East African Notes and Records

history, ethnography, ethnobiology, linguistics...

Thursday, 29 August 2013


by Martin Walsh

The relevant passage in W. E. Taylor's African Aphorisms (1891: 93)
In an earlier note on A click in Swahili I drew attention to the Reverend W. E. Taylor's description of an interjection/ideophone in the Mvita (Mombasa) dialect that contains a dental click (1891: 93). Taylor struggled to render this unusual sound in print, and offered three different transcriptions for it in his African Aphorisms, as can be seen from the entry that is reproduced here. Judging by the eccentric composite -- "Mng'wpc" -- that appeared in Burt's Swahili Grammar and Vocabulary (1910: 13, fn. 1), Taylor was never able to settle on a satisfactory solution. This wasn't really his fault, given that there was little agreement on how to write down clicks and the languages using them at this time.

Abdulaziz 1979
When I posted my note in June 2010 I wasn't aware of any other references to Taylor's Mvita click-bearing interjection, though I did know of the occurrence of a nasalised dental click in similar contexts in different dialects of Digo, spoken to the south of Mombasa island (Walsh 2006). Last night, however, leafing through the edited verses of Mombasa's most illustrious bard, Muyaka bin Haji al-Ghassaniy (c.1776-c.1840), I stumbled across what seems to be an earlier example of its use. Most of Muyaka's poetry is thought to have been written between roughly around 1810 and his death. We owe the survival of his corpus to the extraordinary collaboration between the Mombasa scholar Mwalimu Sikujua bin Abdallah al-Batawi (who collected and wrote down Muyaka's Swahili poems in Arabic script) and the same Rev. Taylor (who transliterated them in Roman script and added linguistic and other notes). The originals are in the School of Oriental and African Studies (SOAS) in London: I owe my own access to Muyaka's poetry to Mohamed Hassan Abdulaziz's study (1979), which I never tire of dipping into.

The poem in which the Mvita interjection appears is in his compilation of the many 'Miscellanous Verses of Muyaka' (1979: 156-335). Abdulaziz's brief gloss on the whole poem is that 'Muyaka here makes fun of certain Wachangamwe women folk (one of the Three Tribes of Mombasa) who are pictured as having maltreated a certain Bacheni' (1979: 287). The Wachangamwe were the inhabitants of Changamwe and thereabouts on the mainland to the west of Mombasa island; in the 19th century they spoke their own variety of the Mvita dialect (Lambert 1958), and were evidently looked down upon by their more cultured island cousins. The identity of Bachemi or Ba Cheni (Baba Cheni, 'Cheni's father'?) seems to have been forgotten: all we have to go on is the poem itself. It is reproduced below, complete with the inconsistencies of spelling in Abdulaziz's edition -- note though that the italicisation of dental /t/ and /d/ is Taylor's. In the book Taylor's transliteration and Abdulaziz's English version are on facing pages (1979: 286-287): here each Swahili verse is followed immediately by its translation. Perhaps understandably, Abdulaziz didn't attempt to translate the poem's title:


Habari ya Wachangamwe niwambiapo mwateka,
Watoshile kamwe kamwe si wake katika rika
Wake washinda waume siombe wakihizika;
Ba Cheni walimshika wakamfanya mng'wa!

Digo mother and child, 1985
When I tell you about the Wachangamwe you first laugh.
They exceed the bounds of acceptable behaviour!
Women can be worse than men when they go astray,
They got hold of Bacheni and bit his lip "Mngwa".

Ni watenzi Wachangamwe hutenda vikitendeka,
Wajipatishile sime zatinda kama kitoka
Ni kweli wake si wamwe, siombe wakipotoka,
Ba Cheni walimshika wakimtinda mng'wa!

The Wachangamwe are people of action, what they do they do well!
Their short swords stuck in their belts, cut clean like hatchets
It's a fact that women are mischievous, beware of them when they get out of hand.
They got hold of Bacheni and cut his lip "Mng'wa."

