Book History 5 (2002) 275-282
[Access article in PDF]
Number Magic in Nigeria
Practitioners of money magic ply their trade quite openly in Nigerian markets and streets. Most run a simple con game called money doubling. The customer puts up a small amount of money, say five naira, which is hidden under a scrap of cloth. Incantations follow, the cloth is whisked away, and ten naira appear. Customers who take the money and depart are well ahead. Many cannot resist playing again, for higher stakes. Eventually a customer has won and risked and won and risked a substantial amount. At this point the doubler gives the customer the cloth package itself, supposedly containing the doubled cash, with the warning not to open it for a specific period of time. When the package is opened, the customer finds the money has disappeared, replaced by scraps of paper. If the money doubler is confronted, the doubler loudly berates the customer for having opened the package too early, thereby ruining the magic. Amused onlookers comment on the greed and ruinous impatience of the customer. The money doubler's angry protestations of innocence are said to be good for business.
Like anyone running a gambling operation, the money doubler operates by understanding the statistics and manipulating them when necessary. Some amount of money has actually got to be doubled, some customers have to walk away richer, in order for his reputation to hold. If there is an [End Page 275] especially irate loser, the doubler offsets the bad public relations by creating a few more winners.
Producing statistics from uncertain data puts the researcher in the money doubler's sandals, and in the customer's as well. We want to invest profitably. We want to avoid accusations of dishonesty or naïveté. We want to make the best of what we have.
I have recently finished a book on the Nigerian literary complex called Bearing Witness: Readers, Writers, and the Novel in Nigeria. For this book I had to generate, gather, and make sense of statistics about a literary field that had been largely unmapped, let alone counted. I strongly believe that this is worth doing, despite the pitfalls and temptations involved. Literary statistics from a poorly documented book culture are like a very rough sketch: some of the lines may be off, but a picture emerges anyway. The following recounts some of my attempts to invest wisely and make the most of what I could find.
In order to do any statistical analysis of Nigerian novels at all, the first job was to create a bibliography. Since the first Nigerian novels (by definition, a novel in English; I excluded the handful of Yoruba novels or Hausa novels) appeared in the early 1950s, it was possible to aim at compiling a complete bibliography and reading all of them. The compilation was interactive; along the way I gave authors and booksellers my list in process and asked for additions and subtractions; I also published an early version of the bibliography in the Journal of Commonwealth Literature, again soliciting and receiving a number of responses. 1 By the time Bearing Witness went to press I had located, read, and coded 476 Nigerian novels, by 261 authors, and these were the basis of the statistics used in the book. I also listed the names of 33 other possible novels that I had been unable to locate, again hoping for responses on these.
Were these all the Nigerian novels? Of course not. In the first place, the number depends on definition. I used sixty pages as a cutoff, for example, which meant that long market pamphlets ended up counting as novels. In the second place, a large and growing percentage of Nigerian novels are self-published, with perhaps five hundred copies run off at a local printer at the author's expense. The author then distributes these to booksellers and hopes for the best. Unless the author is fairly sophisticated, such books never get reported for inclusion in the African Book Publishing Record. If I picked them up, or if the author heard of my...