The oldest maritime sanctuary? Dating the sanctuary at Keros and the Cycladic Early Bronze Age.

Renfrew, Colin ; Boyd, Michael ; Ramsey, Christopher Bronk 等

[ILLUSTRATION OMITTED]

Introduction

The world's earliest sanctuaries are getting older. The remarkable symbolic centre at Gosbekli Tepe in eastern Turkey (Schmidt 2007) dates from the Early Holocene period (some 11 000 years ago); sites in coastal Peru such as Caral (Shady et al. 2001; Shady & Kleihege 2008) date to the very inception of food production there some 6000 years ago. And now the site of Kavos on Keros in the Aegean Sea, here proposed as a sanctuary or place of pilgrimage, can be dated with precision by radiocarbon determinations to the years 2750 to 2300 cal BC. This is not so early as Gobekli or Caral, but it has the special quality of being a maritime centre, the earliest regional maritime symbolic centre yet known anywhere, preceding the great island centre of Delos (also in the Cycladic Islands of Greece) by some 2000 years.

These findings have relevance for the early emergence of 'cult' or 'religion'--if those terms are reserved specifically for places and rituals where deities, conceived as real and powerful supernatural beings, may be adduced (Renfrew 1985: ch. 1, 1994). Religion or cult in that sense is well documented in early state societies, whether in Egypt, Mesopotamia, in the Mycenaean world or in Mesoamerica, where shrines or temples can be identified (e.g. Marcus & Flannery 1996) through their explicit iconography. Here we are speaking rather of 'sanctuaries'--places of assembly or the structured deposition of symbolic artefacts--where the explicit iconography of the divine, with identifiable gods or goddesses, has not yet developed. At Gobekli Tepe or Caral or Kavos on Keros it is not easy for us to define what belief system or 'faith' brought the pilgrims together. That is why, at another more recent pilgrimage centre, Chaco Canyon in New Mexico, it has seemed appropriate to choose a vaguer term: locations of 'High Devotional Expression' (Renfrew 2001).

The Early Bronze Age of the Aegean in the third millennium BC was a crucial period for the emergence of Aegean civilisation and for the development of the strong maritime interactions that helped to bring it about (Renfrew 1972; Broodbank 2000). The recent excavations on the Cycladic island of Keros, both at the sanctuary at Kavos and at the settlement on the islet of Dhaskalio, lying 90m offshore, give new evidence for the scale of such interactions (Renfrew et al. 2007b, 2009; Research Horizons 2011). In particular the new radiocarbon determinations for the settlement at Dhaskalio allow a greater chronological precision for the later part of the Aegean Early Bronze Age and indeed for the Early Cycladic period (Manning 1995).

Dhaskalio Kavos on Keros

The site at Kavos, which lies on the western tip of Keros (Figure 1), first came to light in 1963 as a result of looting. The 'special deposit' (North) was investigated in the same year by Christos Doumas (Doumas 1964) and by Photeini Zapheiropoulou in 1967 (Zapheiropoulou 1968a, 1968b), and was the subject of a detailed investigation in 1987 (Renfrew et al. 2007a). The Cambridge Keros Project of 2006 to 2008 led to the discovery of a new and previously undisturbed special deposit, now termed the Special Deposit South (Figure 2).

As a result of the latter excavation. It is now clear that the vast quantities of broken pottery, broken marble bowls and vessels and fragmented Early Cycladic sculptures ('figurines') recovered from Kavos were deliberately broken in the course of rituals of breakage, and deliberately deposited at the Special Deposit South on Keros (as well as at the later-looted Special Deposit North). Moreover, from a detailed study of the find circumstances, including patterns of breakage, it can be inferred that these products of ritual deposition were not broken locally at Kavos. They were broken elsewhere, presumably on other Cycladic islands, following a use-life of many years, and a residue of the broken fragments was then systematically brought to Keros for ritual disposal. More than 550 figurine fragments and more than 2300 pieces of marble bowls and vessels have now been recovered from the Special Deposit South (Figures 3 & 4), and comparable numbers were originally present in the looted Special Deposit North. Kavos can now be regarded as a place of ritual deposition and in that sense a sanctuary visited from elsewhere by sea.