Ukitakapo hakika kamwulize Mwanasha;
Ati aliyekipika, taa asijaiwasha;
Achondoka kutandika awele kubishabisha.
Habariye isikwisha na kumtenda mn'gwa!

If you want to know the truth go and ask Mwanasha
Who was supposed to have been cooking even before she had lit the lamps.
But when she went to make the beds she was already nagging?
And the whole thing ended with her biting "Mng'wa."

Alipoakimrudi na maneno kumwambia,
Alikitunga mkadi ili kwenda kutembea;
Akema kumradidi una kijana walea,
Awele kumruk'ia na kumkata muomo mng'wa!

When he tried to tell her off and give her a good talking-to
She was stringing the mkadi flower, preparing to go out;
When he went on at her, saying you have a baby that you're (supposed to be) nursing,
She jumped at him and bit off his lip "Mng'wa".

I don't know whether the variant spellings of mng'wa are in Taylor's transliteration or errors introduced during the preparation of Abdulaziz's book. He glosses it as an 'onomatopoeic expression for cutting' (1979: 286, fn. 4), but to me this looks like an over-interpretation influenced by the lines in which it occurs, two of which refer explicitly to cutting. Instead I suggest that it should be read as an approximate rendering of the same interjection-with-a-click-consonant that Taylor tried to describe in African Aphorisms and that he defined as 'an expressive protest against impertinence' (1891: 93). In Muyaka's poem we might think of it as roughly equivalent in meaning to our own 'Hah! So there! Like it or lump it!' Without seeing the relevant archival material it's difficult to know why Taylor himself didn't draw the connection (or why Abdulaziz didn't pick up on it), but it may be that Mwalimu Sikujua's Arabic transcription of the Swahili interjection masked its distinctive enunciation, and that Taylor hadn't yet heard it being spoken. In any event, I've tried my hand at a more idiomatic translation of Muyaka's poem that incorporates this suggested new reading. I'm not an expert in 19th century Swahili poetry or the language in which Muyaka composed his verse: the Mvita that he spoke in Mombasa and the northern dialect forms that he blended in for poetic effect and credibility. But I've drawn on Abdulaziz's text and notes (it's not clear which of these might have come from Taylor), and like him, happily declare the tentativeness of my efforts. So, take it or leave it!

Screw pine, mkadi
You'll laugh when I tell you about the Changamwe
They're just too much and their wives don't act as wives should
The women have the upper hand, Oh beware their shamelessness
They seized Ba Cheni and treated him so!

The Changamwe are doers, and do what they set out to
They wear short swords that cleave like sharp axes
It's true that their wives aren't well-behaved, Oh beware when they misbehave
They seized Ba Cheni and slashed him so!

If you want the truth, ask Mwanasha,
Who was supposed to be cooking, but hadn't lit the lamp,
And carried on quarrelling when she went to make the bed
Which all ended up with her treating him so!

When he came back at her and told her off
She threaded a fragrant flower,* ready to go out
And when he complained that she had a child to look after
She turned on him and cut his lip so!

* A reference to the scented white flower of the Screw pine, Pandanus kirkii (Swahili mkadi)

Despite the superficial similarity, I don't think that there's any link between this Mvita interjection and another expressive nugget that occurs in imprecations in 19th century Swahili poetry, the use of Mngwa- or ngwa- to prefix subjunctive verb forms (Biersteker and Shariff 1995: fn. 254). This has been convincingly explained as a euphemistic contraction of the phrase 'Mungu a-', 'May God...' (Sacleux 1939: 683). Otherwise Sacleux seems to have missed Taylor's interjection and click.


Abdulaziz, Mohamed H. 1979. Muyaka: 19th Century Swahili Popular Poetry. Nairobi: Kenya Literature Bureau.

Biersteker, Ann and Ibrahimu Noor Shariff (eds.) 1995. Mashairi ya Vita vya Kuduhu: War Poetry in Kiswahili Exchanged at the Time of the Battle of Kuduhu. East Lansing: Michigan State University Press.

Burt, F. 1910. Swahili Grammar and Vocabulary. London: Society for Promoting Christian Knowledge.

Lambert, H. E. 1958. Chi-Jomvu and Ki-Ngare: Sub-dialects of the Mombasa Area (Studies in Swahili Dialect III). Kampala: East African Swahili Committee, Kampala College.

Sacleux, Charles 1939. Dictionnaire Swahili-Français. Paris: Institut d'Ethnologie.

Taylor, W. E. 1891. African Aphorisms; or, Saws from Swahili-land. London: Society for Promoting Christian Knowledge.

Walsh, Martin T. 2006. A Click in Digo and its Historical Interpretation. Azania 41: 158-166.

Sunday, 18 August 2013


  by Martin Walsh

My Giriama copper pendant
There's a copper pendant hanging on my wall, one of the few examples of crafts(wo)manship that I've kept from Mombasa. I love its weight (c.75 g) and slinkiness: in my hands it feels like a miniature piece of chainmail. I can't remember now exactly when and where I bought it; maybe in the handicraft shop of Tototo Home Industries on Msanifu Kombo Street; or perhaps Labeka, the well-stocked craft shop on Moi Avenue whose owner was a member of Tototo's board. I worked with Tototo in 1985-88, studying the women's group programme that this Mombasa-based NGO ran with the support of World Education in Boston. The women's groups, scattered throughout what was then Kenya's Coast Province, supplied some of the local handicrafts that were sold in both shops (McCormack et al. 1986: 50-55; Walsh 1989: 375-376), including pendants made by Giriama women.

Kamba girl's apron (from Lindblom 1920)
I wrote a little about these pendants in my study of the Tototo-assisted women's groups and their enterprises (Walsh 1986; abridged in McCormack et al. 1986). They were generally referred to in Giriama (aka Giryama) as ndale, a name that one woman told me properly referred to the copper wire that they were manufactured with. The Reverends Krapf and Rebmann defined ndale as brass (or copper) beads (1887: 299), and noted the expression mudzele wa ndale as the name for a Kamba women's leather apron decorated with the same, and in Rabai the border of beads around a woman's dress (1887: 268). W. E. Taylor (1891: 16) likewise recorded ndale as a Giriama name for brass beads; while making it clear that the generic term for brass, ng'andu, could also refer to copper (ng'andu t'une, 'red brass', 1891: 20, 28). In her Giryama-English Dictionary Florence Deed also defines ndale as 'brass beads' (1964: 71), and has an entry for ngudhi, a 'neck ornament of ndale stitched on leather' (1964: 74).

These names are interesting because they provide clues to the history of these crafts and this particular kind of ornamentation. The Giriama term ndale is clearly related to Kamba nthale, glossed in a modern dictionary as 'brass or iron ornaments affixed to the clothes of women' (Mwau 2006:195), and in an older work as 'little brass or iron cylinders sometimes used in making old women's aprons and as ornaments on the straps of the nthũngĩ', a kind of small woven bag (A.I.M. 1939: 159, 161). According to Charles Hobley only the brass beads on a married woman's apron or kimengo were called nthale: iron beads had their own name (1910: 72). Gerhard Lindblom called these brass beads nzale (1920: 374), a local variant (or mistake) for nthale. He also illustrates a young women's neck collar called nguthi (1920: 377), a name clearly related to the Giriama ngudhi. In this case though the collar is made of iron chains, not beads or cylinders sewn onto leather. The Kamba are inland neighbours and have long been traders among the Giriama and other Mijikenda, and it may be that these items of adornment and their names were taken from them. Or they might have been introduced by the historical Segeju, a group of people closely related to the Kamba, who had a significant impact on Mijikenda society and language in the middle of the last millennium. Linguistic evidence suggests that the Segeju were adept at metal-working and transferred a number of related practices to the Mijikenda and the Giriama in particular (Walsh 2013: 35).

'Gohu elder's amulet', Cantor Art Center, Stanford University
(Accession No. 2012.114)
The ndale pendant in my possession is made of copper beads (twisted from wire) and copper chain threaded together with twine; there are also a few odd brass beads strung together with the copper ones. There's a pendant in the collection of the Cantor Art Center at Stanford University that is very similar in form, but is stitched on leather and includes a central panel of three cowrie shells. The online catalogue describes it as a 'Gohu elder's amulet', the gohu being one of the 'secret societies' that senior Giriama elders could be initiated into. Collected in 1978, it looks a lot older than this, or at least grimy and worn, so much so that it's not clear whether it was made of brass or copper beads and chain. In The Wrath of Koma, Maurice Kambishera Mumba's novel about tradition and tragedy in late colonial Kaumaland, the fictional Jibana 'medicineman' (mganga) Mokoli Mwiru uses a powerful talisman called 'Chiwele', inherited from his grandfather, to access the world of spirits and divine the spiritual affliction and possible courses of action open to his Kauma patient, Old Chembe (Mumba 1987: 6-13, 21-24, 112-113, 143). Chiwele, used here as a personal name, is described as 'a leathery object coated with small cowrie shells, which was dangling from his neck at the end of a short chain' (1987: 6). The cognate term kiwele, plural viwele, was used by modern Giriama producers as the generic name for the women's copper and brass pendants that they made (see below) and that many women first acquired when they were married. This perhaps represents the wider application of a term that was once more restricted in use, and referred originally to the chains hanging down from the amulets. Elsewhere in his novel Mumba refers to ndale as a kind of 'traditional necklace' worn by women (1987: 32, 146), and this indeed seems to be the more widely used name.      

Bomani in 1985
I could probably ferret more information out of the literature, and Mijikenda colleagues and scholars more familiar with their culture could no doubt tell me much more (please do!). I wasn't aware of any of this when I was first working for Tototo. I did, however, learn something about the contemporary making and marketing of Giriama ndale pendants when I was undertaking the third of my women's group case studies in a village some 20 km north of Malindi. Bomani Women's Group had its origins in an adult education class (ngombaru) that started in 1973; Tototo began working with the group in 1978, at the very beginning of its women's group programme. When I arrived to study the group in late 1985 it was best known for its bakery project: for a time this had been the showpiece of Tototo's programme. But the first enterprise that Tototo had fostered in Bomani involved the production and sale of ndale. Here's what I wrote about it in my 1986 report:

  More lucrative [than a group farm] while it lasted was the handicraft trade initiated by Tototo. This was based upon the production of traditional Giriama ndale necklaces for the tourist market. The women bought lengths of ndale copper wire from specialist producers and fashioned these into viwele, heavy pendants, and virangi, with brass and coloured beads added. 1½ feet of wire, bought for 1 sh, was enough to make 2 or 3 viwele, the commonest product, sold to Tototo for 10 sh each, while virangi fetched 20 sh. Women report being able to make up to 100 or 150 viwele in a week and at one point group members engaging in this trade are recorded as making between 35 sh and 500 sh each in a fortnight. Tototo shop records show that in 1978 14,969 sh was paid out to individuals and 920 sh to the group, which took the proceeds from one necklace in every batch an individual produced. This money was ploughed back into the bakery along with the money raised from group subscriptions. The women, however, produced more necklaces than the Tototo shop could sell, and after 2 years the trade came to a halt along with other forms of handicraft production which Tototo had attempted to introduce. Looking back upon this enterprise group members blame Tototo for its failure, an experience similar to Mkwiro's [another group that I studied]. The local, Malindi, market for necklaces remains small -- they are bought by Kamba middlemen and sold to tourists in the town -- and ndale production is not a significant source of income in Bomani; much the position when the group's trade with Tototo started. (1986:

A young member of Bomani Women's Group in 1985
There's more detail in my field notes, though group members gave different accounts of the sequence of events and the role of the various actors. According to Bomani's chairwoman, they were shown how to make ndale in 1978 by Daudi Mtingi, one of the three older men (wazee) who had started the adult education class that gave rise to the women's group. He was already marketing the ndale made by his two wives, who were founder group members, taking the pendants to Kamba middlemen in both Malindi and Mombasa. Tototo started taking them for sale in Mombasa in the same year: the pendants were collected every Thursday or sometimes taken to Mombasa by the group's secretary, Esther Kenga (who later became a Tototo employee).

