Translation And Linguistics – Analysis – Eurasia Review – Eurasia Review

The point of objection concerning the relationship between translation and linguistics has its origin in the fact that the translating activity involves aspects that are beyond its purely linguistic aspects. Indeed, its cultural, ideological, social, economic, professional, etc. aspects make it impossible to explain by the language sciences alone. This heterogeneous nature of the activity imposes the conception of a new autonomous scientific branch, which is translation studies. 1

Translation within linguistics

The birth of scientific reflection on language has led to new formulations regarding the activity of translation.(2)  The twentieth century also saw the birth of the discipline of translation studies as an autonomous field of study, whose specific object, translation, will be the center of approaches as diverse as cultural studies, cognitivism or philosophy. In the interdisciplinary dialogue, linguistics claimed to set itself up as a privileged framework for translation studies, but this enriching dialogue encountered many pitfalls due to insufficient frameworks when it came to explaining phenomena that went beyond languages to approach the phenomena of discourse, even culture, the myth of Babel is no longer considered today as a punishment but as an opportunity to help cultural intercomprehension.

Translation is a complex activity, but fundamentally it is always the transfer of a message from one language system to another. (3) This implies that the very basis of a translation must be a contrastive linguistic study of the systems between which such a transfer takes place. However, contemporary research often tends to neglect the proper linguistic aspect of translation and to focus more on the translator’s activity.(4) 

A linguistic approach to translation can help to understand the linguistic difficulties that make any act of translation complex, whether it is a question of differences in the lexicon between languages, the fact that a language is made up of different ways of speaking it (and is not a homogeneous whole), the fact that textual genres play a central role in the act of translation, or the fact that in a text the meaning of a word chosen by an author depends on the meanings of the words that surround it. (5) Henri Adamczewski argues in an article entitled : “La linguistique, instrument du traducteur : les problemes aspecto-temporels en anglais et en français,” published in the translation periodical Palimpsestes (6): 

“When we speak of the translator, we think first of all of the teacher of translation who, in addition to a proper translation of the source text, must explicitly provide the reasons for his choices. It is easy to see how important a successful grammar of operations is for the teacher in question (which does not mean that the professional translator does not need or want to understand what he or she is doing).

The different definitions of translation show that the possibility of establishing interlinguistic correspondences between units of two given languages ​​is real. We can mention, by way of illustration, some of them. Translation, according to W. Koeller (1972 : 69), (7) is 

a transcoding or substitution operation during which the elements A1, A2, A3 … of the linguistic system Ll are replaced by the elements B1, B2, B3 … of the L2 linguistic system. ”. 

Catford (1965 : 20) (8) offers a simpler definition according to which the translation is 

the replacement of textual material in one language (SL) by equivalent textual material in another language (TL).

But the definition of Darbelnet (1977 : 7) seems to us to be the most complete: (9) 

Translation is the operation which consists in passing from one language to another all the elements of meaning of a passage and nothing but these elements, making sure that they retain their relative importance in the target language as well as their tone, and taking into account the differences between the cultures to which correspond respectively the source language and the target language.

The universals of language

Even if we admit that languages are after all different, we also recognize that they share fundamental universals. Universals of language are traits that are found in all languages and can therefore facilitate translation. (10) Mounin (1963) distinguishes five. (11)

Cosmogonic universals

Cosmogonic universals are traits that are common because all men live on the same planet. These are the cold, the heat, the rain, the wind, the earth, the sky, the fauna, the flora, the day, the night, the cycles of vegetation, etc. From these similarities, we have the basic referential meaning which is the same, the frames of reference to the outside world [which] are the same (Mounin 1963 : 197).(12)

Biological universals

Martinet, quoted by Mounin (ibid.)  notes: (13) 

Since all men live on the same planet and have in common to be men with what that implies of physiological and psychological analogies, one can expect to discover a certain parallelism in the evolution of all the idioms.

These include the desire to eat, drink, sleep, breathe, excrete, temperature, sex, etc. All “these universals,” he writes,necessarily provide referential meanings common – however minimal – to all men, to all languages” (p. 202). (14)

Psychological universals

Serrus, quoted by Mounin (1963 : 203), (15) writes : 

“If there are some very general attitudes common to all the languages of the world, they are linked to the mental type of the human species and we must ask psychology for an explanation.” 

This is constant data, substantially the same in languages. They can be found in dreams and various thoughts.

Linguistic universals

These are the traits common to all the languages ​​of the world. All languages ​​have phonemes and / or morphemes, show both the opposition and the interdependence between the signifier and the signified, express a substance by means of a form, etc. To borrow from Mounin (1963 : 41), (16) it seems that all the languages ​​of the world designate beings in the universe by names and pronouns, processes in the universe by verbs, qualities of beings in the universe by adjectives, qualities of processes and qualifications of the qualities themselves in the universe by adverbs; logical relationships of dependence, attribution, time, place, circumstance, coordination, subordination, either between beings or between processes, or between the two, by prepositions and conjunctions or what takes their place in these languages. In linguistic universals, we can speak of semantic universals and syntactic universals. (17)

Semantic universals are semantic categories that are found in all cultures and therefore in all languages of the world. We can mention the example of colors. There are eleven terms to designate the basic colors : black, white, red, green, blue, yellow, brown, purple, pink, orange and gray, listed in order of importance. Certainly not all languages have all eleven colors, but those with two have the first two, those with three (like many African languages) have the first three, and so on.

