Printer Friendly

Cultural chronology of earthworks in Palau, western Micronesia.

Babeldaob, the largest island in the Palau archipelago, is dominated by clusters of morphologically diverse earth architecture. At European contact, in the late 18th century, the massive structures in the island's interior were unoccupied and failed to play a role in local oral traditions. Determining earthwork age and function relies almost exclusively on the physical evidence unearthed in archaeological and palaeoenvironmental investigations. With a limited amount of inland archaeological data, Palau's early cultural sequences integrated earthwork structures into largely speculative relative chronologies and inferred earthwork function almost solely by morphological analogy (Hijikata 1993; Lucking 1984; Masse et al. 1984; Morgan 1988; Osborne 1966; Yamata 1930). Recent archaeological and paleoenvironmental investigations on the archipelago's volcanic islands (Athens and Ward 2002; Liston 2008a; Phear 2007), specifically those supporting Compact Road construction (Athens and Ward 2005; Athens and Stevenson n.d.; Liston 1999, n.d. c; Wickler et al. 2007), have produced data sufficient to overcome the problems inherent in constructing chronologies in heavily reworked landscapes and to allow for encompassing functional interpretations.

Earth structures were built throughout most of Palau's cultural sequence. What distinguishes the Earthwork Era structures are their significantly greater size and elaboration and their distributional patterning that forms extensive clusters of diverse morphological types. Although a comprehensive earthwork chronology is presented, the focus is on the development of these inland clusters of modified terrain during the Early, Middle, and Late Phases of Palau's Earthwork Era. Also included is a review of the functions, forms, and distribution of earth structures, an examination of changes in material cultural remains--particularly traditional ceramics, and a discussion of the appearance of monumental architecture.

[FIGURE 1 OMITTED]

"Earthwork" is defined as any intentionally altered terrain regardless of its dimensions. "Earthwork complex" is a group of spatially connected earth structures each of which is not necessarily of contemporaneous construction or identical in function. "Abandonment" refers to the final use of the interior earth structures as components of integrated political units and not to the termination of all earthwork construction or use. Palau's earthworks have never been subject to wholesale abandonment as most, if not all, post-Earthwork Era villages are supported by step-terraces and, due to the extent of prehistoric landscape modification, many interior activities continue to be on previously sculpted terrain.

The Palau archipelago comprises some 350 islands spread along a 150 km north to southwest trending arc in the western Caroline Islands of Micronesia (Figure 1). The volcanic island of Babeldaob accounts for nearly three-quarters (333 [km.sup.2]) of Palau's land mass and contains most of its earthworks. At 45 km long and from 0.5 to 13 km wide, there is no place on the island further than 6 to 7 km from the coast. Palau's remaining landforms are platform--like reef islands, atolls, and tectonically uplifted, coralline limestone islands, locally referred to as the "Rock Islands", all unsuitable for terracing. Some earth structures are sculpted into the volcanic portions of the three small islands of Oreor, Ngerekebesang, and Malakal that are close to Babeldaob's southern end.

Babeldaob has an undulating terrain with limited strips of coastal plain or beaches. Splitting the island along its north-south axis are the steep-sided, narrow valleys and heavily eroded, well-rounded peaks of the upland ridge systems, rising to a maximum elevation of 242 m a.s.1. Most of the lowland hills circling the base of the uplands are bordered by thick bands of mangrove forests. Over 90% of Babeldaob's soils are severely leached and highly acidic oxisols, with the remainder mainly alluvial deposits, bog soils, and unconsolidated calcareous sand. Babeldaob is covered in about 74% upland forest and 18% savanna species (Whitesell et al. 1987: 2). Babeldaob, the Rock Islands, and Peleliu, are separated from the surrounding barrier reef complex by a shallow lagoon of some 1200 [km.sup.2]

Palau was probably colonized by Austronesian horticulturists between 4500 and 3100 calBP (Athens and Ward 2001, 2002: 52-5; Clark 2005: 371; Clark et al. 2006; Fitzpatrick 2003; Phear et al. 2003). Little is known about colonization and early settlement, as tectonic subsidence, a fluctuating sea level, and erosional deposits have destroyed or heavily altered most paleoshoreline archaeological sites (Dickinson and Athens 2007). The population soon began to expand into Babeldaob's interior for swidden cultivation and habitation (Masse et al. 2006). A shift to an interior settlement pattern during the ensuing Earthwork Era is characterized by extensive clusters of inland earth architecture that supported the majority of community activities and defined sociopolitical districts (Liston 1999, n.d. c; Liston and Tuggle 1998, 2006). Continuing for the next millennium, the Earthwork Era is divided into the Early, Middle, and Late Phases to distinguish the development, zenith, and fall of these interior districts. After a period of little documented cultural activity, the population returned to coastal settlements identified by massive and elaborate stone architecture and a reliance on wetland cultivation (Masse 1989; Wickler 2005). Earthwork construction was largely regulated to coastal and lowland step-terraces that supported structures and dryland crops.

Palau's earth architecture

Earthworks cover a minimum of 20% of Babeldaob with their impact on the landscape perhaps double that due to the intensive erosion resulting from their construction. (Liston and Tuggle 1998; Liston and Olsudong in prep.). The often visually impressive earth architecture is found in a wide range of morphological types, sizes, and configurations. Structures include modified ridgelines, transverse and lateral ditches, embankments, earth platforms, step-terraces, bermed basins, platform terraces, raised earth paths, leveled plains, ring-ditches, modified gullies or swales, and steep-sided and flat-topped hills called "crowns" (Figures 2 and 3).

Earthwork complexes and individual earth structures can be massive in scale. Step-terraces and ditches descend up to six metres, while the modified hilltops (crowns) rise as much as ten metres above the surrounding landscape. These crowns display a generally small surface area of ca. 20 to 80 metres square, though the surface of Ngerulmud is approximately 180 metres square in area (Liston et al. 1998). To quantify the amount of earth needed to build such structures requires extensive excavation. Phear's (2007: 56-8) excavations in the summit of the Ngemeduu complex suggest that at least 2500 [m.sup.3] of fill was used just for crown construction.

Earthworks are generally not single isolated structures but components of up to 27 [km.sup.2] clusters of modified terrain extending from the coastal hills up to the central ridgeline (Figure 4). Found from one end of Babeldaob to the other, each of the largest ten clusters roughly corresponds to a modern political state. High hills shaped into imposing crown earthworks are strategically located to conceivably control access into the cluster by land (and m some cases from the ocean), and to provide long-distance surveillance from high lookout positions (Liston and Tuggle 2006). Between these clusters are expanses of unshaped terrain occasionally broken by considerably smaller groups of more modest earth features. The extent and distributional patterning of earthworks on the current landscape is a culmination of centuries of occupation in which the initially small groups of structures expanded or incorporated nearby clusters. Just inland of the coastal swamps and mangrove forests are compact groups of step-terraces, many of which date to a later period than the larger and more morphologically diverse inland earthworks.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

Earthworks were built as a single event and in several phases using a combination of sculpting and cut-and-fill techniques. Modifications to the features were probably associated with a significant historical event, a functional change, or a renovation due to structural collapse. The whole pots, pottery caches, and other significant objects deliberately inserted beneath and within some earthwork fill suggest ritual activity associated with construction or with the sociocultural event precipitating the construction (Liston 2008a: 41: Olsudong et al. 2008: Tuggle n.d.). These ritual events could relate to Phear's (2007: 133-7) proposal that the earthwork building process is essential to shaping group identity and social inclusiveness.

Earth structures served as community infrastructure, played ceremonial or ritual roles, and supported cultivated fields and settlements, among other tasks. An integrated system of earthwork usage was likely in effect from the onset of their construction with a complex's individual components (even those of the same structural type) potentially serving different functions from each other and being multifunctional. Evolving with the society over a millennium of interior occupation, earthworks continuously transformed in function, patterning, form, dimensions, and meaning. As no single role is specific to a particular morphological form, discerning a structure's function(s) relies on its temporal and spatial context.

Step-terraces, most following natural land contours, are the principal structural component of earthwork complexes. They supported specialized activity areas, ritual centers, family, clan, or group settlements, and gravesites. Those with high risers served double duty by also blocking access into or out of significant areas. The most advocated explanation for step-terrace construction is their use for dryland taro cultivation (Hijikata 1993; Morgan 1988: 12; Osborne 1966:151-5). However, investigations designed to identify evidence supporting this argument have proved futile (Lucking 1984). Analysis of pollen and phytolith samples collected from Ngaraard's interior structures revealed none of the cultigens which are notoriously difficult to locate archaeologically (Henry et al. 1996: Liston 2008a; Phear 2007, 2008).

Along the coastline, platform terraces contain small pondfields crossed by stone paths (Liston n.d. b). The pondfields do not appear to form complex wetland cultivation systems as do those associated with the later Stonework Era villages. Wide gullies, some paved, are liberally dispersed throughout the earthwork clusters for drainage and irrigation (Liston 2008a: 118). The extent of interior irrigation techniques is uncertain though a picture is beginning to emerge of a system of gullies and ditches truncating ridges to empty into shallow basins and descending the length and width of step-terrace complexes. Raised fields or cultivation beds have not been identified in Palau's prehistoric record.

[FIGURE 4 OMITTED]

Ditches and embankments, some with palisades, are strategically placed across ridgelines, between adjacent crowns, and circling the base of crowns to create confined ritual or social space (Phear 2007: 135) and/or impede access into populated or special use areas. Some ridgelines are cut and elevated to create trails through the steeper terrain. Other ridges are leveled and expanded to support a series of low earth platforms spanning the width of the ridge.

At the highest point of most earthwork complexes, the hilltop was cut or elevated with fill material to form a steep-sided and flat-topped structure called a crown. Many crowns are accompanied by a ring-ditch that in some instances could be a later addition to the pre-existing structure (Figure 5). Capping some crowns are earth knobs and berms, the latter often creating single or double enclosures since eroded to resemble depressions. Hilltop crowns are interspersed inside, and create the perimeter around, earthwork clusters. Squat crowns are positioned to overlook a group of structures such as sets of likely agricultural step-terraces (Liston 2007).

Though their small surface area is impractical for most mundane pursuits, crowns undoubtedly could have supported a single civic or ceremonial building (Phear 2007: 320), perhaps an elite residence. Some crowns, or the knobs and berms capping the features, served as burial grounds for high-status individuals (Tuggle n.d.). Other berms could be altars for ritual activities. Oral traditions, associated with the period after the interior earthwork districts were abandoned, identify high hilltops (many of which are crowns) as ritually or mythologically significant or as smoke-signal towers (klekat), lookouts, and places of refuge or asylum (Kramer 1929: 95; Tellei et al. 2005: 72, 81; see Phear 2007). The place names of a few crowns (e.g. Meklechel a Beab "taro swamp of the rats"), suggest horticulture, which, in such a restricted space, probably had a ceremonial intent. Excavations in the shallow enclosure capping a Ngaraard crown (Phear 2007: 52-65) and the ca. 2.5 m deep enclosure in a Ngiwal crown (Liston 2008b) did not identify evidence of cultivation.

[FIGURE 5 OMITTED]

In the main, the sheer angle and height of most crowns, their commanding view over great expanses of the island and lagoon, their strategic placement within clusters, and their association with palisades and ring-ditches implies a defensive role (Kaschko 1998; Liston and Tuggle 2001, 2006; Osborne 1966: 240). However, the crowns' often monumental size and embellishments suggest that they were significant to society as symbolic structures (Liston 1999: 351-3; Liston and Tuggle 1998; Wickler 2002). Liston and Tuggle (2001, 2006: 65) suggest that the "major defensive element of the Palauan terrace complexes was the appearance of power" with the crowns mirroring the associated polities prestige level.

Individual or grouped elements of stone features, including facings, alignments, pavings, and paths, are commonly associated with interior earthworks (Liston 1999, 2008a, n.d.c; Tuggle n.d.). The basalt, and occasionally coral, structures are generally distinguishable from the later Stonework Era features by the use of smaller, subangular stones and the tendency for single-course pavings rather than raised stone platforms with earth cores. Though present, stone retaining walls are not commonly encountered on Palau's step-terraces due to the stability of the clay soils and earthwork construction techniques. Overall, Earthwork Era stone features do not display highly skilled engineering techniques or focus on monumentality but appear to be the conduit to achieving a functional end, such as a foundation for a superstructure or erosion control. However, in certain circumstances the presence of stonework could denote a stratified society, as only the upper classes may have had paved living surfaces or burial markers. The stone facings embellishing some crowns and high step-terraces appear to be symbolically significant rather than an engineering element.

Stone mortars, large bowls, anthropomorphic faces, altars, petroglyphs, monoliths, and other types of modified stone are interspersed among the earthworks (Hijikata 1995; Osborne 1966; Rieth and Liston n.d.; Van Tilburg 1991). Their traditional association with gods or foreigners and their absence in oral traditions implies a connection with the similarly depicted earthworks. The sculpted faces, referred to as "stones of the gods", are often found in Stonework Era villages where they were likely relocated from the interior once the inland earthwork districts were abandoned (Hijikata 1993: 58; Van Tilburg 1991: 12). Palau's only megalithic complex, Badrulchau, is surrounded by crown complexes in the Ngarchelong earthwork district. Their proximity and relative location suggest the monumental features are contemporaneous.

After placing the interior earthworks into an interpretive framework of a regional settlement model, Liston (1999: 410-3) and Liston and Tuggle (1998, 2006) proposed that the bulk of community activities took place on the earth structures with the clusters defining fortified sociopolitical districts (see Snyder 1997). They suggest that the spatial distribution, size, and morphology of the complexes indicate that, though serving practical uses, the earthworks functioned as symbols of individual chiefly or polity power, to legitimize corporate claims of land and other resources, and to create defensible terrain. To understand the processes and social context of earthwork formation Phear (2007) applied a landscape-historical perspective to her investigations of the monumental earthworks. She concludes that these diachronic structures encompass elements of ritual or sacred significance used to formalize the meanings associated with the land and ancestors (Phear 2007:129-40).

Methods

Establishing an earthwork chronology is difficult because of the complicated depositional histories resulting from cut-and-fill construction, periodic, intense erosional events, and repairs or additions obscuring or destroying earlier structures (Liston 2005: 303). This is compounded by the sheer size of the earthworks and their multifunctional and functionally evolving components requiring extensive investigation of each complex.

In this heavily altered landscape, ascertaining what event an assay is actually dating is dependent on careful site-formation analysis and radiocarbon determinations originating in both securely dated features and secondary depositional units. Paleoenvironmental evidence can serve as "proxy indicators of human impacts on the environment" (Dickinson and Athens 2007: 191) and provide an independent check to the archaeological interpretations. Material culture attributes, particularly traditional pottery, can assign temporal parameters to the earthwork's complex strata by providing relative diagnostic markers much like [sup.14]C assays in the same contexts. These attributes can also assist in identifying site re-use. The new data provided here contribute to the development of Palau's ceramic sequence (Clark 2005: Cochrane and Tuggle n.d.; Desilets et al. 1999; Liston et al. in prep.; Snyder 1989).

Liston (1999: 336-45) constructed a Palauan cultural sequence by superimposing the chronologically organized archaeological sites onto the temporal framework offered by paleoenvironmental investigations, This sequence clearly tied the oscillating pattern of savanna/forest formation identified by Athens and Ward (2002, 2005) with the occupation and abandonment of the interior earthworks. Incorporating the substantial quantity of data accrued during more recent excavations into this sequence both tests and refines the previous chronology. Distinctions between the three phases in the Earthwork Era are based on the number and relative location of assays and sites.

This sequential model is necessarily generalized due to gradual transitions in eras and phases (Phear et al. 2003), the likelihood of localized chronological variations, and the limited data for many earthwork clusters.

Radiometric data

A total of 179 radiocarbon dates, all on wood charcoal, originate in interior contexts. Radiocarbon ages were calibrated using OxCal v. 3.10 software (Bronk Ramsey 2005, atmospheric data from Reimer et al. 2004) at two standard deviations (95.4% probability) using the Intcal04 calibration curve. Application of standard protocols for chronometric hygiene (Anderson 1991; Spriggs and Anderson 1993) adapted to a Palauan context (Liston 2005) removed 48 (27%) determinations from the analysis. As charred material recovered from a secondary context in earthwork sites can be useful in establishing chronologies, the almost universal protocol of only assigning credibility to samples originating in a primary cultural deposit or discrete archaeological feature is not applicable to Palauan earthwork assays. Details about the radiocarbon dates and their contexts are found in individual site reports (see Liston 2005).

Of the 131 accepted assays, 44 (34%) derive from postmolds, garbage pits, burn events, or other subfeatures, 38 (29%) from construction fill, 27 (20%) were collected in primary cultural horizons, and 14 (11%) originate in erosional deposits (Table 1). The context of eight (6%) samples is indeterminable and is thus interpreted as a secondary deposit. Thirty-one earthwork sites, including modified ridges, crowns, earth platforms, ring-ditches, basins, platform terraces, leveled plains, and step-terraces, are represented in the chronology.

Nine earthwork clusters - one on Oreor state and eight on Babeldaob are included in the analysis. The majority of the inland assays (n=73, 56%) are from the intensively investigated earthwork districts in Ngaraard state. Of these, 17 were collected in a small earthwork cluster and the remaining 56 from the larger district. Archaeological work in Ngiwal and Ngatpang has produced 23 (18%) and 20 (15%) assays. Eight (6%) of the assays derive from excavations in Melekeok, three (2%) from the minimal tests in Airai, two (2%) from Ngarchelong, and one (1%) each from Ngchesar and Ngerekebesang, a part of Oreor. Ngardmau, Ngaremlengui, and Aimeliik, the latter two with the largest earthwork complexes on Babeldaob, have not produced any accepted [sup.14]C dates.

Paleoenvironment record

Nine of the 26 wetland cores recovered from Babeldaob, Ngerekebesang, Oreor, and Peleliu's raised limestone reef have undergone full analysis, including radiocarbon dating, sediment examination, charcoal particle analysis, and pollen sequencing (Athens and Ward 2002, 2005; Athens and Stevenson n.d.). Five cores were submitted to both sediment and [sup.14]C analyses, while only pollen sequences were obtained for four cores.

The paleocores document an alternating pattern of savanna growth and decline, paralleled by deforestation and forest regeneration. In general, the sequence indicates that savanna formation began on a mostly forested Babeldaob around 4500 to 4300 calBP (Athens and Ward 2005: 113; Masse et al. 2006: 127). The surge of savanna indicators between about 4400/3000 and 3800/2500 calBP (different cores having different dates for this surge) suggests extensive savanna formation on Babeldaob. An interval of forest regeneration then seems to be documented in the pollen records between roughly 3800/2500 and 3000/1650 calBE This is followed by a resurgence of savanna vegetation between 3000/1650 and 2500/500 calBP with a return to more forested conditions after about 1000/500 calBP (again, different cores having different dates) (Athens and Ward 2005: 113). The disparities in the savanna/forest regeneration cycles between the cores presumably reflect differences in the timing of these events across Babeldaob. The geomorphological evidence reveals abrupt transitions from largely wetland sediments to colluvial material at ca. 2500 calBP (Athens and Ward 2005: 115; Athens and Stevenson n.d.).

Earthwork chronology

Earthwork construction is unlikely to have begun in Palau's Colonization Era (ca. 4300-3200 calBP) but may have started in Palau's Expansion Era (ca. 3200-2400 calBP) as the growing population began to occupy Babeldaob's interior (Table 2). The majority of interior earthworks were built during the Earthwork Era (ca. 2400-1200 calBP) when extensive clusters of modified terrain defined sociopolitical districts. There are far fewer Transitional (ca. 1200-700 calBP) and Stonework Era (ca. 700-150 calBP) interior earthwork dates. Earth construction continued on a severely diminished scale in the latter period with the building of step-terraces to support near-coastal villages.

The Earthwork Era is represented by 27 interior sites (all earthworks) located in seven clusters, 86% (n=113) of the total interior assays, sedimentary disconformities in five paleocores, a largely savanna interior landscape, and changes in pottery attributes. The era is subdivided into three loosely bounded phases - Early (ca. 2400-2150 calBP), Middle (ca. 2150-1500 calBP), and Late (ca. 1500-1200 calBP)--that correspond to the growth, zenith, and decline of interior earthwork occupation.

During the Earthwork Era, limestone caves served as mortuaries (Berger et al. 2008; Fitzpatrick 2003; Rieth and Liston 2001) and the Rock Islands, Angaur, and Peleliu were first semi-permanently then permanently occupied and functioned as bases for short-term activities (Clark 2005; Masse 1989; Wright 2005). Babeldaob's coastal zone is poorly represented with only two assays dating to the Earthwork Era (Mangieri 1998a, 1998b). Regardless, the coastal margin played an integral role in resource procurement, including wetland horticulture, and, where viable, was likely inhabited.

Pre-Earthwork Era

Palau's initially small colonizing population probably had no compelling reason to settle inland areas. During the little understood Colonization Era (ca. 4300-3200 calBP), interior use may have been limited to intermittent bird and bat hunting, supplying transportation routes for travellers and hunters, and gathering wild fruits, medicinal plants, and construction materials. Hill summits may have served in various lookout, communication, and ceremonial capacities. As all of Babeldaob is within 6 to 7 km of the shoreline, forays into the interior did not require overnight camps.

A spike in savanna indicators at ca. 3200 calBR interpreted as deforestation caused by swidden horticulture (Athens and Ward 2005:113), introduces the Expansion Era (ca. 3200-2400 calBP). Marking the gradual shift to a greater reliance on the interior landscape for settlements and swidden cultivation, this era contains eight (6%) of the total inland assays. The [sup.14]C dates derive from pre-earthwork contexts in four interior sites. There is no evidence of earthwork construction at this time.

Early Phase (ca. 2400-2150 calBP)

The inception of the Earthwork Era is documented by sedimentary disconformities in five of the nine dated paleocores, a change in ceramic temper from beach sand to grog, and the expansion of site types and occupied area. Six ridgeline, hill, and coastal lowland sites in four earthwork clusters have produced eight Early Phase (ca. 2400-2150 calBP) radiocarbon determinations or 6% of the total inland assays.

Earthwork construction, possibly of step-terraces, may have begun as early as 2550 calBP on the coastal slopes next to early settlements (Liston n.d. c). The degree and scope of earthmoving over the next few centuries is uncertain. It appears to include leveling and expanding ridgelines, flattening hilltops, and cutting step-terraces into the slopes as the population began to occupy the long, narrow ridges and gentle hills next to their intensifying dryland field systems.

The five paleocores displaying abrupt transitions from wetland to upland sediments at ca. 2500 calBP are strong indicators of the widespread onset of earthwork construction (Athens anti Ward 2002, 2005: Athens and Stevenson n.d.). Two of these cores identify the advent of a savanna landscape a few centuries before or coinciding with the mass wasting events. In Ngaraard, cores extracted in swampland below a modified ridge and from a pond sandwiched between the impressive Ngulitel and Olbakeldrau earthwork complexes date the influx of erosional material to between ca. 2200 and 2500 calBP. Ngaraard's earthwork construction is archaeologically dated to begin at ca. 2250 calBP or a. little earlier, concoinitant with the large-scale erosion documented in the paleocores. The sedimentary unconformities exhibited in two cores recovered in a marsh beneath Ngerekebesang's crown complexes date to 2300-2200 calBP or possibly a little earlier and increased soil erosion is recorded in a core in the Ngardmau district at ca. 2500 calBP. The one archaeological date from the earthworks adjacent to these latter cores is associated with Transitional Era activity rather than initial earthwork construction (Magnuson and Listen 1998).

An additional indicator of significant interior earthmoving is the noticeable shift in temper type from beach sand to grog at 2400 BP. This change is coincident with the deposition of upland erosional material reducing the availability of calcareous sand (Clark 2005: 372).

Suggested motivations behind the transition to a largely interior settlement pattern relate to a limited coastal margin, population pressure, and commensurate competition between groups or villages (Athens and Ward 2005: 99: Listen 1999: 410-3: Listen and Tuggle 2001,2006: Lucking 1084; Masse et al. 1984: 122: Osborne 1966: 153). Osborne (I 966:464-5) speculates that Palau's earth architecture may be a result of cultural diffusion from Southeast Asia, possibly the Philippines. If so, this transported landscape did not extend to other Micronesian island groups, whose insignificant number of earthwolks is confined to small step-terraces. Regardless of their origin, there it no evidence to suggest that the evolution in earthwork design, Configuration. function, and sociocultulal significance it anything other than a result of adaptations to Palau's unique cultural system and environment.

Middle Phase (ca. 2150-1500 calBP)

The Middle Phase of the Earthwork Era begins with the first definitive evidence of earthwork construction at ca. 2250 calBP, grows to contain the majority of the total inland assays (n=77. 59%), and ends with a dramatic decline of interior [sup.14]C determinations. During this ca. 600-year period, represented by 22 cites in six settlement clusters, simple earth structures evolve into large complexes whose distributional patterning defines sociopolitical districts.

The first certain archaeological indicators of earthwork construction are an intact cultural horizon in a Ngaraard platform terrace (Listen 2008a) and two subfeatures descending into the earth platforms lining ridgelines in Ngaraard and Ngiwal (Listen n.d. b: Tuggle n.d.). The cultural deposit is dated to 2320 2040 calBP (B236596) and the two subfeatures to 2330-1990 calBP (WK8297) and 2180- 1950 cal BP (WK 14019). Construction fill in both cites produced Early Phase assays indicating previous occupation and, potentially, prior earthwork construction.

All 41 of Ngaraard's Middle Phase [sup.14]C determinations (representing 60% of the district's inland dates) were recovered along a 3 km length of integrated earthwork components within the ca. 7.5 [km.sup.2] inland district. The calibrated combined date range of the charred remains collected in Ngaraard's intact cultural horizons, burial pits, a possible hearth, support posts, a burn lens, a refuse pit, a pottery cache, and secondary deposits is 2330-1480 BP. The Ngiwal earthwork cluster, with 17 Middle Phase dates (74% of the district's inland assays), closely tracks its northern neighbour with a combined date range of 2180-1560 calBP. The bulk of Ngaraard's Middle Phase assays (61%, n=25) fall between ca. 2050 and 1750 calBP, a temporal patterning mimicked in Ngiwal. Considering the extent of archaeological investigations and the number of assays produced, the chronological similarity in these two clusters suggests that northeastern Babeldaob's earthwork districts reached their zenith about 1900 years ago. Furthermore, the dramatic decline in the Ngaraard and Ngiwal assays after ca. 1550 calBP implies they had essentially disbanded by the end of the Middle Phase. Evidence of conflict and horticulture is found ill both districts in the Late Phase.

Intra-district patterning of the modified terrain, in conjunction with radiocarbon assays, suggests that settlement areas, fields, ceremonial centers, and the like Wele constructed in relation to the high hills since the beginning of interior occupation. Although crowns were enlarged, re-shaped, or embellished with knobs, berms, or stonework during their long period of use. they are not all appendages to existing earthwork complexes (Listen and Tuggle 1998). Multiple assays collected in mass wastage at the base of Ngaraard's imposing Ngemeduu complex record the onset of substantial hilltop modification by ca. 2200 calBP (Listen n.d. a). Phear (2007: 320) identified ca. 2000-year-old structural remains on the already shaped Ngemeduu crown. Modifications to five other hill summits dale to ca. 2100-1700 calBP and possibly earlier, providing a conservative date of ca. 1800 calBP for the onset of crown elevation.

The nature of district occupation during the Middle Phase is uncertain. Step-terraces were cut into almost all sloping terrain and lined stream drainages, Transverse ditches bisected ridges leading to hill summits and longer ridgeline trails. Low crowns perched on some step-terraces and larger ring-ditched crowns bound the complexes into individual clusters of modified terrain. Household units and hamlets or villages, identified by stone paths, parings, and other remnant stonework, structural support posts, and residential debris, were scattered throughout the cluster on leveled ridges and step-terraced hillslopes (Liston n.d. c). It appears settlement areas were widely dispersed within earthwork districts and possibly along shorelines and stream banks below the interior clusters.

Liberally scattered throughout hill and ridge slopes are large pit features that may identify breadfruit (Artocarpus) preserving pits like those observed by Kramer (1926) at the turn of the 19th century. Paleocores document Artocarpus pollen as being "continuously present in some quantity" after 2400 calBP (Athens and Ward 2002: 57) and it is identified in Ngiwal and Ngaraard earthwork district archaeological contexts at a combined date range of 1900-1600 calBP (Murakami n.d.). Currently of little dietary importance, breadfruit may have been a significant contributor to the Earthwork Era diet. Athens and Ward (2005) and Murakami (1999, 2007, n.d.) identify seven other cultigens in Earthwork Era cores or archaeological contexts including Areca catechu, cf. gourds, Cocos nucifera, Colocasia esculenta, Cyrtosperma chamissonis, Syzygium malaccense, and Manilkara udoido. Other identified edibles include Hibiscus tiliaceus, Pandanus spp., Calophyllum inophyllum and possibly an edible species of Bruguiera. Yams (Dioscorea spp.) have not been identified in the prehistoric record. As bone and shell rapidly degrade in Babeldaob's soils, the variety of marine, avian, and other species consumed in interior Babeldaob during the Earthwork Era cannot be identified. It is likely however that the Rock Island and Babeldaob's coastal populations exploited many of the same protein resources.

Earthwork Era interments, possibly restricted to those of status, are found in small structured burial sets on step-terraces (Liston n.d. a; Tuggle 1998: 352-6). The upland gravesites are defined by burial pits that are often oriented east-west and capped with varied combinations of clay, sea sponge mats, an unidentified fibrous material, and small pavings or rock mounds. Some are associated with burial furniture in the form of perishable offerings contained in whole pots and, in one instance, a cache of round basalt cobbles.

Elevated locations, providing symbolic impact, were possibly reserved for the elite or priestly class and defensive functions. The shortage of domestic midden in contrast to the quantity and variety of atypical and specialized features of potentially ritual association implies Ngaraard's Tabelmeduu ridgeline supported the district's paramount village or ceremonial centre (Liston 2008a: 115-6). The whole vessels and pottery caches that are densely distributed among and under the earth platforms lining the flattened ridge are associated with earthwork construction, suggesting their ritual interment (Liston 2008a; Tuggle n.d.). Sandwiched between an earth tier and its supporting earth platform are two whole vessels (one inverted on the other), a two-metre deep stone-filled pit and pottery caches (Liston 2008a: 43-7); intentionally placed in the construction fill above these atypical features are nearly whole pots. Nearby, carefully stacked stone alignments supported an elevated structure associated with the adjacent feature that is tentatively interpreted as an open-fire pottery kiln. If so, the kiln's placement on Tabelmeduu suggests pottery production was highly valued, with master potters allotted socially significant areas in which to practice their craft.

Late Phase (ca. 1500-1200 calBP)

A significant drop in interior radiocarbon determinations and site occupation, accompanied by signs of conflict, indicate the abandonment of inland earthwork districts during the Late Phase (ca. 1500-1200 calBP) of the Earthwork Era. The 28 (21%) Late Phase determinations are found in 10 sites within one small and four large earthwork clusters. The majority (n=24, 86%) of these dates originate in crown or ditch deposits. Despite their provenience, only seven of these assays (four sites) are unquestionably associated with defensive activities.

Occupation of the Ngaraard earthwork cluster decreased to the point that remodeled crowns appear to be the main focal point. Stone-lined pits unearthed in a knob capping the previously constructed Roisingang crown (Tuggle n.d.) and the magnitude of the effort needed to elevate the Ngemeduu crown about four metres (Phear 2007: 57) suggest chiefly interments, ritual activity, or a need to symbolically express polity power during the combined date range of 1530-1170 calBP.

Forest expansion, which would be expected to coincide with diminishing interior use, is not recorded in the paleocores until ca. 900 calBP--about 600 years after the majority of the population had moved out of northern Babeldaob's inland districts. This contradiction in the palaeoenvironmental and archaeological evidence might be resolved with data from Aimeliik and Ngaremlengui--the two largest uninvestigated earthwork districts.

While occupation of northeast Babeldaob's earthwork districts was declining, some interior earth structures located outside of the larger clusters were initially constructed or more intensively occupied. Little is known about these small groups of Late Phase low crowns, basins, and step-terraces.

With careful stratigraphic interpretation, assays from infilled ring-ditches can identify the associated crown's final use as an earthwork district component and, by extension, the fall of district power. Three [sup.14]C dates from the three-metre deep and eight-metre wide ditch circling the base of the Ngebar crown provide corroborating evidence for the timing of this Ngatpang district fortification's defeat (Liston n.d. b). The charred remains of a Pterocarpus indicus post in the ditch's inner margin is the first concrete evidence for palisades in Palau's prehistoric defensive features. The calibrated combined date range of the samples retrieved from the post, a charcoal lens overlying a 35 cm thick, gleyed matrix, and one of 17 infilling matrices is 1420-1170 calBP. Culmination of pondfield cultivation in the nearby Ngermedangeb complex just before ca. 1150 calBP (Liston n.d. b) coincides with the events on Ngebars. Together, the evidence strongly suggests the decline of the Ngatpang earthwork district at ca. 1300 calBP.

The impetus for the move out of the interior includes the interrelated variables of the degradation of terrace soils and the opportunity for shoreline settlements and wetland agriculture due to coastal expansion resulting from centuries of erosion (Athens and Ward 2005: Hunter-Anderson 1991; Liston 1999; Liston and Tuggle 2001, 2006).

Post-Earthwork Era

The Transitional Era (~1200-700 calBP), after the dissolution of the large interior earthwork districts, contains eight (6%) assays recovered from seven sites in two small and two large earthwork clusters. Some cultivation within the largely abandoned districts is indicated while beyond their boundaries the limited construction of inland structures is mostly confined to low crowns. There is little documented cultural activity along Babeldaob's coast (Mangieri 1998b: 292-3; Wickler 2005: 302-7) and a decline in limestone cave interments (Berger et al. 2008; Rieth and Liston 2001). The increase in permanent settlements on the Rock Islands and Peleliu (Beardsley 1997; Clark 2005; Masse 1989) is not significant enough to suggest a wholesale migration of the population out of Babeldaob (Masse et al. 2006: 112). The lack of Transitional Era sites may be partly due to erosional events obscuring or destroying Babeldaob's coastal archaeological remains or to limited sampling in key areas.

A dramatic and abrupt rise in coastal radiocarbon assays beginning at ca. 700 calBP or a little earlier is countered by just two (2%) post-700 BP inland assays. This marks the beginning of the Stonework Era that continues until the first long-term European contact in the late 1700s. Settlements are identified by massive and elaborate stone architecture (Masse et al. 2006: 112-5; Wickler 2002, 2005: 236-45). Compact complexes of these stonework villages were positioned behind a defensive barrier of mangrove forests or swamplands and bounded by wet- and dryland agricultural fields (Liston and Tuggle 2006). Nearby earthworks of an earlier era were integrated into village defenses and ritual activity. The Rock Island's stonework villages were eventually abandoned as only Peleliu, Angaur, Babeldaob, and the latter's neighbouring small volcanic islands were inhabited at European contact (Masse 1989).

Earthwork construction is not terminated outright but is largely confined to the low step-terraces supporting individual household units, community structures, and dryland fields. Stone paths bisecting these step-terraces suggest the latter's pre-Stonework Era construction. However, since only four of the >60 reliable stonework village assays pre-date ca. 700 calBP, it is improbable these terraces are of much greater antiquity than the structures they support. Regardless, the close spatial relationship between interior earthwork districts and coastal stonework village complexes clearly shows their association (Butler 1984: 36; Cordy 1984: 68-9). Unrecorded Earthwork Era coastal hamlets could be the forerunners of these later settlements (Liston 1999: 374-6, 411-2).

The initial move coastward off the interior earthworks may have been to the slopes of ancient, near-coastal earthwork complexes, to sites chosen for their defensive advantage or their religious or ancestral association. Displaying stonework village characteristics, these lowland sites could be the final manifestation of long-occupied settlements or a re-occupation of abandoned earthworks. Oral traditions relate that these villages are precursors to those on the coastal flats. Perhaps only later, once the political order and economy of the new era was firmly established, was there a final transition to coastal stonework villages.

Earthwork Era structures were not entirely abandoned once the population moved coastward. The feature types and architectural style of some stonework found on interior earthwork complexes and their inclusion in oral traditions suggest that they are a Stonework Era addition to previously abandoned earthworks (see Lucking and Parmentier 1990). One example is the chiefly meeting house (bai) platform, head trophy stone (omeroel), and the coral and basalt retaining wall on top of the massive Roisingang earthwork complex (Tuggle n.d.). The high-status or symbolic role of these features indicates their potential association with Chief Ngirairung who, in traditional narrative, crossed Roisingang while relocating stonework villages from Ngaraard's west to east coast.

During the following centuries, Babeldaob's interior reverted back to its comparatively limited pre-Earthwork Era use for transportation routes and resource procurement. The vantage points offered by high hills ensured their continued use as communication posts, lookouts, places of refuge, and perhaps ritual centers of symbolic significance for the new coastal polities (Basilius 2002: 148; Kramer 1929: 95: Parmentier 1987: 273; Tellei et al. 2005: 72, 81). With the arrival of the Japanese in 1914, the interior was again extensively exploited for agricultural pursuits while earthworks behind coastal villages became community cemeteries. Some of these, apparently coincidentally, hold Earthwork Era gravesites.

Material culture

The social and economic transformations related to the shifts between coastal and interior settlement patterns and wet and dryland cultivation should have influenced the stylistic, compositional, and technological character of Palau's material culture. Developing a chronology for material remains is problematic due to the degradation of organic material in the volcanic islands' acidic soils, and the mixing and obscuring of cultural horizons in Babeldaob's heavily eroded and reworked contexts. Other difficulties include the limited data sets associated with Palau's colonization and expansion eras and the potential disparity between contemporaneous Babeldaob and Rock Island assemblages due to variable exchange patterns and a distinct or narrower range of activities on the smaller islands.

Throughout Palau's archaeological record, the most common components of the lithic toolkit are expedient chipped stone implements. Recent technological analysis documents a stable lithic technology based primarily on the bipolar reduction of chert nodules to form simple tools (Williams and Markos 1999; Williams and Pope n.d.). Preliminary investigations suggest the dominance of stone implements early in Palau's cultural sequence with a later shift to a greater reliance on bone, wood, and shell tools that potentially coincides with the move out of earthwork districts (Haslam and Liston 2008). Although a comprehensive study of Palau's formally shaped tools is not yet available, no highly distinctive temporal changes are noted.

Recovery of vertebrate and invertebrate remains is virtually impossible in Babeldaob's acidic oxisols. Degradation of perishable materials is of such magnitude that Pregill and Steadman (2000: 138) claim that the "scarcity of rich, fossiliferous sediment in Palau is thus far unequalled for tropical Pacific islands". Studies of faunal assemblages are therefore confined to Rock Island and coastal middens, none of which are associated spatially and few temporally with Earthwork Era sites. The most common fish recovered from those Rock Island middens that date within the Earthwork Era are parrotfish, threadfin, sea bass, and whiptail breams (Clark 2005; Fitzpatrick and Kataoka 2005). Potentially related to the change in settlement patterns and subsistence technologies are the temporal shifts identified in size and predation of marine resources at the juncture of the Transitional and Stonework Eras (Carucci 1992; Masse 1989). The changes likely indicate overharvesting of some inshore marine resources as the Rock Islands became more populated (Masse et al. 2006: 121-5). With limited data, Masse et al. (2006: 120-1) suggest that the pig (Sus scrofa) was once common in Palau, began to decline in the Transitional Era, and was extirpated in the Stonework Era. Although the dates suggest a correlation between the demise of the pig and the move off the interior earthworks, additional data is needed before the timing or the role of the pig in Palau becomes clear.

Clark (2005) recognized a striking change in pottery manufacture at ca. 2400 BP simultaneously with the onset of the Earthwork Era. He attributes the transition from calcareous-terrigenous sand to grog temper and the associated use of carbonaceous clay to a reduction in available beach sand due to the influx of upland soils and resultant proliferation of mangrove forest. The temporal change in stylistic and compositional pottery attributes documented by Clark (2005) and Wright (2005) at Ulong (a Rock Island) is replicated in Babeldaob's ceramic assemblages (see Desilets et al. 1999 for a review of Palau's ceramic studies). The larger island's more extensive pottery collection exhibits both a wider range of attributes and distinctions in contemporary assemblages related to site or feature type and spatial distribution. Liston et al. (in prep.) will provide detailed documentation of Palau's pottery sequence.

Babeldaob's pre-Earthwork Era ceramic assemblage is not substantial enough to contrast with that of the Earthwork Era but is recognizable in the trend away from calcareousterrigenous sand temper, dark black paste, and vessels walls measuring less than 5 mm thick. Perhaps restricted to the Expansion Era are thin (<3.5 mm) vessels displaying a coiled, footed base (Liston n.d. b).

In addition to shallow bowls, globular bowls with restricted orifices and a backcurving rim are often unearthed in Early Phase deposits. Burnished vessels or those finished with a red or brown slip are also common to early Earthwork Era assemblages. This technique gradually diminishes and is rarely encountered after ca. 1500 calBR Nubbins located just below the vessel's lip are functionally interpreted as handles or suspension supports (Desilets et al. 1999; Liston 2008a) and may be Early Phase pottery markers.

By the Middle Phase, there is a strong tendency towards grog-tempered vessels fired in a reducing atmosphere with a resultant gray-black core sandwiched between orange-brown surfaces. Although highly variable, average body thickness is between 7.0 and 10.5 mm. Earthwork Era vessels are typically direct or inverted bowls--oval, round, shallow, globular, and collared, with straight-sided or interior-thickened rims and slightly rounded bases. Current analyses record an average orifice diameter of 30 cm with a range from 11 to 50 cm and an average vessel depth of 12 cm. Decorative elements are few. Red or orange paint covering either surface or applied in bands to form simple designs (see Phear 2007: 126-8) and impressed or incised geometric patterns close to or on the vessel's rim are occasionally encountered. A single shallow groove circling the rim's base is frequently found in Ngaraard's interior assemblages. A few whole pots dated to the Middle Phase display small (ca. 3-4 mm) holes found in pairs parallel to and just below the lip (Desilets et al. 1999; Liston 2008a). The piercings were possibly used to suspend the vessel or attach a lid. By the Stonework Era, these holes are a centimeter or more in diameter and are arranged in pairs perpendicular to the rim (Hijikata 1995: 268).

Whole or almost complete ceramic vessels are often unearthed in earthwork districts (Liston 2008a: 115-6, n.d. a, b; Olsudong et al. 2008; Tuggle 1998: 340-4, n.d.). The pots are found between burial pits, on small pavings potentially associated with interments, on prepared surfaces and in caches beneath earthwork construction fill (Figure 6), intentionally inserted in the fill, and in the varied sized pits common to the interior districts. In the majority of cases, the vessels' ceremonial or ritual significance is strongly suggested by their association with a mortuary activity or special function site and their deliberate concealment (Clark and Wright 2007). Barring the large number of thin-walled vessels in some Tabelmeduu caches and the potential association of shallow, oval bowls with gravesites, the pots appear to display the same variety of forms, attributes, and decorations encountered in contemporaneous ceramic assemblages. This could signify that in some instances a vessel's meaning shifted from domestic to ritual within their lifetime.

Post-Earthwork Era changes in the pottery sequence are not fully realized until the Stonework Era, perhaps as late as ca. 500 BR In general, post-Earthwork Era assemblages are characterized by open bowls, large jars, and plates or dishes. Rims are thickened or flanged and become progressively larger and denser. Changes in clay sources and firing technology related to the move to the stonework villages is realized in the now fully oxidized, orange-brown or occasionally white, gray, or pink colored pots. Decorated vessels continue to be scarce and are limited to incised geometric designs and painted bands or surfaces. Impressed, stylized human figures on the vessels (Osborne 1979: 146) may date to the Transitional, or less likely, the Stonework Era.

[FIGURE 6 OMITTED]

Without further lithic analysis, the use of material culture to assign temporal parameters to earthwork strata is confined to ceramic assemblages. However, as many of the same attributes are found throughout Palau's pottery sequence, relative dating depends on the characterization of large assemblages to identify the predominant attributes. It is these characteristics that can used to insert the assemblage into the pottery seriation. Future analysis of large, dated assemblages will undoubtedly identify additional conspicuous diagnostic markers, specifically those relating to decorative techniques and styles.

Monumental architecture

Monumental architecture is correlated with increasing sociopolitical complexity, which, in the Pacific, relates to the development of hierarchical chiefdoms (Clark and Martinsson-Wallin 2007; Kirch 1990; Kolb 1994). In Palau, the scale of some structures and the extent of modified terrain attests to a strong ruling body able to organize and command its work force to excavate, transport, and shape large quantities of construction material. The elaboration and magnitude of earth structures far beyond that needed for ordinary activities symbolically expressed the differences in rank, status, and power between chiefs and districts (Liston 1999: 351-3; Liston and Tuggle 1998, 2001, 2006). Growing stratification in the Earthwork Era is also manifested in complex burials and ceremonial or elite centers.

Initial sculpting of ridgelines and hillslopes did not require a significant investment of labor and was not visually impressive. Their construction contributed to the cultural landscape that accrued over several centuries to culminate in extensive clusters of conjoining earthworks. Ngaraard's earthwork district spanned ca. 7.5 [km.sup.2] by ca. 1850 calBP. The potentially twice as large Aimeliik and Ngaremlengui clusters have yet to be dated. Construction of crowns, the most imposing single earthwork component, began by at least 2200 calBP with the leveling of hilltops. Their elevation into towering structures may not have begun until ca. 2000 to 1700 calBP. Burial complexes are dated to the same period or a little earlier while the elite or ceremonial centre on Tabelmeduu ridge was established at ca. 2200 calBE

Monumental architecture appears on Palau in the Middle Phase of the Earthwork Era, between 2000 and 1700 calBP, and continues until the final years of the Late Phase at 1200-1100 calBP. During this latter period, strife may have resulted in a locus on monumental construction with little district expansion. Earth structures of the Transitional Era are more diminutive than their predecessors, congregate in notably smaller clusters, and, within the larger districts, are associated with re-use of the abandoned earth features. The step-terraces serving as Stonework Era structural foundations are not monumental in size or extent. Monumental architecture was re-established in the form of extensive and elaborate stonework features at ca. 700 BP. The stonework, like the earthworks of the previous era, became a symbolic expression of status and power (Liston 1999; Liston and Tuggle 2001, 2006).

The Pacific is replete with massive stone and earth fortifications (Best 1993; Field 1998, Irwin 1994; Parry 1977), tombs (Kirch 1990), boundary markers (Shepardson 2005), elite or ceremonial platforms (Burley 1996; Kolb 1994; Wallin 1993), and vast agricultural landscapes (Kuhklen 2002: Sand 1999). These structures, after several centuries of development, attained monumental proportions between 950-650 BP or a little later (Clark and Martinsson-Wallin 2007; Kirch 2000; Smith 2004). The only monumental feature (or extensively modified landscape) of similar age to those on Palau is a megalithic fortification on Mard Island that is dated to ca. 1800-1700 calBP (Sand and Ouetcho 1993 in Sand 1996: 41). Construction of Palau's monumental earthworks begins 700 to 800 years earlier and ends several centuries before monumentality emerges in most of the remainder of the Pacific. Palau's second phase of monumental construction coincides with the advent of monumentality on many other island groups.

Discussion

Radiocarbon determinations, paleoecological evidence, and the material cultural data present a remarkably consistent chronology of Babeldaob's inland landscape modification. The majority of the radiometric evidence tracks the development of the Ngaraard and Ngatpang districts whose evolution is relatively consistent. As it is unlikely that all the earthwork districts were established simultaneously, developed in tandem, or fell concurrently, this chronology will be revised as additional investigations are conducted.

After colonization, population growth and commensurate pressure on a limited coastal margin led to a growing use of the interior for swidden cultivation and habitation. Earthworks may have been constructed in the last centuries of this Expansion Era (ca. 3200-2400 calBP) inland occupation.

Palau's colonizers probably favored the extensive shorelines bordering the deep inlets of Ngeremeduu, Ngerchemiangl, and Airai Bays--optimal areas for defense, storm protection, and accessing plentiful and diverse resources. Rapid population growth in these settlements could have resulted in their expansion to the adjacent slopes earlier than settlements in less strategically placed locations. Archaeological data from Babeldaob's embayments is limited to Ngeremeduu where step-terrace construction could have begun at ca. 2550 calBP atop hamlets dated to ca. 2950 calBP. Alternatively, the advantages of embayment occupation may have delayed the shift to the interior while settlers in marginal areas with little or no habitable coastline were forced inland earlier in the sequence. If so, Ngaraard and Ngarchelong--located on the island's narrow neck, could be the first in an overlapping succession of earthwork districts, while Aimeliik and Ngaremlengui--bordering the bays, are the final manifestations of the powerful districts.

At ca. 2400 calBP, an increase in inland assays and the initiation of mass wasting events suggests incipient earthwork construction in the already occupied interior. In the Earthwork Era (ca. 2400-1200 calBP), the settlement pattern shifted to a focus on interior occupation and a reliance on mixed arboriculture and dryland crops. Where viable, habitation and cultivation of the coastal margin continued. The Earthwork Era is archaeologically characterized by extensive clusters of earth architecture that supported the majority of community activities and defined sociopolitical districts. This approximately 1200-year long era is divided into three loosely bounded phases. The Early Phase (ca. 2400-2150 calBP) refers to a period of growth and development of a variety of earthwork shapes and functions. The Middle Phase (ca. 2150-1500 calBP) saw the advent of monumental architecture in the form of massive earth structures and large expanses of modified terrain. The Late Phase (ca. 1500-1200 calBP) marks the decline of the powerful interior districts accompanied by the comparatively minor construction of inland earthworks beyond the old district boundaries.

Occupation of Babeldaob's interior decreased with the fall of the Earthwork Era districts. Construction and use of earth architecture continued in a considerably diminished and somewhat altered capacity through the Transitional Era (ca. 1200-700 calBP). Some agricultural activities occurred in the essentially abandoned large districts while small groups of low crowns and step-terraces were constructed outside of the larger clusters. In the Stonework Era (ca. 700-150 calBP), the population returned to a coastal settlement pattern and a subsistence economy based on pondfield cultivation. Step-terraces functioned as structural foundations for stonework village features and associated dry-land cultivation. Ancient terraces flanking the near-coastal polities were integrated into village defenses and ritual activity.

Palau's cultural sequence models are largely defined by changing settlement patterns that are directly related to the nature of the agricultural base and to transformations in Babeldaob's landscape. Expansion into the interior soon after colonization is thought to be a result of a constricted, habitable shoreline. Pressure on the sparse arable land by the growing population and the resultant group rivalry provided the impetus for the shift to a focus on inland occupation with a reliance on dryland cultivation. The morphology and distributional patterning of the inland earthworks show an increasing attention to boundary definition and defence while the monumental scale of some structures and the extent of modified terrain suggest the development of sociopolitical complexity.

The move back to the coast is attributed to inland soil degradation and the expansion of the coastal fats, fresh-water wetlands, and mangrove forests due to coastal sedimentation. The coastal margin had become suitable for pondfield cultivation and provided defensive barricades and ample space for settlement. Warfare and concerns with defence persisted, and are archaeologically manifested in the hierarchical patterning and magnitude of the stonework architecture.

Despite the persuasiveness for these mechanisms as underlying the transformations in Palau's settlement pattern and subsistence economy, they are not entirely in accord with the chronological model. The erosion and alluviation generated by over a millennium of earthmoving would have expanded the shoreline and formed highly productive wetlands long before the archaeological record indicates abandonment of the interior districts. However, it may have taken a considerable length of time for these soils to build up enough to be secure from salt water invasion and thus become suitable for agricultural production. Babeldaob's oxisols rapidly become infertile, are almost impossible to regenerate once depleted of nutrients, and are prone to erosion. Reliance on the long-term sustainability of these soils for subsistence needs seems unlikely. Even if augmented by organics, lime, seaweeds, and other additives, and allowing for long fallows, crop rotation, forest crop production, or mixed gardening, the oxisols would loose their productivity in a few hundred years at most. If primarily motivated by soil degradation and coastal sedimentation, the population could have moved out of the interior far earlier than the end of the Earthwork Era. A sharp decline in interior assays in Ngaraard and Ngiwal at ca. 1500 BP indicates this may be the case in at least parts of Babeldaob.

The more visually impressive earthwork clusters symbolize high-status, political power, and stability. They may have continued to grow and expand after the demise of the relatively smaller, less striking clusters. Aimeliik and Ngaremlengui are Babeldaob's most extensive earthwork districts and display the greatest number of elaborately constructed and monumental complexes. A substantial corpus of radiometric determinations is needed from these two currently uninvestigated districts and the stonework villages located on lowland earthwork complexes. These assays will define the events that transpired between the Earthwork and Stonework Eras so that the processes involved in and the nature of the shift in the socioeconomic focus from the interior to the coast can be understood. Additionally, the data can make significant contributions to distinguishing temporal variations between district extent and components, distributional patterning, and historical succession.

With the wealth of archaeological and paleoenvironmental evidence produced in the Compact Road project and other recent investigations, Babeldaob's massive and extensive earth architecture can now be placed within Palau's cultural sequence. Despite the lack of data from two significant earthwork districts, the key role the earth structures played in Palau's prehistoric society and their longevity is apparent. Many issues remain to be explored, including the nature of the Expansion and Transitional Eras, the development and structure of interior political organization, the role of warfare, the pattern of population growth, and the island's carrying capacity. Other problematic areas include the impact of anthropogenic activities to Palau's island environment, the variables underlying the two dramatic transformations in settlement patterns and subsistence economies, and the relation of monumental architecture to increasing sociocomplexity in Palau and Oceania. This chronological model is meant to provide a framework for addressing these issues in future archaeological investigations of Palau's earthworks.

Acknowledgements

This paper is based largely on data produced by International Archaeological Research Institute, Inc. during the cultural resource management projects associated with construction of the Palau Compact Road by the US Army Corps of Engineers, Pacific Ocean Division (CoE). I extend my deepest gratitude to Chuck Streck, CoE Senior Archaeologist, for providing me with the opportunity to direct the data recovery (1997) and monitoring phases (1997; 2000-2004) of work and to C. Alex Morrison, CoE Resident Engineer, and Albert S. Mathis, CoE Construction Representative, for their support and assistance throughout the endeavor. I appreciate the helpful suggestions and comments given by Steve Athens, Dave Tuggle, Geoff Clark, and Peter White on an earlier version of this paper. I am beholden to Senior Archaeologist Rita Olsudong and the remainder of the Palau Bureau of Arts and Culture staff for their assistance and support of my studies of Palau's earthworks. Finally, I am greatly indebted to the people of Palau who graciously allow me to study their country's archaeological remains.

References

Anderson, A. 1991. The chronology of colonization in New Zealand. Antiquity 65: 767-795.

Athens, J.S. and J. Stevenson n.d. Holocene paleoenvironmental investigations. In: Liston, J, Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase III. Volume XII: Lab Analyses, Discussion, Syntheses. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Athens. J.S. and J.V. Ward 2001. Paleoenvironmental evidence for early human settlement in Palau: The Ngerchau core. In: Stevenson CM, Morin FJ, (eds). Pacific 2000: Proceedings of the Fifth International Conference on Easter Island and the Pacific. Los Osos: Easter Island Foundation. pp. 165-178.

Athens, J.S. and J.V. Ward 2002. Holocene Paleoenvironmental Investigations on Ngerekebesang, Koror, South Babeldaob, and Peleliu Islands, Palau. Prepared for Palau National Communications Corporation, Republic of Palau. International Archaeological Research Institute, Inc., Honolulu.

Athens, J.S. and J.V. Ward 2005. Compaet Road Archaeological Investigations. Babeldaob Island, Palau. Volume IV: The Holocene Paleoenvironment of Palau. (Drafts 1998, 1999). Prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Basilius, U. 2002. Ngchesar oral history. In: Liston, J., J. Tellei, U. Basilius and EK. Rehuher, Archaeological Survey, Monitoring. Oral History of Bore Holes, Rock Quarries, and Coral Dredging Locations for the Compact Road, Babeldaob Island, Republic of Palau. Prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. pp. 141-51.

Berger, L.R., S.E. Churchill, B. De Klerk, and R.L. Quinn 2008. Small-bodied humans from Palau, Micronesia. PloS ONE 3(3): e1780 doi: 10.1371/journal.pone.0001780.

Beardsley, F. 1997. Fishponds, Taro Patches and Shell Middens: Archaeological Investigations on Peleliu, Republic of Palau, Data Recovery and Monitoring for the Palau Rural Water System Program. Prepared for Winzler and Kelly, Consulting Engineers, Republic of Palau. International Archaeological Research Institute, Inc., Honolulu.

Best. S.B. 1993. At the halls of the mountain kings. Fijian and Samoan fortifications: Comparison and analysis. Journal of the Polynesian Society 102: 385-447.

Bronk Ramsey, C. 2005. OxCal Program v 3.10. Oxford: University of Oxford Radiocarbon Unit. http://www.rlaha.ox.ac.uk/oxcal/oxcal.htm

Burley, D.Y. 1996. Sports, status and field monuments in the Polynesian chiefdom of Tonga: The pigeon snaring mounds of northern Ha'apai. Journal of Field Archaeology 23:421-435.

Butler, B.M. 1984. A Preliminary Report on the 1983 Archaeological Survey of Aimeliik State, Republic of Palau. Prepared for Belau National Museum, Inc., Ministry of Social Services Division of Cultural Affairs, Republic of Palau. Center for Archaeological Investigations, Southern Illinois University, Carbondale.

Carucci, J. 1992. Cultural and Natural Patterning in Prehistoric Marine Foodshell from Palau, Micronesia. PhD thesis. Department of Anthropology, Southern Illinois University, Carbondale.

Clark, G.R. 2005. A 3000-year culture sequence from Palau. Western Micronesia. Asian Perspectives 44(2): 349-380.

Clark, G.R., A. Anderson, and D. Wright 2006. Human colonization of the Palau islands, Western Micronesia. Journal of Island and Coastal Archaeology 1: 215-232.

Clark, G.R. and H. Martinsson-Wallin 2007. Monumental architecture in West Polynesia: Origins, chiefs, and archaeological approaches. Archaeology in Oceania 42 Supplement:28-40.

Clark, G.R. and D. Wright 2007. Reading Pacific pots. In: Anderson A, K. Green, F. Leach (eds), Vastly Ingenious: The Archaeology of Pacific Material Culture in Honour of Janet M. Davidson. Otago University Press, New Zealand. pp. 173-189.

Cochrane, E.E. and H.D. Tuggle n.d. Traditional pottery analysis. In: Tuggle, HD, TM Mangieri, and J Liston, Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase IlL Volume IX: Planned Data Recovery Field Reports. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Cordy, R. 1984. Archaeological Survey in Ngerekeai Village, Imeliik, Palau Islands. Prepared for Trust Territory Historic Preservation Plan.

Desilets, M., J. Liston, and H.D. Tuggle 1999. Ceramic analysis. In: Liston, J, Archaeological Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase II. Volume V: Lab Analyses, Syntheses, and Recommendations. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. pp. 181-231.

Dickinson, W.R. and S.J. Athens 2007. Holocene paleoshoreline and paleoenvironmental history of Palau: Implications for human settlement. Journal of Island and Coastal Archaeology 2(2): 175-196.

Field, J.S. 1998. Natural and constructed defenses in Fijian fortifications. Asian Perspectives 37: 32-58.

Fitzpatrick, S.M. 2003. Early human burials in the Western Pacific: Evidence for a c.3000 year old occupation on Palau. Antiquity 77(298): 719-731.

Fitzpatrick, S.M. and O. Kataoka 2005. Prehistoric fishing in Palau, Micronesia: Evidence from the northern Rock Islands. Archaeology in Oceania 40: 1-13.

Haslam, M. and J. Liston 2008. Use of flaked stone artifacts in Palau, Western Micronesia. Asian Perspectives. 47(2).

Henry, J., A.E. Haun and M.A. Kirkendall 1996. Archaeological Mitigation Program: Palau Rural Water System Projects in Chol and Ngkeklau Villages, Ngaraard State, Republic of Palau. Draft prepared for Monterey Mechanical. Paul H. Rosendahl, PhD, Inc. Hilo, Hawai'i.

Hijikata, H. 1993. Collective Works of Hijikata Hisakatsu. Society and Life in Palau: Volume I. Translated and edited by Hisashi Endo. The Sasakawa Peace Foundation, Tokyo. Originally published 1940, Kagaku Nanyo.

Hijikata, H. 1995. Collective Works of Hijikata Hisakatsu: Gods and Religion in Palau, Volume H. Translated and edited by Hisashi Endo. The Sasakawa Peace Foundation, Tokyo. Originally published 1940, Kagaku Nanyo.

Hunter-Anderson, R. 1991. A Review of Traditional Micronesian High Island Horticulture in Belau, Yap, Chuuk, Pohnpei, and Kosrae. Micronesica 24(1): 1-56.

Irwin, G. 1978. Land, pa and polity. New Zealand Archaeological Association Monograph 15.

Kaschko, M.W. 1998. Fortified hilltops, black pottery, and early settlement in Palau. Paper presented at the 63rd Meeting of the Society for American Archaeology, Seattle, Washington, March 25-29

Kirch, P.V. 1990. Monumental architecture and power in Polynesian chiefdoms: A comparison of Tonga and Hawaii. World Archaeology 22: 206-222.

Kirch, P.V. 2000. On the Road of The Winds: An Archaeological History of the Pacific Islands before European Contact. University of California Press, Berkeley.

Kolb, M.J. 1994. Monumentality and the rise of religious authority in precontact Hawai'i. Current Anthropology, Vol. 35(5): 521-547.

Kuhlken, R. 2002. Intensive agricultural landscapes of Oceania. Journal of Cultural Geography, 19: 161-195.

Kramer, A. 1926. In: Thilenius, G. (ed), Ergebnisse der Sudsee-Expedition 1908-1910. II. Ethnographie; B. Mikronesien. Band 3, Teilband 3. Friederichsen, Hamburg (Human Relations Area File typescript, translated by Anonymous).

Kramer, A. 1929. In: Thilenius, G. (ed), Ergebnisse der Sudsee-Expedition 1908-1910. II. Ethnographie; B. Mikronesien. Band 3, Teilband 4. Friederichsen, Hamburg (Human Relations Area File typescript, translated by Anonymous).

Liston, J. 1999. Archaeological Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase II. Volume V: Lab Analyses, Syntheses, Recommendations. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Liston, J. 2005. An assessment of radiocarbon dates from Palau, Western Micronesia. Radiocarbon 47: 295-354.

Liston, J. 2007. Archaeological Data Recovery Investigations at Site IR-1:14, Omsangel Hill, Airai State, Republic of Palau. Prepared for TRB Architects, Republic of Palau. Garcia and Associates, Kailua, Hawai'i.

Liston, J. 2008a. Archaeological Data Recovery at Tabehneduu, Ngaraard Earthwork District, Republic of Palau. Prepared for Ngaraard State, Republic of Palau. Garcia and Associates, Kailua, Hawai'i.

Liston, J. 2008b. Excavation in a Crown Enclosure at Site NI-2:35, Ngiwal, Republic of Palau. Manuscript on file at the Bureau of Arts and Culture, Republic of Palau.

Liston, J. n.d.a. Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase III. Volume X: Monitoring Field Reports. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. (submitted May, 2007)

Liston, J. n.d.b. Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase III. Volume XI: Monitoring Field Reports. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. (submitted May, 2007)

Liston, J. n.d.c. Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase III. Volume XII: Lab Analyses, Discussion, Syntheses. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Liston, J., M.W. Kaschko, and D.J. Welch 1998. Archaeological Inventory Survey of the Capitol Relocation Project, Melekeok, Republic of Palau. Prepared for Architects Hawai'i, Inc., Honolulu. International Archaeological Research Institute, Inc., Honolulu.

Liston, J. and R. Olsudong in prep. Distribution of anthropogenic earthworks on Babeldaob, Republic of Palau. To be submitted to Micronesica.

Liston, J., T.M. Rieth, and H.D. Tuggle in prep. Traditional pottery analysis. In: Liston, J, Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase III Volume XII: Lab Analyses, Discussion, Syntheses. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Liston, J. and H.D. Tuggle 1998. The terraces of Palau: New information on function and age. Paper presented at the 63rd Meeting of the Society for American Archaeology, Seattle, Washington, March 25-29; and at the Society of Hawaiian Archaeology Annual Conference, Kona, Hawaii.

Liston. J. and H.D. Tuggle 2001. Prehistoric warfare in Palau. Paper presented at The Society for American Archaeology 66th Annual Meeting, New Orleans, Louisiana.

Liston, J. and H.D. Tuggle 2006. Prehistoric warfare in Palau. In: Arkush, E. and M.W. Allen (eds), The Archaeology of Walfare: Prehistories of Raiding and Conquest. Gainesville: University Press of Florida. pp. 148-183. Lucking, L.J. 1984. An Archaeological Investigation of Prehistoric Palauan Terraces. PhD dissertation, University of Minnesota, Minneapolis.

Lucking, L.J. and R.J. Parmentier 1990. Terraces and traditions of Uluang: Ethnographic and archaeological perspectives on a prehistoric Belauan site. Micronesica Supplement 2:125-136.

Magnuson, Coral M. and Jolie Liston 1998. Archaeological lnventory Survey Palau Peninsula Resort, Ngerekebesang Island, Korot; Republic of Palau. Prepared for Haas and Haynie, San Francisco. California. International Archaeological Research Institute, Inc., Honolulu.

Mangieri, T.M. 1998a. Ngaraard field reports. In: Liston, J, H.D. Tuggle, T.M. Mangieri. MW Kaschko, and M Desilets, Archaeological Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations Phase II. Volume I: Fieldwork Reports. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. pp. 212-314.

Mangieri, T.M. 1998b. Ngatpang field reports. In: Liston J, T.M. Mangieri, D. Grant. M.W. Kaschko, and H.D. Tuggle, Compact Road Archaeological Investigations. Babeldaob Island, Republic of Palau, Historic Preservation Investigations Phase II. Volume II: Fieldwork Reports. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. pp. 163-236.

Masse, W.B. 1989. The Archaeology and Ecology of Fishing in the Belaeu Islands, Micronesia. PhD Dissertation, Department of Anthropology, Southern Illinois University, Carbondale.

Masse, W.B., J. Liston, J.Carucci, and J.S. Athens 2006. Evaluating the effects of climate change on environrnent, resource depletion, and culture in the Palau Islands between AD 1200 and 1600. Quaternary International 151 : 106-132.

Masse, W.B.. D. Snyder, G.J. Gumerman 1984. Prehistoric and historic settlement in the Palau Islands, Micronesia, New Zealand Journal of Archaeology 6: 107-127.

Morgan, W.N. 1988. Prehistoric Architecture in Micronesia. University of Texas Press, Austin.

Murakami, G. 1999. Identification of Charcoal Taxa. In: Liston, J., Archaeological Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase II. Vohone V: Lab Analyses, Syntheses, Recommendations. Draft prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. pp. 126-140.

Murakami, G. 2007. Identification of Charcoal Taxa, Appendix A. In: Wickler, S.K., D.J. Welch, M.J. Tomonari-Tuggle, J. Liston, and H.D. Tuggle, Intensive Archaeological Survey For the Palau Compact Road, Babeldaob Island, Palau. Historic Preservation Investigations Phase I, Volume I: Scope, Background, Results, Evaluation, and Recommendations. (Draft 1998). Prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. pp. 185-201.

Murakami, G. n.d. Identification of charcoal taxa, In: Liston. J., Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase III. Volume XII: Lab Analyses, Discussion, Syntheses. Draft prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Olsudong, R., C.T. Emesiochel, and R.T. Kloulechad 2008. Sah,age Archaeology at Site ME-4:3, Melekeok, Republic of Palau. Manuscript on file at the Bureau of Arts and Culture, Republic of Palau.

Osborne, D. 1966. The Archaeology of the Palau Islands, an Intensive Survey. Bernice E Bishop Museum Bulletin 230. Bishop Museum Press, Honolulu.

Osborne, D. 1979. Archaeological test excavations, Palau Islands, 1968-1969. Micronesia Supplement 1.

Parmentier, R.J. 1987. The Sacred Remains: Myth, History, and Polity in Belau. The University of Chicago Press, Chicago.

Parry, J. 1977. Ring-ditch Fortifications: Ring-ditch Fortifications in the Rewa delta, Fiji: Air Photo Interpretation and Analysis. Bulletin of the Fiji Museum 3. The Fiji Museum, Suva.

Phear, S. 2007. The Monumental Earthworks of Palau, Micronesia: A Landscape Perspective. BAR International Series 1626.

Phear, S. 2008. Subsistence and island landscape transformations: Investigating monumental earthworks in Ngaraard State, Republic of Palau, Micronesia. In: Clark, G., F. Leach and S. O'Connor (eds), Terra Australis 29--Islands of Inquiry: Colonisation, Seafaring and the Archaeology of Maritime Landscapes. Australia National University E Press, Canberra.

Phear, S., G.R. Clark and A. Anderson 2003. A radiocarbon chronology of Palau, In: Sand, C. (ed), Pacific Archaeology: Assessments and Prospects. Les Cahiers de l'Archeologie en Nouvelle-Caledonie 15. New Caledonia. pp. 255-263.

Pregill, G.K. and D.W. Steadman 2000. Fossil vertebrates from Palau, Micronesia: a resource assessment. Micronesica 33(1/2): 137-152.

Reimer, P.J., M.G.L. Baillie, E. Bard, A. Bayliss, J.W. Beck, C.J.H. Bertrand, P.G. Blackwell, C.E. Buck, G.S. Burr, K.B. Cutler. P.E. Damon, R.L. Edwards, R.G. Fairbanks, M. Friedrich, T.P. Guilderson, A.G. Hogg, K.A. Hughen, B. Kromer, G. McCormac, S. Manning, C. Bronk Ramsey, R.W. Reimer, S. Remmele, J.R. Southon, M. Stuiver, S. Talamo, EW. Taylor, J. van der Plicht, and C.E. Weyhenmeyer 2004 IntCal04 terrestrial radiocarbon age calibration, 0-26 kyr BP. Radiocarbon 46: 1029-1958.

Rieth, T.M. and J. Liston 2001. Archaeological Data Recovery at Ngermereues Ridge, Ngesaol, Koror: Republic of Palau. Prepared for Winzler & Kelly. Consulting Engineers, Republic of Palau. International Archaeological Research Institute, Inc., Honolulu.

Rieth, T.M. and J. Liston n.d. Olbalkeldrau petroglyph. In: Liston, J, Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island. Republic of Palau. Historic Preservation Investigations, Phase III. Volume X: Monitoring Field Reports. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division. Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Sand, C. 1996. Structural remains as markers of complex societies in southern Melanesia during prehistory: the case of the monumental forts of Mare Island (New Caledonia). Indo-Pacific Prehistory Association Bulletin 15, Chiang Mai Papers 2: 37-44.

Sand. C. 1999. From the swamp to the terrace: Intensification of the horticultural practices in New Caledonia from first settlement to European contact. In: Gosden, C. and J. Hather (eds), The Prehistory of Food. One World Archaeology Series. pp. 252-269.

Sand, C. and A. Ouetcho 1993. Etude d'impact de la transversale Kone-Tiwaka sur le patrimoine archeologique. Les Cahiers de l'Archeologie en Nouvelle-Caledonie. New Calidonia Museum Editions, 3:5-55.

Shepardson. B.L. 2005. The role of Rapa Nu'i (Easter Island) statuary as territorial boundary markers. Antiquity 79:169-178

Smith, A. 2004. Are the earliest field monuments of the Pacific landscape serial sites'? In: Attenbrow, V. and R. Fullagar, A Pacific Odyssey." Archaeology and anthropology in the Western Pacific, Papers in Honor of Jim Specht. Records of the Australia Museum, Supplement 29. pp. 133-138.

Snyder. D. 1989. Towards Chronometric Models for Palaun Prehistory: Ceramic Attributes. Unpublished PhD Dissertation, Department of Anthropology, Southern Illinois University, Carbondale.

Snyder, D. 1997. Two millennium of change: A model of settlement system dynamics in Palau, Micronesia. Paper presented at the 62nd Meeting for the Society for American Archaeology, Nashville, Tennessee, April 2-6, 1997.

Spriggs, M. and A. Anderson 1993. Late colonization of East Polynesia. Antiquity, 67(255): 200-217.

Tellei, J., U. Basilius, and F.K. Rehuher 2005. Palau Compact Road Archaeological Investigations, Babeldaob Island, Republic of Palau, Phase I: Intensive Archaeological Survey. Volume III: Oral History Documentation. Prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Tuggle, H.D. 1998. Ngarchelong and Ngaraard field reports. In: Liston, J, H.D. Tuggle, T.M. Mangieri, M.W. Kaschko, and M. Desilets, Archaeological Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations Phase II. Volume I: Fieldwork Reports. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. pp. 81-129, 150-197, 314-424.

Tuggle H.D.n.d. Ngaraard ridgeline fieldwork. In: Tuggle, H.D., T.M. Mangieri, and J. Liston, Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations. Phase III. Volume IX: Planned Data Recovery Field Reports. Draft report prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Van Tilburg, J.A. 1991. Anthropomorphic stone monoliths on the islands of Oreor and Babeldaob, Republic of Belau (Palau), Micronesia. Bishop Museum Occasional Papers 31 : 3-62.

Wallin, P. 1993. Ceremonial Stone Structures. The Archaeology and Ethnohistory of the Marae Complex in the Society Islands. French Polynesia. Aun. 18. Societas Archaologica Upsaliensis: Uppsala.

Whitesell, C.D., C.D. MacLean, M.C. Falanruw, T.G. Cole, and A.H. Ambacher 1987. Vegetation Survey of the Republic of Palau. Pacific Southwest Forest and Range Experiment Station, Berkeley, California, in cooperation with the Pacific Northwest Forest and Range Experiment Station, Portland, Oregon. Forest Service, U.S. Department of Agriculture.

Wickler, S.K. 2002. Terraces and villages: Transformation of the cultural landscape in Palau. In T. Ladefoged and M. Graves (eds), Pacific Landscapes: Archaeological Approaches in Oceania. Easter Island Foundation, Los Osos: Bearsville Press. pp. 63-96.

Wickler, S.K. 2005. Survey area B report, part II. In: Wickler, S.K., D.J. Addison, M.W. Kaschko, T.S. Dye, Intensive Archaeological Survey for the Palau Compact Road, Babeldaob Island. Palau. Volume II: Area Survey Reports. (Draft 1997). Prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu. pp. 113-322.

Wickler, S.K., D.J. Welch, M.J. Tomonari-Tuggle, J. Liston, and H.D. Tuggle 2007. Intensive Archaeological Survey For the Palau Compact Road, Babeldaob Island. Palau. Historic Preservation Investigations Phase I, Volume I: Scope, Background, Results, Evaluation. and Recommendations. (Draft 1998). Prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Williams, S.S. and J.A. Markos 1999. Lithic and shell tools. In: Liston, J, Archaeological Data Recovery for the Compact Road. Babeldaob Island, Republic of Palau. Historic Preservation Investigations Phase II. Volume V: Lab Analyses, Syntheses, and Recommendations. Draft prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute Inc., Honolulu. pp. 90-105.

Williams, S.S. and M.C. Pope n.d. Lithic and shell tools. In: Liston, J., Archaeological Monitoring and Emergency Data Recovery for the Compact Road, Babeldaob Island, Republic of Palau. Historic Preservation Investigations, Phase III. Volume XII: Lab Analyses, Discussion, Syntheses. Draft prepared for U.S. Army Corps of Engineers, Pacific Ocean Division, Hawai'i. International Archaeological Research Institute, Inc., Honolulu.

Wright, D. 2005. The Archaeology of Aulong Island and the Colonisation of Palau. MA thesis, Canberra: Australia National University.

Yamata, I. 1930. Ancient village sites of Palau Island. Geographical Review 6(10). Bernice P. Bishop Museum.

JOLIE LISTON

Garcia and Associates, 146 Hekili Street, Suite 101, Kailua,

Hawaii 96734, USA. uli624@yahoo.com
Table 1. Distribution of inland radiocarbon dates by time period and
district.

                Expansion Era *                     Earthwork Era
                (ca. 3200-2400 BP)               (ca. 2400-1200 BP)

Earthwork                              Early           Middle
Districts                            (ca.2400-       (ca.2150-
                                      2150 BP)        1500 BP)
Aimeluk
Airai                                                   3/2/1
Melekeok                               2/1/1            6/3/1
Ngaraard             2/2/1             2/2/1          41/10/1
Ngarchelong                                             2/2/1
Ngaremlengui
Ngatpang             5/1/1             2/1/1            8/2/1
Ngchesar
Ngerekebesang
  (Oreor)
Ngiwal               1/1/1             2/2/1           17/3/1
TOTAL:               8/4/3             8/6/4          77/22/6

                                 Transitional Era    Stonework Era
                                (ca. 1200-700 BP)   (ca. 700-150 BP)

Earthwork             Late
Districts           (ca.1500-
                    1200 BP)
Aimeliik
Airai
Melekeok
Ngaraard             20/5/2         6/5/2                2/2/2
Ngarchelong
Ngaremlengui
Ngatpang             5/2/1
Ngchesar             1/1/1
Ngerekebesang
  (Oreor)                           1/1/1
Ngiwal               2/2/1          1/1/1
TOTAL:               28/10/5        8/7/4                2/2/1

                     Total

Earthwork
Districts

Aimeluk
Airai                 3/2/1
Melekeok              8/3/1
Ngaraard             73/15/2
Ngarchelong            2/2/1
Ngaremlengui
Ngatpang              20/2/1
Ngchesar               1/1/1
Ngerekebesang
  (Oreor)              1/1/1
Ngiwal                23/5/1
TOTAL:               131/31/9

* Number of: dates/sites/earthwork districts

Table 2. Palau's cultural sequence emphasizing the Earthwork Era.

Time period    Date calBP   Paleoenvironment      Earthworks

Expansion      ca. 3200-    savanna expansion;    none recorded:
Era            2400         forest decline        potentially step-
                                                  terrace construction
                                                  on coastal slopes at
                                                  end of era

Earthwork      ca. 3400-    savanna expansion;    growth, zenith.
Era            1200         forest decline;       decline of large,
                            coastal interior      earthwork
                            sedimentation         districts; primary
                            generated by inland   function -
                            earthmoving           symbolize power of
                                                  chiefs and districts;
                                                  components
                                                  multifunctional -
                                                  boundary markers,
                                                  horticulture,
                                                  irrigation,
                                                  habitation,
                                                  defence, ritual,
                                                  burials, trails,
                                                  drainage

-Early         ca. 2400-    largely savanna       earth construction
               2150         landscape; abrupt     begins, particularly
                            transitions from      step-terraces,
                            wetland to upland     modified ridgelines,
                            sediments             hills and earth
                                                  platforms

-Middle        ca. 2150-    continuation of       expansive earth-
               1500         savanna               works clusters;
                                                  large structures;
                                                  hilltops elevated
                                                  into crowns; zenith
                                                  and abandonment
                                                  of northeast
                                                  Babeldaob's
                                                  districts

- Late         ca. 1500-    continuation of       minor construction
               1200         savanna               of inland earthworks
                                                  beyond older large
                                                  district boundaries;
                                                  crown construction
                                                  intensifies decline
                                                  of large interior
                                                  districts

Transitional   ca. 1200-    forest expansion;     limited use of
Era            700          savanna decline       largely abandoned
                                                  once powerful
                                                  districts; small
                                                  groups of low
                                                  crowns and step-
                                                  terraces constructed
                                                  outside of the older
                                                  large districts

Stonework      ca. 1200-    forest expansion;     step-terraces as
Era            700          savanna decline       structural
                                                  foundations for
                                                  coastal villages and
                                                  associated dryland
                                                  fields; ancient
                                                  terraces flanking
                                                  coastal settlements
                                                  integrated into
                                                  coastal village
                                                  defences and ritual
                                                  activity; interior
                                                  used for paths,
                                                  communication,
                                                  lookouts, resource
                                                  procurement, ritual

Time period    Settlement pattern      Subsistence
                                       economy

Expansion      coastal settlement;     swidden and
Era            inland use for          wetland
               horticulture,           cultivation
               resource
               procurement,
               transportation, and
               various lookout.
               communication, and
               ceremonial capacities

Earthwork      intensive interior      likely mixed
Era            settlement; coastal     dryland
               settlement and          cultivation
               cultivation in viable   arboriculture;
               locations: Rock         some wetland
               Islands semi-           fields;
               permanently then        breadfruit
               permanently             significant?
               occupied

-Early         shift to focus on       shifting
               interior settlement;    cultivation;
               earthworks; Rock        mixed
               Islands semi-           gardening?
               permanently
               inhabited

-Middle        sociopolitical          irrigation of
               districts extend from   interior fields;
               coast to central        some coastal
               ridge; defined by       pondfields
               earthworks clusters;
               separated by
               unshaped terrain;
               settlement areas
               widely dispersed in
               districts; Rock Island
               and likely coastal
               occupation

- Late         gradual movement        inland
               out of interior         cultivation
               focused districts;      continues
               Rock Island and         after district
               likely coastal          decline
               occupation

Transitional   sharp drop in           some interior
Era            interior occupation;    cultivation in
               little documented       irrigated
               activity on             fields;
               Babeldaob's coast;      potential
               settlements on Rock     overharvesting
               Islands and Peleliu;    of some
               villages on slopes      inshore marine
               of lowland              resources
               earthworks?

Stonework      compact complexes       taro cultivated
Era            of coastal villages     in pondfield
               behind a defensive      complexes;
               barrier of mangrove     some dryland
               forests or              agriculture
               swamplands and
               bounded by wet- and
               dryland agricultural
               fields;
               Rock Island villages
               abandoned pre-150
               BP

Time period    Sociocultural            Material culture *

Expansion      population growth;       pottery = generally
Era            limestone cave           thin bodies ([less
               burials                  than or equal to] 5 mm),
                                        sand temper, black
                                        paste; coiled, footed
                                        base

Earthwork      extensive clusters of    pottery = most grog
Era            inland earth             temper;
               and architecture form    carbonaceous clay;
               fortified                reduced firing
               sociopolitical           atmosphere; bowls -
               districts; warfare;      ovah round, shallow,
               hierarchical             globular, and
               structure; burials in    collared; straight-
               Rock Island caves,       sided and interior-
               step-terraces, and       thickened rims; few
               crowns                   simple geometric
                                        designs - paint,
                                        incising, impressions;
                                        average body
                                        thickness 7.5-10.5
                                        mm; whole pots in
                                        ritual contexts.
                                        abundant flake stone
                                        tools: basalt adzes;
                                        shell, bone (likely
                                        wood) tools present-

-Early         elite or ceremonial      bowls - shallow and
               centres -pottery         globular;
               caches/whole vessels     backcurving and
               under/in construction    straight-sided rims;
               fill; limestone cave     red and brown slip;
               burials                  burnishing; nubbins

-Middle        ritual activity -        slip and burnishing
               structured burial sets   through 1 st half of
               w/ burial furniture;     phase; straight-sided
               warfare; monumental      and interior
               architecture in form     thickened rims; grog
               of massive and           temper; paired small
               extensive earthworks     piercings close to
                                        rim; bands of reddish
                                        paint

- Late         chiefly interments       tendency toward
               in crowns; rise in       thicker vessel walls;
               conflict identified in   slightly more
               ring-ditch palisades     frequent incising
               and larger
               defensive/symbolic
               features

Transitional   decline in limestone     bowls predominate;
Era            cave burials; little     potential incised
               evidence for             human figures;
               cultural activity        thickened & flanged
               throughout               rims begin; greater
               archipelago              reliance on bone,
                                        wood, and shell
                                        tools?

Stonework      monumental               oxidized vessels;
Era            architecture in form     deep, straight-sided
               of massive and           jars, open bowls,
               elaborate stone          plates, dishes;
               architecture;            addition of some
               warfare; complex         white, gray, or pink
               hierarchical             colored paste;
               structure                thickened & flanged
                                        rims; decoration
                                        scarce - bands of red
                                        paint; greater reliance
                                        on bone, wood, shell
                                        tools with less stone
                                        implements

* Ceramic attributes highly varied throughout the sequence.
COPYRIGHT 2009 Blackwell Publishing Limited, a company of John Wiley & Sons, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2009 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Liston, Jolie
Publication:Archaeology in Oceania
Geographic Code:8PALA
Date:Jul 1, 2009
Words:14145
Previous Article:The meaning and importance of the Lapita face motif.
Next Article:Legacy of an Ice Age: foreword.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters