Prehistoric Living-Floors and Evidence for them in North Queensland Rockshelters

Abstract

The study of prehistoric living-floors is a comparatively new endeavour in Australia (e.g. see comments in Shawcross and Kaye 1980:120; Mardaga-Campbell and Campbell 1985:105). This is doubtless because most excavations in this country so far have generally been rather small or narrow soundings. In rockshelters especially, large excavations extending to more than 50% of the site are very rare (e.g. Yarrar Shelter, Flood 1970), but a few open-air excavations have uncovered a much larger proportion of site areas (e.g. 100% at Sundown 1, Ranson 1980). However, these excavations were not aimed at the detection and full recording of living-floors. Usually, the norm has been soundings aimed at establishing chronological sequences for various human adaptations and cultural change rather than variations in patterns of behaviour within given sites for specific occupation episodes. In order to find comparative information on living-floors in the Australasian region, one has to turn to recent ethnoarchaeological studies (e.g. Gorecki 1985 at Kuk, Papua New Guinea), even though in terms of data most of these are not directly relevant to the study of living-floors left by Aboriginal Australians. Unfortunately, ethnographic and ethnoarchaeological observations on Aboriginal living-floors and/or activity areas at living sites have only rarely been published (e.g. Hale and Tindale 1933; Gould 1971; Hynes and Chase 1982; Meehan 1982:112-18; Spurling and Hayden 1984). This paper is in part a preliminary report on the early stages of my study of prehistoric living-floors in North Queensland. The project described here aims essentially at investigating various methods of recognising, recovering and analysing Aboriginal living-floors in specific rockshelter conditions (e.g. Pillar and Walkunder Arch Caves, near Chillagoe). It should be pointed out that although the excavation and analysis of living-floors generally involve total or near-total removal of the archaeological layers which contain them, as well as any layers overlying them, I certainly do not advocate this to be carried out systematically and indiscriminantly at all sites in Australia or elsewhere. At this stage, a major part of my research at Pillar Cave is based on extensive area excavations using precise coordinate recording techniques and on a 'palaeoethnographic' analysis (cf. Leroi-Gourhan and Brezillon 1972; Van Noten 1978) of the possible remains of living-floors. It should also be pointed out that the present account does not attempt to present a synthesis on theoretical and methodological aspects of living-floor archaeology. However, as research on living-floors in North Queensland, and indeed in Australia, is still at a pioneering stage, it is advocated that quantitative methods of spatial analysis be tested on fully stratigraphically controlled and rigorously provenanced samples (see also opinions expressed by Johnson 1984:76; Kroll and Isaac 1984:21). The efficiency of statistical methods of spatial analysis also needs to be tested in those specific rockshelter conditions where there appears to be a strong chance of finding living-floors. Furthermore, and as pointed out by Johnson (1984:77), statistical spatial analyses deal with 'discrete' artefacts 'which can be counted as individuals' rather than with generally unquantifiable structures such as hearths, postholes, conjoin networks, orientation of individual specimens, etc.. Statistical methods of spatial analysis will therefore be considered as complementary rather than alternative methods to more 'traditional' archaeological methods such as three-dimensional recording.