Teaching ḥadīth: Ibn Ṭūlūn and the Thulāthiyyāt genre

This week I return to Garret Davidson’s dissertation on post-canonical ḥadīth transmission. Davidson offers a wonderful overview of the genres involved in establishing elevated, i.e. short, chains of transmission and the status that these brought for people who had access to them. While these genres could amount to very different scopes and sizes, it is safe to say that in this context small-scale hadith collections occupied a central position. These often would include the most elevated and thus sought-after chains a well-aged transmitter had to offer to the following generations, and at minimal time to be afforded. This impetus became pervasive in other fields of knowledge as well, at least for a lay audience.

Here, I will not give a summary of Davidson’s classification, however. I reckon for that we will have to wait for the reviews once the book is published. But, continuing where I left off in an earlier post, I will connect his general findings to the specific case of Ibn Ṭūlūn. In this post, the Thulāthiyyāt genre of ḥadīth will be at the center.

The term Thulāthiyyāt denotes originally the shortest traditions found in al-Bukhārī’s canonical ḥadīth work al-Ṣaḥīḥ (and to a smaller degree in other canonical compilations), which link al-Bukhārī to the Prophet by only three intermediaries. Starting in the 12th century, they were frequently compiled separately, again putting emphasis on elevated chains of transmission between the author of a compilation and al-Bukhārī:

By compiling a selection of only the collection’s shortest three-link chains of transmission scholars created a conduit that served to connect them to the Prophet through some of the shortest possible and most venerated chains of transmission (Davidson, 261).

The brevity of these compilations was their greatest strength, since they enabled transmission “in a matter of minutes making it a time efficient way for scholars to hear hadith from the most revered collection of the canon through some of the shortest chains of transmission available” and from large numbers of transmitters (ibid.). This would remain a popular mode of transmission until the 20th century.

Our old friend Ibn Ṭūlūn himself authored only one work of this genre, titled al-ʿUqūd al-luʾluʾiyyāt fī al-aḥādīth al-thulāthiyyāt (The arrangement / the half-score of pearls of traditions with only three links between al-Bukhārī and the Prophet), which does not seem to have survived.

Yet, in at least two of his biographical dictionaries, the genre appears frequently in the context of recitation sessions, attesting to the high status of elevated chains Davidson describes. The two titles Thulāthiyyāt al-Bukhārī and Thulāthiyyāt al-Ṣaḥīḥ are used interchangeably. Whereas in al-Ghuraf al-ʿaliyya fī tarājim mutaʾakhkhirī al-ḥanafiyya, Ibn Ṭūlūn mentions it as part of his education, in the second work, Dhakhāʾir al-qaṣr fī tarājim nubalāʾ al-ʿaṣr, he himself is the host or at least the authority in those sessions where the text is read.

I have not yet finished my reading of the former work. It is bulky with some 940 biographies covering mostly the 6th through 10th centuries. And so far, I have come across only one mention of the work. Yet, in unison with the other biographical work, it epitomizes the place the Thulāthiyyāt is granted in biographies. Moreover, in this case I would argue that the recitation was actually part of the author’s education and not purely for the sake of elevated chains of transmission.

The Thulāthiyyāt is mentioned in the biography of one Ibrāhīm b. Muḥammad b. Sulaymān (885-916), with whom Ibn Ṭūlūn studied (Istanbul, Süleymaniye Library,  MS Şehid Ali Paşa 1924, fols. 20a-22b). As in the following cases, the text is mentioned in conjunction with al-Musalsal bi-l-awaliyya. As Davidson explains it, it “means literally the hadith of serial first transmission, because it became common for it to be the first hadith an authority would transmit to a student coming to hear hadith from him” (Davidson, 125).

The reading of both texts took place when Ibn Ṭūlūn had already reached maturity. According to his own calculations, he must have been in his late twenties at the time. Moreover, in this instance, he received a qualified license to teach these texts:

I recited before him al-Musalsal bi-l-awaliyya with his conditions (sharṭ), then the Thulāthiyyāt al-ṣaḥīḥ, then I heard in his articulation al-Silsil bi-sūrat al-ṣaff, then al-Ḥadīth min riwāyat Abī Ḥanīfa. Then I extrapolated all of it before him in [several] sessions, the last of which fell on Sunday noon, 27.04.908 in the ḥanafī miḥrab of the [Umayyad Mosque]. He wrote me an ijāza for teaching in his own handwriting.

