top of page

Excavating Zabid

Royal Ontario Museum Yemen Project Overview

Ceramic Typology

Edward Keall
Curator Emeritus
Royal Ontario Museum

Circles of buried pots in situ at Fazzah in 1983,

but now completely eroded away by abrasive winds

Ceramic Typology

Ceramics – in the form of broken, discarded bits of pottery (potsherds) form one of the most important ‘fossils’ in helping attribute chronological dates to the stratigraphic layers of an archaeological site where they are found. By far the most useful of indicators of urban life are the ubiquitous bits of broken pottery one can find in excavations and exposed on ground surfaces. The first steps were made in 1982 towards establishing an effective ceramic typology for the study area. Ground surface prospections are known in archaeology as ‘ground truthing’.


A minute percentage of the pottery recovered in the Canadian Mission survey is of foreign origin, predominantly from Egypt, Iraq, Iran, China and Southeast Asia, ranging from the ninth century to the present. Some fragments of pottery from known foreign cultural records – such as lustreware’ from 9th century Abbasid Samarra – were picked up. These are fossil indicators for the locally produced products, which represented the overwhelming majority of the potsherds found. Large-scale production of mostly red-firing pottery, often with simple wavy-line grooved decoration was identified at a 9th century kiln site just east of Zabid. In the early stage of the Project, these decorative designs were unknown to scholarship.


Of the other mass of potsherds picked-up from the surface all over the site, many were totally plain, but others had intricate scratched and punctate surface designs; other pieces were finished with coloured glazes. Through association with the pottery of better-known foreign origin, an effective working typology for the locally produced wares was gradually devised. Sequences of pottery found consistently in trench-work also added to the refinement of this project ceramic typology.


Initially, Instrument Neutron Activation Analysis (INAA) contributed vital clues to the distinction between locally produced and imported pottery. However, the analytical technique that proved to be the most effective in defining local production – and identifying therefore imported fabrics – involved Ceramic Petrography (thin-section analysis). This technique uses polarizing light to characterize mineral particles present in the clays. These provide definitive fingerprints of the clay types. When clay sources of that variety are sourced locally, local production can be affirmed. Through this, and the excavation of kiln products and kiln furniture, it was possible to build up a picture of pottery production in the Zabid region.


One glazed type that had a body not made of the normal iron-rich red clay, but a white ‘stone-paste’ fabric, had a particularly distinctive decorative design. What were the source of this distinctive pottery type? South of Zabid, in the town of Hays, there is still a potter who makes some glazed mixing bowls and items to sell to tourists, and for local kitchen use. Outside of his workshop – in a cemetery opposite his shop – one can find remnants of his ancestors’ work – discarded, used kiln furniture from 16th century production work. This gives us a definitive analytical profile of the mysterious white-bodied wares. For, in the hills behind Hays, there are rock outcrops of granite, potentially producing kaolin – the end-product of decayed granite. Strangely, the potters of 17th century Hays knew of the white clay, but didn’t know of the high temperatures needed to produce porcelain.


An overview of the Zabid pottery typology was published in Ciuk and Keall. It should be acknowledged here that slight revisions need to be made to the dates assigned to some of the pottery types in that monograph, as a result of excavations in wider areas at depth than before, along with the discoveries of other scholars elsewhere in Yemen. But the scheme remains basically solid, and can still be used for a general chronological control of the Zabid pottery corpus.


With a rough chronological control over when the different types were produced, it was possible to start considering the social context in which the vessels were used. A primary feature in the research became what implications were meant by the presence of fine Haysi glazed coffee cups in the Ottoman fort layer, along with dry-stem smoking pipes and the finials for small water-pipes. The findings spurred the quest to study the origins of coffee drinking, and the question of what substances were being smoked in the pipes.


A typical abraded site surface in the Tihamah, with an abundance of potsherds visible


Harvest of diagnostic potsherds on a mat, from the 1982 season of filed reconnaissance


Composite image of potsherds from vessels, and a kiln trivet,

made in the Tihamah (9th to 19th century)


Composite image of potsherds from vessels imported from the Far East (12th to 18th century)

bottom of page