The Pottery-Producing System at Akrotiri: An index of Exchange and Social Activity

Such an understanding requires the study and correlation of interacting archaeological and technological data.

In this paper an attempt will be made to trace the mechanisms of exchange at Akrotiri and the possible effects of the regional exchange system on local production. It seemed appropriate to proceed by studying some of the socio-economic factors which conditioned the form and fashioned the appearance of the vases, and by examining the functional elements of pottery which might have been dictated by the existing economic relationships.

This paper will try to combine these two approaches not by seeing the pots and vessels so much as items of material culture but by considering them as the material product of people's activities.

INTRODUCTION

Pottery represents, clearely, a set of technical, social and economic relationships within human groups and between societies and their environment.

Although much has been learned in recent years about the technical aspects of pottery manufacture at Akrotiri, considerably less is known about the socio-economic context and organization of production, which was indisputably one of the basic industries of the Late Cycladic I city of Akrotiri.

The bulk of pottery found at Akrotiri is locally made and clearly fulfilled a vast range of the settelment's requirements. More than fifty different types of pots can be distinguished (Doumas 1983, 110). The pottery found includes a wide variety of functional types like storage jars, smaller containers (such as amphorae and stirrup jars), pouring vessels, cooking pots, drinking vessels and so on, which all relate to specific activities and which would have been made and distributed with those activities in mind. Given the large number of shapes produced and the relatively high degree of standardization, it has generally been assumed that most, if not all, of Akrotiri pottery was produced by specialized craftsmen in a non-domestic context. Unfortunately neiter potters' workshops nor kilns have been found within the excavated area. The reason may be that the ceramic workshops were located on the periphery of the site which has not yet been excavated. In any event, the ubiquity of the pottery and the consistent repetition of the same types in different sizes suggests production on an industrial scale. The Akrotirian potters seem to have responded to pressures beyond their households, namely to the increasing complexity of regional distribution and exchange systems. We can imagine them as full-time craftsmen working permanently in a high production-rate craft such as pottery manufacture and supporting themselves entirely from the proceeds of their craft (Branigan 1983, 24). In view of the above, one can begin to speak in terms of mass-produced pottery and the existence of organized workshops of craftsmen during Late Cycladic I (1550 - 1500 BC) (Doumas 1983, 112).

But how pottery production was organized at Akrotiri remains an open question. Historically, there is no real documentary evidence dealing with the process of pottery production. Our entire knowledge comes from the ceramic material itself. We have no records concerning the organization and the function of a producing system. To attempt a reconstruction of the whole system, we must study the essential components of a pottery-producing system, which have been categorized by M. Rice into two major sub-systems: technical and socio-economic (Rice 1984, 239 - 242). These two sub-systems overlap and are interlinked. This means that all the factors which make up a pot-making mechanism influence each other and depend in turn on the broader cultural context.

In this paper we intend to investigate only some of the functional elements of pottery which might have been dictated by economic relationships and especially by trading requirements. The study is part of an ongoing project which includes the collection and examination of all available behavioural data in order to reconstruct the Akrotirian Bronze Age pottery-producing system in its socio-economic context.

THE CONTEXT OF POTTERY MAKING

An effective redistributive system was operating within the organized community of Akrotiri, based perhaps on some sort of central authority. It is also possible that this system worked on a more flexible inter-personal basis. Davis and Lewis infer that Cycladic trading exchange was somehow regulated by local administrators, whether merchants or producers (Davis and Lewis 1985, 89). It is indisputable that the main source of wealth accumulation and economic affluence was the maritime and trading activities of the islanders, which were collective in nature. Theran merchants and seamen were themselves the creators, administrators and middlemen of the surplus which they transported to other markets and exchanged for other goods.

It is quite difficult to specify the economic and social structure of LC I Akrotiri, mainly because of the absence of written records. We could, however, make an effort to trace some of the economic relationships within the settlement, which could have been dictated to some extent by some functional elements of the pottery vessels reflecting their distribution in Thera during the Bronze Age.

