Zambezia (1979), VII (i).SHONA LEXICOGRAPHYG. FORTUNEDepartment of African Languages, University of RhodesiaA REVIEW OF the second edition of Hannan's Standard Shona Dictionary1provides an opportunity, not only to evaluate the book itself as a dictionaryin its own right but, that done, to indicate the extent to which it carries for-ward previous work in Shona lexicography. It also provides an opportunityto pay a tribute to its author for his contribution towards the study of theShona language and the development of its written literature.In order to place Hannan's work typologically within the species ofdictionary, an article by the scholar Yakov Malkiel provides a set of charac-teristic dimensions with reference to which dictionaries may be analysed,and in terms of which they vary among themselves.2 These are the dimensionsof range, perspective and presentation.RANGEBy range is meant the 'volume and spread of the material assembled' and ithas three aspects, the density of entries, the number of languages covered,and the degree of concentration on strictly lexical data.3 A dictionary's densityis its approximation to an ideally complete lexicographic record. It is notwithin this reviewer's competence to say how complete or incomplete a recordthe Standard Shona Dictionary is. Writing to introduce his first editionHannan reckoned that the number of entries in the Shona-English part of thedictionary would exceed 20 000, and those in the English-Shona part lessthan half that number.4 Announcing the end of the proof-reading for thesecond edition at the end of November 1973, he estimated there were some81 000 entries in all, of which some 54 000 were in the Shona-English sec-tion.5 The second edition marks a considerable advance in coverage, and acomparison of almost any page with the corresponding section of the first' M. Hannan, S.J. Standard Shona Dictionary (Salisbury, The Rhodesia Litera-ture Bureau, 2nd edit., 1974).* Y. Malkiel, 'A typological classification of dictionaries on the basis of dis-tinctive features', in F. W. Houselander and S. Saporta (eds), Problems in Lexico-graphy (Bloomington, Indiana Univ. Press, 1967), 3-24.3 Ibid., 7-15.4 M. Hannan. 'Shona dictionary', NADA (1958), XXXV, 19.s Ministry of Education, Salisbury, Minutes of the Nineteenth Meeting of theShona Language Committee, 8 Oct. 1973, Item III.2122 SHONA LEXICOGRAPHYedition will bear this out. However, no one pretends that this dictionary isa complete record of all the words in spoken and written Shona. In order toform a judgement on the coverage of the present dictionary and its approachto a complete record, it will be useful to consider one or two points of me-thod. Hannan does not list homophones separately, a procedure which makesfor a reduction in the number of entries. For example the entry introducedby nhanga [LL] KKoMZ n 5 reads: Pumpkin (gen term). 2.Z House in whichsleep the unmarried girls of a family, cp kagero Ko (B); ruvhevhe M. 3.ZHostel for unmarried girls.6 This entry defines at least two distinct homo-phonous nouns.In a contrary sense, however, the number of entries is considerablyincreased by listing all the variant forms of the same word. For example bothmukombe and mukombo (ladle), both nhangande and nhangandi (one smallstick), and the three forms ngwibu, ngwichu and ngwidu (severing) are eachthe heads of separate entries. It is only the first in each case which is fullydeveloped but the others are duly entered and carry a rubric referring backto the main entry. The last example is one of three ideophonic variants whichare particularly prone to occur in phonemically slightly different forms andthereby to generate a multiplicity of entries.It may be asked what is meant by a complete record of all the words ina language. Discussing this difficult and probably unanswerable questionMalkiel refers to the practice of some lexicographers who deliberately smugglelatent words into their lists:In the languages possessing an arsenal of productive suffixes (suchas Slavic or Romance) a lexicographer can quite unobtrusively,manufacture, by the hundreds and thousands, derivatives of his ownmaking. These formations are readily understandable and give theappearance of authenticity; the truth is that speakers, for somereason or other, have not bothered to activate on the same scale allthese grammatical potentialities. It is difficult to detect such camo-uflaged illicit entries, especially in view of legitimate border-linecases (nonces); an unduly geometric design of representative wordfamilies, a dictionary's too heavy saturation with certain recurrentderivational schemes at once arouses the suspicion of an expe-rienced reader.Hannan's awareness of this possibility is expressed in several places inŁ The designations 'H' and 'L' in square brackets indicate the sequence of syllabictones carried by a word or, in the case of verb stems, the tone class they belong to.The designations 'K', 'Ko', 'M', 'Z', 'Ko(B)' indicates respectively the dialect clustersof Karanga, Korekore, Manyika, Zezuru and the Budya variety of Korckore. Theletter 'n' followed by a number indicates a noun and the class it belongs to. Hannanhas omitted to classify 2. as n 9.7 Malkiel, 'A typological classification of dictionaries', 9.G. FORTUNE23a paper he gave in 1959, just after the publication of his first edition.8 Aftermentioning the verbal extensions by which extended or derived verb stemsare formed from simple or basic stems, the derivation of nouns from bothsimple and derived verb stems, and, finally, the tense and aspect signs usedin the inflection of verbs which convey shades of meaning and precisions ofimplications which can be expressed in English only by very much longerforms of speech, he concludes as follows:These few condensed remarks about the Shona verb have, Ihope, left you with the impression that a pure African language isa much finer instrument, and wields a far vaster vocabulary, thanthe feeble hybrid called Fanigalo. In particular it evokes a certainamount of surprise when it is stated that the hastily compiled Stan-dard Shona Dictionary contains over 23 000 entries in the Shona-English section. But when you consider that for every verb stemthat is entered there are from ten to twenty derived forms also inuse, this really means that the Shona vocabulary, not all of whichcan be claimed to have been entered in the new dictionary, reallyexceeds 80 000.In order that the point may be understood, and Hannan's practice in histwo editions made clear, it is necessary to list the main forms of derivationthat exist in Shona. From both the numerical and structural points of viewthere are three main kinds of lexical morpheme9 in Shona: noun stems, verbradicals and ideophones. Hannan lists ideophones as such, verb radicals asstems, that is with terminal vowel /-a/, and noun stems with a primaryprefix.10e.g.thethetheideophonesverb radicalsnoun stemssvetugweju-sung--bat--rume[HL][HL][H][H][HH]-pofu [LL](jumping upwards)(wrenching off)cited as -sunga (tie)cited as -bata (hold)cited as murume n 1 (man,husband)cited as bofu n 5 (blindperson or animal)8 The paper, a copy of which is in my possession, was one of two delivered toa Conference which is not named or dated. The first is on Shona linguistic structureand the second is on Shona traditional literature. Internal evidence seems to show thatthe use of Fanigalo was being advocated at the time as a means of communication be-tween different language groups. The date was probably 1959.9 A commonly accepted description of the morpheme is 'the smallfst meaningfulunit in the structure of a language that is grammatically pertinent'; see H. A. Gleason,An Introduction to Descriptive Linguistics (New York, Holt, Rinehardt and Winston,1961), ch,5.>° Primary prefixes are the constituents with which noun stems most commonlycombine, and with which they indicate normal specimens of the items to which theyrefer. For example the noun mu-rume (man, husband) consists of the two constituents/mu-/, a primary prefix, and /-rume/, a noun stem.24 SHONA LEXICOGRAPHYDerivative forms from ideophones: The formation of both nouns and verbradicals in which ideophones are nuclei are regular and productive processes.Thus from the ideophone svetu we have the derived noun cha-mu-svetu-mu-svetu n 7 (jumping by many in all directions) and the derived verb stems/-svetu-k-l (jump up), /-svetu-dz-/ (make to jump). And from gweju wehave the derived noun cha-mu-gweju-tnu- gweju n 7 (indiscriminate wrench-ing by many) and the derived verb radicals /-gvveju-k-/ (be wrenched off),/-gweju-r-/ (wrench off) and /-giveju-dz-/ (cause to wrench off). The deri-vation of nouns is affected by the addition of appropriate noun prefixes and,normally, reduplication. That of the verbs is effected by the addition of oneor more of a number of suffixal 'verbalizers'.Derivative forms from verb radicals: Verb radicals supply the nuclei forderivative nouns, extended verb radicals and ideophones. Thus from the verbradical /-bat-/ we have the nouns mu-bat-i n 1 (worker, proprietor), mu-bat-on 3 (handle) and many others; the extended radicals /-bat-w-/ (be held),l-bat-ir-/ (hold for), /-bat-is-/ (hold firmly, strengthen) and many others; andthe ideophones bat-e (holding), bat-ei (holding gently), bat-anu (letting go).From the verb radical /-sung-/ (tie) we have the nouns mu-sung-o n 3 (any-thing used for tying things; a string trap), chi-sung-o n 7 (bond, anythingtied) and others; the extended radicals /-sung-w-/ (be tied; be made barren),/-sung-at-/ (tether), /sung-an-/ (be tangled) and many others and theideophones sung-e (tying), sung-ei (tying carefully), sung-unu (untying).Derivative forms from noun stems: Noun stems do not normally providenuclei for verbal or ideophonic constructions. The verb radical /-pofu-mar-/(be blind, cp. bofu (blind person) is a very rare instance of a verb radicalderived from a noun stem. However, noun stems do occur in a great numberof secondary forms by the process of substituting secondary prefixes for pri-mary ones, or that of adding secondary prefixes to primary ones. To quoteHannan again:The stem /-rume/ when combined with the prefix /va-/ givesrise to a noun varume which means 'men'. But chirume means 'man-ly behaviour', whereas rume without any expressed prefix means 'abig or ungainly or uncouth or crude man'. Urume means 'maleseed', while zirume means 'a gigantic man'... [Thus] for everynoun entered in the dictionary there exist at least three other forms,not all of which have been counted in the 23 000 entries. In thisway it would seem that 80 000 is too low a figure for the Shonavocabulary represented by the entries made in the dictionary. Anearer figure might well be 90 000."Entries of ideophones and their derivates: Hannan's practice in regard to thelisting of derivative forms in the first edition differs somewhat from his<> See above, fn.8.G. FORTUNE 25practice in the second. In regard to ideophones the procedure is the same inboth editions. One should note here to his considerable credit that the listingof massive numbers of ideophones is one of the main original contributionsof the Dictionary. He has made it clear that the class of ideophone con-stitutes a major category in the lexicon, comparable to those of the verbradicals and of the noun stems in both number and importance as constituentsin Shona grammatical constructions. Hannan's normal procedure is not onlyto list each ideophone but also the verb stems derived from it. Very oftenthese are three in number, intransitive stems formed by verbalizer /-k-/transitive steins formed by verbalizer /-r-/, and causative stems formed byverbalizers J-ts-/ or /-dz-/-. For example, the entry headed by the ideophonepogo is followed by entries headed respectively by the intransitive derivedstem /-pogoka/, the transitive derived stem /-pogora/, and the causativederived stem /-pogodza/. The entry under pogo lists five meanings, viz.KMZ ideo Breaking in two. 2. KZ Preparing food in abundance.3. KZ Emitting smoke. 4. MZ Making a big fire. cp.bvovo KZ. 5. M Biting long, semi-hard object held inhand (eg tsenza, edible tuber)All of these meanings are repeated under -pogora, viz.MZ v t Break, cp -bhogora KMZ. 2. Z Prepare food in abun-dance. 3. KZ Emit smoke, cp -togora MZ. 4. MZMake a big fire, cp -povora KZ. 5. Bite long, semi-hard object held in hand (eg tsenza, edible tuber)Three meanings are given under -pogoka viz.Z v i Break < pogo. cp -bhogoka KMZ. 2. Be prepared inabundance (food). 3. KZ Smoke, cp -togoka MZ.Under /-pogodza/ the reader is referred to the first four meanings listedunder/-pogora/. While it is probable that these derived verb radicals arenot merely latent forms in the sense described by Malkiel, this reviewerthinks that some more economical way of referring to their existence whenproved could have been used as their appearance as separate entries is largelyredundant.The entry of nouns derived from ideophones is much more economical,and those listed give the impression of having been encountered in the com-pilers' sources rather than having been consciously derived in the mannerof grammarians. Since nouns are listed not according to stem but accordingto class prefix, their other constituent, derived nouns are not normally foundin the vicinity of the nuclear ideophones. Thus the ideophones chachu (eatingdirectly from the cooking pot) and di di di (running with short steps) providethe nuclei for the nouns chamuchachu muchachu n 7 (continual taking offood from cooking pot on the fire by many) and chamudididi n 7 (runningwith short steps). The ideophones tsverere (slipping on slippery surface)and tsvoti (pouring of liquid) provide the nuclei for the nouns matsverere n 6( kid) and matsvoti n 6 (abundance of liquid).26 SHONA LEXICOGRAPHYEntries of verb radicals and their derivates: In regard to derivates from basicverb radicals the first edition made a normal practice of listing all the exten-sions (as derivative suffixes with terminal vowel /-a/, in each entry. As Han-nan wrote in the introduction:Verb forms have been entered as simple stems, that is without anyprefixal formatives; e.g. -bva. The suffixal formatives which, byaddition to the root, give rise to derived species of verb stems, areentered after the simple stem; e.g. -bva. -ana; -ika; -isa; -iwa; -ura;-wa. This means that in addition to the simple form -bva, there arethese derived species of verb stems: -bvana; -bvika; -bvisa; -bviwa;-bvura; and -bvwa. One who wishes to find the meanings of thederived stem, -bviwa, will not find any dictionary entry, -bviwa. Hewill find -iwa, the suffixal formative, which, when added to theroot -bv-, results in the passive species of derived verb stem -bviwa.He will also find, as already stated, the simple -bva and its equiva-lent meaning in English. Some of the derived stems are of rare oc-currence, others are heard more often than the simple stem. Noattempt has been made to indicate in the dictionary which of thederived stems occur frequently.' '-Reference to the entry under -bva reveals that five other 'suffixal for-matives' have been listed in addition to the six listed above, namely -anya.-am, -ata, -aura, -ira. Of these eleven 'suffixal formatives', however, it isdoubtful whether more than five occur with the radical /-bv-/; /-bva/ means'move from' in English, /-bvika/ means 'be capable of coming out', /-bvira/'move from', /-bvisa/ 'take from', /-bvwa/ 'be gone from'. But forms like/-bvana, -bvanya, -bvara, -bvata/ etc. are either non-existent or are unrelatedto /-bva/. Now the entry of every simple verb stem in the first edition isaccompanied mechanically by a similar series of suffixal formatives, usuallybetween ten and twelve in number, indicating the existence of derived verbstems many of which are not only infrequent but non-existent.This almost mechanical repetition of the derivative suffixes, almost be-cause there is variation in the number listed and this points to a degree ofchecking or verification having been done, is the most serious flaw in theentries of verbs in the first edition, and has been avoided in the edition nowunder review. Instead we have now the listing of simple or basic verb stemsas well as of the commoner derived stems. For example, in addition to/-famba/ (move, walk, progress, journey, travel), we have /-jambidzana/ (beon cordial terms, get on together, be compatible) and /-fambisa/ (move,direct, drive. 2. walk quickly). Other derived stems formed by the use ofextensions (or derivative suffixes) feature among the examples under theentry headed by the simple or basic stems. For example, under the head** Standard Shona Dictionary (London, Macmillan, 1st edit., 1959),viii..'*'Ł--G. FORTUNE 27/-posha/ (lend) we have the causative /-poshedza/ and the passive /-poshwa/illustrated. Thus the procedure followed in respect of extended verb stems inthe second edition is much more objective and documented, and avoids thechange of inventing not only latent but non-exsistent forms.Some indication is given in this edition of the use of verb radicals asnuclei in the formation of ideophones. Thus both /-moma/ (settle on andcover as do bees) and mome (settling on and covering) are listed but withoutthe relationship being indicated. Similarly f-monaf (wind around etc.), mono(turning), monono (unwinding) are listed, again without the relationshipbeing indicated. Some ideophonic constructions with verbal nuclei are veryproductive, for example those with ideophonizer /-e/, and, like the use ofcertain of the verbal extensions, for example the passive /-w-/, do not needexemplification. Other ideophonic constructions in which verb radicals areconstituents, for example with the ideophonizers /-au/ (extensive idea) andl-anu/ etc. (reversive idea), are not so predictable, and need entries, or atleast an indication in the entry which treats the verb stem, that such ideo-phonic derivations exist.Nouns which are constructed with verb radicals as nuclei are often butnot consistently indicated under the verbal entry. For example under /-tadzu/(be unable, fail to do, do wrong) the derivation of the nouns mutadzi n 1(one who has done wrong and been found guilty) and rutadzo nil (unsuc-cessful attempt, sin), each with its own entry, is indicated. Here again thechoice of derived nouns for entry has probably been decided not on thegrounds of their latency but of actual occurence in the sources.Entries of noun stems and their derivates: In regard to the entry of derivatesin which noun stems are nuclei, the rather rare cases of verb radicals derivedfrom noun stems have been mentioned, as well as the constructions whicharise out of the secondary prefixes with noun stems. The latter may be con-sidered as derivative forms in an analogous sense. Here again certain forma-tions such as those of diminutives and augmentatives are common andpredictable. Others need to be indicated. For example, in the case of muko-mana n 1 (boy), if the rules for the formation of diminutives and augmenta-tives by using secondary prefixes had been treated in the Introduction, itwould have been unnecessary to list gomana n 5 (big fellow) as a separateentry.Returning to the question posed earlier as to what is meant by acomplete record of all the words in a language, it seems impossible for alexicographer to know what a complete record consists of, and just as im-possible to make one. Strangely enough Fr Hannan believed that he hadrecorded as many as 80 per cent of the words used by Shona-speakers inhis first edition.13 After compiling the second edition in which the ShonaStandard Shona Dictionary (1959), ix.28 SHONA LEXICOGRAPHYentries were two and a half times as many he was more cautious and wrote,calling for further contributions, 'In spite of all that we have done to makethis work as complete and accurate as possible, there must be manyomissions'.14 He invited people to send in contributions found to be missingfrom the Dictionary as a result of work done in their own fields so thatthey could be collected, published annually in NADA and added to futureeditions. As an example of a gap in the record this reviewer finds the diction-ary deficient in words used in the diction of traditional poetry. No doubtmany of these words are passing out of the speech of all but the oldestage grades, and this fact points to another aspect of the difficulty in deter-mining what is a complete record. For with changes in culture the lexiconis proving unstable. The entry of loan words, more accurately referred to asadoptive words, is a more prominent feature in this edition than in theformer. Further editions will give greater prominence to them.'5One reason for the vastly increased coverage in the second edition isthe increase in published Shona literature in the years when it was beingrevised (1967-73), as well as the greater use of what was available before.Written literature appears not to have been used as a source for the firstedition owing to the fact that the compiler was under contract to complete itin the shortest possible time. In the article in NADA in 1957 Fr Hannanrevealed that his list had been compiled from material collected by otherworkers, and these sources are indicated in the Introduction. For the secondedition he had the collaboration of a number of people to fill gaps in thedialectal record, in addition to his team of compilers, as well as the assistanceof the Rhodesia Literature Bureau. However, after the last entry has beenmade, the modern lexicographer still has to recognize, like Dr Johnson, theinevitable deficiencies of his work: 'No dictionary of a living tongue canever be perfect, since while it is hastening to publication some words arebudding and some are falling away', and 'He that undertakes to compile adictionary undertakes that which, if it comprehends the full extent of hisdesign, he knows himself unable to perform. Yet his labours, though deficient,may be useful, and with the hope of this inferior praise he must incite hisactivity and solace his weariness.'16A further reason why any dictionary of Shona will always have tocontent itself to be an approximation to a complete record lies in the numberof complex nominal constructions in current use. No doubt the number of~??d?d Shona Dictionary (1974), ix.ihis feature is dealt with by H. Ghimhundu, 'Some problems relating to themcoiporation of loan words in the lexicon'; see below, 75-91.first Ł t° *le remarks of Dr Samuel Johnson, the great English lexicographer. Thepinrl 1 om the Preface to his A Dictionary of the English Language (1828 edition),TOR ,.from the Advertisement to the 1773 edition; they are quoted in. ^ uenson, A century of Bantu lexicography', African Language Studies (1964),G. FORTUNE 29ideophones, basic noun stems and simple verb radicals is a finite one. Yettheir record constitutes in itself a very formidable task. What are we to sayof the numerous combinations of these three basic types of morpheme asnuclei of larger nominal constructions of which the dictionary records anumber? For example the complex nominal construction mudyandigerewhich as a class 1 nominal means 'a lazy person, pensioner', and as a class 3nominal means 'a person', cp. /-dya/ (eat), ndigere (I being seated).Another side to the density of a dictionary is its coverage in depth,that is the extent to which it goes in recording and documenting differentmeanings of items, including the different connotations which depend ondifferent contexts. The majority of entries are provided with clear and concisedefinitions. Often the English glosses corresponding to a Shona entry haveto be listed and numbered from No. 2 onwards. The first of a series is nevernumbered. Numerical listing of glosses occurs where the head of the entryis homophonous, a number of different words or morphemes having the samephonemic form and the same tone pattern. An example has already beencited, namely nhanga [LL\, which is both a noun of class 5 meaning 'pump-kin', a general term common to all dialects, and in the Zezuru dialect, anoun of class 9 meaning 'a hut in which unmarried girls sleep'. Many ideo-phones are homophones, e.g. gadu, which in Zezuru has the following mean-ings in English: (1) 'striking someone in the face'; (2) 'making a deepincision in someone'. In Manyika this form has two different meanings whichare listed as (3) 'chopping', and (4) 'hoeing hard ground'. This exampleshows that homophonous forms often have regional connotations. Thisindication of the regional and dialectal distribution of words is a secondmajor contribution of the dictionary together with the concurrent indicationof synonyms which occur in other dialects. Thus, in the case of the thirdand forth meanings of gadu, namely 'chopping' and 'hoeing hard ground',which occur in Manyika, the reader is referred to the Zezuru forms ga(chopping) and gau (hoeing hard ground). This information about dialectalincidence and dialectal synonyms is extremely useful, making such formsmutually intelligible and available to a common literature.With the movement of people from their traditional homes where theirclans or sub-clans have their roots, and the growth of towns with new urbanforms of Shona whose influences in country districts is considerable, theterritorial definition of the Shona dialects which dates back some fiftyyears has obviously been profoundly modified, and a new survey will beneeded to describe the sociolinguistic position today. The terms 'Z', 'K','M', and 'Ko' stand for the clusters of dialects defined by Professor C. M.Doke in 1931, Zezuru, Karanga, Manyika and Korekore. The areas to whichthese dialects are referred are shown in the maps on the end papers. It isprobable that narrowly dialectal terms are now restricted to the oldest agegroups.The dialectal dimension stresses space. The time dimension in whichSO SHONA LEXICOGRAPHYphilological analysis and research moves is represented by the starring °^forms which, are reflexes of items listed in Professor Guthrie's ComparativeBantu" and so likely to have been in Shona, showing its relationship toother Bantu languages, for a very long time.Often, but not in the majority of entries, illustrative examples are p*°"vided on no very clearly discernable criterion. Common verbs whose ex-tended forms call for some treatment in the absence of their separate entryare illustrated, as well as those which cover several shades of meaning iQEnglish. An example is /-ramba/ corresponing to 'refuse', 'demur', 'deny% aswell as 'persist'. The latter meaning should have been distinguished as thatof a separate homophone. Another ground on which examples are providedis when the head of an entry is the key word, or the nucleus of a key w<->rd>in a proverb or riddle. For example, the case of J-bata/ (touch, hold, catch,seize, acquire, practise, capture) calls for the citing of the proverb, Chawa-wana batisisa, mudzimu haupe kaviri (Hold on tightly to what you have,the spirit elder does not give twice). Proverbs are indicated by the abbreviartion 'prov'. In other cases examples are cited to confirm the grammaticalidentity of the head of the entry. For example zvino (now) is cited in itsinflected form in the example Uchiri kuHarare nazvino wose (Are you stillin Salisbury even now?) to show that it is indeed a noun of class la. Wherean item is peculiar to Shona culture or unknown in English an example willbe illustrated. For example j-sevaj (dip morsel of sadza into side dish) isexemplified by the sentence: An kuseva sadza mumuto (He is dipping hismorsels of sadza into gravy before conveying them to his mouth).Considerations of economy and the need to compile a dictionary ofconvenient size and price no doubt decided the compliers to limit the numberof illustrative examples. This points to the exercise of discrimination andself-denial in a man so familiar with Shona culture and anxious to do itjustice. Clarity and conciseness are always present. Perhaps in a furtheredition where space can be saved by the omission of predictable derivedforms more space can be given to many items of Shona belief and culturewhich call for further description and exemplification, for example njuzu(water sprite), mudzimu (spirit elder of a family), dare (meeting place formen of village), nhembo (elegaic poetic utterances), nhango (didactic poetry),mutupo (clan name), chidao (principal praise name). Further we should likea treatment of items in relation to the social situation in which they aresuitable and to which they may be restricted. There are a number of dis-tinctive speech styles in Shona which are part of certain social relationships,for example those that link the families of a bridegroom and bride, or which" M. Guthrie, Comparative Bantu : An Introduction to the Comparative Linguis-tics end Prehistory of the Bantu Languages (Farnborough. Hants, Gregg International4 vols, 1968-72).G. FORTUNE 31are appropriate to certain social situations, such as a village court. Theonly indication of this sort in the dictionary at present is the labelling ofcertain expressions as coarse or tactful. The dictionary does not define thesocial relationships or contexts when the use of such words would beoffensive. In some contexts they are obviously in place. Our knowledge ofthe distinctive markers of these distinctive speech styles is as yet rudimentarybut, as these become identified, and to the extent to which they are lexicalitems, they should be incorporated as such in the dictionary where theywill reveal the great inner diversification of Shona speech. Another kindof addition we hope for is the inclusion of more idioms. This reviewer knowsthat Fr Hannan had an ear for these and that he had collected many Š forexample, those including the nuclear word mwoyo (heart). Certain of thesehave been entered but there were many more. He has obviously had to re-strict himself to the limits of this medium-sized dictionary.Two other aspects of the dimension of range were mentioned at thebeginning of this section, namely the number of languages covered and thedegree of concentration on strictly lexical data. What we have is a Shona-English dictionary. In both editions Fr Hannan emphasized that in termsof the task he had accepted and its immediate aims he could provide onlyan index in the English-Shona section. 'The coupling of a full-sized "dic-tionary" and a meagre "vocabulary", typically one fourth or one fifth itssize, which is in reality little more than a makeshift index is not un-common at all because often the two sections have deliberately not beenplanned on the same footing.'18 In this second edition the main English-Shona Index is followed by three appendices which are indices to botanicalnames, to interjectives and to ideophones. The main index and those ofinterjectives and ideophones have an interest additional to their usefulnessin tracing Shona equivalents needed for English words. From a cross-cultural point of view they show graphically the areas of life where Shonavocabulary has proliferated and diversified in response to cultural needs andinterests. For example the treatment of beer occupies two columns, whileeverywhere the stylistic emphasis on graphic representations of actions andevents is shown by the multiplicity of ideophones and their derivates. Forexaniple the Shona equivalents for various postures adopted while sittingoccupies a column, while those expressing various kinds of gait duringwalking occupy almost six.From the description and evaluation so far it will be clear that theconcentration in the dictionary is predominantly lexical. I have expressedmy hope that it may become more encyclopaedic in future editions, giventhe many special features of Shona culture which have no equivalents in»8 Malkiel, 'A typological classification of dictionaries', 12.33 SHONA LEXICOGRAPHYEnglish. Again with a view to the future and with a view to the more econo-mical lexical ordering of the material the following suggestions are made forthe organization of entries involving ideophones, vero radicals and nounstems. They incorporate the criticisms made so far in this review.Much more use could be made of the Introduction to provide a simpleoutline of the linguistic structure of the language and, in particular, theway in which the basic classes of morpheme are related in normal produc-tive processes of derivation. With a little more linguistic insight on tne partof readers, the setting out of the dictionary entries could be made moreelegant, more economical and reflect better the structure of the languageof which it is the lexicon. In the case of ideophones this would mean listing(a) the structures of ideophones and verbalizers which form derived 'deideo-phonic' verbal radicals and (b) the structures of noun prefixes and ideophoneswhich form complex 'deideophonic' nominal constructions. The mostcommon derivations, both verbal, i.e. with verbalizers /-k-, -r-/, andnominal i.e. with prefixes /cha-mu-/ and reduplicated stem, are largelypredictable in their meaning, and no more than an indication of theirexistence would be necessary in an entry with an ideophone headword. Theless common derivations, e.g. with verbaiizer /-/-/ or with noun prefix /ma-/would require both an entry in the appropriate place among the derivativeforms as well as their English meaning. For the rest the typical entry withideophone headword might run as follows.19e.g. simu [HL](1. Dialectal distribution) KKoMZ(2. Linguistic category) ideo(3. English equivalent) Rising*0(4. Synonyms) cp sumu, kwaicu(5. Verbal derivates with) /-k-, -dz-, -is-/(6. Nominal derivates with) /cha-mu-.. . mu- Ł../(7. Examples) Akati simu pasi (He rose from theground)Simuka, tiende (Rise, let us go!)Simudza ruoko (Raise your hand!)Vanhu zvavakati vavhundutswa neshumba,chakava chamusima musimu (When they werestartled by the lion, the people jumped uphelter-skelter)The entry would end, where necessary, with notes on (8) grammatical,stylistic and social restrictions on use; (9) idioms in which the item occurs;and (10) cultural aspects needing interpretation. In the case of simu therewould be little or nothing to add.'Ł The aspects listed on the left of the entries which follow indicate merely the? A A u ' ll " su8gested, the material of each entry should follow. It is not in-tended that they should be made explicit in the entries thpmse.lves.id h e " no nee<* to use *he preposition 'of when giving English equivalents ofG. FORTUNE 33Setting up entries in this way with ideophones as headwords woulddouble the size of the entry but it would render the listing of derivates else-where unnecessary. It would also have the merit of concentrating all theinformation about the ideophone, both in itself and as a nucleus for derivativeforms, in one place, so that a complete grasp of its potentiality as a consti-tuent form in larger constructions, both morphological and syntactic, becomespossible.In the much more complex case of verbs the Introduction should des-cribe (a) the structures of verb radicals and extensions which form extendedverb radicals; (b) the structures of verb radicals and extensions with terminalvowels which form deverbative nouns; and (c) the structures of verb radicalsand 'ideophonizers' which form deverbative ideophones. Once again themost common derivates resulting from regular productive processes wouldneed no more than the indication that they are in use and are not merelylatent. For example in the case of the verb radical /-bik-/ (cook) the exist-ence of the passive extended radical /-bikw-/ or /-bikiw-/ could be indicatedby the mere inclusion of the passive extension /-w-/ or /-iw-/. Similarly theexistence of the agentive deverbative noun mubiki (cook) and of the deverba-tive ideophone bike (cooking) could be indicated merely by the inclusionof the affixes concerned, namely /mu-^-i and /</,-#/. Less common de-rivates and those whose meaning cannot be easily deduced from a produc-tive rule would need not only the inclusion of the affixes concerned in theentry but the provision of English equivalents and illustrative examples.An example of a typical, rather straightforward, verbal entry might beas follows:e.g. -bik- [L](1. Dialectal distribution) KKoMZ(2. Linguistic category) R (viz. verb radical)11(3. English equivalent) Cook in a liquid, brew, stew.(4. Synonyms) (Nil in this case.)(5. Verbal derivates)22 /-an- cook one another, e.g. in folktaleplots; -idz-, help to cook; ik-; -ir-; -is- cook: i.e. causeto get cooked; -w- or -iw-/(6. Nominal derivates) /mu- «« /; mu.<* -o taste; chi- »/Ł> -ocooking pot; chi-m-iro cooking vessel; chi-^-wacooked food; c/zz-1/>-e «/>-e cooking without method;ma- "> -o feast; ma- to -irwo way of cooking; ma- <*-irwoway of being cooked; *» -o cooked food or drink;ma- <" -viia smarting pain; cooked food/zi The term 't' (for transitive) and 'i' (for intranssitive) and should give place ti>the more useful classification of verb radicals with reference to their potentiality foroccurring with primary and secondary object complements which has been proposed byDr N. C. Dembetembe; see below, 'A syntactic classification of non-auxiliary verbs inShona, 49-59.zzCombinations of extensions will also be entered, e.g. /-idz-an-/ etc.34 SHONA LEXICOGRAPHY(7. Ideophonic derivates) /^-e msi cooking lightly;/ i/> -du/ cooking without care,grossly/(8. Examples) Ndibikire sadza (Cook some sadza for me)Teza huni dzokubikisa doro (Gather firewood to brewsome beer) Ndabikwa nezuva (I am perspiring, lit.am stewed, because of the sun) Ndibikidzewo (Pleasehelp me to cook) Bikira muhari iyi (Cook in this pot)Pamabiko zvokudya hazvibikwi natoga, asi vadzimaivanobikidzana (At a feast the food is not cooked byone person alone, the women all help together withthe cooking. Iti zai rangu bikei (Lightly boil my egg)(9. Idioms) Abikwa neshamhu (He was thrashed with a stick)-bika-bika (beat all over, severely; also, cook withoutmetnod or purpose).Grammatical, stylistic or social restrictions on the use of this verb radicalmight follow as well as cultural interpretation. Since /-bik-/ appears to bea neutral item in both these respects, there would probably be nothing to add.The entry of noun stems in the Standard Shona Dictionary departs fromthe procedure adopted for ideophones and verb radicals. Benson writes:The main difficulty in Bantu dictionaries is the method of listingthe nouns. Many Bantu lexicographers in the past and also todayhave argued that for the guidance of the non-expert user of thedictionary alphabetical order by prefixes is preferable. The resultis of course a very considerable number of entries under the initialT?t(?r ^e common prefixes such as mu-, mi-, . . . and so on.Under this system, if it is consistently followed, the plural of anoun should have its own place separate from the singular. Theresult is a great waste of space, and in fact this double entry hasseldom been carried out in full. One cardinal principle which emer-ges from our study is that everything which needs to be said abouta stem or root should be channelled into one single full article,complete with citations if needed. If, as sometimes happens, thesingular and plural have current meanings it is important thattnese should be shown alongside each other.*3Hannan concedes that 'noun forms, as well as verb forms, would havemade the compilation much shorter had they been entered as single stems'.Nevertheless, 'nouns have been entered with their prefix because that is theform in which they are heard or written, and because the use of differentprefixes with the same stem often results in words of quite different mean-ing; e.g. munhu (person) and kunhu (place).'24 For Benson the juxtapostionor examples of the same stem with different prefixes is desirable as throw-ing light on the stem, the common nucleus, combinations of which withStandard Kiku^DictionlV6^081^^'' ""^ ^ " ietminS * hlS cxPerience ondard Kiku^DictionlV8^^' ^ ^ " ietmin"Standard,Shona Diltionary (1974), Introduction, viii.G. FORTUNE 35different prefixes denote its different realizations. Thus it is instructive tocompare munhu (person or human entity), chinhu (thing or material entity),and.kunhu (locality or locative entity). If it was judged suitable to enterverb stems without prefix and to expect readers to be able to ignore verbalinflections in finding them, surely the same would have been possible withthe much simpler structure of the noun? Entry of nominal entries by stemsis the normal practice in dictionaries of Nguni and Sotho," and is the pro-cedure being followed in the Ndebele-English Dictionary currently beingcompiled in the Department of African Languages, University of Rhodesia.Alphabetical listing of noun stems would still allow them to be enteredin combination with their primary prefixes, either prefixed to the stemsbut in lighter type, as Paroz does in the Southern Sotho-English Dictionary,or entered in brackets immediately after, as is the practice of Doke andVilakazi in the Zulu-English Dictionary.An example of this sort of entry might be set up as follows:e.g. fflu-rutne FLHH](1. Dialectal distribution) KKoMZ(2. Linguistic category) n 1(3. English equivalent) Man, husband(4. Synonyms) (Nil in this case)(5. Nominal 'derivates') /rume 5 ma-; chi- "> 7 small, plumpman; male fashion; one of the four divining dice(hakata) used by the diviner (n'anga); ndume 9 male,human or animal; ru- v> 11; ka- io 12; u- u> 14 male(coarse expr)/(6. Nominai constructions, i.e. with /-rume/ as suffix)munhurume 1 male person;mwanarume 1 male child;tsvimborume 9 unmarried male;gukurume 5 cock:gombarume 5 philanderer.(7. Examples) Murume akanaka asiri wdko (A man is attrac-tive when not your husband (prov)) Chirume kufumirapachimwe, kumuromo chakatakura (For a fellow togo early to another means he carries something onbis lips, i.e. has some business to discuss (prov))Rume rimwe harikombi churu (One man, howeverbig, cannot surround an ant-hill, i.e. he needs thecooperation of others (prov)) Karume kangu (Manfeared for his fighting ability).Once again restrictions on the use of this noun, and some cultural interpreta-tion throwing light on its connotations, arising perhaps out of the institutionof polygamy, might follow.BO For example C. M. Doke and B. W. Vilakazi, Zulu-English Dictionary (Johan-nesburg, Witwatersrand Univ. Press, 1948), and A Mabille and _H. Dieterlin, SouthernSotho-English Dictionary revised by R. A. Paroz (Morija, Morija Sesuto Book Depot,1950).36 SHONA LEXICOGRAPHYNouns with zero prefixes will obviously be listed according to theirinitial letter and without any prefix. For example mambo la (chief), jeso 5(devil thorn), shongwe 9 (rocky pinnacle). Nouns with non-syllabic prefixeswhose stems have undergone some prefixal morphophonemic change shouldlogically also be listed according to the first letter of their stems. Forexample banga 5 (knife) should appear under 'p' as (ri-)panga, and hama 9(kinsman) under 'k' as N-kama. But such a procedure would truly be confus-ing and it is probably better to list nouns of these classes under the initialletter of the whole word. As Benson says, referring to the corresponding casein Kikuyu. 'Every language has some element that will not conform to thegeneral pattern.'26PERSPECTIVEThe aim which a compiler of a dictionary has in launching and delimitinghis work will have very practical effects on the type of material selected,its arrangement and its tone. Malkiel lists two basic and complimentary per-spectives, diachronic or the historical ordering of linguistic facts, and syn-chronic, the lexicon as it is today. As already stated the dictionary has asingle historical and comparative strand in that Shona reflexes of CommonBantu stems are starred. For the rest the aim is to make a contemporaryrecord, at the same time not showing an undue haste in the recognitionof adoptives before they become current.The aims of the compiler are very clearly stated in the Introduction.They are (1) to record Shona words in Standard Shona Spelling; (2) toprovide, by a number of examples of the use of words in sentences andphrases, illustrations of the application of the principles of word-divisionon which Standard Spelling is based; (3) to provide as complete a list aspossible, within the time and with the facilities available, of the words usedboth in speech and writing by those whose dialects are known as chiKaranga,chiKorekore, chiManyika and chiZezuru. These dialects or dialect clustersare what is known as Central Shona, and exclude both Eastern Shona orNdau as it is called in Rhodesia, and the Western Kalanga dialects. Wethink, in passing, that Ndau could well have been included among thedialect clusters which contribute to the Dictionary, as it was originallyintended, from the first recommendations on unification, that 'a Dictionarybe prepared, to be as inclusive as possible of words from Zezuru, Karanga,Manyika and Ndau.'27 The reason given for including these dialects at that»Ł Benson, 'A century of Bantu lexicography', 83." Southern Rhodesia, Report on the Unification of the Shona Dialects by ClementM. Doke (Sessional Papers,, C.S.R. 25, 1931), Recommendation 5, 81.G. FORTUNE 37time was that the first essential of the dictionary was to recognize and in-corporate as far as possible words from those dialects which had beenused as literary media.At this point it is desirable to document the stages in the productionof the dictionary by referring to the two bodies which, along with Fr Hannanand his team, were responsible for its compilation. In May 1954 a com-mittee of the Native Affairs Department of Southern Rhodesia, called theInterim Literature Committee, passed a resolution recommending that astandard Roman alphabet orthography be accepted for chiShona for allofficial purposes. This resolution was followed by another requesting theGovernment to appoint a committee to devise a method of writing thesounds of the Shona dialects with the letters of the Roman alphabet, withdue regard to the system at present in use in the schools, and bearing in mindthe problem of word-division and the need for urgency.28 A representativecommittee, called the Shona Language Committee, was appointed, and, atmeetings in July 1954 and February 1955, produced the required standardRoman alphabet to replace the Union Shona orthography approved in 1932which, with its six non-Roman special symbols, had, in the twenty-threeyears of its use, failed to gain general acceptance and was holding up theproduction of Shona literature. At the latter meeting the Committee re-commended1. that a comprehensive Shona-English, English-Shona dictionarybe prepared including (a) an introductory grammatical out-line; (b) comprehensive vocabularies of the Shona dialects,excluding the western group; (c) tone-marking of dictionaryentries only.2. that the compilation of the dictionary be financed by theGovernment as a project of the African Publications Bureau.A competent and elegant small dictionary had already been preparedby one of the members of the Language Committee which had devisedand recommended the 1932 orthography.29 It was necessary to replace thisas soon as possible in order to standardize the 1955 system of spelling. Notime was lost in commencing the work, the Publications Bureau commis-sioned the services of Fr Hannan as lexicographer, and by January 1956 theSecretary of the Shona Language Committee, which had set the guide-linesfor the new dictionary, announced in a circular to all interested persons thatwork on the dictionary had commenced, and 'your assistance is requested28 Ministry of Internal Affairs, Salisbury, Minutes of the Third Meeting InterimLiterature Committee, Native Affairs Department, 19 May 1954, Items 1 and 2.29 H. Barnes, C. R., A Vocabulary of the Dialects of Mashonalttnd in the NewOrthography (London, The Sheldon Press, 1932).38 SHONA LEXICOGRAPHYso that this standard dictionary will contain every possible Shona wordnow in current use'.30At its meeting on Wednesday, 4 January 1956, the Shona LanguageCommittee decided to omit ChiNdau from the dictionary.3' No reason isapparent in the minutes, but the fact that the new 1955 Standard Spellingwas considerably less accommodating to local variations than the 1932 wasprobably responsible.32At the next meeting on 9 January 1957 the Committee expresseditself as very satisfied with the progress made and voted unanimously toextend the contract with Fr Hannan and his assistant, Mr E. Gumbo. Thefirst printing by Macmillan was of 10 000 copies which sold at l0s.6d. percopy, and the date of publication was December 1959. The Shona LanguageCommittee under the Chairmanship of the Revd S. K. Jackson continued tomeet until the dictionary was published, helping with technical advice, untilits last meeting as a Committee of the Department of Native Affairs on 5 July1960. The chairman congratulated Fr Hannan and Mr Gumbo on theirwork and, by implication, the Southern Rhodesia African Literature Bureauwhich had been responsible for arranging both compilation and publication,and which continued to be responsible for distribution. A reprint of another10,000 copies was authorized, and Fr Hannan was invited to prepare arevised edition incorporating words collected during the next few years.It was agreed that tone-marking should be included.At this stage several changes took place. In 1963 the Shona LanguageCommittee became a committee of the Ministry of Education. The RevdS. K. Jackson retired to devote himself to Bible Translation and to adaptthe Bible of the 1955 orthography, while Fr Hannan was voted Chairmanto take his place. He was now anxious that the compilation of the revisedand fuller dictionary should be entrusted to an African graduate.33 At thenext meeting on 8 November 1965 he asked the committee whether therewas any support for a Shona-Shona dictionary and an English-Shoaadictionary with a Shona Index including tones. These could replace thepresent Standard Shona Dictionary when out of print. There had been someresistance to the present dictionary because it had been compiled by aEuropean instead of an African. As there had been no success in inducingan African graduate to undertake the work, the Committee voted for the»° Ministry of Education, Salisbury, Circular Letter signed by J. C. Cheney,secretary Shona Language Committee, 16 Jan. 1956.31 Ibid., Minutes of the Meeting held on Wednesday, 4 Jan. 1956, Item 5.« Compare the latitude allowed by Dolce, Report on the Unification of the $h»naŁ vrl ' 5ara- 158' 'Explanations and Examples in the New Orthography", with therigidity adopted in the Guide to Standard Shona Spelling (Salisbury, ChishawashaJViission Jrress, 1955).33 Ministry of Education. Salisbury, Minutes of a Meeting of the Shona LanguageCommittee held on 30 Nov. 1964.G. FORTUNE 39revision of the Standard Shona Dictionary with the inclusion of tone-markingin such a way as would not add noticeably to the expense.34The revision and redrafting of the second edition began in earnest in1967 and was to last until 1974 when the final pages of the manuscriptwere brought in to the publishers, the Rhodesia Literature Bureau, and thehistoric photograph of Fr Hannan and his team taken which appears onthe dust cover. Seven years of unremitting work saw the work through, butit would have taken much longer had it not been done in close collaborationwith the Bureau staff. It was the Bureau which paid the salaries of theteam through its vote from the Rhodesian Government. The dictionary teamand the editorial staff of the Bureau checked and re-checked each other'swork. The manuscripts submitted to the Bureau, as well as its publications,provided thousands of new entries. When a halt was called to the compila*tion, the Publications Officer, Mr E. W. Krog, who was also Secretary tothe Shona Language Committee, arranged for the publication of the dictionaryby Mardon Printers as the Government was understandably reluctant tospend foreign currency on a project which could be produced within thecountry. The Bureau editors were called in again to collaborate with theproof reading. The 10,000 copies printed were lodged in the Bureau ware-house and the Publications Office initiated a publicity campaign to pro-mofe the sales. Owing to Bureau sponsorship and publication it was possibleto retail the 1,000 page volume at the sub-economic price of Rh$4,50. Todate, of the 10,000 copies printed, over 7,000 have been sold, and urgentthought must be given already to producing a reprint. The whole projectis a magnificent example of hard work, collaboration, sound business senseand local enterprise in support of the growth of written Shona. Those whohave to publish in the divided dialects of the languages to the south of us,Nguni divided into Zulu, Xhosa and Swazi, and Sotho divided intoSouthern Sotho, Pedi and Tswana, look with envy on the achievement, in-complete and precarious as yet, of a standard written form, common to theno less heterogeneous dialects of Shona. The two figures who stand outprominently in this work of integration are those of Doke and Hannan.The work of the first lasted little more than a year. It has proved a solidand enduring foundation, a sound piece of phonetic analysis applied to theproduction of a common viable spelling and writing system. The work ofthe second, one could say, lasted forty years from the time that MichaelHannan began to prepare himself for work among the Shona people untilhis death in 1977. His magnum opus scriptum, his great written work, grewout of his magnum opus factum, the work he did for and among the Shonapeople. His knowledge of the Shona language and culture was put at the34 Ibid.. Minutes of a Meeting of the Shona Language Committee held on 8 Nov.1965, Item 3, Revised Edition of Dictionary.«*O 3HONA LEXICOGRAPHYservice of what might seem a rather narrow aim, a record of words andsentences to illustrate the principles of Standard Shona. He disciplined itinto the system of entries, alphabetically arranged and consistently structured,that we have, its limited orthographic prescriptive aim caught up in a scholar-ly and objective work. His combined knowledge and love of the languagehas filled this enterprise that it is mainly the third aim, abundantly fulfilled,that we are aware of, namely the provision of as complete a record of Shonaspeech as possible 'in the time and with the facilities at our disposal'.PRESENTATIONIn Malkiel's scheme for dictionary classification 'perspective' envisages thebroad policies, while 'presentation' serves as a convenient term for sub-suming narrower preferences such as typographic style, use of specialsymbols and abbreviations, latitude of definition, volume of verbal docu-mentation, graphic illustration and many similar externals.35In presentation the Standard Shona Dictionary is a model of clarityand easy reference. Heads of entries are in bold lower case, and are followedby their tone patterns, distinguished, where necessary, according to dialect.The abbreviation indicating the linguistic category follows, and then theEnglish equivalents. Shades of meaning of the same morpheme are dis-tinguished and numbered in the same way as homophones which are dis-tinct morphemes of quite unrelated meaning Š a procedure which importsa certain imprecision from the semantic point of view. In each sectionof the entry synonyms are indicated with their dialectal incidence, and theresult is a tightly built whole, drawn together by innumerable cross-references. Derivative forms are sometimes indicated according to section,as well as the forms from which the head of the entry is derived. Afterthis, illustrative examples follow, but not in the majority of cases. Manyof the examples are proverbs.As has already been stated, the vast majority of entries in theDictionary are of nouns, verb stems and ideophones. Of the remaining non-nominal substantives, adjective stems are treated like noun stems and listedwith one or two representative prefixes. Enumerative, quantitative andselector stems are listed, however, as stems. The pronouns of each person andclass are listed, and all the demonstratives in both simple and emphaticforms.Noun stems, verb radicals and ideophones are the main root morphemesof the language. Each of the members of the different classes of affix mor-pheme, both inflecting and non-inflecting, are entered and illustrated as well,and the descriptive model of analysis is still, to a large extent, my30 Malkiel, 'A typological classification of dictionaries', 5.G. FORTUNE 4fAnalytical Grammar of Shona.3* Since that work did not deal with thestructure of units larger than