Variation in the number of radiocarbon (14C) dates does not equal change in Neolithic population size over time

Abstract

This study investigates the relationship between radiocarbon (14C) dates and estimations of population size during the Neolithic period, approximately 5950–3650 cal BP, in the Malmö area of southwest Sweden. Utilising a dataset of 1025 radiocarbon samples, of which 732 were selected for detailed analysis, we applied the Lambda (λ) model approach to identify significant periods of change in the number of radiocarbon dates. We also address the complexities introduced by the old wood effect (OWE), which impacts the reliability of radiocarbon dating. Our findings highlight four distinct periods of increase and decrease in the amount of radiocarbon dates during the time period under study. By comparing these results with other data from over 45 years of extensive contract archaeological excavations, we demonstrate that the assumption of a correlation between the number of radiocarbon dates and population size is overly simplistic. Our analysis reveals that a high number of dates does not necessarily guarantee credible population size inferences. These findings suggest that similar conditions could hypothetically apply to other spatiotemporal contexts, thereby calling into question the reliability of radiocarbon dates as sole indicators of population dynamics.