Investigation of the Technology of Manufacture of the Local LBA Theran Pottery: Archaeological Consideration

These techniques seem to be the result not only of a long artistic process but, also a technological evolution. In some cases a knowledge of the level of ceramic technology may clarify issues for which archaeological analysis is insufficient. Hopefully, the collaboration with the Demokritos Laboratory of Archaeometry will throw further light on such questions. Some of the initial problems for which the technological contribution has already proved helpful are the copious use of white in the decoration of Theran vases and the manner of application of bichrome and polychrome decoration by Theran potters. The result is that both the artistic preferences and the technology of the Theran workshops follow the same course - that is the continuation of tradition and the adoption of Minoan elements for adaptation to local needs.

INTRODUCTION

Typological examination of the pottery of the volcanic destruction level at Akrotiri has shown that this is characterized by a considerable variety of decoration techniques. This variety is the result of the Theran potters' many and diverse sources of inspiration, primarily the local tradition, Minoan pottery and the wall-paintings (Marthari 1987). However, the determining of the principal aesthetic trends underlying the evolution and final form of Theran pottery is not the sole means of interpreting its developmental course. For a fuller understanding of this it is necessary also to examine other influential factors (Shepard 1954, 95). One of particular importance is the technology available to the Theran workshops, which was the means through which the potters expressed their aesthetic preferences.

In the investigation of pottery technology, archaeometric methods for the examination of the quality of fabric, the composition of surface coats and decoration pigments, and the firing conditions are particularly useful (Jones 1986, 83-89). To date, there have been very few studies of this nature with regard to Theran pottery (Maniatis and Tite 1978; Noll 1978). The main contribution of Noll's study, which focused on decoration technology, is his comparisons between Theran and Minoan pottery, which he has also studied (Noll et al. 1971; 1974). However, the samples on which his work is based are representative of only a few decoration techniques in relation to the wide range subsequently distinguished and defined as a result of systematic study. To cover this gap, a complete series of samples, 90 in all, representative of all Theran wares defined below, were sent to the Demokritos Laboratory of Archaeometry for examination by Dr Maniatis and Ms Aloupi.

Difficulties hitherto encountered in utilizing the results of archaeometric examination of Theran ceramic material (Doumas 1978, 459-514) are frequently due to the method of sampling. For this reason the present sample was selected on the basis of certain principles so that the technological results would lend themselves more readily to archaeological interpretation. 1. There has been preferential selection of diagnostic sherds: rims, handles, bases, feet, spouts, nipples or other characteristic parts of the vases (Fig. 3). 2. A conscious effort has been made to select sherds representative of the wide range of forms and decoration schemes known. 3. The number of sherds corresponding to each of the groups defined below is, as far as possible, proportional to their frequency of occurrence in the settlement.

The provenances of the samples are two: the buildings of the final phase of occupation of the settlement, and accumulated debris in the open areas which yielded pottery contemporary with that from the buildings. This debris accumulated as a result of organized attempts to repair the damage caused by earthquakes, before the inhabitants were finally forced to abandon the town (Doumas 1974, 111-112).

THE LOCAL THERAN WARES

Local vases comprise approximately 85-90% of the total pottery from the final occupation phase of the settlement. Their fabric is semi-coarse and light-coloured, i.e. buff, often with a greenish or orange tinge. The Theran origin of this fabric has been proved by both chemical (Jones 1978; Kilikoglou 1988, 137-174) and petrological analysis (Einfalt 1978; Williams 1978; Vaughan 1990; for unpublished studies by Williams 1980 and Pittinger 1979, see Jones 1986, 275).

Despite the visible similarity of fabric of this large quantity of pottery, the decoration techniques are diverse. Elsewhere (Marthari 1987) an attempt has been made to classify only the patterned wares, mainly according to style of decoration in relation to their origin. Here we shall consider in total the pottery with the aforementioned fabric, invoking strictly technical criteria for its classification - i.e. surface treatment and decoration technique.

On the basis of surface treatment, this pottery falls into the following three groups, including both plain and patterned wares:

   A.      Washed or self-slipped   (Fig. 1, centre)  

1. Plain (row 1).
2. With red splashes (row 2, 1rst right).
3. With red splashes and dark motifs (bichrome).
4. With dark motifs (row 2, 2nd right).
5. With red and black motifs (bichrome) (row 2, 2nd left). 
6. With dark motifs and white enrichments (bichrome) (Fig. 2, 37).
7. With red and black motifs and white enrichments (polychrome) (row 2, 1st left).

   B.      Dark coated, matt or burnished (Fig. 1, right).

1. Plain matt (row 1).
2. Plain burnished.
3. With white motifs, usually burnished (row 2).
4. With bichrome surface (red and black) and white motifs.

C.      White coated and burnished (Fig. 1, left).

1. Plain (row 1).
2. With dark motifs (row 2).

White-coated with dark motifs (C2, Fig. 6) is difficult to distinguish since the greater part of the white coat has been effaced and tiny preserved patches have only been observed after close scrutiny of the vases. Our conclusion that these vases were, as a rule, decorated with specific motifs or sets of motifs (Marthari 1987, 366, 376) prompted us to search for such a coat even on vases where it was not immediately apparent, and consequently the group was augmented.

The separate groups of vases do not occur with the same frequency in the settlement: some are more common than others. Their statistical pattern and detailed study are problems addressed in my doctoral thesis.

The aims of the ongoing collaboration with the Archaeometry Laboratory of Demokritos are firstly to confirm or refute, through analyses, certain observations and secondly to answer certain questions concerning the manufacture of the vases, more specifically, the processes of preparing and firing the fabric, coats and pigments. The elucidation of such technical issues, in conjunction with other evidence, will enable reconstruction of the way in which the Theran pottery workshops operated; analysis of 29 of the 90 samples has already furnished most useful results (see the catalogue).

FABRIC AND DECORATION

The Theran clay is calcareous and it is to this that it owes its buff colour (Jones 1986, 759). The greenish or orange tinge observed in some cases may perhaps be due to its variable calcite content. However, this is evidently not the only reason, since it has been noticed that those vases of greenish fabric usually have a dark brown or black coat or motifs (e.g. sample no. B3-58), whereas these of orange fabric are usually decorated with a red or light-brown coat or motifs (e.g. sample nos. B3-55, 61, C2-77, 78, 80). This correlation, remarked on a large number of vases, prompted the thought that the same factor influenced the colour of both the fabric and decoration of the pottery. Laboratory examination has proved that this factor is the firing temperature. The samples of these two fabrics are located at the top and bottom ends of the scale of firing temperatures compiled by Aloupi and Maniatis (Table I) for the samples examined to date. Firing temperature for the orange fabric ranges from 700-850° C and for the greenish fabric from 1050-1080° C. The remainder of the fabrics were fired at the intervening temperatures.

With regard to the decoration, my impression on examination by eye is that the dark coats and decoration are of similar composition, which holds also for the white coats and decoration. This impression has been completely verified by analysis. As a rule an iron-rich clay was used for both dark coats and decoration, while talc was used for both white coats and paints. Thus with just two materials, used in various ways and fired in a reducing atmosphere, the Theran potters created most of the wares defined above. In some cases they coated the surface with a suspension of iron-rich clay (slip) and drew the motifs with talc in liquid form (group B3), in others they did exactly the opposite, i.e. the surface was coated with talc and the motifs painted in iron-rich clay (group C2). They also combined the two materials to decorate washed or self-slipped surfaces (group A6).

Table I: Concentrated results on firing temperatures and colours of fabric and decoration. For a full technological description of samples see Aloupi and Maniatis (this volume), Table V.

There are, however, also vases for the decoration of which it is quite clear on examination by eye that two different dark paints were used: red and black (groups A5, A7). One of our questions was how this bichrome effect was created. Laboratory analysis of sample A7-39 (Fig. 2, Aloupi and Maniatis, parallel paper in this volume), on which a swallow with black head and red neck is depicted, has shown that this was achieved by application of the manganese black technique (Noll et al. 1973; Jones 1986, 762-763). In this instance the red derives from an iron-based paint and the black from a suspension of manganese-rich clay. This sherd could have been fired in an oxidizing atmosphere, which was sufficient to produce the red and black colours, even though the exact firing conditions have not yet been clarified. The manganese biack technique has also been detected by Noll (1978, 504).

This technique is simple because it does not require strict control of firing conditions, which is probably the reason it was used even in Neolithic times on the Greek mainland (Jones 1986, 778). In the Cyclades, where there is a rich tradition of vase production with bichrome decoration on a light ground from the MC period (Barber 1974, 34-35; 1978,  376; Overbeck 1984, 111; Davis 1986, 4-5; Marthari 1984, 129; 1990), this technique must also have been known since at least this period, though it was unknown on Crete (Noll 1978, 504). In Crete a completely different technique was used for achieving the bichrome effect on contemporary LM IA vases, demanding considerable skill in execution. The iron reduction technique, normally employed for the dark colours, was carefully elaborated upon, the paint being applied by brush more thinly or thickly, thus achieving shades of red and brown (Popham 1967, 339 Pl. 77b; Jones 1986, 791).

Most of the dark-coated vases with white motifs (B3, Fig. 3) imitate Kamares pottery in its final phase, with regard to the motifs and their syntax, with only minor differentiations. Only a small percentage is decorated in the Cycladic tradition (Marthari 1987, 372-373, 374, 376). I formed the impression that there was an attempted imitation of Kamares pottery on the grounds of technology also. However, the surfaces of the Theran vases vary in colour from red to dark-brown and black, in contrast with the standard black surface of their Kamares counterparts. Furthermore, the Theran vases are almost always burnished, whereas the lustrous dark ground of Kamares pottery is not due to this. Noll has made some preliminary interesting comparisons between Theran and Minoan vases (Noll 1978, 503). The present analysis (Aloupi and Maniatis, parallel paper in this volume) fully explains the reasons for the differences observed. It has been proved that it was the Theran potters' intention to achieve a black colour, like that of Kamares vases, but the coarse and granular nature of the dark Theran coat required particularly high temperatures, above 900° C, to improve its quality. At such temperatures, however, it was difficult to maintain the reducing atmosphere necessary for achieving the black colour, which is why many of the vases were in fact red or brown in colour (Table I).

Here we witness a 'provincialism' of the Theran workshops. Although there is a conscious effort to approach the developed Minoan technology, their persistence in using inferior quality material prevents them from achieving the desired result. To counteract this defect, the coarse surface was burnished, a technique with deep roots in the Cyclades (Coleman 1977, 109; Barber 1978, 374-375).

The white coat and decoration merits particular attention since from my study of Theran pottery as a whole it is apparent that white was extremely popular. In five pottery groups referred to in this paper white plays an important role in diverse ways, which is why laboratory analyses of samples of these groups were requested first. It is used as a surface coat (C1-C2, Fig. 4-6), for painted motifs on dark coat (B3, Fig. 3), in combination with other colours for motifs painted directly on the washed surface (A7, Fig. 7) and as an applied enrichment for rendering details on dad motifs (A6, Fig. 2, 37).

This preference for white becomes even more striking when we compare Theran pottery with wares from other Cycladic islands, such as Milos and Kea, where the use of white is referred to only rarely (Davis 1986, 5). On the contrary, comparisons with Crete indicate that two of the four decoration techniques mentioned above - i.e. B3 and A6 - are similar to those on Kamares pottery and LM IA style respectively. The other two techniques, C1-C2 and A7, are peculiar to Thera (Marthari 1987, 373-378).

In the two latter cases white is used directly on the surface of the vase. With regard to the white coat (Fig. 4-6), this is indeed puzzling. As we have seen, Theran clay is light in colour, due to its high calcite content. One of the advantages of calcareous clays (Maniatis and Tite 1981) is that dark decoration is more strikingly projected, on account of the contrast between motifs and ground. Consequently this cannot be the reason for coating the surface with white, a view further supported by the fact that there are Theran vases with white coat but without painted decoration (C1, Fig. 5).

The archaeological interpretation of this peculiar group, C, is interesting and is based on a detailed typological study and examination of their provenance, a subject which I consider elsewhere. However, laboratory analyses have thrown light on a purely technological aspect of this problem. From examination by eye these vases appear to be light burnished, an observation which has been confirmed by examination under a stereoscopic optical microscope. Aloupi and Maniatis suggest that talc may have been used to achieve this burnish, giving the vases a smooth and soapy surface, an effect which could not be achieved by simple burnishing of the semi-coarse Theran clay. Consequently the talc coat produced finer quality light surfaces in comparison with the washed or self-slipped surfaces of group A vases.

From comparison by eye of the white coat of group C vases and the white of the motifs of groups A6, A7 and B3, the following differences emerge: the white of the decorative elements seems to be thick and quite well preserved, loss being largely due to flaking. The white coat seems to have been a suspension smeared unevenly over the surface with a brush, traces of which are still visible in many instances (Fig. 5, 6). It is of thinner consistency and though effaced in many places this does not seem to be due to flaking. Laboratory analyses (Aloupi and Maniatis, parallel paper in this volume) have shown, firstly, that when used as decoration, the white layer is of the order of 10-15 mm in thickness, and when used as a coat is much thinner. Secondly, the mineral talc is quite foreign to the clay and its adhesion to the vase surface or to clay-based paints is difficult to achieve. The white coat cannot have been a solution of talc in water since this would not have adhered to the surface at all. Thus the talc must have been mixed with some organic medium. Even so, much of the coat would have been removed by burnishing, which explains why the coat is so thin and uneven, particularly in vases of group C.

The manner of preparation of this white coat, and the difficulty of adherence to the surface, in conjunction with the small number of group C vases, give rise to two lines of argument: either only a limited number of these vases were produced by the Theran workshops, because of the aforementioned difficulties, or they were produced by only a very few specialist workshops, possibly just one.

Another interesting fact to emerge from laboratory analysis is that the extensive damage to the coat, particularly in vases of group C2 (Fig. 6) is due to the low temperatures at which these were fired, since talc sinters extensively above 1050° C.

Three of the four samples from group C were fired at temperatures between 700° C and 800° C, which is why the white coat is so fugitive (Table I). This would seem to be the case for most of the vases of this group, which helps explain the difficulty in distinguishing them (see above).

Fortunately the talc beneath the dark decoration is protected by it and thus preserved. Since this can be detected by analysis it will be possible to confirm its presence or absence on the surface of the vases where its existence is suspected yet no visible traces have remained.

Noll has suggested that, since talc has not been found on any other ancient pottery except Minoan, 'its use on Thera shows to what extent the Theran potters followed the technical experience of the mother-country.' (Noll 1978, 503.) The distinguishing of groups in which white is used as coat, samples of which were not available to Noll, and the fact that this also has been proven to be talc, considerably alter his argument. It is quite possible that the Theran potters adopted the use of talc in the preparation of white colour from Crete, but the manner in which they used this borrowed element indicates the extent to which they were able to adapt a foreign technology to their own needs, in this instance the improvement of the light-co loured surfaces of vases of semi-coarse fabric.

The use of talc directly on the surface of the vase, and not applied to the dark painted area as is customary in Crete, for more precise rendering of details on pictorial motifs, such as the underside of the swallow's wings (Fig. 7), is another version of this same trend. In this latter case it highlights the traditional preference of Theran potters for naturalistic representations, in contrast with the decorative spirit of both Kamares and LM IA styles.

CONCLUSION

Typological examination of Theran pottery has shown that this is dominated by three main trends: the developmental continuation of local tradition, the adoption of Minoan traits, both as imitations and free adaptations, and the combination of elements from both these sources, along with other influences, giving rise to completely new creations. These latter are the best expressions of the spirit of Theran culture in the period prior to the final abandonment of the site (Marthari 1987).

The study of ceramic technology has shown that this followed a similar evolutionary schema. Use of traditional techniques, such as manganese black technique, continues alongside the attempted application of Minoan ones, such as that used in the production of Kamares pottery. The mixture of elements from both traditions is apparent in the burnishing of dark-coated vases, which are of Kamares technique in all other respects. Finally, elements borrowed from Minoan technology, such as the talc-based white colour, are used in a completely different manner from that prevailing in Crete, either in order to mask technical weaknesses or to enhance the expression of the Theran aesthetic.

It thus emerges that Therans resorted to Crete on technical matters, on account of the high level attained by Minoan pottery technology, by Aegean standards, during this period. This level is best appreciated if we consider that the high quality Kamares pottery was an MBA creation. The Theran workshops display a remarkable facility in the creative transformation of Minoan techniques to serve their own particular needs, yet were apparently unable to improve their materials in order to approach Minoan standards. Consequently the vases produced were inferior in quality to those they were intended to imitate. Thus Theran ceramic technology, though interesting and inventive, is nevertheless inferior to the Minoan with its rich tradition of high quality achievements.

CATALOGUE

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

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