Language awareness and language learning

This article reviews Language Awareness (LA) as a field of research and practice. It deals with the period from 1990 to the present, asking what LA is, how it has been collectively constructed during this time, what the theoretical underpinnings might be and what it means in practical, methodological terms in the classroom and for society. It is recognized that its multidisciplinary nature and wide scope could lead to fragmentation, but it is argued that the holistic view evident in LA research and practice is a strength, and that its different sub-fields have certain core notions in common which give LA coherence. The paper begins with a brief background sketch and outline, and goes on to discuss the literature on cognitive aspects of LA, such as awareness, attention and noticing. The review then enquires into the characteristics of LA teaching methodology, and what LA is needed for teachers to implement it. Social and political perspectives are then explored in brief reviews of Critical Language Awareness, Inter-/Cross-cultural Awareness, and multilingualism. The paper closes by drawing conclusions and making suggestions for further research.


Introduction
The present review of Language Awareness (LA) focuses on the period from the 1990s to the present, i.e. from the time of such events as the publication of Language awareness in the classroom (James & Garrett 1991), the founding of the Association of Language Awareness (ALA) in the UK in 1992, the first issue of the journal Language Awareness in the same year, the publication of Schmidt's (1990) paper on consciousness in language learning and Fairclough's (1992) edited volume on Critical language awareness. Around then, the tide was starting to turn in favour of the teaching of form within meaning-based instruction, and the view that conscious knowledge facilitates language learning, which caused both research and debate (e.g., Doughty 1991;Long 1991).
The review draws mainly on English-language publications but is international in outlook. Research and practice in a number of countries and settings, relating to a range of languages, is discussed. Although most of it falls obviously within the field of LA, in some cases work of authors who have not themselves used the term LA has been included because it has had a major influence on LA researchers and practitioners or has salient characteristics in itself of an LA approach. The review draws on both theoretically oriented research and debate, and work based in classroom practice. The wide scope of the review does not allow the wealth of research in any one of the sub-fields of LA to be comprehensively covered, and some strands of research are not discussed at all. The aim has nevertheless been to present an overview of LA which is as representative as space allows, at the same time contextualizing and interrogating the field with particular questions in mind. It is hoped that the review will both serve as a useful introduction to readers not acquainted with the field, and contribute to a fruitful debate among LA researchers and practitioners.
The scope of LA is described by James & Garrett (1991) as covering five domains: affective, social, power, cognitive, and performance. As this indicates, LA straddles a cognitive to sociocultural spectrum and involves such apparently distinct areas of research and practice as cognitive linguistics (attention and awareness in language learning), language teaching, language use and intercultural communication (cross-cultural awareness) and this is reflected in the structure of the paper.
I will start with a brief background sketch of LA, followed by a discussion of research on consciousness, awareness and related notions, including a brief discussion of explicit and implicit learning and teaching. I will then move on to an exploration of an LA approach to language teaching, leading to a discussion of the LA of teachers. This is followed by an overview of critical language awareness, and then a presentation of work on intercultural language awareness and multilingualism. Throughout, the review will attempt to identify some of the atheoretical and ideological stances adopted in LA and the theoretical frameworks which underpin them. A summary and discussion of implications will be provided in the concluding section.
It is in the holistic nature of LA that language learning is not easily isolated from other objectives. Jessner (in press), who in addition to a more indepth discussion of LA and multilingualism provides Dr AGNETA M-L. SVALBERG lectures and supervises on Masters and Doctoral programmes in Applied Linguistic & TESOL in the School of Education at the University of Leicester. Before entering tertiary education, she worked as an EFL teacher in different parts of the world. She has published on Language Awareness, and on non-standard varieties of English, and has a particular interest in the learning and teaching of grammar. ■ a brief history and review of LA, points out that its theoretical coverage is vast. But what, exactly, is LA?
The National Council for Language in Education Working Party on Language Awareness declared in 1985 that 'Language awareness is a person's sensitivity to and conscious awareness of the nature of language and its role in human life' (Donmall 1985, cited in Thornbury 1997a; see also James & Garrett 1991, Hawkins 1999, James 1999. The ALA website currently provides the following definition: 'Language Awareness can be defined as explicit knowledge about language, and conscious perception and sensitivity in language learning, language teaching and language use' (emphasis as in original).
Although knowledge about language (KAL) is sometimes used as synonymous with LA (Stainton 1992;Van Lier & Corson 1997), it is also often associated with UK school contexts (Carter 1990;Hawkins 1992) and the early years of the British LA Movement (Hawkins 1999), and particularly with knowledge about grammar. In this review the term LA will be preferred. Near-equivalents in French prise de conscience de la langue/des langues,éveil aux langages oréveil aux langues (éveil referring to an initial stage of LA), in German sprachbewußtsein or sprachreflexion, and in Spanish conciencia lingüística. 1 Perhaps a good start to understanding what arguably gives LA coherence as a field is to consider just a few of the questions it might pose: 'Can we become better language users or learners or teachers if we develop a better understanding? And can we gain other advantages, for example, in our relations with other people and/or cultures, and in our ability to see through language that manipulates or discriminates?' (ALA website) The questions do not refer specifically to language learning and teaching contexts, but to all situations where language is used. For example, popular ideas about language, and their effect in society, such as in the school or workplace, are important LA concerns. LA is thus seen as having an importance and a value in itself, whether or not it facilitates language learning. 2 The founder of the LA movement in the UK, Eric Hawkins (see in particular Hawkins 1981particular Hawkins , 1984, recounts that ALA emerged both out of a concern about literacy levels in L1 and poor performance in the learning of foreign languages, and as a reaction to prejudice (Hawkins 1999). In the UK, the critical dimension of LA was very much present at an early stage when LA proponents' ideas were controversial, clashing with the political establishment and received 'wisdom' about language learning and teaching (Hawkins 1999;James 1999) as, for example, in the government-inspired but subsequently abandoned LINC project (Carter 1990). Among the aspects not appreciated at the time were the project's functional orientation and its views on non-standard English. However, perhaps as a sign of changing attitudes, the LINC materials have recently become available in electronic form. 3 In the 1990s, because the aims of language and literacy teaching were narrowly conceived as 'to produce a literate workforce', there was no place in UK schools for developing a critical understanding of language, according to James (1999: 96); see also a historical overview of LA in the UK by Donmall-Hicks (1997). Wright (1991) also identified resistance to LA, especially among English L1 teachers in the UK (see also Brumfit 1991). According to Mitchell & Hooper (1991), the teachers tended to equate LA with the teaching of grammar, to which they themselves had usually not been exposed. But from the 1990s onwards, there was a gradual shift in schools towards a greater focus on language structure, culminating in the present National Literacy Strategy (see the list of websites at the end of the article), launched in 1997 for primary schools, and in 2001 for secondary. It involves conscious learning about language, but does not commit to an LA approach. As will be evident below, in continental Europe the LA movement has been propelled less by literacy skills concerns and more by sociolinguistic issues. Research and practice often revolve around multilingualism, language attitudes, and citizenship.
The concerns which initially gave rise to the LA movement are not only still with us but are perhaps of even greater salience due to social and economic developments such as migration, globalization, international trade, and international conflict (Jessner in press). The increased interest in LA engendered by such issues has been further enhanced by important developments in the fields of linguistics, applied linguistics and language teachingparticularly in our understanding of the cognitive processes involved in language learning about which more will be said below.

Consciousness/awareness,attention and noticing
To unpack awareness 4 it is necessary to examine the second language acquisition (SLA) research, ■ Language awareness especially cognitive linguistic notions. I will also try to show very briefly how LA relates to the ongoing debate about form-focused instruction.
The best starting point is another elusive notion, attention (Schmidt 2001). According to Posner & Petersen (1990) attention consists of three elements: alertness, detection and orientation. A person who attends to something is by definition alert. Attention, in the form of detection, is a precondition for awareness. Awareness, according to Al-Hejin (2004), causes a change in behaviour or cognitive state and the person is able to report that they became aware and what they became aware of. Attention/detection can occur without leading to awareness. Despite the lack of agreement in the literature as to the precise meaning of attention (Segalowitz & Lightbown 1999), it seems to be generally accepted that some level of it is required for learning. In contrast, awareness is considered to be only facilitative.
