This article reports on ten new accelerator mass spectrometry (AMS) dates from the Chalcolithic period (fifth millennium BC) archaeological type-site of Teleilat Ghassul in Jordan. Early radiocarbon assays from the site proved difficult to integrate with current relative chronological formulations. The ten new AMS dates and follow-up enquiries connected with the early assays suggest that the original dates were up to 500 years too early. A necessary reformulation of regional relative chronologies now views the Ghassul sequence falling between Late Neolithic Jericho and the Beersheban Chalcolithic.

This article reports on 12 new accelerator mass spectrometry (AMS) dates from the latest phases of the Chalcolithic period occupation (late 5th millennium cal BC) at Teleilat Ghassul, type site for the south Levantine Ghassulian Chalcolithic culture. The new AMS dates from Teleilat Ghassul favor an amendment to a previous suggestion (Bourke et al. 2001), that all significant occupation at the site had ceased by 4000/3900 cal BC. This end-date should now be amended to 3900/3800 cal BC. Follow-up statistical modelling sourced to published 14C data drawn from a wide selection of south Levantine Chalcolithic period sites (Bourke 2001; Burton and Levy 2001) raises the possibility that Chalcolithic period occupation had ceased at virtually all major centers by 3800/3700 cal BC. This, in turn, suggests that the new data bearing on the end-date for occupation at Teleilat Ghassul may reflect a more widespread horizon of abandonment in the southern Levant.

ABSTRACT. This article reports on 10 new accelerator mass spectrometry (AMS) radiocarbon dates from early phases of the Early Bronze Age at the long-lived settlement of Pella (modern Tabaqat Fahl) in the north Jordan Valley. The new AMS dates fall between 3400 and 2800 cal BC, and support a recent suggestion that all Chalcolithic period occupation had ceased by 3800/3700 cal BC at the latest (Bourke et al. 2004b). Other recently published Early Bronze Age 14 C data strongly supports this revisionist scenario, suggesting that the earliest phase of the Early Bronze Age (EBA I) occupied much of the 4th millennium cal BC (3800/3700 to 3100/3000 cal BC). As this EB I period in the Jordan Valley is generally viewed as the key precursor phase in the development of urbanism (Joffe 1993), this revisionist chronology has potentially radical significance for understanding both the nature and speed of the move from village settlement towards a complex urban lifeway.

The site of Pella, located in the foothills of the east Jordan valley, was a prosperous city–state throughout the Middle Bronze Age (MBA, ca. 2000–1500 BCE). As part of a widespread trading network, Pella enjoyed extensive socio‐economic relationships with Egypt, Cyprus, and the Aegean, Anatolia, and Babylonia during this period. We report isotopic analysis (87Sr/86Sr, δ18O, and δ13C) from enamel of 22 human permanent second molars of which 13 second lower molars were used for an additional biodistance analysis based on ASUDAS. The multidisciplinary approach investigates the ancestral background of MBA Pella and the degree of temporary or more permanent relocation from other settlements. Ancillary to carbonate isotope analysis for migration investigation, dietary information in the form of δ13Ccarbonate was also collected. δ13Ccarbonate values (mean −12.3‰ ± 0.4 SD) suggest a uniform diet reliant on C3 cereals and legumes as crops and animal fodder, adhering to expected Bronze Age Levantine dietary norms. Two methods are used to identify non‐locals. Using a biospheric baseline, three individuals with non‐local 87Sr/86Sr ratios are identified. Bagplot analysis of both 87Sr/86Sr and δ18O data suggests that three individuals (14%) grew up elsewhere; two individuals who were already identified as 87Sr/86Sr outliers using biospheric data and one more with outlying δ18O values. All individuals identified as non‐locals, using either method, are from one tomb, Tomb 62. The dental nonmetric traits indicated diverse morphology and subsequent ancestry for Tomb 62 (11/13), whereas primary burials (2/13) clustered together. The commingled condition of Tomb 62 material prevented a more exhaustive biodistance analysis, but the tentative results coincide with interpretations of the tomb. Significant movements of populations throughout the Middle Bronze Age are evidenced through funerary rituals and architecture, and this study demonstrates that Pella, thought to be peripheral, nonetheless had some permanent movement evidenced through isotopes and ancestry analysis.

The issue of the nature of the transition from the Late Neolithic to the Early Chalcolithic at Teleilat Ghassul is re-examined in this article. A stratigraphic review of all work at the site suggests a new equivalencing between Hennessy’s ten-phase (A+-I) and the traditional Pontifical Biblical Institute (PBI) five-level (I-IVB) division. It is suggested that no PBI excavation (Mallon, Koeppel, North) sampled Hennessy’s Neolithic phase H-I assemblages. Once the PBI material is set to one side, a detailed comparison between Hennessy’s Neolithic (phase H-I) and Early Chalcolithic (phase G-F) assemblages is undertaken. Architecture is sharply contrasted, but more nuanced views of the ceramic, lithic and ecofactual assemblages emerge. The Late Neolithic assemblages are closely to be identified with earlier Neolithic norms, whereas the Early Chalcolithic assemblages display all the hallmarks of the later Classic Ghassulian culture. Radiometric data old and new is then considered, with the result that the old high dates for the Ghassul Neolithic are set aside, and a new “low” chronology proposed, based on recent short-life samples. Finally, in the light of the considerable differences between the Late Neolithic and Early Chalcolithic assemblages, we raise anew the question of Ghassulian origins.