The Pre-Minoan Landscape of Thera: a Preliminary Statement

The internal topography has so far only been seriously considered by geologists who have concentrated on the broad outline of the volcanic centres. As yet little work has been done on the man-made part of the Minoan landscape, with little consideration of which areas had bare rock surfaces or were suitable for agriculture, or where there were steep slopes or flat land, in order to provide the context for the disposition of Minoan settlements, both known and undiscovered.

During fieldwork the authors have found many sites, some newly exposed, of demonstrable Minoan ground surface with their overlying Minoan volcanic deposits intact, precluding a later origin or contamination. A considerable variety of soil depth and quality, in addition to bare, rocky surfaces, has been noted, together with occasional spreads of Minoan pottery sherds. Several coastal features have also been seen, including sea cliffs now buried by Minoan pumice, allowing an attempt to be made to reconstruct a coastal profile.

So far, this work has consisted of locating sites where Minoan soil can be demonstrated below undisturbed Minoan eruption deposits. The majority of these sites are as yet unrecorded and many have only recently been exposed. This paper records for the first time the broad distribution of current exposures, and offers a tentative reconstruction of the topography of the island in Minoan times. However, the complete survey and recording of all the sites is a much larger task, and could be an ongoing project as new exposures are revealed. The analysis of the information contained within the Minoan soil deposits identified so far represents a separate and larger programme of research, but has considerable potential for understanding the Minoan landscape of Thera and the settlements therein.

AIM OF THE STUDY

The aim of this study is firstly to locate the pre-Minoan eruption land surface on the island of Thera in the Aegean either directly or by means of locating its immediate covering of volcanic deposits; secondly to study this pre-eruption surface and reconstruct the Minoan island topography. We then compare conclusions reached from this evidence with ideas developed previously on the settlement pattern and land use in Minoan times.

It is hoped that the preliminary research outlined here will provide the basis for a future long-term research project which could include the three-dimensional surveying of all of the surviving features of the pre-eruption landscape, and, by using such techniques as remote sensing, the location of the original coastline, the hidden topography and other major landscape features.

INTRODUCTION

The development of landscape archaeology and environmental archaeology has proved a useful complement to site specific work since the early 1970s (Aston 1985; Evans 1975, 1978; Jones 1986; Shackley 1981) and in recent years these techniques have been applied very successfully to sites in the Mediterranean, in particular in Greece, although in the main only preliminary reports have been published to date. Important contributions already completed include work at Myrtos by Warren (1972), on mainland Greece by Bintliff (1977), on Milos by Cherry and Wagstaff (in Renfrew and Wagstaff 1982) and the Minnesota Messenia expedition (McDonald and Rapp 1972).

The Minoan landscape is the only European example of a landscape completely and catastrophically buried in Prehistoric times and remaining so to the present day. The depth of the deposits, the apparent abandonment of the island after the eruption for at least two centuries (the earliest post-eruption, i.e. surface finds of pottery sherds have been dated to 13th/12th century BC, Doumas and Warren 1980) and the consequent lack of folk memory of what the place was like before, with the development of a complex mythology, are all factors which separate Thera from the only comparable Mediterranean sites, i.e. those at Pompeii and Herculaneum in Italy.

Sections around the collapsed caldera and exposures resulting from erosion of Minoan eruption deposits provide an opportunity to see the Minoan landscape in part (Fig. 1), and from this to extrapolate the original form of the island and perhaps something of the earlier settlement pattern (Fig. 3). After critical examination of previous work and their own extensive fieldwork Heiken and McCoy (1984) have produced a coastline reconstruction, together with some major topographical features, and an attempt at a Minoan caldera reconstruction.

THE ERUPTION GEOLOGY

In many places on Thera it is possible to see a similar sequence of Minoan eruption deposits in which there is a consistent series of layers (Pl. 1A, 1B). These deposits commence with Phase 1, relatively coarse-grained lapilli and pumice fragments ('the rose-pink pumice'); this is immediately covered by Phase 2, very much finer grained pumice exhibiting stratified lamination (the 'base surge' deposits), and in turn this is covered by Phase 3, less well sorted pumice deposits with relatively abundant enclosed lava fragments (the 'chaotic' pumice). All of these deposits have been extensively described by many researchers (e.g. Bond and Sparks 1976; Pichler and Friedrich 1980; Pichler and Kussmaul 1980; Heiken and McCoy 1984).

