Bones of contention: reply to Walshe. (Comment).
A recent critique by Walshe of taphonomic analyses made at three key arid zone sites is addressed. We defend our position that the extreme reduction of bone elements in these sites during the Holocene is most likely the product of Aboriginal behaviours rather than natural processes. We argue that the Tasmanian Devil, Sarcophilus, is not implicated in the reduction of bone; indeed there is no convincing evidence that Devils were even present in the Western Desert. We make the case that the likely effects of dingo on these assemblages is not consistent with the observed patterns of bone breakage through time.
Walshe (2000:74) has recently made a case that the highly fragmented and burnt bone assemblages characteristic of the vast majority of Australian archaeological cave sites are indicative of predator activity. The Tasmanian Devil, Sarcophilus, is seen as the "... most significant factor in modification of cultural assemblages" throughout Australia, including the arid zone. With this as her guiding premise she sets about critiquing the faunal studies from the arid zone sites of Puntutjarpa, Intitjikula and Serpent's Glen (Gould 1977, 1996; O'Connor et al. 1998). At these sites we have argued for human agency as a significant contributor to the reduction of the faunal assemblages. Bone reduction is posited to have been a likely consequence of attempting to extract the most in a meat poor environment. Walshe (2000:80) dismisses the `dietary stress hypothesis' and suggests that it was our eurocentric "... expectations of the pre-contact environment ..." that led to its formulation.
In our reply we wish to address the specifics and some of the factual inaccuracies of Walshe's critique in the wider framework of research on the Australian Desert Culture.
The `risk-minimisation model' and the `dietary stress hypothesis' in the context of the Australian Desert Culture
Early expressions of the `risk-minimisation model' (cf. Gould 1974) were based largely on historical and ethnographic data and primary ethnoarchaeological research (Gould 1967, 1969, 1971), and not as Walshe suggests on our own `expectations of a pre-contact environment'.
Multiple studies have emphasised the exceptionally low population densities of Australian desert inhabitants on the global scale, the very high levels of mobility and the vast geographic alliance networks that provided access to neighbouring lands during times of scarcity (see review in Tonkinson 1991). While a nutritionally balanced diet may have been experienced (on the average) by most desert groups (see Smith 2000:65) the overwhelming weight of historical evidence indicates that cyclical regional and long-term droughts constituted unique challenges to desert inhabitants. Extreme forms of residential and territorial mobility were employed to mitigate ensuing resource stress (cf. Gould 1991; Veth 2000). Our recognition of the extremely reduced condition of the bone in cultural deposits from the arid zone sites suggested to us that bone reduction may have been occurring to extract the most from the available protein resources and more particularly to remove fatty bone marrow.
Fatty foods are highly valued by hunter-gatherers from all latitudes not only for their desirable taste but also for nutritional reasons (Jochim 1981). Not only is fat a highly concentrated source of energy but it assists the body to metabolise dietary protein from lean meat (Speth 1983:104). Bone marrow is one of the richest sources of fat, especially on lean species. In times when the animals being hunted are nutritionally stressed it can be particularly important as bone marrow fat is one of the last body fat reserves to be mobilised (Speth 1983:102-104).
The last decade of archaeological research has set out specifically to examine the issue of variation in the productivity of Australian desert landscapes and changing human responses to aridity (cf. Veth et al. 2001; Veth 1993, 1996, 2000; Smith 1988, 1996; O'Connor et al. 1998 and Thorley 1998). It was in the context of this research that we carried out the analyses of the faunal assemblages from Puntutjarpa, Intitjikula and Serpent's Glen.
Bone reduction and carnivore activity
Walshe argues for Tasmanian Devils as the principal agent for bone reduction throughout arid Australia and further, that this explains the bone reduction at Puntutjarpa, Intitjikula and Serpent's Glen. Aside, however, from tooth and maxillary fragments which were so incomplete that no identification could be made other than "large carnivore" at Puntutjarpa (Archer 1977:162), there is no direct evidence for the presence of Sarcophilus or Sarcophilus-related activities at any of the three sites. Archer (1977:162) quotes an extensive zoological literature that in prehistoric times Devils had a "continentally peripheral distribution". Walshe herself admits that there are no sub-fossil records for this species from the Western and Central Deserts. Indeed, Walshe extrapolates her case from evidence for the presence of Tasmanian Devil at Allen's Cave, a site on the Nullabor Plain of South Australia. Not only is this site distant from the Carnarvon Ranges, Warburton and James Range regions, but it also represents a measurably different kind of arid environment from the Western and Central Deserts of Australia. Walshe's (2000:78) assertion that the Tasmanian Devil while "... preferring less arid regions ... were able to tolerate the Western Desert environment until their southern demise around 2000 years ago", is just that--an assertion. At the present time there is no evidence of any sort for the distribution of Sarcophilus extending into the interior arid zone in the Holocene.
We acknowledge that the potential for Devil bones to be `discovered' in cave sites, even when such caves occur in optimal environments for Devils, depends on a variety of factors such as the volume of deposit excavated and the preservation conditions. Even in regions where Devil remains are likely to be naturally deposited and well preserved, such as in the limestone cave regions of Tasmania, they are poorly represented in the archaeological faunal assemblages. At the three Tasmanian cave sites Cave Bay Cave, Mackintosh cave and Kutikina cave where hundreds of thousands of bones were recovered, Tasmanian Devils are represented by only one or two identifiable elements (Bowdler 1984; Geering 1983; McWilliams et al. 1999), and in another cave (Nunamira) with over 44,000 hones recovered they are not represented at all (Cosgrove 1995). However, this is a sampling issue. The point is that even where our excavations are volumetrically small and Devil remains are very sparsely distributed, if we have enough small holes we will pick up a regional signal for this carnivore, and this is precisely what we find in Tasmania. If only one site had been sampled (Nunamira) Devil may have gone undetected, but with multiple sites sampled it is detected despite the fact that its remains are sparse. If we apply the same criteria to the Western and Central Deserts we would make the argument that our sample should be large enough to have recovered Devil remains if this species was active in the region. There are now over 23 sites excavated (some 120[m.sup.2]) in the Central and Western Deserts covering the relevant time period (i.e. the period prior to 2000 BP when the Devil becomes extinct in southern Australia) and as yet there is no direct evidence for this carnivore. Not all these excavations are small holes in large sites (cf. Gould 1977; Thorley 1998; Veth 1993).
This absence of Devil remains in cultural assemblages in the Central and Western Desert cannot be put down simply to their habitat preference for particular types of caves, as Walshe suggests.
Tasmanian Devils prefer smaller confined spaces such as low shelters, ledges and crevices with small openings at the back or burrows rather than the more cavernous spaces. Larger spaces, such as open rockshelters and caves are certainly visited but only to scavenge and leave scats (Walshe 2000:78).
Firstly, it is untrue to suggest that all the arid zone sites that lack Devil remains are open and accessible. The West Cave at Puntutjarpa, which was directly adjacent to the main cave, was exactly the sort of low-ceiling, crevice-like formation that Sarcophilus would presumably have found attractive. However no Devil remains were identified by Archer when he examined the bones from the West Cave deposits.
Walshe suggests that Sarcophilus may not be integrated into "... cultural assemblage(s) through natural mortality due to their specialised use of large cavernous spaces ..." (emphasis ours) (Walshe 2000:78). However many cultural assemblages outside the Western and Central deserts do contain bone remains of Devils (Balme et al. 1978; Bowdler 1984:70; Dortch 1984; O'Connor 1999:46; Rosenfeld et al. 1981:37). Some of these sites, such as Koolan Shelter 2 in the Kimberley, are large, open and accessible shelters lacking ledges, crevices and narrow caverns that would make them suitable as living sites for Devils. An alternative that should perhaps be considered is that the Devil remains in sites like Koolan 2 represent the remains of human prey.
The Koolan 2 shelter has a bone assemblage spanning the periods 26,000 BP-24,000BP and 10,500 BP-630 BP (O'Connor 1999:29) that can be compared with the arid zone assemblages. Although Koolan 2 contains identified remains of Sarcophilus, Thylacine and Dasyurus hallucatus (Northern Quoll), the bone assemblage is much less reduced than that from Serpent's Glen, Puntutjarpa or Intitjikula. Thylacine and Dasyurus hallucatus were also recovered in the Pleistocene and Holocene units and dingo in the Holocene unit of the mainland Kimberley site, Widgingarri Shelter 1 (O'Connor 1999:90). Again, the degree of bone reduction at Widgingarri 1, does not compare with the arid zone sites. Overall, there is a great deal more variability in the size of bone fragments than is seen in the desert assemblages.
