All Important yet Elusive: Looking for Evidence of Cloth-Making at Akrotiri

The emerging image is still vague, but already significant for the impact of weaving on vital aspects of social life and for its importance in detecting Minoan or other influences in the Aegean.

WHERE ARE THE LOOMS?

Weaving was an all-important activity in the Aegean settlements; this is an assumption which hardly needs demonstration: the wall-paintings and the Mycenaean archives provide evidence enough. It belongs to that category of knowledge which is considered acquired and consolidated, even if only by deduction from other archaeological fields. Over and over again Minoan ladies appear in the wall-paintings wearing the most complicated dresses, large multi-coloured skirts, beautifully ornamented bodices, transparent veils. Such elaborate attire, even when enhanced by the process of representation, even allowing in large degree for pictorial convention, constitutes nevertheless the measure of the desires of a society which recognizes itself there. Moreover such textiles were probably within the technical possibilities of both weavers and looms.

Still more important are the series of the palace archives. The carefully marked details of the quantities of wool allocated, as well as of the finished items due, of who works for whom, where, who is in charge and of how many people, point to an elaborate bureaucracy set up by the palace to control and exploit the products of the loom.

However, the material reality of these activities is hidden. They are scarcely visible archaeologically. Almost everything has gone: material, tools, products - or at least the bulk of them. Thus, we are in ignorance of concrete details, weaving techniques, the links with related crafts, and the intermediate stages. In other words, how and where the costumes we see in the Minoan wall-paintings were really made. If someone sets his mind on looking for tangible evidence, he has to set about it the other way around. 'Finding' something means first of all being able to recognize it as such. That means having a rough idea of what one is looking for. For example, the form of the loom, where you might find traces of it, the technical capacity of the whole settlement, and the degree of specialization in other fields. Finally, it means making the most of the evidence from categories of finds clearly related to weaving and spinning activities - even if such finds are minimal and with little significance for the whole process, such as loom weights and whorls. Thus, little by little, the gaps can be filled in and an endeavour made to construct a coherent whole, even if some measure of arbitrariness is inherent in such an attempt.

Akrotiri is the ideal experimental site, an almost unique opportunity. The destruction left everything intact. The archaeological image preserved is one of simultaneity, witness to a single historical moment. Thus we have at our disposal the initial numbers in several series of finds, the numerical relation between them, the initial position of the finds in a room, in a house or in the excavated part of the settlement.

The historical moment to which I am referring is that moment in time just before the destruction. But which finds can be attributed to that brief period? The determination is not easy because, while the use of various objects is roughly contemporaneous, their life span is not. The life of an object differs from one category of find to another. Vases can be used for several years but their position and contents refer to a very short space of time (months or weeks) and a specific utilization.

Two superposed moments might be distinguished in this synchronic image: First, the life (but not use) of the different categories of objects (vases, tools, etc.), different for each category but nevertheless belonging to a relatively limited period of time. They thus form part of the general synchronic image. Secondly, the actual moment of utilization of the objects as evinced by their position and contents (fruits, cereals, nuts, etc.) - much shorter than the first. These conventional delimitations of time are useful in our case including the basic parameter of time, vital for any attempt at reconstructing a technique.

Starting from what we already know from other parts of the Aegean (Carington-Smith 1975, 91ff; Carington-Smith 1977, 114ff.; Davis 1984, 161), we have a rough idea of the type of loom, the preparation of the yarn, the raw materials and the dyes. The type of loom in use at Akrotiri was the warp-weighted loom - an assumption based, as elsewhere, on the fact that large numbers of loom weights were among the finds: where there are loom weights the looms in use must have been warp-weighted looms (Carington-Smith 1975, 400-404; Hoffman 1964, 311ff.). These are vertical looms distinguished by the fact that the warps are not stretched between two beams as in contemporary Egyptian looms (Ling Roth 1913, 14ff.) but between an upper beam and a series of weights attached to the lower end of the warps (Crowfoot 1936, 36ff.). Loom weights thus provide incontrovertible evidence that this kind of loom was in use. This is true for Akrotiri as well.

It is also remarkable that the discoid form is the only type of loom weight found at Akrotiri. It is not just the only one in use in the actual level of destruction inside the houses, but also in the debris, i.e. of the Middle Minoan period. The roughly circular shape becomes oval, piriform or squat and mostly flat-topped (Fig 1). The diameter varies between 0.110 and 0.060 cm with an average of 0.085. They are formed and smoothed by hand from coarse, uncleaned clay with many little pebbles. Imprints of fingers and palms are often clearly visible. Some of them have grooved tops or two small notches on the periphery below the hole.

THE LOOM WEIGHTS

Loom weights are found in Akrotiri in large numbers and this is perhaps the first element which must be emphasized: over 950 have been found so far and, of those, more than 450 inside the West House. Let us remember that the excavated area is probably only a small part of the whole settlement and even that part is not thoroughly excavated. These numbers appear still more significant when compared with those from Ayia Irini in Kea, where 185 loom weights are reported from LC I levels and the excavated area is no smaller than that at Akrotiri (Davis 1984, 161). A second important fact is that loom weights are not found everywhere in the settlement, nor everywhere in a house. They are confined to some houses and some rooms in the house, a feature also shared with Kea (Davis 1984, 162). Thirdly, they belong to a type well-attested in Crete, the flat discoid type, with one or two holes, common in the Aegean in the Middle and Early Late Bronze Age (Carington-Smith 1975, 276-286), at Palaikastro (Sackett and Popham 1965, 304-305), Kea (Davis 1984, 161), Milos (Renfrew 1985, 331) and elsewhere.

The primary function of the loom weights was to stretch the wraps between the upper beam, where they were attached, and the lower end, where the loom weights were tied to the warps and remained hanging. They were thus not permanently fixed as in the Egyptian vertical two-beam loom. The tension of the individual warps could be regulated by changing the weight. We shall therefore examine the two main characteristics of the loom weights from Akrotiri: first, their weight and secondly, the way they were attached to the wraps. These two features will also be used as classification criteria rather than their shape - a non-functional characteristic in this case.

The weight fluctuates between 285 and 125 gr, the average being 190 gr. The main concentration of loom weights lies between 150 and 210 gr (Fig. 2). Since the loom weights were made for a definite use, these differences were intentional and due to concrete technical requirements. Usually a weaver would choose a bigger loom weight if the yarn was thick or if he wanted to attach a large number of wraps to the same weight. The vast majority (125 out of a sample of 160) weight between 150 and 210 gr. The other concentration, around 125 gr, reflects a quite distinct category with definite characteristics; they are substantially smaller, (less than half the size of the others) and they are the only ones with two holes. As the diagram shows, they are the least numerous of all. Then there is a small, third group of substantially heavier loom weights weighting again between 270 and 290 gr which usually show no signs of attachment. The weight of the loom weights does not seem to be related to the way in which they were attached to the warps, except for the group with two holes.

Yarn often leaves marks on the clay of the loom weights and thus it is possible to reconstruct how it was attached to the warps. The important prerequisite was that the knot tying the yarn to the loom weight should not be very tight. This is difficult to achieve precisely because the weight of the loom weights tends to tighten the knot. As the yarn has to be easy to untie so that the weaver can unknot the warps little by little in order to continue weaving, different devices are used to preclude this problem. They are visible from the traces of the yarn on the surface of the loom weights, especially if a loom weight has been used for a long time. Moreover, if the knots were very tight (perhaps because the loom weight was too big), the warp might break and mending would then require more work.

The marks left on the clay provide evidence of three types of fastening:

a)     The warps are divided in two by means of two small notches on the circumference (Fig. 1a, 3, and 4). In this way, the warp threads were separated more evenly, the tension spread over a larger surface and the knots were not very tight. Similar devices are attested in loom weights of Classical and Hellenistic times. The wraps were divided by means of a ring or a wooden rod passing through the hole (McLaughlin 1985, 79ff.). Probably the warps fastened to these loom weights were neither thick nor very thin since fine yarn should not normally be attached directly to the loom weight as the clay might damage the yarn. A feature peculiar to this type are the diagonal marks left in the clay on either side of the hole (Fig. 3). They were probably made by twigs placed between the knots of the yarns around the hole to stop them from becoming too tight.

b)     Many have grooved tops, a feature shared with loom weights from other islands (Kea, Crete). The function of the groove was probably to take a rod to which the loom weights were attached. In some of them the groove is horizontal, in others the groove follows the curve of the periphery (Fig. 1. 2, 5 and 9). The grooved loom weights show no traces of the yarn which means that the warps were not tied directly to the rod but through the hole (Fig 9). This rod might have served simply to regulate the tension and ensure that the loom weights remained in place (hence perhaps the grooved curve). Or, as Peter Warren has argued (Warren 1972, 212), the rod to which the loom weights were fastened acted as a primitive heddle. The warps could be separated mechanically by bringing forward half the warps and then moving the rod so that the shuttle could pass between the two sets of warps in one movement. The shuttle however had to be returned manually to and fro round every warp.

c)     The warps are fastened to the loom weight hole directly; probably this is the case for rather thick yarn or for yarn that would not need tying and untying afterwards. The majority of loom weights are of this type (Fig. 3 and 6).

d)     A very few loom weights have two holes; they are distinguished from the others because they are much lighter (Fig. 1. 4 and 7).

The different technical functions which are reflected in these differences are unknown, however. Since we can safely consider the numbers at our disposal to be the initial ones, we can at least establish the numerical relationship between these groups. Room 3 on the first floor of the West House has yielded about 450 loom weights. The biggest group (a), numbering 175 (almost 40% of the total), is the one formed by loom weights with no specific features. The second (loom weights with two notches on the periphery) comprises 135, i.e. 30% of the whole. This group includes all the marked loom weights (see below). Finally they are 76 grooved loom weights (17%) (Fig. 11). The only conclusion we could very tentatively formulate is that perhaps, since the vast bulk of loom weights are those directly fastened to the warps and the warps might perhaps be of thick yarn, the greatest amount of fabric woven was not the fine materials in the wall-paintings but rather some coarse, everyday textile. Another possible hypothesis would be to imagine different ways of fastening the warps for the same cloth, but that would pre-suppose different thickness and quality for the warps. This cannot be excluded but usually the differences are in the wefts: the warps are mostly identical.

Let us remember that all these loom weights came from a brief period of time just before the destruction and that consequently they were used at roughly the same time. The puzzling fact is that their numbers are surprisingly high: the loom weights form room 3 in the West House are amazingly numerous, far more than would have been needed. Warp-weighted looms required 20-30 weights, according to their length, and this room was large enough for 4 or 5 such looms. Therefore the loom weights found could not have all been in use simultaneously: there are at least three times more than needed. However, even if they were not permanently in use, we must assume that the weavers needed all of them close at hand to be changed frequently perhaps form loom to loom, probably for different kind of textiles. Otherwise it is hard to imagine why they should store such very ordinary things inside a crowded workroom. This is another argument for the different use of the different groups of loom weights; they would not otherwise have needed different warping and therefore different loom weights, the result clearly being different kinds of textiles.

Some loom weights are marked (Fig. 8 and 10). They bear an incised mark of two intersecting straight lines. So far only 9 have been found in two groups in room 3 of the West House. This is another feature probably reflecting some unknown function. Marked loom weights are known from Palaikastro (Sackett and Popham 1965, 304) and Kea (Cummer and Schofield 1984, 53-54). It is also a trait common in Classical and Hellenistic loom weights (Wilson 1930, 127; Immerwahr 1943, 65). No plausible explanation has been offered for these either.

Normally a loom is not kept permanently set up, or such at least seems to be the usual practice in modern Greece (Bouza-Koster 1980, 38). The beams and other items are usually dismantled and stored away when not in use. This does not seem to be the case for the West House at Akrotiri. It seems that here we have a permanent installation. This is implied by the central position of the room on the first floor and the general arrangement of the place (plenty of light and no wall-paintings), suitable for long and tedious work, and also by the nature of other finds (obsidian but no pottery). We should also bear in mind the room needed for other implements (spindles, baskets), material (wool for carding or threads in skeins) and finished textiles. Four or five looms are too voluminous to be stored away easily. This all points to a permanent installation.

SPECIALIZATION OR HOUSEHOLD PRODUCTION?

The presence of a 'textile workshop' is linked to another phenomenon: the fact that loom weights are not found everywhere. Although the excavated area is only part of the town, it is almost certain that not every house had looms. No loom weights were found in Xeste III nor in the House of the Ladies while, apart from those in the West House large numbers were found in Sector A (about 200) and buildings B and D. It appears that we are clearly confronted by a case of specialization. Cloth-making was the work of specialized weavers and that is why traces of looms are not found in every house but only in some of them. This, at first sight, is not surprising considering the degree of specialization in other fields at Akrotiri (Katsa-Tomara 1990).

This situation however does not correspond to the picture we have from Homer, Hesiod and Aristophanes. In Archaic and Classical times textiles were for the most part produced and consumed inside a household (Schaps 1979, 18-20). Weaving was women's main occupation but it was not the only one and was seldom done on a professional basis. Only luxury items were exchanged or marketed (Tzachili 1987). I can think of two explanations. Either we can believe what we have in front of our eyes and, despite later practice, accept that here in Akrotiri weaving and related activities formed a specialized occupation, conducted therefore by a limited number of persons as a full-time occupation and the produce then perhaps exchanged for other services or goods. Or there was a social arrangement whereby many people who did not live in the same house were allowed to utilize the workshop facilities for themselves, either for personal consumption or exchange. Such situations in fact do not occur often in modern Greece (Bouza-Koster 1980, 38). Both explanations readily fit the general image of the West House which seems rather small for the eight or nine persons (and their families) who might weave on the 4 or 5 looms (usually 2 people per loom). There is a third possibility: warp-weighted looms might not have been the only ones in use. Other types, the two-beam vertical or the horizontal could co-exist as has been argued (Carington-Smith 1975, 400-404; Davis 1984, 162). No trace would remain of either. In that case, the warp-weighted would have been a Minoan import, adopted to meet the demand for specialized (i.e. Minoan style) textiles. Specialization would then apply to a type of textile and not to the entire weaving output of Akrotiri. Personally I am inclined to believe the second. It fits better in the general schema of the Homeric and later testimonies.

A few words should be said on the advantages of this type of loom. Controlling the tension of the warps is essential for any type of loom. Any fault there could easily damage the whole appearance of the finished cloth. This tension on the warp-weighted loom could be slightly altered by changing the weight of the loom weights. In the Nordic looms of the same type, the loom weights were weighted before being attached to the warps (Hoffmann 1964, 42). A second advantage is probably the increased possibilities for multi-coloured designs. These designs may be achieved by additional heddles or by dividing and separating a number of warps, of different colour perhaps, and this might be easier since the warps were attached in bunches to the loom weights. Hence, even if the two-beam vertical loom presents certain advantages (rapidity and comfort of the weaver who works seated and beats the wefts down with the pull of gravity), the kind of textiles required led to the adoption or survival of the warp-weighted loom.

HOW WAS THE YARN SPUN?

A major surprise was the absence of spindle whorls, which are usually the second category of numerically important finds related to cloth-making activities. The absence is not total but numbers are insignificant compared with those from other Aegean sites, such as Ayia Irini. In the West House, only 6 spindle whorls have been found and it is not certain if they were in use at the time of destruction. Two of them came from the cavities beneath the ground floor, one from the square in front of the West House. The remarkable fact is that exactly the same situation occurs in Crete. 'The most interesting thing about Minoan whorls is that there are so few of them.' (Carington-Smith 1975, 263). Numerical correspondence between loom weights and whorls in the same settlement does occur in Early Minoan Myrtos (Warren 1972, 212). But in Middle and Late Minoan Crete the numbers of whorls are totally insignificant in comparison with the loom weights. Carington-Smith reports 15 from the Middle Minoan and 11 from the Late Minoan period, while the loom weights from Knossos and Zakros are counted in hundreds and Minoan flocks yielded tons of wool waiting to be spun.

This correspondence between Crete and Thera appears all the more important since the situation is quite different in other sites in the Aegean. In Kea, for example, in house A, the excavators report 122 spindle whorls and 176 loom weights altogether from the three periods of the house (Cummer and Schofield 1984, 49ff.). Both tasks - spinning and weaving - were done in the same place; in Room 27, for instance, 10 whorls and 17 loom weights are reported, from period 1. The same situation appears in the sanctuary at Phylakopi in Milos. C. Renfrew reports 18 whorls and 10 loom weights from different phases but from the same area (Renfrew 1985, 33). One possible explanation would have been another technique of spinning. Perhaps spinning was done by hand as the word χερνήτις in Homer might imply (one who spins by hand, without implements). Another possibility is that spindles and spindle whorls were made of some perishable material.

A few words about the raw materials. Wool was probably the main material used, as palaeozoic research at the site of Akrotiri has demonstrated (Gamble 1978, 748; Trantalidou 1990). The use of flax is also certain. The preparation of flax calls for a lot of water and that is precisely why it would be problematic in Santorini. Nevertheless, the cultivation of flax has been proven by ethno-botanical studies. We should not forget either that the neighboring island of Amorgos was famous in Classical antiquity for its linen. As for the costumes in the wall-paintings, it is difficult to decide which clothes or parts of them were made of wool and which of linen. Linen usually is better suited to fine textiles, whereas wool can be dyed more readily.

SAILS, BED COVERS AND OTHER PRODUCTS OF THE LOOM

So far no remains of textiles have been found. So apart from the luxurious dresses depicted in the wall-paintings, we do not know the final appearance of most of the products of the loom. We can only suppose that there must have been other kinds of cloth, for domestic and non-domestic use - bed covers for example. I think, however, that we are entitled to presume that the quantities of cloth for these needs must have been far greater than for garments. Let us remember Homer's lines when Arete prepares a bed for Odysseus (Odyssey VII, 335-338):

κέκλετο δ΄ Άρήτη λευκώλενος άμφιπόλοισιν

δέμνι ύπ' αίθούση θέμεναι καί ΄ρήγεα καλά

πορφύρε' έμβαλέειν, στορέσαι τ'έφύπερθε τάπητας

χλαίνας τ'ένθέμεναι ούλας καθύπερθεν 'έσασθαι

and white armed Arete bade her maidens lay a bedstead

under cover of the portico, and to lay on its fair

blankets of purple, and to spread thereover coverlets

and on these to put fleecy cloaks for clothing.

Let us also remember that Akrotiri was a town revolving around its maritime life and probably long a short distance trade. That means a lively harbour, preparations for long journeys at sea, boat building and repairing and most probably a large demand for sail cloth. This kind of particularly strong cloth for sails must have been woven somewhere. It must have been important work, extremely long and tedious to do. Concrete evidence does not exist, but the significance of sail cloth is illustrated by the fact that the Greek word for it (ίστία) is the same as the generic name for every product of the loom (ίστός, ίστία). No special name was needed for sail cloth, the most general was enough, as though the main cloth produced was that for sails.

A clear mention of the vital need of linen for sails comes from Athens, many centuries later, Xenophon enumerates the strategic materials for the Athenian sea-power, including linen (Xenophon, Athenaion Politeia II, 11):

Alone, the Athenians are capable of possessing the wealth of Greeks and barbarians. For if a city is rich in timber for ship-building, where will it export it, if it does not reach an understanding with the rulers of the sea? Again, if a city is rich in iron, or copper, or linen, where will it export these, if it does not reach an understanding with the rulers of the sea? Now these are the materials from which I build my ships. From one country comes timber, form another iron, from another copper, from another linen, from yet another wax.

MINOAN FASHION AND THERAN WEAVING

The visible results of weaving are there to be seen in the wall-paintings. The first remark to make is that the elaborate dresses of the Theran ladies are undoubtedly of Minoan style. Still, they present some peculiarities, some distinctive local traits (Televantou 1982, 113ff.). These differences are not in the form or the way the dress was sewn, but more in the appearance of the cloth used, i.e. in the weaving. We may distinguish four categories of divergences:

a)     The weaving patterns are clearly simpler than the Minoan ones. The impression of luxuriousness is given more by the bright, contrasting colours than by elaborate patterns which consist simply of horizontal, vertical and zig-zag stripes and only in one case (a lady from Xeste III) of a flower pattern (S. Marinatos 1976, Pl. 65). Nothing in comparison with the complex combinations of lines and dots, zig-zags, lozenges, foliage, flowers and so on of Minoan textiles (Wace, 1927, 27-34; Zora 1956, 61-79; Sapouna-Sakellaraki 1971, 153-195; Carington-Smith 1975, 305-307).

b)     A separately woven strip is sewn on the edges of the dresses. The primary function of such a band is to make the cloth stronger since the textile is usually likely to start fraying at the edges. These bands are woven by hand on small portable hand-looms often depicted in Classical vase-paintings (Clark 1985, 91ff.). The technique used, the tapestry decoration, allows more complicated patterns to be woven than the textiles from a normal loom. This is not at all the case here. The dresses at Akrotiri are reinforced by bands of a vividly contrasting colour with simple designs but no elaborate patterns (Televantou 1982, Fig. 6).

c)     A quite distinct characteristic from Crete is a special technique named lasia in Hellenistic Egypt. It consists of a series of loops made with extra wefts which gives to the textile a fleecy aspect (Tzachili 1981, 251ff.). This technique was used to make the skirt of the woman bringing water in the West House miniature scene, as well as the robes of the έπιβάται (warriors) of the boats and those of the shepherds (S. Marinatos 1974, Pl. 98, 99, Atlas, 1974, Pl. 7, 9). This technique is not unknown in Crete but very rare. On the other hand, it seems to have been very common in Mycenae and Greece. This technique was used for the textile found at Eretria.

d)     Transparent textiles do not appear in Crete and are perhaps most obviously peculiar to Thera, e.g. the veil of the so-called Iris in the crocus-gatherers fresco (S. Marinatos 1976, Pl. 58). Transparent veils might have been made from flax. The technique is probably the one termed 'sprang' which consists of interlacing wefts on small hand looms (Jenkins and Williams 1985, 414 ff.).

e)     The belts of the dresses are often in the form of ropes which is again a peculiarity to Thera and not to be found in Crete.

As mentioned before, the emphasis in this discussion about relations and the mutual influences between Theran and Cretian weaving must be placed not on the existence of totally different characteristics, but rather on the fact that the divergences are to be found in the different uses of and preferences for well-known weaving techniques and patterns in Crete and Thera as well. The textiles with loops for instance are represented in Cretian iconography but are very rare. On the other hand, they are common enough in the wall-paintings of Akrotiri. It seems a matter of local traditions and preferences. We should perhaps stress the point that looped textiles are not used for the luxurious Minoan dresses. It is shepherds, warriors and women going about their everyday work who wear these garments. However, in the dresses of Minoan type we see another Theran preference: transparent textiles. The general impression is that Crete imposed its ways of dressing, and perhaps weaving technology. But the local weaving patterns survived in the habits of weavers and the preference of the wearers.

The adoption of the Minoan fashion, even if it looks like the frivolity or caprice of a wealthy society, is no less significant for the Minoan or Mycenaean or Cycladic preponderance in the Aegean than any other argument adduced for or against the so-called Minoan thalassocracy. Frivolous societies have, in critical historical moments, proved to be the vital ones (Braudel 1979, 282) and this innocent adoption of Minoan appearance is worth a special chapter in a discussion of Aegean politics. Fashion is an easy and most effective means of influence; who but the most powerful can persuade others to imitate his appearance? Just think, on the other hand, of all the energy put in to resisting alien ways of dressing by societies which felt their autonomy in danger.

Finally, I should like to mention some points which for the moment have to remain unanswered. Who worked at the looms? What place did such work occupy in the activities of individuals and the whole community? If the products have a surplus value, how much was exchanged and how? In other words, how was weaving organized within the society? More weaving seems to have been done in sea ports (Tzachili 1987). As Akrotiri admittedly owed its wealth to trade and exchange, one factor might very well have been the manufacture of textiles which are not only easily exchanged as such, but which may also serve as a means of exchange. Homer again instructs us about their significance. Είματα εύποίητα come third in value for exchange after gold and copper (Odyssey XIII, 367-369):

αύτάρ Όδυσσεύς

άσσον πάντ΄έφόρει, χρυσόν καί άτειρέα χαλκόν

είματά τ' εύποίητα, τά οί Φαίηκες έδωκαν.

and Odysseus brought thither all the treasure, the gold and the stubborn bronze and the finely-wrought raiment which the Phaeacians gave him.

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

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