SNAKES ON THE USANGU PLAINS:

AN INTRODUCTION TO SANGU ETHNOHERPETOLOGY

Martin T. Walsh

Natural Resources Institute, Chatham, and

School of African and Asian Studies, University of Sussex, U.K.

corrected version of a paper originally published in

East Africa Natural History Society Bulletin, 25 (3): 38-43

September 1995

{NB: the page numbers in this version do

not follow those of the published text}

current address:
[email protected]
EANHS Bulletin 25 (3), September 1995 1
______________________________________________________________________________

SNAKES ON THE USANGU PLAINS:

AN INTRODUCTION TO SANGU ETHNOHERPETOLOGY

This paper is based on notes I made while major habitats with specific vegetation
conducting anthropological fieldwork in the associations, ranging from rivers and
Usangu Plains of south-west Tanzania in permanent wetlands, through seasonally
1980-82. I did not make a systematic study inundated flood-plains and grasslands, to
of Sangu ethnoherpetology and was ill- thickets, thorn-bush and woodlands
equipped at that time to identify the snakes dominated by different tree species. The
which I saw and those which were described history of the human population in Usangu
to me (the same applies to other reptiles, and patterns of land use in the Plains have
which are not discussed further here). been described by me (1984) and, with
However, while recently re-reading through particular emphasis upon immigrant groups
my field notes and manuscript dictionary of and pastoral resources, by Charnley (1994).
ishisango, the Sangu language, I found that
there was sufficient information to identify Snakes on the Usangu Plains
some of the named varieties which I was As far as I am aware, no survey of Usangu’s
told about. I have therefore decided to herpetological fauna, and of snakes in
publish this information, together with some particular, has ever been undertaken, though
additional observations and comments, in I assume that some collections have been
the hope that it will provide a useful starting made there. Given the wide range of
point for future researchers. habitats available in Usangu and our general
knowledge of snakes and their distribution
The Usangu Plains in this region, it can be assumed that the
The Usangu Plains lie in the eastern Rift Plains support a wide variety of snake
Valley just south of Ruaha National Park, at species.
an average elevation of 1,000 metres above Snakes are certainly everywhere in
sea-level. They take the form of a shallow evidence, as a number of travellers to
alluvial basin, covering an estimated total Usangu have been quick to notice. Marius
area of 15,560 km2. To the south and west Fortie, an adventurer who crossed Usangu
they are hemmed in by the Southern in September 1934, begins his account of
Highlands and the mountain ranges which this journey by describing how he shot a
rise up from the northern shores of Lake five-foot Puff Adder in two. This
Malawi. The streams and rivers which flow unfortunate serpent had made the mistake of
down from these mountains join in Usangu visiting Fortie while he was in the middle of
to form the Great Ruaha river which a taking a bath in a room in the derelict
meanders out of the Plains to the north-east. compound of Brandt Mission. Later he was
There is one wet season, from December to able to brush this incident aside: “On
April. This is invariably accompanied by entering Usango I saw the trails so
substantial flooding, and the floodwaters crisscrossed with the tracks of the cobras
often remain until the dry season is well and vipers which thrive in those hot sandy
underway. For half of the year, however, plains, that when the puff adder interrupted
from June onwards, most of Usangu takes my bath at Brandt I was not much startled”
on a semi-arid aspect. (1938, p.279). Further on in his account he
The vegetation of Usangu has been provides us with a more comforting
described in outline by Procter (1968) and perspective: “Although snakes are
his classification supported by Charnley numerous, casualties from snake bites are
(1994), who also provides a detailed rare because the puff adder, the most
description and analysis of recent patterns of common venomous reptile, is so sluggish
ecological change on the Usangu Plains. that it seldom strikes unless actually stepped
According to Procter, the Plains host eleven upon” (1938, p.293).
EANHS Bulletin 25 (3), September 1995 2
______________________________________________________________________________

My own experience in Usangu was selling medicine to guard against snake bite
similar, though I did not see any Puff on the basis of this performance.
Adders, which are well camouflaged and Unfortunately, this medicine cannot have
whose numbers have probably been been very effective, because shortly
considerably reduced since the 1930s in the afterwards he was bitten by one of his own
increasingly populous areas of the southern snakes and had to be rushed to hospital in
Plains where people are apt to kill any Mbeya where, it was said, the affected limb
which they come across. However, when was amputated. Sangu observers ascribed
walking the three-hour journey between his misfortune to witchcraft rather than to
Igurusi and Utengule, I invariably saw at the ineffectiveness of his medicine. I
least one snake crossing the dirt track. On assumed that he had failed to milk the snake
forays into the bush around Utengule and of all of its venom.
Luhanga I always found it wise to let one of I noted more instances of snakes being
the villagers lead the way, given their killed. Although snakes are still a common
greater skill in spotting snakes on or by the sight in Usangu it likely, as suggested above
path. In Utengule itself, where I lived, I was in the case of Puff Adders, that their
taught to give a wide berth to any long and populations are declining in inverse
thin object I might see on the path at night, proportions to the increase in human
and got so used to doing this that for many population and land use.
months after returning to England I would
sometimes surprise friends by pulling up Sangu Ethnoherpetology
with a start in front of twigs on the The Sangu (avasango), the indigenous
pavement. inhabitants of the Plains, recognise a
On one occasion, now humorous to number of different kinds of snake (injoxa).
relate, I had to flee from a latrine when a We should not expect to find a one-to-one
large dark snake appeared through a crack correspondence in every case between their
in the earthen wall. Another incident names for these and their scientific
occurred in a friend’s house. Some time designations. This is a common feature of
earlier a large snake had been killed in the folk taxonomies, including the historical and
house while in the process of raiding a nest everyday classificatory schemes of English
of hen’s eggs. My friend declared that (e.g. Berlin, 1992). It is also a feature of
sooner or later the snake’s spouse would other components of Sangu ethnozoology.
also appear. Sure enough, one night as we In some cases one Sangu term covers more
were eating dinner, a large snake slithered than one zoological species and may even
under the front door, scattering all of those refer to species belonging to more than
present, including the mother of the genus. In other cases the Sangu give
household who was sitting on the cement different names to different subspecies or
floor with her children. This snake also colour phases and variants of the same
headed for the hen’s nest, but was species, including animals of different age
dispatched by my friend and another guest and sex (where sexual dimorphism is
with a couple of long sticks. particularly marked). At the same time it
I recorded a few instances of snake bite, must be said that many Sangu names do
but none in which death had resulted. One correspond to single zoological species,
interesting case of serious snake bite especially in the case of large, common
occurred following the appearance of a and/or better-known animals. Small and
Sukuma (on other accounts Fipa) snake- less familiar species, as well as those which
charmer in Utengule on 29 March 1981. He are not easily distinguished by zoologists
drew a large crowd at one of the drinking- working in the field, are more frequently
clubs in the village when he stood in the grouped together. This no doubt also
middle of the open ground and let two applies to the Sangu classification and
sizeable snakes (described in my notes as naming of snakes, as the following
‘cobras’) crawl and intertwine over his examination of indigenous taxonomy
body. He was subsequently reported to be suggests.
EANHS Bulletin 25 (3), September 1995 3
______________________________________________________________________________

All of the Sangu names for snakes borrowed in the past (it may even have been
which I recorded are presented below, borrowed by the Nyakyusa from the Sangu).
together with notes on their identification
and other relevant observations. Unless imulalu (9/10). I would probably not have
otherwise stated the zoological observations recorded this name if it had not been for the
are taken from Easterbrook (n.d.), Hedges fact that I killed one in Utengule on 11 May
(1983), MacKay and MacKay (1985), and 1981. I noticed it on the floor of my house
Branch (1988). at night just as I was preparing to go to bed.
It lay curled up in front of the only door to
inyatu (noun class 9/10), ilipingang’ombe the house, blocking my way out. Lacking a
(5/6). Both of these names appear to refer stick or other suitable implement, I chopped
primarily to the African Rock Python, it in two with a machete. Local opinion
Python sebae. inyatu is cognate with the held that it had probably come into the
term for python in many other Eastern village to escape the seasonal drying up of
Bantu languages, including Swahili chatu. the surrounding Plains and into my house to
It was described to me as non-poisonous escape the increasing cold. My cursory
and I was told that it can sometimes be notes indicate that it had a purple and black
found in water. Although African Rock back and white belly. It was fairly long but
Pythons can inflict painful wounds with only the thickness of a thumb. I was later
their teeth, they usually kill their victims by told that it was very poisonous. However, I
coiling around and asphyxiating them, would now tentatively identify it as a
before attempting to swallow them whole. Purple-glossed Snake, Amblyodipsas sp.,
They rarely attack humans. It is, however a and therefore relatively harmless.
good swimmer, and is known to prey upon
crocodiles, fish and other aquatic creatures. injoxa nyalulemba (9/10). This name
ilipingang’ombe was also given as a name means ‘green snake’ and is probably a
for pythons and likewise was said to be generic term for the Green or Bush Snakes,
sometimes found in water. This is a Philothamnus spp. It was described as a
compound term which literally means ‘the green snake of pencil-thickness which is not
one which stops cattle’ in Sangu, indicating very poisonous and usually flees from
that they are also known to attack livestock. people. I saw one on 28 June 1981 climbing
Its designation in noun class 5/6 implies an a tree near a house at the edge of Utengule.
animal which is large, bad, or both of these, It was very thin and very bright green in
and it may be that ilipingang’ombe is colour. These descriptions fit the Green or
reserved for the larger specimens and inyatu Bush Snakes very well. Different species
for smaller African Rock Pythons. are quite difficult to distinguish and the only
ilipingang’ombe is also the Sangu name for certain way is by scale counts. It would
a rainbow. I was unable to elicit an therefore not be very surprising if the
explanation for this in the field, but it is Sangu, like other East African peoples,
presumably connected with the Nyakyusa grouped them under the same name.
belief that pythons (or at least one mythical
variety of them) cause rainbows when they injoxa nyaluhanjala (9/10). This name
look up to the sky. The same beast is also means ‘firewood snake’, referring to the
believed to live near rivers, and some resemblance between the snake when
Nyakyusa therefore run away from rivers motionless and a relatively thick and long
when they see rainbows (Walsh, in prep.). branch of firewood. The only other
The Nyakyusa are near neighbours to the information which I recorded on this snake
Sangu and many of them are recent is that it is very poisonous and capable of
immigrants to the Plains. The explicit ‘standing up’. This last observation implies
reference of the Sangu name, however, a cobra of some description, following their
suggests that this is or was also their own habit of lifting the forebody and spreading a
belief, though it may have been inherited or hood when threatened. I would tentatively
identify this with the Egyptian Cobra, Naja
EANHS Bulletin 25 (3), September 1995 4
______________________________________________________________________________

haje, which is the right shape and size and Cobra, though of normal dimensions, while
has a variable coloration which might be living in Usangu.
taken to resemble firewood in different
stages of drying and possibly with lichens ilipili (5/6), imili (9/10). This is the Puff
on the bark. The Egyptian Cobra is also Adder, Bitis arietans. The usual form of the
reported to sham death and this could Sangu name is ilipili, which also has
account for the motionlessness which cognates in many closely related Bantu
increases its resemblance to firewood. languages, including the Swahili pili or piri.
It was described to me succinctly as short,
imfwila (9/10), ilifwila (5/6), fat and very poisonous, and not surprisingly
ilipingankwale (5/6). The description of is therefore put in the augmentative noun
the snake which goes by one or other of class. imili is a form of the same name in
these three names (they are equated by the class 9/10, which is the usual class
Sangu themselves) leaves little doubt that designation for snakes. I only heard this
their usual referent is the Black-necked term used once in Utengule. This was when
Spitting Cobra, Naja nigricollis. imfwila a man was reported to be in some pain after
has cognates in other Eastern Bantu having been bitten by a snake while
languages, including Swahili (fira, also working in his rice plot (this was in the
swila), and in which it usually describes the second week of November, 1981). I
same and/or closely related species. I was listened to an argument about the kind of
told that it is a black snake which spits snake which had bitten him: according to
venom into the eyes and also bites, though one side it was an ‘ilipilu’ (presumably
its attacks are not lethal because local misheard for ilipili), and according to the
medicine is available to counteract the other it was imili. This suggests that ilipili
poison (sufficient doses of which are lethal and imili are considered to be different
to humans according to the literature). kinds of snake, though I have no more
ilifwila is the augmentative and emphatic information on this point.
form of the same name, and may be used to
refer to larger specimens. ilipingankwale imoma (9/10), imuhando (9/10, or possibly
means ‘the one which stops the francolin’ umuhando, 3/4). The first of these terms
and suggests that these birds, Francolinus has cognates in other Eastern Bantu
spp., are among its items of prey (Procter, languages, including Swahili (moma), which
1968, records the presence of three species refer to the Gaboon Viper, Bitis gabonica,
of francolin in Usangu, the Red-necked and related species. imoma was described
Spurfowl, F.afer, the Coqui Francolin, to me as being similar to ilipili, the Puff
F.coqui, and Hildebrandt’s Francolin, Adder (B.arietans), being more or less the
F.hildebrandti). I was told that it is large, same colour, but longer and more
has a red throat, and sometimes climbs trees. dangerous. This could conceivably apply to
These are all characteristics of the Black- the Gaboon Viper, although its colouring
necked Spitting Cobra, which in East Africa and pattern are quite distinct to close
usually has a patch of salmon pink observers. In East Africa it is usually found
(sometimes deep red) scales below the neck in forests or on forest margins, and while
which are clearly visible when it rears up to Usangu lacks this kind of habitat, it is
spit. It is possible that ilipingankwale possible that isolated specimens make their
usually refers to larger specimens which way down to the Plains from the forested
have this patch of colour. Otherwise one of escarpments in the south and west. An
my informants specifically identified identification with the Gaboon Viper would
ilipingankwale with ilifwila. The one-time also suit the description of imuhando, which
Southern Highlands Provincial I was told was the most poisonous snake in
Commissioner, J. E. S. Griffiths, recorded Usangu, causing instant death to unlucky
seeing a large spitting cobra on the road to victims. Fortunately for snake-haters it is
Rujewa (diary entry for 20 August 1958), said to be very rare. I was also told that it
and I saw at least one Black-necked Spitting can make a sound like a cock crowing. It
EANHS Bulletin 25 (3), September 1995 5
______________________________________________________________________________

was not linked to the imoma by anyone I poison out. The ‘Black Stone’ is really a
spoke to, though I did not ask if it was. piece of carbonised bone, and these are
imuhando could also conceivably describe supplied to Roman Catholic missionaries by
the Gaboon Viper, which is capable of a mission procure in Belgium (Shorter,
producing a loud and deep hissing noise 1979). As far as I knew no one in Utengule
when disturbed. However, on one occasion had one - the White Fathers were infrequent
I also recorded the compound name imfwila visitors - and people were a little
muhando, suggesting a relationship, real or disappointed that I did not have one myself.
perceived, with the Black-necked Spitting There are presumably many other local
Cobra. Further research is therefore treatments which I did not record.
required to establish the precise reference of A fuller account of these and other
these terms (imoma, imuhando) and whether aspects of Sangu ethnoherpetology must
or not Gaboon Vipers occur in Usangu as I await future research and publication. I
have suggested they may. hope that the notes I have provided in this
paper will form a useful starting-point for
I did not record any other Sangu names for this work.
snakes, except for the mythical serpents
umutila and inyamfwila, which I discuss
elsewhere (Walsh, in prep.). Almost all of Acknowledgements
the terms I did record refer to common My research in Usangu in 1980-82 was
and/or notably poisonous species. This may funded by the then Social Science Research
have been a function of my failure to probe Council of Great Britain, with additional
more deeply, or alternatively may reflect the support from the Smuts Fund and Wolfson
fact that most Sangu only know the names College in the University of Cambridge. I
of the species which they are most likely to am very grateful to my hosts in Utengule
meet and those which they feel are most and all those who provided me with
likely to pose a threat to their lives. A information on snakes and related subjects,
Kukwe (Nyakyusa) woman told me that in particular Eliuter Shinangonele, Betitha
there are only four kinds of snake (i.e. four Mwakalinga, and Ngwila Simuhongole. I
named varieties that she knew of) in would also like to thank Ray Abrahams and
Unyakyusa, in the hills and plains to the Alison Redmayne for their encouragement,
south of Mount Rungwe, south-east of advice and other help over the years since I
Usangu. If, however, a much longer Sangu began work on the Sangu, and Susan
list exists, then it should not be too difficult Charnley for sending me a copy of her
to elicit it. The more difficult task will be to unpublished Ph.D. dissertation. None of
establish the zoological referents of Sangu them, of course, is responsible for the
names. evident shortcomings of this paper.

