The Cyclades in the Middle Bronze Age

A break between EC III (with its varied aspects in different parts of the Cyclades) and early MC is tentatively postulated at the two major sites, though their histories may have been different. At Phylakopi, the break may not have occurred until after the beginning of MH on the Mainland.

The MC settlement pattern seems to represent a consolidation of the new pattern that had been established in EC III when it was probably accompanied by the arrival of new people in the northern Cycladic and the displacement of others.

Sites are strategically placed and/or fortified. The two best known town sites are well planned and substantial and suggest a solidly adequate, if unspectacular way of life. This standard is reflected in the burials. A variety of burial customs are known of which the cist tomb is inherited from the preceding period and others seem related to MH Greece.

The Cyclades appear to have had their own distinctive pottery classes, notably MC burnished and Cycladic White. The former shows connections with Classic Minoan and is widespread though with local variations, the latter seems mainly from Melos.

Against a common background, the island appears to have developed individual traits until both these were submerged under Cretan influence at the beginning of the LBA.

The recent resurgence of activity in the field of Cycladic prehistory has given us, amongst other things, a chance to reassess one of the least understood phases of the Aegean Bronze Age, the Middle Cycladic. In this paper, I propose a definition of that period, describe its main characteristics and discuss briefly some outstanding problems.

DEFINITION

The Second City at Phylakopi is clearly defined stratigraphically between two major destruction levels (Barber 1974, 4-5 & references). Its MBA character is indisputable and connections with MM Crete and MH Greece (op. cit., 5,30) make it a sound basis for the definition of the period.

There are clear indications that Phylakopi II was destroyed in MM IIIB (or possibly a little later) in Cretan terms. But there is no such certainty about the chronological relations of its earlier stages. At present, these seem most reliably related to MH/MM II, possibly preceded by a gap in occupation. Since late Phylakopi I also seems to have MH features, this gap may have occurred within the MH period on the Mainland.

Preliminary reports of the Ayia Irini excavations on Kea suggest a MBA history of the site not inconsistent with that of Phylakopi. Phases D and F (Caskey 1972, 375, 386) may be related to Phylakopi II in terms of the imported pottery. At Kea too there is uncertainty over the earlier stages. Phase D may not have followed immediately on Phase C and there are no imports in these levels certainly earlier than MM IIA (op. cit., 375).

THE BEGINNING OF THE MIDDLE CYCLADIC PERIOD

At Phylakopi an early (pre-MM II Kamares) phase of the Second City was postulated by the original excavators on the basis of certain pottery finds but this was not stratigraphically attested (Atkinson et al. 1904, 258).

The preceding stage (late Phylakopi I = EC III) has some MH characteristics which appear partially paralleled in the earlier MBA at Kea (D) (Barber 1974, 48-50 & references).

At Kea, Phase C has material of the so-called Lefkandi I culture, distantly related to EH III (Popham and Sackett 1968, 8). The characteristic pottery can be seen on other Cycladic islands (especially Syros) as well as in Anatolia and at some sites on Euboae and the Greek Mainland (Howell 1974, 84-86). It has not so far been found at Phylakopi or in the Southern Cyclades and its northerly distribution is emphasized by the location of the finds outside the islands.

A tiny scarp of evidence for possible contemporaneity between the cultures of Lefkandi I (Kea C) and late Phylakopi I comes in the form of an unpublished sherd from among the predominantly Phylakopi I-type material from Paros. This sherd is of a characteristic burnished tankard of Lefkandi I type. The grave finds from Syros, however, are from an otherwise EC II cemetery (Tsountas 1899, pl. 9, nos. 5 & 7).

It thus seems likely the EC III period has at least two aspects (Lefkandi I and late Phylakopi I) which have some geographical significance. There may also be a third, consisting essentially of a suvival of EC II cultural features (Renfrew 1972, 534).

We should recall too, at this point, the clear stratigraphic possibility of a gap in occupation between EC and MC at each of the major excavated sites.

The picture is undoubtedly complex. Additional evidence is reviewed below in connection with Settlements and Imported Pottery.

