Online Ajami Sources and Resources

Where can one find African languages written in Arabic script on the internet?

One need not go to West African or anywhere else to contribute to research of manuscripts that contain instances of African languages written in Ajami (that is, Arabic characters).

What follows is a list of the most prominent collection of Ajami texts and or resources available on the internet that I could find.

NOTE: An easier-to-read and open-access PDF of this post is available here.

Selected Collections in the West

Bibliothèque Nationale de France (BNF)

Manuscrits Arabes

Inventory of the Umarian Library (AKA Fonds Archinard, which includes Umar Tal’s collection from 1890):

Catalogue des manuscrits arabes du fonds de Gironcourt (1878-1960):

Arabic Manuscripts from West Africa: A Catalog of the Herskovits Library Collection at Northwestern

Guide to collection (newer introduction than the above link and includes a link to search their catalogue): Most in Arabic. Some Hausa, Fulfulde, Wolof, Dagbani and Gonja. From Northern Ghana, Mali, Nigeria and Senegal in late 19th and early 20th centuries. Can be searched but not viewed online.

The ‘Umar Falke Collection

The John Paden Collection

The John Hunwick Collection

The University of Ghana Collection

Collected by Wilks. Photocopies of what he deposited at University of Ghana and used for books on Gonja and Wala in northwestern Ghana.

Selected Collections on the Internet

British Library’s Endangered Archive Program (EAP)

Digital preservation of Wolof Ajami manuscripts of Senegal (EAP334)

Same as BU Ajami Library collection for Wolof. Includes Quran that Jeremy investigated in his article.

Digital Preservation of Mandinka Ajami Materials of Casamance, Senegal (EAP1042; same as BU Mandinka)

Only significant collection of Manding manuscripts that is currently accessible online.

Safeguarding Fulfulde ajami manuscripts of Nigerian Jihad poetry by Usman dan Fodio (1754-1817) and contemporaries (EAP387)

Digital Preservation of Mandinka Ajami Materials of Casamance, Senegal (EAP1042)

Project to digitise and preserve the manuscripts of Djenné and surrounding villages (EAP690)

Trinity College Dublin

Use the digital collections link to search for “Arabic manuscript” and then find the following shelf marks which are online and include Soninke and Mandinka Ajami as identified by Darya: IE TCD MS 3499, 3500, 2179, 2689.

They also have mss with Ajami Hausa: e.g. 2258 and 3335.

Leiden University

Leiden University Library, ms OR. 14.052 (with glosses in Soninke) (

Bibliothèque Universitaire des Langues et Civilisations (BULAC)

Manuscrits arabes (

Twenty-seven digitized manuscripts.

MS.ARA.610bis: Fulani Ajami (names of plants)

African Languages Material Archive (ALMA)

Brief introduction and then exclusively Wolof documents from Ngom’s documentation efforts. Not clear how distinct from the other stuff of his that is out there.

African Online Digital Library (AODL)

Diversity and Tolerance in the Islam of West Africa (

Ajami in the Senegambia

Profiles by Ngom of different Ajami scholars along with texts, transliterations and media files.

Northern Factors in Asante History

Some Wangara interviews and pictures connected to Bobo-Dioulasso and Marhaba of Kong.

Qadiri Community of Buh Kunta

No Ajami

Pluralism and Adaptation in the Islamic Practice of Senegal and Ghana (

A continuation of the “Diversity and Tolerance” project listed above; Includes Ajami documents that can be searched via browse)

Discourses of Muslim Scholars in Colonial Ghana

This gallery consists of English translations of Arabic and Hausa texts written by Al-hajj Umar ibn Bakri, a West African Muslim scholar of the late nineteenth and early twentieth centuries. Al-hajj Umar was a member of a Hausa Muslim family from Kano, Nigeria, who resided most of his adult life in Kete-Krachi, a town in what would be in today’s Ghana. His poems, considered some of best examples of Arabic expression in West Africa in this era, comment on European colonial rule, the activities of Muslim leaders, and other topics.

Includes file with approximate translation

Failed Islamic States in Senegambia

English translation of verses from ‘ajami poem of Mohammadou Aliou Tyam celebrating the triumph over and entry into the Bambara capital of Segu in 1861.

West African Online Digital Library

Galleries of photographs, audio interviews, and documents contributed by scholars, with research about best practices for international, multilingual digital repositories. List of galleries with descriptions:


Includes African language interviews and translations that Curtin did.


Potentially interesting photos of typical Jula style mosques etc.

The history and culture of Futa Toro, Senegal and Mauritania:

Africa’s Sources of Knowledge – Digital Library (ASK-DL) at Harvard

Includes Ajami materials for Fulani, Hausa and Swahili. “Bamanankan” is only in N’ko.

West African Arabic Manuscript Database

Started by Charles Stewart

19,000 manuscripts from eleven different collections

Collaboration with Al-Furqan Islamic Heritage Foundation

Used to catalogue Northwestern collections

Archive of the Republic of Niger (AREN) at Boston University

This collection includes some Ajami material, including children’s books in Hausa. While cataloging of this collection is a work in progress, a partial listing of books is available here (

The list lists works that are Arabic or Ajami—could search and put into a post

African Ajami Library at Boston University

Often (or always?) overlaps with the EAP collections

Mandinka Ajami Manuscripts (In Progress) [41]

Fuuta Jalon Pular Ajami Manuscripts [25]

Hausa Manuscripts [6]

Philippe Beaujard Sorabe (Malagasy) Ajami Collection [69]

Wolofal Manuscripts, Senegal [27]

The Oriental Manuscript Projet (OMAR)

Digital repository at University of Freiburg in collaboration with University of Tubingen

2,500 Arabic manuscripts from Mauritania

Digitized from microfilms collected between 1979 and 1997

Library of Congress

2003 digital copies from the Mamma Haidara Library and Cheikh Zayni Baye Library in Boujheba, Mali.

Hill Museum and Manuscripts Library

Sign up for account to view for free

Digital copies of Savama-DCI from between 2013-2016

As of November 6, 2018: 7612 digitized but not catalogued from Aboubacar Bin Said Library; 20628 in process of being digitzed from the Mamma Haidara Library

As of November 6, 2018, 1471 from Mamma Haidara Library (2017-02-24) and 1492 from Aboubacar Bin Said Library (2017-07-31) are digitized, catalogued and viewable.

SOAS Digital Collections

Hausa Language Resources

Includes a History of Samory

Borno and Old Kanembu Islamic Manuscripts


Profiles in Nigerien Artisanry: The Artists of Niamey

The Niger Artisans Project consists of 12 candid conversations with artisans working in Niamey, Niger. The artisans featured here fall into a category more closely related to the corresponding Hausa term ma’aikatan hannu, or people who work with their hands. Each interview includes a transcription in Ajami.

Ajami Texts in BU Libraries

A page that includes bibliographies about BU’s holdings related to Ajami for the following languages: Swahili, Hausa, Kanuri, Joola, Fulani as well as a general Ajami references.

BU Library Guide

BU Library Guide

A guide for researchers and students that lists potential Ajami resources that exist online.

Illinois at Urbana-Champaign Library Guide

“This guide was created to help researchers locate manuscripts originating from West Africa, written in Arabic or in Ajami – a modified Arabic script used to write African languages. Here you will find information on our holdings here at the University of Illinois; electronic resources available for use; as well as archival holdings both in the United States and abroad.”

Super detailed guide covering manuscripts at Illinois, Electronic Resources and a brief overview of physical archives in West Africa and elsewhere.

Ngom 2017 Article

Ngom, F. (2017). West African Manuscripts in Arabic and African Languages and Digital Preservation. In Oxford Research Encyclopedia of African History.

Expansive article that tries to lay out the history and current state of West African manuscripts.

Connections between N’ko and Ajami

Today, one is more likely to come across Manding written in N’ko than in Arabic (or even Latin) script. But did N’ko stifle Ajami or emerge from it?

Robust traditions of Ajami exist and have been relatively well documented and studied for major West African languages such as Wolof, Hausa, and Fulani. It surprising then that thus far there have been few Ajami texts found for major Manding varieties such as Bamanan, Maninka and Jula [1]. Today, in many Manding-speaking areas, one is more likely to come across the language written in N’ko than in Arabic (or even Latin) script.

Map of the Manding language continuum.png
By Coleman DonaldsonOwn work, CC BY-SA 4.0, Link

In the strictest sense, “N’ko” (ߒߞߏ) refers to a non-Latin-, non-Arabic-based script invented in 1949 by a young man named Sulemaana Kantè schooled in the Quranic tradition of his father. Today, it also denotes a broader grassroots social movement in West Africa that labors to promote Kantè’s alphabet for both Manding and African language literacy and education in general. Without any formal support from regional governments or international donors, N’ko literacy has continually spread through dedicated individuals behind local bookshops, radio shows and study groups across the region. They annually produce hundreds of texts in the form of books, newspapers and websites–far more, than what is written in state-backed Latin orthographies for Manding. N’ko’s success thus can potentially be said to have in part stymied any potential modern-day continuation or expansion of Manding Ajami literacy.

Sticker of Sulemaana Kantè with his name written below him in N’ko on a laptop in Wes Africa (Photo credit: Coleman Donaldson)

At the same time, N’ko’s printed texts of today have also preserved at least one concrete connection between itself and Ajami as distinct Manding orthographic traditions. One relatively popular N’ko text today is entitled “Five big questions and their answer.” It is comprised of a transcribed interview (or series of interviews) conducted with Kantè between 1968-1969. In one excerpt he explains his first encounter with written Manding. In 1941 he was presented with a Manding-language history text of the ethnic Fulani of the region, which had been penned in the Arabic script together by his grandfather and maternal uncle. Unable to read the document, despite his native-speaker skills and Quranic education, Kantè was not impressed. In response to his uncle’s remark that it was more of memory-jogging device for the author than a stand-alone text, he recalled thinking:

ߒ ߞߊ߲߫ ߒ ߞߏ߫ ߝߘߊ߬ߝߌ߲߬ ߞߊ߲ ߕߍ߫ ߝߏߦߌ߬ ߘߌ߫ ߸ ߓߊ ߊߟߊ߫ ߡߊ߫ ߛߏ߲߬ ߊ߬ ߛߓߍ߫ ߞߏ ߡߊ߬

“I said African languages are worthless because God has not allowed for them to be written”

Kantè, of course, would later change his thinking on the matter. Ultimately, what hampered his family’s orthographic tinkering and later inspired his own unique solution was the difficulty of representing all of Manding’s contrastive sounds within the graphic inventory of the Arabic script. Using only three short vowel markers and three long vowel markers, Kantè’s uncle and grandfather struggled to devise a system of conventions to represent their native Maninka’s seven vowels with contrastive lengthening, nasalization and tone.

Regardless, Kantè’s N’ko thus points us to the existence of Maninka Ajami manuscripts in the early twentieth century. During the summer of 2016, while doing fieldwork in Kankan, I was shown a photocopy of a Maninka text that presumably stemmed from the generation just before the emergence of Kantè’s N’ko. In the future of our Ajami Lab work, I hope to return to speak with friends and acquaintances to look into the existence of such early twentieth century Ajami documents.

But what of earlier Ajami writings in Maninka? According to the N’ko intellectual and activist Condé (2008), Kantè considered himself an heir to work of Alfa Mahmud Kaba. A skilled leader of the Batè region (of which Kankan is part) of upper Guinea in the mid-nineteenth century, Kaba was also a man of letters who is today remember for his Arabic works and his (presumably oral) translation of Islamic poems into Manding. Kantè was also influenced by a friend of his father, “Diakagbe Taliby Kaba” (Jakagbɛ Talibi Kaba), who also worked to translate Islamic rites and poems into Manding. In fact, he is purported to have devised a writing system for Manding as part of his efforts. Exploring the history of such scholars and any manuscripts that they left behind is part of our work in the Ajami Lab too.

In both cases, such connections invite us to think that it’s not necessarily that N’ko hampered Ajami, but rather that that Ajami gave rise to N’ko.

[1] The exception is Mandinka, spoken at the western edge of the language-dialect continuum. In this case, there are at least two major scholarly works investigating local history through Ajami texts (Giesing & Vydrin 2007; Schaffer 2003).


Condé, Ibrahima. 2008. “The Emergence of N’kophone Literature: From Poetry to the Novel.” Mande Studies 10: 133–48.

Giesing, Cornelia, and V. F Vydrin. 2007. Ta:Rikh Mandinka de Bijini (Guinée-Bissau): La Mémoire Des Mandinka et Sòoninkee Du Kaabu. Leiden; Boston: Brill.

Kántɛ, Sùlemáana. 1968. ߢߌ߬ߣߌ߲߬ߞߊ߬ߟߌ߬ߓߊ ߅ ߣߴߊ߬ ߖߊ߯ߓߌ [Five Big Questions and Their Answer]. Edited by Bàbá Màmádi Jàanɛ.
Schaffer, Matthew. 2003. Djinns, Stars and Warriors: Mandinka Legends from Pakao, Senegal. African Sources for African History 5. Leiden: Brill.

The Revelation: The earliest known use of ʿajamī as a label

The interlinear glosses written in African languages of West Africa often have a descriptive metalinguistic marker, asserting the presence of an Ajami segment.

Our blog has the acronym A-LABEL: we deal a lot with the African languages between the lines. But why “label”? This is because the interlinear glosses written in African languages of West Africa often have a descriptive metalinguistic marker, asserting the presence of an Ajami segment. This marker is typically a graphic variation of the word ʿajamī (عجمي – the form written on our logo) or ʿajam. Another way of the metalinguistic labelling of the Ajami glosses found in the manuscripts is the Arabic phrase fī kalāminā ‘in our language’ (في كلامنا) and its variations (for the use of ʿajamī and fī kalāminā in Soninke and other Mande manuscripts see Ogorodnikova 2017: 122-125).[1]

The earliest Soninke glosses marked with the word ʿajamī are found in the manuscripts written around the early nineteenth century (Darya Ogorodnikova is exploring this question in her PhD thesis expected to see the light of day this summer).   

In Hausa Ajami writing, various stylised forms of ʿajam(ī) were frequently used (Dobronravin 2013: 91) and until last week we could date it to the first half of the nineteenth century, as in the manuscript below, where the word ʿajam (encircled) is written in unconnected letters ʿayn, jīm and mīm:  

Shurūt al-ʿilm ‘The prerequisites of knowledge’. Kano, MS.Murtala. Before (?) mid-19th century.

A week ago, Nikolay Dobronravin discovered Hausa Ajami in a dated manuscript of Maqāmāt al-Ḥarīrī possibly copied in 1166H/1752CE (or 1177H/1763CE, depending on interpretation of the numbers in the colophon).[2] If it is indeed mid-eighteenth century, this would make it the earliest dated Hausa glosses and also the earliest dated label ʿajamī in the Hausa texts. You can see two tokens of ʿajam at the bottom of the page:

Maqāmāt al-Ḥarīrī, BNF/Gallica ark:/12148/btv1b53151891k. Colophon: 1752 or 1763.

Moving further east, the manuscripts with Old Kanembu glosses are generally believed to lack the ʿajamī label. I myself have previously stated as much (see Dobronravin 2013: 91, with a reference to Bondarev). I was wrong.

This conviction of mine was demolished a year ago while I re-investigated the famous dated Borno Qur’an, 1080H/1669CE (Bivar 1960, Bondarev 2006a, 2014). Previously, I had never noticed any instances of the word ʿajamī. Overlooking a word or a whole sentence in this and many other manuscripts is not so shocking – an average page of this 300-folio Qu’ran requires hours and days to make sense of all connections between the main text and surrounding commentaries. In any case, while looking for an Old Kanembu word in Sūrat al-Ahqāf (Q.46), I stumbled upon this:

Qur’an. CSMC, Hamburg, MS.3ImI. Colophon: 1669.

Circled in yellow is our ʿajamī (or ʿajam) in a stylised writing, with the “tail” of the letter mīm happily wagging left rather than submissively hanging down (as in common styles of writing). And this single occurrence of ʿajamī makes it the oldest known case of labelling non-Arabic writing. I am convinced that it was written before 1669 since we know that the Old Kanembu glosses were added to the main Qur’an text first, while the voluminous Arabic commentary of al-Qurṭubī, which wraps around the main text, was done afterwards, in 1080/1669 (see Bondarev 2017: 119-122, 2019).

When I first noticed this ʿajam(ī) I did not understand the meaning of the marked word, krtsāk, and I did not make a connection between this Old Kanembu word and the Qur’anic phrase tanzīl al-kitābi ‘the revelation of the Book (the Qur’an)’. I even fell in the fallacy of chance resemblance, trying to see in the word krtsāk the Ancient Greek χᾰ́ρτης (khártēs) ‘papyrus scroll; book’.  The Greek khártēs did actually make its way into Arabic and became qirṭās to designate papyrus, parchment or paper (Gacek 2001: 114, Gacek 2009). Given the antiquity of Islam in Kanem-Borno (see a succinct overview in Bondarev 2006b), it would not be implausible that the term for ‘writing surface’ was borrowed from Arabic sometime in the early centuries of Islam. Even more so, it would make sense to look more closely at the Old Kanembu krtsāk because in the neighbouring Hausa language the word for paper is takàr̃dā and it is indeed related to the Greek khártēs: it was borrowed from Berber (t(a)– being the feminine prefix in Berber) which in turn was borrowed from Latin charta (Kossmann 2005:77), the loanword from Ancient Greek khártēs.

Well, a closer look revealed to me the revelation. Quite literally. Because Old Kanembu krtsāk is not a translation of the Qur’anic al-kitābi (of the Book) but of the deverbal noun tanzīl ‘the revelation’. This Arabic noun derives from stem II verb nazzala ‘send down; reveal’. A comparison of the same verse of the Qur’an (Q.46:2) in the other manuscripts dismantled the feintest traces of my χᾰ́ρτης fantasy. For example, the manuscript MS.Tahir.Kano clearly tells us that <alqurānka jrksāktī allah kan kaṯīkō> ‘the revelation of the Qur’an is from God’. The item marked in bold is the same deverbal noun as krtsāk in the dated Qur’an MS.3ImI. Both nouns share the same verb root sāk ‘descend’. So it turned out that to translate the Qur’anic notion of ‘sending down, revealing’, Old Kanembu uses an applicative form of the verb sāk which is irgsāk ‘to send down, bring down’. This form is then nominalised either with a prefix k– as in k-rtsākh or with a prefix nz– (written as <j>) as in j-rksāk. The range of graphemic variation of this deverbal noun is great in Old Kanembu manuscripts (the factors being both orthographic and morphological). But whatever the variation, we have the winner for the earliest non-Arabic word in West Africa marked with the label ʿajamī and this word is Old Kanembu: krtsāk ‘revelation’.

There is one unsettling question: why only this instance of the word is marked with ʿajamī? Why no other verses but for Q.46:2 have it? I have checked all occurrences of Qur’anic tanzīl in the dated MS.3ImI. Thus, in 3ImI/Q.26:192 it is krtsak, in 3ImI/39:1 ktsak, in 3ImI/ Q.41:1 kitsāk and in 3ImI/ Q.69:43 kktsak. But none is marked as ʿajamī. I also looked up in the other places of MS.3ImI where tanzīl is mentioned – nothing: no krtsākh no ʿajamī.[3]   

And while diving again into my manuscript data in search of words for ‘descend’, I still could not find any other occurrence of ʿajamī in the other manuscripts. It might well be that ʿajamī in MS.3ImI Q.46:2 is a unique instance of this term – an exception confirming the rule and validating my earlier assumption about the lack of ʿajamī in Old Kanembu manuscripts.   

As for the words ‘paper’ and ‘book’ in Old Kanembu, I’ll leave it for another post.

[1] The paper of the manuscript looks almost a century younger! Manuscripts were often copied with the previous copyists’ colophons and we need to examine the paper first in such suspicious cases. Many thanks to Darya Ogorodnikova for having raised this question.

[2] Our Ajami Lab team is preparing a post on various types of such “labels” across West African manuscripts.

[3] These verses are as follows: MS.3ImI/Q.20:4, 32:2, 36:5, 40:1, 41:42, 45:2, and 56:80.


Bivar, Adrian David H. (1960): “A Dated Kuran from Borno”. In: Nigeria Magazin, 65: 199–205.

Bondarev, Dmitry (2006a): “The language of the glosses in the Bornu Qur’anic manuscripts”. In: Bulletin of SOAS 69, 1, 113-140.

Bondarev, Dmitry (2006b): “An archaic form of Kanuri/Kanembu: a translation tool for Qur’anic studies”. In: Journal of Qur’anic Studies 8, 2, 142-154.

Bondarev, Dmitry (2014): “Multiglossia in West African Manuscripts: A Case of Borno, Nigeria”. In: Jörg B. Quenzer, Dmitry Bondarev and Jan-Ulrich Sobisch (eds): Manuscript Cultures: Mapping the Field. (Studies in Manuscript Cultures, 1). Berlin, Boston: De Gruyter, 113–155.

Bondarev, Dmitry (2017): “Islamic Education and Ample Space Layout in West African Manuscripts”. In: Andrea Brigaglia and Mauro Nobili (eds): The Arts and Crafts of Literacy: Manuscript Cultures in Muslim sub-Saharan Africa (Studies in Manuscript Cultures, 12). Berlin, Boston: De Gruyter. 105–142.

Bondarev, Dmitry (2019): “Tafsīr Sources in Four Annotated Qur’anic Manuscripts from Ancient Borno”. In: in Zulfikar Hirji (ed.), Approaches to the Qur’an in Sub-Saharan Africa. Oxford/London: OUP & IIS.

Dobronravin, Nikolay (2013): “‘Classical Hausa’ Glosses in a Nineteenth-Century Qur’anic Manuscripts: A Case of ‘Translational Reading’ in Sudanic Africa? In: Journal of Qur’anic Studies, 15, 3, 84-122.

Gacek, Adam (2001): The Arabic Manuscript Tradition. A Glossary of Technical Terms and Bibliography (Handbook of Oriental Studies, 58). Leiden, Boston, Cologne: Brill.

Gacek, Adam (2009): Arabic Manuscripts: A Vademecum for Readers. Leiden: Brill.

Kossmann, Maarten (2005): Berber Loanwords in Hausa (Berber Studies, vol. 12). Köln: Rüdiger Köppe.

Ogorodnikova, Darya (2017): “ʿAjamī Annotations in Multilingual Manuscripts from Mande Speaking Areas: Visual and Linguistic Features”. In: Islamic Africa, 8, 111–143.

If you would like to cite the source:  Dmitry Bondarev (2019, April 1). The Revelation: The earliest known use of ʿajamī as a label.

Fur Ajami: an invisible tradition

More on “invisible” Ajami from our colleague Nikolay Dobronravin who already wrote for our Ajami Lab a piece on invisibility of Ajami writing in Nigeria, Chad and Sudan. This time he takes us to Darfur (‘the land of the Fur’) to explore a history of Ajami writing in Fur, a Nilo-Sahran language spoken in the west of the Republic of the Sudan.   

Read his fascinating account of Fur Ajami: The Fur (For, Fūranq bilī, Poor’íŋ belé) variety of Ajami: an invisible tradition.

Following Nikolay’s lead outlined in the above paper, I am heading to London next week to consult Kītāb Dālī , the mysterious Fur law book. Who knows, maybe we are just a short flight from another Ajami discovery.  

Transliterating Ajami into the Latin script

The conventions we use to transliterate African languages from Arabic to Latin script.

One of the hardest things about working on Ajami manuscripts is interpreting the African language that is written in the Arabic script. This is because oftentimes certain contrastive sounds (phonemes, nasalization, tone) are not marked.

To get from the original Arabic to a proper interpretation of the African language, we start by typesetting the Arabic orthographic representation; that is, we type out what is written by hand. Next, we transliterate that Arabic representation into the Latin script.

Of course, we aren’t done then — we still need to attempt to interpret the transliteration into a practical (for us) Latin-based orthography appropriate for the language in question and offer a translation — but for now we’ll stop there because we wanted to announce that we’ve written and posted online the transliteration scheme that we are currently using. We’ve added it to our general Tools & Resources page which basically outlines our whole workflow at this point.

Hope it’s helpful for others working on Ajami manuscripts. Any questions, feedback or issues that you see, please let us know!


How to use Tropy to analyze Ajami manuscripts

Working through Ajami manuscripts can be slow-going. Here is a standard workflow using Tropy that can make it a little easier.

Working through Ajami manuscripts can be slow-going. One important tool in our workflow is the open-source software Tropy from the same people behind everyone’s favorite reference management software Zotero.

I recently uploaded two technical notes that explain our workflow for 1) annotating instances of Ajami in manuscripts and 2) extracting one’s typesetting, transliteration, interpretation and translation of any instances of Ajami.

Using this workflow, one can easily spit out all instances of Ajami into a range of convenient formats that ourselves and other researchers can then explore in whatever way best suits their own research interests into African languages, orthography, scripts, historical sources etc.

I’ve added both documents to our general “Tools and Resources” page as well. Let us know if they’re helpful or something isn’t clear!

Updated abstracts and presentations from Ajami Lab Workshop, 31 Oct-1 Nov, 2018

Presentations and slides from the Ajami Lab’s Nov 2018 workshop


While preparing some new material on types of Ajami manuscripts and specific approaches to study them (based on this talk),  I post here an updated programme of the last year’s workshop. A big thank you to all participants and presenters! Here are the links to slide presentations (following the order of the talks in the programme):

Samby Khalil Magassouba, Bamako and Djibril Dramé, Hamburg
Analysis of a Soninke Ajami poem

Hamadou Boly, Bamako
Une étude analytique de la poésie du cheick Mouhammad Abdoulaye Souadou

