The Dictionnaire français – yipounou / yipounou – français of the Église Évangélique du Sud-Gabon ( CMA ) 1966

The promotion and development of a language involves a wide usage of dictionaries and reference works. The native languages of Gabon are no exception. Although lexicography is still a young discipline in the Gabonese domain, remarkable works have been produced by nonlexicographers. They have therefore left a considerable amount of data available for the development of modern lexicography in Gabon. The different lexicons and dictionaries compiled during this time offer an important treatment of the Gabonese languages. Or, seen from another angle, everything should be done, considering the state of these works, to revise them. For some of the Gabonese languages like Yipunu, Fan and Myene, lexicons, dictionaries, grammars and even encyclopedias already exist. However, several questions remain pending: Why are these dictionaries not used? Is it possible for the non-native speaker to learn the language by means of a dictionary? Is it possible for the native speaker to have quick and easy access to the wanted information? This article therefore have as main aim to give tentative answers to these questions by examining the Dictionnaire français–yipounou/yipounou–français of the CMA.


Introduction
In his survey of the lexicographic work in the Gabonese languages, Mihindou (2001: 7-37) critically discusses some of the existing dictionaries and encyclopedias.He points out the lack of lexicographic principles, the practice of dictionary making in Gabon dating since long before the establishment of metalexicography. 1 The existing lexicographic work would therefore benefit from having a good theoretical basis.
The Dictionnaire français-yipounou/yipounou-français (henceforth DFY/YF) is a translation dictionary, strictly presenting translation equivalents with no further lexicographic treatment.With this dictionary, the compilers wanted to assist the Bapunu people in studying the Bible.Not having any linguistic or lexicographic background, they collaborated with the Bapunu people among whom they were living at that time.The users targeted with the dictionary are all those who believe in the Christian faith.These include a variety of users with different backgrounds.Old, young, and every Mupunu person whose desire it is to understand the Bible is meant to use the dictionary.This specification made by the compilers narrowed the data distribution process to the vocabulary concerning the Christian faith and everything dealing with the church.Knowing that their dictionary would not be of the desired standard, the compilers made provision for suggestions to improve its quality.With the help of these suggestions, they expected to publish a new revised edition of the dictionary.This version was supposed to be more "complete" and more "correct" as they mention in the front matter.However, after almost forty years, no such revised version has been published.
The aim of this article is to characterize the DFY/YF and to consider how, in the light of metalexicographical principles, it can be revised.

2.
The Dictionnaire français-yipounou / yipounou-français In translation dictionaries, the emphasis is often on providing only the translation equivalents of a given language.No further treatment is required although it is possible to find such lexicographic treatment in some translation dictionaries.The DFY/YF is a bidirectional bilingual dictionary arranged alphabetically.As such, it is composed of two main sections that represent the two central lists of the dictionary: the French-Yipunu (F-Y) section and the Yipunu-French (Y-F) section.The dictionary has no back matter and the front matter consists only of some acknowledgements made by the compilers.As the focus of this article will be on macro-and microstructural issues, no further attention will therefore be paid to the content of the front matter.
The fact that the main central list is divided into two sections makes it possible for the dictionary to accommodate two different macrostructures.The dictionary as a whole contains 8 829 entries divided between the central lists.Central list A (F-Y) has 6 541 entries and central list B (Y-F) 2 288 entries.The unequal division between the number of entries in each list is immediately noticeable.The F-Y section contains almost 74% of the entries while the Y-F section contains only 26%.

2.1
The target users The structuring of the macrostructural elements of any dictionary highly depends on the needs of the intended target users of that specific dictionary.The amount of data to be included and the ordering and arrangement are all determined by the needs of a certain user group.It is therefore important for any compiler of a specific dictionary to define the target users.Principles of userfriendliness should determine any effort or attempt of collecting data, compiling a corpus and designing or planning a dictionary.The selection of macrostructural elements and their inclusion in the structure of the dictionary should also be determined by the design plan of the dictionary which should be based on a well-defined target user group (Mabika Mbokou 2002).
According to the missionary compilers, the intended users of the DFY/YF would be persons belonging to the Christian church and interested in acquiring Biblical knowledge.These persons would include two categories of users: young and adult users.If a dictionary is compiled to meet the needs of a wide range of target users, the objective to address any group satisfactorily would be difficult.However, when dealing with a wide range of users, lexicographers tend to turn to polyfunctional dictionaries.This would only have been a partial solution for the compilers of the DFY/YF, because polyfunctional dictionaries cannot meet the needs of very different types of users.And it is the case with this particular dictionary.A child attending a Sunday school class has a completely different need from an adult listening to a sermon.These are two different types of users who each needs different data information.The main object of the compilers of the DFY/YF was to produce a tool to help in performing their evangelical task.Although the DFY/YF forms a good basis for the compilation of further corpora in Yipunu, it was available to any Punu and/or French speaker who wanted to use it.Whether the user was able to access and retrieve the needed information easily was not the main concern.
Translation dictionaries are often the result of a bilingual situation.It is the perceived needs of the users that effects the compilation of a particular dictionary.Thus, the way the languages are treated in any bilingual dictionary should be tested among the users.A quasi-bilingual range of users has used the DFY/ YF, but it is not intended to meet the needs of monolingual French speakers.The lexicographic treatment in this dictionary could have been directed at Yipunu as well as French speakers.In doing so, the compilers could have narrowed the profile of their intended target users, yet, these users still form a very wide range with different needs.
Along with the language issue, another restriction could have been made concerning the age of the users as well as their social status.During 1966 when the dictionary was compiled, not all Bapunu people in the rural areas were able to read or write French, not to mention Yipunu.The dictionary was only accessible to people who could read and write in these languages.This was in contrast to the type of members attending church services.It was the time of the early independence of Gabon when the church was most often attended by old people.When the target users are considered from this perspective, it seems difficult to really say for whom the dictionary was intended.

2.2
The access structure In metalexicography, the access structure of a dictionary is the search route a user follows in an attempt to retrieve the needed information.It is one of the most important structural components of the dictionary.The only access structure that the DFY/YF presents consists of the lemma candidate list.The dictionary was made available in a typed version.Although no bold or italic typescript was used to distinguish between the lemma sign and its translation equivalents, a short dash delimits the one language from the other.The compilers relied on the knowledge of the users to make this distinction.For French mother-tongue users with sound dictionary skills it would have been easy to differentiate between Yipunu and French, and thus to identify the needed lemma.But it would have been more difficult for Yipunu mother-tongue users with no experience of consulting a dictionary.The success of any dictionary, and a bilingual one in particular, highly depends on the ease of the consultation process.If the target users find it easy to retrieve the looked-for information, the dictionary has achieved its main goal.Gouws (2001: 102) explains this as follows: "The rapid and unimpeded access of the user to the relevant data presented in the dictionary has to be regarded as a prerequisite for a successful lexicographic product in a user-driven approach."No integrated texts are found in the front matter of the DFY/YF.Due to this, the access structure of the dictionary is limited.However, the pattern followed throughout the dictionary is applied in a consistent way.It is during the consultation process that the user acquires the skill of retrieving the needed information.He/she is led by the alphabetical ordering and the short dash, and his/her feeling for and knowledge of the language(s).

The ordering and inclusion policy
As has been mentioned above, the DFY/YF is a bilingual dictionary without outer texts as such.The lemmas are arranged alphabetically in both central lists.Nevertheless, some comments concerning the order and the inclusion policy of macro-and microstructural elements should be made.

The macrostructure
To be able to fulfil the needs of its intended target users the macrostructure of any synchronic dictionary should reflect the treated language(s) as they are spoken at the time of its compilation.It should include lexical creations newly introduced into and adopted by the language(s).However, even words concerning general matters of the church and everything associated with the church environment and activities were not fully treated in the DFY/YF.For example, items such as Sunday school, "catechism", or choir "chorale" are not part of the macrostructural element of the central lists.The reason for this is probably that the Bible was the main source for the words included in the dictionary.The following remarks concern the inclusion of lemmas in the macrostructure of the DFY/YF.Firstly, the alphabetical order of the lexical items included as main lemmas is interrupted now and then.In the F-Y section, transitive verbs are often included in their infinitive form.Prepositional phrases are also listed alphabetically along with the preposition that precedes them.
The following extracts from the above examples show the breaks in the alphabetical order: Secondly, not all the letters of the alphabet are treated in the dictionary.For the F-Y section, there are no entries under the letter w, while in the Y-F section, several letters are left untreated.The compilers do not give any explanation in the front matter or anywhere else in the dictionary for this omission.However, this omission can be explained by the fact that the dictionary was directed at the learning of the Bible.The Bible was the first source on which a corpus for this purpose was based.
Thirdly, all the Yipunu verbs are presented in their stem form.The verbal prefix u attached to them is not part of the macrostructural presentation.The advantage of this is that a balance in the number of lemmas under each letter is maintained.If the compilers had grouped the verbs together with their verbal prefixes, half of the dictionary would have been under the letter u.However, this approach could cause a certain confusion for users, because the compilers do not explain the treatment of verbs in the front matter.Within the present lexicographic treatment, they appear like homonyms while they are in fact different words not written or pronounced the same.By following the consistent pattern, a lexicographer or an experienced user would know that the verbs are presented according to the stem form.But the common user has to rely on his/her knowledge of the language to distinguish between the pairs.The lexical items belusa and belusa are presented with no lexicographic indicators.The user can consider them as homographs because they are not pronounced the same, the one meaning "healed", the other "to heal".The main tone for the word meaning "healed" is high, while the one meaning "to heal" has a low tone.The same pattern is followed throughout the entire dictionary.Fourthly, homonyms and/or homographs are given in the central list of the Yipunu section without any lexicographic conventions.Due to the high frequency of tones, there is a great occurrence of homographs in Yipunu.These homographs are treated under the letter to which they belong with no lexicographic indicators, subscripts or numbering.When it comes to homonyms, some of them are presented under a single lemma.The following are examples of homographs where the pronunciation clarifies the difference between them. Example: Fifthly, the choice of compound words listed as macrostructural elements seems to have been determined by the meaning of the headword. Example: laba -voir laba goma -avoir peur laba keri -être triste laba múru -avoir mal à la tête laba ubueji -être heureux laba yisonyi -être embarrassé labana -apparaître, arriver labasana -rencontre quelqu'un la dibandu -c'est pourquoi The word laba means not only "to see" but also "to look".It has been chosen as headword for the following phrases: laba goma avoir peur (to be afraid; literally: to look afraid) laba kari être triste (to be sad; literally: to look sad) laba ubuedji être heureux (to be happy; literally: to look happy) The same pattern can be seen in the following lexical items:

The microstructure
The only microstructural elements found in the DFY/YF are translation equivalents.They are listed in a row without illustrative examples.This pattern is consistently used throughout the dictionary.The different translation equivalents are separated by a comma.Here again, the lexicographic treatment of the translation equivalents is not as effective as it should be.By giving no indications about synonyms and/or polysemous senses, the user is left to guess with which type of information he/she is confronted.It is a hindrance to the access structure, making the consultation process very difficult.Some polysemous senses and meaning extensions are even listed as macrostructural elements.This results in complicating the consultation process.

