Introduction

The idea that speaking should be a part of classroom practice is a very old one. The textbooks for learning Latin 500 years ago in the time of King Henry VIII of England, the so-called Vulgaria, were concerned, as their name suggests, with ordinary, everyday language, and there is evidence to suggest that classrooms at the time were every bit as lively in terms of speaking practice as many communicative classrooms today (for an extended discussion, see Carter and McCarthy Reference Carter and McCarthy2015). So in some senses, not much is new. On the other hand, the recording, analysis and understanding of everyday speech has never been more comprehensive, and the most common form of speaking for most human beings, social conversation, has been the subject of intense scrutiny by researchers. It is also often considered an important goal within second and foreign language learning, however imperfect endeavours to achieve that goal may be.

Underlying this chapter is the conviction that conversation as a social, multi-party activity, as opposed to speaking, which, in the second and foreign language learning context, is a broader and somewhat vaguer notion, often involving monologic talk (e.g. oral examination timed presentations), can and should be taught and learned, a view we share with Slade (Reference Slade1997) and Tsang and Wong (Reference Tsang, Wong, Richards and Renandya2002). Our position is that the strategies that speakers use to manage and participate in conversations (e.g. turn and topic management, reiterating and clarifying) can be incorporated into materials and that students can acquire both the language and the techniques to accomplish everyday conversation successfully. We believe this can be achieved through a combination of an appropriate methodology and through various types of sustained practice. Here we understand methodology to refer to the set of concepts which underlie methods, with methods referring to how practice is brought about in the materials and classroom activities. We do not subscribe to the notion that natural conversational language will simply appear or ‘be generated’ through activities either only loosely based on input material (a criticism mentioned by Shumin Reference Shumin, Richards and Renandya2002) or simply plucked out of the air because the activities seem to be enjoyable or creative, a view often reflecting the triumph of hope over experience.

The Special Nature of Conversation

The introduction above made a separation between conversation and speaking. There is no doubt that all kinds of speaking activities have a legitimate place in the language classroom, and we have already mentioned oral presentations. We could add to those information-gap activities (Doughty and Pica Reference Doughty and Pica1986; Richards Reference 29Richards2006: 18), drama activities (Maley and Duff Reference Maley and Duff2005), discussions (Lam and Wong Reference Lam and Wong2000; Green et al. Reference Green, Christopher, Lam, Richards and Renandya2002), recall protocols and role plays (Yuan Reference Yuan2001; Golato Reference Golato2003), teacher-directed storytelling (Jones Reference Jones2002), interviews with native-speaking informants (Mori Reference Mori2002), task-based activities (Ellis Reference 27Ellis2003), as well as pronunciation practice and activities involving repetition and drilling. Useful as such activities may be as different forms of language practice, clearly none of those types of speaking is what we would recognise as social conversation. And yet the ‘conversation class’ has for many decades occupied a place in syllabuses around the world; very often such classes specify no more in the syllabus than a broad topic launched by the teacher that learners are expected to elaborate on.

Some points which may seem fairly obvious need, nonetheless, to be made and borne in mind when discussing the notion of practice in relation to conversation. Conversation is a social act, it involves more than one participant, it takes place in real time and, in the case of everyday informal talk, it is unpredictable. In the real sense of the word, conversation is creative, in that the product we observe post facto is a jointly constructed process where no party can be sure of the precise nature of the next move as the conversation unfolds (for a broader discussion of creativity in everyday conversation, see Atkins and Carter Reference 26Atkins, Carter, Gee and Handford2012). Conversation is quintessentially co-constructed (Clancy and McCarthy Reference Carter and McCarthy2015). Elsewhere, McCarthy (Reference McCarthy2010) has argued that conversation also meets the demands of creating and maintaining ‘flow’, and more specifically, ‘confluence’, that is to say, a perception of a jointly produced fluency over and above that achieved by any individual speaker. We also know that conversation is subject to conventions of turn-taking (Sacks et al. Reference Sacks, Schlegoff and Jefferson1974) and that strategies are exercised by all parties in terms of appropriate openings and closings, management of topics and, above all, management of the interaction and the state of play in terms of shared knowledge and the relationship among participants. Conversation analysts have long recognised most of these features as central (for a useful summary, see Sidnell Reference Sidnell2010). Such features necessarily involve particular linguistic repertoires, raising the question of how feasible it is to incorporate them into the syllabus and, in the present case, what can meaningfully and usefully be practised, and how?