Attention and awareness come together in noticing, a phenomenon given great prominence by the learner diary study first reported in Schmidt & Frota (1986) and which is central especially to formfocused LA instruction. Schmidt defines noticing as the 'registration [detection] of the occurrence of a stimulus event in conscious awareness and subsequent storage in long term memory' (Schmidt 1994: 179, emphasis in the original; see also critical review in Truscott 1998). To Schmidt, detection without awareness is simply registration. Schmidt (1990: 29) posits that 'noticing is the necessary and sufficient condition for converting input into intake'. He later modified his stance somewhat (see below) but, according to this strong Noticing Hypothesis, implicit learning is not possible as it occurs without awareness. Different types of linguistic features will need to be noticed depending on what is to be learnt. Pragmatic competence, for example, would require some awareness to be focused both on forms as such, and on their functional meanings in the social contexts in which they occur (Schmidt 1993). Tomlin & Villa (1994) disagree with Schmidt on the necessity of awareness for learning, and Al-Hejin (2004) also takes a more cautious stance, but concludes that both attention and awareness (and hence noticing) facilitate learning. This will be the assumption made in the rest of this paper.
Several authors point out that learners will not necessarily notice that which they need to learn next. VanPatten (1990) found that learners had difficulty noticing form and meaning at the same time and makes the general claim that learners notice meaning before form. A study by Lee (2007) indicates that there is a trade-off between focus on form and focus on meaning. A conclusion one can draw is that form with shades of meaning provided by context. It is thus assumed that LA can be (but is not necessarily) the result of consciousness raising.
needs to be in some way foregrounded for learners to direct their attention to it. One way of doing this is by means of input enhancement (see below). Another is by removing any context which could obviate the learner's need to process formal linguistic elements. In VanPatten's Processing Instruction (PI) approach,sentences are decontextualized, including the removal of clarifying adverbials, and presented in tasks which force learners to attend to form in order to extract meaning (VanPatten 1996). VanPatten argues that this leads to a change in learners' processing strategies which endures beyond the task itself and which facilitates learning of the processed form (VanPatten & Sanz 1995;VanPatten 2002b). This is a controversial claim and has caused some debate (e.g., VanPatten 2002a;DeKeyser et al. 2002). In a comparative study, Toth (2006) did not find PI more effective than communicative output.
From a sociocultural perspective, Batstone (2002b) also takes issue with VanPatten's PI approach arguing that co-text, rather than being a hindrance to learning, helps to clarify the meaning of form and thus avoids putting the low level learner in the artificial and impossible position of guessing meaning from form without any clues (see response in VanPatten 2002b). Batstone's concern focuses on the affective dimensions of the learning environment. In Batstone (2002a) he criticizes, from this vantage point, the limitations inherent in communicative tasks. He claims they ignore the learners' inner context in favour of creating an external context where appropriacy is paramount. The stress of trying to produce relevant and coherent output which achieves its communicative ends, and the possibility of losing face if it fails, creates an environment where attention cannot be paid to form and which is thus an obstacle to learning. Batstone (2002a) suggests that teachers should rather discuss the reasons for communicative tasks explicitly with students, to help create a learning culture which is supportive, not face threatening, and which would allow features such as hesitation, repetition and risk-taking. It is argued that detachment from normal communicative norms in favour of a learning focus could facilitate learners' noticing of links between context and form.

Explicit and implicit knowledge and learning
The role of explicit language learning and knowledge has been discussed and debated extensively in the SLA literature as, for example, in a special issue of AILA Review (Hulstijn & Schmidt 1994), the edited volume by N. C. Ellis (1994), and the article by N. C. Ellis (2005). In this section I will merely attempt to clarify its central role in an LA approach, and highlight some of the complexities.
While cognitive psychologists assume awareness to be beneficial, DeKeyser (1994: 83) points out that ■ 'in applied linguistics the burden of proof seems to be on those claiming explicit learning works'. The debate includes Norris & Ortega's (2000) synthesis and analysis of 49 published studies, which concludes that on the available evidence explicit instruction is more effective than implicit instruction, and the critical reanalysis of their work by Doughty (2003), who finds the evidence insufficient (see also Norris & Ortega 2001). The LA stance in this debate is clearly for the effectiveness of explicit learning. Even so, a number of questions remain to be addressed, such as what kind of explicit knowledge is most facilitative, and in relation to what kinds of language features.
Both the level and the quality of awareness appear to be important factors. Abu Radwan (2005) differentiates between awareness as noticing, and awareness at the level of understanding. His and other studies (Leow 1997;Robinson 1995Robinson , 1997Rosa & O'Neill 1999) show learners with higher levels of awareness of the target features performing better than those with lower levels of awareness. How the awareness/explicit knowledge comes about also matters. In Robinson's (1995) study, learning was more likely to occur when the learners themselves had arrived at the rule they verbalized than when it had been provided as part of the instruction.
Another important question is which particular language features should be the target of awareness raising. Gass, Svetics & Lemelin (2003) claim that focused attention has the most effect in more complex areas: for example, more on syntax and least on lexis, where incidental learning seems to be most effective. By way of explanation, Hulstijn & de Graaf (1994) posit that learners can work out simple rules for themselves but that complex rules benefit from explicit instruction, and Gass et al. (2003: 528) concur: 'With more complex rules of grammar, internal devices are insufficient for learning, and focused attention . . . may be a necessary crutch'. The need for focused attention is implied also by Berry (1994), who introduces the notion of saliency and concludes that explicit learning seems to be beneficial for salient features, or features which can be made salient by instruction or materials, while it may be counterproductive for non-salient features; less obvious relationships are instead likely to benefit from a more passive, implicit mode of learning.
The usefulness of explicit instruction also depends on learner characteristics. In Robinson's study, learners with good memory were better able to make use of a rule provided, while more grammar sensitive learners benefited from rule searching (Robinson 1995). Individual learning styles might even subvert task design. As Berry (1994) points out, learners are likely to use a combination of explicit and implicit learning. Robinson (1995) found that some learners reported looking for rules although their learning conditions were meant to be implicit and incidental, while learners who had been instructed to look for rules sometimes did not. Learners' readiness will also intervene, so that more advanced learners might benefit from consciously attending to features they were previously unable to process (Gass et al. 2003: 529).
Despite continued debate among SLA researches (e.g. Doughty 2003), the value of explicit learning to facilitate knowledge, not only of the explicit type, but also the procedural knowledge learners need to become fluent users of the language (N. C. Ellis 1993;Robinson 1997;Housen, Pierrard & Van Daele 2005) is implicit in LA teaching methodology, despite continued debate among SLA researchers (e.g., Doughty 2003). An ongoing issue in LA is, however, how best to stimulate and channel LA. To inform this debate, more research which contrasts the learning effects of different approaches, such as Toth (2006) and Lee (2007), would be particularly useful.
Most of the studies discussed above have dealt with the learning of grammatical form. However, it has been claimed that explicit learning within an LA framework is facilitative in a variety of linguistic fields. This will be discussed below.

Language awareness and teaching methodology
Though it is recognized that some learners are likely to benefit more than others from any given approach (Jones 1997), the point of departure for LA practitioners is that developing a better understanding of the language and of learning/teaching processes will generally enhance language learning/teaching and use. This section deals with what form such awareness raising might take and what assumptions it makes. I will first explore whether LA classroom practice has particular characteristics which might justify talking about an LA approach to language learning/teaching. Secondly, I will briefly review the practical impact of LA on skills teaching. Teaching within a critical language awareness (CLA) framework will be mentioned briefly here but reviewed in more depth later.
3.1 Is there an LA approach to language teaching/learning? One could perhaps argue that any approach to language instruction which aims to raise conscious awareness of how linguistic systems work is an LA approach but in the literature it tends to have certain more specific characteristics. In a discussion of LA in relation to teacher education, Borg (1994: 62, building on Wright & Bolitho 1993 outlines five main features of an LA methodology, summarized here: 1. It involves an ongoing investigation of language as a dynamic phenomenon rather than awareness of a fixed body of established facts.

Language awareness
2. It involves learners in talking analytically about language, often to each other. 3. It considers essential the involvement of learners in exploration and discovery. 4. It aims to develop not only the learners' knowledge about and understanding of language but also their learning skills, thus promoting learner independence. 5. The aim is to involve learners on both a cognitive and an affective level.
While Wright & Bolitho's (1993) paper focuses more specifically on linguistic awareness, Borg (1994) emphasizes the broader educational aims of LA teacher education and LA methodology. Below I will expand on the LA methodology issues in relation to the five characteristics listed.
1. Borg's first feature implies a constructivist view of knowledge. From this follows a learner-rather than teacher-centred approach, involving and indeed relying on extended opportunities for learnerlearner interaction (van Lier 1996;Swain & Lapkin 2002). Borg (1994) calls for learner investigation of language to take the form of exploration and discovery, which often entails alternative answers and perhaps ambiguity. The open-endedness of language as a salient characteristic of everyday interaction, conversation in particular, is addressed by van Lier's work on contingency, where he recommends allowing and creating opportunity for learners to experience and deal with the unplanned, unexpected nature of linguistic interaction (van Lier 1992(van Lier , 1994(van Lier , 1996. At this end of the spectrum, LA work is thus radically distinct from traditional explicit language instruction and can either clash with student and teacher expectations or be perceived as challenging in a positive way (Svalberg 2005).
2. Borg's (1994) second feature has been described by Svalberg (2006) as the talk-about element. It differentiates LA methodology from, for example, the strong version of communicative language teaching in which the emphasis is exclusively on meaning (Howatt 1984). As pointed out by Kramsch (1993: 246): 'Talk about talk is what the classroom does best and yet this potential source of knowledge has not been sufficiently tapped, even in communicativelyoriented classrooms'. From a constructivist perspective, Swain (2006) employs the term languaging for the 'use of language to mediate cognitive activity', and posits that one of the ways of learning a second language is languaging about language (Swain 2000). 5 In Tocalli-Beller & Swain's (2005: 8) view 'Metatalk . . . mediates second language learning because it supports the process of appropriation'.
Languaging naturally involves the use of metalanguage, either as formalterminology, or as informal ways of talking about language structure (Alderson, Clapham & Steele 1997;Berry 2005 special issue of Language Awareness on metalanguage). Hence, LA practitioners need to choose how much and what kind of linguistic terminology to use, and perhaps teach. There is a substantial body of studies on the metalinguistic knowledge and metalanguage of teachers (Mitchell & Hooper 1991;Andrews 2003;Berry 2004) and learners (Fortune & Thorpe 2001;Fortune 2005) but no solid evidence that knowledge of metalanguage as such facilitates language learning. Neither the quantitative study by Alderson et al. (1997) nor Robinson's qualitative investigation (Robinson 2005) was able to establish a relationship between linguistic ability and knowledge of metalanguage. If reflection scaffolded by verbalizing facilitates language learning, formal metalanguage may be a useful tool but informal metalanguage might do the job just as well.
3. The process of talking about and making explicit the workings of language requires learner interaction and engagement, the third feature in Borg's list. There is a body of LA research on the forms of learnerlearner interaction and its learning effects (Fortune & Thorp 2001;Storch 2002;Swain & Lapkin 2002) and also on teacher-learner interaction. In the curriculum outlined in van Lier (1996), awareness, autonomy, and authenticity are key notions, and learner engagement is central. Authenticity refers to a process of making interaction and text authentic for the participants, and may or may not involve the use of so called authentic materials (see also Maun 2006 for a discussion of authenticity). Van Lier (1998: 128) states that 'interactions with learners in classrooms should allow learners to be perceiving, thinking, acting, and interacting persons, rather than passive receivers of knowledge'. Sarangi (1998) also emphasizes the importance of the social and interpersonal aspects of teacher-learner interaction, as do Kress, Ogborn & Martins (1998). Nassaji & Swain (2000) provide support for the effectiveness of consciousness raising feedback, negotiated between teacher and learner, and taking the learner's stage of development into account.
4. In line with the ALA definition of LA, Borg (1994) emphasizes the desirability of also developing awareness of learning/teaching processes as a means of improving teaching and enhancing learner independence (the fourth characteristic). Ewald (2004) reports on a study in which learners explored their own beliefs about and behaviours in group tasks, and how the resulting awareness led them to change behaviours, and potentially to improve learning skills and gain greater learner independence. The author stresses the importance of ■ the learners' insider perspective on language teaching methodology. Other authors (for example, Walsh 2003) discuss how teachers can develop their own awareness of interactions with students (see below). Borg's (1994) LA features seem to leave open the precise implementation of an LA methodology but feature five, which emphasizes that learner engagement needs to be both cognitive and affective, seems more compatible with some pedagogies than others. In areas such as cross-cultural communication (see below), affective issues such as those related to identity are naturally central but Borg implies that affective engagement is essential in all LA work, including grammar and skills work. Robinson (2005) recommends that teachers seek the pupils' perspectives on how they approach writing, and that the affective dimension of writing awareness be taken into account. In a study into the use of authentic and adapted texts, Maun (2006) found that young adults perceived texts as more or less difficult, dense, inviting and so on depending on the visual impact created by formatting, layout and illustrations. Learner engagement in an ideal LA classroom clearly goes well beyond task enjoyment, as is evident also in van Lier's (1996) discussion of an LA curriculum, where some of the key phrases are conscious engagement, depth of processing and commitment to learning.

Most of
In LA work, stimulating affective engagement tends to rely on communicative purpose and meaningfulness of tasks. Wright & Bolitho (1993) point out that among the properties LA aims to develop in learners is a sensitivity to meaning-form connections. It may be, as Thornbury (2001: 38) claims, that 'unless the learner notices the effect that grammatical choices have on meaning, . . . the noticing is not sufficient to have any long-term effects on restructuring'.
To sum up, Borg's five features seem to identify the unifying characteristics of LA methodology which might be reformulated as: description (not prescription), exploration, languaging, engagement and reflection.
Methods and techniques more commonly associated with the LA classroom are, for example, input enhancement; discovery-type, inductive tasks; dictogloss and text reconstruction; and open-ended discussion tasks on authentic or adapted texts (Wright & Bolitho 1993;Thornbury 1997aThornbury , b, 2001. In general, approaches and techniques which make use of or engender conscious knowledge and which stimulate engagement with the language in a specific context, within a constructivist framework, are consistent with an LA pedagogy. A term used to refer to some LA tasks is consciousness raising (Hopkins & Nettle 1994;Hedge 2000;R. Ellis 2002). As Ellis points out, it differs from most other language learning activities in that it does not involve learners in producing output for the purposes of practice (see Svalberg 2005 for examples and discussion). Instead, it denotes tasks which aim to focus learners' attention on particular linguistic features and raise their awareness of how these features work (Thornbury 2001). 6 Borg connects discovery-type, i.e. inductive tasks, with LA methodology and although a few authors consider that consciousness raising can be either inductive or deductive (Mohamed 2004) the inductive type is the most common (Wright 1991;Wright & Bolitho 1993;Thornbury 2001).
Specific techniques, such as dictogloss and other text reconstruction tasks, which have an inbuilt component of student interaction and encourage learners to arrive at and justify their own solutions, are frequently used in the LA classroom (for example, Lasagabaster 2002) and have been the subject of research. Pica (2005) discusses the rationale for task choice and shows samples of how students negotiate meaning and form, as in text reconstruction. García Mayo (2003) compares the effect of multiple choice, dictogloss (involving full text oral input), text reconstruction (involving fragments of written input) and text editing tasks on the quantity and quality of Language Related Episodes (LREs) in pair interaction (see also Fortune & Thorpe 2001).
As already discussed, a starting point for languaging about language is noticing. Sharwood-Smith (1993) uses the term input enhancement for ways in which teachers or materials make particular language features salient in order to promote noticing. The term is sometimes used more specifically to refer to typographical enhancement, i.e. manipulation of fonts such as bolding and underlining. In a review of some of the research on input enhancement, Doughty & Williams (1998) classify it as the most unobtrusive of a range of form-focused instructional techniques. Research on its learning effects has, however, not produced clear cut results. Neither Alanen (1995), White (1998) or Abu Radwan (2005) found any significant learning effect of typographical enhancement. On the other hand, Shook (1994, discussed in Gass et al. 2003 found it had a significant effect on present perfect and relative pronouns, and in Jourdenais et al.'s (1995) study enhancement seemed to produce both an increased quantity and greater accuracy of the target feature in immediate and delayed post-tests. Important factors may be the complexity of the target feature, as discussed above, and the research design. For example, in White (1998) the non-specific nature of the target feature (3rd person pronouns/determiners) is likely to have made enhancement less effective, and think-aloud ■ Language awareness protocols could have had a priming effect in the Jourdenais et al. (1995) study.
In the classroom, input enhancement is not normally used on its own but as one element in a more complex task design, for example, as a starting point for languaging. Learners may be provided with already enhanced input, but more frequently they are asked to carry out text enhancement themselves by, for example, underlining features they identify (see example tasks in Hall & Shepheard 1991;Thornbury 2001). One might argue that the effect of input enhancement should therefore be evaluated in that context.
