecayofA1smudbricksuperstructure.Wefailedtodetecttherobbertrenchin 2001againstthebackgroundoftheoriginalmudbrickdeposit(layers5and9infigure 11b),butin2002wepickeduptheedgeofthetrenchonthenortheastsideofA1/1, IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200223 wherethe ashy deposit that had spilled out across the rooms floor made a sharp edge at the point that the wall stones had been robbed (figure 11b layer 6). Figure 11SimplifiedstratigraphicprofilesthroughchambersA1/1andA1/2.(a) A1/2.Layer1,cuttingandfillforA4(phaseV);layer2,topsoil(phaseV);layer3, compactearth(phaseIII);layer4,compactearth(phaseIII);layer5,clayfrom decayedmudbrick(phaseII);layer6,clayfromdecayedmudbrick(phaseII);layer 7, mixed clay and ash from abandonment of A1/2 (phase II); layer 8, red clay sealing pit (phase II); layer 9, white clay sealing pit (phase II); layer 10, ash fill of pit (phase II).(b)A1/1.Layer1,treetrench(phaseV);layer2,topsoil(phaseV);layer3, compactearth(phaseIII);layer4,compactearth(phaseIII);layer5,clayinfillof robbertrench(phaseIII;notdetectedduringexcavation);layer6,clayinfillof robbertrench(phaseIII);layer7,rubbledeposit(phaseIIor III); layer 8, clay from decayed mudbrick (phase II); layer 9, mixed clay and ash from abandonment of A1/1 (phase II); layer 10, ash fill of basin (phase II); layer 11, clay floor matrix (phase II); layer 12, hard white surface beneath level of A1 wall c. Just one stone was still in position in the northern part of A1/1, though the bottom course wasbetterpreservedinA1/2.Theonlycloselydatableceramicsintherobbertrenchare IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200224 fromsixth-centuryIonicB2cups,butwesuspectthatthestoneswereremovedinthe fourthcenturyandusedinbuildingA3.AtthispointA1smudbricksuperstructurehad presumably decayed, leaving the top of the stone foundations exposed. A3 was built over the south part of A1s ruins, at the highest point on Monte Polizzo. Figure 12The stele and wall e to the east of building A1, facing south Phase III WefoundnofurtherphaseIIIstructuralremainsin2002,butdidrecovermorematerial scatteredinthetopsoillayersaroundA3(figures 15, 16), bringing the number of bronze coins to 9, and limestone dice to 5. All but one of the coins are of the familiar type with a horseandpalmtreeononesideandayouthfulfemaleheadontheother;theninthmay be of the same type, but is badly corroded.43 Drs.DonaldArielandBaruchBrandloftheIsraelAntiquitiesAuthorityhave drawnourattentiontosimilaritiesbetweenthe5-pointed-staramphorastampfoundin zoneAin200044and44suchstamps,withthesymbolsyrslmbetweenthestars points, found mainly in Hellenistic stratum 7 of Yigal Shilohs excavations in Jerusalem.45 TheIAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200225 Figure 13A fragmentary antler lying in front of the stele on A1 wall e Figure 14The jaw and loose teeth of an adolescent human, aged 12-14 years, found in the fill under the paved surface in trench M98 (photo D. Connolly) IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200226 MontePolizzostamplacksthewritingon the Jerusalem examples, and Dr. Ariel informs usthatthefabricandinclusionsintheMontePolizzoexamplehavemoreincommon with the Thasian example that we cited. However, the dates of the Jerusalem, Thasos, and Monte Polizzo examples match closely. Figure 15Phase III artifacts dispersed around structure A3 IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200227 Figure 16Fourth-century coins and dice from A3 (photo D. Connolly) 5.Zone B (figure 17) ExcavationcontinuedinzoneBtoclarifythesixth-centuryandmedievalbuilding sequencesandtofindoutwhetherB1wasafree-standingstructureorpartofalarger complexineachperiod.Thetwobalksshowninfigure17wereleftinplacetoavoid harming the trees of the Monte Polizzo Forest Preserve. Excavation in 2002 concentrated on the areas south and west of B1. Phase II We have now identified four distinct episodes of sixth-century activity. Firstepisode (phase II.a, c. 575-550 BC). The excavated stretch of wall h, 6.32 m long, belongstothisepisode.Wallhdisappearstothenorthwestunderlate-sixth-century wall e, and to the southeast into the unexcavated trench M109. A door 1.08 m wide interrupts h.Wallhwasbuiltintwodifferentstyles.Thestretchesnorthofthedoorandthe southofitasfarasthesouthedgeoftrenchL108aremadeofroughlyhewnsquared limestoneblocks,builttoformcourses(figure18).ThiswasthenormalstyleatMonte Polizzoinallphases;thehillsnaturalmatrixconsistsofalternatinglayersoflimestone IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200228 andsandstonemixedwithlarge,roundedboulders.Thesandstone/bouldermaterialison thesurfaceontheacropolis,andthewallslabswereprobablyquarriedfromalimestone outcrop400mnorthwestoftheacropolis,nearPortellaSantAnna.Typicallytheslabs are5-7cmthickinallphases,butwallh has several stones up to 25 cm thick. Just as h enterstrenchM108,itsconstructionmaterialchangestolargeroundedboulders.Similar boulders were used late in phase IV (twelfth century AD) to block two doorways in room B1/1andforwalljinM108.46UntilweexcavateM109wecannotknowwhetherthe southern part of h was a medieval addition, though that does not seem very likely. Figure 17Remains in zone B. Line #$# shows the profile in figure 22 Thenorthernendofhresteddirectlyonbedrock.Therockslopestothesouth, andwehavenotyetreachedthebottomofhsouthofthedoor.Southofthedoor,we haveexcavated13coursesofthewall(height1.2m),andhavenotreachedthebottom. IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200229 UntilweexposemoreofZoneB,wallhsfunctionsremainunclear.Therearenosigns yetofcross-walls;possiblyitmarkedtheboundaryofaspecialarea,suchasthesacred zone of the acropolis.At the end of the season we uncovered a rough line of flat slabs running west-east through trench L108 toward the door (figure 18). These stones may have supported a clay pathcurvingupthewestslopeoftheacropolisandthroughthedoorinh.Southofthe lineofstonesweidentifiedtwoclaysurfaces,bothpost-datingh.Ontheupperclay surface (L108 layer 11) was a small, round, black deposit of extremely hard charcoal and burning, disappearing into the balk with trench L109. This was probably a hearth. Figure 18Trench L108, facing east, showing the doorway through wall h and, in the foreground, the possible path leading up the west slope of the acropolis The finds from these clay surfaces are rather different from those in other parts of Monte Polizzo (see section 8a below). We found two almost complete indigenous vessels inlayer11,oneagraywaredipperandtheotheraglobularspoutedpotmadefroman unusualorange-graysandwichfabric.47Theclaylayersinthesupposedpathnorthof IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200230 thelineofstonesleadinguptowallh(layers9,13-18)includedPunicamphorasherds, an Etruscan bucchero kantharos handle of around 600 BC,48 three Corinthian or imitation Corinthiansherds,localgraywarebowls,andlocalbowlsofablackishbuccheroid fabric,darkerthantheusualsixth-centurygrayware.Someofthesehadinciseddentedi lupodecorationaswellasthenormalsimpleincisedbands.TheselayershadfewGreek sherds, and