The islands of Kyra Panagia and Yioura, at the northernmost end of the Sporades Archipelago, off the east coast of Greece in the northern Aegean Sea, have not been inhabited at any point in recorded history, which has led to them being called the Deserted Islands. However, archaeological exploration on the islands has revealed the presence of at least intermittent Human occupation in the Mesolithic and Neolithic. The sites of Agios Petros on Kyra Panagia, and Cyclopes Cave on Yioura have yielded obsidian tools from the island of Milos in the southern Aegean, and ceramics which show affinities to items from the Greek mainland, other Aegean islands, western Anatolia, and the Balkans, as well as some which appear to be distinctly local in origin, including white pottery with red decoration and rod-headed figurines.
The presence of obsidian from Milos suggests that the islands were part of a wide-ranging trade network and therefore are an important part of the story of the peopling of the Aegean. Investigation of the pottery assemblage at Cyclopes Cave has yielded a rich assemblage of pottery stiles, which were subsequently shown by geochemical analysis to have originated from different areas, suggesting that the site was visited sporadically by fishermen from different areas. However, as well as ceramics which appeared to have been made elsewhere in the Sporades Islands, the wider Aegean area, or the Greek mainland, some of the material appears to have been local in origin, made somewhere within the Deserted Islands.
The islands of the Sporades Archipelago are separated by short distances, and the currents around them are generally considered favorable to shipping, which would have made them an ideal route for ancient travelers moving between the Greek mainland and western Anatolia.
In a paper published in the journal Archaeological and Anthropological Sciences on 16 January 2023, Archontoula Barouda and Patrick Quinn of the Institute of Archaeology at University College London, and Nikos Efstratiou of the Department of Archaeology at Aristotle University, look at a selection of pottery from Agios Petros on Kyra Panagia, an Early-Middle Neolithic settlement on a sheltered bay, which is considered a potential point of origin for the 'local' pottery found at Cyclopes Cave.
A total of 39 pottery shards were selected from the Agrios Petros material for inclusion in the study. Due to the destructive nature of the tests, these were all small pieces, and did not include any of the rare and precious rod-head figurines, though a variety of different ceramic styles were included. Thin sections were taken from all the samples and subjected to petrological analysis and comparison to ceramics from Cyclopes Cave and elsewhere around the Aegean, as well as to mineralogical samples collected from sites in the Deserted Islands. Thirty seven of the samples were also subjected to neutron activation analysis to determine the proportions of rare elements within them, which can be an effective method for determining the origin of ceramics and other materials.
The results of these tests suggested that the majority of the Agrios Petros material, unlike the majority of the Cyclopes Cave material, was probably local in origin, probably manufactured in or close to the settlement. The manufacturing style used appears to have arrived in the Deserted Islands in the late seventh millennium BC, and to have continued relatively unchanged for the next thousand years, suggesting a strong local identity, despite clearly being on contact with other cultures around the Aegean Sea.
Agios Petros is thought to be one of the oldest settlements in the Sporades Islands, with occupation beginning in the Early Neolithic, based upon ceramics which can be calibrated against the Thessalian Pottery Sequence (a dating system based upon widely traded ceramics from the Greek Neolithic, and radiocarbon dates obtained from the site; 5740-5520 BC from the second phase at the site, and 4860-4500 BC from topsoil at the site. The majority of cultural activity at the site appears to have taken place in the late sixth millennium BC, or the Early Middle Neolithic. The site was first excavated in the late 1960s and early 1970s by archaeologist Dimitris Theochares, who identified three main archaeological material producing units. In the 1980s, Nikos Efstratiou re-investigated the site, who found that the archaeological material extended downwards for 1.3 m, and that although this could be divided into three units, these were part of a cultural sequence, with no sudden changes between the units.
A variety of cultural items have been recovered from the site, such as stone tools, obsidian, and ceramics including the distinctive red-on-white ware and the rod-head figurines. These items show affinities to other material from elsewhere in Greece, Anatolia, and the Balkans. As well as the red-on-white ware, pottery styles include a variety of styles, including open and closed shapes, painted, black-topped, incised pointillé and red monochrome wares. The red-on-white ware is considered unique to the islands, but shows influences from elsewhere, including the typical carinated shapes from Çatal Hüyük West and Hacilar I in Anatolia, as well as Elateia in central Greece, and linear painted motifs known from Early Neolithic sites such as Chaeronea in Boeotia, southern Greece. Pottery with red-on-white decoration was actually quite common in central Greece, and on the island of Euboea, during the Early Neolithic, with the earliest examples from the northern Sporades Islands dating to the late seventh millennium BC.
As well as pottery items, other ceramics are present at Agios Petros. These include the rod-head figurines, several clay ladles, a clay ear-plug. The figurines are similar to the naturalistic figurines produced in Thessaly in the Neolithic, but also show some similarities to figurines from Anatolia and the Balkans. The site also produced a lot of obsidian, particularly from the upper layers. This appears to be more than is likely to have been needed at the site, possibly suggesting the material was traded from there. The material from Agios Petros, particularly the red-on-white pottery, shows close affinities to the material from Cyclopes Cave, strongly suggesting the sites are linked.
A more recent study, led by Patrick Quinn, carried out a neutron activation analysis on 63 fragments of red-on-white pottery, 24 of which were from Agios Petros, for six elements (cobalt, cerium, europium, iron, scandium, and thorium), and compared these to ceramics of the same age from Dimini and Sesklo in Thessaly, finding the Sporadian material to be distinct from the Thessalian.
Barouda et al. selected 37 fragments of pottery for their study, including the full chronological range of the Neolithic occupation of the site, and as many different decorative styles as possible, although the small nature of the fragments means that the exact form of the complete items from which they came is generally unknown. As well as the pieces of pottery, material from one ceramic ladle and one chisel were included in the study, but none of the figurines, as these were seen as to rare and precious. These samples were analysed for 33 elements (aluminium, antimony, arsenic, barium, calcium, cerium, cesium, chromium, cobalt, dysprosium, europium, hafnium, iron, lanthanum, lutetium, manganese, neodymium, nickel, potassium, rubidium, samarium, scandium, sodium, strontium, tantalum, terbium, tin, thorium, uranium, vanadium, ytterbium, zinc, and zirconium).
The samples were further subjected to X-ray diffraction analysis to determine the presence of different mineral phases. For this test about 5 g of each sample was taken, ground into a fine powder, then divided into 0.5 g aliquots which were then refired at 700, 750, 800, 850, 900, 950 and 1000 °C for one hour each, prior to analysis. Sub-samples taken from these were then examined under a scanning electron microscope to determine their clay vitrification microstructure.
In order to determine the origin of the ceramics, the results of the tests were compared to the results of similar tests made on pottery samples from Neolithic sites in Thessaly and Northern Greece, as well as to clay and hard rocks samples collected from sites around the island of Kyra Panagia. The geology of Kyra Panagia comprises folded Cretaceous limestones overlying older Cretaceous conglomerates and sandstones, overlying Jurassic limestones, overlying ophiolitic basement rocks. In the east of the island low-grade metamorphism has produced schistose rocks, principally phyllite, and metavolcanic rocks, ranging from basalt to rhyodacite.
A variety of different mineralogical fabrics, which are determined by both the composition of the raw material and the technology used to turn this into ceramic. were found within the Agios Petros sample-set. The most common of these is referred to by Barouda et al. as the Foliated Limestone Fabric, and is characterized by abundant metamorphosed limestone and smaller amounts of quartzite, bound within a red clay matrix. This fabric shows considerable variation in the size of the limestone particles, with a range of other rock types, such as phyllite and volcanic material, present as rare inclusions. Pottery fragments showing this fabric were decorated in a variety of ways, including red monochrome, black-topped, black-topped painted, black-on-red and red-on-white. This dominant fabric is very similar to the dominant fabric of the pottery fragments from Cyclopes Cave,
A number of less common, but clearly distinct, fabrics were also found in the Agios Petros material. One of these, referred to as the Metapsammite Fabric by Barouda et al., contains a large proportion of rock fragments, apparently from a lightly metamorphosed sandstone, in an iron-rich clay matrix. The fragments of pottery made from this material were decorated with a monochrome glaze and incised pointillé decoration, and were found in the Late Neolithic strata at the site. This fabric has not been reported from Cyclopes Cave, but one pottery fragment with metapsammite inclusions has been found. Another fabric, which Barouda et al. call the Metamorphic and Volcanic Fabric, contains a mixture of sand-sized metamorphic inclusions and volcanic fragments in a non-calcarious clay matrix. Some of the inclusions within this fabric appear very similar to inclusions within the Matapsammite Fabric. The volcanic fragments comprise amphibole and plagioclase feldspar phenocrysts, held within a dark glassy matrix which appears to be basalt. This fabric was found only on two pieces of pottery with incised pointillé decoration, from the uppermost layer at Agios Petros, and has not been reported at Cyclopes Cave, although fragments with volcanic inclusions are fairly common.
Another fragment from the uppermost layer at Agios Petros contained fragments of micritic calcite and low grade metamorphic material, and is referred to at the Micrite and Phyllite Fabric by Barouda et al.. This fabric is again unknown from Cyclopes Cave, but several fragments from there, reported as having a Phyllite Fabric, are essentially the same, except in that they lack micrite inclusions. Finally one fragment from the late Middle Neolithic at Agios Petros is comprised of fragments of grog (recycled pottery) and smaller amounts of phyllite in a non-calcareous clay matrix, something Barouda et al. refer to as the Grog and Phyllite Fabric.
The elements ytterbium and lutetium were shown to have error margins of more than 10% in the elemental analysis, and were therefore excluded from the study, but all the other elements were used, although, for comparison with the Cyclopes Cave material, elements which were not mapped here were ignored in the Agios Petros material. Sufficient variation was present in the Agios Petros divide it into several chemical groups (implying different origins), with a 77% overlap with the Cyclopes Cave material. Notably, there was a strong similarity between the Foliated Limestone Fabric samples from both sites, and most of the variation within the samples assigned to this fabric relating to the amount of limestone used as temper (material added to the clay matrix to prevent shrinkage during firing). When calcium, which makes up more than 10% of the composition of all the Foliated Limestone Fabric samples, was removed from the comparison, the Agios Petros samples could be split into two subgroups, with all the Cyclopes Cave material corresponding to one of these.
The Metapsammite Fabric samples from Agios Petros are (unsurprisingly) geochemically distinct from all other material from the site, but very close to one-another. The Grog and Phyllite sample is very similar to material from Cyclopes Cave, though no samples from Agios Petra resemble the Phyllite of Fine Mica and Quartz, Serpentinite, or Tuff fabric samples from Cyclopes Cave chemically.
Mica samples from some of the samples were petrologically similar to micas from geological samples from the southern part of the island of Kyra Panagia. Some of the rock samples from this area, thought to be a hard rock when collected, were found to be a e limestone with aligned sparry crystals, very similar in composition to the limestone present in pottery samples with the Foliated Limestone Fabric, and when fragments of this were mixed with of fine red non-calcareous terra rossa clay from the Agios Petros bay, something similar to the Foliated Limestone Fabric samples was produced. Samples of metamorphic rock samples from the island produced quartz and mica crystals very similar to those seen in samples of the Metapsammite Fabric. Furthermore, examples of these minerals were found within an alluvial clay from a river valley on the island, which, when mixed with some crushed metamorphic rock, was very similar to the Metapsammite Fabric. Barouda et al. also note that phyllites collected in the field on Kyra Panagia match closely with those found in the Grog and Phyllite and Micrite and Phyllite fabric samples.
Geochemical analysis of terra rosa clay samples from Kyra Panagia produced results which correspond closely to the clays present in the Phyllite Fabric from Cyclopes Cave. The alluvial clay from Kyra Panagia is geochemically similar to the clay used in the Metapsammite Fabric pottery from Agios Petros.
Analysis of field samples from Kyra Panagia combined with data from earlier geological studies of the Deserted Islands suggests that almost all of the pottery fragments from Agios Petros could have been made with clays gathered om Kyra Panagia or Yioura. The Foliated Limestone Fabric, associated with the distinctive red-on-white decoration of the Deserted Islands contains mineral inclusions unlikely to have come from anywhere but the islands, and the Metapsammite Fabric, Grog and Phyllite Fabric and Micrite and Phyllite Fabric are all good matches for local clays and minerals, The Metapsammite Fabric was previously considered problematic, since the minerals within it are a good match for metapsammite rocks on Yioura, these rocks are located on a steep hillside, difficult to access, Barouda et al.'s study has found outcrops of similar rocks on Kyra Panagia which are much more accessible, and therefore considered the likely source of this material. Since Cyclopes Cave is itself located on a steep hillside, and was only apparently only occupied seasonally, it is likely that the majority of the pottery here came from Kyra Panagia.
Although most of the material used to make pottery in the Deserted Islands comes from Kyra Panagia, and Agios Petros seems to have been the only settlement on that island, the no evidence of actual pottery making has actually been found on the island. Nevertheless, Barousa et al. believe it is likely that the pottery was made in or close to Agios Petros. The presence of several different fabrics, and two distinctive geochemical subgroups within the Foliated Limestone Fabric, strongly suggests that different clays were being used by the pottery makers, although it is not possible at the current time to tell quite where the clays were being collected. This in itself raises the possibility that pottery manufacture might have been occurring at more than one location on the island.
These results suggest that much of the pottery from Cyclopes Cave had a more local origin than was previously assumed. The Phyllite Fabric, the Grog and Phyllite Fabric and the Clay and Phyllite Fabric from Cyclopes Cave, previously assumed to be imports from a more distant location, are now considered highly likely to have been made on Kyra Panagia. The Schist Fabric, Polycrystalline Quartz Fabric 1 and Polycrystalline Quartz Fabric 2 from Cyclopes Cave could all also potentially have been produced on Kyra Panagia. The only inclusions in any of the pottery that cannot be explained by a local origin are the volcanic material found in the Metamorphic and Volcanic Fabric, and more rarely in other fabrics, as no volcanic material is known from the Deserted Islands. However, other islands in the Northern Sporades do produce volcanic material, including the islands of Alonnisos and Psathroura. Stone tools made from andesite have also been found at Cyclopes Cave, and it has been suggested that these could have been made from rock sourced from Psathroura.
Based upon the current information, only the Tuff Fabric, Serpentinite Fabric and Fine Mica and Quartz Fabric reported from the Cyclops Cave cannot be explained by locally sourced materials, and may have been manufactured outside the Deserted Islands.
A number of different sources of raw materials appear to have been used to make the pottery found at Agios Petros. The same source of raw material is often used to make pottery of different styles; for example the Foliated Limestone Fabric is associated with pottery decorated in the monochrome, black topped. early red-on-white, and late red-on-white styles, as well as being used in the chisel and ladle from the site. However, while the other decorative styles are also associated with other fabrics, all of the red-on-white pottery has a Foliated Limestone Fabric. It is also noteworthy that the incised pointillé decoration style, whole found on pottery fragments with several different fabrics, is never associated with the Foliated Limestone Fabric.
Most of the ceramic material has been made using fine clays, to which a temper of crushed rock, usually limestone, has been added. This limestone is likely to have been chosen both for its easy availability on the islands, and its functional properties. The limestone tends to form elongated platy flakes when crushed, which, when added to pottery, will tend to prevent cracks from spreading, adding to the strength of the pot. Limestone also has a thermal expansion coefficient similar to that of clay, which is to say that it expands and contracts at the same rate as the clay when heated and cooled, making it less likely to break during firing. Limestone does have the drawback of breaking down at temperatures above about 750°C, although this does not seem to have been seen as a problem by Neolithic potters, with limestone tempers common throughout the Greek Neolithic.
The used of grog (crushed ceramic) as a temper is also widespread in the Greek Neolithic. However, the only sample from Agios Petros in which this was found also had a phyllite temper, suggesting that the grog had been added for some other reason, possibly ritual.
The Fine Mica and Quartz Fabric recorded from Cyclops Cave includes only very fine mineral grains, which were probably present in the original clay and not added as a temper. Fragments with this fabric were decorated with a matt paint, unlike anything else from either of the two sites, but known from Dimini, Platia Magoula Zarkou, and Makrychori in Thessaly, possibly indicating this pottery was imported.
The majority of the pottery appears to have been made by coil-building (making a long thin cylinder of clay by rolling, then making it into a coil with the desired shape, and smoothing over to give an even finish). This can be seen when thin sections of the pottery are examined under a microscope, as the mineral inclusions are arranged into a pattern of concentric rings, corresponding to the original coils. Patterns of markings parallel to the margins of the pottery are thought to represent working with a paddle and anvil, or similar tools. A variety of finishing methods have been used on the pottery, including incising, burnishing, slipping (coving with a thin layer of a clay slurry), painting and partial reduction. The typical red-on-white pottery of the islands is covered by a a fine-grained light-coloured slip, which covered the red clay beneath, and created a 'canvas' onto which decorative patterns could be painted. This slip appears to have been made from the same clay and limestone as the structural mix, albeit in different proportions, and using limestone ground much finer. The slip also appears to be richer in iron and aluminium than the base mixture, possibly suggesting the addition of haematite as a pigment, which would have turned the slip red on firing.
None of the fragments appears to show any loss of clay minerals associated with firing, which suggests that none of the samples was fired to above 800°C. It has been suggested that the pottery of the Deserted Islands is 'incompletely fired', or at least less well fired than contemporary pottery from the Greek mainland. Barouda et al.'s findings agree that the Deserted Island pottery was not highly fired, but they suggest that the fact the pottery has survived for about 7500 years strongly implies that it was sufficiently well fired for its purpose. Since the Foliated Limestone Fabric pottery would have spalled (suffered a breakdown of the limestone fragments in the clay mix) had they it been heated to above about 800°C, it is quite possible that excess heating was actively avoided by the potters, which would in turn imply a good understanding of the raw materials they were using.
All of the samples appear to have been thoroughly oxidised during firing, implying a good control of the atmosphere within the kilns used. The only exception to this is the black-topped samples, which appear to have had the upper part of the vessel deliberately starved of oxygen, again implying a good understanding of the processes being used. This method of achieving black-topped ware has been documented from some sites elsewhere in Greece, but at others the same effect appears to have been achieved by using manganese or graphite as a pigment.
Since no direct evidence for pottery making is known from anywhere in the Deserted Islands, it is hard to assess the methods used by the potters to fire their pots. Nevertheless, based upon the low temperatures at which the pottery was fired, it is likely that this was done in a bonfire or pit. Using such a method it is possible to control the oxygen supply to the pottery being fired by covering it with fuel material to cut of the oxygen supply, or uncovering it to restore it. This method of pottery firing is known from the Neolithic in other parts of Greece, as well as Western Anatolia and the Balkans. What was used as a fuel for the firing is also unclear, though there is evidence that the islands were more wooded during the Neolithic than today, and the remains of Goats, whose dung can be used as a serviceable fuel, are also known from this time.