More generally, Anna Wierzbicka (1988) (18) proposes 61 primitive semantics, that is to say words that would express all ideas in a simple way in all the languages of the world. Some of them are : the substantives (I, you, someone, people, etc.) ; mental predicates (thinking, knowing, wanting, feeling, seeing, hearing); speech (say, word); action, event and movement (doing, arriving, moving); existence and possession (there is, to have); life and death (life, death); time (now, before, after, etc.); space, place (here, there, above, below, far, etc.); logical concepts (maybe, power, because, if); intensifier (very); augmentator (plus); qualifiers (one, two, all, many, etc.); evaluators (good, bad); descriptors (large, small); taxonomy (kind of, part of); similarity (like) and determinatives (this, the same, the other).

The syntactic order of the words (Subject (S), Object (O) and Verb(V)), although it may be different from a language to another, has one constant in most European languages : the order SVO. On the other hand, Japanese has a different order : SOV, as well as most of the languages of the Pacific which are structured in VSO. But linguists, in general, distinguish two sets of basic synatxic order from the position of the subject in the sentence : SVO, VSO, SOV on the one hand and VOS, OVS, OSV on the other. In the first case, the subject precedes the object and in the second, he follows it. However, linguistic studies show that the subject, in most languages in the world, precedes the object in a basic sentence, that is to say a sentence in deep structure. This rule is pretty much universal.

Universals of culture

Aginsky and Aginsky, cited by Mounin (1963 : 214), (19) say that “Certain aspects of cultures, including language, technology, religion, education, power, are found in all cultures ”, without forgetting that many details including“ fire, lever, spear, numbering, incest, taboos, etc.“  (20) are also universal. Thus, the existence of all these universals makes it possible to observe that languages have many references and common denotations, which a priori “allows the passage from any language into any language” (Mounin, 1963 : 222),(21)  hence translation, hence translatability. As such, we are going to conclude that the transition from one language to another is obvious and that all languages are pure and simple nomenclatures.

Translation and linguistics: What is the relationship?

The relationship between linguistics and translation has always provoked lively and interesting debates. (22)  At the center of these debates is an essential question: is translation a branch of linguistics or is it a science in its own right?  (23) It is not our intention to reopen the discussion. However, we would like to point out that even if one considers translation as a science in its own right, it should be recognized that its object of study requires an interdisciplinary approach and that linguistics is its “sister-science” par excellence, since it is the science of language.

In this respect, Maria Tsigou argues quite rightly that: (24)

“To claim that linguistics has nothing to do with translation – a fairly common opinion among translators – is as false as it is misleading. For, during the process of translation, that is, during the transfer of a message from one language to another, there is necessarily a linguistic process going on, and it cannot be otherwise. In the same way, the product of a translation can be the object of study for linguistics in order to draw conclusions about observed language phenomena. Therefore, linguistics and translation are very closely related by nature. Thus, it seems normal to us to accept the principle that one must be a linguist before becoming a translator.”

Although the discussion around the relationship between linguistics and translation studies is not new, Boisseau Maryvonne, Catherine Chauvin, Catherine Delesse and Yvon Keromnes in their work entitled: Linguistique et traductologie: les enjeux d’une relation complexe, (25) elaborate, instead, that if linguistics

“can take advantage of an internal approach to its own system in order to account for its object, the other, translatology, because of the contact of languages and comparison, introduces a cross-linguistic perspective that places it on two sides, both that of questioning the functioning of linguistic systems at the level of speech and that of apprehending a practice whose rules are not, linguistically, always predictable. “

The evolution of translation studies towards “cultural studies” may have masked the permanent presence of linguistics in the debate aimed at defining the contours of this discipline, which was first constituted from linguistics. The diversification of this interdisciplinary field of research has then complicated the issues of their relationship, just as the technological evolutions of the last decades have modified the landscape of linguistics. Taking into account these new parameters, and focusing on various languages (English, French, Italian, German), genres and types of texts, the contributors to the above-mentioned volume identify the epistemological, theoretical, methodological and didactic issues that a renewed practice of interaction between the two disciplines allows. This volume thus offers a diversified perspective on translation as a contact between languages, a perspective also put into perspective by a synthesis of Jacqueline Guillemin-Flescher’s contrastivist approach. (26) It is the various aspects of this complex relationship that this book aims to highlight.

The relations between linguistics and translation lead Roberto Mayoral (2001 : 92), to assert: (27) 

“Practically for any proposition or theoretical model of translation one can find the model (s) of the corresponding linguistic theory.

In addition to the models of Saussurian linguistics and the research of Charles Bally which mark the theses of Vinay and Darbelnet and of Mounin, we can mention other approaches in the Anglo-Saxon field which are present in the translation perspectives. Thus, the analysis of the situational context resulting from the anthropological work of Malinowsky is first used by Firth in linguistics, later by John C. Catford (1965) (28) in translation and will be developed by others such as Hatim and Mason (1990) (29). The semantics of Mel’cuk (1981) (30) are used to explain differences in translation of the same text segment and the variationist current, for its part, has a strong influence in modern theoretical proposals for training translators.

