HARRIET TUBMAN SEMINAR
NOVEMBER 1, 1999
Missionaries and Traders Reports and the Creation of a Slave Biographical Database For the Nigerian Hinterland: Challenges & Prospects
F. J. Kolapo
This is a working paper. The Nigerian Hinterland Project (NHP)has the creation of a slave biographical database for the Nigerian hinterland as one of its principal research activities. The NHPs statement concerning this sub-project states that it will generate a computer-based data retrieval system to enable on-line access to archival and published materials on individual enslaved Africans during the era of the "Slave Route" (UNESCO/SSHRCC Project - Nigerian hinterland 1997, p.872.) This paper therefore discusses one of the beginnings of this research effort. It seeks to explore, understand and highlight possibilities and problems to look ahead for in the process. In light of this, in this paper, I explore implications of a biographical data set that I have generated from a number of missionary sources, as well as from one European trading agents journal. Having selected the primary sources for the purpose, I reduced the biographical information available in them into a table of data sets; all organized by categories. My purpose is to discuss the potential usefulness of this-type material and the methodological problems that must be considered as the more comprehensive database that the NHP has in view is being built. It is hoped that this process will enlighten network members about the project. The presentation should also advertise the data needs of the database project researchers to network members, as well as encourage them to forward useful biographical information they may come across to the NHP.
Limitations of Sources
My original intention for this paper was to create a biographical data set from all available primary documentary evidence for Lokoja and Gbebe, Niger-Benue confluence stations of the CMS (Church Missionary Society) from 1858. Unfortunately, while the CMS documents are available, the diplomatic documents of the British consulate at Lokoja dating from the same year as the CMS documents, are not yet available to me. Materials from trading representatives who served in Lokoja in the second half of the 19th century are also not available to me. Moreover, the reports of three out of six missionary agents who served in these places between 1858 and 1870 ultimately proved to be entirely unproductive of usable slave biographical information. This situation, besides the fact that the time taken up reading up these unproductive sources left only little time before the presentation date, dictated that I changed my plan. I, therefore, decided to consider materials from three regions instead of only one, especially, as the Nigerian Hinterland secretariat already has some amount of half-processed biographical information on one of these regions - Yoruba region. For the third region, Aboh, a published primary source was available that I knew up-front contained biographical information, the collating of which, unlike with the Niger-Benue Confluenceregion, will not be time consuming. In this manner, I tried to assemble a data set that allows for the possibility of simultaneous and systematic comparison of historical information relating to slavery and its allied social-political processes in the three regions.
First, for the Niger-Benue Confluenceregion, I consulted only the documents of CMS missionaries resident in Gbebe and Lokoja. These include all available original manuscripts of Rev. T. C. John, 1866 to 1880, Charles Paul, 1866 to 1879; Simon Benson Priddy, 1862-1864; and James Thomas 1858 - 1879. For the Yoruba region, I used slave biographies at the NHP office collated from the newsletter of the CMS, the Church Missionary Gleaner (CMG). I also used the published diaries of Anna Hinderer, the wife of David Hinderer, a CMS missionary who served at Abeokuta and Ibadan. Biographies from the CMG cover the period 1842 to 1901 and those from Hinderers diaries cover 1852 - 1868. George Jefferies, a CMS missionary at Isehin sent only one report to the Headquarters in 1860, hence, a few biographical snippets from his file complement the latter two sources. For the third region, Aboh, I had the benefit of a non-missionary source a secular and commercial trading agent who wrote a journal of his experience covering a two-year period from 1859 to 1860.
The first obvious point to note is the generally arbitrary manner of the choice of sources consulted. The first line of defense on this count is to explain that the purpose of the analysis done here is essentially exploratory and demonstrative. Thus, conclusions derived from numerical manipulation of the data are not necessarily accurate historical description of the relative slavery conditions of the three regions in the 19th century. This, however, needs to be qualified further. All information that goes into making up the data set come from verifiable historical events. They were products of and reflect historical reality as it was experienced by the persons whose biographies were used to compile the regional data sets. Thus while conclusions from statistical calculations of various variables in the data set here may not reflect the comparative historical reality of the regions, the various variables in the table are descriptive of reality for each local area from which they were derived. The data set in a way is able to simulate a conference of the eyewitnesses of mid and late 19th century events whose records are utilized. We are then able to compare their particular observations concerning slavery and the slave trade for the area where they lived rather systematically and probably objectively.
A justification of the arbitrary source selection for this paper inheres in the fact that should all available slave biographies for the Nigeria be assembled from all extant documents, the database would still be by nature partial. All missionaries, traders and diplomats, and any other category of literate observer or reporter of events in 19th century Nigeria could not have been everywhere at all times. They could not have been able to observe, name, and describe every slave, and to record every knowable detail about the few cases of slaves they were in positions to know. In other words, baring the availability of a record of slave census during this period, a biographical database that is compiled from Nigeria-based documents can hardly be expected to offer statistical absolute figures for regional incidents of the variables it is composed of.
There are other more obvious reasons for my limited and selective choice of sources for this paper. There are limits to what one person can do and to what a paper of this volume can contain. It should be noted that Gwendolyn Halls Lousiana Slave Database has been in the making for about a decade and a half, and the Havard Slave Voyages Database of David Eltis and his colleagues, in its final form, for about a decade (Eltis and others 1999, 3; Hall 1999, 3)
Before considering the possibilities that revealed themselves from the data set, the particular problems relating to the sources I used are next discussed. This concerns the problem of how to categorize and prioritize source documents that would be used to generate information that goes into the database. If the most productive source documents are first consulted, valuable time would be saved and time-wasting reading through documents that might in the end contain only negligible or unuseable information could be foreclosed. Doing this apriori is however as desirable as it is difficult. How does one isolate and eliminate particular files from the start? This was a problem I came across with the original manuscripts of six CMS missionaries who were resident at Gbebe and Lokoja that I had thought would supply an abundance of slave biographical information. To worsen the situation, there was no known published secondary materials to provide any information about the missionaries. Nothing was known about their interests, their levels of training/education, and whether they were recaptives or sons of recaptives. Hence, there was no criteria on which to base any decision to eliminate one or the other of the records of the missionaries or at least to shuffle them to the last place. I had projected my familiarity with some of them to the others, only to find out one after the other that they produced no useful data.
The documents of T. C. John at Lokoja for over a decade do not mention a single identifiable slave, despite his been the leading agent. He wrote about slavery and the evil of slave raids and that was all. T. C. John was a second generation Sierra Leonean returnee of Hausa parentage, was never a slave and never knew his parents as slaves. James Thomas reports makes the most references to individual slaves. He was kidnapped and sold into Atlantic slavery, but was fortunate to be rescued and landed at Sierra Leone as a recaptive. Here he became a Christian and eventually enlisted into the service of the CMS in 1857. He was much less educated than T. C. John or Charles Paul, the third agent. Gbebe and Lokoja were very close to Thomass hometown, and he considered himself to have returned home. These factors perhaps made him more open to dealing with slaves and of having some empathy for their plights, hence, the abundance of slave biographical information he provided. In the degree of usefulness for slave biographical information, Pauls reports fell in-between the latter two. He was, like T. C. John, a well-educated son of liberated Hausa slaves from Sierra Leone. The paradox to explain here is the difference in the reports of the two. The fourth agent whose document contains a few mentions of slaves was Simon Priddy. He was barely literate and seemed to have sent in only two reports from Gbebe. Obadiah Thomas and Thomas Joseph were sons of James Thomas and another CMS agent who died within a year of coming to Gbebe. They served in the late 1870s and did not mention in their reports a single identifiable slave. However, visitors to the Niger after 1870 still had much to say about slaves and slavery in this area. How should one approach these source documents to maximize time? This is a question that will be left floating, for there seems to have been no clear pattern of handling the issue.
The Hinderers embarked on a mission of redeeming as many children, together with their parents as possible, and Anna Hinderer was in charge of raising these orphaned children. Her passion and compassion, matched that of James Thomas, and they are the themes that run through her journal. But the Hinderers lived some distance away from the metropolise of Ibadan, as was the pattern for all CMS mission residences in resident towns. This constituted a major factor that limited the scope of her report. Moreover, because of their profession of an alien religion as well as their opposition to slavery and the slave trade in a thorough slave society, they were at the margins of ongoing events.(On Ibadan, see Falola, The Political Economy of a Pre-colonial African State, Ile-Ife, 1984) The reports of her husband who was more outgoing and who travelled around the villages of Ibadan and paid more visits to nobles of the towns were no doubt fuller and able to yield superior information in amount and in detail. The NHP is still in the process of acquiring the microfilm of his original manuscripts.
William Coles journal was produced for a reading audience, hence, his description of slaves at Aboh revolved around those belonging to some of the nobility who traded with him. Others who came into his view included those on whom acts that he judged to be atrocious had been perpetrated. For the most part, he did not identify the slaves other than as the negro of such and such a noblethe noble himself being appropriately qualified to describe his fiendishness. His journal provides proportionally more anonymous slaves than the others. His situation was not unlike the missionaries, though. As a European trader, his quarters was on the wharf a little further off from the town. He did not have intimate contact with the people, especially, the ordinary ones - the slaves.
These sources and the data set generated from them typify what to expect from documentary evidence of 19th century missionaries and traders in Nigeria. In the long run, the best slave biographical database for the Nigerian hinterland would be produced only when all known documentary sources within and outside Nigeria has been explored. The network approach in use by the NHP should produce this result probably in half the time that Hall and Eltis & Co. took to produce theirs.
Variables in the Data set.
Total Numbers and Percentage Breakdown by Identification and Sex:
The data set contains a total of 152 slave entries. 21% were identified in the sources by names, 28% by a description linking them to somebody or to something, and 51% were entirely anonymous. However, of all these entries, only 4% are not identifiable by sex. 33% were females and 62% were male. While these percentages could not be projected on any Nigerian region because of the gross limitations described for the sources, they do highlight the abundance of biographical variables that the sources contain.
The absolute number of people who had experienced slavery as recorded in the data set can also not be used for any serious analytical purpose. Clearly this data set can not be considered to have embarked on an effort at a census, nor would this goal be possible for the eventual database when produced. In this fundamental respect, this is an exercise that is different from those of Eltis, those of Hall, and those of P. Manning (Slavery and African Life: 1990) in his demographic simulation of the impact of the slave trade for Africa. The other databases that were derived from sources in the Americas and from Europe possess considerable advantages in this regard. The other databases had the advantage of availability of useful information that were kept in official records at various levels of government and commerce. These records were either mandatory or part of the processes by which the slaves were acquired, disposed, controlled, organized, employed or punished. The courts, the churches, the plantations, customs, shipping companies, ship captains, trading merchants, etc., all kept records. These were very handy for the compilation of Halls database, for instance. Hence, Hall could claim that she had a record of virtually all slaves in Louisiana between the late 18th and early 19th century. In the same vein, Eltis & co. could also claim for their database that it contains probably over 90% of all slaving voyages that left British ports, and that the French and Dutch shipping records of slave voyages to and form Africa for the 18th century are totally accounted for in their database.
Given the character of the sources for and purpose of the slave biographical database for the Nigerian hinterland, its goal seems self evident. The goal does not seem to me to include the computation of the absolute slave population figures for any region or area within the Nigerian hinterland. This variable, total figure or number, is therefore one that is discounted for any statistical purpose in the present paper.
Other Less Productive Variables or Field:
The frequency of empty cells in the table indicates variables for which the sources were unproductive of information. Pre-enslavement religion was generally assumed to be traditional where it was not stated to be Christianity or Islam. Pre-enslavement profession is available for only four entries, hence the role of slaves skill in their opportunities, or in being retained or exported out of the region of procurement can not in this data set be known. Pre-enslavement marital or parenthood status is also largely unavailable. Date of enslavement was known for only 20% but they are not representative of the three regions. Nonetheless, they do have some importance with relation to the political and military history of the Yoruba region- during the period. Most other fields for dates, excepting to some extent for death, are even less productive. Thus, fields for the dates of sale or exchange, (after capture or kidnap) redemption, manumission, liberation, and of death returned very minimal information.
The serious drawback constituted by the dismal low incidents of usable date entries is that the data set could not be manipulated to show dynamic sequences in the relationships among the variables. Thus we are denied the opportunity of tasking the data set to generate time sequences for the popularity of different modes of enslavement in and among the three regions. The relative movement of slaves from region to region over time could also not be plotted. This requires that time segments be available during which export or retention dynamics could be compared. Such a dynamic the Harvard database and the Louisiana Slave database displayed because they were not hampered by lack of time data. Again, the ultimate solution to this quandary may lie in no less than being able to gather in all available dates from all sources within or without Nigeria.
Variables of Derivation: Fields of Ethnicity and Pre-enslavement Town:
However, when static comparison of some variables within the same region are in view, absolute or relative figures assume a surprising significance. This is no less evident in this paper than it might be with a full database. Certainly, a larger representation of eyewitness accounts of slaves with Islamic names or with a particular cultural practice in a region where such practices were not known, throws up an important question. Where did the slaves come from? How was the slave identity determined and how did such cultural practice mediate or affect the relationship between the free, the slave and the freed, etc.?
