Degree Name

Doctor of Philosophy


School of Earth and Environmental Sciences


This thesis addresses the use of grinding stones and fragments in Australia through an integrated use-wear and residue analysis of tools from two early occupation sites: Madjedbebe (MJB; formerly known as Malakunanja II), in northern Australia, and Lake Mungo, in western New South Wales. Grinding stones are ubiquitous in Australia and are present in the some of the earliest human occupation sites of Sahul (the Pleistocene landmass comprising Australia and New Guinea), but our knowledge of grinding stones has been overshadowed by a general focus on flaked stone artefacts. Moreover, the function of grinding tools has mostly been inferred on the basis of morphology, and largely restricted to grass seed grinding, which is usually associated with deeply grooved, large sandstone dishes. Previous studies of grinding stones from the region have found little compelling evidence for seed grinding prior to the Pleistocene/Holocene boundary, in part because many grinding stones from Pleistocene contexts occur as fragments with no recurring form and no distinctive grinding grooves. Such tools are often referred to as “amorphous” grinding stones and their function is frequently assumed to be opportunistic, with little understanding of what materials were processed. However, functional analyses of Pleistocene grinding stones have rarely incorporated use-wear and residue analyses and therefore the function of these tools has remained relatively unexplored. Another issue associated with recognising Pleistocene grinding stones is that many are found on deflated and highly eroded surfaces and have been difficult to accurately provenance and date (for example, grinding stones recovered from Cuddie Springs and Lake Mungo). In this thesis, I report on a functional study of 91 grinding stones from MJB and 17 sandstone artefacts from Lake Mungo. Optically stimulated luminescence (OSL) and radiocarbon ages have suggested ages for these artefacts up to 45 ka for MJB and 14 – 25 ka for Lake Mungo. I analysed all specimens for diagnostic traces of use. My use-wear analysis involved the documentation of wear traces on the stone surface, as identified under multiple magnifications and lighting arrangements. The documented wear traces were compared with a use-wear reference library that was created with experimental and ethnographic grinding stones. Experimental specimens included 28 grinding tools made from one of five different sandstone materials used to grind and pound bone, wood, seeds, wheat, haematite and stone for varying amounts of time. The ethnographic tools included 12 arid zone upper stones (hand-stones) made of indurated sandstone and used for processing seeds. My residue analyses involved the removal of adhering material from the tool surface using one of two sampling methods: pipette extractions using multiple solvents, and ultra-sonication with distilled water. I examined removed material microscopically under transmitted light, and biological stains were applied to distinguish organic material. Non-visible residues and biomolecules were detected using a suite of biochemical tests to indicate the presence of fatty acids, proteins and carbohydrate compounds. Residue mixtures were further characterised with Gas Chromatography Mass Spectrometry (GC-MS) to identify specific biomolecules and compared to modern reference material.

The 91 grinding stones from MJB were collected during the 2012 field season. The recovered specimens were made from sandstone (n = 80), quartzite (n = 8), mudstone (n = 2) and volcanic stone (n = 1). These specimens were most frequently from one of three pulses of activity: Pulse 1 (182 – 209 cm below surface (bs)); Pulse 2 (113 – 150 cm bs) and Pulse 3 (10 – 36 cm bs). Unpublished radiocarbon ages produced on charred botanical remains and gastropod shell from the 1989 and 2012 excavations gave bracketing ages of 28.6 – 35.8 ka cal BP (Pulse 1), 9.2 – 18.2 ka cal BP (Pulse 2), and 4.2 – 5.5 ka cal BP (Pulse 3). Of the analysed specimens, 16 had traces consistent with the processing of pigments, 52 had evidence for the processing of plants (including starchy plants and seeds) and four had evidence for the processing of animal tissue. Eleven specimens had traces that indicated the processing of multiple resources. Plant processing tools were identified in all three Pulses, but pigment processing tools were restricted to early Holocene and Pleistocene deposits.

The artefacts analysed from Lake Mungo included 17 sandstone pieces from the central part of the Mungo lunette during 2009 – 2011. A suite of OSL ages has provided bracketing age estimates for the stratigraphic units in which the artefacts were recovered. Ten artefacts are attributed to Unit E (~25 – 14 ka), and four artefacts are attributed to Unit F (~8 ka). Three artefacts from the Golgol lag were of unknown age. Use-wear indicates a likely seed grinding function for 14 of the artefacts. Use-related residues include starch, cellulose and other plant tissues.

Grinding stones are an important artefact class that appear to retain residues at least as commonly as flaked stone, and perhaps in greater abundance on the typically more porous surfaces. I argue that grinding stones provide a unique and vast bank of past resource-use that is only beginning to be fully exploited by archaeologists. The results of this study have provided confirmation of Pleistocene plant processing and seed grinding activities in Sahul, and have indicated a range of other on-site activities that fluctuate in importance through time. The results also indicate the value of employing an integrated approach to functional analysis that includes the examination of stone tool morphology and use-wear together with a forensic study of residues, including morphological, biochemical and other molecular approaches.

The determination of what was ground on stone provides a vast, mostly unexplored data bank with which we can evaluate and assess hypotheses based on other sources of evidence, such as flaked stones. Since grinding technology (like flaked stone technologies) spans most of human history, details of what was ground are likely to provide new insights into understanding behavioural adaptations associated with archaic and modern human evolution. Such insights may include the response of human populations to changing environmental conditions, landscapes and risk, as well as the cultural practices and the use of symbolic expressions.

FoR codes (2008)

210101 Aboriginal and Torres Strait Islander Archaeology, 210102 Archaeological Science, 210104 Archaeology of Australia (excl. Aboriginal and Torres Strait Islander)



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.