To categorise things is a fundamental human and scholarly instinct and activity. And yet it is one not without obstacles, for we soon learn that the world is more complex than we originally thought or that we are confronted with something which refuses to conform to our usual categories. To put it somewhat casually, if our world is populated only by monochrome horses, we probably will have a hard time putting zebras into our usual colour categories—until we make room for them in our classification system. (Of course, even if we fail to do so, this would still allow us to pat them on the head and full-heartedly embrace their difference.)
In this blog, I suggest that similar problems occasionally occur when we try to classify pre-modern Islamicate texts in our OpenITI / KITAB corpus respectively when we try to fit certain pieces of information related to them into our existing metadata categories. Many of these texts were produced within intellectual fields with their own distinctive terminologies and practices which more often use terms and refer to activities without equivalent in our contemporary world. This effectively means that our usual categories are on occasion unable to catch differences for which they were not made. (Some more or less extensive periods of head-patting might be necessary here.)
Now, of course, we might simply call a zebra a black horse and ignore the white stripes, but considering the loss of information that this would entail, it is probably not such a good idea. At KITAB, we are convinced that our metadata should, as far as possible, preserve the terminology of our pre-modern authors and respect distinctions and specificities which they felt were important enough to be reflected in language.
To give a small, but illustrative example. In the OpenITI / KITAB corpus, the term takhrij often occurs in the context of mashyakhat (sg. mashyakha) “lists of shaykhs of an individual” in the sense that a particular mashyakha is stated to be the “mashyakha of X, takhrij Y.” For example, our corpus contains the mashyakha of Muhammad b. Ibrahim al-Bayani in the takhrij of Ibn Rafiʿ al-Sallami. This means that Ibn Rafiʿ al-Sallami contributed to the creation of the mashyakha of al-Bayani through an activity called takhrij (see below).
Although there is variation with regard to their individual make-up, a mashyakha generally is a form of text which mentions the shaykhs of an individual and for each of the shaykhs mentioned adduces one or more hadiths that were transmitted by him or her.
Such lists of shaykhs may also be termed muʿjam (pl. maʿajim). A muʿjam essentially is the same thing as a mashyakha, with the difference that it mostly arranges the shaykhs in alphabetical order.
To a modern reader, this might seem like a strange form of text, but such texts fulfilled clear purposes within the field of knowledge transmission and especially hadith transmission. They showed the position of an individual within the field and allowed to see in condensed form under which scholars that individual had studied and which transmissive chains he therefore “owned.” But they also allowed the students of that individual to connect with the same transmissive chains in a structured and efficient way.
But what about this term takhrij? To what sort of text and associated practices does it refer and how can we account for its specificities in our metadata?
Grammatically speaking, takhrij is the verbal noun of the verb kharraja / yukharriju “to extract.” In the pre-modern scholarly social worlds, in which the term was used in this particular sense, it clearly refers to a distinct form of scholarly activity.
But what exactly is this activity? While it is not easy to pin down the exact nature, it seems to have been of an essentially compilationary character. It furthermore was something one could do for oneself (“kharraja li-nafsihi….”) or for someone else (“kharraja lahu…”). There is a record of siblings who engaged in this activity for each other, fathers for their sons, students for their shaykhs, scholarly peers for each other, to name but a few examples. The individual actively engaging in the activity of takhrij is referred to as mukharrij, whereas the one for whom someone else engages in this activity is termed mukharraj lahu. It seems that some individuals specialised in this form of activity more so than others (introduction to Subki’s Muʿjam, pp. 15–16). The aforementioned Ibn Rafiʿ al-Sallami is a good example. He is credited with the takhrij of more than a dozen mashyakhat for others (Bayani Mashyakha, pp. 20–21).
Depending on the structure and content of the individual mashaykha “list of shaykhs,” the work involved in its creation must have been a more or less labour-intensive combination of several tasks. The first task involved gathering hadith material transmitted by the individual shaykhs and the respective transmissive chains by going through one’s own or someone else’s notes, audience certificates, and ijazat “authorisations to transmit a text” (for example, Abu ʿAbdallah Muhammad al-Razi refers to Abu Tahir al-Silafi as the “mukharrij hadhihi l-fawaʾid min masmuʿati,” Razi Mashyakha, p. 100), making a selection of the gathered material, maybe putting the shaykhs into alphabetical order, maybe adding biographical information for each of them, and so forth.
This essentially compilationary character of the activity is corroborated by other verbs that are used instead of kharraja. Apart from the more neutral verb ʿamila “to make,” most notably, the verb jamaʿa “to gather” occurs. For example, al-Samʿani states that he had put together a muʿjam for his son (“qad kuntu jamaʿtu muʿjam shuyukhihi”) (Samʿani Muntakhab, p. 112). Al-Dhahabi refers to this piece of information with the phrase “his father produced a muʿjam for him” (“ʿamila lahu abuhu muʿjaman”) (Dhahabi Siyar aʿlam al-nubalaʾ, vol. 22, p. 107).
It is interesting to note that some form of personal relationship mostly seems to be playing a role when people engaged in this activity for someone else. As has already been pointed out, they put together mashyakhat / maʿajim for their kin (even milk kinship, see Ibn al-ʿAttar Tuhfat al-talibin, p. 28), teachers, and peers. Even if someone apparently had to be talked into doing the work, the personal relationship seems to matter. An example for the latter case is provided by the introduction to Burhan al-Din al-Jaʿbari’s Mashyakha al-shamiyya, which was put together by al-Qasim b. Muhammad al-Birzali (i.e., al-Birzali was the mukharrij). Al-Birzali recalls the following in the opening passage of the mashyakha: “He [i.e., Burhan al-Din] had a mashyakha that contained all his shaykhs from Baghdad from whom he transmitted hadith and which had been put together for him by Jamal al-Din al-Qalanisi and which hadith students and those who wanted to transmit hadith (ṭalabat al-hadith wa-l-raghibuna fi l-riwaya) read in his presence […]. Then he wanted that another mashyakha be put together for him (yukharraj lahu) which would contain all his shaykhs from Syria so that the material transmitted by him and his shaykhs would be more numerous and his students and all those who sought him out, traveled to him, and visited him would benefit from that. He wrote to me in that regard and sent a companion of his who prodded me (yuharriduni) into putting together the respective material (jamʿ al-matlub).” (MS Leipzig Vollers 719-01, 2v–3r).
The above quote also well illustrates one purpose of the mashyakhat: They allowed to see how each of the shaykhs mentioned in them connected the individual whose mashyakha it was (the so-called “sahib al-mashyakha”) to transmissive chains that lead all the way up to the Prophet and that were needed for the legitimate transmission of knowledge. By virtue of the same logic, if the students of the sahib al-mashyakha read the mashyakha in his presence, they were connected to all of his chains of transmission in a highly efficient and organised way (among other motivations that might play a role, this explains the interest in large numbers expressed in the above quote).
But if a mukharrij put together a mashyakha for someone else, what was the respective contribution of mukharrij and mukharraj lahu? Who should be considered the “author”? Both? Only the mukharrij? In the latter case, are we not missing something important in our metadata? And is “author” even a good term to describe this sort of activity?
The question of what the mukharrij actually did is not necessarily easy to answer. It is rather fascinating to see how often it puzzles the modern editors of mashyakhat who occasionally state they do not know: “As far as the work of the mukharrij is concerned, we have no information that would clarify what his work actually consisted in” (Naʿʿal Mashyakha, p. 22; for a similar statement, see Subki Muʿjam, p. 6). As the editor of Subki’s Muʿjam suggests, things are even more complicated insofar as the fact that passages are written in the first person singular does not necessarily mean they were really written by that person. He highlights a passage by al-Dhahabi in which al-Dhahabi writes about the Muʿjam of al-Samʿani’s son ʿAbd al-Rahim that it is “his father’s words put in the mouth of ʿAbd al-Rahim” (“wa-huwa kalam abihi ʿala lisan ʿAbd al-Rahim”) (see Subki Muʿjam, p. 6; for the quote, see Dhahabi Tarikh al-islam, vol. 12, p. 58).
To give another example of how difficult it is to determine who did what, in the mashyakha of Muhammad b. al-Anjab al-Naʿʿal, the takhrij of which is attributed to al-Mundhiri. This much is stated by al-Naʿʿal. He writes: “Thus I asked God for guidance and put together in this book a large number of the shaykhs that had issued an ijaza to me” (“wa-kharrajtu fi hadha l-kitab jumla min mashayikhi al-mujizin;” see Naʿʿal Mashyakha, p. 55). So, al-Naʿʿal refers—in the literal sense of the sentence—to himself as engaged in takhrij. If so, what role was left to al-Mundhiri, who is explicitly named as the mukharrij?
This is not the appropriate place for an extensive study of the practice of takhrij. What matters here is that from the perspective of corpus management it would be absurd to build a corpus consisting specifically of Islamicate texts, yet at the same time disregard culturally specific scholarly practices which generate what are unique items of metadata for us. Unless we make use of such terms, we will be turning zebras into black horses.
To an extent, this is what happens in contemporary library catalogues. Thus, with regard to takhrij and the distinction in the pre-modern sources between mukharrij and mukharraj lahu, the catalogues characteristically cannot really and fully accommodate the situation created by the pre-modern scholarly practices.
This may be demonstrated by the two worldcat entries for al-Bayani’s above-mentioned mashyakha. In one entry, the work is attributed to Ibn Rafiʿ al-Sallami (i.e., the mukharrij); in the other entry, it is attributed to al-Bayani (=Ibn Imam al-Sakhra) (i.e., the mukharraj lahu).
At KITAB, we are convinced that such information must be recorded in our metadata in a consistent and machine-readable way. If our sources tell us that someone acted as a mukharrij for someone, we should always be able to retrieve both mukharrij and mukharraj lahu, —as well as the fact that they are linked through this particular relationship. We are convinced that being able to search our corpus metadata for information of this kind constitutes real progress book history and we are currently working out a system of how to generally preserve the pre-modern terminology we find in our texts and the various relations between texts and persons signified by terms such as takhrij, mukharrij, and mukharrij lahu.
Of course, it is not entirely wrong to state that the mukharrij is the “author,” but it is not entirely right either. Most certainly, such terms force us to be more specific with regard to our notion of authorship, make room for pre-modern terminology, and rethink what they meant in the highly specific historical scholarly worlds we are dealing with and in which the terms arise.
From the perspective of corpus management, we must be ready to adapt our metadata categories to accommodate specificities if we become aware of them through research. If we level them, our metadata is much poorer than it has to be. By contrast, preserving them along with the pre-modern terminologies themselves will lay the foundations for retrieving richer information from our corpus and will allow users to ask different questions.
While none of this can be implemented overnight, it is important to recognise the issue and begin to address it.
al-Bayani, Muhammad b. Ibrahim: Mashyakhat al-musnid Muhammad b. Ibrahim al-Bayani al-maʿruf bi-Ibn Imam al-Sakhra. Takhrij al-hafiz Ibn Rafiʿ al-Sallami. Bayrut 2004
al-Dhahabi, Shams al-Din Abu ʿAbd Allah Muhammad b. Ahmad: Tarikh al-islam. Bayrut 2003
al-Jaʿbari, Burhan ad-Din Abu Ishaq Ibrahim b. ʿUmar b. Ibrahim b. Khalil al-Rabaʿi: Mashyakha mubaraka Shamiyya li-l-shaykh al-imam al-ʿalim al-ʽamil Burhan ad-Din Abu Ishaq Ibrahim b. ʿUmar b. Ibrahim b. Khalil al-Rabaʿi al-Jaʿbari al-Shafiʿi nazil balad al-Khalil ʿalaihi al-salam ʿan shuyukhihi al-shamiyyin alladhina ajazu lahu. Takhrij al-Qasim b. Muhammad b. Yusuf b. Muhammad al-Birzali. MS Leipzig Vollers 719-01
al-Naʿʿal, Saʾin al-Din Muhammad b. al-Anjab: Mashyakhat al-Naʿʿal al-Baghdadi. Takhrij al-hafiz Rashid al-Din Muhammad b. ʿAbd al-ʿAzim al-Mundhiri. Baghdad 1975
al-Nawawi: Al-Ijaz fi sharh sunan Abi Dawud al-Sijistani. wa-fi muqaddimat tahqiqihi Tuhfat al-talibin fi tarjamat al-imam Muhyi al-Din tasnif ʿAlaʾ al-Din ʿAli b. Ibrahim b. al-ʿAttar. ʿAmman 2007
al-Razi, Abu ʿAbd Allah Muhammad b. Ahmad b. Ibrahim: Mashyakhat al-shaykh al-ajall Abi ʿAbd Allah Muhammad b. Ahmad b. Ibrahim al-Razi al-maʿruf bi-Ibn al-Hattab wa-thabat masmuʿatihi bi-intiqaʾ al-shaykh al-ajall al-imam al-ʿalim al-hafiz Abi Tahir Ahmad b. Muhammad b. Ahmad al-Silafi. Al-Riyad 1994
al-Samʿani, Abu Saʿd ʿAbd al-Karim b. Muhammad b. Mansur al-Tamimi: Al-Muntakhab min muʿjam shuyukh al-imam al-hafiz Abi Saʿd ʿAbd al-Karim b. Muhammad b. Mansur al-Samʿani al-Tamimi. Al-Riyad 1996
al-Subki, Taj al-Din ʿAbd al-Wahhab b. ʿAli: Muʿjam al-shuyukh. Takhrij Shams al-Din al-Salihi. Bayrut 2004