[FIGURE 1 OMITTED]

The settlement on Dhaskalio

The settlement on the adjacent islet of Dhaskalio (Figure 5) remained unexcavated, apart from a few days' work by Doumas, until the excavations of the Cambridge Keros Project of 2006 to 2008 (Figure 6). The assemblages recovered from Kavos and Dhaskalio run in parallel (for the broad framework, see Table 1). The earliest finds at Kavos are of the Keros-Syros culture (Early Cycladic II) continuing into the period of the subsequent Kastri group (Renfrew 1972; Rutter 1984; Sotirakopoulou 1993). But while the deposits at the Special Deposit South on Kavos are disturbed by continuing episodes of deposition, the stratigraphic sequence on the islet of Dhaskalio itself is secure (Renfrew et al. 2009:31-37). The lower strata at Dhaskalio (Dhaskalio Phase A) have a ceramic assemblage, studied by Peggy Sotirakopoulou, which is representative of the Keros-Syros culture. The layers of the succeeding Phase B are characterised by the presence of sherds, notably one-handled tankards, of the Kastri group. These continue in the final phase (Dhaskalio Phase C) where sherds of pale fabric appear, some of a shape first published for Amorgos by Dummler (1886). The prominent Hall and other buildings discovered at the summit of Dhaskalio are of Phase C. This secure sequence from Dhaskalio establishes a sequence which can be extended to the sanctuary at Kavos and which has wider consequences for cultural relations within the Aegean.

[FIGURE 2 OMITTED]

Radiocarbon dating

The excavations at Dhaskalio also yielded a good number of radiocarbon samples, which have been dated and modelled by the Research Laboratory for Archaeology and the History of Art at Oxford University. The combination of a secure stratigraphic context for the samples along with the well-defined ceramic assemblages recovered, together with a calibration based upon Bayesian procedures yields a chronological framework with a precision that permits dating to within a quarter to half a century. Without the support of dendrochronology on stratified charcoal, this must be close to the optimum that can be achieved using ceramic-based markers and the parameters of the calibration curve.

The samples from the settlement at Dhaskalio are listed in Table 2 and the dates given in Table 3. Correction for isotopic fractionation has been carried out using the measured [[delta].sup.13]C values measured on the AMS. The quoted [[delta].sup.13]C values are measured independently on a stable isotope mass spectrometer (to[+ or -]0:3 per mil relative to VPDB). For details of the chemical pre-treatment, target preparation and AMS measurement see Bronk Ramsey et al. 2002 and 2004. The calendar age-ranges in the final column have been generated using the OxCal computer program (v4.1: Bronk Ramsey 2009), using the 'INTCAL09' dataset (Reimer et al. 2009).

[FIGURE 3 OMITTED]

The laboratory was unable to date sample R14 due to low yield. Sample R40 yielded a modern (post-1950) date, and is omitted from further discussion. Sample R49 yielded a date of 6268[+ or -]34 BP, which is more than two millennia earlier than the other samples determined. There is no indication on the site of activity at so early a period, and no interpretation can be offered for this dating. The specimen is therefore omitted from further consideration.

That leaves a series of 15 radiocarbon determinations for further consideration. The careful reconsideration of the stratigraphic position of two of these samples (R 44 and R45) in the light of these radiocarbon determinations led to an adjustment of the phasing for them. This procedure is discussed in greater detail in the final excavation report now in preparation (Bronk Ramsey et al. forthcoming). The resultant phasing is seen in Table 4.

For each radiocarbon determination, comparison to the calibration curve yields a probability distribution. The temporal distribution for the 15 determinations from Dhaskalio arranged according to their revised stratigraphic sequence (i.e. Dhaskalio Phase A, B and C) is seen in Figure 7. The diagram represents the best interpretation that we can achieve for the stratigraphic data. The relationship between the date of the charcoal and that of its context relies on the assumption that the charcoal on the site is often from pruned branches and twigs (based on palaeobotanical analysis, Ntinou forthcoming). We think therefore that the growth of the original wood should be within a decade or two of the deposition of the charcoal, although the possibility that some of the olive wood samples may have derived from roof timbers cannot be excluded. The limitations implied in this assumption underlie the analysis and discussions that follow. The application of a simple three-phase Bayesian model (Bronk Ramsey 2009) allows estimation of the dates for the boundary between Phases A and B, and also the boundary between Phases B and C.

