Year

2017

Degree Name

Doctor of Philosophy

Department

Centre for Archaeological Science

Abstract

The archaeological record is a time-averaged palimpsest of material variably influenced by postdepositional processes. The aim of archaeology is to elucidate and inform on past human behaviour, however, the palimpsest nature of the archaeological record limits the potential for events on the scale of the individual and day-to-day life to be preserved. While some perceive this lack of temporal resolution as a hindrance, it rather presents an opportunity to investigate environmental and behavioural processes at a larger, broader scale. The delimiting and constraining of palimpsests to access this spectrum of temporal scales poses methodological and conceptual complications. It is this challenge that forms the focus of this thesis.

Shell middens, in many ways, magnify the effects of the palimpsest nature of the archaeological record through aspects such as their porosity and frequent lack of clearly visible stratigraphic differentiation. Complex and variable formation processes blur the spatial and temporal relationships of the material contained within a midden deposit. The vertical displacement of midden shell, the time-averaging of previously temporally distinct layers, as well as the muddying of the depositional patterning behind midden formation are all issues that complicate behavioural and palaeoenvironmental interpretations. The midden within the Brremangurey rockshelter, located in the Kimberley region of Western Australia’s far north, embodies this conundrum. During excavation and analysis of the midden, clues were identified that pointed towards a far more complex formation and transformation history than initially thought. Precisely to what extent, though, would be problematic to determine using conventional techniques.

The typical approach to refining the formation processes of shell middens relies on radiocarbon dating. Greater number of samples provide enhanced resolution, but at a considerable cost. Amino acid racemisation (AAR) is a low-cost relative dating technique that has not been widely incorporated into archaeological investigation. In the context of shell midden archaeology, the potential of AAR in resolving issues of site formation and transformation comes from the ability to analyse a large number of samples to establish a high resolution relative chronological sequence of a midden deposit. Recent refinements to the AAR method improve the technique’s accuracy and precision, making it more amenable to the temporal scales at play within these specific archaeological contexts. To test the applicability of this novel use of AAR, this approach was applied to the material excavated from the Brremangurey rockshelter.

The use of AAR dating to establish a high resolution relative chronology of the Brremangurey midden deposit managed to address a range of problems commonly encountered in shell midden archaeology. The temporality and spatial origins of vertically displaced shell could be recognised allowing the integration of ex situ material to the archaeological interpretations. A time-averaged layer was disentangled, and the relative contributions of each phase of deposition to be identified, adding more detail to the formation history of the midden deposit. Most importantly, the application of AAR and ‘Temporal Packaging’ presented a detailed picture of the depositional patterning of the Brremangurey midden deposit.

A complicating aspect of the archaeological record is that the scales of temporal resolution recorded within a palimpsest are hidden. This denies the investigator the opportunity to appropriately adjust their scale of investigation to the scale of evidence the archaeological record can support. Not only were AAR and Temporal Packaging able to refine the chronology of a midden deposit, but also allowed the previously inaccessible temporal resolution contained within a deposit to be defined.

FoR codes (2008)

210101 Aboriginal and Torres Strait Islander Archaeology

Share

COinS
 

Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.