There are a number of serious logical inconsistencies in Walshe's argument for carnivore influence on the Puntutjarpa assemblage. Walshe (2000:78) suggests that "... the vertebrate assemblage from Puntutjarpa is primarily the result of reduction by Tasmanian Devils and in more recent times by dingoes". However, in the same paragraph she states that "At Puntutjarpa the most pronounced bone fragmentation rates were recorded in Zone C, which is the oldest and the most likely to demonstrate devil activity". This statement seems at odds with the following statement that "Dingoes tend to produce smaller fragments due to their less powerful jaws and smaller digestive tracts". If this is indeed the case, then using Walshe's reasoning we should expect more pronounced fragmentation in the upper unit after the introduction of dingoes: instead the reverse trend is identified at Puntutjarpa. If, as we believe, Devils were not active in the Western and Central deserts in the Holocene, the bone fragmentation in the lower levels cannot be explained as a result of carnivore reduction. The arrival of the dingo in the upper levels is not registered. As noted, above, there is no evidence of more uniform reduction of bone or reduction in fragment size in the late Holocene. That this did not occur was one factor that led one author (RG) to look beyond carnivores for a taphonomic agent of reduction. If evidence could be found within the Western and Central Deserts for the presence of Tasmanian Devil, the case for carnivore activity as a decisive taphonomic factor would be stronger, since it could potentially account for the extreme degree of bone reduction prior to the arrival of the dingo. On Walshe's rationale however, it would still fail to explain why the bone in lower level of Puntutjarpa evidences the most pronounced reduction.
At Serpent's Glen, almost all fauna occurs in levels post-dating the extinction of Sarcophilus in southern Australia. If predator activity is responsible then, as Walshe suggests, it is presumably the dingo at work. While this is a possibility, again it has to be said that the levels of bone reduction are unparalleled outside the arid zone and significantly more so than at sites where dingoes are definitely recorded in the bone assemblage, such as at Widgingarri Shelter 1 (O'Connor 1999:87).
While we believe that it is entirely plausible that dingoes have had some impact on the cultural assemblage at Serpent's Glen, we question whether it would take the form that Walshe suggests. Pocock's (1988) detailed study of reduction of the Miriwun bone assemblage, east Kimberley, indicated that there was no change in the degree of fragmentation between the Pleistocene and Holocene units, although dingo was recovered in the late Holocene assemblage. Pocock (1988:96,103) did, however, identify a decrease in bone quantity in the late Holocene that she attributed to the effects of this new predator. She based this view on ethnographic' observations and studies of dingo scavenging behaviour. She notes "... that bone is removed from the archaeological record rather than being destroyed in situ" and concluded that "... dingoes are unlikely to contribute to faunal remains in archaeological sites, or to modify humanly accumulated bone in situ" (Pocock 1988:103). This pattern of scavenging was substantiated in Miriwun by changes in the relative contributions of various faunal elements such as the increase in fish bone which she believed was independent of preservation (Pocock 1988:103). She concludes that "... those predators which are known to have been present at Miriwun are unlikely to have played an important role in the accumulation of bone at Miriwun. The highly fragmented bone is, therefore, probably due to factors other than predators" (1988:95).
In this context, it is worth making the point that archaeological cave assemblages are known which have extreme differences in degree of bone fragmentation through time which cannot be put down to the impact of carnivore predators. For example, Liang Lemdubu, a cave site in the Aru Islands, has a rich and abundant faunal assemblage whose derivation is almost entirely attributable to human agency (Ken Aplin pets. comm.; O'Connor et al. in press).
The abundant faunal assemblage that began accumulating about 27,000 years ago, as a result of human occupation, evidences minimal fragmentation. The fauna in the late Pleistocene levels, however, shows much more intense fragmentation and burning. The increased bone fragmentation in these levels corresponds closely with higher burning indices on the bone and greatly increased stone artefact numbers (O'Connor et al. in press).
Walshe's (2000:76) criticises the "... lack of rigorous evidence independently provided by Gould (1996) and O'Connor et al. (1998) ..." in support of a human agent in bone reduction and yet much of her counter argument relies on selective presentation of data as shown above or as personal anecdote. For example, her principal argument against macropod tooth fragmentation (Walshe 2000:79) was that
From personal field observations of weathered macropod teeth, I know that macropod molars can fragment into at least two and possibly more than six sherds.
This anecdotal reference is unsupported by any data on weathering of macropod teeth or weathering per se. Surprisingly it becomes the basis for the subsequent argument that the minimum number of individuals (MNI) produced by macropod tooth fragments at Puntutjarpa represented only one kangaroo per stratum or 17 individuals over roughly 10,000 years. As a kind of `minimalist' proposition, we can accept Walshe's MNI's, but for a different reason. Until Walshe's data on macropod-tooth breakage and reduction become available, we prefer to view this kind of breakage as a `potentially diagnostic' indicator of a relatively meat-stressed diet that is consistent with the ethnoarchaeological hypothesis presented by one of the authors (RG). One does not actually need a representative sample of a process (in this case, human agency to account for extreme tooth reduction) to build or test a hypothesis--all one needs is a good example that illustrates how the process works. Then testing can proceed. As archaeologists who have used them know, MNIs generally have the effect on faunal assemblages of reducing them to the point where no truly representative samples of ancient human exploitation of animal resources are possible for any level or stratum within a site. For this reason macropod tooth reduction was used as another, potentially better (though far from ideal) index that may indicate cultural reduction of bone. Until better data are forthcoming about weathering as a factor in macropod tooth reduction, the cultural `hypothesis' appears to us to be unchallenged.
In short, we believe that the degree of difference in the faunal assemblages at the arid zone sites versus sites outside the arid zone (where the predators in question are actually found in the bone assemblages) calls for a cultural explanation. Smith (2000) has recently published new evidence that "... bone smashing behaviour was also an outcome of culturally determined behaviour which may or may not be linked to nutritional stress". Interviews with senior Aboriginal people who had been living a traditional lifestyle in the Western Desert until the 1930's revealed that bone smashing/crushing behaviour was carried out in Order to empower hunters. "It was reported that the bones of large animals, preferably kangaroos, were crushed and eaten by men" (Smith 2000:56). Smith's report is substantiated by observations made by Elphinstone (1971) who carried out medical examinations on Aboriginal people living in the Western Desert near Warburton and close to the site of Puntutjarpa Rockshelter in 1958 and 1967. Elphinstone (1971:296) observed that the stools he examined contained substantial amounts of fragmented bone. Whether bone crushing and ingestion was due to cultural practices to extract maximum nutritional efficiency or to empower hunters is to some extent irrelevant and the two are not necessarily mutually exclusive. At any rate it is archaeologically untestable as it will never be possible to discriminate past motives from the crushed bones that constitute the archaeological record. Importantly, this evidence supports the cultural practice of smashing bone and, we would argue, a cultural explanation for the bone fragmentation of archaeological assemblages in the Western Desert.
Sample size and what constitutes a valid comparison
We concur wholeheartedly with Walshe that sample size is critical to any attempt to draw comparisons on an inter- and intra-site level. As with most archaeological assemblages we have had problems with sampling, at all levels. As Walshe points out there is the problem of what percentage of a site is constituted by the excavation and how representative of the whole it will be; there is the problem of how we divide our assemblages into units of time and also how we divide our assemblages into analytical units in order to make our comparisons. We have conflated assemblages over periods of thousands of years in order to gain large enough assemblages to create statistically meaningful samples. We feel, however, that Walshe (2000:80) has misrepresented our studies in several important respects when she states that "Three sites set out over a distance of 1,250 km and 23,000 years in time with intermittent and probably unrelated occupation struggle to constitute a valid comparative sample".
Firstly, the sites in question do not cover 23,000 years. Serpent's Glen does not, as Walshe (2000:74) erroneously states, "... indicate intermittent occupation from about 27,500 to 4,700 BP". In fact, the oldest published date from Serpent's Glen is 23,550 BP (0ZB582) from Spit 34. Above this dated level are six culturally sterile spits. The date of 4,700 BP comes from Spit 27, the first unit with evidence of cultural material above the 23,550 BP date. All units above this contain cultural material, but a marked increase in the quantity of material begins at about Spit 16 and increases again above Spit 10. A Modern date was obtained for Spit 8 and on the information provided by Aboriginal elders, this site was used into historic times. From this the authors (O'Connor et al. 1998:19) suggested that far from "... intermittent occupation between 27,500 BP and 4,700 BP", the site was occupied extremely sparsely in the late Pleistocene at about 23,000 BP and was not reoccupied again until the late Holocene when it continued in use intermittently until historic times. In the context of inter- and intra- site assemblage comparability, this is an important point.
While Serpent's Glen has a sparse lower unit dated to approximately 23,000 BP, the bone assemblage is confined almost exclusively to Spits 14 and above (only 0.7 g of bone occurs below this). Therefore the majority of bone dates to the last 4,700 years and in all likelihood, the last thousand years. Intitjikula has a basal date of approximately 3,000 BP (perhaps as much as 5,000 BP) and Puntutjarpa covers the last 10,000 years, with layers that specifically cover the late Holocene. All three sites therefore have faunal assemblages that date to the Holocene and overlap in the late Holocene.
Secondly, while the sites in question are widely spaced in distance they are arguably more geographically comparable than Allen's Cave on the Nullabor Plain. Allen's Cave represents a measurably different kind of arid environment from the Western and Central Deserts of Australia. This factor is important when we consider the ecological tolerances of Sarcophilus and why it may have included the Nullabor, but not the interior regions of the Central and Western deserts, in its range. It is far more tenable, therefore, to compare economic faunal assemblages from the arid core than from a near-coastal setting.
Anthropogenic versus predator activity?
In her critique, Walshe (2000, 74) claimed that
Gould was struck by the high rate of fragmentation and burning in the Puntutjarpa assemblage and suggests five possible non-human factors in bone destruction. For various reasons ... he dismissed these five agents as unsuitable and selected human agency as the primary cause.
This was an odd claim for Walshe to make, because it was exactly opposite to the results of the Puntutjarpa and Intitjikula faunal analyses. Bone reduction at Puntutjarpa suggested that "... surface weathering and crushing due to soil compaction remain as possible factors and may be taphonomically significant (Gould, 1996:79)". Far from dismissing such taphonomic factors, as claimed repeatedly by Walshe, this study sought to identify and systematically rule out such factors. Further detailed studies of faunal assemblages in arid zone archaeological contexts will be needed to resolve these taphonomic issues and to further test the role of faunal reduction by ancient Aborigines.
In short Walshe has no more demonstrated predator activity at these sites than she has unseated our argument for bone reduction behaviour. While we agree that predator action may be responsible for some of the bone reduction at these sites, we are not convinced by the arguments produced by Walshe that this is the case. We also dispute Walshe's statement that "... recognition of specific human behaviour must come from controlled observations where all other influences are excluded" (2000:79). We fail to see how behavioural studies on people can ever address this question as contemporary human behaviour will not provide an analogue for the type of behaviour we are suggesting. We contend that the answer lies in detailed studies on archaeological faunal assemblages such as that discussed by Walshe.
It is possible to identify the impact of predators on bone assemblages even when the assemblages are both burnt and fragmented but this is not a simple issue. Patterns of fragmentation alone will not resolve the question of what predator contributed to the formation of the bone assemblages. Studies of the range of prey species, skeletal element frequencies, frequencies of carnivore tooth marks by taxon and by skeletal element will all be needed to help resolve this issue. Studies done to date show that on their own none of these indices can be used unequivocally to identify the presence or absence of carnivores (Cosgrove 1995; Marshall and Cosgrove 1990).
Walshe's critique of the bone and tooth reduction patterns at Serpent's Glen, Puntutjarpa and Intitjikula raises a larger, and to our mind, an equally important issue. This is whether or to what degree the ancient Aboriginal inhabitants of arid Australia, lived under conditions of scarcity and uncertainty regarding their basic resources; especially the scheduling of food and water resources. Gould's (1977) original formulation of the `Australian Desert Culture' was not, as some critics have argued, an attempt to generalise about the characteristics of post-Pleistocene human adaptations to Australia's arid zone. It was intended, rather, as a `first cut' at defining the hypothesis that post-Pleistocene conditions in Australia's arid zone imposed conditions of resource scarcity and uncertainty. These conditions were met by ancient Aboriginal populations by means of a repertoire of behavioural strategies that relied heavily upon skills that extracted the most value from limited resources combined with risk-minimizing approaches, such as high mobility, to avoid catastrophic mistakes (Gould, 1991; Veth 2000).
Subsequent archaeological research throughout the arid zone (Veth 1993, 1995, O'Connor and Veth 1996; Smith 1988; Thorley 1998) has demonstrated this variability in behavioural strategies. Cultural elements of skill employed by desert Aborigines in locating and/or processing key resources and in avoiding or mitigating risk factors such as scarcity of water supplies during droughts have prevailed throughout this region, albeit in different ways, for at least the last 10,000-12,000 years. Faunal analysis combined with further work on plant remains, such as current research on chemical residue analysis of grindstones from Puntutjarpa, is intended to identify and evaluate direct indicators of palaeodiet among arid-zone Aborigines. This will allow the stress-response, risk-minimising hypothesis to be tested upon archaeologically visible evidence. In other words, the faunal analysis at Puntutjarpa, Intitjikula and Serpent's Glen should be viewed as `works in progress' rather than as final conclusions of the sort claimed by Walshe.
We would like to thank Dr Ken Aplin, CSIRO, Canberra for discussing several of the points raised in this paper. Jim Allen kindly lent us unpublished reports from his library. We also wish to acknowledge the two anonymous referees who drew out attention to some relevant publications, tempered the tone of our reply and made some very useful suggestions.
Archer, M. 1977 Faunal remains from the excavation at Puntutjarpa Rockshelter. In R. Gould, Puntutjarpa Rockshelter and the Australian Desert Culture Section VIII, pp 158-165. Anthropological Papers of the American Museum of Natural History 54.
Balme, J., Merilees, D. and K. Porter 1978 Late Quaternary mammal remains, spanning about 30,000 years, from excavations in Devil's Lair, Western Australia. Journal of the Royal Society of Western Australia. 61:33-65.
Bowdler, S. 1984 Hunter Hill, Hunter Island. Terra Australis 8. Canberra: Department of Prehistory, Research School of Pacific Studies, The Australian National University.
Cosgrove, R. 1995 The Illusion of Riches: Scale, Resolution and Explanation in Tasmanian Pleistocene Human Behaviour. Oxford: BAR International Series 608.
Dortch, C.E. 1984 Devil's Lair: A Study In Prehistory. Perth: Western Australian Museum.
Elphinstone, J.J. 1971 The health of Australian Aborigines with no previous association with Europeans. Medical Journal of Australia 2:293-302.
Geering, K.E. 1983 Preliminary Analysis of the Faunal Remains for Kutikina Cave, Franklin River, 1981 Excavation. Unpublished Report to the Tasmanian Parks and Wildlife Service, Hobart.
Gould, R. A. 1967 Notes on the hunting, butchering, and sharing of game among the Ngatatjara and their neighbours in the West Australian desert. The Kroeber Anthropological Society Papers 36:41-66.
Gould, R.A. 1969 Yiwara: Foragers of the Australian Desert. London: Collins.
Gould, R.A. 1971 The archaeologist as ethnographer: A Case from the Western Desert of Australia. World Archaeology 3:143-77.
Gould, R.A. 1974 The Australian Desert Culture. Paper presented at the annual meeting of the American Anthropological Association, Mexico City.
Gould, R.A. 1977 Puntutjarpa Rockshelter and the Australian Desert Culture. Anthropological Papers of the American Museum of Natural History 54.
Gould, R.A. 1991 Arid-land foraging as seen from Australia: Adaptive Models and Behavioural Realities. Oceania 62:12-33.
Gould, R.A. 1996 Faunal reduction at Puntutjarpa rockshelter, Warburton Ranges, Western Australia. Archaeology in Oceania 31:72-86.
Jochim, M.A. 1981 Strategies for Survival. New York: Academic Press.
Marshall, B. and R. Cosgrove 1990 Tasmanian Devil (Sarcophilus harrisii) scat-bone: signature criteria and archaeological implications. Archaeology in Oceania 25:102-113.
McWilliams, R., Allen, J., Cosgrove, R. and S. Holdaway 1999 Southern Forests Archaeological Project Data Base, Volume 3 of the SFAP. Melbourne: CD-ROM, School of Archaeological and Historical Studies, La Trobe University.
O'Connor, S. 1999 30,000 Years of Aboriginal Occupation in the Kimberley, Northwest Australia. Terra Australis 14. Canberra: RSPAS, Australian National University, Archaeology and Natural History and Centre for Archaeological Research Publications.
O'Connor, S. and P. Veth 1996 A preliminary report on recent archaeological research in the semi-arid/arid belt of Western Australia. Australian Aboriginal Studies 2:42-50.
O'Connor, S., P. Veth and C. Campbell 1998 Serpent's Glen: a Pleistocene archaeological sequence from the Western Desert. Australian Archaeology 46:12-21.
O'Connor, S., Aplin, K., Spriggs, M., Veth, P. and L.A. Ayliffe In press From savannah to rainforest: changing environments and human occupation at Liang Lemdubu, the Aru islands, Maluku, Indonesia. In A. P. Kershaw, David, B. and N. Tapper (eds) Bridging Wallace's Line: Environmental and Human History and Dynamics of the Southeast Asian-Australian Region 34:270-306. Advances in Geoecology series. Catena Verlag Reiskirchen, GMBH.
Pocock, C. 1988 An analysis of the faunal remains from Miriwun Rockshelter, Ord Basin, East Kimberley. Unpublished Honours thesis. Department of Archaeology, The University of Western Australia, Perth.
Rosenfeld, A., Horton, D. and J. Winter 1981 Early Man in North Queensland. Terra Australis 6. Canberra: Department of Prehistory, Research School of Pacific Studies, Australian National University.
Smith, M. A. 1988 The Pattern and Timing of Prehistoric Settlement in Central Australia. Unpublished PhD Thesis, University of New England, Armidale.
Smith, M. A. 1996 Prehistory and human ecology in central Australia: an archaeological perspective. In S. R. Morton and D. J. Mulvaney (eds) Exploring Central Australia: Society, Environment and the 1894 Horn Expedition, pp. 61-73. Chipping Norton, Surry Beatty and Sons.
Smith, P.A. 2000 Dietary stress or cultural practice: fragmented bones at the Puntutjarpa and Serpent's Glen Rockshelters. Australian Archaeology 51:65-66.
Speth, J.D. 1983 Bison Kills and Bone Counts: Decision Making by Ancient Hunters. Chicago: University of Chicago Press.
Thorley, P. B. 1998a Shifting location, shifting scale: a regional landscape approach to the prehistoric archaeology of the Palmer River catchment, central Australia. Unpublished PhD Thesis, Northern Territory University, Darwin.
Tonkinson, R. 1991. The Mardu Aborigines: Living the Dream in Australia's Desert. Fort Worth, Holt, Rinehart and Winston.
Veth, P. 1993 Islands in the Interior: The Dynamics of Prehistoric Adaptations within the Arid Zone of Australia. International Monographs in Prehistory 3. Ann Arbor, Michigan.
Veth, P. 1995 Aridity and settlement in North West Australia. Antiquity 69:733-46.
Veth, P. 1996 Current archaeological evidence from the Little and Great Sandy Deserts. In P. Veth and P. Hiscock (eds) Archaeology of Northern Australia: regional perspectives, pp. 50-65. Tempus, Archaeology and Material Culture Studies in Anthropology No. 4, Anthropology Museum, The University of Queensland, St Lucia.
Veth, P. 2000 Cycles of aridity and human mobility: risk-minimization amongst late Pleistocene foragers of the Western Desert, Australia. Paper presented to the Annual Meeting of the Society for American Archaeologists, Philadelphia.
Veth, P., Smith, M. and M. Haley 2001 Kaalpi: the archaeology of a sandstone outlier in the Western Desert. Australian Archaeology 52:9-17.
Walshe, K. 2000 Carnivores, taphonomy and dietary stress at Puntutjarpa, Serpent's Glen and Intitjikula. Archaeology in Oceania 35:74-81.
RAG: Department of Anthropology, Brown University, Providence, Rhode Island, USA. Richard_Gould@brown.edu SO'C: Division of Archaeology and Natural History, Australian National University, Canberra, ACT 0200 email@example.com PV: School of Anthropology, Archaeology and Sociology, James Cook University, Townsville, Queensland 4811. Peter. Veth@jcu.edu.au
|Printer friendly Cite/link Email Feedback|
|Author:||Gould, R.A.; O'Connor, S.; Veth, P.|
|Publication:||Archaeology in Oceania|
|Date:||Jul 1, 2002|
|Previous Article:||Late Pleistocene human occupation of inland rainforest, Bird's Head, Papua. (Research Reports).|
|Next Article:||Australian Archaeologist: Collected Papers in Honour of Jim Allen.|