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

Bayesian analysis

In this procedure, the radiocarbon determinations (Figure 7) are reconciled statistically with the stratigraphic order as set out in Table 4. This produces the modal seen in Figure 8. The transition from Phase A to B is seen to lie between 2634 and 2481 cal BC, and the transition from Phase B to C is between 2452 and 2324 cal BC. It is convenient for comparative purposes to calculate the weighted mean, median and 95% probability range for each of the transitions in the model (Table 5).

[FIGURE 7 OMITTED]

Refining the chronology for the region

The large error ranges seen in Table 5 for the start of Phase A and the end of Phase C on the present model are due to the lack of earlier and later dates to represent the preceding and succeeding phases. However, the Oxford Laboratory has conducted a number of analyses in recent years for Cycladic sites, some of which are relevant to this problem.

A series of radiocarbon determinations is available from the Early Cycladic site of Markiani in Amorgos (Renfrew et al. 2006). Those from Markiani II, belonging to the Kampos group, transitional between the Grotta-Pelos and Keros-Syros cultures, are particularly relevant, as this phase is generally agreed to precede the Keros-Syros culture. For the Keros-Syros culture, Markiani Phase III offers five determinations. Although the pottery from these contexts was not abundant, and would not support further chronological subdivision, their attribution is clear enough and we have elected to include them in the present analysis (alternative data runs excluding OXA-3292, which might appear an outlier, or excluding all five dates of the Keros-Syros culture from Markiani, did not significantly change the outcome, and these dates have therefore not been excluded).

[FIGURE 8 OMITTED]

The material from Markiani Phase IV was considered to be related to the Kastri group. But the samples analysed were found in strata containing examples of the two-handled cup, which at Dhaskalio is associated with Phase C rather than with Phase B. So until the pottery from Markiani Phase IV is reconsidered in the light of the more precise Dhaskalio phasing it is probably safer not to involve these determinations in refining the chronology for the later part of the Early Bronze Age.

The same observations can be made for some of the other radiocarbon determinations reviewed by Manning in his recent survey (Manning 2008: 59, fig. 7.5). Here, in addition to the samples already mentioned, the other samples relevant to this period are from Phase IV at the Zas Cave in Naxos. These have been assigned to the 'Kastri phase'. But once more, since the pottery from Zas is not yet fully published. It is difficult to establish whether that would equate with Dhaskalio Phase B or with Phase C, or possibly with both. We propose to omit these here from further analysis. Manning also utilises the earlier Keros determinations (Renfrew et al. 2006; Renfrew 2007) for the chronology of the Keros-Syros culture. These were from the looted area of the Special Deposit North and so we have omitted these in the present analysis since their position in the newly established Dhaskalio phase sequence is not established.

Manning (2008: 56, tab. 7.4; also 2010: 21) does, however, make reference to three analyses from the early Middle Cycladic period from Akrotiri on Thera. These can usefully be utilised in the Bayesian analysis to help refine the end of Dhaskalio Phase C, making the explicit assumption that the Akrotiri early Middle Cycladic context should be later than the Dhaskalio Phase C contexts.

This refined version of the Early Cycladic chronology may be seen in Figures 9 and 10. For the Dhaskalio sequence, the radiocarbon determinations obtained for Markiani for the Kampos group, which there stratigraphically precedes the Keros-Syros culture (here represented by Dhaskalio Phase A), allows the estimation of the beginning of Dhaskalio Phase A at c. 2750 BC. The inclusion of the early Middle Cycladic determinations from Thera allow the conclusion of Phase C to be set fairly precisely at 2300 BC. Using these figures, the Dhaskalio chronology can now be set as in Table 6.

Implications of the refined chronology

It is instructive to compare this newly established calibrated chronology, derived from these new determinations, with the calibrated radiocarbon chronology for the Early Bronze Age Aegean proposed in 1972 (Renfrew 1972:221). This was based largely upon the radiocarbon dates then available from the Early Helladic sites of Eutresis and Lerna, calibrated according to the early calibration curve established by Hans Suess (1967). At that time, the position in the Cyclades of the later Kastri group, which had recently been defined on the basis of finds from the island of Syros, had not been stratigraphically established, although the use of the Kastri group ceramics as a stratigraphic indicator for the later or developed Keros-Syros culture had been proposed.

