Shine Like the Sun:

Chapter 1: Introduction and Objectives



INTRODUCTION

All material objects tend to be interesting to the people who study them, but fine or high-quality Islamic glazed pottery is of interest to a variety of groups, notably art historians, archaeologists, and ceramic technologists.

A broad all-encompassing approach to Islamic ceramics has traditionally been the preserve of art historians, who consider the high-quality wares in the context of the fine and minor arts of the period, utilizing traditional connoisseurship approaches. Although there have been and continue to be important contributions from this group, methodologies for those individuals who have investigated the ceramics have generally not progressed since the days of such researchers as Arthur Upham Pope and other workers in the 1930s.

Much early archaeological work on Islamic sites was aimed at establishing chronologies and provenances for the fine wares studied by art historians. The discovery that these wares were often a relatively small part of the excavated assemblage, if found at all, was most disconcerting. More recent excavations on Islamic sites were aimed at a general understanding of Islamic society, rather than finding fine ceramics. Hence, the now standard methodologies of archaeological ceramic study, including formal analysis, seriation, and study of motif assemblages, were applied only to the large corpora of locally produced wares. The fine ceramics of art-historical importance would never be found in sufficient numbers on a single site to provide a complete sequence, and in some cases may have been represented only by undiagnostic sherds. The pottery of art-historical relevance could, however, be of great interest to archaeologists. Although whole vessels of high artistic merit are uncommon, glazed wares are not necessarily rare, at least in regard to sites that represent the principal residences of the bulk of the population. As a rough guide, Robert Adams, who surveyed and excavated extensively in Iraq and Iran, considered an Islamic site with less than 10% glazed wares to be a small rural site (Adams 1970). Larger percentages would be expected for urban sites, for example, ninth-century Siraf had 17% glazed wares (Hodges and Whitehouse 1982:148). Only in regions of considerable impoverishment of the cultural record do glazed wares become at all rare, examples including most of Jordan, which seems to have largely declined rapidly to a pastoral economy, apart from sites of strategic military importance. For some wares a single place or at least single region of manufacture was suspected, and from there the pottery was distributed across the Islamic world. In some cases this distribution took the wares as far as Europe, South Africa, and EastAsia -- the limits of the Old World (Map 1 includes a number of sites beyond the Middle East that contain Islamic pottery of the early medieval period, mostly from Ho, 1995, and Glover, in press). One could use here the contentious term "trade," for the pottery was often moved as merchandise. However, this pottery would have been distributed in several other ways as well. It was given as diplomatic gifts, and would have been moved as personal property, with artisans, soldiers, and others who travelled freely about the Islamic world. As some dating evidence existed for most of the fine Islamic wares, identification would have also aided in dating of archaeological contexts, particularly in previously unexplored regions (e.g., The Yemen, Mason and Keall 1988a). Unfortunately, these wares may be too poorly preserved, particularly as surface finds, to make precise identification possible. Such are the conditions, particularly for Islamic pottery found on European sites (Hurst 1968).

The ceramic technology of the fine Islamic wares represented some of the most advanced in the medieval world, and has attracted the attention of a number of analysts. Methodologies have ranged from early research using wet chemistry to the latest techniques in electron microscopy. Some technological questions have been answered by practice rather than by analysis, notably, lustre-painting technology. Lustre-painting is a unique contribution of the Islamic world, and is fully explained by Alan Caiger-Smith (1973:25-27; 1985:195-220), a potter who has himself used the technique extensively. Caiger-Smith considers that the technique used in the Islamic world should be more properly called reduced- pigment lustre, to distinguish it from other lustre-painting techniques. However, as reduced-pigment lustre was the only technique used by the potters discussed in the present study, such precision of vocabulary is unnecessary. Lustre-pigment is a compound of silver and copper in a refractory earth, applied to the glazed surface of a previously fired vessel. The painted object is then refired to a red heat in a reducing atmosphere in a small specially constructed kiln used only for this purpose. In the second "firing" the metals are bonded to the glaze as a thin layer with a strongly metallic lustre; the refractory earths can then be brushed off (Caiger-Smith 1973:25-27). This is considered to be a complex and specialized process, which was probably also kept secret. The practice of modern craft potters provides other insights into technology, such as in identifying the colourants of underglaze and in glaze pigments (e.g., cobalt for blue, manganese for purple, antimony for yellow). However, in other cases knowledge derived from practice only produces confusion for non-technologists.

THE PROBLEMS

Problems of where

One of the biggest problems in the study of Islamic glazed wares is the determination of their place of manufacture: their provenance. The most common evidence for production of a particular ceramic type is often simply its presence in abundant frequencies. The fundamental assumption underlying this attribution is entirely unsatisfactory. Indeed, if comparison is made with medieval pottery production in the Saintonge region of France, where luxury wares are known to be produced but are rarely found (Chapelot 1983), then the opposite conclusion could equally be obtained.

There are a number of ways to determine the provenance of a piece of pottery. Some Islamic pottery actually has the provenance written right on it; notable in the present study are the pieces ascribed to Kashan (see Chapter 6). However, use of this evidence relies on the honesty of the painter. Further, a fundamental assumption is made in ascribing uninscribed pieces to the same provenance on the basis of a few pieces, which effectively do not represent a statistically sufficient sample size. Possibly included in this category are pieces signed by individuals with a nisba, an element in the name that indicates the individual's place of origin. A long argument could be made (by people better qualified to dispute it than the writer) concerning whether a nisba indicates that the potter was working in his home town or had the opposite meaning.

A second line of evidence consists of documentary sources contemporary with the production of the pottery, such as travellers' journals. Even if considered reliable, these are usually unspecific about exactly which type of pottery was made at which centre, and could not be used to rule out production at an undocumented centre.

Physical traces of manufacture form a third line of evidence. A waster is a piece of pottery that was sufficiently damaged in the kiln to render it useless, and was discarded on the site. Although technically this could refer to sherds or vessels that appear perfect, it is preferable to use pieces that are self-evidently wasters, such as seriously warped pieces or several vessels stuck together. Strictly speaking, evidence such as this is evidence only that the ceramic type found as a waster was made at the site where it was found; it cannot rule out that the same type was also made elsewhere or indicate which other types may have been made at that site. A particular problem with this line of evidence exists with Lustre-wares. As stated above, the actual lustre-pigment is applied in a second low- temperature firing. As the major part of the stress of the firing process is done before the application of lustre, kiln wasters of Lustre-ware are practically nonexistent. Possible mistakes, such as blackening of the glaze or volatilization of the pigment, would result only in a "second" rather than a waster, and could feasibly have travelled (see Chapter 2). Hence, inadequately published or described accounts of Lustre-painted wasters should be treated with suspicion, even if the sherds were reportedly found in a kiln, particularly when the kiln is said to contain other wares (the author has actually found in a kiln in the Yemen Lustre-painted sherds of pottery that had been made in Basra.

Collectively this evidence can provide indications of varying degrees of reliability for production at a particular centre. However, even in the rare case that this evidence proves that a ceramic type was made at a particular centre, it will not conclusively prove that the type was not also made elsewhere. The only sure way to identify the products of a particular production centre, and to rule out the possibility of production elsewhere, is by analytical characterization and provenance studies. In such studies the products of a kiln must be characterized by the presence of geographically dependent variables. To do this it is necessary to analyse certain kinds of evidence, including wasters, kiln furniture, and other material from kiln sites; the products of current production in a particular region; vessels with their place of manufacture inscribed upon them; and various forms of geological evidence. Techniques of characterization can be divided into those concerned with analysing the chemistry of the body, and petrographic analysis, which relies on the mineralogy and other geographically dependent variables provided by the aplastic inclusions of the body (see Chapter 2).

A number of chemical provenance studies have been undertaken on Islamic pottery, all of which were small in scale and therefore of limited significance. One particular problem with chemical characterization involves stonepaste, the siliceous body also known as "quartz-frit," "artificial paste," "fritware," and so on (see Chapter 2). The recipe for stonepaste can be described as including about ten parts quartz, one part glass, and one part of a fine white clay. The chemistry of this body is overwhelmed by technically determined, not geographically determined, variables. Quartz, the principal constituent, is unlikely to contribute a sufficient amount of the required chemical variability between centres; the small amount of rare fine clay may lead to greater amounts of movement of raw materials than is usual; and the glass component, although possibly made using local quartz and plant ash, is dependent on whatever tradition of processing existed for a particular potter, while the local origin of the glass may not be assumed, given the well-recognized trade in glass cullet, and also in glaze materials (Mason et al., 1992). Chemical characterization of this body must therefore be treated with caution. Petrographic analysis of Islamic pottery has largely been restricted to the work of the author, preparatory to the present study (see bibliography; also Chapter 2).

Problems of how

Although the technology of Islamic pottery has received considerable attention, several major problems still remain. Some are key technological questions. For instance, what are the origins of stonepaste? Where and when was tin-glazing introduced? Where and when was underglaze-painting introduced? Apart from the analytical work itself, it is clear from these questions that it is necessary to approach the problems with a large-scale strategy, considering wares from all the key Islamic ceramic production regions. Another strategic consideration in these technological studies is dating. Clearly from the questions asked, it will be necessary to ensure that the samples are typical and diagnostic of their type, and that the typology has effective dating. The lack of such strategic considerations has been the major drawback of previous analytical works on Islamic pottery. Such works have been characterized by a narrow focus on particular types, and by using samples that are not necessarily diagnostic of datable styles of those types. This focus of earlier research on particular types has left large areas and periods of production unrepresented by any analytical data at all. Hence, a simple methodical classificatory study, providing analysis of significant types in all the important regions, would also be useful.

Problems of when

As mentioned above, a problem with the previous analytical research has been the scarcity of well-dated samples. This problem is not merely a feature of the analytical research, but is fundamental to the study of Islamic ceramics generally.

It may be argued that a major difference between art historians and archaeologists is that art historians are concerned with the exceptional, while archaeologists are concerned with the everyday. Julian Steward, in his major theoretical study of cultural change (Steward 1955), said that only regularities are fit objects for archaeological investigation. Nowadays, many archaeologists, myself among them, would consider the exceptional to be worthy of study, as it is through the study of such artifacts that we may obtain greater insights into the human behaviour of the past. However, for archaeologists it is perhaps the study of regularity that puts the exceptional into context. Art-history books on Islamic pottery are filled with the exceptional, the unusual. To begin the development of a broader classification, it is necessary to look for the regularities, to find what was the main product of a particular type, and to use this as the sequence by which the exceptional may be put into context.

By applying the methodologies of archaeological ceramicists, including formal analysis, seriation, and study of motif assemblages, it should be possible to develop a chronology for the fine Islamic wares. As no single archaeological site excavated to date has had the full range of these wares, no archaeologist working on the material from his or her site has had the opportunity to attempt such a study.

SCOPE

To solve all of the above stated problems for all the pottery across the entire Islamic world would be a tremendous undertaking, a life's work. From the point of view of technological interest, it is clear that the key materials are represented by the very highest status glazed wares. From a similar regard, the need for provenance studies is particularly urgent for these high-quality wares, as they are so widely distributed. If used to constrain the study, such considerations would focus our attention on Iraq, Egypt, Iran, and Syria. The chronological range should begin at the beginning of the Islamic era, as many technological questions must be considered in relationship to pre-Islamic practice. A commonly recognized division of Islamic pottery into those wares made before the onslaught of the Mongols in the thirteenth century, and the predominantly "blue and white" style of later wares may also be applied here. Still, it will be necessary to restrict full investigation to the technologically and art-historically important wares. Although the remainder of the wares produced in these regions may be technologically simple, the work that would be required in characterizing the numerous centres both technically and typologically would be formidable. Hence, the reality of the situation as revealed by this study is that production of the technologically and art- historically important wares represents quite restricted timespans. These include for Iraq c. 700-975, for Egypt c. 975- 1175, for Syria c. 1075-1250, and for Iran c. 1100-1340. This is not to say that other wares will not be studied, as assertions referring to "simple technology" and "limited distribution" must be proven.

OBJECTIVES

The objectives can be divided into three main areas: typology/ chronology, technology, and provenance/ characterization. First a corpus will be assembled of all available and relevant pieces, including those that have been adequately published. A ceramic typology will be developed from this corpus, which will be subjected to seriation and referred to available dating evidence. Diagnostic styles of each type of interest will be subjected to analysis with a scanning electron microscope (SEM) with attached analytical facilities. This will be aimed at determining body microstructure, glaze chemistry, and, where pertinent, pigment chemistry. Similarly, typologically diagnostic samples will be subjected to petrographic analysis to provide provenance-specific characterization, and, where possible, to assign a specific provenance. In each chapter a synthesis of all these findings will be made, to examine the interplay of provenance, technology, and dating. Finally, in two later chapters, we will consider the results of this study, first in the perspective of contemporary events, and then in the form of a general conclusion. But first the full methodology of this study will be outlined.

Through the following pursuit of these objectives a governing principle will be that the procedure and results be archaeologically relevant and scientifically sound. Archaeological science must above all be methodologically sound science; otherwise, any result or conclusion derived from the results is suspect if not automatically worthless. However, it is also essential that the research as a whole be relevant to the issues of archaeological enquiry. At times it may appear that the parallel stress on archaeological relevance leads to dilution of scientific rigour. For instance this may at times occur in relationship to numbers of samples subjected to analysis. In earlier analytical studies of the types of ceramics studied here, large numbers of samples have been subjected to analysis, with the results being averaged and so reported. Although large sample sizes and averaged results are one of the cornerstones of good science, by averaging the products from several different centres made over two hundred years any distinction between centres or development through time is lost. In the present examination, samples are chosen for their diagnostic form and decoration so that they may be put in chronological context, and the SEM study is based on material that has already been characterized or provenanced by petrographic analysis. Availability of such samples may be restricted, so sample sizes may be small, but it is archaeologically more meaningful to sample this restricted number than to add another hundred samples that could not be securely placed in their context. Other compromises between scientific rigour and archaeological relevance may occur through the study, and will be pointed out.


Return to Table of Contents