Précédent Suivant

The problems of analysis of clays and some general observations on possible results

p. 287-289


Texte intégral

1The Oxford Research Laboratory for Archaeology and the History of Art has worked for many years on the analysis of ancient clay vases by optical emission spectroscopy. Advances in the study have been not so much technical (although new techniques are being explored all the time, most recently those by atomic absorption spectrometry) as by means of interpretation, of understanding what is possible, of realising that for many years still we may have to deal mainly with negative identifications rather than positive ones. The principle (see, conveniently, BSA, lviii, 94 ff.) is to determine the chemical composition of the clay from small samples drilled from the vase or sherd. Experience has suggested that nine elements are the most suggestive of similarities and differences in major groups, and the charts of patterns produced by the plotting of the proportions of these elements are used as ready visual aids for comparisons.

2Research has generally proceeded from the analysis of large numbers of samples of pottery from excavated sites within which composition groups are distinguished which, it is hoped, can be related to the composition of local or imported clays. The method can be regarded as the classic one for such research and is basically that still in use in Oxford, in the sister laboratory at the British School at Athens, and, for instance, for the results reported here by M. Dupont in Lyons. Where definition of significant groups has failed this generally seems to have been through insufficient attention to the choice of material, since restudy generally shows that groups which are closely definable stylistically or archaeologically are also closely definable by analysis, and this gives confidence in the whole process. In the light of this observation I have promoted a complementary series of analyses which depend not on the statistical study of what is inevitably somewhat anonymous material, but on the analysis of vases or sherds whose archaeological identity is clear; in other words, to start from the known, rather than the unknown, to define the composition of archaeologically recognisable classes, to discover which are alike, which unlike. The problem of location must then rely on the identification of classes whose home is absolutely certain (Athenian black figure by major artists, for example), or by the identification of the composition of local clays from samples taken from clay beds. The latter method has proved successful in several instances, thought it need not prove successful for every centre, and it requires much prospecting.

3Work now proceeds in collaboration with the Athens laboratory, in trying to establish a «clay map» of the Classical world, through analysis of raw clay samples, analysis of archaeologically definable classes of pottery, and through analysis of assemblages of sherds from individual sites. It will clearly take a very long time and may involve many disappointments, and it clearly will rely very much on the cooperation and good will of excavators and museum curators. At present all results can only be regarded as provisional. Some, resulting from my investigation of already definable classes, have been published in BSA, lxviii, 267 ff.

4For East Greek pottery little progress has so far been made but some of the definable classes of pottery from Al Mina and Naucratis have been tested as well as a number of individual pieces in British collections. What follows is a series of extremely tentative observations based on the few tests so far conducted. I recognise that the number of samples taken for some groups is very small indeed so far, but there is some confidence to be won from the fact that, where a class is closely definable stylistically the analyses have also been found to correspond closely. A good example is the work of the Tocra Painter, whose work is definable purely on grounds of style and has been found only at Tocra where we can be sure that the vases were not made. The similarity of analyses of the three pieces tested (Boardman and Hayes, Tocra, ii, 74) is most striking and reassuring for this technique both of analysis and of interpretation.

  1. Chios. Samples from four different clay beds give a clear pattern within a fair range. Groups of «Chian» pottery from Naucratis, including the polychrome, are remarkably consistent and fall within this range, but pottery from Emporio, also within the range, is slightly different from the Naucratis classes, which suggests that the latter are not from the south of the island but, we might well imagine, rather from the centre and Chios town itself. If any of the Naucratis pottery was produced locally (BSA, li, 55 ff.) it seems not to have been from local clay. The clay of the Tocra Painter vases is decidedly not Chian (the vases were attributed to Chios in Samos, vi. 1, nos. 838-842).

  2. Dr Doumas kindly provided me with clay samples from Rhodes. One sample, from Archangelos, is similar in analysis to several Wild Goat style dinoi from Al Mina (as JHS, lx, 15, fig. 7k and pl. 2k) and to sixth-century figurines from Rhodes (Oxford G4, G9, 1948.311, 1974.313). The type is also that of a group of Bronze Age pottery analysed from Ialysos (Ialysos Type II; Catling, BSA, lviii, 104 ff.).

  3. Fikellura examples tested (including CVA, Oxford ii, pl. 6.1, 3), with one exception (see below), were of a uniform analysis. The same is found in several dish, bowl and jug fragments from Al Mina; and in the plastic terracottas first discussed in detail by Robertson (as BMC Terracottas, nos. 1659-62). Two plates of the «School of Nisyros» (CVA, Oxford, ii, pl. 3.4, 5) are only slightly different.

  4. The deviant Fikellura sample was from a late amphoriskos (Oxford, 1939.3, less micaceous than most of this class) and is matched by clay from Archipolis in Rhodes.

  5. Several bird bowls and a bird jug (the type as JHS, 1x, 15, fig. 7a-e) from Al Mina and Naucratis are of a consistent fabric, similar to (2), above, but:

  6. Other examples are of a different composition, and we shall see other suggestions that bird bowls were made in different centres and cannot be automatically regarded as «Rhodian» as yet.

  7. Rhodian Late Geometric vases (CVA, Oxford, ii, pl. 1.1-3) match black figure situla fragments from Daphnae, fifth-century figurines from Camirus and the Bronze Age Ialysos Type I (see above).

  8. Wild Goat fragments from Naucratis, including several of the black figure style generally now attributed to North Ionia are of a uniform composition (CVA, Oxford, ii, pl. 4.2, 10, 11, 21, 32, 39 et al., and the jug, pl. 2.5, 6). This is closely similar to that of Clazomenian black figure, some other unassigned black figure (including Samos, vi. 1, no. 254) and some kylikes with lip wreathes from Naucratis. A serious gap in our tests so far has been the lack of satisfactory control material from Miletus and Samos.

  9. There appears to be an overall similarity in clay types from the East Greek area so far analysed. The Northampton vase is seriously dissimilar from this range.

  10. Cups with dipinto dedications to Hera in Naucratis are of uniform fabric (Samian?) unlike others studied here, but closest to:

  11. The distinctive Ephesian style, studied by Greenewalt in California Studies vi, 91 ff. Moreover, bird bowls also from Ephesus are of the same composition, suggesting local production though they are stylistically not distinguishable from the familiar series.

  12. Gorgoneion Group plastic vases (the helmeted heads, CVA, Oxford, ii, pl. 8. 1-3, and Reading) are unlike (2) and (3) above, but of a consistent composition closest to (11).

5All these groupings are at present quite tentative although some decided negative results can be claimed and a fair indication of important groups which can be recognised both stylistically and by analysis has been achieved. Experience has told that judgement of fabric by eye can hold most of the time but can never be decisive for an individual piece, and in the East Greek area the visual differences of fabric, even on the same site in different periods, are not such as to lend great confidence to identification by eye alone.

6Stylistic judgements serve well only for the very distinctive wares but in the varieties of Wild Goat decoration, and certainly with the plainer wares where shapes are not distinctive, other criteria have to be sought if origins are to be seriously determined. We are still very far from achieving any sort of control over even the major East Greek centres but, with the cooperation of excavators, museum staff and students in the field, we may hope that later generations of archaeologists may have more secure grounds for assigning East Greek wares found outside the East Greek area than we have today.

Précédent Suivant

Le texte seul est utilisable sous licence Licence OpenEdition Books. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.