TRADUÇÃO NA APRENDIZAGEM DE LÍNGUAS: UMA ABORDAGEM ‘PARA QUÊ’ TRADUCCIÓN EN EL APRENDIZAJE DE LENGUAS: UN ABORDAJE ‘PARA QUÉ’ TRANSLATION IN LANGUAGE LEARNING: A ‘WHAT FOR’ APPROACH

Literature about translation in language learning and teaching shows the prominence of the ‘for and against’ approach, while a ‘what for’ approach would be more profitable. In order to prevent the latter approach from becoming a random list of the potential benefits of the use of translation in language teaching, this essay suggests 1 Ca' Foscarfi University, Venice – Italy. Department of Linguistic and cultural studies. Chair of Language Education Research. E-mail: balboni@unive.it.

b.The pros versus the cons of translation: it is the "Leitmotiv" in most studies on the topic, above all in the introductory paragraphs in an application for European funds by Pym et. al. (2012) the first issue among the "innovative aspects" was "a critical re-analysis of empirical research both for and against the use of translation in language learning", and this demonstrates that the 'versus approach' is still the foundation of the forma mentis of scholars dealing with translation in language teaching 3 .
We think that in the long run the 'versus' approach to translation in language learning and teaching is sterile.Pros and cons have been discussed for some decades now, and it is time to move on in the direction of a 'what for' approach.
A review of literature includes examples of 'what for' reasoning, highlighting the increase produced by translation in the quality of lexicon and in intercultural awareness; some also include grammar and assessment among the 'what' elements, yet these objectives seem difficult to agree with, as we shall discuss later.Yet most essays focus on single objectives or juxtapose a series of them.What is lacking, in our opinion, is the reference to a general model of communicative competence, so that the use of 2 Yet some essays today present translation as a sort of 'new' entry in language teaching research in the 21 st century, to quote a few, Lopriore (2006); Cook (2010); D'Amore (2013); Tsagary; Floros (2013). 3Other examples are Corbacho Sánchez (2004); Arranz (2004); Liao (2006); Di Sabato (2007Sabato ( , 2010Sabato ( , 2011)); Boguslawa (2009); Balboni (2010); Leonardi (2010); Vermes (2010).translation as a pedagogical tool is analysed within a wide framework, and not simply on the bases of the experience or the interests of researchers.

A model of communicative competence
In 1780, the illuminist philosopher Condillac wrote that l'art de raisonner se réduit à une langue bien faite.Defining terms is the basis of well-constructed language.
The terms we are defining here are the two parts of the title of the paragraph: 'model'   and 'communicative competence'. 4  A 'model' is a true, therefore unique and perpetually valid, formulation (linguistic, logical-formal, algebraic, etc.).In linguistics, formulaic definitions have been privileged since the 1950s by Šaumian and Chomsky, and in psychology since the 1960s, by Fodor and Neisser, who worked on schemes that were valid on a purely logical basis, independently from their empirical measurability.
A model has four characteristics: a. it describes and explains all the possible implementations of a phenomenon, a notion, anytime, anywhere; b. it includes all and only the relevant factors of an idea or a phenomenon, so that secondary or unpredictable information do not overload the model; c. it can be simple (it operates on a single plane: the Pythagorean theorem does not require further explanatory models) or complex, hierarchically layered to include inferior levels (the model of communicative competence, below, includes several sub-models, as we shall see).
Simple models are declarations ("all languages have at least three functions: subject, verb, object (SVO)"), complex models connect some declarations ("if in every language there are SVO functions, then all languages include the six possible sequences of these three functions: SVO, SOV, OSV, OVS, VSO, VOS)"; d. simple models are the basis of theoretical sciences, aiming at knowledge per se (linguistics want to describe how language works), complex models are the basis of the procedural knowledge of practical sciences, such as language teaching research.
A model of communicative competence must respond to this brief question: what does 'knowing how to communicate in a language' mean?The answer to the question resides in the many branches of language science: general linguistics, pragmasocio-ethnolinguistics, and sciences that study extra-linguistic communication (kinesics, proxemics, objectemics).On these bases a model can be defined of: a. linguistic competence, which includes the rules that govern language, b. extra-linguistic competence, concerning non-verbal codes, c. contextual competence, i.e. the set of social, cultural and pragmatic rules that govern the use of language in social contexts, in communicative events.
These are competences and, as such, they are mental representations, unrelated to real phenomena.The competence to judge the grammatical structure of a sentence exists in a person even if no sentence is spoken.Yet, sentences that are not spoken do not communicate.Contrary to Chomsky's mental competence, Hymes' communicative competence needs the integration of 'knowing' and 'doing', or knowing 'how to do things with words', to use Austin's words.This process requires a system to translate mental knowledge (competences) into social action (performance), to transform 'knowing the language' into 'knowing how to work language', i.e., how to (a) comprehend, (b) produce, and (c) transform, manipulate, translate and summarise texts.
These elements constitute a model of communicative competence and can be graphically represented as follows: The next question is what the contribution of translation would be to the development of communicative competence.An enormous quantity of literature has been answering this question for the last two decades, but most essays, books, experiences, and best practices are monothematic, since they deal with one point, one element, and one factor of language learning.The model of communicative competence turns single-spot proposals into a full and complex vision of the role of translation in language teaching (BALBONI, 2010a).

