Investigating relationships between motif forms, techniques and rock surfaces in north Australian rock art

Abstract

As with many other parts of the world, rock art research has blossomed in Australia since the mid-1 980s. There are undoubtedly many reasons for this, including the formation of new national and international rock art organisations and a growing sense of confidence in the archaeological ability to characterise rock art practices in both space and time (these two trends are presumably related). Until recently many archaeologists tended largely to ignore rock art as it could not easily be dated, and therefore could not easily be situated within regional models of prehistory. The advent of AMS radiocarbon dating on tiny amounts of carbon, such as is sometimes available in charcoal drawings, organic binders contained within pigments, and stratified oxalates located either immediately beneath or above rock art, changed this by offering archaeologists a means by which the antiquity of artistic expressions could realistically be investigated (e.g. Ilger et al. 1995, 1996; Van de Merwe et al. 1987; Watchman 1993; Watchman and Cole 1993; but see McDonald et al. 1990). Long before the advent of AMS radiocarbon dating, however, many attempts were made to order the art chronologically by dividing it into stylistic and/or technological sets or by reference to assumed evolutionary trends, and thence arguing for their temporal differentiation with little absolute or relative chronological information (e.g. Walsh 1994); indeed, such endeavours continue to this day (e.g. see discussions in Lorblanchet and Bahn 1993). In Australia, this approach has been articulated at various geographical and temporal scales, such as Maynard's (1979) influential scheme dividing Australian rock art into three time periods, the Panaramitee, simple figurative and complex figurative (from earliest to latest), any given region possessing at least one of these three periods, in the understanding that their relative temporal ordering always stayed the same. More recently, Walsh (1994) has argued for a more complex chronological series of rock art phases for the Kimberley of northwestern Australia. Again, each phase consists of a specified set of artistic conventions, such as the use of certain pigment colours, the depiction of certain items of material culture and the employment of certain brush strokes, but at this stage no evidence has been presented for the temporal relationship of types. Other examples of the typological ordering of images to construct chronological models can be found, both in Australia and elsewhere (e.g. see discussion by Watchman 1996). Our aims here are not to criticise the construction of types, typologies or typological systems, but rather to argue that the ensuing temporal ordering of types should be based on explicitly temporal data (see also McDonald 1994). Some of the work presented here was previously addressed in Lecole (1996). In this spirit, this paper presents new data from Wardaman Aboriginal country in northern Australia, where both techniques and rock surfaces have influenced motif forms. There is as yet no evidence for change in motif forms through time.