Central Anatolia and Cilicia chronological chart: 9th - 5th millennia cal BC

Data compiled by Laurens THISSEN
Last updated: 16/03/2007

This chart is based on the CANeW Central Anatolia and Cilicia 14C database

Comments to the Central Anatolia and Cilicia 14C chart

All calibrated data quoted at 1sigma. Calibration curve used is IntCal04 (Reimer et al. 2004). Calibration program used OxCal v3.10 (Bronk Ramsey 1995, 2001, 2005).

Pınarbaşı A: Summing the three dates gives a range of 8550–8230 cal BC, with relative probability sharply increasing towards 8300 cal BC.

Aşıklı Höyük: From the deep sounding carried out in grid squares 4G-H 10 dates are available, ranging between 8200-7500 cal BC at 1sigma, extending over Phases 2H/2G through 2C-A. Below 2J in the sounding a silt deposit overlies three building levels of Level 3 (Phases 3C to 3A, 3C being the earliest). Results from the 14C samples from Level 3 are not yet available. The samples taken by Todd and Kuniholm from the burned deposit in the NW cut at the visible base of the site sum at 8210–7750 cal BC. This burned deposit is stratified 'well above the stratum of the flood deposit', as the latest report on Aşıklı testifies (Esin and Harmankaya 1999, 118). If the flood deposit and the silt layer in the deep sounding are identical, then Aşıklı Level 3 may well pre-date the 8200 cal BC limit indicated by the earliest deep sounding and NW cut dates, and this assumption is reflected in the chart.
The three sets of dates all from Phase 2C–A contexts yield overlapping values when summed. The 21 North Quarter dates range at 7820–7550 cal BC, the 8 SW Quarter dates at 7660–7480 cal BC, and the two dates East of the Enclosure sum at 7740–7530 cal BC. Within all these summed sets, the highest relative probabilities increase strongly towards 7600 cal BC, suggesting an ending of Aşıklı 2C–A at around that time.

Kaletepe: Three dates are available for the Aceramic stage, two from sector P.3.2, and one from the East sector, summing at 8300-7960 cal BC, where the P.3.2 dates are responsible for a high relative probability between 8300–8160 cal BC.

Canhasan III: The exact stratigraphic locations of the samples are not known. The total sum value of the dates (n=16) is set as follows: 7650–7300 cal BC (40.8%), 7250–7200 (1.2%), and 7050–6650 cal BC (26.1%), where samples BM–1660R, 1662R–1667R and HU–11-12 are responsible for the early range, and the remaining dates for the younger time span. This distribution reflects the stratigraphic position of the samples with the exception of OxA–388. Two nearly identical dates HU–11 and HU–12 represent the oldest dates known from the site. The stratigraphic context of these HU–dates is not known. Were we, however, to combine their probability distributions (assuming them to stem from a single event in view of their similarity), then they yield a range of 7600–7545 cal BC with an agreement of 143.7%. This age range may be taken as a terminus ante quem for the beginning of occupation at Canhasan III.

Musular: The Musular SUM is based on 10 dates (R_combining data GrN–25461 and GrN–25611, and exclusive of GrN–23518 due to its large error margins). Though the total SUM range at 1sigma is at 7570–7140 cal BC, 61.2% of that range (=68.2%) is fixed between 7570–7290 cal BC, enabling us to narrow down the probable Musular age range considerably. It is this interval that has been followed on the chart. The single Pottery phase date in the early 6th mill. should be considered as a terminus post quem only.

Suberde: The date from basal Level III (P–1389) is not used since it postdates all stratigraphically later dates. The dates from the middle part of Level III sum at 7460–7070 cal BC (where P–1387 has been left out due to its large error margins), but nothing is known of their exact stratigraphic context. The ensuing, top part of Level III, yielding architecture and early pottery, has two 14C dates P–1386 and P–1385 that may be combined, as they allegedly stem from the fire destruction of Level III [top]. At 1sigma, their posterior probability distribution is 7030–6770 cal BC (agreement of 103.7%).

Çatalhöyük East: The 1sigma range of Çatal Levels Pre XII.D–II covers more than one millennium from 7400–6200 cal BC. Dates from Levels I–0 are as yet not available. As regards the dates from the Mellaart excavations, a more secure 14C sequence, deleting 'old wood' samples a.o., simultaneously taking into account indications for the beginning and end of individual levels, has been carried out by Newton (Newton 1996; Newton and Kuniholm 1999). Following those analyses a more reliable date series of Çatal would consist of the following: Level X: P–782 , P–1370; Level IX: P–1369, P–1372, P–1371; Level VIII: P–1367; Level VII: P–1366; Level VIB: P–1364, P–1362; Level VIA: P–1375, P–827, P–772, P–778, P–769; Level V: none; Level IV: P–774; Level III: none; Level II: P–796. This total of 16 dates groups roughly into two mutually exclusive clusters, viz. an early group between 7200–6600 cal BC covering Levels X–VIB; and a later group covering Levels VIA–II between 6600/6500–6250 cal BC.
This grouping is in rough lines confirmed by the South Area deep sounding carried out by the Hodder team. Summing the 29 dates available from the sounding yields a range of 7100–6550 cal BC. The two dates from a core into the base of the mound (PL–980525A and AA–27982) provisionally labelled as '?pre-Level XII.D' might supply a terminus ante quem for the beginning of settlement at Çatalhöyük East: summing at 7530–7180 cal BC, they combine into a range of 7460–7190 cal BC.

Erbaba: Only the date from basal level III, I–5151 has been used for the chart, where the three other dates have excessive error margins.

Mersin: Leaving out dates W–617 (large error), Rome–808 (too early in the overall sequence), and Rome–1010 (too young) results in two date clusters, viz. an older group of five dates from levels XXXIII–XXVIII early, summing at 6700–6370 cal BC; and a younger cluster of seven dates from levels XXVIII late–XXIII/XXII summing at 6070–5810 cal BC. Rather than assuming a chronological gap between the two groups, the division is probably of a statistical nature only, where the late level XXVIII date (Rome–1226) seems to partly bridge this gap. No radiocarbon dates are available from levels XXII–XVII. There is a single date from level XVI; the two level XIIB ('Latest Chalcolithic') dates have not been used in the chart.

Pınarbaşı B: Two mutually exclusive dates, the SUM split up into 6400–6230 cal BC (37.7%) and 6070–5920 cal BC (30.6%). The third date OxA–5502 dating to the 5th mill. cal BC has not been included in the 14C chart.

Çatalhöyük West: The two dates from Çatal West are the first ones available from this site and they provide a possible terminus ante quem for the start of settlement here, being from the base of the mound. The extension towards 5500 cal BC of the Çatal West range is based on the parallels in material culture between the site and Canhasan I Layers 2B–A.

Canhasan I: Six dates from Canhasan I stem from a possibly single, or short-term event, viz. the burning and collapse of the W and E rooms of House 3 during the last subphase of Layer 2B (P–794, P–795, P–790, P–791, P–792 and P–793). P–793 must be intrusive, as it is too young, and may very well be related to the Middle Chalcolithic Canhasan Layer 1 occupation, as suggested on the chart (see below – Canhasan I layer 1 entry). A combination of the remaining dates fails at 5% in a X2-test due to P–794. Leaving this early date out yields a combined range of 5710–5640 cal BC, with an agreement of 102.9%. No dates being available for Layers 7–3, the beginning of Canhasan I at approximately 6000 cal BC should be considered tentative. Layer 2A is represented by a single date P–789, predating the Layer 2B range gained supra, and is therefore not used in the chart (cf. also the comment in the database).

Canhasan I, Layer 1 (Middle Chalcolithic): Canhasan I, Layer 1 has not provided any radiocarbon dates, and the one date used here (P–793) is in fact stemming from Layer 2B where it is a clear outlier to the other 2B dates (see Canhasan comment supra). We consider P–793 as a possible intrusion from Layer 1 into Layer 2B.

Köşk Höyük: The nine Köşk dendro dates from a single tree span a range of nearly 600 years, from 5300–4720 cal BC. The older occupation has not yielded radiocarbon data yet.

Güvercinkayası: No comment.  

Kaletepe top: No comment.

References