The reversibility of the dictionary
In bilingual lexicography, the reversibility of a bidirectional dictionary is one of the vital components.It is important that the dictionary should be produced in a well-balanced way.The two sections cannot be equal because of the linguistic differences between the two languages, but the discrepancy should not exceed 60%.But, as has been noted, in the DFY/YF, the F-Y section contains 74% of the lemmas while the other 26% is found in the Y-F section.It can be seen that reversibility was not taken into account during the compilation process or the planning phase.Being a translation dictionary, the DFY/YF should have been expected to give a record in the Y-F section of all the lemmas treated in the F-Y section.In other words, all the lemmas treated in the F-Y section should also have been found in the Y-F section.But the fact is that of the 8 829 lemmas treated in the dictionary only 2 288 are found in the Y-F section, while the F-Y section has 6 541 lemmas.This means that the F-Y section contains more than twice the number of entries given in the Y-F section.
To illustrate this the letter u can be taken as an example.There are only seven lemmas under the letter u in the Y-F section.A look at the F-Y section provides far more than seven translation equivalents starting with u.Only on one page of the F-Y section there is a record of 19 translation equivalents starting with u.If the verbs (u-being the verbal prefix) are calculated, the user is left with five other lexical items starting with the letter u, namely usagama, ureyimisa, uregma, udueji and unanguga.These are almost half the number of lemmas found in the Y-F section under the letter u.More examples of lemmas starting with u can be found throughout the F-Y section.Why these words were omitted is unclear, for no reason is given in the front matter.
As far as the reversibility of the DFY/YF is concerned, another inconsistency can be noticed.If the letter u is again used as an example, it is noticed that the first lemmas in the Y-F section are presented as follows: ubuedji -joie, heureux uganyi -battement, palpitation ulelama -bienséant, decent, proper unangama -haut, élevé, grand Looking at the F-Y section, the user should expect to find the above Yipunu lexical items as part of the translation equivalents.But this is not the case: there is no lemma battement in the F-Y section; the lemma bienséant has a completely different translation equivalent uwelimina which is the same equivalent found at the lemma décent; élevé is not found in the F-Y section while élever is included as main lemma; 2 grand has uneni as translation equivalent, a lexical item not listed in the Y-F section; haut has julu as translation equivalent; and palpitation and propre are not listed as part of the lemmas included in the F-Y section.
From all the examples mentioned, only joie and heureux meet the requirements of reversibility.
heureux -ubueji, uboti joie -ubuedji, nsangu However, uboti is not listed as part of the lemmas under the letter u in the Y-F section.Why did the compilers choose not to include uneni, uwelimina and uboti in the central list of the Y-F section?Are the translation equivalents presented as synonyms?Again the user is left to deduce which form to choose during the consultation process.Mashamaite (2001: 114) mentions that "bi-directional refers to a condition whereby a bilingual dictionary is structured in such a manner that the speakers of both languages may use it for either encoding, decoding or any other purpose, including translating".And from the way the lexicographic treatment of the items is effected, it appears that the DFY/YF was intended to serve both French and Yipunu speakers in translation matters.The encoding part of the consultation process was not meant to be fulfilled.With the wide range of users the compilers have left the dictionary open.It is difficult to provide good and relevant translation information.Regardless of the different sections, the lexicographic treatment of each lemma presents a series of equivalents without any http://lexikos.journals.ac.za further information.The user is left with the choice of deciding which translation would satisfy his/her needs.Neither the French user nor the Yipunu user can make such a choice when it comes to synonyms if there are no illustrative examples to direct the decision.
In bilingual lexicography, it is difficult to find equal synonymy because of the differences between the treated languages, and the cultural environment in which they are spoken.All the synonyms for a particular word in a specific language cannot be used equally.Some synonyms will better suit a particular context while they will be less appropriate in another.If the lemma joie from the F-Y section is taken as an example, the translation equivalents are ubuedji and nsangu.In the Y-F section, joie along with heureux is listed as translation equivalents of ubuedji.How can the user know in which context to employ joie and in which heureux when he/she comes across the word ubuedji?It is impossible.The compilers rely too much on the user's knowledge.In this way the DFY/YF fails in satisfying the principle of user-friendliness because it does not properly assist the user in finding the needed information.When the information is found, the dictionary does not provide assistance in interpreting this information, nor in applying this information in a specific usage context.

3.
A new edition of the DFY/YF

The sources
The 1966 edition of the DFY/YF was compiled relying on informants.The compilers were English-speaking missionaries who learnt both treated languages as foreign languages.This is stated in the single page constituting the front matter. 3To build a corpus, the compilers of the dictionary sought help from the Bapunu people among whom they lived.It can be deduced that those who helped in the compilation process were Bapunu members of the congregation.Thus for a new edition to be compiled, the use of a well-devised corpus would be the proper policy to adopt.It would help the compilers in the choice of the data to include as macro-and microstructural elements.The Bible being the first written source of data, the compilers would have to build a significant and relevant corpus that will have to be tested among a wide range of Bapunu speakers.As a start, the compilers could use the knowledge of the informants who helped in the compilation of the first edition.In the meantime, Bapunu mother-tongue speakers will check the data collected in order to obtain a good frequency list.This list will be used as a tool for the inclusion of macrostructural elements.Furthermore, the inclusion policy based on frequency occurrences in the corpus has the advantage of allowing an easy way to refine the selected amount of data.Elements that meet the frequency requirement will thus be included as macro-and microstructural elements.It would be a challenge for the compilers to extract the necessary data from the corpus, but advice from linguists and lexicographers could be used.

