Different levels of interpreting chemical data by prompt gamma activation analysis on polished stone tools in the Carpathian Basin (Hungary)

Abstract Average chemical composition data can be effectively exploited in provenance, i.e. raw material’s origin, research of archaeological polished stone artefacts where direct correlation with potential raw materials is possible. This study summarizes the success of Prompt Gamma Activation Analysis (PGAA) method on two different levels of chemical data interpretation for variable rock types found in Hungary, in the Carpathian Basin and their wider vicinity, i.e. material characterization and provenance identification level. Characterization of different lithologies is possible by variable combinations of some major element oxides (Si, Mg, Ca, Na, K, Al, Fe and Ti). Applying statistical methods, boxplots and Principal Component Analysis, we proved that serpentinites, nephrites and Mg-Si-rich subgroup of ‘white stones’ can be clearly distinguished, while the ‘green stone’ category cannot be further differentiated, partly overlaps with hornfels and significantly overlaps with other rock types (basalts, metadolerites-metagabbros, andesites and metavolcanites). The provenance identification is successful in the case of contact metabasites and HP metaophiolites, while less exact for nephrites, basalts and hornfels. The non-destructive PGAA proved to be very effective in the identification of certain lithologies, while for other polished stone tool raw materials it has medium certainty and requires complementing methods.