Aller au contenu

Classiques Garnier

Comptes rendus

  • Type de publication : Article de revue
  • Revue : Cahiers de lexicologie
    1996 – 1, n° 68
    . varia
  • Auteurs : Humblé (Philippe), Humbley (John)
  • Pages : 193 à 204
  • Réimpression de l’édition de : 1996
  • Revue : Cahiers de lexicologie
  • Thème CLIL : 3147 -- SCIENCES HUMAINES ET SOCIALES, LETTRES -- Lettres et Sciences du langage -- Linguistique, Sciences du langage
  • EAN : 9782812443206
  • ISBN : 978-2-8124-4320-6
  • ISSN : 2262-0346
  • DOI : 10.15122/isbn.978-2-8124-4320-6.p.0195
  • Éditeur : Classiques Garnier
  • Mise en ligne : 09/11/2012
  • Périodicité : Semestrielle
  • Langue : Français
195
COMPTES RENDUS



COLLINS Cobuild Student's Dictionary Bridge Bilingual Portuguese, Harper Collins Publisbers, London, .1995, 673 p. a new bilingual dictionary concept.

Coeulto est bien connu pour ses excellents dictionnaires monolingues anglais. Le COBU/ !.D English-Portuguese Bridge Bilingual est un dictionnaire bilingue d'un type nouveau qui se distingue des dictionnaires déjà sur le marché en ce qu'il dé,+ïnit en portugais le mot anglais, tout en gardant ce mot en anglais dans  !a définition. Cette façon originale de d~nir un mot pour un public dont la langue maternelle n'est pas l'anglais est bienvenue dans un domaine parfois en mal d'initiatives. On pourrait néanmoins se demander si la manière un peu rigide dont cette innovation a été exécutée prend en compte la spécificité de chaque mot, notamment lorsqu'il s âgit d'une publication pour un public linguistiquement bien délimité. En outre, bien que le Bridge Bilingual soit, dans la tradition des dictionnaires COBUlLD, facile à utiliser, certains secteurs plus traditionnels de la lexicographie semblent résister aux tentatives rénovatrices. C'est le cas de la grammaire, présente ici sous une forme qui ne prend peut-être pas su~samment en compte le public du dictionnaire. Quoi qu'il en soit, le Bridge Bilingual, premier dans une série, est un dictionnaire très utile aux débutants et sa conception, même s'il est critiquable sur certains points, a le mérite de remettre en question nombre d'idées reçues.

With the Bridge Bilingual Portuguese Dictionary Collins has just published a new dictionazy in the COBUILD series. This time COBUILD innovates in the bilingual sector of lezicography and the result is certainly Worth a Gloser look : its mixture of monolingual and bilingual dictionazy techniques makes the Brazilian Portuguese Bridge Bilingual undoubtedly both innovative and stimulating. To my knowledge, this is the fast kind of dictionazy ever to have mized both the source language and the tazget language in the body of the actual definition. The first impression is inevitably a Little disorienting and the longer term question is if these innovative procedures entirely suit the needs.
Cah. Lexicol. 68, 199â1, p. 193-207
196 The Bridge Bilingual is the translation of the monolingual Collins Cobuild Student's Dictionary. Roughly everything was translated into Portuguese, except the headword, of course, and the examples, drawn from the COBUIi.D data bank. Excluding the use of Portuguese, the content of the definition does not differ from that of the original monolingual Student's Dictionary. The headword, when used in the body of the definition, remains in English. For instance :
nesty, naetier, nsstiest. l ADJ Algo que é iresty é muho desayraddvel.
Another example :
log, luge, lugging, lugged, VB com OBJ Se voc~ log um objeto pesado, vocd o
carroya com dificuklade; uso intonnal. She lugged the suitcase out into the haAway.
The Bridge Bilingual counts 40 000 references and is preceded by an introduction in Portuguese which explains the use of the dictionary. Most of the entries comprise examples. A second section of five pages given an explanation of the grammatical teems and abbreviations and there is one page on prononciation.
Beginners
This dictionary is cleazly focused upon beginning English students. The use of the student's native language to explain the entry items is the mont apparent illustration of this interest, but other features were also designed with the saure purpose. Verb forets are listed sepazately —given, gave; threw, thrown —and some entries refer to drawings at the end of the volume —car, home, food, etc.
Attention han also been given to cultural information. GCSE, Thanksgiving and many other items are thon explained. Within the entries, meanings have been classified according to their frequency of use. The fast meaning of embark is the metaphorical one, Peru embarked on a massive programme of reform and not the one related to sailing. Although this soeurs more common senne, still dictionazies, especially bilingual ones, mention only the literal meaning : "to go on to a ship".
In many ways, the Bridge Bilingual is a practical dictionary and concerned with a real beginnei s situation. Nevertheless, some of the procedures of the Bridge Bilingual can be questioned.

Comprehension and production
According to the publishers, the audience of this dictionary is alunos universitdrios e profissionais de vbrias atividades (university students and members of different professions). What kind of use are they encouraged to make of this dictionary ?
197 Nothing is specified, but the layout suggests that it should be useful both for understanding and for production of the larget language.
How would it promote understanding ? In the case of nasty, quoted above, the user will deduce that the word means muito desagradâvel. The minimum objectives of a bilingual dictionary aze in this way fulfilled. The question is, however, if this result has been achieved in the most economic and efféctive way. Would it have made any difference if below the word nasty there was the Portuguese translation of the word — muito desagradâvel — instead of Algo que é nasty é muito desagradâvel ? Not significantly, I suspect. The user will spontaneously reduce the entry to : nasty, muito desagradâvel. One could allege that the use of a full sentence demonstrates the use of the word .This was clearly one of the reasons why full sentences were used in the original monolingual edition. It might make sente in this case, because algo (something) suggests correctly chat nasty means muito desagradâvel when it applies to things, whereas applied to persons it might be something else. A simple indication of the type (toisas) muito desagradâvel would have made this cleaz. Coincidentally, the translation of nasty would be the saine in the case of persons, which would reduce the entry to nasty (pessoas, toisas) muito desagradâvel. In this case, the full sentence defmition was Hot really useful.
There aze, of course, a great many Humber of cases in which the Bridge Bilingual procedure works. This is so for a word like begrudge :
begrudge, se você begrudge someone algo, voté sente que essa Pessoa nâo merece isso e sente inveja dela por té-b.

Here the leazner is enable to forma "mental picturé' of begrudge : it is used with a dùect object which is a thing and an indùect object which is a person. The way in which the information is transmitted exempts the leazner from knowing grammaz. In this case the Bridge Bilingual system works well.
The practice of using full sentences to define an item was a major revolution in dictionary making for which the COBtnLD project has to be credited. Nonetheless, the saine advantages were Hot uniformly maintained when the Student's edition was translated into the Bridge Bilingual. In many items, no syntactical particularity is shown by this somewhat verbose way of displaying information. With space restrictions making many dictionazies difficult to use through recourse to incomprehensible abbreviations, symbols and eye-straining print, editors are always very careful as to what they include and ezclude. In the present case, inclusion of more exemples would surely have been more useful chan the full Portuguese ezplanation of items such as :
A kangaroo é um animal australiano de grande porte que se bcomove saltando carrt as peinas traseiras. As fémeas carregam os filhotes Huma brisa bcalizada na barriga.
198 Which means : A kangaroo is a large Australian animal which moues forward by jumping on its back legs. Female kangaroos cary iheir babies in a special pouch on chair stomachs.
In the case of kangaroo, no new information is being transmitted. The whole entry conforms to what an ordinary monolingual Portuguese dictionazy would say on kangazoos and we could even ask why this word is in the dictionary at all. The Portuguese word canguru is very similaz to kangaroo and people will at the most look the word up to know how it is pronounced. Inclusion of this item can be advocated on the grounds chat it is listed among the most frequent words in English, but the space of 25 words or so could have been with only a translation and a phonetic transcription. In determining which English entries should be put into Portuguese cleazly more questions could be asked about why any pazticular entry should be included and with what characteristics. When the ezplanation is in English, the user has the benefit of seeing how an animal litre a kangaroo can be described in English. Without this advantage, inclusion requires soma other justification.
The use of full Portuguese sentences has other consequences, as is illustrated by an item litre belly : Your belly é sua barriga. In this case the Brazilian authors have had the common sense not to translate "the part of your body, especially your stomach, etc." as the original Student's states. However, why not simply say barriga ? Here, the use of a whole sentence is not only superfluous, it gives the dictionary an almost childish tone ill-adapted to the audience.
When oxygen is ezplained as : Oxygen é oxigênio, um gâs, the Bridge Bilingual's indebtedness to an essentially monolingual reference work becomes flagrant. In a bilingual dictionary no one ezpects to coma across chemical information. Bilingual users may even have the impression chat théy are not, in fact, the intended audience. As, in a certain sense, they aze not or not with suKicient thoroughness. On the one band, the reader is given superfluous information, on the other; no ezamples or idioms show the use of the word in question. The inclusion of an idiom litre To run out of oxygen, for ezample, would have been informative for Brazilians who would look up the word oxygen, not to know if its meaning was oxigênio, but to see how it was pronounced and if one should say, e.g., deprived or dispossessed of oxygen.
When it comas to items which are mach more common in Brazil than in Great Britain or in the United States, this Jack of attention to the specific chazacteristics of the audience risks becoming confusing and even funny. The definition of ox could lead the public into thinking chat an ox cannot possibly be a simple boi : An ox é um touro castrado (an oz is a castrated bull).
The reluctance to translate the headword leads in mort cases to an ezcess of words to the detriment of other information. In the case of simple words it may even lead to unnecessary ambiguity. After reading the entry for bakery : é o local onde se assam pries e bobs (is a place where Bread and cakes aze baked), a student of mine suggested it could
199 be a padaria (bakery) as well as a forno (oven). Everything indicates that the translatons stayed too close to the original :
bald, Alguém que é bald tem pouco ou nenhum cabelo no alto da cabeça. (Originalty : on the top of their head.}

At limes the entry word ezists in Portuguese with exactly the saure meaning as in English yet the lezicographers — or the translatons — did not take advantage of this and even seem to have avoided such words. A word like ennoble could easily be translated as enobreeer, without losing anylhing of its meaning. Instead of this, the Bridge Bilingue !
prefers :
ennoble, To ennoble uma toisa ou Pessoa significa tourd-las mais nobres e d~gnas (To ennoble something or someone means to make something or someone more noble and d~gn'rf'ied.)
Why not say : ennobrecer, pessoas e toisas ? It is a naturel mnemonic aid. Omitting a straightforward translation may not only lead to superfluous ezplanations, it tan obscure the meaning. This is the case of
enjoyable, Algo que é en}oyeble dâ prazer, We had an enjoyable dey. (Something that is enjoyable gives you pleasure.)
According to this definition, enjoyable could be the equivalent of agrad6vel, ameno, alegre, prazenteiro, delicioso, prazeroso, maravilhoso, etc. In this instance, where there is a wealth of possible translations, one of the Portuguese words would not only Tender the meaning of the word adequately, but could also narrow down the range of its register and its possible collocations. Brazilian students would know better in what content to use it and how formai or informai the word is.
Every entry is an entity
All this dces not mean that the ezplanation of the lexical items in Portuguese is useless. The issue is the determination of Chose cases where it helps the learner and in what cases it does not. It is assuredly very useful every lime there is no exact equivalent in the larget language. This is so for words like slot, prig, entrant, ultimately, etc. Here, the meaning tan only be circumscribed, and to do this in English would be complicated for learners at a beginner s stage. The explanation in Portuguese builds up an idea of what the word refers to and allows the learner to understand the word in ils content.
This is the case of keep out.
Keep Out, esta avisando às pessoas que nâo devem entrer num deterrninado luger.
200 In Brazil keep out could be translated in various ways, either by perigo, propriedade privada or proibida a entrada, depending on the content. The ezplanation, instead of the translation, allows the user to understand the concept and adapt it linguistically in accordance with the situation.
Slot, é uma abertura estreita Huma mâquina ou recipiente, por exemplo, uma abertura em que você pbe moedas para fazer torr que uma m~quina tancions (Bloc
is a Harrow opening in a machine or container, for example, the hole in which you put money 6o make a machine work.)
The user tractes an almost visuel idea of what a star is and chan adapts the translation (ranhura, fende or aven slot in the case of computers). In such cases, the Bridge Bilingual deserves to be imitated.
Another important category for which the Bridge Bilingual system works well, is
phrasa ! verbs. The explanation in Portuguese shows the learner if the particle remains together with the verb or Hot and it gives an idea of ils position.
Narre uma coisa ou uma Pessoa after outra coisa ou Pessoa, da à primaire o mesmo nome da segunda. (If you naine something or someone after another thing or person, you give the first one the saine Hama as the second one.)
Another example :
Se voté put up o dinheiro for uma coisa, fornece o dinheiro necessério para pagar por ela Qf you put up money for something, you provids the money that is needed to pay for it)
In both these exemples the Portuguese explanation, with the English word embedded, gives valuable additional information. On the negative Bide, a single example of put up hazdly gives an idea of the meaning if chers is no straightforwazd translation in the larget language. One example is Hot enough to ses for oneself if indeed there is no such translation. On the other hand, it could be argued that ariyone who avants to use a word litre put up, probably has a sufficiently good command of the language to use a monolingue ! English dictionary and look up the exemples chers.
Experiments clone could really test to what extent this method of mixing two languages is successful. One of the things it probably depends on is what languages are involved. English and Portuguese, for example, have reasonabIy comparable syntactical systems. I doubt if a similar Finnish version would be very successful. Tests would reveal if students who use this kind of explanation of phrasa ! verbs are indeed becter at using them chan others who look them up in more traditional dictionaries.
In conclusion, every entry is an entity. Ezplaining an English word in Portuguese tan be very useful in soma cases and merely amusing in others. Each entry should be
201
considered in itself, without an ezaggerated concern for symmetry. The fact that some items aze translated does not imply all should be. A number of English words probably have a Wear exact equivalent in Portuguese. This is certainly true for material things and these words would be more readily assimilable through a straight translation. Butter is manteiga, car is carro. Even many abstract words can be translated, certainly at a beginners' level. Love is amor, Lake is pegar, about is sobre, quase, para ch e para Ih, with a few other possibilities which some well chosen ezamples would help to elucidate. More nuanced understanding will be acquired by the leamer through vazied forms of contact with the larget language.

Comprebension and vocabulary
As is usuel for learner's dictionaries, the vocabulary used by the Bridge Bilingual has been controlled to enable gnon-native speaker to understand the tezt of the entries. But as this dictionary is the translation of the monolingual COBUILD, ezamples are taken from real tezts. which makes it more difficult to check vocabulary. Consequently very simple ezplanations are sometimes followed by ezamples with some quise hard words. The exemple for belly is : lions creeping on their bellies, though the verb creep is probably unknown to someone seeking the meaning of belly. There are other ezamples. The entry for behind contains words like limping, sci~edule and paces. If you do not know what behlnd means, you will almost certainly nos know the meaning of these words.

Production and grammar
The dictionary itself contains semi-grammatical items. Able is such an item.
-able SUFFIX Acrescerita-se -able a alguns verbos para former adjetivos que descrevem algo ou alguém como passlveis de sofrerem uma determinada açâo. Por exemplo, uma toisa que é identifiable pole ser identificada. 7tiey are botte immediately recognisable. (-able is added to verbs to foret adjectives describing someone or something steel can have the lhing dons to them which is described by the verb.)
Whilst this is certainly useful information, since it allows readers to entend their vocabulary by a productive item, the question is whether it is in ils place in the alphabetical listing of a dictionary. It seems very unlikely chat the leazner would look up the item spontaneously, at least nobody who is not familier with the meaning of a hypen (-). Either the entry should give very specialised information, or else draw the readei s attention to itself in some other manner. But, apazt from these considerations, let us suppose a Brazilian student tomes across the -able entry. No doubt tee will have some difficulty in understanding — as I had — adjetivos que descrevem algo ou alguém como
202 passtveis de sofrerem uma determinada açdo. Especially the word passtvel and the expression sofrer uma açao are not self-evident.
As so often happens in lezicography, the entry is included primarily for lezicographers to fulfill their task : the entry is there and the information is correct. IYevertheless, from the user's point of view this kind of information does help not very much. It is superficial for chose able to find it. It is too complicated and abstract for those who could profit from its information.
The entry could have been something litre :
Words ending with the suKx-able can be translated in Portuguese by "que gode ser... ' ; rscognisabls is que gode ser reconhecido ; translatabk, que gode ser traduziob; eatablo, que pole ser comido. Generally, this suffix is translated into Portuguese by ü~e suffix -lveF. reconheclvel, iraduztvel, comestivel.

This definition is some ten words longer [han the original, but it has, I [hink, the advantage of being readily usable.
The grammatical section which precedes the actual dictionary seems rather detailed. The ezplanations aze somewhat abstract and some of the items could have used ezamples. The grammatical part of a dictionary is intended only as a support to the dictionary, and not as a sepazate part which can be read independently of it. I assume that the purpose of grammaz in a dictionary is to help whcever has trouble with the use of a word and avants to find in a grammatical cule the answer to a precise problem.
Let us suppose I have difficulties with the word every. I could be unsure whether every has to be used with a plural or a singulaz. A Brazilian user could become confused because two translations are possible : todo and cada. Every person can be translated by codas as pessoas (all of them) and by cada pessoa (each of them). The dictionary says every is a determiner, I might hope to find some supplementary information in the grammatical introduction of the dictionary :
DET significa determiner. Determiners s~o palavras como a, the, my e every, que sâo usadas no inlcio de um sintagma nominal para indicar a que coisa ou pessoa alguém este se referindo.
(Determiners aze words litre a, the, my and every chat aze used at the beginning of a noun to indicate what thing or person you are refera ing to).

The practical value of chia kind of information is limited It basically ezplains what a determiner is, but this was only a problem the dictionary itself created, and it says nothing about the use of determiners in general. Furthermore, we have to understand what sintagma nominal refera to, which is not obvious. And even given this understanding, the grammatical ezplanation does not help us to resolve the real problem which is « how do I use a determiner such as every ». As determiners are not very numerous, it would have
203 been more helpful and simple if the dictionary had indicated the use of each one by means of an example.
In short, insufficient inquiry seems to have been made to determine how this grammatical section would be used in practice. Instead of conceiving the dictionary's grammaz from the user's point of view, the concern was perhaps too much with linguistic correctness and orthodoxy. However, the majority of people do Hot know what a sintagma nominal is and do Hot Head to in order to speak a Foreign language.
A further example of the use of the grammatical section is the reference to suffix.
SUFFIX Chama-se suffix a um grupo de letras, como •able, -er, e -ly que, aaesoentadas ao final de uma palavra, alteram a sua lasse gramatical ou o seu sentido. Por exemplo, acrescentam-se -ly ou -Hess a adje6vos para formar advérbios ou substantivos cujos significados est~o relacionados a esses adjetivos. Pole-se acrescentar -ish a adjetivos para forrnar outras a~etivos que signiticam que alguma toisa tem aquela qualidade sb até oerto ponto. (A sullix is a latter or group of IetOers âke -able, -er, and -ly which are addea to the end of a word in order to alter its word class or its sense. For example, -ly or -Hess tan be added to acÿeetives to foret adverbe or nouns whose meaning is related to these adjectives. -ish tan be added to adjectives to foret other adjectives that mean that something has that quality up m a certain point)

In this paragraph, the examples of suffixes are all in English, but the explanation of the concept suffix is Hot specifically focused on English nor, on the other hand, specifically concemed with the problems a $razilian public could Bave. The explanation in a Portuguese grammaz would indeed Hot be very different. No reference is made to
Portuguese equivalents, which would make the explanation easier to understand. The problem originales in the Tact that the Student's version was aimed at leazners speaking
any native language, from Malay to Dutch. The explanations h2d to remain as general as possible, with all the consequent disadvantages. But this was Hot the case with the Bridge Bilingual, aimed at an exclusively Brazilian audience. It would have been much cleazer and economical to say that a suffix is a sufixo in Portuguese and give soma examples in English. Once again one has the impression of reading a monolingual, instead of a bilingual dictionary. Furthermore for the middle of the road student alteram sua classe gramatical tan mean very little without an example.
There might be an ideology behind this procedure : students have to learn the theory first, and then apply it to prevent them from learning isolated facts instead of a more economic, abstract principle. This theory, applied to native language teaching, is questionable. It probably works for soma people, but Hot for others. In second language leazning, most things are leazned by specific examples. The majority of foreign language learners are Hot linguists but avant to use the dictionary as a tool. Traditional grammar books will provide the general rules. The grammatical part of a dictionary such as the Bridge Bilingual, which addresses itself to a specific public, should offer contrastive
204 information to make it a cool Chat is more adapted to the user. As it now stands, the grammatical part of the dictionary solves problems only the entties themselves created.
Conclusion
On the whole, the Portuguese Bridge Bilingual is a potentially very beneficial initiative. The bilingual dictionary evidently has its drawbacks for leazning a language in a way chat approzimates as closely as possible to the way the native speaks it. The knowledge one draws from a monolingual dictionary is of a different kind to the one drawn from a bilingual one and it is cleaz chat a monolingual. knowledge approximates itself more to a native knowledge than one chat passes through translation. However, equally evidently the use of monolingual dictionazies is not suited to the beginner and bilingualised initiatives such as the Bridge Bilingual, chat undertake to form an intermediate kind of class, precisely, a bridge, have to be welcomed. Nevertheless, this particulaz dictionary is less revolutionary than it might seem at first sight and less than its potential. Many of the flaws of this dictionary are unfortunately the saure as those of other dictionazies.
It Jacks an accurate description of what the tazget audience is. It has not been clearly determined what the use of this dictionary is for its specific public. Is it to produce a second language or merely to understand it ? If it were comprehension, a translation of the lexical items would be sufficient and space could have been cleazed for more entries. As it is, the Bridge Bilingual has too few entries to cope with most novels. If, on the contrary, the intention was to make a production dictionary, more examples needed to be included for most item, and a number of entries could easily have been eliminated.
In this dictionary too, homogeneity has created its own disadvantages. Especially when the audience is so cleazly defined, every entry should be considered in relation to it. Some words aze evident for a Brazilian leazner, others are hazd. Some entries demand more explanation, others less. A bilingual dictionary should be based on contrastive studies and cazeful tazgeting. The people who compile this kind of dictionary can be assumed to have, as in the case of the Bridge Bilingual they evidently have, the requisite knowledge of both languages. This bilingual knowledge however was nôt sufficiently used. The original Student's COBUILD was undoubtedly made by lexicographers, the Bridge Bilingual was made by translatons. Very gond translatons obviously, but with some feaz of being called tradittori.

Philippe HUMBLÉ
Universidade Federal de Santa Catazina
CNPq reseazcher - Brasflia, Brazil

205 Pierre LERAT, Les langues spécialisées, Collection Linguistique nouvelle, Presses universitaires de France, Paris, 1995, 201 p.
Le français dispose depuis pros de quinze ans d'une description approfondie d'un grand secteur de sa langue de spécialité sous la forme du livre de Rostislav KOCOUREK, La langue française de la technique et de la science, qui a connu une nouvelle édition fortement augmentée en 1991. On pourrait se demander alors ce que la récente étude de Pierre LERAT peut apporter de plus. En réalité, la visée de ces deux études est bien différente. D'une part, la portée de cet ouvrage est plus lazge que celle du livre de KocouREx dans la mesure oh les langues spécialisées comprennent non seulement celles des sciences et des techniques, mais aussi celles du droit et d'autres sciences humaines, et que, d'autre part, le degré de spécialisation est défini plus lazgement  ; il ne s'agit pas, dans le livre de P. LERAT, uniquement de discours entre spécialistes, bien au contraire. Mais la justification principale de cet ouvrage est certainement une différence de genre  :celui de KocouREx est un manuel, une somme qui présente le plus grand nombre possible d'exemples, de citations, de points de vue différents, tandis que celui de P. LERAT relève davantage du genre français de l'essai. L'essai à la française comporte une thèse qui est développée paz azgumentation faisant appel à une économie d'exemples et d'illustrations. La thèse qui est développée dans ces pages est que l'étude des langues de spécialité peut et doit étre subordonnée à une analyse linguistique de tous leurs constituants.
Ce recentrage linguistique scientifique est en effet une nécessité, car, lorsqu'on évoque les langues de spécialité, c'est le plus souvent dans un contexte pratique, notamment celui de l'apprentissage d'une langue étrangère, ot~ il convient de limiter le vocabulaire et les principales structures syntaxiques à présenter aux apprenants. On procède donc depuis des années sur des présupposés dont il s'agit aujourd'hui de vérifier le bien-fondé scientifique. Même si cette étude tombe à point nommé, le défi n'en est pas moins de taille, d'autant plus que P. LERAT se situe pazmi ceux qui minimisent les spécificités des langues de spécialité, qu'il refuse de considérer comme autant de sous- systèmes. Àses yeux, il s'agit de l'« usage d'une langue naturelle pour rendre compte techniquement de connaissances spécialisées » (p. 21). C'est pour cette raison d'ailleurs qu'il récuse la dénomination classique (quoique de date récente) de langue de spécialité, préférant langues spécialisées « pour dire en français l'unicité de l'idiome et la particulazité des univers de connaissances » ; ce terme reconstruit aurait l'avantage de « renvoyer au système linguistique pour l'expression et aux professions pour les savoirs ».
Cette dernière formulation lapidaire a l'avantage de présenter on ne peut plus clairement le programme du livre. C'est donc tout d'abord vers les fondements linguistiques que P. LERAT se tourne, en examinant tour à foui les aspects sémiotiques, graphiques, morphologiques, syntaxiques et sémantiques des langues spécialisées (et la liste n'est pas exhaustive), avant d'aborder dans une seconde paztie plus pratique ses nombreuses applications  :traduction, documentation, normalisation, aménagement linguistique, rédaction technique, ingénierie de la connaissance, lexicographie, terminographie et enseignement.
Cette approche permet à l'auteur de couvrir un vaste champ de réflexion linguistique, sans pour autant viser l'ezhaustivité. C'est ainsi que l'apport de l'analyse du discours aux langues spécialisées n'est guère abordée, et les aspects linguistiques de la
206 vulgarisation scientifique et technique ne sont qu'évoqués dans le chapitre sur la rédaction technique. Ce qui importe ici, c'est la « capitalisation théoriquè en linguistique » (p. 15), l'apport de tout l'acquis linguistique appliqué aux langues spécialisées, qu'il considère comme des "plurisystèmes", à l'instar du modèle proposé pour l'orthographe paz Nina CATACH.
Au centre-des langues spécialisées, il y a la terminologie (« La terminologie est par excellence le matériau distinctif du tette spécialisé », p. 62), et P. LERAT incorpore dans cet essai une présentation de sa conception  : il a peut-être plus que quiconque, du moins dans le monde francophone, contribué à donner à cette discipline un statut de branche de la linguistique, en reformulant les postulats des pères de la discipline selon une démarche inspirée de SAUSSUttE et de ses successeurs. Pour P. LERAT, la terminologie, comme les langues de spécialité, n'est qu én partie l'affaire des linguistes, mais elle est digne de tout leur intérêt. La terminologie est une considération transversale dans ce livre, et chaque aspect de l'analyse linguistique a son pendant terminologique.
La division en deux parties que nous avons cru déceler n'a rien de strict. On re- trouve dans plusieurs chapitres consacrés aux applications des considérations de théorie linguistique. C'est ainsi que l'aménagement linguistique est à évaluer, en amont et en aval des actions, à la lumière de la sociolinguistique. De même, le chapitre sur la traduction met en lumière les zones d'ombre d'une approche purement saussurienne, plongée, voire figée dans l'immanence des langues. Caz ce livre, qui jette les bases théoriques d'une étude des langues spécialisées, ne se contente pas de rester dans la théorie  :chaque hypothèse est confrontée aux complexités de la réalité pratique. Et l'auteur ne manque pas de pratique ; il puise ses exemples dans le travail de son Dictionnaire juridique {réalisé avec J.-L. $OURIOUx), dans son exploitation des banques de terminologie, dans son expérience de réviseur du Trésor de la langue française et d'expert auprès du Rint. C'est ainsi qu'il n'est pas convaincu par les logiciels d'extraction de terminologie, car la forme des termes est trop variable, ni par l'intérêt d'immenses corpus numérisés, car on ne se soustrait pas à l'analyse illocutoire. Les grands développements de l'ouvrage, celui sur la féminisation des noms de fonctions paz exemple, visent essentiellement à démontrer que l'insuccès (du moins partiel) de l'action vient de l'insuffisance de l'analyse linguistique préalable —morphologique en particulier — et de la confusion regrettable entre phonétique et phonographématique.
Ce livre n'est pas sans esprit polémique, mais ceci clarifie les positions.
Quel est donc le message de ce livre tonique  ?D'une part, que les langues de spé- cialité concernent un nombre important d'applications, non seulement la rédaction technique, d'ailleurs trop longtemps négligée par les linguistes, mais aussi l'aménagement linguistique, l'ingénierie de la connaissance, sans pazler des secteurs plus classiques comme la documentation et la lexicographie. Puisque ces domaines relèvent des langues spécialisées, et que les langues spécialisées ne se distinguent pas de la langue, ils relèvent aussi, pour leur partie langagière, de la linguistique, et on nous montre comment une analyse linguistique peut éclaircir les problèmes typiques. Il n'apporte pas toutes les réponses, mais il indique des orientations de recherche. Un livre stimulant.
John HUMBLEY
CNRS-INaLF/CTN, Université Paris-Nord