The corpus
The first step for the establishment of a reliable corpus is data collection.A corpus should be representative of the oral repertoire of the lexicon of a language, as well as of the written texts found in that language.The mother-tongue speaker of the treated language(s) must master these repertoires.It should also give a record of the characteristics of the lexical items of that lexicon.As a result, it should not be selective, but objective, for it could serve as basis for the compilation of bigger corpora.The corpus for the compilation of a new edition of the DFY/YF should provide more data than the initial purpose for which it will be built.
In the case of Yipunu, using data from existing linguistic and lexicographic works is necessary.French has a long and well-established lexicographic tradition.No corpus needs to be compiled because corpora already exist, and using existing general French dictionaries would save a considerable amount of time.Thus the compilers will not have to compile two different corpora for the two treated languages.
Although small and specific, with a concentration on Biblical and religious terms, the corpus built for the purpose of the DFY/YF should be relevant and accurate with a scope to expand the range of the contents of the dictionary.It is important that a reliable corpus for the DFY/YF should reflect a true synchronic picture of the Yipunu lexicon.It should also give a fair survey of the lexicological status of the lexical items included as lemmas in the dictionary.On the other hand, the compilation of any corpus does not mean that all the data contained in the corpus has to be included in the dictionary.The typological nature of the specific dictionary, its volume and its content will mostly determine the amount of data to be included.The establishment of a corpus for a new edition of the DFY/YF is no exception.This type of dictionary belongs to the category of LSP dictionaries.As such, it has a limited scope.It will contain all Biblical terms as well as terms related to church activities.Newly entered words should be taken into consideration.
The compilers of the DFY/YF have the benefit of having access to advanced technology in dictionary making.Some of these new kinds of technology are corpus queries.These computer programmes are increasingly used in modern lexicography.They are less time-consuming and some of them are useful in providing frequency counts.They can run several frequency counts of words and phrases.With some programming input, detailed data can be extracted from the corpus.Thus the compilers are able to make correct decisions on inclusion procedures.The use of a corpus query like WordSmith, for example, could be of great help in the compilation process of a version of a dictionary such as the DFY/YF.
Another kind of computational technology that could be helpful in the compilation of a new edition of the DFY/YF is the use of the "ruler", a new lexicographic concept developed by Prinsloo andDe Schryver (2002, 2004) consisting of the creation of an alphabetic scale that limits the amount of work http://lexikos.journals.ac.za done during the data collection and the compilation phase.With this programme, lexicographers working on any dictionary project are able to devise a ruler showing roughly the amount of data each letter of the alphabet of the described language should receive.It is thus possible to know if a particular letter would require more data, and another one less.Whatever the language may be, and however the treated languages may differ, the use of a ruler can set a more realistic balance of letters regarding the right amount of incorporated data.By using such a tool, the compilers of a new version of the DFY/YF could avoid one of the mistakes made concerning the amount of data allocated to the letter u for example.
An additional way to enhance the corpus for the compilation of a new edition of the DFY/YF is by using existing Yipunu dictionaries.Although it should be a restricted corpus, considering existing lexicographic works will help with the inclusion policy in terms of the treatment of microstructural elements.The Grammaire Pounoue et Lexique Pounou (henceforth GPLP) is a good example of an existing reference work that can provide additional data.It is a monodirectional bilingual dictionary of Yipunu spoken in the Nyanga province, mainly in Tchibanga.The data that can be drawn from this dictionary concern the clear grammatical explanation given in the front matter, presenting a section on Yipunu morphology and phonology.The compiler even offers an alphabet of Yipunu, being the basis for the written form of the Yipunu lemmas contained in the central list of the dictionary.
The GPLP was compiled ten years before the old version of the DFY/YF.A look at this dictionary which provides good and relevant data, would have been of great help to the compilers of the DFY/YF.Thus, the lexicographic treatment would not only have been based on the knowledge of informants.Furthermore, to avoid the lack of Yipunu entries under the missing letters as in the first edition, the compilers can refer to research conducted on the language.Although based on the Latin alphabet, some attempts to provide a well-established alphabet for the Gabonese languages were made.The compilers of the new version of the DFY/YF can use one of the existing alphabets.In 1999 the Raponda-Walker Foundation revised Raponda-Walker's alphabet composed in 1932, consisting of 38 letters.It was reduced to 33 letters.This is used by the Foundation to produce schoolbooks (in the Rapidolangue series) in five of the most spoken Gabonese languages.The alphabet 4 used in the Rapidolangue series consists of the following letters and letter combinations: the vowels a, e, e [E], ë [ә], i, o, o [O], u, u [y], and the consonants b, d, dy, dj, f, g, h, j, k, l, m, mb, mp, n, nd, ng, ny or ng [N], p, r, s, t, v, w, z.
In the same year, the Orthography for Gabonese Languages (Orthographie des Langues Gabonaises, OLG) was compiled.The OLG has 35 letters and letter combinations written as follows: the vowels a, e, e t, v, vh [β], w, y, z.The alphabet makes provision for the length of vowels by doubling any of them, if this is necessary in a particular language.Chuwa (1996: 327) states that, when "a lexicographer depends only on informants and on his/her intuition, it leads to a poor choice of entry words".As mentioned in the front matter of the first edition, the DFY/YF is supposed to reflect the Yipunu language ("langue des Bapounou").The variety of Yipunu described in this version does not represent the entire Yipunu lexicon because the Yipunu Bible the compilers used was written in the specific variety of Yipunu spoken in Mouila.For a new edition, the compilers should test the lemma candidate list in other areas where Yipunu is spoken.In doing so, they should check that, where, in the first version, the informants produced a word starting with y, it should, in most cases, be an i.A consideration of earlier lexicographic work could be helpful and all varieties could be taken into account where there are major differences.To confirm that the letter i should have been part of the letters included in the dictionary, modern linguistic research on Yipunu can be referred to: Bonneau 1940, 1947, 1956, Mboumba 1985, and Kwenzi-Mikala 1990, 1998, to name but a few.

3.4
The frame structure of the dictionary As has been said in the previous pages, the frame structure of the actual DFY/YF does not have a proper outer texts section.Its front matter only consists of one page of non-integrated text containing the preface of the dictionary.The frame structure of the dictionary thus only refers to the central lists where the different languages are treated.However, in lexicography, the use of outer texts, that is the front and back matter, is a vital tool for the lexicographer to subscribe to the requirements of user-friendliness.The compilers of the DFY/ YF in many ways fail in fulfilling these principles.In the following paragraphs, a different pattern is proposed that will allow a new edition to enhance the content of the dictionary as well as to facilitate the consultation process.
The DFY/YF being an LSP limited-sized dictionary, there would be no need for the compilers to design an inner structure such as the one found in some bidirectional bilingual dictionaries.Only the front and back matter could in this case fulfil the role of such a section in a new edition of the dictionary.http://lexikos.journals.ac.za

The front matter
An effective way to divide the data in the dictionary would be to incorporate some of the data into outer texts.This approach is effective where it allows the compiler(s) to include a larger amount of data into the dictionary; for example, the use of an expanded mini-grammar in the front matter, where rule formulations and examples of word-formation procedures could help the compiler(s) to incorporate morphological information on the treated language(s).In the case of the DFY/YF, morphological information on Yipunu is needed for a better understanding of the grammar of the language.Such a section could be used to connect information in the outer texts with information in the central list(s) of the dictionary.Thus by means of cross-referencing between the outer texts and the central list, the dictionary actually provides an unambiguous selection of information.This can be illustrated by the treatment of the nominal classes.Because they are important in the system of Yipunu, they should be presented directly after the lemma sign in the Y-F section.Information on nominal classes could be further emphasised in the mini-grammar.In this regard Louw (2004: 53) says: "The consistent implementation of such text-external cross-references will be imperative to the success of the dictionary as an interactive communicative … aid." Furthermore, the access structure of the dictionary could be improved by means of its front matter.Explanations on how to use the dictionary as well as on how to retrieve the needed information should be given within this section.The treatment concerning the verbs could be a good example of the complementary relation existing between the front matter and the central list.The effectiveness of the approach used in the first edition could be enhanced if the compilers could also explain the way simple compound words are used.
The front matter of the dictionary could contain a table of contents helping the user in finding and retrieving the needed information.According to Gouws (2001: 105), such a table of contents can be of great help in the access structure of any dictionary.
The dictionary should also contain a Yipunu pronunciation and grammar outline.The major role of this section should be to explain the method used for the lexicographic treatment in each entry.This should not only concern Yipunu but the compilers should also leave room for French.It may be helpful for Yipunu mother-tongue speakers with less knowledge of French.As the compilers of the first edition of the DFY/YF were non mother-tongue speakers of both treated languages, these issues should be dealt with in the front matter of the dictionary.The front matter should give a record of Yipunu ensuring a better transfer of linguistic information.Thus it should show the difference between the two treated languages.Except for the mini-grammar, the compilers of the DFY/YF should present a detailed alphabetical list of the speech sounds of Yipunu along with their pronunciation.The same should be done for French to show the difference between the two languages.A clear explanation and sample of the way entries are presented within the central list(s) should also be provided in the front matter.

