The Collection And Elucidation Of Vernacular Names: Observations On Shambaa Ethnobotany

THE COLLECTION AND ELUCIDATION OF VERNACULAR NAMES: OBSERVATIONS ON SHAMBAA ETHNOBOTANY

Martin T. Walsh 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, 23 (2): 21-25 June 1993

{NB: the page numbers in this version do not follow those of the published text}

current address: [email protected]

EANHS Bulletin 23 (2), June 1993 ______________________________________________________________________________

1

THE COLLECTION AND ELUCIDATION OF VERNACULAR NAMES: OBSERVATIONS ON SHAMBAA ETHNOBOTANY The following notes provide additional comment to the list of Shambaa tree and shrub names and their uses published by Jon Lovett (1992). The Shambaa (or Shambala, also called Sambaa by some of their neighbours) are a Bantu speaking people who live in the Usambara Mountains in North-east Tanzania. We are fortunate in already possessing more information on Shambaa ethnobotany than for many East African people: in addition to work by Fleuret (1979a; 1979b; 1980) on different aspects of Shambaa plant classification and use, there also exists a cyclostyled dictionary of plant names in Shambaa and two closely related languages, Bondei and Zigua/Nguu (Sangai, 1963). This dictionary gives well over 1,000 names (including cognates) in the three languages, together with their botanical equivalents. Despite its comparatively modest size, the list reproduced by Lovett is an important addition to the literature on Shambaa ethnobotany, and in the following notes I hope to indicate why this is so, as well as make further observations about the Shambaa list and East African ethnobotany in general. The list comprises 53 names provided by Mr Mgaa Sabuni, a resident of Mgwashi village and forest guard at the University of Dar es Salaam's forest reserve at Mazumbai in the West Usambara Mountains, Tanzania. Lovett remarks that subsequent users of Sabuni’s list (using different local guides) have commented that it is not wholly accurate and notes that there is obviously considerable local variation in the application of names to plants that are not commonly used. However, comparison of the Shambaa terms in Sabuni’s list with Sangai's (1963) dictionary reveals a high degree of correspondence (and presumably accuracy). 46 (or 87%) of the 53 names in Sabuni’s list can also be found in Sangai’s

dictionary: which does not mean, of course, that the remainder are wrong. There are, it might be added, some differences in the transcription of Shambaa, and where necessary I have taken account of these. Some of the terms in the Sabuni list are evidently mistranscribed, e.g. "dwaiu" for Sangai's mdwayu, "cherooti" for msheruti, "mbakambaka" for mbwakambwaka, and "mshiwhi" for mshihwi. It is also evident that a few of the terms given by Sabuni have been translated in part or whole from Shambaa into Swahili: thus "mpiga magasa" for the proper Shambaa mtoamaghasa (literally "hand-clapper"). Sabuni's "mti wa paa" (Swahili "antelope tree", identified as Dovyalis abyssinica), not present in Sangai, also seems to fall into this category; likewise his "mweti misitu" or "mweeti of the forest" (given as Rananea melanophloeos), where the unqualified Shambaa name mweeti (also mweleti) refers to Rauvolfia caffra according to Sangai. At the same time, 38 (or 83%) of the 46 terms common to both lists are given substantially similar (or overlapping) botanical identifications, i.e. corresponding at the generic level. Again, this does not mean that the remainder are incorrect: this could only be established by further investigation. To the extent that they corroborate one another, however, this does increase our confidence in the accuracy of both lists. In one important respect, though, the Sabuni list is much less complete than the Sangai dictionary. Where only three (6%) of the names in the former list are given double identifications, more than half of their equivalents in Sangai (24 of the 46, or 52%) are given two or more botanical equivalents, and in 63% of cases (15 out of 24) these multiple identifications are not confined to a single genus. Thus msheruti, mentioned above, is identified as both Philippia benguelensis and Pterolobium stellatum. In this respect the Sabuni list is

EANHS Bulletin 23 (2), June 1993 ______________________________________________________________________________

not inaccurate: it is simply incomplete, though this may help to explain why it has appeared inaccurate to other observers. Lovett suggests that local variation in the application of vernacular names to less commonly used plants may explain the apparent inaccuracy of Sabuni’s list. This is, however, only one a number of possible explanations for the discrepancies in naming which occur. These different explanations can be grouped under four main headings, as follows: 1. Informant error. There are different kinds of errors to be considered. One of these is plain linguistic error, the misquoting of a name by an informant or its mistranscription by the recorder. The translation of vernacular terms into Swahili (often for the benefit of a particular listener, though sometimes as the unintended consequence of a more general process of linguistic change) is a special and fairly common case of this. The misidentification of plants and misapplication of terms also undoubtedly occurs. Although it is not difficult to find young informants from a rural background with a ready command of more than 100 vernacular plant names, knowledge of local vegetation and its uses clearly varies considerably between individuals according to a number of factors (including age and sex), with herbalists and similar specialists often having the widest knowledge. It is also evident that this knowledge will vary from one locality to another depending upon differences in the local vegetation: informants in an unfamiliar environment will be much more prone to making errors than on their home ground. At the same time, and in the absence of any standard other than common agreement, there may also be borderline cases between erroneous identification (beginning with individual errors) and its possible consolidation into unorthodox opinion (which may develop into local orthodoxy). This process may occur when a number of people move from one area to another, and consistently make the same mistakes of identification for the same reasons (e.g. because of the similarity

2

of an unfamiliar plant with one they know from their original home). In this way what begins as collective error can result in linguistic variation. 2. Linguistic Variation. Where the names for plants vary in a regular way from one location to another we can ascribe this to dialect differences. Dialects often form continua of mutual intelligibility and it is not always easy to draw hard and fast boundaries between them. Whereas most languages in East Africa have been classified and described, at least in outline, the description of the dialects which comprise them is not as well advanced. Shambaa (or Shambala: the presence or absence of /l/ varies according to dialect) is no exception. Besha (1989) notes that the Shambaa dialects have not been studied at all, and for the purposes of her own analysis of the language provisionally distinguishes between three main varieties, centring on Mlalo, Lushoto and Korogwe respectively. Given the existence of such differences it is important for researchers or their assistants to transcribe vernacular names as accurately as possible, and for the former to be aware of existing orthographic conventions (which may conceal linguistic variation, as widespread use of standard Kikuyu orthography does in central Kenya). It is perhaps even more important for them to note relevant details about the background of individual informants and their speech. At the very least informants should be identified by their residence or place of origin, as Lovett has done for Mgaa Sabuni. Unfortunately, Sangai's dictionary tells us nothing about his sources or the dialects concerned, though the range of cognates in some cases might lead us to suspect that terms from a number of different dialects are included. Thus Sabuni's "mula" (Parinari excelsa) is given the following Shambaa variants by Sangai: muula, muwa, mbula and hula. Mula, with a short or single vowel, is identified by Sangai as the Zigua version of the name, though it is difficult to judge how reliable either his or Sabuni’s transcription is in this case. Another factor which can complicate

EANHS Bulletin 23 (2), June 1993 ______________________________________________________________________________

this situation is bi- or multilingualism in different dialects and/or languages on the part of informants. This may also give rise to apparent discrepancies which are really normative aspects of linguistic variation. Again, only by carefully recording vernacular names and noting down the relevant contextual information is it possible to determine whether or not this is the case. The failure to take account of linguistic variation is extremely common in the collection of ethnobotanical data and affects even the best-known languages, including Swahili (see Walsh, 1992). 3. Multiple Designations. Everyday processes of linguistic change, incipient or otherwise, may also lead to a situation in which the same plant (or group of plants) has one or more alternative names in the same dialect. These often arise as nicknames used to describe particular features of the plant or its characteristic uses. The best example of this in our sample is the number of different Shambaa names for Isoberlinia scheffleri given by Sangai: mamba, mbaika (the "splitting tree"), mfimbo ("stick tree"), mshembeshembe (glossed as "grain by grain"), msuke, and mtoa-maghasa ("clapping hands"). The glosses given by Sangai make it apparent that some of these are nicknames, a number of them referring to the explosive noise made by the ripe pods of the tree when they burst. One of these names, mbaika, is clearly cognate with the Swahili name mbarika, which is said to be applied not just to I.scheffleri (the Shoesole Tree), but to any tree which makes this noise, including, the Lucky Bean or Mahogany Bean Tree, Afzelia quanzensis, and the Castor Oil Plant, Ricinus communis (Greenway, 1940: 51). The coining of nicknames is clearly one of the forms of linguistic innovation which can result in multiple designations. A word of warning though: the interpretation of such names usually requires a reasonable understanding of the language or dialect concerned. The standard dictionary of Digo ethnobotany, for example, is marred by the careless ascription of meanings (derived from a