MIDDLE CYCLADIC SETTLEMENTS

(Fig. 3; Scholes 1956, 10; Renfrew 1972, 524; Barber 1974, 50)

MC settlements were relatively few, a single one on some of the smaller islands, two or three on the larger. Some of them were certainly fortified. Kea D had a wall and perhaps Phylakopi where, although part of the fortifications have recently been shown to belong to LH IIIB (AR 1975-76 25), it remains possible that part may be earlier, as previously thought. Akroterion Ourion (Vryokastro) on Tenos has wall remains that may be of this period. Other sites, e.g. Rizokastelia on Naxos, Kastri on Amorgos, are acropolis or promontory sites of which the former at least must surely have been chosen for security rather than ease of access.

The common belief that a sharp and sudden contraction in the number of the settlements marks the division between the EC and MC periods in not strictly accurate.

To be sure, the EC site distribution as at present known (Renfrew 1972, 510, Fig.) is somewhat suspect. There are startling gaps on the map - Andros, for instance. Most of the sites are cemeteries. A high proportion are of uncertain date.

But, if the count of dated sites (Fig. 1 & 2) reflects the situation at all accurately, it points to a decline between EC II (Keros - Syros) and EC III (Phylakopi 1 / Lefkandi I) rather than at any other time.

The figures are EC II - 33 sites; EC III - 15; EMC - 20. It seems certain that the decline in settlement numbers and the flight to protected sites took place at or just before the time when intrusive elements (the Lefkandi I culture) appear in the Cyclades. The MC period itself is not one of further retreat but of slow revival and consolidation, with its open local cultural developments and initiatives which are visible most clearly in the pottery.

MIDDLE CYCLADIC POTTERY

The identification of true Cycladic pottery and the establishment of local series have been important aims of recent excavations. Progress is inevitably slow but the potential rewards are great, particularly in the analysis of local trends within the islands themselves.

Remarks here are based on reanalysis of material from earlier excavations at Phylakopi and on preliminary study of finds from recent work, as well as on preliminary reports of the Kea excavations. In the latter two cases, it will be appreciated that further study will inevitably cause modification of ideas and the outline presented here should be regarded as provisional. The bias towards Melian material may also have produced a somewhat misleading picture.

MC Burnished (pl. 1)

Slipped and burnished pottery with a high gloss finish, though on a fairly coarse fabric, is one of the hallmarks of the MC period at both Ayia Irini and Phylakopi (Caskey 1972, 378; Barber 1974, 28). It is particularly common in bowl shapes and shapes which imitate those of Grey Minyan.

At Phylakopi, burnished pottery is found in the First City but it is quite distinct from MC burnished. In general, the surfaces of the earlier class are more crudely finished, tool marks being often visible, and the characteristic high gloss is absent. The Minyan-related shapes are not found.

At Kea, MC burnished might be thought to have antecedents in the highly burnished pottery of Phase C but the shapes of this ware are totally different, as apparently is the fabric.

This class of pottery probably persists throughout the Second City at Phylakopi, though it may have been to some extent replaced by a less common Red Washed class with a matt finish, which also occurs in the Third City, when the former type has died out.

At Kea MC burnished appears most characteristic of the earlier phase of the MBA (Caskey 1972, loc. cit.) and it is not mentioned as a major fabric in the later.

Sometimes there is minor decoration in white, now often visible only in shadow. Occasionally, at Phylakopi, there are elaborate designs involving birds and floral motifs which seem likely to belong to an advanced stage of the Second City.

Kean and Melian burnished wares, though basically similar, can be distinguished from each other by both the fabric and surface treatment.

Cycladic White (pl. 2)

This class also characteristic of the Second City at Phylakopi, was originally called 'Early Mycenaean with Designs in Matt Black'. It has a distinctive whiteish fabric, often with a green or yellow tinge. The quality of the fabric varies considerably being, at its best, moderately fine and of a consistent texture and colour. The fabric of the larger vases is always less fine. Decorated vases are often self slipped and have a smooth and attractive surface. The decoration is in matt paint varying in shade from brown to black. The best period seems to be that of earlier Phylakopi II when the typical beaked jugs are found. Panelled cups, bowls including 'Cycladic' bowls with inturned rims and loop handles, and spouted jars are among common shapes (Atkinson et al. 1904, Pls. 14 - 21).