Lameen Souag, Paris
Preliminary notes on a page of premodern Songhay and Tamasheq poetry from Timbuktu

Ismaila Zangou Barazi and Sidi Imam Miga, Bamako
La pensée lexique dans les manuscrits de Tomboctou (presentation in Arabic)

Darya Ogorodnikova, Hamburg
The scribes of the Soninke Islamic manuscripts and their networks

Djibril Dramé, Hamburg
Soninke of the Islamic Manuscripts: A comparative overview of a written register and modern (spoken) Soninke

Coleman Donaldson, Hamburg
Preliminary ethnographic and historical insights regarding language, literacy and scholarship amongst the Jula of the former Kong Empire

Nikolay Dobronravin, Saint Petersburg
(In)visibility of Ajami in Chad and Sudan and in Nigerian Android applications

Jula lexicographic data on Zenodo

Jula lexical data from Jan 2019

I spent most of January in western Burkina Faso doing research related to linguistic registers and Quranic education and scholarship in Jula-speaking areas.

Whenever, I’m doing fieldwork in the region, I’m always jotting down words that I encounter that strike me as unusual, at least for me that is. Oftentimes, they are quite common expressions, but ones that I hadn’t encountered in a while. Other times, they are words that I don’t recognize, but hope to look up later in various dictionaries. In some cases, they turn out to be words that are absent form such works and in other cases, they are words that I want to integrate into my own lexicographic works or general vocabulary.

In any case, I’ve tried to get more rigorous about how I keep track of all these jottings and so I’ve started typing them up regularly during fieldwork. That might help me remember, but more importantly perhaps, it allows me to produce a clean dataset record for both myself and future researchers (in the off-chance that I stumble across something worthwhile). So without further ado, I present to you a Zenodo record that anyone can consult and download in a range of formats for their own studies and lexicographic research related to Manding.

Would be great if such practices were standard for Peace Corps volunteers and other researchers that are constantly learning and working with languages that often times don’t have robust lexicographic work behind them.

Bonne lecture and don’t hesitate to reach out if you have questions about how I used Toolbox, FLEx or Lexique Pro to produce this record!

Donaldson, Coleman. (2019). Jula lexicographic data collected during January 2019 [Data set]. Zenodo.

Identifying languages of Ajami commentaries

Identifying an African language written in Arabic script may be problematic even for speakers of the language. It is especially challenging when the Ajami items are represented by isolated words in the margins of an Arabic manuscript. Moreover, if the Ajami item is not visually linked to the intended Arabic phrase or word, it is almost impossible to tell what the Ajami item means and in which language it is written.

It all becomes much easier if commentaries in Ajami are linked to their source Arabic content by way of placing the item exactly above (or exactly below) the Arabic word/phrase or by way of connecting the Arabic and Ajami counterparts with lines or by putting a mark signifying what connects to what (Bondarev 2013, 2014, 2017, Ogorodnikova 2017). Frequent use of function words in the Ajami commentaries further simplifies the task of language identification – to the extent that you don’t have to speak the language to correctly guess it! What one needs is a set of fixed correspondences between common Arabic and Ajami function items and frequent lexical items.  

The identification tables presented below have proved to work for non-speakers of the given languages.[1] The tables are still works-in-progress, and all feedback is very welcome.

  1. Ajami identification tables for Old Kanembu, Hausa, Fulfulde, Soninke (Bondarev, Dobronravin, Ogorodnikova)
  2. Ajami identification table for Hausa, Songhai, Soninke (Souag)



Bondarev, Dmitry (2017): “Islamic Education and Ample Space Layout in West African Manuscripts”. In: Andrea Brigaglia and Mauro Nobili (eds): The Arts and Crafts of Literacy: Manuscript Cultures in Muslim sub-Saharan Africa (Studies in Manuscript Cultures, 12). Berlin, Boston: De Gruyter. 105–142.

Bondarev, Dmitry (2014): “Multiglossia in West African Manuscripts: A Case of Borno, Nigeria”. In: Jörg B. Quenzer, Dmitry Bondarev and Jan-Ulrich Sobisch (eds): Manuscript Cultures: Mapping the Field. (Studies in Manuscript Cultures, 1), Berlin, Boston: De Gruyter, 113–155.

Bondarev, Dmitry (2013): “Qur’anic Exegesis in Old Kanembu: Linguistic Precision for Better Interpretation”, In: Tal Tamari and Dmitry Bondarev (eds): Qur’anic Exegesis in African Languages. Special Issue of the Journal of Qur’anic Studies. 15.3, 56–83.

Ogorodnikova, Darya (2017): “ʿAjamī Annotations in Multilingual Manuscripts from Mande Speaking Areas: Visual and Linguistic Features”. In: Islamic Africa, 8, 111–143.

[1] We tested these tables at various workshops at the Centre for the Study of Manuscript Cultures (CSMC), Universität Hamburg, and at two training workshops, as follows: (with Nikolay Dobronravin) “Preservation and cataloguing of Nigerian Islamic Manuscripts”, 24-28 July 2017, Kaduna, Nigeria. A training workshop funded by SFB 950, The Islamic Manuscript Associtation, Arewa House, Kaduna, with participation of Wesleyan Library, USA and St. Petersburg University, Russia; and (with Darya Ogorodnikova) “Working with African Arabic Script Manuscripts” 13-20 August 2017. A training workshop funded by Gerda Henkel Foundation, SFB 950, The Institute for the Study of Islamic Thought in Africa (ISITA), Northwestern University, and GHS. Bondarev and Ogorodnikova participated as lecturers.

Notes from the “Manuscripts and Development” Conference in Bamako

From November 16-17, 2018, myself and Dr. Maria Luisa Russo had the pleasure of attending and presenting at a conference on “Manuscripts and Development” put on by the Arabic Department of the University of Bamako.

Carried out primarily in French and Arabic, the conference had a number of interesting talks focused on various research initiatives being carried out in Mali and elsewhere. Here, I’ll highlight three that stuck out to me:

Dr. Goro N’DIAY (University of Bamako) did an interesting presentation entitled “Les vocabulaires thérapeutiques dans les manuscrits ajami : cas traitement des maladies mentales” that was well received by the audience. It was in Arabic so I couldn’t follow it all, but I recognized a bunch of Manding plant names in the midst of his presentation. Dr. N’Diay’s work is part of an initiative of the University of Bamako’s Arabic Department focused on medicinal plants listed in Ajami manuscripts.

Dr. Abdoul Karim HAMADOU, did an interesting talk in French on one or two Songhai-Arabic lexicons from the Ahmed Baba Institute that are found in MS27538 and MS2457. The date of the manuscripts is unknown but the author’s name is mentioned somewhere within the documents. Dr. Hamadou used some of Paul Marty’s works to make links to the potential author.

Dr. Hamadou presenting on Songhai-Arabic lexicon manuscripts

Dr. Hamadou BOLY did an interesting presentation on one (or two) of the only extant writings of Cheikh Ahmadu Lobbo (founder of the Macina Empire of the 19th century).

The conference also included an exhibit of select manuscripts from either SAVAMA-DCI or the Ahmed Baba Institute.

One of the manuscripts on display at the conference

Looking forward to reading future publications stemming from some this work. Thanks again to the Arabic Department and the University of Bamako for organizing the event!

New Technical Notes for Using Tropy to Analyze Manuscripts (en français)

We just released two working technical notes (in French) to help researchers get set up to use Tropy to analyze Ajami annotations/segments in West African manuscripts:

  1. Comment créer un premier projet Tropy « portable » pour une collection manuscrite
  2. Instructions pour créer un nouveau projet « portable » Tropy pour analyser une nouvelle collection/bibliothèque

We’re working on also putting out some other draft standards/workflows regarding how to use Tropy to “annotate” manuscript images that include instances of Ajami. More news to come!

Ajami Lab Workshop, 31 Oct-1 Nov, 2018

The Ajami Lab is pleased to announce its next workshop which will take place from 31 October to 1 November, 2018, at the Centre for the Study of Manuscript Cultures at the University of Hamburg.

Programme and abstracts available here.

This session is the first of the DFG-funded project “African voices in Islamic manuscripts from Mali: documenting and exploring African languages in Arabic script (Ajami)” and will be held in conjunction with the meeting of the “Old Mande Research Network.”

Welcome to A-label!

This blog is a companion to the research of the Ajami Lab within the Centre for the Study of Manuscript Cultures at the Universität Hamburg. Focused on historical and modern instances of Ajami—that is, sub-Saharan African languages written in Arabic script—we aim here to share updates, anecdotes, and conceptual and methodological tools as we work to catalogue and analyse African Islamic manuscripts. In doing so, the blog will serve both other scholars of Ajami as well as a wider audience of scholars interested in the insights that can be gleaned from these understudied documents and literacy practices.