no Ionic or Attic. Secondepisode(phaseII.aorb,c.550BC).Layer9,theuppermostclaydeposit in the possiblepath,runsfromtrenchL108throughthedoorwayinhintotrenchM108;but afterlayer9formedthedoorwasblockedandathickfillwasdumpedeastofh,raising thegroundsurfacetothelevelofthe top of the wall. Probably at the same time, a stone drainwasbuilt,runningslightlydownhillfromthenortheasttoemptyoutoverh.A dumpofRubbleandclayeysoilweredumpedsouthwestofhtoraisethegroundlevel here(accountingforhsexcellentpreservation)duringepisode2or3.Alargemedieval pitdisturbedpartofthedrain,probablyremovingtwolargestonesfromitssoutheast edge (figures 17, 19). Wecannottellyetwhythedrain was installed, but it might be related to building B2undermedieval B1/1 (see figure 17), if further digging dates B2 to this episode. Wall bbsalignmentisclosetothedrains,andthefewsurvivingstonesfromwallddare perpendiculartoit.Thebedrockwascutbackverticallyimmediatelybehinddd;we suspectthatB2wasarectilinearroomassociatedwiththedrain.Thefallenbuilding blocksfromB2weredugupandreusedinphaseIV,andthepitthatthiscreatedwas partiallyrefilledwithadumpofroundbouldersandsmalllimestoneslabs.Thispit (figure 20) contained medieval pottery and roof tiles. We need to excavate deeper to date B2 securely. Thirdepisode(phaseII.b,c.550-525BC).Alargedepositofgrayashysoil,morethan 5.5macrossandupto50cmthickatthecenter,wasdumpedinzoneB(figure17), coveringpartsofthedrain.49Itscontentsareverylikethepotterydepositexcavatedin 2001inzoneAtrenchesMN100-101:veryfewfinewaresherds,butthousandsof fragmentsofstoragevesselsandbones,abovealldeerantler(seesection8aand IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200231 Appendix1below).However,micromorphologicalanalysisshowsthatthetwodeposits wereformedincompletelydifferentways.ThatinzoneAbuiltupslowly,probablyas clayfrommudbrickswasheddownslopeandgraduallycoveredanabandonedstorage area;whilethezoneBdepositwassweptupfromtheremainsofnumerousfiresin anotherlocation,anddumpedatonetime(seeAppendix2below).Wesuspectthatthe zoneB deposit was dumped here from zone A. We excavated a large ashy deposit just 5 m from the mounds edge in trench L106 in 2001,50 but this deposit otherwise had little in common with the ashy dump under B1/2. Bothdepositscontainedsmallfragmentsofiron,bronze,andworkedbone,but zoneBalsoincludedalavagrindstone(figure21),atinychainofbronzelinks,andan unevenlimestonedie.Wehavefoundfivelate-fourth-centurylimestonediceinzoneA (see above), but the zone B die was securely stratified under the phase II.c paved floor of B1/2, and cannot be intrusive.51

Fourth episode (phase II.c, c. 525-500 BC). Around 525 BC, the one-room structure B1/2 was built over the top of the dump of ash, storage sherds, and antlers. The retaining walls onthesamealignmentintrenchesK-L106-107alsobelongtothefourthepisode.We cannotbesurefromthesmallareasexcavatedin2001whetherthespacesnorthwestof B1/2wereopen.52OnlymedievallayerssurviveinM106recovered,restingdirectlyon bedrock.IronAgelayerssurviveimmediatelywestofthebalknorthofB1/2walla,but excavations in 2002 only exposed their surface. B1/2 had a paved floor, but only the northeast half of this survives, the rest having beenlosttoerosion(figure22).Wallsk-lfilledmuchofB1/2(internal dimensions 6.5 x 2.6m).In2001weassumedthatk,l,andbwerethree of the walls of a stone bin, with thefourthwallintheunexcavatedquadrantofB1/2,andthatthebinwasemptiedand reusedinphaseIV.However,butthe2002excavationexposedtheedgeofasecond stone structure, up to three courses high, continuing to the southwest on the same line as wallk.ThismayhavemajorrepercussionsfortheinterpretationofB1/2,butwecannot say much about it yet, since it lies in the unexcavated quadrant of B1/2. It either predates theashydump,inwhichcaseiteitherbelongswiththesecondbuildingepisodeinzone IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200232 B,orwasdugdownintothedump.B1/2andwallsk-lmayhavebeenbuilttocontinue the earlier structures as yet undefined functions in more formalized ways. B1/2waspeacefullyabandoned:sherdsfromtheabandonmentdeposit53seemto be fragments rather than broken whole vessels, and there is no trace of fire. We found no decayedmudbricklikethethickclaylayersfoundinA1.Weshouldconcludefromthis eitherthatB1/2 had all-stone walls, which the medieval occupants of zone B dug up and reusedtobuildtheirownhouse,orthatB1/2wasanopen-airenclosurewithlowwalls, rather than a roofed room. Phase IV WeexcavatedasmalltrenchtobedrockinthewesterncornerofB1/1toconfirmthe theorythatthisroomwasamedievaladditiontoIronAgeB1/2.Figure 23, facing west, showsthatthisisindeedthecase:whereaswalld(attheleftofthepicture)restson bedrock,withpureIronAgelayers16and 24-28 providing a terminus ante quem for its construction,wallainB1/1 (at the right) can be fixed firmly in phase IV. In the western part of B1/1 erosion had destroyed the medieval paved floor that we excavated in 2000 in theeasternpartoftheroom,butthesamefillofmedievaldebrisandrooftilesthatlay underthefloorintheeasternpartoftheroomstillsurvivedinthewest.54Thefinds includedalate-tenth-centurycarinatedbowlfragment.55Severalofthevesselsrecovered in 2000 from B1/1s abandonment deposit have now been reconstructed (figures 24, 25), and probably date to the twelfth century. Sofar,therearenosignsthatB1waspartofalargercomplex.Northofthe building,in trench M106, the poorly preserved wall aa snaked along the bedrock, almost certainlydividingexteriorspaceratherthanfurtherrooms.Tothesouth,B1/1wallb restedontopofearlier(probablyIronAge)wallsbbanddd,andadeepmedievalpit filled trench N107 (figure 19), perhaps dug to recover Iron Age building stones. Excavationin2002showedthattheareasouthofroomB1/2has a fuller record, with three episodes of activity. IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200233 Figure 19Deposits in M108, facing northeast. B1/2 wall b is at the left, with the ashy dump of phase II episode 3 and the earlier phase II drain visible beneath it. In the background medieval wall cc can be seen, and in the foreground the fill of the medieval pit, containing roof tiles, dug down beneath the level of the Iron Age drain IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200234 Figure 20North face of balk between N107 and N108, showing medieval pit (scale = 50 cm) First episode (probably eleventh century):There was an open area in trenches M-N 108. Ithadaclayfloor with a few paving slabs in N108. B1 then collapsed, leaving a heap of broken roof tiles up to 30 cm thick on the first medieval surface (figure 26). Secondepisode(eleventhortwelfthcentury):Alargepitwascutthroughtheearlier medievalandIronAgelayersinM108,inplaces(especiallytothesouth)reaching bedrock. The pit damaged the Iron Age drain, displacing at least two large stones from its southeast face (figures 17, 19). The pit was then filled in. Thirdepisode(twelfthcentury):ThespacesouthofB1/2wassubdivided.Wallj,made ofroundfieldstones,wasbuiltoverthepitinM108,andtheroughwallccaddedin N108.Wallccsmixedconstructionstylescanbeseeninfigure26.Theareabetweenj and B1/2 b was partly