Thursday, 23 December 2010


In the good old days of descriptive ethnography children's games and pastimes were treated as a serious matter. The now-defunct handbook Notes and Queries on Anthropology declared games of all kinds to be "worthy of special study", and advised fieldworkers to join in and learn to play them, as well as recording them as fully as possible (1929: 321; 1951: 334). When I began fieldwork in Usangu in 1980, Alison Redmayne encouraged me to collect children's riddles for the linguistic and cultural information they contained, much as she and other Oxford-trained anthropologists of her generation had done. But after a tentative start I switched my attention to the verbal 'games' that adults played, making no more than occasional notes on children's pastimes whenever I came across them. Sometimes they were difficult to ignore, like the boys who playfully parodied the ethnographer-as-photographer with their own clay models of my camera.

I discovered Iona and Peter Opie's The Lore and Language of Schoolchildren (1959) in the mid-1990s and haven't tired of dipping into their work ever since, not least because it brings forth memories of my own childhood in the north-west of England. In 1998 I went through my Sangu dictionary and fieldnotes from Usangu plucking out the snippets that I'd written about children's toys, games and other pastimes. Here are some lightly edited extracts from my compilation:

Counting songs
In 1981 I recorded one childhood counting song from two adult Sangu, a man and a woman, who sang it in unison:
1. mbwe
2. yangaya
3. yapusi
4. yalima
5. tumbwe-tumbwe
6. manguluma
7. kinengu
8. nengulya
9. kilosa
10. chumilya
These apparently nonsensical 'numbers' bear no obvious relation to the Sangu cardinals in everyday use, nor other known Sangu counting systems (a cominatorial verbal sequence in use in the late 19th century and a finger counting system which is still employed). The tenth term, chumilya, appears to be a playful variant of Swahili kumi, 'ten', or one of its Northeast Coast Bantu cognates (/ch/ is not an inherited Sangu phoneme, nor is it usually retained in loanwords except in some proper names). The ninth term is possibly derived from the place name Kilosa, located in Usagara on one of the old caravan routes to the coast, though why it should be so derived is obscure (/ki/ also has a very restricted distribution in Sangu, and is probably indicative of a loan in this case: this also applies to the term for seven, kinengu). The other terms in the list share the regular phonological characteristics of Sangu, and playful etymologies might be suggested for some of them, though this would be no more than guesswork.

The Sangu version of jacks (imdodo) that I witnessed on numerous occasions in Utengule in 1980-81 was played with twelve small stones as the jacks and a small round fruit of the ilihuluhulu (Capparis tomentosa) shrub as the ball. The game is played as follows. A small hollow is made in the earth and the stones are placed inside it. The fruit is then thrown up in the air and before it is caught a single stone must be scooped up in the other hand. If more than one stone is scooped up then the extra stone(s) must be returned to the hollow on the next throw. When all twelve stone have been removed one by one in this way, they are returned to the hollow and the same procedure is repeated, this time the stones being removed in twos. Next time they are removed in threes, then fours, and so on, until all twelve stones have to be scooped up in one go. If the player errs at any point (for example by failing to catch the fruit) then it is the turn of another player (or the same player if she is playing alone) to start from the beginning again. This game is most commonly played by girls. A twelve-stone game was also observed and photographed by Kubik in 1976 in the vicinity of Mahango-Mswiswi (1978: 103). This is a version of the game called 'fivestones' by Opie and Opie (1997: 56-72).

Where's the cow?
The following guessing game is one of the most common Sangu children's games, and I watched it being played in both Utengule and Luhanga in 1981. The game is played by two opponents (often accompanied and assisted by other children) who face one another and take turns to guess in which hand the other is concealing a small stone. The stone is referred to as ing'ombe, a cow, and each turn begins with the following exchange:

Player concealing the cow: hilili, hilili (the name of the game)
Player trying to guess where it is: ing'ombe, ing'ombe, 'cow, cow'
First player again: ing'ombe yili kwi?, 'where's the cow?'

If a player guesses correctly, then he or she is entitled to advance his or her counter (also a stone) towards a goal drawn on the ground. This goal takes the form of a circle or small hollow in the middle of a pitch of typically five or six concentric circles (sometimes not completed at the sides, so that the pitch comprises two bands of lines on either side of the goal). The game starts with the two counters at the outer edge of the pitch, being moved progressively towards the centre (across individual lines) at each successful guess. The winner is the player or team whose stone counter reaches the central goal first. When I first recorded this game being played, in Luhanga, children whose turn it was to conceal the 'cow' frequently attempted to cheat by dropping the stone behind their back, so that it was in neither hand. The game is called hilili, presumably after its opening formula, the etymology of which is opaque. The same term is also used in an extended sense by many Sangu speakers to refer to puzzles and riddles in general. Hilili can perhaps be thought of as the quintessential Sangu children's game, not only because of the wider use of its name, but also because it is cast in a bovine idiom.

If I was to rewrite these notes I'd say more about the actual games I watched and the contexts in which they were played. There's a nice account by Marius Fortie (1938: 302-303) of a series of contests and games which he observed being played in the west of Usangu in September 1934, having provided the prize money himself. That's one way of getting people to play games, although it wasn't Fortie's intention to record them in any great detail.  Studying children's games isn't child's play, and I'm not aware of any attempt in East Africa to undertake the kind of research that the Opies did in England, Scotland and Wales, though there are scattered sources from which a compilation might be begun, including the now largely forgotten literature on string figures. In November-December 2002 my Hehe-speaking research assistant, Justin John Kitinye, filled six school exercise books (436 pp.) with descriptions of the games that he knew and had played as a child. I've barely begun to translate these and reflect on their significance, but am very much looking forward to it.


Fortie, Marius 1938. Black and Beautiful: A Life in Safari Land. London: Robert Hale.

Kubik, Gerhard 1978. Recording utamaduni in Tanzania - a field report from Iringa and Mbeya regions, Oct 10 - Dec 14, 1976. Review of Ethnology 5 (11-14): 81-107.

Opie, Iona and Peter Opie 1959. The Lore and Language of Schoolchildren. Oxford: Clarendon Press.

Opie, Iona and Peter Opie  1969. Children's Games in Street and Playground. Oxford: Oxford University Press.

Opie, Iona and Peter Opie 1985. The Singing Game. Oxford: Oxford University Press.

Opie, Iona and Peter Opie 1997. Children's Games with Things. Oxford: Oxford University Press.

Redmayne, Alison 1970. Riddles and riddling among the Hehe of Tanzania. Anthropos 65: 719-813.

The Royal Anthropological Institute 1929. Notes and Queries on Anthropology (5th edition). London: The Royal Anthropological Institute.

The Royal Anthropological Institute 1951. Notes and Queries on Anthropology (6th edition). London: Routledge and Kegan Paul Ltd.

Walsh, Martin 1985. Shisango Dictionary. Unpublished ms. (computer printout), Cambridge, June 1985.

Walsh, Martin 1998. Children's toys, games and other pastimes in Usangu. Unpublished ms. (draft).

Sunday, 12 December 2010


When I was a boy I loved role-playing "Cowboys and Indians". I always wanted to be a "Red Indian" (Native American), and was especially fond of mock-fighting with a bow and arrows. In one overgrown English garden I fashioned my own flightless arrows from the straight stems of Goldenrod (Solidago sp.), and have looked at this plant wistfully ever since. In many parts of East Africa archery is much more than child's play or a weekend sport, and its technology is correspondingly complex. Like many other aspects of African technology, it is also woefully understudied, as I discovered when I set about investigating Mbeere practice in (what was then) Embu District with the help of Silas Kibwece and other local research assistants in 1992-93. I've never found time to finish writing up the results, though I have posted some notes online (Nyaga 1992; Kibwece 1993; Walsh 1993). Last year I tried photographing a selection of the materials that I'd purchased from rural markets or had made for me at the time (unfortunately I've managed to lose my collection of Mbeere bows). Some of the results are shown below. I prefer the close-ups.
Mbeere iron arrowheads
Rusting Mbeere arrowhead
Painted flight bindings on arrowshafts
Decorated flight bindings
Mbeere arrows with iron arrowheads (minus fletching, eaten by insects)
Mbeere wooden arrows with detachable points
Detachable wooden points, barbed and plain
Splayed shafts and detachable points of wooden arrows
Close-up of splayed shaft and wooden point
Fletching of wooden arrows
Fletching and nock of arrow showing bindings
Close-up of nock
Mbeere leather quiver
Top end of leather quiver showing decoration
Cover of leather quiver
Open quiver showing cover and arrows inside
Close-up of open quiver and cover
Close-up of open quiver with arrows inside

Kibwece, Silas 1993. Mbeere Archery. Unpublished manuscript notes on Mbeere archery written by Silas Kibwece in May 1993 for Martin Walsh, Embu. [Answers to a list of questions asked about Mbeere archery and its technology (to fill in gaps in Walsh 1993).]

Nyaga, Alfred 1992. Mbeere Hunting, Trapping and Fishing. Unpublished manuscript notes on Mbeere hunting, trapping and fishing practices written by Alfred Nyaga in December 1992 for Martin Walsh, Embu.

Walsh, Martin 1993. Mbeere Archery and its Technology: A Preliminary Description and Analysis. Unpublished ms., draft, May 1993.

Sunday, 5 December 2010


Taita bag, sketch by James Walsh (1989)
One of the things that fascinates me about public institutions and other large organisations is the way in which otherwise subordinate departments, offices and teams can develop their own collective identity and ways of working that subvert official procedures and create spaces for autonomous (but not necessarily subversive) action. I saw this when studying a local government department in north-west England in 1979, and was especially conscious of it when working as the field manager of a UK-funded project in Tanzania two decades later (more of which, perhaps, in future posts). A nice historical example of this particular kind of institutional resistance-cum-creativity was the practice that developed among British colonial officers of keeping a "goat bag" or hidden fund for use in emergencies and to cover other contingencies that were unlikely to receive official approval. Here's Elspeth Huxley's account of its origin:
  It was in the NFD [Northern Frontier District] that that useful Kenyan institution the goat-bag was born. In the days of its conception, tax was paid in goats instead of money. Most of the goats were fed to KAR [King's African Rifles] askaris. Every one had to be accounted for to the Treasury in Nairobi. But that department's officials overlooked the fact that in any given flock of goats, births as well as deaths will occur. The district officer who started the first goat-bag did not overlook it, and gradually built up a flock that had no official existence, and that could be converted into cash by selling the animals. He also discovered that by drying and marketing the skins, his unofficial fund could be augmented. Every DC [District Commissioner] in the country was continually being confronted by a need for cash to meet unexpected demands unlikely to be sanctioned by the Treasury. The goat-bag proved to be the answer. It was not long before every DC in the country had latched on to the idea. Each commissioner kept a meticulous account of how the money was spent, which he locked away in his confidential safe, so that when the auditors came round on their annual examination of the station's accounts, the secrets of the goat-bag were concealed from their eyes.
   Every DC could give examples of the uses of the goat-bag; here is a single one. On the road between the Tanganyikan border and Nairobi, some unknown person halted his car to fire at a zebra standing on the skyline, missed and drove on. The bullet proceeded on its way until it dropped through the roof of a hut and into the head of a young Maasai girl, killing her stone-dead. Her family, according to custom, demanded blood-money: but who was to pay? In the Maasai view there was no doubt: the Government. The Treasury disclaimed all responsibility. The elders came angrily to the DC at Kajiado, who feared serious trouble should the claim not be met. The Treasury remained adamant. Luckily, the goat-bag at Kajiado was a fat one. The DC handed over twelve head of cattle and the crisis passed. (1985: 166-167)

This last anecdote was given to Huxley by Robin Wainwright (1985: 254, fn. 5), who was Kajiado DC in 1945-46.  Here's another example from the same period, related by L. S. van Aardt in a recent letter to Old Africa magazine:

  After the war I joined the Tanganyika Agricultural Department but was seconded to the Kenya Government to do locust control in the Northern Frontier District. I spent the most blissful time of my life stationed at Garissa. Abundant game covered the area. Since water was scarce, the game lived mostly near the Tana River or around a seasonal water hole at Kolbio on the Somaliland Border. The D.C. and one Policeman ran the boma. The former, Symes Thompson, used to smuggle in Joffes gin from Somaliland, which he sold to the policeman and I for five shillings a bottle. When his request for money to build a swimming pool was refused, he instructed the policeman to arrest some well-known scoundrels and made them dig a suitable hole. He bought materials using money from the "Goat Bag." (van Aardt 2010)

Construction of a sheep and goat dip (FAO)
A quick search of Google Books produces other examples in the literature on colonial East Africa, and even further afield. I don't know whether anyone has ever tried to pull this documentation together, but it would be a great research project, especially if combined with work in the archives and interviews with former colonial officers. This begs the question whether the tradition of the goat bag simply died out with colonialism or survived in some postcolonial contexts. Has it simply morphed into personal corruption? Or been reinvented by contemporary officials with more enlightened interests than the lining of their back pockets? One place to look is surely at the history and ethnography of community fundraising or harambee in independent Kenya, with its well-known propensity for serving both the collective good and the selfish desire of corrupt individuals and groups. The example given by van Aardt suggests that the potential for corruption was always present anyway in colonial practice. And there's another subject for research.

Another place to look is within modern aid programmes and projects, institutions that inherited some of the functions (and personnel) of the colonial technocracy (I can feel another dissertation proposal coming on). In my own work on projects I've never seen a goat bag sensu stricto, but have experienced 'creative accounting', by which I mean the 'bending' of laid-down procedures and rules (e.g. strictures regarding what might or might not be purchased under particular budget headings). When this is done in good faith, for the benefit of the project and its officially-sanctioned goals, we can see the spirit of the old goat bag in action. But when it's done primarily for private gain, the jury of public opinion is likely to be less forgiving -- as Elspeth Huxley was when reporting the suspicion that one former DC had made off with a government-owned lawnmower (1985: 146-147). There are, of course, many shades of grey between official audit and public morality, as the UK parliamentary expenses scandal has so amply illustrated.

The most outrageous case of creative project accounting that I've seen was in the drylands of Tharaka, where the eastern foothills of Mount Kenya slope down to the Tana River, and a British TCO (Technical Cooperation Officer) had evidently dug in the footsteps of Symes Thompson. First, let me quote from a potted history of the project in question:

The Ministry of Livestock Development, supported by the British Government’s Overseas Development Administration (ODA) looked at developing a different type of dual-purpose goat more suitable for arid and semi-arid lands (ASAL). The approach taken at the Marimanti Breeding Station in Tharaka-Nithi District (1983-1989) was to acquire several hundred Galla goats from northern Kenya and select for growth and mothering ability (Skea, 1989). The station was well-funded and as soon as this ended the manager left and the farm quickly fell into disrepair. The goats unfortunately developed beznoites [i.e. besnoitiosis], a disease which is hard to control, forcing many goats to be culled. Eventually all the goats were sold or stolen and the buildings are now used as a district headquarters. (Peacock 2007: 7)

Source: Bill Forse, Where There Is No Vet (Macmillan, 1999)
Marimanti is remote enough to deter all but the most determined auditor from visiting. Fieldworkers are another matter, and I pitched up at the Marimanti station in March 1993 for a meeting with the Kenyan manager of the Goat and Sheep Breeding Project (felicitously abbreviated to GASP). Ernest Njuguna Mbogo proved an excellent host and interviewee, and I came away with detailed notes on agricultural development and livestock production in Tharaka. But the setting of our interview was surreal: we sat by the side of the dilapidated project "goat dip", constructed by a former TCO, ostensibly for the purpose of bathing the doomed hybrid goats in insecticide. At least that's what project accounts submitted to the BDDEA (British Development Division in Eastern Africa) office in Nairobi had declared, or so it was said. Except that this particular goat dip bore a distinct resemblance to an empty swimming pool: its not-so-caprine dimensions, the well-finished tiles, what looked for all the world like a diving-board, and the poolside chairs on which we were lounging...


van Aardt, L. S. 2010. Old Eldoret: early days in East Africa. Letter to Old Africa magazine (published in full on the Editor's blog).

Huxley, Elspeth 1985. Out in the Midday Sun: My Kenya. London: Chatto & Windus.

Peacock, Christie 2007. The Goat Model: a proven approach to reducing poverty among smallholder farmers in Africa by developing profitable goat enterprises and sustainable support services (FARM-Africa Working Paper No. 9). London: FARM-Africa.