The joint mention of both texts occurs repeatedly in Ibn Ṭūlūn’s accounts of biographical trajectories of other ḥadīth transmitters. Whereas this makes sense for the Musalsal (since it is defined by being the first one transmitted), the elevated status of the Thulāthiyyāt is emphasized by being a close second. As the quote shows, they are followed by several more ḥadīth works.

This is strong evidence for a teaching context. Even more so since Ibn Ṭūlūn continues to list five more works he studied with Ibrāhīm, receiving another ijāza for giving fatwas on 09.04.911. That indicates a longterm commitment on both sides to this teacher-student relationship (rather taraddud than mulāzama, I guess). The session also coincided with his appointment as imām of the Yūnusiyya Khānqāh in the quarter of Sharaf al-Aʿlā (on 08.04.908).

Turning to the second work, Dhakhāʾir al-qaṣr fī tarājim nubalāʾ al-ʿaṣr, here Ibn Ṭūlūn emphasizes his role as a teacher and authority. This is little surprising since contextual evidence suggests he authored the work during the very last years of his life. It is much smaller in scope than al-Ghuraf with some 160 biographies and a much more decisively local selection of biographees.

In at least 15 biographees, Ibn Ṭūlūn mentions the Thulāthiyyāt al-Bukhārī / al-Ṣaḥīḥ, exclusively when he was the transmitter. In this case, Davidson’s argument that most hadith transmission should not be considered an educational  but a devotional endeavor seems to hold true. Only in one case, an ijāza for teaching is issued and then only after repeated auditions.

This impression is substantiated when we look at the sites of the sessions. The only institution featured in the sample where Ibn Ṭūlūn had a post, is the Ottoman Salīmiyya Mosque (four entries for Ṣafar 942; one for 05.05.949). Admittedly, it is the one site mentioned most often but none of his other teaching posts are at all mentioned. Instead, the second-most important site is the garden of his cousin Burhān al-Dīn Ibn Qindīl (three entries for 13.05.940).

Another case in point is that the account of these recitations is usually immediately followed by a list of which of Ibn Ṭūlūn’s works these people heard from him. Since the majority of those recitations took place in villages around Damascus, most importantly a shrine for Abraham in Barza (maqām Khalīl), and since those works were usually hadith compilations, this suggests that also the Thulāthiyyāt were recited in a devotional context.

“Speed reading” of hadith

I have recently been reading Garrett Davidson’s exciting PhD thesis “Carrying on the Tradition: An Intellectual and Social History of Post-Canonical Hadith Transmission“. It attempts – very successfully – a reassessment of the methods, purposes, and underlying logic of hadith transmission once the canon was fixed with the compilation of the canonical six books.

It contains a plethora of points pertinent to Ibn Tulun’s oeuvre, and I will address some more in the future, such as how the devotional character of reciting hadith as well as chains of transmission relate to several of his compilations on animals, plants or food. Even though he is often characterized as a historian, hadith was like the mortar that held the edifice of his written corpus together.

The one point I want to address today is the phenomenon of speed reading (sard) which  preserved the appearance of oral transmission while trying to fit as much text into a restricted period of time (pp. 107-112). Davidson mentions several outstanding examples of that practice when readers went through the entirety of Bukhari’s or other canonical hadith compilations in only a few days.

Interestingly, Ibn Tulun also mentions most of the same feats in his biographical dictionary Dhakhāʾir al-Qaṣr fī Tarājim Nubalāʾ al-ʿAṣr. They appear in the biography of one Muhammad b. Ibrahim b. ʿAlī al-Nābulusī al-Qudsī, who was also known as al-Hijazi (MS Gotha, Orient A 1779, fols. 57a-b). While Garrett cites multiple sources like the Meccan historian al-Fāsī, Ibn Rajab, al-Kattani, or Ibn Hajar himself, Ibn Ṭūlūn combines all this information in this inconspicuous place.

Among those examples is that of Ibn Ḥajar’s “most famous act of speed transmission[…,] his having raced through al-Ṭabarānī’s al-Mu’jam al-saghir, a collection that consisted of more than one thousand five hundred hadith, in a single audition session between the noon and afternoon prayers” and Bukhari’s Sahih in ten sessions (pp. 109-110).  Or, in Ibn Tulun’s own words:

the fastest thing happened to Abū al-Faḍl Ibn Ḥajar on his Syrian journey when he recited the al-Muʿjam al-ṣaghīr by Abū al-Qasim al-Ṭabarānī in one session between the noon and afternoon prayer[ recitation]. This book is a volume which is organized in about 1.500 ḥadīth, because he published one unique ḥadīth for each of his 1.000 shaykhs. And he transmitted the Ṣaḥīḥ in the Khānqāh al-Baybarsiyya in 10 majālis of 4 hours each.

Otherwise, Ibn Tulun says about the biographee that he heard some of Bukhari’s Sahih with him but recited by a member of the al-Shuwaykī family. While this took place in the Takiyya Salimiyya, he also attended shortly Ibn Tulun’s sessions (durus) in the Umariyya madrasa just next door. He also attended the teachings of shaykh Sharaf al-Dīn al-Ḥijāzī al-Ḥanbalī but fell out with him later. Almost half of the biography, however, is an account of feats in speed reading.

The practice seems to have been continued into Ibn Tulun’s own times. The biographee is doubtful that such fast readings could be done at all but Ibn Tulun (obviously) has examples at hand. He himself had seen in the thabat of the khatib Shihab al-Din al-Himsi that he had recited the whole Sahih al-Bukhari in six sessions in Ramadan 882. Ibn Tulun’s teacher had another example in Jamal al-Din al-Askari who went through the whole work in three and a half days at the end of Rabi II 880.

Ibn Tulun himself seems to have endorsed speed reading. The former’s cousin Zayn al-Din al-Askari recited the Sahih before Ibn Tulun in five days and Abu al-Abbas Ahmad b. Ahmad al-Sufi’s work named here only as al-Mu’taqad before him in three days, “the last of which was 15 Ayyār 1806 according to the Alexandrian and Roman (rumi) calendar, which is one of the longest days of the year”. So while Ibn Tulun did believe in the benefits of speed reading, he was also aware that certain extrinsic conditions could be favorable in doing such feats.

Finally, speed reading appears to have had a pervasive function in communal devotional contexts at the time. In 873 (1468-69), Damascenes suffered through a drought and, thus, a dearth in food prices. The price for wheat increased five-fold  to more than 2.000 dirham per sack (ghirara). Ibn Tulun’s near-contemporary Ali al-Busrawi describes in his chronicle that people were so desperate that they gave away their children and themselves resorted to eating “mayta” – carrion or corpses? – just to stay alive.

In reaction to their plight, people convened during the holy nights around Mid-Sha’ban praying for rain. Al-Busrawi also states that in the course of just two nights, they recited the complete Sahihs of both al-Bukhari and Muslim – twice each night! According to the account, those measures proved successful: the very next day, rainfalls began and simultaneously wheat prices plummeted by half.

Initially, I suspected the Sahih recitations (and some other aspects of the account) to be tropes. Yet, there are other explanations, one of them being speed reading. Nonetheless, if we compare this instance with the acclaimed feats mentioned before, the double recitations still seem out of the ordinary (or even the extraordinary). The only explanation I can imagine would be that in this context several reciters worked simultaneously, splitting up the two Sahihs between them. The auditory experience must have been incredible.

Ibn Tulun manuscripts in Leiden

Among other libraries, Leiden University hosts several autographs by Ibn Tulun (for an introduction of the Oriental MS collections, see here). The collection is quite extraordinary among those, since every Ibn Tulun text is bound individually, even though most of them are of modest size. Their page numbers range between single and low double-digits. Judging by my experience that would make them ideal candidates for publication in majmu’as (and I address that issue in an article submitted to the Journal of Islamic Manuscripts).

Unfortunately, I have not yet been able to consult the whole collection in person. So far, I had to rely on the catalogue made by Carlo von Landberg shortly after the acquisition in the 1880s and the Handlist of P. Vorhoove. In both, the Ibn Tulun manuscripts are clustered within a range of about twenty call numbers (132-146 in Landberg, 2503-2520 in the Handlist). While this could be attributed to their common author, it is also possible that this cluster was retained from the collection of the Cairene seller Amin al-Madani and ultimately from their original state in autograph majmu’as.

Be that as it may, Leiden has uploaded one of those manuscripts, MS Or. 2512, which contains the short text Tuḥfat al-ḥabīb fīmā warada fī al-kathīb. This is, in fact, the title as it is given in Ibn Tulun’s own work list, whereas the Leiden catalogue gives a slightly different title that is written on the first recto of the manuscript: Tuḥfat al-ḥabīb bi-akhbār al-kathīb.

This is apparently owed to the author’s phrasing that this text contains “mā akhbaranā shaykhunā al-muḥaddith…” and perhaps also to titling conventions at the time when this title was added (the hand is different and more modern than the author’s). Yet, both titles refer to the same work as Ibn Tulun mentions the “warada” title on the verso of fol. 1.

Initially, I was surprised that the writing already begins on the recto of fol. 1. For Ibn Tulun this is highly unusual, especially since the text appears to be a (very well preserved) fair copy. To elaborate, it contains no other handwriting than the author’s, and that is restricted to an even and regular text block of 23 lines per page. The margins seem wider than I have seen in other, more annotated Ibn Tulun manuscripts.

Getting back to my surprise, a quick check on the verso made sure that the text begins only here. Instead, the writing on the recto, albeit also in Ibn Tulun’s hand, seems to be an audition certificate, giving part of the isnad, some of the attendants and the date of the text’s recitation: 9.11.936/15.07.1530. ّIt also indicates that all attendants received a certificate for transmission (ijāzat riwāya) from the author.

The catalogue gives that date as the date of the text’s completion but I am uncertain whether that conclusion can be so easily made. I am not even sure whether we can assume that this specific manuscript was completed before that date. The clean state of it suggests rather that it might be a fair copy of the original, which would have been in existence by 936/1530.

The text itself is a collection of reports and sayings about Moses (al-kathīb) and his shrine (maqām/ziyāra) near the village Masjid al-Qadam, south of Damascus. The most interesting thing about it – for me – is that it appears in Ibn Tulun’s biographical dictionary Dhakhāʾir al-Qaṣr fī tarājim nubalāʾ al-ʿaṣr.

At least three of the biographees included in this work heard recitations of this work. One of those heard it recited at the “Ziyārat al-Kathīb close to the village Masjid al-Qadam” on the same day as the one given in the samāʿ. This Muḥammad b. Mūsā b. ʿAbduh al-Qubaybātī al-Ḥanbalī known as Ibn Qayṣar went on to write a mulakhkhiṣ of this work, giving a rare proof to Ibn Tulun’s reception through emulation.

 

The real manuscripts of Dar al-Kutub

In December 2017, I attended a workshop on Arabic codicology at the French Archaeological Institute in Cairo (IFAO). It was taught by Elise Franssen and co-organized by Mathieu Eychenne and Abbès Zouache.

In this context, we also paid a visit to the Dar al-Kutub, the Egyptian National Library, which holds one of the largest collections of Arabic manuscripts in the world. I also visited its department for manuscript on two more days during my stay. Noah Gardiner has written an exhaustive guide to research and especially the manuscript catalogues of the Dar al-Kutub.

It was my second visit to this institution, after I had spent a week there in November 2016. Before this first visit I had heard from friends and colleagues that with such little time it would not make sense to even try to get access to the physical manuscripts, and thus I restricted my research to the microfilm section, which is situated in a different part of the city (Bab al-Khalq).

To be sure, this visit paid other dividends but it made it apparent to me several questions that could not be answered without consulting the manuscripts themselves. And research at the Nile branch where the manuscripts are kept turned out to be much more pleasant than anticipated. The staff is in general very forthcoming and the head of the section is knowledgeable in manuscripts and himself an experienced editor (muḥaqqiq).

Some things should, however, be considered prior to one’s visit. These are not always congruent with the procedure at the Bab al-Khalq branch, and actually less strict. For instance, computers are allowed. The first thing you should prepare is a list of manuscripts you want to see with call numbers, titles and author names. Secondly, bring your passport or another kind of ID. You will have to leave it at the entrance.

You will need your institution to provide you with a letter of recommendation in Arabic. Best, if it already lists the manuscripts you aim to see. If your institution is not able to do that, one alternative option is to ask at the IFAO if they provide a letter.

The other thing you will need is time. There is the issue of finding and bringing the manuscripts. You can only have one on your desk so these intervals appear repeatedly. And on the first day you might be invited for a review of your letter of recommendation as well. The best time to arrive is around 10:00 a.m. Although the section opens at 9:00, the staff often arrives later than that.

With a closing time at 2:30 p.m., you might need at least a week if you are interested in the content of a longer work. Time is also required if you prefer to get images of the manuscript. In my experience, they will usually be made within the day but not long before closing time. And if you don’t want the full manuscript, calculate an extra day to get also those pages photographed that were missed on the first try (Do not make the mistake I made!).

Finally, the images are of a good quality albeit not of a resolution as high as those you get from, let’s say, Chester Beatty Library. But then you pay a pittance compared to that library. The only downside is the watermark that all images from Dar al-Kutub carry. Unfortunately, that leads to some annotations being not as legible as in the manuscript itself.

During my admittedly short stay, the other visitors could be counted on one hand. So if you need manuscripts from the Dar al-Kutub, the time to get them is now.

Did Ibn Ṭūlūn read Ibn Iyās? A manuscript from Paris

One obvious course in establishing a scholar’s importance as well as his intellectual biography is the search for manuscript copies of his works. Whereas this step of my project is mostly finished, a recent attendance at a codicology workshop at the IFAO in Cairo did bring up this topic once more.

I am grateful to IFAO’s own Robin Seignobos who pointed me to MS Paris, BnF, Or. 3973, which contains a text on Nile tides ascribed to one Shams al-Dīn Muḥammad al-Dimashqī al-Ḥanafī (fols. 170-189). He also made me aware of David Wasserstein’s article “Tradition manuscrite, authenticité, chronologie et développement de l’oeuvre littéraire d’Ibn Iyās” (Journal Asiatique 280 [1993]: 81-114) which provides a more detailed description of the manuscript and the text.

Finally, Robin speculates that this text, “Nuzhat al-khāṭir wa-bahjat al-nāẓir fī ziyādat al-nīl wa-nuquṣānuhu wa-muntahā ziyādatihi wa-awānihi“, was authored by Ibn Ṭūlūn.  If this is the case, this would be the second Ibn Ṭūlūn text in the Bibliotheque nationale de France. The other one is a short prayer likewise included in a composite manuscript (MS ar. 1945, fols. 67b-68a) which references itself as “ṣalāt Ibn Ṭūlūn” (fol. 68a).

Yet, in both cases the question remains whether the author really was “our” Muḥammad Ibn Ṭūlūn. Both manuscripts are later copies of an original according to their colophons. It is interesting that the given author name, in MS 3973, lacks the most distinct part of Ibn Ṭūlūn’s writerly identity, whereas in MS 1945 it is the only element of appellation.

Here, however, I will rather talk about the topic of the text on the Nile to substantiate Robin’s suggestion. Ibn Ṭūlūn wrote texts on Egypt (al-Ajwiba al-jalliyya ʿan al-asʾila al-miṣriyya, Ḥawar al-ʿuyūn fī tārīkh Aḥmad Ibn Ṭūlūn, ʿAjab al-dahr fī tadhyīl min malik Maṣr) and he wrote texts on recurring natural phenomena (e.g. in his Nukat al-tārīkhiyya, nos. 19, 22 or his works on animals or plants) but there is no evidence in his work list nor his autograph manuscripts that he ever combined both.

Moreover, as far as we know he never traveled to Egypt to see the Nile with his own two eyes. (In turn, this might help explain why his name is given as it is in a compilation that is clearly Egyptian. For an Egyptian audience, he was simply a Ḥanafī and Damascene.)

I should mention that the Nuzha is vocalized throughout. While this can be attributed to the manuscript being a later copy, it might point to a very different readership that Ibn Ṭūlūn’s autographs originally had.

That being said, the Nuzhat al-khāṭir shares some characteristics with Ibn Ṭūlūn’s known writings. But in some ways it stands out amongst them: The basmallah is followed by an exceptionally short Tamḥīd: “al-ḥamdu li-llāh”. Elsewhere, this section often covers several lines of texts. Instead, it is followed here by several lines of rhymed prose.

Then, the text starts in earnest after the ubiquitous “wa-baʿd”. In general, Ibn Ṭūlūn would almost instantly follow this with the title of the work and often an indicator towards its genre (most often “taʿlīq”). This element is missing in this text, which could discredit it immediately as a false ascription or possibly the work of another Damascene Ḥanafī scholar by the name of Muḥammad. However, the following introduction sounds very much like something that Ibn Ṭūlūn could have written:

فقد وقفت على كراريس مخرومة الاوائل متعلقة باخبار المبارك الاعلام جامعها و يمكن الوقوف على اسمه من ديباجة تاريخ له و سماه بدايع الزهور في وقايع الدهور وجدتها محتوية على لطائف و حوادث و غرائب و تواريخ في زيادة النيل و نقصانه
fa-qad waqaftu ʿalā karārīs makhrūma al-awāʾil mutaʿlliqa bi-akhbār al-mubārak lā aʿlam jāmiʿahā wa-yumkin al-wuqūf ʿalā ismihi min dībājat tārīkh lahu samāhu “Badāʾiʿ al-zuhūr fī waqāʾiʿ al-duhūr” wajadtuhā muḥtawiyyatan ʿalā laṭāʾif wa-ḥawādith wa-gharāʾib wa-tawārīkh fī ziyādat al-Nīl wa-nuqaṣṣānihi
I have come upon some incomplete quires attached to “Akhbār al-Mubārak”.*  I do not know whether the work is complete. It is possible to gather the ism (the author’s name or the work’s title?) from the poem on history (or date?). He called it “Badāʾiʿ al-zuhūr fī waqāʾiʿ al-duhūr”. I found it comprises subtleties, events, oddities, and dates of the Nile’s high and low tides. 
* I am not sure whether this is a title; the name Mubārak could relate to one of several traditionaries (e.g. ID: 14138, 3960, 14173, 14212, 14292, 14410, 14457, or 14081 in the onomasticon arabicum).

The name of the excerpted work will be well-known to most people familiar with the later Mamluk period. It is none other than the great chronicle of Abū al-Barakāt Muḥammad Ibn Iyās al-Jarkasī al-Ḥanafī (1448-1524), which is perhaps the most important Egyptian source on the Ottoman conquests.

According to Brinner’s article in the Encyclopaedia of Islam 2, his oeuvre remained relatively unimportant in his own time, which is supported by the state in which it was found by the author of the present treatise. If this indeed was the case, the work’s reception seems more probable in the first decades after his death than much later.

As far as I know, no connection between him and his Damascene counterpart Ibn Ṭūlūn has so far been known. Yet, Ibn Ṭūlūn would have been the person to acknowledge the value of an obscure fragment he found in the back of a manuscript.

The way the text is introduced further points to Ibn Ṭūlūn as the author: Instead of an author or abstract text, the author refers to a distinct manuscript, apparently a majmūʿa containing at least two texts. What follows, is in form and content quite similar to many of Ibn Ṭūlūn’s works, namely commented excerpts of another author’s text, which he extracted and compiled around a specific topic.

The text is structured by source and thereby to a certain extent also by theme.  It begins with legends or sayings about the Nile and its springs that is followed by a description of its geography (one of the sources there is al-Masʿūdī). Then the text addresses the measurement of the nile floods (starting fol. 176b). A strict annalistic order begins only on fol. 181b and seems to be taken from a work by Ibn al-Jawzī. From there to the end, nile floods from 208 through 922 AH are described.

Unfortunately, I must admit that I do not know Ibn Iyās’ Badāʾiʿ well enough to say how much of this treatment of the Nile is taken from his work. Yet, to me it seems that such an argument would not be found in a chronicle of a basically political-social nature.

However, the structure is reminiscent of Ibn Ṭūlūn’s bio-topographical work on a cemetery in Damascus, Ghāyat al-bayān fī tarjamat al-shaykh Arslān. The biography as such is followed by a description of the cemetery and concluded with several biographies and events in which it played a role. Therefore, I would say that it is probable that this work on the Nile be added to Ibn Ṭūlūn’s established corpus.

That leaves the question why this text does not carry the characteristic description as a taʿlīq. One tentative answer could be that it was not written in a scholarly or educational context. That might also explain why it is not mentioned in Ibn Ṭūlūn’s work list.