Cover Page
The handle http://hdl.handle.net/1887/55980 holds various files of this Leiden University dissertation.
Author: Bagci, Y.
Title: Coloured Ceramics of the Caliphs: A new look at the Abbasid pottery finds from the old Gözlükule excavations at Tarsus
Issue Date: 2017-11-29
CHAPTER 4
ABBASID CERAMICS: THE CURRENT STATE OF KNOWLEDGE
After having set the historical context of Tarsus in the previous chapter, this one shifts its focus to the current state of the arts on ceramics of the Abbasid period. The relations of the Gözlükule ceramics with similar ones are examined to understand the Tarsus pottery evidence in the broader context. What are the characteristics and chronologies of the pottery classes of Abbasid times and how far were these ceramics distributed? In order to answer these questions, major Abbasid settlements offering well-stratified pottery finds or assemblages associated with this period are presented and discussed in relation to their ceramic material. These include sites in the Near East and Egypt as well as in East Africa. The overview of Abbasid sites, which follows in the next pages, aims to sum up the actual stage of research about pottery of the Abbasid period because a similar comprehensive survey has not yet been put together. In the second part of this chapter, major ceramic classes of the 9th - 10th century including glazed and unglazed pottery are discussed. For each class, a history of the production, information on the production centres and distribution patterns, a chronological range of the production and the morphological repertoire are offered by discussing and comparing the pottery evidence of different sites and areas.
I/ Overview of the main sites of the Abbasid period
Before investigating the relationships of the Gözlükule ceramic material with similar ones, a survey of some important archaeological projects of the early Islamic period is offered below. These excavations and surveys produced key information on the characterisation of the pottery repertoire of the Abbasid period in the Middle East and other regions. This brief presentation follows a geographical survey on a West-East axis, starting from Turkey and going eastwards to Iran and then surveying the main Abbasid sites in the Levant, Egypt and Kenya. The different archaeological excavations and surveys mentioned here are limited to those carried out after the 1970s (Fig. 4.1.1).
Sites in Turkey (Tab. 4.1.1-2a)
Apart from the excavations on the mound of Gözlükule, several other excavations were conducted in Tarsus including rescue excavations. As these were discussed in the previous chapter, these different projects in Tarsus are not mentioned here.
The site of Hisn al-Tinat (“Castle of Figs” in Arabic) was a medieval fortified settlement adjacent to the mound of Kinet Höyük which lies in modern day Dörtyol, in the Hatay (Antiochene) province in southern Turkey (Eger 2010, 52). This small military structure was excavated between 2006 and 2011 by Asa Eger and his team.
The settlement was constructed in three phases dating from the 8th to the 12th centuries when the site was abandoned. Eger ranged the chronological sequence from the middle of the 8th century to the 9th century (early Islamic), the 10th century (early Islamic/ middle Byzantine) and 11th - 12th centuries C.E. (middle Byzantine) (Eger 2010, 58). The excavations of the small fortress yielded an important amount of ceramics of the Abbasid period which have not yet been fully published. This pottery assemblage included glazed and unglazed ceramics such as Polychrome Painted Glazed Ware, Splash Glazed Ware, few Iraqi imports, Unglazed Buff Wares and examples of Brittle Ware (Eger 2010, 65).
Tilbeshar is situated 20 km south-east of the city of Gaziantep in southeast Turkey. The site which was excavated by Rıfat Ergeç and Christine Kepinski- Lecomte was occupied from the Bronze Age through the Ayyubid period (c. 2400- 2300 B.C.E and 12th - 13th C.E.) (Kepinski-Lecomte and Ergeç 1999, 245). The ceramic material came from surveys around the mound and spanned in date between the Hellenistic and Ayyubid periods (Rousset 1998, 177). The body of ceramics of the Abbasid period included, glazed fine wares: White Opaque Glazed Ware with green splash belonging to the “Samarra Horizon Ceramic Repertoire”, Polychrome Painted Glazed Ware, Splash Glazed Ware, Polychrome Sgraffito Ware; unglazed common wares: Unglazed Buff Ware, Unglazed Moulded Ware; cooking ware: Brittle Ware.
These ceramics were dated by typological comparanda due to the lack of stratigraphic context. Marie-Odile Rousset proposed for some ceramic classes a dating that may be too early (e.g., Polychrome Painted Glazed Ware, Splash Glazed Ware) (Rousset 1998, 173). Rousset suggested an interruption of the settlement between the Roman and Islamic periods on the basis of the absence of Byzantine forms in the pottery corpus (Rousset 1998, 177).
Zeugma is close to the modern city of Gaziantep in southeast Turkey. Given the considerable expansion of the city in the Roman period, several archaeological projects focusing essentially on the Greco-Roman levels were undertaken on different locations in and around the city. Among various international projects such as the rescue excavations by the Turkish team directed by Rıfat Ergeç and Mehmet Önal, the French team supervised by Catherine Abadie-Reynal and the British team directed by Rob Early, very few investigated on the post-antique strata (Abadie-Reynal and Ergeç, 1997-1998; Aylward 2013; Ergeç and Önal 1998). Despite the lack of a firm stratigraphy, the British pottery specialist Philip Kenrick was invited to work with the team from Oxford and recorded early Islamic pottery in the salvage excavations.
Apparently, the Islamic trench was rather small and was excavated hastily (Kenrick 2013, 66). In his report, Kenrick distinguished the ceramics of the Abbasid period.
The provided dating was occasionally a bit early, because it repeated the “high chronology” of the Abbasid polychrome glazed wares proposed by some publications.
Glazed and unglazed pottery of the Abbasid period included Polychrome Painted Glazed Ware, Splash Glazed Ware, vessels of a light utility character and plain wares referred to as Unglazed Buff Wares, as well as Brittle Ware cooking pots (Kenrick 2013, 66).
Surveys in Turkey
The 2005 Amuq survey (“the Antioch plain”) in southern Turkey near modern Antioch (Antakya) was directed by the Dutch archaeologist Fokke Gerritsen and brought together an international team of archaeologists including Rana Özbal, Tasha Vorderstrasse, Asa Eger and Andrea de Georgi. The project investigated on the continuity and change of the ancient landscape and the settlement patterns on a longue durée, from the Neolithic through the modern periods (Gerritsen et al. 2008). Out of the 287 sites recorded in the survey, 67 of them were assigned to the early Islamic period by Asa Eger (Gerritsen et al. 2008, 266). The quantity of pottery dated to this period was not specified and the presentation of this ceramic data was short. The ceramics which were recorded corresponded to the most common pottery classes of the Abbasid period such as Polychrome Painted Glazed Ware (“Yellow Glazed Ware”), Polychrome Sgraffito Ware and Unglazed Buff Wares and Brittle Ware examples (Gerritsen et al. 2008, app. 1, fig. 16).
Keban Dam projects - The 1967-1975 Keban Dam Rescue Project was as a turning point for archaeology in Turkey (Dissard 2011). It was conducted in eastern Turkey to record and study the history and archaeology of the Keban region which is part of the Elazığ province, before it disappeared under the waters of the dam (Dissard 2011, 4).
These salvage excavations and surveys were carried out over an extensive region scored by the northern section of the Euphrates river. Among the numerous archaeological sites recorded in this project, Kurban Höyük will be our focus.
Kurban Höyük - In addition to the ground-breaking Kurban Höyük excavations of the Chicago Oriental Institute, the 1980-1984 Kurban Höyük survey was carried out in the lower Karababa basin, in the north of modern Şanlıurfa in south-eastern Turkey (Wilkinson 1990). A team of archaeologists, including Tony Wilkinson, investigated the settlement patterns and the land use between the Neolithic and the medieval periods (c. 8000 B.C.E. and 1200 C.E.) (Wilkinson 1990). The surveyed sites produced a limited amount of early Islamic ceramics. In particular, site 6 (near Saskan village) yielded a corpus of ceramics of the Abbasid period. This unglazed and glazed pottery group contained Polychrome Painted Glazed Ware, Splash Glazed Ware, Polychrome Sgraffito Ware, Opaque White Glazed Ware Imitation with green splash and Unglazed Buff Wares and Brittle Ware examples (Wilkinson 1990, 127-128).
Lidar Höyük survey - The site of Lidar Höyük which is located in the Karababa Basin near Samsat and Urfa in south-eastern Turkey belonged to the “Asağı Fırat projesi”
(the under Euphrates project). The excavations on the mound and the survey around it were directed by the German archaeologist Harald Hauptmann between 1979 and 1987 (Gerber 1996, 303). Christoph Gerber published the ceramics. Pottery of the Abbasid period was recorded in small quantities in Incirli and Gazbek and mainly consisted of common wares and cooking wares such as Unglazed Buff Wares and Brittle Ware (Gerber 1996, 306).
Birecik and Carcemish Dam survey - After the construction of the dams of Keban, Karakaya and Atatürk, the 1989 Birecik and Carcemish Dam survey was carried out by Guillermo Algaze, Ray Breuninger and James Knudstadt (Algaze et al. 1994).
The area which was investigated excluded the ancient site of Carcemish, but comprised the region near the modern town of Halfeti (Algaze et al. 1994, 3). The early Islamic/Abbasid occupation in this area was confined to two sites (Zeugma and Tarlabașı) (Algaze et al. 1994, 22). Small amounts of glazed wares such as Splash Glazed Ware and Polychrome Sgraffito Ware as well as sherds of Unglazed Buff Ware were recorded (Algaze et al. 1994, 22).
Sites in Syria (Tab. 4.1.1-2b)
The site of al Hadir-Qinnasrin is located in the south of Aleppo and was first surveyed by an international team including the Oriental Institute of Chicago, the Sorbonne and the General Directorate of Antiquities and Museums of Syria in 1998. In 2004, new excavations were initiated by a Syrian-French team directed by Marie-Odile Rousset and Asad al-Youssef. These new excavations unearthed an extensive early Islamic settlement dated between mid 7th - 10th centuries continuing into the middle Islamic periods (c. 12th century) (Rousset 2012). The ceramic assemblage of al-Hadir which was well-stratified is discussed further down as a reference assemblage to the Gözlükule pottery corpus. In general, the ceramic material of al-Hadir included glazed
and unglazed wares typical of the 9th -10th century Abbasid period (Rousset 2012, 99- 103).
The ancient city of Raqqa is in northern Syria at the intersection between the Euphrates and the Balikh rivers. The area under archaeological prospection included the Hellenistic Nicephorion, the Byzantine city whose walls were restored under Justinian, and al-Rafiqa, the Abbasid city built following the model of Baghdad (Heidemann 2006, 33). Raqqa was first known thanks to the archaeological inspections of Sarre and Herzfeld in the beginning of the century (Sarre and Herzfeld 1911-1920). The city constituted one of the rare archaeological sites excavated with a special focus on the Abbasid period. Due to its political and commercial importance, several archaeological studies focused on the city of al-Rafiqa. Numerous workshops such as pottery and glass ateliers were uncovered in the industrial area situated in the periphery of the city in the site of Tall Aswad and Tall Abu Ali. These discoveries were significant for understanding the industries of the Abbasid period. The excavations revealed potters’ quarters producing unglazed and glazed ceramics corresponding to the broad repertoire of pottery manufactured in the Abbasid period (François and Shaddoud 2013; Miglus and Becker 1999). The Abbasid pottery recorded at Raqqa included Polychrome Painted Glazed Ware, White Opaque Glazed Ware Imitation with turquoise/green splash, Splash Glazed Ware and Polychrome Sgraffito Ware, Unglazed Buff Wares (moulded and wheel-thrown) and Brittle Ware (François and Shaddoud 2013; Gonnella 1999, 55-60; Miglus and Stepniowski 1999, 37-38; Watson 1999a, 81-83).
The site of Rahba-Mayadin was excavated by a Syrian-French team lead by Thierry Bianquis and Qassem Toueir between 1976-1980 (Rousset, 1998, 14). These twin sites in the middle Euphrates region developed particularly in the Islamic period between the 9th and 14th centuries. Rahba preceded Mayadin, although both were occupied during the Abbasid period. The excavations revealed several buildings and a large citadel. The ceramic finds of Rahba constituted a “pre-Samarra” pottery assemblage which in the following phase included the polychrome glazed wares of the “Samarra Horizon Ceramic Repertoire” (Rousset 1996, 356). The ceramic finds from the following periods were recorded, including glazed ceramic production in the 11th century C.E. (Rousset 1996, 358).
Qasr al-Hayr al-Sharqi is situated in semi-arid Syrian steppe in central Syria.
The site was first excavated between 1964 and 1971 by an American team lead by Oleg Grabar and the University of Michigan. The archaeological research was primarily designed to understand the function and the dating of this large-scale fortified site which was generically called “Umayyad desert-castles” (Grabar et al.
1978). After excavations and restorations carried out by a Syrian team, the excavations resumed with Syrian and Swiss efforts under the direction of Denis Genequand and Walid al-As‘ad between 2002 and 2008 (Genequand 2002, 2004 and 2008; Genequand et al. 2005, 2006, 2007). The foundation of the city was dated to 728-729 by an inscription which is now lost (Grabar et al. 1978, 12). The site consists of two main structures including a palace and a larger walled area with living quarters, industrial complexes and a mosque but also baths and two large monuments with a central courtyard (Genequand 2002, 2004 and 2008; Genequand et al. 2005, 2006, 2007; Grabar et al. 1978, 149). As indicated by the pottery evidence, the site was inhabited in the 9th - 10th centuries when it was abandoned and re-occupied between the 12th and 14th centuries. Only a selection of the ceramic material was published, from several houses (house A and B). For the house A, it provided a narrow chronology between the 8th and the last quarter of the 9th century which could
be fixed by fine glazed pottery belonging to the “Samarra Horizon Ceramic Repertoire” (Genequand et al. 2007, 144-147). From the house B, only the Brittle Ware was published and indicated a similar chronological range between 8th and 9th century (Genequand et al. 2010, 210-14). Most of the ceramic repertoire of the Abbasid period of Tarsus-Gözlükule was also found here including the various ranges of polychrome glazed table wares, common wares and cooking wares.