3

superficial comparison of Digo with Swahili) to many of the plant names (Glover et al. 1969). Another, though less common, source of alternative designations occurs when the same species has clearly distinguishable forms: thus young and mature Doum Palms (Hyphaene compressa) are given different names in many Bantu languages. Other linguistic processes which may lead to the existence of multiple designations have already been referred to above: these include bi- or multilingualism and the adoption of names from other languages, whether as a result of direct borrowing or the social and linguistic assimilation of the speakers of these languages. 4. Taxonomic Under-differentiation. This is a frequently remarked feature of vernacular taxonomies, especially when the oral transmission of botanical knowledge is the rule. The same name may be applied to more than one plant for different reasons. The most common reason is because of the perceived morphological similarity – and in some cases failure to recognise the difference – between species which are distinguished botanically (though not always easily so). Thus the Shambaa mkuyu, identified solely as Ficus sur in Sabuni’s list, is applied to at least three other members of the same genus according to Sangai (and likewise the cognate term in other East African Bantu languages). Similarly, mkumba describes at least three Macaranga species (M.capensis, M.conglomerata and M.kilimandscharica according to Sangai), while mshai is applied to three different species of Albizia (A.adianthifolia, A.gummifera and A.schimperiana). Plants grouped together because of their morphological similarity need not, of course, be members of the same genus or even family, according to the morphological criteria which are applied in any particular case. Morphologically dissimilar plants may also be described together under the same name because they share similar cultural uses and functions: a principle of classification which is very different from the Linnaean. Examples of

EANHS Bulletin 23 (2), June 1993 ______________________________________________________________________________

this principle being applied are difficult to discern in Sangai’s dictionary because it does not give detailed information on plant uses, though Sabuni’s list implies that Dicranolepsis usambarica and Peddiea fischeri are called by the same Shambaa name (mkisigizi) because their bark is used in making similar kinds of ropes. It is likely, however, that many more examples of the application of such functional criteria can be found in Shambaa ethnobotany, broadly similar to those described for the Mbeere of central Kenya by Riley and Brokensha (1988). There are therefore a range of possible explanations for the discrepancies between different lists such as those provided by Sabuni and other informants. In order to be certain which explanation applies in a particular case, a lot more ethnobotanical data has to be collected and published. This is as true among the Shambaa, who appear at first sight to have been well served in this respect, as it is among other peoples in East Africa. A single list from a single informant, however well-informed, is clearly insufficient as a database, and only becomes valuable when it can be compared with and evaluated against other lists, collected from other informants in different places and with the relevant details added. The existence of a considerable body of information, such as that provided by Sangai, should not discourage further research, but encourage the collection and publication of new data so that it can be checked and refined. As in the Shambaa case, this material can often be improved even further by more careful attention to the sociolinguistic context (which are the dialects involved? where and from whom were different names recorded?) and the elicitation of much more information on the local uses of plants, as well as the details of how they are recognised and why they are classified as they are. Readers should be encouraged, like Jon Lovett, to use the EANHS Bulletin to publish material of this kind.

4

Note In the absence of adequate reference materials I have made no attempt to check or update the botanical names given by Sangai (1963).

References Besha, R. M. 1989. A Study of Tense and Aspect in Shambala. Berlin: Dietrich Reimer Verlag. Fleuret, A. K. 1979a. The Role of Wild Foliage Plants in the Diet: A Case Study from Lushoto, Tanzania. Ecology of Food and Nutrition, 8: 87-93. Fleuret, A. K. 1979b. Methods for Evaluation of the Role of Fruits and Wild Greens in Shambaa Diet: A Case Study. Medical Anthropology, 3: 249269. Fleuret, A. K. 1980. Non-food Uses of Plants in Usambara. Economic Botany, 34: 320-333. Glover, P. E., F. C. Magogo and Ali Bandari Hamisi 1969. A DigoBotanical-Digo Glossary from the Shimba Hills, Kenya. Nairobi: Kenya National Parks. Greenway, P. J. 1940. A Swahili-BotanicalEnglish Dictionary of Plant Names. Dar es Salaam: Government Printer. Lovett, J. C. 1992. Some Local Names and Uses of Trees and Shrubs in the University Forest Reserve at Mazumbai in the West Usambara Mountains, Tanzania. EANHS Bulletin, 22 (2): 2427. Riley, B. W. and D. Brokensha 1988. The Mbeere in Kenya (Vol.II: Botanical Identities and Uses). Lanham: University Press of America. Sangai, G. R. W. 1963. Dictionary of Native Plant Names in the Bondei, Shambaa and Zigua Languages with their English and Botanical Equivalents. Nairobi: East African Herbarium (cyclostyled). Walsh, M. T. 1992. Swahili Ethnobotany and Carved Doors. Azania, 27 (in press).

EANHS Bulletin 23 (2), June 1993 ______________________________________________________________________________

__________________________________ Martin T. Walsh, P.O. Box 99187, Mombasa, and School of African and Asian Studies, University of Sussex, Falmer, Brighton.

5