Technological change in Wardaman Country: a report on the 1999 field season. (Research Report).

Clarkson, Chris

Introduction

In this paper I report on the aims, field procedures and preliminary results of four months of archaeological fieldwork conducted in 1999 in Wardaman Country in the Northern Territory. The study region is located among the black-soil plains, sandstone outcrops and mesas of the semi-arid zone, about 120 km southwest of Katherine (Figure 1). The project focuses on documenting changes in techno-logical provisioning strategies employed by hunter-gatherers inhabiting this region over the last 10,000 years. This latest phase of fieldwork builds on an earlier season and forms part of a doctoral thesis at the Australian National University.

[FIGURE 1 OMITTED]

Project aims

The purpose of the 1998 and 1999 field seasons was to gather data to further develop models of changing Aboriginal land-use in northern Australia over the last 10,000 years. Analysis of stone artefact sequences from several rockshelters in Wardaman Country has already revealed major changes in technology over this period, including changes to the organisation of procurement and the introduction of new forms of implement manufacture (Attenbrow et al 1995; Clarkson and David 1995; Cundy 1990; Mulvaney 1969; Sanders 1975).

This project aims to understand these changes in terms of the ecology of hunter-gatherer land-use, and to develop understanding of the strategic role of technological organisation in Wardaman Country. This involves understanding the way in which people bridged the gap between the constant need for tools and the scheduling and distribution of subsistence opportunities (Kuhn 1995). Understanding technological change therefore involves exploring the various ways in which people provisioned them-selves and certain key locations in the landscape with raw materials to maintain a constant supply of tools and/or tool-making potential.

In documenting changing land-use practices, the project aims to incorporate multiple lines of evidence, such as changes in reduction technology, resource acquisition, procurement, use and discard, as well as changes in foraging range, diet breadth, and intensity of site occupation. Understanding the spatial dimensions of behaviour is an integral aspect of this research, and aims at characterising change within the context of the broader foraging and land-use system. The approach adopted has so far involved analysis of spatially segregated rockshelter deposits, open site surveys and GIS analysis, phytolith analysis, construction of generalised reduction sequences, and stone sourcing studies. Some of the procedures employed in each of these facets of the project, and some preliminary results, are discussed below.

Field surveys

Spatial analysis of stone artefact manufacture was undertaken to understand the admixture of different provisioning strategies implemented in different environments. This requires understanding how the economics of stone tool production varied, given different constraints on raw material utility and abundance across the landscape.

The 1999 field surveys managed to complete a total of 12 one-kilometre-square strategically located survey quadrats, resulting in the location of 188 sites and the recording of a wide range of attributes on over 4000 stone artefacts. The 1999 season augmented an earlier one in which 65 km of transects were surveyed, 109 sites located and 5920 artefacts recorded. Survey quadrats and transects obtained a sample of land-use practices and reduction behaviour for the four distinct land units that exist in the study area. Multiple collections of knapping floors were also made for analysis of reduction sequences.

Survey quadrats were placed at differing distances to critical resources, such as permanent water, shelter, lithic material, and level ground above seasonally inundated watercourses. The study also identified areas of differential abundance of raw materials and hence differing degrees of stone artefact transport, reduction and curation. Two types of zones were identified: supply zones, where raw materials were procured, and receiver zones, into which raw materials were imported. This was an important stage in identifying different segments of land-use systems and the differing provisioning strategies employed across the landscape.

A set of up to 28 attributes designed to document the changing morphology and reduction strategies was measured on each artefact located in the surveys. A minimum of 100 artefacts were recorded in each unit, with up to several thousand artefacts measured in a single quadrat. In cases of extreme abundance, sites were sampled by measuring artefacts within a number of 2x2 m squares positioned over high-density scatters within the site (Figure 2).

[FIGURE 2 OMITTED]

Excavations

Three stratified rockshelters were excavated in the study area as part of the 1999 field season. These were excavated to add to the sample of rockshelters previously excavated (Garnawala 1 and 2, Yiwarlarlay, Gordol-ya, Mennge-ya and Nimji (Ingaladdi)). The new rockshelters chosen for excavation were positioned next to major water sources or next to sources of flakeable stone in order to examine the effects of differential access to critical resources on knapping strategies, raw material rationing and the curation of stone tools. Rockshelters were chosen for excavation on the basis of their size and potential for stratified sub-surface deposits.