[FIGURE 9 OMITTED]

It is remarkable that the chronology proposed as early as 1972 needs so little adjustment to conform with the newly available data. The dates now obtained for Dhaskalio Phases A and B appear to be comparable with the estimate then reached for the Keros-Syros culture. Those for Dhaskalio Phase C are comparable with the earlier part of the estimate then reached for the Phylakopi I culture. If the assumption is made that the end of Dhaskalio Phase C corresponds to the inception of the Middle Cycladic period, the end of the Early Cycladic period would now appear to fall some two centuries earlier than suggested in 1972.

[FIGURE 10 OMITTED]

It should also be noted that the alleged 'gap' in the Early Cycladic sequence (Rutter 1983, 1984), which has for long been a matter of debate, is no longer evident. Rutter had argued that a gap of up to 150 years was evident in the Cycladic sequence between the Kastri group and the Phylakopi I culture. He himself however had suggested that this gap would be likely to disappear with the excavation of an unbroken stratigraphic sequence (1984: 95). This is now the case, and the evidence of Dhaskalio indicates a striking continuity between the successive phases there, with no suggestion of a gap. The good series of radiocarbon dates from Kolonna on the Saronic island of Aegina (Wild et al. 2010) harmonises well with the dates here, suggesting that the end of Dhaskalio Phase C falls a century or so earlier than the Early Helladic/Middle Helladic transition at Kolonna.

It is to be hoped that the chronology presented here can now form a sound basis for the chronology of the later part of the Early Bronze Age in the southern Cyclades. However, such a claim must await the final publication of the Dhaskalio pottery sequence (Sotirakopoulou forthcoming), with its full analysis of the relationship between the latest Dhaskalio pottery and that of the Phylakopi I culture. It may well be possible at that point definitively to propose that Dhaskalio Phase C represents the final phase of the Cycladic Early Bronze Age sequence.

History of the Kavos sanctuary

Turning again to the sanctuary at Kavos, it is clear that the floruit of the site was during the Phase A occupation of the settlement at Dhaskalio, with strong continuing use during Dhaskalio Phase B, and only sporadic activity later, in Dhaskalio Phase C. That places the floruit of the sanctuary at Kavos from c. 2750 to 2550 BC, with its continued use during Dhaskalio Phase B from c. 2550 to 2400 BC, and with a further period of infrequent use of some 100 years. The wider significance here is for the development of symbolic interaction in the Aegean. Ritual activities can be recognised on a modest scale in the Aegean in earlier periods, but the sanctuary on Keros seems to be the first time that ritual activity was practised not only on a local but on a wider regional scale. The sanctuary at Keros came to act as a symbolic attractor for the entire region of the Aegean occupied by the Cycladic Islands, and perhaps more widely. Moreover at this time the symbolism embodied in the typical Cycladic folded-arm figurine was seen also on Crete and on the Greek mainland. It may not be appropriate to speak here of a religious cult (which raises problems of definition). But clearly Keros was, from around 2750 BC, serving as a regional ritual centre. This is the first time that such a phenomenon can be recognised in the Aegean, although in one form or another it persisted from that time, with the archaic sanctuary on the Cycladic island of Delos being a notable example flourishing two millennia later. The origins of that ritual practice can now be dated with precision.

Congregation before deism

The primary purpose of this paper was to set out the more precise chronology for the Aegean Early Bronze Age resulting from the new radiocarbon determinations. But it also offers an opportunity to note the wide variety of early places of congregation, from Gobekli Tepe to Chaco Canyon, and to draw attention to the special role which seafaring may have in the early annals of pilgrimage. The site at Kavos, although it does not have the monumentality of Gosbekli Tepe, or of Tarxien in Malta, or indeed of Stonehenge, can now be claimed as the oldest maritime sanctuary in the world--maritime in the sense that. It is only accessible by sea.

The wider problem here is for the study of early religion. The difficulties in using the term 'religion' have been well brought out recently in a volume devoted to social and cognitive life at Catalhoyuk (Hodder 2010)--another Anatolian site but a couple of thousand years more recent than Gobekli Tepe. Maurice Bloch (2010) argues persuasively that the term 'religion' cannot be validly applied at Catal. We would likewise argue that its proper use requires the evidence for explicit reference to deities, such as can be claimed for the Late Cycladic shrine at Phylakopi in Melos (Renfrew 1985)--some 1500 years more recent than Kavos on Keros--or for the temples of Early Dynastic Egypt or Mesopotamia. The sanctuaries at Gobekli Tepe, Caral or Kavos on Keros cannot be identified as temples or shrines with such an explicit iconography. Yet, even if they are not temples or shrines in this sense, they can be regarded as places of convocation and pilgrimage, as locations of high devotional expression.

So it may now be appropriate to define a ritual behaviour of great antiquity (congregation), which we would not want to dignify with the designation 'religion', where that term implies the veneration of specific deities, of supernatural powers which can be separately identified (i.e. deism). Yet sites of congregation and convocation, whether marked by monumental meeting places (as at Gobekli Tepe or Stonehenge) or by the structured deposition of symbolic artefacts (as at Kavos on Keros) may be regarded as sanctuaries, as places of pilgrimage. The nature of the belief systems that motivated their devotees or participants requires further consideration. But it seems useful to establish, in a preliminary way, that such rituals of assembly and of deposition were practised millennia before the well-defined deities of organised religion can be identified.

Acknowledgements

Acknowledgement is made to the Oxford Radiocarbon Accelerator Unit (NERC Radiocarbon Facility), and for the Cambridge Keros Project, to the Greek Archaeological Service, the British School at Athens, the McDonald Institute, the Institute for Aegean Prehistory (INSTAP), the Stavros S. Niarchos Foundation, the Leventis Foundation, the Leverhulme Trust, the British Academy, the Society of Antiquaries of London, the Balzan Foundation and the N.P. Goulandris Museum of Cydadic Art.

Received: 14 July 2011; Accepted: 30 September 2011; Revised: 4 October 2011

References

BLOCH, M. 2010. Is there religion at Catalhoyuk ... or are there just houses?, in I. Hodder (ed.) Religion in the emergence of civilisation Catalhoyuk as a case study: 146-62. Cambridge: Cambridge University Press.

BRONK RAMSEY, C. 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51.1: 337-60.

BRONK RAMSEY, C., T.F.G. HIGHAM, D.C. OWEN, A.W.G. PIKE, & R.E.M. HEDGES. 2002. Radiocarbon dates from the Oxford AMS System: Datelist 31. Archaeometry 44(3 Suppl. 1): 1-149.

BRONK RAMSEY, C., T.F.G. HIGHAM & P. LEACH. 2004. Towards high-precision AMS: progress and limitations. Radiocarbon 46.1: 17-24.

BRONK RAMSEY, C., C. RENFREW & M.J. BOYD. Forthcoming. The radiocarbnn determinations, in C. Renfrew, O. Philaniotou, N.J. Brodie, G. Gavalas & M.J. Boyd (ed.) The sanctuary at Keros and the origins of Aegean ritual practice: excavations at Dhaskalio and Dhaskalio Kavos, Keros 2006-2008. Volume L Cambridge: McDonald Institute for Archaeological Research.

BROODBANK, C. 2000. An island archaeology of the early Cyclades. Cambridge: Cambridge University Press.

DOUMAS, C. 1964. [TEXT NOT REPRODUCIBLE IN ASCII.]. Archaiologikon Deltion 19:409--412.

DUMMLER, F. 1886. Mitteilungen von den griechischen Inseln. Mitteilungen des Deutschen Archaologischen Instituts (Athenisehe Abteilung) 2:15-46.

HODDER, I. (ed.) 2010. Religion in the emergence of civilisation: Catalhoyuk as a case study. Cambridge: Cambridge University Press.

MANNING, S.W. 1995. The absolute chronology of the Aegean Early Bronze Age: archaeology, history and radiocarbon (Monographs in Mediterranean Archaeology 1). Sheffield: Sheffield Academic Press.

-- 2008. Some initial wobbly steps towards a Late Neolithic to Early Bronze III radiocarbon chronology for the Cyclades, in N. Brodie, J. Doole, G. Gavalas & C. Renfrew (ed.) Horizon. A colloquium on the prehistory of the Cyclades. Cambridge: McDonald Institute for Archaeological Research.

