Printer Friendly

Contours of conversion: the geography of Islamization in Syria, 600-1500.


When Khalid b. al-Walld invaded Syria in 13/634, the region was inhabited by a religiously mixed population with multiple kinds of Christianity present alongside Judaism and paganism. When the Ottoman sultan Selim I conquered Mamluk Syria nine centuries later in 922/1516, the region's religious diversity looked distinctively more Muslim, with Sunnis of four legal schools sharing the land with Druze, Nusayris, Ismailis, and Twelver Shiites, in addition to reduced populations of Christians and Jews. The process of Islamization whereby regions such as Syria slowly shifted from areas without Muslims to those where Muslims

formed the majority is one of the more dramatic transformations of the medieval world.

Both the mechanisms and the contours of Syria's Islamization are poorly understood. In part this is due to the absence of surviving demographic data before the Ottoman tax census records of the sixteenth century. After Selim I's conquest, the bulk of Syria was divided between provincial governments (sg. eyalet) based in Aleppo, Tripoli, and Damascus, although portions of eastern Syria around al-Raqqa were assigned to Ruha (modern Urfa), while areas historically regarded as northern Syria were incorporated into the eyalet of DQlqadir or the Ramadanid principality. (1) The tax registers (sg. defter) produced by the new provincial governments, many of which survive, identify religious minorities due to the differential taxation applied to non-Muslims. (2) These records demonstrate that in the sixteenth century the Muslim population of Syria formed an overwhelming majority in the countryside and a large majority in most towns and cities. (3) Much scholarly debate centers upon how quickly the majority of the population adopted Islam.

But demographic change was only one component of a multi-faceted process of Islamization in Syria. Many medieval Muslim jurists seem to have regarded widespread conversion to Islam as irrelevant to Islamic society, while increased enforcement of regulations upon non-Muslims to demonstrate the superiority of Islam appears more important to their notion of Islamization. The Umayyad caliph 'Abd al-Malik b. Marwan (r. 65-86/685-705) is generally credited with replacing Byzantine coins with aniconic "Islamic" coins and with building the Dome of the Rock in Jerusalem, showing that state-sponsored Islamization had numismatic and architectural implications. Governmental Islamization was only one component of the broader, and far slower, process by which Syria became in some sense "more Islamic." That process transformed urban environments, as mosques replaced churches, synagogues, and temples as the foci of cities. (4) Islamization included the conversion of landscapes, as monasteries fell into ruin and Muslim shrines sprang up instead. (5) Concepts of areas as "primarily" one religion or another shifted with Islamization, as did social expectations regarding typical relationships between members of different religious groups. Islamization was a complex and multi-dimensional process that spanned many centuries.

This lengthier process was also not one-directional. Muslims converted to Christianity as well as vice versa. Ruined non-Muslim religious sites could sometimes be rebuilt. (6) Al-MuqaddasI (fl. late tenth century) acknowledged that despite his high praise for Syria's many advantages, "some [of its people! have apostasized." (7) Yaqut al-Hamawi (d. 626/1229) mentioned a village named 'Imm between Aleppo and Antioch, "in which today everyone is Christian," but he quotes the Risala of Ibn Butlan from the eleventh century to say that two centuries earlier it had a mosque. (8) In certain contexts Islam was not the only religion supported by the state, as Muslim rulers sometimes provided stipends to Jewish and Christian religious authorities in addition to the ulema. (9) Furthermore, the Byzantine reconquest and the Crusades reintroduced non-Muslim rule to portions of Syria from 358/969 to 690/1291, so that even state support for Islam could not be taken for granted. Indeed, under Frankish rule a large enough number of Muslims sought to become Christian that canon law needed to be developed in order to handle difficult social questions regarding marriage and slavery in such cases. (10) As Benjamin Kedar concludes, "in the Frankish Levant, passages from Islam to Christianity and vice versa were not rare at all." (11)

The boundaries of Syria in medieval Arabic geographical thought were different from today. Yaqut presented the most common definition of this region as extending from the Euphrates to the town of al-'Arish on the Egyptian coastline southeast of Gaza, and from the Arabian desert to the Mediterranean Sea (see map on previous page). (12) This definition leaves open how far into modern Turkey the region was thought to extend, and before the Byzantine reconquest of the tenth century the northwestern border of Syria was simply considered to be the boundary of Byzantine control, sometimes even including Malatya on the upper Euphrates as the northern edge of Syria. (13) On the other hand, Yaqut does not include any major city north of Manbij and Aleppo, noting only in passing the border regions (thughur) of al-MassIsa, Tarsus, Adhana, and Mar'ash. (14) This article will take as the northern border of Syria the Taurus Mountains of southeastern Anatolia, excluding Malatya but including the border towns mentioned by Yaqut.

The classic study of Islamization remains Richard Bulliet's Conversion to Islam in the Middle Period (1979). Bulliet analyzed biographical dictionaries to graph the adoption of specifically Muslim names in several different regions across the Islamic world, making certain approximate assumptions about length of generations and age of conversion. He argued that these distinctively Islamic names first appear within a lineage for a convert to Islam or for his children, and graphing the incidence of Islamic names for the ulema in a biographical dictionary gives a curve (p. 19) that can be taken as the conversion curve for the region as a whole. (15) The result is a summative S-curve that displays the Muslim proportion of the population as monotonically increasing, whose slope represents the rate of conversion, first slow, then increasing to a midpoint, and then decreasing to level off again as the number of late adopters decrease. He suggests (pp. 109, 112) that conversions peaked from the late-eighth to the mid-tenth century, and that the rise of Shiite groups such as Druze, Nusayris, and Ismailis was occasioned by the late conversion of mountain Christians. Nevertheless, he acknowledges (pp. 110, 112) that his proposed conversion curve is difficult to correlate with the political, social, or religious history of Syria, and he concludes, "Syria does not present a tidy, easily understandable picture."

Somewhat more recently, Nehemia Levtzion (1990: 290) summarized what is known about the contours of the Islamization of Syria and Palestine before the Ottoman conquest, based primarily on secondary scholarship with reference to primary sources by al-Baladhurl (d. 279/892) and Michael the Syrian (d. 1199). This account derives primarily from narrative historical sources, whether used by Levtzion or by the other scholars he cites, and although narrative sources are helpful for connecting otherwise isolated data, their interests are typically circumscribed in ways that limit their utility for the purpose of describing regional Islamization. Michael the Syrian, for example, is interested almost exclusively in the secular rulers and in his own denomination of Christianity, and thus says very little about other Christian groups such as the Chalcedonians, much less the Jewish population of Syria. Al-Baladhurl's Futuh al-buldan primarily collects traditions about the seventh-century conquests, and only mentions non-Muslims to the degree that they figure in such traditions, without any attempt to discuss the state of non-Muslims in Syria in his own lifetime. Narrative sources need to be supplemented by additional evidence to provide a wider picture of the Islamization of Syria.

Studies of Islamization in Syria since 1990 have focused on specific themes, restricted source materials, and narrower time frames. Bethany Walker (2013) synthesized the archaeological evidence for Islamization into the ninth century at a site in central Jordan. Uri Simonsohn (2013) examined legal sources from the early Islamic period to clarify the process of personal conversions, especially reversed and repeated conversions. Nancy Khalek's Damascus after the Muslim Conquest (2011) focuses on the transformation of a capital city to the end of the Umayyad dynasty, while Amikam Elad (1995) examined pilgrimage to holy places in Jerusalem, primarily but not exclusively in the first couple of centuries of Islam. R. Stephen Humphreys (2010a) argued that Christianity continued to prosper under the Umayyad dynasty, bringing together a range of literary, economic, and archaeological sources. For a later period, the conversion of Syria's religious topography was analyzed by Daniella Talmon-Heller (2007a) for the eleventh to thirteenth centuries, indicating that Islamization was not merely an early Islamic phenomenon. These studies, each important for its scope, do not provide or permit the synthesis of a trajectory of Islamization, especially after the Umayyad period ending in 132/750.

One body of evidence that allows us to provide a first sketch of the contours of the Islamization in Syria over the longue duree, from the conquests of the seventh century to the Ottoman annexation of Syria in the sixteenth century, consists of the geographical texts composed by administrators, travelers, and belles-lettrists describing the region of Syria in the medieval period. An eclectic body of Islamic geographical literature, primarily in Arabic but partly in Persian, preserves indications of the progress of Islamization at different periods. The complexity of this literature requires methodological nuance to interpret it, but properly understood it is a valuable body of evidence for the development of Syrian society.

This article analyzes ten Arabic geographical works, as well as one work in Persian and one in Hebrew. On the basis of these works it sketches a trajectory of Islamization in Syria until the Ottoman conquest. The Islamic conquest of Syria began the process of Islamization with the rapid installation of mosques and garrisons in the major cities and along the coastline, while the first inhabitants of Syria to adopt Islam were many of the Arabs who already lived in the region before the seventh century. By contrast, there is little evidence for rural Islamization before the tenth century, and the evidence that exists suggests that before this period rural Islam in Syria was primarily a nomad's religion. The Byzantine conquests of the tenth to eleventh centuries and the subsequent Crusades reintroduced Christian rule in Syria, which resulted in certain segments of this region being known for Christian populations more than others, such as the area north of the Ghuta around Damascus and the coastline. Rule by Christians and the confiscation of certain urban mosques may also have lent urgency to the process of founding Muslim rural shrines, although in many cases the earliest shrines that are known to have interested Muslims were dedicated to pre-Muslim figures, and in some cases were maintained by Jews or Christians. When the Mamluks from Cairo expelled the last Crusaders, they also devastated the coastline, leaving Christianity primarily attested in northern Syria. (16) Even under Mamluk rule, however, certain villages located along major roads north of Damascus which were still entirely or predominantly Christian would have reminded Muslim travelers that Syria was not an exclusively Muslim territory.


The geographical literature that describes Syria under Muslim rule is an important body of sources for the long process of Islamization. (17) This literature does not form a single genre, but rather exists in several forms. Al-Baladhuri's Futuh al-buldan is primarily a work of history and traditions that describes places only by virtue of their having been conquered by the Muslims. The geographers of the Balkhi school, such as Ibn Hawqal (d. ca. 362/973) and al-Muqaddasi, divided their works by regions, each provided with a map. By far the most extensive geographical work is the Mu'jam al-buldan of Yaqut al-Hamawi, arranged as a dictionary with place names in alphabetical order. The works of Benjamin of Tudela (d. 1173), Ibn Jubayr (d. 614/1217), and Ibn Battuta (d. 770s/1370s) are travelogues intended to convey geographical information. Nevertheless, the authors in the geographical discourse utilized earlier works in different genres and freely quoted other authors, as is typical for medieval Islamic scholarship.

Methodological Challenges

The evidence provided by the geographical literature is complicated by several factors. Geographical works pay selective attention to certain non-Muslim groups more than others, and therefore cannot be used to infer relative demographic strength. Thus al-Baladhuri's Futuh al-buldan only mentions Syrian Jews briefly with respect to Damascus, Tripoli, Hints, and Qaysariyya, (18) but even after the massacre ordered by the Byzantine emperor Heraclius in 629, Tiberias was an important center of the Jewish population. (19) The greater interest in Christians than in Jews, in al-Baladhuri and later Muslim geographers, is probably due to political opposition with the Christian empire of Constantinople rather than to demographic realities. On the other hand, the Samaritan population of Filastin is singled out for attention by al-Baladhuri and later geographers, which likely reflects their presence exclusively in this region. (20) No indication of the relative strength between Jews, Samaritans, and Christians can be inferred from these references.