In summary, the picture of LA pedagogy which emerges is one which aims to engender LA by learner engagement with the language, including the use of languaging, to construct knowledge about the language in any of its domains affective, social, power, cognitive, and performance. I will return to a discussion of the power domain in the section on CLA. The next two sections will review LA as it has been applied to the learning/teaching of the so called four skills.

Listening and speaking
McCarthy & O'Keefe (2004) point out, in a review of research in the teaching of speaking, that important advances have been made in our understanding of spoken language. It is striking therefore how much less LA work has been done on speaking and listening compared to reading and writing. Nevertheless, there are a number of interesting studies. To start with listening, Kim (1995) approaches it from a cognitive perspective, asking what features L2 learners are able to attend to in streams of speech. The study thus deals with awareness at the level of noticing. For the participating Korean University students, noticing was facilitated by phonetic prominence (for example, primary or secondary stress), and by initial or final position in an intonational phrase. Rate of speech also had an effect so that more words were noticed at a slower rate. Kim proposes an implicational scale of aural processing by which learners develop from being unable to identify even phonetically prominent key words, through to key word identification, to words surrounding key words in a phrase, to the grammatical/semantic relationships of clauses, and finally an ability to perceive relationships between clauses. Rather than each learner being at a specific stage, Kim suggests that learners function within a range of phases depending on features of the text and its delivery (for example, its pace). The author speculates that slowing down input, thus making more of it available for processing, could push learners from lexically to more syntactically driven processing. However, research is needed to test this, and also whether raising EFL learners' conscious awareness in this area would be beneficial for listening comprehension.
Simply drawing learners' attention to prosodic features may not be enough to improve listening comprehension and intake. Like Kim (1995) above, Pennington & Ellis (2000) investigated learners' aural cognition and found that the Cantonese speakers in their experimental study did not recall prosodic information of the English sentences to which they had been exposed. They improved slightly when their attention was explicitly drawn to intonation, but only for sentences with contrastive versus neutral stress.
Combining noticing with explanation may be more effective. A computer assisted, quasiexperimental study involving Spanish speaking EFL learners (Ramirez Verdugo 2006) explained pitch contours of native English speakers to the learners and contrasted them with the learners' own production. The overall pedagogic framework built on VanPatten's processing instruction framework and structured input activities (1996, 2002b). As an important part of the training, pitch contours were displayed on computer screens. The intervention resulted in both heightened awareness of the role of intonation and improved prosodic performance. The authors recommend research into whether such training results in more long term independent noticing and learning of prosodic features.
Computer technology was used also by Coniam (2002) in an approach to increasing non-native speaking teachers' awareness of suprasegmentals in English. The software allowed digitization of audio recordings, producing displays in wave form. The trainees matched the displays with utterances and speakers (native English speakers and Hong Kong Chinese), and the qualitative differences between speakers were discussed in class in terms of rhythm and timing.
The studies above dealt with listening and pronunciation skills. In contrast, Nakatani (2005) investigated the effect of awareness raising on young, Japanese adults' use of oral communication strategies (OCS) such as maintenance of fluency and negotiation of meaning. The author attributes a significant improvement in test scores for the trained group to greater awareness of OCS and consequent increased use of such strategies. The trained group produced longer utterances and used more achievement strategies, and did not abandon the message as often as the control. Slade (1997) and Jones (2001Jones ( , 2002 advocate explicit instruction in conversational story telling. Jones (2001) describes a consciousness raising approach in some detail. The learners compare a written narrative and a transcript of a (non-authentic) oral narrative. They then answer questions on the transcript, designed to draw their attention to generic structure and features of spoken language, before engaging in their own story telling.

Reading and writing
Research on LA and the four skills includes a large body of work on young children's literacy development in L1 and L2. A review of earlier research on metalinguistic awareness relating to ■ different cultures and writing systems, to evaluate claims about how children learn to read and write, is to be found in Downing (1986). It concludes that LA literacy work needs to take children's metacognitive development into account, and points out the need for research in a range of linguistic and cultural contexts. Francis (1999Francis ( , 2002 studied the metalinguistic awareness of Náhuatl/Spanish bilingual children in Mexico, and its relationship to literacy development. The ability to discriminate between two language codes, and to separate them, seemed to be linked to development of reflective, metalinguistic ability. The 2002 study suggested that metalinguistic awareness might be particularly facilitative in relation to writing. Findings from studies into the effect of phoneme awareness on literacy development are not unequivocal. In a quantitative, longitudinal study by Layton, Robinson & Lawson (1998) phonological awareness training of pre-schoolers in Australia failed to have a significant effect on reading and writing. Nor was the reading performance of the 8-10 year old children in Layton et al.'s study improved by the syntactic awareness they had gained through explicit instruction. In contrast, Elbro & Petersen (2004) found phonological awareness training of kindergarten children of dyslexic parents effective. The children received training for 17 weeks and it was found that the treatment had long-term benefits as late as grade 7. The contradictory results in regard to awareness training seem to confirm that the quality of awareness is significant. Teacher-led, explicit instruction and learner-centered discovery type tasks could potentially produce qualitatively different knowledge, as could the frequency and density of the input, and the length of treatment. Different quality of noticing may be an essential feature in terms of the reflections and conclusions it is able to stimulate in the learner, as implied by Tocalli-Beller & Swain (2005). A small-scale study reported by Qi & Lapkin (2001) suggests that quality of noticing is in turn influenced by L2 proficiency. In reformulation exercises, lower level learners might, for example, find it harder to establish the nature of a noticed gap or be less ready to accept a reformulation than more advanced learners.
Contradictory research findings could also be due to research methodology. In a critical review, Troia (1999) lists a number of experimental studies involving phonological awareness training of children, which claim that phonological awareness correlates highly with reading and spelling achievement, and examines the methodology employed in these studies. The review concludes that there is a trade-off between internal validity and ecological validity. Normal classrooms are not controlled or controllable environments and, by the same token, findings in controlled environments do not necessarily apply to classroom settings.
Literacy development may be influenced not by one particular skill but by a range of factors in combination. For example, in the case of the pre-school child, metalinguistic awareness and pre-conventional reading and invented spelling development could work together (Lazo, Pumfrey & Peers 1997). In a study with Spanish pre-school and 1st grade children in Spain, Goikoetxea (2005) claims that there is a link between pre-reading spelling and phonological awareness development. In some school contexts, linguistic background could also be a factor. In a recent study, Lesaux, Lipka & Siegel (2006) found that native English-speaking children achieved significantly higher scores on syntactic awareness and verbal working memory than ESL children. Whether this advantage could be partially offset by syntactic awareness training specifically for ESL students remains to be investigated. A very different, and interesting, aspect of children's literacy is addressed by Kümmerling-Meibauer (1999) who discusses what an understanding of irony in textpicture relationships requires native English-speaking children to know or be made aware of.
In academic writing instruction, a genre-based approach is popular among LA practitioners. Distinctions made in Tolchinsky (2000) seem to imply that it is inherently prescriptive, but this need not be the case. As shown in Henry & Roseberry (1999), and Clynes & Henry (2004), raising awareness of genre can in fact mean that learners are trained to apply genre analysis to text themselves. Johns (1997: 92-105) dedicates a whole chapter to the student as researcher in what she calls a socioliterate genrebased approach to academic writing. Recognizing the difficulties of the analyst role for non-native speakers, Clynes & Henry (2004), in line with Johns (1997), attempted to implement training on a relatively simple genre, anchored in local culture -the Brunei Malay wedding invitation. The idea was that the students would benefit from being able to draw on already existing schemata. The training enabled the students to identify moves successfully, but they were less able to establish and formulate the purpose of moves within the text, and the overall communicative purpose. Careful scaffolding seems to be necessary.
Common awareness-raising techniques in the teaching of writing are dictogloss (Thornbury 1997a;Lasagabaster 2002), discussed above, and reformulation. In the latter, the teacher recasts the student's written text to make it more target-like, while respecting the student's intended meaning. This production-first-and-model-second sequence is the reverse of more traditional instruction, where teacher-led input is followed by production (fluencyto-accuracy vs. accuracy-to-fluency; Thornbury 1997b: 328). Thornbury points out that working with whole texts in this way allows consciousness raising on a number of levels, including discourse, ■ Language awareness syntax, lexis, within a form-focused or meaningfocused syllabus. Reformulation has been investigated in a number of studies (Qi & Lapkin 2001;Swain & Lapkin 2002;Adams 2003). As part of their body of work on output, Tocalli-Beller & Swain (2005) studied how French immersion students' noticing the gap between their own written output and the reformulated text generated cognitive conflict episodes, and also the learning effect of these episodes. They found that cognitive conflict, more than agreement, stimulated restructuring of the learners' knowledge. Emphasizing the distinction between cultural or affective conflict on the one hand, and cognitive conflict on the other, they assert that learners are likely to benefit from the latter if it gives them opportunity to talk about the errors and how to correct them. Similarly, Rijlaarsdam & Couzijn (2000) conclude in relation to peer teaching of writing in secondary school, that activities which required students to reflect on their own and their peers' writing led to heightened awareness and helped them to revise their work. In this study the reflection did not, however, take the form of verbal interaction. Instead, the learners provided written feedback on their peer's writing, and self-reflected on this and on the feedback they received on their own work. They also wrote learner reports for the purposes of the study.
The examples above indicate that an LA approach can facilitate learning in different contexts, at different ages, and for a range of purposes, from early literacy development to academic writing and enculturation. Writing instruction within a critical pedagogy framework will be addressed in the section on CLA. The next section will examine the potential usefulness of teachers implementing an LA approach, and how it might be achieved.

Teachers and language awareness
In recent years there has been an increased emphasis world-wide on language teachers' content knowledge (Andrews & McNeill 2005: 160). Thornbury (1997a: xii) quotes from the Examination report for the British-based RSA/UCLES Diploma in TEFL examinations for the year 1991/1992: 'It is a matter of concern that so many teachers of English seem to have such a limited knowledge of the language they are teaching.' In the UK, a mandatory Literacy Test for Primary Teachers, who teach English L1, was introduced in 2001 (Hislam & Cajkler 2005). One can perhaps assume that much of the knowledge expected of teachers by Education authorities worldwide relates to grammar but, as the discussion below will show, teachers need a much broader range of explicit knowledge and sensitivities. 7 7 See also Andrews (forthcoming) on teachers' LA and special issue of Language Awareness on 'Teacher language awareness' Teachers' LA is said to be 'the knowledge that teachers have of the underlying systems of language that enables them to teach effectively' (Thornbury 1997a: x). This indicates that whatever the teaching approach, teachers' LA is different from that of other expert language users; they need not only to know about the language, but also to reflect on their knowledge and on underlying systems (Andrews 2001(Andrews , 2006. Regarding the extent of teachers' LA, teachers thus take on three roles in regard to language, those of user, analyst and teacher (Edge 1988, cited in Cots & Arno 2005. Hence, they require a high level of language proficiency, plus linguistic and pedagogic knowledge (Wright 1991(Wright , 2002. Andrews & McNeill (2005), in a study of three good language teachers, in Hong Kong and the UK, attempt to identify what characterizes such teachers' LA. The good language teachers in the study had a willingness to engage with language content, a desire for self-development, an intuitive understanding of the importance of noticing and the use of input enhancement, and an understanding of learners' difficulties helped by their confidence in their own LA, and paired with an awareness of their limitations in this regard (see also Hislam & Cajkler 2005).
The trend towards a sociocultural and discourse perspective on language has added to the complexity of content knowledge that language teacher education wishes to achieve. Contributors to the volume edited by Trappes-Lomax & Ferguson (2002) suggest that teachers need heightened LA about the nature and value of languaging; the social dimensions of language (for example, varieties or identity); language as (pedagogic) subject and as object; the reflexivity of learner language; and what discourse provides the best learning opportunities (see also the edited volume Hawkins 2004). For teachers to make informed choices regarding the teaching/learning of pragmatics, Kasper (1997) suggests that they need to be aware not only of native speakers' use of language but also of the pragmatics of the classroom, including cross-cultural issues. In typical FL contexts, such an approach would seem particularly well justified. In Brazil, a brief course on pragmatics failed to produce a change in teachers' classroom practices (Lana Chavez de Castro 2005) possibly because pragmatics had not been discussed in relation to naturally occurring classroom situations. Van Lier's work on contingency (1992,1994) suggests that more opportunities for social and interpersonal communication need to be created in the classroom. In that regard, Kress et al. (1998) make the point that language is normally accompanied by (or accompanies) other modes of communication, such as objects, gestures, or images. Such spontaneous (Andrews 2003), which includes a literature review on teacher cognition (Borg 2003). ■ interaction, however, places very particular demands on the sophistication of the teacher's LA, and presents interesting challenges for LA teacher educators.
How can such complex LA be promoted effectively in teacher education? Many authors stress the importance of integrating the three elements of user, analyst and teacher knowledge on teacher education courses. Borg (1994) argues against a strongly linguistic orientation and for an integration of pedagogical issues. Andrews & McNeill (2005) likewise call for a balance between linguistically oriented training and methodology, and suggest that LA development opportunities should be available at all stages of a teacher's career, including by selfaccess. The argument for an integration of pedagogy and linguistics is also supported by contributions to Bartels' (2005) edited volume, which explore whether applied linguistics knowledge once acquired is or is not transferred into the language classroom. It becomes clear that knowledge transfer cannot be taken for granted. In Burns & Knox's (2005) study, for example, MA students' training in systemic functional grammar did not have the expected impact on their teaching. Summing up the findings in Bartels (2005), the editor concludes that LA is more likely to transfer to classroom practice if there is task similarity between the teacher training and the classroom experience (Bartels 2005a). In this light, the findings in Andrews & Bunton's (2006) study are not surprising: a relevant degree as defined by the Hong Kong government, such as in Linguistics, was not a good predictor of teacher LA, particularly when it came to preparing language teaching materials. Borg (1994) recommends that the training approach should reflect the practical and reflexive nature of the LA methodology it promotes. More recently, contributors to Cots & Nussbaum (2002) show how one might work analytically with naturally occurring and contextualized language in language teacher education. The volume contains analysis and discussion of a range of sample activities designed and used in Spain.
Teachers' LA and their previous experiences of language learning and/or teaching have a major influence on their pedagogic choices (Borg 2005;Cots & Arno 2005). As already mentioned, UK teacher trainees tend to have a low level of LA (Cajkler & Hislam 2002) and have had little exposure to form-focused language teaching. On the other hand, in countries where teacher trainees' awareness of form is generally high, the inductive, exploratory nature of LA work often goes against their previous experience and expectations. For both groups, the process of change from user to analyst, during which already held knowledge and beliefs are challenged, can be difficult and destabilizing (Wright 1991). Hislam & Cajkler (2005) stress that trainees need time to reflect on and absorb LA, and opportunities to try out methodologies in a safe environment. Burley & Pomphrey (2002 report on an intercomprehension approach to teacher development in which teachers of English and of Modern Languages learnt from each other. The dialogue across subjects changed the teachers' perceptions so that some of the English teachers benefited from having their literature focused views of English enriched by the Modern Languages teachers' more linguistic approach to language, and vice versa.
In this section, a discussion of what characterizes an LA approach to language teaching concluded that description, exploration, languaging, engagement and reflection were salient features. The methodology was then explored in terms of classroom techniques, and skills teaching. A discussion of teachers' LA found that language teachers in general need awareness in a wide range of areas, but that LA as a teaching approach makes particularly challenging demands in this respect.
In the following two sections, the importance of considering the social, cultural and political contexts in which LA work takes place will be explored.

Critical language awareness
The exploratory approach in which language is 'a dynamic phenomenon rather than . . . a fixed body of established facts' (Borg 1994: 62, as discussed above) is fully exploited in CLA, which deals with language as discourse, i.e. with verbal interaction as social practice.  consider discourse from three perspectives: social context (situationinstitution -society), sociocognitive processes (production -interpretation) and text (spokenwritten). Rather than simple, conscious knowledge, the awareness CLA seeks to engender is the socially situated process which Freire (1982, cited in Clark et al. 1991 called conscientisation (see also Freire 1985).
Hence, just as the subject matter of CLA is wider than more linguistically oriented LA, so are its aims and objectives. Other LA approaches are criticized by CLA proponents for presenting the existing sociolinguistic situation and ideologies embedded in the discourse as 'natural', thereby contributing to their perpetuation rather than, as CLA aims to do, to social change (Clark et al. 1990Fairclough 1992a). A particular focus of interest and analysis is how power relationships construct, and are constructed by, discourses. For example, Corson (1997) discusses the ideology of correctness prevalent in schools which helps to maintain dominant/non-dominant social structures. The notion of appropriateness might seem more socially aware than correctness but is nevertheless rejected by Fairclough (1992a), who claims that rather than providing learners with opportunities, ■ Language awareness appropriateness cements the hegemony of a particular variety. The model is said to mask the complexity of how, when and why varieties are used and the fact that conventions are constantly contested by individuals and groups. In contrast, CLA aims to empower learners to make meaningful choices, including ones that introduce change.
The social change and emancipation sought by CLA practitioners is towards greater equality and democracy. The wide remit of CLA is evident in the Norton & Toohey (2004) edited volume, which includes papers on language and gender (Pavlenko 2004), language testing (Shohamy 2004) and the teaching of form (Morgan 2004) from a critical perspective. Another important source for understanding CLA aims is Fairclough's edited volume (1992). In it, the editor himself argues that the nature of contemporary society makes CLA more necessary than ever in order to create citizens for an effective democracy. It is seen, by Janks & Ivanič (1992: 305), as contributing to emancipatory discourse i.e. 'using language, along with other aspects of social practice, in a way which works towards greater freedom and respect for all people, including ourselves.' In the political arena, Fairclough (1999) identifies a narrowing down of political discourse as party political agendas become less and less diverse. In a healthy democracy, access and input to political discourse, and some control of where it goes, needs to be as broad as possible; effective participation, he argues, requires critical awareness of the discourses involved. Wallace (1997: 242) calls CLA 'the pedagogical arm of critical discourse analysis', situating it within a framework of Critical Pedagogy (Wallace 1999) which, as she points out, draws both on educational and social theory (writers such as Apple, Bernstein, Freire, Foucault, Giroux and Habermas), and on linguistics (for example, Clark, Halliday, Fairclough, Fowler and Kress) (Wallace 1997: 305; see also Phipps & Guilherme 2004). CLA practitioners clearly consider that it is both important and possible to carry out their potentially subversive work from within mainstream education. Fairclough (1999) is highly critical of the focus in Higher Education on the transmission of skills, such as communication skills, in the narrow service of economic needs. In a similar vein, Sealey (1999) critiques National Literacy Strategy documents for English primary schools for their lack of acknowledgement of the social and political nature of language and discourse, which a CLA approach could provide. In a Language Awareness special issue on CLA (Clark & Ivanič 1999), Fairclough (1999 addresses the challenges and opportunities of the 'new global capitalism' with its inherent discourse of flexibility, and argues that a critical awareness of the role of discourse in social practice is required in such a society, not only for personal success but also in order to drive social change; it is the role of language education to foster such awareness. An example of how discourse can simultaneously engage and marginalize learners in a culturally diverse classroom is found in Duff (2004), who emphasizes the need for greater awareness of such issues among teachers and learners. Another recurring theme in CLA is the importance of a historical perspective (Janks & Ivanič 1992). Bhatt & Martin-Jones (1992) point out the dangers of well-intentioned but uncritical awareness work in multilingual classes and stress that unless they are given a historical and social dimension some activities risk presenting minority languages as exotic. They recommend that discussions include how language attitudes and values are formed socially and politically, and how minority languages and their speakers are represented.
Similar issues are raised by research on dialect awareness, including social and geographical varieties. Central is the dominance of English over other languages, or Standard English over less privileged varieties. Corson (1997) discusses the effect of using different sociocultural and geographical varieties of English, particularly in school settings, and calls for more explicit discussions of power and social justice issues in teacher education. Children should be empowered to make choices based on an understanding of social contexts. Some modest progress in the fostering of CLA by schools is suggested by Smitherman (2004), who claims that attitudes towards African American Language have become more positive. The Afro-Caribbean Language and Literacy Project (ILEA 1990) produced materials for multilingual classrooms which aimed both to raise awareness about power issues and language, and to help students achieve their best in written Standard English. The material has been published (ILEA 1990) and an informative review is available on the web (Harris 2004). An American perspective is presented by contributors to Villanueva & Smitherman (2003), an edited volume which emerged from a Language Knowledge and Awareness Survey research project.
Discourse, the subject of inquiry, is constructed by and serves to construct social relationships and identities, and so CLA is inherently constructivist (Norton & Toohey 2004: 1). This is reflected in an emphasis on students' own experiences, social settings and perceptions.  suggest that in the classroom, descriptive language knowledge and skills, and critical awareness of social relationships and power, need to be brought together in purposeful discourse. In an example activity, children discuss what kinds of people are or are not included in history books, such as people they know, which leads into writing a family history. Interesting examples of CLA pedagogy, which include purposeful discourse (though the term is not used) are Janks (1991) on a CLA approach to teaching grammar in South Africa, ■ and Morgan's (2004) account of teaching modality to recent Chinese immigrants in Canada. Morgan considers the grammar lesson a site where identity is not only represented but also constructed, and at the same time manages to demonstrate the place the teaching of form might take in CLA (see also Carter, Hughes & McCarthy 2000;Goatly 2000;Wallace 2003: 39).
Classroom practitioners of critical pedagogy face the dilemma, as explained by Peyton Young (2000), that text production and interpretation can never be neutral as they are mediated by the unequal power and resource distributions of societies. This of course applies not only to the discourses which are the target of CLA work, but equally to the critical researcher's, teacher's and learner's discourses. Hence any representation of language (e.g., descriptions of grammar) is suspect. Morgan (2004: 173) concludes, however, that avoiding representation would do a disservice to students for whom they are 'important resources with which to shape counter-discursive practices in ESL classrooms'.
Much published work on CLA relates to academic reading and writing, and builds on Fairclough's model, which includes ideational and interpersonal meanings, reader/writer relationships and thoughts, and their views and assumptions about the discourse (Fairclough 1989). Clark & Ivanič (1991: 170) set out three principles for LA work in this critical framework, one of which is that 'LA work should focus on the way text, socio-cognitive processes and socio-cultural context are interrelated'. This interrelatedness is evident in Pittard & Martlew's (2000) ethnographic study of a writer's engagement with different writing tasks. They conclude that social aspects of writing could not be separated from cognitive aspects; individual and social dimensions of meaning making were equally important. In Allwright, Clark & Marshall-Lee's (1996) discussion of teaching EAP to University students, the academic discourse community is referred to as a site of struggle. The writer can choose to comply or resist conventions related to who they are allowed to be, what they express and how. In order to exercise this choice, they need to become aware of the options and the likely effects and consequences. Adopting a similar approach, Hyland (2002) discusses the common misconception that first person pronouns should not be used and recommends awarenessraising activities on real academic text in the students' subject discipline as a means of helping them express their own voice. This focus on discourse as the outcome of choice, with purpose and audience in mind, is evident also in Rowley-Jovliet & Carter-Thomas's (2003) study of genre awareness and rhetorical appropriacy. From their analysis of NS and NNS scientists' use in English of syntactic resources to manipulate information structure they conclude that the NS are more sensitive to context and genre, and more skilled in adapting to either spoken or written delivery. Consciousness raising of these issues in oral and written language is recommended for both L2 learners and EAP/ESP instructors.
Many CLA practitioners might agree with Janks & Ivanič (1992: 320) that 'CLA should underlie all language teaching and learning' but its uptake could be limited by its confrontational nature. This concerns Wallace (1999: 104), who is critical of the three principles around which she claims critical pedagogy is usually constructed: teaching as emancipatory, difference-oriented and oppositional, on the grounds that they 'overstate the importance of a confrontational stance to establishment discourse'. She suggests that critical pedagogy should value commonality and resistance rather than difference and opposition, and should foster 'an understanding of the nature of disadvantage and injustice beyond that personally experienced' (Wallace 1999: 104). Such an approach might make CLA more widely applicable in mainstream education.
From this reviewer's perspective, criticality in LA work can perhaps best be seen as a matter of degree and focus, rather than as either present or absent. Outside CLA circles, but largely due to their influence, a critical approach might in that sense already be more common than is immediately apparent. A field usually associated with a degree of criticality is inter/cross-cultural awareness, to be discussed below.

LA in a multicultural and multilingual society
Not all approaches to inter/cross-cultural communication focus on language. In the work reviewed below, however, including some work on multilingualism, language is central and teacher and learner awareness pivotal.
The relationship between foreign language teaching and cross-cultural awareness is the subject of Kramsch (1993) (see also Kramsch 1995). Hinkel (1999) addresses culture in classroom interaction, its effect on writing and acquisition of pragmatic competence, and in relation to teaching materials. Referring to the teaching of Modern Languages in the UK, Phipps & González (2004) argue for a critical constructivist approach in which languaging (see above) plays a central role in shaping intercultural being. 'Unless we actually centre languages at the heart of life, at the heart of intercultural being, grounding a fresh curriculum in common experience, then the marginalization will continue and our lives and curiosities for other worlds and other ways of being will continue to be eroded in the name, even more ironically, of the global market' (Phipps & González 2004: 166).
A range of culture-awareness themes can be distinguished in LA research, such as cultures of ■ Language awareness learning/classroom culture (often related to Asian learners of English), culture and writing (in particular academic writing), culture and the media, and dominant vs. marginalized cultural encounters, to mention but a few. An example of the first strand is Cortazzi & Jin's (1999) often cited work on Chinese learners (also Jin & Cortazzi 1998). They discuss the mismatch between the cultures of learning of Chinese students of English and their English (native-speaker) teachers. Expectations on teacher and learner roles, textbooks, and classroom interaction differ and may clash. Students in Cortazzi & Jin's research stress that they are intellectually active while their western teachers perceive them as being passive. Much of the work in this area is concerned with countering existing stereotypes. There is perhaps a danger that despite the researchers' intentions, work such as that by Cortazzi & Jin may be interpreted as showing what a particular group of learners are 'really' like, in which case one stereotype has merely been exchanged for another.
How does one overcome or counter stereotypes? Lessard-Clouston (1996) observes that opportunities to interact with another cultural group do not necessarily produce cultural awareness. Instead, interaction can reinforce stereotypes and prejudice. Explicit training is needed, according to the author, as it affords time for reflection and development of critical awareness. Lack of sociolinguistic awareness among Chinese students in the UK is discussed by Marr (2005), who found that his participants were disappointed and surprised at the regional accents of their lecturers, whom they had expected to provide them with a model of 'good' English, i.e. RP (received pronunciation).
As exemplified by this and other studies above, the target of research has often been the learner of English. In de Courcy (1997), in contrast, the participants are English speaking learners of Chinese and their Chinese native-speaker teachers. The setting of the study is an Australian university, where the students were enrolled in a Chinese immersion programme. de Courcy describes how learners and teacher functioned each according to their own cultural script, and the problems that arose out of this. It is an interesting account of genuine bewilderment at culturally different behaviour. Although at least one teacher's outlook changed in the course of the programme, it also becomes clear that cultural awareness does not necessarily lead to a willingness to modify one's behaviour patterns.
Another strand of research with cultural and awareness dimensions is represented by Shi (2006), who investigates textual appropriation, currently a much-debated topic as it relates to plagiarism in tertiary education. The study involved native English-speaking and German, Chinese, Japanese and Korean undergraduate students at a North American university. Shi found in interviews that the students separated cultural hurdles from linguistic challenges, and suggested that it would be useful for both students and teachers to view textual appropriation issues from these two viewpoints. Shi advocates a postmodern pedagogic approach which would involve imitation, effective borrowing and inferential thinking skills to develop students' authorship skills and cites a number of recent studies and authors which take a similar position, principally Minock (1995).
It may be controversial to put the Jaworski et al. (2003) study of travel programmes on British TV down as an example of encounters of dominant with marginalized cultures. Nevertheless, it could be argued that the non-English cultures and languages in the programme are being marginalized. The researchers present a critical analysis of out-group code-crossing (such as the use of basic phrase book expressions in the local language) in travel programmes on British television, and what it reveals about ideological stances and identity positions of the participants and programme makers (see also Gieve & Norton in press on a similar topic). More transparently to do with dominance and marginalization, Lindberg (2003) discusses the need for native speakers of Swedish to be aware of their responsibility as the more competent language users in inter-ethnic, gatekeeper encounters, and what the native speaker can do to facilitate successful communication. The author reports on attempts to empower recent immigrants by enlisting native-speaker volunteers to provide conversation opportunities and provides an analysis of those encounters. Another teaching approach used was text reconstruction, which, perhaps surprisingly, benefited the least educated learners the most.
In Europe, a number of projects aiming to raise cultural and linguistic awareness among school children and their teachers and parents have been implemented in recent years. Helot & Young (2006) report on a three-year project in a small French primary school in Alsace (see also Young & Helot 2004). In collaboration with the parents, Saturday sessions were run at which the children were exposed to the languages and cultures represented by children in the school and in the community. Based on analysis of video footage and observation data, the project was found to be successful in giving the children a more positive attitude to the languages and cultures, and to their peers of different backgrounds. The parents' participation was an important factor in enhancing the status of the less dominant groups and languages .
A larger project, which has served as the catalyst and model for other, subsequent projects, is EVLANG ('L'eveil aux langues dans l'ecole primaire'). Initially a three-year project (1997)(1998)(1999)(2000)(2001), it has since continued as an organization (see website). It involved teachers, children and researchers in Austria, France, Italy, Spain and Switzerland ■ (Candelier 2001(Candelier , 2005. The project addressed researchers' concern about western ethnocentrism and the dominance of western culture in a culturally and linguistically diverse society by producing materials which exposed 5th-and 6th-grade children in 150 classrooms across Europe to a great range of languages and related tasks. As the name suggests, it was concerned with an initial, 'awakening' stage of LA, not to teach language but to raise awareness of how languages work. The assumption was that this would both facilitate subsequent learning of languages and make students more accepting towards a diversity of languages, their speakers and cultures. The tasks designed for the project involved over 40 languages and proceeded in three stages: becoming aware of knowledge already held (including misconceptions or lack of knowledge), creating new knowledge by solving cognitive problems, and becoming aware of what knowledge had been gained and how (Masats, undated). Topic areas included language structure, language and culture, language contact, variation, languages in a global context, bilingualism, and the status of languages. They attempted to avoid stereotyping, instead striving to engender acceptance and appreciation of diversity. According to Candelier (2005) an evaluation produced mixed results, but generally the effects on attitude were deemed very positive, and there were indications that linguistic sensitivity had also been enhanced.
EVLANG has inspired a number of subsequent projects in Europe and beyond, involving large numbers of researchers, communities and schools. Some education authorities have considered the approach effective enough to integrate it into the curriculum (Candelier 2005). In Switzerland, teaching materials in the form of 30 tasks for children from pre-school through primary were commissioned in a project called 'Eveil aux langages et ouverture aux langues a l'école' (EOLE), and LA is now part of some teacher education programmes in Switzerland (Perregaux et al. 2002;Perregaux 2006; see also http://www.romsem. unibas.ch/sprachenkonzept/Annexe_11.html).
In Greece, the Athens Institute of Pedagogy under the auspices of the Ministry of Education published materials and introduced LA in 150 schools. In Austria, the Center for School Development has made EVLANG materials electronically available to schools. On a smaller scale, EVLANG and subsequent projects have, according to Candelier (2005), had some impact on teaching programmes in Portugal and France, but  lament the lack of guidance on how teachers should implement the (limited) LA aims of the French primary curriculum.
The main spin-off from EVLANG was the 'Janua Linguarum Reserata' project (2000Reserata' project ( -2003, usually referred to as JALING, which extended to cover all of primary and secondary, involving 10 countries in eastern and western Europe (Candelier 2004(Candelier , 2005. Despite generally positive outcomes, Candelier (2005) also refers to some obstacles to a wider implementation of Language Awakening type programmes in schools: there is strong parental pressure to provide one foreign language, usually English, which normally takes all the time available for language education, and there is sometimes a perception that LA is a soft option which does not involve any 'real' learning.
A number of in-school LA projects have been run, and are still being implemented under the auspices of the Council of Europe through the Centre Européen pour les Langues Vivantes (CELV) in Graz, Austria. Among them is the 2004-2007 project, 'Language Educator Awareness' (LEA) with the subheading 'Developing plurilingual and pluricultural awareness in language teacher education' 8 (see website). Another current CELV project is 'Linguistic diversity and literacy in a global perspectivean intercultural exchange with African countries' (2004)(2005)(2006)(2007) in collaboration with UNESCO and others (see UNESCO website).
That there is a need for multi-cultural and multilingual LA work not only in Europe is highlighted by the statement on the UNESCO website that: 'Today there is a serious lack of adequate teaching resources in Africa and in Europe, especially in regards to languages of small migrant communities, nonstate languages and non-territorial languages' (see website).
Like the researchers and practitioners in the projects discussed, Jessner (2006) advocates a crosslinguistic approach to LA, particularly with multilingual students who can capitalize on commonalities between L1, L2, L3 and so on, and also take the differences between all their languages into account. Being multilingual, Jessner asserts, has cognitive effects in itself such as enhanced metalinguistic awareness and an enhanced multilingual capacity to monitor, and it positively affects divergent and creative thinking, pragmatic competence, communicative sensitivity and translation skills. She points out that, as a consequence, all learning is affected by prior linguistic and metalinguistic knowledge, and L3 learning is therefore different from L2 learning. Research is needed to understand these differences, and the possible implications for teaching a third language. Jessner's claims are backed up by findings from Kemp's (2006) study in which an adapted Modern Language Aptitude Test was administered to non-native speakers of English. The effect of multilingualism on test scores was 'extraordinary' and ■ Language awareness cumulative; the more languages the higher the scores. The effect of multilingualism on metalinguistic and metacognitive awareness of multilinguals is an exciting field for further research (see a review of the field in Cenoz 2003).
In the UK, CILT -the National Centre for Languages -presents a very up-beat picture of community languages in England, Scotland and Wales in a report on a 2005 survey (see website), which contrasts with the critical view taken by, for example, Bhatt & Martin-Jones (1992), referred to above. Lasagabaster (2003) provides an extensive review of research into attitudes to languages around the world, as well as an account of his own study of attitudes to Basque, Spanish and English among students at the Universidad del País Vasco. In the context of Europe, there are concerns both about the demands of an increasingly more complex linguistic environment and about the negative effects of English as a Lingua Franca.
Increased multilingualism is only one of the suggested and possibly complementary ways of handling this situation. The fostering of intercomprehension is another. The concept is discussed in some detail by Doyé (2005), who also provides a definition: 'intercomprehension is a form of communication in which each person uses his or her own language and understands that of the other' (2005: 7). It applies to both spoken and written language and aims to develop interpretive discourse skills in typologically related languages. This could, it has been suggested, act as a counterweight to English Lingua Franca by increasing the use and perceived value of less dominant languages, thereby providing more opportunities for the expression of their cultures. The European Awareness and Intercomprehension (EU + I) project has as its stated aim to contribute to LA in Europe by developing an intercomprehension learning/teaching methodology and materials (see EU + I website).
A number of other project sites provide samples of activities and materials, for example, EuroComRom (Romance Languages), IGLO (Germanic Languages), Intercomprehension in Slavonic Languages, and ILTE (Intercomprehension in Language Teacher Education). The EuroComCentre provides links to all (see website).
Language policies in a de-facto multilingual European Union are discussed by Nelde (2001), who points out that there is no one-size-fits-all solution to the competing needs and demands of majority and minority language communities. Cots & Nussbaum (2002a: 58-60) provide an account of LA as content and medium in the Spanish and Catalonian curricula, and their broad LA aims for L1 and FL learning and teaching.
In France, Blanchet (2006) argues against the notions that multilingualism is divisive, or that it requires a rare talent, and for a culture of multilingualism and the teaching and nurturing of 'les langues de France' referred to in the title, that is languages spoken where schools are located, including local and home languages such as Breton and Berber. Perregaux (2006) adopts a similarly positive stance on literacy in multilingual pre-school children where the teacher cannot assume shared knowledge, experiences and aspirations, or languages, but can assume a rich variety. The author discusses an approach to creating a common culture through coconstructed, shared writing where more than one language may be drawn on (for example, use of particular vocabulary items), reflecting the linguistic repertoires of the children. The children compose the story orally and the adult writes it down demonstrating and facilitating the transition from oral to written. Throughout the process, the children are encouraged to draw and reflect on the resources of their different experiences and languages.
In summary, the different languages and varieties making up a person's linguistic competence are increasingly seen, in the LA literature, as constituting a repertoire rather than a number of language sets. In the LA classroom, social, cultural and linguistic awareness, and language learning are often treated as overlapping and mutually facilitating.

Conclusions and the way ahead
The aim of this review has been to present an overall picture of where LA as research and practice is today. Because of this wide scope, more narrowly focused reviews could usefully be conducted to interrogate each of the areas addressed above. Nevertheless, an impression of the field has emerged, which I will try to summarize below.
Research into awareness and related notions has enhanced our understanding of its complexities, in particular how awareness and attention relate to noticing. Classroom-based research shows that attempts to guide learners' noticing towards target structures, for example through input enhancement and processing instruction, are promising but precise causal relationships are complex and difficult to determine. Whatever their facilitative effect these techniques might have, embedding them in interactive learning contexts would seem necessary in order to allow meaning-form connections to be firmly established. The important role of classroom interaction, whether structured or informal, in scaffolding knowledge creation is an exciting area of LA research and theory building.
A range of LA approaches to teaching were discussed above: a theoretical framework suggested by Borg (1994); three broad curriculum principlesautonomy, awareness and authenticity -developed by van Lier (1996); and a practical pedagogic schema for language awakening training, provided by EVLANG and subsequent projects. The extent to which any of these sit within a larger framework of ■ Critical Pedagogy depends crucially on the beliefs and preferences of the teacher. The work of educators and researchers in cross-cultural LA and multilingualism highlight the important sociocultural and citizenship aspects of LA, in CLA combined with a particular focus on power issues. Work on cross-cultural LA is particularly concerned with the engagement of minority and dominant groups with each others' languages and cultures. Regarding language teachers, there seems to be a consensus that to use LA effectively, they need both a high level of language awareness and first-hand learner experience of an LA approach, for example, from their teacher education courses.
There is perhaps a danger that the multiple foci of LA could lead to fragmentation. Depending on the orientation of the research or classroom practice, LA work will draw to differing degrees on sociocultural, critical, linguistic, cognitive and general educational theory. Informal discussions with participants at ALA conferences in 2004 and 2006 suggest that there are no ready answers as to what the various orientations within LA have in common, and yet there is a sense of coherence. I propose that features emerge from the literature reviewed above which explain this common ground of the LA community.
A shared concern, I would argue, of LA practitioners and researchers, is the notion of engagement with language. As collectively constructed over the last 15-20 years, LA does not refer to a purely intellectual awareness and is not passive. In the work discussed above, LA both engenders engagement with language and is constructed through it. An important manifestation of engagement is languaging, a prime site of knowledge construction. The engagement can be intellectual, affective, social or political or, usually, a combination of the above. Researchers are concerned much less with memorized or encyclopaedic knowledge than with LA as it encourages, facilitates, discourages or hinders particular types of engagement with language, be it by language learners, gatekeepers, the general public or other groups. Thus, LA has a common core of stances and concerns which give it an important degree of coherence, and within which LA researchers and practitioners situate themselves. Nevertheless, it has so far been under-theorized, and its coherence needs to be articulated.
A multitude of research avenues are suggested by the review, some of which I will mention here. More longitudinal research is needed on possible language learning effects of LA. In the realm of multilingual awareness, is learning/teaching of L2 and L3 the same, or different, as Jessner claims? If the latter, how are they different? The effect of particular awarenessraising techniques on language acquisition is still open to investigation. Languaging, as a process and its effects, has not been fully explored. What kind of metalanguage facilitates reflection and verbalization in language learning? The link between learning, how the knowledge has been acquired and the resulting quality of the knowledge is not well understood.
Noticing should be further explored; for example, the relationship between noticing and subsequent production, and between noticing and learner styles, together with the affective variables which promote or hinder noticing. LA needs to be studied in different social, cultural and material contexts. Little LA research appears to have been conducted in Asian or African settings, for example. Similarly, a few languages tend to dominate and the repertoire needs to be widened. Could LA be an effective approach to stimulating the use of Arabic as an intellectual and literary language in Lebanon, where English and French arguably tend to dominate? Could intercomprehension be fostered between related languages, such as on the Indian subcontinent? What demands do different approaches to LA, in different contexts, make on teachers' language proficiency, and how can teachers be equipped to meet those demands? Finally, the notion of engagement suggests a number of useful directions. What do native speakers do or not do to allow L2 speakers to engage with the L1 of the host country? What might encourage or dissuade people from engaging with a language/languages? In order to promote LA in a largely monolingual context such as that dominant in the UK, it would be useful to know when engaging with a foreign language is considered legitimate, and when it is perhaps seen as pretence, mimicking or even open to ridicule. I would agree with Gnutzmann (1997) that the great strength of LA is its holistic approach and therefore its ability to house different ideological positions at the same time, which makes it particularly suited to dealing with the complexity of language. The present situation, globally, is characterized by population movements, internal and external political conflicts, tensions between language, faith and other communities, potential and actual power abuse. In dealing with these issues there is a great need for enhanced literacies, and multilingual and multicultural competence and tolerance to which LA is well placed to make a major contribution.