According to another group member, Mary Ngonyo, they bought strings of ndale beads that had already been made by folding small lengths of copper ('red') or brass ('white') wire around pieces of thread (nyuzi) using a special instrument, and it was these strings that they bought at the price of a shilling for each length of around 1½ feet. The women then added coloured beads (virangi), bought in the local shops, to make neck pieces, and chains (viwele) to make the pendants, which were rather easier to produce. At first they just took them to Malindi for sale, but found it difficult to sell them. It was Esther Kenga who first wrote to and visited Tototo. There were around 20 women in Bomani Women's Group at this time. They were joined by women from the nearby village of Madzayani when they heard that there was a market for ndale, though after Tototo stopped taking the pendants they left to form their own group. As well as ndale, Tototo also persuaded the Bomani women to produce small woven bags, baskets, and straw hats for sale. But all this came to an end when Tototo told them to stop. By this time running the bakery had become the group's main focus. Some women continued to make both copper and brass ndale as well as other items of women's jewellery, and two or three years before I did my research Kamba buyers had begun coming directly to the village to purchase them. The group's chairwoman still kept a few pendants to show guests and in 1985 had presented some to a group of official visitors from Zimbabwe.

Feeding the next generation, Bomani 1985
The failure of collective marketing is an all too familiar tale. Selling local handicrafts was a great idea, but supply and demand weren't always well matched. Tototo's handicraft operation had problems of its own: staff often lacked key marketing and other business skills, and the shop wasn't particularly well run. To make matters worse, some managers were suspected to be dipping their hands into the till or fiddling the accounts -- widespread practices in an era of rampant corruption at every level of society. It's ironic that just as field staff were advising women's groups on how best to manage their various enterprises, their colleagues in Mombasa were sometimes struggling to manage the NGO's own affairs. Although Tototo received extensive help from World Education and other partners, the shop lost the comparative advantage that it once had, and the organisation itself was eventually closed down. Nevertheless, in the heyday of pendant production and marketing a fair number of these attractive handicrafts must have found their way into the hands -- and perhaps even around the necks -- of visitors to the coast. I imagine that they are still made in at least small quantities, though I can't see any evidence of this on the internet.

Kamba girl's necklace (from Lindblom 1920)
Disappointed though they were by Tototo's actions, the ndale episode wasn't an entirely negative one in the history of Bomani women's group and the evolution of local gender relations (and as it happens, greater disappointments were to come, detailed in my 1986 study). There are two particularly heartwarming stories that deserve to be told and remembered in this context.  One is the role that Daudi Mtingi played in fostering and supporting the adult education class, an associated nursery class, and then the women's group -- as well as introducing them to ndale production and marketing, he also provided significant assistance to the bakery business by injecting capital into it at a time when it was floundering, and then selflessly handing the profits over to the group. The second is a story that I've already told in brief in an article about the gendered control and use of income from group activities, which explains why women's income generation doesn't necessarily translate into their empowerment within joint households (Walsh 1987). Following a long personal struggle to attain economic independence and the right to educate her son, Bomani's chairwoman, Jumwa Kiti, encouraged group members to invest their profits from pendant sales in goats, and kept their money in safekeeping until they had accumulated enough to do so. This acted as a check on husbands' ability to commandeer their wives' ndale profits, and though it didn't stop some of them from later appropriating their wives' livestock, it did enable a number of group members to accumulate and exercise a degree of control over assets that wouldn't otherwise have been possible.