Other features of literary texts also complicate the use of geographical works. Geographical literature often lists places, but lists of villages are necessarily not comprehensive, nor can these lists be presumed to be representative. Furthermore, different authors have diverse interests that influence the selection criteria, so unless the author has a clear interest in recording religion, the absence of a reference to a particular religious community or expected religious edifice does not indicate its absence from the location. For example, while Ibn Battuta frequently mentions mosques in his descriptions of the places he visited, no matter how briefly, Yaqut only infrequently refers to mosques in his geographical dictionary. The late ninth-century historian and geographer al-Yaqubi (1918: 86-87) does not mention any mosque in Syria other than the Umayyad mosque of Damascus, being primarily interested in recording the tribal affiliations of the various Arab populations. By contrast, Benjamin of Tudela (1907) is principally interested in recording the distance between cities and the size of the Jewish populations that lived there. Finally, as with most fields of medieval scholarship, information included in the geographical work may have been borrowed from an earlier source without attribution, which makes it challenging to identify the period to which any given assertion may pertain. (21) The result is that these works cannot simply be transformed into a database on which statistical analysis can be performed; rather, each text must be read as a literary and linguistic performance, yet one intended to communicate certain facts about the world. (22)

In a way, some of these challenges turn out to be surprising opportunities for writing the history of Islamization. Literary geographical texts sometimes indicate whether the authors regarded certain details as surprising or typical, which partially compensates for the lack of comprehensive or representative data. Even the adoption of earlier texts' words without modification or attribution, although bearing a different relationship to the author's experience than new composition, typically reveals what the author regards as well said and plausible enough. Such literary phenomena provide hints to the evolving expectations and assumptions regarding the religious landscape of Syria, which are as much a part of the process of Islamization as progressive personal conversions or architectural repurposing; yet such attitudes and conceptions would be largely absent from census records. With a nuanced literary approach, geographical texts can be useful sources for a full-orbed account of Islamization.


The earliest stage of Islamization of Syria reported by the geographical texts is the conversion of nomadic Arabs, resulting in a distinction between often nomadic Muslims and primarily sedentary non-Muslims. The core of the first Muslim community in Syria was formed by the conquering Arab armies. Al-Baladhuri reported that the Muslim commanders appealed to the largely Christian Arabs who already lived in Syria on the basis of their common ancestry, with mixed results. Jabala, the chief of the Banu Ghassan, rejected Islam (although one account says he converted and then apostasized) and moved to territory still under Byzantine control, while the Arabs near Qinnasrln and Aleppo proved more agreeable, with many of them accepting Islam. (23) The degree to which the Islam of the conquest period was viewed as "an Arab society" is perhaps indicated by the account of the Banu Taghlib in the region of Diyar RabFa to the northeast--the tribe remained Christian, but instead of paying jizya it paid double the normal Muslim sadaqa. (24) Al-Baladhuri presented the ruler as saying, "Since it is not the tax of the unbelievers (aclaj), we shall pay it and retain our faith." (25) The chief of the Banu Ghassan is said to have made a similar offer to pay sadaqa instead of jizya, but in his case it was rejected. (26) Syrian Arabs who had accepted Islam were already sufficiently numerous for the military commander Abu 'U bayda to station a garrison of them in the city of Balis on the Euphrates within three years of the battle of Yarmuk. (27) It is unclear how quickly the nomadic and semi-sedentary Arabs of Syria converted to the new religion, but the swifter adoption of Islam by Arabs than by non-Arabs likely caused the religious boundaries between Muslims and non-Muslims in the countryside to approximate the divide between nomads and sedentary farmers. (28)

The Islamization of the sedentary population seems to have begun in major cities and coastal towns, due to the presence of Muslim governors and garrisons. (29) As presented by the geographers of subsequent centuries, the circumstances of many cities' surrender or conquest provided a location to be used for the mosque, whether part of the city's cathedral or a new site. (30) A quarter of the main church of Hims was made into a mosque, (31) while the cathedral of St. John the Baptist in Damascus had a portion set aside for Muslim prayers. (32) Aleppo's mosque was a new construction, (33) as was that of Latakia on the coast. (34) The link between coastal garrisons and mosques is made explicit by an account reported by al-Baladhuri that the caliph 'Uthman directed his cousin Mu'awiya, then governor of Syria, to garrison the coastal towns, to build new mosques, and to enlarge existing mosques. (35) Thus, al-Baladhuri referred to mosques in the cities of 'Asqalan "in the days of Ibn al-Zubayr" (d. 73/692), the newly founded district capital al-Ramla by 101/720, al-Massisa by 84/703, and Tarsus by 172/788. (36) He also indicated that Mu'awiya transferred Muslim populations to the coastal cities of Tyre, Acre, and Antioch. (37) Later geographers mentioned Jerusalem's Dome of the Rock, built by the caliph 'Abd al-Malik b. Marwan in 72/691-2, (38) although a sizeable earlier mosque in the city was mentioned by a Latin pilgrim who visited the city around 680. (39) Al-Harawl and Yaqut al-Hamawi would include Bethlehem outside Jerusalem as a city that acquired a mosque under the second caliph cUmar b. al-Khattab, but the fact that al-Muqaddasi did not mention such an edifice, although he was a native of the area and seems disposed to mention all the mosques he knew, suggests that it was a more recent creation. (40)

How long did Christians and Muslims share sanctuaries? According to al-Baladhuri, the transformation of the Cathedral of St. John the Baptist in Damascus into the Umayyad Mosque took place during the reign of the caliph al-Walid b. cAbd al-Malik (d. 96/715), indicating that the mosque in the capital city of the caliphate was shared for a few generations, continuing even after the construction of the Dome of the Rock in Jerusalem. (41) The cathedral of Hims was divided between Christians and Muslims even longer, although it is not possible to say with certainty exactly how long on the basis of the geographical literature. Ibn Hawqal's statement, "In [Hims] is a church, part of which is the Friday mosque, and half of it belongs to the Christians," (42) may indicate that it was still divided in the middle of the tenth century. On the other hand, his pupil al-MuqaddasI later in the century only refers to the division at the time of the conquest. (43) A story reported later by Yaqut refers to a young Muslim "playing in the church with a ball, and it happened that the ball entered the mosque," showing that the building was certainly divided, perhaps as late as the 170s/ca. 790. (44) The early construction of mosques in the major cities of Syria added an Islamic focus to urban centers, but did not necessarily exclude or replace non-Muslims. (45)


Outside of the major cities, the slower progress of Islamization is shown by the delayed diffusion of rural mosques and Muslim shrines. In the tenth century Ibn Hawqal remarked that the region of Filastln had "around twenty minbars [i.e., mosques] despite its small size." (46) On the one hand, twenty mosques would cover all the major cities but only a few of the many villages. That only very large villages possessed mosques is indicated not long afterward by al-MuqaddasI, who said of Filastln, "In this district are large villages with their own mosques, and these are more populous and more flourishing than most of the cities of al-Jazira." (47) Indeed, al-Muqaddasi listed twenty-one mosques in the region of Filastln. (48) Of these, it is clear that only two (Jabal Zayta and al-Yaqin) are properly rural, and seven are described as being in villages (qura), namely, Hebron, Ludd, Kafr Saba, 'Aqir, Yubna, 'Amawas, and Kafr Sallam.

On the other hand, Ibn Hawqal's concessive clause ("despite its small size") seems to indicate that in the diffusion of mosques Filastln was more Islamized than Ibn Hawqal expected, which may hint that throughout the rest of Syria at that time there were not many village or rural mosques. Since the number of Muslims that a mosque could serve might range from tens to thousands, it is impossible to estimate the Muslim population of Filastln based on this figure, but it does suggest that in the mid-tenth century only the largest villages would have had a Muslim architectural presence. (49) This suggests that after the initial conquests Islamization was a process of diffusion from the cities to the villages and countryside.

Away from the coast, the roads connecting major cities overland were meeting-places for Muslims and non-Muslims, and thus conduits of Islamization. Al-Muqaddasi listed six "large villages possessed of their own mosques," of which he located three (Kafr Saba, 'Aqir, and Kafr Sallam) on main roads. (50) Of the other three, two were also on important roads: Yubna is between Yafa and 'Asqalan on the coastal road, and 'Amawas is between al-Ramla and Jerusalem. Of the six villages, only Ludd is not on a major road, but in the late tenth century it was essentially a suburb of the district capital al-Ramla; according to al-Muqaddasi, it "lies about a mile from al-Ramla. There is here a great mosque wherein large numbers of people assemble from the capital, and from the villages around." (51) By contrast, he reported no mosques in villages or cities away from roads, the coast, or the outskirts of the capital al-Ramla.

If the district of Filastin was probably the Syrian district with the greatest concentration of mosques, al-Muqaddasi is explicit that it also had a greater number of rural shrines than other districts, especially in the neighborhood of Jerusalem, and he asserted that he had listed most of them. (52) The most striking feature of the list he gave, however, is how few of them celebrate specifically Muslim figures. (53) The majority of these shrines and holy sites pertain to figures of ancient Jewish history who were venerated in common by Jews, Christians, and Muslims (although often not Samaritans): Abraham, Isaac, Jacob, Rachel, Job, Moses, Saul, David, Uriah, Solomon, and Jeremiah, some with multiple sites. A smaller number pertain to Jesus, Mary, and John the Baptist's father Zakariyya. (54) Of distinctively Muslim sites possibly outside of major cities, he referred only to "'Umar's mosques," which presumably would include the mosque in commemoration of the caliph on Jabal Zayta outside Jerusalem and al-Yaqin Mosque outside Hebron. Even his vague reference to the "shrines of the prophets" shows a Muslim approach to pre-Islamic history. It is therefore unclear whether the many shrines to pre-Islamic personages were in any way distinctively Islamic, or whether they were shared between Muslims and non-Muslims, perhaps even in the possession of the latter.

Later geographical sources record increasing numbers of shrines. (55) The Persian poet and philosopher Nasir-i Khusraw, whose account of his travels tends toward brevity, indicated that he turned aside from his travels down the coastal road in order to visit "a mountain where various prophets' shrines are located" between Acre and Tiberias. (56) But the majority of these religious sites were dedicated to historical figures shared by Jews, Christians, and Muslims alike: Esau, five sons of Jacob, the father-in-law and wife of Moses, the mother of Moses, Joshua b. Nun, Jonah, and Ezra. (57) Another shrine was dedicated to the legendary founder of Acre, namely cAkk, while two others of presumably Muslim origin were dedicated to the pre-Islamic figures Hud and Dhu 1-Kifl. A mosque known as the "Jasmine Mosque" was situated to the west of Tiberias, while the only shrine dedicated to an earlier Muslim was the tomb of Abu Hurayra south of Tiberias, and the Persian traveler indicated that pilgrimages to it were impossible because the children of the local Shiite population harassed would be visitors. (58) Even more significantly, he described and explained the "Spring of the Cows" ('ayn al-baqar) outside Acre without mentioning the mashhad of 'Ali, which geographers from 'Ali al-Harawi onward mention there. (59)

Over a century later, 'Ali al-Harawi recorded clear instances of shrines shared between Muslims and non-Muslims. For example, he mentioned outside of the Jewish Gate of Aleppo a stone "at which votive offerings are made and upon which rosewater and sweet fragrances are poured. Muslims, Jews and Christians hold it in regard. It is said that beneath it is the tomb of one of the prophets ... or the saints. God knows best." (60) He also referred to shrines dedicated to pre-Islamic figures such as Joshua b. Nun, Alexander the Great, and the mother of John the Baptist, in greater number than indicated by al-Muqaddasi. (61) On the other hand, in his record of the pilgrim shrines he encountered in his travels through Syria in the late twelfth century, shrines to figures of early Islamic history have greatly increased. In the countryside around Damascus he recorded the tombs of Dihya al-Kalbi (who also has a tomb near Tiberias and a tomb at al-Fustat), Hujr b. 'Adi, Zumayl b. Rabia, Rabia b. 'Amr, Khalid b. Sa'id, Sa'd b. Ubada al-Ansari (although al-Harawi rejected the validity of this tomb), Umm Kulthum, Mudrik, Kannaz, Shaykh Sulayman al-Daranl, Abu Muslim al-Khawlanl, Umm 'Atika, and Suhayb al-Rumi (the last two likewise rejected by al-Harawi). (62) By al-HarawI's time in the late twelfth century, the Ghuta around Damascus had the highest concentration of rural shrines. By contrast, he complained that due to Crusader rule, "in Ascalon's cemetery are many saints, and Successors whose tombs are unknown; the same with Gaza, Acre, Tyre and Sidon and all of the towns of the coastal plain." (63) And it is noteworthy that he was unable to list any Muslim shrine in the hinterland around Aleppo, despite wishing to exalt the city of his final patron. (64)

Muslim devotion to pre-Islamic figures and prayers at shrines devoted to them were not problematic, as from a common Muslim perspective the prophets of old had preached Islam, but the Jews and Christians had corrupted the message. Far from a religious difficulty, shared shrines provided an opportunity for Muslims to have pilgrimage sites maintained by non-Muslims, and an opportunity for non-Muslims to convert more easily to Islam without forsaking the loca sancta and past holy figures upon which they relied. The progression of reports from al-Muqaddasi to Nasir-i Khusraw to al-Harawi, however, seems to indicate that the creation of specifically Muslim holy sites in rural areas was a slow process, perhaps only beginning in the tenth century around the time of the Byzantine reconquest, and by no means complete by the end of the Crusader period. In particular, the late appearance of the tombs of Companions in rural areas even around Damascus, and often with questions regarding their authenticity, suggests that funereal veneration of the Companions was a late stage in the development of a specifically Islamic landscape. (65) The dedication of shrines to figures revered by multiple religions was an important step in the conversion of the rural population, but it was five centuries before distinctively Islamic shrines are attested in most of Syrian countryside.


When Nikephoros Phokas entered Syria at the head of a Byzantine army in 350/962, establishing Antioch as a Byzantine outpost in 358/969, Syria was already accustomed to marauding armies. Neglected as a province after the caliphal capital moved from Damascus to Baghdad in the eighth century, by the later ninth century central 'Abbasid authority was waning and Syria was contested between various Muslim rulers such as the Tulunids, Ikhshidids, and Hamdanids. The new element introduced by the Byzantine reconquest of northern Syria was rule by Christians in a territory from which it had been absent for over three centuries. From the 350s/960s until the Crusader stronghold at Acre was captured by the Mamluk armies of Egypt in 690/1291, Syria was divided between multiple Muslim and Christian rulers, with a brief hiatus between the Byzantine loss of Antioch in 477/1084 and the Crusader conquest of the same city in 491/1098. This division of Syria widened the gap in Islamization between different portions of the region, reversing the coastal cities' early adoption of mosques while perhaps encouraging the development of rural Islamic shrines.

The earliest geographers to observe the Byzantine reconquest painted the invaders as little more than raiders and deplored the sorry state of Islam that permitted them to succeed. Ibn Hawqal recorded Greek attacks on Hims, Aleppo, Qinnasrln, Jabala, Hisn Barzuya, Antioch, al-Hadath, Mar'ash, al-Haruniyya, al-Iskandaruna, and he blamed the enemy's success on failures of Muslim religious zeal. (66) He lamented over Antioch,

   The enemy has overcome it and possessed it, and before its conquest
   it had become disordered in the hands of the Muslims, and now it is
   more severely disordered and abased.... Around it there are
   sultans, Bedouin, lords, and kings, for each of whom his today
   distracts him from attention to his tomorrow, and what is forbidden
   him and his vanities distract him from what God Most High enjoined
   and the governance and leadership incumbent upon him. (67)

The Byzantine invasion of Hints he credited to their confusion (khabal) and luxury (yasar), while Hisn Barzuya's surrender he ascribed to, among other factors, their "lack of faith" (qillat al-iman). (68) Al-Muqaddasi devoted less space to the presence of Byzantines than his predecessor, because he consciously excluded from his description those portions of Syria ruled by the Christians, (69) but he described a climate of fear among the Muslims in Syria: "The people live in dread of the [Byzantine army], as if they were in a foreign land, for their frontiers have been ravaged, and their border defenses shattered." (70) The Muslim geographers of the tenth century depicted the division of Syria between Christian and Muslim rulers as a religious catastrophe.

Nevertheless, these same geographers indicated that most Muslims in the areas now under Christian rule were content to accept the new system. Ibn Hawqal concluded his section on Syria with a pessimistic prediction that Byzantine rule and the jizya levied on Muslims would lead many of the people of Syria to abandon Islam: "Most of its people remained, while they accepted jizya from them, and I think they will convert to Christianity, disdaining the humiliation of the jizya and greedy to obtain provisions for honor and comfort." (71) Al-Muqaddasi confirmed Ibn Hawqal's dire predictions: "Some [of the people] have apostasized, while others pay tribute (jizya), putting obedience to created man before obedience to the Lord of Heaven. The general public is ignorant and churlish, showing no zeal for the [struggle], no rancour towards enemies." (72) The new Christian rulers were evidently inclined to treat their Muslim subjects much as their Muslim predecessors had treated Christian subjects: the term jizya probably refers to a Muslim-specific head tax, analogous to the poll tax on non-Muslims levied by the caliphs. Although the Byzantine army reportedly destroyed mosques, such attacks seem to have occurred as an element of capturing and plundering a city, and there is no indication that the new Christian overlords prevented local Muslim rulers from repairing mosques. (73)

Muslim travelers through the Crusader states in the twelfth century adopted a curiously mixed attitude toward the Frankish rulers. (74) Christian rule in lands formerly under Muslim control was clearly regarded as a shameful fact, as 'All al-Haraw! complained to Prophet Muhammad when the latter appeared to him in a dream in the mosque in 'Asqalan in 570/1174, before Salah al-Din's conquest of the city. (75) Ibn Jubayr polemicized against Muslims who chose to remain in lands ruled by non-Muslims: "There can be no excuse in the eyes of God for a Muslim to stay in any infidel country." (76) On the other hand, both al-Harawi and Ibn Jubayr often emphasized how little the Franks had interfered with Muslim religious practice. Thus, al-Harawi indicated three times that the Crusaders did not damage various aspects of the Muslim sanctuaries in Jerusalem and Bethlehem. (77) Ibn Jubayr, for his part, ascribed to God's intervention the preservation of part of the main mosque at Acre and part of 'Ayn al-Baqar outside the city. (78) He also indicated that the Muslims of Tyre were treated better than those of Acre, and that they lived under a written guarantee of safety (aman),19 while Muslim peasants enjoyed greater security under Frankish rule than they would in lands ruled by Muslims. (80) Both of these authors clearly expected greater harassment from the Franks than they received. (81)

Ibn Jubayr in particular laid out the Crusaders' treatment of their Muslim subjects, and his description mirrors the treatment of Christians by Muslim rulers. When the Franks captured Acre, "Mosques became churches and minarets bell-towers," but just as the Byzantine cathedrals of Damascus and Hims were divided between Christians and Muslims in the seventh century, so Acre's Friday mosque and the shrine at 'Ayn al-Baqar were shared after the Crusaders' conquest of the city. (82) He explicitly likened the tax levied upon Muslims under Frankish rule with that levied upon Christians under Muslim rule: "The Christians impose a tax on the Muslims in their land which gives them full security; and likewise the Christian merchants pay a tax upon their goods in Muslim lands. Agreement exists between them, and there is equal treatment in all cases." (83) In his polemic against Muslims choosing to dwell in lands ruled by non-Muslims, Ibn Jubayr alluded to the head-tax upon Muslims as "the abasement and destitution of the capitation." (84) He gave the rate of the jizya as 1.25 dinars, as well as half of the crops. (85) However, a lighter tax could be assessed on travelers: Ibn Jubayr indicated that when he first entered lands under Crusader rule, the jizya he paid was levied primarily upon Muslims from the Maghrib, due to a prior attack by a group of western Muslims on a castle of the Franks, and that other Muslims were not taxed in this manner. (86) The Andalusi traveler had earlier marveled at the security of both Christian and Muslim travelers despite the ongoing wars between Salah al-Din and the Crusaders, (87) and he was surprised, and horrified, at the level of social integration that could be achieved between Christians and Muslims in Syria.

Despite Ibn Jubayr's assertions of the security of non-combatants in Syria even with the ongoing wars, geographical texts written during Syria's divided period frequently mention raiding by both Muslims and Christians. Ibn Hawqal complained of raids not only by the Byzantine army, but also by the Bedouin who surrounded Hims after the Byzantine invasion. (88) He indicated that Byzantine incursions led to an increased use of the inland route from Damascus northward, where a sign of the degeneracy of the age was "the dominance of the Bedouin over the governors," and he expected all travel to cease. (89) Ibn Jubayr himself noted that the khans where travelers lodged in Syria were all heavily fortified, and that the road between Hims and Damascus was largely uninhabited except for a few large villages at the caravan stops. (90) He mentioned Frankish raiding possibilities from Hisn al-Akrad to Hims or Hama, as well as on the road from Damascus to the coast. (91) Although he presented Salah al-Din's capture of Nabulus in 580/1184 as a glorious conquest for Islam, the attack on the unwalled village was clearly more in the nature of a plundering raid. (92) While both Muslim and Crusader rulers may have avoided plundering merchant caravans, on which they found it easier to assess commercial taxes, the rural and semi-rural settled population probably fared worse at their hands.

Nevertheless, significant demographic shifts happened during Syria's divided period. The Byzantine reconquest brought Greek rule back to Syria, but it also brought Armenians who settled in northern Syria and along the Cilician coast. Ibn Hawqal already mentioned that when the Byzantines conquered Malatya in 319/931, they peopled it with Armenians. (93) Al-MuqaddasI noted Armenian control of Jabal al-Lukkam on the northern edge of Syria. (94) Two centuries later Yaqut indicated Armenian control not only over Cilician cities such as 'Ayn Zarba, al-Haruniyya, Sisiyya, and Tarsus, but also a dominantly Armenian population in the fortress of Tal Bashir two days north of Aleppo and the surrounding district of Nahr al-Jawz between Aleppo and al-BIra (modem Birecik) on the Euphrates. (95) Armenians would remain a substantial portion of the population in this region throughout Ottoman times.

On the other hand, the Shiite population of Syria grew during and after the "Shiite century," in part due to Fatimid interests in Syria. Writing after the first invasions but before the Fatimid dynasty held any possessions there, al-Muqaddasi mentioned Shiite populations only in Tiberias, 'Amman, half of Nabulus, and half of Qadas north of Tiberias. (96) A half-century later Nasir-i Khusraw referred to the Shiite population and Egyptian garrison in Tripoli, as well as other Shiite groups in Tyre and in the countryside west of Tiberias, where they prevented Sunni pilgrims from visiting the tomb of Abu Hurayra. (97) Ibn Jubayr remarked on the Ismaili castles in Mount Lebanon, and formerly in al-Bab between Aleppo and the Euphrates. (98) Indeed, he indicated the broad diffusion of Shiites in Syria: "They are more numerous than the Sunnis, and have filled the land with their doctrines." (99) As an exception in northern Syria he praised Manbij, whose population he indicated was entirely Sunni, "so that through them the town is undefiled by those dissident sects and corrupt beliefs that are found in most of this country." (100) The Ismaili castles in the mountains of western Syria and Lebanon would become a refrain of later geographers. (101)

Although most Muslim geographers pay little attention to the Jewish population, the travelogue of Benjamin of Tudela gives approximate Jewish populations for many cities of Syria. Although no other geographical work documents the presence of Jewish communities so extensively, such indications as do exist seem to indicate a marked decline in the Jewish population of Filastln and the coastlands in this period, relative to much larger Jewish populations in the inland portions of central and northern Syria. Thus, al-Muqaddasi had complained of the greater numbers of Jews and Christians than Muslims in his native Jerusalem in the late tenth century, but two centuries later Benjamin of Tudela indicated a population of only 200 Jewish men in the city. (102) Al-BaladhurI mentioned that Mucawiya had settled Tripoli with Jews in the seventh century, but Benjamin of Tudela in the twelfth century only indicated that a recent earthquake had killed many Jews and gentiles; his text does not indicate how many Jews remained in the city. (103) Al-Baladhuri's figure of 20,000 Jews in Qaysariyya, alongside 700,000 soldiers and 30,000 Samaritans, is clearly exaggerated even beyond the extent of the likely influx of rural refugees, but Benjamin of Tudela's indication of about 10 Jewish men in the city still indicates a dramatic decline. (104) No city under Crusader rule had a larger Jewish population than Tyre, with about 500 Jewish households. (105) The figures given for Jewish populations of cities still under Muslim rule stand in stark contrast: 3,000 for Damascus, 5,000 for Aleppo, even 2,000 each for Tadmur on the edge of the desert and Rahba on the Euphrates. (106) This discrepancy and the fact that the Yeshiva of Jerusalem was headquartered in Damascus indicate a preference for medieval Syrian Jews to relocate outside of Crusader control, resulting in a much smaller Jewish population for coastal cities and Filastln. (107)

The religious character of the rural population also shifted markedly in this period. Ibn Jubayr's remarks on Crusader rule over Muslims presumed that many of the Muslims were peasants, a point he made explicit with regard to the population around Baniyas between Damascus and Tyre. (108) Ibn Jubayr's remarks should not be taken to indicate that the rural population throughout Syria had largely converted to Islam, however; the contrary is indicated by his reference to the entirely Christian town of Qara south of Hims, as well as his description of the respect with which Christian peasants treated Muslim hermits. (109) The peasants' veneration for Muslim holy men no doubt contributed to their progressive conversion to Islam, as did the rural shrines shared between Muslims and non-Muslims, such as 'Ayn al-Baqar outside Acre mentioned by Ibn Jubayr and Yaqut, the tomb of the unknown prophet outside Aleppo, and the shrines visited by Nasir-i Khusraw between Acre and Tiberias. (110) The proliferation of shrines dedicated to specifically Muslim figures during this period, such as those enumerated by al-HarawI, may have been partly due to the loss of many urban mosques to Byzantine and Crusader conquests, but it was also partly due to the increase of Shiite Muslims who sometimes took over an urban mosque and sometimes dedicated rural shrines to 'Ali b. Abi Talib. (111) Nasir-i Khusraw described specifically Shiite shrines around Tripoli: "The people of this city are all Shi'ites, and the Shi'ites have built nice mosques in every land. They have edifices there like caravanserais, which they call mashhads, but no one lives in them. Outside the city of Tripoli there is not a single structure except for a couple of mashhads." (112)

Nor was the convergence of religious topography limited to rural settings. Ibn Jubayr mentioned that the main mosque of Acre was shared between Christians and Muslims, (113) and al-Harawi noted that the shrine around the tombs of Abraham, Isaac, Jacob, and Sarah was administered by Greek Christians in the days of Frankish rule, although he as a Muslim visited it. (114) Earlier Nasir-i Khusraw had commented that Muslims as well as Christians and Jews came as pilgrims to Jerusalem. (115) During Crusader rule the Dome of the Rock in Jerusalem was converted into a church, and al-Harawi's account of his visit to the building mentions icons of Solomon and Christ. (116) Despite, or perhaps because of, this sharing of the Umayyad edifice, al-Harawi is the earliest geographer to describe the Rock as the place from which Muhammad ascended on his night journey; Ibn Hawqal and al-Muqaddasi had identified the stone as "the Rock of Moses." (117)

If the appearance of Muslim peasants broke down the older division between Muslim nomads and the non-Muslim sedentary population, the partition of Syria between Christian and Muslim rule generated new axes of Islamization between different regions within Syria. The Byzantine reconquest and the accompanying influx of Armenians probably resulted in the northern edges of Syria containing a higher proportion of Christians than areas further south, such as the Ghuta around Damascus. Ibn Jubayr noted that most villages of the Ghuta had a bathhouse, while Yaqut described the village of 'Imm between Aleppo and Antioch as entirely Christian in his day, although he quoted Ibn Butlan's statement that it had a mosque in the eleventh century. (118) If the Byzantine reconquest introduced a north-south axis, the Crusades resulted in an east-west distinction between coastal areas and inland. This is particularly significant since in the earlier centuries of Islam, the garrisons on the Syrian coast had resulted in an earlier Islamization in precisely the areas where the Crusaders now ruled. But al-Harawi complained of the inability to identify the tombs of early Muslims in the coastal cities due to the Frankish regime. (119) The noteworthiness of the level of rural Islamization in the Ghuta to Ibn Jubayr may reflect the fact that Damascus was one of the few cities of Syria not sacked by the Byzantines or the Crusaders.


The Mamluk conquests of the last Crusader states on the Levantine coast reunited Syria under Muslim rule, although the Armenian kingdom of Cilicia continued to exist as a troublesome vassal of the sultans in Cairo until 776/1375. But three centuries of divided rule had left Syria much more regionally diverse and the non-Muslim populations much reduced. While Crusader rule of the coast had encouraged Christian residence there, the Mamluk conquests were accompanied by deliberate destruction of the coastal settlements. As a result, most references to Christians in the geographical works of the Mamluk period situate them in northern Syria, whether in urban or rural settings. Most cities of northern Syria had a non-Muslim population, and some large towns on major roads continued to be entirely or primarily Christian at least into the fourteenth century.

Unlike the first Muslim conquerors of Syria over six centuries earlier, the new Mamluk rulers did not fortify the coast, but rather depopulated it to a significant degree and rendered it indefensible. The purpose was evidently not to prevent a new Crusader invasion, but to prevent the Crusaders from being able to hold anything on the mainland. Thus, according to Ibn Battuta, the Mamluk sultan Baybars (r. 658-676/1260-1277) destroyed the walls around Antioch when he captured the city in 666/1268. (120) Tripoli was demolished by the Mamluks in 688/1289 and refounded away from the coast. (121) Jubayl and Beirut may have fared somewhat better, with Abu l-Fida' indicating that the former had a Friday mosque and Ibn Battuta indicating the same for the latter. (122) On the other hand, the ports that were the last Crusader strongholds, Tyre and Acre, were completely ruined after the Mamluk conquest, according to Abu l-Fida' who participated in the capture of the latter. (123) Between them, the formerly important port of Sayda was in the fourteenth century merely a small village, while further south Qaysariyya, Arsuf, and 'Asqalan were also ruined. (124) Abu l-Fida' seems to indicate that the main beneficiaries of this coastal destruction were Gaza, which became the main port for traders from the Hijaz, and the Cilician port of Ayas in the Armenian kingdom, which became the preferred port for Christian merchants from Europe; between them, Yafa was the only port he identified as still active. (125) The coastal areas, where the early Islamization was most thoroughly reversed by the period of Crusader rule, were devastated rather than reconverted by the conquerors from Cairo, who generally moved their governmental centers inland.

Under Greek and Crusader rule, the coastline and northern Syria probably had a higher proportion of Christians than other regions. After the Mamluk devastation of the coastline, northern Syria remained the place where Christians were most frequently mentioned by geographers. One interesting exception is al-Shawbak between 'Amman and Ayla in southeastern Syria (today part of Jordan), which Abu l-Fida' described as mostly inhabited by Christians still in the fourteenth century. (126) Otherwise, references to Christian populations in Syria in the geographical works of the Mamluk period all refer to northern Syria. It is indicative in this regard that al-Dimashqi describes the wildest party he knew of as Easter at Hama, for which Christians would gather from all over northern Syria: Hims, Shayzar, Salamiyya, Kafr Tab, Abu Qubays, Masyaf, Ma'arrat al-Nu'man, Tizin, al-Bab, Buza'a, al-Fu'a, and Aleppo. (127) The annual festival at Dayr al-Farus outside Latakia likewise made an impression on al-Dimashqi, Abu l-Fida', and Ibn Battuta. (128) Urban Christian populations are indicated at Antarsus on the coast and in Damascus, where Ibn Battuta recounted the combined prayer procession of Jews, Muslims, and Christians in response to the arrival of the Black Death. (129) It thus appears that most if not all of the urban centers of northern Syria had, or were plausibly reputed to have, Christian segments of the population into the fourteenth century.

Particular rural areas also became known for Christians, such as the "Lake of the Christians" north of Famiya (ancient Apamea), whose eels were enjoyed by Christians, according to al-Dimashqi. (130) On the other hand, Ibn Battuta adapted Ibn Jubayr's description of the villages in the Ghuta around Damascus, adding Friday mosques and markets to the earlier traveler's bathhouses as indicative of what most villages near Damascus possessed in the fourteenth century. (131) In rural areas, roads were important locations of non-Muslim visibility to the Muslim population. One of the main roads from Iraq to Damascus passed through al-Sukhna, east of Hims, and when Ibn Battuta passed that way in the mid-fourteenth century he described it as "a fine town, most of whose inhabitants are Christian infidels." (132) Closer to Damascus, the town of Qara was described by Yaqut as the first stopping point on the straight road from Hims to Damascus, which is the main north-south road of central Syria. (133) Yaqut also indicated tersely that "all of its people are Christians," while a generation earlier Ibn Jubayr had said at greater length that the village "belongs to Christians who dwell there under treaty and in which there are no Muslims." (134) In the fourteenth century Abu l-Fida' modified the description of Qara's population to "predominantly Christian," indicating that this large town on a main road probably had its first Muslim inhabitants under Mamluk rule. (135) Further north, the same author indicated the need to pass through the entirely Christian village of Yaghra near Antioch in order to reach the towns of Darbasak and Baghras, the former of which had a mosque and minbar. (136) As late as the middle of Mamluk rule, Muslim travelers in central and northern Syria might be forced to spend the night in villages with only a small Muslim presence.

Following the Ottoman conquest of the early sixteenth century, tax registers survive that permit a more detailed summary of the state of religious diversity in Syria. These records indicate a small and almost exclusively urban Jewish population, comprising perhaps 2.6% of the population of Aleppo in 924/1518, 6% of Damascus in ca. 950/1543, and around twenty years earlier 11% of Sidon, 1.7% of Beirut, and 2% of Ba'labakk, northwest of Damascus. (137) Further south, Jewish populations made up around 21% of Jerusalem in 932/1526, 24% of Safad, north of Tiberias, in the same year, 10% of Gaza in the same year, 2% of Hebron in ca. 945/1538-9, and 6% of Nabulus in the same register. (138) The centers of Jewish population partially shifted back to the coast and to Jerusalem after the end of Crusader rule, with the addition of the newly prominent Jewish center of Safad. But this shift primarily reflects a decrease in the Jewish populations of inland Syria: the number of Jewish households in Jerusalem in 932/1526 agrees almost exactly with the figure given by Benjamin of Tudela over three centuries earlier. (139) By contrast, the number of Jewish households in Damascus was only a little over one-sixth what the twelfth-century traveler reported, and the Jewish population of Aleppo was a mere 6% of Benjamin of Tudela's figure. (140) Over the course of the sixteenth century, Safad would rise to prominence as the Jewish capital of Syria, until in the 970s/1560s the city had almost twice as many Jewish households as Damascus did, and its total population was almost evenly split between Jews and Muslims. (141)

The Christian population of Syria in the first century of Ottoman rule was increasingly marginalized. In contrast to geographers' reports of substantial Christian populations in northern Syria under the early Mamluks, the Ottoman tax register of 943/1536 for the province of Aleppo presents a total population that was about 98% Muslim and 2% Christian. (142) In that register, Christians were significant minorities in the areas of Jabal al-Aqra' on the coast (8%), al-Shughflr south of Antioch (8%), Shayzar northwest of Hama (8%), and al-Qusayr south of Antioch (5%); only in the district of al-Suwayda' on the Mediterranean coast southeast of Antioch did the Christian population approach the Muslim population (49%). (143) A sizeable rural village such as Zaytuniyya near al-Suwayda', with 82 households and 42 bachelors, could still be 87% Christian, but the village of Yaghra, which Abu l-Fida' had reported as entirely Christian, was entirely Muslim two centuries later. (144) Areas of Syria further north, which had significant Armenian populations in the Mamluk period, were incorporated into the Ramadanid principality or the Dulqadir province and thus not included in the register for the province of Aleppo, but the Christian population around Aleppo is far below what one might expect from the geographers' characterizations. Perhaps the instability of the fifteenth century, when the area around Aleppo was again a frontier zone between the Mamluk rulers in Cairo and their Turkmen neighbors to the north and east, encouraged religious minorities to convert to Islam or emigrate to more stable regions.

Within the province of Damascus, the bulk of the Christian population was rural, although substantial numbers of urban Christians inhabited Gaza, Jerusalem, Beirut, Ba'labakk, and the city of Damascus itself. (145) The town of Qara, the important stopping-place between Damascus and Hims, was slightly over half Muslim by ca. 930/1523, and the surrounding countryside was a little over 60% Muslim. (146) Some villages remained entirely Christian, and in certain towns such as al-Karak, southeast of the Dead Sea, Christians outnumbered Muslims two to one in the middle of the sixteenth century. (147) On the other hand, large areas of the countryside were now entirely Muslim, with no non-Muslim inhabitants in the tax registers, such as almost the entire triangle from Damascus to Beirut and south to Tyre. (148) The areas where non-Muslims outnumbered Muslims were shrinking and were more remote at the beginning of Ottoman rule than when incorporated into the Mamluk empire. The Syria registered by the Ottoman census-takers of the sixteenth century was a Muslim land, in population as well as in government.


When the Umayyad prince Sulayman b. 'Abd al-Malik (d. 99/717) began to construct al-Ramla as the new capital for the district of Filastln in the early eighth century, his expense manager was a Christian scribe from the nearby Roman town of Ludd. (149) Al-Baladhuri, who reported this story in the ninth century, evidently saw nothing out of the ordinary in a Muslim ruler employing a skilled non-Muslim. Later geographers transformed this account in ways that reveal the shifting place of non-Muslims, and particularly Christians, in Syria's society. Al-Muqaddasi in the late tenth century omits the reference to the Christian accountant, but his report of the minaret of al-Ramla being built by Hisham b. 'Abd al-Malik (r. 105-25/724-43) from marble columns that he extorted from the Christians of Ludd, who were hiding them in preparation for enhancing their own church, presents church construction as unremarkable in the latter days of the Umayyad dynasty. (150) In the early thirteenth century, Yaqut al-Hamawi revisited Sulayman's construction of al-Ramla, but in his account the Christian accountant is transformed into a threatening scribe whose desire to obtain the house beside the church is thwarted, whereupon he suggests to Sulayman that he build al-Ramla in order to destroy the church. (151) Finally, in the fourteenth century Abu 1-Fida' expanded on the animosity by stating that the Umayyad prince destroyed (akhrabaha) Ludd and founded alRamla. 152 Nevertheless, in the early Ottoman tax register, Ludd was approximately as large as al-Ramla, and was still around 40% Christian. (153) Even where non-Muslims continued to be a substantial portion of the population in Syria, the non-Muslim role in the construction of a district capital was simply erased.

Despite the hermeneutical challenges it poses, the medieval Muslim geographical literature is a rich body of source material for social history, and particularly for the history of Islamization. These works remind us that Islamization was more than just the progressive conversion to Islam of the populace of Syria, but included the construction or conversion of mosques and the diffusion into the countryside. In the earliest period, Islam seems to be the religion only of the ruling elites and garrisons in cities and coastal towns, and some of the nomadic Arabs in the countryside. The evidence for sedentary rural Muslim populations, and for Muslim shrines outside of the cities, begins only in the geographies of the tenth century, and grows quickly in the subsequent period. The invasions of the Byzantines and the Crusaders divided Syria between Christian and Muslim rule, which reversed the trend of Islamization in areas under their rule, particularly in northern Syria and along the coast. Frankish governance in particular seems to have encouraged Jews to settle outside of Filastin and inland, under Muslim rule. The division of government created two axes of differential Islamization within Syria, one north-south and the other coastal-inland. With the Mamluk devastation of the coast in the process of conquest, geographical works of the period came to mention Christian populations and institutions primarily in the urban and rural areas of northern Syria. Although Islam was now as much a peasant's religion as a ruler's, pockets of Syria remained entirely or largely non-Muslim to the end of Mamluk rule.

In light of the long duration process of Islamization, Bulliet's conversion curve seems too steep and too early to represent the population of Syria as a whole. This is unsurprising for a graph based on the naming practices of the ancestors of ulema, which likely represent families that have been Muslim on average longer than families without Islamic religious experts. (154) The one-dimensional increase of Muslims as a proportion of the population, an artifact of Bulliet's use of a summative S-curve, has been shown to be false for portions of Syria that came under Byzantine or Crusader rule, and the late proliferation of specifically Islamic rural shrines in the twelfth century, which was likely a contributing cause rather than an effect of the conversion of the rural population, reveals that Bulliet's date of 1010 (1979: 131) for the "essential completion" of "the primary conversion process" is too early. The use of geographical works has also permitted a more locally nuanced account of the process of Islamization within one region considered by Bulliet, showing that different parts of Syria experienced very different trajectories of Islamization. (155)

This study largely supplements the work of Levtzion and enriches our understanding of the process of conversion, where he was compelled by his sources to speak in generalities and, with regard to Islamization in particular, primarily about shifting government attitudes. The effect of religious spaces shared among Muslims, Jews, and Christians on the Islamization of Syria adds a new dimension to Levtzion's account (1990: 297-99) of conversion to Islam largely in terms of governmental pressures and mob violence. The partial reversal of Islamization due to Byzantine and Frankish rule also puts a significant question mark next to his assertion (p. 299) that Middle Eastern Christians preferred Muslim to Christian rule. Some Christians who were particularly negatively affected by Byzantine or Crusader rulers likely did applaud the Mamluk conquests, but the areas that remained under Muslim rule experienced rural Islamization faster than areas conquered by Christians.

Research into Islamization has often focused on trying to identify an "age of conversions" in which the majority of a population had adopted Islam, sometimes even referring to the achievement of demographic majority as a "tipping point." This study has omitted any speculations regarding the date at which Islam became the religion of the demographic majority for three reasons. One is simply the lack of evidence that could validate such speculation. More importantly, such speculation ascribes an unwarranted significance to demographic precision and presumes an erroneous shape of Islamization. Premodern authors remarked on the relative size of religious groups, not with respect to large regions such as Syria, but with respect to specific locales such as particular cities or towns. (156) This is typical for phenomena that had no imperial significance but could affect daily life. However, it is difficult to see how the experience of living in a locale with large groups of non-Muslims would be noticeably different whether Muslims constituted 45% or 55% of the total population. No "tipping point" was proposed in this study because the conversion of 50.1% of the population had no significance. That is the second reason. Thirdly, Islamization was a complex process with different dynamics in different periods. To suggest an approximate date for a Muslim demographic majority, in Syria or any sub-region, would imply that Islamization was merely demographic, and only quantitatively different at different times. Islamization includes demographic change, but the qualitatively different character of processes of Islamization in different periods, from early conversions of Arab nomads and garrisons of major cities to the late medieval proliferation of Islamic shrines, cannot be reduced to proportions of populations.


Abu 1-Fida' Isma'Il b. 'Ali. 1840. Geographie d'Aboulfeda. Ed. J. T. Reinaud and W. MacGuckin de Slane. 2 vols. Paris: Imprimerie Royale.

Adamnan. 1958. De locis sanctis. Ed. D. Meehan. Dublin: Dublin Institute for Advanced Studies.

Antrim, Zayde. 2012. Routes and Realms: The Power of Place in the Early Islamic World. Oxford: Oxford Univ. Press.

Bakhit, Muhammad Adnan. 1982a. The Christian Population of the Province of Damascus in the Sixteenth Century. In Christians and Jews in the Ottoman Empire: The Functioning of a Plural Society, ed. B. Braude and B. Lewis. Vol. 2: 19-66. New York: Holmes and Meier.

--. 1982b. The Ottoman Province of Damascus in the Sixteenth Century. Beirut: Librairie du Liban.

Bakhit, Muhammad 'Adnan, and Nufan Raja Hamud, eds. 1989. Daftar mufassal Liwa 'Ajlun: Tabu daftari raqm 970, Istanbul. Amman: al-Jami'a al-Urduniyya.

--, eds. 1991. Daftar mufassal Liwa 'Ajlun: Tabu daftari raqm 185 Anqara sanat hijri 1005 muwafiq miladi 1596. Amman: al-Jami'a al-Urduniyya.

al-Baladhuri, Ahmad b. Yahya. 1916. The Origins of the Islamic State. Tr. P. K. Hitti. New York: Columbia Univ. Press.

--. 1957. Futuh al-buldan. Ed. 'A. A. A. al-Tabba' and CU. A. al-Tabba'. Beirut: Dar al-Nashr li-l-Jami'iyyin.

Benjamin of Tudela. 1907. The Itinerary of Benjamin of Tudela: Critical Text, Translation and Commentary. Ed. M. N. Adler. New York: P. Feldheim.

Bulliet, Richard W. 1979. Conversion to Islam in the Medieval Period: An Essay in Quantitative History. Cambridge, Mass.: Harvard Univ. Press.

Chabot, Jean Baptiste, ed. 1917. Chronicon ad annum Christi 1234 pertinens. 2 vols. Paris: E Typographeo Reipublicae.

Cohen, Amnon, and Bernard Lewis. 1978. Population and Revenue in the Towns of Palestine in the Sixteenth Century. Princeton: Princeton Univ. Press. al-Dimashqi, Shams al-Dln Muhammad b. Abl Talib. 1866. Cosmographie de Chems-ed-Din Abou Abdallah Mohammed ed-Dimichqui. Ed. C. M. Fraehn and A. F. Mehren. St. Petersburg: Academie Imperiale des Sciences.

--. 1874. Manuel de la cosmographie du Moyen Age. Tr. A. F. Mehren. Copenhagen: C. A. Reitzel.

Elad, Amikam. 1995. Medieval Jerusalem and Islamic Worship: Holy Places, Ceremonies, Pilgrimage. Leiden: E. J. Brill.

Ellenblum, Ronnie. 1998. Frankish Rural Settlement in the Latin Kingdom of Jerusalem. Cambridge: Cambridge Univ. Press.

Gil, Moshe. 1992. A History of Palestine, 634-1099. Cambridge: Cambridge Univ. Press.

Grabar, Oleg. 2006. The Dome of the Rock. Cambridge, Mass: Belknap Press of Harvard Univ. Press.

Haneda, Masashi, and Torn Miura, eds. 1994. Islamic Urban Studies: Historical Review and Perspectives. London: Kegan Paul International.

al-Harawi, 'Ali b. Abi Bakr. 1953. Guide des lieux de pelerinage. Ed. and tr. J. Sourdel-Thomine. 2 vols. Damascus: Institut Francais de Damas.

--. 2004. A Lonely Wayfarer's Guide to Pilgrimage: (Ali ibn Abl Bakr al-Harawi's Kitab al-Isharat ila Ma'rifat al-Ziyarat. Tr. J. W. Meri. Princeton: Darwin Press.

Harrison, Alwyn. 2012. Behind the Curve: Bulliet and Conversion to Islam in Al-Andalus Revisited. Al-Masaq: Islam and the Medieval Mediterranean 24,1: 35-51.

Humphreys, R. Stephen. 2010a. Christian Communities in Early Islamic Syria and Northern Jazira: The Dynamics of Adaptation. In Money, Power and Politics in Early Islamic Syria, ed. J. Haldon. Pp. 45-56. Surrey: Ashgate.

--. 2010b. Syria. In The New Cambridge History of Islam, ed. C. F. Robinson. Pp. 506-40. Cambridge: Cambridge Univ. Press.

Ibn Battuta, Muhammad. 1958. Travels of Ibn Battuta, A.D. 1325-1354. 5 vols. Tr. H. A. R. Gibb and C. F. Beckingham. Cambridge: Hakluyt Society.

--. 1964. Rihlat Ibn Battuta al-musamma Tuhfat al-nuzzar fighara'ib al-amsar wa-'aja'ib al-asfar. 2 vols. Misr: al-Maktaba al-Tijariyya al-Kubra.

Ibn Hawqal, Muhammad b. 'Ail. 1964a. Kitab Surat al-ard. Beirut: Dar Maktabat al-Hayah.

--. 1964b. Configuration de la terre (Kitab surat al-ard). Tr. J. H. Kramer and G. Wiet. 2 vols. Beirut: Commission Internationale pour la Traduction des Chefs-d'oeuvre.

Ibn Jubayr, Muhammad b. Ahmad. 1952. The Travels of Ibn Jubayr. Tr. R. J. C. Broadhurst. London: Jonathan Cape.

--. 1981. Risalat i'tibar al-nasik fidhikr al-athar al-karima wa-l-manasik al-ma'ruf bi-Rihlat Ibn Jubayr. Rev. ed. Beirut: Dar wa-Maktabat al-Hilal.

Ibn Munqidh, Usama. 1930. Kitab al-I'tibar. Ed. P. K. Hitti. Princeton: Princeton Univ. Press.

--. 2008. The Book of Contemplation: Islam and the Crusades. Tr. P. M. Cobb. London: Penguin, 2008.

Kedar, Benjamin Z. 1985. Muslim Conversion in Canon Law. In Proceedings of the Sixth International Congress of Medieval Canon Law: Berkeley, California, 28 July-2 August 1980, ed. S. Kuttner and K. Pennington. Pp. 321-32. Vatican City: Biblioteca Apostolica Vaticana.

--. 1990. The Subjected Muslims of the Frankish Levant. In Muslims under Latin Rule, 1100-1300, ed. J. M. Powell. Pp. 135-74. Princeton: Princeton Univ. Press, 1990.

--. 1997a. Multidirectional Conversion in the Frankish Levant. In Varieties of Religious Conversion in the Middle Ages, ed. J. Muldoon. Pp. 190-99. Gainesville: Univ. Press of Florida.

--. 1997b. Some New Sources on Palestinian Muslims before and during the Crusades. In Die Kreuzfahrerstaaten als multikulturelle Gesellschaft: Einwanderer und Minderheiten im 12. und 13. Jahrhundert, ed. H. E. Mayer. Pp. 129-40. Munich: R. Oldenbourg.

Kennedy, Hugh. 1985. From Polis to Madina: Urban Change in Late Antique and Early Islamic Syria. Past and Present 106 (February): 3-27.

Khalek, Nancy. 2011. Damascus after the Muslim Conquest: Text and Image in Early Islam. Oxford: Oxford Univ. Press.

Kister, M. J. 1997. Land Property and Jihad: A Discussion of Some Early Traditions. In idem, Concepts and Ideas at the Dawn of Islam. Pp. 270-311. Aldershot: Ashgate.

Lapidus, Ira M. 1967. Muslim Cities in the Later Middle Ages. Cambridge, Mass.: Harvard Univ. Press.

--. 1969. Muslim Cities and Islamic Societies. In Middle Eastern Cities: A Symposium on Ancient, Islamic, and Contemporary Middle Eastern Urbanism, ed. I. M. Lapidus. Pp. 47-74. Berkeley and Los Angeles: Univ. of California Press. el-Leithy, Tamer. 2005. Coptic Culture and Conversion in Medieval Cairo, 1293-1524 A.D. Ph.D. diss., Princeton Univ.

Le Strange, Guy. 1890. Palestine under the Moslems: A Description of Syria and the Holy Land from A.D. 650 to 1500. London: Alexander P. Watt.

Levtzion, Nehemia. 1990. Conversion to Islam in Syria and Palestine and the Survival of Christian Communities. In Conversion and Continuity: Indigenous Christian Communities in Islamic Lands. Eighth to Eighteenth Centuries, ed. M. Gervers and R. J. Bikhazi. Pp. 289-311. Toronto: Pontifical Institute of Mediaeval Studies.

MacEvitt, Christopher. 2008. The Crusades and the Christian World of the East: Rough Tolerance. Philadelphia: Univ. of Pennsylvania Press, 2008.

Meri, Josef W. 2002. The Cult of Saints among Muslims and Jews in Medieval Syria. Oxford: Oxford Univ. Press.

Miquel, Andre, tr. 1963. Al-Muqaddasi, Ahsan at-taqasim fi ma'rifat al-aqalim (la meilleure repartition pour la connaissance des provinces). 4 vols. Damascus: Institut Francais de Damas.

--. 1967. La Geographie humaine du monde musulman. Paris: Ecole des Hautes Etudes en Sciences Sociales.

Morony, Michael G. 1990. The Age of Conversions: A Reassessment. In Conversion and Continuity: Indigenous Christian Communities in Islamic Lands. Eighth to Eighteenth Centuries, ed. M. Gervers and R. J. Bikhazi. Pp. 135-50. Toronto: Pontifical Institute of Mediaeval Studies.

al-Muqaddasi, Muhammad b. Ahmad. 1906. Descriptio imperii moslemici. Ed. M. J. de Goeje. 2nd ed. Leiden: E. J. Brill.

--. 1994. The Best Divisions for Knowledge of the Regions: A Translation of Ahsan al-taqasim fi ma'rifat al-aqalim. Tr. B. A. Collins. Reading, UK: Centre for Muslim Contribution to Civilization.

Nasir-i Khusraw. 1975. Safarnamah-i hakim Nasir-i Khusraw Qubadiyani Marvazi. Ed. M. D. Siyaqi. Tehran: Anjuman-i Asar-i Milli.

--. 1986. Naser-e Khosraw's Book of Travels. Tr. W. M. Thackston. Albany: Bibliotheca Persica, 1986.

Nasrallah, Joseph. 1992. De la Cathedrale de Damas a la mosque omayyade. In La Syrie de Byzance a l'Islam: VIIe-VIIIe siecles, ed. P. Canivet and J.-P. Rey-Coquais. Pp. 139-44. Damascus: Institut Francais de Damas, 1992.

Sanjian, Avedis K., tr. 1969. Colophons of Armenian Manuscripts, 1301-1480. Cambridge, Mass.: Harvard Univ. Press.

Sener, Murat, and Salih Dutoglu, eds. 2010. 397 numarali Haleb livasi mufassal tahrir defteri (94311536). Ankara: T. C. Basbakanlik Devlet Arsivleri Genel Mudurlugu.

Simonsohn, Uriel. 2013. Conversion to Islam: A Case Study for the Use of Legal Sources. History Compass 11,8: 647-62.

Talmon-Heller, Daniella. 2007a. Graves, Relics and Sanctuaries: The Evolution of Syrian Sacred Topography (Eleventh-Thirteenth Centuries). ARAM 19: 601-20.

--. 2007b. Islamic Piety in Medieval Syria: Mosques, Cemeteries and Sermons under the Zan gids and Ayyubids (1146-1260). Leiden: Brill.

Tannous, Jack B. V. 2008. Between Christology and Kalam? The Life and Letters of George, Bishop of the Arab Tribes. In Malphono w-Rabo d-Malphone: Studies in Honor of Sebastian P. Brock, ed. G. A. Kiraz. Pp. 671-716. Piscataway, N.J.: Gorgias Press, 2008.

Walker, Bethany J. 2013. The Islamization of Central Jordan in the 7th-9th Centuries: Lessons Learned from Hisban. Jerusalem Studies in Arabic and Islam 40: 143-75.

al-Ya'qubi, Ahmad b. Abi Ya'qub. 1918. Kitab al-Buldan. Ed. M. Sadiq. Najaf: al-Maktaba al-Murtadiyya wa-Matba'atuha al-Haydariyya, 1918.

Yaqut b. 'Abd Allah al-Hamawi. 1990. Mucjam al-buldan. Ed. F. 'A. 'A. al-Jundi. 7 vols. Beirut: Dar al-Kutub al-'Ilmiyya.


Oklahoma State University

Earlier versions of this article were greatly improved by suggestions from Michael Cook, the late Patricia Crone, Peter Brown, Christian Sahner, the Princeton Islamic Studies Colloquium, and the JAOS anonymous reviewers. The author records his gratitude for their corrections and recommendations, while acknowledging that all remaining errors are his own.

(1.) An account of the Ottoman conquest is found in Bakhit 1982b: 1-34.

(2.) Scholarship on the Ottoman defiers remains uneven. The 1536 census records of Aleppo have been published (Sener and Dutoglu 2010); I have not found any edition or analysis of defters from Tripoli. Analyses of the defters from Damascus and 'Ajlun, two of the nine districts (sg. sanjak) of the province of Damascus, have been published in Bakhit 1982b; Bakhit and Hamud 1989 and 1991. Fifteen defters for four additional sanjaks of the province of Damascus were analyzed in Cohen and Lewis 1978. Finally, Bakhit 1982a is a synthesis of the Christian portion of the non-Muslim population of the province of Damascus as a whole.

(3.) Multiple tax registers were compiled at different times for different portions of Syria, making it impossible to speak of proportions of the total population of the region at one time. Tapu defteri 401 for the district of Damascus in ca. 950/1543 seems to indicate a population approximately 90% Muslim, 9% Christian, and 1% Jewish, based on tables in Bakhit 1982b: 37-89. The tax registers do not record total population, but rather households (khana) and bachelors (mujarrad) for each religious group. Thus, proportions of the population are necessarily approximate, depending on the unknown average number of people per household in each community. Most scholars use a figure of about five people per household to estimate total population, but as long as the household size did not vary significantly across religious boundaries, the precise multiplier should not greatly affect calculations of the proportion of a population that was non-Muslim.

(4.) The classic studies of urbanism in Islamic society are Kennedy 1985; Lapidus 1967. For a more complete historiography to 1994, and a critique of "Islamic city" as a category, see Haneda and Miura 1994.

(5.) The lengthy process of the conversion of holy sites is discussed by Talmon-Heller 2007b: 188-90.

(6.) The twelfth-century Andalusi traveler Ibn Jubayr (1981: 253-54; 1952: 323-24) reported the conversion of a traveling companion in Syria to Christianity. A fifteenth-century Turkmen ruler of the Aqquyunlu confederation is credited with building a church in eastern Anatolia (Sanjian 1969: 205).

(7.) Al-Muqaddasi 1994: 139; 1906: 152.

(8.) Yaqut al-Hamawi 1990, 4: 177 (the date of 540/1145 for the Risala is incorrect, being about a century too late).

(9.) For an example of Fatimid financial support for Palestinian Jews in the late tenth century, see Gil 1992: 551. For an example of Zangi giving a pair of church bells to a Syrian Orthodox church in al-Ruha, see Chabot 1917, 2: 136.

(10.) Kedar 1985: 326.

(11.) Kedar 1997a: 196.

(12.) Yaqut al-Hamawl 1990, 3: 355.

(13.) Ibn Hawqal 1964a: 154; 1964b, 1: 164-65.

(14.) Yaqut al-Hamawl 1990, 3: 354.

(15.) Alwyn Harrison (2012: 38) suggested that Bulliet's graphs are often misunderstood to refer to the total population, when in fact they refer only to the percentages "of those who would convert--of the ultimate unquantifiable total of converts" and therefore "There is thus no way to extrapolate any quantifiable data regarding conversion, or to identify the point at which Muslims became a numerical majority and the ahl al-dhimma a minority." This interpretation picks up on certain nuances of Bulliet's language, but Bulliet himself (1979: 1) seems to slip into identifying the conversion curve with broader demographics, for example in his conclusion of "a causal relationship between the conversion of a majority of a region's population and the dissolution of central Islamic government in that region" (emphasis mine). It is also unclear how the "stage in the conversion process" to which Bulliet frequently ascribes causal force would be comparable across countries if in one region it represented a large majority of the population and in another conversion rates dwindled after the Muslim population reached around 20%. Michael Morony (1990: 136-37, 138) critiques Bulliet's use of the "conversion curve" as a cause rather than a consequence, but asserts that Bulliet's curves represent the conversion of the population as a whole, albeit with some reservations.

(16.) The adverb is necessary: Arabic geographical texts do not devote much space to Mount Lebanon, which continued to have a substantial Maronite Christian population to the present; see Levtzion 1990: 306-7.

(17.) For a recent study of the literary aspects of this "discourse of place," see Antrim 2012.

(18.) Al-Baladhuri 1957: 170, 174, 183, 187, 192; 1916: 190-91, 195, 206, 211, 217.

(19.) Gil 1992: 8-10,70-71.

(20.) Al-Baladhuri 1957: 215-17; 1916: 244-45.

(21.) Antrim (2012: 72) indicates that some geographers use sources without acknowledgment after having specifically criticized them elsewhere.

(22.) The performative aspect of geographical literature is cogently stated by Antrim 2012: 3. Antrim critiques Miguel 1967 for neglecting this literary dimension and simply mining the sources for information. Guy Le Strange (1890) had earlier synthesized the geographical literature with regard to Syria, but in dicing up the sources, he rendered it impossible to engage with the texts as literature. In the latter trait, however, he merely follows in the footsteps of Yaqut al-Hamawi.

(23.) Al-Baladhuri 1957: 185-86, 198; 1916: 209, 224.

(24.) Al-Baladhuri 1957: 249-52; 1916: 284-86. Richard Bulliet (1979: 106) also suggests that the early Syrian ulema included in later biographical dictionaries were mostly Arab. However, his conclusion is based upon the preponderance of ulema from inland as opposed to coastal Syria, and he assumes that people from an area including Damascus should be presumed to be Arab. The precedence of the Bedouin in conversion to Islam is also indicated, with bibliography, by Humphreys (2010a: 49).

(25.) Al-Baladhuri 1916: 285; 1957: 250.1 have emended Hitti's translation.

(26.) Al-Baladhuri 1957: 185; 1916: 209.

(27.) Al-Baladhuri 1957: 205; 1916: 232.

(28.) Early Muslim aversion to farming is indicated by a number of hadith analyzed by M. J. Kister (1997, 4: 270-86). That nomads did not overwhelmingly adopt Islam at the time of the initial conquests is indicated by the existence of a Syrian Orthodox bishop "of the Arabs" consecrated in 686 who died in 724 (Tannous 2008).

(29.) For example, regarding the garrisons of the sea coast under 'Umar b. al-Khattab or 'Uthman b. 'Affan, see al-Baladhuri 1957: 173; 1916: 195. In this regard it is revealing that several of the hadith analyzed by Kister (1997, 4: 286-90) presume that Muslims are urban dwellers.

(30.) Walker (2013: 148-49) indicates that the conversion of a church into a mosque was quite rare in central Jordan; abandoned churches would become private residences far more commonly.

(31.) Al-Baladhuri 1957: 189; 1916: 201; al-Muqaddasi 1906: 156; 1994: 144; Yaqut al-Hamawi 1990, 2: 348.

(32.) Khalek 2011: 96-97. Joseph Nasrallah (1992) cites the standard sources for the received narrative, although with no critical engagement.

(33.) Al-Baladhuri 1957: 200; 1916: 226.

(34.) Al-Baladhuri 1957: 181; 1916: 204.

(35.) Al-Baladhuri 1957: 175; 1916: 196.

(36.) Al-Baladhuri 1957: 195, 226, 232; 1916: 220, 255, 262.

(37.) Al-Baladhuri 1957: 160-61, 201; 1916: 180, 228.

(38.) For example, al-Muqaddasi 1906: 169; 1994: 154. A recent architectural history of the building is Grabar 2006.

(39.) Adamnan 1958: 42-43.

(40.) Al-Muqaddasi 1906: 172; 1994: 156; al-Harawi 1953, 1: 29; 2: 70; 2004: 76-77; Yaqut al-Hamawi 1990, 1:618.

(41.) Al-Baladhuri 1957: 171; 1916: 191-92.

(42.) Ibn Hawqal 1964a: 162; 1964b, 1: 173.

(43.) Al-Muqaddasi 1906: 156; 1994: 144.

(44.) Yaqut al-Hamawi 1990, 2: 349.

(45.) Lapidus (1969: 57) asserted that in this period across the Muslim-ruled world "Muslim cities were isolated in Christian, Zoroastrian, or pagan countrysides," but this view neglects both the nomadic Arab Muslims and the fact that cities taken over by Muslim conquerors continued to have a non-Muslim majority for an indeterminate period.

(46.) Ibn Hawqal 1964a: 159; 1964b, 1: 169.

(47.) Al-Muqaddasi 1906: 176; 1994: 160, where "al-Jazira" is translated as "the Arabian Peninsula," but the region referred to by this phrase is more commonly upper Mesopotamia. Andre Miquel's French translation (1963: 208) likewise rendered "al-Jazira" with reference to Arabia, and while it is the more likely interpretation, I have reverted to transliterating the Arabic to preserve the ambiguity.

(48.) Al-Muqaddasi mentioned mosques in al-Ramla, Dajun nearby, Jerusalem, Jabal Zayta (Mount of Olives) outside Jerusalem, Hebron, al-Yaqin outside Hebron, Gaza, 'Asqalan, Yafa, Arsuf nearby, Qaysariyya, Nabulus, Jericho, 'Amman, Ludd outside al-Ramla, Kafr Saba on the road to Damascus, 'Aqir on the road to Mecca, Yubna near the coast, 'Amawas between al-Ramla and Jerusalem, and Kafr Sallam near Qaysariyya on the coastal road; see al-Muqaddasi 1906: 165-66, 168-77, 182; 1994: 151, 153-60, 165.

(49.) Humphreys (2010b: 533-34) suggests a more rapid Islamization of Syria south of Hims than in the north or in al-Jazira. While this may be the case, the restriction of mosques to cities and a few large villages implies regions devoid of Muslim inhabitants. If al-Muqaddasi's list of mosques is comprehensive, then the city of Bayt Jibril between Hebron and 'Asqalan not only lacked a mosque, but was not within 20 km of a mosque. The hill country between Jerusalem and Nabulus also lacked any mosques, in precisely the area found by Ellenblum (1998: 283) to be dominated by Christian settlements two centuries later.

(50.) Al-Muqaddasi 1906: 176-77; 1994: 160.

(51.) Al-Muqaddasi 1994: 160; 1906: 176.

(52.) Al-Muqaddasi 1906: 184; 1994: 167.

(53.) His list is found in al-Muqaddasi 1906: 151; 1994: 138.

(54.) Josef Men (2002: 195-201, 210-12, 243-50) discusses such shared shrines; Christopher MacEvitt (2008: 126-30, 132-34) discusses the sharing of churches between Franks and Middle Eastern Christians, which could be every bit as awkward.

(55.) For discussion of the increasing number of Muslim shrines in medieval Syria, see Talmon-Fleller 2007b: 190-95; Meri 2002: 257-62.

(56.) Nasir-i Khusraw 1986: 17; 1975: 26.

(57.) Nasir-i Khusraw 1975: 26-31; 1986: 17-19.

(58.) Nasir-i Khusraw 1975: 29-31; 1986: 18-19.

(59.) Nasir-i Khusraw 1975: 25-26; 1986: 17.

(60.) Al-Harawl 2004: 12-13; 1953, 1: 4; 2: 9.

(61.) Al-Harawi 1953, 1: 7, 23; 2: 14, 16, 59; 2004: 16, 66.

(62.) Al-Harawi 1953, 1: 11-13; 2: 27-32; 2004: 24-30. For Dihya al-Kalbi's other tombs, see al-Harawi 1953, 1: 20, 37; 2: 52, 87; 2004: 40, 98.

(63.) Al-Harawi 2004: 82; 1953, 1: 33; 2: 76.

(64.) Al-Harawi (1953, 1: 5-6; 2: 10-11; 2004: 12-14) mentioned various Muslim shrines in and around Aleppo itself, but apart from a tomb in the village of Ruhln that he identified as belonging to Quss b. Sa'ida al-Iyadi, all of the other shrines were dedicated to Jewish or Christian figures. Indeed, he reported that even the shrine of Ruhin was alternatively identified as belonging to Christian figures.

(65.) For a contrary view, Nancy Khalek (2011: 123) acknowledges that a specifically Islamic sacred landscape "was slowly being built up" over centuries, but contends that "Tombs of fallen Companions who had served in the conquests were the first elements of that new environment." Meri (2002: 257-58) suggests that early monuments to fallen Companions had very little connection to later medieval pilgrimage shrines.

(66.) Ibn Hawqal 1964a: 162-65, 167; 1964b, 1: 173-77, 179-80.

(67.) Ibn Hawqal 1964a: 165.

(68.) Ibn Hawqal 1964a: 162, 164; 1964b, 1: 173, 176.

(69.) Al-Muqaddasi 1906: 152; 1994: 140.

(70.) Al-Muqaddasi 1994: 139; 1906: 152.

(71.) Ibn Hawqal 1964a: 172.

(72.) Al-Muqaddasi 1994: 139; 1906: 152. The translator's rendering of jihad as "holy strife" is problematic and has been changed here to the more neutral term "struggle." Andre Miquel (1963: 154 n. 52) interpreted this sentence as an "allusion aux Juifs et au Chretiens, sujets proteges," in defiance of the context.

(73.) For example, in the city of Aleppo; see Ibn Hawqal 1964a: 163; 1964b, 1: 174.

(74.) For one synthesis of the evidence for Muslim subjects of the Crusader states, see Kedar 1990. Kedar (p. 145) presumes that the majority of the population under Frankish rule was Muslim, while Christopher MacEvitt (2008: 12), who analyzes the Frankish treatment of their Eastern Christian subjects, asserts without citation that Christians were the majority of the population in "northern Syria," a term that for him includes northwestern Mesopotamia.

(75.) Al-Harawi 1953, 1: 32; 2: 76; 2004: 82.

(76.) Ibn Jubayr 1952: 321; 1981: 252.

(77.) Al-Harawi 1953, 1: 25-26, 29; 2: 63-65, 70; 2004: 72, 76.

(78.) Ibn Jubayr 1981: 249; 1952: 318-19.

(79.) Ibn Jubayr 1981: 250, 252; 1952: 319, 321.

(80.) Ibn Jubayr 1981: 247-48; 1952: 316-17.

(81.) These views counterbalance the evidence of Usama Ibn Munqidh, whose Kitab al-lctibar mentions nonMuslims primarily in narratives of battles with Franks. But his book is almost exclusively concerned with elites, whether Muslim or Frankish, and battles loom large among the anecdotes related. Even Ibn Munqidh, however, indicated that the Franks who had been in Syria longer harassed Muslims less; see Ibn Munqidh 1930: 134-35, 140; 2008: 147, 153.

(82.) Ibn Jubayr 1952: 318; 1981: 249. Ibn Jubayr's description of sharing the mosque over 'Ayn al-Baqar seems to refute al-Harawi's contention (1953,1: 22; 2: 57; 2004: 44) that the Franks intended to make it a church, but were thwarted by 'Ali b. Abi Talib mystically killing their night-watchmen. Yaqut (1990, 4: 199) likewise alluded to the fact that 'Ayn al-Baqar was venerated by Muslims and non-Muslims alike, including Jews in his list of worshippers there.

(83.) Ibn Jubayr 1952: 301; 1981: 235.

(84.) Ibn Jubayr 1952: 322; 1981: 252.

(85.) Ibn Jubayr 1981: 247; 1952: 316.

(86.) Ibn Jubayr 1981: 247; 1952: 316.

(87.) Ibn Jubayr 1981: 234-35; 1952: 300-301.

(88.) Ibn Hawqal 1964a: 162-63; 1964b, 1: 173.

(89.) Ibn Hawqal 1964a: 165; 1964b, 1: 176.

(90.) Ibn Jubayr 1981: 205, 209-10; 1952: 264, 269.

(91.) Ibn Jubayr 1981: 206, 209, 246; 1952: 265, 268, 315.

(92.) Ibn Jubayr 1981: 245; 1952: 314.

(93.) Ibn Hawqal 1964a: 166; 1964b, 1: 179.

(94.) Al-Muqaddasi 1906: 189; 1994: 172.

(95.) Yaqut al-Hamawi 1990, 2: 47, 213; 3: 338; 4: 33, 201; 5: 446.

(96.) Al-Muqaddasi 1906: 179; 1994: 162-63.

(97.) Nasir-i Khusraw 1975: 21, 24, 30-31; 1986: 13, 16, 19.

(98.) Ibn Jubayr 1981: 202, 206; 1952: 259-60, 264.

(99.) Ibn Jubayr 1952: 291; 1981: 227.

(100.) Ibn Jubayr 1952: 259; 1981: 201.

(101.) Yaqut al-Hamawi 1990, 2:77, 119; 3: 243; 4: 535; 5: 168; al-Dimashqi 1866: 200, 202-3, 208, 209; 1874: 269, 274, 282-84, 286; Ibn Battuta 1964, 1: 44-45; 1958, 1: 106; Abu 1-Fida' 1840, 1: 229-30; 2,2: 7.

(102.) Al-Muqaddasi 1906: 167; 1994: 152; Benjamin of Tudela 1907: 22, JD. As for the latter, the numbers given in the Hebrew are usually of yahadlm, which is ambiguous as to whether it includes women or refers only to Jewish men. Despite the term being translated as "Jews" throughout, it is suggested (1907: 16 n. 2) that only male heads of households were in view.

(103.) Al-Baladhuri 1957: 174; 1916: 195; Benjamin of Tudela 1907: 17, U'.

(104.) Al-Baladhuri 1957: 192; 1916: 217; Benjamin of Tudela 1907: 20, XD.

(105.) Benjamin of Tudela 1907: 18, D.

(106.) Benjamin of Tudela 1907: 30-32, 34.

(107.) Benjamin of Tudela 1907: 30.

(108.) Ibn Jubayr 1981: 246-48; 1952: 315-17.

(109.) Ibn Jubayr 1981: 209, 233-34; 1952: 269, 300. Ellenblum (1998: 283) also concluded on the basis of archaeological evidence that in the twelfth century Palestine still had large rural areas dominated by Christians and other rural areas dominated by Muslims. The former included western Galilee around Acre and the hill country north of Jerusalem, while the latter included eastern Galilee and central Samaria. Kedar (1997b: 138) pointed out that Ellenblum's finding of religious segregation in Crusader-ruled Palestine is consistent with the absence of evidence for personal ties between Muslim peasants and the Franks according to an Islamic hagiographie text. It may, however, be objected that Kedar's interpretation is an argument from silence; reporting such relationships may not have served the hagiographer's purpose.

(110.) See nn. 57-58, 60, and 82, above. Although Ibn Jubayr (1981: 254; 1952: 324) mentions very few shrines, he likewise mentions tombs dedicated to ancient Jewish figures around Tiberias.

(111.) 'Ayn al-Baqar outside Acre was dedicated to 'All b. Abi Talib, for example (al-Harawi 1953, 1: 22; 2: 57; 2004: 44). Ibn Battuta (1964, 1: 39; 1958, 1: 94) mentioned a Shiite mosque in the village of Sarmin southwest of Aleppo.

(112.) Nasir-i Khusraw 1986: 13; 1975: 21.

(113.) Ibn Jubayr 1981: 249; 1952: 318.

(114.) Al-Harawi 1953, 1: 30-31; 2: 72-73; 2004: 78.

(115.) Nasir-i Khusraw 1975: 34-35, 62-63; 1986: 21, 37-38.

(116.) Al-Harawi 1953, 1: 24-25; 2: 62-63; 2004: 70; Grabar 2006: 160-69. The Dome of the Rock is likely the "small mosque that the Franks had converted into a church" mentioned by Usama Ibn Munqidh (2008: 147; 1930: 134-35), in which he would pray with permission from the Templars.

(117.) Ibn Hawqal 1964a: 158; 1964b, 1: 168; al-Muqaddasi 1906: 151; 1994: 138; al-Harawi 1953, 1: 24; 2: 62; 2004: 70. Miquel (1963: 147-48 n. 15) supplied other possible identifications of the Rock of Moses, without ruling out the identification as the Dome of the Rock. Indeed, al-Muqaddasi (1906: 169; 1994: 154) identified a separate "Dome of the Ascent" (sc. of Muhammad; qubbat al-mi'raj) near the Dome of the Rock, while Nasir-i Khusraw (1975: 43; 1986: 26) claimed that Muhammad ascended from al-Aqsa Mosque, which he distinguished from the Dome of the Rock.

(118.) Ibn Jubayr 1981: 224; 1952: 288; Yaqut al-Hamawi 1990, 4: 177.

(119.) He made a similar complaint regarding Crusader-ruled Jerusalem, which he visited before it was conquered by Salah al-Din (al-Harawi 1953, 1: 28, 33; 2: 68, 76; 2004: 74, 82).

(120.) Ibn Battuta 1964, 1: 43; 1958, 1: 103.

(121.) Al-Dimashqi 1866: 207; 1874: 282; Abu l-Fida' 1840, 1: 253; 2,2: 30; Ibn Battuta 1964, 1: 37; 1958, 1: 88. Sources disagree as to how far inland the new foundation was, between one mile (Abu l-Fida') and five miles (al-Dimashqi), which indicates that these reports are not all copied from the same source.

(122.) Abu l-Fida' 1840, 1: 247; 2,2: 26; Ibn Battuta 1964, 1: 36; 1958, 1: 85. Abu l-Fida' mentioned Beirut, but only to quote from older sources, so his text gives no indication of its current size.

(123.) Abu l-Fida' 1840, 1: 243; 2,2: 20, 22; Ibn Battuta 1964, I: 35; 1958, 1: 83.

(124.) Abu l-Fida' 1840, 1: 239, 249; 2,2: 17, 26; Ibn Battuta 1964, 1: 34; 1958, 1: 81.

(125.) Abu l-Fida' 1840, 1: 239, 249; 2,2: 16, 17, 27.

(126.) Yaqut al-Hamawi 1990, 3: 420; Abu l-Fida' 1840, 1: 247; 2,2: 25.

(127.) Al-Dimashqi 1866: 280; 1874: 408.

(128.) Al-Dimashqi 1866: 209; 1874: 285; Abu l-Fida' 1840, 1: 257; 2,2: 35; Ibn Battuta 1964, 1: 49; 1958, 1: 115.

(129.) Al-Dimashqi 1866: 207-8; 1874: 283; Ibn Battuta 1964, 1: 60-61; 1958, 1: 144. A major textual variant in the text of al-Dimashqi questions whether the church and monastery are in Antarsus or in Anafa, a coastal village. The translator followed a Paris manuscript that ascribed them to the coastal village of Anafa northeast of Jubayl but still placed another early monastery in Antarsus.

(130.) Al-Dimashqi 1866: 205; 1874: 279; Abu l-Fida' 1840, 1: 41; 2,1: 51. Certain manuscripts of al-Dimashqi omit the reference to Christian fishermen, only referring to the eels.

(131.) Ibn Jubayr 1981: 224; 1952: 288; Ibn Battuta 1964, 1: 63; 1958, 1: 148.

(132.) Ibn Battuta 1964, 2: 175; 1958, 4: 916.

(133.) Yaqut al-Hamawi 1990, 4; 334-35.

(134.) Ibn Jubayr 1952: 269; 1981: 209.

(135.) Abu l-Fida' 1840, 1: 229; 2,2: 6.

(136.) Abu l-Fida' 1840, 1: 261; 2,2: 38.

(137.) Cohen and Lewis 1978: 16; Sener and Dutoglu 2010: 15 n. 39; Bakhit 1982b: 49. Regrettably, in most other cases the editors of the defter of Aleppo give "total figures" arrived at by simply adding together households, bachelors, religious figures, and other tax-exempt individuals. These totals could still give proportions of a population belonging to different religions, if the proportion of households to bachelors was constant, but the variability is larger and less controlled.

(138.) Cohen and Lewis 1978: 94, 111, 128, 149,161. Samaritans made up about 2.5% each of Gaza and Nabulus in registers of 932/1526 and ca. 945/1538-9, respectively.

(139.) The defter of 932/1525-6 lists 199 Jewish households out of a total of 934 households, while Benjamin of Tudela (1907: 22, JD) mentioned 200 Jewish men; cf. Cohen and Lewis 1978: 94.

(140.) Bakhit 1982b: 49; Benjamin of Tudela 1907: 30, 32,; Sener and Dutoglu 2010: 15 n. 39.

(141.) Cohen and Lewis 1978: 161; Bakhit 1982b: 49.

(142.) Sener and Dutoglu 2010: 17. The Jewish population of the city of Aleppo seems to have been deliberately omitted from this defter, since it appears with somewhat fewer than 300 households in registers from before 932/1526 and after 978/1570, but even including them would give a Jewish population of 0.5% of the province as a whole (ibid.: 15, 17).

(143.) Sener and Dutoglu 2010: 16-17. The town of al-Qusayr near Antioch should not be confused with the more famous town southwest of Hims, which was in the province of Tripoli. I have not been able to consult Ottoman tax registers from Tripoli.

(144.) Abu l-Fida' 1840, 1: 261; 2,2: 38; Sener and Dutoglu 2010: 240, 266.

(145.) Cohen and Lewis 1978: 93-94, 128; Bakhit 1982b: 49, 55, 80. Approximately as many Christian households were located in villages around Gaza and Jerusalem as in the cities themselves, as recorded by Bakhit (1982a: 52-56), which must qualify the assertion of Cohen and Lewis (1978: 16) that non-Muslims were primarily town-dwellers.

(146.) Bakhit 1982a: 22; 1982b: 37.

(147.) For example, the large town of 'Aqura outside Ba'labakk was entirely Christian with 119 households and 11 bachelors in ca. 930/1523. Al-Shawbak, which was identified as entirely Christian in the Mamluk period, had very few Christians in the early Ottoman period; evidently many of them had relocated to Gaza (Bakhit 1982a: 32, 44, 45, 55).

(148.) The only exceptions in this triangle seem to be the nahiyas of al-Zabadanx and Shuf al-Bayad, which had about 20% and 10% Christian populations respectively in ca. 950/1543 (Bakhit 1982a: 41, 74).

(149.) Al-Baladhuri 1957: 195; 1916: 220.

(150.) Al-Muqaddasi 1906: 165; 1994: 151.

(151.) Yaqut al-Hamawi 1990, 3: 79. It is not clear in this account whether the scribe should still be thought of as a Christian or not.

(152.) Abu 1-Fida' 1840, 1: 241; 2,2: 18.

(153.) Bakhit 1982a: 56.

(154.) The elite bias of biographical dictionaries has been pointed out by critics of Bulliet, and their complaints discussed by Morony 1990: 138; Harrison 2012: 38-39. These commentators omitted the specifically Muslim dimension of ulema elitism, since first-generation converts to Islam were rarely included among the ulema after the seventh century. While Harrison is correct to point out that basically all premodem literary sources are elitist, the geographical works used in this article may be less sharply elitist than biographical dictionaries.

(155.) Morony (1990: 138) suggested refining Bulliet's conclusions for Iran by considering intra-regional differences.

(156.) Tamer el-Leithy (2005: 27) pointed out both medieval Muslims' general lack of interest in relative religious demography and the important but often overlooked lack of political significance associated with demographic strength. El-Leithy's remarks are a very necessary corrective, even if relative demographic strength might have had other social and cultural impacts despite medieval historians' neglect, such as which holidays drew large crowds or whether the presence of non-Muslims was considered common or bizarre.
COPYRIGHT 2015 American Oriental Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Carlson, Thomas A.
Publication:The Journal of the American Oriental Society
Article Type:Critical essay
Geographic Code:7SYRI
Date:Oct 1, 2015
Previous Article:Settlement, economy, and demography under Assyrian rule in the West: the territories of the former Kingdom of Israel as a test case.
Next Article:Some parascriptural dimensions of the "Tale of Harut wa-Marut".

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |