The African Dimension of the Oxford English Dictionary

African languages have contributed and continue to contribute a great many vocabulary items to English, both directly and via the intermediary of other non-African languages. The OED Additions Series contains 63 words of African origin, most of which have come into the language during the last years. The second edition of the OED contains about 275 words of African origin, drawn from 30 languages. The treatment of these words is somewhat uneven, owing to two factors: the unavailability to the editors of relevant information at the time of compilation of OED1; and supremacist attitudes, which caused entries for words of this kind to be shorter and less detailed, and affected their definition and description. In the third edition of the OED, words of all kinds should receive the same degree of attention, which implies that data collection from the reading of primary sources should include all varieties of World English; description and definition should be undertaken from a neutral standpoint; and etymological research and documentation must be as full as is practicable. All items of African origin will be sent to linguistic specialists (the bulk of this work has been done in the DSAE for items of South African origin). The checking of other etymologies will bring to light items whose African origin was not previously indicated (for example, borrowings from Brazilian Portuguese). In a sample from the letter M, there has been a gain in the numbers of African etymolOgies and the accuracy of treatment. Loanwords of all origins will be given as much attention as our resources allow, but there may be more ground to make up in the African sector of the lexicon.


1.
African sources in DEDI The first edition of the OED excluded the word African.This is what the editor, James Murray, wrote in his original Preface to Volume 1 (a Preface that was not reprinted in the completed OEDl): The word African was one of the earliest instances in which the question of admission or exclusion arose with regard to an important adjective derived from a geographical proper name.After much careful consideration, and consultation with advisers, it was decided (perhaps by a too rigid application of first principles) to omit the word, as having really no more claims to inclusion than Algerian, Austrian, or Bulgarian.But, when American was reached, some months afterwards, it was seen that Americanize and Americanism must of necessity be included, and that these (with "the Americanizing of our institutions") could not be explained without treating American, and explaining its restricted application to the United States.American was accordingly admitted.Then the question arose, whether the exclusion of African was consistent with the inclusion of American: but the question came too late; African had been actually omit-ted, on its own merits.And the inconsistency is only on the surface; American is included, not on its own account, but to help to the better explanation of derived words.(Murray 1884: ix.) We would not now accept this principle, and of course the second edition of the OED (Simpson and Weiner 1989) contains Algerian, Austrian, and Bulgarian, as well as African, which were all added in the Supplement (Burchfield 1972).
This rather astonishing omission is in a way symbolic of a tendency to underestimate the role of Africa in the English language at that time, of which I shall presently say more.
The contribution of the African countries to the English language continues to be made.In the OED Additions Series Oxford University Press publishes work in progress on words and senses since the completion of the Second Edition in 1989.So far, Voliunes 1 and 2 (Simpson and Weiner 1993) have appeared.Volume 3 is in production, and further material is in preparation.In this body of edited material (which contain in all about 12,000 vocabulary items) there are 63 words of African origin.Most of them have entered English during the last 25 years.They are drawn from at least 30 languages.Not many of these words are part of the central vocabulary of World English.Some are the names of ethnic groups and languages and many are terms as yet restricted to a particular area of Africa.But as the use of English in some of these areas increases, we can expect such loanwords to assume a greater importance.
It is estimated that OED2 contains 275 words of African origin, drawn from 80 languages.The treatment of these words in the dictionary is somewhat uneven, owing to two main factors that influenced the compilation of the first edition (DEDI) both of which are familiar in lexicography; namely, the unavailability of information and a supremacist attitude.

Unavailability of relevant information
Relevant information may have been unavailable to the editors of DEDI because they did not have access to scholars and experts and their books.
Recalling that the first edition was compiled between 1879 and 1928, we should not be surprised if (especially while the first half of the alphabet was being compiled, up till about 1905) the editors were unable to obtain the information they needed and ought to have deployed in their entries.This had a particular effect on three parts of the entry; the documentation, the definition, and the etymology.

Effect on the documentation
Few quotations were available, giving the impression that the word was of infrequent occurrence and marginal to the language.For example, the word impi (a Zulu regiment) had, in the first edition, only three quotations: 1879 Close by in DEDI is the plant name imphee (now usually spelt imfe).The first edition had only two examples, one of 1880 from an Australian source, and one of 1893 from a book on Indian produce.By contrast, DSAE has no fewer than ten pre-l900 examples, the earliest dating from 1828.

Effect on the definition
Largely because of the scanty documentation, the definition was often very brief, sometimes combining into one section a range of meanings which, in an indigenous English word, would have merited a series of distinct sense sections.For example, in DEDI the word induna is defined simply as "An officer under the king or chief of the Zulus, Matabele, and other South African tribes".
The earliest example is ante 1875.When we turn to the DSAE, we find that there are three senses which date from before 1900: "1.a.A headman, councillor, or officer under a chief' (1835 onwards), "b.transf A foreman, head servant, mine overseer, or policeman" (1857 onwards), and "2.fig.One in authority" (1897).

Effect on the etymology
In the absence of accurate linguistic information, etymologies often fall well below the normal standards in both completeness and accuracy.For example, the word cola (the West African nut-bearing tree) has the etymology "Kola, Kolla, Goora, in Negro languages of W. Africa".This is evidently unacceptable.
The languages are not named; the region is far too large for the origin to be localized; and one has little confidence in the forms and their phonetic accuracy."Negro languages" is also an objectionable expression, an example of the supremacism discussed below.
Another example is the word macute (a form of currency).In DEDI nothing more was said than "origin unknown".Some addenda and corrigenda, published at the end of the completed DEDI, included a note on the origin of this word from "the Reverend W. Holman Bently, writing from the Congo Free State", which, though more informative, is still lacking in precision and reliability.This note was inserted into the etymology in OED2, following the inten-tion of the compilers of the addenda, but in the draft OE[)3 text it has been moved to the quotation section and replaced by information taken from a much more recent scholarly study of Portuguese Africanisms.

A supremacist attitude
Supremacism is undesirable in a dictionary; but, to be fair to OEDt•s editors, such an attitude must have been largely unconscious, and shared with the vast majority of their British contemporaries.Moreover, I would see it rather as the (slightly) more excusable implicit assumption of the cultural superiority of inherited western European traditions, than as a crude or doctrinaire racism.However, it undoubtedly contributed to the inadequacies that we have noted above.For example, the shortage of documentation may.have been partly due to a reading programme which devoted less effort (proportionately) to sources concerned with Africa than to those concerned with Britain, Europe, or North America.The brevity of definitions and absence of finer sense discriminations may have been due to a feeling that the vocabulary of the outposts of the Empire was not as important as that of the well-established English-speaking areas.And the skimpiness of the etymologies could have been due to an unconscious prejudice in favour of Indo-European and other languages with a long scholarly tradition.Such attitudes may be detected in an entry like that for Eboe (the ethnonym, now usually spelt Ibo) which reads as follows: A name applied in the W. Indies and U.s. to the Negroes from Benin.Hence attrib., eboe-tree (Dipteryx eboensis), a tree of Central America, yielding eboe oil, a name also given to the oil of a different tree; see quot.
Normal entry structure is suspended in the presentation of this information.The cursoriness of the wording seems to convey a distinctly dismissive attitude towards a whole nexus of interesting questions relating to the cultures of both the West Indies and West Africa. (The 1972 volume of the OED Supplement, incidentally, merely added "The same term as IBO in West Africa, but used with wider application in the W. Indies".) Another example is the entry for cam-wood: [According to some, ad.native African name kambi.]The hard red wood of 8aphia nitida (N.O.Leguminosae), imported from West Africa.
The etymology is hardly a model of exact scholarship.Notice also the implicit Britocentrism of the last four words.

African words in OED3
The third edition of the OED, as is now well known, is in progress and due for completion in 2005.Whereas the second edition represented the integration of OEDI and its four-volume Supplement, for the third edition all aspects of the dictionary are being revised.The new words which are at present being publishedseparately will all be included in this edition.In working on the third edition, we should, in theory, bestow the same degree of attention on words of all kinds, whatever variety of English they may belong to or have arisen in.This has implications for data collecting, defining, and etymological work.

Implications for data collecting
The collection of data, from the reading of primary sources, potentially includes all varieties of World English.We have a British and a North American Reading Programme, which take in not only British and American English, but also Caribbean, South Asian, West and East African, and East Asian varieties.But in fact the analysis of lexical data from a variety of English outside Britain is better done by local scholars than by British lexicographers.Once a dictionary covering one such variety is complete, it is relatively easy for us to scale down its very detailed treabnent to the dimensions of the OED, ensuring that the overall perspective remains intact.Hence, the Dictionary of South African English on Historical Principles is not only a landmark in the study of World Englishes but will also make an enormous difference to OED3's effectiveness in covering English of African origin.

Implications for defining
OED3's description and definition of vocabulary will be undertaken from a more neutral standpoint.It is, of course, impossible to write a dictionary without making a number of cultural assumptions.However, it is possible to try to accord every variety equal treabnent, within the limitations of the materials available, and it ought to be possible to avoid the grosser kinds of cultural superiority complex.

Implications for etymological work
Etymological research and documentation are to be as full as is practicable.We are sending the majority of etymologies that have any significant non-English input to be checked by people who are proficient in the relevant languages.In the case of very frequently cited languages, we obtain guidance on overall pol-icy and style, and on exceptionally difficult words, from experts, but have the bulk of the forms checked by graduates who are proficient in those languages.We shall therefore send most of the items of African origin to linguistic specialists for correction and amplification.(In this regard, again, it is very helpful to have the DSAE available, since most of the work on items of South African origin (at least 12%) has already been done by those editors and checked by local experts.)Furthermore, we confidently expect that this checking will bring to light items not recognized hitherto as of African origin.There are a number of borrowings from Brazilian Portuguese which, when traced through Portuguese etymological sources, turn out to be of African origin.For example, the word macumba (the Brazilian religiOUS system), whose etymology in OED2 reads Simply "Portuguese", is almost certainly derived from a language of the Zalre region.Recent scholarship makes it very likely that the word mahogany is from the Yoruba m'oganwo, which in Nigeria is applied to a different but not dissimilar tree.
A further stage of checking is in store for etymologies which have been extensively rewritten in-house and newly provided with foreign-language etyma.These are all flagged in the database and will be sent to experts after the main phase of revision.This process will probably bring yet more African borrowings to light.

3.
Results so far The revision of the OED began with the letter M, because it was from this point that our extensive quotation files had been systematically sorted into the categories needed for efficient revision.Revision begins with the documentation, definitions, and sense-structure of the scientific and general vocabulary.Etymological work on a given range of words is done after the work of the science revisers and general revisers has been keyboarded, so that all changes thrown up by new forms, new dating, and new ordering can be taken into account.Already in the part of the letter M that has been etymologically revised we have seen: (a) a gain in the number of African etymologies, and (b) an improvement in the accuracy of information given for such words.Examples of (a) are machila (hammock-like conveyance), formerly ascribed (via Portuguese) to Tamil, but now believed to originate in a Mo~ambican language; macumba, mentioned above as probably Zalrean; and many-root, a West Indian plant name previously considered to be a purely English compound, but now believed to come from a Twi word influenced by folk-etymology.An example of (b) is marimba (musical instrument) previously said to be from "Congo", but now ascribed to a specific Kimbundu word.There are also other words which look as if an African origin, not previously mentioned, is likely; for example malcimbo bark, which entered English from Spanish or Portuguese, looks quite likely to tum out to have an African etymon.It is not, of course, that we are giving preferential treatment to Reproduced by Sabinet Gateway under licence granted by the Publisher ( dated 2011) http://lexikos.journals.ac.za words of African or any other origin outside the historic English-speaking areas.Loanwords and local words of all origins will be given as much attention as our resources allow.But it may be the case that there are more deficiencies to be made good, and more cultural slants to be adjusted, in this sector of the lexicon than in many others.
Sleutelwoorde: AFRII<ATALE, ENGElS, OXFORD ENGLISH DICTIONARY, OED ADDI-TIONS SERIES, LEKSIKOGRAFIFSE DOKUMENTASIE, DEFINISIE, ETIMOLOGIE, LEEN-WOORDE, MEERDERWAARDIGE HOUDING, DERDE UITGAWE VAN DIE OED, LEESPRO-GRAM, DICTIONARY OF SOUTH AFRICAN ENGLISH ON HISTORICAL PRINCIPLES Daily Telegraph, 1885 Harper's Magazine, and 1888 Rider Haggard Maiwa's Revenge, none of them African publications.This entry was published in 1900.The same entry has now been treated in appropriate detail in the new Dictionary of South African English on Historical Principles (Silva 1996).If we look at the DSAE to see what the potential documentation at that date was, we find five examples up to 1900, all different from those in DEDI, and all taken from people writing in Africa itself, not in Britain or America; moreover three of them are markedly older than OED's examples, dating from 1836, 1846, and 1862.