Apamea which is situated on the right bank of the Orontes river in north- western Syria, was a large Roman city with continuous occupation into the medieval period. The Belgian excavations were intermittently conducted between the 1930s and 2010 and created the Belgian centre for archaeological research at Apamea in Syria (“Centre Belge de recherches archéologiques à Apamée de Syrie”). The 2001-2010 campaigns were directed by Didier Viviers. Among substantial urban Roman structures including an agora, a large Roman theatre, pagan (a “tycheion”) and Christian buildings, baths and residential areas, the excavations produced medieval layers including early Islamic structures (Viviers and Vokaer 2015). Although the early Islamic/Abbasid settlement was unpublished, a fair amount of medieval ceramic material were found in the north-eastern quarter of the site and its central sector where large Roman villas were located (Rogers 1972, 256; Rogers 1984, 261). This body of glazed pottery was unfortunately unstratified because of later disturbances such as relocation of soil for fills or accumulated debris from floor scrapings (Rogers 1984, 261). The ceramics of the Abbasid period found in these mixed contexts form a relatively small assemblage composed of unglazed and glazed pottery. The Unglazed Buff Wares (including the moulded type) were found in small quantities, whereas Brittle Ware ceramics were well-represented. The glazed wares consisted of Splash Glazed Ware, White Opaque Glazed Ware Imitation and Polychrome Painted Glazed Ware (“Mustard and Cress Ware” was the most common class of glazed wares), Glazed Moulded Ware and Monochrome Green Ware with stamped decoration (represented by very few sherds) (Rogers 1984, passim).
Balis-Meskene was the ancient Barbalissos known as a military and port city which lies on the right side of the Euphrates river. The site was excavated before the World War Two, and in the 1970s, by a French team but the results were never fully published. Apparently, some ceramics of the Abbasid period were found in deep soundings (Leisten 2000, 37, n.3). Between 1996 and 1998, new excavations were conducted with the collaborative efforts of the University of Tübingen and Princeton University on the western part of the city. The new archaeological excavations were designed to investigate the pre-Ayyubid occupation of the site, including the relation between the Byzantine and Islamic layers and the layout of the settlement (Leisten 2000, 37-38). The pottery assemblage came from four different areas, including one from a castle. This ceramic corpus was of substantial size (c. 16000 sherds) dated to the 9th - 10th and 12th - 13th centuries C.E. (Leisten 2000, 47). The wares of the Abbasid period included the plain and moulded type of Unglazed Buff Wares which were found in great quantities. Brittle Ware examples were recorded as well. The assemblage included a rather modest amount of glazed wares mainly comprising Polychrome Painted Glazed Ware, Splash Glazed and Polychrome Sgraffito Wares (Leisten 2000, 53). The pottery finds of the Abbasid period were never fully published with the exception of one article dealing with Unglazed Moulded Ware of the Abbasid and Ayyubid periods (Mulder 2014). These carefully excavated Islamic layers are largely unknown.
The site of Madinat al-Far is situated 70 km to the north of Raqqa. The first excavations were directed by Nassib Saliby between 1981 and 1982, then by Claus-
Peter Haase from 1987 until circa 1999. Because of the association of the site with the fortress of Maslama which was built by the Umayyad caliph Abd al-Malik according to the Arabic sources, the archaeological research of the German-Syrian expedition was mainly designed to understand the Umayyad occupation in the area (Bartl 1994, 121; Haase 2008, 395). The site proved to be occupied in the Abbasid period as was attested by the pottery evidence (mid 8th – mid 10th century). The pottery material came from several contexts including brick made cisterns which were located in the walled-city centre. Moreover, there was other ceramic material collected from sondages outside the city centre. The ceramic corpus of the Abbasid period included Polychrome Painted Glazed Ware, Splash Glazed and Polychrome Sgraffito Wares;
Unglazed Buff Wares (wheel-made and moulded) and Brittle Ware cooking pots (Bartl 1994, passim).
Resafa which was an important religious and military center in the Roman and late Antique period, is situated in the northern Syrian steppe land, in the southwest of Raqqa. The site was first discovered by Sarre and Herzfeld and then, it was excavated by a German team directed by J. Kollwitz and K. Otto-Dorn in the 1950s (Sack 1996, 3). From 1977 onward, Thilo Ulbert continued the archaeological enterprise in and around the site (Sack 1996, 3). The city which was constructed from gypsum and raw bricks and surrounded by a massive enclosure presents a fairly elaborate urban plan (Brands 2011, 59). The most intensive building activity there spanned between the end of the 5th century and the beginning of the 6th century (Brands 2011, 62). Some of the important monuments consist of the church of Saint Sergios (Basilika A), the martyrion and the large underground cisterns amongst others (Brands 2011, 59). In the 8th century, the Umayyad caliph Hisham constructed several palaces in and around the city. The material culture (coins and ceramics) that may be associated with Abbasid occupation was of limited amount and may have reflected “a collapse of the urban economy” (Brands 2011, 75). Dorothee Sack conducted excavations in the Great Mosque of Hisham, which was a mosque constructed adjacent to Saint Sergius’
Church in the 8th century; Nusa Logar studied the ceramics found there (Logar 1996;
Sack 1996). The ceramic material was dated by comparison to similar finds due to frequent disturbance of the stratigraphy (Logar 1996, 77). Some of the common classes of Abbasid pottery were found: Brittle Ware, Unglazed Buff Wares (including the moulded type which was occasionally dated later), Monochrome Green Wares, Splash Glazed Ware, Polychrome Painted Glazed Ware and few Lusterware pieces (Logar 1996, 77, 84-86).
The Balikh survey in North Syria - The survey around the region scored by the Balikh river was conducted in 1983 in the frame of the excavations of Tell Hammam et Turkman directed by the Dutch archaeologist Peter Akkermans. This survey aimed to trace the settlement patterns of the Balikh area where 200 settlements were identified chronologically ranging between the aceramic Neolithic and the Islamic periods (Akkermans 1989; Bartl 1994, 1). The surveyed area spanned between the Turkish border southwards to the Balikh lying East of the city of Raqqa in northern Syria. The German archaeologist Karin Bartl published the Islamic ceramics finds.
The early Islamic pottery included Unglazed Buff Wares including moulded examples, Brittle Ware, Monochrome Glazed Wares, Splash Glazed Ware and Polychrome Painted Glazed Ware (“Ware mit mehrfarbiger Glasurmalerei oder Unterglasur”) (Bartl 1994, 27).
Sites in Iraq (Tab. 4.1.1-2c)
The site of the castle of Ana is an island located in north-western Iraq on the western stream of the Euphrates. The site was excavated by Alastair Northedge part of a rescue project in 1981 and 1982. The chronological framework spanned from the Neo-Assyrian to modern times. The Abbasid period was represented in several trenches (Northedge et al. 1988, 27). The pottery was classified in two consecutive phases including the “early Abbasid” and the “Samarra Horizon Ceramic Repertoire”
periods (Northedge et al. 1988, 80-94). While the earlier phase (c. mid 8th – mid 9th centuries) mainly produced examples of Unglazed Buff Ware referred to as “Thin Wares”, the later phase (c. mid 9th – mid 10th centuries) yielded the whole “Samarra Horizon Ceramic Repertoire” as well as pieces of Unglazed Buff Ware and examples of Brittle Ware (Northedge et al. 1988, 82-83).
The Abbasid capital Samarra lies north of Baghdad in central Iraq on the eastern shore of the Tigris (caliphal residence: 839-892 C.E.). The site was first excavated by the German archaeologist Friedrich Sarre and Ernst Herzfeld between 1911 and 1913. These excavations produced the ground for a general typology of the Abbasid ceramic repertoire referred to as the “Samarra Horizon Ceramic Repertoire”
(Sarre 1925). Qualifying a broad range of polychrome glazed table wares that were first introduced under the Abbasids, the main classes of this ceramic repertoire included Lustrewares, White Opaque Glazed Ware with cobalt blue decoration, Glazed Relief Ware, Splashed and Polychrome Sgraffito Ware.
Between 1983 and 1989, the British archaeologists Alastair Northedge and Robert Falkner undertook a survey in several places of the Abbasid city of Samarra including the area between Qasr al-Ja‘far and the north of al-Mutawakkil’ s new city and the Mosque of Abu Dulaf and the remains to the east of the octagonal enclosure of al-Qadisiyya (Northedge and Falkner 1987, 146). This survey aimed to understand the occupation of the site, in particular those of the non-Abbasid period. Furthermore, following the accounts of Abbasid historians and geographers, some of the architectural remains were identified such as the unfinished city called al-Qadisiyya among others, characterised by its octagonal shape. The survey did not help to refine the dating of the Abbasid wares but offered evidence to date the occupation of the site (ranging broadly between 3000-2000 B.C.E. to the 12th - 13th century C.E.) (Northedge and Falkner 1987, 162). The 9th - 10th century pottery was represented by Monochrome Green Wares and White Opaque Glazed Ware Imitation with splashed decoration, Unglazed Buff Ware and Brittle Ware (Northedge and Falkner 1987, 163- 164).
Tulul al-Ukhaidir was an Abbasid castle built in southern of Iraq, to the West of the Euphrates. The site was surveyed and excavated by the German archaeologists Jürgen Schmidt and Barbara Finster between 1973 and 1976 (Finster and Schmidt 1976). The ceramic evidence revealed a broad spectrum of Unglazed Buff Wares and Brittle Ware cooking pots belonging to the period before the “Samarra Horizon Ceramic Repertoire”, when polychrome glazing was not recorded (Finster and Schmidt 1976, 92-93, 111).
Tell Abu Sarifa is a small mound situated in southern Iraq between the Tigris and the Euphrates close to the Sumerian city of Nippur (Mac Adams 1970, 87). The excavations were undertaken by the American archaeologist Robert Mac Adams. The pottery evidence yielded the whole repertoire of unglazed and glazed wares of the Abbasid period. The scholar referred to Unglazed Buff Ware as “Thin Wares” (Mac Adams 1970, 100). The scholar suggested that Tell Abu Sarifa reflected the ceramic industry and economy of a modest village on the basis of the “poor quality” of the
glazed fine wares such as Polychrome Sgraffito Ware and the limited occurrence of imported glazed ceramics like Lustrewares (Mac Adams 1970, 105).
Sites in Iran (Tab. 4.1.1-2d)
Susa is located at the eastern bank of the Euphrates in western Iran near the Iraqi border. The site was excavated several times mainly under the auspices of French archaeological missions from the 1920s onward. For example, these included the 1929-1933 campaigns of Roland de Mecquenem, the 1946-1967 campaigns of Roman Ghirshman, the 1968-1979 campaigns of Jean Perrot and the 1972-1978 campaigns of Monique Kervran and Remy Boucharlat (Ghirshman 1964; de Mecquenem 1925;
Kervran 1977). The site was composed of several parts including the Apadana, the Acropolis, the Royal City and the Artisans’ City ranging from the Sukkalmah period (c. 1880-1450 B.C.E.) to the Islamic period (c. 13th century C.E.) (Rosen Ayalon 1974).
The Islamic material was published by various scholars including Raymond Koechlin, Myriam Rosen Ayalon, Monique Kervran and Claire Hardy Guilbert (Hardy Guilbert 1984; Kervran 1977; Koechlin 1928; Rosen Ayalon 1974). The latest campaign which was directed by Monique Kervran contributed to refinement of the typo-chronology of the ceramics of the Abbasid period. The ceramic finds included classes of “the Samarra Horizon Ceramic Repertoire” including the wide range of polychrome glazed pottery and examples of Unglazed Buff Wares (Kervran 1977).
Siraf is located on the Persian Gulf in southern Iran. The city constituted one of the largest Abbasid emporia controlling the Indian Ocean trade between mid 8th - 11th centuries C.E. The site was predominantly investigated by British projects. The first excavations were undertaken by the British archaeologist David Whitehouse between 1966 and 1971, followed by a typological study of Seth Priestman, in 2007- 2009 (Whitehouse 1972; Priestman 2011). In addition to yielding the whole pottery repertoire of the Abbasid period, the excavations of Siraf were very efficient in setting the relative chronologies for these different classes of ceramics (Whitehouse 1979).
Furthermore, the site was a pottery production centre. Unglazed Buff Ware production was attested by the recovery of kilns, and wasters of Opaque Glazed White Wares suggested in situ glazed ceramic production as well (Whitehouse 1971, 12, 14-15).
Nishapur which was a Sassanid city founded in the 3th B.C.E is placed among one the most important cities of the Khurasan province in eastern Iran (Wilkinson 1973, XXXIII). The city hosted a significant Muslim population after the spread of Islam. The excavations of the Metropolitan Museum of New York, undertaken by Joseph Upton, Walter Hauser, and Charles K. Wilkinson in 1935-1940 and 1947, uncovered a substantial Islamic settlement spread around several “tepe” (mounds) where occupation apparently continued intermittently until the modern period (Wilkinson 1973, XXXIII). Substantial amounts of medieval ceramics including early Islamic material were found. The site also revealed pottery workshops producing Unglazed Buff Wares (Wilkinson 1973, 355). Some selected pieces dated between the 9th and the 11th century were published by Wilkinson without clear stratigraphic distinction (Wilkinson 1973).
The publication Nishapur Revisited consisted of the results of the new Irano- French excavations carried out in the earliest part of Nishapur called Qohandez (citadel) by the collaborative efforts of the Musée du Louvre, the CNRS and the MAEE (Rante and Collinet 2013). These new excavations were conducted on several parts of the mound to resolve some chronological and topographic issues including
the date of the foundation of the city (Sasanian or Islamic), the time span of the settlement and its geographical limits and the dating of the artefacts associated with these different strata (Rante and Collinet 2013, 6). While the pottery finds were subjected to petrographic and chemical analysis, the stratigraphic sequence was dated with thermoluminescence (TL) and archaeomagnetic analysis. Subsequently, several medieval phases could be identified. The material culture of the first phase of level III dated between the late 8th and 11th centuries, showed a change characterised by a peak in glazed table wares. The majority of these glazed table wares seemed to be manufactured as of the late 9th century in Nishapur (Rante and Collinet 2013, 205).
The early Islamic ceramics included Iraqi imports such as White Opaque Glazed Ware with cobalt blue decoration, the imitation of this ware, Splash Glazed Ware and Polychrome Sgraffito Ware, Unglazed Buff Wares including few examples of Unglazed Moulded Ware (Rante and Collinet 2013, 174-183). The Abbasid layer (c.
9th - 10th centuries) and the Samanid layer (c. 10th - 11th centuries) could not be distinguished in the stratigraphic sequence.
