Translation is one of the greatest resources teachers have at their fingertips, since there is a wealth of ready-prepared, authentic materials which have great generative potential if exploited well. However, it is frequently discredited, for, among other things, being 'uncommunicative', or for focusing purely on accuracy, and teachers have been programmed to view translation as 'bad' classroom practice. My aim in this article is to show that translation does have a place in the classroom and that its implementation Carl be beneficial for learners. Besides, students do, and always will, translate into their L1, no matter how often we exhort them not to - Atkinson (in Harbord, 1992) calls it a "learner-preferred strategy . . . and an inevitable part of second language acquisition". Perhaps we should stop working against this tendency and turn it to our advantage.

Less didactically, over 60% of the world is bi- or multi-lingual, so translation is an everyday activity for many people, with extremely practical applications. The fact that translation has been frowned upon implies that we take a very Euro-centric view of language: since (most) European countries are monolingual (at least officially), we seem to think this is the norm, not the exception, and that 'good' practice should reflect this. This is hugely unjust to the millions of people who speak two - or more -languages daily.

Furthermore, the majority of my students are professionals or University students who have (primarily passive) contact with English in their jobs or studies. To assume that these students, outside my classroom, read in English without translating, is, I think, a gross misjudgement. Using translation in class and encouraging learners to develop useful, applicable techniques to deal with it is surely more beneficial to learners' real needs than, say, role-play or cloze exercises.

As international integration grows, translation will become more and more necessary for non-native speakers of English. I think it would be unfair of me to withhold this valuable technique from my learners.

Translation in the development of language teaching

Translation has always played a role in language teaching. Until the end of the eighteenth century, learning a foreign language implied learning Latin and was based around bi-lingual word lists and parallel texts. When 'modem' languages began to be taught at the end of the century, the same approach was followed, and the structures of English, French, Italian and so on were presented in relation to the structure of Latin. Speaking was not the aim of learning the language: the focus was on the translation of model sentences, chosen to exemplify the structural idiosyncrasies of the language system. Richards and Rodgers (1986) give this example of language having no surrender-value whatsoever for the student:

The philosopher pulled the lower jaw of the hen.

My sons have bought the mirrors of the Duke.

The cat of my aunt is more treacherous than the dog of your uncle.

Titone 1968:28 (in Richards and Rodgers, 1986)

This approach became known as the Grammar-Translation method, since students were presented with rules and then applied them in translation. The theoretical underpinning was "memorising rules and facts in order to understand and manipulate the morphology and the syntax of the foreign language" (Stern, in Richards and Rodgers, 1986). The importance of structure in this approach has resulted in the view that translation is unpedagogical and uncommunicative, which still persists today. The 'Communicative Approach', which most teachers today would profess to use (in some form or other) in the classroom is ambiguous in its approach to translation. Purists argue that all interaction should be undertaken in the L2, although some advocates claim that translation can - and should - be used (judiciously) in class. Howatt (in Richards and Rodgers, 1986), for example, writes: "translation may be used where students need or benefit from it."

Other approaches which use L! in practice include 'Community Language Learning' (where students are provided with language after saying it in the mother tongue) and 'Suggestopaedia' (where learners are frequently exposed to bilingual word-lists or parallel Ll-L2 texts).

While preparing this article, I looked at a number of Advanced-level coursebooks to see how they approached translation. I was surprised to find that none of those I consulted included the technique as a classroom activity, although a number of lower-level coursebooks did. Here is one example:

Translate

• I've been to the United States.
• I went there in 1987.
• Have you ever tried Indian food?
• I live in the capital city of my country. I've lived here all my life

This would imply two things. Firstly, materials are designed with international markets in mind, including multi-lingual classes, where using translation would be unwieldy. Secondly, learners who have reached an advanced level of proficiency in English are considered not to need translation as a tool for acquisition, since they are already able to manipulate the language system successfully and do not need L! reassurance as they gain new language.

A summary of translation theory and problems translation poses

There is a great deal of theory behind the different concepts of translation. I will now consider some of the issues central to the translation process. It is worthwhile looking at these issues since I can confirm that the problems they give rise to are all apparent in my learners' production. The errors they commit in speaking and writing are all unquestionably due to translating (often unconsciously) directly from their L!

Decoding and recoding

This Saussurian principle dictates that language is conveyed through a combination of signs, these being the letters or sounds the writer/speaker uses to convey meaning. These signs vary from language to language, so in order to translate a message the translator must, firstly, decode the signs of the source text to understand the message, secondly, analyse the message for meaning, thirdly, transpose the content into the target language and, finally, reformulate the message using the signs of the target language. The model below summarises the process:

SOURCE LANGUAGE TEXT		RECEPTOR LANGUAGE TRANSLATION
ANALYSIS		RESTRUCTURING
	TRANSFER
		(Nida, in Bassnett-McGuire, 1991)

Obviously, the closer the source language is to the target language, the easier it will be to transfer meaning successfully. Difficulties arise, though, when the L! and the L2 are more distant, or when the "meaning-patterns" (Sapir, in Manko, 1998) of the L! differ from the L2, in which case they are likely to influence target production.