Conclusion
There are other dimensions of Sangu References
ethnoherpetology which I have not Berlin, B. 1992. Ethnobiological
described above. Among these are Classification: Principles of
indigenous treatments for snake bite. One Categorization of Plants and Animals in
treatment I was told about used the roots of Traditional Societies. Princeton, New
a short grass called ilidilimbuli (said to be Jersey: Princeton University Press.
known as nsindu in Nyakyusa). These are
chewed and rubbed onto the bite, and can Branch, B. 1988. Field Guide to the Snakes
also be carried around as a protective charm and Other Reptiles of Southern Africa.
against snake bite. Some people drink their London: New Holland.
own urine as a cure for snake bite. Others
extolled the virtues of the ‘Black Stone’, Charnley, S. 1994. Cattle, Commons, and
which is applied to the site of (bleeding) Culture: The Political Ecology of
punctures and stings and reputed to draw the Environmental Change on a Tanzanian
EANHS Bulletin 25 (3), September 1995 6
______________________________________________________________________________

Rangeland, unpublished Ph.D. Ministry of Agriculture, Livestock and

dissertation, Stanford University. Natural Resources, Forodhani, P.O.Box
2283, Zanzibar, Tanzania).
Easterbrook, M. S. n.d. A Book of East
African Snakes. Nairobi: Westland
Sundries.

Fortie, M. 1938. Black and Beautiful: A Life

in Safari Land. London: Robert Hale.

Griffiths, J. E. S. 1958-59. Personal diary, 4

August 1958 to 11 March 1959 (written
during tenure as Southern Highlands
Provincial Commissioner), Mss.Afr.r. 190,
Rhodes House Library, Oxford.

Hedges, N. G. 1983. Reptiles and

Amphibians of East Africa. Nairobi:
Kenya Literature Bureau.

MacKay, A. and MacKay, J. 1985.

Poisonous Snakes of Eastern Africa and
the Treatment of their Bites. Nairobi:
privately published.

Procter, J. 1968. ‘The Birds of the Usangu

Plains: A First Checklist, with Notes on
Ecology and Distribution’, Tanzania Notes
and Records, 69, 1-14.

Shorter, A. 1979. Priest in the Village:

Experiences of an African Community.
London: Geoffrey Chapman.

Walsh, M. T. 1984. The Misinterpretation of

Chiefly Power in Usangu, South-west
Tanzania, unpublished Ph.D. dissertation,
University of Cambridge.

Walsh, M. T. 1985. Shisango Dictionary,

unpublished manuscript.

Walsh, M. T. (in prep.). ‘Snakes Between

the Lakes: The Mythical Serpents and
Magical Snakes of South-west Tanzania’,
unpublished paper.

___________________________________
Martin T. Walsh, Natural Resources
Institute, Chatham, and School of African
and Asian Studies, University of Sussex,
U.K. (current [1995] address: Zanzibar
Cash Crops Farming Systems Project,