words, the dictionary is deficient in the in-formation it gives on such things as phrase-inflecting and clause-inflectingmorphemes. Earlier in this review it was stated that the Introduction tothe Dictionary should include a description of the main kinds of derivation.It should also include a description of the main structures larger than theword, namely phrases, inflected phrases and clauses, their structure andtheir potentiality for entering larger constructions as constituents. Thisbackground information is needed to describe and identify more adequatelymany of the affixes with extensive syntactic functions such as the inflectingmorphemes of substantive phrases and verb phrases. It is hoped that asecond edition of the Analytical Grammar of Shona will soon be publishedto provide the extended hierarchical treatment of Shona grammatical con-structions on all levels from the word to the sentence, and, if it provesacceptable, provide the Dictionary with the architectonic dimensions neededto see each item in its own proper structural domain. For lack of these per-spectives the identification of a number of items is faulty or fumbling,for example the description of the very different structures nokuda anddzokono as 'adv conj', namely adverbial conjunctives.The linguistic structure is one architectonic principle which, dulydigested, must order, inform and illustrate the identification and positioningof the morphemes of a language listed in a dictionary. Complementary tothis a dictionary should also identify the items in terms of what Gleasoncalls the structure of content, that most intimate system of inter-relatedmeanings and values which is at the heart of a culture.37 It is with referenceto bodying out this structure that I have suggested that the definitions ofmeaning might become more encyclopaedic and written from within theShona world view.PREVIOUS DICTIONARIES AND VOCABULARIESDictionaries are a cannibalistic species which usually incorporate and digesttheir predecessors. Hannan has acknowledged a number of the earlierdictionaries and vocabularies as a principal source of his first edition, andit will be of interest to conclude this review by listing them according todate of issue and dialect.»e G. Fortune, An Analytical Grammar of Shona (London, Longmans, Green,1955).»* Gleason, An Introduction to Descriptive Linguistics, 3: The speaker compre-hends what he is talking about in terms of an organising structure. This structurecauses him to select certain features for description and determines the ways in whichhe will interrelate them. It also cuts the situation up into portions in a characteristicway. These selected features .. . form patterns which recur, and which are at leastpartially predictable. These recurrent patterns are the structure of content'42 SHONA LEXICOGRAPHYThe first vocabulary incorporating a list of Shona words was made bythe pioneer Bantu comparativist, Bleek, in his The Languages of Mosatn*bique.3e The dialect represented was called Sofala, no doubt the modernShangwe spoken along the coast between the Pungwe and Save rivers. Thefirst vocabulary to deal with the languages of this country was Weale's Mata-bele and Makalaka Vocabulary.39 The Shona dialect represented is Karanga,but the book shows signs of having been hastily put together and inconsist-ently spelt and proof-read. Weale writes with reference to the design of thebook:The idea in giving these two languages in the same book is to givethe reader a chance of studying the one while merely glancing atthe other, it would be a great mistake to attempt to learn the twoat the same time, the Matabele will be found easier to learn thanthe Makalaka and in a great many cases far more useful.40The next to appear was the English-Mashona Dictionary by the RevdA. M. Hartmann, S.J., one of the chaplains of the Pioneer Column.4' Thoughso small in size and consisting of only about 1,000 entries, its illustrativeexamples are sometimes more ambitious than those of the Standard ShonaDictionary and they reflect the events of the time. The spelling adopted, anaspect in which both Weale and Hartmann were pioneers, was devised onthe principle: vowels as in German, consonants as in English. Hartmannfound difficulty in distinguishing between voiced and voiceless sounds andso the results, though quite intelligible, look strange today. For examplethe entry Exterminate v.a. Kill, uraya petsa. All the mice have been exter-minated since the arrival of the white men in our country; we can no longerenjoy them, magonzo osse akapetsgwa gubvira guzwiga gwawarungu munyikamedu; atichagona guadshga. In modern spelling this would read: Makonzoose akapedzwa kubvira kusvika kwavaRungu munyika medu; hatichagonakuadya. Hartmann's Dictionary was preceded by his Outline of a Grammarof the Mashona Language.*3- He called the form of Shona which he describedthe Gomo dialect, no doubt Zezuru as spoken among the Shawasha.In 1897 W. A. Elliot published his Dictionary of the Tebele andShuna Languages.*3 He had lived among the Ndebele for fourteen yearsas a member of the London Missionary Society before producing it, andit is a much more ambitious project than either of its predecessors and someSB W. H. I. Bleek, The Languages of Mosambique (London, Harrison, 1856).*»M. E. Weale (Late D Troop B.S-A.Cos. Police), Matabele and MakalakaVocabulary. Intended for the Use of Prospectors and Farmers in Mashonaland (CapeTown, Murray and St Leger, 1903, 32pp).4olbid., 32«i Published in Cape Town, Juta, 1894, vi, 74pp.42 Published in Cape Town, F. Y. St. Leger, 1893, 69pp.*» Published in London, David Nutt, 1897, xv, 440pp.G. FORTUNE 43of its successors. It was divided into three vocabularies. The first listed Eng-lish terms with 'Tebele' and 'Shuna' equivalents in parallel columns (pp.l-179), about 6,000 entries. The second part consisted of the 'Teb.-Eng.' section(pp. 180-257), and the third the 'Shuna-English' (pp. 258-398). The voca-bularies were preceded by a grammatical outline of the languages andfollowed by a series of phrases divided into sections dealing with Travel,Cattle, Work, Buying and the like, arranged in three columns. Elliot's workis noteworthy in that it was 'an attempt to present a written basis for theShona language as a whole, from which the peculiarities of the differentdialects may be observed'. His examples are drawn from Kalanga andKaranga, and his aim, to prepare a really satisfactory dictionary of thelanguage of the Shona as a whole, was not to be taken up again until thework of Fr Hannan. As an example of the difficulty in making such a workcomplete, he quotes (p.vi) the amusing experience of a Dr Young who,after the publication of his Analytical Concordance to the Bible, involvingthirty years labour, and 'striving in every way to avoid errors', told how acritic pointed out to him that the words 'Holy Ghost' had been omitted.The next vocabularly to appear was compiled by Mrs H. E. Springerof Old Umtali entitled A Hand-Book of Chikaranga** in which pages 1-24form a grammatical outline and the rest are divided into two vocabularies,English-Chikaranga (pp. 25-81), and Chikaranga-English (pp. 82-106).Though Mrs Springer's work was concerned largely with Manyika, shechose to call the language Chikaranga because she was sure it was thesame language as had been spoken in the Makaranga kingdom described bythe Portuguese, and because no term could be discovered among nativespeakers to apply to the language of the whole country. It is interesting tonote that the Revd Fr F. Marconnes was, for similar reasons, to call hisinteresting grammar, published at the request of Doke in 1931, A Grammarof Central Karanga." Mrs Springer tells us that the Revd John White ofSalisbury allowed her to consult the vocabulary he had compiled, and thatfrom it she extracted some 200 words known but seldom used in Manica-land. She looked forward to the publication of a common Bible to be usedby speakers of all dialects as 'there is no more difference between the differ-ent dialects here than between the dialects of the different parts of the UnitedStates'.46 With this in view she kept in mind the suitability of her Shonaentries for Bible translation.In 1906 the first edition of the work which was to be the main referencebook in Zezuru for many years appeared, the English-Chiswina Dictionarywith an Outline Chiswina Grammar by E. Beihler, S. J. It was divided into** Published in Cincinnati, Jennings and Graham, 1905, 106pp.45 Published in Johannesburg, Witwatersrand Univ. Press, 1931.*e Springer, A Hand-Book of Chikaranga, 40.44 SHONA LEXICOGRAPHYa grammatical section with index (56 pages), an English-Chiswina Dictionary(pp. 57-263) and a Chiswina Vocabulary (pp. 1-120), the latter being an Indexto the Dictionary. An interleaved copy in the possession of the present re-viewer states that the English-Chiswina section contains 4,500 words and theChiswina-English 3,425 words. Subsequent and enlarged editions of this workwere published in 1913 and 1927.47 In the third edition the grammar re-ceived more attention and the work was entitled on the cover ChiswinaGrammar with English-Chiswina Dictionary and Chiswina-English Voca-lary. However, the title page retained the wording of the first edition. Doubtsabout the suitability of the name of the language appear in the PreliminaryRemarks where, after giving the names and distribution of local dialects,the compiler adds, 'The Native Department of S. Rhodesia is of opinion thatthe use of the term "Chiswina" should be officially discouraged, and theterm "Chishona" substituted for it.' Accordingly the second part of thedictionary in both the second and the third editions is headed 'Mashonaor Chiswina Vocabulary'. In a fourth edition which appeared in 1950*8the obnoxious term 'Swina' is dropped completely, and the term 'Shona' issubstituted throughout. This edition was printed to provide a work ofreference in the absence of any other at that time. It was printed in theoriginal pre-1931 spelling as there were doubts whether the 1931 NewOrthography would come into general use. Indeed, one of the reasons whythe 1931 orthography did not come into general use may have been thepopularity of 'Beihler' among members of the Native Affairs Department,later the Department of Internal Affairs.After Beihler's first edition the next dictionary to appear was ADictionary with Notes of the Grammar of the Mashona Language commonlycalled Chiswina*9 Pages i-xliv were devoted to a grammatical outline,pp. 45-134 to the English-Chiswina Dictionary, and pp. 135-206 to theChiswiaa-English Dictionary. No author's name is mentioned but it isknown that the book was the work of Fr H. Buck and compiled at StAugustine's Mission, Penhalonga. The language treated is Manyika.*y Published in Roermond, J. J. Romen and Sons, 1906. 263, 120 pp. The secondin Holland, while the third, with additions by the Revd Fr"j. Seed, S.J., was printedin Natal by the Marianhill Mission Press. The publishers of all three editions werethe Jesuit Fathers, Chishawasha.«B The title on the cover reads Beihler's Shona Dictionary. The title page readsA Shona Dictionary with an Outline Shona Grammar Originally Compiled by the RevdB. Beihler, S.J. (Chishawasha, The Jesuit Fathers; and Gape Town, Longmans, Green.1950, xii, 337pp.).*o Published in London, S.P.G.K., 1911, jdiv, 206pp.G. FORTUNE 4SThe first vocabulary of Ndau appeared in 1951 entitled Chindau-Englishand English-Chindau Vocabulary. Published by the Rhodesian Branch of theAmerican Board Mission, it appeared anonymously like Fr Buck'sdictionary. It was, in the main, the work of the Revd G. A. Wilder whohad begun to make a collection of Ndau words as far back as 1888. However,as the Mission among the Ndau employed Zulu almost exclusively for itswork at the beginning, not much attention had been paid to Ndau, andthat is the reason given for the relatively late appearance of the book. Itcontains about 4,000 Ndau entries. Unlike the earlier works the sectionlisting the Shona entries is placed first.The same year Mrs C. S. Louw published her Manual of the ChikarangaLanguage*0 arranged in four parts: Grammar (pp. 3-104); Key to the Exer-cises (pp. 105-28); Useful Conversational Sentences (pp. 129-44); andVocabulary in two sections, English-Chikaranga (pp. 149-290) and Chika-ranga-English (pp. 291-397). The Karanga vocabulary contains over 8,000words. The purpose of the Manual wasto provide for the European student a gradual introduction to thegrammar of this important branch of the group of African Nativelanguages known as the Bantu family . . . The Vocabulary of PartIV does not claim to be either an exhaustive or correct dictionary.Such words have been included as could be collected from thenatives, and meanings assigned them, which, it is hoped, will befound to be generally correct.The compiler appeals for criticisms and additions and asks others to 'co-operate by suggestion and advice towards the issue in the future of a correctand complete Dictionary of the Chikaranga language.51At this point in time, with small dictionaries and vocabularies producedin four of the Shona dialects in which missionary work had been undertaken,new compilations cease. Doke says that a little English-Chikaranga Dictionarywas produced in 1926, being some 50 pages of excerpts from Mrs Louw'slarger book.52 And we have seen that an edition of Beihler's Dictionarywith the addition of some 800 entries appeared in 1927. The perspectiveof all these works, to use Malkiel's term, had been to assist foreignersto speak local forms of Shona. Almost all of them had been compiled bymissionaries, and words had been gathered during the labour of translationof the Bible and other religious and educational material into the differentdialects in which the several missionary societies worked. Each dictionaryhad its own peculiar form of spelling as indeed had each version of theScriptures.»o Published in Bulawayo, Philpott and Collins, 1915.»i Ibid., v-vii.»«C. M. Doke, Bantu : Modern Grammatical, Phonetical and LesdcographicalStudies since 1860 (London, International African Institute, 1945), 98.46 3HO.NA LEXICOGRAPHYThe desire for a common version of the Bible and, by implication, acommon reading and writing system for Shona as a whole, had been voicedat missionary conferences as early as 1906. However, this desire was notto become effective until 1928 when the Southern Rhodesian Governmentdecided to support the use of African languages during the early years ofeducation, and urged the missionary societies to 'put their heads together toget a common language suitable to the needs of the different Mashonatribes, so that larger and less expensive editions of higher class text-booksbe printed'.53The result of the collaboration between the missions and Governmentwas, of course, the introduction of the New Unified Orthography for schoolsin 1932, and of Standard Shona for the country generally in 1955. Thebook which exemplfied the Unified Orthography was the last to appearin this series of dictionaries and vocabularies before the Standard ShonaDictionary of 1959 and was entitled A Vocabulary of the Dialects ofMashonaland in the New Orthography by the Revd Bertram H. Barnes, C.R."Barnes had expressed his idea of the dictionary that would be neededto serve the Shona language in an article for 1928 entitled 'A campaignagainst Babel. The unification of the dialects of Mashonaland'. His words areworth quoting since they express an ideal which was only partly realizedin his own work, but has been more fully realized in the dictionaries com-piled by Fr Hannan and his colleagues.This divergence of vocabulary [between the dialects] then,oeing less serious than has been thought, it seems that the solutionŽay not be the selection of any one dialect to the exclusion of others,out the inclusion of all.i . iW can ** be d°ne? Let us have a dictionary which shall in-ae m the vocabularies in one alphabetical order, printed in thecommon orthography... The various dialectal areas will allit\nA VL Use the dictionary, for their own words will be infoitnJi Ł same time they wiI1 be getting familiar with the wordsŽ"1 °ther dialects. Books for use in school will use the com-ortnography, and will draw their words from the dictionary,m~K>n pool. Here in Manyikaland we shall, of course, preferfind "ntu Words> but we shall not be limited to them and we shallof tfJ* t.words in the dictionary. We shall gradually learn sometives WOt^et,,words' and may find ft useful to- use them as alterna-not reoef £ Up a book Printed for another area, and it willin thpV US at once by *ts unfamiliarity, because it will be printed« common orthography. We shall be able to read it straightW- M. Leggate ,?nerMy Ł £oloi»al Secretary of the Southern Rhodesian Government,J?, G- S. Louw '«T° «anve Education, to the Rhodesia Scientific Association, quoted««°rferia MissionŽ r? the vernacuI^r in missions', Proceedings of the Southern54 Published Z flZ 41 Ł Ł Ł 1928 (Salisbury. "he Conference, 19283), 51.In London. The Sheldon Press, 1932, x, 214pp.G. FORTUNE 47off, very likely reading it with a Manyika accent, but anywayactually reading it. We shall find strange words, but they will notremain strange, for we can look them up in our common dictionaryand find out what they stand for in our own dialect.In a generation or two we shall have advanced perceptiblytowards a common language, not by the road of conquest, but bythe better road of peaceful interpenetration. Is it not true that weEnglish of all parts of the English-speaking world have been broughtto share a common speech very largely by the unifying influence ofthe dictionary? The Bible and Shakespeare have done much, but itmay be argued that Dr. Johnson has done even more.00In the work of both Fr Barnes and Fr Hannan, Shona lexicographyentered a new phase. It was no longer a tool for the foreigner to learn asecond language, or rather, a dialect of a second language. It became arecord of the Shona lexicon for its own sake, and as an aid towards a com-mon literature.56tsNADA (1928), VI, 50.so Other dictionaries with specialist aims have appeared and are listed here tocomplete the record:H. Wild, A Southern Rhodesian Botanical Dictionary of Native and EnglishPlant Names (Salisbury, Government Printer, [1953]), and A Rhodesian BotanicalDictionary of African and English Plant Names, revised and enlarged by H. M. Biegeland S. Mavi (Salisbury, Government Printer. 2nd edit., 1972, xii, 281pp.)L. H. DeWoIf and C. Mazobere, Theological Dictionary (Salisbury, Epworth Theo-logical College, mimeo 1963, 49pp.).I. M. Scovill and S. Nehumayi, Pocket Shona Dictionary (Old Umtali, RhodesiaMission Press, 1965, 102 pp., tone-marked).D. Dale, S.J., A Basic EnglishŠShona Dictionary (Gwelo, Mambo Press inassociation with Rhodesia Literature Bureau, 1975, xii, 212pp., illustrated).