"Creating an Internet Archive for French West African History"
by James A. Jones, Ph.D.
West Chester University of Pennsylvania

Presented at the 45th Annual Meeting of the African Studies Association,
Washington DC (December 6, 2002).

Go to Archive Table of Contents


This paper examines issues involved in organizing a collaborative effort by scholars to share notes using the Internet. The main issues are the organization of the data, the site administrator's function, and who "owns" the content. It is based on my experience creating and a web site that presents a significant part of my notes from research conducted for the book, Industrial Labor in the Colonial World: The African Workers of the Chemin de Fer Dakar-Niger (Portsmouth, NH: Heinemann, 2002).


I am a historian of modern Africa. Sources on human activity are the raw material of my research, and government documents are staple items for the practice of my craft. Despite their flaws, government documents offer fairly detailed data on most historical moments in the modern era. At the very least, they are useful for creating detailed timelines that help organize data obtained from other sources.

Unfortunately, the sources for West Africa history are widely dispersed since the French decentralized their colonial records in the late 1970s, and colonial-era government documents are stored in national archives according to a scheme that requires extensive knowledge of French colonial history. For instance, while most documents from the history of Burkina Faso are housed in that country, many are located in Mali because the colony of Upper Volta was administered from Bamako until 1914, and parts of Burkina Faso were attached to the French Soudan from 1932 to 1947.1

In another example, documents on the Dakar-Niger railroad, the subject of my most recent book, are distributed in a number of archives located in at least three countries, and filed under eight different company names including: Chemin de Fer de Dakar à St. Louis, Chemin de Fer de Kayes à Bafoulabé, Chemin de Fer de Kayes au Niger, Chemin de Fer de Kayes à Ambidédi, Chemin de Fer de Thiès à Kayes, Chemin de Fer Thiès-Kayes-Niger, Chemin de Fer Thiès-Niger, and Chemin de Fer de Dakar au Niger.

I believe that the Internet offers a way improve access to these archives. Besides providing a means to "publish" information at low cost, it eliminates problems of access associated with a document's location, and offers groups of scholars the means to pool their efforts.


I have created a prototype of an on-line archive called the "French West African Archives Web Site." It resides on a server at my university in a sub-directory of my African studies materials, and contains 210 text files (2.26M) and 44 images (5.65M). Each text file corresponds to a dossier from one of four archives: the national archives of France, Senegal, and Mali, and the Archives of Afrique Occidentale Française. Each file contains notes on the documents in the dossier, and includes bibliographic citations down to the level of page position. The on-line archive also features a fifth section containing interview transcripts from Senegal and Mali, and the sixth section containing notes from theses, out-of- print secondary literature, and relevant articles on the Web. The site also offers definitions, a glossary of place names, practical information about each archive, and topic suggestions for student research papers.

In the proposal for a grant that supported the creation of this site, I described it as a way to give my students a taste of the experience of conducting original research on West African history. For most of them, money and language present insurmountable barriers to making an actual trip, but this site gives them access to information in the form that I found it once I'd completed my research and began to write. Although it can't teach them everything that they need to know about how to do research in Africa--for instance how to find the right street corner from which to catch a "bush taxi" to a town outside of the capital--the on-line archive offers a way to model the thought process that goes into analyzing archival sources.

When I conducted research in France and West Africa, I did it all by hand, like generations of scholars before me. I took handwritten notes and made carbon copies to send home from time to time. Unlike earlier scholars, I began my research just as laptop computers became affordable for graduate students, and each evening I typed the day's notes into a computer. As I typed, I was careful to preserve the correct archival references and the exact content of my notes. Once the notes were digitized, I sorted them by subject and chronology, created timelines, and identified cross references. Once that was completed, I began to write.

To give just one example of the usefulness of this approach, in fall 1990 I found a reference to a marabout named Seydou Nourou Tall in a document located in Aix-en-Provence.2 Then in May 1992, I found a document in Senegal that mentioned his support for the administration during a period of labor unrest.3 The following week, I located another sequence of documents concerning his appearance in court earlier that same year.4 Four years later, when I saw his name mentioned in an article by Frederick Cooper,5 it only took a few minutes to locate the other references and recall who he was.


In undergraduate courses, I assign a document for all of my students to read, and use that as a basis for classroom discussion. In graduate courses, I ask students to chose one of the documents in the collection, read it, and annotate it using information from other documents on the web site plus secondary sources and journal articles. Finally, they are expected to write a paper that explains their annotations and identifies the larger historical context in which the document was created.

For instance, one set of notes is entitled "Correspondence from a small military post along the railroad, spring 1888." The post was Galougo, located in modern-day Mali along the Upper Senegal River between Kayes and Bafoulabé. In the spring of 1888, a military detachment had just completed the construction of an 80-meter iron and masonry bridge. Their commander, Captain Mahmadou Racine, was one of only a few African officers in the French army during the conquest period. He was preoccupied with trying to locate enough workers to dismantle construction equipment and transport it to the end of the line. He also faced jealousy from French military officers, desertion by African workers, resistance from a local chief, and demands from his superiors to resettle Africans along the railroad line.6 In short, the correspondence from this one post provides a glimpse of a wide range of issues introduced by the French conquest.

(Clockwise from top left)
1. Map of the first stretch of the Kayes-Niger railroad, adapted from Jones, Industrial Labor in the Colonial World, map 2.1.
2. The bridge at Galougo, from Joseph Galliéni, Deux campagnes au Soudan Francais ... 1886-1888 (Paris: Hachette, 1891), 380.
3. A voie Decauville in use, from Galliéni, Deux campagnes au Soudan Francais, 159.


To facilitate their annotation, students read Myron Echenberg's book on the tirailleurs sénégalais (which provides some details on the life of Mahmadou Racine),7 and they have access to my notes on Denise Bouche's Les Villages de Liberté en Afrique Noire Française on the Web site.8 They can also use our university's print library, on-line journal articles from J-STOR, LEXIS Academic, and other web-based information services.

I expect a good paper to locate Galougo along the Upper Senegal River southeast of Kayes, and identify voie decauville as a type of narrow gauge railway used on construction sites in the 1880s. It should also explain the larger historical context of the French effort to conquer the Niger Valley, using the railroad to transport supplies from Kayes for the assault against Segou, the Umarian capital, in 1890.


The basic idea is that scholars of French West African history make their notes available to each other by contributing their notes or summaries of individual documents to an on-line archive. To do this, we need a way to convert notes to web pages, a place to post the pages, and a scheme for indexing the material. In addition, we need to address the question of who owns the material and to explain how it was created (and hence, the limitations of its usefulness) to other researchers.

First, let me address any concerns that an on-line archive will encourage weak research. There is no argument with the fact that the on-line archive will, in nearly all cases, offer something that is inferior to the original documents. Budgetary restrictions make it unlikely that an on-line archive will ever feature high-quality scanned images of archival documents, and nothing will replace the assistance that a knowledgeable reference librarian can offer. Using the on-line archive, however, has the potential to serve as a "finding aid" to enable more efficient use of time and money in overseas archives. In that sense, if properly used, the effect of an on-line archive will be similar to that of telephones, photocopiers and fax machines, all of which make it easier to obtain information from archives before one visits them.

Assuming that each contributor is able to type up his or her research notes, the next step is to place the notes on the Web. There is no reason why the notes have to reside on the same server, as long as the home page is located somewhere conspicuous (perhaps on the ASA or MANSA server). To be most effective, every file should use a common header that identifies the archive where the original documents are located, the title and date(s) of the documents, the archive's reference number, and the name of the person who took the notes.

Within each file, notes should use standardized spellings for African place names and proper names, as well as French terms such as cercle (not "district") and gouverneur rather than "governor." Standardization across all files in the archive will be harder to achieve, since each contributor will have his or her own system for transcribing foreign words. Either we must be prepared to tolerate irregularities, or else contributors must be prepared to submit their work to a site administrator for editing.

At its simplest, the job of the site administrator will be to post contributions for those who have no server of their own, update links from the site index to new files, and respond to questions about the site. Depending on the administrator's dedication, other tasks may include creating uniform document headers, expanding abbreviations, providing alternative spellings, and updating the search path for software used to index the site.

Site administration would be greatly simplified if authors add keywords to their notes, either in a separate section at the beginning of the document, or at appropriate points within the document itself. For instance, notes on a document about the tirailleurs sénégalais might interest someone who searches for words like soldier, military, and army. The author should add them as keywords if they are not already in the notes.


Once the web pages have been created and placed on a server, they need to be organized for use by researchers. The system used in modern archives is derived from the organization of the colonial bureaucracy, a complex, dynamic structure created by French legislators over the course of a century. The structure became frozen at independence and is preserved in the arrangement of dossiers in the archives bearing titles such as "Soudan political affairs," "Senegal health service," "Guinea native justice" and the like. Although this official system is rather arcane, many researchers are familiar with it and printed finding aids already exist,9 so it should not be abandoned.

That does not prevent an industrious site administrator from devising something more useful. If contributors can be convinced to provide some additional information along with their submissions, and to make their submissions by email, then it will be fairly easy to mount and maintain a coherent on-line archive. When contributors submit their notes, they should provide the following information:


For instance, here is the content of my disclaimer page:

These notes were collected in the process of writing a dissertation, and are placed on the Web for educational purposes only. No representation or warranty is made or implied regarding their accuracy or their suitability for your particular purpose. To the extent you use or implement this information in your own setting, you do so at your own risk.

The information provided herewith is solely for your own use and cannot be sold. In no event will the author or West Chester University be liable for any damages whatsoever, whether direct, consequential, incidental, special, or any claim for attorney fees, arising out of the use of or inability to use the information provided herewith.

Please keep in mind as you use these notes that they were translated by a native-English speaker from handwritten notes made from French-language archival documents. In addition, nearly all of the documents produced before World War I were hand-written, and not always easy to decipher.

Finally, keep in mind that the author of these notes collected them during an investigation into the social impact of a colonial-era railroad, so these notes do not necessarily contain all of the information available in each document.


There are some other practical questions for which I do not have the answers. I would welcome input from this audience.

