This paper reviews the late Holocene archaeology of Central Australia. The last 1500 years saw significant changes in the archaeological record in this part of the Australian arid zone, with shifts in settlement pattern, site histories, resource use, tool inventories and rock art. Much of the evidence points to regional population growth, beginning 1500-1000 cal. BP and coinciding with expansion of summer-rainfall grassland and more frequent palaeoflood events. Hunter-gatherer groups appear to have increased their use of marginal or outlying areas as these became seasonally accessible. Responses to the demographic changes, especially in the better-watered ranges, include more extended occupation of existing sites, more processing of acacia and grass seeds, and an increase in territoriality reflected in the greater differentiation of rock art complexes after 1500 cal. BP. The archaeological changes are not scaled commensurately with the modest environmental shifts at this time, indicating that human-environment interactions were not linear. A human-environment threshold may have been breached 1500-1000 years ago, with existing socio-economic or historical factors acting to amplify the effects of small environmental changes. However, it remains difficult to fully characterize the nature of these human-environment interactions, despite the fine-grained archaeological record now available. An unresolved problem for this emerging picture of climatic amelioration and population growth is that Aboriginal settlement in Central Australia was expanding at a time when ENSO-driven variability appears to have been at its highest.

One of the largest concentrations of rock paintings in Australia is found in the rugged Kimberley region in the northwest of the continent. A temporal sequence of visually distinctive figurative styles is presumed to span periods of cultural change and major climatic events. As the nature and course of these changes are poorly understood, this paper investigates the relationships between continuity and change in the stylistic attributes of the selected anthropomorphic figures in the rock art assemblage. Some previous Kimberley rock art researchers have argued for an abrupt discontinuity in the art assemblage between the Wararrajai Gwion (the most recent of the Gwion styles) and Painted Hand Periods (formally Clothes Peg Figure and Clawed Hand Periods respectively), while others have argued for more gradual change. Based on the study of 204 rock art sites from 15 site complexes, which included a total of 7,579 motifs and 3,685 identifiable anthropomorphic figures, we identify the core characteristics of anthropomorphic figures in each of the established stylistic periods and show that there is no evidence to support notions of an abrupt discontinuity of art through time. Rather, attribute preferences changed gradually, existing as clades of variation rather than discrete units, with identifiable threads of continuity and periods when certain attributes (core characteristics) are preferentially adopted. A quantitative analysis supports our interpretation.

The continuing discussion about Indigenous perspectives and understandings of rock-art has implications for archaeological research, especially in Central Australia where rock-art is a significant part of Aboriginal life today. We present an Indigenous view of outcomes of research projects generally and an outline of acurrent collaborative rock-art study in Central Australia. We investigate the context in which rock-art was produced in order to develop an understanding of the relationship between the landscape, the motif assemblages and other past human activities.

In October 1962 Norman Tindale was fl own to the Geosurveys Hill area, deep in the Simpson dunefield (Figure 1), to follow up reports of prehistoric occupation exposed on interdunal pans in the area (Anon. 1962). Tindale, then at the South Australian Museum, had been invited to make the trip by Reg Sprigg, Managing Director of Geosurveys of Australia Ltd, one of several companies prospecting the Simpson Desert for oil and gas in the 1960s (Sprigg 1993). Geosurveys staff had noticed that 'long lines of stones on a claypan disappear under sand hills' (Sprigg 1993). Left alone in the desert in the late afternoon, Tindale, then nearly 62, was a hardy, self-reliant field archaeologist: ‘I took stock of my camp,' he wrote in his journal, 'got together some firewood against the night, chose a place to sleep and then made a hasty reconnaissance of the claypan.' Later he 'fed on chops grilled in ashes, made tea and then with a flashlight searched for and found stone implements ... on the claypan' (Tindale 1962:7). Although he had planned for several days of fieldwork, Tindale only had a few hours of daylight in the area as the plane returned the next morning to collect him before impending rain made local clay pans too soft to land on.

The earliest accounts of Central Australian rock art were recorded in the journals of the explorer Ernest Giles, who set out in 1872 to explore the unknown interior of Australia. Mounted on horses, he and his two companions followed the dry sand bed of the Finke River into the rugged and relatively well watered central ranges. before turning south-west across the mulga-covered plains which run west to the sand dune country where surface water is found only after rains. Several months into the trip, after days without water, the party's progress was impeded by a vast salt lake whose surface proved treacherously boggy. forcing the riders to turn hack to what Giles quaintly recorded as 'a little pass and glen where we knew that water was to be got' (1995:49). The water, described as 'thick and dirty with a nauseous flavour', was found by digging a deep trench into the sand (Giles 199):50). Subsurface water seeped slowly into the bottom of the trench leaving lime for the men to explore their surroundings as they waited to water their horses.

Distinctive mulberry paintings found in northern Australia, particularly those of the Kimberley region, have been argued to represent some of the oldest surviving rock art on the continent. Significant research efforts continue to focus on resolving the age of these motifs, but comparatively little attention has been given to understanding their physical composition and potential source(s). In a pilot investigation, we conclude that (at least) two mineralogically distinct mulberry pigments occur in 'Gwion' motifs and demonstrate that their major components can be indicatively chemically differentiated, non-invasively. Characterization of a 'quarried' mulberry ochre source demonstrates that these pigments occur locally as natural minerals.

The archaeology of Glen Thirsty, a desert well in the Amadeus Basin, Central Australia, illustrates the changing relationship between the ranges and desert lowlands during the last 1500 years. Historical records and Aboriginal accounts of the site document the regional importance of Glen Thirsty as one of the few wells in this part of the desert. Archaeological excavations and rock art research show that despite its proximity to Puritjarra with its long, late Pleistocene record of occupation, Glen Thirsty only became an important focus of occupation after 1500 BP. Several lines of evidence independently suggest the establishment and consolidation of a new cultural and economic landscape in the Glen Thirsty area around this time. Growing population pressure and shifts in patterns of land-use and economy in the Central Australian ranges may have provided the impetus for more intensive use of the Glen Thirsty area, although the timing of this was constrained by climatic factors. As a rain-fed well in the lower part of the Amadeus Basin, Glen Thirsty is sensitive to shifts in palaeoclimate and its history reflects changes in regional rainfall patterns during the late Holocene.

Palaeoclimatic research suggests that the period between 13 000 and 6000 BP is likely to have seen some of the most favourable conditions for hunter–gatherer settlement in the interior of the continent. The northern monsoon system reestablished itself around 13 000 BP, creating a seasonally wetter regime with greater runoff and denser vegetation than that which prevailed during much of the mid-late Holocene (Wyrwoll and Miller 2001; Hesse et al. 2004). This period also appears to have lacked the high inter-annual or inter-decadel variability that characterises the current ENSO cycle, which began around 4000– 5000 BP (Gagan et al. 2004; Moy et al. 2002). We could expect to find evidence, at this time, for an expansion and some intensification of Aboriginal settlement in central Australia — but the period is poorly documented in archaeological sequences across arid Australia. In central Australia, only two sites of this age have been located: Puritjarra, where use of the rockshelter intensified and diversified after 13 000 BP and again after 7500 BP (Smith 2006); and Kulpi Mara, where there is a discrete occupation layer dating between 12 000 and 13 000 BP (Thorley 1998a). Further afield, Puntutjarpa rock-shelter, near Warburton, has a rich occupation layer dating to 6000–7000 BP with more ephemeral traces of occupation at 10 200 BP (Gould 1977). We report excavations at NEP22, a small site on the Watarrka plateau (now part of Watarrka-Kings Canyon National Park) (Figure 1), that shows the presence of chipped stone artefacts in a buried sand plain at 11 440 BP. These finds are sparse, but confirm human use of the relatively well-watered country in the George Gill Range, a key drought fall-back region for desert populations in the ethnographic period.

A new find of an engraved 'archaic face' in the Toomba Range, on the northeastern edge of the Simpson Desert, provides additional evidence for the production of these distinctive motifs on the eastern side of the arid zone (Figure 1). This supplements an earlier report of an engraved archaic face at Carbine Creek, 100km to the northeast of the Toomba Range (Morwood 1978, 1985). Together, these two engravings with characteristic bas-relief facial features extend the known distribution of archaic faces and suggest that sometime in the past people shared aspects of a common visual vocabulary across the entire breadth of the arid zone, north of the Tropic of Capricorn.

The production of rock art by Indigenous Australians continued throughout recent centuries, often documenting the arrival of Asians and Europeans in various parts of Australia, but this very recent rock art has until now not been explored in detail from a continent-wide perspective. In this chapter, we outline the nature of this imagery, the issues associated with identifying contact period rock art, and the results of case studies in Wollemi National Park, near Sydney (New South Wales), the Pilbara region, near Roebourne (Western Australia), the deserts and ranges of Central Australia, and Western Arnhem Land (Northern Territory).