History of Sunni Hadith
Download as PDF (best formatting)
a] This is one of the six major collections (Sihāh Sittah):
The Five Book canon, which is first noted in the sixth/twelfth century, incorporates the Jami’of al-Tirmidh_ (d. 279/892). Finally the Six Book canon, which hails from the same period, adds either the Sunan of Ibn Majah (d. 273/887), the Sunan of al-Daraqutni (d. 385/995) or the Muwatta of Malik b. Anas (d. 179/796). Later hadith compendia often included other collections as well. None of these books, however, has enjoyed the esteem of al-Bukhari’s and Muslim’s works.
Note that it took about two hundred years for the Hadith Canon to be established. Bukhari died in 870, yet his Hadith collection purportedly presents an authoritative report of the life and sayings of Muhammad who is reported by Muslim sources to have perished in 632. There is thus a problem of temporal separation between the supposed narration or action and its written record.
b] Powers observes that there is no consensus among Muslim scholars that all Bukhari’s hadiths are sahih. Furthermore, there are several different versions of his corpus:
Although all of the Hadiths in al-Bukhāri are claimed to be ṣaḥīḥ¸not all scholars are agreed. Nor have these collections come down in a single unchallenged edition. AlBukhari’s text, for example, “exists in several ‘narrations’ (riwayat), of which the version handed down by al-Kushmayhani (d. 389) on the authority of Bukhari’s pupil al-Firabri is the one most frequently accepted by the ulema.”
Quiring-Zoche also observes that al-Yunini noted ‘discrepancies between the texts of al-Kusmihani and Ibn Hamiuwayhi…’ regarding versions of the Sahih. She states that al-Yunini did not establish his own recension, but rather ‘a text with variants and a critical apparatus.’ She also states that al-Yunini’s personal copy of the Sahih, which consisted of two volumes, was lost, one volume in the sixteenth century, the other sometime after 1909.
c] Reports of Bukhari’s life seem to be late, and arguably embellished. For example, Brown states: ‘It is reported that al-Bukhari devoted 16 years to sifting the hadiths he included in his Sahih from a collection of nearly 600,000 narrations.’ His source for this is the Egyptian Taqi al-Din al-Subki, 1284-1355 AD, whereas Bukhari died in 870!
According to Abdul-Jabbar, Bukhari stayed in Mecca after his haj aged 16, and then travelled to various parts of the Empire. Maujood has someone called Sahl ibn a-Sirree (this may be a variant of as-Suri) quoting Bukhari as stating: ‘I visited Ash-Sham (Syria and surrounding regions) and Egypt; the Arabian Peninsula twice; and I resided in Al-Hijaaz for six years, and I cannot count the number of times I entered Kufah and Baghdad with other Hadeeth scholars.’
However, Majood’s sources for this are Taareekh Baghdad, written by al-Khatib al-Baghdadi, d. 1071, who may have finished as-Suri’s work. This becomes more complicated when we look at The Creed of The Imaam of Hadeeth Al-Bukhari & of The Great Scholars From Whom He Narrated, pp. 9-10, which states:
A1-Laalikaa’ee1 (d. 418H) said in his Sharh Usool I’tiqaad
Ahlus-Sunnah (2/172):
‘The Creed (I’tiqaad,) of Aboo ‘Abdullaah Muhammad ibn Ismaa’eel al-Bukhaaree and the Group from the Salaf about whom he narrates.’
Ahmad ibn Muhammad ibn Hafs al-Harwee said: Muhammad ibn Ahmad ibn Salamah narrated to us, saying: Abul-Husayn Muhammad ibn Imraan ibn Moosaa al-Jarjaanee narrated to us saying: I heard Aboo Muhammad ‘Abdur-Rahmaan ibn Muhammad ibn ‘Abdur-Rahmaan al-Bukhaaree from ash-Shaash – saying I heard Aboo ‘Abdullaah Muhammad ibn Ismaa’eel al-Bukhaaree saying:
I met more than a thousand men amongst the people of knowledge from the people of al-Hijaaz, al-Makkah, al-Madeenah, al-Koofah, al-Basrah, Waasit, Baghdaad, Shaam and Egypt. I met them numerous times, generation after generation and then generation after generation. I met them while they were ample and widespread for over forty-six years; the people of ash-Shaam, Egypt and al-Jazeerah twice, (the people of) al-Basrah four times in a number of years. (Those of) al-Hijaaz (over a period) of six years and I cannot enumerate how many times I entered (upon the people of) al-Koofah and Baghdaad along with the muhaddithoon of Khuraasaan...
While not wanting to be skeptical for its own sake, one wonders how we can test this claim, and if it actually goes back to Bukhari himself.
Majood’s other source is Siyar A’lam al-Nubala, by Muhammad ibn Ahmad ibn `Uthman ibn Qayyim `Abu `Abd Allah Shams ad-Din al-Dhahabi, d. 1348! Again, the late date is significant.
d] Another problem concerns extant copies of Bukhari’s corpus. There are indications about the age of Bukhari’s transmission: ‘When compared with the enduring transmission of the Sahih from al- Bukhari’s most famous student, al-Firabri, his other student Hammad b. Shakir’s (d. 290/902–3) recension of the text contained two hundred fewer narrations. Ibrahim b. Ma’qil al-Nasafi’s (d. 295/907–8) was three hundred less.’ When we examine the date of Firabri’s death - Muhammad b. Yusuf al-Firabri (d. 320/932) – we see how far from the original we are.
Melchert also notes ‘the meagre circulation of the Sahih before the mid-tenth century C.E…’ Linked to this is his comments on the reliability and transmission of the Sahih:
Quotations and internal evidence together suggest that al-Tarikh alkabir and al-Awsat are securely attributed to Bukhari but underwent amendment both by him and others, mostly over the middle third of the ninth century. The Sahih is a little less securely attributed. First, there is the evidence that it did not begin to circulate significantly until well into the tenth century. Tirmidhi’s evidently limited access to it has been mentioned already. No commentary appeared until the mid-tenth century, although commentaries on Muslim’s Sahih began to appear in the first generation after him. Five men are named as transmitting it from Bukhari, only two of whose recensions are preserved at all in the extant commentaries, mainly those of Ibrahim ibn Ma’qil al-Nasafi (d. 295/908?) and MuIammad ibn Yusuf ibn Matar al-Firabri or Farabri (d. 320/932). Firabri’s reported boast that ‘Ninety-thousand men heard Kitab al- Sahih of MuIammad ibn Isma’il, but I am the only one left to relate it’ is not credible… Secondly, there is evidence of textual instability into the tenth century.
The weight of Bukhari’s work becomes even more inconclusive when we consider the manuscript evidence:
As for manuscript evidence, the earliest extant fragment of the Sahih was found by Alphonse Mingana to present hadith in a different order from that of the familiar text, with occasional variations of spelling, addition or omission of pronouns, and the other sorts of variations that the commentaries preserve. Jonathan Brown is dismissive: ‘Mingana’s partial manuscript of the Sahih consisted of only three chapters. We have no evidence that the ordering of the remaining ninety-four chapters was irregular’. To the contrary, I would say, it seems highly unlikely that the three chapters preserved in Mingana’s Manchester manuscript were exactly the three characterized by such irregularities, the others being identical to the familiar text. Moreover, we actually do have a great deal of evidence, mainly in the commentaries, that such irregularities did characterize the whole of the Sahih. Mingana’s manuscript should be welcomed as confirmation of the commentary tradition.
The basis for contemporary copies of the Sahih only go back to the fourteenth century A.D.:
Only versions of Firabri’s recension, not even the recension of Nasafi as well, were available to al-Yunini (d. 701/1302), whose work formed the textual basis of the so-called Sultani edition of 1311-1313 (mid-1890s), the main basis in turn of subsequent editions and the closest we have to a standard text today.
The actual manuscript to which Mingana referred is ‘Mingana Arab. Isl. 225’, which according to Mingana ‘contains an important text of the second part of the famous collection of Islamic Traditions by Bukhari. The kitabs which it contains are only the Zakat, the Saum, and the Hajj.’ Mingana observes the palæographical evidence for its dating:
The MS. is unfortunately incomplete at the end, and so bears no date, but on palæographical grounds it cannot be later than A.D. 1000, and is probably earlier, and so may easily be ascribed to not more than about a hundred years after the author’s death, which took place in A.D. 870. This is borne out also by the fact that the title-page contains inscriptions, one of which is dated Ramadan A.H. 464 (A.D. 1072), and the other Dhu’l Hijjah 574 (A.D. 1178).
Mingana goes on to note: ‘The MS. probably contains the oldest text of Bukhari in existence.’ This means that the oldest extant manuscript of the collection usually seen as second only to the Qur’an exists only as a partial fragment. A further consideration should be noted:
A much more important feature is that every section begins with the sentence: “Bukhari has informed us, saying” …which would imply that it was not Bukhari himself who wrote the text of his famous book, but one of his disciples. One is tempted to go farther, and to state in this connection that it was not an immediate disciple of Bukhari who first committed it to writing, but a hearer of one of these immediate disciples. What renders this hypothesis almost certain is the fact that sometimes a chain of two authorities separates Bukhari, the author, from the man who first put the book down in writing. So the first bob of Kitab az-Zakdt begins as follows: “We have been informed by Abu Zaid Muhammad b. Ahmad, who said that Muhammad b. Yusuf told us, saying that Bukhari informed us saying: ‘We have been told by Abu ‘Asim Dahhak b. Mukhallad,’” etc.
After comparing the text with other editions, Mingana concludes: ‘…the text of Bukhari underwent many vicissitudes before it reached its present standardized form.’ This surely raises further questions about the credibility of Bukhari’s Sahih as we presently possess it. The Schoenberg Database of Manuscripts records Schoenberg 31539 Al-Jami Al-Sahih Part 36, 20 Folios, 26 Lines, and dated to 1138.
Interestingly, the Encyclopaedia of Hadith declares that among the sources it used for its edition of Sahih Bukhari was the following: ‘The Dar al-Shab (Cairo) edition, which was baed [sic] on the Sultaniyya edition, which was in turn drawn from the celebrated and authoritative copy of Imam al-Yunini.’ The rest of its sources are commentaries from the fourteenth and thirteenth centuries A.H., i.e. the twentieth and nineteenth centuries A.D. Significantly, Quiring-Zoche observes: ‘This Yuniniyya was solely used by al-Qastellani to a decisive extent. It is only through his commentary on al-Buhari that al-Yunini’s redaction of the Sahih has been known to us.’ Shihab al-Din Abu’l-Abbas Ahmad ibn Muhammad ibn Abu Bakr al-Qastallani al-Qutaybi al-Shafi’i lived 1448-1517, and his commentary is called Irshad al-Sari fi Sharh al-Bukhari.
In conclusion, it seems that extant mss. of Bukhari are based on al-Yunini (d. 701/1302, so (fourteenth century AD), possibly by way of al-Qastallani (fifteenth-sixteenth centuries AD), Asqalani’s Fatḥ al-Bārī fī Sharḥ Ṣaḥīḥ al-Bukhārī (fifteenth century AD), and Umdat al Qari by Badr al Ayni (also fifteenth century AD). The Maknaz edition uses a text from 873 A.H. – 1468 A.D.
a] It is often stated that this is the earliest hadith collation: ‘An example of an early collection of hadith is the sahifa of Hammam b. Munabbih (d. circa 130/747), a disciple of Abu Hurayra, which includes 138 hadiths…’ However, its authenticity is questioned:
…it is contended here that what goes by the title of Hammam’s sahifa is for the main part the handiwork of none other than ‘Azq. He used the strand to support a number of partly brand-new, partly other, older traditions of his own making or copied from others, which he had already circulated with the help of strands of an older vintage. ‘Azq.’s Ma’mar/Hammam/Abu Hurayra strand was, in turn, copied by other, younger collectors in order to serve them as a convenient prop for their own traditions, mainly shawahid…
In his evaluation of the transmission via Ma’mar/Hammam/Abu Hurayra, the modern editor of the ‘corpus’, Muhammad Hamidullah (d. 2002), failed to appreciate that there is an anomaly concerning the recorded death dates of the first three transmitters, something which made it well-nigh impossible to attribute any historicity to a strand of this sort. Consider the following data:
Abu Hurayra is reported to have died in 57/677 or 58/678 or 59/679. For Hammam b. Munabbih’s year of death two conflicting clusters of dates are given, no less than thirty years apart. On the one hand, it is recorded in Ibn Sa’d (V, p. 396) that it was in 101 or 102/719-20; however, on the other hand, in the quotations of the same Ibn Sa’d in Mz., Tahdhib, XXX, p. 300, and Ij., Tahdhib, XI, p. 67, the date given is 131/749, while Dhahabi in Siyar, V, p. 312, states that Hammam died in 132/750. The problem is then aggravated by the additional information in IS that Hammam is supposed to have died before his older brother Wahb (d. 110/728 or 114/732), so the apodictical solution of the editor of Mz.’s Tahdhib, namely that the Sachau edition of IS is mistaken, is, for lack of a better term, too hasty. In the final analysis it is impossible to say which year from any of the two clusters for Hammam’s death is the most feasible.
In short, no date seems to work, when the historicity — if any — of this strand is evaluated. Be that as it may, if the second cluster of dates (131 or 132) is postulated to be the correct one, Hammam must have listened to hadith recited by a man who died some three quarters of a century earlier. The narrative embellishment describing Hammam as being so old that his eyebrows rested upon his eyes is a widely used topical digression to underline Hammam’s status as a mu’ammar. However, Dhahabi’s rijal lexicon of centenarians, entitled Ahl al-mi’a fa-saida, does not list Hammam in any case.
And if the first date (101 or 102/719-20) is taken as point of departure, Ma’mar is supposed to have learned the traditions from his informant Hammam some nine years before he actually commenced his hadith gathering, which, as the appurtenant sources inform us, was in the year Hasan al-Basri died, namely 110/728. Hammam occurs twice in spidery bundles with Sufyan b. ‘Uyayna as key figure, where he sits between that man’s informant ‘Amr b. Dinar (d. 126/744) and Muawiya b. Abi Sufyan (d. 41/661). In these two, Hammam’s year of death is apparently assumed to have been the earlier of the two, 102/720. If Hammam’s late year of death is taken literally, creating every opportunity for Ma’mar to have heard him personally, and if it is at the same time maintained that he was a transmitter of Abu Hurayra, he must have reached an age which requires an act of faith to accept. A solution seems in any case to lie beyond the reach of the modern isnad analyst.
Summing up, one can say that the historicity of the transmission of the so-called Sahifat Hammam cannot be maintained with any reliable measure of certainty. References to young children learning hadith with ancient hadith masters are admittedly legion, but they are all situated in a later stage in the development of hadith instruction, a stage which originated several centuries later than the time we are here and now digging into, namely the second half of the first/seventh century and the first few decades of the second/eighth century. Besides, when one reads through all the traditions of the Sahifa, the overall impression one gains, is of a late collection, displaying a stylistic finish only found in relatively late traditions. Moreover, they present virtually all the Prophet’s direct speech with a few qudsi sayings directly attributed to God thrown in, only a few have Muhammad being asked questions to which he provided answers. The whole corpus is supported by a supposedly very early but probably historically untenable isnad strand.
b] Morris makes these observations about the Sahifah in terms of actual dating, manuscript evidence and variants:
It’s sometimes billed as the oldest extant hadith collection. That does not mean that our surviving manuscripts are the earliest to contain hadiths – the current scholarly edition by Hamidullah is based on manuscripts from the twelfth century and later –; it means that the text itself, preserved in later volumes, may represent the earliest collection.
Between variants of the Sahīfah, differences in wording are few and trifling. However, as Marston Speight has shown, there are many and significant differences between the hadiths in the Sahīfah and variants of the same hadiths transmitted elsewhere. This is what we’d expect. The ninth-century turn to writing and publication ‘fixed’ a body of oral tradition that was previously fluid and adaptive: quite different accounts of the same story could then be preserved in parallel.
The point is that the Sahīfah was rigorously transmitted once it had been written, but it was only written in the ninth century… It would appear that ‘the Sahīfah of Hammām’ is more properly the Sahīfah of ‘Abd al-Razzāq, and a thoroughly ninth-century work.
Musanaf of Abdul Razzaq
a] ‘Abd ar-Razzaq was born in 744 A.D. – so over a hundred years after Muhammad’s death according to Muslim sources - and died in 827 – nearly two hundred years after Muhammad is claimed to have died. Note this: ‘This book is written by Hafiz Al-San’ani, Musannaf of Abd al-Razzaq. Its original menuscript [sic] was lost and mixed up with other books, and could not be differentiated for almost 1100 years until it was arranged and edited by the Indian Scholar Maulana Habib al-Rahman al-Azmi…’ Motzki also notes its manuscript history:
The Musannaf of ‘Abd al-Razzaq ibn Hammam al- San’ani (d. 211/827) has been available since 1972 in an eleven-volume edition prepared by Habib al-Rahman al-A’zami and published by al-Majlis al-’Ilmi, Beirut…
1. The manuscript Murad Mulla (Istanbul) is the basic text. It consists of five sections and is -by al- A’zami’s estimation - complete, aside from small losses at the beginning of the first and fifth sections. This judgement can only apply to the part of the text covered by this manuscript, because the end of the work is missing. This manuscript dates from the year 747I1346-7. It represents the basis for volumes one through ten, page 145 of the edition.
2. For the rest, the manuscript Fayed Allah Effendi (Istanbul), from the year 606/1209-10, was used.
The dates here are thus fourteenth and thirteenth centuries. It follows that extant copies of this work are extremely late. The reconstructed nature of the latter also raises questions.
b] Abd al-Razzaq was from Yemen, a student of Mālik:
He brought into circulation large numbers of traditions, which he copied from earlier CLs after providing them with isnād strands of his own making... He became blind at the turn of the second/eighth century...
It is tempting indeed to speculate that ‘Azq., cooped up in Yemen as he was, had far less opportunity to meet hadīth masters from whom he could learn a thing or two, or that he did not bother to look for such masters. So he simply produced a huge portion of his Muṣannaf himself, all this support d by some of his favourite SSs, and all this on a far more extensive scale than the two other collectors mentioned above. All three mawālī have enriched Islam: the first two did it mainly through transmission of what was already there, and the third one by cleverly introducing many brand-new hadīths...
‘Azq. apparently also made use of the Muhammad b. ‘Amr b. ‘Alqama/Abū Salama / Abū Hurayra strand in order to circulate traditions which are probably of his own making.
Motzki also notes that people such as Hanbal and An-Nasai had problems with Abdul Razzaq because of issues of reliability:
Furthermore, he is supposed to have dictated texts from memory. Because of this, Ahmad ibn Hanbal deemed the traditions of people who studied with him in this period to be da’if (unreliable).e3 Later scholars such as Ibn al-Salah (d. 643/1245-6) joined him in this opinion and-following al-Nasa’i (d. 303/915-6) insisted that texts deriving from ‘Abd al-Razzaq be tested, whether to distinguish the later from the earlier, good transmission, or because they generally distrusted him and only wanted to accept the traditions attested elsewhere as well.
In the light of this, we can understand why this never became an authoritative collection, and why Bukhari felt the need to sift ahadith.
c] Note also the comment of Lucas, p. 290: ‘‘Abdal- Muṣannaf ... draws heavily upon the collections of Maʿmar b. Rāshid (d. 153/770), Sufyān al-Tawrī (d. 161/778), and Ibn Jurayj (150/767)...’ Even his sources are late.
Muwatta Imam Malik
a] Guillaume states the following (bold emphasis ours):
The earliest date which Muhammadans give for the collection of hadith is contained in the following tradition, said to rest on the authority of Malik b. Anas (94-179) –
‘Umar b. Abdu-l-’Aziz wrote to Abu Bakr. B. Muhammad b. ‘Amr with the order: ‘See what hadith of the prophet of God are extant or ancient customs (sunna madiya) or hadith known to ‘Amra, and write them down; for I stand in dread of the disappearance of knowledge and of the death of them that possess it’ This Aba Bakr. B. Muhammad was one of the Ansar whom ‘Umar II appointed judge at Medina, and ‘Amra was his aunt. Of the statement Sir William Muir writes: ‘About a hundred years after Muhammad, the Caliph ‘Umar II issued circular orders for the formal collection of all extant tradition. The task thus begun continued to be vigorously prosecuted; but we possess no authentic remains of any compilation of an earlier date than the middle or end of the second century of the Hijra.’
It would seem that this writer accepts the statement at its face value, but the fact that no authentic remains of this alleged first-century compilation are extant, and that the indefatigable students and compilers of tradition in the third century make no mention of an effort to trace such early documents, suggest very strongly that the tradition is not based on fact. It is difficult, if not impossible, to suggest a cogent reason why such an early collection, if it existed, should never have been mentioned by later scholars whose life-work it was to recover the genuine hadith of the apostolic period. For this reason the hadith must be regarded as an invention designed to connect the pious caliph, whose zeal for the sunna was gratefully recognized by theologians, with the tradition literature of Islam. This seems the more likely, as another tradition connects Ibn Shihab Al Zuhri with ‘Umar II in this work. Moreover, Malik’s statement is only to be found in Al Shaibanl’s recension of the Muwatta. It is absent from the other versions.
Note what Guillaume states about the various recensions/versions of Al-Muwatta – there was, initially at least, no fixed, canonical edition of the text. Interestingly, Introduction to Translation of Malik’s Muwatta, translated by `A’isha `Abdarahman at-Tarjumana and Ya`qub Johnson, has this to say:
Malik (full name Malik bin Anas bin Malik bin Abu Amir Al-Asbahi) was born in 93 A.H. and died in 179 A.H. He lived most of his life in Madinah, the city in which the Prophet (pbuh) settled in. He was a preeminent scholar of Islam, and is the originator of the Maliki judicial school of thought. He is reputed to have had over one thousand students. During Malik’s lifetime, he steadily revised his Muwatta, so it reflects over forty years of his learning and knowledge. It contains a few thousand hadith.
So, according to this, Malik, the author of Al-Muwatta, was born nearly a century after Muhammad (711 AD), and Al-Muwatta was completed by the time he died – 179 AH (795 AD), so nearly two centuries after their Prophet (died 632 AD). Note that the Muslim authors of the translation above state that Malik ‘steadily revised his Muwatta’ over forty years; why, if the narrations were strong and valid, would he need to do so?
b] There are questions as to how historically reliable it is. Goldhizer has this to say on the subject:
The Muwatta’ cannot be regarded as the first great collection of traditions in Islam, nor does not appear to have been considered as such in Muslim literature. Despite the great prestige which it has enjoyed, from its appearance to this day, in the east and west of the Islamic world - the history of its origin has been surrounded with large number of pious legends - and despite the great reverence shown to the name of the author, the great imam dar al-hijra, it did not originally gain its authority as a canonical work of tradition… with the exception of the Maghribi schools, this work has no place amongst the ‘six books’ … and only the reverence of later generations, who were no longer in close touch with the origins and had the urge to widen the circle of canonic literature, occasionally included in that category.
The work of Malik is in fact not in the proper sense a collection of traditions, forming a counterpart to the sahihs of the next century, nor one which could, from the point of view of the literary historian, be mentioned as a member of the same literary group. It is a corpus juris, not a corpus traditionum… Its intention is not to sift and collect the ‘healthy’ elements of traditions circulating in the Islamic world but to illustrate the law, ritual and religious practice, by the ijma’ recognized in Medinian Islam... Inasmuch as the book has anything in common with a collection of traditions it lies .in the sunna rather than the hadith. Occasionally Malik does not cite one single tradition in a paragraph but only cites fatwas by recognized authorities in actual or casuistically pointed cases in order to conclude with his own assenting opinion and by stating Medinian usage and consensus. A transmitter of the hadith school would have put forward not fatwas, but hadiths going back.to the Prophet...
The variants in copies of the Muwatta must be recognized:
An unfavourable impression of the reliability of Islamic tradition in the second century is gained if one considers that the version in which various authorities hand down the Muwatta’, all directly, or indirectly, in the name of Malik, differ from each other in their text and contents, as well as in plan and order, to such a degree that One might be tempted to think of them as mutually differing and by no means as identical writings.
Rauf also recognizes this:
The Muwatta9 was revised several times over forty years by its author, who flourished in Medina, having studied earlier with renowned scholars there, and in turn taught those revised works to his disciples. Malik’s revised work survived in some different versions through his disciples, notably Yahya b. Yahya al-Laythi of Cordoba (d. 232/848), and of Muhammad b. al-Hasan al-Shaybanl (d. 189/804), the well known Hanafi authority. Yahya’s version is the more popular.
We note the continual revision of the work and the existence of different versions.
c] We should also observe that there is some uncertainty about Malik’s life in general:
His year of birth is variously recorded as 90/709, 93/712, or 94/713. However, when certain reports on his age at a particular point in time are taken literally, it is likely that he was born at a date considerably later than the birth dates given here. Beside doubts about his supposed year of birth published elsewhere, there is a relevant anecdote describing how some one in 134/751 was asked about the natural successor of the mawlā Rabī’a b. Abī ‘Abd ar-Rahmān, always referred to as Rabī’a ar-Ra’y, i.e. the man well-known for his independent legal thinking, who held sway in the mosque of Medina. After mentioning Yahyā b. Sa’īd al- Ansārī (d. 143-6/760-3), the person questioned ventured: ‘And then there was that boy (fatā), Mālik b. Anas’. This permits the speculation that Mālik was in fact born, say, some twenty years before 134/751, a calculation which depicts him at the time of his demise as some sixty-five years old, an age arguably more plausible, at least for the second/eighth century, than the age of ca. eighty-five years computed from his alleged year of birth given above. Furthermore, there is a report attributed to the early rijāl expert ‘Alī ibn al-Madīnī in which he clearly stated that Mālik never knew Ibn Ishāq personally or that he heard traditions with him. This is hard to reconcile with the age of Mālik, if we take his alleged early birth in the nineties as point of departure.
d] The reliability of his transmission is questioned:
Once Mālik’s reputation for ‘uluww ar-riwāya became established in this way, it seems never to have caused hadīth scholars to express words of doubt until the present day. And this belief and acceptance was concomitant with the circulation of Iraqi and Egyptian diving SSs, many of which formed themselves in to undatable spiders. All of these centred in Mālik’s alleged masters such as Nāfi, Mūsā b. ‘Uqba, or others. Mālik could furthermore be found to have resorted on various occasions to the formation of apocryphal family isnāds, a conspicuous feature within the corpus of isnād structures that he used to support his traditions with. In the following, a number of family isnāds will be scrutinized, sometimes headed by one of those ‘longeval’ and obscure authorities, whom Mālik seemed to ‘monopolize’. This seems Mālik’s answer to the methods developed by his Iraqi counter parts in Kūfa and Basra. In those cities particularly longliving, fictitious persons, or historical persons who were claimed to have died at incredibly advanced — and thus probably fictitious — ages, were inserted in isnād strands that were meant to bridge the whole first/seventh century.
This needs to be emphasized. Malik was not above ‘creative’ license in producing isnāds, notably the insertion of his own family members into the mix:
The devising of apocryphal family isnāds (i.e. on the authority of someone/his father/his grandfather and so forth)…. These family isnāds did not catch on immediately with all hadith collectors, but in the end even Malik b. Anas (d. 179/795), Medina’s most influential and prolific CL, produced family isnāds listing various of his own family members, to wit fictitious uncles. And also Bukhari and Muslim, initially somewhat hesitant to adopt them, ultimately made use of them to authenticate their traditions.
This is a very serious matter; if Malik, Bukhari and Muslim all invented people – especially family members, whether to explain where they got their narrations or to embellish their genealogies to show some connection to their Prophet, then how historically trustworthy is the entire Hadith corpus? The whole concept of isnād becomes irrelevant as an historical foundation for transmission, since essentially elements of it, at least, are apocryphal – even forged.
e] This is even more emphasized when we consider that Malik altered his own age to qualify as one of the hearers in the chain of transmission:
The CL himself projecting his own year of birth back into the past so as to make himself a believable transmitter of his spokesman. Most eminent examples of CLs who, by pulling out all the stops, resorted to stretching their own birth into the past so as to make their contacts with particular ancient hadīth masters historically acceptable were Mālik b. Anas (d. 179/795) and the mawlā Sufyān b. ‘Uyayna (d. 198/814), who blithely emphasized time and again that they were just about old enough to have received traditions respectively from Ibn ‘Umar’s mawlā Nāfī’ (d. 117/735) and Zuhrī (d. 124/742).
Ironically, one of the criteria in Bukhari’s evaluation of hadiths was the character of the transmitter: ‘The hadith critic will have some direct experience of the people he narrates from, experience in turn narrated to later scholars. For example, the scholar might have caught a certain narrator lying. Or … that, at a certain age, a narrator’s memory had started to fail him.’ According to this criterion, we should not trust Malik and his Muwatta. How genuine are his isnāds, and indeed, the matn of his hadiths, if he concocted narrators?
f] How old is the earliest extant manuscript of Malik’s Muwatta? There is a copy of the Muwatta in Dublin’s Chester Beatty Library, Mss. 3001. It is described as ‘The second ‘third’ of a celebrated treatise on Islamic jurisprudence’, and is dated to 277/890. According to Islamic Awareness, the earliest extant manuscript is a fragment entitled PERF No. 731 in the Austrian National Library, Vienna, ‘Second half of 2nd century of hijra. This papyrus fragment is dated to Mālik’s (d. 179 AH / 795 CE) own time.’ Their source for this is ‘N. Abbott, Studies In Arabic Literary Papyri: Qur’anic Commentary and Tradition, 1967, Volume II, University of Chicago Press: Chicago (USA), p. 114.’ Elsewhere, they quote her (p. 127) as stating:
Thus the paleography, the scribal practices, the text, the order of the traditions and the isnad terminology in the papyrus show a remarkable degree of conformity with the scholarly practices of Mālik and his contemporaries. On the strength of this internal evidence the papyrus folio can be safely assigned to Mālik’s own day.... The codex represented by our folio therefore originated sometime during the quarter century or so that elapsed between the writing of the Shaibani and the Laithi recensions and hence must represent one of the many lost recensions of that interval. Inasmuch as the papyrus text shows only minor variations from the printed text of the Laithi vulgate it is even possible that it represents the vulgate text as it was before it received....
This omits her words: ‘It could have been produced by his secretary-copyists, dictation and reading masters, advanced pupils, or admiring fellow scholars. As already pointed out, the text is not in the Shaibani recension but is essentially that of the vulgate as transmitted by the Spanish judge and jurist Yaḥyā ibn Yaḥyā al-Laythī (d. 234/848), who heard the Muwaṭṭaʾ from Mālik shortly before the latter’s death in the year 179/795.’ It should be noted that a 2008 thesis by Karim Samji observes concerning the fragment that:
Punctuation includes O with a diagonal dividing line (e.g., Recto, Line 9). The marginalia of Recto, Line 21, contains a circle with no intersection. The earliest case for this device in the shape of an inverted heart, with or without a vertical line intersecting the middle, in lieu of the circle with or without a dot, is dated to the first half of the third/ninth century.
Significantly, the footnote for this observation is ‘Nabia Abbott, Studies in Arabic Literary Papyri, vol. 1: Historical Texts (Chicago: University of Chicago, 1957), 61’, where the reference is to PERF 665 The Sirah of Ibn Hisham. Samji’s work dates PERF No. 731 (which he refers to as Ms Or. P173 to the ‘Third/ninth century’. He does this on the authority of ‘Petra Sijpesteijn (Personal e-mail correspondence (12 February 2008)’. Sijpesteijn is professor of Arabic and a noted papyrologist who has studied the papyri at the Austrian National Library first hand. By contrast, Abbott informs us that she had studied photostats of PERF No. 731. Therefore, the evidence indicates that this fragment actually should be dated to the ninth century – and so is not ‘early’.
Musnad of Abu Dawud Sulaymān ibn Dāwūd b. al-Djarud Al-Tayalisi
The scholar Sulaymān ibn Dāwūd al-Djarud Al-Tayalisi was born in 133/751 and died in 203 or 204/819-820. His musnad has been suggested as one of the earliest hadith collections. Juynboll writes about Al-Tayalisi:
He was renowned for his memory; he boasted that he could recite 30,000 traditions without pause. He was especially acclaimed for the many traditions he said that he had learned with Shu’ba b. al-Hadjdjadj (d. 160/776 [q.v]), the greatest Basran traditionist of his time... Al-TayalisI is particularly famous for his collection of traditions, called the Musnad. It is preserved in an ancient Haydarabad edition of 1321/1903 which contains 2,768 traditions, some two dozen of which strike one by their remarkable length...
When his Musnad is compared with the collections entitled Musannaf made by his younger colleagues ‘Abd al-Razzak al-San’ani (d. 211/827 [q.v.]) and Abu Bakr b. Abī Shayba (d. 235/849 [q.v.]), each of which does indeed contain several tens of thousands of reports, the surmise is tenable that al-Tayalisi’s Musnad, as we have it now, is a collection from which all the reports supported by less well-attested isnad strands have been removed according to the isnad criteria of post-classical times: it does not contain Companion reports (mawkufat), hardly any Successor reports of the type called mursal [q.v.] and no personal opinions of early fukahā that go by the term akwal, all three categories occurring in abundance in the Musannafs mentioned. Although this is not substantiated in the Haydarabad edition or anywhere else as far as can be ascertained, al-Tayalisi’s Musnad as we have it now may be an anonymously abridged version of an original corpus several times the size of the present Musnad, a corpus that is apparently no longer extant.
Boekhoff-van der Voort also suggests that ‘Abū Dāwūd al-Ṭayālisī adjusted his tradition over time.’ In the chapter examining ‘The Raid of the Hudhayl: Ibn Shihāb Al-Zuhrī’s Version of the Event’, and its different versions in the Traditions, the author suggests the following:
Finally, another possible explanation is that either Abū Dāwūd al-Ṭayālisī or Yūnus ibn Ḥabīb transmitted the story orally instead of through writing or dictation. Oral transmission – probably combined with written notes – could cause differences such as a different order in the elements, omission of elements, different formulation…
The revealing comments about oral tradition in this field raise further questions about the accuracy and reliability of transmission. Rauf also notes that the collection has been criticized for inaccuracy: ‘The Musnad of Sulayman b. Da’ud al-Tayalisi (d. 204/818), which contains 2,767 hadith, is believed to be the first musnad. Critics say that it includes some errors.’
Schoenberg (59499) lists a manuscript of ‘Abu Dau Tayaisi’ entitled Al-Musnad, dated to 1250 A.D., from Khuda Bakhsh Oriental Public Library, Patna, Bihar. The latter renders it as ‘Musnadu Abî Dâ’ûd At Tayâlisî’, and describes it as ‘An old and extremely rare copy of Musnad Tayalasi, containing a collection of Musnad Hadis’. They add: ‘The copy, though not dated, was evidently written in or before the 7th century A.H. No other MS. copy of the work seems to be extant.’ The seventh century A.H. is equivalent to the thirteenth century A.D. The basis for their judgment seems to isnad, rather than paleographical concerns. They further add that the Hyderabad edition is exclusively based on this manuscript.
If we consider the quality of his narrations, and the suggestion that these have been edited by someone(s), then we must question the quality of the work in general. The fact that it was never considered as being one of the canonical collections is suggestive of its standing.
Sunan of Abu Dawud Sulaymān ibn al-Ash’ath al-Azdi as-Sijistani
a] Abu Dawud was born in 817 and died 889. His work, the Sunan, consisted of ‘4800 traditions from a mass of 500,000, and that it contains sound traditions, those which seem to be so, and those which are nearly so.’
b] What is interesting is its transmission history: ‘The Sunan was transmitted through several lines, some versions being said to contain material not found in others. Al-Lu’u’i’s version is the one which has gained most favour. A number of editions of the Sunan have been printed in the East…’ Hence, there seems to be no fixed ‘canonical’ text of the Sunan.
c] A prominent publisher of Hadith in Saudi Arabia and the West has stated that the following regarding the Sunan:
In verification of the Arabic text, we used the edition published along with the commentary ‘Awn al-Ma ‘bud as the main source, while comparing it to that of Al-Khattabi in Ma’alam As-Sunan, as well as a number of other valuable printed editions of Sunan Abü Dãwüd.
There are some discrepancies of variation in some of the manuscripts and reported versions, as well as published editions. Sometimes there is an additional word here or there, or one hadith or chapter is cited earlier or later in sequence.
In cases of additional words or phrases found in one or few of the manuscripts and editions, the addition has been marked by square brackets []. This method is visible in the English translation as well, and whenever it was deemed necessary to insert an explanatory term, then parenthesis ( ) were used for that purpose.
Note the variants in the manuscripts. ‘Awn al-Ma ‘bud was a commentary on the Sunan by Shams-ul-haq Azimabadi, who died in 1911 A.D. The authors state: ‘The most famously cited commentary today, is that of Al-’Allämah Abu Tayyib Muhammad Shams Al-Haqq Al-’Azimãbãdi, entitled; ‘Awn Al-Ma ‘bud. This work contains comments taken from his larger collection, entitled: Ghãyat Al-Maqsud, some of the larger collection is published.’ The reference to Al-Khattabi in Ma’alam As-Sunan is explained on an earlier page:
The most famous of commentaries on the Sunan of Abu Däwud is that of Al-Khattabi. He is Abu Sulaiman Hamd bin Muhammad bin Ibrahim Al-Khattäb Al-Khattäbi Al-Busti. He heard from the previously mentioned Abu Sa’eed Ibn Al-A’rãbi in Makkah, and Abu Bakr Ibn Dasah in Al-Basrah, as well as other scholars. He died in the year 388 after Hijrah.
His commentary is on an abridged selection of chapters and narrations of the Sunan, and it is said that his commentary is the first commentary on a Hadith book, hence its great rank and importance in the field of Hadith commentary. The name of his commentary is Ma’alam As-Sunan.
Note that Al-Khattabi died in 388 A.H. – i.e. the tenth century A.D.! Note also that his commentary is both selective and seems to be based on oral tradition, as indicated by a previous statement:
These narrations and statements of Al-Khattäbi are taken from his introduction to Ma’ãlam As-Sunan. The narrators in the chain for the last statement were all graded trustworthy by Al-Baghdadi in Tarikh Baghdad, and each of them are confirmed to have heard from the one he is reporting from. Abu ‘Umar Az-Zahid was called “Tha’lab’s boy,” and Tha’lab is Ahmad bin Yaya - Ash-Shaibäni of Al-Kufah - that Al-Khattäbi mentioned.
Robson makes this revealing observation:
The Sunan, like other books, was handed down to succeeding generations by chains of authority. In dealing with the transmission I have been mainly dependent on details given by Abu Bakr b. Khair (502-575) in his Fihrisa, but this is supplemented by further details to be found in a volume containing five works published in Haidarabad in 1328, in the British Museum Catalogue (No. MDLXV), and in the John Rylands Catalogue 8 (No. 130), with a reference to Ahlwardt’s Catalogue 9 (No. 1246) where an isnad is given for the first juz’ only.
Robson comments that ‘Nowadays it would seem that the version of Lu’lu’i alone is used.’ He goes on to state:
Abu ‘Ali Muhammad b. Ahmad b. ‘Amr al-Lu’lu’i al-Basri is the one whose version has survived and is printed at the present day. He is said to have been the last through whom the Sunan was transmitted, having got it in Muharram, 275, the year in which Abi Dawud died. He has two transmitters, Abu ‘Abdallah al-Husain b. Bakr. b. Muhammad b. Wazzan al-Basri, about whom I have been unable to find any information, and Abui ‘Umar al-Qasim b. Ja’far b. ‘Abd al-Wahid al-Hashimi (322-413), who held the office of qadi in Basra and had a reputation as a reliable authority on Tradition.
d] In regard to manuscripts, Robson mentions ‘H. 1, p. 7; H. 2, p. 16; H. 3, pp. 6, 37; and the MSS. Museum and the John Rylands Library have identical chains to Abu Hafs ‘Umar b. Muhammad b. Ma’mar b. Tabarzadh (516-607). Abu Hafs got the Sunan from the shaikhs Ibrahim Mansir al-Karkhl (d. 537 or 538) and Abul Fath Muflih Abul Fath b. Ahmad b. Muhammad al-Daumi.’ This is what Mingana states about Rylands 130 manuscript: ‘The writer of the present work… who died in 275/888, had compiled his traditions in a way perhaps lacking in method’ and then refers to someone else ‘two centuries later… usually called, who died in 463/1070, digested and edited the work in an elaborate arrangement of parts or sections. This manuscript is his edition…’
Hence, this manuscript dates from the eleventh century A.D. A later mss. – thirteenth century -is MS. Marsh 292, dated 604 A.H./1207 A.D., and kept in the Bodleian Library in Oxford University. Schoenberg (119193) lists a copy from 950, as part of the Chester Beatty library. Arberry lists it (3123) as: ‘A fragment of the second book of a celebrated collection of Traditions’, but describes it as ‘Undated, 4/10th century’ so perhaps this is more of an estimate.
The Muṣannaf of Ibn Abū Shaybah
Abu Bakr ‘Abd Allah B. Muhammad B. Ibrahim (= Abu Shayba) B. ‘Uthman Al-’Absi Al-Kufi 159-235/775-849), was an Iraqi traditionist. Note that he was born about one hundred and fifty years after Muhammad’s purported death, dying over two hundred years after the event. His work can scarcely be called early. Notably, he opposed Abu Hanifa:
An interesting point to note is that in his life, Shaykh Ibn Abi Shayba showed enmity (opposition) to the Great Imam Abu Hanifa (RA) as he (Ibn Abi Shayba) named one of the longest Chapters of his book, Al-Musannaf the title “Book of the Refutation of Abu Hanifa” in which he proceeded to list about 125 Hadiths of the Prophet (PBUH) which Imam Abu Hanifa apparently contradicted.
The ‘main sources’ listed for this biography are ‘al-Dhahabi, Siyar 9:394-396 #1841 and Tadhkira al-Huffaz 2:423’. Muhammad ibn Ahmad ibn `Uthman ibn Qayyim `Abu `Abd Allah Shams ad-Din al-Dhahabi died in 1348, and the latter is dated to thirteenth century. Scott informs us about the Musannaf:
Abū Bakr Ibn Abī Shayba was an exemplary member of the “Companions of ḥadīth”. His primary scholarly activity consisted of collecting tens of thousands of Companion and Successor opinions, complete with isnāds, along with a substantial corpus of prophetic ḥadīth. Most of his major teachers were prominent Iraqi ḥadīth scholars whom Ibn Qutayba (d. 276/889) classified as “Companions of ḥadīth” in his Kitāb al-maʿārif. Ibn Abī Shayba wrote a brief refutation of approximately 120 opinions attributed to the central “Companion of raʾy”, Abū Ḥanīfa (d. 150/767) that is included in his Muṣannafs. In 234/848-9, after the infamous ‘Abbāsid inquisition (miḥna) over the nature of the Qurʾān, Ibn Abī Shayba heeded the invitation of the Caliph al-Mutawakkil (r. 232-47/847-61) to promote anti-Mutazili ḥadīths in the Mosque of Ruṣāfa, a quarter of Baghdad on the eastern bank of the Tigris. Ibn Abī Shayba’s Muṣannaf was compiled and preserved by the Cordovan ḥadīth-champion Baqī b. Makhlad (d. 276/889), an iconoclastic scholar who refused to conform to the teachings of any single jurist and exercised ijtihād on the basis of his trove of transmitted materials.
Note the polemical nature of the work – against Hanifa and the Mutazilites, and its late compilation by someone in Spain, rather than Iraq. Therefore, how reliable is it? It was never considered as being one of the authoritative collections. Scott also notes that the Musannaf is rather limited: ‘My empirical analysis of the legal chapters of Ibn Abī Shayba’s Muṣannaf reveals that only one in eleven reports is a prophetic ḥadīth.’ Scott further suggested that ‘the Alexandrian trader, ʿAbd al-Raḥmān b. Makkī (d. 599/1202-3), who appears in Ibn Ḥajār’s isnād of the Muṣannaf, acquired it on one of his trips to al-Andalus and brought it to Egypt…’ This puts its most ancient extant edition at the thirteenth century A.D.
The Musnad of Ibn Ḥanbal
Hanbal was born in Baghdad in 164 A.H. (780 A.D.), and died in 241 (855). He is famous for opposing the idea that the Qur’an was created. Brown dates the musnad era to the late 8th and early 9th centuries. He writes:
The most famous musnad is that of Ibn Hanbal, which consists of about 27,700 hadiths (anywhere from one fourth to one third of which are repetitions of hadiths via different narrations) and was actually assembled into final form by the scholar’s son. Ibn Hanbal claimed he had sifted the contents of his Musnad from over 750,000 hadiths and intended it to be a reference for students of Islamic law. Although he acknowledged that the book contained unreliable hadiths, he supposedly claimed that all its hadiths were admissible in discussions about the Prophet’s Sunna – if it was not in his Musnad, he claimed, it could not be a proof in law.
In the light of his attitude to unsound narrations, we can understand why this musnad is not one of the authoritative collections. Schoenberg 5 (9495) lists a copy of the Musnad as dating to 1236. This is from the Khuda Bakhsh Oriental Public Library, Patna, Bihar. It is a partial copy.
Ṣaḥīḥ Muslim
a] The Sahih of Muslim b. al-Hajjaj (d. 261/875) is usually considered the second most important hadith collection after Bukhari. We read of Muslim b. al-Hajjaj: ‘Al-Imâm, Al-Hâfiz, Al-Hujjah ‘Abul-Hussain Muslim bin Al-Hajjâj bin Ward bin Koshâh Al-Qushayrî An-Nîshâpûrî was born in 202 or 204 or 206 AH in Nîshâpûr and educated in the same town. He heard Ahâdlth the first time, at age eighteen, from Yahya bin Yahya Tamîmî.’ Interestingly, there are regional divergences about the relative position of Bukhari and Muslims: ‘Imâm Abu ‘Ali bin Husain Nîshâpûrî and some scholars of the Maghrib (the Muslim countries of North Africa) give precedence to Sahih Muslim but the majority of scholars insist that it is Sahih Bukhâri which should be given precedence.’
Others explained the difference in attitude to the two Sahih works as reflecting divergent standards of accuracy:
Al-Nawawi remarks that Muslim’s Sahih has an advantage over Al-Bukhari’s because he gives all the lines of transmission when he mentions a tradition, whereas Al-Bukhari repeats traditions in different places, sometimes giving one line of transmission and sometimes another. Al-Dhahabi quotes Ibn ‘Uqda to the effect that Al-Bukhari sometimes makes the mistake of mentioning a man on one occasion by his name and on another by his kunya and imagining that there are two men, whereas Muslim rarely makes an error.
It is an interesting admission that Bukhari is not inerrant, and one wonders if this was one reason for the emergence of Sahih Muslim - as a corrective?
b] Guillaume notes both the exalted status of this work and the reason for its position:
…Muslim and his work… has always been bracketed with the Sahih of Bukhari, and they are cited as The Two Sahihs (Sahihan). But inasmuch as the ground of the authority of the Sahihan was their acceptance by the general consent of the Islamic community, and they had not been subjected to any systematic critical examination, some dissentient voices have been raised against them from the earliest times down to the present day. Like the customs they sought to authorize by appeal to apostolic custom and precept, they owe their position to ijma’, not to their inherent virtue and faultlessness.
Hence, it is not so much the intrinsic worth of the Sahihan which commends them, but rather the general reverence they enjoy in Muslim society. Calder also questions the dating and thus reliability of such works, given redactional activity:
The acknowledgement of organic texts, pseudepigraphy, and long-term redactional activity as features of some third-century material must affect assessment even of material which has not (yet) been shown to be organic. A case of much relevance to the arguments put forward in this book is that of the great collections of hadith. Apparently the product of the devoted and orderly activity of a single person, works like the Sahihs of Bukhari and Muslim should probably be recognized as emerging into final form at least one generation later than the dates recorded for the deaths of the putative authors.
c] Motzki critiques such a view, yet his analysis still suggests an element of redaction and variant copies, by the very fact that he refers to a canonical ‘vulgate’ edition:
Based on his analysis of a partial fifth/eleventh-century manuscript of Sahih al-Bukhari, Alphonse Mingana concluded that the text was still in a relatively fluid form at that point in time. Yet there is little available evidence suggesting that, beyond the normal permutations of manuscript transmission for texts as large and detailed as the Sahihayn, either al-Bukhari’s or Muslim’s books were altered substantially after their deaths.
The Sahihayn are two massive works, and the vagaries of manuscript transmission introduced the possibility of frequent variation even for a text transmitted intact from its author. Several generations of editors, such as Abu Dharr al-Harawi (d. 430/1038), al-Saghani (d. 650/1252) and the Egyptian Hanbali al-Yunini (d. 658/1260), thus played important roles in collating different transmissions of Sahih al-Bukhari into vulgate editions. Such editorial review, however, was endemic to the pre-print world and does not reflect any instability specific to the Sahihayn.
Note the reference to ‘normal permutations of manuscript transmission’ and ‘different transmissions’ and ‘vulgate editions’. Even if we were to accept that the differences between the manuscripts were minimal, the boast of some agents of Muslim apologetics of the absolute reliability of the Sahih hadith are clearly questionable, to say the least.
d] In terms of transmission, Robson notes that one of the Sahih’s most famous commentators is ‘Muhyi al-Din Abu Zakariya’ Yahya b. Sharaf al-Nawawi (631-676/1233-1277) was a very famous jurist and traditionist whose work is well known.’ However, Robson also notes that al-Nawawi’s commentary displays either a broken isnad or little regard for it:
He confines himself to his own line of transmission, which he declares to be the authoritative one in his neighbourhood, and he mentions the line through Al-Qalanisi, but does not pursue it beyond the beginning of the fifth century. He quotes Ibn al-Salah (577-643/ 1181-1245)2 as indicating that by his time and for a long period before it, the connected transmission of the Sahih is not important. It is enough for practical purposes to have a written copy which has been guaranteed by two authorities who have compared it with numerous others.
The other two sources of information include Abu Bakr b. Khair, Fihrisa:
Abu Bakr Muhammad b. Khair b. ‘Umar b. Khalifa (502-575/1109-1179) belonged to Seville. There is no record of his travelling abroad, but he met many authorities on Tradition in Spain and corresponded with others. He learned traditions either by word of mouth or by correspondence from over a hundred shaikhs, a list of whose names he compiled.
Note both the late dating and that he learned traditions by other than direct contact with the documents. The other figure was Muhammad b. Ibrahim al-Shalahi, Kitab al- imta’, [al-imta’], wal-intifa’. Robson’s data leaves us with a great deal of uncertainty:
The identity of the author of Kitab al- imta’ wal-intifa ‘is unknown. The form Al-Shalahi is doubtful, and it is not clear whether the name which appears at the end of the MS. With the date 701/1302 is that of the author or of the copyist… The author derives most of his information regarding the transmission of Muslim’s Sahih from Abul Qasim al-’Azafi who got it from his father Abul ‘Abbas al-’Azafi (557-633/ 1162-1236). I have found no information about Abul Qasim. His father is recognized as a traditionist, being called a musnid by Al-Tinbukti. He heard traditions from a number of authorities among whom were Abu. Muhammad ‘Abdallah b. Muhammad b. ‘Ali b. ‘Ubaidallah al-Hajri al-Mari (505-591/1112-1195), and Abu Bakr b. Khair; but Ibn Khair’s name does not appear in any of the chains in the MS.
Note the lack of information on Abul Qasim, and the appropriation of hadith data. A further observation must be on the late dating – twelfth and thirteenth centuries.
e] The oldest source that the Encyclopaedia of Hadith uses for its edition of Sahih Muslim is ‘The Tahrir edition of 1384, which is baeds [sic] on the Sultaniyya edition of 1329.’ In the UK, MS. Marsh 648 stored at Bodleian Library, Oxford University, contains Sahih Muslim, and has a commentary by ‘Sibṭ ibn al-ʻAjamī, 1352-1438 AD’, which indicates that the manuscript is no earlier than the fifteenth century A.D. Another mss. containing the Sahih is Islamic Arabic 1070, stored at Cadbury Research Library, University of Birmingham, and dated probably to the fifteenth century A.D. Mss. Or. 1714, containing the Sahih, is dated to 1309 A.D., and is stored at University Library, Cambridge University. Schoenberg lists a copy (36526) dated to 1050, which had belonged to Bertram, 4th Earl of Ashburnham. Another copy also belonging to the Earl (185703), entitled Musnad as-Sahih is given the same date.
Jami’ at-Tirmidhi
a] This work the jami’ (compilation) sometimes known as the Sunan of Tirmidhi:
After Sunan Abü Dawud, Imam AbU ‘Elsa At-Tirmidhi’s Jãmi’ (compilation) – better known as “Sunan At-Tirmidhi” is considered the most authentic among the Four Sunan. In fact, some scholars even considered Sunan At-Tirmidhi to be the best out of all of the Six Books, not based upon a criteria of authenticity, but rather because of how well organized it is, making it easy for the average person to find what he is looking for, and all of the additional areas of knowledge that the author has included, which are not found in the other titles among the Six.
The publishers of the work just referenced note variations in the text:
The original text of Sunan At-Tirmidhi has been widely published in the Islamic world, and a team of scholars has reviewed the famous publications and manuscripts in verification of the text for our publication. Finally, there were three main texts relied upon for verification, and these are the text published in India, with the commentary Tusifat Al-Ahwadhi by Shaikh ‘Abdur-Rahmãn Al-Mubarakpuri; the text published in Beirut, with the commentary ‘Aridat A1-Ahwadhi by Al-Oath Ibn Al-’Arabi; and the text published in Tunisia which is based upon the text verified by Shaikh Ahmad Shãkir and Muhammad Fuw’ãd ‘Abdul-Baqi.
There are slight discrepancies of variation in some of the manuscripts and published editions. Sometimes there is an additional word here or there, or one Hadith or chapter is cited earlier or later in sequence in one manuscript. In cases of additional words or phrases found in one or few of the manuscripts and editions, the addition has been marked by square brackets [ ]. This method is visible in the English translation as well, and whenever it was deemed necessary to insert an explanatory term, then parenthesis ( ) were used for that purpose.
b] Tirmidhi is alleged to have had a powerful memory, as the following story suggests:
An interesting story is told to illustrate his power of committing traditions to memory. On the way to Mecca he met a shaikh from whose traditions he had copied out two juz’. Thinking he had these notes with him, he began to question the shaikh about his traditions, but he was surprised to find that, instead of his notes, he had brought some blank sheets of paper. He continued his questions with these sheets in his hand, and after a time the shaikh noticed that they were blank and rebuked him, whereupon Tirmidhi assured him that he knew his traditions by heart. The shaikh was unconvinced of his genuineness even when he recited his traditions to him, so Tirmidhi asked him to recite some others. The shaikh recited 40 traditions and Tirmidhi repeated them without making a single mistake, thus showing his remarkable powers of committing traditions to memory.
Whilst we should not dismiss this story out offhand, we wonder if the technique of demythologization should not be applied to it. This is accentuated by a claim the same author (Dhahabi) makes about Tirmidhi going or even being born blind: ‘He is commonly referred to as al-darir (blind), his blindness usually being attributed to his weeping over Bukhari’s death. Dhahabi remarks that he remained blind for two years; but the suggestion has been made that he was born blind, a suggestion difficult to accept.’
c] Another interesting point is that: ‘The Jami’ has deservedly been considered important, although Muslims were slow to give it general recognition.’ Given its standing now, the lack of general reception requires some explanation. This is also linked to questions about its transmission which Robson highlights:
It is not clear whether all those who are mentioned transmitted the whole of the Jami’, as it is merely said that they transmitted from Tirmidhi without anything being said of how much they transmitted. While it is possible that most or all of them were transmitters of the Jami’, there are only two of those mentioned above of whose transmissions of this work I have found records. In addition, Abu Bakr b. Khair (502-575) gives in his Fihrisa one chain each from Abu Dharr Muhammad b. Ibrahim b. Muhammad al-Tirmidhi and Abu Muhammad al-Hasan b. Ibrahim al-Qattan, transmitters of whom I have found no mention elsewhere.
d] The Encyclopaedia of Hadith lists as its sources the following works for Tirmidhi:
The edition by Ahmad Shakir and others. (Cairo, 1356/1937)
Aridat al-Ahwadhi, the commentary by Abu Bakr Ibn al-Arabi. (Cairo, 1350/1931)
Tuhfat al-Ahwadhi, the commentary by al-Mubarakfuri. (Beirut, 1410/1990)
Manuscript No. 648 Hadith preserved at the Egyptian National Library, dated 726.
Note again how recent are the dates for such works. The Egyptian manuscript, dated ‘726’, obviously refers to its A.H. dating, corresponding to c. 1325 A.D. There is thus a gap of 693 years (taking the death of Muhammad at 632) between the manuscript and the events it describes. Hence, the oldest manuscript on which the Encyclopaedia of Hadith edition on Tirmidhi is based goes back to the fourteenth century A.D. To give some idea of historical (UK) parallel, it is as if the first extant history of the Declaration of Arbroath on Scots sovereignty in 1320 was only issued in 2013! The Maknaz Edition uses a mss. dated to 626/1229 (copied by Mustafa b. al-Hajji Qutilmish).
Sunan An-Nasai
a] The author of this collection is ‘Abu ‘Abdur-Rahman bin Shu’aib ‘Ali bin Sinan bin Bahr An-Nasai. The name “An-Nasai” is an ascription to Nasa’ of Khurasan.’ He was born c. 214/215 A.H., i.e. c. 829 A.D., and died 303/915. Some hold that he studied under Bukhari. Others dispute this: ‘There is some doubt as to whether al-Nasai studied with al-Bukhari: al-Nawawi affirms this while al-Dhahabi says that al-Nasai never transmitted from al- Bukhari…’ Wensinck informs us that other aspects of his life are uncertain:
Very little is known about him. He is said to have made extensive travels in order to hear traditions, to have settled in Egypt, afterwards in Damascus, and to have died in consequence of ill-treatment to which he was exposed at Damascus or, according to others, at Ramla, in consequence of his feelings in favour of ‘Ali and against the Umayyads. On account of this unnatural death he is called a martyr. His tomb is at Mecca.
b] There are some questions about authorship:
…Imam An-Nasai compiled his Sunan Al-Kubra first, then sometime later, smaller Sunan, which later was referred to as Al-Mujtaba and is also called Mujtana. Both the books have a same meaning: "the selected," and it is not clear who first referred to the smaller Sunan with either of these names. In fact, some of them also called it the Sahih.
Scholars differ over whether Imam An-Nasai himself compiled Sunan As-Sughra or Al-Mujtaba - or it was a compilation of his student Ibn As-Sunni. The fact that since the smaller Sunan or Al-Mujtaba is generally known to be reported from An-Nasai by Ibn As-Sunni, it has led some to believe that it is, in reality, the work of Ibn As-Sunni. Imam Ad-Dhahabi (Siyar A’lam An-Nubala) and whoever followed him held this view, saying that we only know of his Sunan through the narration of Ibn As-Sunni.
This raises further issues about historical reliability.
c] There are also problems in the history of transmission, in terms of records and identity:
Although Ibn Hajar says that all the men in his list transmitted the Sunan, he himself, as will be seen later, does not claim to have received transmission of the work through more than four of them. Indeed, records do not seem to have been preserved of the transmission from half of the men who have been mentioned. Another matter to be noted is that when records are kept of the transmission, care is not always taken to make it clear whether the work in question is Al-sunan al-kubra, or Al-sunan al-sugbra. This causes a certain amount of indefiniteness in the account of the transmission.
There are also issues of either historical credibility or copyist errors:
There are two instances where the transmission is quite impossible. Ibn Khair says that Abul Hasan Yunus (447-532) transmitted the Sunan from four men, one of whom was Abu ‘Abdallah Ziyad (347-430). The only suggestions which can be made are that either Ibn Bashkuwal has made an error in the dates he gives, or that an error has crept into Ibn Khair’s text. The other matter is much more serious, as it is repeated frequently. Ibn al-Sunni’s text is transmitted through Duni (d. 501) from Kassar who died in 385 according to Subki, or in 380 according to Ibn al-’Imad. He would need to have been very long-lived for this to be possible. Yet Ibn Hajar and H. 1-5 give this transmission without any hesitation. The explanation may be that Ibn Hajar, or a copyist, has inadvertently omitted a link and that the others have followed this uncritically.
There is a mss. of ‘Sunan an-Nasa’i dated 525 A.H. (1123 A.D.) and preserved in Maktaba al-Aqsa in Jerusalem’. Hence, this manuscript dates to the twelfth century.
Sunan Ibn Majah
a] This is the final part of the canonical six major collections. The author is described as follows:
Abu ‘Abdullah Muhammad bin Yazid bin ‘Abdullah Rab’i Al-Qazvini nicknamed Ibn Majah. He was a non-Arab… Various explanations have been given for his nickname. Allamah Zubaidi, writing in Tajul-’Urus, has given several explanations for the nickname, one explanation being that Majah was his mother’s name. Imam Nawawi gives weight to this explanation. Shah ‘Abdul-’Aziz Dehlavi in Bostanul-Muhadithin says: (The correct opinion is that Majah was his mother.) That is why the Arabic word for son (Ibn) is written with the Arabic letter alif to indicate that Ibn Majah qualifies Muhammad, not ‘Abdullah. Anyhow, some scholars believe that Majah was his father’s name. That is also the opinion of Hafiz Ibn Hajar.
As to his birth and death, we learn this:
He was born in 209 AH corresponding to 824 CE. Yaqut bin ‘Abdullah Al-Hamavi, quoting Ja’far bin Idris’ Tarikh Qazvin, wrote: Abu ‘Abdullah died in the year 273 AH and 1 heard him say “I was born in 209 AH.” …The Imam died on Monday, 22 Ramadan 273 AH corresponding to 887 CE, aged 64…. Some scholars said that Ibn Maijah died in the year 275 A H.
He was also a commentator and historian, but none of these works are extant: ‘While he is remembered for his compilation of traditions he had wider interests and is said to have written a commentary on the Qur’an and a history of Qazwin, but only his Smart seems to have survived.’
b] Significantly, Sunan Ibn Majah seemed to take a long time to be counted among the Six Collections: ‘Sunan Ibn Majah began to be included in the Six Books by the end of the fifth century AH.’ Robson also notes that the value of the Sunan was questioned:
Different views have been expressed about the value of the Sunan. Ibn Majah is said to have submitted his work to Abu Zur’a (d. 264) who, after looking into it, remarked that if this work fell into people’s hands all the jami’s, or most of them, would become worthless, adding that there were not as many as thirty traditions whose isnad contained a weakness. Dahabi displays modified rapture in his estimate of the value of the Sunan, saying it would have been a fine book had it not been spoiled to some extent by weak traditions, yet he quotes Abu Ya’la al-Halili (d. 446) as saying that Ibn Majah was a great authority, as everyone whose authority was quoted agreed. Ibn Hajar says his book is a good collection with many bābs and remarkable features, but contains some very weak traditions, and quotes al-Sari to the effect that when Ibn Majah gives a unique tradition it is generally weak, and that there are many munkar traditions.
Perhaps this was why canonical acceptance of the Sunan was delayed, as Goldhizer indicates:
…the general recognition of the ‘six books’ had not yet prevailed in the first half of the fourth century…. there was a tendency even at that time to extend the circle of canonical collections of traditions beyond the two Sahihs, but it did not yet include all the ‘six books’. Secondly, that at the end of the fifth or beginning of the sixth century al-Tirmidh, and Ibn Maja were already included in this group. Ibn Hazm (d. 456) still had some doubts about al-Tirmidhi. Doubts were maintained longest about Ibn Maja because of the many weak (da’if) traditions which he incorporated into his corpus traditionum.
Indeed, doubts about Ibn Majah continued for some time, up about the twelfth or thirteenth centuries A.D.:
…there are signs that doubts about Ibn Maja remained alive for yet another century. The Spanish scholar Razin b. Mu’awiya from Saragossa who lived in Mecca (d. 535) wrote a compendium of the six sahih books but Ibn Maja was not used as a source for his work; the author used the Muwatta’, in addition to the five books. Also Muhammad ‘Abd al-Haqq al-Azdi, called Ibn al-Kharrat, from Seville (d. 581) allotted no place to Ibn Maja among the sources of his compilation al-Ahkam al-Kubra, which he based on the recognized canonical collections. Muhammad b. Ab, Muhammad Uthman al-Hazimi from Hamadan (d. 584) only knows al-a’imma al-khamsa. The attempt to gain a place amongst the canonical authorities for Ibn Maja had already been made at that time; it was instigated by Abu’-Fadl Muhammad b. Tahir al-Maqdisi (d. 507) but met with only partial success.
Robson, who was not normally too sceptical judging by his other writings about the Hadith, expressed problems with this Sunan:
But when one examines the chain it is impossible to avoid misgivings. Qattan, whose date of birth does not seem to be recorded, died seventy-two years after Ibn Majah’s death. He must either have died a very old man, or have received the Swum from Ibn Majah at a very early age. Abu Talha died sixty-four or sixty-five years after Qattan’s death, and so he might have been at least in his teens when he received the Swum. Muqawwimi, however, was only about nine or ten years old when Abu Talha died, and Abu Zur’a was only three when Muqawwimi died. This precocious child is first said to have received the Sunan from Muqawwimi by ijaza if not by hearing, a procedure which is quite normal, as ijaza was given to young children, but Ibn Hajar says it later became clear that he had heard the whole of it from Muqawwimi. How a child of three or less could do this it is difficult to understand. Muqawwimi is said by Ibn Hajar to have used ahbarana in telling how he received the Swum from Abu Talha, although he was no more than ten when Abu Talha died.
After some more disquieting examples, he concludes: ‘The general impression received from such considerations is that there are either some weak links in the transmission, or that people of later times have been careless in their recording and have omitted some links.’ Interestingly, brown suggests the Sunan became canonical for reasons of expediency, rather than accuracy of record:
Examining the canonical collections of Sunni ḥadîth, however, we find that authenticity was not a consistent priority. The canonization of Ibn Mâjah’s Sunan illustrates that the ḥadîth canon was formed in part for reasons other than textual authenticity as defined by Sunni ḥadîth criticism. Although advocates of Ibn Mâjah’s Sunan lauded its author for his selectivity and critical rigor, luminaries of the Sunni ḥadîth tradition across the centuries have lambasted the book for the unreliability of its contents. According to the testimony of influential participants in the Sunni study of ḥadîth, the book was admitted into the canon not because of its reliability but because it vastly expanded the number of useful ḥadîths in the canonical body.
Brown suggests that it was only in the thirteenth century A.D. that the Sunan gained increasing acceptance in the canon.
c] The manuscript tradition of the Sunan is illuminating as to its dating:
There is a very old and authorized manuscript of Sunan Ibn Majah at dates back to year 601 AH. The manuscripts had writings of a number of scholars including Ibn Qudamah and Imam Mizzi, in this manuscript, Ibn Majah’s work is called As-Sunan li Ibn Majah. This is the correct title of the work…
All this is in addition to the fact that the author did not title his work Sunan Ibn Majah. Therefore, books and opinions of their authors in giving titles to their books should be respected. It follows then that, Ibn Majah’s book should be called ‘As-Sunan by Ibn Majah’, especially when printing the work.
It is, however, strange that though this book has been published several times, the correct title did not appear in the cover of any of these editions. What is more astounding is the fact that one of the editors of the book, Dr. Muhammad Mustafa A’zami adopted the authorized manuscript in which this title appeared. Yet, he disregarded this title.
What is meant by ‘authorized manuscript’ is not explained. The date given – 601 A.H. – corresponds to 1204 A.D. (the year of the Latin conquest of Constantinople). Among the sources for the Encyclopedia of Hadith is ‘Manuscript Taymur Pasha No. 522 Hadith, preserved at the Egyptian National Library, copied in 561 AH.’ This corresponds to 1122 A.D. Another author claims that ‘the oldest known manuscript copy of Sunan ibn Majah. It is dated 485 AH and now located in the Suleymaniyye manuscript library in Istanbul, Turkey.’ This corresponds to 1092, so the oldest mss. for Ibn Majah dates from the eleventh century A.D.
SUMMARY
We shall begin with the Six Books:
Sahih Bukhari - Mingana Arab. Isl. 225 – c. 1000, so tenth-eleventh centuries.
Extant editions of Bukhari are based on al-Yunini (d. 701/1302, so (fourteenth century AD), possibly by way of al-Qastallani (fifteenth-sixteenth centuries AD).
Asqalani’s Fatḥ al-Bārī fī Sharḥ Ṣaḥīḥ al-Bukhārī (fifteenth century AD), and Umdat al Qari by Badr al Ayni (also fifteenth century AD).
The Maknaz edition uses a text from 873/1468 - also fifteenth century AD.
Sahih Muslim - Tahrir edition of 1384, based on the Sultaniyya edition of 1329.
MS. Marsh 648 stored at Bodleian Library, Oxford University, contains Sahih Muslim, and has a commentary by ‘Sibṭ ibn al-ʻAjamī, 1352-1438 AD’, which indicates that the manuscript is no earlier than the fifteenth century A.D.
Islamic Arabic 1070, stored at Cadbury Research Library, University of Birmingham, and dated probably to the fifteenth century A.D.
Mss. Or. 1714, containing the Sahih, is dated to 1309 A.D., and is stored at University Library, Cambridge University.
Schoenberg lists a copy (36526) dated to 1050, which had belonged to Bertram, 4th Earl of Ashburnham.
Another copy also belonging to the Earl (185703), entitled Musnad as-Sahih is given the same date. So the oldest copy is eleventh century A.D.
Jami’ At-Tirmidhi - Manuscript No. 648 Hadith preserved at the Egyptian National Library, dated 726/1325 - fourteenth century A.D.
Maknaz Edition uses a mss. dated to 626/1229 (copied by Mustafa b. al-Hajji Qutilmish) - thirteenth century A.D.
Abu Dawud - Rylands 130 manuscript, dates from 1070 A.D. - the eleventh century.
MS. Marsh 292, dated 604 A.H./1207 A.D., and kept in the Bodleian Library in Oxford University - thirteenth century.
Schoenberg (119193) lists a copy from 950, as part of the Chester Beatty library. Arberry describes it as ‘Undated, 4/10th century’ so perhaps this is more of an estimate.
An-Nasai - mss. of ‘Sunan an-Nasa’i dated 525 A.H. (1123 A.D.) and preserved in Maktaba al-Aqsa in Jerusalem’ - twelfth century A.D.
Ibn Majah - ‘authorized’ mss. dated 601/1204 - thirteenth century A.D.
Manuscript Taymur Pasha No. 522 Hadith, preserved at the Egyptian National Library, copied in 561/1122 - twelfth century A.D.
Another mss. in the Suleymaniyye manuscript library in Istanbul, dated 485/1092, so the oldest mss. for Ibn Majah dates from the eleventh century A.D.
The other books:
Muwatta - Dublin’s Chester Beatty Library, Mss. 3001, dated to 277/890. PERF No. 731 in the Austrian National Library, Vienna, dated second half of 2nd century A.H. by Nabia Abbott, but more recently dated to first half of the third/ninth century by Professor Petra Sijpesteijn.
Taylisi - Al-Musnad, dated to 1250 A.D., from Khuda Bakhsh Oriental Public Library, Patna, Bihar.
Abi Shaybah - thirteenth century
Hanbal – 1236 - thirteenth century, from Khuda Bakhsh Oriental Public Library, Patna, Bihar, a partial copy.
Musanaf of Abdul Razzaq – The basic manuscript Murad Mulla (Istanbul) dates from the year 747I1346-7.
The manuscript Fayed Allah Effendi (Istanbul), from the year 606/1209-10. The dates here are fourteenth and thirteenth centuries.
Sahifa Hamman B. Munabbih – twelfth century.
CONCLUSION
It is amazing that so many articles and academic books on the Hadith, whether by Muslims or not, for whatever reason fail to address the question of what manuscripts are extant. Usually, studies of Biblical or Classical material show great concern on equivalent matter, and we need only consider the title as well as the theme of the book by the renowned Biblical scholar Bruce Metzger called The Text of the New Testament. In many cases, even non-Muslim scholars seem to content to be guided by the Islamic concept of isnad, rather than the normal academic approach of examining the date of manuscripts and working back from that starting point. The issue becomes particularly important when we consider the question of Islamic Origins. Apart from non-canonical seerah material, the main Muslim source – and certainly the canonical one – is the Hadith. The canonical Sunni Six Collections only emerged two centuries after the events they relate, but the gap is even larger when we take into account manuscript age. The oldest manuscript for Bukhari is tenth-eleventh centuries; for Muslim, eleventh century; for Tirmidhi, thirteenth century; for Abu Dawud, possibly tenth, or eleventh century; for An-Nasai, twelfth century; for Ibn Majah, eleventh century. Surely this makes reconstruction of Islamic Origins even more questionable, but it certainly raises questions as to why the manuscript evidence should be so sparse when Sunni Islam was empowered?