Some Changes in the BA Pottery Production at Akrotiri and their Possible Implications

Three aspects of the pottery of the MC to LC transition period are examined: technology, morphology as transformed by specialized function, and decoration. During this period an 'advanced' technology (the wheel) was introduced at Akrotiri, there was an explosion in the range of ceramic shapes, and the novelties in decoration adopted from Crete were treated differently in order to conform to the traditional Cycladic tectonic style. These transformations in the pottery, along with evidence from other aspects of the material culture at Akrotiri (architecture, wall-paintings, stone industry, etc.) may be indicative of the process of adaptation of its inhabitants to the new needs of the LC period.

The interpretation of variation in the material evidence of a culture involves two levels of scientific analysis: that concerned with matters of ideology and theories of social change, and that which deals with stylistic or technological changes in the artefacts.

It is a fundamental goal of the archaeologist to point out 'inequalities' in the material culture, as these have impartially been described, in order to understand the character and peculiarities of the community involved. Both the approach to the data and the manner in which conclusions are reached are a function of the researcher's methodology.

In this paper the descriptive taxonomical approach already existing for the data will be retained and an evolutionary backbone assumed as an overall framework for the social transformations. Thus, rather than presenting a clear description of a particular case of transformations, it poses a series of questions arising from the study of the pottery at the site.

The MC strata at Akrotiri generally consists of a levelling fill underlying the LC town, both houses and streets. For this reason the material is not in situ (the only exception to date are the pits beneath the floors of the West House) and can rarely yield other than typological information. There is no way in which phases may be distinguished, such as those dating the rebuilding activities in the Second City at Phylakopi, for instance.

The pottery consists of several thousand sherds and only about 15 complete vases (Papagiannopoulou 1987).

This situation imposes certain a priori limitations on its study. For example, there is little possibility of examining details of manufacture, gathering quantifiable data, or for close correlation of the different sherds to pottery shapes. It is possible, however, to define the chronological limits of the MC settlement at Akrotiri, its extent in relation to the LC town and the character of its pottery.

Concerning the dating of the MC settlement, the pottery belongs to the mature (middle) and late phases. There is an apparent scarcity of early MC material, such as fan-spouted jugs, cups with convex-concave profile, the barrel jar (Atkinson et al. 1904, Pl. XI). This phase (early MC) is, however, known on Thera from Ftellos (Marthari 1982, 86 - 100, Fig. 4).

For the moment this lack of evidence cannot be interpreted as a gap in the Akrotiri sequence, between the EC and MC periods, since pottery found in the so-called 'Fire deposit' at the site includes some forms of the EC IIIB phase which also continue in the MC phase (S. Marinatos 1970, Fig. 2 and 4).

It is a fact, however, that the rich MC material from Akrotiri includes types characteristic of the middle to late phases of the period elsewhere in the Cyclades.

Regarding the extent of the MC settlement, this seems to have been about the same as that of the LC town. As yet there is insufficient evidence as to the construction of the buildings and their relative density. The possibility of uncovering extensive stretches of MC walls seems rather small, since the bedrock slopes upwards to the north and all traces at the base of this incline have been obliterated by the former winter torrent. MC layers may have survived to the east of the bed of this ravine, where the ground slopes abruptly.

Finally, as to the actual pottery, macroscopic analysis has shown that the clays used were the same as for LC wares. This has also been verified by chemical analysis (Kilikoglou-Papagiannopoulou, in progress).

The three major categories of surface treatment occurring on comparable wares from other MC settlements, Ayia Irini and Phylakopi, are represented at Akrotiri: Burnished ware, Cycladic White ware and Bichrome ware (black and red). Burnished ware is black or red in colour, depending on firing conditions, the clay fabric being the same (Whitbread-Papagiannopoulou, unpublished). It should be added that most of the pottery belongs to the plain wares, of which some whole examples have been preserved.

Some categories of decorated pottery are also represented, though less frequently: Slipped ware, Light on Dark ware and Polychrome (imitation Kamares) ware.

Table I shows the MC pottery shapes found to date at Akrotiri and the wares in which they occur. Those underlined are introductions from the Minoan repertoire. Those shown by a fine black line are MC shapes for which there are whole examples in Thera. Sherds have been drawn in those cases where the shape is ambiguous. Where reconstruction of the shape is possible, but no whole sample exists, LC parallels are shown with broad black outline.

This table will certainly be augmented as examination of the sherd material progresses. Nevertheless, some points have already emerged:

1. There is quite a rich variety of shapes in the MC period, particularly in the later phase when the local repertoire was enriched through introduction of Minoan shapes (those underlined in Table I).
2. Until the very end of the period local shapes outnumber imported ones and both categories are commonly manufactured in local wares (Cycladic White, Burnished, Bichrome).
3. For each shape there is a relatively strong correlation in one or two wares. For example, panelled cups never occur in Burnished ware, Cycladic cups only rarely occur in Bichrome ware and a type of breasted jug (Table I, 15) only occurs in Cycladic White ware. At present it is impossible to estimate the relative frequency of each type in each ware, since examination of the sherd material has not been completed. However, it might prove interesting to investigate whether there is any patterning in the frequencies of the different wares, in order to single out areas and modes of production.
4. Though there are few whole pots from the MC period, from those existing there does not seem to be any standardization of size.
5. Finally, unlike Phylakopi Second City and Ayia Irini periods IV and V, there are no shapes imitating Minyan wares.

Concerning motif variation and decoration syntax, MC potters observed certain rules which remained largely unchanged up to the end of the MC period, when foreign ideas were slowly absorbed into the local style. No whole vases from the mature MC phase, when these changes might be more clearly followed, have been recovered at Akrotiri. However some whole vases have been found in the Karageorgis quarries on Thera (Marthari, personal communication) (Table I, 3, 13, 21) and there is sherd material from the site.

Table II shows some of the most frequent motifs occurring on the two main decorated wares, Cycladic White and Bichrome (CW, BW). These are: in BW: goat, loop, foliate band, grape, crocus, disk and dots, pomegranate and tree. In CW ware: linked disks, hatched lozenges, running spirals, closed quirk, wavy lines, foot motif, open quirk, grape, hook, hatched-leaved rosettes, filled semicircles, eyed spirals, enclosed dots, pomegranate, vetch, bird and chevrons.

From Table II it becomes clear that floral and animal representations appear more frequently on Bichrome ware than Cycladic White ware, which is usually decorated with geometric designs. It is therefore evident that Bichrome pottery appeared at a later stage in the MC period than Cycladic White.

Some of the pictorial motifs, such as the pomegranate, bird, grape and possibly the goat, occur on both CW and BW during the MC period. Perhaps these had some particular significance in the motif repertoire of the Theran potters, since they continued in use during the LC period too.

Below, the pottery of both the MC and LC periods at Akrotiri will be considered as an entity and transformations observed in three particular spheres will be examined and discussed. These are: technology, function and decoration.

1.    TECHNOLOGY

From the few whole vessels recovered from the soundings beneath the floor of the West House, as well as from excavations elsewhere on the island (Karageorgis quarries), some details may be gathered concerning their manufacture.

It seems that, unlike the situation prevailing in Crete, most of the pottery in the Cyclades, in this case Thera, was still made by hand. Channel-spouted jugs, fan-spouted jugs, Cycladic bowls and even bird jugs (the Cycladic potter's finest creation) were all made without the aid of a wheel. This is deduced from the bases, which bear no sign of use of a cutting string during rotation, and the lack of rilling, from centrifugal force, on the inside of the walls.

Some shapes, such as the ribbed vessel, kymbe or lamp, which imitate stone prototypes, are more easily made by hand, since their shape is difficult to achieve on the wheel. Conversely, shapes which are symmetrical in form, such as Cycladic cups, panelled cups and jugs are the first to be made by the new technique.

Here the scarcity of whole pots of mature MC date from the settlement is regrettable and evidence from the Karageorgis quarries will be used to remedy this (Marthari, personal communication).

