On Islamic Historiography
By Islamic Historiography I mean written material concerning the events of the early period of Islam written by Muslim historians. This material is essential for any major research on Islam but has been continuously discredited by predominantly Western scholars. Therefore, before the study of these texts, an outline of their characteristics and a short discussion about the criticisms of these texts and their authors is indispensable.
Among the problems proclaimed in the criticisms are: the gap between the historical events and their recording, the fact that early historical compilations have not survived and have been paraphrased or summarized in later digests, the problem of the oral origin of many reports, the task of the historian, the incompatibility of non-Islamic sources, forged reports, political influences on historiography, the purpose of historiography and the originality of the historian.
In this paper the criticisms concerning the Islamic historiography and the answers of the some historians to these criticisms will be surveyed.
The origin, the terminology and the form of the early Islamic historiography
According to Robinson, Arabs produced very little written material before Islam and relied instead on orality.
It seems logical to conclude that the enormous volume of written work which was produced after Islam must be ascribed primarily to the emphasis in various Qur´ānic verses on writing and the stories in this book about the previous peoples and prophets, which encouraged the Muslims to narrate, and reflect and investigate about the origins of those narrations, examples are, the next two verses:
…By the pen and what they write with it…. (Qur’ān 68:1)
Relate these allegorical stories (to the people) perhaps they might think. (Qur’ān 7:176).3
The second important impetus seems to have been the traditions of the Prophet of Islam which were to be preserved for the future generations. Islamic Tradition informs us that the Prophet of Islam discouraged his followers, in the initial stages of his mission, to write about him in order to prevent any confusion between his sayings and the Qur´ān.
However the reports about the alteration of this attitude in a later stadium encouraged the biographers to write Sīra or biographical collections at the end of the first and beginning of the second Islamic century. The campaigns of the Prophet (Maġāzī) and the conquests (Futūḫ)  were the other historical works, produced in the period between the first works and the later great compilations.
The collections with the modern name for history, Ta’riḥ, appeared in the 2nd/9th century.
Their source material consisted of Aḥbār which according to Rosenthal means both information and the events and corresponds to history in the sense of story, anecdote (ḫekāyat). Later, when the term was used together with āthār, it became synonymous to hadīth.
The other sources were the above mentioned Sīra, Maġāzī and Futūḫ works, the books of aḥbāriyyūn and genealogical works and oral accounts.
Thus, the first historical works, as the ordered record of the events of the past, began as a mixture of the above mentioned genres. This is the same multi-faceted character that Robinson says history used to have:
“…coming via Latin from the Greek historia, generally meant ´inquiry´; it earlier described a variety of genres, including geography, folklore and ethnography, in addition to what we would commonly understand to be history.”
And the way Rosenthal defines history:
History in the narrow sense.., should be defined as the literary description of any sustained human activity either of groups or individuals which is reflected in, or has influence upon the development of a given group or individual….for the modern mind, the general concept of history may, in theory, be extended to include all animate or inanimate matters. 
While he also mentions that:
Muslim historiography includes those works which Muslims, at a given moment of their literary history, considered historical works and which, at the same time, contain a reasonable amount of material which can be classified ashistorical according to our definition of history, as given above. 
Thus, history is made up of many elements which together have certain meaning for certain people. This is by no means the denial of general definitions of or theories about history, rather, the emphasis is on the meaning of a certain concept, object or idea in a specific context.
Not only the combination of aḥbār and āthār became synonymous to ḫadīth, but also the form of historical narratives took the form of ḫadīth. According to Dūrī two perspectives existed among the early compilers: the ḫadīth perspective and the tribal perspective. Very soon, the first perspective prevailed which explains why the Islamic historiography has maintained the form of ḫadīth, thus, beginning with an isnād or chain of transmitters, continued by the report (ḥabar).
The problems concerning the Islamic historiography
Islamic history books and Muslim historians have been the subject of both praise and critique. There are problems concerning the historical texts and those concerning the narrators both historians and their transmitters.
One problem ascribed to Islamic historiography is the fact that there is a gap between the time of the events of the early period of Islam and their historiography. Is this gap so long that it can in fact disqualify the whole historiography? It seems that this gap was not considered to be very important when the Western scholars first came into contact with the Islamic sources of the second and third century of Islamic era. Perhaps this was caused by their earlier experiences with other historiographies. The later recording of the events in Islam had its precedents in other historiographies. For example, according to Robinson: The gap between event and record in early Islam is relatively narrow compared with our source material for the ancient Israelites, which usually dates from several centuries after the facts they purport to relate.
Thus the problem of late compilation does not seem to be restricted to Islamic historiography.
Besides, some gaps might never be filled. The possibility that the gap is filled partly if a certain manuscript is found always exists, but I presume that the gaps in historical narrations might never be completely filled and even if they did, they would not answer all the inquiries of the modern researchers. There are then two options left: either to abandon using the available sources or to carry out research with the material which is in our possession. I reckon no researcher in the history of Islam, even those who consider all the material in the Islamic historiography as corrupt, who have chosen the first option.
The other problem concerning the Islamic historical texts is that the above mentioned early works of maġāzī and futūḫ, have either not survived or survived in the form of citations or paraphrases in later digests and compilations.
But Humphreys states:
The classical compilations of the 3rd/9th and 4th/10th century are not our only resources for early Islamic history, to be sure. They are supplemented in important ways by certain universal chronicles, biographical dictionaries, and encyclopaedias written in later centuries, for these often preserve otherwise unknown citations from early writings.
Thus, fortunately, many early historical collections can be recovered from the later compilations and there have even been attempts, mainly from the Western scholars to recover the original texts from the available sources through source criticism. In source criticism divergent accounts which have appeared in different times or those which have different chain of transmitters are compared. The problem which arises is that they lead to different results.
But there have never been attempts, even with the modern facilities, to compare ´all´ the available material on a specific subject. The most historical comparisons, as Leites informs us, are restricted to two or a few Sunnī works. Even an extent project might not eradicate all the differences and contradictions which are natural when different sources are involved.
Furthermore, the transmission might have a twofold of problems: the problem of prejudice and the problem of authenticity. Concerning the first part, prejudices can never be ruled out, but they are not always difficult to detect, like the example of Šī‘ite inclinations of Naṣr ibn Muzāḫim. Concerning the authenticity, unless harder evidence rejects the report, the narration must be accepted as authentic and it can be used for scientific historical research. But also the harder evidence is subject to scrutiny: What is hard evidence? Or harder evidence? Who is going to determine this?
As to the problem of orality in Arabic culture which seems to have continued after Islam and even when the historians were gathering their material, it is claimed that the oral origin of narration discredits historiography. For example Humphreys thinks that these sources can only be used when we learn more about the oral tradition, the circumstances in which it was committed to writing and the degree of its alteration before they achieved their definitive formin the 3rd/9th century.
Such expectations from oral transmissions are difficult even impossible to satisfy. Does this mean that the sources can not be used then? Not as far as orality is concerned according to Robinson who argues that history need not be based exclusively or even mostly on written material: No less an authority than Herodotus…managed to produce a very respectable history mainly on the basis of sightseeing and oral history.
With other words only reliable material which contradicts the historiography of Herodotus can discredit his historiography. It is interesting to mention that most historians who discredit Islamic historiography, base their research on Greek and Syriac sources which according to Humphreys are composed a century …(later).. and their sources of information are obscure at best.
Robinson’s idea complements this part:
If written history shares features with oral tradition, it is still much less pliable than oral tradition, which generally drops out of circulation as soon as it loses its relevance. Revision to history can certainly be made, and new versions can eclipse old ones, but history is only rarely obliterated. In part this is because stone, clay tablets, papyrus, animal skins, paper and other writing materials outlive human memories, but in equal part because it has its guardians- historians- in whose hands material is constantly recycled for later generations.
Thus, the oral nature of transmission does not automatically discredit a historiography unless the core of the report is lost or the historian uses unreliable or an extremely narrow range of transmitters. It seems that this problem was felt early by Muslim historians although perhaps not soon enough. Two reasons, fabricated traditions and ideological disputes, instigated the development of methods to scrutinize the transmitter. In the science of (Jarḥ wa al- ta‘dīl) or ḥadīttransmitter criticism, transmitters were divided into categories and transmissions from the reliable sources gained predominance. Since this process came about only later the major problem concerning the transmitters was the authenticity of the transmission, thus, whether a certain companion had really heard or said a tradition). It must be emphasized here that the early collections of (predominantly Sunnī) traditions were collected after they underwent two processes: first, the separation of unique, obscure or impossible details from the core text and second, the precise investigation of every name in the isnāds. This separation without doubt removed many details which were not in accordance with the ideological trends of the time, but it is not possible to decide which of the different accounts of the famous reports, like that of Ġadīr ḥum, for example, is authentic. Some of the later historians, like Ṭabarī, were ḫadīth specialists themselves and it is evident from their works, that they have been selective in their source material. But this has not always been the case and historians also used material which was not used by ḫadīth specialists, for example the narrations of the storytellers or quṣṣaṣ. Hoyland explains
… while they (historians and ḫadīth specialists) might both use the same source material the historian would be more willing to manipulate the text of reports. He would combine, harmonize, expand, abridge, paraphrase or interpret them,…with a view to the particular position that he sought to advocate.
Using the works of others could lead to another problem according to some critics: the lack of historical insight and originality. Thus, Noth observes the contradictions in one historian´s presentation of events in other compilations, in precisely the same respects. But Dūrī and Humphreys, who both have an eye on the specific characteristics of Islamic historiography, believe that in the early period of Islamic historiography, it was not the task of the historian to interpret his material. Dūrī writes that since history was regarded as a form of knowledge (‘ilm) and not individual judgement (ra’y), the reputation of the historian depended on presenting varying accounts or well-known impartial accounts, while dates and isnāds were fundamental element for assuring accuracy.  This matter is confirmed by Humphreys who states that the historians´ task was decisively not to interpret or evaluate the past: If a would-be historian spoke about these matters in his own words, he would inevitably be regarded as no serious scholar but as a mere propagandist for one or another faction.
Concerning originality, Hoyland´s idea is that the compiler would bring variations in the borrowed material and shape them as he wished and thus, many derivative compilations are in fact more original than they might appear at first sight. So the historians could be considered neither outright plagiarists nor totally original. He thinks that even modern history cannot easily be distinguished from fiction:
It seems to me preferable to regard history and fiction as lying on the same continuum…with scholarly compilations situated at one end and legendary sagas at the other, all to some degree possessing factual reference and semblances of verification, but all to some degree possessing animated by the imagination. This is not to say that Muslim historians knowingly fabricated material, or used fabricated materials- on the contrary, most of the writers … would have been convinced of the truth of what they wrote- but rather that history requires the mediation of fiction in its treatments of the past.
The theological nature of Islamic historiography is another point of criticism. Not only most reports about the Prophet or the companions and events were considered to have been forged later but scholars like Wansborough believed that Muslims had tried to create a specific theology of history, or tended to put a halo around the founder of their religion. It has now become evident that the theory of later invention is untenable since new material has shown the early compilation of many works. But it is true that as Robinson mentions Muslim and Christian historians were more concerned about the history of ideas than about ´what happened´. Without denying that some of the traditions concerning the Prophet of Islam might be biased, Harald Motski writes that this does not make the sources on the whole useless for historical use. Robinson even thinks that these biases in the history can have social and political functions and criticize a social order, a ruler or a state.
Despite the above mentioned problems which are attributed to Islamic historiography, my strongest motive in using Islamic historiography is the conviction that no other source may or can replace a people’s own historiography. It is always possible to take adventures like: history of Christianity from Buddhist point of view of from Jewish point of view or to wait for more archeological material to be discovered. But when it comes to reliable accounts on history I share this opinion with scholars like Humphreys who writes:
The true contemporary sources (papyri, archaeology, and Christian writings) are tantalizing indeed, but are either fragmentary or represent very specific or even eccentric perspective. An adequate and convincing reconstruction of Islam´s first century from these materials alone is simply not possible. That leaves us with the Muslim Arabic literary sources.
It is possible to gather external evidence and compare their information with internal historiography of any people, like the work of Patricia Crone and Michael Cook in Hagarism who tried to reconstruct history exclusively on the basis of non-Muslim sources. Such an attempt shows nothing but how a people was perceived by others, often from remote distances, locally and culturally. Valuable as these sources can be Hoyland does not cease to mention that:
When Blaise Pascal (1623-62) wrote that since Muḥammad worked no miracles and was not foretold he could not be a true prophet, he was simply echoing the judgement of John of Damascus (wr.ca. 730) passed more than 900 years earlier. Similarly, the explanation of Muḥammad´s revelation as the result of epileptic fits, found in numerous thirteenth-century and later texts, was already given by the Byzantine monk and chronicler Theophanes the Confessor (d.818).
There is no doubt, however, that any historiography must be studied cautiously and many elements have to be taken into consideration. The influence of the politics on historians´ works is one of these elements. There are reports that for example, the Umayyads kept many works, among which of al-Zuhrī (d.124/741), in their library and that he was coerced into writing down his works. From the other side, being pro-‘Alid would influence the reputation of the transmitter and consequently the historian. Zamān reports that Šī‘ī ḥadīt transmissions were only accepted if they were transmitted through a proto-Sunnī scholar. He also mentions that ‘Abbāsid caliphs even forged ‘Abbasid isnāds to give their rule the necessary legitimacy, although they were not acknowledged as reliable transmitters and caution is necessary when one encounters a transmission attributed to a ruler.
The important argument brought up by Fred M. Donner is that lack of orthodoxy in the Islam from the very beginning and the extant Islamic territory made a unique reading of the ´tradition´ and ´history´ impossible. With other words political influences could not shape a certain position or point of view from India to Spain.
This survey was meant to investigate whether Islamic historiography could form the basis of an historical research. Not only its characteristics but also some problems attributed to it were discussed and my conclusions are summarized as follows:
While in pre-Islamic `Arabia, oral transmission of historical events was prevalent, ordered writing of the Qur´ān and the biography of the Prophet opened the way for the production of more than 590 historical books in one millennium. The preliminary productions dating from the first Islamic century were heavily influenced by the ḫadīth specialists and acquired ḫadīth form, beginning with a chain of narrators followed by the report. In the period between the events and the compilation of the first history books oral transmission was predominant but some of the books of this period can be recovered from later compilations and digests. In later compilations, transmitters´criticism and ḫadīth criticism filtered many reports. The oral nature of the transmission does not make a historiography useless and the theological nature of the historiography is not unexpected from a religious historiography.
No other source material can or may replace a people’s own account of the historical events although outsiders´ account are useful to conceive how a people was perceived by another.
The repetition of historians of contradictory reports of others has led to the conclusion by some scholars that there no originality in Islamic historiography. It has to be borne in mind that each historiography has its own characteristics. An early Muslim historian had to separate his own idea from his transmission to avoid being regarded as a propagandist.
The influence of politics and ideological inclinations and the possibility of forgery necessitate prudential treatment of the historical material, although they do not discredit whole works.
Hoyland’s idea about extremist statements about Islamic historiography, of Gustav Von Grunebaum in 1946 that the overall objectivity of Arabic historiography is remarkable, and of Patricia Crone The works of first compilers…are…mere piles of disparate traditions reflecting no one personality, school, time or place, is that they are out of date. He believes that instead of the obsession with the question of authenticity scholars must pay attention to the manner of transmission of an account as well as to its facticity.
Muslim historians considered their work as a serious science and although they used much more material than the ḫadīth transmitters they were careful to use reliable material. This is not to say that fiction or storytellers played no role in their works, rather, that the spirit of the ḫadīth science had made them cautious about their sources and handling of material.
So generally speaking I have to agree with Zamān’s idea that the problems with Islamic historiography are problems with almost all ancient historiographies but they do not inhibit historical understanding.
 Robinson, 2003,8. Exceptions to this are according to Dūrī the inscriptions and carvings in southern Arabia datable from the eighth century B.C. to seventh century A.D.,Dūrī,1983, 14-15.
 About 590 historical works in the first millennium, according to Faruqi, 1979, 1.
 The Qur´ān translations are from Ṭ. Ṣaffārzadeh, 2005.
 Kamālī, 2003, 77, who refers to M.Šaltūt where he quotes a ḥadīt from the Prophet from Ṣaḥīḥ Muslim´s ḥadīt collection narrated from Abī Sa‘īd al-Hudrī: lā taktubū ‘annī….
 Van Dalen and others, TA’RĪKH, Encyclopaedia of Islam, 2nd ed., Brill Online, accessed 28/10/2009, 22-23.
 According to Rosenthal Ta’rīh, ..is a noun which exists in Arabic and South Arabian..meaning ‘date’ and ‘era’. It is mentioned first in Arabic literature together with the stories of the introduction of the Muslim era and is used in a papyrus dated in the year 22 of hijra. ..It then acquired the meaning of ‘historical work’, and afterwards that of ‘ history’, exactly as history or Geschichte may mean…In the meaning of ‘ historical works’, it is used from the second century of hijra onward, in Rosenthal, 1968, 11-14
 Rosenthal, 1968, 11.
 Dūrī, 1983, 61.
 Robinson, 2003, 6.
 Rosenthal, 1968, 10-11
 Ibid, 17
 Dūrī, 1983, 22-23
 As mentioned by H. Motzki, 2000,.XI.
 Robinson, 2003, 50.
 Sīra works are the exceptions. According to Dūrī with the works of ‘Urwa ibn al-Zubayr (d.94/712) and Muslim ibn Šihāb al.Zuhrī (d.124/741), in A.A. Dūrī, 1983, p.30. According to Jafri, with the works of Abān (d. ca. 100/718-19), the son of the Caliph ‘Utmān; ‘Urwa b. Az-Zubayr b. Al-‘Awām (d. 94/712-13); Wahb b. Munabbih (d. 94/728-9) which have been studied by Nabia Abbot and others in S.H.M. Jafri, 1989, p.30, Jafrī gives this reference for Abbot´s studies: STUDIES IN ARABIC LITERARY PAPYRI (Chicago, 1957-72), I, 5-31;II, 5-64
 Humphreys, 1991, 69-71. Faruqi mentions one example, the pact between ‘Amr ibn al-‘Āṣ and Mu‘āwiya to mobilize public opinion to demand the vengeance for the murder of ‘Utmān from ‘Alī and its text is quoted by Ibn Sa`d (d.230/844-5) showing the name of the scribe as Wardān and the year 38 A.H, in Faruqi, 1979, 188, his reference being Ibn Sa´d, ṬABAQĀT, vol.IV, 254.
 Leites, in Motzki, 2000, 51.
 Humphreys, 1991, 69-70.
 Robinson, 2003, 13.
 Like Sancisi-Weerdenburh and Kuhrt who have their serious objections against the image of Xerxes, the king of Persia in the historiography of Herodotus, reference in the bibliography.
 Humphreys, 2006, 2.
 Robinson, 2003, 13.
 The first work being KITĀB AL-ṬABAQĀT AL-KABĪR of Muḥammad ibn Sa‘d (d.230/845) according to Lucas, 2004, 7.
 Lucas, 2004, 73
 Hoyland, 2006, 21-22.
 Noth, 1994,7.
 Dūrī, 1983, 73-75.
 Humphreys, 1991, p.74.
 Hoyland, in J. Bray, 2006, 32.
 Ibid. 40.
 Ibid., pp.17-18
 Robinon, 2003, 12.
 Motzki, 2000, 170-171
 Robinson, 2003, 13.
 Hamphreys, 1991, 69.
 Motzki, 2000, XIII, he refers to P.Crone and M.Cook, 1977.
 Hoyland, in H.Motzki, 2000, 276.
 Although Durī thinks that these might be echoes of disputations which arose after al-Zuhrī´s time, Dūrī,1983.
 Zaman, 1997,77 and note 30.
 Ibid. pp., 126-127.
 Donner, 1998, p.27.
 Hoyland, in Bray, 2006, 19.
 Zaman, 1997, 25
About the author:
Simin Rafati studied Islamic Theology and Comparative Religion in Leiden University. Her thesis “Al-Qaida and Hizbullah: their compliance with Sunni and Shi’ ite jurisprudence of Jihad”, won the prize for best thesis of Islamic Theology in 2009. She is an independent scholar and translator and translated “Jip and Janneke” children’s stories and “Checkpoint Jerusalem”.