The 1980s witnessed major changes in the direction of archaeological enquiry in Australia. Until then, prehistories were largely constructed around static models based on concepts of 'traditions', where socio-cultural dynamics were usually seen in terms of changes from one relatively stable period to another. Viewed through such lenses, socio-cultural innovation was generally treated as a response to external stimuli - migration or diffusion, or as a result of environmental stress. With the works of researchers such as Hughes and Lampert and, more recently, Lourandos, these dominant interpretative frameworks changed. Some of these authors argued that Aboriginal prehistory witnessed significant alterations during the mid to late Holocene, that these changes involved demographic alterations, and that socio-political forces were instrumental to these changes. It was also suggested that these changes may have involved a re-structuring of territorial networks, although no-one had yet investigated this issue with appropriate data at hand.
A major problem nevertheless remained. A number of authors questioned the very data-bases upon which these views were based, hence placing the latter in doubt. These doubts stemmed around the argument that the apparent archaeological changes did not reflect human behavioural changes, but merely the differential preservation of archaeological materials in unstable environmental contexts.
This thesis addresses some of these issues, and is divided into two major components. The first component presents the results of new excavations, and reviews previously published archaeological data, from one part of Australia, southeast Cape York Peninsula. My aims are to determine whether or not the types of changes Lourandos and others propounded are indeed observable from the region's archaeological record. If so, do these changes represent 'real' socio-cultural trends, or are they merely the products of post-depositional taphonomic processes. I conclude that unprecedented socio-cultural changes did take place c.35(X)-2500BP, involving major increases in stone artefact, bone and ochre deposition rates and sedimentation rates. In addition, new site types began to be created, and these involved changes in settlement-subsistence strategies. Furthermore, major increases in cave painting activity are also documented at this time.
Taking this as the starting point for the second component of my thesis, I then ask if these changes also involved major modifications in the structure of prehistoric socio-cultural networks. To address this question, I investigate the distribution of rock art conventions through space and time. I conclude that during the late Pleistocene and early Holocene, rock art was relatively homogeneous, whereas during the mid and late Holocene it was highly regionalised. This regionalisation is evident in at least two levels. Firstly, there is a strong north-south division; to the north of the Walsh-Mitchell Rivers, the rock art is predominantly figurative in form, executed in infill or outline and infill, and the paintings are often created in two colours, with white being rarely used. To the south, on the other hand, die paintings are almost always monochrome, they are predominantly outlined, overwhelmingly consisting of non-figurative and track designs, and white is more common. The change-over between these two geographical blocs is sudden, showing no evidence of clinal change.
Secondly, this broad spatial division is highly regionalised. Within each zone, smaller 'stylistic' sub-regions can be identified. In the Princess Charlotte Bay and Flinders Island Group, moth-butterfly designs and zoomorphs with crescent heads are predominant (such paintings do not occur anywhere else). In die Koolburra Plateau, echidna-human therianthropes are common, but do not appear to occur to the east of the Kennedy River. At Laura, a very broad range of conventions was used, but to the immediate south, between the Palmer and Mitchell Rivers, there is a very limited range of motif forms, with the vast majority of paintings being infilled anthropomorphs showing no evidence of internal elaboration. Macropods are totally absent from this area. At Bare Hill (Davies Creek), relatively small anthropomorphs with upturned arms are common, but again these are very rare elsewhere.
To the south, the Rookwood-Mungana-Chillagoe area contains very large numbers of radiating lines. Grid patterns are common at Ngarrabullgan, while at Lawn Hill, to the west of the study region proper, non-figurative designs contain numerous sinuous lines unlike those found further to the east.
I therefore conclude by suggesting that social networks became highly regionalised during the mid to late Holocene. A social model is presented to explain this process of regionalisation. It is stressed, however, that this thesis primarily aims at exploring the 'inter-regional' structure of the archaeological record, not to test possible explanatory (socio-cultural) scenarios. My concluding model is based on the conviction that social and cultural systems are historical products. To understand such systems - as well as to understand the archaeological record - we must treat social practice as continuously becoming, as resulting from socio-political processes which are themselves both products and producers of the past.
The model presented thus argues that the mid to late Holocene witnessed a regionalisation of socio-cultural networks in southeast Cape York Peninsula and beyond. This process of regionalisation is argued to have involved major changes in the configuration of social landscapes (including inter-regional behaviour), involving a 'closure' of territorial structures, after c.35(X)-2500BP. By 'closure', I mean that social interaction became more formalised. This may have acted as a means of dispute management, in response to increasing populations in north Queensland during the mid to late Holocene.