Aegean Catchment (E Greece, S Balkans and W Turkey)
14C databases:
10,000 - 5500 cal BC
Data compiled by Agathe REINGRUBER and Laurens THISSEN, and originally commissioned by Clemens Lichter, DAI Istanbul, within the framework of the international workshop "How Did Farming Reach Europe?", Istanbul 20-22 May 2004
Last updated: 30/06/2005
Preliminary note
Dates are presented, wherever possible, in stratigraphic order, with earliest levels first. Dates within stratigraphic units ('phases', 'levels' or 'layers') are presented with earliest dates first. The abbreviation 'nd' stands for no data available.
In the Provenance column, small capital 'D.' (respectively a triangle in the pdf-files) represents depth below surface. All data are calculated with the conventional Libby half-life value of 5568 yr. The cal BC 1sigma ranges are single ranges, and are established with OxCal v3.10 (Bronk Ramsey 1995, 2001, 2005),
using the most recent calibration curve IntCal04 (Reimer et al. 2004).
The Aegean catchment area is determined by the rivers flowing into the Aegean Sea. Within this area, the present file tabulates absolute dates from four geographical units: I. Eastern Greece including the islands; II. Bulgaria south of the Stara Planina (the Balkan range), and inland Turkish Thrace; III. the Former Yugoslav Republic of Macedonia (FYROM); and IV. Western (Aegean) Turkey. It follows that, for Greece, the dates from Sidari, Grava, Megalakkos and the Preveza Region are not included; nor, for Bulgaria, the dates from Poljanica-Platoto.
Again, many people were helpful in providing us with 14C dates and other useful information. A special word of thanks is due to Aneta Bakamska and J. van der Plicht (Galabnik data), Zafer Derin (Ulucak Höyük data), Nikos Efstratiou (Knossos data), Kostas Gallis and Giorgos Toufexis (Platia Magoula Zarkou data), Bernd Kromer (Argissa and Otzaki data), Hermann Parzinger, Jochen Görsdorf, Mehmet Özdoğan and Necmi Karul (Aşağı Pınar and Hoca Çeşme data), Adamantios Sampson (Maroulas data), Mies Wijnen and Harald Hauptmann (Sesklo data), and Marion Lichardus-Itten (Kovacevo data). Any responsibility for the data presentation and interpretation remains, of course, with the authors.
I. Eastern Greece
Comment ACHILLEION
No dates are available for Achilleion subphase IIIa. Cases of discrepancies in dates BP between the first presentations of some of the Achilleion dates (Lawn 1975, and especially Linick 1977) and the final report are indicated here, and are left unexplained by Gimbutas (1989, chapter 3). There are also occasional differences in phase attribution and provenance between these publications. We follow the reassignments of the final publication.
Not included in the list above is P-2128 (7270±80 BP, see Lawn 1975, 200), being a combination of two separate samples from different quadrants; it is also not published in the final report.
Comment AGIOS PETROS
The re-evaluation date BM-2020R supersedes the original date BM-2020 (cf. Bowman et al. 1990). A second sample from the site (BM-2021R) has not been included here, as it postdates the 5500 cal BC limit.
Comment ARGISSA MAGOULA
In 1958 Milojcic sent 12 charcoal samples from what he considered
Preceramic
contexts to the Heidelberg laboratory. They received entry numbers H-889 up to H-900. We thank Bernd Kromer for this information. Results are known from only six of these samples, two of which, H-?, 5870±120 BP and H-?-?, 5630±150 BP (cf. Coleman 1992, 209f.), have not been included here, since they postdate 5500 cal BC. The UCLA-dates on animal bones should be treated with extreme caution: they were published in 1973, well before the introduction of the AMS dating technique in the late 70s (see Protsch and Berger 1973, and discussion in Bloedow 1992/3, 51ff.). Their provenance is, moreover, not known (not included here is UCLA-1657B, 5000±1100 BP [Coleman 1992, 210]). Milojcic (1973, 250) gives a wrong date (7560±90 BP) for GrN-4145.
The 'Phase' indications as presented here are based on the reassessment of the Argissa site (Reingruber 2002); in brackets the labels as originally assigned by Milojcic. The sample H-889-3080 was taken from the rim of pit alpha - it can either belong to the pit itself which was dug in from spit 27c dating to EN II, or else it pertains to the bordering area in
Delta 8/9, in which case it would belong to spit 31, dating to EN I.
Comment CHAIRONEIA
No further information available.
Comment CYCLOPS CAVE-YOURA
Pending final publication we give priority to a strictly chronological order of the Youra dates. Within this sequence, we were able to differentiate into four groups whose calibrated ranges are mutually exclusive on the 1sigma level.
Phase attributions are those used in Sampson et al. 2003, 123, viz. 'Lower Mesolithic' (around 8700–8300 BC) and 'Upper Mesolithic' (around 7500–7000 BC). Moundrea-Agrafioti 2003, 132 mentions that according to Sampson layers yielding dates between 9000–8200 BP (8200–7200 cal BC) are 'Late Mesolithic'. According to Facorellis et al. 1998, 973 both layer 6 (trench C-west), and layer 15 (trench C-east) yielding DEM-547, included a few sherds from the Neolithic.
Due to measurements on molluscs from Youra the marine reservoir effect has been determined for the first time in the Aegean. It is shown that the mean apparent age of the marine shell was c. 520 years older than dates on charcoal (cf. Facorellis et al. 1998, 970).The cal BC single ranges given here provide the calibrations as most recently carried out (references nos. 2 and 4), and take the marine reservoir effect into account.
A total of 30 dates has been published. Not included are one date from the Late Neolithic (DEM-525), two from the Chalcolithic (DEM-267 and DEM-521), one from Roman (DEM-268) and one from modern (DEM-392) times since they postdate the 5500 cal BC limit. Not included either is the date DEM-345 (6730±38 BP), which, being on a mollusc, yields a 1sigma range of 5196-5054 cal BC (see Facorellis et al. 1998, Tab. 1).
Comment ELATEIA
All samples pretreated with acid and alkali. GrN-2933 considered suspect (Vogel and Waterbolk 1963, 183). GrN-3502 and its parts is stated by Vogel and Waterbolk as stemming from trench 2 (1963, 183), but Weinberg (1962, 207) has trench 1, which we follow here. According to Weinberg (1962, 206) GrN-3037 does not belong to the fill of the bothros, but rather to the underlying destruction level.
Comment FRANCHTHI CAVE
Mesolithic phasing is after Perlčs 1990, e.g., 'P. VII' meaning lithics phase VII according to Perlčs. The labels 'Interphase 0/1' and 'FCP' (=Franchthi Ceramic Phase) are on the basis of Vitelli 1993. Eight dates from the late Palaeolithic (P-1668, P-1520, P-2232, P-1923, I-6129, Beta-2514, I-6139, P-2231) and four dates from the Late Neolithic (P-1921, P-1920, P-1661, and P-1630) have not been included here, since pre-, resp. postdating our temporal framework (cf. Jacobsen and Farrand 1987, Plate 71).
Comment FRANCHTHI KOILADA BAY
One date (AA-31) has not been included in the list, as postdating the 5500 cal BC mark (see references).
Comment HALAI
Sample material was generally undersized and 'not well-consolidated because of the dampness of the soil' (Coleman et al. 1999, 296). The four oldest dates may 'represent or have been contaminated by older organic material' (ibid.), but one of them, A-7272, should be combined with A-7622 and Beta-66803, as stemming from a 'single body of carbonized material' (ibid.).
Comment KLISOURA
A second date (Gd–3790: 6230±30 BP) from trench B has not been included here, since it postdates the 5500 BC framework.
Comment KNOSSOS
All BM dates listed here stem from the 1960 sounding in area AC in the Central Court (Evans 1964),
whereas the OxA date derives from a new sounding carried out by Nikos Efstratiou. More 14C dates are available from Knossos,
but they have not been included in the table, as they integrally are beyond the 5500 cal BC mark apart from BM–126 (7000±180 BP) attributed to level V. No dates are available from levels VIII–VII. BM–436 was re-assigned to level IX by Evans (1994, 5), which is in deviation of all previous publications of the date (e.g., Evans 1968, 269, 272; Evans 1971, 117), and not followed here.
Comment MAROULAS on KYTHNOS
We are very grateful to Adamantios Sampson for providing us two new dates from Maroulas. In view of the deviating calibrations provided by (2) for DEM-? and DEM-??, the sample material might have been molluscs. If so, the younger ranges are the more likely ones. Two dates (Gd-11655 and Gd-11653) 'were obtained on epigenetic carbonates and do not, directly, pertain to occupational events but give the age of the episodes when the humidity and temperature favoured migration and concentration of carbonates' (Sampson et al. 2002, 62).
Comment NEA NIKOMEDEIA
Q-655, GX-679, P-1202 and P-1203A were analysed during the early and mid-1960s, whereas the OxA samples on bone and seeds were processed in 1988 and 1993, making use of archive material from the original excavations.
Comment OTZAKI MAGOULA
According to Bernd Kromer, a total of four samples, all with the annotation 'cardium', was sent to the Heidelberg Laboratory (entries H-901/13, H-902/14, H-904/16, H-905/17). Only two of them have been analysed, but their exact ages are not cited in the reference given; they are estimated from the graph compiled by Milojcic.
Comment PLATIA MAGOULA ZARKOU
There are no dates available for the lowest level, the occupation of the site probably dating back to the transition from Early to Middle Neolithic (Giorgos Toufexis, pers. comm., May 2004).
Comment SERVIA
The revision BM-1885R supersedes the original measurement BM-1885 (6360±190 BP) Likewise, BM-1887R replaces BM-1887 (6420±120 BP) (cf. Bowman et al. 1990). Not included in this list are the three Early Bronze Age dates from Servia (BM-1886R, BM-1888R, and BM-1108) (Bowman et al. 1990, 72).
No comment.
Comment SESKLO
Dates are available from all three areas, viz. Sesklo A (Acropolis), Sesklo B (SW of Acropolis), and Sesklo C (W of Acropolis). One Sesklo-date from the Late Neolithic (P-1671) has not been included here, because it postdates the 5500 cal BC limit (cf. Lawn 1973). The original
'Preceramic'-label is now certainly erroneous. Even in these deposits pottery is occurring, although in low quantities (Mies Wijnen, pers. comm., 26 February 2004), a notion confirmed by Harald Hauptmann who reports having seen sherds sticking out of the profile of the 'Preceramic' deposits in 1958 (Harald Hauptmann, pers. comm.).
Comment SESKLO TL-dates
Another seven dates assigned to the 'Middle Neolithic' but yielding values around 4500/4100±500/300 BC, and 15 dates from the 'Late Neolithic' are not included here, since they postdate the 5500 BC limit (see Liritzis and Galloway 1982, 458). The dating method applied was the quartz (feldspar) inclusion technique.
Comment THEOPETRA CAVE
Within this sequence five groups can be distinguished whose calibrated ranges are mutually exclusive on the 1sigma level.
Not included in this list are some 15 dates from the Middle and late Upper Palaeolithic period, nor five dates from the Late Neolithic (DEM-122, DEM-361, DEM-591, DEM-913, and DEM-916), one date from the Chalcolithic period (DEM-141), one date from the Early Bronze Age (DEM-250), and nine AD dates (see Kyparissi-Apostolika 1999, 236 and Facorellis et al. 2001 for details).
Three OSL-dates were taken from sherds belonging to trench Delta 7 but could not be assigned with any certainty to a Mesolithic or a Neolithic context. They show results between 10,670 and 5140 BC (Liritzis et al. 2002, 41–42).
II. Bulgaria and Turkish Thrace
Comment AŞAĞI PINAR
The compilers are very grateful to Hermann Parzinger,
Jochen Görsdorf and Mehmet Özdoğan for communicating the details of the Aşağı Pınar data, and for the permission to publish them on the CANeW website. 18 more dates are available from the site dating Aşağı Pınar levels 5-3. Since they postdate the 5500 cal BC limit, they have not been included here.
Comment CAVDAR
Levels VI-II are asigned to the Early Neolithic (Karanovo I). Level I (Karanovo II/III) did not yield any dates. There are some major discrepancies between the dates quoted by Georgiev and the most recent treatment by Görsdorf and Bojadziev, as is indicated. We tend to follow the later source.
No comment.
No comment.
Comment GALABNIK
The sample for GrN-19786 was actually too small, hence the large error factor (J. van der Plicht, pers. comm., 18 August 1993).
Comment KOVACEVO
We are very grateful to Marion Lichardus-Itten for allowing us to include some as yet unpublished dates from Kovacevo. The period indications are based on the ceramic development and are to be regarded as preliminary.
Comment KREMENIK
Levels I-IV are synchronized by Georgiev to Anzabegovo II-III, and Karanovo I-II (Georgiev et al. 1986, 149f.).
Comment OKRAZNA BOLNICA
Level V is assigned to the Karanovo I phase, level IV to Karanovo II (Görsdorf and Bojadziev 1996, 127).
No comment.
Comment TELL AZMAK
According to Kohl and Quitta (1966, 32) most charcoal samples, if not from grain and seeds, were burned wooden posts. Likewise, they comment that 'Dates for Bln-293 and Bln-291 appear to be too old, but may mean that oak timber used for house construction was sometimes much older than age of settlement' (Kohl and Quitta 1966, 35). Azmak basal level I has five consecutive building horizons and is assigned to the Early Neolithic (Karanovo I); Azmak level II already belongs to the Middle Neolithic stage (Karanovo II/III). The Eneolithic data from Azmak have not been included here.
Comment TELL KARANOVO
The exact level attributions follow those given in the final excavation report (Görsdorf 1997). The three dates from level II without specification (Bln-152, Bln-201, Bln-234) derive from the earlier series of excavations by V. Mikov and G.I. Georgiev between 1947-1957. The remaining dates are from the new investigations, under way since 1984. Dates from Karanovo III and further have not been included here, since postdating the 5500 cal BC limit.
III. Former Yugoslav Republic of Macedonia
Preliminary note
The sites from the Pelagonian Plain are not included, despite the presence of dates fitting the timeframe of this database. More particularly, this concerns the sites of Cuka (Srdoc et al. 1977, 474), Mogila (Srdoc et al. 1977, 474f.), Porodin, and Veluška Tumba (Valastro et al. 1977, 322). All sites concerned would rather date to the later 6th millennium cal BC (cf. Thissen 2000b, 203ff.).
Comment ANZABEGOVO
There are many discrepancies concerning the Anza 14C dates published in several articles and the final report, as has been pointed out by Milojcic (1978:556) and Nandris (1979, 263). For the sake of clarity we adhere to the list given in the final report (Gimbutas 1976, 30). All dates are from the American soundings, except for the two dates on bone, which derive from Garašanin's excavation.
Three remaining dates from level IVb have not been included, as postdating the 5500 cal BC limit (LJ-2178, 6100±250 BP - NB. 7050±150 BP according to [3]!); LJ-2411, 6070±190 BP; LJ-2329, 6210±60 BP).
Comment ANZABEGOVO TL-dates
A total of 10 sherds were measured by M. Aitken and J. Huxtable at Oxford, but six were rejected as poor. The dating method applied was the fine-grain technique (Gimbutas 1976, 32). No standard deviations were mentioned.
Comment VRSNIK-TARINCI
All three samples derive from a square structure with pisé walls full of charred grain, interpreted as a grain storage (Garašanin and Garašanin 1960-61, 36).
IV. Western Turkey
Comment HOCA ÇEŞME
The compilers are very grateful to Necmi Karul and Mehmet Özdoğan for allowing us to publish here the additional Hoca Çeşme dates from the Heidelberg Lab. We are indebted to Mehmet Özdoğan also for communicating the exact provenances of the samples. It should also be noted that the phase assignation of several samples has changed slightly with Karul's definitive working out of the Hoca Çeşme sequence (Karul 2000, 60).
Comment ULUCAK HÖYÜK
We are very grateful to Zafer Derin and the excavation team of Ulucak for providing these dates. They are the first dates ever available from West Anatolia. Ulucak is a mound of 4.5 ha, with a total deposit of 9m. The excavators note a difference in building methods between phases V (consisting of levels Va-c) and IV (levels IVa-h) (Çilingiroğlu et al. 2004).
No comment.
References