Previous analysis of pottery from Cyclopes Cave in isolation led to the conclusion that the variety of material found was indicative of pottery being imported from sites around the Aegean. Barouda et al.'s study of the material from Agios Petros reaches a reverse conclusion, that most of the pottery from both sites was, in fact, local in origin. Despite the clear importation of obsidian from Minos, the inhabitants of Agios Petros do not seem to have been part of a wide network trading in ceramic items, with only a few Late Neolithic items from Cyclopes Cave being apparently incompatible with local manufacture. With the exception of these few pieces, which may have been left by travelers from Thessaly passing across the Northern Aegean, none of the material appears to have been transported more than about 3-5 km from where it is likely to have been made, presumably being used at Cyclopes Cave when the site was occupied annually for some seasonal activity.
Although the pottery of the Deserted Islands has its own distinctive character, it also shows stylistic and technological similarities to pottery from other areas in Anatolia, the Balkans and Greece, which probably both relates to the origins of these people and their continued contact with other peoples around the Aegean.
The ceramics from both sites appear to have been produced on Kyra Panagia, and probably in or close to the settlement of Agios Petros, although this conclusion is essentially based upon the absence of any other known settlements on the island. No evidence of actual pottery-making has been found by any survey carried out on either of the islands, although this is not inconsistent with the predicted method of manufacturing, with bonfires and/or small pits not always expected to find traces that can be found after thousands of years. The existence of several chemical groups within the ceramics, including two within the dominant Foliated Limestone Fabric, could potentially indicate several different manufactories on the islands, using material from slightly different sources, although it might just indicate that the availability of clay was limited, and that the potters switched to different sources as local reserves became extinct.
The presence of a variety of decorative styles and fabrics within the ceramics of the Deserted Islands could be explained in a number of ways. Firstly, these could be the work of artisans following different traditions, albeit within a fairly localized area. Alternatively, different fabrics and styles may have been used to make different objects, or objects used for different purposes. The limited, and fragmentary, nature of much of the material makes this hard to assess, although Barouda et al. do note that the Foliated Limestone Fabric is associated with a variety of different objects. The fact that the different fabrics appear to persist for long periods of time implies that the potters did not simply use the material at random, but were working to strict, pre-determined recipes. The presence of incised pointillé decoration only in the uppermost layers at Agios Petros, combined with the fact that it is never combined with red-on-white ware or the Foliated Limestone Fabric, may suggest the arrival of new cultural influences on the islands, probably from the Eastern Aegean, where this form of pottery is most common. There may also be a connection between this and the arrival of pottery thought to be of non-local origin in the Late Neolithic at Cyclopes Cave.
As well as the Neolithic pottery, Cyclopes Cave has produced earlier, Mesolithic items, including t fish bones, hooks, and chipped stones, implying that the Northern Sporades were inhabited before the Neolithic, and it has been suggested that there may also have been a Palaeolithic occupation of the area, although the evidence for this is uncertain. Given this, the sudden appearance of a fully developed pottery tradition in the Late Early Neolithic at Agios Petros, and the Early Middle Neolithic at Cyclopes Cave, without any intervening experimental stage is quite surprising, and presumably attests to the technology arriving fully formed from elsewhere, before surviving relatively unaltered in the islands for several thousand years. The precise origin of this technology is unclear; the pottery bears similarity to styles from the Greek Mainland, Anatolia and the Balkans, but cannot be more directly linked to any one tradition, having instead a somewhat 'multicultural' feel. Barouda et al. suggest that this apparent uniqueness is likely to be linked to other, as yet unknown, sites in the Sporades or around the Northern Aegean.
About 2000 years after the appearance of pottery in the Deserted Islands, the pottery makers appear to have left, never to return. The islands appear to have been briefly occupied a few times during the Early and Middle Bronze Age, then abandoned. Today the islands form part of the Alonissos and the Northern Sporades National Marine Park, the largest protected marine conservation area in Europe, and are therefore protected from new developments and unlikely to be re-occupied, but in the sixth and seventh millennia BC they appear to have been home to a flourishing and unique culture, with its own distinctive craft traditions.
See also...