It's quite likely that the pendant hanging on my wall was made by one of those women, and possible that it contributed in its own small way to her greater empowerment and the improved welfare of her children. It might equally have been a token of hope that turned to disappointment, as her newly-acquired wealth was seized by a greedy husband or her expectations of greater profit were dashed by Tototo. And whatever the case, it may well have helped to line the pockets of one of that organisation's less scrupulous employees. I can only guess. But I feel much more satisfied having these and other possible histories to ponder than I would have done without knowing anything about my shining pendant's past.  


A.I.M. 1939. A Kikamba-English Dictionary. Compiled by: The Language Committee of the Africa Inland Mission in Ukamba (third unrevised edition, 1970). Nairobi: The Literary Centre of Kenya for The Afrolit Association.

Deed, Florence 1964. Giryama-English Dictionary. Nairobi: East African Literature Bureau.

Hobley, C. W. 1910. Ethnology of A-Kamba and Other East African Bantu Tribes. Cambridge: Cambridge University Press.

Krapf, L. and Rebmann, J. 1887. A Nika-English Dictionary (edited by T. H. Sparshott). London: Society for Promoting Christian Knowledge.

Lindblom, Gerhard 1920. The Akamba in British East Africa: An Ethnological Monograph. Uppsala: Appelbergs Boktryckeri Aktiebolag.

McCormack, Jeanne, Martin Walsh and Candace Nelson 1986. Women's Group Enterprises: A Study of the Structure of Opportunity on the Kenya Coast. Boston: World Education, Inc.

Mumba, Maurice Kambishera 1987. The Wrath of Koma. Nairobi: Heinemann Kenya.

Taylor, W. E. 1891. Giryama Vocabulary and Collections. London: Society for Promoting Christian Knowledge.

Walsh, Martin 1986. Interim Report for a Study of Income Generation and its Effects among Women’s Groups in Kenya’s Coast Province. Report to World Education Inc., Boston.

Walsh, Martin 1987. Buying power? Some outcomes of income for women. Reports (World Education Inc., Boston) 27: 14-16.

Walsh, Martin 1989. Tototo Home Industries: assistance strategies for the future. In C. K. Mann, M. S. Grindle and P. Shipton (eds.) Seeking Solutions: Framework and Cases for Small Enterprise Development Programs. West Hartford: Kumarian Press, Inc. 365-385.

Walsh, Martin 2013. The Segeju Complex? Linguistic evidence for the precolonial making of the Mijikenda. In Rebecca Gearhart and Linda Giles (eds.) Contesting Identities: The Mijikenda and Their Neighbors in Kenyan Coastal Society. Trenton, New Jersey: Africa World Press. 25-51.

Saturday, 17 August 2013


by Martin Walsh

I've just finished The Wrath of Koma, Maurice Kambishera Mumba's entertaining tale about tradition and tragedy among the Kauma, published by Heinemann Kenya in 1987. The back-cover blurb of the paperback introduces the story:

Old Chembe, a Kauma man from the Am[i]dzichenda people of the Kenya coast, is gravely worried about his son, Dundu, who appears to be going mad. He travels across many ridges to see Mokoli Mwiru, the famous medicineman, for a diagnosis of his son's problem. Mokoli Mwiru gives his diagnosis: Dundu, born after intercession to the ancestral sprits, is under the wrathful attack of the spirits who resent the ingratitude of Chembe's family. Mokoli Mwiru prescribes a ceremony of appeasement.

Old Chembe, weak and irresolute, botches the ceremony... What tragic fate awaits him, his son and his larger family as the unappeased spirits seeks their revenge?

I won't spoil the plot here. Although I bought this short novel in Mombasa shortly after it came out, this is the first time that I've read it from cover to cover. It may not be great literature, but I've enjoyed Mumba's story-telling, the tales-within-tales, the dissolution of the text into poetry, and his wry description of rural life and morality in the waning years of colonial rule. The Wrath of Koma is especially interesting for its anthropology, and not just because the Kauma are among the least written about of the Mijikenda. I've mined Mumba's tale in the past for ethnographic and in particular ethnoornithological tidbits (as in my 'Birds of omen and little flying animals with wings'), but there's much more in it that deserves to be exposed and analysed anthropologically. This includes passages that revolve around the gendered perceptions and motivations of spirit possession and the sometimes violent reactions of men to the claims made upon them. Hearing the distant sound of an exorcism ritual (kupunga pepo), the protagonist reflects on his own experience:

As the sound of the pepo drums faded in the background, becoming less and less audible as he plodded on, Old Chembe thought of the costs he had incurred in having pepo [possessory spirits] exorcised from his four wives.

  There were moments when Old Chembe wondered whether his wives had not conspired to feign that they were possessed by pepo so that they could acquire items which he had refused to buy for them. But Old Chembe was not like his eastern neighbour Jaramba, who once flogged his wives with his walking stick to exorcise the demons in them--none of them 'dared' to be possessed by pepo again after that! (p.29)

The text goes on to detail how Jaramba came to administer this punishment, and how Old Chembe acceded to requests for the treatment of his possessed wives despite his scepticism. This is literature grounded in the observation of everyday life, a fictionalised account that conjures up the so-called 'deprivation theory' of spiritual affliction (Lewis 1971: 72-89), and invites further analysis with reference to current understandings of gender-based violence and its genesis in the imbalances of power and gendered patterns of inequality in the domestic and public spheres. It's a pity that The Wrath of Koma isn't more widely known. Maurice Mumba was the Town Clerk of Mombasa when it was published, and didn't follow it up with another novel. I don't know what has become of him since. When I lived in Mombasa I actually bought three copies of his book, one to give to another anthropologist. Although I only used it for many years as a work of reference, I'm glad that I finally got round to reading it.


Lewis, Ioan M. 1971. Ecstatic Religion: An Anthropological Study of Spirit Possession and Shamanism. Harmondsworth: Penguin Books.

Mumba, Maurice Kambishera 1987. The Wrath of Koma. Nairobi: Heinemann Kenya.

Walsh, Martin 1992. Birds of omen and little flying animals with wings. East Africa Natural History Society Bulletin 22 (1): 2-9.

Saturday, 10 August 2013


by Martin Walsh

I'm not a great fan of cockroaches, at least not the urban varieties, and it's hard to forget some encounters. I wasn't too thrilled by the evil antennae that emerged from a dry plughole to welcome me to the house that I'd borrowed on the University of Dar es Salaam campus in July 1980. I didn't jump for joy when I switched on the kitchen light one night (circa 1986) in Guraya, Mombasa, and was greeted by the sight of a chaotic mass of cockroaches of every shape and size swarming over the floor. And I didn't smack my lips when the body parts of a drowned roach surfaced in my soup one lunchtime in the Bella Vista off Moi Avenue (this was well before the seedy Italian restaurant became the bar / nightclub that was recently subject to a deadly grenade attack).

After living for a few years in the sweaty centre of Mombasa (Guraya, then Kibokoni), where I became an expert in spotting cooked cockroach legs, we moved out to the suburban calm of Nyali, with its Bougainvillea blooms and sea breezes. But the roaches still came. That is, until I copied down the following recipe:

Cockroach Killer recipe and chemist's label for the key ingredient
I was given this by a friend and colleague in Nairobi, who told me that she'd been given it by an older couple in the highlands out west (I've forgotten where exactly; Kericho perhaps). It worked so well, especially when placed in the kitchen cupboards, that if I saw an adult cockroach in the house -- the younger generations disappeared completely -- I was able to accuse our Belgian neighbours of being its source (we were living in interconnected maisonettes). I even wondered how much money might be made out of the recipe if production could be scaled up and Cockroach Killer balls marketed commercially. But thanks to the internet, I can now see that boric acid was registered as an insecticide many years ago, and that there are similar home recipes all over the web. So here's the one that worked for me.