Accessing the Pre-Islamic Past
Since Arabic-Islamic historiography only came into being in the seventh century, access to the pre-Islamic past required having recourse to the knowledge stored by groups whose collective memory extended further back.173
Most Arabic-Islamic scholars had no ideological qualms to make use of non-Muslim source material. Non-Muslim sources were not only cited in religious polemics, e.g. the refutations of Christianity by Ibn H azm (d. 456/1064) or al-Shahrastani (d. 548/1153), who drew on Christian material of antique and late antique provenance such as the gospels, the Nicene Creed, etcTh4 They also provided Arabic-Islamic historiographers with large quantities of material. In his chapter on the Romans, al-Ya'qubi (d. after 292/905) paraphrases the Nicene Creedi75 and summarizes the essential doctrines on the divine formulated by ancient Greek philosophers, concluding that there had been some clear-sighted and sensible thinkers (ahl al-naz,ar) among themTh6 To write about the Roman presence in the Near East, al-Tabari (d. 310/923) drew on ‘scholars among the people of the book from Palestine’177 and ‘the testimony of Christians’. 17® The Andalusian historiographer Ahmad b. Muhammad al-Razi (d. 344/955) seems to have employed local Christian source material to reconstruct the ancient history of the Iberian Peninsula. 179 Doing research on the Romans’ origins, al-Mas'udi (d. 345/956) consulted the Torah and other books of the Hebrews.180 His passages on Roman history are based on Christian sacred texts           and Christian historiography, including a Melkite chronicle he had seen in Antioch,182 the ‘histories of Jewish and Christian legal authorities’,^3 as well as several books written by Christian authors of Melkite, Maronite, Jacobite, and Nestorian confession in Arabic, some of whom he knew personally.^4 The Andalusian scholar Ibn Juljul (d. after 284/994) regarded the kitab Hurushiyush as an ‘outstanding historical account’ (tarikh 'ajib) of ‘great merits’ (fawaid Azima).16*’ Al-Blrnnl (d. c.442/1050) states that he received a list of Roman emperors thanks to H amza al-Isfahani on the authority of al-Wakf al-qadi who ‘took it from a book that belonged to the ruler of the Byzantines’.m
Later works of universal history employed non-Muslim material less frequently. Ibn al-Athir (d. 630/1233), for example, drew on al-Tabari as his main source for Roman history in a preliminary exposition of the latter, then ventured to write a second, fuller version that seems to be based largely on al-Mas‘udl.187 Abu l-Fida (d. 732/1331), in turn, followed the model furnished by Ibn al-Athir.188 The list of Frankish kings that became available to al-Mas'udl in 336/947, was reproduced by al-Bakri (d. 487/1094) and al-Nuwayri (d. 733/1333).189 Able to draw on earlier works of Arabic-Islamic scholarship, later historiographers could thus afford to ignore material provided directly by non-Muslims. This should not imply that they never made use of non-Muslim sources. Ibn Khaldun (d. 808/1406), for example, drew on the Arabic version of Orosius’ Historiae adversuspaganos, the tenth-century Jewish chronicle Josippon, as well as the Coptic historiographer Ibn al-'Amld (d. 672/1273) to reconstruct Roman history. 190
With one exception, the historiographers dealt with in this study either made use of non-Muslim sources without further comment or occasionally even praised their merits. Ibn Khaldun, however, seems to have felt obliged to discuss the necessity of using such material. In a passage on the Hasmoneans and the kingdom of Judah, Ibn Khaldun defends his use of the Jewish-Roman historian Flavius Josephus, falsely believing that the latter had authored the actual source at his disposal, i.e. the tenth-century Jewish chronicle Josippon.191
I have summarized it [the historical narrative] here as I have found it there [in the chronicle], since I have not encountered anything comparable to it. For the people (al-qawm) is better informed about its own history, provided that it does not contradict what is preferred to it [i.e. divine revelation]. As he [the prophet Muhammad], peace be upon him, said: “Do not believe the people of the book.“ However, he also said: “Do not accuse them of lying”. But in spite of this, he himself had recourse to the history of the Jews and the stories of the prophets which contain elements of divine revelation from God. For the latter said afterwards: “Say: We believe in what has been revealed to us and to you.” [Sura 29:46] As concerns data on events which have been recorded by an eyewitness (mustanida ila l-hiss), one document is sufficient, if its authenticity imposes itself on the intellect. However, it is necessary to link this data with their earlier historical records, to complete our understanding of their affairs from the beginning to the end.   
Thus, Ibn Khaldun justified his use of pre-Islamic and non-Muslim sources by pointing to the prophet’s contradictory handling of pre-Islamic material of Jewish and Christian origin and by playing the Qur’an off against Muhammad. Ibn Khaldun again addresses the reliability of non-Muslim sources, when he claims that the kitab Hurushiyush was more trustworthy than the Josippon because two Muslims had translated it. 193 Considering that preceding historiographers had no qualms to use non-Muslim sources, it is curious that Ibn Khaldun saw the necessity of defending the use of such sources and of demonstrating the reliability of a translation by pointing to the translator’s religious affiliation. Whether this necessity was imagined or real, if it arose from his biography or a general rigidification of Muslim scholarship in the fourteenth century, remains to be clarified.^4 Given that later scholars such as al-Qalqashand! (d. 821/1418) and al-MaqrlzI (d. 845/1442) continued to cite non-Muslim sources without any apparent qualms, one should not generalize too quickly.195
Accessing the pre-Islamic past did not only constitute a challenge because it i nvolved using sources written in other languages than Arabic by non-Muslims. Since the first year of the Islamic calendar is equivalent to the year 622 of the Christian era, Arabic-Islamic historiographers also faced the problem of dating events and periods that did not form part of the Muslim calendar. The resulting difficulties can be elucidated by showing how they dealt with the chronology of Roman history.
Al-Ya'qubi (d. after 292/905) failed to define when Roman imperial rule began and simply divided his list of Roman rulers into two chapters that deal with pagan and Christianized rulers respectively. Each ruler is credited with a certain number of years of rule. The list’s chronological position within world history can be understood with reference to the preceding chapter on Ptolemaic rulers as well as to the ensuing chapter on the Persians that occasionally correlates the rule of Persian and Roman rulers. Retrospective calculation from the only chronological point of reference provides the possibility of establishing a very imprecise form of absolute chronology. Dealing with the twenty-nine-year rule of Justin II (Yustus), al-Ya'qubi states that Muhammad was born in this period.^
The issue of chronology was successfully tackled by al-Mas'udi (d. 345/956) who had already complained about the many contradictory specifications of ruling years attributed to Roman emperors in his earlier muruj al-dhahab.™ At the beginning of his list of Roman rulers included in his later kitab al-tanbih, al-Mas'udi states that he wrote the lines in question in 345/956 during the rule of the Byzantine emperor Constantine VII Porphyrogennetos (Qustantin b. Lawun b. Basil). He then specifies the number of years that have passed since the first Roman ruler, i.e. Caius Iulius Caesar (Ghaiyus Qaysar), assumed power. According to al-Mas'udl, this added up to a number of 966 years and one month. During this period, pagan rulers held power for 374 years and three months while Christianized rulers governed for 591 years and ten months.199 Whereas the mathematics are correct, the calculation contains historical errors. If al-Mas'udl calculated in solar years of approximately 365 days, Caesar would have taken power in the year 10 BCE, i.e. thirty-four years after his death in 44 BCE. Constantine would have accepted Christianity in 365 CE, i.e. twenty-eight years after his death in 337 CE. If he calculated in lunar years of approximately 354 days, Caesar would have taken power in 20 CE, and Constantine would have converted around 392 CE. Although al-Mas'udl did not err very much with regard to the chronological framework of Roman history accepted by scholars today, the example serves to elucidate the dimension of the problem faced by Arabic-Islamic historiographers. Al-Mas'udl also drew on alternative systems of dating. He states that Jesus was born 5,506 years after Adam and thirty years after the death of Cleopatra,2°° and correlates the reigns of Justinian I (Yustinus) and the Persian ruler Chosroes I (Anushirwan).°i He reaches safe chronological ground in the chapter dedicated to ‘rulers of al-Rum from the hijra to the year 325’7°2
Al-Birnni (d. c.442/1050) sought a systematic solution to the chronological problem. Written around 390/1000, his computational study entitled Vestiges of the Past (al-athar al-baqiya) puts the chronological question at the centre of enquiries^ Al-Blrum divides world history into several eras. Roman history is given space in the era of Augustus, the era of Antoninus, and the era of Diocletian.204 In addition, he supplies three lists of Roman rulers that deal with pagan rulers, Christian rulers, and rulers of Constantinople respectively. These lists allow calculating the number of years that have passed since the first ruler in the list.205 Additionally, al-Blrunl explains how to calculate dates in different computational systems^6
For other scholars the computional problem was far from solved. Al-Bakri (d. 487/1094), for example, based his chronology of Roman history upon references found in the kitab Hurushiyush as well as in the writings of Ibn Khurdadhbah, al-Tabari, and al-Mas'udl. This resulted in a random assortment of relative chronological specifications provided in different unrelated passages and in the following order. Carthage (Qartajina) was founded seventy-two years earlier than Rome in the era of the Israelite king David. The Messiah was born in the forty-second year of Augustus’ reign and 369 years after Alexander the Great. Titus destroyed Jerusalem 1,060 years after its construction. Rome celebrated its thousandth anniversary when Philippus Arabs ruled in his second year. It was founded 400 years before the emperors held power. It was the residence of forty-nine or twenty-nine rulers for 437 years up to the reign of Constantine I, etc.207 One can hardly speak of a systematic chronology.
Abu l-Fida (d. 732/1331) finally offered a practical solution for those dealing with pre-Islamic chronology by establishing a calculation table made up of several chronological points of reference taken from the Judaeo-Christian and Roman traditions. Situated at the beginning of his universal history, the table allows calculating the time spans between Adam’s expulsion from Paradise, the Deluge, Abraham’s birth, Moses’ death, the beginning of Nebuchadnezzar’s reign, Darius’ defeat at the hands of Alexander, Cleopatra’s defeat at the hands of Augustus, Jesus’ birth, the reign of Diocletian, and, finally, the hijra.2(,s
-  al-Mas'udl, al-tanbih, ed. de Goeje, pp. 160—1.
-  al-Mas'udl, muruj, ed./trans. Pellat, § 719, p. 34 (AR), pp. 270—1 (FR): ‘wa-ra’aytu fi madinatAntakiyya fi ba d tawarikh al-Rum al-malakiyya . . . ’.
-  Ibid., § 726, p. 38 (AR), 272—3 (FR): ‘tawarikh ashab al-shara’f min ahl al-kutub’.
-  al-Mascudl, al-tanbih, ed. de Goeje, pp. 154—5, trans. Carra de Vaux, pp. 212—13.
-  185 As cited in Ibn Abi Usaybia, tabaqat al-atibba, ed. Rida, p. 494.
-  186 al-Biruni, al-athar, ed./trans. Sachau, p. 97 (AR): naqalaha min kitab li-malik al-Rum’, p. 106 (EN).
-  Ibn al-Athir, al-kamil, ed. Tornberg, vol. 1, pp. 230^3 (Leiden), pp. 323—39 (Beirut). SeeChapter 4.2.3.
-  Gibb, Abu l-Fida’’ (1960), p. 118.
-  189 al-Masudi, muruj, ed./trans. Pellat, § 912, p. 146 (AR), p. 344 (FR); al-Bakri, al-masalik, ed.van Leeuwen and Ferre, § 567, p. 340; al-Nuwayri, nihayat al-arab, ed. al-Tawil and Hashim, vol. 15, pp. 222—3.
-  Ibn Khaldun, tarikh, ed. Zakkar and Shahada, vol. 2, pp. 229, 283. Cf. Issawi, ‘Ibn Khaldun’(1998), pp. 51—77, esp. 62—6; Branco, Storie (2009), pp. 202—3.
-  Issawi, ‘Ibn Khaldun’ (1998), pp. 63^; cf. Josippon ed. Flusser.
-  Ibn Khaldun, tarikh, ed. Zakkar and Shahada, vol. 2, pp. 134—5: ‘fa-lakhkhastuha huna ka-mawajadtuha f!hi li-annan! lam aqif ala shay’ f!ha li-siwahu, wa-l-qawm a lam bi-akhbarihim idha lamyuaridha ma yuqaddam alayha. wa-ka-ma qala salla llahu alayhi wa-sallam: la tusaddiqu ahl al-kitab.fa-qad qala wa-la tukadhdhibuhum. ma a anna dhalika innama huwa raji ila akhbar al-Yahud wa-qisasal-anbiya’ allat! kana f!ha al-tanz!l min inda llah, li-qawlihi ba da dhalika: wa-qulu amanna bi-lladh!unzila ilayna wa-unzila ilaykum. wa-amma al-khabar an al-waqiat al-mustanida ila l-hiss fa-khabaral-wahid kafin f!hi idha ghalaba ala l-zann sihhatuhu, fa-yanbagh! an nulhiqa hadhihi al-akhbar bi-mataqaddama min akhbarihim li-takmul lana ahwaluhum min awwal amrihim ila akhirihi . . . ’.
-  Ibid., vol. 2, p. 234.
-  Lakhsassi, ‘Ibn Khaldun’ (2002), p. 360, for example, speaks of a ‘victorious ultra-orthodoxy’which ‘started to take the upper hand in the eighth/fourteenth century’.
-  Cf. their reception of the kitab Hurushiyush, in kitab Hurushiyush, ed. Penelas (introduccion), pp. 77, 79-81.
-  al-Ya'qubl, tarikh, ed. al-Muhanna, vol. 1, pp. 186—99.
-  Ibid., vol. 1, p. 198.
-  al-Mas'udl, muruj, ed./trans. Pellat, § 733, p. 40 (AR), p. 274 (FR).
-  al-Mas'udl, al-tanbih, ed. de Goeje, p. 123. 5 Ibid., p. 124. 201 Ibid., p. 153. 202 Ibid., p. 156.