While fashions in language teaching ebbed and flowed during the twentieth century, certain basic assumptions were accepted by most language teachers, mostly traceable to the 'Great Reform' of the late nineteenth century (Hawkins, 1987). Though these assumptions have affected many generations of students and teachers, they are rarely discussed or presented to new teachers but are taken for granted as the foundation-stones of language teaching. Among them are the ideas that spoken language is more basic than written, that explicit discussion of grammar should be avoided, and that language should be practiced as a whole, rather than as separate parts.

Part and parcel of this tradition is discouraging the use of the L1 in the classroom. This can be phrased in stronger or weaker forms. At its strongest it is Ban the L1 from the classroom. Only in circumstances where the teacher does not speak the students' L1 and the students have different L1s could this be achieved. At weakest it is Minimise the L1 in the classroom, that is to say, use it as little as possible. A more optimistic version is Maximise the L2 in the classroom, emphasising the usefulness of the L2 rather than the harm of the first. To generalise a specific remark from Polio and Duff (1994, p.324), 'teachers have some sense, then, that using the TL as much as possible is important'. However the assumption is phrased, the L2 is seen as positive, the L1 as negative. The L1 is not something to be utilised in teaching, but should be set to one side.

Most teaching methods since the 1880s have adopted this Direct Method avoidance of the L1. According to Howatt (1984, p.289), 'the monolingual principle, the unique contribution of the twentieth century to classroom language teaching, remains the bed­rock notion from which the others ultimately derive'. Stern (1992, p.281) feels that the 'intra-lingual' position in teaching is so strong 'many writers do not even consider crosslingual objectives'. Audiolingualism, for instance, recommended 'rendering English inactive while the new language is being learnt' (Brooks, 1964, p.142). Recent methods do not so much forbid the L1 as ignore its existence. Communicative language teaching and task-based learn­ing methods have no necessary relationship with the L1 yet, as we shall see, the only times the L1 is mentioned is to give advice how to minimise its use. The main theoretical treatments of task-based learning do not for example have any locatable mentions of the classroom use of the L1 (Crookes & Gass, 1993; Nunan, 1989; Skehan, 1998). Most descriptions of methods treat the ideal classroom as having as little of the L1 as possible, essentially by omitting any reference to it. Perhaps the only exception is the grammar/translation method, which has little or no public support.

This avoidance of the L1 lies behind many teaching techniques even if it is seldom spelled out. Most teaching manuals take the avoidance of the L1 as so obvious that no class­room use of the L1 is ever mentioned, say by Halliwell and Jones (1991). The L1 occurs in Scrivenor (1994, p.192) only in the list of problems – 'students using their own language'. Even writers who are less enthusiastic about avoiding the L1 take issue primarily with the extent to which this is imposed, for example Macaro (1997). Duff and Polio (1990) wind up their discussion of the high variability of L2 use in the classroom by listing suggestions for enhancing the proportion of the L2 component, not for utilising the L1 component. People arguing for the L1 to be mixed with the L2 in the classroom on a deliberate and consistent basis are few and far between, to be reported below. This anti-L1 attitude was clearly a mainstream element in twentieth century language teaching methodology.

This is not to say that teachers do not actually use the L1 every day. Naturam expelles furca, tamen usque recurret: like nature, the L1 creeps back in, however many times you throw it out with a pitch-fork. Lucas and Katz (1994, p.558) report that even in English-Only US classrooms, 'the use of the native language is so compelling that it emerges even when policies and assumptions mitigate against it.' The UK National Curriculum still needs to remind teachers 120 years after the Great Reform 'The target language is the normal means of communication' (DES, 1990). Teachers resort to the L1 despite their best intentions and often feeling guilty for straying from the L2 path. Those interviewed by Mitchell (1988, p.28) 'seemed almost to feel they were making an admission of professional misconduct in "confessing" to low levels of FL use'.

One reason why this position has not been more chall­enged may be the influence of EFL, which for historical or political reasons, has concentrated on classes where the teacher does not speak the language of the students and where the students often speak several L1s (Atkinson, 1993). The avoidance of the L1 is a practical necessity in much EFL, even if it reinforces the political dominance of English (Phillipson, 1992).

So a core belief arising out of nineteenth century theories of language teaching is probably held by the majority of the teaching profession in some form. If it has proved successful for over a hundred years, perhaps it should be left alone: if it works, don't fix it. The use of the L1 can be dismissed as yesterday's question: let us get on with the burning issues of today. In a survey of advice offered by 19 Local Education Authority advisors in the UK, 'not a single respondent expressed any pedagogical value in a teacher referring to the learner's own language' (Macaro, 1997, p.29). Yet dismissing the L1 out of hand restricts the possibilities for language teaching. Whatever the advantages of demonstrating 'real' classroom communication through the L2, there is no logical necessity why communicative tasks should avoid the L1. The six types of task used in task-based learning are described in Willis (1996, p.26-27) without mentioning the L1 but are followed by the by now familiar advice 'Don't ban mother-tongue use but encourage attempts to use the target language' (Willis, 1996, p.130). Avoidance of the L1 may well be totally justified for all time and there may be no cause ever to question it again. Yet there have been periods in the past when it was not seen as a self-evident truth and a minority of people in every period who have rejected it.

Let us try and find some of the reasons for avoiding the L1, difficult as they are to find in the literature. To simplify matters, the discussion assumes a teacher who can speak the L1 of the students.

The original justification was probably the way in which monolingual children acquire their first language. If the only completely successful method of acquiring a language is that used by L1 children, teaching should be based on the characteristics of L1 acquisition, as many teaching methods have claimed since the Great Reform. For example Total Physical Response 'simulates at a speeded up pace the stages an infant experiences in acquiring its first language' (Asher, 1986, p.17). So the fact that monolingual L1 children do not have another language means L2 learners should not rely on their other language. The comparison of L1 and L2 acquisition is a vast question. L2 learners have more mature minds, greater social development, a larger short term memory capacity and other differences from young children (Singleton, 1989); above all L2 learners already know how 'how to mean' (Halliday, 1975). The non-existent other language in L1 acquisition is in the class of unalterable differences from L2 learning. The L1 monolingual child does not have another language by definition; it is the one element that teaching could never duplicate. A more effective argument would be based on young bilingual children, as Dodson (1985) points out. The argument for avoiding the L1 based on L1 acquisition is not in itself convincing. It seems tantamount to suggesting that, since babies do not play golf, we should not teach golf to adults.

A related argument is the belief that L2 acquisition is usually unsuccessful, seen in a quotation such as 'Very few L2 learners appear to be fully successful in the way that native speakers are' (Towell & Hawkins, 1994, p.14). The L2 learner is a failure for not achieving the same competence as the child, that is to say becoming a native speaker. The goals of L1 acquisition and L2 learning are treated as identical, except that the L2 learner seldom gets there. As argued elsewhere (Cook, 1997a; 1999), this attitude sees L2 users as failing to achieve membership of a group that they can never belong to, as shadows of native speakers, not L2 users in their own right. L1 children achieve native speaker competence in one language; L2 users achieve competence in more than one language (Cook, 1997b). Whether L2 learners are successful or not has to be measured against the standards of L2 users not those of native speakers and so L1 'success' in becoming native speakers is different from L2 'success' in becoming L2 users.

The argument from language compartmentalisation

A second argument is that successful L2 acquisition depends on keeping the L2 separate from the L1. This implies the goal of teaching is coordinate bilingualism in which the two languages form distinct systems in the mind rather than compound bilingualism in which they form a single compound system (Weinreich, 1953). L2 learning should happen solely through the L2 rather than being linked to the L1. The mid-twentieth century rationale for this was transfer theories such as Contrastive Analysis (Lado, 1957); if the major problems in L2 learning come from the L1, then let us eliminate it as much as we can. This compartmentalisation is particularly evident in the many twentieth century attempts to teach meaning without recourse to the L1. Teachers explain the L2 word, define or mime its meaning, show pictures, and so on, without translating, in the long-term hope that this builds up the L2 as a separate system. Yet the two languages are inter­woven in the L2 user's mind in vocabulary (Beauvillain & Grainger, 1987), in syntax (Cook, 1994), in phonology (Obler, 1982) and in pragmatics (Locastro, 1987). L2 users are more flexible in their ways of thinking and are less governed by cultural stereo­types (Cook, 1997b). The L2 meanings do not exist separately from the L1 meanings in the learner's mind, regardless of whether they are part of the same vocabulary store or parts of different stores mediated by a single conceptual system (Cook, 1997b). A L2 is not just adding rooms to your house by building on an extension at the back: it is rebuilding all the internal walls. Trying to put languages in separate compartments in the mind is doomed to failure since they are connected in many ways. The uniqueness of L2 use is seen in code-switching in which both languages are simultaneously on line. One language is switched to another according to speech function, rules of discourse and syntactic properties of the sentence (Cook, 1996). Code-switching is a highly skilled activity—the 'bilingual mode' of language in which L1 and L2 are used simultaneously rather than the 'mono­lingual mode' in which they are used separately (Grosjean, 1989). It forms part of normal L2 use in many L2 situations outside the classroom where both participants share two languages. Such non-compartmentalised L2 use may well be part of the external L2 goals of teaching and could form part of the classroom if not nipped in the bud. The L1 plays an integral role in L2 learning as well as L2 use. Teachers using group­work have often lamented the tendency for students to use their L1. Vygotskyan-style research has, however, documented how this forms a valuable part of learning as a social enterprise and of the 'scaffolding' support that the learners need to build up the L2: 'L1 is used as a powerful tool of semiotic mediation between learners … and within individuals…' (Anton & DiCamilla, 1998, p.415). Surveys of students' strategies show the importance of this L1 use, for example the 73% of students who 'ask classmates for meaning' (Schmitt, 1997). The theory of cultural learning sees collaborative dialogue as the essential means by which human beings learn (Tomasello 1999). We learn by trying to see the world from the viewpoint of others. As Stern (1992, p.282) puts it, 'The L1-L2 connection is an indisputable fact of life'. Keeping the languages visibly separate in language teaching is contradicted by the invisible processes in the students' minds. Language teaching that works with this fact of life is more likely to be successful than teaching that works against it. Many likely L2 goals for students involve mediation between two languages rather than staying entirely in the L2. Students trained in coordinate bilingualism will, for instance, find it difficult to carry out the jobs of interpreters, business negotiators or travel representatives. Nor indeed can a separate L2 achieve the internal goals of language teaching; if the aim of learning a language is to improve the students' minds cognitively, emotionally or socially, the L2 had better not be insulated from the rest of the mind.

Clearly the learner needs to encounter the language in order to learn it. One of the functions of teaching is to provide samples of the L2 for the students. The argument suggests that the teacher can maximise the provision of useful L2 examples by avoiding the L1. A further step is to insist that the L2 should be the language of real communication during the class rather than the L1. A typical communicative teaching view is that 'Many learners are likely to remain unconvinced by our attempts to make them accept the foreign language as an effective means of satisfying their communicative needs, if we abandon it ourselves as soon as such needs arise in the immediate classroom situation' (Littlewood, 1981, p.45). The L2 will remain a set of odd and arbitrary conventions if the students do not experience it in meaningful ways. This is the basis for such claims as 'The natural use of the target language for virtually all communication is a sure sign of a good modern language course' (DES, 1990, p.58). A teacher who uses the L1 for classroom interaction is depriving the students of the only true experience of the L2 they may ever encounter. The teacher is wasting a golden opportunity if they say 'What's the time?' or 'Put your home­work on my desk' in the L1.

The language of classroom interaction is, however, a genre of its own, as Willis (1996, p.17) reminds us. Classrooms use the interaction sequence Initiation, Response and Feed­back (IRF) (Sinclair and Coulthard, 1975), which is peculiar to teaching whichever language it is done in. Using the L2 for IRF demonstrates the language of class­room interaction and trains the students to participate in other class­rooms. But it hardly in itself provides exposure to the full range of external language the students need.

The teacher also uses the L2 for social interaction with the students about the weather, the world, yesterday's baseball match or whatever they are interested in. This provides the students with a range of natural samples of the L2 in action that go further than the language of teaching. However, it is likely to be restricted in conversational topics, roles, and language functions, constrained by the different roles of students and teachers.

Despite these quibbles, so far as external goals of teaching are concerned, it is clearly beneficial to expose the students to as much L2 as possible. No-one will quarrel with providing models of real language use for the students. Nor would anyone deny that it is important for the students to develop strategies for working out the meanings of the L2 from realistic classroom contexts. But this is not necessarily incompatible with the use of the L1 in the classroom. Overall, accepting that students should meet natural L2 communication in the classroom supports maximising the L2 rather than avoiding the L1. Willis (1996, p.49) gives the typical good advice 'explain to students that if they want to communicate in the target language they need to practice'. The maximal provision of L2 input does not deny the L1 a role in learning and teaching. Having a large amount of meaningful L2 use, including samples of language relating to external goals, does not preclude using the L1.

To sum up, none of the three arguments from L1 learning, compartmentalisation of languages and the provision of L2 use strongly support the view that the L1 should be avoided. They rely partly on a comparison of things that are ultimately incommensurate – L2 learning is not L1 acquisition and L2 users are not the same as L1 users – partly on extending the maximal provision of L2 to an avoidance of the L1. Useful as it may be to use the L1 sparingly, this has no straightforward theoretical rationale. The pressure from this, mostly unacknowledged, anti-L1 attitude has prevented language teaching from looking rationally at ways in which the L1 can be involved in the classroom. It has tended to put an L2 strait­jacket on the classroom which has stifled any systematic use of the L1. If avoiding the L1 is merely sensible advice that students should encounter as much of the L2 as possible, new avenues are opened for language teaching which involve the active systematic use of the L1.

How can the L1 be better integrated into teaching? One step is to license what teach­ers now feel guilty about doing, namely falling back on the L1. At least teachers can be given absolution for using the L1 – even if in English secondary schools this would actually break the law of the land – and feel slightly more comfortable with their daily practice.

The next step is to consider overall teaching methods that make use of the L1 actively within the classroom. Some have been mentioned in the context of L2 user goals in Cook (1999). None of them have probably been practiced on a large scale nor do any of them represent a complete approach that can apply to a variety of situations. Their common factor is trying to use the L1 positively in the classroom rather than seeing it as a regrettable fact of life that has to be endured.

In alternating language approaches the students are at one moment learning their L2, at another using their L1. An important criterion is reciprocality: both languages are involved, not one taken for granted. The difference from bilingual classes with groups with the same L1 is that they require mixed groups of speakers of two languages. One possibility is computer 'e-pals' who exchange messages in their respective languages, say through the Tandem system. A more radical approach is alternation between languages for parts of the school day. The Key School Two-way Model teaches classes of mixed English and Spanish speakers through English in the morning and Spanish in the afternoon (Rhodes, Christian & Barfield, 1997). The Alternate Days approach teaches the standard curriculum subjects to children with L1 Pilipino using English and Pilipino on alternate days (Tucker, Otanes & Sibayan, 1971). Dual Language Programs strike a balance between two languages in the school curriculum, ranging from 90% in the minority language versus 10% in the majority language in the pre-school year to 70% versus 30% in second grade (Montague, 1997).