The only significant variation on this sequence is that of the thickness of the different layers, but where the pumice is still present and undisturbed since the eruption it is possible to recognize each unit in its normal position, and thereby confirm that the underlying surface is the Minoan ground surface. Although when complete the sequence of Phases 1, 2, and 3, i.e. 'rose-pink', 'base-surge' and 'chaotic' pumice, is invariably seen, there are substantial variations in thickness of all the units from place to place. In some, for example on Therasia, the 'rose-pink' unit is extremely thin (down to little more than 30 cm), whereas south of Phira it reaches several metres in thickness. The 'base-surge' unit is also of variable thickness, but where the bottom of this unit is exposed the 'rose-pink' will occur in undisturbed sequence, so that recognition of undisturbed pumice is made relatively easy. The only notable example of internal changes in sequences which by the above criteria are undisturbed occurs on the steep slopes below Profitis Ilias and above Mesa Gonia. Here the 'rose-pink' pumice lies directly over Minoan soil on the marble basement, but is itself composed of numerous layers showing reversed grading (Pl. 2A, 2B). These reverse-graded units result from the cascades of loose pumice falling off the steep slopes during the eruption's first plinian phase, and are not the result of subsequent erosion after the eruption ceased. This is demonstrated by the overlying 'base-surge' unit which is still in situ.

Because the consistent sequence described above is universally found on the island, even when the underlying Minoan surface is not exposed, it is nevertheless possible to infer its presence, and even estimate its approximate depth when this sequence is recognized in its undisturbed state.

There are many examples of pumice sequences which are apparently derived from reworked primary deposits. These are particularly thick and conspicuous on the lower slopes of the outer parts of the islands, especially below the main mountains, e.g. Profitis Ilias and Megalo Vouno. In places this material could be supposed to be in situ primary 'chaotic' deposit, but careful examination will usually reveal an abnormal sequence and/or contamination and, therefore, it can be demonstrated that any underlying soil or ground surface may well be post-Minoan, since the precise time of deposition of these reworked deposits cannot be easily determined and must in any case post-date the beginning of the Minoan eruption.

It is notable that the upper part of some of these apparently reworked deposits exhibits evidence of hot deposition, notably in the form of magnetic alignments in included volcanic bombs and fumarolic pipes, both indicating deposition of the upper pumice levels whilst still hot, and therefore that the reworking must have been going on during the episode of the Minoan eruption (Wright 1978, 132). Further evidence of this contemporaneous reworking is seen below Profitis Ilias, as described above for the 'rose-pink' unit.

Bombs (large blocks of pre-existing lava from the vent) deposited on very steep mountain sides indicate that soft pumice deposits must once have cloaked the bare rock now exposed, allowing the enclosed bombs to settle gently on the slope rather as an erratic settles from a melting glacier, and to remain perched in positions in which they could hardly have come to rest from free-fall (Pl. 3). This demonstrates that the steep slopes were once covered not only by the 'rose-pink' pumice which has accumulated at their foot, but also by the overlying units which normally include these large bombs. Further evidence for the complete coverage of the mountain tops is seen in the widespread occurrence of clearly recognizable fragments of pumice from the 'rose-pink' unit, which lie, sometimes in considerable quantities, on soil comparable to the Minoan example in hollows on the rocky surface. This is particularly common on the top of Mesa Vouno and Profitis Ilias.

Although there is considerable evidence for much of the stripping of pumice from the slopes having happened during the various phases of the eruption, as cited above, there is clear historical evidence, in the form of early photographs, that modern erosion, either natural or assisted by man, has substantially increased the areas of bare hillside (Hiller von Gaertringen 1904, vol. 3, Fig. 38; our Pl. 4). In this example the terraced fields below Sellada are now completely removed and the site of the former pumice deposits is occupied by the road from Kamari to Old Thera.

The recognition of the overlying Minoan eruption deposits in situ, as described above, is an essential prerequisite without which it would be difficult, if not impossible, to distinguish with certainty an authentic Minoan ground surface from any other well-developed soil horizon or exposed rock surface.

THE PRE-ERUPTION MINOAN SURFACE

During our fieldwork between 1980 and 1988 we have observed the Minoan ground surface in many places, indicated on Fig. 1, in addition to the numerous inaccessible exposures to be seen, notably on caldera cliffs. These newly recorded examples occur as sections at the sides of roads and paths and in working and abandoned quarries. Perhaps surprisingly there are many examples of exposed surfaces of Minoan soil, mainly in quarries after removal of the overlying pumice and also in the floor of some paths, forming the present ground surface, for example in the valley behind Perissa.

In addition to the currently exposed examples, we have recorded several temporary exposures in 1987 which resulted from the short-lived fashion for 'stone-washed' denims. This caused many small excavations in the 'rose-pink' unit, which contains particles of ideal size for use in treatment of textiles, and incidentally revealed the Minoan ground surface for a matter of a few months, for example above Perissa and Pyrgos and near Mesa Gonia.

Elsewhere the immediately overlying geological sequence has been observed and indicates the proximity of the Minoan surface even though no Minoan surface is exposed, such as on the high ground to the west of Kamari village.

