THE TRANSMISSION OF IBN SAʿD’S BIOGRAPHICAL DICTIONARY KITĀB AL-ṬABAQĀT AL-KABĪR

This article develops the concepts and tools for the systematic study of the mechanics of survival for medieval Islamic books. These concepts and tools are then applied to studying the history of the earliest extant biographical dictionary of the Islamic tradition: Ibn Saʿd’s Ṭabaqāt. First, the book’s transmitters and their historical contexts are investigated using a large number of transmission chains. Then, conclusions are extracted from this data concerning the book’s authorship, the survival process of its many versions, and the trajectories of its geographical diffusion at different phases of its long life.


About the Kitāb al-Ṭabaqāt al-kabīr
The Kitāb al-Ṭabaqāt al-kabīr (literally, 'The Great Book of Strata', henceforth KTK) was compiled by the Baghdadi ḥadīṯ transmitter and historian Muḥammad Ibn Saʿd (d.230/845). 1The book belongs to the Islamic genre of biographical dictionaries of ḥadīṯ transmitters (tarāǧim).Within that tradition, it belongs to a specific sub-genre made up of lists of biographies of ḥadīṯ transmitters (muḥaddiṯīn) organized by generation.Such works are usually called 'books of strata' or kutub alṭabaqāt.Ibn Saʿd's KTK stands out among its contemporaries in this genre, and even among historically minded compositions of the late second and early third Islamic centuries because the latter are basically lists of names, short lineages, dates of birth and/or death, 2 whereas the KTK has full biographies organized according to a number of criteria. 31 This article arises from research undertaken for my unpublished doctoral dissertation A History of Ibn Saʿd's Biographical Dictionary Kitāb al-Ṭabaqāt al-Kabīr (Santa Barbara: University of California at Santa Barbara, 2009).For the most up-to-date biography of Muḥammad Ibn Saʿd, see A.N. Atassi, History, 34-95. 2 Surviving examples of such compositions in the ṭabaqāt of muḥaddiṯīn genre are: Ḫalīfa b.Ḫayyaṭ al-ʿUṣfurī's (d.240/850) Ṭabaqāt, and Ibn Saʿd's Kitāb al-Ṭabaqāt al-Ṣaġīr (still in manuscript).
3 For a good description of the KTK see J. W. Fück 'Ibn Saʿd'.The first modern edition of the book was issued in eight volumes (plus a volume of indices) in Leiden by E. Sachau.The first two volumes constitute a biography of the Being the earliest surviving biographical dictionary, and later a staple of the Sunnī tradition, it is surprising that Ibn Saʿd's KTK has not received the attention it deserves, or at least as much attention as al-Ṭabarī's Tārīḫ, for example. 4This paper will remedy some of this 'injustice' by tracing the history of survival and transmission of the KTK. 5 In the process, we will also explore what it means to study the history of a medieval Islamic book and how the notions of transmission and survival fit into that history.

Sources, data, and methodology
The aim here is to study the 'survival dynamics' of the KTK through an investigation of its communication circuit in each generation.The elements of the communication circuit (at least the ones that can be accessed from the available data) are the KTK author(s), its transmitters/teachers, its copyists/students, and its readers/users.Our first task then is to establish a pool of candidates for these roles, and assign one or more roles in the circuit to each person in that pool.For this we need to locate the KTK's extant manuscripts and extract their different chains of transmission, and to locate the later compilations that contain Saʿdī reports and extract the transmission chains of such reports.
There is no single complete manuscript of the KTK, only fragments of it, with some overlaps. 6Therefore, for these manuscripts, it is important to determine whether they represent a single recension of the work, a number of overlapping recensions, or widely different ones that cannot be, or should not be, reconciled.Fortunately, this work was done for us by the successive editors of the printed editions of the KTK. 7Next, several transmission trees of the KTK's recensions represented by the extant manuscripts are drawn. 8Studying the transmission chains of Saʿdī reports within later compilations helps add more branches to these trees. 9Using biographical information of the persons involved in the aforementioned transmission trees (manuscripts and other recensions), we can study the temporal and geographical diffusion of the KTK.
Moreover, comparing these reports to corresponding ones in the printed edition of the KTK helps to establish the existence and character of other recensions, compared to the one available to us, and to give an approximate date to their disappearance from circulation; thus describing the process of crystallization of the book. 10Counting the frequency of Saʿdī reports in different compilations helps draw a picture of the KTK's 7 Several editions appeared in the Arab world that were based on the Leiden edition; namely the editions of Dār Bayrūt, 1957; Dār Ṣādir 1960; Dār Bayrūt lil-Ṭibāʿah wa-l-Našr, 1978; Dār al-Taḥrīr, 1968.In 1983, Ziyād  For a detailed investigation of the overlapping of the extant manuscripts, see the introductions to the different volumes of the Leiden edition.In fact, we show here that the manuscripts and the Saʿdī reports in compilations written after the fifth/eleventh century come from the fusion of two recensions.
8 Due to space restrictions, these trees are not included in this paper, only a list of the major transmitters organized in generations is given.Readers interested in diagrams of these trees are referred to in Atassi, History, Appendices I and II.
9 How can we distinguish between a book-transmission chain and a reporttransmission chain?I noticed that a good number of reports in later compilations share a portion of their transmission chains with those of the extant manuscripts; i.e. the portion covering the period from Ibn Saʿd's time to the fifth/eleventh century.Therefore, when encountering a large number of such reports, I assumed that they were drawn from copies of the same recensions as those of the manuscripts.For example, we can confirm this assumption for Ibn Ḥaǧar al-ʿAsqalānī's works because he tells us the sources of his copies of the KTK in his in al-Muʿǧam al-mufahras, 1: 168-70.
literary diffusion according to genre. 11This diffusion is an indication of how different generations perceived and classified the KTK.This classification, combined with the transmitters' historical context, should orient later investigations concerning the reasons behind the KTK's survival; and hence how and why its authority as a book of tradition was gradually established.Finally, comparing borrowings with extant manuscripts should give us an idea about the accuracy of book transmission within the medieval Islamic culture, which is, as we have mentioned, related to the rise of what we called the textbook.

The sample of compilations
Having combed a hundred or so medieval compilations looking for Saʿdī material, I noticed the existence of two major time periods according to the number of compilations that contained Saʿdī reports and the number of such reports within each compilation.Beyond the sixth/twelfth century, compilations containing Saʿdī reports increased dramatically and so did the number of such reports in each compilation.Therefore, for this period I only included in my study the compilations that supplied the transmission chains of their Saʿdī reports.I ignored the compilations that borrowed from Ibn Saʿd's works without specifying which one or how it was obtained.Before this date, I included all the compilations containing Saʿdī material that I could find, except when several of them belonged to the same compiler and featured similar numbers of Saʿdī reports.In the latter case, I selected a representative compilation of the compiler's work which were then grouped into six genres: tarāǧim (biographies) books 12 , sīra and maġāzī books 13 , history (or historiography) books 14 , ḥadīṯ 11 The counting was done electronically with the help of digitized versions of the books consulted and the help of al-Maktaba al-Šāmila; see Atassi, History,  208-11.  1By tarāǧim books I understand books that contain a succession of indivisible parts (tarǧama, or biography) each containing information relating to one person.In this category I include books from the ṭabaqāt genre such as Ḫalīfa b.Ḫayyāṭ's Ṭabaqāt, ansāb books such as Balāḏurī's Ansāb al-ašrāf, and biographical compilations such as al-Ḫaṭīb's Tārīḫ Baġdād.
13 By sīra and maġāzī books I understand biographies of Muḥammad (sīra), monographs about his battles (maġāzī), and books glorifying his personality traits and his acts (šamāʾil and faḍāʾil books). 14 Transmitters of the KTK According to the chains of transmission of the KTK's extant manuscripts, 17 the material contained in these manuscripts is the fusion of two recensions, the first transmitted by Abū Muḥammad al-Ḥāriṯ b.Muḥammad Ibn Abī Usāma al-Tamīmī (186/802-282/895), 18 and the 17 For details about the transmitters of the KTK inferred from Saʿdī reports that were included in later compilations see A.N. Atassi, History, 211-250 and references therein; see also Appendix II for transmitters of the extant manuscripts only.
18 He resided in Baghdad and was probably a copyist and a tutor for hire.He has a musnad compilation (ḥadīṯs organized according to selected transmitters, usually the first after Muḥammad) attributed to his name; but generally he was second transmitted by Abū ʿAlī al-Ḥusayn b.Muḥammad b. ʿAbd al-Raḥmān Ibn Fahm (211/826-289/901) who is the more problematic of the two. 19Both transmitters were second-tier muḥaddiṯs and possibly teachers by vocation.In the second generation, Abū al-Ḥasan Aḥmad b.Maʿrūf al-Ḫaššāb (d.321 or 322/933 or 934) transmitted on the authority of both Ibn Abī Usāma and Ibn Fahm.He was an obscure muḥaddiṯ from Baghdad.It is difficult to ascertain his profession from the designation al-Ḫaššāb (literally, 'the carpenter' or 'wood handler/cutter').However, it would not be far-fetched for the muḥaddiṯs of the pre-madrasa era to teach ḥadīṯ and related material as an avocation.20Also in the second generation is Abū Ayyūb Isḥāq b.Sulaymān al-Ǧallāb (d.334/945), another minor muḥaddiṯ from Baghdad, whose profession could have been a carrier given his designation al-Ǧallāb.He transmitted on the authority of Ibn Abī Usāma only. 21Al-Ǧallāb's role as a transmitter of the KTK is inferred from transmission chains of Saʿdī reports in later compilations; especially Ibn ʿAsākir's Tārīḫ madīnat Dimashq.
The third generation is even more problematic than the first two for it contains one person only; namely Abū ʿUmar Muḥammad b.Ḥayyuwayh al-Ḫazzāz (295/907-382/992).Both manuscripts and Saʿdī reports give us this one transmitter.He lived in Baghdad and, according not a major figure of the Baghdadi ḥadīṯ scene.It is noteworthy that the sīra part of the extant manuscripts is transmitted by Ibn Abī Usāma alone, the eldest of the two transmitters of the KTK.This lends credence to the claim that the sīra part of the Leiden edition of the KTK used to be circulated as a separate book.
to his designation (ḫazzāz), he might have been a maker of silk yarn.We have no complete manuscript of the KTK with only Ibn Abī Usāma or Ibn Fahm in the chain of transmission.However, all available manuscripts include Ibn Ḥayyuwayh in their transmission-chains as the only transmitter at the third level after the author. 22It is possible that Ibn Maʿrūf had collected the entire KTK before Ibn Ḥayyuwayh; but it is the latter who seems to have propagated it.Al-Baġdādī mentions that Ibn Ḥayyuwayh 'heard plenty and wrote [i.e.copied] all his life and transmitted large compilations such as the Ṭabaqāt of Ibn Saʿd, the Maġāzī of al-Wāqidī, the compilations of Abī Bakr b. al-Anbārī, the Maġāzī of Saʿīd al-Umawī, the History of Ibn Abī Ḫayṯama, and many others'. 23One of the manuscripts' transmission-chains states that Ibn Ḥayyuwayh copied the corresponding section of the KTK while the text was being recited back to Ibn Maʿrūf in the month of Šaʿbān of the year 318/930.This means that Ibn Ḥayyuwayh was then twenty years old and that Ibn Maʿrūf was at the end of his life.We notice here the same pattern we observed in the transmission of the KTK from Ibn Saʿd to Ibn Fahm; i.e. a young student tries to get the teacher's book as early as possible in his career and as late as possible in the teacher's life.This was a common practice among muḥaddiṯs because it lowered the number of transmitters between the last in a chain and the Prophet. 24We must also remark that collecting and transmitting such large works possibly needed full time dedication.It is difficult however, given the dearth of information about these transmitters, to ascertain their professions, and whether or not they practiced teaching.
In the fourth generation we encounter three transmitters of the KTK, all of whom seem to have been teachers by vocation. 22If it were not for earlier books that mentioned Ibn Saʿd and his KTK with numerous borrowings that matched the KTK verbatim, I would have suggested considering Ibn Ḥayyuwayh as the 'real' compiler of the KTK.Nonetheless, it is possible that he had an impact on the KTK in terms of selection of recensions, organization of reports, and addition of some information.For a discussion of Ibn Ḥayyuwayh's partition of the KTK in twenty four parts (aǧzāʾ), as well as other known partitions, see A.N. Atassi, History, 239-41. 23Al-Ḫaṭīb al-Baġdādī, Tārīḫ Baghdād, 3: 121, no.1139. 24Receb Şentürk, Narrative Social Structure, 1-28.
(355/965-435/1043).According to al-Ḫaṭīb's Tārīḫ Baġdād, al-Ǧawharī resided in Darb al-Zaʿfarānī, where many muḥaddiṯs used to live.Al-Ḏahabī's Siyar aʿlām al-nubalāʾ adds that 'he was steeped in transmission, he transmitted abundantly, and held many dictation sessions'. 25Al-Barmakī resided in Baghdad and was a Ḥanbalī muftī, with a teaching circle (ḥalqa) at the al-Manṣūr mosque. 26Al-Ḫaṭīb also alludes to the fact that al-Azharī taught large compilations, such as the KTK, when he says: 'we heard from him large compilations and long books'. 27n the fifth generation, we know of five transmitters of the KTK; three of them through the manuscripts and two through Saʿdī reports.All of these transmitters were from Baghdad, and most of them seem to have been teachers.For example, Abū Bakr al-Qāḍī Muḥammad b. ʿAbd al-Bāqī (442/1050-535/1140) was a scholar and a teacher. 28Abū Naṣr Muḥmmad b. al-Ḥasan (434/1042-510/1116) had two teaching circles in Baghdad, which he took over after his father, one of them being at the famous al-Manṣūr's mosque. 29To this generation belongs al-Ḫaṭīb al-Baġdādī (d.463/1071), the compiler of the famous Tārīḫ Baġdād. 30In the sixth generation, we know of five transmitters, all from Baghdad.The manuscripts give us only one, but the most renowned.He is Abū Muḥammad ʿAbd Allāh b.Duhbul b.Kāra (d.599/1202). 31To this generation belongs Ibn ʿAsākir (499/1105-571/1176), the compiler of the famous Tārīḫ madīnat Dimašq. 32The sixth generation is practically the last of the known Baghdadi generations of KTK transmitters. 3325 Al-Ḏahabī, Siyar aʿlām al-nubalāʾ, 18: 68. 26Al-Ḫaṭīb al-Baġdādī, Tārīḫ Baghdād, 6: 139, no.3180.The mosque of al-Manṣūr, which should be located close to al-Manṣūr's palace (Qaṣr al-Ḫuld), was the main mosque on the western side (i.e. the old city) of the Tigris.Important teachers of all disciplines had teaching circles in that mosque. 27 In Syria and Egypt, we know that transmitters of the KTK were also full-time teachers, while famous compilers remained mostly users of the KTK.While many Baghdadi transmitters of Ibn Saʿd's work taught in the neighborhood of al-Karḫ (south of Baghdad where the aforementioned Darb al-Zaʿfarān was located) or in al-Manṣūr's mosque in the walled city, their Syrian and Egyptian counterparts taught in institutions sponsored by the ruling elite, such as the network of madrasas patronized by the Mamlūk rulers and their amīrs.Moreover, while the Baghdadi transmitters were possibly religious scholars by avocation, their Syrian and Egyptian counterparts were professional scholars, judges, and members of the religious elite 40 Islamic 'tradition' has a well-known fragmentary nature.The prevalent way of transmitting this tradition continued to be the individual report, which consisted of a transmission chain attached to the report's text.Instead of continuous narratives, what emerged are compilations of reports which preserved the fragmentary nature of the original reports, and made possible their own re-fragmentation.Medieval Muslim compilers tended to fragment the works of their predecessors into individual reports (the same report could even be fragmented into many smaller ones to suit the needs of the user), and then include these fragments into their own works.Compiling and fragmenting knowledge were two distinct and opposing processes always active in the production and transmission of medieval Islamic knowledge.It is puzzling, but it seems that students of medieval knowledge had an aversion toward teaching books that they collected in their travels.Instead, they fragmented what they learned and wrote their own compilations which they later taught.In the current study, I suggest that people who chose to teach others' compilations tended not to write any of their own. 39It is probable that this mosque and the neighboring district of Bāb Ḥarb, at whose cemetery some of these transmitters were buried, had strong ḥanbalī affiliations.It is possible that the ḥanbalites' strong attachment to tradition may explain their interest in the KTK, given that it was one of the earliest works to deal with early Islamic history.This intellectual, and maybe social, aspect of the KTK's history still needs further investigation. 40Such information is included in the biographies of the different transmitters referenced in this paper when each of them is mentioned for the first time.See, for example, footnotes 40-4.
Attempting to explain this observation, I suggest that in medieval Muslim societies, intellectual prestige was built through the writing of compilations and legal texts, dictating them rather than teaching older compilations.Legal texts required competency, but compilations only required fragmenting older works and reassembling them.In this atmosphere of enhancing one's reputation as a scholar by absorbing and building upon the works of predecessors, the survival of older books becomes quite difficult: for that to happen, a group of dedicated transmitters, whose task is to popularize a selected group of works, has to exist.What would then make transmitting rather than compiling attractive to these teachers?This is a hard question to answer, but the transmission of entire books transformed these books into authoritative sources of tradition by virtue of a process of selection, at the heart of which were those dedicated teachers.In fact, such dedicated transmitters defined and preserved the 'canonical' books of tradition.This exact process transformed the KTK into an authoritative source of the Islamic tradition.

Methods of transmission of the KTK
It is noteworthy that by and from the ninth/late fourteenth century, the transmission of the KTK happened mostly by iǧāza.The clearest example is Ibn Ḥaǧar, who obtained five different permissions to use the KTK.It was also common for calculating parents to take their young boys (at age three or four) to hear a famous and old teacher for a while and then obtain a permission from this teacher for their son.This was the case, for example, of Ibn Sayyid al-Nās (671/1272-763/1361) who, while a child, obtained a permission from Abū l-Faraǧ al-Ḥarrānī (587/1191-672/1273). 41 The use of iǧāza in the transmission of the KTK was known since the third/tenth century, and, according to Ibn Ḥaǧar, even Ibn Ḥayyuwayh in the fourth/tenth century obtained parts of the KTK by an iǧāza from Ibn Maʿrūf al-Ḫaššāb.Tracking the use of iǧāza in the transmission-chains of Ibn Sayyid al-Nas and Ibn Ḥaǧar, we notice a steady increase in this usage as time progressed.By Ibn Ḥaǧar's time (the ninth/fifteenth century), it was possible to obtain an iǧāza by mail and without even seeing the person granting it. 42This is an indication that the KTK had acquired such stability in its form that one could acquire a copy of it and then authenticate that copy through one or multiple iǧāzas from different teachers.It was not required for the grantor of the iǧāza to have heard the entire book from a teacher either, only a status of scholarship and a reputation of trustworthiness sufficed for the chain of authentication to be valid and to carry the weight of samāʿ (hearing), the ultimate source of authenticity. 43y the ninth century, the KTK had become fixed.No one could alter its content or form without attracting the attention of scholars and copyists both in Syria and in Egypt, who were capable of detecting such a change.The KTK had become a staple of the Islamic tradition, and possibly even textbook.Not many books attained a level at which survival was no longer an issue and did not depend on the efforts of a few dedicated transmitters.Beyond the seventh/fourteenth century, the survival of the KTK was assured by the increase in the number of students copying it, as well as by the multiplication of copies later authenticated by permissions from reputed scholars.

Authorship of the KTK
The bibliographer Ibn al-Nadīm (d.385/ 995 or 388/998) in his Fihrist claims that Ibn Saʿd has only one book, which coincides with the sīra part of the printed edition of the KTK. 44However, Ibn al-Nadīm also added that Ibn Saʿd had 'compiled his books', thus insinuating that Ibn Saʿd might have had more than one book. 45  4 Idem.Ibn al-Nadīm also claims that these alleged works were a mere reworking of al-Wāqidī's compilations (Ibn Saʿd's main teacher and source of reports).
tābiʿūn'. 46Since the bulk of the KTK is composed of biographical information about the two classes of persons identified by Ibn al-Nadīm as Ibn Saʿd's area of expertise, it is then possible that the latter wrote something about that topic in order to establish his authority.Ibn al-Nadīm also attributes a book of ṭabaqāt to Ibn Saʿd's teacher and main source, al-Wāqidī (d.207/823). 47Given that he is the only bibliographer who has ever made such a claim, and since he considered that Ibn Saʿd's works were mere plagiarism of al-Wāqidī's work, it is possible that he attributed the ṭabaqāt work (one of possible two) to the teacher rather than to the student.Finally, when listing the books of which he was aware and whose authors were not known to him, Ibn al-Nadīm names a certain Kitāb al-ṭabaqāt and attributes it to a certain Muḥammad b.Saʿd. 48It seems to me that Ibn al-Nadīm either did not double check his sources or intentionally downplayed Ibn Saʿd's importance. 49f Ibn al-Nadīm cannot always be trusted in ascribing books to their rightful authors, it is necessary to use other sources to confirm that our Ibn Saʿd had written a work of ṭabaqāt that can be confidently identified with the KTK.This was indeed possible since the third/ninth centurygenealogist al-Balādhurī (d.279/892) in his Ansāb al-Ashrāf mentions in passing that 'Muḥammad b.Saʿd, the scribe of al-Wāqidī,' has to his name a book of 'ṭabaqāt of muḥaddiṯīn and fuqahāʾ,' 50 from which he has extensively borrowed.The borrowed material exists in the KTK, which proves that the third/ninth century compiler Muḥammad b.Saʿd is indeed the author of the KTK.In fact, we have in our hand a recension of the KTK which is different from the recension used in al-Balāḏurī's book. 5146 Idem. 47Ibid, Ḏikr mā waǧadtu min alkutub al-muṣannafa fī l-ādāb li-qawm lam yuʿraf ḥāluhum ʿalā l-iṣtiqṣāʾ.49 In comparing Ibn al-Nadīm's biography of al-Wāqidī and the latter's two biographies in the KTK, we are led to conclude that Ibn al-Nadīm's biography of al-Wāqidī is a type of summary of the two biographies given in the KTK.Ibn al-Nadīm also mentions that his source was none other than Ibn Saʿd, al-Wāqidī's scribe; see ibid., al-Fann al-awwal min al-maqāla al-ṯāliṯa: Aḫbār al-Wāqidī.
50 Balāḏurī, Ansāb al-ašrāf, 2: 263.Another third/ninth century author, Wakīʿ (d.306/918), in his Aḫbār al-quḍāt mentions, also in passing, that 'Muḥammad b.Saʿd, the scribe of al-Wāqidī,' has a book of ṭabaqāt attributed to his name.Wakīʿ,2: 397;3: 269. 51 In A.N. Atassi, History, 106-108 and 164-5, I suggest that Ibn Saʿd started writing the KTK sometime after 207/823, finished the bulk of it The Egyptian author Ibn Ḫallikān (d.681/1282), in his Wafayāt al-aʿyān, 52 mentions that Ibn Saʿd's Ṭabaqāt was a large (kabīr) book of fifteen volumes.Moreover, we learn there that there existed another work of ṭabaqāt that is a shorter (ṣuġra) version of the first.Is this a play on adjectives, or is al-kubrā really different from the KTK?We have previously concluded, when discussing Ibn al-Nadīm's claims, that Ibn Saʿd's Tārīḫ and Sīra (the first two volumes of the Leiden edition of the KTK) are most likely one and the same book.But it is curious that al-Ḏahabī mentions the Tārīḫ as if it were separate from the Ṭabaqāt.Cooperson thinks that the Sīra book 'may have been intended to stand as a separate text'. 56We also know that the manuscripts upon which the sometime around 213/828 (and started teaching it, which accounts of Ibn Abī Usāma's recension); and kept editing and adding new material to it until 228/842, or until shortly before he died (I dated the writing of Ibn Fahm's recension to around the interval 226/840-230/845). 52Ibn Ḫallikān, Wafayāt al-a‛yān, 4: 160, no.645.In fact, al-Ḫaṭīb al-Baġdādī mentions that Ibn Saʿd has compiled a 'large' (kabīr) book in the ṭabaqāt genre.Al-Ḫaṭīb, Tārīḫ Baġdād, 5: 321, no. 2844. 53  Leiden team depended for their edition were either transmitted or approved by al-Dimyāṭī, 57 al-Ḏahabī's teacher who taught him al-Ṭabaqāt al-kubrā.It is then possible that when the Sīra was added to Kitāb al-ṭabaqāt al-kabīr, the two together became known as Kitāb al-Ṭabaqāt al-kubrā.This lumping together of the Sīra and the Ṭabaqāt in one book may have been the work of al-Dimyāṭī.It is also possible that the two books, despite being separate entities, were transmitted together by the same teachers (al-Dimyāṭī, for example), and were thereafter treated as one book.
In al-Ḫaṭīb al-Baġdādī's Tārīḫ Baġdād, we encounter a report that matches verbatim the biography of Ibn Saʿd that appears in the printed edition of the KTK at the end of the section dedicated to Baghdadi transmitters. 58However, the isnād says explicitly that Ibn Fahm, a major transmitter of the KTK manuscripts, was the writer of the biography.It seems that Ibn Fahm has added it after the death of his teacher.It seems normal that the student pays homage to his teacher by informing the reader about him.However, there is more.The best example of a biography that Ibn Saʿd could not have written is that of Aḥmad b.Ḥanbal (d.241/855). 59First of all, Ibn Ḥanbal died ten years after Ibn Saʿd's death.Second, the biography mentions that 'Ibn Ḥanbal was summoned to appear before al-Mutawakkil and was later offered money which he refused to take'. 60The ʿAbbasid caliph al-Mutawakkil took office in 232/847, two years after Ibn Saʿd's death.Therefore, Ibn Saʿd could not have known this information.Moreover, the biography contains a description of Ibn Ḥanbal's funeral.There are also many biographical entries dedicated to persons who died after 230/845.Their author is possibly Ibn Fahm, but other transmitters of the book should not be dropped from consideration. 61ow, we must deal with the question of who put together the recension represented in the KTK's printed edition from different available recensions.Although all the transmission trees converge to a focal point at Ibn Ḥayyuwayh (295/907-382/992), I think that Ibn 57 See A.N. Atassi, History, 211-4 for an extensive discussion of the eighth generation of transmitters. 58Al-Ḫaṭīb al-Baġdādī, Tārīḫ Baġdād, 5: 370, no.876;.Ibn Saʿd, KTK, 7: 258. 59Ibn Saʿd, KTK, 7: 253. 60Idem. 61For an extensive discussion of biographies contained in the printed edition of the KTK, but that were possibly added after Ibn Saʿd's death, see A.N. Atassi, History, 113-29.
Maʿrūf al-Ḫaššāb started the process. 62All reports coming from Ibn Fahm were related by Ibn Maʿrūf only, without any exception.Moreover, we have not detected any Saʿdī report transmitted by Ibn Fahm with a chain different from that of the extant manuscripts.Therefore, it seems that Ibn Fahm bequeathed his recension of the KTK only to an otherwise ordinary student, namely Ibn Maʿrūf.Furthermore, Ibn Maʿrūf also transmitted reports from Ibn Abī Usāma, who also passed on a large number of Saʿdī reports, if not the entire KTK, to many students such as Wakīʿ and al-Ṭabarī.These reports came, as we will show in the next section, from Ibn Abī Usāma's own recension of the KTK.Why then would Ibn Maʿrūf be the only person interested in collecting two different recensions and passing them on to future generations?If Ibn Maʿrūf was interested in teaching the KTK, why did he then bequeath his collection or recensions only to Ibn Ḥayyuwayh, who later took charge of its distribution on a large scale?Ibn Ḥayyuwayh also collected parts, or all, of Ibn Abī Usāma's recension from al-Ǧallāb.What impact did Ibn Ḥayyuwayh, or for that matter Ibn Maʿrūf, have on the KTK, in addition to transmitting it?
The discovery of Ibn Haǧar's detailed transmission chain of the KTK puts everything back into question. 63It shows that Ibn Maʿrūf transmitted the two recensions of the KTK (those of Ibn Fahm and Ibn Abī Usāma) with sizeable lacunae, even at the biography level.It also shows that Ibn Ḥayyuwayh used most of Ibn Maʿrūf's material except for certain sections that he obtained from al-Ǧallāb.In fact, we can say the same thing about Ibn Maʿrūf.That is, he had the complete recensions of Ibn Fahm and Ibn Abī Usāma but preferred to combine them, just as Ibn Ḥayyuwayh did.It seems that there is enough room for arguing that the KTK was actually put together by Ibn Maʿrūf and later improved upon by Ibn Ḥayyuwayh.This conclusion is confirmed by the analysis of individual Saʿdī reports in compilations written before the fifth/eleventh century, as the next section will show.Ibn Abī Usāma's recension differs in many instances from the one available to us, because Ibn Fahm's recension was the one relied upon in our version of the KTK and not that of Ibn Abī Usāma. 64Therefore, we can say that the work of Ibn Maʿrūf and Ibn Ḥayyuwayh was a process of selection and fusion of the two recensions of the KTK available to them.Finally, we have showed earlier that Ibn Fahm, and possibly Ibn Abī Usāma, had added to the KTK.It is possible then to say that all members of these three generations of transmitters had an impact on the form and content of the KTK. 65wards the definitive text of the KTK Since our first encounter with Saʿdī reports, we notice that expecting a verbatim match between the reports found in a consulted compilation and the corresponding report in the printed KTK is unrealistic.The differences range from minor differences in word selection to major rewording of the report (while at the same time preserving certain core sentences).Other minor changes involve the order of a number of reports in a sequence, or changing the last transmitter (i.e. the source of the compiler).Major changes involve truncation of a long report, fragmentation of several reports and regrouping of selected fragments, grouping of several reports, and finally an extensive rewording of one or more reports.These changes can be consciously induced by the compilers or due to differences between the recensions used in the compilations consulted. 66e can distinguish three phases in the history of the KTK's recensions.The first phase stretches from the book's compilation by Ibn Saʿd early in the third/ninth century until the writing of Tārīḫ Baġdād by al-Ḫaṭīb al-Baġdādī in the first half of the fifth century.This is a period of relative obscurity and possible openness of the book.Any additions and/or modifications to the KTK must date to this phase.During this phase, we can talk about the possible existence of six recensions of the 64 For the analysis of Ibn Abī Usāma's recension and how it differs from the Leiden edition of the KTK see A. N. Atassi,History,166,168,[169][170] This agrees with Schoeler's conclusion, The Oral and the Written, 45, that the sources of these compilations (for example of Malik's Muwaṭṭā, of Ṭabarī's Tārīḫ and Quranic commentary, and of Abū l-Faraǧ al-Isfahānī's Kitāb al-Aġānī) are in most cases lessons given by the šayḫs on the basis of written notes (jottings), that they read or recited and which the pupils heard and wrote down (or took notes of).Most of them were not written works in book form, which authors definitively composed and published.Most of them were not purely oral transmission, meaning that the šayḫ and his audience did not keep the transmitted material exclusively in their memories.66 A lengthy and detailed discussion of the different recensions of the KTK that may have been used by later compilers is given in A. N. Atassi, History, 146-93.KTK that exhibit differences from the printed edition. 67The two most important recensions of which we have numerous quotes are those of al-Balāḏurī and Ibn Abī Usāma as we have seen this recension was not fully incorporated in the available manuscripts.68 There are also two possible recensions of unknown provenance: one used by Abū l-Qāsim al-Baġawī, 69 and the other used by al-Bayhaqī (d.458/1066).70 The remaining two possible recensions can be attributed to the Baghdadi transmitters ( 1 72 This period was covered in the manuscripts by the following transmitters: Ibn Abī Usāma and Ibn Fahm; Ibn Maʿrūf and al-Ǧallāb; Ibn Ḥayyuwayh; and finally al-Ǧawharī.The book has crystallized during this period with only one recension surviving, i.e. the one compiled by Ibn Ḥayyuwayh based on Ibn Abī Usāma's and Ibn Fahm's recensions.This recension of the KTK was actually the only one to have survived.Although many persons acquired the KTK from Ibn Saʿd or from Ibn Abī Usāma, very few of them decided to teach it to future generations.Most Saʿdī reports encountered between the third/ninth and fifth/eleventh centuries were transmitted individually, not as part of a wholesale transmission of the KTK.It is remarkable and worthy of investigating that Ibn Maʿrūf al-Ḫaššāb learned the KTK from Ibn Abī Usāma and Ibn Fahm then taught it to Ibn Ḥayyuwayh, who collected the material and divided it into systematic sections and then taught it to al-Ǧawharī, al-Azharī and few others.Beyond al-Ǧawharī's generation, many persons will be involved in teaching the KTK.In summary, we can say that the KTK crystallized by the process of dying out of all other recensions and the fusion together of Ibn Abī Usāma's and Ibn Fahm's recensions in a book that found generations of dedicated teachers.
The second phase stretches from the fifth century to the seventh century, ending with al-Dimyāṭī.During this phase the definitive text of the book spread outside Baghdad to Syria and Egypt, the two main centers of its later teaching.This phase saw an accurate and precise transmission of the KTK through the dictation-writing procedure.All our extant manuscripts go back to the end of this period.The third phase stretches from the seventh to the fourteenth centuries.During this phase the KTK continued being transmitted with the old dictation-writing procedure, but also saw the transformation of manuscripts into commodities bought, sold and inherited.This is how the extant manuscripts reached us.
Cairo after a long stay in Damascus; al-Ḏahabī actually went to Cairo to learn the KTK under al-Dimyāṭī; Ibn Abī al-Maǧd (707/1307-800/1397), a famous preacher and teacher in Damascus, was invited to teach in Cairo by an official of the Mamlūk establishment, Ibn Ḥagar al-ʿAsqalānī, who acquired the KTK through multiple channels (mostly from Damascus), later settled and taught in Cairo.This is not a surprise since power shifted from Baghdad to Cairo during the reigns of the Ayyūbids (564/1168-659/1260) and the Baḥrī Mamlūks (648/1250-784/1382). 74

Literary diffusion of Saʿdī reports
Although we differentiated between the KTK and individual Saʿdī reports, the diffusion of Saʿdī reports is an accurate measure of the diffusion of the KTK since most Saʿdī reports came from the KTK, and after the fifth/eleventh century most of them came from one recension of the KTK.The most fruitful in terms of producing Saʿdī reports are tarāǧim books.Sīra and maġāzī books and historiography books produced less Saʿdī reports than I originally expected.Ḥadīṯ compilations produced the least amount of information about the KTK or about Ibn Saʿd (books of ḥadīṯ criticism only produced short quotes and some clarifications).In fact, very few ḥadīṯs were transmitted on Ibn Saʿd's authority.Most of them come from one source, i.e. one of Ibn Saʿd's students, namely al-Ḥāriṯ b.Abī Usāma who was also a transmitter of the KTK. 75The majority of Saʿdī reports were biographical in nature.It came as a surprise to me that Saʿdī reports were less represented in historiography and ḥadīṯ books than in biographical dictionaries.It is a common practice in our field, when having general, collegial discussions of topics related to early Islamic periods, to talk in equal terms about historiographies and about biographical dictionaries; the latter usually being valuable sources of historical information.Moreover, given the lengthy biographies of the KTK, it is always considered a book of historiography.The previous results constitute a strong reminder that the two genres, namely historiography and biography, are not to be confused.They are actually very different in nature and often serve very distinct purposes.It seems that traditionalists 74 Atassi, History, 236, 245-8. 75Muʿǧam al-šuyūḫ/mašyaḫāt books only contain chains of transmission and not reports; therefore, this category will be dropped from the analysis of the KTK's literary diffusion.
have always regarded the KTK as a source for biographical information that is best suited for writing other biographical dictionaries.
Even as a biographical dictionary, the KTK is different from the ones dedicated to ḥadīṯ transmitters such as Ḫalīfa's Ṭabaqāt, Buḫārī's al-Tārīḫ al-kabīr, or al-Ṭabarānī's al-Muʿǧam al-kabīr.The latter books are terse and usually focus on the trustworthiness of transmitters.Biographies written by Ibn Saʿd are longer, contain more biographical and historical information, and follow a general model.At least for the biographies of Muḥammad's companions and the Medinan ḥadīṯ transmitters, the model seems to be Ibn Saʿd's biography of Muḥammad, since it is organized thematically rather than chronologically.These biographies are best described as hagiographies; the epic life-stories of the founders of the ḥadīṯ movement.After all, most of the book is dedicated to the companions and the Medinan transmitters; only two shorter sections are dedicated to Kufan and Basran transmitters; and even shorter sections relate to all other transmitters from the rest of the ʿAbbasid empire.It is no surprise then that most borrowings from the KTK come from the sections dedicated to Muḥammad's companions.

Conclusions
For the KTK, the paper has showed Ibn Saʿd was indeed its original compiler, but it also showed that three successive generations of transmitters had contributed to, or modified, it.Many recensions of the book circulated until the fourth/tenth century when a well known Baghdādī teacher called Ibn Ḥayyuwayh produced an authoritative recension.Beyond the fifth/eleventh century, only this recension dominated the market until modern times.Studying the geographic and temporal diffusion of the KTK, it became clear that its real popularity was ushered in by al-Ḫaṭīb al-Baġdādī's (fourth/tenth century) intensive borrowing from it in his Tārīḫ Baġdād.It was the Damascene scholar Ibn ʿAsākir (sixth/twelfth century) who brought the book from Baghdad to Damascus and extensively borrowed from it in his Tārīḫ Dimašq, thus popularizing it in the Muslim west.It is possible that he found in it a great help for his quest to implement the ǧihād agenda of Nūr al-Dīn Zankī against the crusaders.The KTK was taught exclusively in Baghdad until the early seventh/thirteenth century when almost simultaneously it started being taught in Aleppo, Damascus, and Cairo where it reached the zenith of its popularity.Studying the KTK's transmission methods showed that, by the ninth/fifteenth century, it was mostly transmitted by iǧāza (authentication, permission to teach).Starting from the third/ninth century, this usage increased as time progressed.By the ninth/fifteenth century, it was possible to obtain an iǧāza by mail.This is an indication that the KTK had acquired a stable form and had possibly become a textbook.Finally, the paper observed that tarāǧim books (biographical dictionaries) showed the most frequency of occurrence of Saʿdī reports.Sīra books and historiography books produced less Saʿdī reports.Ḥadīṯ compilations produced the least number of such reports.It seems that traditionalists have always regarded the KTK as a source for biographical information that is best suited for writing other biographical dictionaries.8 vols.(Leiden: Brill, 1905-17).
M. Manṣūr published the part missing from the Medinan ṭabaqāt.In 1998, Dār al-Kutub al-‛Ilmiyya, with M. ʿAbd al-Qādir ʿAtā as editor, published the more complete, but a worst, version of the KTK.In 1994, Muḥammad Ṣāmil al-Salamī published the fifth stratum of the companions.The fourth stratum of companions appeared in 1995 in a volume edited by ʿAbd al-ʿAzīz al-Sallūmī.Finally, in 2001 Maktabat al-Ḫānǧī in Cairo published the most complete version of the KTK edited by ʿAlī Muḥammad ʿUmar.

use recensions different from those in the printed edition of the KTK.
By history books I understand books of reports organized in any format other than the tarāǧim format.Such books include Ḫayyāṭ's Tārīḫ, Ibn Ḥabīb's al-Books of the Shīʿa tradition, sīra and mašyaḫāt lists are almost equally thinly represented in the sample.This imbalance may seem a great obstacle facing any serious conclusion as to the frequency of Saʿdī reports as a function of genre.However, the representation of different genres in my sample reflects their real representation in the entire Islamic tradition.Books of tarāǧim, ḥadīṯ and historiography are the most common.Sīra books are few and well known given the obvious limitation on their multiplication (i.e. the limited number of reports about Muḥammad's life and person).Ṣaḥīḥ or the thematic monographs of Ibn Abī al-Dunyā; the ḥadīṯs transmitted by one rāwī like the masānīd; or any book listing ḥadīṯs without any other kind of reports.In this category I include ḥadīṯ criticism (ǧarḥ wa-taʿdīl) books such as Ibn Ḥanbal's ʿIlal, Ibn Šāhīn's Tārīḫ Asmāʾ al-ṯiqāt, Ibn Mākūlāh's al-Ikmāl, and Ibn Ḥibbān's al-Ṯiqāt and his al-Ḍuʿafāʾ, and al-Ḏahabī's al-Muġnī fī al-ḍuʿafāʾ.
History, 229, for a discussion of whether al-Ḫaṭīb taught the KTK or not, and his probable role in introducing it to Damascus.31ForIbnKāra'smention in the available manuscripts see ibid, 222, 244-245, 247.We also know from Ibn Ḥaǧar al-ʿAsqalānī's transmission chain of the KTK that Ibn Kāra taught the book to a certain Ibn al-Ḥaǧǧāǧ.32Seeibid,232-3foradiscussion of Ibn ʿAsākir's popularization of the KTK in Syria.33Infact,IbnḤaǧar al-ʿAsqalānī, in al-Muʿǧam, 1: 168-70.suppliesuswith Naṣr taught at al-Manṣūr's mosque.39Althoughal-Madrasa al-Niẓāmiyya was built in 459/1066, none of the aforementioned teachers taught there because the vizier Niẓām al-Mulk, the founder of the school, prohibited non-Šāfi‛ites from teaching at al-Niẓāmiyya.
Šaraf al-Dīn Abū Muḥammad ʿAbd al-Muʾmin b.Ḫalaf al-Dimyāṭī is a famous Egyptian teacher who resided in Cairo.56M.Cooperson, 'Ibn Saʿd,' 201.This claim finds additional support in the fact that the manuscript of Kitāb at-Ṭabaqāt al-Ṣaġīr (Süleymaniye Library, Özel 216) does not include the Sīra or any abridgement of it; which could mean that the original Ṭabaqāt project that materialized in the KTK did not include a Sīra part.