A continental narrative: human environment interactions across Australia using time-series analysis

Abstract

The relationship of prehistoric hunter-gatherers and climatic change has long been a focus of archaeologists and palaeoclimatologists alike. One of the key limitations to these studies has been the dissociation between the spatial, temporal and resolution characteristics of the indices, which makes any correlation difficult. Palaeoclimatic records are commonly continuous, high-resolution and undertaken at regional scales, while archaeological data are discontinuous, have variable resolution, and are often highly localised. This thesis applies time-series analysis to archaeological radiocarbon data as one approach to resolve this disparity. With large enough datasets, the technique can develop a regional and continuous proxy of human activity in the past, and enable direct comparison with palaeoclimatic records. The approach can also be readily applied to archaeological questions associated with demography and societal development through time. The use of radiocarbon data as a proxy is not new, being first applied in the 1980s, however it has proliferated globally in the archaeological literature in the last decade or so. In Australia, the use of radiocarbon data as a hunter-gatherer proxy was first used by Smith and Sharp (1993) and Holdaway and Porch (1995) to undertake a general review of the Sahul archaeological record and explore the nature of occupation in Pleistocene Tasmania, respectively. Since then, in contrast to global trends, the use of 'dates as data' has only been sporadically adopted by researchers to archaeological issues (e.g. Lourandos & David, 1998; Turney & Hobbs, 2006; Ulm & Hall, 1996). This thesis consists of 16 publications, which over the last seven years has systematically explored, applied and tested the technique in an Australian archaeological context. Part of this work included developing a continental dataset of 5,044 radiocarbon ages from 1,748 archaeological sites. Currently, this is the most comprehensive spatial and temporal dataset for any archaeological indices in Australia; and is publically available. Other publications include the interrogation of the dataset at a range of spatial and temporal scales, and compare past human activity with climatic change, most notably the Last Glacial Maximum and El Nino Southern Oscillation. Two key achievements include, the development of protocols and procedures for using time-series analysis (Williams, 2012) - a methods paper that remains widely cited, adopted and/or critiqued in the archaeological time-series literature; and a continental model of demography, including quantitative population estimates, for Australia over 50,000 years (Williams, 2013). Ultimately, the thesis provides a series of continental and regional models of demographic change for researchers to utilise as a framework, ground-truth, and refine into the future. It significantly progresses the use of radiocarbon dates as a mainstream and reliable proxy for exploring hunter-gatherer behaviour in Australia; and provides a number of future directions for both time-series analysis and Australian research more generally. Globally, the analyses undertaken here remain at the forefront of the technique, and are routinely used, applied and critiqued across the world.