Association for the Publication of African Historical Sources

Issue No. 9 - October 1997

History Department, Michigan State University
East Lansing, Michigan 48824

Note from David Robinson, Michigan State University, October 1997.

Finally you are getting the newsletter promised at the last ASA. Not having the patience or skill of John Reid, who left two years ago for a post at Ohio State University, Lima Campus, I am simply putting the material together in a "normal" format.

Two books have recently come out from the MSU Press: those by Eren Giray and by Jean Boyd and Beverly Mack (described below). The volume on El Fellati prepared by Sydney Kanya-Forstner and Paul Lovejoy should be ready by the ASA, and perhaps Adam Jones' edition of the Benin portions of Dapper. These are coming from the African Studies Program at Wisconsin, thanks to David Henige; the ASP has become the mainstay for these works in the last few years.

I have heard recently from Liz Gunner that her revisions of the manuscript on Isaiah Shembe are virtually complete; I am not sure, as of this date, about whether MSU Press is still committed to the publication of this volume. Tim Geysbeek continues to work away on his dissertation and his volume of Mandingo traditions, in conjunction with David Conrad. These and other projects are mentioned below.

Beverly Mack has generously provided a detailed commentary on the production of the Nana Asma'u volume, which takes up most of this newsletter and will show you how much labor and dedication were committed to theCollected Works. Her commentary is so useful that it may be wise for it to be published in a journal such asSudanic Africa or Islam et Societes or History in Africa.

The ASA was uncommonly slow in getting its program out this year. For some reason our roundtable was submitted only recently for approval to the program committee, and I do not know the time slot to which it will be assigned. I have had no response about whether we have a business meeting slot or not. We will have at least the roundtable, at which David Henige will preside and Sydney Kanya-Forstner, Robin Law and Paul Lovejoy will provide discussion; hopefully Beverly Mack and Murray Last, who agreed to participate but will probably not appear on the APAHS program because of other commitments, will join in as well. As of this printing I cannot say when the roundtable will occur.

A text prize will be awarded this year as well. Presumably the committee has completed its deliberations by this time.


Andreas Eckert, Universitat Hamburg

Andreas Eckert is expecting to publish a book on Douala petitions, c.1860-c.1960. This project involves the translation and the publication in French and English of about 30 petitions (chosen among 60 petitions) of different length (ranging from 1 to 15 pages) that various Douala individuals or groups have sent to different local and metropolitan institutions (for example: the Queen; the German Kaiser; the British Prime Minister; the German Parliament (Reichstag); the Versailles Peace Conference; the League of Nations; United Nations Trusteeship Council; German and French local administration /administrators in Cameroon; various missionary societies). The petitions recorded are written in English, German, and French and cover the period between 1865 to 1960. Most of the documents can be found in colonial archives in Cameroon, France and Germany, while a few have been discovered in private archives in Douala. The petitions deal both with national and international affairs (e.g. Mandate status of Cameroon) as well as with local questions concerning Douala/the Cameroon littoral (most important issue: land questions).

John Wright, University of Natal, Pietermaritzburg

John Wright is working on volume five of the James Stuart Archive of Recorded Oral Evidence Relating to the history of Zulu and neighboring peoples. The evidence of another 40 or so of Stuart's informants will go into the volume, to add to the evidence of the 119 informants whose statements have appeared in the four volumes of the JSA so far published. The testimonies which were recorded by Stuart primarily in English have been placed on computer. The translation of the testimonies which were recorded in Zulu is under way.

Dadari Project

Sydney Kanya-Forstner, Paul Lovejoy, John Hunwick and Sean O'Fahey are collaborating in a research project aiming at gathering materials about Muhammad Dadari. Dadari was a Fulani cleric from the Western Sudan who became a pupil of Shaikh Usman Dan Fodio and spent many years in the Sokoto Caliphate before travelling east to the Nilotic Sudan in search of the Mahdi. Dadari who seems to have acquired a significant following in the region by the time Muhammad Ahmad proclaimed himself Mahdi in 1881, joined the Mahdi in 1882 and was also instrumental in persuading Hayatu Ibn Sa'id, a grandson of Caliph Muhammad Bello, to accept the Mahdi as well. When the Mahdi Muhammad Ahmad died in 1885, Muhammad Dadari was among the first to proclaim his allegiance to the Khalifa Abdallahi and became one of his closest advisers. Muhammad Dadari himself died in 1891, but one of his sons remained with the Khalifa, and a grandson is apparently still living at Mai Wurno in Sudan.