Different language teaching contexts
A question arises from the final words of the paragraph above: "the role of translation in language teaching".Yet, the word 'language' is an umbrella term which needs some analysis in order to find how translation can be used in teaching the mother tongue, second, foreign and classical languages, as the nature and role of translation changes according to the language teaching context.
We shall survey the different context starting with one which is seldom taken into consideration when discussing about translation in language teaching: diachronic translation in the mother tongue (or in the language of instruction, which in some context is not the mother tongue): a. L1, the mother tongue: in this context the translation of ancient texts into today's variety of the same language can be profitable to help students reflect on their L1.There is some philological research on the topic (STOPPELLI, 2014), but as far as we know there is no language teaching research on "diachronic translation" (BALBONI, 2010b), for example, translating Machiavelli into today's Italian or Cervantes into today's Spanish; we shall not go into this topic any longer as this issue is about nonnative language teaching, but it is apparent that the linguistic, sociolinguistic and cultural awareness raising from this hands-on activity is enormous and no other class activity can provide it with the same time/effort/result ratio.b.L2, second languages, that is, non-native languages spoken in the context where a student lives and studies it: a French student of Portuguese with an Erasmus scholarship in Coimbra.

Translation means at least two different things in this context:
-translating from L1 to L2 focuses on the ability to transfer (translatum is the supine form of transferre, to transfer) L1 meanings into the appropriate (socio-cultural parameter), efficacious (pragmalinguistic parameter) and correct (linguistic parameter) L2 forms; in other words, it is an L2 activity; --vice versa, translating from the L2 into the L1 focuses on full comprehension of the L2 and on an accurate analysis of meanings and connotations in the L1.In other words, it is both an L2 receptive task and a L1 productive activity.
-Both are useful activities, provided the competence in L2 is advanced enough to make the task productive in terms of language acquisition and awareness.c.FL, foreign languages, i.e.French studied in Brazil.
All professional translators translate from FL to L1, although interpreting may be bidirectional; when translating from L1 to FL, professionals have their translations proofread by FL native speakers.
In FL to L1 translation, FL comprehension must be deep and subtle, but the main focus is on L1 competence, that is in transferring all the meanings (linguistic, sociolinguistic, cultural, pragmatic, aesthetic if any) from the FL into the mother tongue, which is the real beneficiary of this activity in terms of language awareness.L1 to FL translation is unusual in the real world, sometimes it is used in language courses, but as we shall see there seems to be no time/effort/result ratio unless the FL level is C1 or, better than C2.And the FL text produced by translation is doomed to be a hypothesis, until a native speaker of the FL has read and accepted it.Maybe, reading and comprehending a poem by Sapphus or by Catullus would be far more motivating, as it could focus on some specific language points, and arrive eventually at translating these poems into the students' native language.This can make students aware of : -the literary quality of the text: ancient literary texts used different rhetoric and stylistic devices than today, and translating these marks of literality can be a fascinating challenge, of great impact on the students' knowledge about their own L1; -the extreme difficulty to transfer the same amount of emotion and aesthetic pleasure into their mother tongues.The way of expressing emotions and the very emotions to be freely expressed in the classical world are quite different from today; -the difficulty and sometimes the impossibility to translate cultural elements: in the 21 st century labelling Sapphus as 'lesbian' is misleading, because bisexuality in Greece had another status than today; translating Cicero's letters where Tiro is referred to as servus, slave, requires for sophisticated choice of words so that the modern reader does not think of Tiro as the way peplum films depicted slaves (as for the cultural dimension in translation, HOUSE, 2016).
f. Language of instruction: so far 'L1' or 'mother tongue' or 'native language' have been used to refer both to the students' L1 and to the language of instruction, of the school system.In most European and American contexts the equation L1 = Language of Instruction is almost true (although migration phenomena are dramatically changing the situation), but in multilingual countries, e.g.Indian schools using English or Maghreb schools where Classical Arabic or French are used, it would be useful to introduce the distinction, because translating into the Language of Instruction is functional for scientific, sociological, informative texts, but literary texts should also be translated into the L1, and comparison between the same student's translations into the mother tongue and into the language of instruction could grant a very high level of translative awareness6 .g. Languages for specific purposes: the languages for the specific purpose of non-ambiguous communication in science, technology, law, and so on; the language of literature, whose specific purpose is -to use Roland Barthes's wordsle plaisir du text.All these varieties of language require special translation abilities, and in these cases translation is not a tool for language learning but a teaching goal per se.
Yet, some translation from special languages can be very useful with advanced students, to improve their awareness of the special characteristics, both formal and cultural, of scientific and literary texts, and of the 'realistic' language of the cinema, which is extremely artefact to sound extremely authentic7 .Yet, such special language translation experiences should be episodes in a normal language course not aimed at training professional special language translators8 .This surveyalthough very syntheticshould be effective enough in showing that translation has many different roles according to the teaching contexts in which it is used.

The model of communicative competence and the use of translation in language teaching
The model shown in the diagram in point 1 shows two boxes, the mental world where competences reside and the social worlds where communication is performed.
They are connected by the cognitive abilities and the language skills.Translation activities have two effects on the components of communicative competence: on abilities (comprehension, translation, production) and on mental competences.For half a century, these effects have been considered negative.Today a lot of literature claims there can be positive effects under special conditions, which can be summarised like this: a. students must be proficient.Most European researchers indicate B2 as the threshold level for the first translation activities, B1 being the threshold level for everyday survival communication.Translation as an ability to be developed pertains to B2 and over; b. translation must be just one of the many techniques used in the classroom; c. the aim of translation must be language learning and language/culture awareness, not translation in itself.
We will consider the benefits and problems caused by translation on the mental competences on the next sections.

Linguistic competence
Mental competences are, in fact, clusters of competences:

phonological, graphemic, morpho-syntactical, textual and semantic competences
The grammar-translation approach claimed that translating developed both morpho-syntax and lexiconand a lot of literature, mutatis mutandis, still does.

mind world a. Morphology and syntax
The focus on form movement often adopts translation as one of the techniques that can reinforce morpho-syntactical competence (e.g., KÄLLKVIST, 2008;WHYATT, 2009;but  -focus on a clear objective, while translation is the most complex and less focused activity: the elements of linguistic competence but phonology are simultaneously present, together with socio-cultural and pragmatic elements; morpho-syntax, which is an enormous field, is just one of the many aspects of a text a translator deals with, so it is not a relevant technique to work on morphosyntax; -be accepted by students as efficient to reach an objective: students are well aware that work on pronouns or verb forms or comparatives or coordination can be done with direct, quick, clear exercises in very many ways (BALBONI, 2012, describes over a hundred techniques), and do not accept to do difficult, wide range, time consuming translations in order to work on some examples of pronouns or verb forms and so on; -be comparable: in a class of 20, a morphological or syntactical exercise on clear objectives can be corrected quickly, and the correct form is apparent to everyone; translations are not easily comparable, unless they are limited to short decontextualized sentences.There may be many correct variables, on all the elements of language, and this prevents students from focusing and comparing morpho-syntactical objectives; -be economic: i.e. the time/effort/result ratio must be positive.Translation is highly time demanding both to carry out, and to correct and to comment; Morphology and syntax can be practiced and developed through a number of teaching techniques which are precise in their object, quick to administer and correct, clear to students who know they are practicing the past tense, subordinate clauses and so on.Why use a difficult, time consuming activity such as translation to practice, develop and assess morpho-syntax?How can students focus morpho-syntax in texts where they also have lexical, textual, socio-cultural problems to face?In our opinion, translation may highlight some morpho-syntactical problems, but it is not a good technique to focus morpho-syntactical features.

b. Lexicon
Literature points out the other important component of linguistic competence that can be supported through translation: lexicon9 .Lexicon has three components: -denomination, that is the objective signifié of a word: 'house' means a building intended for people to live in, 'cat' is a four-legged mammal of the felis catus species; -connotation, i.e. the emotional or cultural nuances linked to a word: 'home' is a 'house', but it is more than just a building; and 'mansion', 'palace', 'hut', 'cabin' are 'houses' but are culturally different, even among Englishspeaking countries; 'pussycat', 'kitten' are not just 'cats'; -metaphorical or figurative meaning strictly linked both to a culture ('to be home and dry', 'Italy is my home country' and so on) and to a period ('house of cards' has acquired a new meaning after the world wide popular TV series describing the fight for power in Washington); 'John is a cat, not a dog' tells a lot about John's character, and 'it rains cats and dogs' has nothing to do with feles cati.
Working with advanced students and translating from the target language into the mother tongue (or into a well-mastered language of instruction) is maybe the best way to work on connotation and on figurative meaningas a matter of fact it is almost the only technique that can be used to focus on these characteristics of lexicon and to let students become lexicon aware, both in the target language, be it modern or classical, and in their own L1 or in very well mastered L2.
Andrea Camilleri has been the Italian best seller author for the last three decades; he uses an invented language, half Sicilian half Italian, and refers to implicit A second textual element which emerges immediately to translators is hypotactical versus paratactical text structure, as the choice between hypotaxis (subordination) and parataxis (coordination) varies dramatically among cultures.
Oriental texts are spirals, Italian and German texts are jagged lines, English texts are sequences of segments: In translation, the hypotactic Italian sequence of subordinate clauses embedded into the principal one and into other subordinates must be reassembled as a paratactic sequence of coordinated sentences in an English text: this demands for morphosyntactical cohesion devices to be changed, (and that makes it a good exercise on the topic, which is clear and apparent to students, and it can be compared) but above all demands for a global view of the whole paragraph (which in Italian can easily be over 10 lines long…) so that the new text conveys the same meaning although it has quite a different structureand this makes translation the best technique for awareness of hypo/paratactical text structures, a C1-C2 objective, not for low level learners.
The differences above concern geolinguistic and geocultural aspects, but there is another extremely important language variety where translation is the most efficient activity, professional and scientific language.Each speech community, i.e. each group of specialists in a field, has its own rules: German art critics, engineers, doctors or teachers follow strict stylistic and rhetoric rules that show their belonging to the speech communities of German art critics, engineers, doctors or teachers.Yet, Indian or Argentinean art critics, engineers, doctors or teachers follow other stylistic rules, such as translating a scientific text from German into Hindi or into South American Spanish means translating the original style into the target speech community's style.This means, for instance, modifying the use of impersonal, passive, stative verb forms, the subordination or coordination cohesive structures, the use of Greek and Latin prefixes and suffixes, the pragmalinguistic force of communicative moves and of communicative acts, and so on.Scientific translation is not a problem of terminology, it is a problem of style and culture, and only translation activities can make advanced students aware of this sociolinguistic and ethnolinguistic feature of scientific discourse.
Translation is the best teaching activity to develop textual awareness, above all as far as the cultural rules of communicative genres are concerned.Yet, textual awareness is an objective for advanced students.

Extralinguistic competence
kinesic competence: gestures, expressions proxemic competence: interpersonal distance, postures objectemic competence: communicative objects Gestures, expressions, interpersonal distance, posture, objects, gifts and so on have communicative value.They are often perceived as natural or universal, yet they are as cultural as language: the same gesture or status symbol or posture may have quite different meanings in different cultures.
Translators do not translate gestures of postures, of course.But in narrative texts, they translate descriptions of gestures, expressions, postures, and so on, and they must be aware that, for instance, the two gestures used in the Western Europe and America to mind world mean OK may have different meanings and they must translate meaning, even though this means non-literal translation.If "John raised his thumb to say OK", it is translated literally, South Asian readers interpret it "he stretched his middle finger"; if the OK was made by thumb and index forming an 'O', Russians would read it as "fuck you".
When a narration tells that "she received the gift and unwrapped it smiling", a translator must consider whether the target culture considers unwrapping a gift as a sign of appreciation or of mere interest in the contents.And must explain the meaning, adding a few words which are not found in the original text.The same must be done for "he arrived, wearing his Harvard School of Law tie", whose meaning is transparent for Northern Americans, but is quite opaque for the rest of the world.
The reflections above are quite obvious for translators and translatologists, but are important for teachers, because there is no technique as efficient as narrative translation in making students aware of the polysemy and the cultural nature of gestures, postures, expressions, gifts, status symbols, clothes, and so on.The main problem is that only native speakers in the target language or almost native like L2 speakers can provide correct judgments in this area of communicative competence.A solution for students is easily available today through tandem mind world translation, that is cooperative distance translation (MEWES, 2010;CERVINI et al., 2013;and passim D'ANGELO, 2012, describe this methodology).
In point 2.b, we hinted at the ordeal of translating Camilleri.The problem of course is not the sociolinguistic dimension, Camilleri's pseudo-Sicilian dialect; the real problem concerns pragmatics and communicative moves, for instance the way respect or disrespect, agreement or challenge etc. are 'told' by the choice of words, the description of gestures, the use of metaphors.An example: Detective Montalbano pays a respectful visit to the local mafia boss, and they smoothly talk about fruits falling near the tree or rolling down far from it, yet the implicit meaning is that the boss's nephew has rolled too far away from his grandfather's way of carrying out mafia business, and as a consequence the boss himself will provide to have his nephew handed into the policebut all of this is never mentioned, it is all left to figurative speech, to silence, to

Abilities and skills
Comprenhending, speaking, dialoguing Summarising, note taking Transforming texts: translating, paraphrasing Translating is one of the text transformation abilities and skills: 'ability' refers to the cognitive abilities, the processes required in comprehension, production and so on, 'skill' is the implementation of abilities in a language (WIDDOWSON, 1998).
Translation is of course important in a theoretical model, as important as all abilities and skills are (apart from comprehension, which is the condition for language acquisition and learning).Yet needs analysis shows that very few language learners really need to develop the ability of translating for actual use in their future lives.In this mind world essay we are not referring to such students, but to all the students who will never be translators, apart from occasional translations functional to momentary pragmatic needs.
Translation activities during a normal high level course help making students aware of the processes of deep comprehension in the non-native languagei.e.
connotative, symbolic, cultural, figurative comprehensionand of production in their mother tongue.This means that occasional translation activities are necessary, not just welcome, at high levels of language instruction.And the most difficult, that is the most challenging and fruitful texts to do this are literary texts and films.
This means that translation helps develop other abilities in two languages.
An unacceptable 'what for': translation used for assessment Duff (1989, p. 5), one of the first who had the courage to support translation as a language learning tool in the Eighties, complained that "it tends to be used not for language teaching, but for testing".And a lot of scholars, above all Asians, still suggest this use for translation (ITO, 2004;LAI, 2008; the whole Part II of TSAGARI; FLOROS, 2013); other researchers are more cautious (KÄLLKVIST, 1998;PORCELLI, 2007;VERMES, 2010;GARIDEL;NIETO, 2014).
Källkvist's case study (1998) shows that the number of lexical errors in translation is higher than in free compositions by the same students, and so it is unreliable even for one of the most important objectives of translation in language learning, i.e. the development of lexicon.
Garidel and Nieto's case study (2014) poses another question: what is the meaning of 'error' when used to evaluate a translation?They carry out a wide literature review on the topic, showing that ambiguity is the common feature of most definitions of translation error.
Yet there are some basics of language testing and evaluation that prevent translation to be used for assessment: a. affective filter: anxiety pollutes data, that become unreliable.Translation is maybe the most complex language skill, and complexity goes with anxiety; only dictation, which has a dramatic impact of the time factor on performance is worse than translation as far as acceptability is concerned (see Carrol's parameters quoted in 3.1); b. relevance of the object: what does translation measure?Does it measure competence or performance?A misunderstood word produces a wrong translation, but the problem is not in the translation ability or in the productive ability in the target language, it is in the comprehension of the source language.
Yet, the same error may result from dictionary misuse, and it concerns the translation process and not comprehension or production.In other cases the error may be of intercultural origin, yet it does appear as a language error; c. testing measures specific learning objectives: how can the tester know whether the errors are in mental competences (lexicon, socio-pragmatics, culture, etc.), in the process (and which part of it: comprehension, translation, production?), or if they are mistakes, that are mere performance errors?How can the tester know which perfect solutions are due to the help of dictionaries, as it often happens with Greek and Latin?
There are lots of testing techniques that focus on the single aspects of communicative competence, and they are combined in certification batteries to provide a 'picture' of the competence of a person at a certain stage.It is not necessary to use a complex, anxiogenous technique such as translation to do what can be done more easily and better with lots of techniques well known to teachers and, above all, to students.

Conclusions
Many decades in the history of language teaching have been spent to convince course designers, authors of teaching materials, language teachers and language testers that translation should not be used in language teaching.Our essay aims at showing that only the latter group, language testers, should avoid translation as a teaching instrument.
In all language teaching contexts translation can be useful and even necessary, provided it is adequate to the competence level of the students and it is used to practice d. LF, lingua franca, which today means English; Greek, Latin, Italian and French once where LFs and Chinese might become relevant in the future.LF speakers are not particularly interested in correctness: if a speaker does not differentiate from and since in duration forms or uses informations in the plural form nobody cares (provided these mistakes are noticed…), because the main aim of LFs is pragmatic efficacy and, if possible, sociolinguistic and intercultural appropriateness, while linguistic correctness is optional as far as it does not interfere with communication.Almost all authors consider translation an activity for high levels of proficiency, and LFs aim at low or middle level proficiency.The only benefit translation grants in the LF context (which means: in the teaching of English as a LF) is training students to be quicker, producing linguistically inaccurate but pragmatically valid translation and interpreting in negotiations.The quality of such an oral translation is measured on its capability of letting multilinguals share enough information to agree on a document, to discuss a contract, and to let them reach their pragmatic social goals.In the LF context, written translation is predominately granted by automatic computer translation softwares, and an interesting classroom activity may be 'correcting' computer mediated translations 5 .e. Classical languages, the context where translation is still the most widely used teaching activity, often consisting of the translation of short decontextualized sentences.Oral skills are not taught in classical languages, the only skills that can be developed are reading comprehension and translation from the CL into the student's L1.Asking students to translate just to train their morphosyntactical and lexical competence means using the most difficult skill, translation, to do what can be more easily, economically and effectively done with other techniques (LINDGREN; BLUMBERG; LANGSETH, 2010).
mafia and antimafia codes and valueslanguage, value and codes which are deeply rooted in Sicily and took decades to be understood by Italian readers.Let alone, foreign readers.All translators working on Camilleri agree that his denotative lexicon is easy and simple, yet its connotations and figurative meanings are sometimes beyond translatability.In 2005, Penguin Books published the English translation of Il profumo della notte as The Smell of the Night and in 2007, Picador published it as The Scent of the Night.Gutkowski (2009) describes the problem in her book Does the night smell the same in Italy and in English speaking countries?Camilleri in English.An essay on translation.c.Textuality Text types are universal, an argumentative text uses the same processes in all languages; text genres are culturally bound implementations of a type, for example, allPhD dissertations are argumentative texts, that is all of them present a thesis, study literature to find homotheses and antitheses, and try to find a synthesis, yet each national tradition has special formats and special styles to be used in this genre: in Italian the personal pronoun 'I' is forbidden in a dissertation, while in many cultures it is required as a sign of personal responsibility in what the PhD student writes.The same happens, for instance, with narrative texts, which are a universal type described by narratology, for instance jokes are narrative texts, but they are a genre which is implemented in extremely different ways in different cultures.Translation activities are the best way to make students aware of genre cultural features, both in the target language and in the mother tongue.
pragmalinguistic competence: functions, acts and moves (inter)cultural competence: models, values, social rules Boguslawa (2009b) provided a good synthesis of the problem in "building L2 communicative confidence through interlingual tasks; towards function-focused L2 learning".The notion of communicative confidence is quite interesting because it includes sociolinguistic appropriateness, pragmalinguistic force and effectiveness, intercultural awarenessand there is nothing like translation to build communicative confidence in an advanced level in a non-native language.
minimal gestures and smiling and frowning… Translating this dialogue is really a pragmalinguistic and intercultural challenge.But only translation activities can open up the awareness of very advanced students as far as the complexity of translation is concerned: translative awareness, as mentioned above.
this position is discussed and supported, explicitly or implicitly, Carrol (1980)introduced an acronym, RACE, to summarise the four main features of teaching techniques.They should: