Engagement with language: interrogating a construct

In ‘Language awareness and language learning’ (Svalberg, 2007) I concluded that the otherwise diverse and multidisciplinary field of language awareness (LA) is given coherence by its focus on engagement with language. I argued that LA is seen as active and not merely as a state of conscious awareness or sensitivity. This paper goes one step further by developing and testing the construct. It discusses what engagement with language might consist of and contrasts with and how it can be identified. Cognitive, social, and affective aspects of engagement, including notions such as attention, autonomy, and agency, are posited. Having arrived at first at a working definition and then an expanded and refined definition, the construct is applied to some classroom and student interview data. Some tentative conclusions about English for Speakers of Other Languages (ESOL) students' engagement with language, what motivates, facilitates, and hinders it, are drawn. The main purpose of the paper is, however, to try and determine the usefulness or otherwise of the construct as such, how it relates to other constructs in the literature, and where it might be taken in future research.


Introduction
In a recent review paper of Language Awareness (LA) I stated that despite being a very 'broad church', LA is given coherence in both classroom practice and research by its focus on engagement with language: 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 . . . . The engagement can be intellectual, affective, social or political or, usually, a combination. 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. (Svalberg, 2007, p. 302) The problems with such a statement are all too obvious. What does engagement with language mean? Do we really need another term? Is it a useful concept, or just a fairly empty and ambiguous buzzword? The present paper is an attempt to answer these questions.
The everyday contexts in which the term 'engagement' is used are suggestive: cogs engage; armies engage in battle; people get engaged (to be married). It seems to suggest physical or metaphorical close contact and some force driving things forward or pushing in a direction. It is an attractive metaphor, and a popular one in the literature on education (pupil engagement, literacy engagement) and in other areas such as management (workplace engagement). Its meaning is, however, usually assumed rather than defined, raising the possibility that one writer might not mean the same as another.
My own interest in engagement was triggered by a number of recent publications on learner variables such as motivation and learning strategies (e.g. Dörnyei, 2003;Pavicic Takac, 2008) and a similar trend in postgraduate research proposals. Research into motivation attempts to understand what drives engagement and studies of learning strategies deal with certain forms of engagement, how they can be encouraged and what their effect might be on learning. I would argue, therefore, that researchers and practitioners alike are currently trying to come to grips with 'engagement' as the place where learning happens.
In the field of LA, the term 'engagement' is often not defined. Some linguists to whom 'engagement' is central use the term in its everyday sense. For example, Hyland (2003) and Batstone (2002) have investigated learner engagement, and Borg (2007) has looked at teachers' research engagement. But the term has recently been used by Storch (2008) more specifically to denote instances of noticing as evidenced in language-related episodes (LREs). Drawing in particular on the works of Leow (1997) and Kuiken and Vedder (2002) who referred to 'simple and elaborate noticing' (and also 'perfunctory' and 'substantive' noticing;Qi & Lapkin, 2001) she distinguished between two levels of engagement, limited and elaborate. The term 'engagement with language' is thus used by Storch to 'describe the quality of the learners' metatalk ' (2008, p. 99) and is linked to the quality of attention on which, as Storch (2008) has pointed out, little research has been done. Storch's definition is useful but limits engagement to instances which result in metatalk, and it gives no indication of how dimensions other than the cognitive might be included.
Engagement seems an intuitively apt term for what happens in LREs but is potentially a richer notion than this might imply. As a construct, the notion of engagement could provide ways of explaining why some linguistic or language-related behaviours and attitudes seem to facilitate language learning and learning about language/s more than others. I would argue that in addition to its cognitive aspects it crucially involves a range of social and affective phenomena and that it is this complexity which justifies the construct.
In order to facilitate a clearer understanding and make it researchable and able to be evaluated as a construct, 'engagement with language' needs detailed interrogation. Its elements need to be laid bare, and that is what I will attempt to do below. The approach has two strands. First a methodology employed by Ellis (2004) to interrogate another notion, 'L2 explicit knowledge', is explained and then applied. Figure 1 is my summary of the steps he follows.
For the purposes of this paper, X in Figure 1 is 'engagement with language'. The first part of the analysis and discussion will follow Ellis's methodology as far as the expanded definition. Having reached a more in-depth understanding of the construct, the paper continues with a brief discussion of the relationship between engagement and LA.
In the second part of the paper the various components of the construct will be applied to some interview and observation data. This is for illustration only, as the data was originally collected for a different purpose. Armstrong and Svalberg (2005) interviewed English for Speakers of Other Languages (ESOL) teachers and their adult students and observed them in class. For the present paper, one set of field notes from the classroom observations and one of the interview sessions, with two ESOL students, were subjected to an analysis of 'engagement with language'. The discussion below leads to suggestions of factors which might encourage or discourage engagement with language. An evaluation of the construct and suggestions for ways ahead conclude the paper.

What is 'engagement with language'?
The working definition from which this investigation starts is as follows: In the context of language learning and use, 'engagement with language' (Engagement) is a cognitive, and/or affective, and/or social state and a process in which the learner is the agent and the language is the object and may be the vehicle (means of communication).
For convenience, I will from time to time refer to 'engagement with language' as Engagement, and capital E will be used also in the verb ('Engage' rather than 'engage with language').
In the definition, I distinguish between when the language is the object of Engagement and when it is a vehicle of communication. The term 'object' has been preferred to other possibilities such as 'form' or 'system' to signal the inclusion of social and other nongrammatical aspects of language. Likewise, 'vehicle' has been chosen in preference to, for example, 'use', as it seems to carry fewer limiting connotations. In the definition, language being the object is seen as primary, while language as a vehicle is a secondary and therefore not necessary feature of engagement. The reasons should become clear in the following discussion.
An example of 'language as object' in the classroom would be learners doing language exercises or tasks, perhaps discussing solutions in the L1 or target language (and thus giving rise to the much-studied LREs). Collaborative dialogue (Swain, 2000) or what Swain has later referred to as 'languaging' (Swain, 2006) is thus one manifestation of Engagement. A classroom example of language as vehicle would be learners doing a communicative task in the target language without paying attention to aspects other than communicative efficiency. In many contexts the two aspects of Engagement will both be present. I would argue, however, that language can also be used without Engagement. Somebody speaking/writing spontaneously and automatically for purely communicative ends may be socially engaged but not Engaged with language. It cannot be assumed therefore that use of language is in and by itself evidence of Engagement in the sense discussed here. Also, Engagement is gradable, i.e. one can be more or less Engaged, and it would seem to possess a number of qualities.
These are implied by the working definition, in the statement that Engagement involves particular cognitive states, and/or affective orientations, and/or social attitudes, as well as actions and behaviours. In other words, it should be analysed as both a state and a process. The cognitive, affective, and social aspects of Engagement overlap and are likely to affect each other, but the separation is made to allow a more in-depth analysis below.

How is engagement different from similar notions?
Though they are by no means synonyms, three notions which seem semantically close to Engagement are 'involvement', 'commitment', and 'motivation'. In Table 1, I have attempted to establish to what extent and in what way Engagement differs from or is similar to its semantic neighbours. 1 Table 1 represents an attempt to unpack the meanings of 'engagement' and the terms semantically close to it, as they are commonly used. Along the horizontal top row, I have chosen to compare it with terms which are common currency, like 'involvement', rather than less common notions such as 'investment' (Norton Peirce, 1995) on which there may be less intuitive agreement (but see comments below). The boxes containing a dot (•) each mark a feature implied by the term at the top of the column. For example, the table makes the claim that the terms 'engagement' and 'involvement' both imply some level of alertness on the part of the engaged or involved individual. It does not claim that committed or motivated individuals cannot be alert, only that the terms 'commitment' and 'motivation' do not imply anything about alertness.
The comparison with 'involvement', in the table, implies also that while focused attention is necessary for engagement, involvement can be unfocused. Likewise, in this analysis only 'engagement' involves a process of taking personal ownership of knowledge, turning it into so-called action knowledge (Little, 2007, p. 19, discussing Barnes, 1976 on which further action can be based. It is posited that 'involvement' is not necessarily affective. The affective dimension is more obviously a component of 'commitment' and 'motivation', which are similar to 'engagement' in that they rely on a positive attitude and willingness, and are to some degree purposeful.  Van Lier (2007, p. 48) has explained autonomy as 'the feeling of being the agent of one's own actions', while Little (2007, p. 15, citing Holec, 1981 has described it as 'the ability to take charge of one's own learning'. I have assumed that autonomy is an affective disposition and that a motivated or engaged person is characterised by some degree of autonomy but that this is not necessarily the case with a person who is merely involved. The two features unique to 'engagement' in Table 1 are the cognitive features 'focused attention' and 'action knowledge' (making knowledge one's own). The picture emerging is of individuals who are actively constructing their knowledge not only by mental processes but also equally by being socially active and taking initiatives. In this analysis, all the nine features marked in Table 1 are implied by the term 'engagement'.
In Table 2, the key characteristics, or qualities, of the more specific construct, Engagement (with language), are identified. Following the working definition, it is analysed as comprising cognitive, and/or affective, and/or social states and processes. Table 2 attempts to clarify the relationship between Engagement on the one hand and notions such as attention, reflection, willingness, and interaction on the other.
The characteristics are only meaningful if they can be operationalised -the 'Criteria for identifying it' step in Figure 1. Table 3 is an attempt to indicate how it could be done. In Table 3, 'L' stands for 'learner' but would not necessarily be a learner in the formal sense; L could be any individual potentially engaging with language. The table suggests for each kind of Engagement some questions which could be asked in order to establish its degree and qualities. A person might, for example, be cognitively but not socially Engaged or be Engaged to different degrees in different spheres.
The posited characteristics of Engagement and their operationalisation (Tables 2 and 3) may not be exhaustive, but they offer a principled and manageable framework for the interrogation and analysis of the level of engagement and the identification of the characteristics of that engagement. They can be said to validate the earlier semantic unpacking and comparison (Table 1) to the extent that there are clear links between them. For example, Table  1 suggests that a cognitively Engaged person is alert, pays focused attention and constructs their own knowledge. Table 3 operationalises that by asking questions around how alert, focused, reflective, and critical/analytical the person is. Affective and social Engagement are operationalised in similar ways.  An expanded, more precise definition can now be attempted.

Expanded definition of Engagement with Language:
In the context of language learning and use, Engagement with language (Engagement) is a cognitive, and/or affective, and/or social state and a process in which the learner is the agent and language is object (and sometimes vehicle).
r Cognitively, the Engaged individual is alert, pays focused attention and constructs their own knowledge.
r Affectively, the Engaged individual has a positive, purposeful, willing, and autonomous disposition towards the object (language, the language and/or what it represents) r Socially, the Engaged individual is interactive and initiating.
The definition is not meant to be 'the last word' on the subject. Instead, it raises a number of questions and needs to be further refined by being debated and challenged. For example, in terms of determining when Engagement is or is not present, are some features of Engagement necessary while others are optional? Is Engagement always cognitive to some degree but not necessarily affective and/or social? Is it possible to be socially Engaged without the presence of some interactivity? If Engagement is a matter of degree, how does one establish a minimum level of Engagement? I have assumed above that from a learning point of view, Engagement matters. If so, are all kinds of Engagement equally effective? One purpose of the above analysis is to make such questions researchable. A final question, to which I will return below, is whether this specification of the construct can help establish factors which facilitate or hinder Engagement with language.

Engagement with language and LA
Having determined what Engagement means and acknowledging that this is a work in progress, the purpose of the remainder of this paper is to try and determine how useful (or otherwise) the construct might be to LA and where it might be taken in future LA research. LA can be considered from two different angles in relation to Engagement: as the likely outcome of Engagement with language and as a resource which feeds Engagement. LA is most obviously the outcome of Engagement when language is the object. The degree to which language as vehicle contributes to LA may depend on the user's analytical skills and what other demands there are on their attentional resources. Figure 2 attempts to illustrate these relationships.
According to Figure 2, once some LA has been acquired it is a resource which can be drawn on when Engaging with language, either as object and vehicle or as object only. Such Engagement is likely to enhance or add to the person's LA and thus to create a richer store of LA for further Engagement and knowledge creation. The model in Figure 2 implies that LA research and practice are intimately and crucially concerned with the characteristics and qualities of Engagement with language. A particularly interesting area of investigation might be how this cycle is affected by multilingualism (cf. Singleton & Aronin, 2007). The model would need to be tested in empirical (or experimental) research, but this is beyond the scope of the present paper. Below I will only try to illustrate with some sample data how the construct might help us understand some of what happens in language classrooms.

Some ESOL learners' Engagement with language
The criteria above for identifying Engagement (Table 3) indicate the need for micro-analysis of conversation data or audio/video recording of classroom interaction, but for the purposes of this paper I have, as previously indicated, merely reanalysed some already available data. Three adult ESOL student groups were observed, by two researchers taking field notes, in 12 class sessions of about 50 minutes each. Subsequently, eight of the observed ESOL students and their four ESOL class teachers were interviewed. The focus of the observations was on how the teachers managed grammar instruction and on the learners' engagement. The student interviews explored attitudes to the teaching methods and materials used and rationales for such perceptions. 2 Below I will discuss some illustrative extracts from the observation field notes, and from a transcript of one 30-minute pair interview with ESOL students from the same class as in the observation extracts. I believe the data sheds some light on the role of Engagement in the creation and use of LA (LA as outcome and resource). Each extract has a heading indicating what it is meant to illustrate. The section concludes with suggestions for factors which might facilitate or hinder Engagement.

Cognitive Engagement
I have taken it as given that the development of LA, and the learning of a language or languages, normally relies on a combination of cognitive, affective, and social factors. Below I will attempt to be more specific by isolating instances of the three aspects of Engagement referred to above. While this separation is somewhat artificial (there are clearly strong interrelationships) I believe it can help provide a rich analysis.
In the field notes, numbers were assigned to the students in the order they entered the classroom. In the extract below, 'T' is the teacher; 'S' indicates a single student and 'SS' two or more students.
In creating classroom conditions which are conducive to cognitive Engagement with language, the teacher has a key role. In the sample observation the teacher kept her voice at a fairly low pitch and low volume. This could be calming and help students concentrate, though a livelier voice could help some stay alert. When she was monitoring students' work, the teacher often kneeled down to reach eye level with the seated students, as in the example extract below. Creating a degree of social intimacy in this way might facilitate focused attention.

Extract 1 (observation) -creating the conditions for cognitive Engagement:
S1 and S7 discuss alternatives with T ('newer'/'more new') in private conference. T encourages SS to put the adjectives in sentences. T reads some and laughs with SS. S tries out an example structure on the T. Discussion ensues one-to-one [T and S].
Above, student 1 and student 7 were working together. In this episode the students had been given an exercise to do in pairs on using the comparative of adjectives. They were discussing which form of 'new' to use. The interaction was not tape-recorded, but parts of it could be overheard.
These hushed discussions between a student or students and the teacher were quite frequent. The teacher rarely provided solutions but favoured guided discovery, which encouraged focused attention on the students' part, though this might have been for brief spells only. Also, it was difficult to tell whether in replying to the questions students drew inferences or relied on memory. The former would, presumably, require a higher degree of cognitive Engagement than the latter.
As adults come with an educational history, often including more than one language, it is not surprising that many have a fair degree of LA, including declarative grammar knowledge. This is evident in the following interview extract. (Student names are pseudonyms; 'I' is the interviewer.) Extract 2 (interview) -LA facilitating Engagement in the form of noticing and reflection:

I:
What do you think about the grammar teaching? Max: Seems very useful because, I think that just a simple grammar is not enough because if I want to read a newspaper, sometimes there is not just a simple sentence, will and write a lot of clause or compound, complex so I want to try out, to really take the maybe the subject and the object and which one is adjective which one is something, it's very complicated I think. But I think the grammar is the most important tool. More understand a lot of different piece of writing.

I:
So would you like more, or do you think it's enough as it is?
Max: For me, I want more. In the writing, I can't just write a subject verb object and nothing else.
Max was acutely aware that he did not have the grammar needed to express the meanings he wanted. He noticed the complexity in the input which he had chosen as his target model, i.e. newspaper text. More than simple familiarity with grammar terms, his declarative knowledge seems to amount to LA, facilitating Engagement in the form of noticing and reflection. Max's frustration at not having sufficient resources could be either discouraging or motivating. In his case, the impression was that this affective dimension intensified his cognitive Engagement with English.
When Max was asked if he wanted to have his mistakes corrected by the teacher, he indicated that corrections helped him notice.

Extract 3 (interview) -correction facilitating cognitive Engagement:
Max: I think it is very important because if something written is wrong then tutor correct for me, then I will notice 'oh this is wrong', so next time I will notice and I think I will guess and improve . . . about this.

Affective Engagement
In the analysis above, it was said that affective Engagement could be operationalised as degree of willingness, purposefulness, and autonomy. The good-humoured mood in the lesson -partly due to the teacher's approach, described above -contributed to the impression of willingness. Sporadically, there was low-key laughter in the class. A particular episode stands out. In the extract below, the students are doing a listening comprehension exercise which involves matching pictures and questions with what they hear. The teaching points are comparative adjectives and use of 'all' and 'both'. The facilitator of the learners' Engagement is probably their enjoyment of the competitive task.

Extract 4 (observation) -affective Engagement/willingness: Tape:
T says what [the tape] contains. (One conversation around each picture; SS pick out the question on the work sheet.) SS appear to enjoy their 'successes'. There is quiet laughter and silent victory signs.
The students appeared to enter into the spirit of the task, making it a game. They were engaging willingly and deriving some pleasure from their achievements. The competitive element (perhaps not intended by the teacher) seemed to contribute to both affective and cognitive Engagement.
Along with the teacher's approach and the task design, the topic and type of text may have an effect on learners' willingness to Engage. In the sample interview, Max (below) appeared particularly keen to read and listen to factual texts such as newspapers and news reports. The learner worked purposefully on expanding his vocabulary. The facilitator seems to have been the intrinsic interest of topics and texts used in the classroom (and also available outside).

Extract 5 (interview) -affective/purposeful Engagement: I:
And is that what you enjoy most, learning to read and write or? Max: Yes, to read what happened in this world. Because from just a couple of weeks ago we study a lot class this afternoon -it's about Britain and the world. It is very interesting.
It encouraged me to, more hardworking to study English to try understand more and build up more vocabulary.

I:
So do you like learning about the country? Max: Yes.

I:
The history and so on? You enjoy that. Max: Yes.
The intrinsic interest of the material used in class spurred Max on to Engage, particularly with the vocabulary.
The other student interviewee, Jenny, seemed to have a positive orientation towards the target language and what it represented. The following utterance follows immediately after the previous one. The facilitator here seems to be Jenny's self-perception.
Extract 6 (interview) -affective Engagement/willingness: Jenny: I proud when I can speak English.
Learning the language is clearly more to this student than a practical necessity. Nevertheless, meeting objective needs and feeling motivated and satisfied go hand in hand. To a question about what they thought of the course content, the reply of the same student is as follows. The immediate relevance of the language being taught provided Jenny with a purpose for Engaging with it. Though other types of data would be needed to know the depth of the Engagement process, the extract shows that the student's own difficulties with telephone conversations made her keen to participate in classroom activities on this topic, which for her were obviously purposeful.

Extract 7 (interview) -affective/purposeful Engagement
The sense of relevance and purpose made the two students Engage with the language not only in the classroom but also independently at home.

Extract 8 (interview) -affective Engagement/autonomy: I:
So, you like the worksheets? To take home. Jenny: Yes, yes, I can do it. Max: Excuse me. Another, is not an activity, it's very useful is role-play.

I:
Role play is useful? Yeah. Do you remember something you like especially? Can you give examples? Jenny: Something like that? Example? About listening, with the tape. Before I couldn't understand nothing. But it's better. Very better. Because all the time teacher try for me, and I try at home, TV and radio. It's better than before, very better.

I:
You are improving all the time. Jenny: Very important.
For most ESOL students, the English language is key to a more secure and possibly more fulfilling future, and so it is not surprising to find that some are highly self-motivated and try to further their learning in a number of ways. It was often evident in both the interview and the observation that Engagement was the learner's choice. Max even took the opportunity to follow additional courses.

Extract 9 (interview) -affective Engagement/autonomy: I:
I: What do you think -how much do you learn from your teacher, your classmates or by yourself? Max: I think the half from the teacher, after, for example, if I've got time I back to home, I will do some revision from the tutor then try to find out more, encourage more to read, so maybe it is half by myself.

Social Engagement
In the previous analysis (Tables 2 and 3), social Engagement was essentially linked to interaction and to learners' initiation and maintenance (or not) of it. Some of the extracts already discussed in terms of affective Engagement could also be analysed from the social point of view. Below, I will however examine evidence of social Engagement in other parts of the interview data. Key questions are firstly with whom the language learners/users need, want and are likely to interact and in what contexts. The students in the sample interview mentioned interaction in medical and shopping contexts and with neighbours; job interviews; reading government documents, including solicitors' letters; and (as above) understanding people on the telephone.
Max cared particularly about the quality of his interaction outside the classroom and was aware of the importance of appropriate language.
Extract 10 (interview) -social Engagement: maintaining interaction: Max: Yes, definitely. Because sometimes English is very is very difficult for me, because it is, like speaking a formal and an informal, I should and I must understand which way is formal and which way is informal and which which place I can use these, and . . . .
Not mastering colloquial varieties could make social interaction difficult for example in the workplace or with neighbours. Another challenge mentioned was the natural speed of delivery which can make understanding difficult (Flowerdew & Miller, 1996;Graham, 2006), particularly in telephone conversations. There are many situations outside the classroom where there is a lack of thinking time. In the job interview, there is added pressure on the interviewee because of the power held by the interviewer.
In contrast, features of classroom interaction with potential to encourage and facilitate Engagement are the availability of thinking time and the more egalitarian relationships between interlocutors (cf. Batstone, 2002). Individual learner variables nevertheless intervene. Max is keen to use peer interaction as a means of learning.

Extract 11 (interview) -social Engagement/readiness to interact: I:
What do you think of working in pairs and groups in the class? Max: It's a good way to communicate. First you can get a communication with another classmate, and you can not sure, just express maybe after then you get something, like you can use each other, try to find out from the talking. Jenny: Practice for talking and mainly practice for talking, practice for listening, accents, . . . very difficult . . . . English people, another . . . , young people, something is different, yeah . . . is happy to . . . the accent.
'You can use each other, try to find out from the talking' encapsulates the notion of 'languaging' (Swain, 2006), a social manifestation of Engagement which Max finds helpful.
Jenny is less able to Engage in the framework of peer interaction, apparently because of differences in language proficiency (a power differential), and prefers the teacher's expert scaffolding.

Extract 12 (interview) -social Engagement/readiness to interact: I:
OK. So you think that pair and group work help you with this? Would you like to do more group and pair work or less, or what do you think? Jenny: In the middle, a little bit . . . the teacher told is better, but yes this one practice is better, not too bad. Not a long time, because together I haven't a lot talking, just a little talk is finished. Me, I don't understand lots of, just a little talking. But with teacher help me I can speak longer.
As the two extracts show, Max is the more proficient of the pair of interviewees. He appreciates the opportunity to try things out with peers. For Jenny, both speaking and understanding in a group of more proficient students are very demanding. She needs the teacher to scaffold her. Despite the more accommodating nature of a classroom (as compared to external settings), it can apparently not be taken for granted that peer interaction will favour Engagement -in particular in mixed-ability groups.
It was clear that the students' desire to be able to interact with a range of people in a variety of situations was not just for practical expediency. Max (below) wished to be regarded as a good neighbour. He sees himself acting as member of a community where target language is a vehicle and is aware of the effect misunderstandings might have on his relationships with neighbours and cares about the quality of interaction.
Extract 13 (interview) -social Engagement/maintaining interaction: Max: Because for me I think by living in this country for ever I should understand language . . . . I should use the language to communicate with the foreign people and I should understand a lot of piece of writing like government or official . . . like recycling or something. I live in the area -I didn't until now think my neighbours will say 'this is foreigner, not good'.
Max is anxious not to create a negative impression among his neighbours, for example by not being able to understand the rules for recycling household rubbish. Reading comprehension thus becomes especially important, but communicating with 'the foreign people' is also important ('foreign' here probably means in relation to him).
The language proficiency of adult ESOL students has an impact not only on the individual but also on their families. Obvious needs are for example to find and keep jobs, but Jenny (below) had a less obvious need. She was concerned that she did not know English well enough to be able to help her son, probably with schoolwork. In the extract below, Jenny presents herself as part of a family. She has a need to Engage (with English) in interaction with her child; this is thus interaction for teaching rather than learning. (On an affective level, it seems to also provide a purpose for Jenny's Engagement in the classroom.) Jenny may be trying to say that while the grammar she gets in class is sufficient to develop her own proficiency, she needs more declarative knowledge about English grammar in order to help her son. This is quite possibly a strong facilitator of her Engagement with grammar learning activities in class. Further on in the interview, Jenny also suggested that if she had had more than one child at school, this would have given her opportunity to practise English at home, her assumption being that the children would naturally prefer to speak in English with each other. Hence, the presence of school-age children in the immediate or extended family might encourage and facilitate Engagement with the language by older family members both in a classroom setting and at home.

Extract 14 (interview) -social Engagement/readiness to interact
The sample data was not designed to address the effect of differences in gender, language, and cultural and educational backgrounds might have on Engagement with language in ESOL classrooms. A related issue is the link between identity and Engagement. Richards (2006, p. 63, discussing classroom conversation data from Willis, 1992) has provided a striking example of how a teacher dismisses and effectively stops a learner's attempt to introduce a facet of their identity which is clearly essential to the learner. It could be useful to understand better what effect such incidents might have on the learner's Engagement.

Facilitating or impeding engagement
Returning to the quote at the start of the paper, I have claimed that LA researchers and practitioners are concerned with 'LA as it encourages, facilitates, discourages or hinders particular types of engagement with language' (Svalberg, 2007, p. 302). Table 4 represents a first attempt to identify factors which do so. The qualities which would be affected have been repeated first (from Table 2) for convenience. Factors which might influence presence or degree of Engagement positively are called 'facilitators' and the negative factors 'impediments'. It is clear in Table 4 that cognitive, social, and affective factors encroach on each other; e.g. cognitive Engagement can arguably be affected by emotional state. Similar overlaps can be detected in regard to affective or social Engagement. For example, social interactive processes can be affected by perceived power differentials, which are also likely to influence affective factors such as trust. But for the purpose of analysis it seems useful to separate facilitators and impediments according to the three types of Engagement outcomes they affect. Table 4 posits that the factors listed are likely to affect Engagement and that they are individual and context-dependent. The same feature might either facilitate or impede Engagement; e.g. background music is distracting for some but helps others concentrate on the task; a competitive activity could inhibit Engagement in one context but stimulate it in another.

Conclusion
This paper was prompted by the perception that for researchers and teachers, Engagement is a central notion. I have tried to show how Engagement can be defined and operationalised and how it relates to LA: I have argued that LA can be simultaneously a result of and a resource for Engagement. The analysis attempts to provide a principled way of establishing, measuring, and interpreting the presence/absence, degree, and nature of Engagement. To the extent that this has succeeded, it should help make Engagement researchable. The discussion of the sample data shows, in my view, some of its potential to provide rich descriptions and analyses.
Engagement is clearly related to, but also different from, a number of other notions in the research literature. Its affective-state quality, 'positive orientation', is related to motivation (Dörnyei, 2003;Gardner, 1985), while 'behavioural readiness', a social-state feature, is perhaps closer to aspects of investment (Norton Peirce, 1995). Both the motivation and investment constructs, however, deal more with the reasons why Engagement does or does not occur than with what Engagement is. Thus, as one would expect, facilitators and impediments of Engagement (Table 4) relate to the above in an even more transparent manner, e.g. self-perception to motivation (Dörnyei, 2009) and power differentials to investment (Norton Peirce, 1995). The reader may well be able to see other links to notions in the literature.
This paper has attempted to provide an essentially dynamic model of Engagement itself, i.e. a model of what 'goes on' when there is Engagement, and in doing so to bring together, in a principled way, a variety of aspects of a highly complex environment (notably the language classroom; e.g. Larsen-Freeman, 1997). It has drawn on the works of Storch (2008) and others (e.g. 'languaging'; Swain, 2006) for the cognitive aspects and notions such as agency (Ahearn, 2001;Swain, 2006) and autonomy (Holec, 1981;Little, 2007). It is up to the reader to decide whether this makes Engagement too all-encompassing to be useful or if it may be a useful construct precisely because of its complexity.
Several research avenues are possible. Systematic testing of sub-components of the Engagement construct could be as valuable as more holistic research. Questions which might be investigated within an Engagement framework include under what conditions it produces useful LA outcomes. For example, memorised grammar rules (a kind of LA) may be of limited use to the learner. Could this be to do partly with the kinds of Engagement typically required for memorisation? What qualities, or combinations of qualities, should learners' Engagement have in order to produce learning? 3 Research on different cognitive, affective, and social aspects of Engagement might eventually be synthesised. Following Ellis's (2004) methodology (Figure 1) it is hoped that for any researcher wishing to conduct research on Engagement, the model presented above might help inform the choice of research design and measurement instruments, as well as the approach to data analysis, and that it will also stimulate a critical debate on related issues.

Notes
1. I am indebted to many of my doctoral students who have discussed this with me and helped me get to this stage of analysis. My initial set of features was somewhat modified on the basis of their input. 2. The data, collected for, and discussed in Armstrong and Svalberg (2005) is used here with the kind permission of my co-researcher. The overall research focus was the relationship between the aspirations of the National Adult ESOL Core Curriculum in the United Kingdom in regard to grammar and the actual classroom practice and dynamics.