Languishing in a hospital bed recently, my feverish mind swirled with memories of earlier hospitalisations - my own and others' - many of them in East Africa. The institution I came to know best, as both patient and visitor, was
Mombasa Hospital, memorably situated on the sea front next to
Fort Jesus. If I close my eyes I can still feel the strong breeze blowing off the Indian Ocean into an upstairs ward. And I'll never forget my first visit there, the last time that I saw Jim Allen.
James de Vere Allen was an independent scholar who played a key role in the establishment of the
Lamu Museum. He became an authority on coastal history and culture, and is perhaps best known now as the author of the posthumously published
Swahili Origins: Swahili Culture and the Shungwaya Phenomenon (
1993). When I first came to live in Mombasa in 1985 Jim was ensconced on his farm in Kwale, to the south-west of the town. For a long time I knew of him only through other researchers, all of whom described him in glowing terms. I only recall meeting Jim twice, once when he came to find me in the Jubilee Building on Moi Avenue, where I worked. A diary entry for 28 April 1988 records that I met him in the morning (was this the same occasion?) and gave him a copy of
William Frank's
Habari na Desturi za Waribe (
1953) and (I presume from the cursory entry) Krapf and Rebmann's
Nika-English Dictionary (1887) - the latter a photocopy for him to copy and then return. I also gave him my unpublished paper on '
Mijikenda origins' (written in March 1987, but not published until
1992); we must have discussed this at some point, as well as my research on the historical influence of the Segeju on the Mijikenda, but I can't now remember when.
I didn't know that Jim was seriously ill until May 1990, a month before he died. I went to see him in Mombasa Hospital on Thursday 7th June. He looked terribly weak, but found the energy to talk briefly about our different views of Segeju history. We agreed to disagree and continue the conversation when he had "turned the corner" and got better. I squeezed his frail hand and left. That was the last time I saw Jim: he died the following Wednesday, the 13th June 1990. Although I didn't know Jim well, I felt cheated by his death and its rude interruption of our discussion of coastal history. I've regretted ever since that we didn't spend more time together.
As it happens, Jim had already drafted his magnum opus,
Swahili Origins. A year or so before he died, knowing that he was seriously ill, he asked the anthropologist John Middleton to edit the sprawling typescript.
The resulting book was published in
1993 with a somewhat apologetic preface by Middleton explaining its genesis and warning the reader of its author's unorthodoxy. Like other researchers I was and remain very critical of
Swahili Origins. Here's part of what I wrote in
a review article at the time:
Before his untimely death in 1990 [...]
, Jim Allen had already made an outstanding contribution to Swahili studies. Among a variety of achievements, he played a key part in helping to overturn the traditional image, which still lingers in popular literature and tourist brochures, of Swahili culture as an exotic blend of Oriental and African influences in which the former have consistently dominated the latter. Linguists had long realised that Swahili was not a creole or pidgin combining Arabic, Persian and Bantu elements, but a Bantu language with a significant proportion of recent Arabic and other loanwords. Allen was among the first to trace the cultural and historical corollaries of this linguistic observation. In a series of articles published from the 1970s onwards he attacked the traditional emphasis upon exogenous, oriental influences ([an emphasis]
encouraged, it must be said, by many Swahili speakers themselves), and highlighted instead the indigenous, African, roots of Swahili culture. The detailed implications of this paradigm shift are still being worked out by younger researchers, many of whom benefited directly from Jim Allen's personal interest and encouragement.
The manuscript of Swahili Origins, having been edited by John Middleton and published posthumously, presents Allen's view of Swahili genesis and his alternative to the orientalist interpretation of Swahili history. The core of this gloriously detailed account is a hypothesis first outlined in his papers published in 1983 and 1984. Allen argues that Swahili society and culture, as well as various institutions among the Mijikenda and other neighbouring people, originated in the interaction between a polyethnic blend of pastoralists and their clients, including indigenous hunter-gatherers and Bantu-speaking farmers. The dominant alliance of pastoralists included Cushitic speakers and the Segeju (whose contemporary representatives fall into two related groups, one speaking Daisu, a Central Kenya Bantu language, and the other speaking dialects of Swahili and/or Mijikenda). The culmination of this alliance was a unified state, called Great Shungwaya by Allen (after the name of the mythical homeland of various Swahili, Mijikenda and other coastal traditions) and estimated by him to have flourished some time between the ninth and fourteenth centuries. Its sequel is hypothesised to have been a series of so-called successor states, relatively familiar to us from the historical record and ushering in the ethnic map of the region as we now know it.
This hypothesis and Allen's idiosyncratic use of different historical sources to support it have already been widely criticised, indeed in some cases dismissed out of hand (as by Spear, 1983). In Swahili Origins Allen has expanded and amplified many details of his account, but done absolutely nothing to answer his critics, who would see Great Shungwaya as no more than a fantasy of his fertile imagination. It is all too easy to find fault with Allen's arguments and cast him as a kind of Erich von Däniken of Swahili studies. It is perhaps more useful to consider why Allen, a master of historical detail, should have clung to such a flawed hypothesis, and what subsequent influence it has had. Despite his work for the Lamu Museum and his living for many years on the Kenya coast, Allen was essentially an armchair historian, which helps to explain both his skill in uncovering obscure documentary references and his largely uncritical interpretation of them. Ironically his successful critique of the orientalist thesis of Swahili origins encouraged his promotion of an inflated account of the indigenous alternative, constructing an imaginary state to counter the traditional picture of coastal civilization as a ready-made import from overseas. To this extent Allen was an unwitting prisoner of the tradition which he sought to overthrow, and his understanding of the variety of everyday coastal life remained subordinate to a continuing fascination with the classical tradition of urban Swahili culture. (
1993: 143-144)
Flawed though Jim Allen's thesis is, I remain captivated by the detail in his work, as a source for further exploration if nothing else. For many years I wondered what had been omitted from the original typescript. John Middleton described his editorial work as follows:
My actual 'editing' has comprised several actions. First, the shortening of the original manuscript by over a quarter, almost entirely by the deletion of repetitious sentences and paragraphs. Secondly, the omission of a long final chapter that was intended to be the summary of a second part of the book, seemingly not written: this chapter was clearly cobbled together during Jim's final illness, and is inconsequential, without footnotes, and wanders away from the Swahili: it is not in any sense a conclusion to the whole. Thirdly, the omission of four appendices that are marginal to the main account. In brief, I have shortened, but not 'edited' or 'revised', and I have tried to retain as many of Jim's comments on others' work and approaches as I could. This is a very personal work and I have tried to keep that characteristic. (
1993: ix)
Then, in autumn 2004, I met another of Jim Allen's friends at a conference in Oxford, the historian Ed Steinhart, and he told me that he had a copy of the unedited draft of
Swahili Origins. On his return to the States Ed scanned the typescript and announced its availability to interested scholars on the
H-Africa discussion network. Both of us also made electronic copies for deposit in different libraries (
the University of Florida, Rhodes House in Oxford, the British Institute in Nairobi). Comparison of this draft with the published work shows that aside from making the major excisions that he noted, John Middleton (who died himself in February 2009) had indeed remained largely faithful to Jim's text. The final chapter that he left out was entitled '"Galla incursions" and the demise of Shungwaya': readers can judge for themselves whether this omission and that of the appendices was justified. The version of this chapter in Ed Steinhart's typescript includes foootnotes; John Middleton's comments about their absence suggests that he was working with a different text. I must confess a personal interest in this chapter and especially its footnotes, because that's where Jim made reference to my draft paper on '
Mijikenda origins' and arguments about the influence of the Segeju on the Mijikenda. But seeing my name in his typescript is little comfort. I'd much rather that he'd turned that corner and stuck around to continue our conversation.
References
Allen, James de Vere 1983.
Shungwaya, the Mijikenda, and the traditions.
International Journal of African Historical Studies 16 (3): 455-485.
Allen, James de Vere 1984. Shungwaya, the Segeju and Somali history. In T. Labahn (ed.)
Proceedings of the Second International Congress of Somali Studies (Volume II:
Archaeology and History). Hamburg: Helmut Buske. 35-72.
Allen, James de Vere 1993.
Swahili Origins: Swahili Culture and the Shungwaya Phenomenon [edited by John Middleton]. London: James Currey.
Frank, William 1953.
Habari na Desturi za Waribe. London: Sheldon Press.
Horton, Mark 1990. [Obituary of] James de Vere Allen 1936-1990.
Azania 25: 108-109.
Krapf, L. and J. Rebmann 1887.
Nika-English Dictionary. London: Society for Promoting Christian Knowledge.
Middleton, John 1993. Preface. In
Allen 1993: vii-ix.
Spear, Thomas 1983.
A Reply by Thomas Spear [appended to
Allen 1983].
International Journal of African Historical Studies 16 (3): 484.
Walsh, Martin 1992.
Mijikenda origins: a review of the evidence.
Transafrican Journal of History 21: 1-18.
Walsh, Martin 1993.
Becoming Swahili and Mijikenda [a review article].
Azania 28: 143-147.