paved. A second tile deposit marked the end of this episode. We have not yet studied the 2002 finds closely, but they seem similar to those from B1/1, suggestingoccupationacrosstheeleventhandtwelfthcenturies.Therearenotracesof fire in the medieval layers. The incorporation of so many tile fragments in the matrices of thefloorsofthefirstepisodeandinlevelingfillsinB1/1,alongwithfindsoflate-tenth-IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200235 centurysherdsinmixedcontexts,stronglysuggeststhattherewereearliermedieval buildings nearby. Figure 21Grindstone in B1/2, facing north, with B1/2 wall a in background Figure 22Simplified profile through B1 IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200236 Figure 23B1/1 walls a (at right) and d (left), facing west. Beneath wall a are phase II layers 16, 24-28 6.Zone C (figures 4, 5, 28) We continued excavating in zone C to recover more of the early-sixth-century destruction inbuildingC1andtoexaminetheoriesthatsixth-centuryindigenouscommunitieslaid outtownsongridplans,imitatingGreekpractices.Wecarriedoutamagnetometer surveytoassessthistheoryin2001,butcouldnotdetectwalls,56andtheslopeistoo steepforgroundpenetratingradar.Weplantotryfurthernon-invasivetechniquesin 2003,butbeforethemainseasonin2002,wealsoexcavateda57-meterlong,1-meter wide trench (the Great Trench, or GT) from grid square M109 to Q120, linking zone B totheTusaHouse(figure27).WedividedtheGreatTrenchinto5sections.SectionsI and II, between M109 and the 2001 trench N/O 113/114, were 8.75 m long; and sections III-V,betweenN/O113/114andtheTusaHouse,were11.57mlong. IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200237 Figure 24White-glazed medieval bowl from B1/1, excavated 2000 (restoration A. Gjefle; photo D. Connolly) Figure 25Green-brownglazedmedievalbowlfromB1/1,excavated2000 (restoration A. Gjefle; photo D. Connolly) IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200238 Phase II WeimmediatelycamedownonsterilebedrockthroughmostofGTsectionI,butatthe southendexposedthreecoursesofGTIwalla(figure5),anIronAgewallrunning northeast-southwestatadifferentanglefromB1andtheTusaHouse(seefigure47 below). In section IV we exposed building C2, on the same alignment as B1 and the Tusa House.Wewidenedthetrenchto2mhere,andexcavatedaslicethroughthebuilding. Like B1, C2 may have been built in the sixth century BC, then reused in the eleventh and twelfthcenturiesAD.PartsofC2swallscanbeseeninatreetrenchsouthwestofthe Great Trench, showing that C2 was at least 7 m long. There is a third wall to the north. In sectionV,neartheTusaHouse,deepsoilhadwasheddownfromtheridge,andwedid not reach phase II levels. Figure 26Roof-tile layer 13, trench N108, facing north. Wall cc is at right; B1/2 wall b at the top left WeextendedtheexcavationofC1tothewholeofgrid square O113 (figure 28). Asin2001,wefoundthickrubblelayerscoveringburnedbuildingdebris(particularly charcoalandclay)andbeneaththat, a rich destruction deposit containing many shattered IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200239 wholevesselsandmuchcharcoalandbone.Thefinds included a broken lava grindstone, abronzering,theneckofanimitationCorinthianoinochoe,agraywareminiature amphorawholeexceptforahandle(figure29),fragmentsoftwoIoniccups,andmany grayware bowls, some of them decorated with incised triangles. In the center of area C1/1 wasmostofasmashedamphora(figure30)andbeneathit,acompleteplainbasin, shatteredwhenalimestonecolumnbaseorrollerfellonit(figure31).Theassemblage lacksIonicB2cupsandAtticblackglaze,andprobablydatestophaseII.a(c.575-550 BC), earlier than the destruction deposits from House I. Figure 27The Great Trench, looking south toward the Tusa House from M109 C1 has a complex stratigraphic sequence. The south end of wall a and the paved floor seem to be contemporary with the massive destruction layers postdating both; but as figure 32 shows, the north part of wall a goes off at a different angle from the south part, was more loosely built, and sits on top of the destruction layer. The whole excavatedIAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200240 Figure 28Building C1, showing the paved floor excavated in 2001 in the west part of C1/1, and the destruction deposit excavated in 2002 in the east part. North of C1, in GT II, are medieval walls a and b stretchofwallbrestsonthedestructiondeposits,whilewallcpredatesthem.The situationisnotyetclear,butitlooksasifC1washastilyrebuiltafterburningdown.If thelimestonecylindershowninfigure31isacolumnbaseratherthanaroller, modification of the destruction deposit before rebuilding C1 would explain how a column drumendedupontopofabasin.ItisalsopossiblethatthethickburnedlayersinC1 weredumpedherefromadifferentdestroyedbuilding,butthatseemslesslikely. However,wehavefoundnodistinctfloorsurfacebelongingwiththesecondphaseof wall a and wall b. IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200241 Figure 29Amphoriskos, C1 layer 24 (photo D. Connolly) About25mduewest(i.e.,downslope)ofC1,weobservedtheedgeofa substantialwallonthesurface,runningroughlynorthwest-southeastbetweentreelines. We picked up traces of this running for more than 70 m to the northwest. We opened 8 x 2m trench J/K 112/113 to examine the most southerly stretch visible (see figure 5). This wall, a, dates to the sixth century. Three courses survive on the upslope side, and just one courseonthedownslope.Itis80cmwide,whichisunusuallylargeforMontePolizzo (housewallsaretypically60cmwide),stilltoothintohavehadseriousdefensive functions.Itisalsoroughlybuilt,fromamixtureofflatlimestoneblocksandround fieldstones (figure 33). Wall a may have served to mark the acropolis off from the rest of thesettlement.Layer9,theoccupationdepositassociatedwithit,containedmostly graywarebutalsoafewCorinthianandSikeliotesherdsandmanyanimalbones.A secondwall,b,wasaddedontop of layer 9, perpendicular to a. Only the edge of b was visibleinthetrench.Itwasmadeentirelyfromfieldstones.J/K112/113wallsaandb definitelydatetophaseII, but we cannot be sure from such a small sample whether they belong with phase IIa or IIb. IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200242 Figure 30C1 layer 20, facing east. At center is amphora base in situ, with column drum just visible to its right Figure 31C1 layer 24, showing basin shattered by fallen column drum (cf. figure 30) IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200243 Phase IV TheGreatTrenchalsocastlightonthemedieval village. The phase IV wall noted in the edge of trench N/O 113/114 in 2001 proved to be poorly preserved, as did a second wall immediately to the north (walls GT II a and b in figure 28). But if structure C2 was first builtintheIronAge,itwasrenovatedinphaseIV,likeB1/2.Themedievalrubble includedabeautifulgreen-glazedbowl(figure34),withearly-twelfth-centuryparallels fromMonteIatoandSegesta.57Mr.GiovanniScimemi,who dug with Vincenzo Tusa at Monte Polizzo in 1970 as a student, tells us that the uppermost walls of the Tusa House in P/Q120/121weremedieval.ByreusingtheIronAgeruins,themedievalsettlersofthe west slope of the acropolis may have created a regularly laid-out village. Figure 32C1 wall a, facing west. Left of the scale, wall a is contemporary with the paved floor; right of the scale, wall a rests on top of the destruction layer 24, postdating the paved floor IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200244 Figure 33J/K 112/113, facing northeast, with the enclosure wall in the foreground Figure 34GT IV medieval green bowl (restoration, A. Gjefle; photo D. Connolly) IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200245 7.Zone D (figures 4, 5, 35) Phase II NorthofM/N98themodernsurfacedrops3-4m,thenaroughlylevelplateaustretches fromgridrows 96/97 through row 92. We dug here to see whether this area was part of theritualcenteraroundA1.Weexcavated80m2,findingpartsoftwostructures.Apart fromahandfulofmedievalandmodernartifactsfromthetopsoilandtreetrenches,all activityheredatestophaseII,andprobablyII.c.Althoughthereisnostratigraphic evidence for their relative dates, we are assuming that D2 was built after D1 (and perhaps evenafterD1hadbeenabandoned),becauseD2/1wallbblockedadoorbetweenD1/2 walls e and h. Figure 35Plan of zone D, showing walls and pottery in the abandonment deposit Thescarcityoffallenbuildingstonesandthehardclaylayersabovethefloor depositsmustmeanthatD1andD2hadmudbrickupperwallsonlimestonefoundations IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200246 4-6courseshigh. Walls d and f and the northern part of wall c are poorly preserved, but a, b, and e contained a mixture of stone types. A tree trench damaged g, but the surviving sectionisunusualinbeingmadeoflargeblocksofdressedsandstone(visibleatthe bottom of figure 39), more like altar A2 than any house walls on the acropolis. The floor of D1/2 had been dug down into the soft bedrock about 10 cm below the bottomcourseofwalld.D1/2measuredatleast6x5m.Ifthedoorinwallewas centrallyplaced,D1/2was6x9.5m.WehavenotexcavatedenoughofD1/2tosay whether it was an internal or an external space, but if it was roofed, there must be one or morecolumnbasesintheunexcavatedpartsoftrenchesI95andJ95.Thefindsinthe destructionlayerincludedanunusualtaperingcylindricallimestoneweight,height approximately20cm,withoneholedrilleddownitslongaxis,and4morethroughits shortend.Theabandonmentlayercontainedsomecharcoalandfragmentsofburned mudbrick, and fragmentary storage vessels dominated the finds. Figure 36Collapsed pithos in room D1/1 D1/1wasalsoalargespace.Tworectangularstonesplacedtogethermayhave been the base for a central pillar; if so, the room measured 7.5 x 5.2 m. Disturbed areas of rubbleatthenorthandeastedgesoftheexcavatedpart of trench K94 (the extreme east IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200247 edgeoftheexcavatedarea)maybelongtodestroyedwalls,butfurtherexcavationis required.D1/1sfloorwaspartlypaved.Amassiveabandonmentdepositrestedonit, containing421kg.ofstoragevesselfragments(figure35),includingonelargeand apparentlycompletevesselthathad collapsed on itself (figure 36). This deposit is similar tothestorageareaoutsidetheeastwallofHouseI,whichcontainednearly500kg.of pithos sherds.58 We also found three large hollow clay weights, shattered but complete, in D1/1 (figure 37). Figure 37Clay weight from room D1/1 (restoration A. Gjefle, photo D. Connolly) Despiteitssize,D1/1containedonly3smallfragmentsofbronzeand2ofiron, and few bones compared to other phase II deposits at Monte Polizzo. The scarce fineware sherdsincludedthebaseofaCorinthianaryballossandwichedamongpithossherds (figure38),asmallAtticblackglazesherd,andalamp.59Thenumerousamphora fragmentsincludeEtruscan,Punic,andwestGreekvessels.Thesecannotbedatedas closelyasfinewares,butsomeofthePunicshapeshavecloseparallelsfromlate-sixth- andfifth-centurylevelsatMotya.60Theabandonmentdepositalsoincludedbodysherds from Punic torpedo amphoras, normally dated to the fifth and fourth centuries. One of the Etruscan amphoras may be of Pys type 3A/B, rarely found on indigenous sites, but IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200248 Figure 38Base of a Corinthian aryballos trapped between pithos sherds, D1/1 normally dated 625-525 BC.61 D1/1 seems to be a storage space, abandoned in phase II.c. There is much less evidence for burning than in D1/2. RoomD2/1wasmuchsmaller,atjust4.2x2.8m(internaldimensions).Atree trenchhaddestroyeditswesternend,reducingtheareaavailableforexcavationstill further. We quartered the room and excavated two of the quadrants (figure 39). Although we excavated less than 5 m2, we recovered almost 100 kg. of pithos fragments, including PunicandEtruscanamphorasverylikethosefromD1.ButD2/1differedfromD1/2in havingmoresignsofburning,andinyieldingthesamenumberofmetalfragmentsfrom a much smaller area, including the only precious metal so far found at Monte Polizzo, the endofasilverpin(figure40).D2/1alsoproducedseveralverythinsheetsofwhite limestone, which may be roof material. IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200249 Figure 39Room D2/1, divided into quadrants for excavation, facing northeast 8.Discussion 8.a)Phase II Religion Wehaveonlyexcavated215m2inzoneA,whichisamuchsmallerareathanhasbeen uncoveredatthecomparableindigenousreligiouscenters(seefigure45);andthe discovery of structure A5 and A1 pavement e at the end of the 2002 season shows that we still have surprises in store. But the evidence already collected suggests a lot about sixth-century practices at Monte Polizzo.First,weshouldclarifyourterminology.Webelieveitisappropriatetocallthe activitiesinandaroundA1religious.Therehavebeentwobroadtrendsinthe anthropologyofreligion.ThefirstbeganwithWeberandDurkheim,andsoughtto clarifyatranshistoricalanalyticalcategoryofreligion,whichwillallowsystematic comparisonsbetweendifferenttimesandplaces,buildingtowardageneraltheoryof whatreligiondoesandhowitdoesit.62ThemostwidelycitedexampleisClifford IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200250 Geertzsproposalthatareligionis:(1)asystemofsymbolswhichactsto(2)establish powerful,pervasive,andlong-lastingmoodsandmotivationsinmen[sic]by(3) formulatingconceptionsofageneralorderofexistenceand(4)clothingthese conceptionswithsuchanauraoffactualitythat(5)themoodsandmotivationsseem uniquelyrealistic.Geertzunpackedthisdefinitionwithcare,suggestingthatIn religiousbeliefandpractice a groups ethos is rendered intellectually reasonable by being showntorepresentawayoflifeideallyadaptedtotheactualstateofaffairstheworld-viewdescribes,whiletheworld-viewisrenderedemotionallyconvincingbybeing presentedasanimageofanactualstateofaffairspeculiarlywellarrangedto accommodatesuchawayoflife.63Thesecondtrend,rooted in deconstruction, saw the beliefsandsymbolsthatthefirstwouldclassifyasreligiousaspartsoflargersetsof authorizing practices. In a widely read formulation, Talal Asad asserted that there cannot beauniversaldefinitionofreligion,notonlybecauseitsconstituentelementsand relationshipsarehistoricallyspecific,butbecausethatdefinitionisitselfthehistorical productofdiscursiveprocesses.64Asadscorethesisisthatwesternsocialscientists