Towards a Multilingual Dictionary of Academic

Vocabulary is an essential component of study, of reading and writing skills. At present, English is the primary language of tertiary education in South Africa, and this puts students from other language backgrounds at a disadvantage. The proposed dictionary will therefore be aimed at second language English users who are entry level students at university. It will be in four languages: English, Afrikaans, Zulu and Sepedi. I.S.P. Nation and J. Coady have compiled an academic core vocabulary which tertiary students in all subject fields need to master in order to be able to encode and decode tertiary level texts, and this has been the basis for the planned dictionary. However, this list needs to be compared to and updated with material from a South African academic corpus. A preliminary comparative study has been conducted on the basis of authentic South African data, and a revised list has been created. The dictionary will not be subject-specific, but will deal with a general sub technical vocabulary. Full definitions will be provided in all four languages, as well as translation equivalents, with English as the pivot language. Where a lexical gap occurs, translation equivalents will be coined with the help of the relevant experts. Thus the dictionary will also participate in much-needed corpus development of the African languages. It should become a valuable reference tool for both teacher and student.


Introduction
In 1997 Peter Titlestad, Head of the Department of English, University of Pretoria, initiated a dictionary project with the aim of producing an explanatory multilingual dictionary of academic words.This dictionary is being produced in collaboration with Adelia Carstens (Department of Afrikaans, University of Pretoria) and Danie Prinsloo (Department of African Languages, University of Pretoria), with two research assistants, the authors of this article.The dictionary will be non-subject-specific, dealing rather with an academic core vocabulary, and could therefore be classified as a hybrid type (neither general nor scientific), which is intended to function as a teaching and tutoring aid for university students at entry level.The languages used will be English, Afrikaans, Sepedi, and Zulu, with English as the pivot language.
Today English is the main language used for tertiary education in South Africa and this can present a major problem to the increasing number of educationally disadvantaged university students for whom English is a second or even a third language.Even first language English students tend to find the jump from the vocabulary required to both read and produce texts at school level to that required at university level, a problem.One can therefore easily understand the difficulties experienced by non-mother-tongue speakers of English when faced with academic discourse for the first time.This dictionary is an attempt to provide such students access to a core academic vocabulary to partially bridge the gap between the language requirements at school and those at university, Essential to this question of accessibility is vocabulary: in a study conducted by Paul Nation and James Coady it was found that of all the factors involved in the skill of reading, "vocabulary was the most important and had the strongest effect" (Nation and Coady 1990: 98).Furthermore, it was pointed out that "sentences are more readable if they contain words that are of high frequency in occurrence and that are shorter rather than longer".Other measures that influence readability are "the degree to which a word calls up other words quickly -association value -and concrete versus abstractness" (Nation and Coady 1990: 98).The average university text does not usually display these features.However, by helping students to understand and learn a core vocabulary of frequently occurring university words, we hope to help them with the difficult task of decoding tertiary level texts, and with the production of their own texts using this crucial vocabulary.
The focus here has been on entry-level (first and even second year) tertiary students for whom English is not the mother tongue.We have chosen to concentrate on the largest language groups for the other languages, namely Zulu, Afrikaans and Sepedi.These are the most significant in the region where the study is focused.
Of the 26 991 students enrolled at the University of Pretoria in 1998,15542 are Afrikaans-speaking, 4 261 are English-speaking and 6741 speak languages other than Afrikaans and English.
Further justification for our use of Sepedi and Zulu is derived from the fact that of the top five languages in South Africa, Sepedi is the largest representative of the Sotho group of languages, while Zulu is the largest of the Nguni languages (see Appendix A).
Modifying Bergenholtz and Tarp's matrix (1995: 21) for the definition of dictionary-user types, we have defined the target users of this dictionary as students with incomplete encyclopaedic knowledge and low second-language proficiency.Encyclopaedic knowledge implies a knowledge of, in our case, general academic discourse.Therefore this dictionary can be classified as a leamer's dictionary.
The target group is specifically South African, and therefore we have to deal with unique situations compared to those that helped British and other dictionary makers compile their leamer's dictionaries.We do not seek to simply add to the number of excellent monolingual leamer's dictionaries such as the Oxford Advanced Learner's Dictionary, the Longman Dictionary of Contemporary English or the Collins COBUILD English Language Dictionary, but instead we seek to confront our South African multilingual situation and produce useful definitions as well as translation equivalents.
The format that the dictionary will take, will be modelled on those of successfulleamer's dictionaries using definitions in very accessible language, clear symbols and a user-friendly referencing system.Layout will be a problem due to space constraints so it is imperative that each entry be clearly indicated, with possible colour-coding to highlight the different languages.
In fact, the macrostructure of the dictionary will be determined by its quadrilingual nature.Multilingual dictionaries run the risk of reaching excessive proportions.We hope to counteract this problem by using English as the language of access.Sepedi, Afrikaans and Zulu will be treated as target languages, but reverse lists will be provided in the back matter to serve as indices.
In this way we will provide a means of access through all the languages and avoid complete unidirectionality.

Macrostructure
Gabriele Stein says that "when dealing with a lexical knowledge base, one would be interested in the concepts with which learners should be familiar if they want to talk about all of those things that would be covered by the concepts underlying the actual English words" (1995: 25).For this reason we would like to include words with clear definitions to facilitate understanding and which will be crucial to respond correctly to examination papers, assignments and soon.
As a starting point for delimiting the macrostructure we have used LS.P.
Nation's Teaching and Learning Vocabulary: A University Word List, compiled at Victoria University, Wellington, New Zealand and consisting of the 811 most frequently used words in academic texts.One of the problems we encountered was with the contents of the list (see Appendix B).The list, going back to 1990, is slightly out-of-date, and certain words are of doubtful validity.
We will be supplementing and updating this list with terms used specifically in the contemporary South African academic context.To accomplish this, a study has been conducted by scanning authentic South African study material to form a computer database.With the assistance of Jeffrey Wagemakers, a visiting lexicology student from the Free University of Amsterdam in the Netherlands, we scanned study guides and examination papers from all the Faculties at the University of Pretoria.A computer program called Omnipage Pro was used, and the words were saved as ASCn files, which were retrieved into Microsoft Word.The resulting corpus consists of 92 000 words.
Obviously this corpus is too small and too limited in breadth and depth.Firstly, one university cannot be regarded as representative of all South African universities, and secondly, examination papers and study guides alone are not sufficiently representative of academic discourse.Published academic literature such as text books and academic journals would have to be used as well.This corpus is an interim solution.We are only experimenting, using material from the University of Pretoria for tentative results.
The advantage of having a computer corpus is that we were able to compile frequency lists.We first extracted all words with a frequency of three and more.The result was our preliminary list which we then compared to the Nation list.The result of this comparative study is shown in Appendix C.
In this list we have usually included words from the Nation list which coincided with words in the list compiled in our study.The frequencies in the last column enable us to determine whether a particular word is possibly a valid addition to our final list.From both lists we aim to remove all words with nonacademic senses or which are subject-specific.Examples are: bubbles and tractor which we removed from the Nation list, and chloride, auction, couplet, and court which we deleted from our list.
Chloride, couplet and court are subject-specific, while auction, bubbles and tractor are of doubtful validity in general academic discourse.
If a word occurs in both lists, this is considered to be fairly strong justification to include it in our final list.However, the absence of a word from either one list or the other will not necessarily be seen as grounds for exclusion.All the same, neither frequency nor the fact that a word occurs in both lists will necessarily be seen as a sufficient reason if there appears to be no other justification for inclusion or exclusion.For example, anthropology and carbon were in both lists with a fairly high frequency.But carbon is subject-specific and anthropology is simply the name of a university subject.
In addition, comparing Our list and the Nation list confirmed the existence of words with specific variant meanings in South African English academic usage.An example is motivate, which in South Africa is used in the sense "to support an argument", as illustrated in the following question taken from a first year Cultural History examination paper: Is genealogical research merely a pleasant pastime or is it of value to the community at large?Motivate your answer.
As students we were unaware that this use of motivate was not standard English until one of our English lecturers pointed it out.The standard English senses of motivate given in the Concise Oxford Dictionary are: 1 supply a motive to; be the motive of. 2 cause (a person) to act in a particular way. 3 stimulate the interest of (a person in an activity) Motivate had a frequency of 36 in the South African corpus, which is high for such a small corpus.Such examples of specifically South African senses will be marked accordingly.However, they are the exception rather than the rule, and we shall generally aim to maintain a certain standard of academic English, which Peter Titlestad believes exists as an important recognisable international form of English.We hope that the end result of this work will be a sound academic core vocabulary, more up-to-date than the Nation list and with added authentic South African material.

Microstructure
Our English definitions are formulated by consulting three successful international advanced leamer's dictionaries: the Oxford Advanced Learner's Dictionary, the Longman Dictionary of Contemporary English and the Collins COBUILD English Language Dictionary.The advantage of modelling definitions on these dictionaries is that they use a restricted defining vocabulary based on research into the needs of second-language English learners.Also apparent from the style of definition was the use of the pronoun you to refer to the third person instead of the impersonal (and these days somewhat formal) one, as well as the occasional use of the headword in definitions.
When defining there is, of course, the problem of "adequacy" in terms of linguistic correctness and "appropriateness" in terms of the needs and skills of the target group.Too much linguistic information might confuse the prospective user.There are, however, certain usages requiring explicit indications as to the typical syntactic and semantic contexts in which they occur, or to which they are restricted.While grammatical specifications are given in dictionaries, they are often in a coded form which even mother-tongue speakers find hard to follow.On the other hand, example sentences are more helpful in illustrating how the headword is used.Such examples of use will therefore be given, along with relevant collocations.This, in addition to the conventional grammatical information, will enable users to produce text as well as to translate it.This makes the dictionary more prod,uctive from a pedagogical point of view.
We are not simply providing translation equivalents as found in the South African Multi-Language Dictionary and Phrasebook, and the Concise Trilingual Dictionary Series.Although ambitious and fraught with pitfalls, we shall provide full definitions across all four languages.The challenge is to produce a quadrilingual dictionary which is comprehensively translatory and explanatory.Another problem exists in cases where one or more of the languages have more than one translation equivalent of a term.If the translation equivalents represent different polysemic distinctions then each of the two equivalents must be given with their own definitions.However, the less commonly used equivalent will, if possible, be given as a synonym, as with the Sepedi definition of contrast (see Appendix D).
In a subject-specific dictionary only one meaning of a term usually needs to be included.However, in general academic language terms are often polysemic and there is also the problem of homonyms.It would appear to be almost impossible to link different meaning distinctions in the different languages with one another.This is quite evident simply by glancing at the length of the definitions.Finer distinctions that exist in one language do not necessarily exist in another.If we look at the example of define there are three senses given for English, while in the Afrikaans and Sepedi definitions only one sense is given.In a bilingual dictionary there are set ways of dealing with this problem, but in a quadrilingual dictionary the situation is more complicated and challenging, and will require creative and innovative solutions.
Another serious hurdle to be confronted when compiling a South African multilingual dictionary is the occasional absence of word-for-word translation equivalents across languages.If a complete lexical gap should occur in one of the target languages (as is anticipated), a surrogate equivalent may be coined as an interim solution.Such a lexical item would be marked to indicate its nonlexicalised status.This borders on the corpus development of the previously marginalised African languages.This would involve close co-operation among a wide range of experts such as the Terminology Division of the National Language Service.Furthermore, it is a controversial and immense task.When more work has been done in this regard, the dictionary will have to be revised.

Conclusion
As Gabriele Stein (1995: 2) points out: "There is still a great and urgent need for developing adequate teaching materials to facilitate vocabulary expansion and differentiation."By developing a multilingual dictionary of academic words we will be paving the way towards helping students and tutors bridge the gap between the actual and the expected vocabulary required from a tertiary level student in a predominantly English and sometimes Afrikaans environment.The dictionary aims to provide students with a general knowledge basis of the "foreign language" -academic English.A good reference guide would lower the learning curve of the student and be a useful aid to lecturers.
This would be a definite starting-point for a task that needs to begin somewhere.There has been a lot of talk about such projects, and very little concrete action.Since this is a ground-breaking dictionary, its production will be an ongoing process.With the opportunities that desk-top publishing provides, we may be able to produce pilot versions to test sections of the dictionary with students before the final product is released.The primary aim is after all to produce a dictionary which will be a relevant and useful tool for its target users: students in a multilingual South Africa.
The last thing we wish to produce is yet another vague document which ignores all the glaring problems and explains words that have no practical relevance.Even if this dictionary in itself may seem limited in scope and may seem to be of little value to, say, a student of chemistry or engineering because it is non-subject-specific, we are aware that much is done at present to create such subject-specific dictionaries.Our goal lies elsewhere.
We aim to create a more general tool that will also advance the development of dictionary material in the African languages, making information more accessible to all.

APpendixB
Teaching and Learning Vocabulary -Towards a University Word List: A-D (Nation 1990: 235-236) The numbers after the words group them according to frequency and range.So the words with 1 following them are in the most frequent group and have the widest range.Reproduced by Sabinet Gateway under licence granted by the Publisher ( dated 2011)  hlalosa; hlatholla V. go ya hlaloso ya selo se se itsego gore motho a tie a kgone go tseba go re se bolela ka eng goba se mabapi Ie eng: Sebodu ke motho wa go tswafa.

A
Reproduced by Sabinet Gateway under licence granted by the Publisher ( dated 2011) http://lexikos.journals.ac.zaReproduced by Sabinet Gateway under licence granted by the Publisher ( dated 2011) http://lexikos.journals.ac.za fapantsha V. go laetsa mokgwa wo e lego go re dilo di ka fapana ka gona mabapi Ie seemo goba sebopego sa tsona; = phapantsho: A ke 0 leke go fapantsha dilo tse pedi tse di latelago : setulo Ie tafola.phapano N. go laetsa mokgwa wo e lego go re dilo di ka fapana ka gona mabapi Ie seemo goba sebopego sa tsona; = pharologanyo: Go no Ie phaphano ye kgolo magareng ga go kitima Ie sepela.DEFINEdefine V. 1. to state precisely the meaning of something (words, terms, phrases etc.): define the biological term.2. to state or describe the nature, extent, properties or essential qualities of something: The powers of a judge are defined by law; It's hard to define exactly how Dickens must have felt.3. [usually passive] to show a line, shape, a feature, an outline etc. clearly: The outline of the man was clearly defined by the light behind him; Boundaries between farms must be clearly defined in order to avoid conflict ... definable ADJ.definieer W. noukeurig omskryf: 'n begrip in die Sielkunde definieer; Definieer die begrip 'ideo logie' in 'n paragraafvan veertig woorde.
. 1. to point out or consider the differences between two or more people or things.2. to show a clear difference when close together or when compared: John's actions contrasted sharply with his promises ... contrasting ADJ.very different in style, colour, etc: contrasting ideas, patterns etc. contrast 2 N. 1. a great difference between two or more people or things which is clear when you compare them. 2. a) the state of one thing being very different to something else b) a difference that is clearly seen when things are compared or put close together; to show that you are referring to a very different situation from the one you have just mentioned: by contrast; in contrast; in contrast to.c) a thing that shows such a difference: The way you sang tonight was quite a contrast to Monday's performance.kontrasteer W. 1. teenoor mekaar afsteek: Die tradisionele kleredrag van die Tsongas kontrasteer sterk met die van die Ndebeles.2. teenoor iets anders stel: Kontrasteer die tydgees van die Barok tydperk met die van die Rokoko.kontras N. groot verskil tussen twee mense of dinge in 'n bepaalde konteks: Die kontras tussen lig en donker is baie groot in 5y sk;lderye. http://lexikos.journals.ac.za