Shine Like the Sun APPENDIX C: ARCHAEOLOGICAL SITES AND THEIR POTTERY

Shine Like the Sun:
Appendix C: Archaeological Sites and their Pottery

INTRODUCTION
This Appendix provides a somewhat uneven account of the sites from which pottery was sampled for the present study, and sites that have important dating or other evidence. Some maps are provided for sites in which the account might be confusing, or which are otherwise unavailable. As noted in the section on conventions at the beginning of this book, all dates are regretfully given only in AD or Common Era unless otherwise signified.
ABU SUDEIRAH, Iraq (see Kish)
‘ACHARNEH, Syria
This site has recently been awarded to Michel Fortin of Université Laval (it can be transcribed from the Arabic of the excavation permit as Tell al-‘Asharna, according to the MESA system of transliteration). It is on the Orontes, just downstream from Shayzar. Trial excavations of this site, aimed primarily at the Bronze and Iron-Age levels, also produced evidence of significant architecture of the medieval period with associated pottery in primary contexts. The presence of the robust architectural remains may support the identification of ‘Acharneh as Tell ibn Macher, the site of a castle founded by Tancred of Antioch in 1111 to harass the Munqidhite occupants of Shayzar (Deschamps 1973; see also Runciman 1951, v.2:118). There seems to be no historical record of its abandonment, but the site would have been outflanked by the conquest of Afamiya by Nur al-Din in 1149.
The pottery from the medieval occupation provides a useful corpus of Syrian stonepaste-bodied Incised-wares and Polychrome-relief ("Laqabi") wares of Group One as presented in this volume (the group which includes the "Tell Minis" Lustre-ware type, see Chapter 5), and also some Fustat clay- bodied Underglaze-painted wares equal in date to Fustat Lustre-ware Group 3 or 4 (see Chapter 4). The dating of the Syrian wares is proposed as c. 1075-1125 and the Egyptian wares as c. 1075-1125 or 1125-1175. Four of the more diagnostic stonepaste wares have been sectioned and are of the "Tell Minis" Petrofabric, which provides further circumstantial evidence to link this group to western Syria. The Fustat wares are representative of a large group of this type found on Crusader sites, and in this case can be positively linked to production in Fustat.
ANI, Turkey
The ancient Armenian capital of Akhuryan on the right bank of the Arpa- ay River, which forms the modern border between Turkey and Armenia (Barthold 1960). The city became a key prize in conflicts between Muslim rulers, local Armenian or Georgian leaders, and the Byzantines, leading to its destruction by the Seljuk Alp Arslan in 1064. Ani was not, however, subsequently abandoned. Occupation continued, and n became an important fortress under the Georgians in the twelfth century. It is considered to have declined following Mongol conquest in 1239, and popular history claims the abandonment of the city after an earthquake in 1319, although coins struck in Ani are known as late as the fifteenth century. It is considered to be a possible site for the production of stonepaste ceramics (Rogers 1989). Pottery from the site was supplied for analysis by Esin Atil.
AQABA (al-‘Aqaba), Jordan
The early Islamic city of Ayla was a new foundation in the vicinity of the Byzantine city of Aila or Ailana, probably under the caliph Uthman ibn Affan in about 650. It became an important staging point in the Hajj from Egypt, North Africa, and Syria, and so was tremendously wealthy. During its history it has had to contend with a number of calamities, including sacking by the bedouin Banu Jarah in 1024, earthquakes in 748, 1068, and 1071, and final destruction by a Crusader army under Baldwin of Jerusalem in 1116. Subsequently settlement focused around the castle of Aqabat Ayla, later shortened to Aqaba (Whitcomb 1994).
Excavations since 1987 at Aqaba by Donald Whitcomb of the Oriental Institute, Chicago, have revealed the general plan of the oblong, 170 m by 145 m settlement, and five successive stratigraphic phases (Whitcomb 1987, 1988, 1989, 1991, 1994). The earliest, Phase A (650-750), represents first delineation of the city and is associated with Byzantine- influenced ceramics, particularly locally produced amphorae. Phase B (750-850) represents a period of major growth following the earthquake of 748, and is associated with the first imports from Iraq and Egyptian Semi-glazed ware. Phase C (850-950) represents orderly occupation associated with "Samarran" imports from Iraq. Phase D (950-1050) represents an intensification of occupation and narrowing of streets combined with midden accumulation, all associated with early Fatimid lustre-wares and Chinese qingbai and Yue imports. Finally, Phase E (1050-1116) represents desultory occupation following the earthquake of 1068. Of more specific chronologies, an unusually late tenth century date is argued for the Turquoise-glazed Hib type (Whitcomb 1991).
ATHENS, Greece
Excavations of an ancient cistern to the north of the Hephaisteion in Athens, reused in the Byzantine period as a waste dump, produced a sherd of Islamic pottery (Frantz 1938). It is described as being "imitation porcelain," with a hard white body and a white glaze and is described as the "base of an open dish with ring foot and wide shallow fluting on the outside",. The piece was considered to be Egyptian by Ernst Kuhnel (1934:432, n. 2), and I would agree with that identification on the basis of available evidence, the piece probably being Incised ware, possibly of Group One. It is said to be from a layer within the cistern containing extensive burning, and where two coins of Alexios I (1081-1118) were found, although fragments of a pithos from this fill join a vessel found nearby that contained a hoard of coins of Manuel I (1143- 1180).
BASRA (al-Basra), Iraq
The traditional account of the founding of Basra is that it was one of two major new settlements founded in Iraq on the orders of Caliph Umar b. al-Khattab during the early years of the Islamic conquest. Located first as a camp on some Sasanian ruins, al-Khurayba, at the junction of two canals drawn from the west bank of the Shatt al-‘Arab, the site retained the Muslim's preferred landward connection with the Holy Cities of Arabia, while being a convenient staging point for the expected penetration of Iran. The precise date for the foundation of Basra is probably between 635-6 (AH 14) and 637-8 (AH 16) (Donner 1981:229-30). Estimates for its population in the eighth and ninth century hey-day vary from 200,000 to 600,000 (Pellat 1960).
Details about the early configuration of Basra do not even approach those that exist for its sister city, Kufa, but there are some descriptions from the tenth century (Le Strange, 1905: 44-6). Flanking the western bank of the two canals, the city formed a semi-circle with a single gate towards the desert, the Bab al-Badiyah. Here the Mirbad quarter was a busy convening point for caravans. The most important monuments were also on the landward side of the settlement. These included shrines from the time of ‘Ali ibn ‘Abi Talib's residence, including those of Talhah ibn ‘Ubaydull h and al-Zubayr ibn al- ‘Awwam, Companions of the Prophet, who died during the first civil war between the Umayyads and the Hashimites in 656.
Following the growth of Baghdad under the ‘Abbasids, Basra lost some of its importance as an administrative centre, although it benefitted from its geographical location as the main port for Iraq on the Gulf. It is evident from the presence there of scholars of some repute that the city enjoyed considerable prestige in the ninth century (Pellat 1960). However, the city witnessed a number of violent upheavals, including the Zanj revolt of 871, and particularly the Qaramita sack of 923, which lasted 17 days.
By 1052 Nasir-i Khusrau describes Basra as populous, with the wall in good repair, although many parts of the town were in ruins (Schefer 1970:235-36). By the beginning of the twelfth century the city wall was two kilometres short of the previous one (Le Strange, 1905: 46). Ibn Batutta (in the mid- 14th century A.D.) described, as he often did, a city largely in ruins. Whatever the true merit of the remark, it is clear that the future settlement of New Basra was beginning to grow on the site of al-Ubulla, where the more southerly of the two canals forming the "Island of Basra" joined the Shatt al-‘Arab. Old Basra had ceased to be important. The name of the current settlement in the area of Old Basra, al-Zubayr, is taken from the shrine already referred to in the vicinity of the old Bab al- Badiyah.
Evidence for ceramic production at the site comprises a group of sherds and rod-shaped kiln furniture in the collection of the Metropolitan Museum of Art, New York (accession numbers 52.130.1 to .27; at some time subsequent to accession a piece was broken off one sherd, and this fragment was also numbered within the sequence, and some were renumbered; this is why the accession record states "26," while the accession numbers go up to 27). The record for this material attributes the finds to a Mr. N. M. Baker, and states:
On August 7, 1952, [he] gave to the Near East Dept. 26 pieces of pottery (fragments) and glazing material, from the site of old Basra, Iraq. This site is about 10 miles west of the present city of Basra. All these fragments are from a potters kiln, dating about 9-10th century AD.
Charles Wilkinson would have been the curator responsible at the time, and it is clear that he realized the significance of the material (Wilkinson 1973:180). Although there are problems with find-spot identifications provided by dealers, we have no reason to doubt Mr. Baker's information. As we can not be absolutely certain that the pottery was actually made in the kiln in which it was found, the strength of the material, and the analyses undertaken upon it, relies on the kiln furniture. The pottery consists primarily of bowls and small cups. These are not self-evidently wasters, but may have been broken during processing and left in the area of the kiln. Many of them do, however, have what might be called kiln detritus fixed to their glazes, including comminuted ceramic, ash, and carbon. Many entirely lack glazes, which is unusual for this type of material, and in others the glaze is seriously deteriorated, although this may be a result of post-depositional factors. Where the glaze does persist, it is most often of an opaque white type, with splashes of turquoise or purple. The "glazing material" mentioned on the card comprises five chunks of glass, one coloured blue, two green, and two turquoise.
CAESAREA, Israel (see Qaysariyya)
CORINTH, Greece
Exacations at Corinth produced significant amounts of Islamic pottery dated collectively to the eleventh and twelfth century heyday of medieval Corinth (Stevenson 1947: figs 147-150). These are consistently miss-identified as "Persian", but include at least one Basra Group Six lustre ware, some Fustat Group One lustre wares, underglaze painted wares equal to Lustre Group Four, and Incised wares which may be either of Egyptian or Syrian origin.
DHAHRAN, Saudi Arabia
Pottery from the vicinity of Dhahran on the Gulf Coast of Saudi Arabia forms part of a private collection to which the author has had access. The material is predominantly of early ninth-century date, comprising Blue-painted, Moulded, and Turquoise-glazed wares, with a few Iraqi underglaze-painted wares of the late fourteenth century. As such it would appear to represent a useful grouping, perhaps from a site with a short period of occupation.
FUSTAT (al-Fustat) and Cairo (al-Qahira), Egypt
The standard history has the foundation of Fustat in 641 as a military encampment built by the invading Arab Muslim army, for the seige of the Byzantine fortress of Babylon (Creswell 1969:19-23; Scanlon 1965; Kubiak 1987). It was located where a sandstone ridge nears the Nile, enabling construction above the floods and meandering of the river. The camp became an important Islamic administrative centre, and increasing settlement led to development of a series of administrative suburbs, stretching north-east along the ridge, and upwind from Fustat. The first of these suburbs was founded under the ‘Abbasids, then the largely autonomous Tulunid dynasty of governors, and finally the invading Fatimid caliphal dynasty founded Cairo in 969. During this time Fustat remained as the commercial heart of this urban complex, but suffered a series of calamities. This included the running street-warfare of al- Hakim's reign (in 1020), between his African corps, and the Berber and Turkish corps'; and a major famine in 1024. A famine lasting seven years, from 1065 to 1072 was so terrible that people were reduced to eating dogs, cats, and even each other (Canard 1965). During the collapse of the Fatimid caliphate in 1171 Fustat was allegedly burnt, and there was considerable street-warfare throughout. Subsequently Cairo became the centre of urban growth, leading to the modern city. Fustat remained as the port for Cairo, and was included in a great wall for the city built by the Ayyubid Salah al-Din. With the gradual westward shift of the Nile, and construction of port facilities downstream by the Mamluks, Fustat declined further, and by 1427 was described as ruinous. However, occupation persisted to some degree as it does to this day, even before the massive growth of modern Cairo.
After abandonment of much of the site, the remains were continually dug over in pursuit of building stone, phosphatic earth, and antiquities for the market (Kubiak 1987: 29-30). Many museums, including the ROM, the Ashmolean, the BM, and the MMA have collections from these diggings, which were available for study. First excavations at the site were under Bahgat (1914), and were able to define the unplanned nature of the street-pattern, at least within the confines of the excavations, but provided little in the way of stratigraphy. Amongst more recent excavations are those by the American Research Centre in Egypt (ARCE), directed by George Scanlon. These have also made a contribution to knowledge of urban planning, but only a very broad stratigraphic ceramic sequence has been presented in the published reports to date. Material from the ARCE diggings are in the collections of the ROM and Fogg museum (currently stored at the ROM), but no stratigraphic information accompanied the finds. Many types that are not of art-historical interest, such as those of the Lead- glazed classes, are represented in the corpus only by material from the ARCE site.
GENOA, Italy
Excavations at the convent of San Silvestro produced a number of pieces of Islamic pottery, the pieces of concern here being attributed to Syria ("Raqqa wares") by the excavators (Pringle 1977, 1978). Two of these are illustrated, from Phase M, dated to 1170-1200 by the excavators (Andrews and Pringle 1977). The first (Pringle 1977, no. 168) has an opaque turquoise glaze and a friable white body, with a form closely parallelled by Egyptian Incised wares of Group Two. The second (Pringle 1977, no. 169) has an "azure" opaque glaze on the interior, with transparent green on the exterior, and a broad flat rim. Again I have noted similar pottery among the corpus of Egyptian Incised wares.
GHUBAYRA, Iran
A representative collection of pottery from these excavations (Bivar and Fehérvári 1974) is held in the ROM. Pottery sampled includes various twelfth to fourteenth century types, including Lustre-painted wares and an extensive corpus of monochrome Incised wares. Later pottery includes some buff clay-bodied, Underglaze-painted wares, which may be of late Muzaffarid occupation, but one is of "Cizhou" style, which may be dated to the latter part of the fifteenth century (Mason 1996).
GHULAYFIQA, Yemen (also Ghul faqa)
A port site on the Red Sea coast of the Yemen, Ghulayfiqa was studied during field work by Ed Keall (1983; see also Stone 1985:35). It appears to have been important during the twelfth and thirteenth centuries judging by the imports found at the site, many of which were thin-sectioned.
GURGAN, Iran
The site of old Gurgan, present-day Gunbadh-i Qabus, was an important city during the twelfth and thriteenth centuries, as attested by the surviving monumental gunbad itself. Being largely abandoned, the site has been the source of many finds of pottery, especially Lustre-wares, and was considered to have a unique locally produced style of Lustre-ware(Bahrami 1949). Evidence of production of stonepaste wares certainly does appear to exist (Bahrami 1949, pl. 8, left). Further evidence of production at Gurgan was provided by excavations by Kiani (1984), predominantly in the form of kilns. Pottery available for analysis from Gurgan comprises a group of sherds in the Ashmolean Museum, Oxford.
HALABIYA, Syria
This is an important Byzantine site, and fragments of stonepaste wasters of twelfth-century date were apparently found at the site by a respected colleague. Unfortunately he has completly forgotten anything useful about these samples, and only the handwritten "found at Halabiya near Deir az-Zor" on the envelope supports the provenance. These were subjected to petrographic analysis and form a distinct group.
HAMA (Hamat), Syria
Hama is a very important site at which major excavations were undertaken by a Danish team (Ingholt 1934; Riis and Poulsen 1957). Unfortunately, the site appears to have been riddled with pits which posed an intractable problem for the excavators; the published reports are not entirely clear about what findings were obtained; and the pottery isfor the most part published without profile drawings or photographs of the vessel reverse. Hence, what was an important and extensively excavated site is of very limited use.
HARRAN, Turkey
Fragments of what that would appear to be Syrian stonepaste- bodied Group Seven and possibly Group Six were excavated at the Harran citadel (Rice 1952, fig. 14.2, 12), and have been used towards dating of the pottery elsewhere (Northedge et al. 1988:102).