Different Minoan ground surfaces observed have comprised thick brown loamy soil, thin stony soils, barren, relatively unweathered volcanic ashes and lavas, and metamorphic basement rock, i.e. examples of all the rock types and surfaces seen on the island today. Soils, now very firmly cemented and resistant to plant growth, are seen in several places, such as the extensive exposure in the valley above Perissa (Pl. 5) and along the old road from Imerovigli to Mikros Profitis Ilias. Elsewhere soils of various types can be seen, for example in the new exposure (made in 1988) along the west side of the road leading to Oia, below Mikros Profitis Ilias and at many places along the cliff-side path to Oia, in the Phira, Megalochori and Oia quarries, on the slopes below Profitis Ilias, on the path to Pyrgos. and on the island of Therasia (Fig. 1).

In some places, e.g. Oia quarry, Megalochori quarry and the northern side of the Akrotiri peninsula, the old ground surface contains sherds of Minoan pottery in situ, but as yet we have not noted any Minoan buildings, roads or field boundaries. We have been particularly careful in our fieldwork to ensure that any pottery sherds noted were either definitely within the soil, i.e. observed in situ, or embedded in the surface of any exposed areas (Pl. 6). In addition to these sealed Minoan pottery finds, extensive scatters of presumed Minoan pottery occur at Megalochori and Megalo Vouno (on the caldera slopes).

Topographically the Minoan island was in part much as today, for example the mountains of Profitis Ilias and Mesa Vouno and Mikros Profitis Ilias were all features of the Minoan landscape. Elsewhere the landscape has changed dramatically with the addition of substantial thicknesses of pumice to the valley floors and in particular around the outer coastline, which has been extended in places by more than one kilometre as at Monolithos (Fig. 2). However, in detail the topography was locally different; for example, the northern half of the island was dominated by several volcanic cones which considerably affected the thickness of pumice deposited from place to place and gave a more rugged appearance than anywhere on the islands today, with the possible exception of Nea Kameni. The sections on the caldera rim (Pl. 7) exhibit a striking similarity to the rocky scenery depicted in the Spring fresco from Akrotiri.

On the small scale, Minoan topographic features have completely disappeared under the blanket of pumice; small valleys, such as that under Oia, and low hills, such as that revealed in the Megalochori quarry which has produced Minoan pottery (Pl. 6) and was probably a settlement site, have been obliterated, leaving a modern ground surface with relatively flat featureless topography. The only real interruption to this surface comes from the deep erosion gullies (known as potamos valleys) and the ubiquitous man-made terraces.

THE PRE-MINOAN COASTLINE

There is evidently little difference between the modern cliff line in some places, as around the high mountain masses of Mesa Vouno and Mikros Profitis Ilias (Koloumbos), and that of Minoan times. Elsewhere however, the Minoan coastline has been lost beneath thick pumice deposits and the modern coast extended seawards. The former cliffs remain, however, and can be seen running inland from Perissa and Kamari, and around the marble mass of Gavrilos. It is evident that in pre-eruption times the embayment now represented by Perivolos was at least partly occupied by the sea, and that the site now occupied by the town of Emporion was closer to the Minoan shore. Perhaps this would have enhanced its possible importance as a trading centre during Minoan times, as suggested by its modern place name (Fig. 2).

Although the Minoan coast was inundated by pumice and the cliffs overwhelmed in many places, there are today two localities where the original Minoan cliffs are still visible. These are at the south-western point of Therasia and below the town of Oia at the north-west end of Thera. The cliff near the northern end of Therasia is obscured, and the cliff which faced the peninsula of Akrotiri must have disappeared with the enlargement of the caldera during the Minoan eruption. The visible sections indicate far lower cliffs than exist at present in the same areas (Pl. 8A, 8B).

Whereas today there are many extensive areas of relatively flat land, for example the site of the airport and the coastal strip to the north, the area north of Oia and the western side of Therasia, it is clear that in Minoan times relatively little if any of this land was available for occupation. It is probable that the original Minoan cliff line ran some one to two kilometres inland from the present shore, approximately following the break of slope which occurs at around the 100 metre contour line (Fig. 2).

Just as Heiken and McCoy have concluded that the Minoan island was considerably smaller in outline (but probably larger in area), we now suggest that the topography has changed from cliff-dominated shoreline to the present softer landscape with extensive areas of gently sloping ground and many long beaches. The surface area of the present island is approximately 90 square kilometres. This area includes the smaller islands of Nea and Palaea Kameni and Aspronisi as well as Therasia and Thera, and has been estimated in this study by two methods, i.e. counting kilometre grid squares (with coastal intersects each counting as 0.5) and by weighing paper cut-out shapes and comparing with a standard area. Both techniques give the same figure, i.e. 90 square kilometres. We have reconstructed an outline of the Minoan island which is closely comparable with that of Heiken and McCoy from which we calculate that the area would have been about 103 square kilometres (Fig. 3). This includes the area of a possible northern caldera for which evidence was cited at the Third International Congress by Druitt, Friedrich and co-workers, and Heiken and McCoy.

