Wonderwerk Cave has yielded one of the longest and most complete Holocene Later Stone Age (LSA) records for the arid interior of South Africa. This paper presents the results of a new radiocarbon dating program for Excavation 1 that is explored within a Bayesian model of all existing Wonderwerk Cave radiocarbon (14C) dates for the Holocene. The proposed model, using Phases within an OxCal Sequence model, provides robust age estimates for changes in the technological and paleoenvironmental record at the site. The more precise dates allow a comparison of the timing of climate shifts across the interior of southern Africa and begin to allow us to identify whether hiatuses in human occupation, or cultural shifts, are synchronous across broader areas of the subcontinent, or not.

The southern African Later Stone Age sequence is widely considered to be well dated based on radiocarbon dates from dozens of archaeological sites, and apparently shows more or less synchronous cultural shifts across an extensive area. Yet, closer examination reveals the inadequacy of many of the decades-old and uncalibrated individual site chronologies that underpin this regional chronology, making robust comparisons of the chronology of technological change across this region impossible. Here, we present 26 new AMS 14C dates and Bayesian modeled chronologies for two important archaeological cave sites in southernmost Africa, Nelson Bay Cave and Byneskranskop 1. The results provide more robust age estimates for these cultural and paleoenvironmental sequences and revise interpretations of these sites in several instances. This project demonstrates the necessity of redating key sites, and the value of currently underutilized methods, including calibration and Bayesian modeling, for southern African archaeology.