For example, the data set is able to highlight slave routes or pathways, or at least, sources of supply for particular regions could be more clearly seen. Preferences for slaves of certain ethnicity, region, or of a particular description in any chosen area can also be demonstrated. For example, 62% of all slaves of Niger Benue confluence provenance served within the region in the latter half of the 19th century. 34% of them served at Aboh in the Delta region. None was found in the Yoruba region. On the other hand, there was only one case of a slave deriving from the Niger Delta region seen in the Niger-Benue confluence area.
Chart 1. Slaves of Niger/Benue Region Origin: Exported and Retained Proportion and Known Sex Ratio.
Chart 2. Proportion of Slaves Who Served in the Niger-Benue Confluence Region by Area of Derivation
Chart 2 displays the record of slaves who served in the Niger-Benue confluence area whose area of derivation was known (about 25%). Almost all of them were from the region.
Whatever might be said concerning the acceptability of the statistical absolutes here, a clear pattern emerges in all the cases that stimulates questions. Was the direction of slave export on the lower Niger during the 2nd half of the 19th century only southwards? For how long was this the trend? What was the military-political situation of the region that exports about 25% of the slaves of its area of origin who our reporters learnt about during the later part of the 19th century? An examination of the data for slaves serving in Aboh is also interesting, though, unfortunately, also incomplete. Forty-one percent (9 out of 22) individual slaves recorded as serving in the Aboh area have their area of origin recorded. Of this, 8 (36% of all mentioned slaves) derived from the Niger Benue confluence region, while the 1 (4%) left was described as Ibo. The 13 (59%) whose areas of origin were not recorded were most likely of Iboorigin.
Chart 3. Derivation of Slaves who Served in the Niger Delta Region
Whichever way one might look at it, the Niger Delta region is clearly linked to the Niger Benue confluence region essentially by a north-south slave traffic in the latter part of the 19th century. This fits mid19th century labor need pattern for legitimate goods producing Niger Delta area. It also fits the known absence of legitimate goods production in the Niger-Benue confluence area of slave export southwards. (Lynn & ). When eventually the more comprehensive database derived from all extant written record in and outside Nigeria and covering the period from before the 19th century becomes available, how would this type of relationships play out on the charts, one would like to speculate. This will definitely inform the ongoing discussion about the identification and significance of slave ethnicities both on the Nigerian coast and in the Americas.
Modes of Enslavement Variable:
Again, at the static comparative level, this variable yields some interesting results for all the regions together and individually. Mode of enslavement is recorded for 18 or 39.13% of the entries listed for the Niger-Benue Confluenceconfluence region. Twenty-eight cases (60.86%) had no known way by which enslavement was caused. Of the known cases, 27% (5) was by raiding, 22% (4) by kidnap, 16% (3) through war, 11% (2) by extra-judicial seizure, 11% (2) through birth, 5% (1) each by judicial and commercial methods. However, for the Niger-Benue confluence area, there is a difficulty in ascertaining whether many who were listed as having been enslaved through capture at war or by kidnap were not subsequently distributed by purchase. The observers only seemed to have been interested in the original mode of enslavement, rather than subsequent modes by which slaves were distributed. The history of this area certainly shows that commercial process of enslavement was as important as the military or violent ones (Kolapo, "Military Turbulence, Population Displacement and Commerce on a Slaving Frontier of the Sokoto Caliphate: Nupe c.1810-1857," 1999 chapt.4)
Chart 4. Known Cases of Mode of Enslavement For The Entire Data Set
The data for the Yoruba region shows that 8 incidents (35%) were by capture at war and 4 (17%) were by extra-judicial seizure. Unfortunately, over 52% (23) of the entries did not have their mode of enslavement recorded. Nonetheless, the overrepresentation of the violent means by which slaves were procured in Yoruba region in the mid 19th century reflects the military turbulence of the time to which the missionaries were eye witnesses. Coles record for the Niger Delta region is the most deficient for this variable. (Cf. Harris, J. S. Some Aspects of Slavery in Southeastern Nigeria, Journal of Negro History, 27, (1942) 37-54; Northrup, Trade Without Rulers, 1978, 65-80). A statistical illustration of relationships among the different mode of enslavement for the three regions will also be deceptive. However, the data do suggest that all the regions have incidents that reflect all the recorded types of enslavement processes.
Variables on Prices
Exchange, redemption and manumission prices, and the number of times slaves were sold from one master/mistress to another are also woefully lacking. If these were available, the data set would have afforded a different set of values against which to assess Eltiss and Lovejoy and Richardsons coastal price data (Eltis, Economic Growth and the Ending of the Transatlantic Slave Trade, London, 1987; Lovejoy and Richardson, "The Initial crisis of adaptation: the impact of British abolition on the Atlantic slave trade in West Africa, 1808 - 1820" in Law, R. C., From Slave Trade to Legitimate Commerce: Commercial Transition in Nineteenth-century West Africa, Cambridge, 1995.) Should a substantial percentage of exported slaves be found to have served productively for a year or more within the Nigerian area, how would this affect the calculation of profitability for the slave owners and merchants?
Enslavement Parenthood Status: Married, Father, Mother.
Of the 7 incidents of slaves or ex-slaves who clearly had parenthood status while still in slavery, 2 were married to their masters, 2 male in the Niger Delta region rose to position of prominence and became local notables with many wives, and 3 were single mothers. Again, the data can only be interpreted to reflect the use to which slave women were put to in all the three regions. The cases in the Niger Delta region of slaves becoming kings, seem to confirm the 'trade and politics' thesis championed for the Niger Delta by Kenneth Dike (Trade and Politics in the Niger Delta: 1830-1885, Oxford, 1956). The subjects of the particular biographical entries in the data set were able to change classes because of the wealth they derived from commerce. This fits with the scenario described by Dike for Bonny where legitimate production opened a floodgate of wealth to previously oppressed slaves and others of servile origin, with a revolutionary political effect.
Gwendolyn Halls Louisiana Slave database includes 113 fields of variables per biographical item. No doubt the 26 fields contained in this data set could be expanded further. Additional fields could contain information regarding returnees and recaptives, relationship between the redeemed and the redemptor, domicile after redemption or manumission, and post-freedom profession. It is expected that the discussion will come up with more fields that the NHP can consider in its future work on this project.
The difficulty that Nigeria-based missionary and traders sources might present to the researcher who tries to derive comparative absolute regional figures from them calls for an attack plan. Perhaps, identifying those sources most productive of the most useful of the desired variables would assist in prioritizing search efforts. Africa-derived sources that could produce any quality of absolute figures should obviously come in for first consideration above other types. The first that comes to mind is the record of the Slave Trade Commission established in Sierra Leone in 1819 (cf. Northrup, Trade Without Rulers, 235-40; Curtin and Vansina, "Sources of the Ninetieth century Atlantic Slave trade," Journal of African History, 1964, 185-208), and the 1848 census record in Sierra Leone. If reports from Sierra Leone in the CMS newsletter, the CMG, is anything to go by, the manuscripts of the CMS Sierra Leone mission in the years preceding the Saro exodus to Nigeria would include substantial information on slaves from the Nigerian area.
Sierra Leonean documents covering the 1820s to 1840 should thus assume a position of priority, whether these be diplomatic or missionary documents. Thereafter documents written by missionary agents in Nigeria who had been slaves, and those by expatriate missionary agents who were moved into serving in Nigeria in part as a result of their Christian evangelical cum anti-abolitionist ethos should be considered. Finally, within these bounds, it is necessary to start with the earliest documents. This is on the assumption that the documents are available and waiting to be sorted and shuffled in this manner. When gestures are directed at network members to give out whatever they have, the general outline of priorities could be made known to them. However, it may be less useful for their particular agendas, in which case, whatever relevant biographical information they have and from whatever source and for whatever periods are welcome.
It should be noted that the second section that the slave biographical database consists of, direct quotation from all sources on any individual identified during the era of the slave Route is not considered at all in this paper(SSHRCC/UNESCO Nigerian Hinterland Project, 872). Suffice to remark that the variables in the table presented here are derived from their corresponding individual biographical quotations. Aso the ethnography of names is retained as recorded in the original sources.
Chart 5. Mode of Enslavement War. Region of Capture & Age at Capture.
Chart 6. Types of Slave Employment in Egga/Bida Region. 1870s
(For a detailed discussion of the project, see P. E. Lovejoy, "Biography as Source Material: Towards a Biographical Archive of Enslaved Africans" in Robin Law, Source Material For Studying the slave Trade and the African Diaspora: Papers From A Conference of the Centre of Commonwealth Studies. University of Sterling April 1996, pp. 119-140)