-- 2010. Chronology and terminology, in E.H. Cline (ed.) The Oxford handbook of the Bronze Age Aegean (ca. 3000--1000 BC): 11-28. New York: Oxford University Press.

MARCUS, J. & K.V. FLANNERY. 1996. Zapotec civilization. London: Thames & Hudson.

NTINOU, M. Forthcoming. The vegetation and the use of timber at the Early Cycladic settlement of Dhaskalio, Kavos: the wood charcoal analysis, in C. Renfrew, O. Philaniotou, N.J. Brodie, G. Gavalas & M.J. Boyd (ed.) The sanctuary at Keros and the origins of Aegean ritual practice: excavations at Dhaskalio and Dhaskalio Kavos, Keros 2006-2008. Volume I. Cambridge: McDonald Institute for Archaeological Research.

REIMER, P.J., M.G.L. BAILLIE, E. BARD, A. BAYLISS, J.W. BECK, P.G. BLACKWELL, C. BRONK RAMSEY, C.E. BUCK, G.S. BURR, R.L. EDWARDS, M. FRIEDRICH, P.M. GROOTES, T.P. GUILDERSON, I. HAJDAS, T.J. HEATON, A.G. HOGG, K.A. HUGHEN, K.E KAISER, B. KROMER, F.G. McCORMAC, S.W. MANNING, R.W. REIMER, D.A. RICHARDS, J.R. SOUTHON, S. TALAMO, C.S.M. TURNEY, J. VAN DER PLICHT & C.E. WEYHENMEYER. 2009. Intcal09 and Marine09 radiocarbon age calibration curves, 0-50,000 years cal BP. Radiocarbon 51.4: 1111-50.

RENFREW, C. 1972. The emergence of civilisation: the Cyclades and the Aegean in the third millennium BC. London: Methuen.

-- 1985. The archaeology of cult: the sanctuary at Phylakopi (British School at Athens Supplementary Volumes 18). London: Thames & Hudson.

-- 1994. The archaeology of religion, in C. Renfrew & E.B.W. Zubrow (ed.) The ancient mind: 47-54. Cambridge: Cambridge University Press.

-- 2001. Production and consumption in a sacred economy: the material correlates of high devotional expression at Chaco Canyon. American Antiquity 66 (1): 14-25.

-- 2007. Appendix C. Radiocarbon dates and absolute chronology, in C. Renfrew, C. Doumas, L. Marangou & G. Gavalas (ed.) Keros Dhaskalio Kavos the investigations of 1987-88. Cambridge: McDonald Institute for Archaeological Research.

RENFREW, C., R. HOUSLEY & S. MANNING. 2006. The absolute dating of the settlement, in L. Marangou, C. Renfrew, C. Doumas & G. Gavalas (ed.) Markiani, Amorgos an Early Bronze Age fortified settlement overview of the 1985-1991 investigations: 71-80. London: British School at Athens.

RENFREW, C., C. DOUMMS, L. MARANGOU & G. GAVALAS (ed.). 2007a. Keros Dhaskalio Kavos the investigations of 1987-88. Cambridge: McDonald Institute for Archaeological Research.

RENFREW, C., O. PHILAN1OTOU, N. BRODIE & G. GAVALAS, E. MARGARITIS, C. FRENCH & P. SOTIRAKOPOULOU. 2007b. Keros: Dhaskalio and Kavos, Early Cycladic stronghold and ritual centre. Preliminary report of the 2006 and 2007 excavation seasons. Annual of the British School at Athens 102: 103-136.

RENFREW C., O. PHILANIOTOU, N. BRODIE & G. GAVALAS. 2009. The Early Cycladic settlement at Dhaskalio, Keros: preliminary report of the 2008 excavation season. Annual of the British School at Athens 104: 27-47.

Research Horizons. Island of broken figurines. June 2011: 26-27. Available at: http://www.cam.ac.uk/research/features/island-ofbroken-figurines/(accessed 8 July 2011).

RUTTER, J. 1983. Some observations on the Cyclades in the later third and carly second millennia. American Journal of Archaeology 87: 69-76.

-- 1984. The Early Cycladic III gap: what it is and how to go about rilling it without making it go away, in J.A. MacGillivray & R.L.N. Barber (ed.) Prehistoric Cyclades: contributions to a workshop on Cycladic chronology: 95-107. Edinburgh: Department of Archaeology, University of Edinburgh.

SCHMIDT, K. 2007. Gobekli Tepe, in Badisches Landesmusuem Karlsruhe, Die altesten Monumente der Menschheia 74-77. Stuttgart: Theiss.

SHADY, R.S.S., J. HAAS & W. CRAMER. 2001. Dating Caral, a preceramic site in the Supe Valley on the central coast of Peru. Science 292: 723-26.

SHADY, R. & C. KLEIHEGE. 2008. Caral. Lima: Universidad de San Martin de Porres.

SOTIRAKOPOULOU, P. 1993. The chronology of the 'Kastri group' reconsidered. Annual of the British School at Athens 88: 5-20.

SOTIRAKOPOULOU, P. Forthcoming. The pottery from Dhaskalio, in C. Renfrew, O. Philaniotou, N.J. Brodie, G. Gavalas & M.J. Boyd (ed.) The sanctuary at Keros and the origins of Aegean ritual practice: excavations at Dhaskalio and Dhaskalio Kavos, Keros 2006-2008. Volume IV. Cambridge: McDonald Institute for Archaeological Research.

SUESS, H.E. 1967. Bristlecone Pine calibration of the radiocarbon timescale from 4100 BC to 1500 BC, in Radioactive dating and methods of low level counting. 143. Vienna: International Atomic Energy Agency.

WILD, E.M., W. GAUSS, G. FORSTENPOINTNER, M. LINDBLOM, R. SMETANA, P. STEIER, U. THANHEISER & F. WENINGER. 2010. [sup.14]C dating of the Early to Late Bronze Agc stratigraphic sequence of Aegina Kolonna, Greece. Nuclear Instruments and Methods in Physics Research B 268: 1013-1021.

ZAPHEIROPOULOU, P. 1968a. [TEXT NOT REPRODUCIBLE IN ASCII.]. Archaiologikon Deltion 23:381-83.

-- 1968b. Cycladic finds from Keros. Athens Annals of Archaeology 1: 97-100.

Colin Renfrew (1), Michael Boyd (1) & Christopher Bronk Ramsey (2)

(1) McDonald Institute for Archaeological Research, University of Cambridge, Downing Street, Cambridge CB2 3ER, UK (Email: acr10@cam.ac. uk; mjb235@cam.ac.uk)

(2) Research Laboratory for Archaeology and the History of Art, University of Oxford, Dyson Perrins Building, South Parks Road, Oxford OX1 3QY, UK (Email: christopher, ramsey@rlaha.ox.ac.uk)

Table 1. Early Cycladic culture names and phases.

Dhaskalio phase Markiani phase Culture name

C -- (Dhaskalio C)

B IV Early Kastri group

A III Keros-Syros culture

-- II Kampos group

-- I Grotta-Pelos culture

Dhaskalio phase Numerical phase Cycladic comparanda

C ECIII Early Phylakopi I (Melos)
 Arkesine (Amorgos)

B ECII/III Kastri (Syros) Ayia Irini
 III (Kea) Panormos
 (Naxos)

A ECII Skarkos (Ios) Phylakopi
 A2 (Melos) Ayia Irini II
 (Kea)

-- ECI/II Kampos cemetery (Paros)
 Simigdalas cemetery
 (Ano Kouphonisi)

-- ECI Phylakopi A1 (Melos)
 Grotta (Naxos)
 Lakkoudes cemetery
 (Naxos)

Table 2. Samples analysed by the Oxford Laboratory.

Sample Trench Layer Ceramic phase

R42 II 39 A
R44 III 2 A
R4 I 32 B
R7 I 44 B
R8 I 18 B
R40 I 34 B
R41 I 26 B
R43 II 43 B
R11 VI 23 C
R13 VI 28 C
R14 VI 35 C
R15 VII 3 C
R19 VII 32 C
R26 VII 41 C
R45 VI 34 C
R46 VI 33 C
R49 VII 38 C
R51 VII 39 C

