12 mins read

Witnessing the dawn of urbanisation in Europe?

The islet of Dhaskalio, as seen from Keros. During the early Bronze Age a causeway probably linked the two.

In a world obsessed with cutting communication times and securing natural resources, it can be hard to understand the ancient allure of Dhaskalio. Situated at the remote tip of a sparsely inhabited island, the site seemingly had little to draw visitors. Yet they came in sufficient numbers to create a type of settlement previously unseen in Europe. Colin Renfrew, Michael Boyd, and Evi Margaritis explained why to Matthew Symonds.

Today, the tiny islet of Dhaskalio pierces the waters of the Aegean 90m offshore from the Cycladic island of Keros. Despite lying at the heart of the Aegean, and enjoying glorious views over the surrounding island chain, neither Dhaskalio nor Keros are currently inhabited by humans. Dhaskalio itself is little more than a rocky mass, clad in scrub vegetation, which rises steeply from the sea. It is a mere 200m in length. Although Keros is rather larger, reaching about 7km at its longest point, it is only a slightly better bet from an agricultural point of view. Most of the land is too marginal to sustain crops, while the absence of marble or metal deposits seemingly left the island with little to attract human interest. Rather being condemned to obscurity, though, both Keros and Dhaskalio had attention lavished on them in the early Bronze Age that was out of all proportion to their modest resources.

The broken head of a marble statue from the southern special deposit on Keros.

The earliest indications that Keros was seen to be in some way special can be dated back to c.2700 BC, when groups of people began voyaging to the island. How precisely these mariners travelled – whether by raft or by boat – is unclear, as is the place (or, more likely, places) where they embarked. What is certain is that stowed aboard their vessels were fragments of pre-broken artefacts such as pots, marble bowls, and marble figurines. To modern eyes such a cargo seems a curious one, but these goods played a key role at what can be described as the earliest maritime religious sanctuary in the world. This status does not seem to have applied to the entirety of Keros; instead, the ritual deposition of the shattered sculptures occurred at two places within bays flanking the westernmost point of Keros. Back then, this probably comprised an imposing promontory reaching out into Aegean. It is this promontory that rising water levels have transformed into the modern islet of Dhaskalio.

Dhaskalio was transformed by constructing terraces from tonnes of imported marble.

While most of the early visitors wished to add their broken marble figurines to one of the special deposits, presumably as part of an act of devotion, at some stage an additional cargo that produced a more tangible benefit was transported to Keros. North of the northern bay containing part of the sanctuary, the land protruded into the sea once more. This area, known as the Kavos Promontory, offered a shelf of flat ground where the winds racing off the Aegean were at their fiercest. Such conditions were perfect for achieving the high temperature needed to smelt metal, and so copper ore was ferried to the island and processed there. When precisely this sideline developed is unclear, but it is certain that, over time, the visitors’ eyes were drawn to the rocky outcrop of Dhaskalio, moored to ‘mainland’ Keros by a narrow natural causeway. By around 2500 BC, the sides of Dhaskalio were clad in terraces made of imported marble, on which a sophisticated settlement was founded. This precociously early development preceded the famous Bronze Age palaces of Crete by several centuries. Now excavation at the site is revealing the extraordinary story of how a marginal site was sustained by supply networks at the very dawn of European urbanism.

A special place

The origins of modern exploration of the site lie in those fragmented marble figurines having drawn visitors to Keros once more, but this time to loot rather than deposit them. In the 1960s, the site came to the attention of Christos Doumas of the Greek Archaeological Service, who mentioned the problem to a young doctoral student studying the Cycladic islands at Cambridge. ‘My involvement goes all the way back to 1963,’ remembers Colin Renfrew, co-director of the project and Senior Fellow of the McDonald Institute for Archaeological Research. ‘Prof. Doumas – as he now is – suggested I visited the island, where there was broken pottery, marble bowls, and a few broken figurines strewn about the place. I undertook surface collection, and then later that year Prof. Doumas did a small excavation and found a few traces of buildings. Later, his colleague Mrs Zapheiropoulou excavated further and formed the view that the artefacts had been broken by the looters.’

Colin Renfrew’s association with Keros stretches back to 1963. This photo was taken en route to Keros in 1987, when Colin (left), Lila Marangou (centre), and Christos Doumas (right) undertook surface exploration of the island.

‘In 1987, Christos Doumas, Lila Marangou, and I received a permit to undertake surface exploration. Survey at that time found some indication of marble bits and pieces further to the south of the looted deposit. Some years later, I thought it would be worthwhile to excavate the area properly, so in 2006 I led a project looking at both the looted site and this area to the south. It proved to be another – undisturbed – special deposit, and it became clear that the material had been deliberately broken elsewhere before being carried from other islands, probably in a respectful manner, and deposited at the site. In 2007 and 2008, we had a good look at the islet of Dhaskalio, which we have concentrated on since work resumed in 2016.’

Excavations have revealed that the sanctuary was the focus of activity from c.2750-2550 BC, before attention turned to Dhaskalio. From around 2550-2400 BC, work was under way refashioning the rocky outcrop into a settlement with buildings supported on marble terraces, created using tonnes of stone imported by sea. This material was probably quarried from the south-eastern tip of Naxos, which lay 10km distant. Meanwhile, the frequency of objects being added to the special deposits at the bays gradually dwindled. Finally, between c.2400- 2300 BC a set of monumental buildings was built on the summit of Dhaskalio, by which time the promontory had completely supplanted the sanctuary in terms of importance. It was not to last, though, and occupation at the site petered out in the years after 2300 BC. Excavation has been teasing out the connection between the sanctuary and the settlement.

‘Clearly Dhaskalio and the sanctuary are a pair’, explains Colin, ‘but you only find the broken bowls and sculptures in the special deposits. There are absolutely no fragments of sculptures from the settlement, and their complete absence is very striking. On the other hand, this does not mean that the settlement was just an ordinary one: it may well have had symbolic status and functions. We have come to the conclusion that the numbers of people at the settlement increased significantly at certain times, perhaps during festivals of some kind, when people came to make their offerings in the special deposits.’

Metalworking seems to have been a feature of everyday life at Dhaskalio. Moulds – including this one, probably for a dagger – have been excavated from a workshop along with over 100 fragments of baking pans, 14 blowpipe fragments, and quantities of spilled copper.

‘So the activity at Dhaskalio is related to what was happening at the sanctuary. Indeed, in a way it was the sanctuary’s counterpart, and it is quite possible to imagine that this was the place where people involved in the ceremonies were living. Obviously, if you’re travelling by boat without mechanised power, you don’t arrive and then immediately leave – you perhaps stay for several days. That visitors had time to engage in metallurgy definitely supports this view. Visits to the island were presumably seen as a special activity, and the way that the terraces on Dhaskalio were constructed to create rising walls in receding planes must have made the site look very impressive to those approaching by sea from the north. While the settlement and sanctuary have different characters, they both seem to have a symbolic significance.’

This is an extract from the full article in featured in issue 91 of Current World ArchaeologyClick here for more information about subscribing to the magazine.


  1. I’m reading Pococke’s “India in Greece”. Pretty obvious the people responsible for these constructions came from India. For instance, he derives the name “Cyclades” from Goclopes, that is, the chiefs from Gocla country on the banks of the Yamuna river in India. Gokul is a town just south of Mathura on the Yamuna from where these people came, still existing after 5000 years, remarkably. This name Dhaskalio can no doubt be derived from a Sanskrit word, as Pococke does with hundreds of place names in classical Greece.

Leave a Reply

Your email address will not be published.