• Table of Contents
    • Methodological Remarks: Project al-Najjari
    • Category Remarks: Printed Books

Methodological Remarks: Project al-Najjari

Adam Mestyan & Kathryn A. Schwartz

I. Printed titles

How do we reconstruct the titles in al-Najjari’s book collection? What do we mean by “reconstruction”? What is a title? 

In this short description we outline our methodology in creating our al-Najjārī Google Sheet database. This is almost identical to the description of the process of making the database. The process itself, however, is not linear. We have several waves of returning to data, cross-checking each others’ work, etc. (see Work Flow).

There are three phases of identification: “the title,” “the work,” “the edition.” Once we determine these identifications, we classify editions according to criteria of “certainty”: absolute and relative certainty. Only absolute or relatively certain editions appear in the database. 

I.1. “The title”

First, Mestyan enters onto the database what he has transcribed by hand from the al-Najjārī inventory daftar in the archives: the title; that it is printed; the publication type (government – mīrī or external – barrānī); the value; and any other available data. This is done in Arabic and in ALA-LC/English. Each title is given an identification number which runs in accordance to its appearance in the inventory, and the information corresponding to each title flows across one row. Rezk Nouri then double-checks this preliminary data.

This is the extent of our rough data. The “title” is the usually two to three words that Muḥammad Qindīl, the assessor (kashshāf) of the books in al-Najjārī’s estate, indicated as a title, to which he rendered a particular monetary value in the original inventory.

In a few cases, the “title” is actually a conglomeration of titles – usually smaller essays and treatises – which were printed and likely bound together. Thus Muḥammad Qindīl added only one value for this grouping. In this case, a master identification number is provided to the whole conglomeration, and separate, sub-Google Sheets are created in which the titles are disaggregated into separate rows.

Let us move forward with the example of a straightforward “title,” as opposed to a conglomeration of titles. We will take up the example of the third entry in the daftar inventory, “Maʻāhid al-Tanṣīṣ.” Mestyan numbered this title as n1t3. He copied “Maʻāhid al-Tanṣīṣ”, that its value is 70 qirsh, that it is printed, and that it is mīrī.

I.2. “The work”

In the next phase, Mestyan, Schwartz, and Swanick each take take over identifying these titles, based on agreed pre-assignment. 

The one in charge of a title attempts to identify the work, which best corresponds to the title data from the inventory. We do this through various searches in online and printed catalogues (see Sources); and based on our knowledge in Arabic literature and science.

Our example “Maʻāhid al-Tanṣīṣ” is easy since there is only one work with this title: Maʻāhid al-Tanṣīṣ, a literary commentary on an earlier work by ʻAbd al-Raḥīm ibn ʻAbd al-Raḥmān ibn Aḥmad al-ʻAbbāsī, a sixteenth-century author (d. hijrī 963).

I.3. “The edition”

Once the one in charge identifies the work, that researcher next tries to identify the printed copies which 1) correspond to the known bibliographical information of “the title,” and 2) were printed before 1870, the year when al-Najjārī died.

Based on these criteria, the researcher may find two to three, or even more, printed editions of one work. 

I.3.1 Absolute certainty

If there is only one printed edition, and/or only one that corresponds to the inventory title data (for instance, a mīrī print), the researcher then tries to find a physical copy of the text to observe, or a digitally-posted scanned copy of the text. They then enter into the Google Sheet database a large number of bibliographical and extra-bibliographical features of that edition in English/ALA-LC transliteration (see Categories).

In the case of our example, there is only one edition of Maʻāhid al-Tanṣīṣ produced from the Egyptian government printing press before 1870: Sharḥ Shawāhid al-Talkhīṣ al-Musammā Maʻāhid al-Tanṣīṣ. Following the scanned copy of this edition from 1857 (hijrī 1274), they hand-gather the various features which they then enter into the database.

In this case, we are absolutely certain that a copy of this particular edition was among al-Najjārī’s books.

I.3.2 Relative certainty

If there are more printed editions of the title before 1870, and the researcher in charge of the title cannot identify the particular edition based on the features from the inventory daftar, the researcher then collects these editions.

In this case, we allow a degree of arbitrariness, and the researcher selects a printed edition based on its printing date closest to al-Najjārī’s death. This is a weak criterion, so we call this degree of certainty “relative certainty.” 

For instance, if there had been 1846, 1851, and 1866 government press editions of Maʻāhid al-Tanṣīṣ, we would have only gathered information from the 1866 edition in our database.

However, in the remarks section associated with the identification number, we often indicate the other available editions of the work.

I.3.3 Inaccessible copy, identification problems with title/work/edition

We did not enter into our database any description of a printed work for which we had no physical or scanned access, or for which we have ongoing identification problems. However, in the remarks section, we provide our ideas and best guesses.


Category Remarks: Printed Books

Mestyan & Kathryn A. Schwartz

The references to printed books in al-Najjari’s 1870 inventory are only titles, written in shorthand, with notes about whether they were made by Egyptian government presses or “external” presses, how many volumes the inventory-maker counted, and their appraised value. But within and beyond each book’s entry lies further information about the wider social context in which it emerged and circulated. 

Our search to identify a digitized or physical copy of a particular edition, which was with “absolute or relative certainty” part of al-Najjari’s collection, enables us to explore this wider context. We do this by describing the meta-data of the absolute or relatively certain edition in categories that we devised. 

This is time-consuming work! But by enhancing the inventory’s database through these hand-gathered categories, we seek to uncover networks, patterns, and trends about Cairene print culture during this foundational era. For example, did individuals specialize into separate roles for bringing books to press? Did printed books increase or decrease in value over time? And which factors influenced the original sale price of a book, such as the publisher, the type of technology used to publish it, and the number of pages it contained? 

To begin to answer these questions, and others, whenever we consult a given text, we enter information about it into 64 columns of categories in our database. Of them, roughly half are devoted to recording the same data two ways: one in the original language, and the other in ALA-LC (American Library Association – Library of Congress) standards of transliteration, to ultimately support building a website that will function in Arabic and English. Some columns carry transcriptions of the information logged in al-Najjari’s inventory. Others keep track of the library which houses the edition in question, where relevant, or of the scanned PDFs that we consulted on digital library websites.  

But in what follows, we elaborate on some of the more interesting enhanced categories that we collect data for: 

  • Title on copy 

(ALA-LC title of C)

In most cases, the title listed in the inventory is shorthand for a more expansive title. We therefore record the full title listed on the copy of the text that we consult. 

For example, the title listed in the inventory for one book (n1t18) is: Sirr al-Layālī. The digital copy of this book that we discovered online features its full title, Sirr al-Layāl fī al-Qalb wa-l-Ibdāl, so we also record this full title. 

  • Author of copy 

(Author of C in ALA-LC)

The inventory does not list the authors of books, unless their names happen to appear in the title, so we gather data for this category. 

‘Authorship,’ as a concept, is difficult to pin down. Firstly, the possibility of living from the fruits of one’s pen was not something that had taken hold in the Ottoman Empire by 1870. Instead, people presented themselves as composers of texts. Secondly, one could take on several roles in producing a text that could arguably be authorial – like, for example, composing the introduction to the book, editing it, or contributing commentary. 

To satisfy our bibliographic concern for collecting all names relevant to authorship, and the competing pressure not over-populate any given entry, we take authorship to mean the primary composer(s). In cases when secondary composer(s) exist, we include them under the author category if, and only if, their names do not fit into any of our other categories. 

For example, the (n1t58) 1847 Tarjamat al-Julistān al-Fārisī al-ʻIbārah al-Mushīr ilā Maḥāsin al-Ādāb bi-Alṭaf Ishārah has, as its original composer in Persian, al-Saʻdī. However, the introduction and colophon of the book make clear that the text’s translator into Arabic, Jabrāʼīl ibn Yūsuf al-Mukhallaʻ, also introduced and elaborated on al-Saʻdī’s original text. Because we can list Jabrāʼīl ibn Yūsuf al-Mukhallaʻ as the translator for this text, we do not include his name under the author category. 

An exception to this arises with regard to any book listed as a sharḥ. Here, we include under the author category the name of both the original composer, and the exegete. 

Another exception arises when we locate an edition that may, for one reason or another, contain one or more texts printed in the margins of the title that we are concerned with. In such cases, we include under the author category the names of the composers of all texts. 

  • Translator of copy 

(Translator in ALA-LC)

We collect the names of translators of books who are explicitly listed as such, most often by being recognized as the text’s mutarjim.

We take this policy as a practical measure, so as not to over-populate any given entry. As an intellectual matter, there are many cases in which it is difficult to separate translating from authoring.

Following the example above, of (n1t58) Tarjamat al-Julistān al-Fārisī al-ʻIbārah al-Mushīr ilā Maḥāsin al-Ādāb bi-Alṭaf Ishārah, we include Jabrāʼīl ibn Yūsuf al-Mukhallaʻ’s name under the translator category alone. 

  • Official sale value of copy 

(Official Sale Value of C in English (Qirsh))

On rare occasions, books include a printed note about their official sale value. We collect this data in qirsh whenever it is available. 

After we finish collecting data from the books themselves, we will also turn to book catalogs produced by publishers and orientalists from the 1830s through to the 1870s, to log the information listed within these other sources about the official sale value of books.

  • Source of official sale value of copy 

(Source of Official Sale Value in English)

Whenever we obtain data on the official sale value of a given edition, we also record the source of this data. 

  • Paid price of copy 

(Paid Price of C in English)

On rare occasions, we have evidence of a copy’s sale. Some books, for instance, include a handwritten note about how much someone paid for the book when they acquired it. We collect this data in qirsh whenever it is available. 

  • Place of publication of copy 

(Place of Publication in C in ALA-LC)

We record where the books were published, at the level of detail offered by the book itself. Usually this is at the city level, although sometimes we are able to record the street where the press was located. 

On occasion, we use our own judgment to name the place of publication when we feel confident about it, even when no place is listed explicitly. We place these judgments in brackets. 

  • Press name of publisher of copy 

(Printer of C in ALA-LC)

We record the name of the press that printed a given edition as it is listed. 

Oftentimes, across different books, the same press could go by a range of names. The Bulaq Press, for example, might go by the names: Maṭba‘at Būlāq, Dār al-Ṭibā‘ah al-‘Āmirah, Dār al-Ṭibāʻah al-Miṣrīyah, or Al-Maṭba‘ah al-Kubrā, among others. 

  • Date of publication of copy, in ḥijrī and mīlādī 

(Date of publication of C in Hijrī years) 

(Date of publication of C in Gregorian years)

Books printed from Cairo often list the lunar month in which they were printed, in addition to the ḥijrī year. We record this information. We also use it for conversions to the mīlādī year, which is calculated on the ‘Philosophia Islamica’ website, before we record this too in the data set. 

Sometimes, the ḥijrī date listed in a book corresponds to what could be either the end or the beginning of a mīlādī year. When it is impossible to identify the precise mīlādī year, we list the two mīlādī years separated by a ‘/.’ 

When a given title was published over multiple volumes, spanning several years, we list the range of the mīlādī years separated by a ‘-.’

  • Whether this title was in the Khedival Library catalogues of 1872 and 1893 

(Data in K1 or K2)

To better understand the nature of al-Najjari’s collection, and the comprehensiveness of the Khedival Library which was established in 1870, we will cross-check al-Najjari’s inventory against early printed catalogues from the Khedival Library. 

  • Seals of owners in copy 

(Seals of Owners)

Whenever a book bears the seal of a contemporary owner, or the stamp of a later owner of historical consequence, like an important scholar, we attempt to transcribe it, and we also capture its image. 

  • Inscriptions of ownership in copy 

(Inscriptions of Ownerships)

When there is a contemporary ownership inscription in a book, we attempt to transcribe it, and we also capture its image. 

  • Taqrīz/Tārīkh composer in copy 

(Taqrīẓ/tārīkh author)

Usually at the end of a book, and sometimes at the beginning, there is/are blurbs of endorsement for the text, and/or chronograms composed on the occasion of its publication. We record the names of those who composed them. 

  • Taqriz/Tarikh date in copy 

(Taqrīẓ/tārīkh date)

We also record the date given for such compositions.  

  • Individual identified in copy as multazim/‘alā dhimmah/‘alā nafaqah 

(Individual multazim / ‘ala dhimmah/nafaqah in ALA-LC)

Many of the books printed in Cairo during this period name, in their colophons, those who were in a business, and/or contractual relationship with the press. They might have initiated the production of, and/or provided the funding for, the printing of a given title. The names of these figures usually follow the designation of ‘multazim,’ or appear after the expression ‘‘alā dhimmah’ or ‘‘alā nafaqah.’ We record these names. 

  • Whether the copy is a lithograph 


Al-Najjari’s inventory makes no mention of whether a title was printed lithographically, or typographically. We therefore note this information in our database, upon consulting the edition under consideration. 

When the text of a book is printed typographically, but it also includes lithographic images or engravings, we also make note of that. 

  • Copyist of copy 

(Name of copyist in ALA-LC)

When the edition is a lithography, it sometimes lists the name of the person who copied the text out by hand. We record that person’s name. 

  • Corrector of copy 

(Corrector in ALA-LC)

The category of corrector, like that of author, is difficult to pin down, as one could serve the role of corrector for a printing in a number of capacities, like correcting a printing for typos, or cross-checking the printing against manuscripts. 

We simply record the name(s) of the edition’s corrector(s), when listed explicitly, most often via the term muṣaḥḥiḥ.

  • Range of numbered pages in copy 

(Number of paginated pages)

We record the range of the numbered pages of the edition. That is, the range between the first page number listed and the last page number listed. 

Sometimes, the numbering begins from the fihrist at the beginning of the book, and then begins all over again at the start of the text proper. In such cases, we start our range with the fihrist, and effectively ignore the recommenced count at the start of the text proper. We do this because our goal is not perfect accuracy in capturing a given edition’s page count – indeed, many front and end pages are unnumbered, and the first and last pages of a given copy are most vulnerable to being missing – but rather, we wish to gain a rough sense of the copy’s size. 

When a copy runs across multiple volumes, we record the page range for each volume consulted.