havebundledcertainthingstogetherasreligiousbecausetheyareactingwithina culturallyspecificmodernistpowerstructure,whichseekstodistinguishitsown intellectualsystem(whichbreaksrealityapartintolaw,economics,politics,etc.)from older ones which lumped all these things together.65 Figure 40Fragment of a silver pin from room D2/1 (photo D. Connolly) IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200251 Alongsideboththeseapproaches,wemightsetthedominantideaamong historiansofreligion,who accuse anthropologists of all stripes of reductionism. As many historiansseeit,theclassicapproachesseektoreducereligiontoareflectionof somethingelse,reflectingsocialnormsbackandseekingtojustifythem,whilethe deconstructivistapproachesdenytherealityofreligioninfavoroftherealityofpower. Bothlittertheiranalyseswithquotationmarks,andrarelyattemptto understand religion astherealityitspractitionersexperienced.Byeschewingbroadcomparisons,historians claimtoreachmuchfullerunderstandingsconcentratingofpractitionersactionsand beliefs within their specific historical context.66 Notsurprisingly,archaeologistshavefavoredthefirstanthropologicalapproach. Theyhaveevenmoredifficultiesthansociologistsandhistorians,becausetheydealwith muteartifactsratherthanparticipantobservationorverbalreports.Theeasiestwayto copewiththisistosetupauniversaldefinitionofreligion,andtoseekmaterial correlates.Inourpreviousreport,wedrewonRenfrewandBahnswidelyused discussionofthearchaeologyofreligion.67Thishasthetwinmeritsofemphasizing Webersthesisoftheotherworldlycoreofreligion,orientinghumansinthehere-and-nowtowardinvisiblebeings,andofbeingsomethingarchaeologistscanoperationalize.68 Itremainstobeseenwhetherarchaeologistscanoperationalizeeitherthedeconstructive anthropological approach or the historical approach to religion. IntheWeberiansenseofreligionandfollowingthecriterialaidoutbyRenfrew andBahn,acropoliszoneAmustbeaccountedareligioussite.A1seemsdesignedto focusattentionontheboundariesbetweenthisworldandanother,bringingparticipants togetherintranscendentcommunication.A1occupiedthehighestpointonMonte Polizzo,usedanarchitecturalformevokingroundhousesthathadbeennormaluntilthe eighthcenturybutwerenolongerused,and(atleastinitslatestphase)hadaspecial enclosure,markedoffbyA1walld.ThestelestillinpositionjusteastofA1perhaps symbolizedthedivine,andaverylargestoneblockfoundoutofpositionin2001may havebeenanotherphaseIIstele.69TheroundclayhearthinA1/1andpitinA1/2,both heaped with ash and the latter (and perhaps also the former) sealed with baked clay when A1wasabandoned,suggestspecialactivity;sotootheevidenceforanimalsacrificeat thetwostoneplatformsthatwehavecalledaltars.Themanyfragmentsofwinecups IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200252 suggest that intoxication was an important part of the activities in zone A. The apparently openspacemorethan25mwidebetweenA2andB1/2isunusual;sotoothecontrast betweenmassiveconcentrationsofheavy-dutystoragevesselsaroundA2andinzoneD (andalsointhedumpunderB1/2,ifthatdidindeedcomefromzoneA)andthe preponderanceoffinewaresinA1.WesuggestthatA1wasapointofreligiouscontact with a divine sphere, and advance some tentative theories about how this contact worked. Hellenization and religious architectureHellenization,inthesenseofnativeadoptionofGreekpractices,pantheon,andbeliefs, hasdominated discussion of Iron Age Sicilian religion.70 Much of this has been fueled by findsatSabucina,whereSicansflirtedwithfeaturesofGreekreligiousarchitecture acrosstheseventhandsixthcenturies,combiningporchesandcolumnswithtraditional localroundhuts,aswellasmakingremarkablemodelsofshrinesthatsimilarly manipulate Aegean forms.71 WefocushereontwoaspectsofreligionatMontePolizzo:inthissection, architecturalforms,andinthenext,sacrifice.In both cases, we see connections between indigenousSicilianandAegeanpracticesbutemphasizethelocalcontextmorestrongly. Weconcludethatproperanalysisneedstobemoreempiricallydetailedandmore conceptually precise than has often been the case. Thereareintriguingparallelsinreligiousarchitecturebetweenthosepartsofthe Aegeanmostinvolvedintradeand colonization and western Sicily.72 The worship of the gods produced few archaeologically visible traces or distinctive architecture in either area before about 750 BC. Then, between 750 and 700, attention-focusing devices appeared in bothregions.InthecentralAegean,whereapsidalhouseswerenormalintheeighth century,largerversionsofthesewerebuiltforthegods,andincreasinglyusedwallsand otherdevicestodistinguishsanctuariesfromnon-religiousspace.73InwesternSicily,a similarprocessunfolded,exceptthatthegodshouseswerelargerversionsoftheround hutsthatwerenormaldwellingsintheeighthcentury.74PossiblyGreekstaughtwest Sicilianstheideaofbuildingelaborateversionsofhousesashomesforthegods;but giventheformaldifferencesbetweenAegeanandwestSiciliantemples,thenear-certaintythatthe first round hut-shrines in west Sicily predate the first Greek settlements IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200253 atHimeraandSelinous,andtheabsenceofmonumentaltemplesfromGreekcitiesin Sicilyuntilthelateseventhcentury,thisseemsimplausible.Mostlikely,wesuggest, peopleintheAegeanandwestSicilyrespondedtosimilarproblemsinsimilarways:as populationgrewinbothareasandcompetitionforresourcesincreased,peoplechoseto invest more heavily in worshipping the gods.IntheAegeanandwestSicily,rectilinearhousesreplacedcurvilinearonesinthe seventhcentury.75IntheAegean,rectilineartempleswithdistinctDoricandIonic architectural orders replaced apsidal temples by 600. In west Sicily, however, round hut-shrineswerecommontill500.Sixth-centuryshrinesinSicilythusevokedancestral tradition in ways that contemporary Aegean temples did not. It is interesting to speculate onwhyindigenousSicilianreligionmighthaveremainedfocusedonancestorswhilethe emphasisintheAegeanmovedtowardadistinctrealmofOlympiangods,butthereis currently no way to ground the issues empirically.76 IntheAegean,roundreligiousbuildingswererare,butdidexist.StructureVIII from Lathouriza near Athens (c. 700 BC) is the earliest known example.77 Round shrines becomemorecommoninthelatesixthcentury,butmostdatetothefourthcentury.78 Two generations ago Fernand Robert argued that Greeks sometimes used round religious buildingsforchthoniccults,aimedatspiritsandheroeslivingunderground.79He concludedthat round buildings often had sacred pits (bothroi) rather than altars (bomoi), because pouring the blood from a sacrifice into a pit aided its passage to the dead beneath theground,whilesplashingthebloodontoan altar did not. He also suggested that there werespecialformsofchthonicsacrificeemphasizingblood,includingtheholokaustes, wherethecompletevictimwasburned,insteadofdividingthebestpartsamongthe worshippersandonlyburningtheinediblepartsforthegods.Robert derived the circular form from the enclosure walls of Bronze Age burial mounds, and the focus on blood and pits from the need to feed the dead. Robertscharacterizationofroundchthonicshrinescertainlycorrespondswith A1sform,thepitinA1/2,andsomeofitsburnedbones.Ifthehumanjawfoundin trenchM98camefromanearbygrave,thatmightevenhavebeenthefocusofthecult. OtherSicilianhutshrinesalsohavesimilaritiestoRobertsmodel.AtColleMadore,pit 40waseitherdugimmediatelyoutsidehutAoragainsttheinnerfaceofthewallofhut IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200254 B,datedc.550-525BC,inwhichcaseitspositionaswellasitsdatematchthepitin MontePolizzoA1/2.Pit40containedash,charcoal,andanimalbones.80Thesecond floorsinPolizzellohutshrinesAandB,probablydatingtotheearlysixthcentury,also containedpits,sealedbysherdsfromlargepots.HutDhadunasortadibothros at its center. These pits contained ash, charcoal, animal bones, fragments of bronze and amber, and small vases.81 Butweshouldalsobeawareofdifferences,bothbetweentheSicilianstructures, andbetweentheSicilianstructuresasagroupandtheGreekexamples.Claybasinsand roundhearthsareevenmoreprominentthanpitsinSicily,thoughtheydonotfeatureat all in Roberts model or most excavated tholoi in Greece. The basin in A1/1 has parallels inMontagnolihutshrines 1 and 7, although these were larger (diameters 0.9 and 1.0-1.4 mrespectively),earlier,anddecoratedwithringsofimpressedcircles.82Polizzellohut shrineA,datingtothelater-seventhcentury,hadanaltarecircolareapiattaformedi pietrelleatitscenter,andsemi-circularstructureChadaround clay basin, heaped with ashandsmallanimalbones.83LikeA1,thefirsttempiettoatSabucina(DeMirosblue phase),aseventh-centuryrectangularroomwithabench,combinedpitsandbasins, although neither of the round hut shrines that succeeded it had these features.84 WithintheSiciliangroup,A1hasseveralunusualfeatures(seefigure45):the other hut shrines are mostly larger (at 8-10 m diameter compared to A1s 6.4 m), single-roomedwithbenches,earlier(goingbackintothelate-eighthcentury),andhavericher finds.Noneofthehutshrineshasproducedrichvotiveslikethosefoundinarchaic Aegean sanctuaries, leading to a general conclusion that Elymian gods did not want gifts. As table 1 shows, metal and worked stone and bone finds are about twice as concentrated around A1 as in zone B,85 but the finds from A1 look like discards and accidental losses, not dedications. This suggests that more metal was used in A1 than in zone B, but either itwasnotactuallygiventothegods,orwhenA1wasabandonedpeoplefeltfreeto removeitems.A1,then,isconsistentwiththetheorythatvotiveswerenotpartof indigenous cult; however, the evidence from other sites is more complicated. The objects foundinthehutshrinesat Montagnoli and Sabucina also seem to have been used, rather thandedicated,althoughthefifth-century(yellowphase)shrineatSabucinahadapit filledwithvotives,ash,andpigsjawbones,andtwopitswithvasesandsmallanimal IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200255 bones(includingastragals)werenotedatMontagnoli.86AtColleMadoretoolittle survivesfromtheroundhutsatthesummitto draw a conclusion, but the sacello had no votivesapartfromDepositA,probablyafoundationdepositfortheshrine,containing objects spanning the whole period 850-550 BC. The brief Polizzello report speaks of pits with metal and pottery votives, and even an enclosure for offerings. Finally, a remarkable deposit at Montagnola di Marineo contained three bronze helmets, two bronze schinieri, a possiblebronzeshieldbossandironweapons,ajugcontainingremainsofatleasttwo lambsorgoatkids,andanivoryplaqueintheshapeofaram,alldatingc.525-500.87 Overall,thepictureismixed:votiveswerenotasimportantasinarchaicGreek religion, butdidoccur.SabucinaandMontagnoli,whichendedinsuddendestructions,hadsome richfinds;MontePolizzoA1 and the Caltabellotta hut shrine, both abandoned peacefully in the early-fifth century, have the poorest finds.88 WediscussedA1suniquedivisionintothreesmallcompartmentsinsection4 above. At this point, the evidence suggests that A1 was originally single-roomed, and that wallsbandcwereaddedinasecondphase,alongwithexternalstructuresd,e,the possiblewallintheM98/99balk,andthealtar(ifsuchitis)inareaA1/4.Ifthiswasan open-airaltar,ithasfewparallelsonindigenoussites.OnlyMontagnolihasapossible open-airaltarinofromtheeighth-/seventh-centuryphaseintrenchM40,andaprobable smallaltarfromtheearly-sixth-centuryphasewithhut1.89However,itseemsalmost certain that Monte Polizzo A2 was an open-air stone structure that served as the focus for burnedanimalsacrifices.TheleadingauthorityonGreekreligionsaysthatthemost essentialelement[inasanctuary],moreessentialthanthecultstone,tree,andspring, [was] the altar, bomos, on which the fire [was] kindled.90 A2 may have been a direct and important borrowing from Greek religious practice.ButiftheSiciliandataaremessy,theAegeanmaterialisevenmoreso.New excavationshaveshownthatroundcultbuildingsinfacthadmanyuses,particularlyas diningrooms,andFredCooperandSarahMorrishavecriticizedmorphological fallacieslikeRoberts,linkingshapetoaspecificfunction.91Heroescouldhavethe samekindsofaltarsasOlympiangods,andtheterminologyofheroicandOlympian religionwasless clear-cut than it seemed in Roberts day.92 Even the distinction between blood sacrifices flowing into the earth and sacrifices to the gods in which smoke rose into IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200256 theskyhasfragmented:aninscriptionfromSelinousdatingc.450specifiesthata sacrificeshouldbeperformedasisnormaltothegods,butthenadds Let him slaughter (the victim so that the blood flows) into the earth.93 Aegeanroundbuildingsandpitscouldhavechthonic associations, but did not do souniformly.Further,ifwestSiciliansborrowedGreekideasaboutchthoniccult,they massivelyreinterpretedthem,regularlyplacingtheirhutshrinesatthehighestpoint available,asfarfromtheunderworldasitwaspossibletoget.94Butperhapsmost importantly, the chronology of round shrines in the Aegean and Sicily is inconsistent with thetransmissionofformsfromtheformerregiontothelatter.Thedatessuggestthatof roundchthonicshrineswithpitsweretransmittedfromoneculturetoanother,themost plausible direction is from Sicily to the Aegean. The most plausible scenario is that Greek settlersadoptedsomeindigenousreligiousactivities,includingroundbuildings,but adoptedthemtotheirownpurposes,including(butnotrestrictedto)chthoniccult. Agrigento may have been the major point of interaction.95 Hellenization, drinking, and sacrificeTheseconddimensionwewishtodiscussisreligiousconsumption.Winedrinkingwent onatallthehutshrines.ThefindsfromHouseIandzonesBandzoneCatMonte PolizzoshowthatGreekkylikesandwineamphoraswerenotrestrictedtoreligious settings, and surveys have found Greek wine cups even on the tiniest rural sites.96 Vessel shapes and the prominence of strainers suggest that beer, not wine, was the main drink in thetenthandninthcenturiesBC.GreekcupsandjugsandPhoenicianandGreek amphorassuggestthatwinewascommonlydrunkineasternSicilybythelateeighth century,andinthewestbythelateseventh.Atpresentwecannotsaywhether PhoeniciansandGreeksreintroducedwineintheeighthandseventhcentury,whether SicilianshadbeendrinkingitcontinuouslysincetheBronzeAge,orwhetherSicilians grewtheirowngrapesintheBronzeAge,abandonedthemintheEarlyIronAge,then startedcultivatingthemagainaround700orlater.Asinglevitissylvestrisgrapeseed, probablywildratherthancultivated,hasbeenfoundinahouseofthisperiodat Morgantina;97theearliestdefiniteevidencefromSicilyforvitisviniferis,thecultivated winegrape,isacarbonizedgrapeseedfoundin2002intheashydumpunderMonte IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200257 PolizzoB1/2(figure41;seeAppendix3),firmlydatedc.550-525.Eventhisdoesnot guaranteethatindigenousSiciliansweregrowingwinegrapesinthesixthcentury,of course;therewasaflourishingraisintradeinclassicaltimes,andshipwreckevidence fromTektas Burnu shows that imported amphoras could contain seeds as well as wine.98 Butitdoesmakesixth-centurylocalwineproductionalongsideimportsinGreek, Punic,andEtruscanamphorasverylikely.Interestingly,insouthernFrancevitis sylvestrisappearsconsistentlyonabout10percentofLateBronzeandEarlyIronAge sites,whilevitisviniferisiscompletelyunknownbefore600BC,butappearson26 percentofallsixth-centurysites.99ThecombinationoffindsfromMontePolizzo southernFrancemaymeanthatvinecultivationwasestablishedrelativelysuddenly acrossthewestMediterraneaninthesixthcentury,presumablybyGreektradersof settlers. Figure 41Carbonized vitis viniferis seed from ash dump B1/2 layer 15, c. 550-525 BC. Scale 1 cm (drawing H-P. Stika; see Appendix 3) IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200258 Tara Hnatiuks study of more than 30,000 bone fragments from zones A and B100 also provides important information about sacrifice and feasting. Most of the bones come from concentrations immediately north of A2, around the possible altar in A1/4, and from the dump under B1/2.101 3,649outof18,042bonefragmentsstudiedfromzoneAwereidentifiable(20 percent).Table2showstherepresentationofthemajorspecies.Thefigureof37.5 percentforcervuselaphus(reddeer)ismostunusual.Venisonmayhavebeenan importantfoodsourceinancientSicily;roeandfallowdeerboneswereprominentina first-century AD farmhouse at Montallegro, and a large deposit of fallow deer bones was found in a third-century BC shop at Morgantina.102 The Morgantina deposit, like those at MontePolizzo,containedmostlyheadandfootfragments,andtheexcavatorsplausibly interpreteditasbutchersdebris.Neitheroftheselaterdepositshadsignificantamounts ofreddeer;butthisspeciesisknownfromnon-religioussixth-centurysitesinwest Sicily.AtEntella,10ofthe241identifiedbones(4.1percent)fromlayersdatedc.600-450inafeastingareaattheedgeofacemetery were cervus elaphus, and at Monte Iato threeantlershavebeenfoundintheGreekcourtyardhouse,datingaround475.103At MontePolizzo,acompleteantlerwasfoundinsideabrokenstoragejaratthePortella SantAnnain2000,andalargefragmentinzoneDin2002;and11-21ofthe232 identifiable fragments from the Profile in 1998 (4.8-9.1 percent) were cervus elaphus.104

But none of these contexts begins to compare with the quantity of finds from zone A.TheonlydepositthatdoescompareisthedumpunderB1/2,where4,835fragments out of 12,537 analyzed could be identified (38.6 percent). Red deer make up no less than 82.3percentofthecollection(table3;figure 42). As noted on p. XX above, the pottery foundinthedepositnorthofA2andthedumpunderB1/2issosimilarthatthelatter depositmayhave originally come from zone A. The micromorphology of the B1/2 dump is consistent with this (see Appendix 2 below). The quantitative pattern, and the specifics oftheA2andB1/2bonesdescribedbelow,suggestthatwhiledeerweresometimesan importantfoodsourceinancientSicily,theyalsohadaspecialreligioussignificancein the sixth century BC. Theminimumnumberofindividualdeerpresentis just 7, but the total number of fragmentsandtheirweightsuggestthatthetruenumberisinthedozens.Althoughthe IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200259 statisticsprobablyoverrepresentdeer,becauseantlers(whichmakeup96percentofthe deerbones)arebotheasilyidentifiableandeasilyfragmented,theprominenceofdeer antlers in these two deposits is nevertheless highly unusual.105 Figure 42Antler fragments in B1/2 ashy dump layer 15, c. 550-525 BC The state of fusion of phalanges I and II in the antlers shows that most of the deer inthesedepositswerekilledaroundNovember(although the presence of two fetal tibias suggeststhatsomedoeswerekilledinthelatespring).106Pollenrecoveredfromthe Profile and House I in 1998-2000 indicates that there was little forest cover around Monte Polizzointhesixthcentury,andsoprobablyveryfewdeerimmediatelyroundthe settlement.107Thereweredeerparksinsome Greek sanctuaries, but there is no evidence foranythinglikethisinSicily,andtheseanimalswereinanycasenormallyprotectedby the divinity, not sacrificed.108 The most likely scenario is that deer that hunting parties set offfromMontePolizzotothewoodedmountainsofnorth-centralSicilyasautumn turnedtowinter(thesametimeasthehuntingseasoninthecontemporarynorthern hemisphere),killeddeer,andcarriedthecarcassesbackwholetoMontePolizzo.There, theydismemberedthemonaltarA2,andperhapsalsoonthestonestructureinspace A1/4.Theironcleaverfoundnearthisstructuremayhavebeenusedtochopupthe sacrificialvictims,andtheironarrowheadfoundimmediatelywestofthe structure could IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200260 evenhavefallenfromthebodyofadeer.109Peoplethentookthemeatyjointsonlong bonesaway,eithertoseparatefeastingplaces,or(asoftenhappenedinGreece)totheir homes.Buttheyburnedsomeoftheheadsandfeetonthealtars,leavingthedebris scatteredaroundzoneA.6.6percentofthebonesstudiedtodatewereburned;and84 percent of these burned bones came from around A2 and the possible altar in space A1/4. Morethanhalftheseboneswereburnedathightemperatures,over400C,whichis consistentwiththefiredamagetoaltarA2.11067percentoftheboneswerehead fragmentsandafurther10percentfromthefeet,withjust16percentfromtheaxial skeleton and 7 percent from limbs.111 Archaeologistshavenotedantlersatothersixth-centurywestSicilianhitshrines. NofaunalstudywasundertakenatPolizzello,butthebriefreportneverthelessmentions antlersfromhutshrinesAandB;DiRosasanalysis of the bones from pit 40 under two poorlypreservedhutshrinesfromColleMadorefound6fragmentsofcervuselaphus (2.3percentoftheassemblage); and the report on bones from the Malophoros sanctuary atSelinouscommentsthatcomparedtosheep,goat,andcow,Piscarsi sono i cervidi, deiqualifiguranopartidelcapo(corna)edegliarti.112Thesefindssuggestthatdeer played a religion all across west Sicily. TheantlersreceivedspecialtreatmentatMontePolizzo.Whiletheskeletal elementsshowednosignsofcutting,theantlershadnumerouscutmarks,not only from the separation of the antler from the skull, but also where the tine and beam met.113 A few fragmentshaddefinitelybeenworked;figure43 shows an example from the dump under B1/2, in which a knife was used to carve a groove around a tine. Butifwearerightthatdeerwereimportantinsixth-centuryreligion,whatwas theirrole?Severalscholarshavediscussedthewellknownseventh-andsixth-century plastic handles and cordon-decorated vases from Segesta, Poggioreale, Entella, Naro, and Polizzellowhichseemtoshowpeoplewithhorns.Leightonsuggeststhattheymay cleverlycombinecurvedandgeometricmotifsinsuchawayastosuggestfacesand hornedanimals:perhapsanintentionallyambiguouspersonificationofadivinitywith bothhumanandanimaltraits.114Figure44showsthemostfamousindigenousSicilian vasepainting,anoinochoedatingc.600-550BCfoundinatombatPolizzellointhe 1920s.Somescholarsbelievethatitshowsawarriorinamainlandtypeofhatwitha IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200261 widebrim,butputtingitintothecontextoftheotherhornedhuman representations and the Monte Polizzo faunal evidence, we suggest that it shows a dancing man wearing deer antlers.Thecutmarksonthefindsareconsistentwithchoppingantlersdowntoasize thatcouldbeusedinthisway,115andthegrooveinfigure46couldbeforattachingthe antler to a headdress. Reconstruction of ritual and belief from archaeological data is always a hazardous process,butbringingthevariousstrandsofevidencetogether,wesuggestthatthe iconographycelebratedanimportantmomentinwestSicilianrituals,inwhichthe boundariesbetweenanimals,men,andgodsbrokedown.Afterbutcheringthedeer carcasses, perhaps feasting on their meat nearby, burning the skulls and feet on the altars, anddrinkingwine,thatthehighpointoftheritualsmayhavecomewhensomeorallof thecelebrantssteppedintoliminalpositionsbetweenthisandotherworlds,atwhich point they literally walked with the god(s).116 Figure 44Early-sixth-century oinochoe from Polizzello IfwearecorrectinourinterpretationoftheritesatMontePolizzo,religious practiceherewasquitedifferentfrommainstreamGreekbehavior.TheonlyGreekcults IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200262 that seem to have any connection at all are those of Apollo and particularly Artemis, both deitiesassociatedwithhunting.IntheIliadArtemiswaspotniatheron,mistressofthe animals,117 and in the Odyssey Artemis with her arrows striding down from a high peakTaygetos towering ridge or Erymanthos thrilled to race with the wild boar or bounding deer, and nymphs of the hill race with her, daughters of Zeus whose shield is storm and thunder, ranging the hills in sport, and Letos heart exults as head and shoulders over the rest her daughter shines, unmistakableshe outshines them all, though all are lovely.118 Figurinesandotherrepresentationsofdeerarecommonineighth-andseventh-century Greeksanctuaries,and14latearchaicandclassicalofferingsareknowndepictingdeer withfemales,10ofthemfromknownsanctuariesofArtemis.119Sixth-andfifth-century AthenianvasepaintingsusedthebowanddeerasvisualcuesforArtemis,andbythe thirdcenturythepursuitofdeerwashermainassociation.120ApoeminthePalatine AnthologydescribesLykormashangingadeershideandhornsinasanctuaryof Artemis.121PausaniassaysthatinRomantimes,deerandgazelles(alongwithother animals)werethrownontoArtemisLaphriassacrificialbonfireatPatras,andmost scholarsassumethattheRoman-eracustomofoffering her cakes shaped like deer was a substitutionforofferingreal deer.122 But the most famous story goes back at least to the fifthcenturyBC:afterAgamemnonkilledastaginArtemissacredgroveatAulis,the goddessdemandedthathesacrificehisdaughterIphigeniabeforehecouldobtain favorablewindstosailtoTroy,onlytorelentatthelastmomentandallow him to kill a doe in her place.123 ButdespitethestrengthoftheGreekliterarytraditionidentifyingArtemiswith deer, deer-hunting, and deer-sacrifice, the archaeological evidence for this translating into thesacrificeofdeerisverythin.Theremainsoftwodeerandagazellewerefoundin ArtemissanctuaryatEphesus,andantlers(alongwithboarstusks)atKalydonand IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200263 Lousoi.124ReddeeraremoreprominentinthesanctuaryofArtemisandApolloat Kalapodi,whichhasaremarkablerecordspanningnearly2,000years,beginningc. 1150 BC.Herereddeerfluctuatedbetween9.0percentofthetotalassemblageofidentified mammal bones (c. 1050-950 BC) and 2.9 percent (c. 900-700). Red deer were always the mostcommonwildanimals.Astable4shows,theirprominenceinthetotalmammal assemblage was primarily a function of the prominence of wild animals as a group, which sharplydeclinedafter900BC.Surveydataandpollendiagramssuggestbetween1150 and 900 BC population densities fell lower in mainland Greece than at any time since the earlythirdmillenniumBC.Theinverserelationshipbetweenwildanimalsandhuman populationdensity,andthesteadydeclineintherepresentationofreddeerwithinthe wild animal group as human population rose, suggests that the history of deer sacrifice at Kalapodimayhavebeendrivenasmuchbypatternsinhumandemographyasbyritual considerations.125Webadlyneedlong-termpatternsfromothersanctuariestocompare with Kalapodi. Traces of deer have been found at the sanctuaries of Athena at Lindos and Tegea, Hera at Perachora, and Demeter at Knossos,126 also suggesting that in practice the link between deer and Artemis was not as strong as the literary sources imply. NoneoftheArtemissanctuarieshassuchprominentdeerremainsasMonte Polizzo,andevidenceforspecialemphasisonantlersisevenlesscommon.At Kalapodi, only39ofthe757reddeerboneswereantlers(5.2percent).127Theexcavatorsat Ephesusfoundakindofgoat-hornaltar,parallelingliteraryaccountsofsuchanaltarto ArtemisbrotherApolloonDelos;andanotherhornaltarwasfoundintheseventh-centurytempleofanunidentifieddeityatDreros.128TucheltandHggsuggestthat characteristicpartsofsacrificedanimals,suchasgoathornanddeerantler,were preserved in sanctuaries as a kind of monument to piety, but all the excavated evidence in Greece comes from goat rather than deer.129 The Monte Polizzo finds have certain similarities with Greek round cult buildings, andperhapsalsowiththeworshipofArtemis.Butin both cases, the evidence for Greek practices is messier than the normative statements in the literary sources, and the clearest Greekevidencedatesafterthesixthcentury.130MostdiscussionsofHellenizationhave goneonwithinwhat,insection1,wecalledthetraditionalframework.Thisposited discrete,bounded,materialcultures,directlylinkedtoElymian,Greek,orSicanidentity. IAN MORRIS, TRINITY JACKMAN, EMMA BLAKE, BRIEN GARNAND, AND SEBASTIANO TUSA MAAR 47, 200264 ButtheevidencefromtheAegeanshowsawiderangeofbehavior.Therewerecentral tendencies,whichallowedGreekstoidentifythemselvesasdistinctfromEgyptiansor Persians,andtotellacultofArtemisfromoneofZeus,butthesewereconstantly recreatedandrenegotiatedinpractice.Thishastwoconsequencesforarchaeological analysis. First, we must work from precise statistical summaries of the observed evidence in different regions and times, not from sweeping normative generalizations based on late literarysources.Withoutprecision,itishardtoknowwhatcouldpossiblyfalsifyany particulartheoryaboutthetransmissionofGreekculture.Second,wemustthinkabout contactsbetweenPhoenicianorGreeksettlersandnativeSiciliansintermsofgive-and-take, with different individuals and groups trying to make sense of the situation, and in so doing, sometimes creating wholly new situations: whateverassociationsofplaceandculturemayexistmustbetakenasproblemsfor anthropologicalresearchratherthanthegivengroundonetakesasthepointof departure;culturalgeneralizations(likeethnicandnationalones) must be understood ascomplexandcontingentresultsofongoinghistoricalandpolitic