David Conrad

Conrad's project involves translations from "market tapes", local performances of historical Mande traditions recorded by entrepreneurs and sold in local markets. The material will be published in two volumes. One will focus on oral tradition from Upper Guinea pertaining to the epic ancestors of the Sundjata era. Another will be on the nineteenth-century heroes of Mande epic tradition. Both volumes will comprise individual introduction and an extensive glossary of all Maninka terms mentioned in the texts.


Eren Giray
Nsiirin! Nsiirin!Jula folktales from West Africa (MSU Press, 1997). Eren Gray's book is a compilation of tales extracted from storytelling occasions involving a number of raconteur families coming from different areas of Burkina Faso and West Africa, yet recounting their oral traditions in Jula. As such, the stories represent a natural retelling of the tellers themselves as well as that of the compiler and translator. The book has received the endorsement of the American Folklore Institute.

Jean Boyd and Beverly Mack
The Collected Works of Nana Asma'u 1793-1864 (MSU Press, 1997). This is a marvelous resource of over 700 pages. After introductory materials, the authors/editors provide translations of more than 60 poems and other materials written (and recited in most instances) by Nana Asma'u, the daughter of Uthman dan Fodio. They then provide Roman transcriptions and Arabic-script facsimiles of the works, as well as glossaries and indexes, to make this a fully usable resource for students and scholars of Hausaland and its environs, and of the languages in which Nana wrote (Arabic, Fulfulde and Hausa). This will be an excellent companion to Jean Boyd's biography of Nana, and will show the enormous influence and pedagogical innovation of this important female Muslim scholar. See Beverly Mack's comments on the production immediately following.


Beverly Mack. 18 July 1997.
Arabic and Hausa-related production problems in the preparation of CRC (camera-ready copy) for The Collected Works of Nana Asma'u 1793-1864(Mack and Boyd, MSU Press, 1997).

Adam Jones and especially Knut Vikor have provided extensive guidelines for scholars working with Arabic manuscripts and preparing them for camera ready copy (see Jones' 16 October 1995 notes, and Knut Vikor's 30 October 1995 email). Most of the technology they describe is suited to use with Macs, and for a long time it has been Mac users who have been best able to deal with Arabic script and Arabic diacriticals in the transliterated form. The comments offered here reflect experience with a PC using WP 5.1 in Dos beginning in 1990. At the time, the massive size of our collection, and the fact of needing to reproduce Arabic, Hausa, and Fulfulde in WP 5.1, meant that we faced a different set of problems than those considered by others in the field using Macs. Without an upgrade to WP 6.2 Windows very late in the project's life, this project could not have been completed satisfactorily. The dilemmas we encountered are described below (with computer-related issues in bold), but for those who prefer a summation, here are our recommendations for scholars commencing work on comparable materials, and with comparable software (i.e. WP 6.2 Windows):

  1. considering the unfamiliarity of most academic presses with Arabic, and other African language typeface issues, assume that you will have to provide camera-ready copy;

  2. for Arabic manuscripts, consider whether in-putting the Arabic will compromise the presentation of the document, and if it will, then opt for using photographs of the manuscripts;

  3. for manuscripts inajami, assume there is no option except photographic reproduction;

  4. for transliterated Arabic or any other language originally presented inajami, find, install, and master use of the software program that allows for computer regeneration of it;

  5. be sure you have a laser printer that can reproduce hardcopy of what you see on your screen;

  6. choose a "standard" transliterated form of each foreign term the first time it is used in your own narrative, and keep a record of it for universal search/replace-spell check function at the end of the project (cited passages may differ, so it would be useful to select a form most commonly used by your sources, if that is possible);

  7. after research, and before commencing the narrative of your book, set the ms. format, including typeface, font size for text and notes, footnote form, margins, pagination placement and style, header format, bibliography style, indexing format, and any other particular functions you might need to use, such as tables, columns, etc.;

  8. become familiar with creation, compression, and expansion of master documents;

  9. establish glossary and index lists from the commencement of the writing;

  10. become connected through an email account that allows for transmission of large documents, and is able to preserve their appearance;

  11. if you have questions at any time on any aspect of the reproduction of the book, consult a press that is likely to be marketing your work;

  12. while typing, use the "save" function frequently.

Our work began with concern for the research, and a commitment to making Asma'u's works available to other scholars. Never did we dream of being capable of producing camera-ready copy (and we were right; we could not). The process of creating a publishable volume of texts and translations was so fraught with problems that we would never have embarked on it had we known what lay in store. The work resulted in the English translation of 66 long works in three languages, reproduced with the original manuscripts in a volume of 754 pages of small (ten pitch) typeface, with extensive footnotes (in eight pitch), and includes an introduction, six maps, three glossaries, two appendices, bibliography and index. Much of what we did would ordinarily have been done by the editorial staff at a publishing house. We worked on the assumption that we would be able to send this finished manuscript to a publishing house for editorial revision of a nature that we were not qualified to provide. Toward the end of the project, when we negotiated a contract, we were told that we needed to provide CRC because the book's multi-linguistic nature made it cost-prohibitive for the publisher, who did not have qualified editors to review and revise across such a range of languages and formats. At that point, the only reason we carried on was that we had come that close to finishing -- and because we found technical assistance in the computer wizardry of Amy Barnes here in Lawrence, Ks. Without her expertise, the book would not have been produced in anything close to CRC.

It took over a year beyond the point of finishing the writing to produce the CRC required by the press, and then eight months for CRC to be reproduced and marketed by the press. Needless to say, such terms of publication of foreign language texts and translations are discouraging to scholars, many of whom are ill-prepared to do double or triple duty as computer wizards and professional editors. It would be miraculous if we scholars were able to convince publishing houses that THEIR EDITORS need to re-tool in order to handle what they are requiring of each of us, but certainly down-sizing will happen before this does. On that cheerful note I end this summation. What follows is narrative of the production of this book, with a focus on the tremendous changes in technology we encountered between 1990 and 1996, for those who are interested in the details.

At the outset of this project Jean Boyd and I, living on different continents, commenced the project of compiling, translating/transcribing, and annotating a collection of 61 previously unknown long poetic works by a nineteenth century Muslim author. The works existed in copied manuscript form that had been preserved by the author's family over the past century and a half. The Arabic, Fulfulde, and Hausa language manuscripts were all in Arabic/ajami script; in addition, some Hausa works had been transcribed into Roman script. Some of the works in the collection had been translated into English by Jean Boyd and the author's descendants. In addition, the collection included two works known to be forgeries (one of them in both Arabic and Fulfulde, the other only in Fulfulde), a list of the author's students, and two works written to honor the author after her death: an elegy and a praise song. The collection was extensive in size (some of the works had over 300 couplets), multifarious in nature, and trilingual in two script types.

: The situation presented above required technology that could deal successfully with the following immediate aspects of the project: 1.) transmission of poetic works intact (margins, line spaces, tabs); and 2.) sufficient memory to coordinate compilation of 400 pages of text, (minimally 10,000 words). This problem was nothing that money could not solve: so I spent the next sixteen months writing, submitting, and awaiting response on a grant from the National Endowment for the Humanities (1). During this time we translated as we could, and communicated translations by snail mail. The grant was successful, and its budget included funds for a computer (and components) with a hard drive, software, modem, a letter-quality laser that might in the future be able to produce Arabic script, and funding for a Compuserve account that would transmit poetic works intact, unlike standard email transmissions, which had jumbled messages in the transmission.

: From the outset we investigated software for Arabic, Fulfulde, and Hausa. At the time, none that we knew of existed for PCs for the last two. Arabic software programs required fluency in Arabic, knowledge of a different keyboard, and decisions about how the manuscripts in their original could most accurately be rendered in print. We realized that to attempt to transfer the Arabic on the manuscripts into a typed, standard form would not only require our decision about which Arabic form to follow, but also would inevitably result in the distortion of the manuscripts, in which Arabic was rendered in a style distinctly connected to nineteenth century northern Nigeria, and unlike standard Arabic script of the Middle East. From an historian's perspective, such distortion was less than desirable, because it would render the manuscripts unusable as historical documents. Further, such transcription would distort the poetic work itself. To attempt to interpret manuscript Arabic and render it as standard Arabic script would have compounded whatever puzzles exist in the interpretation of the manuscripts in the first place, and any hope of accurate translation would be lost. Our aim was to provide original texts as a complement to our translations of them. The final deciding factor in choosing to provide facsimile forms of the manuscripts lay in the fact that only a few of the works were originally in Arabic; the rest were in Fulfulde or Hausa, for which the prospect of ajami presentation through software was impossible. Thus, we decided that all original manuscripts were to be included in the volume in facsimile form, left as we worked from them, for the benefit of the expert who might rework them. We thought our problem concerning Arabic, Fulfulde, and Hausa software was solved, as as far as ajami goes, it was.

For our purposes Arabic and Fulfulde poetic works were not rendered in Roman script, but those written in Hausa were. All the works originally written in Hausa ajami had been transliterated into Hausa Roman to facilitate the translation process (2). From this situation we faced the next dilemma, that of finding a software program that could reproduce Hausa diacriticals. After investigation it was decided that a software company in Madison, Wisconsin could make such a program for us. This was an unanticipated delay and expense. Once the package was delivered, it took another month to determine that there was a glitch in the program, reconfigure, redeliver, and finally master operation. The software would run on WP 5.1, but not on Windows. This was not an issue in 1990-91, but it would become important at the end of the project. Cost: $200 and two months' time

Our aim in this volume was to provide for the reader as much information about the works and their historical origins as possible. With this in mind, we added various features of our introductions to each work. The translation and historical footnotes explaining certain terms throughout the work were, we felt, not enough; to these we added to each poem: working poem numbers as identification, dates of authorship, language of the original, source of the manuscript, an introduction, related texts, and significant features (literary or otherwise). Thus, each poem became a microcosm of the entire volume, as we sought to provide as much historical information as possible, setting the material in context for the reader, and guiding the reader in seeing the work's connection to the whole.

This stage of production involved answering questions: in what order should we present these works ? If the working numbers represent chronological order of their creation, how would we explain works that were re-written in a different language (i.e. for a different audience), and where would we place such a work(s) in the chronology ? As for dates, should we use the Muslim calendar dates that are true to the works (and cited in some of them), or the Gregorian calendar dates of composition that would be more useful to the average reader ? If the work was written in several languages should we record all of them as language of composition ? What if we had several sources of manuscripts of the same work ? What exactly needed to be in an introduction to the poem to clarify its circumstances for a general audience, removed in time and culture ? How much should we say about related texts and significant features ? (i.e. was there room for literary analysis here ? At first we thought so, but obviously not).

While these are creative decisions, technology had to be used actively in implementing them. We had to decide page placement, font size, font type, use of bold and underlining, etc. for the indicators of each feature. When genealogies were part of the introductions, they had to be laid out carefully to fit on the page. We did what was possibly worse than ignoring this aspect of production: we failed to coordinate decisions about formatting, which rendered the manuscript impossible to refine through universal replace commands. Nevertheless, we anticipated this would not be a problem, because we thought, "THIS WILL BE HANDLED BY THE PRESS".

Then there were the footnotes, an important part of this project. The book was getting too long, and footnotes are supposed to be in smaller print; this was a dilemma, since we could not manipulate footnote size on the WP 5.1 software we used until switching to Windows at the very end. So we used the footnote font size (8 pitch) as our main font size throughout, changing it to a larger version for any part of the narrative that was not footnote. Still, we expected this would present no problem, because we thought, "THIS WILL BE HANDLED BY THE PRESS".

Throughout this stage of the process we collected names, places, and terms that would need to be included in an appendix. We should have collected them in one, alphabetized, columned list from the start. The appendix grew into three separate appendices, and presented us with many of instances in which we would need foreign-language diacritical markings, which we were still unable to provide. Nevertheless, we expected that THIS WILL BE HANDLED BY THE PRESS.

Pulling the book together required a general introduction. People do not read introductions, and yet we knew that without some background, a reader would miss much of the richness of this material. So we wrote a general introduction purposely lacking the re-invention of the wheel with regard to Sokoto jihad history. We called it (unceremoniously) "The Essential Asma'u", implying that it contained everything one needed to know about Nana Asma'u but might be afraid to ask. We hoped it shouted "Please read". This was one of the few technologically easy parts, since it was narrative, but even here we needed columns for date lines.

The book also needed maps for places that no longer exist, to illustrate the battles, sites, and perambulations of the jihad forces. Jean Boyd knew a cartographer in Britain who had worked on comparable maps, and employed him for the production of six original maps. Ironically, after our having provided and paid for professional quality maps, we learned at the end that this and only this WOULD HAVE BEEN HANDLED BY THE PRESS.

The appendices described above held more than just diacritical problems. We had to decide how to format their appearance (columns, bold, italics, etc.), alphabetize by hand, and account for the linguistic origin of each term. Of course we expected, "THIS WILL BE HANDLED BY THE PRESS".

Beyond the appendices lay two other leviathans: the bibliography and the index. The bibliography presented the piece de resistance of software usage [notice the absence of accents on that foreign phrase]: hanging tabs, to underline or not to underline, to italicize or not to italicize, to follow British bibliographic listings (just first initials, place of publication omitted) or American. Bigger questions loomed: social science format or humanities ? THIS WILL BE HANDLED BY THE PRESS (we thought). And there was not just one bibliography, but several, to account for books still in print, books published only in Arabic, British colonial archive works, books that are available only in the archives of the Fodio clan in Sokoto...we narrowed it down to two categories, published and unpublished works cited. THIS WILL BE HANDLED BY THE PRESS (we thought).

Finally, the index. Hell on earth. An index for 400 narrative pages, over 20,000 words in four languages, three typefaces (counting Arabic and Hausa diacriticals). YOU decide the categories next time, gentle reader. There is an index-generating capacity in Windows that we should have used from the outset, but it does not supply the categories ! Nevertheless, we expected that THIS WILL BE HANDLED BY THE PRESS.

The final checking of camera ready copy for such a volume is a nightmare. Spell check is useless when foreign words and names are involved. Working between British and American cultures meant that many of the English words had several right spellings, and needed to be used consistently. Varieties of English spellings are merely the tip of the linguistic iceberg when dealing with varieties of accurate Roman script transliterations of Arabic and Hausa terms. Throughout the original manuscripts, as well as in sources cited we found varieties of spellings: Hunayn/Honein (Ar.) = Hunaini (H.); Thaikif (Ar.) = Sakifu (H.); Nakhla (Ar.) = Nahilatu (H.). Furthermore, there is no agreement in the scholarly community about standard form of terms transliterated from Arabic script in Arabic, much less for Hausa ajami, and presses were no help either: the advice we received from Brill in Leiden contradicted advice from U.S. publishers. In examining existing publications with comparable terms we found no agreement whatever. For instance, "Aisha" is rendering differently by Hiskett (1975), Muir (1923), Nurbakhash (1990), Schimmel (1994) Waddy (1980), and the Encyclopedia of Islam (1960--). Such varieties of spellings include, of course, varieties of diacritical markings for both Arabic and Hausa, further complicating the situation. For Fulfulde, there is even more variety of diacriticals (i.e. the written version of Fulfulde is even less standardized) than in Arabic and Hausa.

At this point the technical limitations of our computer became significant. By this stage it was 1995, and I had found software programs for both Hausa and Arabic in Roman script, but they only ran in Windows. So before installing these programs, I had to purchase and install more memory in order to run them. After this delay, it took more time to learn the keystrokes for the software; several people and several weeks were involved in this process. Altogether the time and expense devoted to converting the computer to accommodate Arabic and Hausa diacriticals amounted to at least two months and $300.

Having installed Windows to run the Arabic and Hausa software had another effect: it abruptly cut direct communication of typed versions of the manuscript with Jean Boyd, since her computer would not run Windows or the software for diacriticals. It also meant that the appendix with all the works in Hausa Roman that I had already typed in could not be merged with the full manuscript, since that older software for Hausa diacriticals would NOT run in Windows. Furthermore, Jean's husband had improvised key strokes in an attempt to provided Arabic diacriticals in WP 5.1 throughout the text. All these needed to be re-done by hand once the new Arabic diacritical software was installed.

Thus, the final stages of the production of this book involved many human hours to coordinate vestiges of drafts typed at various stages of technological standards because PC technology changed so dramatically between the time we embarked on this project and its end. A major problems remains, however, and that is for scholars in the field to find a way to standardize transliterations of Arabic, Hausa, and other languages written in Roman as well as Arabic script.

1) But it was not this simple: first there were innumerable telephone calls to editors and scholars, through which we explained the dilemma, inquired about production approaches, and learned a great deal about how little we -- or the publishing world at large -- were prepared to render in print such a formidable collection of original manuscripts. the grant proposal was a last resort, not an initial idea.

2) To verify the accuracy of these transliterations, each was approved by an official "reading" in the field by a scholar qualify to speak to the veracity of the work in its new form.

H-Net, Humanities & Social Sciences OnLine
Send comments, suggestions and submissions via our Contact Page
This site is maintained at Michigan State University
Copyright ©1997, H-Net, Humanities and Social Sciences OnLine
Connect to H-Net's Home Page