xiii. The Early Islamic Period, 7th-11th Centuries
Early Islamic pottery has been found in two main regions of Persia: Ḵūzestān and the Persian Gulf, on one hand, and the Persian plateau, including Khorasan, on the other. For most of this century, however, study of all Islamic pottery of the first four hundred years has been dominated by the finds from Sāmarrā in Mesopotamia, which had been the capital of the ʿAbbasid caliphate from 221/836, when it was founded by al-Moʿtaṣem, until 269/882, when it was abandoned by al-Moʿtamed. In 1911-13 a German expedition led by Ernst Herzfeld and Friedrich Sarre conducted excavations at the site, where large amounts of pottery were found. In 1925 Sarre published a monograph on this pottery, which included, in addition to a wide range of Islamic wares, some Chinese and supposedly Chinese material: white porcelains, green ware, and wares striped or mottled with green and brown. Originally almost all were thought to date from the brief period during which the city was the capital. As a result, scholars concluded that Islamic potters had developed a new range of table wares in the 3rd/9th century as a direct response to the introduction of Chinese ceramics into western Asia (Lane, pp. 10-16). In recent decades, however, excavations of similar wares at Susa in Ḵūzestān and Sīrāf on the Persian Gulf, as well as at a number of sites in Khorasan, Afghanistan, and Transoxania, have forced a reconsideration of the accepted chronological framework based on the finds from Sāmarrā.
The Sāmarrā ceramics. Five main categories of Islamic ceramics were found at Sāmarrā.
1. Monochrome green-glazed ware is characterized by a yellow fabric; a translucent bluish-green glaze; and often applied, incised, or gouged ornament. The best-known type is a large jar with barbotine decoration, examples of which are known from Sāmarrā, Susa, Sīrāf, and other sites (cf. Sarre, pl. VI; Lane, pl. 3; Whitehouse, 1979, pl. IIIa). Pottery of this general kind was already widely used in the Sasanian period; at Sīrāf it was still the most common variety of glazed pottery at the beginning of the 3rd/9th century (Whitehouse, 1979; idem, forthcoming).
2. “Splashed” and incised glazed wares are usually of a red fabric coated with white slip and a transparent glaze. The glaze is stained with splashes, stripes, and spots of brown, green, and sometimes purple. These stains may be arranged in patterns, but more frequently they seem haphazard. One of the few datable vessels of this type is a jar from Susa, which contained a hoard of coins, the latest of which had been struck in 344/955-56 (Allotte de la Fuye, pp. 6-17, correcting Koechlin, p. 76). Although this ware is often labeled sgraffiato (incised) in the literature, only a small proportion of the splashed pottery has incised ornament, generally consisting of simple, often hastily executed leaf scrolls and half-palmettes. Splashed and incised wares have been found all over Islamic western Asia, from Syria to Khorasan (Lane, pp. 25-27); in Egypt; and in the former Byzantine empire (Allan, pp. 18-20, pl. 3).
3. Glazed relief ware includes plates, cups, and other small objects pressed in molds and covered with monochrome lead glazes, sometimes enlivened with metallic luster (see below). The molded decoration includes geometric designs composed of interlace bands and half-palmettes with a distinct Sasanian flavor, sometimes accompanied by Kufic inscriptions (cf. calligraphy). The discovery of 2nd/8th-century molded ware at Fosṭāṭ (Old Cairo) in Egypt supports the view that it was first made there (Scanlon, p. 104). Furthermore, one inscribed “condiment dish” in this ware attests that potters moved from one region to another, perhaps bringing new styles or techniques; the dish is signed “. . . made by Abū Naṣr al-Baṣrī [i.e., from Baṣra] in Meṣr [Egypt]” (Lane, pl. 4E). This conclusion is supported by Yaʿqūbī’s report in 278/891 (Boldān, p. 264) of a transfer of “makers of pottery (ḵazaf)” from Baṣra and Kūfa to Sāmarrā.
4. Opaque white-glazed ware was usually coated with a glaze to which tin oxide had been added. Well-preserved examples have a glossy surface that recalls white porcelain. The shapes, too, often resemble those of Chinese vessels, and it is widely believed that this glaze was developed in imitation of imported porcelain. Whether or not this hypothesis is correct, Islamic potters were not content simply to produce plain white pottery, and many vessels were ornamented with blue, green, purple, and yellow, either singly or in various combinations or with luster. Among the more common motifs are half-palmettes, palm trees, stripes and splashes, geometric designs, and Kufic inscriptions. Provisional analysis of the finds from Sīrāf suggests that at that site plain white vessels appeared first, followed by those painted in blue, green, or both and then by those with polychrome ornament and luster (Whitehouse, 1979, pp. 51-56.)
5. Lusterware. Luster is produced by fixing a thin film of metallic oxide on the surface of a ceramic object that has already been glazed and fired. After much controversy, it now seems likely that this technique was invented in Egypt by glassmakers. The earliest apparently datable piece of lustered glass is a cup from Fosṭāṭ, now in the Museum of Islamic Art, Cairo, which bears the name ʿAbd-al-Ṣamad b. ʿAlī, presumably the same man who governed Egypt for one month in 155/771-72 and died in 185/801 (Pinder-Wilson and Scanlon, pp. 28-29 figs. 40-42). Subsequently potters learned to apply luster to both lead-glazed relief ware and objects with opaque white glazes. The technique essentially involved combining a metallic oxide (e.g., silver oxide, which produces a yellow stain, or copper oxide, which produces a gold color) with sulphur and an earthy vehicle like ocher; grinding the mixture fine and storing it in vinegar, wine lees, or urine to dissolve the oxide; and painting the resulting liquid on the object, which was then fired a second time, at a lower temperature and in a “reducing atmosphere,” that is, with little oxygen. After the vessel was removed from the kiln, the ocher was rubbed off, leaving a lustrous stain. Several sites in Persia and elsewhere have yielded early Islamic lusterwares painted in different styles and in combinations of yellow, golden brown, ruby, and olive green (e.g., Kervran, pp. 84-85; Whitehouse, 1979, pl. IIIb).
The problem of traditional ceramic chronology. At Sāmarrā the finds included lustered wall tiles from the palace of Jawsaq al-Ḵāqānī, al-Moʿtaṣem’s residence. The ornament includes several familiar elements: half-palmettes, Sasanian wing motifs, and leaf scrolls. Some of the tiles are painted with birds encircled by wreaths. A second, larger group of luster-painted tiles, set into the frame of the meḥrāb (niche) at the Great Mosque of Qayrawān in Tunisia, has much in common with the finds from Sāmarrā. On the basis of a reference by the 9th/15th-century writer Ebn Nājī (d. 837/1433; II, p. 97; cf. Marçais, pp. 9-10), quoted from a lost work by ʿAtīq b. Ḵalaf Tojībī (d. 422/1031), they are usually believed to have been imported from Iraq for the renovations of the Aghlabid Abū Ebrāhīm Aḥmad, which were completed in 248/862-63 (Creswell, p. 314). Taking these two groups of tiles as his starting point, Ernst Kühnel proposed a hypothetical development of luster ceramics in Iraq: The earliest pieces were ornamented in polychrome; in about 246/860 a bichrome palette composed of brown and yellow came into use; and soon after the abandonment of Sāmarrā as capital monochrome luster was introduced. The tiles from the palace of Jawsaq al-Ḵāqānī were not found in place, however, and it is therefore not certain that they formed part of the original decoration. The reports about the Qayrawān tiles also leave room for doubt about the accepted dating (Hansman, pp. 145-46).
The conclusion that new wares were developed in the Islamic world in the 3rd/9th century as a result of the importation of ceramics from China was based partly on the assumption that Sāmarrā was occupied for only fifty years. Yet, although Sāmarrā ceased to be the capital in 279/892, silver coins continued to be minted there until 341/952-53 (Miles). Furthermore, according to Ebn Ḥawqal, who probably visited the area in ca. 358/969 (pp. 243-44, 247; tr. Kramers, pp. 236, 239) and Maqdesī (Moqaddasī, pp. 122-23), who wrote in about 375/985, parts of it were still inhabited. As the excavations of 1911-13 were conducted without regard for stratigraphy, all that can properly be said about an object from the site is that it may date from 221-375/836-985, but it may be even later. On the basis of the Sāmarrā finds alone there is thus no way of knowing whether new types were introduced all at once or at intervals over a period of a century and a half; for further information, it is necessary to turn to related finds from Susa, Sīrāf, and other sites.
Until recently knowledge of early Chinese imports to western Asia was restricted to literary references and to the fragments found at Sāmarrā, supplemented by finds from Susa, Fosṭāṭ, and Nīšāpūr in Khorasan (see čīnī). According to the Ghaznavid historian Bayhaqī (writing in 451/1059; ed. Fayyāż, p. 538), Chinese ceramics were known in the Islamic world as early as the reign of Hārūn al-Rašīd (r. 170-93/786-809); ʿAlī b. ʿĪsā, whom he appointed governor of Khorasan in 180/796, sent him a gift of “two hundred pieces of Imperial porcelain, including basins and bowls and other things the like of which had never been seen before at a royal (pādšāhī) court, and two thousand other Chinese ceramic vessels, including covered dishes, large bowls, and large and small pottery jars.” Ṯaʿālebī (d. 429/1038; pp. 220-21; tr. p. 141) and his contemporary Bīrūnī (p. 226) distinguished two types of Chinese ceramics: the “apricot-colored” and the “cream-colored.” If the apricot in question is the green Damascus apricot, then Ṯaʿālebī may have been describing Chinese green (“Yüeh”) stoneware and white porcelain. Although little, if any, known early Islamic pottery seems to have been made in imitation of Chinese green ware, it has been assumed that Islamic white-glazed pottery was patterned after white porcelain and that splashed or mottled vessels were based on Chinese san- ts’ai (sancai, lit. “three-color”) earthenware. Tang (618-907 C.E.) splashed wares were predominantly used for tomb furniture, however, and are thus unlikely to have been exported in significant quantities to western Asia. Furthermore, finds from dated tombs show that san-ts’ai wares ceased to be made in the mid-2nd/8th century; much of the so-called “Chinese” mottled pottery from Sāmarrā has no parallels from China itself. At present, therefore, it appears that the similarity between Chinese and Islamic mottled wares may be largely fortuitous. In the white-glazed pottery, however, Chinese influence is unmistakable. Although Robert Adams raised the possibility that white glazes were a local development in southern Iraq (p. 110), it is obvious that a large number of vessels from Sāmarrā and elsewhere are imitations of Chinese bowls with radial ridges terminating in indentations on the rim. There are two possible explanations: Either opaque white glazes were already being made in Iraq before Chinese white ware was first known and copied there, or, more probably, opaque white-glazed ware was developed in imitation of Chinese imports.
Evidence for ceramic chronology in southern Persia. The most detailed information on the chronology of early Islamic pottery in Ḵūzestān and the Persian Gulf comes from excavations at Susa and Sīrāf. In the most recent study of material from Susa Monique Kervran described a stratified sequence from the “Apadana tepe.” She divided the Islamic sequence into three “levels,” labeled (from early to late) III-I. In level III most of the pottery resembled earlier, Sasanian material. It included both a fine unglazed ware made from pale-yellow clay and decorated with meticulously incised and molded ornament and a ware with monochrome green glaze. Level I, on the other hand, produced the full range of pottery associated with the “Sāmarrā horizon.” Kervran dated level III to the period ca. 29-132/650-750, level II to ca. 132-84/750-800 +, and level I to the “caliphal” occupation at Sāmarrā (221-79/836-92). Although there is no reason to question the sequence at Susa, the chronology is open to doubt, in the light of the stratigraphic sequence at Sīrāf.
As the leading entrepôt in the Persian Gulf from ca. 184/800 until the 5th/11th century, Sīrāf handled much of the eastern merchandise going to Iraq and Persia. The platform of the principal mosque, built ca. 184-94/800-10, contained pieces of Chinese stoneware storage jars and stoneware bowls with underglaze-painted ornament. White ware was completely absent, however. Later deposits, probably of the 3rd/9th century, yielded many Chinese cream or white stoneware bowls with ridges and indentations on the rim, the prototypes of Islamic white-glazed bowls; other types included green-ware ewers and mottled sherds of uncertain origin.
The Islamic material corresponding to that from level I in the Apadana tepe at Susa comes from Sīrāf period II, which may be subdivided stratigraphically as follows: IIa, introduction of white-glazed pottery, either plain or decorated only with blue; IIb, introduction of splashed ware (without incised ornament) and of green and brown decoration on white-glazed ware; IIc, introduction of splashed ware with incised ornament and of luster decoration on white-glazed pottery. The first and second phases belong to the 3rd/9th century; the third stage may have begun as late as the early 4th/10th century (Whitehouse, 1979, pp. 56-59). The evidence from Sīrāf thus tends to contradict the view that there was an explosive development of Islamic pottery in the 3rd/9th century. Indeed, the earliest splashed ware with incised decoration and white ware with luster decoration may date from the 4th/10th century. A further development at Sīrāf is worthy of note: the introduction, apparently ca. 416-42/1025-50, of a type of incised pottery with a hard red fabric covered white slip and a shiny transparent glaze (see Whitehouse, 1979, pl. IVb). The very distinctive decoration consists of concentric zones of vegetal and epigraphic motifs. This “style III” incised pottery (as it is called in the publications of Sīrāf) is commonly found throughout the Persian Gulf and along the western coasts of the Indian Ocean.
Evidence for ceramic chronology on the Persian plateau. Just as Sāmarrā long dominated the study of early Islamic pottery in Iraq and adjoining regions, so Nīšāpūr has dominated the study of early Islamic pottery on the Persian plateau. The city flourished in both the Sasanian and early Islamic periods; in 509/1115 and 540/1145-46 it was damaged by earthquakes and shortly thereafter, in 549/1154, was sacked by the Ḡozz Turks. The coup de grace was delivered by the Mongols in 618/1221. A large quantity of early Islamic pottery was excavated at the site by Joseph Upton, Walter Hauser, and Charles K. Wilkinson in 1935-1940 and 1947 and published by Wilkinson in 1973. Even larger quantities of Persian pottery have been attributed to Nīšāpūr (with varying degrees of reliability) by dealers.
Wilkinson classified the pottery from the site in twelve categories: 1) “buff ware,” with decoration in black, yellow, and green slips under a colorless glaze; 2) “color-splashed” ware; 3) pottery decorated in black on white slip under a colorless glaze; 4) pottery decorated in combinations of black, red, green, yellow, and brown on white slip under a colorless glaze; 5) slip-painted ware with colored slip; 6) opaque white ware and its imitations; 7) opaque yellow-glazed ware; 8) “ware with yellow-staining black,” that is, decorated in black on white slip under a transparent yellowish glaze; 9) pottery with a monochrome glaze, usually green; 10) Chinese ceramics; 11) alkaline-glazed ware; 12) unglazed ware.
The most common varieties of glazed pottery from the site are color-splashed ware (category 2), buff ware, decorated with colored slips under a colorless glaze (category 1), and black-on-white ware (category 3). The buff-ware vessels are covered over their entire surfaces with rich and varied ornament: birds, animals, human figures, palmettes, leaf scrolls, and inscriptions. In contrast the decoration on black-on-white dishes and bowls is smaller, more restrained, and usually restricted to the center and the rim; among the more common motifs are birds (often with leaf sprays in their beaks), leaf scrolls, and inscriptions (sometimes of great elegance).
The types of pottery found at Nīšāpūr also occur over much of the Persian plateau and a large area of Transoxania; comparable wares have been found, for example, at Gorgān, Ray, and Sīrjān (Morgan and Lethaby) in Persia; Mary (Marv) and Afrasiyab (Afrāsīāb) in the Soviet Union (Tashkhodzhaev); and Laškarī Bāzār in Afghanistan (Gardin). Some of this material, particularly color-splashed (category 2) and white-glazed (category 6) types, is also closely related to wares from Iraq, Ḵūzestān, and the Persian Gulf coast. The local white-glazed pottery, however, usually has a poor finish, and the range of colors used for decoration is limited to green and purple or black. Imports from Iraq include lusterware, which was imitated locally by painting with slips under colorless glaze (category 6; Wilkinson, pp. 181-82, pls. 199-204).
The largest collection of excavated material in the southern part of the Persian plateau, from Sīrjān, has much in common with the finds from Nīšāpūr; it includes pottery with painted decoration over white or reddish-brown slip, splashed glazes (sometimes with incised decoration), and monochrome glazes (also sometimes with incised decoration). Neither Nīšāpūr nor Sīrjān produced a well-dated sequence of pottery types, however, and the chronology of early Islamic pottery on the plateau is thus heavily dependent on information from Laškarī Bāzār. Recently Terry Allen (1988, pp. 60-66) has argued that some of the buildings at that site and at nearby Bost are of Samanid (279-395/892-1005), rather than Ghaznavid (387-582/997-1186), origin. If so, some of the Islamic glazed pottery probably dates from the same period, despite Gardin’s conclusion that the sequence did not begin before ca. 391/1000 (pp. 179-204). In particular, Allen believes that three-color slip-painted pottery (Gardin’s group II = Wilkinson’s group 6) decorated in red, olive green, and black, thought to be a local imitation of lusterware, came into use before ca. 287/900 (1989, pp. 60-62; cf. Wilkinson, pp. 181-82). He argues that three-color lusterware “appears to have gone out of production well before 1000” and concludes from the stylistic unity of Mesopotamian polychrome luster that it may not have continued much beyond the end of the 3rd/9th century; the earliest examples of imitation luster on the plateau should not then be datable later than ca. 287/900. This chronology finds some support in a survey of the coins from Nīšāpūr (Allen, p. 61; cf. Wilkinson, pp. xxx-xl) and in Lisa Volov’s work on slip ware from Khorasan, in which she has concluded on epigraphic grounds that some of it is datable to the 3rd/9th century.
Unglazed ceramics. Despite the attention devoted to the various categories of early Islamic glazed pottery in Persia, the most common single ware was unglazed and frequently without decoration, made for everyday use. Such ceramics account for more than 90 percent of all 4th/10th-century pottery found at Sīrāf and 80 percent from Sīrjān. In the past, such wares were neglected by archeologists and collectors; more recently they have begun to receive the scientific attention they deserve, and information on local production is now available from Susa, Nīšāpūr, and Sīrjān. However, it is in apparent conflict with the suggestion made above that the earliest luster ware may be somewhat later than is generally supposed.
R. M. Adams, “Tell Abū Sarīfa. A Sassanian-Islamic Ceramic Sequence from South Central Iraq,” Ars Orientalis 8, 1970, pp. 87-119.
J. W. Allan, “Incised Wares of Iran and Anatolia in the 11th and 12th Centuries,” Keramos 64, 1974, pp. 15-22.
T. Allen, “Notes on Bust,” Iran 26, 1988, pp. 55-68.
Idem, “Notes on Bust (Continued),” Iran 27, 1989, pp. 57-66.
F. M. Allotte de la Fuye, Inventaire des monnaies trouvées à Suse, MMAP 20, Paris, 1928.
M. R. Ayalon, La poterie islamique, MDAFI 30, Paris, 1974.
Abū Rayḥān Moḥammad Bīrūnī, Ketāb al-jamāher fī maʿrefat al-jawāher, Beirut, n.d.
Qāsem b. and Īsā Ebn Nājī, Maʿālem al-īmān, Tunis, 1320/1902.
K. A. C. Creswell, Early Muslim Architecture II, Oxford, 1940.
J.-C. Gardin, Lashkari Bazar II: Les trouvailles. Céramiques et monnaies de Lashkari Bazar et de Bust, MDAFA 18, Paris, 1967.
Idem, “La céramique islamique,” in R. Boucharlat and O. Lecompte, Fouilles de Tureng Tepe sous la direction de Jean Deshayes I: Les périodes sassanides et islamiques, Paris, 1987, pp. 121-70.
J. F. Hansman, “Dating Evidence for the Earliest Islamic Lustre Pottery,” Annali dell’Istituto Orientale di Napoli 42 (N.S. 32), 1982, pp. 141-47.
M. Kervran, “Les niveaux islamiques du secteur oriental du tépé de l’Apadana II. Le matériel céramique,” CDAFI 7, 1977, pp. 73-165.
R. Koechlin, Les céramiques musulmanes de Suse au Musée du Louvre, Mémoires de la Mission Archéologique en Perse 19, Paris, 1928.
E. Kühnel, “Die ʿAbbāsidischen Lüsterfayencen,” Ars Islamica 1, 1934, pp. 149-59.
A. Lane, Early Islamic Pottery, London, 1947.
G. Marçais, Les faïences à reflets métalliques de la grande mosquée de Kairawān, Paris, 1928.
G. C. Miles, “The Sāmarrā Mint,” Ars Orientalis 1, 1954, pp. 187-91.
P. Morgan and J. Lethaby, “Excavated Ceramics from Sīrjān,” in J. Allan and C. Robert, eds., Syria and Iran. Three Studies in Medieval Ceramics, Oxford, 1987, pp. 231-72.
A. Northedge and R. Falkner, “The 1986 Survey Season at Sāmarrā,” Iraq 49, 1987, pp. 143-73.
H. Philon, Early Islamic Ceramics, Ninth to Late Twelfth Century, London, 1980.
R. Pinder-Wilson and G. T. Scanlon, “Glass Finds from Fustat,” Journal of Glass Studies 15, 1973, pp. 12-30.
F. Sarre, Die Keramik von Samarra, Die Ausgrabungen von Samarra 2, Berlin, 1925.
G. Scanlon, “Fusṭāṭ Expedition: Preliminary Report 1965, Part I,” Journal of the American Research Center in Egypt 5, 1966, pp. 83-112.
Abū Manṣūr ʿAbd-al-Malek b. Moḥammad Ṯaʿālebī Nīšābūrī, Laṭāʾef al-maʿāref, ed. E. Abyārī and H. K. Ṣayrafī, Cairo, 1960; tr. C. E. Bosworth, The Book of Curious and Entertaining Information. The LaṭÂ¡āʾif al-maʿārif of Thaʿālibī, Edinburgh, 1968.
S. S. Tashkhodzhaev, Keramika Samarkanda, Tashkent, 1967.
L. Volov, “Plaited Kufic on Samanid Epigraphic Pottery,” Ars Orientalis 6, 1966, pp. 107-34.
D. Whitcomb, “Khirbet al-Mafjar Reconsidered. The Ceramic Evidence,” Bulletin of the American Schools of Oriental Research 271, 1988, pp. 51-67.
D. Whitehouse, “Islamic Glazed Pottery in Iraq and the Persian Gulf. The Ninth and Tenth Centuries,” Annali dell’Istituto Orientale di Napoli 39 (n.s. 29), 1979, pp. 45-61.
Idem, Siraf XI: Sasanian and Islamic Glazed Pottery, London (forthcoming).
C. K. Wilkinson, Nishapur. Pottery of the Early Islamic Period, New York, 1973.
A. Williamson, “Regional Distribution of Mediaeval Persian Pottery in the Light of Recent Investigations,” in J. Allan and C. Roberts, eds., Syria and Iran. Three Studies in Medieval Ceramics, Oxford, 1987, pp. 11-22.
Originally Published: December 15, 1991
Last Updated: October 11, 2011
This article is available in print.
Vol. V, Fasc. 3, pp. 308-311