Among the pottery discovered there is an example of a hole-mouthed jar (Fig. 1). This shape is Minoan, but the pot is manufactured in local CW ware. It is also interesting that it was made on the wheel. Although to judge from one example requires caution, it may be that the wheel was introduced to the island from Crete. This vase represents one of the earliest experimentations on the wheel on Thera and it is interesting that the shape is foreign, while the fabric and surface treatment are local. Is this a case of a local potter manufacturing a new shape in a new technique learned abroad (in Crete), or of an itinerant Minoan potter manufacturing, on his wheel, a familiar shape in a foreign style? It is possible that relations with Crete promoted experimentation in pottery production, though internal island conditions may equally well have prompted people to seek new ways of producing their ceramic containers. Reference will be made to these internal conditions below, when the surface treatment of the pottery is discussed.

At this point it should be emphasized that by the end of the MC period the majority of vessels were wheel-made. Their walls are of uniform thickness and, on close inspection, traces of rilling can be seen on their interior.

Unfortunately, there is not a sufficiently satisfactory sample of whole pots in each shape to draw reliable conclusions on shape standardization and size variation, though the general impression is that standardization in production is well under way by the end of the MC period and size differentiation tends to disappear.

All this evidence is indicative of an 'advance' in pottery manufacture in the direction of mass production and specialization. It is perhaps too bold to add that this specialization heralds a change in the level of pottery production overall, yet there is another very crucial element in the pottery of this period of transition from MC to LC. This is the rarity of potter's marks on LC pottery, which are found occasionally on the few extant MC pots. (This is a provisional observation and must be taken as such, since such comparisons are premature at this stage of research. It arises from the general impression that, given the quantity of intact LC vases, the number of potter's marks preserved would be considerably greater than those known from whole MC pots, which are only 15.)

The signing of a pot, whether to denote its provenance or destination, its volume or the potter's signature, endows it with an additional importance, perhaps signifying that it was considered a single artefact by the potter. This no longer applies when several pots are thrown per day on the wheel. Bikaki has demonstrated a parallel explosion of potter's marks on Kea, Period IV, though she attributes this to general demand rather than individual need.

It would be interesting, in this connection, to clarify whether the introduction of Minoan shapes was followed by the introduction of the wheel to the Cyclades and whether this new technology meant a change in the level of pottery production. If the Minoans did introduce the wheel to the islands, it must be proven that almost all Minoan shapes were wheel-made. If, however, it emerges that Cycladic shapes start independently to be manufactured by fast rotation, then the influx of Minoan shapes and the introduction of the wheel must be regarded as concurrent but separate phenomena.

Whatever the case, by the LC period the wheel was employed to the virtual exclusion of other techniques. The few pots which continued to be made by hand were always traditional Cycladic shapes, such as the breasted jug or jug with backwards tilted neck, that occurs only rarely. These, as well as some kymbe, miniature juglets and ribbed vases, which were still made without a wheel, must come from local workshops reluctant to abandon the traditional technique, as long as some old wares were still in demand.

Comparison of the evidence with that from other Cycladic islands would be most useful, since this would clarify whether there was a universal change in the islands in this period or whether it was more localized.

2.   FUNCTION - SHAPE DIFFERENTIATION

In the area of function, as this is expressed by the shape of a pot, one definite trend in the transition from MC to LC is the increase in shape specialization. The shape repertoire at Akrotiri during the LC period includes some one hundred different shapes, an impressive increase from the approximately thirty shapes known from the MC repertoire. Even allowing for some MC shapes which cannot be reconstructed, the multiplication of shapes is striking.

This development came about through minor or major changes in the rim profile of some basic local pot types, as well as the introduction of some Minoan shapes. As a consequence there was a diversification in vase use, certain shapes being appropriate for specific purposes.

The most obvious example of this development is the jug type. Throughout the MC period the fan-spouted jug (Table I, 12), the channel-spouted jug (Table I, 13) and the bird jug (Table I, 14) were the three most common jug types at Akrotiri. (The breasted jug (Table I, 15) and types known only from the Milian MC repertoire at the moment, may prove to have been popular too.) Even allowing for the addition of other types in the future, the MC repertoire was certainly far more limited than the fifteen types of jugs known for LC Akrotiri. To mention some of those occurring most frequently: the cut-away spouted jug, trefoil-mouthed, flat-rimmed, open-mouthed with or without collared neck, eyed, high-necked, globular with tubular spout, funnel-mouthed etc.

The same standardization is observed for the cups also. The hemispherical cup (Table I, 5) and the panelled cup (Table I, 4) were the two main types of drinking vessel in the MC period, the 'Cycladic' cup (Table I, 2, 3) being rather unsuitable for this purpose. By the end of the MC period another two types were added to those already existing, the Keftiu and semi-globular cups (Table I, 6, 7), borrowed from the Minoan repertoire.

This analysis may be extended to the multiplication of LC pithos shapes, jar shapes etc. It simply shows that there was an increased need for pots with specific functions around the end of the MC period and that by the LC period each shape had gradually been standardized in size and morphology. On the one hand this process resulted in each household having many more vessels and on the other, production was intensified to meet this demand.

Here again, lack of evidence of pottery workshops at the settlement only allows speculation. Unless their absence is taken to mean, as is most probable, that these quarters existed on the periphery of the settlement; another indication of 'village' production rather than 'household' workshops.

Some attention should be paid here to the ritual pots. There is clear conservatism in those forms of local MC pottery that carry some symbolic meaning. For example, the breasted jug (Table I, 15) changed very little, if at all, in the LC period. Of course the local repertoire of ritual vessels was enriched with new forms: the 'magical' pot, zoomorphic rhyta, peg-topped rhyta, etc. The resultant pottery repertoire was mainly a mixture of local and Minoan shapes. The survival of unaltered local shapes for ritual use suggests strong continuity of tradition. It seems that the indigenous population at Akrotiri consciously decided which adoptions were necessary and the areas in which these would occur.

Finally, it should be emphasized that Akrotiri is an excellent example of a site where material-technical equipment met the needs created by changes in the socio-political sphere. MC Akrotiri must have presented a quite different picture from its EC precursor. The town must have grown larger and contacts with the outside world more intensive. Akrotiri became a major settlement in the Cyclades, either due to agglomeration of EC villages; or increased trade with Crete; or the establishment of much safer seafaring conditions in the Aegean; or the forging of political alliances with other Cycladic islands, particularly Milos; or a combination of all these factors, and possibly others not archaeologically identifiable. All affected the pottery production there, no matter how conservative the pottery is. This conservatism will be discussed below in connection with pottery decoration. However, the morphology of the pottery changed quite dramatically by the LC period, verifying that Akrotiri was a strong community, ready to make the accommodations demanded by the new needs which developed at that time.

3.   DECORATION - SURFACE TREATMENT

In addition to technology and function, it is useful to explore another area of information gathered from the pottery at Akrotiri. This is the decoration, which is probably less susceptible to radical changes since it is less dependent on function.

However, it is not the intention to discuss here theoretical viewpoints on the reasons pottery styles change and the variability of such changes, but rather to highlight some elements of the decoration, in order to show how ingeniously the islanders at Akrotiri used 'Minoan novelties' introduced from Crete.

Regardless of differences in opinion as to the mechanism involved in the decoration of a pot, it is generally agreed that the style of a family of similarly decorated ceramic containers uses the same system of designs because this denotes the same symbolic meaning for the users of the pots. It is difficult for the archaeologist to decipher this meaning, since he lacks oral information from the potter. Therefore, by using criteria such as the elements of the decoration or their syntax and arrangement on the vessel, he tries to document any visible change in the pottery group being studied.

In trying to understand the changes that occurred at Akrotiri during the MC - LC transition period, it has become evident that more information than the simple enumeration of design elements which infiltrated the repertoire from Crete at this time is required. This enumeration cannot explain why these particular motifs were chosen, or why they were applied in a different way. The whole system of decoration which these motifs entered, what Furumark called the syntax, must be examined.

It has already been pointed out (Marthari 1987, 359 - 379) that at LC Akrotiri two languages or two style traditions were in use. One is a direct descendent of Cycladic pottery decoration, while the other is influenced by prototypes best expressed on Minoan pottery.

The main characteristic of Cycladic style, already established by the end of the EC period, is the vertical axis in the placement of motifs: the different elements of the decoration were placed around the circumference of the vase, leaving empty spaces between. Thus there was never continuity of design, but rather an imaginary vertical arrangement.

This syntax of decoration originated in the EC - MC vertical band and was continued by the vertically placed MC S-spiral, the lozenge, the rosette and even decoration in a panel. The most fundamental effect of this arrangement is that the elements exist in a sort of vacuum and there is no possibility of their being linked in a larger configuration. This isolation of design elements has another by-product, namely that their morphology is independent of the overall design and the initial geometric character of the elements thus remains unchanged.

In Minoan style, however, the potter uses either the horizontal axis of the pot, for the application of the 'zonal' decoration, or the entire body surface, for 'unity' decoration. In both cases the different elements of the design are linked together; in the first case horizontally, in the second in different directions, creating torsional, radiating or other arrangements.

It seems that the Theran potters only borrowed 'zonal' syntax from their neighbours.

Wherever 'unity' decoration is applied on Theran pottery this seems to follow the art of wall-painting, examples of which abounded in the settlement, rather than radiating, torsional or other arrangements of floral and geometric motifs used by the Minoans. The most common examples of pots with unity decoration from the site are the bird jugs, kymbe, flower pots and some pithoi, all decorated with pictorial motifs. There are examples where motifs intended to be used in unity decoration, such as the lily, are arranged horizontally on the vase, thus emphasizing its horizontal axis (S. Marinatos 1969, 125). In other cases the principle of the MC panel is kept, by the addition to the pictorial representation of a vertical foliate band, separating the surface into panels (S. Marinatos 1972, Pl. 60a). The crocus, vetch, reed and all other floral motifs used on Theran pottery are placed vertically or horizontally, accentuating the main axes of the vase surface and preferably leaving empty space in between, so that they exist in isolation and not in 'unity'.

Of course, even in LM IA Crete, 'unity' decoration covering the whole body of the pot was less popular than 'zonal' decorations, but there the unity designs were regularly used for the broad upper zone of the vases (Furumark 1941, 156).

The Theran potter easily understood the 'zonal' system of decoration because this was more akin to the one-directional system already in use before the influx of Minoan vases to the settlement in great quantities. He also adopted the difference in the width existing in the field division of the Minoan 'zonal' system, but instead of separating the pot surface into different zones he used the entire surface as one zone, leaving room only for multiple horizontal bands or dotted lines or blobs on the lower body. Division into multiple zones of more or less equal width is used mainly when the ripple motif is applied. This is particularly common on conical (conical rhyton) and ovoid or piriform forms (three-handled jar, small hole-mouthed jar).

Tabulated information of this kind can lead to very interesting results concerning the ways in which Theran potters adapted Minoan decorative innovations to their structure of decoration. This is attempted in Table III for three LC I Theran pottery shapes. Zonal decoration is used in most cases. Where unity decoration is applied, this is in a different way (separation in panels) from the Minoan principle. In some cases the traditional Cycladic vertical syntax is preferred (oval-mouthed amphora). The case of the bridge-spouted jug (large) is indicative of the uneasy situation with which Theran potters were sometimes faced. Here the body surface was used as one zone, to conform with the Minoan style on which there was a broad upper zone. However, the local potter, being unfamiliar with combining decorative elements to form designs and the use of these in configurations, devised another solution. He augmented the size of the running spiral, the zonal motif par excellence on Thera, and covered the remaining space thus formed with dots, blobs, semicircles, multiple tangents and additional white details to make their design less awkward.

The corpus of Theran pottery is open to this and to other quantifiable information, which will throw light on the transformation that took place during this period in the Cyclades. This information may be used more effectively when the evidence from the preceding MC period is better understood. MC pottery, with its technological particularities, morphological features and decorational principles is essential to the evaluation of developments in the LC period.

During the LC period Thera belonged to a much wider community and shared many cultural elements with its powerful allies, particularly Crete. Nevertheless, there existed a dynamic indigenous population at Akrotiri which struggled to adopt new technologies (the potter's wheel), exploit a newly acquired wealth (specialized shapes in the pottery repertoire) and conserve its ancestral tradition (structure of pottery decoration). It is hoped that through combined information on the material evidence produced by the settlement at Akrotiri we shall come closer to forming a picture of its life during the Bronze Age.

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

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