The tenth century AD was a period of political, economic and religious change in western Asia as the Buyid amirs, of Shia allegiance, expanded their control from Persia and imposed rule over Baghdad in 945. These changes, along with the destruction of the port city of Sîrâf by an earthquake in 977, restructured exchange networks in the Persian Gulf and the western Indian Ocean (Beaujard Reference Beaujard2012). The Fatimids had considerable influence over Yemen, via the Sulayhid dynasty, and, from 1083, over Aden via another Ismaili dynasty, the Zurayids. Merchants from Egypt and Yemen, often Jews, were active in the Red Sea and between the Red Sea and India (Goitein & Friedman Reference Goitein and Friedman2007). It was within this context that the port of Sharma was created ex nihilo in southern Arabia, probably by Sîrafîs. Sharma was briefly mentioned by al-Muqaddasî around 985, and later by al-Idrîsî around 1150.

Located 50km east of al-Shîr on the Hadramawt coast of Yemen, the site of Sharma was discovered in 1996. The publication under review, edited by Axelle Rougeulle, provides an account of the research conducted at the site by French teams from the CNRS between 2001 and 2005.

The site of a seasonal settlement at the start of the Christian era, Sharma was fully occupied by the end of the tenth century. Excavations have shown that it was “a transit entrepôt, a cluster of warehouses [. . .] entirely devoted to the maritime trade” (p. 5). Sharma flourished until the middle of the twelfth century. It was then abandoned and was only briefly reused toward the end of the thirteenth and into the early fourteenth century. Sharma was thus active for only a relatively short time.

Various quarters of the site have been identified (Chapter II). The town included a mosque, and one building may have been the house of the chief merchant. The site's stratigraphy and assemblages have led researchers to distinguish six phases of occupation. Phase III was marked by the building of a defensive wall, dated to the mid eleventh century (later dismantled). The construction of this wall may be linked to Sulayhid expeditions along the Hadramawt coast c. 1063, a period that also corresponds to the fall of the Buyids in Baghdad. Phase IV (second half of the eleventh to early twelfth centuries) saw a major development of the entrepôt, after the Seljuks took control of western Asia—by ousting the Buyids—and looted Suhâr (Oman) in 1053. A new wall was later erected, although Sharma fell into decline during Phase V (the second half of the twelfth century). Sharma's decline and later abandonment may have been linked to the aggressive policy of Qays, located in the Persian Gulf, or due to the rise of Shihr and Mirbât, on the southern Arabian coast, or perhaps to an Ayûbbid intervention, conducted from Aden (discussed by Rougeulle et al. in Section III/2). Defensive structures have been found, including forts; some structures may go back to the South Arabian period, around the end of the first millennium BC.

The site has yielded various finds, in particular remnants of fragile cargoes. Some ceramics were locally produced, but most (70 per cent) were imported, coming from all parts of the Indian Ocean, extending from China to East Africa. Glazed ceramics from Iran indicate the origin of Sharma's founders and occupiers. Sassano-Islamic pottery and sgraffiatos from the Persian Gulf have been excavated, along with ceramics from Sind, southern India and north-western India, suggesting that Indian merchants were present at the port. The earliest Chinese ceramics, presented by Zhao Bing in Section III/4, go back to the third quarter of the tenth century and are present in Phase 1. The proportion of Chinese ceramics (4 per cent) is higher here than at other sites (Chinese pottery is rare at al-Shihr, and it is obvious that Sharma and al-Shihr imported foreign products independently, as discussed by Hardy-Guilbert in Section V/1). These ceramics include various types of stonewares and porcelains. East Africa accounts for 16.2 per cent of non-glazed ceramics, revealing contacts with this region and probably with the Comoros (it is likely that slaves were imported). These contacts have been confirmed by finds of copal, preponderant among the resins identified at Sharma (described by Regert et al. in Section III/9).

Glass finds, presented by Foy in Section III/5, are relatively abundant. They include vessels traded both as containers and as commodities in their own right. Some forms appear to have come from Iran, while a few flasks with marbled decoration and a single goblet may have been imported from Egypt or Syria.

Discoveries of mustard ware—produced in the region of Zabîd or Aden—reveal a brief reoccupation of the site during the thirteenth or fourteenth century. A few sherds of Chinese stoneware date to this period (Section IV/1). The site was again reoccupied during the eighteenth and nineteenth centuries. Sharma has also yielded archaeobotanical data, discussed by Dabrowski et al. in Section III/10, especially cereals, which were probably imported: wheat, rice and sorghum.

In addition to work at the port, research conducted inland of Sharma has led to the discovery of sites dating to various periods (Section V/1). Yadghat, where pottery was produced, dates back to the Sharma period; some imported ceramics have also been unearthed. Vallet (Section V/2) discusses how the development of Sharma points to alliances with the tribal groups who controlled the region.

Prior to the CNRS fieldwork, the place of southern Arabia within international trade networks of the early second millennium AD was poorly understood. The excavations at Sharma have shed light on the roles played in the Indian Ocean by southern Arabia and the Persian Gulf during the first centuries of the second millennium; in contrast, Sharma shows few links with the Red Sea. The results inform us about the types of merchandise traded—at least those leaving traces in the archaeological record—and reveal links with East Africa, including the probable presence of Swahilis (free and enslaved), and perhaps of Comorians and even of Malagasy (some of the copal found at Sharma may have come from Madagascar). Data also confirm Iranian influences along the East African coast at this time (see Horton & Middleton Reference Horton and Middleton2000: 56).

This detailed and high-quality publication illuminates the roles played by southern Arabia and the Persian Gulf in wider Indian Ocean trade networks during the early second millennium AD, and is of outstanding value for an understanding of trade and wider developments across the region at that time.