Changes in fabric are probably related to changes in shape and decoration and seem to have chronological significance. The later group have biscuit of a less uniform colour: it is often grey or orange-brownish with white flecks, the latter type merging into the characteristic local fabric of the Third City. In the supposedly later stages, thin-walled cups, often with everted rims and of a 'clinky' fabric, are particularly common. Simple spiral decoration is often found on them. In general, the later decoration is a great deal more careless, especially on the larger vessels.

This is not a prominent category on Kea and one or two sherds which look to be imports of this kind stand out among the material illustrated (Caskey 1972, 381 (Cycladic Matt-painted ware), D 62, 63, pl. 84).

A few pieces have been elsewhere, at Lerna, Knossos and on Naxos (Barber 1974, 31, n. 81).

Black and Red Style (Atkinson et al. 1904, pls. 20, 21)

This decorative style is found on both varieties of Cycladic White, as well as on its Third City successor. It is formed from a combination of decorative elements of the Burnished and Cycladic White classes. The main motifs are based on red burnished elements, usually though not always circles. The bird is the most common motif. The rest of the decoration is in matt brown to black paint.

Cycladic White sherds decorated in this style have been found recently in a floor deposit with burnished pottery of the earlier stage of the Second City. There is no evidence that the style antedates this phase and it seems more common in later contexts, both on Melos and abroad. The basic decorative system continues in the Third City, though the dependence on burnished elements dies out.

The style appears local to Melos, though several examples are known from foreign contexts (Davis 1976).

Coarse and Plain Wares

A number of plain and coarse shapes have been described briefly and illustrated from Kea. New material from Phylakopi is being studied. Little is available form the ealier excavations at the site.

On Melos, there is considerable variety of fabric but it is difficult to establish hard and fast divisions.

A good deal of the pottery had a very thin white wash applied and occasionally very elementary decoration in white, though the vases must still be classified as coarse. There seems to have been a particular fondness for jars with everted rims and vessels of bridge-spouted jar shape, though few actual spouts have been found.

There is a wise, if fairly basic range of jugs, cups and other domestic vessels like tripod vases, pans and lamps. The conical cup, in quantity, seems to be a distinguisheing feature of the early Third City.

A small group of strap-handled cups is in a very thin brittle grey-black fabric with splashed decoration. There is an appreciable quantity of a heavier but otherwise rather similar variety which often has simple white banded decoration. The oval basin is a regular shape.

Other fabrics

The first report of the Phylakopi excavations remarked a variety of 'Geometric' pottery with decoration in matt paint as being a feature of the earlier phase of the Second City. This contrasted with the Geometric pottery of late Phylakopi I which had designs in lustrous paint. There were also some slight but mostly unspecified differences in motifs between the two classes (Atkinson et al. 1904, 105).

Although it has for some time been recognized that the lustrous-matt distinction is impossible to apply, the identity of the first group is never in doubt since the fabric is clearly recognizable. It seems characteristic of late Phylakopi I and may well be entirely residual in Phylakopi II.

The second group is as hard to recognize as the former is easy. It is possible to define a smallish group with a soft coarseish grainy fabric, a softish white slip and very simple designs in totally matt paint but this could easily be contemporary with the bulk of the MC material and seems to have no specially significant attributes.

Imported pottery

Kea D and F are marked by Minoan imports of MM II - III (Caskey 1972, 375, 386). Phylakopi II produced pottery of MM I - III (and perhaps later), though apparently very little of MM I (Renfrew 1972, 198 (table)). Polychrome ware (i.e. MM II) was absent from the final phase (Atkinson et al. 1904, 261).

Classic Grey Minyan (MH II) is common at both sites, at Kea in D rather than F (Annual of the British School at Athens 17 (1910-11), 16; Caskey 1972, 376, 382, 387). Mainland Matt-painted is not common at Kea (Caskey op. sit., 376, 387) and the same is true of Phylakopi.

IMPLICATIONS OF THE IMPORTED POTTERY

The precise chronological relations of Phylakopi II in terms of imports have not been greatly illuminated by recent excavations. The debris of the destroyed city forms a thick layer over much, if not all of the site and the pottery in it appears to span the whole of the life of the Second City. MM I pottery is not common. MM II Kamares ware is present as are sherds of MM III, including early dark-on-light. Some LM Ia imports and imitations may belong to the end of the Second City.

The main problem of MC chronology upon which the imported pottery bears has already been raised. It concerns the nature and relative position of the earliest phase of the MBA in the Cyclades. While a detailed survey is not appropriate here, it may be useful to summarize the evidence, which inclines in different directions.

Point 1 - 3 apply to both major excavated sites.

1. There is a clear stratigraphical break between EC and MC.
2. There is a lack or scarcity of EB III - MB I imports.
3. The earliest MB phases are related to MH / MM II.
4. In the case of Kea, this leaves a considerable gap between the early - EH III - related Lefkandi I culture of Phase C and the MM IIa imports in Phase D.
5. The character of late Phylakopi I (before the break) is crucial. It has features distinctly related to MH Greece and early MC Kea (see above). But the stratum is sealed below the main MC levels (above, 4).
6. The first phase of the tripartite division of the Second (MC) City at Phylakopi as originally defined, has no sound stratigraphical basis.
7. The continuity in pottery styles on which it was based can also be doubted (above).

The histories of the two sites may have been quite different. For Phylakopi, one can argue a substantial degree of contemporaneity between the late First City and the MH period, in which case overlap between Lefkandi I and Phylakopi I would become improbable.

There may have been a gap in occupation at Phylakopi within the early MH period, long enough to disturb but not totally destroy cultural continuity; short enough to allow for MH connections both before and after.

If there was a break at Kea, it may have been considerable longer.

The resolution if this difficult problem will not come without further excavation.

OTHER EVIDENCE

It is not easy to fill out a picture of the MB Cyclades since other classes of finds are virtually unpublished.

The architecture of the settlements suggests developments in scope from the less ambitious EC period. Local materials are always used, schist slabs at Kea, rounded boulders from the seashore at Melos, with drystone walling and mortar bonding respectively. Ashlar masonry was not used. Considerable regularity is apparent in the planning of houses and lanes. There is no sign of central or palace buildings but no site is completely accessible because of later overlay. The fortifications described above represent a considerable degree of planning and organization of resources.

MC burials are ill known. Rock cut tombs are thought to have been used at Phylakopi (Atkinson et al. 1904, 234) but there is no absolutely clear association of MC material with this form of tomb. The use of cists certainly continued from EC times in several places (e.g. Naxos, Papathanasopoulos 1963, 130) but it is not clear whether this practice was widespread. If it was, it is surprising that more have not been found. The only sign of larger built tombs is on Kea (Caskey 1972, 383) where pit graves as well as cists and urn burials have been found. The latter occur also at Phylakopi, though in contexts of late Phylakopi I.

Kea is the only site to provide evidence of grave goods apart from pottery. Stone vases, beads of stone and occasionally gold, seals, whorls and other minor objects were found. The report suggests that the majority of burials were accompanied only by pottery and that one or two relatively rich burials in built tombs accounted for most of the other small finds.

MC THERA

On this occasion it is appropriate to add a few words specifically about the MBA occupation of Akrotiri, although this is not yet well known and most of the material so far excavated relates to the Third City of Phylakopi.

Some of the published finds point to occupation as far back as EC III (Marinatos 1970, figs. 4 - 6). Local burnished pottery as well as MH and MM imports confirm MC habitation (Marinatos 1972, 44 and references; 1974, 31 & Pl. 67b). Marinatos (1972, 44) postulated an earthquake destruction of the MC town.

Some fragments of bird vases may have been imported from Melos (Marinatos 1969, 44, fig. 30) and at least one Theran import was found in the earlier excavations at Phylakopi (Atkinson et al. 1904, 120. fig. 92), though the context is unknown. None of these need be earlier than LBA.

The relationship of the so-called 'nippled ewers' from Thera to the beaked jugs from Phylakopi is not clear. There are some similarities in shape, in plastic additions and in some decorative elements but the Phylakopi cases seem, in general, to be distinctly earlier.

Signs of MC culture are therefore slight but significant.

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

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