Language: I provided my notes in English, but included extensive quotations in French and provided translations for technical terms. For my students, this is essential, since few can read French, but for more serious researchers, the choice of language should not be an issue. The title of the document is a different matter altogether, and it (plus all critical terms) should appear in their original language to reduce errors due to translation. Thus, "chef de canton" should remain unchanged, and not be translated to "cantonal chief" or "district officer" (although the translation can be added in parentheses). For the same reason, proper names should appear just as they do in the original document, with alternate spellings in parentheses.

Error-checking: One of the most important concerns should be to produce notes that are as error-free as possible. Unfortunately, I know from experience that this is impossible, and I also know that while a few users will take the time to email me with errors they've found on my web site, I am the one who is most likely to find errors. That explains why my disclaimer contains such foreboding legal language, and why I recommend to anyone who does research on French West Africa for publication to consult the original documents.

As I use my notes to write articles and books, I try to verify them using the original documents wherever possible. In the process, however, I have discovered another problem. As I took my notes, I carefully indicated exact quotations with appropriate punctuation, but in many instances, I paraphrased the author's words in order to save time and space. Later, when I employed my notes, I rewrote my paraphrased material, and occasionally found that I had inadvertently restored it to a form that resembled the author's original words. While all of this might seem completely obscure, in the past year several well- known scholars have learned the hard way that even inadvertent plagiarism is not tolerable.10 The only solution that has occurred to me is for contributors to make clear which words are exact quotations from the original, and for researches to consult the original documents before writing.

Ownership: Although it is hard to imagine that notes of French West African history have significant monetary value, the work that goes into producing on-line notes certainly has value. That raises the question of "ownership" of the material, for which I have no answer. My hope is that scholars will make contributions to the site because they share a desire to promote interest in French West Africa, and because they understand that scholars have less right to "ownership" than the people whose lives provided the subject matter for the documents. Rather than try to establish fee schedules, royalty payments, and a way to bill site users, I have decided to make my notes available to other scholars without restriction. Until someone devises an alternative that can be implemented without undue technical expertise, I will continue to do so.

What happens when a researcher finds something useful at the web site--a statistic or other brief item--and wants to use it in a publication? Should the researcher cite the original document or the on-line notes? Like the previous problem, this one can be settled by referring to original documents, but when those documents reside on the banks of the Niger River at the end of a journey costing several thousand dollars, the temptation to rely on the web site will be great.

In the larger scheme of things, the creation of an on-line archive will reduce the amount of time and money that researchers spend in countries that possess archives. The effect is unlikely to be noticeable in Paris or Aix-en-Provence, and only slightly more noticeable in Dakar, but it may be significant in capitals like Conakry or Bamako where the tourist industry is less well- developed.

On the other hand, an on-line archive can help preserve fragile archival documents by reducing the need to handle them. It will certainly enhance the ability of researchers to plan their research strategy, and provide an opportunity to practice research skills before venturing out into the field. Finally, by making otherwise inaccessible data more available, the on-line archive will stimulate interest in the study of French West Africa and colonial history in general.

Reference Notes

1. The two cercles were Wahiguya and Tugan.

2. Colonie du Soudan Français, "Rapport Politique Annuel" (1933), 69, in ANFOM Affaires Politiques, carton 2802, dossier 3.

3. Gouverneur Général de l'Afrique Occidentale Française to Gouverneur du Soudan Français, copy of letter nÝ86/C (Dakar, 16 September 1937), in AOF K 18 (1).

4. See, for example, "Declaration fait par Seydou Nourou TALL le 19 Mai 1937" in ANS 5 M 487.

5. Renseignements, 29 October 1947, AOF K 457 (179), cited in Frederick Cooper, "'Our strike': equality, anticolonial politics and the 1947-48 railway strike in French West Africa" in The Journal of African History, volume 37, nÝ1 (Jan 1996), 94.

6. "Correspondence du Commandant du Camp à Galougo" (December 31, 1887 to March 27, 1888) in ANS 15 G 118.

7. Myron Echenberg, Colonial Conscripts: The Tirailleurs Sénégalais in French West Africa, 1857- 1960 (Portsmouth NH: Heinemann, 1991), 38-39.

8. Denise Bouche, Les Villages de Liberté en Afrique Noire Française, 1887-1910 (Paris: Mouton & Cie. et École Pratique des Hautes Études, 1968).

9. Examples include Etienne Taillemite, Les Archives de la Marine conservées aux Archives nationales (Vincennes: Service historique de la Marine, 1980); Peter Claus Hartmann, Archives, Bibliothèques et Centres de Documentation a Paris (Paris & New York: K.G. Saur, 1976), Patricia Carson, Materials for West African History in French Archives (University of London, The Athlone Press, 1968); Saliou M'Baye, Guide des Archives de l'Afrique Occidentale Française (Dakar: Archives du Sénégal, 1990); and I. Bekény, Mali: Reorganisation des Archives Paris: UNESCO NÝ2231, December 1970).

10. Steven Ambrose and Doris Kearns Goodwin were both accused of plagiarism in 2002 for doing something similar.