Geoarchaeology of burnt mounds: site formation processes, use patterns, and duration

Abstract

Burnt mounds, accumulations of fire-cracked stone and fuel residues dating largely from the Bronze Age, are a widespread and numerous site type across Britain and Ireland. However, the function, duration, and depositional history of this site type remains unknown. This study examines the formation processes and duration of burnt mounds using a new multi-proxy geoarchaeological methodology comprising micromorphology, x-ray fluorescence, loss on ignition, and soil pH assessment alongside radiocarbon dating and statistical exploration. The sampled material covers nine sites across northern Britain, spanning Northumberland, Wester Ross, and Orkney.

This work shows that the burnt mounds assessed are complex and multi-faceted, consisting primarily of fuel residues; wood and grasses in Northumberland and Wester Ross, and peat, turf, and seaweed in Orkney. Many sites contain discrete deposits of bone, unburnt plant tissues, and rubified sediments, and the use of earth-ovens is suggested at one site by repeated layers of charred grasses and massive rubified peds. Mounds were deposited in small increments, each representing either one discrete firing event or the admixing of several firing events. Principal component analysis of quantitative multi-element and sedimentological data was able to group burnt mound deposits by fuel type, which was verified by micromorphology.

Micromorphological and sedimentological analyses identified widespread natural sediment incursion and taphonomic processes across all mounds. These incursions often represent hiatuses in deposition and indicate the recurrent, cyclical, or even seasonal use of these sites. Although manifested differently due to local environmental conditions, natural formation processes show recurrent abandonment and cyclical patterns of use at sites through the deposition of material of similar types.

Limited seasonal indicators such as desiccation features, soil formation, and rain-splash crust formation suggest hiatuses in wetter seasons and activity in drier seasons. Available radiocarbon dates suggest that the depositional duration of most mounds was around 100 years, although outliers exist.

Together, these data indicate for the first time that burnt mounds were not the product of single large events such as monumental feasting, but rather represent a more gradual accumulation of small-scale activities by small communities or groups that were cyclical in nature. This has implications for wider interpretation, both of burnt mounds and the socioeconomic systems that they represent, which is discussed.

Equally, this study argues that burnt mounds warrant more direct archaeological attention as they contain well-stratified deposits that hold palaeoenvironmental and geoarchaeological records of human-environment interactions over the short to medium term. Consequently, this study puts forward a best practice excavation and sampling guide to inform on future work, angled towards the policy and developer-funded sectors.