The back matter
The compilers of the DFY/YF should add back matter within the frame structure of their dictionary.Should a new edition be compiled, the back matter could be used to incorporate data on Hebrew culture in Biblical times.Here the names of Biblical characters and Jewish customs and ceremonies could also be included and translated into Yipunu.The information contained in the glossary presented at the back of the 1992 version of the Yipunu Bible could furthermore be given in the back matter.

The macrostructure
Although some adjustments should be made, all the lemmas contained in the first edition of the DFY/YF should be included as macrostructural elements in a new edition.The change will appear in the orthography.The compilers should make use of the Orthography for Gabonese Languages (Orthographie des Langues Gabonaises, OLG) for the written transcription of the lemmas.With the establishment of a well-designed corpus, the new edition should give a fair record of the lexicon of the treated languages, and more particularly of the Yipunu lexicon.Moreover, the missing letters could receive a new lexicographic treatment and the discrepancy between the amounts of data could be reduced.
In the Y-F section, homonyms and homographs should be properly treated allowing the user to make the distinctions between the two categories.One way could be to treat homonyms and homographs similarly with numeral indicators.The fact that homographs have a greater occurrence in Yipunu will not affect the lexicographic treatment.The compilers should only ensure that the user grasps the difference between the two words. Example:
The verb form should be preceded by a hyphen as indication for the user that he/she is dealing with a verb and not a noun.This approach will have the advantage of helping the user in such a way that he/she will not confuse homonyms and what can appear to be homonyms because of the presentation.

The microstructure
As far as the microstructural elements are concerned, the translation equivalents should be given together with illustrative examples.In this way, it could guide the user in the choice of the right word for the right usage context.Therefore, all the synonyms should be treated according to their senses.The use of lexicographic conventions such as bold, italics and numbering should be applied in order to help the user access the needed information.These types of markers should also help the compilers in avoiding the use of any confusing pattern.
To treat a fair amount of the Yipunu lexicon, the compilers should include other varieties as well as the pronunciation and the nominal class of the nouns.The microstructure of the dictionary should also name the parts of speech in order to provide the right lexicological status to the data given in the central lists.Because of its typology, there is no need for definitions in the DFY/YF.

Conclusion
The benefit of already having an existing lexicographic work is that the lexicographer will not have to start from the very beginning.There are great possibilities in extending and improving the existing work by using a proper lexicographic methodology.A dictionary such as the DFY/YF provides a good basis for a theoretically sound work.However, for a dictionary to be used, the lexicographer should combine theory and practice in order to respond to the need of the target users.Although practical work provides good data information, the relevancy of that information is made possible by a theory adjusted to a particular context.By mixing theory and practice, the compilers of a new version of the DFY/YF will give the Bapunu a dictionary that can meet their needs.In the meantime, it can help in promoting, developing and standardizing Yipunu.

2.
Élevé is the past participle of the verb élever which is included in its infinitive form.The motivation behind the choice of one form to the other is unclear. http://lexikos.journals.ac.za ://lexikos.journals.ac.za http