A European variant on alternating languages is found in Reciprocal Language Teaching (Hawkins, 1987). Pairs or groups of students learn each other's languages on alternating occasions. A reciprocal course for English teachers of French and French teachers of English alternated languages each day so that the students on Monday were the teachers on Tuesday (Cook, 1989). This applied not just to the classroom but also to everything that was said during the day and any social event in the evening. The course took place in England and France on alternate years. No constraints were placed on the teaching other than the language-of-the-day requirement. The unique feature of reciprocal teaching is the exchange of learning and teaching roles as well as the alternation of languages.

While alternating language methods recognise the student as an L2 user, they still compartmentalise the two languages. Dolson & Lindholm (1998) for example give as one of the criterial features of the Two-Way Immersion Model 'the program involves periods of instruction during which only one language is used.' In some ways they represent a doubled-up way of avoiding the L1; while the L2 changes, there is not necessarily any change in the role of the L1 in the teaching. The same is true of immersion education, in which in a sense languages alternate between the school and the world outside but not necessarily within the classroom itself. Alternating languages approaches are limited by requiring two more or less balanced groups of L1 speakers. When the two languages are spoken by a majority and a minority group within the same country, the unequal relationships between the two groups also need to be taken into account.

B) Methods that actively create links between L1 and L2

There have nevertheless been teaching methods that have favoured using both languages within the same lesson.

i) the New Concurrent Method

In the New Concurrent Method the teacher switches from one language to another at key points according to particular rules (Jacobson, 1990). Teaching English to Spanish-speaking children, the teacher can switch to Spanish when concepts are important, when the students are getting distracted, or when a student should be praised or reprimanded. Or the teacher may switch to English when revising a lesson that has already been given in Spanish. This method acknowledges code-switching as a normal L2 activity and encourages the students to see themselves as true L2 users, at home in both languages. Hence the language classroom becomes a real L2 use situation in which both languages are concurrent, not a pretend L2 monolingual situation. Jacobson's switch-points resemble the patterns in real-life code-switching, adapted to the class­room.

More radical concurrent teaching uses cognates as the entry point for the student (Giaque & Ely, 1990). In the early lessons the teacher and students use the L1 to supply vocabulary items they don't know, for example 'Je am having difficulté with this learning activité'. This 'franglais' fades out rapidly so that after two weeks the teacher is talking at least 50% French. Code-switching is happening within the same sentence rather than between sentences as in Jacobson's proposal. In general the role of the L1 in concurrent teaching is to foster L2 learning through a more natural L2-using situation.

ii) Community Language Learning (CLL)

The core of the beginners lesson in CLL, alias counselling-learning, is the students talking to each other spontaneously in the L2 via the mediation of the L1 (Curran, 1976). At stage 1 the student says something in their L1; it is translated by the teacher into the L2 and repeated by the student in the L2; the other students overhear both the L1 and the L2 versions of the sentence. As the students progress, they depend less and less on L1 translations. Though the follow-up techniques in CLL are conventional, this central technique uses the L1 as the vehicle for giving L2 meaning in whole sentences. It sees the L1 as the initiator of meaning and attaches the L2 to the L1.

iii) Dodson's Bilingual Method

Dodson's Bilingual Method requires the teacher to read a L2 sentence aloud several times and to give its meaning in the L1 (a technique termed 'interpreting' rather than 'translating'). Next the students 'imitate' by repeating the sentence, first in chorus and then individually (Dodson, 1967). The teacher tests their understanding of the meaning by saying the L1 sentence while pointing to a picture, requiring them to answer in the L2. The method has been developed for helping English-speaking children with Welsh (Dodson, 1985). In general the role of the L1 in the Bilingual Method is to help the students get hold of the meaning of the language. Like CLL, translation is used only to convey meaning and consists of whole sentences. Here, however, the process starts with a teacher's L2 sentence translated into the L1, while in CLL it starts with the student's self-created L1 sentence, which is then translated into the L2.

4. Ways of using the L1 positively in teaching

Let us see how the L1 could be used more positively by building on existing classroom practice, which has been well described in the studies to be cited during this section, namely Macaro (1997) who questioned 271 modern language teachers and 196 pupils in English schools, Franklin (1990) who looked at 201 modern language teachers in Scotland and Polio and Duff (1994) who concentrated on six US university class­rooms 

If there is no over-riding obligation to avoid the L1, each use can be looked at on its merits. One factor to consider is efficiency: can something be done more effectively through the L1? A second factor is learning: will L2 learning be helped by using the L1 alongside the L2? The third factor is naturalness: do the participants feel more comfortable about some functions or topics in the first language rather than the second, as studies in code-switching have shown? The fourth factor is external relevance: will use of both languages help the students master specific L2 uses they may need in the world beyond the classroom? Against these four factors has to be set the potential loss of L2 experience. Despite reservations about avoiding the L1, it is clearly useful to have large quantities of the L2, everything else being equal.

A Teacher conveying meaning

The age-old problem for the teacher is how to convey the new meanings of the second language to the learner, whether for words, sentences or language functions. Some mid-twentieth century teaching methods treated this as the core element in methodology, for example the film-strips of audio-visual teaching or the situations in situational teaching.

-   teacher using L1 for conveying and checking meaning of words or sentences

The use of the L1 for conveying word and sentence meaning recognises that the two languages are closely linked in the mind, as in the overall methods described in the last section. 39% of Scottish teachers (Franklin, 1990) used the L1 for explaining meanings, 8% the L2, the remainder using the L2 'with difficulty'. Trying to explain hospitable or nuclear family in lesson three of an English beginners' course (Mohamed & Acklam, 1992, p.18) may be a waste of time. Explaining la biologie or la physique (Terrell, Rogers, Barnes,  & Wolff-Hessini, 1993, p.25) to a French class using cartoons of a frog and a mathematical formula is unlikely to be successful (unless of course it relies on the surreptitious L2/L1 equation of cognates in the students' minds). This is not to say that teaching should relate all meaning to the L1. Many differences of vocabulary and meaning cannot be covered by giving a translation equivalent, say the deceptive English-Norwegian translation pairs hate=hate and love=elske. (Johansson, 1998). This L1 use extends to the checking of comprehension via the L1, also popular among teachers. Using the L1 to convey meaning may be efficient, help learning and feel natural in the L2 use environment of the classroom.

-   teacher using L1 for explaining grammar

Explicit grammar teaching, discouraged during most of the twentieth century, has had some life breathed into it recently through the advocates of language awareness and of Focus on Form (FonF), who claim it may be used when it arises naturally out of classroom activities rather than being the starting point (Long, 1991). However, while the contributors to Doughty & Williams (1998) raise many questions about how to implement FonF, none of them asks which language should be used, presumably accepting the usual L2 default. Whether the L1 or the L2 is best for explaining grammar is a practical issue. Most studies of cognitive processing suggest that even advanced L2 users are less efficient at absorbing information from the L2 than from the L1 (Cook, 1997b). Hardly surprisingly teachers are not enthusiastic about carrying out grammar explanation in the L2 (Macaro, 1997). 88% of Scottish teachers used the L1 (Franklin, 1990) and all six teachers in Polio and Duff (1994). Given that Lesson 2 of a French beginners course Panorama (Girardet & Cridlig, 1996) includes 'La conjugaison pronominale', 'Construction avec l'infinitif' and 'Les adjectifs possessifs et demonstratifs', what else are they supposed to do? The main overall argument for using the L1 for grammar is then efficiency of understanding by the students.²

B Teacher organising the class

The ways that the teacher organises the class also involves a choice of language. Some of the possibilities include:

-   teacher using the L1 for organising tasks

To carry out a task, the students have to understand what they have to do. The first page of Unit 1 of Atlas 1 (Nunan, 1995), a 'beginning' English course, starts with '1 Look at the picture and practice the conversation. 2. Learn key classroom terms. Label the pictures with these words …' Unless translated into the L1, these instructions are unlikely to be more than words on the page, partially comprehensible through the teacher's skill at demonstration. Hardly surprisingly, the most common L1 use found by Polio and Duff (1994, p.32) was 'words related to the academic context'. Interestingly Deux Mondes (Terrell et al, 1993, xxiv) in reverse uses L2 French words in the advice to English students, such as 'Getting started with the Etapes'. Some teachers resort to the L1 after they have tried in vain to get the activity going in the L2 (Macaro, 1997). Franklin (1990) found 68% preferred the L2 for activity instructions, 8% the L1. Again the argument for the L1 is efficiency, leading to more effective learning.

-   teacher maintaining discipline through the L1

The need to maintain control over secondary school classes often calls for the L1. Saying 'Shut up or you will get a detention' in the L1 is a sign of a serious threat rather than practising imperative and conditional constructions. One class reported their teacher slipped into the L1 'if it's something really bad!' (Macaro, 1997). 45% of teachers prefer the L1 for discipline, 15% the L2 (Franklin, 1990, p.21). Again the reason for the L1 may partly be efficiency of comprehension, partly to show it is real rather than pretend.

-  teacher getting contact with individual students through the L1

84% of English teachers try to provide feedback in the L2 (Macaro, 1997). Telling a student how well they have done in their own language may make this more 'real'. The L1 is predominantly used for correction of written homework, for instance by 56% of Scottish teachers (Franklin, 1990). The teacher may also switch to the L1 to make personal remarks to a student, for instance when a student has a coughing fit (Polio & Duff, 1994, p.318). The main benefit of the L1 for personal contact is naturalness. The teacher is treating the students as their real selves rather than dealing with assumed L2 personas.

-   teacher using the L1 for testing

At one level, running tests is an everyday classroom routine, for which the L1 is preferred by 51% of Scottish teachers (Franklin, 1990). At another level it is the public examinations that students may have to take. A vexed question in the UK is the extension of L1 avoidance to examinations through the dictum that 'The form of a Target Language exam assumes also rubrics in the TL' (SCAA, 1994). The students are tested not only on their ability to do the test but also on their understanding of the instructions. On the one hand this may not fully challenge the students' L2 abilities, on the other it may constrain the complexity of the tests that can be set through the limited language that can be used.

One alternative is to treat the examination as a test of L2 use involving both languages. The Institute of Linguists (1988, p.2) have developed examinations in languages for international communication in which candidates 'are called on to mediate between speakers and/or writers of these two languages' through tasks in which both languages have to be handled. A recent Diploma level Italian examination for instance requires the students to adopt the role of a journalist preparing an article on young people in Europe. Their first task is to summarise in English the contents of two Italian newspaper articles for their editor to read; their second is to write two letters in Italian requesting the views of two authors on the information they have uncovered. To succeed, students have to use both languages: they are tested on whether they can use the L2 effectively, not on how close they are to monolingual native speakers. The argument for using the L1 in testing is then sheer efficiency.

C Students using L1 within the classroom

The student's own use of the L1 has mostly been minimised by the teaching manuals, which provide many suggestions on how this may be achieved. Unlike the discussion so far, student uses of the L1 do not necessarily mean that the teacher has to know the L1 since they can largely take place without the teacher's control, dangerous as this may be if the teacher knows nothing of their L1.

-   students using L1 as part of the main learning activity

The word 'translation' has so far been avoided as much as possible because of its pejorative overtones in teaching. Translation as a teaching technique is a different matter from translation as a goal of language teaching. A certain proportion of students may intend to translate or interpret from one language to the other, whether as professional translators or as business people or whatever. Like code-switching, translation is a unique attribute of L2 users and a normal part of many of their lives (Malakoff & Hakuta, 1991), for example in the world-wide practice of minority children acting as interpreters for adults who cannot speak the majority language.

Stern (1992) approves translation teaching techniques which go from the L2 to the L1 as they do not presuppose the knowledge that the exercise is supposedly teaching. The chief problem may be using translation as linguistic detective work rather than as a communicative exercise to convey in one language what has been expressed in another. Once the avoidance of the L1 is relaxed, there is no intrinsic reason why translation is wrong, even if it has other snags. As Malakoff & Hakuta (1991, p.163) say, 'translation provides an easy avenue to enhance linguistic awareness and pride in bilingualism.' Usawa (1997) surveys several positive approaches to translation.

If the L1 and the L2 co-exist in the same mind, both languages can be used at the same time without necessarily converting chunks of one into chunks of the other. The students are not treated as monolinguals in either language but are working through both, for example in exploiting L1 and L2 cognates, as in the New Concurrent Method. The basis for using the L1 as part of the main activity is its potential contribution to learning and its relationship to external L2 user goals.

-   students using L1 within classroom activities

Advice to teachers on groupwork and pairwork has mostly stressed minimising the use of the L1; 'If they are talking in small groups it can be quite difficult to get some classes – particularly the less disciplined or motivated ones – to keep to the target language' (Ur, 1996, p.121). Yet code-switching is a normal feature of L2 use when the participants share two languages. Without the distrust of the L1, there is no reason why students should not code-switch in the class­room. Furthermore, as we saw earlier, L1 provides scaffolding for the students to help each other. L1 use 'is a normal psycholinguistic process that facilitates L2 production and allows the learners both to initiate and sustain verbal interaction with one other' (Brooks & Donato, 1994, p.268). They may explain the task to each other, negotiate the roles they are going to take, or check their understanding or production of language against their peers through the L1. These purposes for the L1 clearly fit well with the overall rationale for the task-based learning approach, even if so far they have been discouraged or ignored.

Several other possibilities exist for students to use the L1 in learning, both in the classroom and outside, particularly as a way in to the meanings of the L2. One is the use of bilingual dictionaries, which 85% of students find useful (Schmitt, 1997). Another possibility is dual language texts on facing pages, seen in the Penguin Parallel Texts series for adults, for example Roberts (1999), and in the Mantra books for ethnic minority children in England, for example Sanjivinie (1995). A third is the use of L2 films with L1 subtitles, sometimes found as an option in video techniques in CD-ROMs. The students exploit L1 principally for getting the meanings of the L2.

To conclude this paper, rather than the L1 creeping in as a guilt-making necessity, it can be deliberately and systematically used in the classroom. Some suggestions here are:

·  to provide a short-cut for giving instructions and explanations where the cost of the L2 is too great

·  to build up the inter-linked L1 and L2 knowledge in the students' minds

·  to carry out learning tasks through collaborative dialogue with fellow-students

·  to develop L2 activities such as code-switching for later real-life use

In all of these the classroom is treated as a situation of L2 use where two languages are permanently present. They are only a hint of the techniques that teachers can develop once they are free from their inhibitions about using the L1.

Howatt (1984, p.289) suggested that 'if there is another "language teaching revolution" round the corner, it will have to assemble a convincing set of arguments to support some alternative (bilingual?) principle of equal power'. This article has suggested ways of introducing the L1 into the classroom to produce students who are able to operate with two language systems, as genuine L2 users, not imitation natives. Bringing the L1 back from exile may lead not only to the improvement of existing teaching methods but also to innovations in methodology. In particular it may liberate the task-based learning approach so that it can foster the students' natural collaborative efforts in the classroom through their L1 as well as their L2. While this paper has tried to deal with the L1 issue on its own, the discussion forms part of a wider approach to language teaching that is emerging, based on the uniqueness of the L2 user (Cook, 1998; 1999).

Vivian Cook
Vivian Cook worked in EFL in Ealing Technical College in London, then as Director of the Language Service at North East London Polytechnic and now teaches at Essex University [addendum 2019; he is now emeritus professor at Newcastle University: Vivian.Cook@ncl.ac.uk]. He is chiefly known for L2 learning research, for for his book on Chomsky and for research into the English writing system and street signs. He was founder and first President of the European Second Language Association (EUROSLA).

Footnotes

1. I am grateful to Ignazia Posadinu, John Roberts, Phil Brew and my MA students for reactions to earlier drafts and to the anonymous reviewers of CMLR for helpful suggestions.

2. A little-mentioned problem with grammatical explanation is whether the grammar should come from the L1 or the L2 cultures. The logic of avoiding the L1 would mean that only the grammar of L2 grammarians is appropriate, creating additional problems when the grammatical traditions are different, say between English and Japanese.

References

Anton, M. & DiCamilla, F. (1998). Socio-cognitive functions of L1 collaborative interaction in the L2 classroom. Canadian Modern Language Review, 54, 3, 414–342

Asher, J.J. (1986). Learning Another Language Through Actions: The Complete Teachers' Guide-hook. 3^rd edition. Los Gatos, California: Sky Oaks Publications

Atkinson, D. (1993). Teaching in the target language: a problem in the current orthodoxy. Language Learning Journal, 8, 2–5

Beauvillain, C. & Grainger, J. (1987). Accessing interlexical homographs: some limitations of a language-selective access. Journal of Memory and Language, 26, 658–672

Brooks, F.B. & Donato, R. (1994). Vygotskyan approaches to understanding foreign language discourse during communicative tasks. Hispania, 77, 262–274

Brooks, N. (1964), Language and Language Learning. New York: Harcourt Brace

Cook, V.J (1983). Some assumptions in the design of courses. University of Trier Papers, Series B, no 94

Cook, V.J. (1989). Reciprocal language teaching: another alternative. Modern English Teacher, 16, 3/4, 48–53

Cook, V.J. (1994). Timed grammaticality judgements of the head parameter in L2 learning. In G. Bartelt (ed.), The Dynamics of Language Processes (pp.15–31). Tübingen: Gunter Narr

Cook, V.J. (1996). Second Language Learning and Language Teaching, Edward Arnold. Second edition [5th edition 2016]

Cook, V.J. (1997a). Monolingual bias in second language acquisition research. Revista Canaria de Estudios Ingleses, 34, 35–50, 1997

Cook, V.J. (1997b). The consequences of bilingualism for cognitive processing. In A. de Groot & J.F. Kroll (eds.), Tutorials in Bilingualism: Psycholinguistic Perspectives  (pp.279‑299) Hillsdale, N.J.: Erlbaum

Cook, V.J. (1998). Relating SLA Research to Language Teaching Materials. Canadian Journal of Applied Linguistics, 1, 1–2, 9–27

Cook, V.J. (1999). Going beyond the native speaker in language teaching. TESOL Quarterly, 33, 2, 185–209

Crookes, G. & Gass, S. (1993). Tasks in a Pedagogical Context: Integrating Theory and Practice. Clevedon: Multilingual Matters

Curran, C.A. (1976). Counselling-Learning in Second Languages. Apple River Illinois: Apple River Press

Department of Education (DES) (1990) Modern Foreign Languages for Ages 11 to 16. London: Department of Education and Science and the Welsh Office

Dodson, C.J. (1967). Language Teaching and the Bilingual Method. London: Pitman

Dodson, C.J. (ed.) (1985). Bilingual Education: Evaluation, Assessment and Methodology. Cardiff: University of Wales Press

Dolson, D.P & Lindholm, K. (1998). World-class education for children in California: a comparison of the Two-Way Bilingual immersion and European School Models. In To be found

Duff, P.A & Polio, C.G. (1990). How much foreign language is there in the foreign language classroom. Modern Language Journal, 74, ii, 154–166

Franklin, C.E.M. (1990). Teaching in the target language. Language Learning Journal, Sept, 20–24

Giauque, G.S. & Ely, C.M. (1990). Code-switching in beginning foreign language teaching. In Jacobson, R. & Faltis, C. (ed.) 174–184

Girardet, J. & Cridlig, J-M. (1996). Panorama. Paris: CLE International

Grosjean, F. (1989). Neurolinguists, beware! The bilingual is not two monolinguals in one person. Brain and Language, 36, 3–15

Halliday, M.A.K. (1975). Learning How to Mean. London: Edward Arnold

Halliwell, S. & Jones, B. (1991). On Target: Teaching in the Target Language. London: CILTR

Harbord, J. (1992). The use of the mother tongue in the classroom. English Language Teaching Journal, 64, 4, 350–355

Hawkins, E. (1987). Modern Languages in the Curriculum. Second edition, Cambridge: CUP

Howatt, A. (1984). A History of English Language Teaching, Oxford: Oxford University Press

Institute of Linguists (1988). Examinations in English for International Purposes. London: Institute of Linguists

Jacobson, R. (1990). Allocating two languages as a key feature of a bilingual methodology. In Jacobson & Faltis (eds.), 3–17

Jacobson, R. & Faltis, C. (Eds.) (1990). Language Distribution Issues in Bilingual Schooling. Clevedon: Multilingual Matters

Johansson, S. (1998). Loving and hating in English and Norwegian: a corpus-based contrastive study. In D. Albrechtsen, B. Henrikson, I.M. Mees & E. Poulsen (eds.) Perspectives on Foreign and Second Language Pedagogy (pp.93–106). Odense University Press

Lado, R. (1957). Linguistics Across Cultures. Ann Arbor: University of Michigan Press

Littlewood, W. (1981). Communicative Language Teaching: An Introduction. Cambridge: Cambridge University Press

Locastro, V. (1987). Aizuchi: A Japanese conversational routine. In: L.E. Smith (ed.), Discourse across Cultures (pp.101–113). New York: Prentice Hall

Long, M. (1991). Focus on form: A design feature in language teaching methodology. In K. de Bot, R. Ginsberg, & C. Kramsch (Eds.), Foreign Language Research in Cross-cultural Perspective (pp. 39–52). Amsterdam: John Benjamins.

Lucas, T. & Katz, A. (1994). Reframing the debate: the roles of native languages in English-Only programs for language minority students. TESOL Quarterly, 28, 3, 537–561

Macaro, E. (1997). Target Language, Collaborative Learning and Autonomy. Clevedon: Multilingual Matters

Malakoff, M. & Hakuta, K. (1991). Translation skills and metalinguistic awareness in bilinguals. In Bialystok, E. (ed.), Language Processing in Bilingual Children (pp.141–166), Cambridge: Cambridge University Press

Mitchell, R. (1988). Communicative Language Teaching in Practice. London: CILTR

Mohamed, S. & Acklam, R. (1992). The Beginners' Choice. Harlow: Longman

Montague, N.S. (1997). Critical components for dual language programs. Bilingual Research Journal, 21, 4 

Nunan, D. (1989). Designing Tasks for the Communicative Classroom. Cambridge: Cambridge University Press.

Nunan, D. (1995). Atlas 1. Boston: Heinle and Heinle

Obler, L.K. (1982). The parsimonious bilingual. In Obler, L. & Menn, L. (eds.), Exceptional Language and Linguistics, Academic Press

Office for Standards in Education (OFSTED) (1993). Handbook: Modern Foreign Languages, Key Stage 3. London: Her Majesty's Stationery Office (HMSO)

Phillipson, R. (1992). Linguistic Imperialism. Oxford: Oxford University Press

Polio, C.P. & Duff, P.A. (1994). Teachers' language use in university foreign language classrooms: a qualitative analysis of English and target language alternation. Modern Language Journal, 78, iii, 311–326

Rhodes, N.C., Christian, D. & Barfield, S. (1997). Innovations in immersion: the Key School two-way model. In R.K. Johnson & M. Swain (eds.) Immersion Education: International Perspectives (pp.265–283). Cambridge: Cambridge University Press

Roberts, N.J. (ed). Short Stories in Italian. Harmondsworth, Penguin

Sanjivinie (1995). Itchykadana: Asian Nursery Rhymes. London: Mantra Publishing

Schmitt, N. (1997). Vocabulary Learning Strategies. In Schmitt, N. & McCarthy, M. (eds.) Vocabulary: Description, Acquisition and Pedagogy (pp.199–227). Cambridge: Cambridge University Press

School Curriculum and Assessment Authority (SCAA) (1994). Modern Foreign Languages in the National Curriculum: Draft Proposals. London: SCAA

Scrivenor, J. (1994). Learning Teaching. Oxford: Heinemann

Skehan, P. (1998). A Cognitive Approach to Language Learning. Oxford: Oxford University Press

Sinclair, J. & Coulthard, M. (1975). Towards an Analysis of Discourse. Oxford: Oxford University Press

Singleton, D. (1989). Language Acquisition: The Age Factor. Clevedon: Multilingual Matters

Stern, H.H. (1992). Issues and Options in Language Teaching. Oxford: Oxford University Press

Terrell, T.D., Rogers, M.B., Barnes, B.K. & Wolff-Hessini, M. (1993). Deux Mondes: A Communicative Approach. New York: McGraw-Hill

Tomasello, M. (1999). The Cultural Origins of Human Cognition. Boston: Harvard UP

Towell, R. & Hawkins, R. (1994). Approaches to Second Language Acquisition. Clevedon: Multilingual Matters

Tucker, G.R., Otanes, F.E. & Sibayan, B.P. (1971). An alternating days approach to bilingual education. In J.E. Alatis (ed.) 21^st Annual Round Table Meeting on Linguistics and Language Studies (pp.281–300). Georgetown U.P.

Ur, P. (1996). A Course in Language Teaching. Cambridge: Cambridge University Press

Usawa, K. (1997). Problem-solving in the translation processes of Japanese ESL learners. Canadian Modern Language Review. 53, 3, 491–505

Weinreich, U. (1953). Languages in Contact. The Hague: Mouton

Willis, J. (1996). A Framework for Task-Based Language Learning. Harlow: Longman

Other online papers