Reconstructing the full temporal range of archaeological phenomena from sparse data.

0553266 - BC 2022 RIV US eng J - Článek v odborném periodiku
Key, A.J.M. - Roberts, D.L. - Jaric, Ivan
Reconstructing the full temporal range of archaeological phenomena from sparse data.
Journal of Archaeological Science. Roč. 135, Nov (2021), č. článku 105479. ISSN 0305-4403. E-ISSN 1095-9238
Grant CEP: GA AV ČR(CZ) Fellowship J. E. Purkyně
Grant ostatní: AV ČR(CZ) Fellowship J. E. Purkyně
Program: Fellowship J. E. Purkyně
Institucionální podpora: RVO:60077344
Klíčová slova: Temporal distribution * Modelling * End date and origin inference * Optimal linear estimation * Weibull * Computational archaeology * Archaeological record
Obor OECD: Archaeology
Impakt faktor: 3.508, rok: 2021
Způsob publikování: Omezený přístup
https://doi.org/10.1016/j.jas.2021.105479

Archaeologists rarely discover the first or last known occurrences of past cultural phenomena, yet 'start' and 'end' dates are central to our understanding of past human behaviour., therein lays a paradox long known within the discipline. Optimal linear estimation (OLE) has recently been used to reconstruct the full temporal range of prehistoric archaeological technologies using only the partial records available. That is, OLE has been used to reconstruct the portions of the archaeological record not yet evidenced through artefact discoveries. Here we present OLE to a wider archaeological audience and outline for the first time the model's assumptions as they pertain to archaeological phenomena. We demonstrate OLE to be an accessible, user-friendly and methodologically transparent temporal range estimation method applied via a single set of equations. Further, we present five additional frequentist techniques that enable archaeologists to account for observation reliability, search effort and extreme data scarcity when inferring temporal ranges. These methods allow archaeologists to gain a more accurate understanding of the temporal range of past human behaviour.
Trvalý link: http://hdl.handle.net/11104/0328248