Material and Techniques of the Minoan Ceramics of Thera and Crete

In many respects, the artists of Thera followed the tradition of Minoan Crete, e.g. they used calcium rich clays for the production of the ceramic body, they adopted the technique of "sinter layer" for the production of the "Urfirnis", the technique of "Glanzton" layers and the use of talc as a white pigment in vase painting. But they also employed their own methods. There are differences which seem to open some ways to differentiate between the production on both islands.

INTRODUCTION

By the kind assistance of the late Prof. Marinatos, during a visit to the excavations of Akrotiri in 1973, the author had the opportunity to study samples of the ceramic material which has been excavated at Akrotiri in such an impressive amount and quality ⁽¹⁾. The aim of this work was to get information on the raw materials used and the techniques of manufacturing. Of special interest seems the question, to what extent potters and artists on Thera learnt from the experience of Minoan production on Crete or went their own ways. For comparision with Crete the results of a series of my own studies on the techniques of ancient pottery were available (Noll et al. 1975A; Noll 1976). It has been shown that the decoration and its technology are the most informative. This is also evident from a comparative study of the ceramics of Thera and Crete.

Our sherds are representative of the following types of Akrotiri ceramics:

(A)      A ware coated with a dark brown "Urfirnis" and a white decoration painted on it (pl. 1).

(B)      A ware without such coating. The decoration has been painted directly onto the surface of the sherd in black, brown and red (pl. 2).

(C)      A ware bearing a decoration comparable to that of (B) but with white spots and stripes on the dark parts (pl. 3).

Stylistically the decoration of type (A) corresponds to that of the Cretan Kamares ware, (B) to that of the late Minoan LM I ware. (C) is supposed to belong to an intermediate phase between both. Marinatos has shown that the decoration of the Theran (B) type ware represents the Cretan floral style and comes to an end when that has reached its maturity. It is on the supposition that the styles of ceramic decoration developed on Thera and Crete simultaneously, that the conclusions about the time of the destruction of the Minoan settlement on Thera have rested.

THE CERAMIC BODY

One of the most striking features of the Theran ceramics is its high calcium content. It varies between about 10 and 18 % CaO. If the compositions, not considering the contents of iron oxide and alcalies, are plotted into the system SiO₂ / Al₂O₃ / CaO + MgO they fall into the triangle quartz / anorthite / diopside (fig. 1).

The Minoan ceramics of Crete as well as Mycenaean and Attic wares and even the clays employed in the manufacture of pottery on Crete and Aegina nowadays lie in the same area.

The phase content is uniform, too. Quartz, plagioclase, diopside, calcium carbonate which has not been converted into silicates during the firing process, mica, iron (III) oxide and sometimes small amounts of gehlenite have been found. Part of the iron is camouflaged in diopside as iron (II) ion.

The preferred use of calcium rich raw materials for the ceramic production is not confined to Aegean countries. We found it also in the sequence of the neolithic-chalcolithic cultures of Great Mesopotamia (Noll 1976A). Therefore the question arises, is this an accidental matter or is it caused by an intentional selection from among the raw materials which were available to the ancient potters? The following facts could favour the possibility that a selection has been made:

1. Calcium acts as a fluxing agent. It enables the potter to reach a sufficient degree of sintering as well as body strength even at the relatively low temperatures which could be obtained in the primitive kilns.
2. As shown by Tite (1975), the structure of the sherd remains essentially unchanged in the region of about 850 up to 1050° C. This is the range of temperature within which the ancient potters were able to fire. Therefore an exact control of temperature was not particularly critical.
3. The thermal expansion coefficient of the sherd is increased by the calcium content and approaches that of the paint layer. This becomes the more important the more the paint layer has been vitrified. Thus the danger that the paint layer splits off during cooling is reduced.
4. Finally, a calcium content of the sherd brightens its color, probably by camouflaging the iron (II) ions in diopside. Therefore the contrast between the colors of the paint and the surface of the sherd is increased.

Of course these interrelations were not known to the ancient potters, but it seems probable that they had gained an empirical knowledge of the advantages brought by the calcium-rich raw material.

Looking for chemical differences between the ceramics of Thera and Crete our interest was directed to the minor constituents. By means of optical emission spectroscopy the contents of Sr, Ba, Mn, Cr, V, Ni, Cu and Zr were determined semiquantitatively. The result was that on the average Ni, but also Cr and V are enriched in the Cretan ceramics. By means of neutron activation analysis which was performed in the Gesellschaft für Kernforschung by the kind help of H.C. Einfalt, Mineralogisches lnsitut der Universität, Karlsruhe, it was shown that in a total of 11 sherds Co allowed a clear distinction between Cretan and Theran material (Thera: 12 - 16 ppm, Crete: 21 - 48 ppm). Further and more extended research seems necessary to confirm if cobalt can play the role of something like a geochemical "Leitelement".

The following findings may also be helpful for a mineralogical differentiation:

1. By thin section microscopy it becomes evident that the Theran ceramics contain more and coarser grains of rock fragments than the pottery of Crete. The egg-shell ware in particular was made from a finely slurried clay.
2. X-ray diffraction shows that in the Theran material higher percentages of calcite can be found than in that of Crete. That means that generally the potters on Thera fired at lower temperatures than potters on Crete. The same becomes evident from the figures for the loss of ignition at 1000^o C: the average of 6 Cretan sherds was 4,2 %, of 6 Theran sherds 11,0 %. Further indications for the levels of firing temperatures will be given by a study of the paint layers.

Provided that a broader statistical test basis confirms these findings, they may prove helpful for characterizing and distinguishing the ceramics of both islands.

Finally it should be mentioned that from the phase content it is also possible to gain some quantitative information on the firing temperatures. In a study by Peters & Jenni (1973) data for the temperatures are available at which in calcium rich clays diopside is formed and calcite is decomposed. The formation of diopside begins at about 850^o C, its presence being therefore an indication that this temperature has been reached. The decomposition of calcite in mixtures with clay minerals begins at about 650 - 700^o C. But the bulk of the carbonate disappears at about 800^o C so that the firing temperatures of calcite-rich sherds must have been below this figure.

THE DECORATION

It has been shown that valuable information on the techniques of ceramic painting can be gained by a study of the chemistry, phase content and morphological features of the paint layers (Noll et al. 1975A ; Noll 1976). The methods used are energy dispersive X-ray micro analysis, X-ray diffraction by powder methods and scanning electron microscopy (Noll et al. 1977).

In this way it could be shown that the so called "Urfirnis" covering the sherd of the Kamares ware is a silicate "sinter layer" produced by firing iron and potassium rich clays in a reducing atmosphere at relatively high temperatures (950 - 1100^o). The type (A) ware of Thera bears such a coating too. It does not differ chemically from that of the Cretan Kamares ware (Fig. 2). Potassium and iron (II) ions act as a fluxing agent. Therefore, and because of the high firing temperatures, the sinter layer of the Cretan Kamares ware is relatively rich in fused mass and in bubbles (pl. 4). In the sinter layers of the corresponding Theran ceramics these features are less pronounced (pl. 5).

During firing, iron (II / III) spinels are formed which are carriers of the dark colors. In the Cretan sinter layer of Kamares type hercynite has been identified, in the corresponding Theran ware maghemite, besides smaller percentages of hematite.

This may be the reason why the sinter layers of Theran ceramics generally have a brownish color. Iron (III) oxides were formed either by an incomplete reduction or by reoxydation during cooling.

White paint is produced on both islands by using talc as a white pigment. The plate-like shape of the mineral particles remained unchanged during firing (Pls. 6, 7), but X-ray diffraction patterns show the interferences of protoenstatite, i.e. the product of a thermal transformation of talc. This means that the white painting has been fired too. The transformation of talc into protoenstatite begins at about 900^o C, the degree of order of the protoenstatite lattice being improved in the range up to 1100^o C. Therefore the firing temperature can be estimated on the basis of the X-ray patterns. The result is that the Theran material of the MM A-type has been fired at about 900^o C, the material of Crete between about 950 to 1100^o C.

To date, talc as a white pigment has been found in no other ancient ceramic paints. It is typical of the Minoan technique. Its use on Thera shows to what extent the Theran potters followed the technical experience of the mother-country.

Red paints, often applied on the Cretan Kamares ware, were produced by iron oxide ochres and, like the white paints, fired together with the ceramic body and the sinter layer. In contrast to the material of the sinter layer the ochre is practically free from potassium and retained its powdery structure during firing without sintering. It became black during reduction like the sinter layer. But because of its porosity the color changed to red when the reduction was followed by oxidation.

At the same time the sinter layer remained black ("iron reduction reoxidation technique": Noll et al. 1975A; Noll 1976). This technique was invented by the Minoan potters on Crete. The absence of red paint on the Theran (A) type ware indicates that the potters there did not make use of this method.

Simultaneously with the stylistic change from MM III to LM I an alteration in techniques of ceramic painting also took place. Now black, brown and red paints are applied to the plain uncoated vessel. For a red pigment iron rich ochres were used (pl. 8). Black was produced again by the reduction of iron and potassium rich clays. But as is shown by spectra: (Fig. 2) the dark paintings of this period differ from the Kamares sinter layer in having a higher potassium content. From this it can be concluded that at this time the clay mineral illite was concentrated by settling and only the fine grained fractions were used. It is characteristic for this type of paint that its material is homogeneous and practically free from bubbles (pls. 9, 10). The surface is smooth and has a high gloss. A. Winter has reproduced the technique (1959). It became evident that the quality of the ancient material could be reproduced if the firing temperature were not higher than about 850^o C, i.e. less than with the sinter layers. Thus the texture of the paint layer consisting of parallel orientated clay mineral flakes remained practically unchanged and made an important contribution to the high gloss. Such paint layers have been called "Glanzton" layers (Hofmann 1962; Oberlies 1968). They were invented as early as the Halaf culture (Noll 1976). The Minoan potters invented it once again and initiated a development which dominated ceramic painting techniques up to the end of the classical Attic period.

The Theran potters adopted this technique and rapidly learnt to decorate their ceramics with fine bright paints.

On the other hand there are indications of a certain autonomy on the part of the Theran craftsmen. They also used the manganese black technique (Noll et al. 1973) in order to produce black paint. This technique has not yet been observed on Crete. It is characterized by the use of manganese rich ochres which were added to the clay dispersions before painting. During firing manganese iron oxides of different types are formed as black pigments. In three black paints of Akrotiri manganese spinels and hausmannite were found. They are formed in an oxidizing fire. It was easy therefore to produce manganese black and iron oxide red side by side since no changes of firing atmosphere were necessary.

The manganese technique is a more specialized technique of ceramic painting. Since it appears on Cyprus at the end of the seventeenth century B.C. it may be assumed that the Minoan potters on Thera acquired the technique from there.

As in the Cypriot wares the manganese black layers on Thera contain small rounded pigment particles whose shape and size is typical of ochres (pl. 11). The appearance of this technique on Thera indicates that in spite of their dependency on Cretan tradition the Minoan potters on this island went their own way too.  

- (1). At the same time the material of the wall paintings could be studied. A report on this subject has been given in Noll et al. 1975.

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

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