With the constitution of Linguistics as a scientific discipline starting from Ferdinand de Saussure, theorists bent on studying the phenomena of translation in the light of the contributions of “hard” linguistics, an adjective that we use to refer to the linguistics of language, as opposed to its “soft“ counterpart i.e. linguistics of speech. The translation is no longer seen as an art but “as a discipline in which one strives to systematize the process of the translating operation” (Larose, 1989 : 9) (31). The proliferation of translations, international exchanges after World War II encouraged research.

Vinay and Darbelnet describe the connection of translation studies to linguistics. They also look at other things that are important to complement their approaches to translation and that are included in a translation, such as stylistics, rhetoric or psychology. They define the strategies that a translator can use. They speak of situational equivalence when they stress the responsibility of the translator vis-à-vis the context of the ST, but they admit that the very situation of the translation involves a certain freedom of choice: (32) 

“Remember that when translating, the translator brings together two linguistic systems, one of which is expressed and fixed, the other is still potential and adaptable” (Vinay and Darbelent, 1958 : 46).

According to Vinay and Darbelnet, translation is a passage from the source language to the target language which results in “a text that is both correct and idiomatic without the translator having to worry about something other than linguistic easements “ (ibid : 48) (33).  Vinay and Darbelnet list other procedures that the translator can use when a literal translation cannot be done : borrowing, tracing, crossover, modulation, equivalence, and adaptation.

Translation from a linguistic point of view

Most linguistic theories involve several levels of analysis of a text. For example, one can analyze a text from a phonological point of view : the organized system of sounds in a language. We can also analyze it from a morphological point of view : the way we can (or cannot) analyze the words of a language as semantic units. Let’s not forget syntax (the analysis of words organized into sentences), semantics (the analysis of the meaning of words and sentences), pragmatics (the results obtained from sentences), and discourse (the analysis of sentences in the context of whole texts). Translation is often thought of as a transfer between the structure of a source language and the structure of a target language. What is the nature of these structures that are transferred ?  One could answer (and this may seem obvious to many people) that we transfer a structure of meaning. Semantics is therefore the appropriate level of analysis in the context of translation.

However, for Jarjoura Hadrane, linguistics and translation have generally a problematic cohabitation issue: (34) 

But the legitimacy of this mission of theorizing and structuring, which linguistics assumed very early on in the training of translation and interpretation, and which contributed considerably to the birth and development of what everyone today calls translation studies, was questioned from the start (Mounin, 1963 : 10-17). The main argument put forward against this legitimacy consists in recalling the very nature of the translating operation : although it refers to the systems of the two languages in contact, its fundamental objective is to reformulate in the target language a speech act produced in the source language. However, a speech act cannot be reduced to its purely linguistic elements, it refers to multiple cultural, cognitive and contextual dimensions (Lederer, 1994). It cannot therefore constitute an object of study reserved exclusively for linguistics, which, according to Ferdinand de Saussure, would have the principal task of studying language “which is social in its essence and independent of the individual”, and not speech, “an individual part of language”, the study of which remains “secondary” (De Saussure, 1979 : 37). If theorization and explicitation are indispensable to translation and interpretation, they must be carried out outside of linguistics and serve as constituent elements of a translatology recognized as a specific and autonomous discipline.

And thought and language form a complex dialectic. 

Reflective thinking, Vygotsky’s inner language, cannot be defined independently of language since it is its only manifestation and instrument. Language has the distinction of being both natural and cultural. Natural, since human beings are, so to speak, genetically programmed for language. Cultural, in that the development of language skills of a subject is inseparable from learning. The development of language in children depends on the socio-cultural context in which the acquisition takes place. This context determines the linguistic system (the first language) which will be adopted by the subject in order to communicate his thoughts but above all shaping a certain vision of the world (Humboldt) (35), and therefore a way of thinking. As Meschonnic reminds us : “it is thought that is maternal, not language” (Meschonnic, 1999 : 59).(36) 

According to Mounin, translation operations make it possible to observe and to analyze the behavior of languages brought into contact. This original method, offered to study the structures of language allows, according to Mounin [1963 : 4], to verify in particular that: (37) 

“If the systems – phonological, lexical, morphological, syntactic – constituted by languages are indeed systems, that is to say sets so interdependent in all their parts that any modification on a single point [any interference, here] can gradually alter the whole. Or for to check, moreover, if such or such of these systems, or parts of systems, the morphology for example, are impenetrable one from the other language to language.”

We understand by this that, in any given language, each element is part of a coherent whole in the sense that a word is taken as such, because it is determined by a signifier and a signified. This word, when combined with others by means of a syntax specific to that language, contributes to the formulation of a statement. All the words that make up this statement are thus interdependent in order to perpetuate its meaning. 

Nevertheless, Maurice Pergnier believes that the vast majority of linguists, of all schools, even when they make direct forays into the field of translation, are superbly ignorant of the contributions of translational research and systematically fail to address their problems, which in turn can only reinforce translationalists’ prejudice against linguistics. These prejudices – as the few researchers who have a foot on both sides of the river know – can take purely phantasmatic forms, as when linguistics is made responsible for the bad habits contracted in the scholastic practice of translation, bad habits that are detrimental to its practice as well as to its theoretical understanding.