Initial survey identified four distinct phases in the citadel (Lloyd and Brice 1951), with the final phase being Crusader modification attributed to the final years of the County of Edessa (1098-1146). Consideration of the historical information suggests a far more complex picture (Rice 1952). Excavation of the citadel's southeast gate revealed an original construction dated by inscription to 1059, with a subsequent rebuilding of the gate, and a final enclosure of the ruined gate within a wall. The relevant pottery was found in the debris deposited against the second phase, and is presumably thereby sealed by the later wall. Rice (1952) suggests that the debris is from an earthquake, and two such occurrences are recorded, in 1114 and 1157, while extensive rebuilding is recorded for 1191. However, the citadel is also recorded as having been the subject of siege by the Mongols in 1259, while the entire town was depopulated and destroyed by them after 1271. The last piece of epigraphic evidence from the citadel is a Mamluk inscription dated by Rice (1952) to 1315. Rice prefers to see the last phase as essentially Ayyubid despite the Mamluk inscription, but little remains of the wall sealing the southeast gate, and it would not be possible to argue that this cannot be Mamluk. A terminus ante quem of 1157 (the date of the earthquake) for this pottery would be untenable unless the published pottery sketches are completely misleading; hence it may be preferable to attribute this destruction to 1259 (the date of the siege).
HIRA (al-Hira), Iraq
In the fifth and sixth centuries, Hira thrived as the capital of the Arab Lakhmid dynasty, the seat of a bishop, and a major trading centre, with ships from India and China discharging their cargoes. The Lakhmids exercised considerable authority, not just locally, but also in the Sasanian court. In 602 the last Lakhmid was replaced by a Persian governor. Perhaps because of this imposition, Hira capitulated to the Muslims in 633 without any show of resistance. The historian al-Baladhuri states that the Muslim invaders were able to ride around the site on horseback, which seems to suggest the open nature of the site as suggested by excavation. The establishment of the new centre of Kufa rather eclipsed Hira, and occupation is thought to have been abandoned in the ninth century.
The ruin field as it existed at the time of excavation was a large area comprising numerous small mounds, rather than one large tell. Excavations by a team from Oxford in 1931 were financed by Gerald Reitlinger, and directed by Reitlinger and D. Talbot Rice (Rice 1932, 1934). Trenches dug into the small mounds and intervening areas indicate a number of larger structures, with small houses scattered in between. At Mound I a large fortified building was uncovered, with Sasanian foundations, but final occupation was dated by coins of 769, 773, and 783, plus a further twenty unidentifiable coins of the eighth century. At the bipartite Mound III-IV, a number of smaller but well-constructed courtyard houses with elaborate carved plaster were excavated; they contained coins of the eighth century, including one of 779 and another of 762. The settlement appears to have gradually moved towards the east, and so was not all occupied at the same time.
Glazed pottery from these excavations, predominantly Opaque-glazed, Turquoise-glazed, and Lead-glazed wares, are currently stored in the collection of the Ashmolean Museum. Unfortunately, no information exists regarding their provenance within the excavations, so it is not possible to state categorically that the pottery was associated with the coins described above. However, the weight of coin evidence of quite a narrow time-span suggests that the bulk of Islamic finds are of a similar date.
ISTANBUL, Turkey
The city of Istanbul was, of course, Constantinople, the capital of the Byzantine Empire, at the time we are concerned with in this book. Two excavations are of ineterst to us here, those at Sarachane and those at the Palace of the Emperors.
Sarachane
Excavations at Sarachane provide a group of stonepaste wares from the Islamic world, described by John Hayes (1992). They generally have a fine paste relative to wares Hayes saw in Cyprus considered to be Syrian, and he suggests an Iranian origin. However it is probably more likely that the Cypriot wares were a coarser or more porous-bodied Syrian product, perhaps from Damascus, and the bulk of the Sarachane ware appear Syrian. They are found in a series of deposits all stratigraphicaly sealed to varying degrees, and can be dated by associated coins and the relatively well-dated local ceramic sequence.
Deposit 51: A dump inside a twelfth century cistern, circa 1150-1175. Finds include No. 51.1, a Laqabi ware (Hayes 1992, pl. 9g) with rather a wide rim and tall foot for Syrian Group 1 and probably closer to Group 3 or later. Find No. 51.2 is incised with a colourless glaze (Hayes 1992, fig. 81) and has a similar form to No. 51.1. Find No. 51.3 is incised with a colourless glaze (Hayes 1992, pl. 9h) and looks like a good Group 1 or 2 form. Find No. 51.4 is incised with a turquoise glaze (Hayes 1992, pl. 9i) and appears to be a Syrian Group 1 or 2 conical bowl. A few further "scraps" of a similar nature were found in this deposit.
Deposit 52: This overlies Deposit 51, dated to circa 1200, and contains a coin of Manuel 1 (1143-80). Find No. 52.1 is colourless with radial blue stains; No. 52.2 has a turquoise glaze and may be a plate similar to those of Fustat Lustre Group One, but appear unknown as incised; No. 52.3 is a base fragment, possibly part of No. 52.1, and is colourless with blue stains and relief similar to Fustat Group 2; collectively over ten sherds.
Deposit 53: This context is sealed by church destruction fills, dated to late twelfth century. Find No. 53.1 has a colourless glaze and has a very similar form to No. 51.1; also a further eight fragments with turquoise glaze, two blue glazed, one colourless/white (Hayes 1992, fig 16.8), and three green coarse bodied wares (Egyptian?) with incisions.
Deposit 55: This context is dated to circa 1200 or slightly later, but includes much survival material from eleventh century and some sixth to seventh century. Find No. 55.1 has a "rather rough-textured light yellow fabric" and a thick dark leaf-green glaze, is incised, and has very close parallels to some Egyptian Group One lustre-ware forms although I haven't seen any incised wares. A short flat rim frag may belong to this piece (Hayes 1992, pl. 9f) and would also appear suitable to Fustat Group One or Two. Also two further similar fragments; and one "normal fine fabric" with a colourless glaze.
Deposit 56: This context is the post 1204 destruction levels in narthex and at west end of church. Find No. 56.1 has an off-white to pale green glaze, foot may be related to Egyptian Group 3 but precise parallel of rest of form and rim. Also further two blue glaze releif decorated, three white or colourless glazeed wares, and four blue glazed.
Deposit 60: This context represents early 13th century destruction fill and robber pits. Find 60.1 is green incised and not sufficiently diagnostic; it could be Egyptian or Syrian
Generally the Sarachane pottery would appear to be earlier than the contexts would suggest. This is particulalrly odd in that many pieces would appear to be from primary contexts, in that they are large parts of vessels. Given that this is a church, it may be suggested that the ceramics had a function in the church similar to that in many Italian churches, that is, as architectural ornament. Hence their date of incoproration may relate to a date of construction rather than destruction of the building.
Palace
Excavations at the site of the palace of the Byzantine Emperors took place in the 1930s (Stevenson 1947). Finds include two nearly complete turquoise carved relief vessels with wide flat rims (Stevenson 1947, pl. 26, figs. 3-4). One of these pieces (Stevenson 1947, fig. 3) seems closely related to the Sarachane Laqabi bowl, and was found in Stage V (Stevenson 1947:46- 56). This stage includes coins of Constantine X (1059-67), and what the excavators consider an "intrusive" coin of Alexius III (1195-1203). Second piece was in level that included coins of Manuel I (1143-80) and Alexius III (1195-1203), while a coin of Manuel I (1143-80) was in the same context as the Islamic vessel. Locally-made vessels were often in a similar state of preservation and were considered typical of this time-range.
ISTAKHR, Iran
The site of Istakhr, in Iran, contains a significant assemblage of Iraqi pottery (Schmidt 1939:105-19). Apart from contributing some key pieces to the typological study, individual context assemblages from this site strongly support the sequence suggested for that typology. This material is stored at the Oriental Institute, Chicago.
JERUSALEM (Israel/Palestine)
The relevant material from Jersualem is from excavations in the Armenian Garden in the southwest area of the city. The excavation report (Tushingham 1985) provides a sequence of occupation; the strata of pertinence to this study are those following the Byzantine occupation. There is evidence of occupation immediately subsequent to the Byzantine phase, including pottery and coins of Umayyad and ‘Abbasid date, but the excavators consider the area to have been subsequently abandoned. During this period of abandonment it appears that parts of the city wall and some of the site were destroyed by surface wash. At some later date the wall was repaired, the washed-out gullies were filled, and in the Ayyubid period buildings were constructed. These comprise a series of small structures next to the wall, a narrow road parallel to the wall, and a larger complex of buildings on the other side of the road which appear to have relied on earlier Byzantine structures. There was some continuity of occupation, and then another period of abandonment in which the city wall again appears to be missing in places. Following this the building complex was occupied again for a short time during the Mamluk period.
There is a wealth of coin evidence from the excavations, with contexts excellently recorded in the Wheeler-Kenyon method (the excavations having been begun by Dame Kathleen), and fully reported by Doug Tushingham (1985). The publication, however, seems to rely more strongly on historical evidence that relates to the city wall, citing the record of construction by Salah al-Din in 1191-1192 and al-Mu‘azzam ‘Isa in 1212-1214, and al-Mu‘azzam's subsequent de- fortification of Jerusalem in 1219 and 1227. Tushingham thereby tries to put the entire Ayyubid occupation into the short period of 1212 to 1227. However, it may be noted that elsewhere in the area of the Armenian Garden (smaller trenches were opened to the north and south of the main excavation) the city walls remained intact. Hence, the missing part of the wall may have been very small and easily patched at any time, and not necessarily of sufficient note to have an historical reference. If one relies solely on the coin evidence (extracted from a correlation of coin lists and context numbers) then it may be noted that the only coins found in the make-up fill are of the Zangid Nur al-Din Mahmud (1162-1175). Further, it appears that most of the coins directly associated with the "Ayyubid" occupation are Zangid, with a large proportion of the Ayyubid coins actually originating in the Ottoman-period garden make-up, which was probably brought in from off-site. Some Ayyubid coins were found sealed in occupation layers, including one of Salah al-Din (c. 1190) another of al-Aziz Uthman (1193-1199), found in a pit sealed by a floor, and another of al-Adil abu Bakr (1199-1218), found in the sediment at the bottom of a cistern together with a Crusader coin of 1152-1187; none of Mamluk date were found. The coin evidence certainly supports Tushingham's attribution of the bulk of this occupation to this period, yet it is difficult to accept that the first re-use of this area was not in the mid-twelfth century.
Following this re-examination of the evidence, the Jerusalem excavations suggest a dating for the relevant ceramics to c. 1150-1200, with these wares being in an abraded state of tertiary deposition at the time of the make-up associated with Zangid coins and prior to the time of the Ayyubid occupation after 1191.
KISH, Iraq.
Three mounds in the area of the ancient city of Kish were examined in 1930-1931 as part of an Oxford University and Chicago Field Museum expedition directed by L. C. Watelin (Reitlinger 1935). These are called Abu Sudeirah, Sha‘al Ghazna, and Tell al-Su‘aidan.
Abu Sudeirah received most attention. It was a long tell, up to 11 m high and almost a kilometre in length, and formed the easternmost site of a complex of mounds associated with the site of Kish. This was originally constructed along the banks of the Shatt al-Nil canal. A trench into the side of the site revealed seven layers of occupation upon possibly Babylonian foundations, which explained the presence of inscribed baked bricks of Nebuchadnezzar on the surface of the tell. The bulk of the tell is of Islamic date, with occupation judged to range from the late tenth to the fourteenth century, based on comparative pottery from other sites rather than any internal evidence. Abu Sudeirah does however provide stratigraphic relationships for the pottery.
Sha‘al Ghazna, also in the area of Kish, and Tell al- Su‘aidan, a site 13 km distant, received smaller trial trenches. Both revealed Babylonian buildings, reoccupied in the eleventh century, again having been dated, it appears, by ceramic analogy.
The earliest Islamic levels were dominated by Slip-incised wares. The generally earlier "Uberlaufglasuren" style, which is incised more deeply, is more symmetrical, and is characterized by "acanthus-scroll" motifs and fish-scale panel filling, is scarcely represented. The "Uberlaufflecken" group has weak undecided scratching, the whole surface being covered with meaningless scribbles. In general both were found contemporaneously. Very few examples of Lustre-ware were found. The next phase of occupation is dominated by early Underglaze-painted pottery.
KO KHO KHAO, Thailand
This is one of a number of sites on the Isthmus of Kra in southern Thailand which appear to have been occupied for a very short period at the beginning of the ninth century. The excavators, from the Field Museum, Chicago, postulate that for some reason the normal sea route from the Indian Ocean to China was obstructed, requiring a short overland route (Bronson and Ho 1991). Finds include extensive amounts of Chinese wares, upon which the close dating of the site is based, and large amounts of Islamic blown glass, an exotic commodity in China. Of more relevance to this monograph are the few fragmentary yet diagnostic pieces of Basra Blue-painted ware.
KWAGANDAGANDA, Natal, South Africa
This site near Durban comprises an Early Iron Age settlement dated by radiocarbon determination to the early seventh to eleventh century. Excavations here produced a small sherd of Turquoise-glazed ware securely dated to the ninth century (Whitelaw 1994).
LASAIL, Oman
Lasail is a copper extraction and processing site of some antiquity in the Omani highlands, with an Islamic component (Whitcomb 1975, Weisgerber 1981). Pottery from the site was collected by Ed Keall and is presently stored in the ROM.
MA‘ARRAT AL-NU‘MAN, Syria
The significance of this site may have no bearing on the pottery claimed to be from the region. By this I refer to a group of pottery fragments donated to the Ashmolean Museum in 1980 by Dr. H. Bartels. These are are said to have come from "Tell Minis, near Ma‘arrat al-Nu‘man", but it has become tradition at the Ashmolean to call them the Ma‘arrat al-Nu‘man sherds. Where they actually came from will probably never be known.
MADINAT AL-ZAHRA, Spain
This city was founded by the Umayyad caliph of Cordova, Abd al-Rahman III, in 936. During a civil war the city was sacked and destroyed in 1010, apparently being used subsequently only as a source of building stone (Jimenez 1986). If these dates are reliable, they usefully bracket the finds of later Basra Lustre-painted wares found there (Ettinghausen 1954).
NIPPUR (Niffar), Iraq
This ancient site had a final Islamic phase that was abandoned in about 800 judging from the total lack of any coins on the site post-dating this (E. J. Keall and M. Gibson, personal communications). Islamic and pre-Islamic pottery has been recovered from the recent excavations at Nippur by the Oriental Institute; these finds are stored in the ROM.
NISHAPUR, Iran
A very important city in the Khurasan province of north-east Iran, Nishapur was the site of extensive fieldwork by Charles Wilkinson (1973). The site is particularly noted for the lead- glazed wares of the ninth to eleventh centuries, but generally little or no new dating evidence is provided for any of these wares. Stonepaste bodied wares are known to have been produced at Nishapur, and Wilkinson suggests a very early date.
The site at the "East Kilns" revealed three kilns all associated with the production of "Alkali-glazed Ware" which includes wares with stonepaste bodies. The largest of these kilns contained peculiar "shafts" which at least in part reflect an earlier phase of the kiln (Wilkinson 1973:xxxvii-xi). Fragments of lead glazed wares of the ninth to eleventh centuries, including wasters, were found at the "bottom" of the shafts. In one of these shafts (the exact position does not seem to be reported) were found three silver coins, one of Mas‘ud I of Ghazna (r. 1030-41) and two of Caliph al-Qasim (r. 1031- 1074). If these coins dated the stonepaste-bodied wares made in the kiln it would make them extremely early, but of course what it really does is provide a terminus post quem. I would argue that the kilns had been modified from earlier use for lead-glazed wares. It might, however, be argued that a kiln is unlikely to be in use for about one hundred and fifty years, which would be the time between these coins and the more commonly given earliest date for the introduction of stonepaste into Iran. A probably more relevant but less evidenced terminus ante quem would be provided by Wilkinson's argument that the kilns were abandoned after the earthquake of 1145 and sack by Ghuzz Turks in 1153 (Wilkinson 1973:263). Following these events the inhabitants of Nishapur were resettled in the former administrative suburb of Shadyakh (Wilkinson 1973:xxviii), which would put the kilns over 2 km east of the city. Hence, although I would argue that Wilkinson's evidence does not seriously indicate a date before 1100 for the production of stonepaste wares in Iran, it does suggest that these wares were made significantly before 1145-1153.
Pottery most commonly associated with these kilns are stonepaste-bodied with alkali glazes, coloured cobalt-blue or copper-turquoise, or which are colourless with streaks of cobalt-blue. They are often releif decorated with incision or moulding. Profile drawings of the East Kiln pottery finds are not provided, but decoration of one (Wilkinson 1973, cat. no. 11.54) is comparable to Kashan Group One. Profile drawings are provided of a few whole vessels excavated elsewhere in Nishapur, and those which most closely resemble the kiln finds (Wilkinson 1973, cat. nos. 11.1, 11.3, 11.33) comprise cono- segmental bowls and flat-rimmed dishes equivalent to those of Kashan Group One. Some undrawn but photographed whole pieces also appear equivalent (Wilkinson 1973, cat. nos. 11.35-11.37). Although I have had access to the pottery from Nishapur stored at the MMA, as permission to sample for petrographic analysis was not provided I decided not to undertake an incomplete study.
The only pottery subjected to petrographic analysis from Wilkinson's fieldwork comprises a group of sherds from two fifteenth or early sixteenth century kilns in the area of the town. Pottery recovered from these kilns is presently in the collection of the MMA. The accession information for the sherds states that one kiln was "near the river", and the other was "near the road". From the "river" kiln is a waster of a clay-bodied bowl, while late fifteenth or early sixteenth century stonepaste wasters were recovered from the "road" kiln (Mason 1996). These samples provide characterization for a substantial body of fifteenth- and sixteenth-century pottery (Mason et al. 1996), while extensive production of clay-bodied wares from earlier periods can be correlated to these by analysis of wares of the suitable type. However, none of the evidence of extensive stonepaste production in the twelfth century uncovered by Wilkinson was made available for analysis, nor has any sample of this type been found to be the same as the later petrofabric.
PAVIA, Italy
Excavations at the Torre Civica have produced a number of pieces of Islamic pottery (Blake 1978). Illustrated fragments are from Phase X, dated by the excavators to the late eleventh century. Unfortunately, the forms are largely undiagnostic, monochrome wares with stonepaste bodies. One, however, has been tested for lead by T. Mannoni, and it contains this element in significant quantities. Hence, it must be either Egyptian or Syrian Group One. The date would be suitable to either one, but would provide valuable evidence for dating of the Syrian group if so attributed.
QAYRAWAN (al-Qayrawan), Tunisia
The Friday Mosque of Qayrawan contains a number of interesting lustre-painted tiles. These comprise polychrome wares of styles directly comparable to those best known at Samarra', and monochrome lustre of styles elsewhere known only from Susa (Marcais 1928). The installation of tiles has been dated to 862-863 by reference to a historical document (Marcais 1928:9-13, Creswell 1969: 313-14); although ambiguity within the text and its lateness (the writer, Ibn Naji, died in 1494) suggest caution, it is still a date worth considering. The document states that tiles were brought from Baghdad, and that additional tiles were made by a man from Baghdad, presumably at the site. The designs and motif assemblages of the polychrome tiles are suitable to Group Three of this study. The monochrome tiles are Group Four (BOG4b) although they are in the second subgroup, that which does not show "ancestry" to Group Five. If the Qairouan date reflects a stage when Group Three material was still available, but Group Four was current, then this would suggest a transition from Group Three to Four in about 850.
QAYSARIYYA, Israel (also Caesarea)
This city existed essentially because, for economic and strategic reasons, Herod the Great wanted one there. Hence an enormous artificial harbour and city were built and named after the Roman Emperor between 22 BC and 9 BC. Although considered depopulated shortly after the Muslim conquest, the existence of the artificial harbour continued to make the site significant and it was maintained as a ribat or border settlement. Eventually it regained its role as a prosperous port and agricultural centre, but this would end with the Crusader conquest of 1101, in which all the men were massacred and the women sold into slavery. The site eventually became a prosperous Crusader city, but was finally conquered and destroyed by the Mamluk sultan Baybars in 1265.
The largely abandoned site became a focus of fieldwork for archaeologists from Europe, North America, and Israel. Pottery from the site has provided useful dating evidence and some has also been made available for analysis, including possible examples of Egyptian Semi-glazed wares considered by the excavators to be "probably . . . ninth century" (Holum et al. 1988, fig. 157). The division between pre-Crusader and Crusader levels provides a very useful horizon, as in many sites in the region. Pottery from pre-Crusader levels includes a number of Egyptian Lustre-painted Group One and Two examples (Brosh 1986, pl. V.3). Pottery from Crusader levels includes numerous Underglaze-painted wares equal to Lustre- wares of Egyptian Group Three and particularly Group Four (Pringle 1985; Holum et al. 1988, fig. 163, centre; Brosh 1986, fig. 3.12-13, pls. 2a-b, 3-4; Blakely 1987, fig. 46).
RAQQA (al-Raqqa), Syria
The Hellenistic foundation of Nikephorion was renamed al- Raqqa after the Islamic conquest (see Fig. C.1), and rapidly became an important regional capital (Meinecke 1994). Under the early ‘Abbasids a new massively fortified city was created, historically dated to between 770 and 775, and called al-Rafiqa (the Companion). The combined size of Raqqa and Rafiqa was probably surpassed only by Baghdad, which made it a logical location for the capital of Harun al-Rashid from 796 to 808. Under Rashid an extensive palace complex of 10 km2 was constructed to the north of Raqqa/Rafiqa, almost entirely made of mud-brick. Following the death of Rashid, the capital was switched to Baghdad, and the palaces were largely abandoned. Raqqa/Rafika continued as an important regional centre, and palatial residences were constructed in the area between the two cities. Following 942 supremacy in the region switched to Aleppo and Mawsil, and Raqqa declined considerably, with a major reduction in population, including the complete abandonment of the ancient foundation of Raqqa, and the assumption of the name "Raqqa" by Rafiqa. Following conquest by Imad al-Din Zangi in 1135, and subsequently under the Ayyubid dynasty, Raqqa recovered some importance, becoming once again a regional centre with associated building, although occupation seems to have been largely restricted to the eastern half of the ‘Abbasid city. This is the period of production of the so-called "Raqqa" wares, with excavated kilns partly within the city wall to the south of the ‘Abbasid Great Mosque (Sauvaget 1948). The city is reputed to have been destroyed in the Mongol invasion of 1259, but perhaps more importantly the site was abandoned as a defensive station by the Mamluks in 1265 for strategic reasons. In the sixteenth century Raqqa was restored as a regional administrative centre by the Ottomans, but occupation was highly restricted. It was again abandoned in the seventeenth century. Finally the site was once more restored as a regional capital in the nineteenth century, and has grown until it effectively covers the entire site at its medieval maximum.
Excavations at Raqqa have been going on since 1944, and have been aimed chiefly at the palace buildings, with a sustained campaign by the German Archaeological Institute in Damascus under Michael Meinecke from 1982 (Meinecke 1994, 1989). Specific excavations have included the Western Palace, a residential building; the North Complex of barracks; the East Complex of recreational space; and the Eastern Palace, rese rve d for "re pre sent ativ e' pur pos es. Exc avat ions hav e also bee n und erta ken at Tell As wa d, one of a stri ng of mo unds with large piles of ashes spread along the northern perimeter of Raqqa. Numismatic evidence apparently points to use in the time of Harun al-Rashid.
Pottery samples from these excavations that were available for study and analysis include a selection of sherds from the Tell Aswad excavations, all dated to c. 800, and pieces of the twelfth century, including "Tell Minis" and "Raqqa" styles, the latter including wasters from the area of the ‘Abbasid Mosque, the Western Palace, and the East Complex. What these wasters were doing in what the published evidence states to be an abandoned area some two kilometres from the published kiln- sites of the period is unknown. Conceivably, presently unknown kilns of the twelfth century are also situated in this area, although presumably the quartz source would have been cobbles from the Euphrates flood plain, so the known site would be more logical. Of course the palace sites may have received rubbish deposits, but it seems a long way to go to dump some wasters. There are, however, other rationales for locating kilns, such as being downwind from the settlement, or perhaps from the residence of the Ayyubid prince al-Malik al- ‘Adil Abu Bakr, who lived in the city from 1201 to 1228.
RAYY (al-Rayy), Iran
The ancient site of Ragha, this site is located in an important geographical zone where agriculture is possible between the Alborz Mountains and the Dasht-i Kavir desert, and routes to passes through the Alborz cross the great Khurasan Road. Improved well-development allowed the modern capital of Iran to grow slightly uphill from the site, and the ruins of Rayy are covered by Tehran's southern industrial suburbs. This geographical location led to considerable importance in the Islamic period, but sectarian violence and then the Mongol invasion resulted in the site becoming largely abandoned (Minorsky 1995).
Rayy was the source of a great deal of ceramics and other materials coming onto the art market, and thus attracted the attentions of archaeologists. Excavations in 1934 and 1936 were undertaken by a joint expedition of the universities of Philadelphia and Chicago (Keall 1979). Unfortunately, nothing from the excavations nor any subsequent work with the finds has produced any useful chronological information for Islamic pottery. However, the ceramics, presently stored at the University Museum, Philadelphia, provide a useful corpus of ceramics from the region, including the most extensive collection of Iranian Lustre-wares and the only Mina'i from excavations.
Principle areas of excavation (see Fig. C.2) include Chesmeh Ali (pottery with RCH prefixes in Appendix A, the waster was probably also from this area), the "Government- Quarter" (RG prefixes), and the Citadel.
SAMARQAND, Uzbekistan
This important city is situated in the fertile basin of the Zarafshan River, often called Soghdia. At this precise location routes from India, China, and Iran all intersect, making Samarqand one of the greatest cities of the Silk Road. It formed the regional capital at least since the days of Alexander, when it was called Maracanda, and throughout the earlier Islamic periods it was important, often being the capital of a much larger region. It appears to have suffered some depredations under the Mongols, but returned to pre-eminence under Timur at the end of the fourteenth century (Schaeder [and Bosworth] 1995). The Timurid rebirth of the city took place to the south of the tell of the earlier town, which is now called Afrasiyab.
Evidence of clay-bodied ceramic production from the eighth through to the thirteenth century is well attested from Afrasiyab (Shishkina and Pavchinskaja 1992:31-72); it is also suggested that there was early stonepaste ceramic production (Rogers 1989: note 3). Characterization of the products of Samarqand is based on analysis of pottery of the early fifteenth century. Evidence of ceramic production at Samarqand in this period includes archaeological traces of production (Pugachenkova 1950:94) and a reference to Damascus potters at Samarqand by the Spanish envoy to Timur's court, Ruy Gonzalez de Clavijo: "Trade has always been fostered by Timur with a view of making his capital the fairest of cities. . . . Thus from Damascus he carried away with him . . . the craftsmen in glass and porcelain, who are known to be the best in the world" (Clavijo 1928: 287-88). Although no wasters were available for analysis, sherds and significant parts of vessels of both stonepaste and fired clay were available from excavations or reliable find-spots in Samarqand, including Afrasiyab and the Citadel (Mason 1996). The overwhelming supremacy of one petrofabric group for each ceramic body type among the ceramics at Samarqand leads to their tentative but confident attribution to this site. Only one fragment of an early slip-painted ware of a style associated with Samarqand was available for analysis, and this was found to be identical to the petrofabric defined for the fifteenth-century wares. No early stonepaste wares have been found to be identical to the fifteenth-century stonepaste petrofabric, although two sherds were available for analysis (see Appendix A).
SAMARRA', Iraq
This site has long been recognized as the seat of the ‘Abbasid Caliphal court and administrative centre of the vast ‘Abbasid empire between 836 and 892. Although no stratigraphic evidence was presented, the excavators of Samarra' (Sarre 1923) claimed that their excavations were only in levels occupied during the caliphal residence. This dating was followed by many subsequent authors (e.g., Lane 1947), but was questioned by Miles's study of coin evidence (Miles 1954), which indicated a mint in operation until 953. This led to a new trend of questioning the whole dating of Sarre's finds. In retrospect it may actually seem somewhat absurd to assume that Sarre was ignorant of occupation at other periods across the 40 km long ruin field, particularly given the well-published historical evidence (Le Strange 1905), and the presence of a modern town within the confines. Despite this newer trend, an open mind about Sarre's assertion that the material came only from buildings occupied during the caliphal residence has been maintained by the author, particularly as this assertion is endorsed by Adams (1970).
All the Samarra' pottery that has been studied and sampled for the present study is from Sarre's excavations, which was dispersed to various museums after the First World War, including the ROM and the BM. The results of this study would seem to support Sarre's dating. Apart from attribution of the finds generally to the caliphal occupation, Sarre distinguished between two Slip-incised or "sgraffito" styles, an early "Uberlaufglasuren" and later "Uberlauf-flecken" (Sarre 1923:76-77).
SIRAF, Iran
Following Whitehouse (1968, 1969, 1970, 1971, 1972, 1974), it can be usefully repeated that the earliest historical reference to Siraf and the India trade is to be found in the mid-ninth century geography of Ibn al-Faqih. According to Ibn al-Balkhi, the early tenth century was a period of great prosperity for Siraf. However a major earthquake in 977 may have caused its merchants to locate elsewhere, and after the fall of the Buyid Dynasty around 1055 most traffic diverted to Qais. In 1218 Yaqut delivers his usual sign of disapproval by describing the site as ruinous, with impoverished residents. During the hey- day of Siraf as an international entrepot on the Persian Gulf, its merchants were recorded in the Red Sea, East Africa, Madagascar, India, Malaya, and China.
The site identified as Siraf occupies a shallow south-facing bay, 4 km across, terminated by low sandy deltaic wadi deposits. The coastal plain is about 500 m wide, and is bounded on the north side of the site by a sandstone ridge. Traces of settlement are found all along the shore, but excavations were essentially in the western half. Although the region is rather arid at present, a survey of the hinterland showed that it could have maintained considerably more cultivation in the past. Seven hundred hectares may have been cultivated in the vicinity of Siraf, at least 72% probably irrigated by aqueducts. However, Whitehouse (1974:2-5) considers that this would still not have made the city self- sufficient in farm crops, and that there would have been a need to bring in food-stuffs from the highlands of Iran. The town also received water from the hinterland via aqueducts (Whitehouse 1972: 68).
Pottery from Siraf has been studied in some detail, with close studies of the material in the ROM and Ashmolean Museum find allocations, and briefer but thorough examination of the BM allocation.
Siraf sites
Site A, Sounding: Excavated in the first season to obtain a stratified pottery sequence (Whitehouse 1968:5-9). The earliest deposits consist of about 1 m of earth and sand, containing Hibs and stoneware. The second phase constitutes four subphases of a building with associated Opaque-glazed types: phase 2a with introduction of Plain and Blue-painted; 2c with turquoise or brown Splashed, and 2d with turquoise and brown Splashed. This is followed by an abandonment phase with squatter occupation and agricultural soil. Finally, domestic deposits appear, probably from adjacent buildings of fourteenth to fifteenth century date.
Site B, Congregational Mosque and Sasanian Fort: Excavated in first to sixth seasons inclusive (Whitehouse 1968:9-11, 1969:41-48, 1970:2-9, 1971:2-5; Sasanian fort and outer enclosure in 1972: 70-71, 1974: 7-9). The earliest levels of this site comprise a substantial fort, probably approximately square and about 62 m across, with a gatehouse flanked by semi- circular headed towers, elliptical towers at the corners, and a ditch on the landward side. Also included in this phase is the Outer Enclosure, a complex of buildings soundly constructed but in disorderly arrangement, delimited on two sides by a 1 m thick curtain wall. Five phases were discerned in the Outer Enclosure, 1 and 2 considered Sasanian, while phase 4 floors seal lead coins of eighth or very early ninth century. Excavations under the Fort revealed earlier structures and types of pottery also found in the Fort and Outer Enclosure.
Above the Fort and Outer Enclosure was the Congregational Mosque, also consisting of five phases. The first consists of the original hypostyle mosque with a minaret next to the main entrance. Fill for the platform of the first phase contained Islamic coins of eighth century style, and copper and silver coins, the latest c. 780 (Whitehouse 1974:6). It should be cautioned that these deposits were tertiary, and entirely derived from earlier strata, and that the lead coins may well have been demonetized by the time of their deposit (Lowick 1985:3: Allen [1982] has speculated that the coins were lost by the construction workers on site). In the second mosque phase the hypostyle mosque was enlarged. The platform fill for the enlargement also included Hibs and lead coins dated to 803-804, but no Opaque-glazed types. The third phase saw the Extension and Ablution areas built, which are associated with a coin hoard of 1024, and pre-date introduction of Hatched Slip-incised ware. Phase 4 saw the Extension remodelled, and is associated with Hatched Slip-incised ware, but lacks stonepaste ceramics. Phase 5 comprised a major restoration of the main mosque enclosure, and is associated with stonepaste ceramics. This was thought to pre-date the impoverished community described by Yaqut in 1218, but pottery found in the fourth season had an inscription considered to be a date equivalent to 1263-1264.
On the street outside the mosque was a bazaar, consisting of irregular shops and stalls built against the mosque wall. The earliest phase here sealed lead coins of 803-804 (cf. Tampoe 1989:89; Whitehouse 1974:7), and may be associated with two sherds of Plain Opaque-glazed pottery. The final phase is associated with Hatched Slip-incised ware, but lacks stonepaste ceramics. Above the Congregational Mosque was the smaller Late Mosque, built with its mihrab directly over that of the Congregational Mosque. This structure is perhaps fifteenth century, and was excavated and removed in the second season.
Site C, Bazaar: Excavated in first, fourth, and fifth seasons (Whitehouse 1968:11-12, 1971:10-12, 1972:84). This site consists of a street running parallel to the shore, a large courtyard structure on the north side facing rows of shops across a covered market, a public hammam, and a mosque. The mosque had a complex history with four periods of construction. A well in the hammam contained a complete Hatched Slip-incised ware bowl at its bottom, and Iranian Lustre-painted and other stonepaste wares from collapse deposits. The road contained 1 m of superimposed surfaces.
Site D, Pottery Works: Excavated in first, fourth, and fifth seasons (Whitehouse 1968:12-13, 1971:12-15, 1972:84). Excavations in the fourth season revealed a courtyard building (16.9 m ×17 m), converted to use as a pottery, with kilns and workshops, and four associated yards also containing kilns and workshops, collectively forming a block (45 m × 42 m) bounded on three sides by streets and the sea to the south. In the fifth season, blocks adjacent to the pottery were also found to contain kilns and workshops, so this was evidently a potters' quarter. More than thirty kilns were excavated, mostly updraft kilns with a lower fire-chamber and an upper chamber for the pots, while others were of the single chamber type. Some had rod-shaped kiln furniture. The only published wasters from this site are of coarse bowls, but fine "Egg-shell" wares and large jars are also often attributed to the site (Whitehouse 1971:12- 15; 1984:25). Occasionally, claims of glazed ware production are made: "a smaller quantity of tin glazed ware and perhaps also pottery with a green alkaline glaze" (Whitehouse 1971:12), but no evidence has been published. Although the presence of updraft kilns and kiln-rods suggest glazed ware, such a claim must be backed with harder evidence. The Siraf Petrofabric defined by analysis of unglazed wasters from this trench has been found only in poorly preserved splashed Lead-glazed wares.
Site E, Fifteenth-Century Buildings: Complex of buildings excavated in second and fifth seasons (Whitehouse 1969:54- 57; 1972:85). The first phase consists of "the square building" of mortared stone construction, with later but associated buildings of irregular rubble and mud construction. Pottery includes unglazed-painted pottery and Timurid Underglaze- painted pottery of "Indian Ocean" Petrofabric. Pottery and coins date occupation to the fifteenth century. The house at site E in particular contained a dump of nearly whole vessels of the same general style, perhaps indicating that they were originally a shipment en route to somewhere else. Three coins from the occupation layers date from Ghazan Mahmud (c. 1296-1304) to Abu Said (c. 1320-1330), with a possible further Chinese coin thought to be of Ying Tsung (1436-1439) (Whitehouse 1969:56). The mechanisms by which nearly whole ceramics are deposited must be considered to differ from those mechanisms depositing coins in floor trample, so the correlation must require some temporal lee-way. Together with other pottery from East Africa, these samples define the "Indian Ocean" stonepaste petrofabric (see Mason 1996).
Site F, Residential Quarter: Group of houses excavated in second, third, fourth, and fifth seasons (Whitehouse 1969:48- 53, 1970:9-15, 1971:5-10, 1972:84). All houses are of the courtyard type, with evidence of at least one upper storey. An orthogonal grid pattern of unpaved streets and alleys had been present from earliest occupation. Some houses, including House W, were abandoned before introduction of Hatched Slip-incised ware, others before stonepaste. The site was reoccupied in the fourteenth century.
Site G, Shrine: Excavated in third and fifth seasons (Whitehouse 1970:15, 1972:84).The site contains a mosque probably of late thirteenth century date, built in a twelfth century cemetery, with Safavid rebuilding.
Site H, Imamzadeh: Excavated in fifth season (Whitehouse 1972:82). The small mausoleum and associated structures of this site are considered to the date to fourteenth to sixteenth centuries.
Site J, Military Complex on West Promontory: Group of buildings associated with defensive wall, excavated in fifth and sixth seasons (Whitehouse 1972:78). Buildings directly attached to the wall include a hammam and other buildings with courtyards. The complex is considered military in use by Whitehouse on the basis of their proximity to the wall, and their strategic location on the western promontory. A large building (51 m × 38 m) constructed in the middle of the complex, after demolition of previous buildings, is also considered to be military, perhaps an armoury. Finds from below this building show that it was constructed in the tenth century, perhaps in answer to threats from Suhar in Oman and the Qaramita of Bahrain (Whitehouse 1974:18-21).
Site K, A Palatial Residence: Excavated in fifth and sixth seasons (Whitehouse, 1972: 72-74, 1974: 9-12).Thi palace is one of at least four large residences along the ridge dividing the Shilau Valley from the coastal plain. The smallest of these is almost twice the size of the largest house excavated or surveyed from the coastal plain. Collapse deposits are devoid of Hatched Slip-incised wares and stonepaste-bodied wares.
Site L, West Gate: Excavated in fifth season (Whitehouse 1972:78). Defensive wall with remnant of gate or portal.
Site M, Mosque, Cemetery, and Buildings by the Shore: Excavations here in the sixth season consisted of a line of exploratory trenches across a depression, and others nearer the shore. The depression was a cemetery, perhaps at the edge of an early phase of the settlement that continued in use after building developed around it. In Trench M [A] four buildings were found at the edge of the cemetery, roughly of the same alignment and general nature as those at Site F, and were perhaps part of the same planned quarter, between the Congregational Mosque and the Site M cemetery. Building 2 differed significantly in having separate entrances to the street and side-alleys for the rooms about the courtyard. Trenches nearer the shore revealed a series of buildings, including a mosque, and a later defensive wall, which had incorporated walls of some buildings and necessitated the destruction of others, although the mosque was apparently already ruinous. Whitehouse generally considers this defence to be an answer to Omani and Qaramita threats of the mid-tenth century (Whitehouse 1974:12-18).
Site N, Basilica-like Building: Excavated in the sixth season (Whitehouse 1974:22), a building (10.15 m square) with central "nave" and side "apses" behind colonnades. Considered to date to the ninth century, this "basilica" is hypothesized to be a Christian church.
Site O, Monumental Cemetery: Ten monumental tombs excavated in the sixth season (Whitehouse 1974:23-25). The tombs were mostly for collective burial and lack ground-floor access, with one for single burials having ground-floor access. There were also numerous single burials around the tombs. The collective tombs would not be considered respectable Muslim practice, and may have originally been Sasanian, but grave goods indicate use in the ninth and tenth centuries.
Site P, Mosque/Workshop and Buildings by the Shore: Two trenches were excavated in the sixth season (Whitehouse 1974:18). Trench P [1] contained a mosque, with later modification for industrial use, and subsequent building of the defensive wall. Trench P [2] contained a mosque with three phases of rebuilding.
Chronology and Siraf
Early in the excavations at Siraf, it was clear that despite normal taphonomic complexities involved in the excavation of a large urban site of several hundreds of years occupation, and with disturbed deposits, it was possible to discern stratigraphic separation of the "Samarran" pottery. There is some degree of expectation that the excavations should be capable of producing a precise chronology for the pottery, but there are several problems confronting this approach.
The history of Siraf was initially based on the historical evidence, outlined above, which was interpreted as indicating Siraf's rise in the early ninth century, its decline in the late tenth and eleventh century, and its nadir by the time of Yaqut's visit in 1218. Such a profile of steady growth, sustained prosperity, and decline, does not seem to have been subsequently questioned, although at best such brief historical references must be considered incomplete. In reality Siraf's dependence on trade and lack of naturally tenable agriculture would have made it susceptible to troubles caused by the Qaramita and the Suhar Omanis, and other unrecorded problems, which may well have led to fluctuations in the Siraf economy. This is one possible explanation for the dearth of Moulded-relief ware, the low frequencies of Basra Lustre- and Blue-painted, particularly when compared to Plain Opaque- glazed wares. Even the more remote and isolated sites in the wide distribution of this pottery have higher abundances among their fine wares. It would appear likely that there were times during the ninth and tenth centuries when there was less of the current high-quality wares coming into Siraf. Such fluctuations in the local economy might also help explain the abandonment of the mosque at Site M, and the conversion of the mosque at Site P. Such ecomomic depressions would explain the abandonment and collapse of some houses before the introduction of Hatched Slip-incised ware while others continue later. These collapses usually show signs of abandonment, and then collapse, rather than evidence of catastrophic collapse of the 977 earthquake that is suggested by Tampoe (1989: 79). Given this uneven record of occupation, it may be that these scarce types may even exist in more respectable abundances in presently unexcavated parts of the site.
It is argued here that while the stratigraphic sequences at Siraf serve as useful indicators of the dates of materials contained within them, it is dangerous to coerce the data into a precise chronology. The revised chronology of Tampoe (1989) has overly emphasized the coin evidence provided by the lead fals found in many contexts at Siraf. Coin evidence can be overly stressed at any time, but particularly when involving these lead coins, which were probably demonetized in the early ninth century, and would have then been considered practically worthless. The tertiary nature of Site B platform fill and various other "make-up" deposits on other sites should also forewarn caution. These comprise material derived from elsewhere on the site, along with their attendant finds, rather than being primary or secondary deposits. Tampoe uses this falsely precise chronology to provide an early date for the introduction of "Opaque White-Glazed" pottery, and stretches the archaeological data yet further to support it. For instance it is unsupportable to state that the fill of a well is contemporary with the house that is associated with the use of the well, as suggested by Tampoe (1989: 89), referring to House E at Site F. Other stratigraphically questionable evidence comprises that at the Site C Mosque tertiary fill, where three coins of mid- eighth century date are associated with "Opaque White Glazed." The tertiary fill is sealed by a contemporary floor that contained a coin ten years earlier than two of those below the floor. Whitehouse considers these to be residual, and the evidence does not really discount this.
SIRJAN, Iran
The site of Sirjan stands in the largest basin of this area of the southern Zagros range. Agriculture was predominantly qanat-based, but the size of the basin and the intersection of important routes in the area across and along the parallel chains of ridges that dominate this part of Iran explain the existence of a city since at least the Sasanian period. From the ninth until the early eleventh century the city grew in importance, at times being the capital of Kerman province, but subsequently Sirjan went into a slow decline until eventually the site was abandoned entirely. The modern city of Sirjan is about 15 kms to the northwest.
Fieldwork in the region was begun in 1968 by the late Andrew Williamson, who actually located the site and eventually undertook excavations within it. The archive and pottery from the excavations and survey are stored at the Ashmolean Museum, Oxford, and following Williamson's tragic death in 1975 these were written up by others (Morgan and Leatherby 1987). Pottery from Sirjan and the region provided some useful samples for analysis, including evidence of production in Sirjan.
SUSA, Iran (also al-Sus, or currently Shush)
Susa has been the domain of French archaeologists since the first excavations by Loftus in 1851. In these early excavations the bulk of material was bulldozed aside in pursuit of the Achaemenid foundations. More recent excavations have provided a useful corpus of early Islamic pottery (Kervran 1977). Unfortunately, the only publication of this pottery is a journal paper, which presented the conclusions of a study relating the typology to the site stratigraphy; there was no full publication of the ceramic corpus in relationship to its stratigraphy. Most other publications of the site (Kervran 1979; Boucharlat et al. 1987; Hardy-Guilbert 1984) appear to be primarily concerned with local wares. Pottery from early excavations at Susa, presently stored at the Louvre, were available for analysis and study, while published examples of pottery from Susa (Kervran 1977) have been included in the typological study.
TELL ABU SARIFA, Iraq
Although pottery from this site was not available for analysis, the excavations are one of the few attempts to come to grips with the Sasanian to Islamic ceramic sequence, and so provide archaeological data which must be considered in this study.
Publication of this site consists of one paper by R. McC. Adams (1970). Prior to the 1969 excavation, this tell was one of numerous unnamed sites of Sasanian to Early Islamic date in this region of south central Iraq. The tell covers approximately 3 hectares, and received two main trenches (each 5 m × 10 m), which were sunk into the highest point of the tell, down to natural deposits at a depth of 4.8 m below the surface. A further 10 m × 5m trench was sunk into the Islamic strata, with minor trenches elsewhere. The strata comprise six main phases, with a seventh phase represented by a deflated surface with later sherds. Level I comprises about 1.6m of mixed earth with scarce Parthian pottery. Levels II to III consist of occupation layers of Sasanian date. Levels IV to VI consist of occupation layers of early Islamic date. Adams courageously offered dates for the Islamic levels of 650-800, 800-950, 950- 1100, and 1100-1150 for the deflated strata, but is not dissembling in his admittance of the almost total lack of evidence for these dates. It seems slightly odd that Level V is equated with the Samarra pottery, while according to his tables it is Level IV that contains the Lustre- and Blue-painted wares. This does, however, support an early dating for the Blue- painted wares.
Unfortunately, the stratigraphy appears to have been riddled with pits, many of which were not recognized until obviously intrusive material was found, or pit-profiles were recognized in the baulks. Hence many individual pieces were "reattributed" to their "correct" levels. In one particular case, fragments of a Slip-incised bowl were found in two separate phases, while Adams considered the type more suitable to a completely different phase. As Adams states, the small proportion of glazed wares (generally far less than 10%) in what is a small rural site, coupled with the almost complete absence of numismatic evidence, makes this site less useful for providing a chronology for Islamic glazed wares. Adams considered that the excavations did provide useful evidence concerning the relative chronology of one class of ceramics. He divided Slip-incised wares into two main chronological divisions, although without a distinct transition between. The earliest group carried "dense, all-over colour patterns, with closely grouped streaks" while the incision was fine and diffuse, barely visible, with simple patterns. Effectively the splashed colours where intended to provide the major effect. The later group have more isolated patterns of dots, while the incised decoration is more complex and deliberate, and with varying thickness of line.
TELL ASWAD, Syria (see Raqqa)
TILLE HOYUK, Turkey
A small site on the right bank of the Euphrates in southeast Turkey (Moore 1993), probably representing a defensive work. Excavations at the site revealed a medieval occupation comprising three levels, with a phase of abandonment in the third level. Pottery from the site includes a number of diagnostic Syrian pieces of Groups One ("Tell-Minis"), Two and Three ("S-back"), with diagnostic Group One pieces being largely restricted to Levels One and Two, Group Three pieces restricted to Levels Two and Three, and Group Three pieces restricted to Level Three. Using the chronology of this study, this would suggest a limited timespan of c. 1075-1150 for the bulk of occupation at the site. Coin evidence from the site comprises seven coins. If we ignore one coin, the coins create a useful sequence with an ‘Abbasid piece of c. 740 possibly indicating early use of the site (although there appears to be no pottery of this period), four worn Byzantine bronze coins collectively of 1042-1070, mostly in Level One, dating the beginning of the main occupation; and a Venetian silver coin of 1382-1400, pierced for suspension, belonging to the reoccupation of the site. The coin being ignored is a French coin of the twelfth or thirteenth century from a pit said to predate Level One. I could find no further reference to this pit in the text and it is not in the published sections, but it appears to be on the southwest margin where overlying strata is thin. Given that otherwise the coin evidence agrees admirably with the ceramics, it might be possible to suggest that there is some problem with this find. The excavators date the pottery by reference to the 1947 edition of Lane's Early Islamic Pottery, and so could not have been aware of the anomaly.
ZABID, Yemen
Agriculture in the Tihamah, the Red Sea coastal plain of Yemen, is primarily dependent on the seasonal monsoonal rains which flood from the mountains down the wadis. When controlled by quite modest water-managment this can lead to significant agricultural productivity. Zabid is situated in the midst of what appears to be the largest of these irrigated areas on the Tihama, and lies across the principal route along the plain. The city was founded in the ninth century, and became an important regional capital, at times being the seasonal capital for the entire Yemen.
Systematic fieldwork in the region was begun in 1980 (Keall 1983), with excavations continuing in the city of Zabid since 1987. Analysis of pottery from the region initially concentrated on defining local products (Mason and Keall 1988a), but imported pottery of significance to this volume, mainly from Iraq and Egypt, has also been sampled.

Return to Table of Contents