THE SETTLEMENT PATTERN AND LAND USE

Settlements from the Minoan period on Thera have been discussed by Hope Simpson and Dickinson (1979), Doumas (1983) and Wagstaff (1978). The detection of these sites over more than a century has inevitably been confined to the surviving parts of the original island, i.e. from not more than around 72% of the possible area. This is the area, 65 square kilometres, of the modern island which was in existence in Minoan times; a further 25 square kilometres has been added by the eruption, but around 43 square kilometres was also lost in the caldera enlargement. Usually settlements have been recognized from buildings, walls, burials and pottery found under the mantle of overlying pumice, and from pottery found in sections exposing the old land surface. It is perhaps worth emphasizing this, because not only has relatively little survived of the Minoan island, but only an infinitesimally small part of the old ground surface has been exposed, despite all the quarrying activity, and therefore only a thin and arbitrary section through the deposit is visible on which to base any discussion of settlement patterns. However, for there to be so many sites from such a small part of the original island and from so few exposures could well indicate a very densely settled island in Minoan times. Even if the settlement pattern was somewhat irregular in its distribution, with few if any sites in the rugged northern parts of the island, there should still be a significant number of as yet unknown sites in the surviving part of the island. The density of sites already known, mainly in the Akrotiri area, could well be matched elsewhere (Fig. 3). On the geologically similar island of Milos, research (Renfrew and Wagstaff 1982, 22) has indicated around 20 Bronze Age sites in an area of 33 square kilometres; a comparable density of settlements on Thera would predict 62 sites on the Minoan island of which around 39 might be expected to have survived.

No doubt new sites could be found from more systematic and intensive fieldwork even from the relatively small areas available. For example we might expect there to be settlements on the relatively level areas of the Minoan island, for example on the slopes to the north of Oia and on the west side of Therasia.

At Oia a broad gentle slope of perhaps a kilometre or more in width is indicated, and is exposed as an extensive Minoan ground surface with deep soil in the quarry north of the town. Its seaward limit is indicated by the old cliff line exposed on the shore.

Another implication of the different Minoan topography is that the buried coast could contain numerous sites for ports and harbours. Doumas implied this in the Akrotiri region with a 'suggested prehistoric harbour' to the west of the site (Doumas 1983, 13). Around the main mountain mass of Profitis Ilias there are possibilities of harbours under Kamari and near Emporion. The sheltered caldera would have been most likely to have been used as a port, as other such land-locked embayments are elsewhere in the Aegean (e.g. Milos). We suggest this as the most likely area for the major settlement and most important port on the island in pre-eruption times, and the topography suggested by Heiken and McCoy would allow for this possibility.

FUTURE RESEARCH

A number of obvious tasks result from the above fieldwork and observations. Firstly, there is need for systematic fieldwork and recording of the various existing exposures of the pre-Minoan eruption land surface in all its dimensions, i.e. height above sea level, aspect, type of surface and so on. A programme to discover and record the surviving visible Minoan cliff sections and coastline could accompany the detection of the original surface beneath the eruption deposits, using remote sensing equipment. This should enable the production of a three-dimensional model of the island with its former coastline, inlets, bays, and cliffs, together with some idea of the buried topography.

Secondly, where the Minoan surface is visible, its character should be analysed, especially in the case of the soils. As far as we are aware, this sort of analysis has hardly begun, with the notable exception of the recent work on Santorini of Susan Limbrey (1975), while to date this has been confined to the site at Akrotiri. It is anticipated that much could be learned about the pre-eruption landscape, its agriculture and economy.

Armed with this information, some assessment of the former settlement pattern could be attempted. We would surely realize how little of the early landscape is available to us and appreciate the possibility that the Akrotiri site may have been neither the largest nor the most important of the pre-eruption settlements.

Addendum

Evidence was produced at the Third Thera Congress for a water-filled northern caldera during Minoan times. This consisted of observations of inward-sloping Minoan ground surfaces on the caldera rim of Therasia and near Oia by Heken and McCoy, and of Minoan pumice and soil at a position well down the caldera cliffs below Phira town by Druitt. The other evidence was that of abundant blocks of algal limestone (stromatolites) scattered over the Minoan pumice and in some places (Oia quarry) in situ, in the area near Oia. All this evidence points to the probable existence of a partly flooded caldera in the north of the island in Minoan times, and therefore indicates the relative importance of the flatter and more easily cultivatable land to the south and west.

-----------------------------------------------

--------------------------------------------------