The little space given to translation by linguistics in its theorizations thus contrasts with the hold that many non-linguistic translators wrongly attribute to it. But it is true that the general lack of interest shown by linguistics in the problems of translation studies contributes to the perpetuation of inappropriate approaches to both the practice and the theory of translation, to the detriment of linguistic research itself.

Pergnier goes on to argue, with much strength, that: (38)

“The misunderstanding that continues to mark the relations between linguistics and translation owes much less to the ill will of rival schools than to an epistemological reality that is too little recognized on both sides, namely that the problematic of translation is addressed to a linguistics of speech, while linguistics for its part (with notable exceptions, of course) continues to be essentially a linguistics of language. This difference can be summarized as follows : translation studies reason about texts while linguistic studies reason about sign systems. Most linguists who have not worked on the theoretical and practical problems of translation think that this is a simple difference of perspective, without any impact on the substance of things, since – they implicitly assume – texts are composed of signs, and that what applies to the linguistic segments considered in the language system must apply ipso facto to the sets that integrate these segments into speech. Thus, they have a natural propensity to restrict the study of translation to a comparative or contrastive approach to the languages involved. Translators, for their part, even if they do not repudiate this approach, argue that it has very little place in their practice, the success of which is subject to quite different considerations, having to do with the nature of the texts, the way in which words are inserted into the dynamics of these texts, and where questions of word equivalence are posed much more in relation to conceptual structures than in relation to the structures of languages. It is not translators, but translation itself, which is challenging the models proposed by linguistics of language. The divorce between translation and linguistics only highlights the inability to think jointly about the facts of language and the facts of speech in an integrating theory.”

Linguistics and translation: A relationship marked by mutual ignorance

The relationship between linguistics and translation has long been marked by mutual ignorance, if not haughty exclusion. Until the late 1960s, linguistics was equated with structuralism and generative theory. Some authors have attempted to base their general, methodological reflections in translation on linguistics (Catford 1965, Mounin 1976, Koptjevskaja-Mamm 1989, etc.). A conception of translation as transfer, comparison of structures, independent of any pragmatic, sociolinguistic or discursive dimension, then predominated. At the same time, this somewhat mechanistic view of languages ​​and translations was reinforced by the utopia of machine translation. 

This formal stage was shaken up by a stage that could be called ethno-semantic : on the one hand there was a return of the repressed in linguistics, that is to say of the problematic of meaning and, on the other hand, the apprehension of the cultural aspects of meaning thanks to certain anthropological works (Boas, Malinowski, Sapir, Lévi-Strauss, etc.). Nida illustrates this route quite well : starting from transformational grammar (1964), he comes to component and semic analyzes, gradually integrating social and cultural dimensions (with Taber 1969). Contrastive linguistics itself has undergone evolutions : sometimes heir to a pure and hard tradition, sometimes aliant beyond proposition (Vinay-Darbelnet 1958 ; Guillemin-Flescher 1981), sometimes placing itself within a precise theoretical framework –  see for example the comparative systematics of Garnier (1985) applying the psychomechanics of language due to G. Guillaume.

The abundance of linguistic theories and their long influence on translation can no doubt be explained by the very conception of the latter. First, doesn’t the most common definition that considers translation to be the passage of a message from a source language to a target language imply that the latter is a purely linguistic phenomenon ? In any case, Jakobson’s design leaves no room for doubt :

(1) Intralingual translation, or rewording is an interpretation of verbal signs by means of other signs of the same language.

(2) Interlingual translation or translation proper is an interpretation of verbal signs by means of some other language.

(3) Intersemiotic translation or transmutation is an interpretation of verbal signs by means of signs of nonverbal sign systems (Jakobson 1987 : 429).(39) 

Interlingual translation, which interests us, is defined here by Jakobson as the interpretation of source linguistic signs by other target linguistic signs.

The object of linguistics is the study of the knowledge that speaking subjects have of the language. At this level Baylon & Fabre (1999 : 17)  (40) distinguish two opposing conceptions of linguistics: 

  • Linguistics as a description of languages which considers a language as a system of linguistic signs ; and 
  • Linguistics as a study of the functioning of language as a system of rules. 

According to Chomsky’s linguistic theory, there are general features common to all languages that generative grammar should endeavor to make explicit.

According to Chomsky (Chomsky, 1969 : 96), (41) the study of the universal conditions that prescribe the form of everything that is human language constitutes “general grammar”. These universal conditions, we do not learn them ; rather they provide the principles of organization that allow you to learn a language, and that must exist in order to move from data to knowledge.

As for sociolinguistics, which can be considered as a branch of linguistics, it is interested in the relationships between society and language. It studies, among other things, linguistic variation as a manifestation of belonging to a social class, group, etc. The organization of the message has a social implication that linguistic analysis can elucidate. For sociolinguistics, the understanding of a statement goes beyond the linguistic framework and encompasses social factors.

The many theories of translation based on linguistics and / or sociolinguistics are not sufficient to analyze the relationship between language and culture, because most of these approaches revolve around the concept of equivalence, the content of which varies from one approach to another. Hence the need for approaches that encompass not only linguistic factors, but also cultural factors.

For Catford, translation is an operation between languages, that is, a process of substituting a text in one language for another text in another language (1965 : 1). (42) This conception of translation leads Catford to posit equivalence as being at the center of the practice and theory of translation :

“A central problem of translation-practice is that of finding TL [target language] translation equivalents. A central task of translation theory is that of defining the nature and conditions of translation equivalence. “ (Catford 1965 : 21). (43)

Catford distinguishes two types of equivalence : textual equivalence and formal correspondence. Textual equivalence is any form of target text which, on observation, can be said to be the equivalent of a form of source text (1965 : 27), (44) while there is formal correspondence when the different categories of the target language occupy the same place as those of the source language.

Catford also distinguishes the reduced translation (“restricted translation ”), as opposed to“ total translation ”, defined as“ replacement of SL textual material by equivalent TL textual material, at one level ” (1965 : 22). (45)  This notion of reduced translation designates equivalence at the phonological, graphological, grammatical or lexical levels. This type of translation is of very little interest for translation which, as the therocians will agree thereafter, relates in general to texts.

According to Catford, translation may be impossible, and he distinguishes two situations : linguistic untranslatable and cultural untranslatable. Linguistic untranslability results from the lack of equivalents in the target language, and cultural untranslability refers to the absence of cultural elements of the source language in the culture of the target language. After analysis, Catford brings back cultural untranslatable to linguistic untranslatable, because he says:

“to talk of “cultural untranslatability“ may be just another way of talking about colloquial untranslatability : the impossibility of finding an equivalent collocation in the TL. And this would be a type of linguistic untranslatability. “ (Catford 1965 : 101). (46)

Of all the linguistic theories of translation, Catford’s has been the least successful, because it focuses too much on the linguistic system rather than its use. Despite the distinction between formal correspondence and textual equivalence that Catford establishes, he fails to perceive that this difference arises from the close connection between language and culture, and that, therefore, translation cannot be reduced to a transfer purely linguistic. The “translation shifts” that Catford sees are a description of the results of the process, rather than a theorization that can be used in the translating activity. Catford’s approach represents theories with a linguistic conception and mechanistic approach to translation which not only does not correspond to the practice, but very often leads to the impossibility of translation between two languages.

Contrastive linguistics and translation studies approach

Today, when we talk about translation we are talking about contrastive linguistics and translation studies. These are two very close areas and we will try to clarify the differences and similarities between the two.

Contrastive linguistics consists in opposing two different linguistic systems in order to be able to find the interferences. It is to make a rigorous and systematic comparison of two languages ​​and especially of their structural differences. Traditionally, contrastive linguistics has aimed to bring out the structural similarities and divergences between two linguistic systems and to look at all possible equivalents.

For translation studies approach each analysis is unique and translation analysis is an end in itself and not a means. It is, mainly, interested in how units are translated.

Contrastive linguistics seeks to explain linguistic phenomena by making use of contrasts between languages, whereas translation studies approach focuses its activity on the process of translation. Gallagher believes that there is an interdependent relationship between contrastive linguistics and translation studies approach. He writes :

“Indeed, just as translation studies are an auxiliary discipline of contrastive linguistics, so contrastive linguistics is an auxiliary discipline of translation studies. “ (Gallagher, 2003 : 58).(47) 

Eriksson notes that translation analysis is not a method but the goal of the work of the translation specialist. It is the role of translation specialists to see how one can solve specific translation problems, such as dialects, images, idioms, etc. in a specific text, (Eriksson, 2000 : 15). (48)

The dissociation between the translational point of view and the linguistic point of view can be found in the same theorist. We will only take as proof the case of the most illustrious French author in the field, namely Georges Mounin. We know of two equally remarkable books by him, which contradictorily reflect two sides of the author himself, as well as two divergent views on translation. In Les belles infidèles (a work that is not unknown but too little known), Georges Mounin is the poet, the fine literator and the translator; in Les problèmes théoriques de la traduction (written a decade later), he is the functionalist linguist. The result is two radically different approaches to translation that, curiously, do not seem to meet anywhere. One could say, by reducing these works to categories that did not exist when they were written, and in which they are precursors, that in the first one, Mounin works as a translator, while in the second one he works as a contrastivist linguist. The theoretical problems of translation have had a more glorious fate than Les belles infidèles.

Linguistics and translation issues

The practice of translation is not a simple task. (49) It requires a lot of knowledge and exceptional skills, which include daily learning and professional competence, in order to avoid empirical problems that interpose misunderstanding of the translated message.  Although translation is known as an ancient profession, it is a vital part of our human society where no one can be independent of another. The difficulties of translation practice have two slopes, namely: the textual slope and the translator’s slope. We can also add the failure of technology in the amateur translator.  That is to say, the Gordian knots of translation that come from the technicality of the language and the incompetence of the translator.   

One of the first to take an interest in the problem of translation from the perspective of textual linguistics was de Beaugrande (1978). Early on, he saw the extreme importance of the role played by the notion of textual linguistics, both conceptually and methodologically, in the study of translation strategies and techniques. In his description of the three types of strategies guiding the translation process, de Beaugrande explains that they respond to the constraints imposed by the text. However, the author is quick to point out that, logically, the strategies activated by the context will not solve all the problems. In his opinion, it is inappropriate to demand that a translation theory provide solutions to all translation problems. On the other hand, it should be able to offer the principles and strategies necessary to tackle them (de Beaugrande 1978 : 14). (50)

According to J.P Vinay and J. Darbelnet, there are three linguistic problems that false friends (faux amis) cause in English-French and French-English translation.  They are semantic, stylistic and structural (pp.70-72, 170): (51) 

The semantic aspect : The meaning of words varies from one language to another.   The trap of translation is that the target language and the source language share some words of the same spelling. 

The stylistic aspect : As far as the stylistic aspect is concerned, the wrong associations/“false friends“ (faux amis) appear in the order of evoking or referring to a different environment in the target text because the style varies from one language to another in intellectual, psychological, literary, technical, scientific, commercial and specialized value.  But, stylistic misnomers can possess almost the same meaning. 

Structural/phraseological/syntactic aspect : The wrong associations/“false friends“ (faux amis) show themselves in the global meaning which is divergent to the meaning of the sentence structure.  The structural aspect of false friends unites the lexicality and syntax of a sentence which make the phraseological structure.  

Whether working on a translation of a technical document or a sworn translation, there are five types of translation problems : lexical-semantic, grammatical, syntactic, rhetorical, pragmatic and cultural. And that’s not counting the administrative, computer and nervous problems… (52)

1. Lexical-semantic problems

Lexical-semantic problems are those that can be solved by consulting dictionaries, glossaries, terminology banks and experts. This is the case for terminological alternations, neologisms, semantic gaps, contextual synonymy and antonymy (which concern polysemous units : synonymy/antonymy concerns only one meaning and it is the context that allows one to know which meaning is to be taken into consideration), semantic continguity (a process of coherence that functions thanks to the recall of common semantic features between two or more terms) and lexical networks.

2. Grammatical problems

Grammatical problems concern for example questions of temporality, aspect (aspect indicates the way in which the process or state expressed by the verb is considered from the point of view of its development – as opposed to time), pronouns, and whether or not the subject pronoun is explicit.

3. Syntactic problems

Syntactic problems can arise from syntactic parallelisms, rection, passive voice, focalization (point of view according to which a narrative is organized) or rhetorical figures of construction such as hyperbate (inversion of the natural order of discourse) and anaphora (repetition of the same word or segment at the head of a verse or sentence).

4. Rhetorical problems

Rhetorical problems are related to the identification and recreation of figures of speech (simile, metaphor, metonimic, sinecdoche, oxymoron, paradox, etc.) and diction.

5. Pragmatic problems: the example of marketing translation

Pragmatic problems include differences in the use of “you” and “you”, idiomatic phrases, locutions, sayings, irony, humor and sarcasm. But these difficulties can also include other challenges, such as, in an English-French marketing translation, the translation of the personal pronoun “you” : the translator may have to struggle to define whether to use the informal pronoun “tu” or formal “vous” – not always an obvious decision.

Corpus linguistics and translation

Methodologies and, in general, the achievements of modern applied linguistics, seem to play an increasingly important role in the training of trainee translators. This is one of the consequences of the linguistic reorientation that translation studies have undergone in the last few years, which is undoubtedly related to its emancipation from philological translation.

In fact, we can see that specialized translation is taking over the place it already occupied in the academic research field and that, at the same time, researchers are turning more and more to the methodologies and tools developed by corpus linguistics, specialized lexicology, etc. We are thus developing an autonomous theoretical work space, which is not reduced to linguistics, however applied it may be, but which is based, without complexes, on it. (53)

One of the conditions for the creation of a methodology that can be transmitted to trainee translators seems to us to be the presentation of a coherent (and simple) combination of notions “inherited” from general or applied linguistics, and the needs and processes typical of specialized translation proper : it is therefore essential to define and fix a certain number of fundamental linguistic notions that the teacher will be led to use later on.

In recent decades, corpus linguistics has revolutionized the language sciences by focusing on the study of observable facts on large sets of texts. It finds applications in many sub-disciplines of language sciences, such as language teaching or the study of the language of second language learners. (54)

Corpus linguistics began to be considered as a new discipline at the end of the 80s of the last century and the beginning of the 90s, in particular thanks to the definitive work published by John Sinclair in 1991 (55). One can think that, in the fifties and sixty, the critique of Chomsky (1957 : 159) (56), which rejects the use of the corpus by asserting the following, was conceived without too much difficulty:

“Any natural corpus will be skewed. Some sentences won’t occur because they are obvious […], false, […] impolite. […] the description […] would be no more than a mere list. “

This criticism is no longer justified today for several reasons. Chomsky needed to distinguish himself from structuralists and behaviorists, and we can therefore consider that this criticism was more a posture aiming at marking his difference from a linguistics that he considered to be only a collection of butterflies. Moreover, the size of the paper corpora at first, and then of the first electronic corpora, did not allow them to be truly representative of a language. It is therefore understandable that at the time, relying on the intuition of the native speaker-listener seemed to be a reasonable solution. 

For almost fifteen years, corpus linguists have been interested in translation in various ways. At the same time as the question of the usefulness of corpora in translation was being raised, a whole school of corpus linguists began to develop in the 1990s, notably with the CULT (Corpus Use and Learning to Translate) and TaLC (Teaching and Language Corpora) conferences. In his founding article, Aston (1999 : 289)  (57) demonstrates the role that corpora can play in improving the quality and speed of the translation process. He immediately addresses the issue of translator training by showing how corpora can enable future translators to develop their interpretative skills and translation strategies, while improving their sensitivity to the problems posed by translation. All research in this field has subsequently developed in this direction. Numerous researchers have demonstrated the usefulness of corpora in translation, particularly in solving questions of terminology and phraseology, but also questions of style according to genres and fields of specialization, or in evaluating the quality of translations (see, among others, Zanettin 1998 (58), Varantola 2000, Bowker 2001 (59), Maia 2002 (60), Bowker & Pearson 2002 (61), Zanettin et al. 2003 (62), Beeby et al. 2009 (63), Kübler, & Aston 2010 (64);  Kübler 2011 (65) ).

On the side of translators, there have been attempts to describe the language of translated texts as a third code that could be differentiated from the original language using the methodology of corpus linguistics (Baker 1999 (66), Olohan 2004 (67), and Mauranen 2007 (68) ). However, these assumptions are not always validated and do not always take into account criteria of very different orders, such as text genres, fields of specialization, but also the training of translators and the translation tools they use.

Today, linguists from contrastive analysis are increasingly interested in using corpora to validate their observations on translation choices. However, these observations are often made on literary parallel corpora, which is not completely adapted to the problem of pragmatic translation.

We can therefore say that not only is linguistics questioning and trying to provide answers to what translation is, but also, and above all, that corpus linguistics represents an epistemological and methodological turning point in the description and understanding of what translation is.

Applied linguistics and machine translation

Most applied sciences are born of the needs of the society in which they develop. Without doubt, there are few examples of a pure science leading to the birth of an applied science, without the impulse of a social need. It is the need which, in our time, allows to release the financial means essential to a research which is a little complex and thus voracious of time and talent. Such is the case of applied linguistics. Neither the scientific curiosity nor the ingenuity of phoneticians have done more for the study of certain laws of language susceptible of practical applications, than, for example, the research of the best conditions of telephone transmissions; the studies of the engineers of telecommunications on the transmission of signals led to the theory of information, where linguistics finds new ideas and new means for the disinterested and even philosophical study of language. These are only the beginnings.

It appears today more and more clearly that the application of automatism to the mechanisms of thought leads us towards a revolution of the means of communication and conservation of the explored and conscious contents of the human mind, comparable to the revolution of the printing press in the sixteenth century. (69) For all those whose professional and intellectual lives revolve around the written or spoken word, or whose communications with their contemporaries or with the future depend on the written word, this twentieth-century revolution calls for a re-examination of the methods and purposes of their communications. For those who are interested in the study of language to varying degrees and for different purposes, a new constellation of disciplines is taking shape on the borders of cybernetics and new linguistics : its content and boundaries still vary according to country and school, but in Moscow as in Washington it is given the same name, that of applied linguistics.

Once the machine is able to translate sentences from one language into another, and has sufficient memory to do so on a massive scale, it will also be able to retrieve and classify concepts, i.e. to retrieve information without using artificial and more or less arbitrary codings, as is currently being sought everywhere for documentation purposes. The translation machine will undoubtedly provide a fruitful means of unpacking and classifying knowledge, based on the words in the dictionary. In translation, whether automatic or human, the word or semantic unit evokes the idea, which in turn evokes the word of the output language. This mechanism presupposes a well-organized memory, which can be used to classify the titles of a library, as well as the analysis of the works listed in a bibliography. The universe of well-classified discourse will thus serve as a guide to the reader through the shelves of a library. Translation leads us to complete solutions to the problem of information retrieval.

One should not be overly influenced at the outset by considerations about the type of machine that will eventually be used to translate. Today there are many electronic calculators with different characteristics. The fact that the machine to be used for translation will probably be a machine with a highly developed memory (with a large capacity and ultra-fast access) and a relatively less powerful logic device, will not prevent us from making use, for the time being, of existing machines, designed to solve administrative or scientific problems of a completely different nature than those of translation. It will be especially important to make sure that the machines employed are part of sets adaptable to various purposes and whose parts can be combined at will to answer these purposes – concern which coincides besides with that of the large manufacturers anxious to extend the applications of their material. Above all, it will be necessary to use equipment that is widely and easily available, so as to be able to develop as much as possible according to an overall plan the preliminary works, those in which the close collaboration of man and machine requires the use of machines that can wait for their slower partner.

For machines, there is no need to take a premature stand between various theories or draft theories of language. (70) It is very tempting to think that a general explanation of the facts of language could greatly accelerate research for the automation of translation and information retrieval. However, language reflects the universe of representations, concrete or abstract, particular or general ; and the information communicated by it varies in particularity or generality according to the intentions of the speaker. This is why translation, which is a transposition of information – just as much as the search for information, which is based on the detailed or analytical recording of the facts sought – cannot, without loss of information, deviate much from the degree of particularity intended by the author of a text. It is therefore very difficult to conceive how a general theory that is anything other than a general taxonomy, a methodical inventory, will ever allow a significant simplification of a problem at the base of which we always find the word, that is to say, either the grammatical tool, or the signifier establishing a relation between a spirit and a signified.

One of the reasons why translation has become topical again is that it lies at the very heart of this knot of problems, which have never yet been completely separated by a definitive analysis. The research undertaken in order to automate the translation of scientific and technical texts allows, and requires, the isolation of the various elements of the transmitted message in the discourse – phonic, graphic, lexical-semiotic, aesthetic elements, etc. They require, at least at the beginning, that we study graphemes independently of phonemes, and that we take written language as a basis – even if this effort must underline the inconsistencies of the alphabetical transcriptions of certain languages, written for too long for the simplicity of their orthography. Then, beyond these distinctions made at the level of the constituent signals of the message, there are others which become necessary and possible: the work in view of automatic translation has set in motion analyses and experiments allowing the isolation of the cognitive elements of the discourse from the stylistic or aesthetic elements; one will thus have to distinguish between the signals, the cognitive meaning, and those additional aspects of the meaning which are of the nature of aesthetic suggestion rather than of information. The methods of analytical experimentation required by the automation of language bring to linguistics means of scientific investigation allowing a more rigorous separation between information and poetry, in the analysis of written messages. In the same way they involve a rigorous application of the logical analysis for the determination of the authentic information content of a message.

The great Gordian knot for some translators is the trust they place in technology. However, technologies cannot reason like human beings.   For example, a contextual translation where a word is polysemic drives the machine crazy because it will be unable to choose the right word. Therefore, the practice of translation is a very intellectual work that surpasses automation because of the technicality of the language and the need for professional competence. Human translators are the best translators, not the technologies that some translation amateurs trust. 

With the evolution of systems, the realization of an automatic draft appears less and less absurd and familiarization with the technique of post-editing should be part of any translation course, even on an ancillary basis, since it accentuates the critical distance from the target text and exercises the acuity of the reviser.

Conclusion

It’s an open secret : one can be an excellent translator without having ever studied linguistics, just as one can be a great writer without having ever learned grammar and, above all, without having ever read a single book on stylistics or having ever attended courses or columns of professors and critics who pontificate year round on the art and the manner of producing masterpieces.

The different definitions of the translation show that the possibility of establishing interlinguistic correspondences between units of two given languages ​​is real. We can mention, yet another time for the sake of emphasis, by way of illustration, some of them. 

Translation, according to W. Koeller (1972 : 69), (71)  is “a transcoding or substitution operation during which the elements A1, A2, A3 … of the linguistic system Ll are replaced by the elements B1, B2, B3 … of the L2 linguistic system. ”

Catford (1965 : 20) (72) proposes a simpler definition according to which the translation is “the replacement of textual material in one language (SL) by equivalent textual material in another language (TL). “

But the definition of Darbelnet (1977 : 7)  (73) seems to us to be the most complete:

“Translation is the operation which consists in passing from one language into another all the elements of meaning of a passage and only these elements while ensuring that in the target language they retain their relative importance as well as their tone, and taking into account the differences between the cultures to which the source language and the target language correspond respectively.”

We note, in passing, that all the definitions we have mentioned are linguistic definitions of translation. Linguists have developed a theoretical framework for translation, which should, according to them, be part of linguistics to be more practical, more reflective, in a word more scientific.

Reconciling linguistics and literature, even if, in terms of translation, each tends to pull the blanket to its side : “translation is a specialized function of literature” claimed Octavio Paz; (74) “the translating activity is a linguist’s business” already claimed Roman Jakobson. (75) In addition, the linguistic approach to translation usually adopted by translator-essayists (computational linguistics in the service of simultaneous translation, the so-called “push-button” translation) as stated by Mounin: 

The scourge of Esperanto and Volapuck no longer haunts us, but the translation machine is watching us, which will translate faster and fairer than us, say the prophets of doom – and here comes the push-button translation.“  (76)

This state of things does not, however, work for reconciliation and denotes a partial view of the discipline. 

It is true that the theories of translation were opposed to a certain conception of linguistics, but the evolution of the two disciplines – from a normative and prescriptive point of view to a more descriptive and explanatory perspective – can provide a theoretical reflection on facts of language and facts of speech which are undoubtedly inseparable and united.

There are essential skills in the uses of the language : translating, writing, making people understand and discovering, the mastery of which does not seem to depend directly on the knowledge that is produced about them. But we also know that translating, writing, making people understand and making people discover are nonetheless the product of hard work, of discipline in every sense of the word. Now this work of the enunciator, whether he is a translator, writer or mediator of knowledge, this constant discipline that he imposes on himself to signify and not just repeat the stereotypes of his time, this work of the he enunciator is also the work of the language which is at the heart of any linguistics worthy of the name. Efficient and useful linguistics must reproduce this language work. Not blindly ; it would then only redouble it, clumsily paraphrase it the way of a schoolboy without imagination, but explaining it. Ideally, linguistics should, in order to make language work explicit, decompose language in such a way that it can be recomposed to so to speak à la carte. A process which, you guessed it, could easily follow that of the translation.

You can follow Professor Mohamed Chtatou on Twitter : @Ayurinu

Endnotes: