SOURCE: "Source Criticism in the Uthmaniyya of Al-Jahiz," in The Muslim World, Vol. LXX, No. 2, April, 1980, pp. 134-41.

[In the essay below, Zahniser considers the methods that al-Jahiz used to evaluate the validity of his sources.]

In a previous article published in this journal on the Uthmaniyya of Abu Uthman Amr b. Bahr al-Jahiz (d. 869), the celebrated Basrian Mutazilite litterateur, reference was made to the fact that the author wrote a series of treatises for the caliph al-Mamun (reigned 813-833) on the imamate, and that his Uthmaniyya was probably one of them.¹ It presents the views of a sect called the Uthmaniyya, views al-Jahiz himself seems to have shared. In it he argues for the superior virtue of Abu Bakr and, thus, his exclusive right to the leadership of the Muslim community.

An attempt is made here to elucidate al-Jahiz' assessment of the reliability of information from traditional sources for purposes of establishing the relative merits of the individuals surrounding the Prophet Muhammad. The importance of such an assessment is indicated in the third paragraph of the treatise:

If we examine carefully their historical data (akhbar) and consider the number of their Traditions (ahadith) as well as the authorities who transmitted them, if we look at the quality of their chains of authority (asanicf), [we find] the tradition (khabar) supporting the priority of Abu Bakr's conversion more general in its application with a greater number of transmitters and a sounder chain of transmission, and, therefore, both more widely accepted and more obvious in meaning. In addition to this, there is the trustworthy poetry and the widely diffused (mustafid) historical material relating to events both during the life of the Messenger of God…and after his death. There is no difference in value as evidence between poetry and historical narrative (akhbar) as long as the circumstances of its origin and transmission preclude a conspiring together to fabricate it.²

Al-Jahiz maintains that "one who wants the tradition he cites to be useful and convincing as proof must go deeply into the disciplines of tradition (khabar) collection and criticism. If he submits himself to these disciplines and has a sound mind, he will save himself and others a lot of trouble."³ According to another passage: "When we claim that a Tradition (hadith) is weak and corrupt and you suspect our judgment and fear lest we be prejudiced or in error, consult the specialists in Tradition because they have the answers to that about which we contend and knowledge of that in this area which confuses us."⁴

Not all traditions are reliable, the existence of contra dictory traditions implying the existence of fraudulent traditions. Therefore, for any traditional material to be reliable for argument "the circumstances of its origin and transmission [must] preclude a conspiring together to fabricate it."⁵ Such traditions al-Jahiz terms "undeniable" (la yukadhdhab mithluhu),⁶ or "compelling" (khabar mudtarr ilayhi),⁷ identifying the criteria which such a tradition must satisfy: 1) the tradition must be widely attested so that the very number of its transmitters testifies against its having been fabricated; 2) those who transmit the tradition must do so for different reasons; and 3) those who transmit should transmit from authorities in their own tradition. That is, the earliest authorities as well as later transmitters must have held diverse points of view.⁸ For example, in the following passage al-Jahiz defends the authenticity of the epithet al-Siddiq (the Trustworthy) by showing that it is not related by any particular group in order to enhance their own position by exalting his.

But how can the name al-Siddiq be a forgery when most of those who used it [did not have some axe to grind. They] were not members of some sect enhancing their esteem by him, or devotees of some intellectual discipline trying to make known his excellence [in that discipline], nor were they people related to Abu Bakr seeking to establish his preeminence and, hence, their own.⁹

More important to al-Jahiz than the reliability of a par ticular transmitter is the probability of a given tradition's having been forged.

Al-Jahiz identifies three kinds of tradition: 1) mustafid ("widespread"), the term used for traditions which satisfy the criteria described immediately above;¹⁰shadhdh ("anomalous, irregular, exceptionelle"), or unattested, unique items of historical information which do not qualify for use as evidence; and 3) all other traditions. Traditions in the last group, while not shadhdh, are not beyond reproach as evidence in deciding matters of importance such as that under discussion in the Uthmaniyya. In fact, in discussing the Tradition called Hadith Ghadir Khumm, "O God, be an enemy to Ali's enemy and a friend to his friend," al-Jahiz argues that no traditional material may be admitted as proof over which there is disagreement as to content, transmission or interpretation.¹¹

Furthermore, traditions in which the Prophet indicates the superiority of a particular Companion are of weak authority even when they are judged to be sound by the above criteria because of the possibility that understood exceptions or special applications may pertain to them. For example, when the Prophet said, "There is no one in the world who speaks more honestly than Abu Dharr,"¹² he did not need to add "except me" since this exception was understood by all who heard him. The Shia assume that Ali's exclusion is understood and the Uthmaniyya assume that Abu Bakr is excepted.¹³ While the sayings of the Prophet Muhammad concerning the excellence of his Companions should not be overlooked, a final judgment on the superiority of any one of them over the others must not be made without examining the historical records of their great deeds.

Obviously al-Jahiz' tradition-criticism would eliminate large numbers of prophetic Tradition accepted by critics such as Muslim and al-Bukhari, who gave primary consideration to the reliability of the individual transmitters.

Moving from a discussion of historical Traditions in general and prophetic Traditions in particular to al-Jahiz' use of both in the development of his argument, we notice that he relies heavily upon material from historical sources.

Abd al-Salam Harun's index of the Traditions cited in the Uthmaniyya contains a total of ninety-six. Since a number of his entries consist of different wordings of another Tradition, it is possible to reduce the number to eighty, of which seven are Traditions to which the Rawafid¹⁴ are represented as appealing in support of their position in the dispute,¹⁵ thirteen are given to illustrate the principle of understood exceptions mentioned above, and fifty-nine are cited in support of the claims of the Uthmaniyya. Some of these fifty-nine al-Jahiz admits are substandard and not sufficiently reliable for use as proof. He refers to two as completely reliable, to another as mustafid, and to another as undeniable. That is, only four of his fifty-nine Traditions are declared by him to meet his criteria for reliable traditions. Forty five of the remaining fifty-five Traditions are ones to which Abu Uthman appeals as evidence without comment as to if, how, or why they are sufficient for proof; most of these are related in some way to specific incidents in the life of the Prophet or connected with his death and its aftermath.¹⁶

We set out to trace these eighty Traditions in the prin cipal collections of Islamic Traditions, consulting first the works indexed in the concordance compiled by Wensinck and others,¹⁷ but also examining other works, including four historical sources.¹⁸ These efforts revealed that the Traditions cited by al-Jahiz are not only inconsistent in wording within the Uthmaniyya itself, but differ significantly from the counterparts that we found in works devoted specifically and primarily to collecting Traditions. This made it difficult to be sure that all the corresponding Traditions had been located in the concordance.¹⁹ There is some evidence in other sources for sixty-seven of the eighty Traditions, twentysix of which appeared in only one source.

Of the fifty-nine traditions which Abu Uthman uses in making his case, only thirty appear in the exact words of their counterparts in one or more of the sources examined. Of these thirty, fourteen are found only in the historical sources. Nineteen of the thirty are among the Traditions connected with events in the Prophet's biography. This is congruent with al-Jahiz' claim that an appeal to the record of the great deeds of Abu Bakr and Ali is less subject to uncertainty as proof than an appeal to the words of the Prophet.²⁰

As a result of all this, one would expect al-Jahiz to rely upon historical sources in developing his argument. In fact he does precisely this, citing three of the most prominent figures in the field of sira and maghazi literature by name: Ibn Ishaq,²¹ al-Waqidi²² and al-Zuhri.²³ An examination of the material in the Uthmaniyya that relates to historical events during the life of the Prophet and surrounding his death reveals a striking similarity to the corresponding materials in the Sira of Ibn Ishaq²⁴ and al-Waqidi's Kitab al-Maghazi²⁵

The historical material is naturally divided into two periods: up to and including the Hijra and from the Hijra to the death of Muhammad. For the former period we have no treatment in the Maghazi; for the latter we have both the Maghazi and the Sira providing material for comparison. The last sickness of the Prophet, his death and its aftermath are covered in the Sirah and not in the Maghazi.

The material in the Uthmaniyya that falls within the periods covered by Ibn Ishaq alone includes two passages where the two sources exhibit striking word-for-word agreement.²⁶ In four other passages there is less correspondence between the two sources, but in these passages differences can be attributed to paraphrase, abbreviation or purposeful omission on the part of al-Jahiz.²⁷ The material for this period also includes a number of passages which are either without parallel in Ibn Ishaq or which are different enough to suggest dependence upon a separate historical tradition.²⁸

When we turn to incidents treated by Ibn Ishaq, al-Waqidi and al-Jahiz we find that several passages are omitted by Ibn Ishaq where the Uthmaniyya and the Maghazi agree virtually word-for-word.²⁹ For incidents in the life of Muhammad which are covered by both the Sira and the Maghazi, al-Jahiz' wording is closer to that of al-Waqidi than to that of Ibn Ishaq.³⁰ In two passages, both about the Battle of Badr, al-Jahiz follows Ibn Ishaq more closely.³¹ We found only one case where an entire incident was missing from the Sira and the Maghazi but included in the Uthmaniyya—the incident at Ghadir Khumm, a prime passage for the Shia to cite in support of their case.³²

Thus, the Uthmaniyya contains far more passages from the Meccan period for which no parallel exists in the Sira or the Maghazi. In addition, al-Jahiz tends to favor al-Waqidi's material for the period covered by the Maghazi. And since al-Waqidi's material is known to be distinctive,³³ we may be justified in suggesting that material in the early period not found in the Sira has high probability of having appeared in al-Waqidi's treatment.³⁴

At any rate, in terms of al-Jahiz' use of sources from history and Tradition to support his argument in the Uthmaniyya, it is clear that he prefers the historical to the words of the Prophet and that he draws on Ibn Ishaq and al-Waqidi for a large portion of this material.

A third source of information to which al-Jahiz and his opponents, the Rawafid, appeal in developing their arguments is the Quran. Both he and they are convinced that certain passages in the Quran refer to their respective champions, Abu Bakr and Ali. Since neither is mentioned by name in the Holy Book, references claimed for either must be understood by interpretation. In refuting the contention of the Rawafid that S. 5:55/60, "Your friend is only God and his Messenger, and the believers who perform the prayer and pay the alms while bowing down,"³⁵ refers to Ali, al-Jahiz indicates his criteria for determining that a particular verse refers to Ali or any other person.

There are two possibilities regarding this verse: either the plain outward meaning of the words agrees with the way [the Rawafid] want to interpret the verse; or, it was revealed in connection with some important and widely recognized incident involving Ali similar to that of Abu Bakr in the Cave. If we cannot establish either of these two possibilities as true, then the only other possibility to which we can turn to support their interpretation is that the Prophet said to the people, "This refers to Ali; so admit to him his rightful recognition and recognize his consequent virtue." But if this last alternative were in fact true, the interpreters would not have differed in their exposition of it, and Ibn Abbas would not have given the view he gave.³⁶

That is, if it is not plain from the text itself that Ali is referred to, as would be the case if his name were mentioned in it, if the verse is not connected with some specific incident which is known to have involved Ali, and if there is no Tradition from the Prophet which has wide recognition, satisfying al-Jahiz' criteria for reliability indicated in section I above, then it cannot be established that the verse refers to Ali. Since the first possibility is excluded for any verse of the Quran, and the third requires a consensus that does not exist, the second remains the only way of associating a particular verse with either Abu Bakr or Ali.

The two passages from the Quran which carry the greatest weight for al-Jahiz as proof of Abu Bakr's superiority are passages relating to specific events in the life of the Prophet. One is the story of Abu Bakr's accompanying Muhammad in the Cave at the start of the Hijra,³⁷ and the other is the incident of Abu Bakr's restoring Mistah to favor and support.³⁸ Of the first, al-Jahiz writes: "If Abu Bakr's only noteworthy action were that to which this verse refers, he would be above everyone else in position, virtue and close companionship to the Prophet."³⁹

Here again, when appealing to the Quranic evidence for the superior virtue of Abu Bakr, we find al-Jahiz arguing from the same premise from which he argued when dealing with evidence derived from the Traditions of the Prophet: proof should be limited to reasoning from indisputable evidence which, in the case under discussion, amounts to what is related about well-attested events during the period of the Prophet's life and the events connected with his death.

In fact, it might be argued that al-Jahiz wishes to apply the same criteria to tradition, including hadith, akhbar and tafsir, which were applied to the texts and recensions of the Quran. The Seven Readings (al-qiraat alsab) both possess the consensus of the Companions (ijma al-sahaba) and are mutawatir ("successive," widespread), a designation similar to al-Jahiz' mustafid). The Mushafs of Ibn Masud and Ubayy, for example, are considered shadhdh ("irregular") because they possess only the ijma and not the tawatur. It has been forbidden to use them since 935.⁴⁰ These criteria are similar to the cardinal points of al-Jahiz' tradition criticism: there must be both consensus among specialists and such widespread attestation that forgery is absolutely precluded. In fact, in another treatise, al-Jahiz laments the fact that the sayings, miracles, and other glorious deeds of the Prophet had not been collected and standardized as had the Quran.⁴¹

In summary, then, al-Jahiz' source criticism requires of transmitted data that it be widely recognized in diverse enough circles to preclude the possibility of fabrication. This historical method leads to the affirmation of the value of sira and maghazi sources for use as evidence in serious theological discussion. It calls into question the elaborate system constructed by the Muhaddithun for evaluating and verifying transmitted information, focusing on the two criteria of wide circulation among divergent groups and consensus among specialists rather than on considerations related to the quality of each link in the chain of transmitted data.

Notes

¹ A. H. Mathias Zahniser, "Source Criticism in the Uthmaniyya of al-Jahiz," M.W., LXIX (1979), 8-17. The series of books is mentioned in Amr b. Bahr al-Jahiz, Al-Bayan wa l-Tabyin, ed. Hasan al-Sandubi, 3 vols., Maktabat al-Jahiz, 2 (Cairo, 1926-27), III, 374.

² Amr b. Bahr al-Jahiz, Al-Uthmaniyya, ed. Abd al-Salam Muhammad Harun, Maktabat al-Jahiz, 3 (Cairo, 1955), p. 3.

³ Ibid., p. 135. Al-Jahiz uses khabar (and sometimes athar) to refer to transmitted historical data in prose form and hadith to refer to the words of the Prophet. Not all akhbar (athar) are ahadith, but all ahadith are akhbar. When we wish to make a distinction between these terms here, we shall use a capital "T" to indicate prophetic Tradition (hadith).

⁴Ibid., p. 151.

⁵Ibid, p. 3.

⁶Ibid., p. 244.

⁷Ibid., p. 271.

⁸Ibid, p. 116.

⁹Ibid., p. 124.

¹⁰Other approximate synonyms include shai ("diffused") and mujma alayhi ("accepted by consensus"). Ibid., pp. 115-16.

¹¹Ibid., p. 148.

¹²Ibid, p. 140.

¹³Ibid., pp. 137-42.

¹⁴ The Rawafid are the Shii opponents of the Uthmaniyya in this dispute. See W. Montgomery Watt, The Formative Period of Islamic Thought (Edinburgh: Edinburgh University Press, 1973), pp. 157-62, 167.

¹⁵ Appendix C of our dissertation consists of a list of all the Traditions cited in the Uthmaniyya and the sources in which parallels or counterparts were found. A. H. Mathias Zahniser, "The Uthmaniyah of al-Jahiz: An Analysis of Content, Method and Sources" (unpubl. Ph. D. dissertation, Johns Hopkins University, 1973), pp. 205-12.

¹⁶ Suffering in pre-Hijra Mecca (4), Isra and miraj (3), Muhammad orders Hassan b. Thabit to defame the Quraysh (1), the Hijra (2), building the first mosque in Medina (1), Badr (2), Ghadir Khumm (1), Uhud (4), Hudaybiyya (4), the Ditch (1), Conquest of Mecca (1), Hunayn (1), Tabuk (1), the pilgrimage of year nine (1), last illness and death of the Prophet (8), and the Ridda (1).

¹⁷ Arent Jan Wensinck et al., Concordance et indices de la tradition Musulmane, 7 vols. (Leiden: E. J. Brill, 1933-1970).

¹⁸ The sources examined include Abu Dawud, al-Bukhari, al-Darimi, Ibn Hanbal, Ibn Maja, Ibn Sad (al-Tabaqat), Malik, Muslim, al-Nasai, Ibn Hisham (Sira), al-Tabari, Muhibb al-Din (Al-Riyad al-Nadira), al-Tirmidhi and al-Waqidi.

¹⁹ This suspicion was confirmed by the fact that an examination of sources which lent themselves to perusal revealed parallel traditions with similar meaning but significantly different wording.

²⁰Uthmaniyya, pp. 6 and 74.

²¹Ibid, p. 27.

²²Ibid., p. 27.

²³Ibid, p. 33.

²⁴Abd al-Malik b. Hisham, al-Sira al-Nabawiyya, ed. Mustafa al-Saqqa, et al, 2nd ed., 2 vols. (Cairo, 1955).

²⁵Muhammad b. Umar al-Waqidi, Kitab al-Maghazi, ed. Marsden Jones (London: Oxford University Press, 1966).

²⁶Uthmaniyya, pp. 30 (7 lines) and 32-35 (34 lines), the similarities between the two sources are rendered even more striking by the fact that the minor differences in wording favor al-Jahiz as a composer of literary prose.

²⁷Ibid, pp. 52 (Sira, I, 492), 38 (Sira, I, 263), 69 (Sira, I, 398), 52 (Sira, I, 492), 43, 44 (Sira, I, 483).

²⁸Ibid, pp. 29 (2 lines), 31 (1 line), 37 (3 lines), 38 (4 lines) and 41 (5 lines); 37 (Sira, I, 350), 51 (Sira, I, 484), 79 (Sira, II, 655), 83 (Sira, II, 663) and 85 (Sira, II, 649).

²⁹Ibid., pp. 62 (Maghazi, p. 257), 63 (Maghazi, p. 257), 65 (Maghazi, p. 1121), 67-68 (Maghazi, pp. 107-9 [36 lines]).

³⁰Ibid., pp. 64 (Sira, II, 312; Maghazi, p. 581), 70 (Sira, II, 316; Maghazi, pp. 611-12), 71 (Sira, II, 93-94; Maghazi, pp. 292 and 297) and 72 (Sira, II, 396; Maghazi, p. 792).

³¹Ibid., pp. 53 (Sira, I, 620 and 626; Maghazi, pp. 297 and 300) and 56 (Sira, I, 614; Maghazi, p. 293).

³²Ibid., pp. 134, 144 and 162.

³³ See Alfred Guillaume, "Introduction," in The Life of Muhammad: A Translation of [Ibn] Ishaq's Sir at Rasul Allah (London-New York-Toronto: Oxford University Press, 1955), p. xxxii; J. Horovitz, "The Earliest Biographies of the Prophet, IV," Islamic Culture, II (1928), 519; and Marsden Jones, "Muqaddimat al-Tahqiq," in Maghazi, pp. 30-34.

³⁴ Several books listed by Muhammad b. Ishaq b. al-Nadim, Fihrist, ed. G. Flugel, I (Leipzig: F. C. W. Vogel, 1871), 104, could have contained such accounts: Kitab al-Tarikh wa 'l-Maghazi wa 'l-Mabath, K. Akhbar Makka, K. al-Sira, K. Wafat al-Nabi, etc.

³⁵ Our translation.

³⁶Uthmaniyya, p. 120.

³⁷ S. 9:40 (Uthmaniyya, pp. 100-12/Sira, I, 482-86).

³⁸ S. 24:18 (Uthmaniyya, pp. 112-13/Sira, II, 303/Maghazi, p. 434).

³⁹Uthmaniyya, p. 111.

⁴⁰Encyclopaedia of Islam, II, 1023.

⁴¹ Amr b. Bahr al-Jahiz, "Kitab Hujaj al-Nubuwwa," in Rasail al-Jahiz, ed. Hasan al-Sandubi (Cairo, 1933), p. 119.