Printer Friendly

Diversity in lithic raw material sources on New Britain, Papua New Guinea.

Abstract

The nature of chert exposures in the Passismanua area of West New Britain, Papua New Guinea is reviewed in light of reports of worked seams of chert in five caves. Extraction of chert at one cave, Ale, began within the last 3,000 years, but such exposures have been used from the late Pleistocene onwards. The nature and quality of the exposures would often have placed severe constraints on the production of flaked tools. The chert sources are compared with those of obsidian on the north side of New Britain, highlighting the potential advantages and problems of each. A small group of finely made stemmed chert tools is identified as potentially valuables similar to stemmed obsidian tools of the Willaumez Peninsula obsidian source region. While the chert examples differ in aspects of technology and form, they share with the obsidian forms the concept of bifacially-worked stems and were made during the same period. This is seen as indicating social relationships between the two areas during the middle Holocene comparable to that recently proposed between Manus and New Britain.

Keywords: obsidian, New Britain, chert, stemmed tools, mid-Holocene, sources

**********

The nature of lithic raw material sources can have a significant influence on the production, use and circulation of flaked stone tools (e.g. Andrefsky 1994; Bamforth 1990; Torrence 1986). In the Papua New Guinea region, studies of sources of flakeable stone have focused on exposures of obsidian in Manus and West New Britain provinces (Fullagar et al. 1991; Fullagar and Torrence 1991; Torrence et al. 1992, 2000, 2004a). As Pavlides (2004: 98, 2006: 207) has pointed out, however, secondary exposures as cobbles and boulders in the beds of water courses and on beaches were also major sources of flakeable stone in many parts of Papua New Guinea (e.g. Sheppard 1996; Evans and Mountain 2005; Ford, this issue). In this paper I am concerned with a third kind of exposure that is often less visible: that of chert in the Passismanua area of interior West New Britain (Fig. 1). The chert industries of this region have been on record for over 40 years (Chowning and Goodale 1966; Shutler and Kess 1969), and Pavlides has provided a picture of the production and use of chert tools from the late Pleistocene onwards (Pavlides 1999, 2004, 2006; Pavlides and Gosden 1994). Chert has been reported in stream beds within the Passismanua area (Crowning and Goodale 1966: 150; Goodale 1966; Pavlides 2004: 98; Bulmer 2005: fig. 2), and the use of 'bedrock' primary sources in sinkholes has been proposed on the basis of extensive deposits of 'rubble' and other debris likely to derive from extracting and processing the raw material at several locations (Pavlides 2006: 211). There are, however, no detailed accounts of chert exposures that have served as source locations (cf. Pavlides 1999: 145).

Important new data relating to primary sources now provide insights into the nature and distribution of chert exposures, and permit a better understanding of issues relating to raw material procurement, processing and use in this region. The paper presents these data in terms of the availability, accessibility and quality of these sources, and compares them with the obsidian sources on Willaumez Peninsula and at Mopir on the north side of New Britain (Specht 1981; Fullagar et al. 1991; Torrence et al. 1992). The comparisons raise a number of significant issues regarding raw material procurement and use in each area, and identify the possibility of a category of chert valuables in the middle Holocene comparable to stemmed tools of the Willaumez Peninsula region.

The chert sources--description

The Passismanua chert derives from the Yalam Limestone geological formation that covers about 7200 [km.sup.2] (16%) of the surface of New Britain. This limestone is mostly of marine origin and was formed during the Miocene (Davies 1973; Ryburn 1974, 1975: 11-12, 1976: 11). It extends in several major sections from Lamogai Plateau in the west to the Gazelle Peninsula in the east (Fig. 2). The western section includes both the Lamogai Plateau and the Passismanua area and covers almost 3,000 [km.sup.2]. In the Passismanua area and part of the Lamogai Plateau, it forms a dissected karst plateau of massive coral-algal limestone with numerous dolines and saucer-shaped depressions resulting from doline collapse. Chert occurs throughout this plateau. The Alimbit, Palicks and Andru Rivers and numerous streams provide the main surface drainage, though there is also an extensive network of underground water courses. The plateau rises from ca 200 m at its southern edge 5-10 km from the Kandrian coast to ca 1000 m on its inland margin, where it joins older volcanic formations to form the Whiteman Range (Ryburn 1975, 1976). On the north side of the Lamogai Plateau the limestone has a soft, marly facies with volcanic detritus and a range of calcareous rocks (Ryburn 1976: 5, 11). There are currently no reports of chert in this facies.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

Chert exposures occur in two forms. The beds of many streams are cutting down into the limestone and exposing nodules of chert that eventually become loose and available for collection or can be broken off. This has happened in the Sipsa stream bed at the men's washpool near Auwa hamlet, where flakes and chunks have been removed from nodules embedded in the limestone. Loose nodules and struck pieces were also observed in lower sections of the Sipsa, and suggest the exposure of nodules in multiple localities. Similar observations were made in other stream beds between Auwa and Akiuli, which lies about 8 km south from Auwa (Fig. 3). These occurrences are not to be confused with chert artefacts found in stream beds (Chowning and Goodale 1966: 150; Goodale 1966: 26-27). These represent material dislodged from the soils through which the stream beds have cut, and chert items are often visible in the stream banks. The caption to a photo of artefacts in a stream bed near Silop explicitly states that the artefacts were "embedded in the earth" (Bulmer 2005: fig. 2), indicating that either the stream bed was still cutting through soils, or stream banks containing artefacts had collapsed and the soil had not yet been washed away. Where nodules are eroding naturally from the limestone stream bed, there is usually no evidence for chert in the adjacent stream bank profiles.

[FIGURE 3 OMITTED]

The second form of chert exposure is within caves and sinkholes, where seams are exposed by erosion of the limestone walls; to date, such seams have not been noted on open-air limestone faces. The presence of such seams was proposed on the basis of the large quantity of 'rubble' at several archaeological sites. This rubble is interpreted as waste from testing chert nodules and the production of artefacts (Pavlides 2006: 211). Excavations at Auwa hamlet (site code FGT) produced particularly large quantities that are thought to come from a nearby sinkhole, though this has not been investigated because its entrance was used as a latrine and rubbish dump at the time of fieldwork. However, five examples of such primary sources can now be confirmed. The first four were located by a French-sponsored caving expedition to the Passismanua area in 2001, during which Roman Hapka (2009), a speleologist with archaeological training made observations on the presence of chert seams and artefacts in several caves. Four of these had seams of chert exposed on at least one wall: Helena cave, about 3 km northwest from Auwa; Pomalngin cave, about 4 km southwest from Auwa; Singlip cave, about 1 km northwest of Pomalngin; and an unnamed locality recorded as R1 where an underground river emerges on the surface (Fig. 3). There are no details about the seams at first three localities, but just inside the opening of R1 there is a worked seam of chert that appears to be blackish-brown in colour and slopes from water level (in the photo, about waist deep) upwards to at least one metre above the water (Fig. 4).

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

At Helena, Pomalngin and Singlip caves, the French cavers noted extensive deposits of chert debitage and artefacts including 'axes', flakes, scrapers, blades, cores and "outils divers" (Hapka 2009: table 1). Each cave appears to have been both an extraction and reduction site, though some material might be worn out tools that were replaced by new ones at the source. As the expedition was not conducting formal research, only general observations were recorded and none of the material was collected (Roman Hapka, personal communication).

The fifth example is in Ale cave (site code FKX) near Aliwo hamlet, about 5 km northeast and inland from Kandrian. I first recorded Ale as a potential archaeological site in 1979, and returned to it with the late Baiva Ivuyo in 2002. Ale is situated near the top of a small limestone rise at the southern limit of the Yalam Limestone (Fig. 3). It has a low side entrance leading into an irregular-shaped chamber that ranges from 4 m to 7 m across and is about 18 m long. The interior is dark and dank, with rainwater entering through the side entrance and from a small opening in the cave ceiling about 14 m above the floor. This opening is currently about 400 mm in diameter, and presumably with time and further erosion it will expand and the roof will collapse. Limestone boulders from earlier roof and entrance collapse partly cover the floor, which slopes steeply downwards to a tunnel at the rear of the cave. On the northern side of the main cave entrance there is a vertical shaft about 3 m deep that leads to a lower chamber through which flows a water course. Neither of these additional chambers was investigated.

Three seams of chert are exposed on the north wall of the cave, at about 4 m, 9 m and 10.5 m above the sloping cave floor (Fig. 5). Access to these seams was probably by climbing up tree branches leant against the cave wall, though footholds on the irregular surface of the cave wall also provide access to at least the lowest seam. The seams are currently about 3 m long, and range in thickness from about 50 mm to 150 mm. They present irregular, fractured surfaces that together with chert debris over the cave floor indicate human extraction activity. The limestone around the chert seams appears to have been excavated, presumably to gain access to chert concealed within it; this 'quarrying' penetrates about 250 mm.

We excavated a trench 1.5 by 1 m in an area with chert debris on the surface, about 14 m from the cave mouth and 4 m from the tunnel. The sticky clay sediments were too wet to sieve and this, combined with the darkness inside the cave, meant that bone, stone and other materials had to be located by feel and torchlight, and then taken to Kandrian for cleaning and sorting. Four main layers were recognised:

1. Surface layer of very dark brown to black clay containing charcoal, bones, chert pieces and fragments of marine shell.

2. Reworked grey-brown tephra lacking evidence of human activity.

3. Sticky, dark brown clay with charcoal, densely packed limestone rubble and pieces of chert debris and artefacts, including retouched pieces, three hammer stones of volcanic rock and one plain sherd.

4. A basal layer consisting of yellow-brown tephra, the upper part of which (4A) is clearly reworked, and the lower part (4B) is undisturbed. The tephra originated from the W-K2 eruption of 3480-3150 cal. BP (Torombe 2006; Petrie and Torrence 2008: table 5).

The excavation ended at 160 cm below ground surface on limestone bedrock, the surface of which was highly polished, presumably as a result of water flow. In view of the position of Ale near the top of the limestone rise, it is unlikely that a stream once ran through the cave, so perhaps the water flow was intermittent as result of seepage through the limestone before the cave entrance formed. As the W-K2 tephra could not have entered the cave before the main entrance was formed, it is possible that the end of water flow, deposition of tephra and opening of the entrance were related or occurred closely in time.

Three AMS radiocarbon dates on charcoal (Table 1) are calibrated by the CALIB 5.01 program using the Intcal04.14C curve, with the calibrated ages cited here as two-sigma ranges (Reimer et al. 2004; Stuiver and Reimer 1993). Sample Wk-12115 (2760-2460 cal. BP) from the bottom of layer 3, just above the W-K2 tephra, provides a minimum age for the start of chert extraction. The presence of debitage in layer 1 possibly indicates that extraction continued until 1310-1090 cal. BP (Wk-12113). This seems a long period (ca 1200-1500 years) for continuous exploitation of the three seams, even with quarrying into the cave wall. Perhaps extraction was intermittent, with only small quantities removed at any one time. The seams, however, are still not completely exhausted, because one of our guides in 2002 was able to break off small but potentially usable pieces with relative ease.

The basal date of Wk-12115 raises another, quite different issue. The proposed date of 3480-3150 cal. BP for the W-K2 volcanic event is at least 400-600 years earlier than this sample. There is no evidence of human activity in the reworked upper section of the tephra (layer 4A), and the question arises whether this indicates the absence of people from the region for many years after the W-K2 event, or whether the cave and its chert seams were simply not found until the time of Wk-12115. Some support for the former possibility comes from three other dating results from Auwa/FGT and Yombon Airstrip/FIF: FGT/I: Beta-1545, 2575 [+ or -] 100 bp (Specht et al. 1981: 14); FGT/7: Beta-47048, 2570 [+ or -] 90 bp (Pavlides 1999: appendix 1); and FIF/3: Beta62320, 2540 [+ or -] 70 bp (Pavlides 2004: fig. 2, where the age is uncorrected for [delta][sup.]13 C). As at Ale, in each case the dates relate to first signs of human activity at the sites after the W-K2 event. The four results are statistically the same at the 95% confidence level, and may indicate that the Passismanua area was abandoned after the W-K2 event for several hundred years longer than on the Willaumez Peninsula and Garua Island (Petrie and Torrence 2008: table 7). This is important for its implications regarding memories of place and the location of chert sources (cf. Torrence and Doelman 2007: 53-55). If the Passismanua area were uninhabited for such a long time, it is highly unlikely that people returning to the area knew the location of chert exposures, and they would have had to undertake the discovery process all over again. It is also possible that people living on the coast, where the W-K2 event had very little impact, made special forays into the limestone country to obtain chert, and this may explain the presence of plain pottery of coastal origin at Auwa (Specht et al. 1981: 13; Glenn Summerhayes, personal communication). Such a long period of abandonment, however, would raise doubts about cultural continuity implied by the presence of stemmed chert tools in both pre-and post-W-K2 tephra levels (Kononenko et al. 2010).

The chert sources--discussion

Each kind of chert source is dependent on time and the processes of erosion, particularly in dolines that require at least partial collapse to provide an opening, or for an underground river to force its way to the surface or down into a subterranean channel. Neither of these is a predictable event. The new data confirm Pavlides' suggestion (2004: 98) that primary chert sources occur across a large part of the Passismanua region: the straight-line distance from Auwa and Helena cave to Ale cave is about 22-24 kin. Although the number and location of past chert sources is not known, chert was probably widely available throughout the ~3,000 k[m.sup.2] area of the Yalam Limestone in this area, with perhaps several hundred or more caves with seams of chert yet to be recorded. Pavlides (2008) also reports chert sources on Lamogai Plateau, but apparently of lower quality than the Passismanua chert.

Fresh exposures of Passismanua chert vary considerably in colour across the light grey, greyish-orange, greyish-yellow, yellowish-brown, light brown and reddish brown ranges of the Munsell rock colour chart (Munsell 2009: 4-5; cf. Goodale 1966: 28-29). Many archaeological pieces have a yellowish-orange, off-white or greyish-yellow patina resulting from burial in tephra-derived soils. Some items recovered from water courses have a distinctive reddishpurple stain derived from immersion in water (Goodale 1966: 28; Chowning and Goodale 1966: 150); no excavated pieces have this stain. The colour of the R1 chert seam appears to be unusually dark. No chert pieces of this colour have been recorded among the thousands of pieces excavated to date, but a reddish-brown chert occurs in the Chowning-Goodale collection (Goodale 1966: 28) and in surface collections between Auwa and Akiuli (Fig. 3). The R1 colour may be distorted, or perhaps material from this source was used in a very restricted area where no archaeological work has yet been carried out.

The likely number and distribution of seams in caves and stream bed scatters were probably sufficient to ensure that irrespective of people's subsistence base, their daily and seasonal movements would have taken them close to known sources of chert. Replenishing supplies of raw material would have been easily embedded in daily activities that required travel, and special scheduling was unnecessary (cf. Pavlides 2006: 214). In 1979 I recorded at least four stream bed sources within one hour's walk from Auwa, in addition to probable sinkhole sources at Auwa and Sisisel (Pavlides 1999: 213). There is no reason to believe that this kind of density did not exist in the past. There are two corollaries to this. Firstly, there was little need to 'provision the person' (cf. Kulm 1995: 22) as individuals travelled to acquire or produce food, engage in social activities, and so forth. The caves would have provided temporary shelter while people replenished their supplies of raw material. This could explain the presence in Helena cave of ground stone axe blades that suggest activities other than simply extraction or reduction (Hapka 2009: table 1). Secondly, the large number of sources across the landscape would have allowed people to spread extraction between several localities, rather than relying on one only. This would have conserved individual seams in caves and ensured an ongoing supply, and could explain the long history of exploitation at Ale. As these sources need to be exposed by partial or complete collapse of dolines, each new opening was probably inspected for chert seams immediately it was noticed. Stream bed sources of nodules, however, are relatively impermanent, as during the wet season the streams become raging torrents and flush loose nodules downstream. The chert from these sources was probably best used soon after discovery. The identification of these sources would have been an ongoing task built into any travel involving the crossing of water courses.

Only when people moved out of the Yalam Limestone region would they have had problems with supply. A transect survey from Auwa to Kandrian in 1979 revealed a marked decline in the frequency of chert artefacts on landforms of the Johanna Beds (Ryburn 1975, 1976), which lie between the Yalam Limestone and the coastal Pleistocene raised coral reef limestone (Fig. 3). The Johanna Beds are formed by Pliocene lagoon floor sediments that do not contain chert, and today there are very few settlements on these beds on account of the lack of reliable water supplies. It is not surprising, therefore, that in 1979 two days of survey around Angelek village, which is situated on the Johanna Beds, produced only one location with chert artefacts.

[FIGURE 6 OMITTED]

The seams at Ale and R1 show that the abundance of chert varies between localities. The discontinuous seams consist of slabs or blocks that have broken along joint lines within the limestone. This probably made it impossible to predict the size and quality of each piece until it was extracted. Several pieces of chert excavated at various sites have irregular joint planes that prevented controlled flaking and presumably led to their rejection (Figs 6, 7A). The seam in R1 shows similar jointed blockiness, and this may be a feature common to all chert seams. Furthermore, the variable thickness of the seams could have presented problems, as the removal of cortex or surface irregularities on thin pieces would quickly reduce their viability (cf. Fig. 7B). In brief, the length of a seam may look impressive, but the size and nature of the pieces that can be extracted and their flaking quality cannot be relied upon.

This restriction on size also applied to stream bed nodules, which are rarely longer than 150-200 mm. While this may be adequate for most artefacts, the chert often has a cortex up to 5 mm thick that must be removed to determine whether there is usable chert inside. Nodules might have too much cortex for some purposes, or consist entirely of soft, chalky rock totally unsuitable for artefact production. The irregular shape of many nodules further restricts the size of artefacts that can be produced (Fig. 7C).

Pavlides' excavations revealed variations in the quantity of 'rubble' at her sites (Table 2). This is not simply the result of different excavation areas, as Auwa with 3 m2 has more than 150 times the amount of rubble found in 5 m2 at Yombon Airstrip (comparison is based on excavated area since available data do not permit calculation of density). The rubble at Auwa and Sisisel included a high proportion of cortex likely to be from seams, which Pavlides (2006: 211-213) interprets as indicating proximity to primary chert exposures, with the 'rubble' representing waste from extraction and testing. If this is correct, then the 'wastage' rate, the ratio by weight of unusable to usable raw material, was enormous: 20:1 at Auwa, and 125:1 at Sisisel. Even if some 'rubble' was of natural origin and some chert artefacts were discarded elsewhere, wastage was obviously high and usable chert at times was probably available in relatively small quantities. The number and size of cortical artefacts from Pavlides' sites probably reflect this (Tables 3,4).

[FIGURE 7 OMITTED]

Cortex is present in high proportions throughout the sequence, and the mean maximum and axial lengths of cortical pieces are consistently larger than those of noncortical pieces in every period. Pavlides (2006) interprets the quantity of cortex as a general indication of the reduction stage, but its presence may indicate that it was not possible or even necessary to remove it completely as this could have made the tool too small for its intended function. Moreover, where the objective was simply to produce a usable edge, the presence of cortex on part of the tool would have been irrelevant.

These constraints, however, did not prevent the production of numerous unifacial and bifacial tools with stems or side notches forming a 'waist' during the middle Holocene (Pavlides 2006: 215-219). Most of these are less than 100 mm long (Pavlides 1999: table 9.27; cf. Shutler and Kess 1969: table 1), but larger ones do occur. The extant length of a broken stemmed tool from Auwa is 122 mm and its original length was probably closer to 200 mm (Fig. 7D). A biface excavated at Auwa is 124 mm long, though the limitation of the raw material is reflected by patches of cortex on both surfaces that show that the original block was only 44 mm thick (Fig. 7E, 7F). On the other hand, the quality of nodules or blocks was sometimes sufficient to permit production of stemmed tools completely lacking cortex that are only 5-12 mm thick (Figs 8A, 8B; Kononenko et al. 2010: table 1, fig. 6).

The transfer of chert through exchange networks was clearly not hindered by these issues, as chert artefacts have been found on the coast and islands around Kandrian, where chert does not occur naturally (Specht 2009). Chert tools have also been found in the Arawe Islands, 55 km to the west of Kandrian (Gosden et al. 1994) and about 25 km southeast from the nearest outcrops of Yalam Limestone (Ryburn 1976). The export of chert included finished bifacial and waisted forms to the Kandrian area (Specht 2009: fig. 2.3), and stemmed tools to the Arawe Islands (C. Pavlides, La Trobe University, and Rev. Jamieson, Kumbun Mission, personal communications).

Comparison between obsidian and chert sources

The New Britain obsidian source regions have been far better studied than the chert ones, with a growing literature on their distribution, geochemistry, age of formation, reduction strategies, production of special stemmed forms, and tool functions (Specht 1981; Fullagar et al. 1991; Torrence et al. 1992, 2004a; Fullagar 1992, 1993; Bird et al. 1997; Kealhofer et al. 1999; Araho et al. 2002; Rath and Torrence 2003; Torrence 2004b, 2005; Kononenko 2007, 2011; Kononenko and Torrence 2009).

Table 5 summarises the main features of the obsidian sources compared against those of the Passismanua chert. Whereas the chert is found in areas of marine-derived limestone, obsidian is the product of specific volcanic settings that produced silica-rich glass in central Willaumez Peninsula (Lowder and Carmichael 1970) and at Mopir to the south of Mt Witori (Specht and Hollis 1982; Fullagar et al. 1991). In contrast to the chert sources, which are the products of long duration geological processes, the obsidian flows were formed as specific events, with some occurring within the last 20,000 years (Torrence et al. 2004a); further flows may yet occur.

The extent of each obsidian source region is probably around 40-50 k[m.sup.2] (Fullagar et al. 1991: 112). While this pales into insignificance in comparison with the ~3,000 k[m.sup.2] of the Yalam Limestone over which chert is likely to occur, the obsidian sources have many advantages. They are concentrated, highly visible and most are easily accessible as surface outcrops and secondary deposits on beaches and in gullies, though at several localities boulders were recovered by digging pits (Specht 1981: 344-346). Moreover, after 35,000-40,000 years of exploitation, the two source regions still contain thousands of tonnes of extractable obsidian; there is no evidence that any particular exposure was ever exhausted. Their use could be continuous and long-term without taking conservation measures or having to seek out new exposures.

The obsidian sources were probably available immediately after their formation, once the flows had cooled. Even after severe volcanic events that left the Willaumez Peninsula unoccupied for up to several centuries (Petrie and Torrence 2008: table 7), access to obsidian was probably still possible before the region became habitable again. This contrasts with Mopir, where the W-K2 eruption caused significant landscape changes that suddenly 'shifted' the obsidian source well inland and apparently interrupted access to it (Torrence et al. 1996: fig. 3, 2004a: 60; Summerhayes et al. 1998; Torrence 2004a: table 10.1). This and other Holocene eruptions appear to have had little or no direct effect on chert sources, although they may have affected the Passismanua area in other ways to prevent or deter occupation for a lengthy period after the W-K2 event.

The quality of obsidian in both source regions is generally high, with the exception of the Hamilton subsource, though some flows of Baki obsidian have layers of impurities separating bands of flakeable obsidian (Torrence et al. 1992: table 1). Cortex is sometimes present, especially on cobbles from the Baki and Gulu sub-sources, but generally it is not a major issue as cortex-free pieces can be removed from the large boulders on beaches and in gullies. Thus, unlike chert, there was less need to test blocks of obsidian or remove thick layers of cortex, and wastage would have been comparatively minimal. On the other hand, the concentrated distribution of the obsidian meant that individuals and groups travelling away from the sources would have had to plan carefully to ensure an adequate supply of multi-purpose tools or raw material, or both, until they could return to the source areas (Torrence 1992: 121, 2004a). In contrast, those living in the Passismanua area could be reasonably confident of finding an adequate supply of raw material while travelling within their own territory, and only needed to be well-provisioned if they ventured beyond it.

The concentrated and permanent nature of the obsidian sources probably led to the long-term attachment of their controllers/owners to specific localities and their products, through which they could develop, maintain and perhaps expand social networks based on or supported by the exchange of obsidian. In contrast, those living in the Passismanua area were probably attached more in the long-term to a territory than to a locality, and had less opportunity to monopolise or restrict access to chert.

No chert artefacts likely to derive from the Yalam Limestone have been recorded at sites on the Willaumez Peninsula, but obsidian from that source region as well as from Mopir was transported to the Passismanua area from the early Holocene onwards (Specht et al. 1981: 14; 1983: 92), and to the Arawe Islands from at least the middle Holocene (Gosden et al. 1994: 108). This persistent presence throughout the Holocene of obsidian in sites on the south side of the Whiteman Range, and the total absence of chert at the Willaumez Peninsula sites suggest that the Passismanua people did not acquire obsidian by direct access to the source regions, but obtained it through down-the-line exchange networks. Presumably this was across the island via the Lamogai Plateau, which has been a regular trade route in historical times (Pavlides 1988: 53), rather than through the extremely rugged and uninhabited Whiteman Range between the Passismanua area and Willaumez Peninsula (cf. Gilliard and Lecroy 1967). Whatever goods moved northwards against the southwardmoving obsidian, they did not include chert. Perhaps the people living near the vast reserves of obsidian had no need to import chert. The same might be said, in reverse, about the importation of obsidian into the chert-using areas: was it really necessary in a utilitarian sense, or was it required for purposes that for culturally-defined reasons could not be met by chert?

The sharpness, colour, translucency and shininess of the obsidian appear to have combined to make it a highly desired good in exchanges (Torrence 2004b, 2005: 364365). It was transported over large distances from the late Pleistocene onwards, and accompanied the human expansion into Remote Oceania (Summerhayes and Allen 1993; Araho et al. 2002; Specht 2002; Torrence et al. 2004b). The early and middle Holocene saw the production of two of finely made stemmed tools on prismatic blades (Type 1) and on kombewa and other flakes (Type 2) that probably constituted valuables or prestige goods (Araho et al. 2002; Rath and Torrence 2003; Torrence 2004b, 2005). A special feature of these putative valuables is their large size, which arguably reflects the availability of large blocks and the presence of specialist skills that allowed the production of tools up to ca 200 mm long (Araho et al. 2002; Torrence 2004b: table 2; Specht 2005; Torrence et al. 2009: table 1).

[FIGURE 8 OMITTED]

To date, these obsidian tools have dominated the discussion of valuables, but some stemmed bifacial chert tools may also belong in this category (Fig. 8). Reflecting the constraints of the raw material, these chert artefacts are generally smaller than those of obsidian, none appear to be made on kombewa flakes, and there are no stemmed prismatic blades similar to the Type 1 forms produced at the obsidian sources. The chert stemmed tools are notable, however, for their completeness, thinness and complete removal of cortex; they are clearly the products of a significant level of skill beyond that needed to produce a simple usable edge. While they lack the brilliance and shininess of obsidian, they display a similar concern for symmetry and balance. Perhaps these, too, were a form of valuable, with their value deriving from the artisan's capacity to produce such forms from a raw material that was not always user-friendly.

The production of Type 2 obsidian tools began on Willaumez Peninsula during the early Holocene, and they were common until the W-K2 eruption, after which stemmed tools were rare, of much reduced size and in different forms (Torrence et al. 2000; Araho et al. 2002; Torrence 2004a; Kononenko et al. 2010). The Passismanua stemmed chert tools have a rounded or paddle-shaped blade, lack the crescentic blades and delicate pointed terminals of the finest Type 2 obsidian forms (Araho et al. 2002; Torrence 2004b; Torrence et al. 2009: 120-121). Their production on non-kombewa flakes presumably reflects the limitations of the raw material. Their apparent absence from early Holocene levels in the Passismanua area may be a sampling issue, although the double-notched tool recovered by Pavlides from an early Holocene context at the Yombon Airstrip site (Pavlides 1999: figs. 8.12-13) may indicate their presence at that time. They were definitely produced in the middle Holocene (Pavlides 1999: plates 9.17 to 9.20, 2006), and as with the obsidian forms, a few examples occur in post-W-K2 contexts (Kononenko et al. 2010). It is tempting to see the production of similar tools in the two areas during the same period as the result of social connections between the areas involving the transfer of ideas and skills between the two areas rather than simply the movement of objects, with the chert artisans simply imitating imported obsidian forms. While some of the differences between the chert and obsidian forms probably reflect the potentials of the raw materials, the differences in overall form could be a function of different requirements, implying variations in the purposes of production.

Torrence (e.g. 2004a, 2004b; Torrence and Summerhayes 1997; Torrence and Swadling 2008; Torrence et al. 2009) has emphasised the importance of social relations for access to the Willaumez Peninsula obsidian sources. Furthermore, Torrence and Doelman (2007: 61) explain the dominance of the Kutau-Bao obsidian during Lapita pottery times as probably reflecting some form of strict control over access to the sub-sources, whereas before and after that period "the degree of monopoly" was less. If that were the case, then forms of control over access to sources could have varied through time. We can only guess how they might have been organised, but it might have involved a relatively small number of groups living close to the sub-sources, as in recent times (Specht 1981). The dispersed distribution of the chert exposures would also have suited management by the local groups closest to each one, but this would have resulted in numerous controlling groups spread across the Passismanua area. If people moved out of their local area, they would have had little or no knowledge about the location of exposures belonging to other groups. Even if exposures were identified in someone else's territory, fear of local spirits and other supernatural forces associated with particular resources would always have been a threat to those who did not know how to appease them (cf. Tacon 1991:199 for an Australian example). In recent times, malignant spirits were viewed as being present throughout the Passismanua area, particularly in caves and sinkholes (cf. Goodale 1995: 30-31, 72-73, 133), and similar concerns may well have existed in the distant past. Furthermore, in 1979-1981 fear of sorcery between the Kaulong and Sengseng people who live in the Passismanua area (cf. Goodale 1995: 24), and between them and coastal groups, was still a potent force that deterred unwelcome visits to each other's territory.