Sites in Jordan and Palestine (Tab. 4.1.1-2e)
Tiberias which is mid-way on the west coast of the Sea of Galilee in Israel, was intermittently excavated for the past 50 years. Recently, David Stacey, an experienced field archaeologist of Jordan-Palestine, produced a very insightful publication dealing with the early Islamic layers (Stacey 2004, 1, 9-10). These strata were uncovered in 1973 and 1974 as part of a salvage excavation in the south of the ancient site of Tiberias where a hotel was going to be erected (Stacey 2004, IX). These excavations were directed by Gideon Foerster and supported by the Israel Antiquities Authority, the Ministry of Housing and the Tiberias Municipality, in collaboration with the Institute of Archaeology of the Hebrew University and the Israel Exploration Society.
The early Islamic pottery assemblage of this site came from a selected context called area C. This area was located at the north of a Roman monumental arch, within the Byzantine city walls and it was continuously occupied between the 7th and the 12th centuries (Stacey 2004, 21). The ceramic assemblage included 9th - 10th century pottery classes, Lustrewares, according to the author, both from Iraq and Egypt, some glazed wares related to Polychrome Painted Glazed Ware, Polychrome Sgraffito Ware, Opaque Glazed Ware Imitations, Unglazed Buff Wares and Soft-stone Imitations (Stacey 2004, passim).
Caesarea Maritima is one of the major ancient ports of present day Israel, situated on the Mediterranean coast, 15 km north of the current Israeli capital. Large scale excavations were conducted in the site between 1992 and 1997 with the collaboration of Combined Caesarea Expeditions and the Israeli Antiquity Authority.
A substantial amount of post-antique ceramics were recovered from several places, within the old city walls, outside of it, south of the fortifications, under the southern wall of the old city, among others. Apparently, many pieces of the assemblage came from relatively well-dated contexts such as sealed deposits dated with coins, inscriptions, and other datable items. The pottery of the post-Byzantine periods was dated according to both dynastic frameworks and calendar dating, the early Islamic (640-1101 C.E.), the Crusader and Mamluk periods (1101-1291 C.E.) (Arnon 2008, 9). The publication contained many inconsistencies particularly dealing with the 8th to 11th century phases which makes it very difficult to date the ceramic finds associated
with these layers.79 Further, the phasing of the ceramic catalogue did not correlate with the main text. Even though continuous occupation was mentioned, the chronology included a gap between the mid 9th and late 9th century.
The body of ceramics of the early Islamic period contained the “the Samarra Horizon Ceramic Repertoire”: Lustrewares, Opaque Glazed Ware with cobalt blue decoration, Glazed Relief Ware, Splash Glazed Ware, Polychrome Sgraffito Ware, White Opaque Ware Imitation, some glazed wares related to Polychrome Painted Glazed Ware, Soft-stone Imitations, Unglazed Moulded Ware and Monochrome Green Ware with stamped decoration (Arnon 2008, passim). It is unfortunate that this large corpus of pottery was not accurately dated.
Khirbat al-Mafjar is renowned for being one of the very exquisite examples of an Umayyad palace in the Palestinian desert (Hamilton 1959). The building was apparently abandoned after an earthquake dated to circa 747 C.E. The preliminary excavations at the site were conducted by Dimitri Baramki under the auspices of the Department of Antiquities of Palestine between 1937 and 1948. These archaeological explorations yielded a substantive amount of Unglazed Buff Wares and other glazed pottery of the Abbasid period such as some Polychrome Painted Glazed Ware pieces (Baramki 1944, 66-72, 80-81, 90-91, 98-99). These first investigations were followed by Robert Hamilton. Hamilton studied primarily the buildings of the palace that he interpreted according to a narrative drawn from Islamic texts (Hamilton 1959).
More recently, new excavations were carried out by an American-Palestinian team under the direction of Hamdan Taha and Donald Whitcomb between 2006 and 2013 (Whitcomb and Taha 2013). Whitcomb’s reinterpretation of the excavations revealed that the earthquake that occurred in 747 C.E. which was conceived as a hiatus in occupation brought about less damage than it was initially assumed and that habitation on the site continued with a slight reduction through the Abbasid, Ayyubid and Mamluk periods (Whitcomb 1988, 65). With regards to the ceramics, Whitcomb correlated the pottery with their contexts and found what he qualified as the transitional types of ceramic, the phase associated with the earthquake (750-800 C.E.) (Whitcomb and Taha 2013, 60). These types were identified on the basis of comparanda with Mahesh Ware recorded at Aqaba, a variant of Unglazed Buff Ware characteristic of the early Abbasid period in southern Palestine dated c. 750-800 or later). The Abbasid period (800-950 C.E.) was the major occupation of the site and produced the wide range of unglazed and glazed pottery introduced during this period including Unglazed Buff Ware and Unglazed Moulded Ware (Whitcomb and Taha 2013, 60).80
After the excavations initiated by the British archaeologist Crystal Bennet between 1975 and 1979, the archaeological investigations on the citadel of Amman resumed with the efforts of Alastair Northedge between 1979 and 1981. Other archaeological projects were conducted simultaneously including the Spanish excavations directed by Martin Almagro starting in 1974 (Northedge 1992, 17-18).
The ceramic material of the Abbasid contexts in the citadel was briefly described and dated. This pottery included mainly examples of glazed table wares such as
79 In the main text, the phase VII was dated to the late 8th and the early 9th centuries and phase VI was dated mid 9th - mid 10th centuries in the ceramic catalogue (Arnon 2008, 82, 166). But in the preface, the phase VII was attributed to mid 8th century and late 9th and the phase VI to late 9th mid 10th century (Arnon 2008, preface).
80 Withcomb calls Unglazed Ware “Cream Wares referred to Mafjar ware by the excavators”
(Withcomb and Taha 2013, 60). The terminology difference demonstrates that there is the need to establish more standardised names for the ceramics of the Abbasid period.
Polychrome Painted Glazed Ware, Splash Glazed Ware and Polychrome Sgraffito Ware and for the common pottery pieces of Unglazed Buff Ware (Northedge 1992, 159).
Another important archaeological project in Jordan corresponds to the excavations of the port city of Aqaba (early Islamic Ayla) which is located in south- western of Jordan, on the débouché of the Red Sea. This project was directed by Donald Whitcomb between 1986 and 1993 and primarily focused on the medieval Islamic layers spanning between the 7th and the 12th centuries C.E. (Whitcomb 1994).
With regard to the ceramic material, the excavations yielded the most common pottery classes of the Abbasid period including White Opaque Glazed Wares and variants of Polychrome Painted Glazed Ware and Unglazed Buff Ware (Whitcomb 1994).
Pella lies in northwestern Jordan, in the south of the Sea of Galilee. It was excavated since 1979 by the Sydney-Wooster Joint expedition. The site is an extensive multi-period settlement which started around 8000 B.C.E. and continued without major interruption.81 The early Islamic settlement was uncovered on several areas. Umayyad occupation was recorded in a large residential district on the main mound and in Byzantine churches all of which was destroyed by a major earthquake.
The 749 earthquake left an extraordinary archaeological record. Umayyad houses which were fully equipped with their dwellers were found intact (Walmsley 2006).
