Figure 1: Location of the Stelida peak sanctuary (Dieter Depnering)
So here we are, once more reporting on the 2019 season of the Stelida Naxos Archaeological Project
[SNAP], co-directed by Dr. Tristan Carter (Professor, McMaster
University) and Dr Demetris Athanasoulis (Director, Ephorate of Cycladic
Antiquities). So why are we talking about work that happened two years
ago? Well in part there hasn’t been a whole heap to report about given
the travel and work restrictions of Covid-19, but it is mainly because
we have been waiting for a new paper to be published to tell you about
the exciting new discoveries at Stelida, that arguably make it the most
important research-active site on Naxos, if not the entire Cyclades (and
I speak as a member of the Keros project…).
As anyone who has been following our work since the project’s inception in 2013 should know, Stelida
is a major chert source that was exploited throughout prehistory for
the production of stone tools. Two years of geo-archaeological survey
were followed by excavation, with over 40 sondages dug on both
flanks of the hill, and down to the modern coastal plain, with
stratified deposits of up to 5 metres depth containing thousands of
tools whose form, and mode of manufacture suggested the site’s use by
hunter-gatherers from the Lower Palaeolithic to Mesolithic, most likely
by both Homo sapiens and Neanderthals. In 2019 we published our first major paper
on the site’s chronology based upon luminescence dating of Trench 1, a 3
m+ sequence on the upper western flanks of the hill, with dates
extending back to at least 200,000 years old, i.e., the Middle
Pleistocene.
The excavations at Stelida can be challenging. The soils are very
alkaline, which prohibit the survival of earlier prehistoric organics
(plants, animal and/or human remains), and virtually everything we have
excavated is in secondary context. The latter issue is due to a
combination of local topography (relatively steep hillslopes), and the
ravages of deep-time (Ice Age) climatic fluctuation, with our
stratigraphic studies detailing periods of massive erosion due to
increased precipitation. This means that (a) we almost never have an
insight to in situ Palaeolithic activity, and (b) the technique
we use to date the last time the artefact-bearing soils were exposed to
the sun, is detailing the period at which point these deposits came to
rest post-erosion, rather than when the tools were first made, i.e., all
our dates are minimum ages (terminus ante quem).
To address these issues, we decided in 2019 to open a handful of
trenches on the uppermost, flattest areas of the hill, the logic being
that anything we found in these loci should theoretically be in situ.
We had hitherto largely ignored this part of the site because our
2013-14 survey suggested that much of the upper reaches were very
exposed, with outcropping chert rather than well-preserved soil depths,
and a trench excavated in 2017 on the hill’s western shoulder revealed a
75 cm stratigraphic sequence of almost pure lithics, the original soil
having deflated/been blown, or washed away. As much as there was a
research logic to excavating up top, there was also a pragmatic side to
this work, in that two of the supervisors had finished existing deep
trenches in the first two weeks of the season, and I did not want to
open new sondages on the flanks that we could not finish in the
last month of work. The first couple of trenches – as predicted –
produced precious little, the 1×1 metre units on the hill’s saddle being
only 5-10 cm deep before hitting bedrock, and almost no artefacts, the
original soil and stone tools likely having washed away in the Ice Age.
Another trench halfway along the ridge produced the same negative
evidence, so we finally moved up to the southern highest peak, where it
was hoped that some chert outcrops might have protected prehistoric
deposits (Figure 1). They had, but not what we were expecting…
Figure 2: Kristine Mallinson and her team initiate excavation of Trench 44 (Maria Cummings).
As soon as Trench 44 (Figure 2) was opened we found
artefacts, but not just the Mesolithic and Upper Palaeolithic tools that
the survey predicted we would find in this locale, but also quantities
of pottery. It quickly became evident that a dominant form of this
ceramic assemblage was a small drinking vessel, typical of later Bronze
Age Crete, the ‘Minoan conical cup’ (Figure 3). Another sondage
(Trench 47) was rapidly opened nearby, with even greater amounts of the
same kinds of pottery recovered, along with an array of other finds
including stone and metalwork. The significance of these new discoveries
was rapidly apparent, and we took what we hope was the right ethical
decision to stop excavating two weeks early in this area, as we simply
did not have the support system in place to professionally deal with
such material, not least the lack of a conservator, someone we had never
had a need for given the nature of our Palaeolithic archaeology.
Nonetheless, the quantity, character, and location of these finds
allowed us to argue strongly that this part of the site comprises a
‘Minoan type peak sanctuary’, one of only a handful of such sites known
outside of Crete. This hypothesis, along with supporting evidence, has
just been published in the Journal of Greek Archaeology (Carter et al 2021); we here provide a precis of our discoveries.
Figure 3: Minoan style conical cup from the Stelida peak sanctuary
The ceramic assemblage relates to the preparation and consumption of
food and drink, with scores of simple, mass-produced, undecorated
handleless (conical and other) cups, plus lesser quantities of jugs,
cooking vessels and a very few storage vessels (pithoi). This pottery is
overwhelmingly ‘Minoan’ in style, though an initial study of these
vessels’ fabric suggests that most of it was made on Naxos, i.e., it
represents a local emulation and appropriation of Cretan products and
practices. The other finds include over a thousand beach pebbles of
various raw materials, items that we believe were collected en route
by the worshippers to be dedicated at the site, a practice known from
many other peak sanctuaries, such as Atsipadhes in western Crete.