Sample Material (species) Sampling method

R42 charcoal (Angiosperm) flotation
R44 plant remains (Prunus Amygdalus) flotation
R4 charcoal (Olea europaea) excavation
R7 charcoal (Hedera helix) excavation
R8 charcoal (Olea europaea) excavation
R40 charcoal (Pistacia lentiscus) flotation
R41 charcoal (Olea europaea) flotation
R43 charcoal (Prunus Amygdalus) flotation
R11 charcoal (Juniperus cf. phoenicia) excavation
R13 charcoal (Juniperus cf. phoenicia) excavation
R14 charcoal (Olea europaea) excavation
R15 charcoal (Juniperus cf. phoenicia) excavation
R19 charcoal (Olea europaea) excavation
R26 charcoal (Olea europaea) excavation
R45 charcoal (Phillyrea or Rhamnus) flotation
R46 charcoal (Olea europaea) flotation
R49 charcoal (Fagaceae) flotation
R51 charcoal (Ericaeae) flotation

Table 3. Results of radiocarbon dating. The dates are uncalibrated in
radiocarbon years BP (Before Present AD 1950) using the half life of
5568 years. * Modern date (post-1950) reported as F14C
(fraction of modern).

OxA Sample [[delta].sup.13]C Date BP

OxA-22754 R42 -26.32 4065 [+ or -] 30
OxA-22756 R44 -25.24 3933 [+ or -] 29
OxA-22745 R4 -23.63 4021 [+ or -] 29
OxA-22746 R7 -23.34 3876 [+ or -] 28
OxA-22747 R8 -23.96 4033 [+ or -] 30
OxA-22752 R40 -24.30 * 1.17301 [+ or -] 0.00339
OxA-22753 R41 -22.50 3849 [+ or -] 31
OxA-22755 R43 -25.32 3921 [+ or -] 31
OxA-22748 R11 -22.18 3919 [+ or -] 28
OxA-22749 R13 -23.97 3923 [+ or -] 29
OxA-22750 R15 -22.51 3841 [+ or -] 29
OxA-22751 R19 -25.56 3904 [+ or -] 30
OxA-22761 R26 -23.00 3870 [+ or -] 30
OxA-22757 R45 -23.93 4164 [+ or -] 30
OxA-22758 R46 -22.67 3852 [+ or -] 29
OxA-22759 R49 -24.84 6268 [+ or -] 34
OxA-22760 R51 -24.62 3837 [+ or -] 30

Table 4. Samples from Dhaskalio ordered according
to final stratigraphic phasing.

Sample Ceramic phase Date BP

R42 A 4065 [+ or -] 30
R45 A 4164 [+ or -] 30
R4 B 4021 [+ or -] 29
R7 B 3876 [+ or -] 28
R8 B 4033 [+ or -] 30
R41 B 3849 [+ or -] 31
R43 B 3921 [+ or -] 31
R44 B 3933 [+ or -] 29
R11 C 3919 [+ or -] 28
R13 C 3923 [+ or -] 29
R15 C 3841 [+ or -] 29
R19 C 3904 [+ or -] 30
R26 C 3870 [+ or -] 30
R46 C 3852 [+ or -] 29
R51 C 3837 [+ or -] 30

Table 5. Chronology of the phases in years BC.

 95%
 Weighted probability
Transition mean Median range

Start of Dhaskalio Phase A 2789 BC 2748 BC 3124-2581 BC
Transition Dhaskalio Phase A
 to Phase B 2554 BC 2550 BC 2634-2481 BC
Transition Dhaskalio Phase B
 to Phase C 2392 BC 2395 BC 2452-2324 BC
End of Dhaskalio Phase C 2328 BC 2326 BC 2432-2238 BC

Table 6. Chronology of the phases in years BC after refinement.

 95%
 Weighted probability
Transition mean Median range

Beginning of Dhaskalio Phase A 2746 BC 2742 BC 2885-2622 BC
Dhaskalio Phase A/B transition 2540 BC 2536 BC 2609-2482 BC
Dhaskalio Phase B/C transition 2391 BC 2392 BC 2451-2322 BC
End of Dhaskalio Phase C 2290 BC 2292 BC 2387-2193 BC

 Approximate
Transition consensus

Beginning of Dhaskalio Phase A c. 2750 BC
Dhaskalio Phase A/B transition c. 2550 BC
Dhaskalio Phase B/C transition c. 2400 BC
End of Dhaskalio Phase C c. 2300 BC