[Dossier d’Archéologie polynésienne ] ne2 Bilan de la recherche archéologique en Polynésie frangaise 2001 - 2002 Testes réunis par HENRI MARCHES! Service de la Culture et du Patrimoine Punsauia - 2003(DE PAPETOA a ‘COMMUNE DE MOOREA - COMMUNE: HE ANCIENT «HOUSE SOCIETY>> OF THE ‘OPUNOHU VALLEY, MO’OREA: _ Overview of an archaeological project 2000-2002 Jenwiren Kau and Parnice V, Kinci™ a Résumé? Les auiteurs présentent les travaux orchéologiques conduits sur des sites d'habitation du secteur Tupaururu de la sallée de ‘Opunohu. Le choix des campagnes de filles 2001-2002 s'est porté sur trois complexes archéologiques, SeMo 170-17, 120 et 133, représentotifs des divers modesd accupa- tion de Vespace. Les objectifs de étude sont multiples: la réalisation d'une cartographie des sites 1a fouilte des structures afin de documenterleur fonction et leurréledans les activités de production et de consommation: lésaluation des variations architecturales etleurs éléments caractristiques, identification de la culture materieleassociée. Ces éléments permetient d’@laborer des hypotheses surlestatut social des unités d’habitation. Enfin.ilest possible d’éablirune chronologie précise de occupation & partis des datations radicarbones obtenues, Pour chaque ite, celles-ci permettent denvisager la possibilité de deux phases d’oocupation. l'une initiale au XV* siéele, susie dune seconde phase aux XVF-XVIF siécles. 1. INTRODUCTION In his seminal work, The Way of the Masks, Claude Lévi-Strauss (19821 introduced the notion of sociéiés & maison («house societies»] as afundamental kindof social grouping, quite istinc from the «lineage» structures found in many \raditional societies. In his words, the whousen in such societies consists of personne morale détentrice d'un domeine composé 3 la fois de biens matériels et immatériels, qui se perpétue par la transmission de son nom, de sa fortune et de ses titres en ligne réelle ou fietve, tenue pour léitime & lo seule condition que cette continuité puisco <‘exprimer dans le langage de la parenté ou de Vallance, e, le plus souvent, des deux ensemble (1979: 47}. Subsequently, mary ethnographere of Austronesian-speaking cultures distributed throughout Oceania and istand Southeast Asia have recognized the relevance ofthis concept for understancing traditional social organization in this part ofthe wortd log, Carsten and Hugh-Jones, eds., 1995: Fox, ed, 1993: Macdonald, ed. 1987: McKinnon 1961, 1995: Waterson 1990). Recently, the house society concept has been taken up by Polynesian archaeologists ard prehistorians (kirch 1996, 2000; Green 1998; Green and Pavley 1998, 1998: Anderson 2001). Because the concept of house society closely links social ‘organization to architectonic space (the physical structures occupied by a secial group, along with their estate and immediate landscape, it offers much potential for developing a «prehistoric ethnology» of Oceanic societies. In theirrezent book on the reconstruction of Ancestral Polynesian societies, Kirch ond Green (2001 209-2061 discuss some criterial features of house societies in Ocearia. They note that the concept ofa “house’ as both secial group and physical dwelling is an emic, lexically marked category.2 Secend, both the house and the social unit have endurance and persistence over time; there are social mechanisms for recruitment of members to the "house’ thus ensuring ‘Archaeological Research Facity, Unversity ofC 2 Redaction: Trrara Marie 3InPalyresian seceties this is frequently seme variant ofthe Prote-Potyresin tem "kage orn, BerkeleyCOMMUNE DE MOOREA - COMMUNE ASSOCIEE DE PAPETOA) 2 ‘continuity, Third, the "house" is a fixed-property holding unit, but in addition, the ‘house’ group holds significent movable praperty [such as canoes, or fishing nets). Morecver, various rights, privileges, and titles pertain to the house’, even the ownership of specific ancestral genealogies end founding myths. Fourth, the ‘house’ ise vehicle for the transmission of proper names, both of the physical dwelling and its estate, and of social groups. Fifth, ‘houses’ are ofien ranked relative to each other, in part due to a historical process of fission. In sorme Oceenic societies such differential ranking is weakly developed, hence heterarchical, but in others (such as that of the Society Isiands] quite pronounced and hierarchical 1.1, Houses and Households in the Society Istands ‘Across the broad Polynesian social canvas, the Society Islands have iong been noted for thelr cegree of socal complexity ‘and hierarchy. Sshline included them with Hawai, Tonge, and Samoa in his «Greup I» of «most structurally complex ranking systems», and explicitly notes that chietly households «were larger and more elaborately ornamented than those of others», and had «erylarge guest houses» attached (1958: 40]. Goleman, ikewise, counts the Society Islands among his «strattied» cases, and observes that the «status system was complex, allowing for a gradation of rank based on toth genealogical primegeniture and on achievement [1979 : 178] In proto-historie Ma'chi society during the two-to-three centuries immediately prior to European cantact, then, one may anticipate that ancestral Polynesian concepts ofthe ‘house’ ha¢ undergone significant local elaborations, in concert with developing social stratification and hierarchy, The received ethnohistory of Ma’chi saciety indeed supports this view, although it must be stressed that there is no consistent agreement on many aspects of traditional social structure, Ferdon, summarizing various early European accounts of housing, states that a «Tahitan’s house tended to reflect his economic status as well as his station inife», with chiefly houschelds being marked by a greater varaty of structures, and by superior construction (1981 -78) Oliver's ‘more detiitive and nuanced synthesis of Ma'ohethnohistory also quotes a variety of early servers on the substantial range of variation in dwellings and their grounds (1974 : 62-76). In particular, Olver cites Elis in reference to larger houses erected «only forthe leading chiefs» [1974 (71 sarenones | OTE Loss ‘SOURCE | Functional [Fare Generalized frm for ause ‘Types Fare t’eto Sleeping house Handy 1990; Ortiac 2000 Fare wtu Cooking house Handy 1930; Oriec 2000 Fare tamara Tapumen's eating house Fareboa ‘Shed for beating tarkcloth ‘Davies 1851; Ortiac 2000 Fare tupapa's House forthe dead Davies 1851; Handy 1990; Otc 2000 Fare taruma Out house to keep lumber in: also ahouse cccupied by att Davies 1851 Fare rau. fare pora Smal shed or shelter Handy 1990, Fare arii Chiefs residence Handy 1990 Fare oo Chis portable house Handy 1930 Fare 30a) ‘ela maison du ciel», special house inwhich a young ciel isinstalled ‘rtiac 2000, Farwo'a House for citdren of achiel Orla 2000 Fare ite Warrior's house Handy 920: Ortiae 2000 Handy 1900; Ortiae 2000, Handy 1990: Orlie 2000 Hove of ‘grand prdtre sorcier™ ‘rise 2000, ‘Theotre, house for dancing Handy 190 Farau ve'a.toreve's | Cance shed or hanger Handy 1990; Orie 2000 ‘Arehfectural | Fare hav pope, Rectangularhouse, Types: fare hau paru, fore hautia | house wih straight ends Davies 1851; Handy 1990: Onli 2000 Fare pote’, Fare potx'a | Round-ended nouse house with apsidal ends Handy 1930; Ortiac 2000, Fare auth ‘Shes covered with Cordyline Leaves Davies 1851 Fare ohy Smait movable round house, ‘Used for sheltering thngs Doves 1851 Fore nau House thatched with ceconut leat mats | Handy1930 Farearetu House thatched with grass Handy 1910 Fare taua House on llesor tits Handy 1930, Table 1 The Emic Classification of Ma’ani Houses. «This largely due to the apid changes following European contact, and the need therefore, to reconstruct the pre-historic society on the bisis ol fragmentary ard inconsistent European accoun', augmented by late «salvage» ethnography.COMMUNE D€ MOCOREA - COMMUNE ASSOCIEE CE PAPETOAL ‘The range of architectural variation in Ma‘ohi dwelling structures has been reviewed by a number of scholars, including Ferden [1901 and Oliver {1974I, but alzo Handy [19921 and Orliac (20001. is typical in many parts of Polynesia, the emic or folk-classification of houses seems to have been complex, with both functional and architectural categories, as summarized in Table 1 Despite the obvious importance of habitation structures in Society islands archaeology —and their potentiat to yield critical information on both the long-term development of Ma'ohi caciely and its structure in the pre-European period such sites have been largely neglected or ignored by investigators. The trend was set early on by Emory [1933 : 44 4g), whose exploratory monagraph on Seciety Islands «stone remains» concentrated almost exclusively on marae, with «house sites» mentioned only in passing. In their pioneering application of «settlement paitern archaeology in the ‘Opunohu Valley, Mo'orea, Roger Green and Janet Davidson (Green, et al, 19671 did excavate three round-ended house structures, demonstrating the potential of household archaeology. However, they did not attempt to excavate at the full range of habitation features revealed by Green's surface survey, and the promise oftheir initial work was not fol- lowed up by other investigators, who continued to emphasize either marae or the perpetual hunt for «early» sites Garanger 1967; Emory and Sinoto 1965}. For example, in all ofthe extensive field research carried out under the auspices of the Département Archéolegie of the former C.P.SH,, we are aware of only a single report on the excavation ofa set of ‘occupation terraces, in the Vaihiria Valley (Cristino etal, 1988), Clearly, the archaectogy of houses and habitation sites, is avast lacuna in Sociaty Islands research, crying out for study. 2. ARCHAEOLOGY AND PREHISTORY IN THE ‘OPUNOHU VALLEY, MO'OREA ‘The largest valley on the second-largest stand in the Sociely archipelago, the ‘Opunchu holds a prominent place inthe ‘annals of Ma‘ohi tradition, ethnohistory, and archaeology. Oral tracitions relate that a century or so prior to European contact the ‘Opunohu was conquered by the holders of the powerful Marama title, and at the time of Cook's voyages vas the seat of @ prominent chief of Marama line named Mahine (Green 1967 : 220-221; Oliver 1974 : 1204-1208]. Cook anchered in the ‘Opunohu Bay on his third and final voyage, providing an opportunity for the artist John Webber to render views of the valley both from the sea and from inland (Joppien and Smith 1988 : 986-92) Slightly later, the valley wes a refuge for Porare Ii during the period of post-contact wars, and alse the setting for one ofthe frst dramatic conversions to Christianity lat the coastal village of Papetoai; Oliver 1974 : 1340). ‘The first archaeclogical descriptions from the ‘Opunohu are by Emory (1993 : 105-1071. [..]In 1960, Roger C. Green carried out the first systematic setitement-pattern survey in Polynesia in the “Opunohu, implementing the then innovative approach of Harvard professor Gordan Willey (Green 1961]. As nated above, Green and his collezques also excavated at three round-ended houses, aswell as at several mnaraein the valley (Green et al. 1967). Although Sinoto carried out tourism-related restoration of several marae in the early 1970s, no further research was conducted until Dana Lepotsky of the University of California, Berkeley, began renewed werk on archaeological and paleoecologicel evidence for prehistoric agricultural practices and land use {Lepotsky 1994: Lepoisky et al 2000). At the same time, Oakes [1994] excavated the first example of a rectangular house structure in the valley. Although they did not carry out new fieldwork, Green and Descantes (1989; Descantes 1990, 1993] synthesized the data collected during Green's criginal 1960 survey, providing an important baseline for future field research, Drawing upon all of this collected research, Green (1996) offered a retrospective synthesis of ‘Opunohu prehistory, including a four-phase culturet sequence. 3. RESEARCH DESIGN AND METHODOLOGY The overall goal of the ‘Opurohu valley archaeological project was to examine inter- and intra-household variability, at a range of late prehistoric household sites. Specifically, the work was structured to understand local lor household Level) economic and socio-political organization within the ‘Opunchu valley and their relationship to regional complexity within the overall Society sland chiefdom. Ethnohistoric documents suggest that most economic production within the Society Island archipelago was organized at the household level [Oliver 1974]. However, there is virtually no archaoolo~ gical data available for reconstructing the structure of household activities and the relationship of domestic preduction to the broader political economy of the Society Island chietdoms. ‘To examine variation within and among tate prehistoric Mach/houses, extensive areal excavations were carried out in the ‘Opunohu valley betwen 2000-2001 at three site complexes, each associated with a prehistoric house site. The project utilized household archaeology techniques, to identify and examine differences between commoner eng elite residential architecture, land use, material culture, and domestic production and consumption. Three rectangular houses [fare haupapel and iwo round-ended houses (fare pote’el were excavated in three neighboring architectural ‘complexes within different geoarephical contexts in the Tupauruury section of the ‘Opunohu valley /Figure 1, The‘COMMUNE DE MO"OREA - COMMUNE ASSOCIEE DE PAPETOA) 2 objective was to furnish detailed archaeological data from households deriving from 2 continuum of social classes within @ community. Following this, the excavated residences ranged in contex, size, and architectural elaboration, Ethnohistoric texts document that fare haupape (rectangular houses] and fare pote’ [round ended houses} were the two principal domestic structures that were used as sleeping houses (Handy 1930 : 9: the former were commoner sleeping houses, the latter, elite sleeping houses, The minimal archaeological investigation of prehistoric Maohi houses suggests that these structures were used for a wider variety of purposes: thus, Green [1996] posits that fare pote’e functioned as either elite sleeping houses, meeting houses, or god houses associated with merae complexes, while fare°haupape may have functioned as commoner sleeping houses. or specialized craft activity areas, and were ‘not necessarily always of low social status [see Cakes 1994]. Further extensive areal excavations of both rectangular and round-ended house sites were required to determine their range of material variation and function. This is the main research problem that our project set cut to address. Our excavations at Opunohu valley residential complexes provide empirical means for delineating whether the disparities in material culture and domestic activities in the house sites correlates to sociceconomic distinctions between household groups or diachronic shifts in residential ‘activities. Ongoing laboratory analyses will provide further data for understanding the articulation of household wealth and status with household preduction, consumption, and cratt specialization. 3.1. Specific Research Questions ‘Our excavations at ‘Opunohu complexes ScMo-170/171, 120, and 123, atlow for detailed comparative analysis of the material remains associated with the fare haupape and fare pote’e house types. Our specific research goals were as feltews,(1]to create accurate maps of each residential complex, including all surface architecture end excavated ‘areas, (2) to complete large scale horizontal excavations of several fare haupape and fare pote’e house sites, and test pit excavations of associated structures, to document site function [3] to assess the variability in household architec ture, associated features, and material culture, and their relationship to household production and consumption, and household status and [4], to accurately date the household complexes in order to build a precise local chronology for ‘Opunohu valley settlement, 3.2, Site Selection and Excavation Methodology [1 We completed extensive horizontal excavations at five house sites, exposing large areal blocks in house interiors, as well 2s on the exterior habitation terraces or tving flats, Excavation sample sizes range from 38 m* to 56m. ‘The excavated house sites were located during a four week re-survey of the upper Tupauruuru sector-ef the valley carried out by Kahn in the summer of 1999 [Kahn and Kirch 1999]. Kahn relied on site descriptions and maps fram la ( os ln ) het Lae trent Pe ae eal lex a = ¥ Figure 1 - Map of Mo'orea based on the BRGM Geological Survey 125000 [R.¢. Maury and A. Dez 2000)‘COMMUNE DE MOOREA - COMMUNE ASSOCIEE DE PAPETOA Green's earlier ‘Opunohu Valley survey (Green et al 1967, Green and Descantes 1989] to Locate over 27 site complexes with potential for excavation. Compioxes ScMo-170/171, 120, and 123 were chesen irom this set, based primarily on their welt preserved architectural remains, in addition, the spatial arrangements of the house structures and associated architectural features, such as their proximity to one another and similarity in architectural style, suggested that these residential complexes each formed the basic domestic unit of a household |utuafarel, end thai these neighboring complexes in aggregate formed an interrelated neighborhood ar cormmurity (mata’eina’al (Oliver 1988 : 42-44: Davies 1851 :136, 305). Because the three complexes vary with respect to geographical location, architectural elaboration, and ‘umber of associated architectural features, they provide an excellent data set for contratled comparison. [.) 4, RESULTS OF THE 2000-2001 HOUSEHOLD EXCAVATIONS IN THE ‘OPUNOHU VALLEY We present our preliminary results from the excavation of five house structures in thrae neighboring house complaxes [ScMo- ‘70/171, SeMo-123, ScMo-120) in the Upper Tupauruuru region of the ‘Opunohu Valley (Figure 1, 4.1, House Complex SeMo-170/171 We interpret the site 170/171 residential complex as a neighborhood, or utuafere, with four house sites in close proximity, straddling a ridge (Figure 2). A fare pote'e ‘or round ended-house site |structure 170] is located atthe highest peint on this ridge, followed by three fare haupape or rectan- gular house sites (171A, 178, 171C], and by ‘several stone faced terraces (1710, 171E1. A simple enclosure with coveral dyke stone uprights (771Fis situated at the base of the complex, at the lower promontory of the ridge slope. We interpret this structure as a simple marae or shrine lef. Green and Descantes 1989 :126). This small structure is unlikely to have been the marae for the eccupant of structure 170, which is likely tohave been one of the larger marae in the immediate vicinity. 4.1.1, Exeavations at the SeMo-170 Fare Pote’e ‘A-curbstone outline of unworked basalt slabs demarcates this moderately large round-ended house. The fare pote’e curbstone outline is disturbed, and may have been robbed of stones for the construction of 2 later-dating rectangular house. The fare pate’e occupies an eleva- ted, stone-faced terrace, which ies upstope and to the south of a large earthen terrace. ‘A pavement occupies the space between the terrace facing and the fare pote’e curbstone outline. A backrest stone and small basalt upright are found on this paved area exterior to the house. Also be Phe. sm ‘ AS, te ‘Opunohu Valley ‘ < ScMo-170-171 * Beaton / ‘Legend = BoeattUpright Sab El Exeavated Area © To0e ‘eaen Stone Alignment = Downslope me E Uprighis & % ? > Figure 2 Pian view of ScMo - 170/171 with excavated areas,[COMMUNE DE MO"DREA.-CONMUNE ASSOCIEE DE PAFETOAL 26 exterior to the house along the northernmost extent cf the elevated terrace, are two single-course alignments of waterwern boulders, Overall, 58 m# were excavated at SeMo-170, 43 m* within the house interior, 10 m# outside of the house on the ele- vated platform, and 5 m* en the earthen terrace located in front of and downslope from the terrace. A single cultural deposit [Layer B] was encountered in the excavations, overtying construction fill (Layer 6 fill) which was leveled to construct the house terrace. Radiocarbon sample Beta-153427 was taken from an in-situ combustion feature at the Interface between the bottom of the construction fill for the house terrace (bottom of Layer B fil) and the top of the underiying sterile deposit [Layer Cl [see Table 2] We interpret this sample as dating to the intial clearing of the site for construction of the house and elevated house terrace, probably in the '5"* century AD. TaN. |Site Provenance ‘Genventonal | S°> | — calbrated Ape “eage reas | “oo atte oP. Ben [1 [Un Nive EVor, sab-aes | WOT A “aT a0. 027 Cembe n-itu charcoal lens ‘AD. 1600-1620 BIG interface; init se elearng. Gen] TB [Unt nv7 e101, Feawwre 8 | Boe “Ea [Ro eos ‘ssa ‘earth oven, heuse exterior. ‘RO. 1630-1680 ‘80. 170-1800 (A0.1940.1980 Ba THC | Unt nore Featerey; aoa THAAD. 100-1820 ssa ‘combustion testure (scoop ‘AD. 1570-1620 earn], nous it Bea [TAC [ Unit NOD E¥8, Feature S| GOED TBI [ RO asa 1340. hearth, nouse exter. TANS” [THA] charcoat trem earth oven | 2020 Be [RO TE an [excavated by Oakes 1994) ‘AD, 1610-1670 ‘AD. 1780-1800 TANS: [THA | Charcoal Wom under fause | 210867 EE [AD. 1520-1550 7087 ‘eur stones Lexcavated by ‘80. 1690-1690 Oakes 1994) 01920-1980 Bee [PH] Unt NP Et Feature 16 2D “ea [AD. 1590-1550 vsti? ‘combustion feature, heuse ‘20. 1630-1680 Inveror. A0.17041810 | ‘A0.1920.1950 Bets. 2A | Uni WTOSETTG Feature 10: TBE PAD. 1510-1600 vests insstu charcoal lens at 8 ‘80. 1610-1650 AUC interface, init site cleanng DP Table 2 - Radiocarbon Ace Determinations by House Site, In general, the house interior was clean and devoid of cooking features, pt features, or charcoal concentrations. A moderate number of basal lakos [n= 594, 1.8 lakas/m? density and a complete adz, with 2 reverse triangular cross section and morphology conforming tothe Dutt 44 type, were recovered in the fare pate'e interior (Duff 1956, 1959) in addition to several posthole features (Tables 3 and 4] An earth oven and several gits (possibly breadfrut fermentation pits] were exposed in he exterior living area below and infront of the house platform (Table 3. The high incidence of Feature Type 120 1B 170 1718. 1c Inte Ext | int. Ext | tnt, Ext | Inte Ext | nts Ext (a es Ts Se 3 2 6 a Postmold_ | PitiPost mold Pit Pit/Combustion Feature ‘Stake hole a ‘Arvorphous burring depesks z zo ‘Stone-outlined hearth t Esrthoven 7 7 int interiar of Pose Ext, eter of house, Table 3 Counes of Sub-Surtace Features Exposed at individual House Sites S ae, § Table 4, column 170-WS reports artifact counts forthe & m? excavated aleng the northertmost extent ef the house terrace, which we have Interpreted neon in-sits ads preduction workshop. Celumn 179 does rot inclde the 170-WS ceunts but reports atic! ceuntsfor the esto} the unis excavated at 170, ath wihin the house and n the exterior areas.‘COMMUNE CE MOOREA - COMMUNE ASSOCIEE DE PAPETOA! Fa ‘exterior house features and the discovery ofa pt feature with two separate compartments demarcated by 8 stone lining ‘suggest formal segmentation of the domestic space, indicative of an elite household. The frequency and type of sub- surface features indicate that domestic tasks such as cooking and food storage were carried out at this rosidence. Basalt flakes and detitage were recovered irom the 179 house interior and exterior areas at a moderate density (16.5, {akes/ m'), indicative of activities rotating to adz use and maintenance, A unique discovery was the high frequency of basalt flakes and flaking debitage in association with two single course surface alignments, about § m north of the 170 fare pate's on the elevated house plattarm. The basalt artifact assem- blage recovered from the & m* excavated here [designated as site 170-WS [workshop] column, Table 3} differed from the other excavated assemblages because they included adzes and adz fragments from all stages of the production ‘sequence. The 170-WS area had the highest density of basalt artifacts per unit excavated, 136.9 lakes/m') in contrast to the moderately low density found at the other areas excavated within the fare pote’einterior and exterior {16.6/m?l. Both large and smell lakes were recovered, in contrast tothe largely small flakes foundin the other house excavations. This area also had the highest density of adz flakes, at 23%. We tentatively interpret 170-WS a8 an in-situ adz production workshop. The high adi flake density alse indicates that adz rejuvenation or maintenance activities took place here. Our interpretation must be further supported with data froma detailed reduction stage analysis ofall recovered tools and debitage, which is in progress. However, we suggest that this workshop, which would have been visible to the rest of the neighborhood situated downslope, reflects prefe- rential access to basalt sources and/or craft production activities, and thus, symbolically represents the material wealth and status of this house” group, Artifact Type 1208 123A 170 rows ine ane Ae 7 7 7 ° ° 2 Reworked adz , o 1 ° ° ° ‘Agz fragment 4 a 2 2 ° 1 Act lake 3h 9 2 18 1 5 ‘gz blank ° 0 ° 3 ° 0 ‘Adz preform ° a ° 3 0 2 Worked prismatic bas 5 2 1 4 ' 2 Bacal fake 666 982 594 647 230) a6 Shatter 1341 850 0 a 120 wo Retouched take 1 1 ° 1 0 9 Utlized flake 0 a 4 1 2 2 Hammerstone 1 2 0 0 1 Whetstone 1 0 ° ° o ° Core ° 3 2 2 a 4 TOTALS 72056 1250 823 1095 356 593 Table & Basalt Arifact Countaby Ske, ‘ScMo-170 is unique in many ways st is the only fare pate’e within the rosidontial complex; itis located at the most elevated geographic locale; ang, itis the largest and most architecturally elaborate house in the complex, with a woll-constructod exterior pavement, elevated house platform, and a backrest stone and upright on the platform. Given these features, we hypothesize that ScMo-170 functioned as an elite residential area for the neighborhood, possibly the «origin hause» or residence of the seniar member of this particular ascent group. The arrangements of artifacts end sub-surface features, and the activities that hey reoresent, support this hypothesis. In particular, evidence for the formal segmentation of domestic space and access to basalt and/or craft activities which are both highly indicative of an elite residential area, were found at ScMo-170 but were lacking atthe other excavated structures, 4.1.2. Excavations atthe SeMe-171B Fare Haupape This small rectangular house is located 48 m dewnsiope of 170, and 24 m downslope af 171A, a large fare haupape excavated by Oakes (1994) (Figure 2). The western, eastern, and coutharn limits of 718 are outlined by simple alignments of unworked curbstones. The northern fare haypape wall forms the facing for the house terrace. Wie excavated 28 m? at site 1718, ineluding 28 rr? in the house interior and 10 m? in exterior areae adjacent to the house. The excavations revealed deposits relating to a single period of cultural occupation, overlying sediments that were filled in to buildup the house terrace and modify the natural ridge slope toa lat. ving surface. Asmall number of basalt flakes were recovered in the house interior, in addition to several postholes and a small fre pit. The mor~ shalogy of this shallow fire pit corresponds ta historic descriptions of combustion features which functioned to light, heat, or protect houses fram insects [Orliac and Orliae 1980: dj, end also conforms to informal hearths recovered(COMMUNE DE MOOREA COMMUNE ASSOCIEE DE PAPETOA 2 earlier ‘Opunobu Valley excavations that were interpreted as being used to heat the house interior [Green and Green 1964 : 168]. Most sub-surface features and basalt artifacts were recovered in the exterior areas adjacent to the rec- tangular house. Notably, 2 large earth even (umul was recovered 1 m outside of the southern curbstone house wall This large oval pit [85 x 65 cml of moderately shallow depth (25 cm) had a dense concentration of charcoal, ash, and fire cracked vesicular bazalt rocks, overlaying a thin lons of burnt, oxidized scil. Radiocarbon sample Beta-153428 was taken from a dense concentration of charcoal found just above the lens of oxidized soil at the bottom of this feature Table 2). Several othor features were recavered to the south of the house exterior. These include a small area paved with waterworn basalt cobbles, adjacent to a locale with a posthole and a series of small informal fire pits. Iti Ukely that this wasa cooking area, and it may have been associated with an informal cooking shed. Few basalt artilacts were recovered in the house interior, and those recovered in the exterior areas were concen- trated along the southern living ‘lat adjacent to the house. Excavations at site 171B produced the lowest density of basalt tools and debitage from all the house sites excavated, at 9. lakes/m? (Table 4). The majority of the basalt flakes ‘and shatter were recovered from the southern side of the fare haupape, just exterior to the house curbstone outline ‘and in asscciation with the small exterior pavement ‘The low density of artifacts relating to adz production and maintenance actWvities at this site, atong with the small size of the fare haupape, and the lack of interior artifacts and features other than a simple fre pi, lead us to interpret this structure as a sleeping house (/are fa ‘oto for persons of somewhat lower status than the occupants of ScMo-170. It appears ac if the areas exterior to the hause were used for a variety of domestic activities, including cooking. food storage, and a¢z maintenance. A small sub-surface pavement of water-worn basalt stones was associated with this exterior activity area along the southern extent of the exterior living flat. 4.1.3. Excavations at the ScMo-171C Fare Haupape “This smell rectengular house was delineated by single alignment of unworked basalt curbetone slabs. Fare haupape 171C was the smallest house in the complex, measuring 4.§ x5.5 m. Two stone-faced terraces are found downstope, to the west. Intotal, 46 m! were excavated at site 171C, including 27.5 m¥ along the house interior and 18.5 m* in the exterior areas adjacenttte the house. Cultural deposits relating to two occupations were encountered. & thin, diffuse historic occupa- tion was recovered in the top 5 em of the site, Several small stake holes and a few pieces of window glass were found in these deposits. The main period of accupation [Layer Bl associated with the initial construction of the rectangular house and the majority of artifacts and sub-suriace features recovered, was found beneath the diffuse historic deposit and reprecents the prehistoric occupation ofthe site. The features and artifacts discussed below were retrieved from the prehistoric cultural deposit (Layer B), ‘Asmall fire pit and several postholes were exposed in the house interior (Table 3. This combustion feature resembled that found in the interior of 171B end appears te have been used for providing light, haat, or keeping insects out of the house interior. Radiocarbon sample Beta-152429 was taken from the interior of this combustion feature /Table 2/ anc dates to mid-15" 1016" centuries, > Plater ‘5cHO-171 Calter excavation‘COMMUNE DE MOOREA - COMMUNE ASSOC! DE PAPETONL a ‘An adz preform and a concentration of basalt flakes were recovered adjacent to the posthelein the mid-line ofthe house. A wider range of features were recovered just cutside the house, including Large pits, 3 simple cooking hearth with a stone outline, and a series of postholes. Thus itis likely that there were a series of informal shelters or struc tures on this adjacent terrace to the west. Radiocarbon sample Beta-15420, which dates tothe 15" century, was taken fromm an exterior hearth, which was eutlined by 2 circle of rocks and had a series of small vesicular basalt rocks at its basal boundary, just below a thin tens of charcoal and ash. This hearth feature resembles cooking features used for arilling or roasting taro, bananas, or other foods ever fie [Ortiac and Orliac 1980 : 64), ‘A moderately tow density (12.9 flakes/m* of basalt artifacts and flakes were recovered from the house interior and exterior (Table 4). A wider range of sub-surface features were recovered in areas exterior tothe house. These included 2 series of postholes, and a pit that may have functioned for breadiruit fermentation. Given the small size of this rectangular house. its non-elaborate architecture, and its lower elevation in relation to structures 170, 1718, and 171B, we hypothesized that 171C wes occupied by a heusehold group of lower status. The type and quantity of sub-surface features and artifacts recovered indicates that a range of domestic activities were carried ut at the site, including food preparation, storage, cooking, and minimal adz production/maintenance. The majerity of these activities, other than adz production/maintenance, took place solely in areas exterior to the house structure. We interpret site 171C as having functioned as a steeping house (fare ta'atol, possibly with an informal fare tutu or cookshed attached. [..) 4.1.6, Overview of the ScMo-170/171 Complex ‘Our excavations confirm that the ScMo-170/171 complex represents a Ma'ohi neighborhood or utuafare, of the late Atiro'o to Mareme Phases. The consistency among the four radiocarbon dates indicate that the structuresin the complex ‘were contemporanzously occupied, between the 15-17 centuries AD. We interpret site ScMo-170.as the «origin house» or residence of the head of o local ascent group, or ‘house’. Multiple lines of evidence architectural, geographic context, artifact frequency, and the type and form of sub-surface features) indicate that the occupants of this fare pote’e wore not only of high status, but of the highest status within the complex. The use of space within and around the house also cemonstrate a certain conformity to prescribed social rules. The location of an in-situ adz production ‘workshop, in close associated with the fare pote’e on the house terrace, represents preferential access to important raw materials and goods jhigh quaiity basalt and specialist labor] and may also represent craft production financed by the house inhabitants, We surmise that the 170 fare pote’e was the residence of a lawer-ranked chief Ito’ofa or ra‘atral, ‘who served asthe head ofthe larger ‘house’ ascent group which occupied the 170/171 complex (Oliver 1974 : 749-777: cl. Green 1996; Orliac 1982 : 285-288). ur excavations confirm that the fare haupape 1718 and 171C served primarily as sleeping houses, with 8 variety of domestic activities performed in and arauhd them, including food preparation, cooking, and storage. These house occupants had restricted access te certain goods and raw materials [e.g,, ad? manufacture) and had smaller houses with non-elaborated architecture. We believe that the data support an interpretation that the occupants of house structures 1718 and 171C wore of relatively lower status than those of 170 Iti Ukely that the occupants of 1718 and 1710 ‘were either junior members of the ‘house’ group under the direction of tne headman who lived at the 170 fare pote’e, or were attendants for the occupants of that structure, Further ethnohistorical research may help to differentiate between these hypotheses. 4.2. House Complex ScMo-120 Site complex ScMo-120 is located on a small, isolated ridge opposite complex 170/171. It exhibits elaborate archi- tectural features, including an oval-ended house site {120B] located on an elevated terrace. A pertially peved lower terrace [120A] and a faced terrace (120A LT] are found below the fare pote’e structure to the northwest. A complex ahu-bearing marae (120E) and e raised stone platform or paepae 1200) are found above the round-ended houce, on elevated faced terraces (Figure 3. Unlike ScMo-170, the round-ended house complex at ScMo-120 was adjacent to, but at 2 lower elevation than the marae structure, ScMo-1208 offers an excellent contrast with complex 170/171, because the ScMo-1208 fare pote’eis of medium size and is slightly smaller than ScMo-170, but is Located on amore elaborately constructed terrace. Further, thet70 fare pote’e was associated with 2 much simpler marae than that found above ScMo-1208. 4.2.1. Excavations at House Site SMo-120B We excavated 54.5 mt of this site, including 41.5 min a large horizontal block excavation of the house interior, and 13 rv along the exterior portion of the house terrace. In comparison with the other excavated structures, the culturalOMNUNEDE MO‘OREA - COMMUNE ASSOCKEEE PAPETOAI 0 deposit from the interior of the fare pote'e was remarkably devoid of cultural materials, and lacked concentrated ‘amounts of charcoal. Interior sub-surface features included postholes, a pt feature, and an in-situ fire feature, located ‘along the southernmost portion o! the house (Table 3]. The latter is interpreted as a single use fire pit, which likely ‘served to provide warmth er light for the structures’ interior. Charcoal fragments from this fire pit were submitted as radiocarbon date Beta-165117 (Table 2/ and indicate that the 1208 structure was most likely ozcupied between the 16-17" centuries AD. Site 120B yielded o significant number of basalt artifacts from the house interier, including high frequencies of adz flakes [0.6 flakes/m*} and moderate amounts of reworked adzes and broken adz fragments (0.09 flakes/m'l iTable 4). In the course of our excavations, two large sub-curface boulders were encountered within the house interior, sur- rounded by concentrations of a¢z flakes and simple basalt flekes on the adjacent house floor. We interpret these flat topped boulders and their associated lithic scatters as activity areas, where individuals sat and used adzes, and/or reworked the adz bevels. These data, as well as the recovery of a whetstone and an adz fragment that was reused as a polishing stone (Table 4), suggest that this was a locate where adzes were being used rather than produced. Our ‘excavations also Indicate some spatiat separation of activities within the structure, with the majority of edz-related ‘work in the middle portion ofthe house near the sub-surface boulders and a storage/wel-lit area in the southernmost portion of the structure. Formal cooking features such as earth ovens or Fre pits with stone borders were not encountered in the excavations of the house exterior. In exterior arees just to the west of the house, several postholes end pit features were recovered. The evidence suggests that storage shelters or small superstructures were erected in this area of the tarrace. The house terrace to the north of the fare pote’e lacked any features, although several units were excavated in this sector. Cultural deposits just exterior to the house had a noticeably higher concen- tration of charceal than the house interior, suggesting either that the interior of the house was periogi- cally swept clean, or that certain activities were restricted to the house exterior, in order tomaintain cleanliness within the structure. Several lines of evidence support a specialized function for this structure. First, cooking features were not encountered, in contrast ria to the excavations at ScMo-170, = ten 171B, and 171C. Second, the basalt ee a assemblages included high den- sat sities of adz flakes and moderate amounts of adz fragments and reworked adzes, concentrated around two sub-surface boulders In the house interior. Third, the nature of the basalt assemblage, Figure 3- Plan view of SeMo - 120 complex with excavated areas, including the smatl size of the ‘Opunohu Valley ‘ScMo-120 ComplexCONMUNE DE MOOREA - COMMUNE ASSOCIEE DEPAPETON H >Plate2 ‘SchO-129 B alter excavation recovered flakes, the overall low ratio of flakes to shatter, the absence of adz blanks and/or preforms, and the high frequency of polished adz flakes, suagests that 1208 was a locale of ad2 use, rather than production (Table 4). Finally, there are several indications that the house interior was intontionally kept clean, perhaps due ta its ritualized nature ‘These excavation data point towards an interpretation of fare pote’e 1208 asa locale for specialized activities, which is, supported by its close association with an isolated, complex marae site. We believe that the bestinterpretation of the archaeological evidence is that fare pote'e 1208 was a lacale where wood was sculpted, Ethnohistoric sources [Corney 1915; Henry 1928] describe a myriad of ritually important sculpted wood objects [including unu or posts for offering platforms, house posts, and ged images] found on or near marae sites. The production of such items would have been ritually charged, and is likely to have been carried out at a specialized local such as fare pote‘e 1208. itis possible that site 1208 represents 9 fare-ia manah, described by Henry as «Farther out, in front of the marae, was a good-sized house, called the fare ia-manaha (house of sacred treasuresl, where the keepers of the maraedwelt, and in which were kept the images of the gods, the sacred apparel, the drums, and other most sacred appendages. Here alse the sacred objects for the marae ceremonies were made Henry 1928: 138)» 4.3. House Complex ScMo-123 ‘This site complex is located northeast of ScMo-120, on the adjacent ridge top. The complex is comprised of five clustered stone structures. These include a rectangular house {123A of moderate size, delineated by unworked basalt curbstones and situated on a raised, partially paved terrace with stone facing (Figure di. A possible ramp or entryway is located at the northeast extent ofthis paved terrace. Two large ahu bearing marae 1238, 123C! are found upstope of the rectangular houses, A substantial earthen terrace separates the fare haupape from the two marae. The southern Limit ofthe site is demarcated in areas by 2 single course alignment of large boulders; the ridge slopes moderately steeply to the SW of this alignment. Given the often-accepted notion that fare haupape served as commoner sleeping houses, which has recently been challenged [Orliac 2000; Oakes 1994), site 123A was chosen for excavation because its context suggested that it may have had a specialized functicn. 4.3.1. Exeavations at House Site ScMo-123A A total of 47.5 m were excavated at ScMo-123A, Including 135 m* within the house interior and 34 m? on the exterior terrace. A radiocarbon age determination was made on a wood charcoal sample from an in-situ burn located at the interface between Layer 8 fill [construction fill underlying the Layer B cultural deposit} and the underlying Layer C (basal deposit). We interpret this in-situ burn as relating to the initia site clearance and terrace construction. The radiocarbon age determination places the initial construction ofthe fare haupape and house terrace within the 16-17" centuries, during the later Atiro’o or Marama Phases (Table 2) * Only marae 1298 is presented in our map ofthe complex See Green and Descantes 1989 209 fora map tha includes marae 124(COMMUNE DE Ma"OREA - COMMUNE ASSOCIEE DE PAPETOA| 2 ‘Opunohu Valley ScMo-123 Complex by Pay Figure 4- Plan view of the ScMo~ 123 complex with excavated zreas. ‘Several postholes features were located in the house interior, which lacked ary evidence for sub-surface combustion features, The fare haupape interior contained maderate amounts of basalt flakes and shatter (8.0 Ilakes/m? and 17.9 flakes/er? respectively) (see Table 4). Notable artifacts recovered include an unfiniched bifacially worked basalt flake that resembles a coconut grater and three adz fragments. Most of the basalt tcols and debitage were found along the ‘western edge af the house, near the paved area, suggesting that this was a locale for stone tool use or reworking, ‘Sub-surlace features recovered on the terrace exterior to the house include a simple in-situ fire feature, several postholes, and numerous pits of varying farms (Table 2). The high density of exterior postholes suggests that there ‘were several small informal structures present on the 123A terrace adjacent to the fare haupape, to the west, north, south, and east. A series of small in-situ burnt sediment lenses and dense charccal deposits were recovered at the south-eastern sector of the terrace and are indicative of a series of small informal structures or posts between the house structure and the marae, some of which may have been burnt. A series of pits and pestholes concentrated in the mid-eastern sector of the house terrace suggest an area used for storage activities. ‘The lack of well-defined earth ovens and the paucity of informal cooking features in areas exterior to the ScMo- 123A structure suggest that this fore haupepe was not used as a slesping house, or ac a structure for day-to-day domestic activities. The high number of sub-surface features exterior to the nouse vary trom patterns found at the other excavated areas and indicate that a number of superstructures and/or posts were erected between the house ‘and the up-siope marae. This lends further support to our interpretation that ScMo-123A had a specialized function, Further laboratory analysis and archival research are necessary before a more detailed explanation of site function can be offered. 5.0 RADIOCARBON DATING Inorderto.assess the age and time span of occupation of the excavated houses, we submitted six samples of charcoal for AMS radiocarbon dating, te the Beta Analytic Laboratory. James Coil of the U. C. Berkeley Oceanic Laboratory identified the charcoal in each sample in crder to select wood from short-lived taxa, thus minimizing the «ald woodCCONMUNE BE MOOREA - COMMUNE ASSOZICE DE PADETOA 3 ‘Atmospheric data from Stuiver etal. 1998); OxCal ve.8 Bronk Ramsey (2002); cub #4 8d: 12 preb usp [chron] 153427 | 420 a408P 153429 | 400 2408P 159430 | 460 »4oBP 152428 | 250 =408P 6260 280 2408P 7087 290 260BP ses | 240 408P ress | 320 .408P Calibrated date "200C81AD 1490CatAD 6coCaIAD sBo0caiAD 2o00CatAO > Figure 5- Oxcal multiplot lor the dated heuto sited jal. problem. in all cases, fragments of Hibiscus tiliaceus (aurau) were selected as the mast appropriate dating mat This is a soft-wood, anda relatively short-lived weedy shrub that does not have an inbuilt age that could bias the radio- carbon determinations. We are therefore confident that the «C ages obtained should accurately reflect the true age of site occupation and use. Laboratory methods reported by the Beta Analytic Lab ware stanard. The charcoal samples were pretreated and examined for rootlets, followed by a hot acid wash to remove carbonates, rinsing to neutratity, another acid wash, followed by another rinsing to neutrality. The Oxcal program was used to calibrate conventional radiocarbon ages [atmospheric data from Stuiver et al. [1998], calibraied using Oxcal, version 3.8, Bronk Remsey [zooal), Results are presented in Table 2. which also includes the two samples dated by Oakes [1994) from her excavation of fare haupape site 171A. Figure'5is an OXCAL plot of ll radiocerbon sample ages, shown es probability distributions. 4s frequently occurs with radiocarbon samples dating to the last half-millennium, there are multiple «intercepts» ‘when one plots the radiocarbon ages against 2 calibration curve, resulting in the multiple calibrated ages shown in the right-hand column of Table 2. In some cases, these include calibration ranges falling into the post-contact period (as for cites 171A, 171B, and 120). However, on independent archaeological evidence, such as the absence of historic-period ‘artifacts at 1718 and 120, we can reject these late phase caliorations. In general, all of our radiocarbon datos fall within the late Atira‘o to Marama phases ef the ‘Opunohu sequence as {efined by Green (1996), supporting an interpretation of maximal population density in the centuries immediately prior to European contact. However, our dates also suggest the possibility of two phases of occupation at the large site ScMo-170/171 complex, with an initia phase dating to the 15" century at siructures 170 and 171C}, followed by @ second phase in the 16-17" centuries lwith the construction of structures 171A and 171B), Sites 129A and 120B also date to this later phase. 6. CONCLUSIONS ‘As noted above, the raciocarbon data cleariy indicate substantia settlement in the interior of the “Cpunohu Valley dating to the 15-17" centuries, corresponding tothe late Aliro‘e to Marama Phases ofthe valley's cultural sequence {sensu Green 1996: Lepofsky 1994; Contra Descantes 1990, 1993; Green et al, 1967) Its worth pointing out that nane of these sites inthe upper Tupauruuru sector of the valley had an substantial re-occupation during the post-European ‘contact Marama to Perrare Phases contra Descantes 1990, 993). Site 171C had only minirral evidence for a brief period of re-occupationin the post-European contact period, represented by low numbers of historic artifacts la few pieces of Lote or window glass and metal and the recovery ofa small stakehole anda single posthole inthe top Sern ofthe site ‘deposits, above the main occupation Level. We interpret this as representing brief re-utilization of the site sometime softer 1767.COMMUNE DE MO'OREA - COMMUNE ASSOCIEE D= PAFETOA! 3% Kirch,P.V.and RC, Green, 2001 ~ Hawaii, Ancestral Poly- nesia : An Essay in Historical Anthropology. Cambridge Cambridge University Press. Lévistrauss, €. 1979 - Nobles Sawages. In Culture, science, et dévelopgement : Contribution & une histoire de homme : mélangesen I’honneur de Charies Morazé, pp. 41- 55, Toulouse : Privat, Lepotsky, D., 1994 ~ Prehistoric Agricultural Intensification in Society atends, French Polynesia. Unpublished PhO Dis sertation, University of California, Berkeley. Lepofsky, D. K. Lertzman, and P.V. Kireh,, 2000 ~ Stratigr3~ phic and Paleobotanical Evidence for Prehistoric Human Induced Environmental Disturbance on Mo‘erea, French Polynesia. Pacilic Science 503}: 253-273. Macdonald, C., ed. 1987 - De ia hutte av palais: Societés ‘maison’ en Asie du Sud-Est insulaire. Paris : CNRS. Maury, RC. and Le Dez, A.,Carte gécl. dela Polynésie frar= aise [1/25 000), feille Mo'erea. Ortéans BRCM. McKinnon, 5.1991 ~ From a Shattered Sun : Hierarchy, Gender, and Alliance in the Tanimbar Islands. Madison : Un= versity of Wisconsin Press, ‘McKinnon, S., 1995 - Houses and hierarchy :The view froma South Moluccan society, In J. Carsten and S. Hugh-Jones, eds, About the House :Lévi-Sirauss and Beyond, pp. 170-88 Cambridge : Cambridge University Press. Oakes, N.,1994 - The Late Prehistorie Machi Fare Haupapi ‘An Examination of Household Organization in Me‘orea, French Polynesia, Unpublished MA thesis, Department of ‘Archaeology, Simon Fraser University. Oliver D, 1988 - Return to Tahiti: Bligh’s Second Breadiruit Voyage. Henclulu: University of Hawaii Press. Oliver D., 1976 — Ancient Tafitian Society. Honolulu : Uni- versity of Hawai Press. DOrlise, ©. 1982 - Matériaux pour étude des habitations pro- tohistoriques 2 Tahit, Polynésie trancaise. These de Dacio- rat, Paris, Université Paris | Panthéon Sorbonne. Orliae, .,2000 - Fare et habitat é Tahiti Marseille: Editions Parenthavee. Ortiac, C. and M, Ortlac, 1980 Les structures de combustion et eur interprétation archéologique : Quelques exemples en Pelynésie, Journal de la Société des Océanistes, 66-67 : 6 7 Sahtins, M.,1958 - Seciol Stratification in Polynesia. Soatto American Ethnotogical Society. Stuiver M., Reimer, P.J. and Eraziunas, TF, 1998 ~ High: procision radicarton age calibration for terrestrial ang ‘marine samples. Radiocarbon 4olg] 127-151. Sutton, D.G,, ed., 1990 ~ The Archaeology of the Kainga : A ‘Study of Precontact Maori Undefended Settlements at Pouerua, Northland, New Zealand. Auckland : Auckland Univorsity Proce. Taomla, J.N.E., 2000 ~ Househeld Units in the Analysis of Prehistoric Social Complexity, Southern Cook Istands. Asian Perspectives 39 : 139-164 Van Bilder, Cand P.V. Kirch, 1997 - Household archaeology inthe Ahupua'a ef Kipape ond Nakachu. In P.V. Kirch,ed., No ‘Mea Kahiko Kahikinul, Studies in the Archaeology of Kahiki ‘nu Maui pp. 45-60. Special Publication No. 1.Archaeologicat Research Facility, University of Californie, Berkeley. Waterson, R., 1990 - Tho Living House + An Anthropology of Architecture in South East Asia Watson-Sutpll Publications, New York. \Weisler, M. and PV. Kirch., 1985 - The structure of settle- mont space in a Polynesian chiofdom + Kawela, Molokai, Hawaiian Islands. New Zealand Journal of Archeecloay 7 : 129-188.