Arguably the most significant of our other finds was a complete stone
ladle of banded marble from Trench 47 (Figures 4-5).
These vessels have long been associated with ritual activities,
primarily at a group of sites in north-central Crete – Knossos, Juktas
and Archanes – plus a handful of high-profile non-Cretan loci, including
peak sanctuaries on Kythera, and Kea, plus a grave at Mycenae. Given
their small size, shallow interiors, and heart-shaped (cordiform)
outline, it has been suggested that they were used for blood libations.
Figure 4: Stone ladle from Trench 47 when first discovered (M. Pareja)
Figure 5: Stone ladle when cleaned (S. Crewson)
The other finds included a few bronze items, including the bottom
half of a male anthropomorphic figurine, plus some small bronze strips
(all of which again have good comparanda from Cretan peak sanctuaries),
plus our first assemblage of animal bones, attesting to the on-site
consumption of sheep/goat, and cattle. Charred plant remains also
survive in these uppermost, more recent strata, with considerable
evidence for burning in the form of ashy deposits and burnt stone. The
lack of cereal grains suggests that most of the plant material relates
to fuel; while the wood charcoal has yet to be studied, analysis of
phytolith samples indicates the presence of local grasses, and bushy
herbs such as vetches and thyme, the latter potentially being chosen for
its aromatic properties as much as its flammability. There is also
evidence for burnt palms, though these may have been in the form of
baskets, or matting that had been deposited at the site, rather than
being brought here as fuel alone.
The establishment of fires atop the hill may have been
multi-functional, in part for the cooking of foodstuffs consumed in
ritual activities, but also for communicative purposes, either smoke
signals during the day, or as a nighttime beacon. Indeed, the setting of
the sanctuary at Stelida is impressive, with a viewshed analysis using
geographic information systems (GIS) showing that on a clear day anyone
standing at the southern peak would not only have a dominant view over
western Naxos, and the surrounding islands (Ios, Paros, Thera), but as
far south as Crete, and as far north as Kea, Euboea and across to Ikaria
and Chios (Figure 6).
Figure 6: Viewshed from the summit of Stelida; islands outlined in green are visible (C. Lopez)
Central to the site, is a small (~6 × 7 m) stone structure, the actual peak sanctuary,
alas most of this building is covered by a modern communications tower,
however its outline is visible, a double walled construction of faced
chert blocks, with rubble between, some 1 m in width, potentially
suggesting a tower-like support. The construction of buildings at peak
sanctuaries is something we associate with the Neopalatial, or Second
Palace period on Crete, when mountain/hill-top ritual seems to shift
from an inclusive, rural tradition evidenced across the island, to a
more exclusive practice associated with (and/or controlled by) some of
the palaces. The dating of the Stelida pottery suggests that our
sanctuary indeed falls within this later period, and potentially
represents a link between north-central Cretan populations and Naxos, an
argument that has previously been suggested for the political dynamics
underpinning the construction of a peak sanctuary at Agios Georgios on
Kythera. More specifically, the pottery is thus far dated Middle Minoan
IIIB – Late Minoan IA in Cretan terms, which predates the Theran
eruption. We do, however, have tiny shards of volcanic glass from the
soil (as detailed in the phytolith study) indicating that the ash cloud
from this cataclysmic event did cover the site. There are also a handful
of pieces of pumice, which again could theoretically have been
deposited naturally at Stelida, though might conceivably have been left
there during post-eruption ritual activities (the peak sanctuaries on
Kythera, and Kea are known to have existed in this Late Minoan IB
period).
On reflection, the discovery of a peak sanctuary at Stelida might not
come as a complete surprise. While Naxos was never viewed as a
significant participant in Cycladic – Minoan socio-economic relations in
the later Bronze Age (with the so-called ‘western strong’ of Thera,
Melos and Kea being viewed as the major players of the era), Andreas
Vlachopoulos’ recent reappraisal of the finds from Grotta, the major
Bronze Age port of Naxos, suggests that certain Naxians were in fact
very well connected with Crete. In turn, we have the work by Olga
Philaniotou and Robin Barber at Mikre Vigla, a few kilometers down the
coast from Stelida, where terracotta figurines, plaster, and Minoan
style pottery has long suggested a ritual hilltop site of Cretan style.
With Stelida dominating the horizon for anyone looking southwards from
Grotta, some 1.5 km distant, and the highest peak in the immediate
vicinity, Stelida fulfils many of the criteria we have come to associate
with the location of peak sanctuaries.
The 2019 discoveries, that we can finally share with you
post-publication, take SNAP in an exciting new direction. This is not to
say that we are turning our back on the Palaeolithic activity, far from
it, but it does provide us with new research questions (and a number of
us have also long worked on Minoan archaeology), and perhaps more
importantly, new space for junior scholars. Thus, one long-term team
member, Kristine Mallinson (Missouri), has shifted her PhD project from
mainland Bronze Age studies, to a Stelida-focused piece, working on the
conical cups and issues of Cycladic-Minoan relations. In turn, Shannon
Crewson (McMaster) is developing a museological-oriented PhD on how one
represents deep-time archaeology to the public, working through some of
the recurrent themes we see at Stelida, from Pleistocene to Anthropocene
(resource extraction, communication and vista, social gathering inter alia).
Now we look forward to further work on the site in collaboration with
Dr Athanasoulis, more studies, more publications, and for now, more
grant applications…
Professor Tristan Carter
Department of Anthropology, McMaster University; co-director, SNAP
