Early June witnessed the start of the second excavation season for the Stélida Naxos Archaeological Project [SNAP] on the eponymous hill and chert source located on what today is the island’s north-west coast, a couple of kilometres south of the harbour town Chora. SNAP has been running for the past three summers, initially as an independent CIG survey (2013-14), then an official collaboration with the Cycladic Ephorate of Antiquities when we turned to excavation mode. Today the project’s co-directors are myself, Dr. Tristan Carter of McMaster University, and Dr. Demetris Athanasoulis, the head of the Cycladic Ephorate, while the team – the largest yet – comprised some 25-35 students and scholars from Canada, Greece, the UK, France, Serbia and the US.

Last year we commenced work on the hill’s western flanks, both up close to the natural chert outcrops where the raw material was being extracted and initially worked, and lower down on what today is a narrow flatter coastal strip. In 2016 we revisited these areas – as none of the previous year’s sondages had been excavated to natural – as well as expanding our work to neighbouring plots, whereby our summer’s work involved the stratigraphic excavation of some 10 trenches. These are relatively limited areas being exposed, nothing larger than a 2×2m sondage, as our focus tends to be deposit-oriented, rather than the architectural foci of most excavations. Indeed we expect to find no built features for Stélida is an early prehistoric quarry and stone tool workshop, that we estimate to have been exploited – likely intermittently – from the Lower Palaeolithic to the Mesolithic, i.e. an estimated ≥250,000 – 9,000 years ago.

The archaeology of Stélida is dominated by the by-products of stone tool manufacture, together with a small quantity of hammerstones, a number of which are made of emery, a raw material native to Naxos, albeit from the north-eastern part of the island, the best known outcrops being around Apeiranthos and Koronos, some 15km linear distance from the site. These artefacts can be found in overwhelming quantities - a large bag-full from every zembil sieved! – and represents one the major challenges facing our work, both methodologically in terms of how one studies so many thousands of items, and pragmatically in terms of having to carry such a weight of material off-site each day, and the space required to store it all. Most of the artefacts recovered can be conceptualised as production debris, with true end-products under-represented if not completely absent from our stratigraphic units, the logic being that having tested the raw material quality, the knappes would have then roughed out cores and/or tool-blanks which would then be transported off-site to their seasonal camps at which point the finished implement would be fashioned. As such it has been a struggle at times to date our deposits using a classic tool-typology approach, with our assemblages quite distinct from other excavated earlier prehistoric sites of the region – as for example the Franchthi and Klisoura caves of the Argolid.

While tool-types and knapping traditions can provide us with a broad-stroke chronology for our material, i.e. ‘Mesolithic’, or ‘Upper Palaeolithic’, the fine-tuned dating system we require to answer many of our research questions can only be achieved through the generation of absolute dates via scientific techniques. To this end we had Dr. Christelle Lahaye (Université Bordeaux Montaigne) visit to take a series of samples for optically stimulated luminescence [OSL] dating, a chronometric technique that is well established for earlier Palaeolithic archaeology; indeed Dr. Lahaye has recently been applying it to some of the classic Neanderthal sites of SW France, while elsewhere in Greece it has produced dates of over 400,000 years old. The sampling itself took place by moonlight… as exposure to the sun would ruin the process, having initially studied in detail the stratigraphy of our main sondages alongside our geo-archaeologists Dr. Dan Contreras (CNRS Aix-en-Marseille) and Justin Holcomb (Boston University).

Having top-notch scientists and methods at hand is alas not the end of the problem however, as Stélida provides us with yet further challenges due to the dynamic nature of its landscape. During the period within which Stélida was being exploited by early modern humans, Neanderthals and earlier hominins (a claim we make based on the types of tools we find at the site), i.e. the Middle Pleistocene – Early Holocene in geological terms, the landscape and environment would have changed radically due to the fluctuating periods of glaciation and warmer inter-glacial eras. During the cold periods the environment was probably quite stable, with steppe-like low scrubby brush covering the hill; however, as temperatures rose and precipitation increased, the shallow roots of those bushes would have provided a poor anchor to the soil, which in the context of heavier rainfalls would have begun to wash downslope, only stabilising once again when warmer climes facilitated the growth of more deep-rooted larger plants and trees. The net result is that we are excavating the remains of a highly dynamic landscape, a steep-sided hillslope that has undergone various periods of major erosion, followed by epochs of stability. Most of what we have dug over the past two years can be interpreted as colluvial (hill-wash) deposits, where everything we find is in secondary context, having originated from upslope knapping floors. That said, the assemblages within many of these strata are quite homogenous, i.e. we get great swathes of Upper Palaeolithic material moving quite rapidly en masse, rather than everything being completely mixed and undateable. In between these colluvial fills we also believe that we can recognise periods of stability, palaeosol development and even surfaces. These claimed different kinds of deposits are being analysed by Dr. Panagiotis (‘Takis’) Karkanas, a world-renowned micromorphologist and Director of the American School’s Weiner Laboratory, who sampled some of our excavation sections in late June.

It does remain however that in certain trenches these erosional episodes provide us with stratigraphic inversion. For instance, our 3m deep Trench 1 provides us with a series of deposits that almost always contain small quantities of Upper Palaeolithic material in association with far larger amounts of what appear to be much earlier Lower Palaeolithic tools. Our working interpretation is that we have a series of stable periods dating to the later Pleistocene, during which Upper Palaeolithic hunter-gatherers worked at the site leaving knapping debris on the surface of the hill. These stable periods were then bookended by periods of climatic and environmental instability leading to major periods of erosion that led to large quantities of soil – containing the remains of much earlier Lower Palaeolithic activity – washing down from the upper parts of the hill and covering the Upper Palaeolithic knapping surfaces, and so-on. As such, we are quite prepared for a series of OSL dates from this trench that do not increase in age as we descend the stratigraphic sequence; this will be yet another challenge to explain to the scientific community in detail as to how such dates are still highly meaningful, and can attest to different periods of deep-time activity at Stélida.

The major discovery of the season was a hearth that we believe to be of Upper Palaeolithic date, a feature that is in situ and located in front of one of the major chert outcrops at the top of the hill, sealed by some 75cm of artefact-packed colluvial deposit. Here finally we have something which is both easily dateable (via OSL, thermoluminescence, and/or carbon-14), and will provide us with a date of an actual series of activities, rather than our usual terminus ante quem results (‘material from deposit X is at least date Y’). We also took a great many soil samples for water sieving with the hope that we will recover charred plant materials, indices of both Upper Palaeolithic cuisine and the local environment, as Stélida’s acidic soils have otherwise destroyed all other organic finds from the other trenches.

For those of you attending the annual meetings of the Archaeological Institute of America in Toronto this January, you can find out more about our 2016 discoveries when we present a paper in a session dedicated to the fieldwork carried out under the aegis of the Canadian Institute in Greece; see you there!

Tristan Carter
SNAP co-director