This destruction layer was also an important chronological demarcator between the Umayyad and the Abbasid periods. Major Abbasid constructions were found in the northeast of the mound among others (Walmsley 1992, 379). The stratigraphic sequence has shown that 9th century pottery deposits constituted a major break with the earlier layers. With regards to the ceramic material, this phase was characterised by the appearance of Unglazed Buff Wares (“thin-walled Samarra style pale cream jars and strainer jugs”), to be followed in the mid 9th century, by polychrome glazed wares such as Splash Glazed Ware (Walmsley 1995, 668). The assemblage did not contain any wares of the “Samarra Horizon Ceramic Repertoire” such as Lustrewares.
These new polychrome glazed wares were imitations of Egyptian wares (Walmsley 1995, 664). A variant or a local imitation of the so-called “Coptic Glazed Ware”
characterised by bowls with a carinated body and a flat base, and decoration painted in green, black and yellow was recorded among the mid 9th century ceramic finds (Walmsley 1995, 664).
Sites in Egypt (Tab. 4.12-3e)
Fustat which was established as the new capital of Egypt after the Muslim conquest, flourished as a trade centre between the 7th and the 14th centuries C.E. (Jomier 1965, 957-959). Retaining most of its medieval structure, the city became the object of numerous archaeological explorations dealing with the Islamic period. Fustat constitutes one of the most important sites for Islamic archaeology. In fact, the 1912- 1924 excavations of the Egyptian archaeologist Ali Bahgat Bey are often considered to be a starting point for the modern discipline of Islamic archaeology in Egypt (Treptow 2013). Another ground breaking project were the excavations carried out by George Scanlon between 1964 and 1980 under the auspices of the American Research Centre in Cairo which focused on several areas of the medieval Islamic city (Kubiak and Scanlon 1989). And finally, the French campaigns directed by Roland-Pierre Gayraud in Istabl ‘Antar between 1985 and 2005 focused on some residential areas
81 www.sydney.edu.au.
but more importantly on the extensive necropolis lying at the fringes of the city (Gayraud 2003).
The finds from Istabl ‘Antar were in fact well dated, with a number of evidence including money, glass stamps and the burnt deposit because of the fire set out in the city by the last Umayyad caliph Marwan II in 750 (Gayraud et al. 2009, 172). The closed deposits of several pits enabled to refine the dating of the different types of early Islamic ceramics. The evidence of these pits could date the introduction of the glazing technology into the phases after 750 C.E. Furthermore, the development of the formal repertoire pertaining to more orientalising traditions (opaque glazes, flared bowls with ring bases) was dated to last third of the 9th century (Gayraud et al. 2009, 190). The chronological sequence of this assemblage showed that the technological innovations on ceramics common to the 9th - 10th century (polychrome glazing and new morphological repertoire) appeared simultaneously in Egypt and in Iraq. Among other pottery classes, the ceramic evidence uncovered at Fustat included Monochrome Lustreware, Splash Glazed Ware, Glazed Relief Ware, Opaque White Glazed Ware Imitations, Polychrome Sgraffito Ware, Polychrome Painted Glazed Ware and variants of Unglazed Buff Wares (Gayraud et al. 2009;
Kubiak and Scanlon 1973; Scanlon 1984).
Sites in Africa (Tab. 4.1.1-2e)
Lastly, it is relevant to mention two sites located in the Lamu Archipelago in Kenya.
Manda and Shanga were excavated by the British archaeologists Neville Chittick between 1970 and 1978 and Mark Horton between 1980 and 1988 (Chittick 1974 and 1984; Horton 1996). By exploring the settlements and the material culture of medieval Arab colonies, the archaeological finds including the ceramics demonstrated strong commercial ties with the Abbasid emporium Siraf in southern Iran. The pottery evidence included Abbasid White Opaque Glazed Wares like Lustrewares and White Opaque Glazed Ware with turquoise splash, Splash Glazed Ware and Polychrome Sgraffito Ware and very few examples of Unglazed Buff Ware and Brittle Ware (Horton 1996, 276-277, 295-299; Priestman 2011, 108). The ceramic sequence of Shanga was very insightful for understanding the diffusion of Abbasid pottery classes which generally supported the dating of the ceramic evidence from Siraf (Horton 1996, 277).
Conclusive remarks
New progress - The current state of knowledge about the material culture of the Abbasid period, and more particularly ceramics is built on a number of archaeological projects from the Near East and Egypt and fewer sites outside this core area. The emerging picture is far from being complete but, over the past decades, has relatively contributed to a better understanding of the pottery production, diffusion and consumption of the Abbasid period including the “Samarra Horizon Ceramic Repertoire” as well as Unglazed Buff Wares and Brittle Ware.
This improvement was the result of numerous archaeological investigations in Syria and south-eastern Turkey until the break out of war. In addition to revealing ceramic workshops in urban centres, these studies provided new information about the diffusion and consumption of these wares in rural areas (e.g., the Balikh survey in Syria, the Amuq survey in Turkey). This situation was slightly different in Iraq where the endless, intermittent conflicts obstructed new archaeological investigations.
The archaeological studies of Samarra were numerous (e.g., surveys of Samarra of Northedge, Kennett and Falkner) because it was the capital of the Abbasid Caliphate and its medieval remains were still visible and somewhat undisturbed. The rural areas in southern Iraq and Basra, the largest emporia on the Persian Gulf, are still poorly known. The lands East of Iran generally remained unknown to Western literature because of language barriers but are of great importance for acquiring information about the eastern part of the Abbasid Caliphate.82 How far can the effects of the technological innovation in ceramics be traced? Further research will shed light on that.
The archaeological inquiries in the southern Levant and Egypt have been growing over the past decades and contributed significantly to a fuller knowledge of the eastern Mediterranean in the Abbasid period. Surprisingly, common wares such as Unglazed Buff Ware are studied as much as glazed wares.
Unsolved issues - Despite these advances, some major problems still persist. The overwhelming problem is centred on the dating of ceramics. One of the main issues of early Islamic pottery, more particularly, the wide range of polychrome glazed wares of the Abbasid period, referred to as “Samarra Horizon Ceramic Repertoire” is the problem of absolute chronologies. In fact, despite the persistence of the debate on the high and low chronology of this group of fine wares (discussed below), the excavations in the Iranian sites Susa allowed to confirm the progressive introduction of the different wares of this pottery repertoire which stretched over a century (Kervran 1977). In addition to Susa, the excavation at Siraf provided relative chronologies for the respective phases of the introduction and circulation of the different wares. However, absolute dates are difficult to establish mainly because of the numismatic evidence. The coins are in fact scarce in assemblages dating between the mid 9th and the 11th centuries (Heidemann 2011, 55). To illustrate this situation, Alastair Northedge gave the example of the Samarra excavations which produced only two dirhams (Northedge 1996, 230). The general rarity of coins blurred the stratigraphy even more by raising the possibility of the association of earlier coins with later pottery (Northedge 1996, 230). Furthermore, the precise dating of other finds such as glass and metal are not well-established either, which does not help to solve the chronological problem. In this context, scientific methods such as radiocarbon or thermoluminescence dating could be helpful. These techniques were employed rarely even though this situation is slowly changing.
One should also be alert with respect to the relative chronologies established on the basis of comparanda with similar pottery finds. Alan Walmsley underlined the problem of repeating erroneous dates frequently attributed during old excavations. For instance, Unglazed Buff Ware examples uncovered on the Jordanian site of Mount Nebo were dated three centuries earlier than what they were (Walmsley 2001, 306).
As a solution to misinterpretations, he suggested to consider first and foremost the morphological components of ceramics in order to detect specific technological practices and changes (Walmsley 2001, 306). This call for a more thorough study of the technological aspects of pottery is well-founded. This perspective would also require a multi-disciplinary approach using experimental archaeology on the one hand but also science based methods such as petrography or chemical analysis on the other.
After this brief survey of important archaeological investigations of the Abbasid period, the following part focuses on the ceramic material recorded from
82 To this I would suggest to add the southern Caucasus including Georgia and Armenia.
these different sites and surveyed areas. It situates the early Islamic ceramic corpus of Gözlükule in a broader framework by discussing the consumption, distribution and production of the main classes of pottery and their respective chronologies.
II/ Major pottery classes of the Abbasid period: production, distribution and consumption
For each ceramic class, this part provides a discussion about the history and the characterisation of the production, information about the distribution patterns and comparanda with similar material from other sites. The different wares are described in a chronological order, starting with the earliest ones. Glazed ceramics are presented first followed by unglazed wares.
General characteristics of ceramics of the 9th and 10th century
As seen previously, the industrial revolution in the Abbasid period had a fundamental impact on ceramic production. The development of new pottery technologies resulted in the introduction of new wares with different shapes and decorations that became highly consumed goods diffused in the Dār al-Islām spanning from North Africa to Uzbekistan (Watson 2004, 47). The morphological repertoire broadened and lead glaze became a common coating for ceramics and lamps. Glazed table wares emerged as a new type of commodity from the 9th century onwards (Watson 2004, 45).
The “Samarra Horizon Pottery” or the “Samarra Horizon Ceramic Repertoire”
are the names given to the most common glazed table wares introduced as part of this major economic and technological change. This appellation refers to the spot where these polychrome glazed dishes were first found, namely, the excavations of the Abbasid palatial complex of Samarra which were undertaken by Herzfeld and Sarre in 1911-1913 (Sarre 1925). This naming which prevails Samarra as the sole origin of polychrome glazing is naturally problematic and restrictive. In fact, these different classes of glazed table wares (discussed below) were used over a longer period, extending beyond the times of the occupation of Samarra by the caliphs (836-892).
Nonetheless, this term is generally accepted and used until new information emerges.
Simply put, the “Samarra Horizon Ceramic Repertoire” qualifies a broad range of polychrome glazed ceramics that were introduced in the 9th - 10th centuries under the advent of the Abbasids. The polychrome glazed ceramics include a group of lead glazed wares: Polychrome Splash Glazed Ware (Splash Glazed Ware), Polychrome Sgraffito Ware, Relief-Moulded Ware (Glazed Relief Ware); and a group of opaque glazed wares: Lustrewares (Monochrome and Polychrome Lustreware), White Opaque Glazed Ware with cobalt blue decoration, White Opaque Glazed Ware with cobalt blue and green or brown decoration and Plain Opaque White Glazed Ware (Northedge 2001, 210).
Although the relative chronologies of the “Samarra Horizon Ceramic Repertoire” are generally accepted, scholars cannot agree on the absolute chronologies. Two ways of thinking have been formulated. The “high chronology”
associates this ceramic repertoire with the Abbasid revolution and dates it to the 8th century. This idea started with pottery specialist Florence E. Day in the 1940s and is followed by archaeologist Monique Kervran (Northedge 2001, 210). The “low chronology” which is defended by British scholars such as John Hansman and David Whitehouse sets this phenomenon from the first half of the 9th century to the end of the 10th century (Northedge 2001, 210) (Table 4.2.1). It was already noted elsewhere in this thesis: important historical events do not generally have direct consequences
on material culture. In fact, ceramics frequently have their own rhythm with respect to change (Vroom 2003, 26). Subsequently, I would propose that the idea that the
“Abbasid revolution” triggered an immediate and ground-breaking innovation on pottery technology (the creation of the white opaque glaze and the wide use of polychrome glazing on daily vessels) does not seem so plausible. I believe that the dating proposed in the “low chronology” appears to be much more consistent.
In contrast to the lead glazed types, the opaque glazed wares were a complete innovation. Although it is not entirely clear, Chinese ceramics (Tang Ware) were probably the main inspiration in the manufacture of these new wares (Hallett 2000).
Hence, Lustrewares, White Opaque Glazed Ware with cobalt blue decoration, White Opaque Glazed Ware with cobalt blue and green or brown decoration, and Plain Opaque White Glazed Ware became a widely consumed commodity object, circulating along sea routes (see distribution maps below). Oliver Watson suggested that the clientele of these ceramics may be two social groups, the wealthy upper class who could afford luxurious pieces made of silver and metal, and the “bourgeoisie”, the new urban class emerging as part of the economic and social changes brought by the advent of Islam, who could mainly buy the less costly earthenware vessels (growth of urbanisation, commercialization of the economy, commodification, etc.) (Watson 2004, 47). I believe that this idea seems very likely considering the historical background of the Near East in Abbasid times, when they were experiencing a
“Golden Age” (see chapter one).
Opaque Glazed Wares are called in reference to the opaque, generally, white, glaze that they are covered with. Not so long ago, archaeologist Seth Priestman established a broad typological definition of White Opaque Glazed Wares: these ceramics were “characterised by a finely-levigated cream-coloured calcareous body with limited coarse inclusions, and a thick, opacified glaze that covers the entire interior and exterior surfaces, except for the foot where the vessels were stacked in firing” (Priestman 2011, 89). This new characterisation broadened the repertoire of White Opaque Glazed Wares while new types were identified within this large group, on the basis of a substantial amount of pottery finds from archaeological work conducted in Iran, Manda and Shanga (Kenya). Subsequently, Priestman distinguishes nine major types of White Opaque Glazed Ware (essentially sixteen83): OPAQ.W1, OPAQ.D, OPAQ.MGL, OPAQ.C, OPAQ.PS, OPAQ.PL, OPAQ.MRL, OPAQ.W2, OPAQ.TBS (Priestman 2011, 91-94) (see footnote for full explanations of the abbreviations).
In the typology of the Gözlükule medieval ceramic corpus, the Opaque Glazed Wares consist of Polychrome Lustreware, Monochrome Lustreware, White Opaque Glazed Ware with cobalt blue decoration, White Opaque Glazed Ware imitation, White Opaque Glazed Ware with turquoise and/or other colour decoration.
1 White Opaque Glazed Ware with cobalt blue decoration (c. 9th century C.E.)
83 These 16 types are classified as OPAQ (referring to Opaque Glazed Ware), plus an additional letter standing for the initials of the surface treatment technique(s) is added. These correspond to OPAQ.W1 (Monochrome White, Iraqi), OPAQ.TS (Turquoise Splashed on White), OPAQ.D (Degraded or Non- Identified), OPAQ.MGL (Monochrome Gold Lustre), OPAQ.C (Cobalt on White), OPAQ.T (Monochrome Turquoise), OPAQ.PS (Polychrome Splashed), OPAQ.PL (Polychrome Lustre), OPAQ.B (White on Monochrome Black), OPAQ.CP (Color Painted, Unglazed), OPAQ.MRL (Monochrome Ruby Luster), OPAQ.BW (Black on White), OPAQ.W2 (Monochrome White 2, Sirjan), OPAQ.WC (White on Monochrome Cobalt), OPAQ.BT (Black on Turquoise), OPAQ.TBS (Turquoise and Black Splashed on White) (Priestman 2011, 91-94).
A history of the production
The art historical premise inherited from Arthur Lane regarded White Opaque Glazed Ware with cobalt blue decoration and related wares as the symbol of a genuine Islamic tradition.84 Many scholars suggested that the raison d’être of this ware was a response to the increasing demand for Chinese ceramics (Allan 1991, 6). This is also partly because this pottery class was one of the earliest classes of the “Samarra Horizon Ceramic Repertoire”, dated generally to the 9th century on the basis of the ceramic sequences of Siraf and Susa in Iran (Northedge and Kennet 1994, 25). The late 8th century dating could be too early as it will be discussed below.
Development of the production
White Opaque Glazed Ware with cobalt blue decoration like Lustrewares were made of a very fine clay of light yellow colour currently associated with a provenance in southern Iraq, in Basra (Northedge 2001, 210). In comparison with the widely distributed Lustrewares, White Opaque Glazed Ware with cobalt blue decoration showed a rather less extensive distribution area, whereas its morphological repertoire followed the pottery forms of Lustrewares (see distribution map below). Most of the information about the technological development of this class came from studies conducted on museum ceramics (Hallett 2000; Mason 2004). Jessica Hallett who compared the glaze composition of this class with Lustrewares suggested that the variation of the glaze composition could mark an earlier stage of production, while Lustrewares bore a more standardised recipe (Hallett 2000, 153). Furthermore, SEM (Scanning Electron Microscope) analysis by Nigel Woods on some sherds from an archaeological assemblage has shown that slip did not contain any tin (Priestman 2011, 96). The idea of slip being employed on the earliest tin opacified glazes was put forward by Robert Mason previously (Mason and Tite 1997). Hallett’s idea seems to be reasonable in this context.
Robert Mason developed a seriation model for the major classes of Abbasid pottery (“Samarra Horizon Ceramic Repertoire”) by examining a large body of early Islamic ceramics coming from excavations, museum collections and surveys. This seriation was based on vessel shapes, decoration styles and the petrographic study of fabrics (Mason 1997).85 Even though this research was rigidly criticised, Mason’s work constituted one of the very few studies dealing with Abbasid pottery. In his research, Mason distinguished four stages in the development of the bowls of White Opaque Glazed Ware with cobalt blue decoration summarised below (Mason 1997).
The first phase was characterised by a bowl associated with a pre-Islamic, Sasanian shape. This bowl displayed a rounded body with simple rims and a low, thick ring foot (Mason 1997, 22). The decorative repertoire consisted of stylized vegetal motifs referred to by Mason as “sprig” or “harp” (Mason 1997, fig. BB1, BB.4). The second phase of the seriation was marked by a new shape that showed no direct continuity with the preceding one. The bowls had a short ring foot and everted and curved rims with a triangular profile. The decoration included three-lined calligraphy, the “sprig” motif of the initial phase and arcs (Mason 1997, fig. BB.5-6).
Three vessel forms appeared in the third phase. The same base (low ring foot) was used for three different rims. The first was a simple rim reminiscent of the first phase,
84 “Fortunately, the Baghdad potters did not rest content as imitators. They were tempted to paint on the fine white surface, using cobalt blue, copper-green, manganese-purple, and sometimes antimony yellow.” (Lane 1947,13).
85 For more details about the white opaque glazed ware assemblage he studied and Priestman’s critique, see: Lustrewares.
the second was slightly everted, and the third was a short everted rounded rim. In line with the diversification of the vessel shapes, the decoration repertoire broadened.
Geometric motifs became more common and included circles, triangles, thin frames filled with points (Mason 1997, fig. BB.11, BB.7, BB.14). The three-lined calligraphy shortened into two lines (Mason 1997, fig. BB.9). In the fourth phase, the rims remained recurved while the foot grew taller and the base narrower. The decoration displayed geometric or vegetal forms filled with hatchings such as triangles or leaves;
an eight lobed flower motif is present as well (Mason 1997, fig. BB.16-18) (Fig.
4.2.1).
Production
Although other production centres in the vicinity of Basra may exist, Basra is considered as the main centre for production as shown by the studies of Mason and Hallett (Mason 1997; Hallett 2000).86 In the German and Syrian excavations carried out in Raqqa before 2000, White Opaque Glazed Ware with cobalt blue decoration was not recorded. The pottery kilns recently excavated in the south of Tall Abou Ali provided a different picture by yielding some examples of this class (François and Shaddoud 2013, 37). Some fragments of White Opaque Glazed Ware with cobalt blue decoration and some sherds painted with green and blue were found in the area of the pottery kilns together with other ceramics (François and Shaddoud 2013, 37, pl. 25.
11-12). Due to the absence of wasters, we are not sure whether this class was manufactured in these kilns. However, these finds raised the issue of multiple production centres and the issues of imitation. These ceramics which were found in small quantities may have been employed as a model for the production of other glazed table wares.
Distribution
White Opaque Glazed Ware with cobalt blue decoration was distributed in the extent of the Abbasid Caliphate in the Near East including modern day Turkey, Syria, Iraq, Israel and Iran (Fig. 4.2.2a). This class was indeed recorded in Iran (Siraf, Istakhr, Nishapur and Susa).87 It was found in Iraq (Kish, al-Hira and Samarra).88 Sites in Turkey yielded some examples (Tarsus, Al-Mina, Antioch-on-the-Orontes).89 Finds were recorded in Syria (al-Hadir, Raqqa, Rahba, Qasr al-Hayr al-Sharqi and Hama).90 It was found in Israel (Caesarea) and in Jordan (Aqaba).91 It was present in India (Sinjan);92 in the United Arab Emirates (Jazirat al-Uyla and Kush);93 in Saudi Arabia, (Dahran);94 in Yemen (Zabid East);95 in East Africa (Manda and Shanga);96and in Thailand (Ko Kho Khao) (Fig. 4.2.2b).97
86 See: Lustrewares for the discussion of the work of Hallett and Mason.
87 Iran, in Siraf and Istakhr (Mason 1997, 25-26, 36); Nishapur, Susa (Wilkinson 1973 chapter 7;
Kervran 1977, fig. 43-1).
88 In Iraq, in Kish, al-Hira (Mason 1997, 36), and Samarra (Sarre 1925, 1).
89 In Turkey, in Tarsus (Day 1941a, 146), Al-Mina (Lane 1937, 31), Antioch-on-the-Orontes (Waagé 1948, 91, fig. 53: 1-11).
90 In Syria, in al-Hadir, (Rousset 2012, 99, fig. 6), Raqqa (François and Shaddoud 2013, 37), Qasr al- Hayr al-Sharqi (Grabar 1978, tome I, 114, tome II 242-243), Hama (Riis and Poulsen 1957, 127, fig.
387); Rahba (Rousset 1996,143).
91 Caesarea (Arnon 2008, 35-36); Aqaba (Hallett 2000, app. 1.nos: 44-46).
92 Sinjan (Priestman 2011, 94).
93 In the United Arab Emirates, in Jazirat al-Uyla (Kennet 1994, 192, fig. 12.23-26), and Kush (Priestman 2011, 108).
94 In Saudi-Arabia, Dahran (Mason 1997, 35-36).
95 In Yemen, in Huseiniyah, Zabid East (Mason 1997, 36).