The exchange system and the determination of the quantities exchanged are two key elements underlying the internal and external economic relations of a society organized for redistribution.

Exchange is obviously a commercial activity and, hence, part of production; processes of production are inextricably tied to exchange and transportation. Pottery production is also clearly linked to the exchange mechanism and the needs for transportation of purchased goods (Knapp 1985, 11).

THE ROLE OF POTTERY IN EXCHANGES

The prerequisite for any redistributive and exchange system handling commodities is a means of determining quantity. Consequently the invention of units of quantity and of a numerical system to count them was of capital importance for an exchange-geared society such as that of Akrotiri. Units of measure for subsistence and other commodities in the Middle Bronze Age are implied by the hieroglyphic inscriptions and by the Linear A and Linear B tablets, where almost every line ends with an ideogram designating the commodity and a number indicating quantity (Evans 1921 - 35, 282; Ventris and Chadwick 1973, 31; Renfrew 1972, 410). In spite of the absence of any written records at Akrotiri, the archaeological evidence reveals that concepts of measurements, both of weight and number, had been formulated. Standard measures may already have been in operation by the end of the Middle Bronze Age, such as those documented by a graduated series of weights - made in disc form - found at the site (Petruso 1978, 1, 547 ff.; Michailidou 1990). But weight is not the only parameter for the measurement of quantity. Capacity is also a means of determining the quantity of commodities. The existence of units of capacity in Late Bronze Age times is documented by the linear tablets archives (Ventris and Chadwick 1973, 55; Palmer 1963, 12 - 14) and inscriptions in the Minoan Linear A script, engraved on some pithoi from Epano Zakro and Knossos (Best 1972, 82, Brice 1961, II, 6, ii, iii, Pl. XXIIIa/XXIII). The notation of units of a liquid measure for VINUM on these pithoi clearly implies the great significance of pottery vessels in determining the quantity of their contents.

It must be recognized that pottery vessels, like any other ustensils, were made to be used, and that function plays a very important role in determining a vessel's characteristics. As Riley has pointed out, 'a pot's function directly influences all stages of its cycle from production to distribution through to use and discard.' (Riley 1984, 60). The intended function affects the choice of clay, the forming technique, and the shape and the size of the pots. The size is indicative of the quantity every vessel can contain. Large storage jars (pithoi) would be needed to store commodities whereas, for ease of transport, smaller containers would be used to move and apportion produce. In fact, the length of a man's arm limits the pot size to a capacity of about twenty litres; that is also the maximum a man can comfortably carry (Balfet 1984, 185). In certain circumstances, different sizes of pots were identified with a specific quantity of content which, in turn, could be recognized from the distinctive size and shape of the pots - the smallest unit of volume is indicated in the tablets by ... , clearly by the measure of a 'cup' (Ventris and Chadwick 1973, 58).

The various sizes of vase should thus represent recognizable quantities of commodities, which is a fundamental element in the function of exchange. The Akrotirian merchants handling a commodity - say wine - in amphorae should have been able to calculate easily the amount of wine transported from the number of containers they carried in their ships, since the capacity of each amphora when filled was known to be 14 - 18 litres (Fig. 1). (We could draw a parallel here with the current practice in Greece of selling oil in 17 kg tins).

We may therefore assume that the shape, capacity and sometimes decoration of the vases are indicative of the commodity and the quantity every vessel contains. Since individual transactions would normally cover different quantities of a given commodity, various 'standardized' types of vessels were needed to meet traders' requirements. It is only natural that potters should comply with those requirements.

FRACTIONAL QUANTITIES IN AKROTIRIAN POTTERY

It has already been pointed out that the consistency in 'repetition throwing' in Theran pottery was one of the factors which led to the standardization of types and sizes. The same type of vase occurs in different sizes, presenting a convincing picture of a gradation in volumes. This observation prompted us to examine whether the visual proportions are confirmed by measurements of the vessels' capacity. Furthermore, it would be very interesting to investigate whether there was some kind of standard ratio between the standardized quantitites which might have been dictated by (or merely reflect) a measuring system developed locally or introduced from elsewhere.

In a previous paper, due to the published shortly, a first attempt was made to trace the system on which the successive sizes of the same type of vase were scaled. Important indications as to how this system was organized come from measurements of the volume of three types of pottery vessels: bridge-spouted jars, ovoid funnel-mouthed pithoi and open-mouthed jugs.

The choice of these particular types was dictated, firstly because they are considered to be carrying-vessels, and thus likely to give us ratios of volume, since they were the most suitable means of exchange; and secondly, because each type consists of many different sized vases, something not commonly observed in other types of pottery.

The volumetric analysis has revealed a tendency for volumes of different vases to be concentrated around specific numbers, which represent volumes expressed in cubic centimeters (cm³). This concentration has determined the classification of vases in different volumetric groups. Although the limits are in some extreme cases not evident, we have delineated the groups so as to achieve the smallest standard deviation of volumes within each group. Trying to divide each category of vessels into groups, we also took into consideration the visual impression, which in some cases in clearer than the result of the measurement.

FIRST TYPE: BRIDGE-SPOUTED JARS

Bridge-spouted jars comprise 112 pots of five different sizes; the smallest vase has a capacity of 460 cm³ and the biggest one of 15,000 cm³ (Fig. 2). In this category, we can distinguish five groups according to their capacity as presented graphically in Fig. 3 (see also Fig. 2).

The first group (A) consists of forty-two vases of an average volume of about 625 cm³ (460 - 850 cm³). In the second group (B), we have thirty-four vases; the average volume is 2305 cm³ (1500 - 3200 cm³). The third group (C) consists of sixteen vases of an average volume of 4306 cm³. The fourth group (D) is the smallest, with just six pots; the average volume of this group is 6993 cm³ (6300 - 7500 cm³). The fifth group (E) comprises the largest vases in this caterory. It consists of eleven pots with an average volume of 12,836 cm³ (11,400 - 15,000 cm³).

Assuming that the average volume of group E vases represents a unitary quantity, then the successive 'sub-units' show the following fractional values of the major denomination: 1 : 1/2 : 1/3 : 1/6 : 1/20.

Group D is half the size of the major unit E; group C is two-thirds as big as group D; group B is one-third of group C and finally the smallest size, group A, is three or four times smaller than B (Table I, see also Fig. 4).

The relationships can also be expressed as 1 X 3.33 X 2 X 1.5 X 2, this series representing the ratios of adjacent sub-units in ascending order. The ratios appear to be regular with the exception of the figure of 3.33 which indicates the ratio between the groups A and B and is somewhat questionable. This is probably due to the divergence of the volumes within group B. New excavated elements might throw some light on this relationship.

SECOND TYPE: OPEN-MOUTH JUGS

The second category, the open-mouth jugs, comprises seventeen vases, which appear to belong to four different-sized groups (Figs. 5, 6). The smallest pot has a capacity of 580 cm³ and the biggest one of 1900 cm³.

The first group (A), which consists of the smaller pots, has an average volume of about 663 cm³ (580 - 740 cm³). The second group (B) has an average volume of 850 cm³ (820 - 960 cm³). The average volume of the third group (C) increases to 1190 cm³ (1150 - 1240 cm³). The fourth group (D) comprises the biggest pots in this category; the average volume calculated is 1,860 cm³ (1,850 - 1,900 cm³) (Table II).

The differences in volume between the four groups of the open-mouthed jugs are smaller then the differences observed between the groups of the bridge-spouted jars (first category). We can observe, however, that the smallest groups in both categories have almost the same capacity: the first 624 cm³ and the second 663 cm³. This brings in mind the method used in Linear B tablets, where capacity is measured in two series, one for dry measures and the other for liquid measures, and the two lowest units are common to both series.

Supposing now that the fourth group (D) is the major unit in this set, the ratios in descending order are: 1 : 2/3 : 1/2 : 1/3 (Fig. 6). That is to say, the ratios in ascending order are: 1 X 1.5 X 0.75 X 1.5

THIRD TYPE: OVOID FUNNEL-MOUTHED PITHOI

The third type, the ovoid funnel-mouthed pithos, comprises thirty-six vases, which are divided into six groups of different sizes and also have proportionate values (Fig. 8). The smallest pot in this category has a capacity of 1000 cm³ (1 litre) and the biggest one of 32,000 cm³ (32 litres) (Fig. 7).

The first group (A) consists of the smallest vases and the average volume is 1787.5 cm³ (1000 - 2350 cm³). The second group (B) has an average volume of 3408.9 cm³ (2800 - 4000 cm³). The average volume of the third group (C) rises to 6640 cm³ (6000 - 7700 cm³). The fourth group (D) has an average volume of 9500 cm³ (8000 - 11,000 cm³). The fifth group (E) has an average volume of 18,328.5 cm³ (16,300 - 20,100 cm³)and the last one (F) comprises the biggest vases of this category, has an average volume of 28,428.5 cm³ (26,000 - 32,000 cm³), (Table III).

The ratios in descending order are: 1 : 2/3 : 1/3 : 1/4 : 1/8 : 1/16 (Fig. 9). The relationships can also be expressed as: 1 X 2 X 2 X 1.33 X 2 X 1.5, which represents the ratios of adjacent sub-units in ascending order.

METHODS OF MEASUREMENT

The results of the above measurements confirm the proportional graduation of the volumes in these three types of vases and may indicate that such a regular gradation is perhaps due to a method of capacity measuring. In fact the three series of graduated volumes appear to represent measures of capacity. The groups which can be distinguished in each type of vase, represent fractional quantities of the major denomination. These fractions result from continuous division by 2 and/or continuous division by 3. Thus we have two sets of fractions:

1; 1/2 ; 1/4 ; 1/8 ; 1/16 ; ?1/32 and 1; 1/3 ; 1/6 ; ?1/12.

The relationships within the three types of vase we have examined are expressed as fractions in both series.

A possible measuring system, which would satisfy a vast range of requirements, could be derived form combining these two methods. The method of continuous division by 2 is very simple in conception and very easy to use. On the other hand, the insertion of units resulting from continuous division by 3, while more complicated, gives factional sums with smaller intervals and facilitates measurements in small-sized vases.

The conception of this method of measurement is definitely related to the system of recording fractional quantities in Linear A script, where odd amounts are expressed as fractions of the primary unit, e.g. 1 + 1/2 + 1/8 + 1/16 = 1  11/16 (Ventris and Chadwick 1973, 54). As Bennett has pointed out, the Linear A fractions imply that odd amounts were reckoned by pouring the residue once only into a series of smaller vessels scaled successively at 1/2, 1/4, 1/8, etc. of the primary unit (Bennett 1950, 221).

This enables us to assume that some types of vessels, which show serial gradation of size - such as those mentioned above - were manufactured by Akrotirian potters to be used as sets of graduated measures of capacity. It is also possible that each fractional measure required was used only once, and the combination of different measures somehow recorded.

As we can see in Fig. 10, the fractional measures used at Akrotiri correspond to the tentative values of Linear A fractional signs. Such a similarity in conception must also be linked to the methods of calculation used, which were probably taken over from Linear A as well. We may suggest a direct link between the two systems of measurements. Akrotirians of the Late Cycladic period probably adopted the fundamental principles of accounting used in the Linear A script.

Taking into consideration the history of mutual relations between Crete and the Cyclades, the adoption of a common system of weights and measures must have had the effect of making communications between regions easier and better (Petruso 1978, 549; Parise 1986, 303).

The remaining problem is to determine the absolute values of the Linear A fractional signs, in order to compare them with the actual fractional quantities in Akrotiri. Furthermore, it would also be desirable to find the primary unit to which these quantitites are reduced.

Although such determination cannot be easily achieved, some more evidence can be provided by measuring the volume of large vessels found at Akrotiri and comparing the results of these new measurements with the available material from Minoan Crete and other Aegean sites. It is likely that the capacities of storage jars bear some relation to the standard units of measure in Minoan book-keeping. We must not expect such consistency in wheel-made vessels, but there is a good chance that, if enough specimens can be measured, a rough average will emerge.

It should be noted from an inscription in Linear A Script engraved on a pithos excavated at Epano Zakro, that the pithos was meant to contain twenty-two units of wine (Best 1972, 82). Jan Stronk estimated its contents at 556 litres ± 11%, which means that the unit to be reckoned with contains 25.2 litres ± 11% (25.2 ± 2.8 litres) (1972, 85). Ventris and Chadwick estimated the Mycenaean dry unit at 13 - 14 litres, which is approximately the capacity of the most frequent size of Mycenaean stirrup-jar (1973, 60). More recent investigations by Miss Mabel Lang have shown that this figure should be reduced by at least 20% (Lang 1964, 99 - 105). Similar assumptions might perhaps be made for liquid measures. In that case the main liquid unit, which was estimated at 36 litres, would decrease to 28.8 litres, or even less.

This assumption, if correct, would be of great significance for the results of capacity measurement at Akrotiri, since the major denomination of the ovoid funnel-mouthed pithoi has been calculated at 28.4 litres and the major 'sub-division' in the series of the bridge-spouted jars has a capacity of approximately 12 litres. One can imagine these types of vessels as two series of measures of capacity, one dry and one liquid. Further evidence, however, will be needed for more definite conclusions in this respect.

CONCLUSION

In trying to reconstruct an ancient system of absolute capacity values by measuring the volume of excavated pottery, one must allow a rather generous range of tolerances. Although the classification of various sizes was probably based on calculations made in units which were measures of content - i.e. the quantity which filled a bridge-spouted jar of 12 litres - it would be quite difficult for potters to achieve the exact size required every time without the aid of some mechanical means to control symmetry and wall thickness. It seems possible that the Bronze Age potters has specific size of vessel in mind and tried to reproduce it using a specific type and amount of clay. Hence, we should expect variations in dimensions, shapes and size, because of variations in the precision of execution, for some potters would be more skilful than others, and standards of manufacture obviously varied from craftsmen to craftsmen. Many variations in the repetition of types and size may also occur because of unforseen circumstances during the throwing process. For instance, instead of destroying the entire pot if the clay of the rim contained a piece of grit which prevented him from simply finishing or folding the rim as he normally would, a potter might merely produce a smaller pot by cutting off the rim (van As 1984, 136). Therefore, when we attempt to reconstruct a series of capacity measures, we should bear in mind that pots intended to have the same capacity would vary somewhat. Even if there is no noticeable external difference between the pots meant to contain the same quantity of a commodity, it can easily be shown that differences in their capacity can actually reach 1 - 2 litres. In our case the deviation from the required size appears to be as much as 10 - 20 %.

It has been my aim throughout this study to investigate some of the components which constitute the socio-economic subsystem of pottery making at Akrotiri and to produce a framework which will serve as a basis for a more thorough analysis in the future.

From this research, it may be concluded that the demand for the determination of standard quantitites of commodities exchanged led the craftsmen and merchants of Akrotiri to adopt the above-mentioned principles of measuring, which affected the mechanization of the pottery-producting system. These principles have been shown to have had an important and distinct impact on the creation of shapes and sizes by influencing the development of throwing techniques. Furthermore, the gradation system which underlies pot size is closely related to the exchange mechanism, which in turn dictates, up to a point, the capacity of the vessels.

The establishment of regular routes within the Aegean (Davis 1979, 143) led to increased movement of goods; consequently a regular exchange of local, luxury and surplus goods, including metals, would have become feasible as a result of the advances in transport technology. The increased demand for standardized exchanges, inextricably linked to commercial transactions, might have been one of the main factors which led to the standardization of pottery production, both determining and determined by the exchange system. Thus, the whole intercommunicating network of ceramic production and exchange would have depended on the specific regional economic conditions and would reflect the socio-economic structure of the prehistoric city of Akrotiri.

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

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