Conclusions

The new data on chert sources in the Passismanua area broadly support the predictions of Pavlides (1999, 2006), but introduce several significant new perspectives. The widespread availability of chert is tempered by the unpredictability of its quality and quantity, and the life-span of individual exposures; separately and collectively these would have imposed constraints on production that differed markedly from those at the obsidian sources. How much they influenced behaviour patterns that are likely to be expressed in the archaeological record remains to be seen, but they do invite caution in the interpretation of some aspects of the reduction process. Pavlides (2006: 211) discusses the quantity of cortical pieces as related both to proximity to a source and to the stage of reduction (cf. Symons 2003 for obsidian). In many cases, however, the presence of cortex could relate directly to the nature of the raw material: in some cases, full removal of cortex could have destroyed the usefulness of particular pieces, and in other cases complete removal was not necessary. What is more important, however, is the presence of complete, cortex-free stemmed tools that I have tentatively identified as valuables comparable to stemmed obsidian tools. Their value probably resided as much in the skills required for their production as in their finished forms, perhaps even requiring specialist artisans. Although the currently known distribution of these tools is more restricted than that of the obsidian stemmed forms (Torrence et al. 2009: fig. 1), it covers a substantial part of southern New Britain. Finally, unlike the virtually unlimited reserves of obsidian at Mopir and on Willaumez Peninsula, the chert exposures required careful management to ensure their long-term viability, though the discovery of new exposures was probably always a high priority.

In addition to extending the discussion of stemmed tools as forms of valuables, this review also opens opportunities for exploring new dimensions of social interaction and the development or transfer of skills in prehistoric New Britain. I have previously argued for the existence of "networks of social relationships linking communities" to explain the distribution across New Britain of stemmed obsidian tools attributed to the middle Holocene (Specht 2005: 374). One Type 2 obsidian tool, greatly reduced by reworking, was found near Avalngin hamlet, about 5 km southeast from Auwa (Specht 2005: 378). This suggests that the makers of the chert stemmed tools almost certainly were familiar with those of obsidian. It invites the question whether the development of the two forms was connected through the exchange of skills rather than just objects, or possibly even the movement of specialist artisans between the two areas. A similar transfer of ideas and skills, again during the middle Holocene but over a much larger distance, has been suggested between New Britain and Manus in the context of other stemmed obsidian tools, particularly those with hammer-dressing (Torrence and Swadling 2008; Torrence et al. 2009). If the Passismanua evidence does support a similar transfer of ideas and/or skills, it would support the picture of a regionally interconnected word through which people, goods and ideas could move well before the emergence of Lapita pottery (cf. Torrence et al. 2009:119; Swadling et al. 2008).

Finally, the new data on chert exposures clearly reveal the potential of such sources to provide a more detailed picture of extraction, reduction, use and discard than has been possible hitherto. There is also a clear need for a more complete description and classification of the waisted and stemmed forms, to allow more controlled comparisons with the obsidian forms. The French descriptions of the Helena, Pomalngin and Singlip caves suggest that these may be profitable locations to begin this work.

Acknowledgements

This paper draws on my work in the Kandrian-Passismanua area in 1979-1981 and 2002. I thank the Australian Research Grants Committee for funding in 1979-1981, and the Australian Museum for funding in 2002. I thank the PNG National Museum and Art Gallery for research affiliation, and the PNG National Research Institute for research permits; John Normu, John Namuno, Blaise Vatete and Matthew Kabui (West New Britain Cultural Centre) for research support and other assistance over many years; Jack Kaik and others of Aliwo village for access to Ale cave and for other assistance, and Bob Lisio (then president of the Kandrian Local Level Government) for liaison and transport assistance in 2002; Professor Hugh Davies and Marisa Torombe (University of Papua New Guinea) for tephra analyses; Roman Hapka (Switzerland) for copies of his field notes and publications, and for permission to reproduce his photo of the R1 chert seam; Christina Pavlides (La Trobe University) for discussions about Auwa and other potential exposures, and for information on finds in the Arawe Islands. My special thanks go to the late Baiva Ivuyo of the PNG National Museum and Art Gallery for his professional contributions to this and other field projects on which we worked together. I am also grateful to Ombone Kaiku and Theodore Mawe (PNG National Museum and Art Gallery), Ian Lilley (University of Queensland), and Foss Leach and Janet Davidson (New Zealand) for their participation in various aspects of the 1979-1981 field seasons. Finally, Robin Torrence and an unnamed referee provided constructive suggestions that have improved the paper.

References

Andrefsky, W. 1994. The geological occurrence of lithic material and stone tool production strategies. Geoarchaeology: An International Journal 9(5):375-391.

Araho, N., Torrence, R. and White, J.P. 2002. Valuable and useful: Mid-Holocene Stemmed Obsidian Artefacts from West New Britain, Papua New Guinea. Proceedings of the Prehistoric Society 68:61-81.

Bamforth, D.B. 1990. Settlement, raw material, and lithic procurement in the central Mojave Desert. Journal of Anthropological Archaeology 9:70-104.

Bird, R., Torrence, R., Summerhayes, G.R. and Bailey, G. 1997. New Britain obsidian sources. Archaeology in Oceania 32(1): 61-7.

Bulmer, S. 2005. The stones of Pasismanua revisited. In C. Gross, H.D. Lyons and D.A. Counts (eds), A Polymath Anthropologist: Essays in honour of Ann Chowning, pp. 23-34. Auckland: Department of Anthropology, University of Auckland. Research in Anthropology and Linguistics, Monograph 6.

Chowning, A. and J. Goodale 1966. A flint industry from southwest New Britain, Territory of New Guinea. Asian Perspectives 9:150-153.

Davies, H.L. 1973. Gazelle Peninsula, New Britain. Sheet SB/562.1:250,000 Geological Series--Explanatory Notes. Canberra: Australian Government Publishing Service.

Evans, B. and M-J. Mountain 2005. Pasin bilong tumbuna: archaeological evidence for early human activity in the highlands of New Guinea. In A. Pawley, R. Attenborough, J. Golson and R. Hide (eds), Papuan Pasts: Cultural, linguistic and biological histories of Papuan-speaking peoples, pp. 363386. Canberra: Pacific Linguistics, Australian National University. Pacific Linguistics 572.

Fullagar, R., 1992. Lithically Lapita: functional analysis of flaked stone assemblages from West New Britain Province, PNG. In JC. Galipaud (ed.), Poterie Lapita et Peuplement. Actes du Colloque Lapita Noumea Janvier 1992, pp. 135-143. Noumea: ORSTOM.

Fullagar, R. 1993. Flaked stone tools and plant food production: a preliminary report on obsidian tools from Talasea, West New Britain, PNG. In P. Anderson, S. Beyries, M. Otte, and H. Plisson (eds), Traces et Fonction: Les gestes retrouves, pp. 331-337. Liege: Editions ERAUL. Colloque International de Liege 50.

Fullagar, R., Summerhayes, G., Ivuyo, B. and Specht, J. 1991. Obsidian sources at Mopir, West New Britain Province, Papua New Guinea. Archaeology in Oceania 26(3): 110-114.

Fullagar, R. and R. Torrence 1991. Obsidian exploitation at Umleang, Lou Island. In J. Alien and C. Gosden (eds), Report of the Lapita Homeland Project, pp. 113-143. Canberra, Department of Prehistory, Research School of Pacific Studies, Australian National University. Occasional Papers in Prehistory 20.

Gilliard, E.T. and M. Lecroy 1967. Results of the 1958-1959 Gilliard New Britain expedition. 4. Annotated list of birds of the Whiteman Mountains, New Britain. Bulletin of the American Museum of Natural History 135:173-216.

Goodale, J.C. 1966. Imlohe and the mysteries of the Passismanua, Southwest New Britain. Expedition 8(3):20-31.

Goodale, J.C. 1995. To Sing with Pigs is Human. Seattle: University of Washington Press.

Gosden, C., Webb, J., Marshall, B. and Summerhayes, G.R. 1994. Lolmo Cave: A mid- to late-Holocene site, the Arawe Islands, West New Britain Province, Papua New Guinea. Asian Perspectives 33(1):97-119.

Hapka, R. 2009. Speleo-Archaeology in Papua New Guinea: The discovery of Palaeolithic flint mines on the shores of the Soloman [sic] Sea. In W.B. White (ed.), Proceeding, 15th International Congress of Speleology. Vol. 1: Symposia, Part 1. Kerrville, Texas, USA, July 19-26, 2009, pp. 97-104. [No place of publication]: International Union of Speleology.

Kealhofer, L., Torrence, R. and Fullagar, R. 1999. Integrating phytoliths within use-wear/residue studies of stone tools. Journal of Archaeological Science 26:527-546.

Kononenko, N. 2007. The contribution of use-wear/residue studies of obsidian artefacts for understanding changes in settlement and subsistence patterns in West New Britain, Papua New Guinea. Indo-Pacific Prehistory Association Bulletin 27:135-142.

Kononenko, N. 2011. Experimental and archaeological studies of use-wear and residues on obsidian artefacts from Papua New Guinea. Technical Reports of the Australian Museum, Online 21:1-244.

Kononenko, N. and R. Torrence 2009. Tattooing in Melanesia: local invention or Lapita introduction? Antiquity 83. Project Gallery, accessed on-line 9 August 2009 at http://antiquity. ac.uk/projgall/kononenko/.

Kononenko, N., J. Specht and R. Torrence 2010. Persistent traditions in the face of natural disasters: stemmed and waisted stone tools in Late Holocene New Britain, Papua New Guinea. Australian Archaeology 70:17-28.

Kuhn, S.L. 1995. Mousterian Lithic Technology: An ecological perspective. Princeton (N J): Princeton University Press.

Lowder, G.G and I.S.E. Carmichael 1970. The volcanoes and caldera of Talasea, New Britain: geology and petrology. Geological Society of America Bulletin 81:17-38.

Munsell Color 2009. Geological Rock-Color Chart, with genuine Munsell color chips. Revised edition. Grand Rapids (MI): Munsell Color.

Pavlides, C. 1988. Trade and exchange in the Arawe Islands, West New Britain, Papua New Guinea. Unpublished BA (Hons) thesis, La Trobe University, Bundoora, Victoria.

Pavlides, C. 1999. The Story of Imlo: The Organisation of Flaked Stone Technologies from the Lowland Tropical Rainforest of West New Britain, Papua New Guinea. Unpublished PhD thesis, La Trobe University, Bundoora, Victoria.

Pavlides, C. 2004. From Misisil Cave to Eliva Hamlet: rediscovering the Pleistocene in interior West New Britain. Records of the Australian Museum, Supplement 29:97-108.

Pavlides, C. 2006. Life before Lapita: new developments in Melanesia's long-term history. In I. Lilley (ed.), Archaeology of Oceania: Australia and the Pacific Islands, pp. 205-227. Oxford: Blackwell Publishing.

Pavlides, C. 2008. On the inside: Coping with the inland environments of West New Britain. Unpublished paper presented to the annual conference of the Association of Social Anthropologists in Oceania, Canberra, February 2008.

Pavlides, C. and C. Gosden 1994. 35,000 year-old sites in the rainforests of West New Britain, Papua New Guinea, Antiquity 68:604-610.

Petrie, C.A. and R. Torrence 2008. Assessing the effects of volcanic disasters on human settlement in the Willaumez Peninsula, Papua New Guinea: a Bayesian approach to radiocarbon calibration. The Holocene 18(5):729-744.

Rath, P. and R. Torrence 2003. Producing value: stemmed tools from Garua Island, Papua New Guinea. Australian Archaeology 57:119-127.

Reimer, P.J., M.G.L. Baillie, E. Bard, A. Bayliss, J.W. Beek, C.J.H. Bertrand, R.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, F.G. McCormac, S.W. Manning, C.B. Ramsey, R.W. Reimer, S. Remmele, J.R. Southon, M. Stuiver, S. Talamo, F.W. Taylor, J. van der Plicht and C.E. Weyhenmeyer, 2004. IntCal04 terrestrial radiocarbon age calibration, 26-0 ka BP. Radiocarbon 46: 1029-1058.

Ryburn, R.J. 1974. Pomio, New Britain. Sheet SB/56-6. International Index. 1:250,000 Geological Series--Explanatory Notes. Canberra: Australian Government Publishing Service.

Ryburn, R.J. 1975. Talasea-Gasmata, New Britain. Sheet SB/56-5 & SB/56-9.1:250,000 Geological Series--Explanatory Notes. Canberra: Australian Government Publishing Service.

Ryburn, R.J. 1976. Cape Raoult-Arawe, Papua New Guinea. Sheet SB/55-8 & SB/55-12 International Index. 1:250,000 Geological Series--Explanatory Notes. Canberra: Australian Government Publishing Service.

Sheppard, P.J. 1996. Hard rock: archaeological implications of chert sourcing in Near and Remote Oceania. In J. Davidson, G. Irwin, F. Leach, A. Pawley and D. Brown (eds), Oceanic Culture History: Essays in honour of Roger Green, pp. 99-115. New Zealand Journal of Archaeology Special Publication.

Shutler, R. and C.A. Kess 1969. A lithic industry from New Britain, Territory of New Guinea, with possible areal and chronological relationships. Bulletin of the Institute of Ethnology, Academia Sinica 27:129-140.

Specht, J. 1981. Obsidian sources at Talasea, West New Britain, Papua New Guinea. Journal of the Polynesian Society 90(3):337-356.

Specht, J. 2002. Obsidian, colonising and exchange. In S. Bedford, C. Sand and D. Burley (eds), Fifty Years in the Field. Essays in honour and celebration of Richard Shutler Jr's archaeological career, pp. 37-49. Auckland: New Zealand Archaeological Association. Monograph 25.

Specht, J. 2005. Obsidian stemmed tools in New Britain: aspects of their role and value in mid-Holocene Papua New Guinea. In I. Macfarlane, M.-J. Mountain and R. Paton (eds), Many Exchanges: archaeology, history, community and the work of Isabel McBryde, pp. 373-392. Aboriginal History Monograph 11.

Specht, J. 2009. The aceramic to ceramic boundary in the Bismarck Archipelago. In P.J. Sheppard, T. Thomas and G.R. Summerhayes (eds), Lapita: Ancestors and descendants pp. 11-34. Auckland: New Zealand Archaeological Association Monograph 28.

Specht, J. and J.D. Hollis 1982. A new obsidian source in West New Britain, Papua New Guinea. Mankind 13(5):424-27.

Specht, J., Lilley, I. and Normu, J. 1981. Radiocarbon Dates from West New Britain, Papua New Guinea. Australian Archaeology 12:13-15.

Specht, J., Lilley, I. and Normu, J. 1983. More on Radiocarbon Dates from West New Britain, Papua New Guinea. Australian Archaeology 16:92-95.

Stuiver, M. and P.J. Reimer 1993. Extended 14C database and revised CALIB radiocarbon calibration program. Radiocarbon 35(1):215-230.

Summerhayes, G. and J. Allen 1993. The transport of Mopir obsidian to Late Pleistocene New Ireland. Archaeology in Oceania 28(3): 144-148.

Summerhayes, G.R., J.R. Bird, R. Fullagar, C. Gosden, J. Specht and R. Torrence 1998. Application of PIXE-PIGME to archaeological analysis of changing patterns of obsidian use in West New Britain, Papua New Guinea. In M.S. Shackley (ed.), Archaeological Obsidian Studies: Method and Theory, pp. 129158. Plenum Press, New York.

Swadling, P., P. Wiessner and A. Tumu 2008. Prehistoric stone artefacts from Enga and the implication of links between the highlands, lowlands and islands for early agriculture in Papua New Guinea. Journal de la Societe des Oceanistes 126127:271-292.

Symons, J. 2003. Obsidian artefacts and land-use in the midHolocene of the Willaumez Peninsula, Papua New Guinea. Australian Archaeology 57:128-134.

Tacon, P.S.C., 1991. The power of stone: symbolic aspects of stone use and tool development in western Arnhem Land, Australia. Antiquity 65:192-207.

Torombe, M.W. 2006. Holocene tephras from Kandrian, the Huon Peninsula, and Mount Giluwe, Papua New Guinea. Unpublished BSc (Hons) thesis, University of Papua New Guinea, Waigani.

Torrence, R. 1986. Production and Exchange of Stone Tools. Prehistoric obsidian in the Aegean. Cambridge: Cambridge University Press.

Torrence, R. 1992. What is Lapita about obsidian? A view from the Talasea sources. In J-C. Galipaud (ed.), Poterie Lapita et Peuplement. Actes du Colloque Lapita Noumea Janvier 1992, pp. 111-26. Noumea: ORSTOM.

Torrence, R. 2004a. Now you see it, now you don't: changing obsidian source use in the Willaumez Peninsula, Papua New Guinea. In J. Cherry, C. Scarre and S. Shennan (eds), Explaining Social Change: studies in honour of Colin Renfrew, pp. 115-125. Cambridge: MacDonald Institute.

Torrence, R. 2004b. Pre-Lapita valuables in Island Melanesia. Records of the Australian Museum, Supplement 29:163-172.

Torrence, R. 2005. Valued stone--how so? In I. Macfarlane, M-J. Mountain and R. Paton (eds), Many Exchanges: archaeology, history, community and the work of Isabel McBryde, pp. 357372. Canberra: Aboriginal History Inc. Monograph 11.

Torrence, R., R. Bonetti, A. Guglielmetti, A. Manzoni and M. Oddone 2004a. Importance of source availability: A case study from Papua New Guinea. Mediterranean Archaeology and Archaeometry International Journal 4(2):53-65.

Torrence, R. and T. Doelman 2007. Problems of scale: evaluating the effects of volcanic disasters on cultural change in the Willaumez Peninsula, Papua New Guinea. In J. Grattan, J. and R. Torrence (eds), Living under the Shadow: The archaeological, cultural and environmental impact of volcanic eruptions, pp. 42-66. Walnut Creek, California: Left Coast Press.

Torrence, R., V. Neall, T. Doelman, E. Rhodes, C. McKee, H. Davies, R. Bonetti, A. Guglielmetti, A. Manzoni, M. Oddone, J. Parr, C. Wallace 2004b. Pleistocene colonisation of the Bismarck Archipelago: new evidence from West New Britain. Archaeology in Oceania 39(3): 101-130.

Torrence, R., C. Pavlides, E Jackson and J. Webb 2000. Volcanic disasters and cultural discontinuities in Holocene time, in West New Britain, Papua New Guinea. In W.G. McGuire, D.R. Griffiths, P.L. Hancock and I.S. Stewart (eds), The Archaeology of Geological Catastrophes, pp. 225-244. London: The Geological Society of London. Special Publication 171.

Torrence, R., Specht, J., Fullagar, R. and Bird, R. 1992. From Pleistocene to Present: Obsidian sources in West New Britain, Papua New Guinea. Records of the Australian Museum, Supplement 15:83-98.

Torrence, R., J. Specht, R. Fullagar and G. Summerhayes 1996. Which obsidian is worth it? A view from the West New Britain sources. In J. Davidson, G. Irwin, F. Leach, A. Pawley and D. Brown (eds), Oceanic Culture History: Essays in Honour of Roger Green, pp. 211-224. Dunedin: New Zealand Journal of Archaeology Special Publication.

Torrence, R. and G.R. Summerhayes 1997. Sociality and the short distance trader: intra-regional obsidian exchange in the Wiilaumez Peninsula region, Papua New Guinea. Archaeology in Oceania 32(1):74-84.

Torrence, R. and P. Swadling 2008. Social networks and the spread of Lapita. Antiquity 82:600-616.

Torrence, R., P. Swadling, N. Kononenko, W. Ambrose, P. Rath and M.D. Glascock 2009. Mid-Holocene social interaction in Melanesia: New evidence from hammer-dressed obsidian stemmed tools. Asian Perspectives 48:118-147.

Australian Museum, 6 College Street, Sydney NSW 2010, Australia and School of Philosophical and Historical Inquiry, University of Sydney, NSW 2006, Australia. jspecht@bigpond.com
Table 1. Radiocarbon dates for Ale cave (site FKX), near Kandrian,
West New Britain Province, Papua New Guinea. All samples were
charcoal and were dated by the AMS technique. They were calibrated
using the Int.Cal.04 curve of CALIB 5.0.1 and are cited as 2
sigma ranges.

Context        Lab. No.   [delta][sup.C]

Layer l base   Wk-12113   -27.8 [+ or -] 0.2[per thousand]

Layer 3 top    Wk-12114   -27.2 [+ or -] 0.2[per thousand]

Layer 3 base   Wk-12115   -26.4 [+ or -] 0.2[per thousand]

Context        CRA                Cal. BP at 26   Cal. ranges   Prob.

Layer l base   1307 [+ or -] 49   1307-1089       1307-1167     0.935
                                                  1163-1127     0.046
                                                  1109-1089     0.019

Layer 3 top    1430 [+ or -] 57   1509-1263       1509-1495     0.008
                                                  1490-1466     0.017
                                                  1417-1263     0.975

Layer 3 base   2540 [+ or -] 41   2751-2487       2751-2650     0.386
                                                  2645-2487     0.614

Table 2. Weights (in grams) of chert and 'rubble' by period
at six sites excavated by Pavlides in the Passismanua area.
Data from tables 7.1, 8.1, 9.1, 10.1 and 11.1 in Pavlides
(1999).

Period             Material   Auwa FGT   Eliva FYV   Airstrip FIF

Recent             Chert      674        178         223
                   Rubble     9,037      484         0
Late Holocene      Chert      266        269         71
                   Rubble     3,600      2,245       0
Middle Holocene    Chert      513        3,251       2,406
                   Rubble     15,500     4,032       160
Early Holocene     Chert      14,556     429         474
                   Rubble     299,000    2,372       37
Late Pleistocene   Chert      0          443         83
                   Rubble     0          0           0
Totals             Chert      16,009     4,750       3,257
                   Rubble     327,137    9,133       197

Period             Material   Asiu FYW   Sisisel FGV   Dulago FGW

Recent             Chert      0          340           47
                   Rubble     0          30,765        0
Late Holocene      Chert      57         245           0
                   Rubble     380        46,870        0
Middle Holocene    Chert      1,434      162           6
                   Rubble     4,450      16,500        0
Early Holocene     Chert      238        0             0
                   Rubble     14,225     0             0
Late Pleistocene   Chert      0          0             0
                   Rubble     0          0             0
Totals             Chert      1,728      748           53
                   Rubble     19,055     94,135        0

Period             Material   Totals

Recent             Chert      1461
                   Rubble     40,286
Late Holocene      Chert      907
                   Rubble     53,095
Middle Holocene    Chert      7,772
                   Rubble     40,642
Early Holocene     Chert      15,698
                   Rubble     315,634
Late Pleistocene   Chert      526
                   Rubble     0
Totals             Chert      26,364
                   Rubble     449,657

Table 3. Frequency of non-cortical versus cortical items by
period (number (percent)).at six sites excavated by
Pavlides in the Passismanua area. Data from tables 7.3,
8.7, 9.9, 10.8, 11.10 in Pavlides (1999).

Period        Auwa       Eliva      Airstrip   Asiu      Sisisel
              FGT        FYV        FIF        FYW       FGV

Recent        82 (57)    23 (55)    41 (67)    n/a       56 (81)
Late          27 (51)    2 (2)      3 (13)     6 (29)    14 (35)
Holocene
Middle        36 (46)    138 (36)   130 (23)   28 (41)   3 (100)
Holocene
Early         470 (48)   43 (42)    43 (19)    4 (15)    n/a
Holocene
Late          n/a        4 (33)     7 (41)     n/a       n/a
Pleistocene
Totals        615 (49)   210 (33)   224 (27    38 (33)   73 (65)

Period        Dulago
              FGW

Recent        8 (80)
Late          n/a
Holocene
Middle        0
Holocene
Early         n/a
Holocene
Late          n/a
Pleistocene
Totals        8 (67)

Table 4. Frequency of non-cortical versus cortical items by
mean maximum and axial lengths through the Holocene at
six sites excavated by Pavlides in the Passismanua area.
Data from tables 8.20 and 8.21, 9.24 and 9.25, 10.15 and
10.16, and 11.21 and 11.22 in Pavlides (1999).

Period                Cortex         Maximum
                      No cortex      length

Recent                Cortical       29.6 [+ or -] 16.1
                      Non-cortical   19.7 [+ or -] 7.0
Late Holocene         Cortical       28.1 [+ or -] 15.4
                      Non-cortical   19.1 [+ or -] 7.8
Middle Holocene       Cortical       43.3 [+ or -] 15.7
                      Non-cortical   36.6 [+ or -] 13.5
Early Holocene--FGT   Cortical       58.9 [+ or -] 22.8
                      Non-cortical   38.1 [+ or -] 18.0
Early Holocene--      Cortical       49.9 [+ or -] 8.5
  other sites         Non-cortical   40.1 [+ or -] 14.3
Late Pleistocene      Cortical       n/a
                      Non-cortical   n/a

Period                Axial
                      length

Recent                28.0 [+ or -] 14.0
                      18.5 [+ or -] 6.8
Late Holocene         25.4 [+ or -] 10.5
                      18.7 [+ or -] 8.2
Middle Holocene       34.9 [+ or -] 16.1
                      28.3 [+ or -] 13.7
Early Holocene--FGT   51.9 [+ or -] 20.7
                      32.4 [+ or -] 17.0
Early Holocene--      39.5 [+ or -] 10.1
  other sites         34.7 [+ or -] 14.1
Late Pleistocene      n/a
                      n/a

Table 5. Comparison between selected characteristics of the
obsidian sources of Willaumez Peninsula and the chert
exposures of the Passismanua area.

Source                 Obsidian               Chert
characteristics

Distribution           Concentrated in        Highly dispersed and
                       well-defined 40-50     localized over ~3,000
                       [km.sup.2] area; of    [km.sup.2], within
                       volcanic origin and    specific facies of Yalam
                       formed by specific     Limestone and formed
                       events in the          by long duration
                       Pleistocene and        processes
                       Holocene

Visibility             High, except where     Low in dolines and
                       digging was required   caves, periodically
                                              high in stream beds

Accessibility          Generally good, but    Generally good,
                       occasionally           little direct impact
                       interrupted by         from volcanic
                       volcanic events.       events. Inland only.
                       Both coastal and       Stream beds easy to
                       inland; beach          access but depend on
                       exposures of flows,    erosion, not
                       though flows           reliable in long-
                       difficult to quarry;   term due to flushing
                       boulders available     by floods; seams in
                       on beaches and in      caves, etc
                       gullies, some          accessible only
                       digging for boulders   after partial doline
                       at some inland         collapse
                       locations

Quantity/Use-life      Virtually unlimited    Highly variable, but
                       supplies with          never abundant, many
                       continuous use in      exposures of limited
                       most contexts, but     extent, patchy and
                       locally secondary      unpredictable;
                       contexts may be        limited size
                       restricted; large      potential;
                       boulders common, no    alternation of
                       size restriction;      sources could extend
                       abundance means no     use-life, but need
                       need to search for     to search for new
                       new sources            sources

Quality                Generally highly       Unreliable, cortex
                       dependable with        always present;
                       excellent flaking      nodules need
                       qualities, even        testing, likely to
                       where there are        have thick, chalky
                       compositional or       cortex; seams likely
                       thickness variations   to be jointed, with
                       in a flow; generally   impurities;
                       cortex not a           controlled flaking
                       problem; low wastage   often impossible;
                       rate due to cortex     high wastage
                       removal                relative to quantity
                                              of usable chert

Functional value       Extremely sharp but    Less sharp but with
                       very brittle; edges    stronger edges that
                       not easily re-         can be resharpened
                       sharpened by           by retouch; suitable
                       retouch; best for      for percussive tasks
                       non-percussive tool    and other
                       functions; has         activities, except
                       qualities that         where very sharp
                       transcend purely       edges are required;
                       utilitarian            flaking qualities
                       functions--large       less important than
                       stemmed blades and     size of workable
                       stemmed kombewa        piece; bifacially
                       flake tools used as    flaked stemmed tools
                       a valuables?           used as valuables?

Control                Long-term,             Individual exposures
                       continuing local       controlled by local
                       ownership and/or       groups? Multiple
                       control over access;   controlling groups
                       controls still         across Passismanua?
                       operating today        Fear factor as
                                              control of other
                                              group's exposures?

Trade/exchange value   Very high, ample       Probably high
                       supplies; export       potential, but
                       finished items or      export of finished
                       raw material with      items only as
                       little quality         quality checks
                       checking required;     essential for
                       movement over long     nodules with cortex;
                       distances since late   limited supplies
                       Pleistocene, with      perhaps restricted
                       potential as a         opportunities for
                       highly valued good;    trade, up to 25-30
                       still highly valued    km from sources,
                       today                  from at least middle
                                              Holocene onwards

Impact on land-use     Away from sources,     Less emphasis on
strategies             scheduling and         provisioning
                       planning essential     individuals as an
                       to provision both      exposure is likely
                       individuals and        to be close-by;
                       places, though much    daily replenishing,
                       of the peninsula is    less planning
                       within half a day's    required?
                       walk or canoe ride
                       to a source
COPYRIGHT 2011 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 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Specht, Jim
Publication:Archaeology in Oceania
Article Type:Report
Geographic Code:8PAPU
Date:Jul 1, 2011
Words:10211
Previous Article:Learning the lithic landscape: using raw material sources to investigate pleistocene colonisation in the Ivane Valley, Papua New Guinea.
Next Article:Morphometric analyses of Batissa violacea shells from Emo (OAC), Gulf Province, Papua New Guinea.
Topics:

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