The first rockshelter (Jugali-ya) was located in a deep sandstone gorge, close to a spring-fed waterhole and about 10 km from a major jasper quarry. A 1x1 m excavation was undertaken in the centre of the floor below paintings of anthropomorphic figures and depictions of the major animal totems of the site (Figure 3). The excavation was over 1 m deep and yielded 8729 stone artefacts, including a wide range of rare stone-tool types and evidence of manufacturing activities. Analysis is presently underway on the stone artefacts from this site. Four dates have so far been obtained to suggest a mid-to-early Holocene age for the site. Phytoliths have also been extracted from one square and are currently being analysed by Lynley Wallis. This analysis aims to produce a sequence of local environmental changes, but it may also reveal information about the use of various plants at the site.

[FIGURE 3 OMITTED]

The second site (Gindun) excavated is located within a vast quartzite quarry, conservatively estimated at containing over two million artefacts. The rockshelter sits immediately beneath the quartzite ridge upon which most of the quarrying activity has taken place (Figure 4). Excavation of two 1x1 m pits was undertaken to a depth of over a metre. Twelve well-defined stratigraphic units exist. Four radiocarbon dates suggest the site is only 1000 years old, but the site is nevertheless of great value in defining the nature of major late-Holocene changes in stone artefact manufacture. It is particularly useful in dating the emergence of the production of the long quartzite blades that were a feature of long-distance trade networks in recent times--an industry that as yet remains undated (Allen 1997; Paton 1994). Work has just commenced on sorting and analysing this site.

[FIGURE 4 OMITTED]

The third site (Gidglee) is also located in a sandstone gorge close to permanent water but about 20 km from a stone source (see Figure 4). The shelter is a narrow overhang with rock-art and bush food plants in abundance. A 1.5 x 0.5 m trench was excavated to over 1 m in depth until massive rubble was reached. Seven stratigraphic units are apparent. A date of 1300BP was obtained for a hearth two-thirds of the way down the deposit. The site is unlikely to be much older than 2000 years but will nevertheless help to document spatial and temporal changes in provisioning in the region. Only very preliminary analysis has so far been carried out on this site.

Conjoining

Ten large conjoin sets were collected from four different raw materials. The conjoin sets were found lying together as discrete knapping floors on fairly stable sedimentary surfaces. In most cases, preliminary refitting was employed in the field to establish the potential of the collections to yield comprehensive core reconstruction. In other cases, the potential of the collections awaits determination. The conjoin sets were all located in association with a raw material source and so will be more instructive of the initial procurement of stone flakes than of later stages of core reduction undertaken at greater distances to stone sources. As cores become increasingly rare with distance from source, conjoin sets were not found outside of the quarries themselves. The existing conjoin sets should nevertheless offer a picture of the complete manufacturing process in the form of single core flaking episodes.

Sourcing

PIXE/PIGME analysis was undertaken by Glenn Summerhayes at the Australian Institute of Nuclear Science and Energy at Lucas Heights in Sydney. Thirty chert samples collected from around the study region were submitted for quantification of elemental composition. The purpose of this analysis was to characterise variability between the different sources of chert in the region and to determine whether such techniques might reliably distinguish artefacts deriving from each of those sources in future. Results of these tests have been obtained but have so far only undergone preliminary analysis. If different sources may be reliably distinguished using this technique, further samples may be submitted from archaeological contexts to determine the foraging range of prehistoric groups and the nature and degree of curation of artefacts deriving from each of those sources.

Preliminary results

Preliminary analysis of the survey data and excavated stone artefacts is revealing major changes in provisioning strategies employed by prehistoric knappers over both time and space. Furthermore, changes in such strategies employed through time were not uni-directional but changed back and forth in relation to major fluctuations in environmental variability and aridity. From preliminary analysis of three excavated assemblages, it appears that this occurred as a gradual shift, from an early-to-mid-Holocene provisioning of places with cores at around 4000-5000 years ago, to the provisioning of individuals with transportable tool-kit by around 3000 years ago (Kuhn 1995). This was followed by a partial reversion back towards the provisioning of places by the time European materials are beginning to appear in the deposits.

This shift had several steps, beginning with an initial period of raw material supply in the form of cores to the sites. This was followed by a change towards the increased transportation of cores away from sites, although core rotation still occurred at these sites. Following this, flakes began to be transported instead of cores, and their use lives extended, as indicated by an increase in retouching. Subsequently (and perhaps following a short occupational hiatus), new modes of flake production and shaping appeared at around 3000BP, involving the introduction of points and blades and the recycling of broken flakes as a form of raw material procurement and tool reshaping. The appearance of these components suggests an increased reliance on highly standardised, portable and multifunctional tool-kits.

The last step in the process involved a reintroduction in the provisioning of sites with cores and a decline in the use of portable tool-kits. It is tentatively suggested that climatic change may have been a major factor stimulating these changes in land-use and provisioning. Climate change appears to have taken the form of a period of increased variability and decline in effective precipitation between about 3000 and 3500 years ago, as documented in pollen cores for northern Australia. This period of aridity and fluctuation is generally argued to be associated with the onset of ENSO (El Nino/Southern Oscillation) conditions and a weakening of the northern monsoon (Markgraf et al 1992; McGlone et al 1992; Schulmeister and Lees 1995). As aridity increased it is possible that levels of risk associated with subsistence failure resulted in an increase in residential mobility. Only once climatic amelioration took place at around 1500BP did things relax towards a state of relative sedentism.

The survey data collected in 1998 and 1999 also provide information about changes in the provisioning strategies employed within different land-use segments. For example, in zones of abundant stone material and permanent water, artefact production tends to be unstandardised, with reliance on core production and little conservation of materials or curation of tool-kit components such as points, tulas and pointed blades. In these zones there appears to be a greater reliance on place provisioning, often involving the importation of materials located at great distances from the site. In those zones where stone material and water are scarce, on the other hand, stone artefacts tend to be highly curated, recycled and resharpened. Cores are rare and transported items tend to be lightweight and standardised in form. Assemblages at such locations are more in keeping with individual provisioning strategies.

In conclusion, the 1999 field season was successful in locating new sites of tremendous potential in understanding changing land-use and provisioning in Wardaman Country. Work is well under way on the analysis and write-up of results, and it is hoped that syntheses of this material will soon appear in a more substantial form.

ACKNOWLEDGMENTS

Special thanks to AIATSIS for generously funding two field seasons and subsequent radiocarbon dating. Thanks also to the Wenner-Gren Foundation for Anthropological Research for partly funding the 1999 season. This work could not have been accomplished without the help of Peter Hiscock, Annie Clarke, Alex Mackay, Kelvin Hawke, Garry Estcourt, Darren Rousel, Catriona Murray, Bill Harney, July Blootcher, Lilly Ginginna and Mick Pearson. Thanks also to Graeme Ward for the encouragement to submit this report.

REFERENCES

Allen, H. 1997 The Distribution of Large Blades: Evidence for Recent Changes in Aboriginal Ceremonial Exchange Networks. In P. McConvell and N. Evans (eds), Archaeology and Linguistics: Aboriginal Australia in Global Perspective, Oxford University Press, Melbourne.

Attenbrow, V., B. David and J. Flood 1995 Mennge-ya and the Origins of Points: New Insights into the Appearance of Points in the Semi-arid Zone of the Northern Territory, Archaeology in Oceania 30, 105-19.

Clarkson, C. and B. David 1995 The Antiquity of Blades and Points Revisited: Investigating the Emergence of Systematic Blade Production South-west of Arnhem Land, Northern Australia, The Artefact 18, 22-43.

Cundy, B.J. 1990 An Analysis of the Ingaladdi Assemblage: Critique of the Understanding of Lithic Technology, PhD thesis, Australian National University.

Hiscock, E 1994 Technological Responses to Risk in Holocene Australia, Journal of World Prehistory 8, 267-92.

Kuhn, S. 1995 Mousterian Lithic Technology, Princeton University Press.

Markgraf, V., J.R. Dodson, A.P. Kershaw, M.S. McGlone and N. Nicholls 1992 Evolution of Late Pleistocene and Holocene Climates in the Circum-South Pacific Land Areas, Climate Dynamics 6, 193-211.

McGlone, M.S., A.P. Kershaw and V. Markgraf 1992 El Nino/Southern Oscillation Climatic Variability in Australasian and South American Paleoenvironmental Records. In H.F. Diaz and V. Markgraf (eds), El Nino: Historical and Peleoclimatic Aspects of the Southern Oscillation, Cambridge University Press, Cambridge, 435-62.

Mulvaney, D.J. 1969 The Prehistory of Australia, Penguin, Ringwood.

Paton, R. 1994 Speaking through Stones, a Study from Northern Australia, World Archaeology 26, 172-84.

Sanders, B. 1975 Scrapers from Ingaladdi, MA (Qual.) thesis, Australian National University.

Schulmeister, J. and B.G. Lees 1995 Pollen Evidence from Tropical Australia for the Onset of ENSO-dominated Climate at c. 4000 BP, The Holocene 5, 10-18.

Chris Clarkson
School of Archaeology and Anthropology
Australian National University
[email protected]