Madjedbebe: Stone Tools and Plant Foods
Stone tools and charred plants offer glimpses of food preparation at Madjedbebe, a rockshelter in Mirarr Country in northern Australia.
A rockshelter and its chronology
Madjedbebe lies on the western edge of the Arnhem Land plateau. Clarkson and colleagues reported an occupation beginning around 65,000 years ago, using sediment dating and evidence about the position of artefacts. This is an archaeological estimate, not an exact arrival date for people across Australia. The researchers examined concerns about disturbance by testing sediment layers and refitting stone fragments.
The relationship between objects and sediment matters because an artefact displaced into an older layer could give a misleading impression of its age. Refitting involves matching broken pieces that once belonged together and comparing their excavated positions. The researchers found refits within the lower and middle dense artefact bands, but none connecting those two bands. They interpreted this pattern, together with intact hearths and sediment analyses, as evidence that the deposits retained meaningful distinctions between occupation phases. Excavation study.
Tools as evidence of work
The earliest dense occupation deposits contained flaked stone, grinding stones and hatchet heads with ground edges, alongside a hearth and worked ochre. These finds reveal several kinds of material processing. They do not identify who performed each task or establish a daily timetable.
The assemblage also includes cores and flakes, recording the working of stone as well as the presence of finished tools. Grinding surfaces require further investigation: a stone’s appearance alone does not establish everything processed on it. Microscopic wear and surviving residues help distinguish activities. The excavation study illustrates red pigment residues on a grinding stone, showing why grinding equipment should not automatically be treated as evidence of food preparation alone. Clarkson and colleagues.
Gathering and preparing plants
Florin and colleagues studied burnt plant remains from the occupation phase dated approximately 65,000–53,000 years ago. Identifications included fruit and nut fragments, underground storage organs such as roots or tubers, and palm tissue. Their interpretation connects this assemblage with gathering and food preparation, including cooking and extraction of edible parts. This suggests that everyday subsistence involved both obtaining food and spending time preparing it.
Different foods presented different preparation problems. The researchers interpret some fruits and easily opened nuts as relatively accessible foods, while roots, tubers and palm stems required cooking and sometimes peeling. Palm pith was probably also pounded, and extracting pandanus kernels demanded substantial effort. These interpretations point to sequences of work rather than a single act of gathering. Botanical knowledge and preparation techniques helped make a wider range of plant resources usable. Plant-food study.
What survives is only part of the diet
The plant study recovered no edible seeds as macrofossils; evidence for seed grinding instead came from residues and wear on grinding stones. Burnt remains cannot show the proportions of foods eaten: raw foods and meals consumed elsewhere may leave little trace. The finds support a varied plant diet, but cannot reconstruct a complete menu or division of household labour.
Identification also depends on comparison. Researchers examined the anatomy and form of ancient fragments under microscopes and compared them with modern regional plants. The available reference collection and similarities between related plants limited how precisely specimens could be named. Preservation adds another filter: different plant tissues survive differently, and burning does not preserve every meal equally. The authors therefore did not calculate dietary proportions from the recovered fragments. Discussion and methods.