Translation, therefore, goes beyond simply linguistic equation of meaning, since in practice anyone involved in the process must be aware of the paralinguistic implications of the message being conveyed. A good example is the translation of the English pronoun you. Although context often indicates the relationship between speaker and hearer, English makes no distinction between number, nor shows levels of formality, using periphrasis to overcome this. Thus the translator, recoding you into, say, French, must decide whether tu or vous is more appropriate. In Portuguese, this pragmatic choice is further complicated by the existence of eight possible renderings of you.

Strategic decisions

The translator has to take strategic decisions. These encompass the semantic and pragmatic features of the text, and address one of the most controversial issues of translation: how close should the target text be to the source text? Translators must decide whether their version should be nearer the source text, and thus risk L2 fluency, or be a free-standing example of discourse in the target language, and risk being an unrecognisable version of the source text. Hervey and Higgins (1992) claim, however, that to think in terms of a translation being either 'literal' or 'free' is untenable, since all translations fall somewhere on the cline below (adapted from Hervey & Higgins, 1992) :

Literal

Faithful

"Core"

Idiomatic

Free

Linguistic Equivalence

This concept rules that every lexical item carries with it a set of culture-bound connotations, which makes the process of translating one message accurately into another very difficult. Native speakers are aware of these connotations and so there is little scope for serious misunderstanding between them. However, when non-native speakers use language, they may be unaware that some L2 expressions are not fully equivalent to an L1 term, which can lead to misunderstanding, or, in extreme cases, offence. For example, the Portuguese adjective simples when applied to a person is a positive quality, implying a down-to-earth, uncomplicated character; the semiotically similar English lexeme simple carries different connotations.

As well as being interlingual, non-equivalence may also occur on an intralingual level, for example the different concept of bathroom for British and American speakers.

Overcoming problems of equivalence involves using either a different item or combination of items, in order to ensure what Jakobsen (in Bassnett-McGuire, 1991) terms 'adequate interpretation' in the target language. In practice, this means the translator will either subinterpret the text (giving the target audience less information than the source audience) or superinterpret the text (giving the target audience more information).

Loss and gain

This occurs when the target language has no direct equivalent of a source language concept. A frequently-quoted example is the range of Eskimo terms for snow (now discredited in Pullum's The Great Eskimo Vocabulary Hoax, quoted by Swan, in Schmitt and McCarthy, 1997), or the difficulties of expressing time in languages of the Indian subcontinent, where time is viewed as revolving around a point in the present, not as past, present and future as in European languages (for example, the Hindi word kal means either yesterday or tomorrow, depending on the context). The fact that there are no dichotomous translations for a lot of L 1 -L2 (and, obviously, L2-L1) items raises many difficulties for learners. This manifests itself frequently in the classroom. For example, in practice I have noticed that my learners have difficulties rendering some areas of the Portuguese language system, e.g., fazer (do/make) into English, or the variety of words for coffee. Conversely, verbs like glitter, sparkle, gleam, shine, etc., or stare, peek, peer, glance, glimpse, etc. are impossible to translate into Portuguese without using periphrasis.

Meaning

In translation, meaning is either 'attitudinal' or 'associative'. Attitudinal meaning is used to show the attitude of the writer/speaker to something, and is indicated in his/her choice of words. Hervey and Higgins (1992) give the example of the choices between the police, the fuzz and the pigs, and the different attitudes expressed by the choice of word used. Associative meaning shows associations made by the reader/listener. For example, saying He's a nurse or She's a soldier is syntactically and lexically correct, but the reader/listener may feel it to be semantically 'wrong'. Associative meaning can also be extended to names, dates and places too. For example, November 5th has high associative meaning in Britain, as does 25th April in Portugal.

Untranslatability

This occurs when there is no exact syntactical or lexical equivalent in the target language. Examples would include English phrasal verbs, rendered by single lexemes in other languages, or the Portuguese reiterative reply, where the short answer 'echoes' (reiterates) the main verb in the question:

Portuguese: Gostas? Gosto

English: Do you like it? *I like.

Obviously, these problems can be easily remedied, and, in most cases, learners are unaware of the fact that theorists classify them as untranslatable.

Disadvantages of using translation as a teaching resource

Using translation in the classroom has been associated with a number of drawbacks, the most frequently quoted being:

• It focuses solely on two skills (reading and writing).
  
  
• It is an unnatural process, since "no two languages are ever sufficiently similar to be considered as representing the same social reality" (Sapir, in Bassnett-McGuire, 199 1).
  
  
• Translation is usually product-focused, with emphasis on final versions. This gives students no scope for discussing meaning, especially if the translation is set for homework. Students see it as a mechanical, dull activity because they are not aware of the benefits of translation as a process.
  
  
• It places high emphasis on accuracy, contradicting current methodological thinking. This brings into question the whole problem of how to evaluate translation. Evaluation is highly subjective, since there is no universal canon dictating which source and target structures equate. It is therefore difficult to say what is correct and what incorrect, although, presumably, rendering cat as chien would be considered an error. It is easy to focus on the weaknesses of the translation, rather than on its strengths.
  
  
• Traditionally, there have been two ways to focus on the effectiveness of a translation: either its closeness to the source text, or its value as a free-standing text in the target language. It seems to me that, in order to evaluate fairly, the teacher needs to be less dogmatic than this, bearing in mind the literal-free cline above. Popovic (in Bassnett-McGuire, 199 1) claims that within several target texts there will be an 'invariant core' of features present in all. The teacher can use this as a basis for evaluation.
  
  
• The teacher is required to have a high level of competence in the students' language. Also, translation can only be used successfully in monolingual teaching situations. There are a number of 'paratextual features' in any source text which will cause the translator difficulties, especially in very genre-specific texts, such as poetry or technical manuals. In the classroom, this manifests itself as features of L1 appearing in L2. For example, when writing formal letters, students may use the valediction "*I sign myself respectfully" rather than the standard "Yours faithfully/sincerely", because the former is closer to the conventional form of Portuguese.
  
  
• Many texts used are decontextualised and graduation is often overlooked. Advanced students may be able to cope with most, if not all, texts on a linguistic level, but may have more problems decoding and recoding messages between source and target texts.
  
  
• Translation can only be used in a limited number of situations. It can be a useful practice activity, (especially if the text is chosen for the linguistic features it contains), and also for testing or evaluation, but cannot be used for, say, presenting new language.
  
  
• To translate successfully requires a deep knowledge of both source and target culture. The transposition of ideas and concepts becomes very difficult if the target culture is unknown, or only partly known. (Fortunately, my students are aware of target anglophone culture(s), through TV, films and music, and also through classroom exposure.)
  
  
• Students risk becoming over-reliant on translation and will be unable to acquire new language fully until they have translated it.

Advantages of using translation as a teaching resource

Applied linguistics has taken a great interest in translation recently, and in the classroom it benefits students because:

• It is ideal for studying the language system: its focus can be altered to make it genre, lexis, structure or function-specific.
  
  
• It encourages learners to think about meaning and form concurrently. This is because in translation they have already been provided with what they have to say, but need to say it in the target language.
  
  
• There is a wealth of ready-prepared, fully authentic texts at our fingertips, with examples of all kinds of real language in use for learners to notice. This is surely a better way for learners to build up a holistic picture of the language system than giving them strings of language manipulation exercises. If necessary, or with low levels, texts can be made semi-authentic to highlight specific features of the target language, especially if translation is being used as a practice or evaluation activity, or for remedial and revision work.
  
  
• As mentioned, it reflects my learners' contact with English outside the classroom, since they use the language predominantly in its written form.
  
  
• Students' world knowledge is improved through authentic exposure to the target culture(s).
  
  
• It can be used at all levels, and from the first lesson if required.
  
  
• It is a very humanistic approach, since all students are able to follow the development of the lesson.
  
  
• It helps students notice non-equivalent linguistic, semantic and pragmatic features of the second language. In doing so, their attention is drawn to the differences between two language systems. This can dispel some of the mysteries of L2, as students are made aware that its structures have a familiar equivalent in their language. Learners can become more prepared to take risks in the target language when they realise that other people use is as a means of communication, just as they themselves do in their own language. This can help students lower their affective filter, in turn making the acquisition of new language easier.
  
  
• Using translation can help teachers draw students' attention to ingrained errors, such as false cognates, word order or time-tense distinctions.
  
  
• It gives students excellent practice in the subskills of reading and writing. For example, using L2 to L1 translation will improve, among other things, the following L2 reading subskills:

1. recognising the script of the L2 (if different from L1)
2. identification and familiarisation with style, register and appropriacy
3. deducing the meaning and use of unfamiliar lexical items and meaning-patterns in the second language (including metaphorical and non-standard language)
4. understanding conceptual meaning
5. understanding the communicative function of clauses, sentences and paragraphs
6. familiarisation with cohesion
7. familiarisation with coherence
8. identifying important/relevant information
9. deducing meaning from context
10. heightening awareness of genre and identifying sources
11. heightening awareness of different schemata
12. skimming
13. scanning

The writing subskills that follow will also be developed when learners approach L1 to L2 translations:

1. focusing on new information
2. guiding the reader through the message of the text
3. layout and script manipulation
4. replicating the function of stress and intonation
5. selection and use of register
6. selection and use of style
7. spelling
8. cohesion
9. coherence
10. 0 sequencing ideas
11. 1 connecting clauses, sentences and paragraphs
12. 2 referencing (adapted from Stokes, 1993/1994)

• Translation can help make students' writing more sophisticated if they think of what they want to say in L1 and then, with help, transpose those ideas into L2. This supports the argument that students are rarely willing to take risks, or that once they reach a certain level of linguistic development they stagnate and refuse to stretch themselves.
  
  
• Students are unaware of the implications of decoding and recoding a message, or of any of the problems such as equivalence, or loss and gain. Making students more aware of these concepts will (hopefully) sensitise them to the pitfalls of word-for-word translation which frequently occurs in their writing. Students need to be focused on the pragmatic functions of language, rather than on the linguistic features it displays, and, in my view, translation fulfils this objective well.
  
  
• Students need to be made aware that some items show no L1-L2 equivalence at all, and that they will have to sacrifice some of the connotative weight of the L1 (or L2) expression. Examples from (British) English could include trying to render Christmas pantomime or Wimbledon into another language.
  
  
• Making translation more process-focused provides learners with a more holistic view of the L2. Translation can be made a truly integrative activity, practising all four skills, and encouraging students to manipulate the target language together, through drafting, editing and reviewing, which makes translation less mechanical and product-focused.
  
  
• Encouraging students to bring in texts which interest them or which they 'need' will increase motivation and ensure that their needs are constantly met.

Conclusions: implications for the practising teacher

Translation can be a very valuable classroom activity. It is especially -beneficial in the monolingual classroom and can be tailored in such a way that it is highly practical, learner-focused and process-based.

It seems to me that translation can be a highly effective way of drawing learners' attention to the linguistic, semantic and pragmatic features of the target language. I think I should therefore make an effort to give my students greater access to translation as a classroom activity if they want it.

Many of the linguistic errors produced by students are put down to L1 transfer, but I feel it would be more exact to call this phenomenon L1 -L2 mistranslation. Sensitising students to this will hopefully help them iron out at least the more common of these errors. Allowing students simply to 'notice' the salient features of the target language system is not enough to make students competent L2 users (doing so would probably take years in a non-English speaking environment). It is surely more efficient (both for us as teachers and for our learners) to make some of the unique features of the English language system (phrasal verbs, modality, word-final "-s", and so on) more accessible to students by any means necessary, including explicit contrast between L1 and L2 systems. In the course of my reading I came across the term 'refraction' propounded as an alternative description of L1 influence, which is, in my view, an excellent description, implying an imperfect but clearly-discernible image.

In the classroom, then, when we use translation as a technique, there are number of strategies we can usefully adopt. They are:

• To make students more aware of the equivalent affect of what they translate (i.e. whether the effect of the target text will be equal to that of the source text).
• To ensure that texts used are interesting, relevant and, as far as possible, authentic. Using shorter texts in the classroom and longer ones at home which can then be discussed in class will maximise the time spent in class and will encourage spontaneous speech, negotiation and discussion.
• To make translation a process-based activity, including all students at all stages of the process. This will include giving time to plan, reflect, discuss, review and edit their work, and also encourage meaningful, independent interaction in English.
• To try and provide students with learner-centred, cognitive translation activities to help them notice the differences (and similarities) between L1 and L2 meaning-patterns (in the short term), and of the language system as a whole (in the long term). Hopefully, this will help them acquire the patterns of the L2 and lessen the influence of the L1 on their developing interlanguage.

Bibliography

Aitken, R., 1992, Teaching tenses (Longman)
Bassnett-McGuire, S, 199 I, Translation studies (Routledge)
Duff, A., 1989 Translation (OUP)
Harbord, J., "The use of the mother tongue in the classroom" in ELT Journal, Vol. 46/4 (October 1992)
Hervey, S. & Higgins, I., 1992, Thinking translation (Routledge)
Manko, I., 1998, "Cognitive vs Communicative" in IH Journal, April 1998.
Richards, J. C. & Rodgers, T. S., 1986, Approaches and Methods in language teaching (CUP)
Schmitt, N. & McCarthy, M. (eds.), 1997, Vocabulary (OUP)
Soars, J. & Soars, L., I99 I, Headway Pre-Intermediate (OUP)
Stokes, J., l993/4, "Writing and teaching writing" in IHTT RSA/UCLES DTEFLA Distance Training Programme