Artefacts in time: Archaeological structures identified by dense radiocarbon sampling

0532965 - ARÚ 2021 RIV CZ eng A - Abstrakt
Demján, Peter
Artefacts in time: Archaeological structures identified by dense radiocarbon sampling.
26th EAA Virtual Annual Meeting. Abstract book. Prague: European Association of Archaeologists, 2020 - (Kleinová, K.). s. 207-208. ISBN 978-80-907270-7-6.
[Virtual Annual Meeting of the European Association of Archaeologists /26./. 24.08.2020-30.08.2020, online]
Grant CEP: GA MŠMT EF16_019/0000728
Klíčová slova: radiocarbon dating * clustering * bayesian modeling * chronological modeling
Obor OECD: Archaeology
https://submissions.e-a-a.org/eaa2020/repository/preview.php?Abstract=1611

Calibrated radiocarbon determinations are routinely used to gain absolute chronological information about archaeological artefacts, ranging from single finds to larger contexts such as settlement features or whole phases of occupation. This is done by associating the calibrated date with a specific event (such as felling of a tree) which led to the construction of the artefact (e.g. the post of a house). There are however find situations, where even though we are able to determine that some human activity took place based on an accumulation of its associated residues (e.g. pottery fragments, charcoals, cultivated plant seeds), we lack more complex structures (e.g. houses, pits, strata) that would indicate whether it's the result of a single event, multiple separate events or a continuous habitation. Thanks to increasing availability of high precision radiocarbon dating, we are now able to view the whole series of dates from a single site or context as a probability distribution and use statistical methods to reveal significant irregularities in it, which can be interpreted as artefacts of human activity. In this presentation we will examine the thesis that by performing cluster analysis of a set of radiocarbon dates based on their distance in time, we can determine the minimal amount of separate events that would explain their distribution. We will also look at ways to eliminate the possibility that the observed gaps between the events are a product of irregular sampling or the calibration process. Events identified in this way can then be treated as evidence of archaeological structures such as settlement phases and used for Bayesian modeling to further precise our interpretation of the find situation.
Trvalý link: http://hdl.handle.net/11104/0311484