The above is the broader picture as regards conversation. It is also clear from corpus linguistic investigations that a specific linguistic repertoire is associated with it, in the sense of lexis and syntax, which, we argue, is a crucial element in what Hinkel (Reference Hinkel2006: 115) refers to as ‘a sufficient lexico-grammatical repertoire for meaningful communication to take place’. Buttery and McCarthy (Reference Buttery, McCarthy, Gee and Handford2012) have demonstrated the existence of a set of vocabulary with a high frequency of occurrence in spoken interaction, which distinguishes it from writing, although there is considerable overlap between the vocabulary found in both modes (for further discussion of the vocabulary of conversation, see also McCarten Reference McCarten2007). Even more notable in recent years has been the recognition of a grammar of speaking and in particular a grammar of conversation (Carter and McCarthy Reference Carter and McCarthy1995; Biber et al. Reference Biber, Johansson, Leech, Conrad and Finegan1999; Leech Reference Leech2000; Rühlemann Reference Rühlemann2006; Carter and McCarthy Reference Carter and McCarthy2015).

Both the lexicon and the grammar of conversation are characterised by their focus on successful interaction and the relationship between speakers and listeners, as opposed to acting merely in support of message topic and content. A further, related insight from corpus linguistics has been the growing evidence of the ubiquity of chunks in conversational language. Chunks, or formulaic utterances, lexical bundles, clusters, prefabricated items, as they are variously known (Wray Reference Wray2000; Schmitt Reference Schmitt2004; Biber Reference Biber2009; Martinez and Schmitt Reference Martinez and Schmitt2012), have been claimed to account for up to and exceeding a half of the linguistic material in conversational transcripts (Erman and Warren Reference Erman and Warren2000). Chunks are, of their nature, a fusion of lexis and grammar; such that what we might teach and practise for the purposes of fostering natural conversation should better be termed the lexico-grammar of conversation. The fact that chunks are all-pervasive, pre-fabricated and automatically retrieved (i.e. not assembled at the time of utterance) has implications for practice in relation to the development of fluency (Hunston Reference Hunston, Reppen, Fitzmaurice and Biber2002; McCarthy Reference McCarthy2010). Automaticity has long been recognised as a significant factor in second language acquisition (Gatbonton and Segalowitz Reference Gatbonton and Segalowitz1988; DeKeyser Reference DeKeyser2010), and there can be no doubt that chunks make a major contribution to the production of fluent strings of speech and interaction.

The insights discussed above have come mostly from the analysis of spoken corpora, a resource not available to the present authors when we began teaching conversation classes 40 or more years ago. Now we find it hard to attach much credibility to conversation classes, speaking syllabuses and materials for the teaching of speaking that fail to take into account the insights from corpora which tell us how the delicate fabric of successful conversation is woven. But it behoves us to lay out how such insights can be (and have been) translated into teaching materials and into the types of practice which will best lead to consolidation of the input and to acquisition of the skills and linguistic repertoires involved in conversational interaction. This we undertake in the following sections.

The Special Nature of Methodology for the Teaching of Conversation

In the previous sections, we claimed that conversation was, in a sense, special within the general notion of ‘speaking’. Following from this is the attention to appropriate methodology which the teaching of conversation calls for. The present authors worked for many years within the traditional PPP paradigm of Presentation–Practice–Production (Richards Reference 29Richards2006: 7) (a paradigm also discussed in Chapters 2 and 5). The input, typically sentences chosen as epitomes of a target structure or lexical item(s), was presented to the class and then drilled to a satisfactory level of accuracy before being produced in controlled or freer activities. Tsang and Wong (Reference Tsang, Wong, Richards and Renandya2002) comment on such preoccupation with form and lack of attention to conversational processes and refer to these as the ‘counterproductive aspects of a traditional conversation class’ (p. 213).

The three Ps are a tried-and-tested formula. Yet they exhibit troublesome flaws in regard to the teaching of social conversation. Firstly, conversation cannot be presented in exemplary sentences and can only be illustrated through extended texts, either in the form of concocted dialogues, edited conversational transcripts or raw, unedited ones. Artificially created dialogues have been compared with naturally occurring talk and often been found wanting (Burns, Gollin and Joyce Reference Burns, Gollin and Joyce1997; Carter Reference Carter1998; Gilmore Reference Gilmore2004). The unpredictability of real conversation also means that target items may occur only here and there, or only once or twice in the chosen conversation; therefore, the methodology calls for measures to draw attention to the target items. Practice, needless to say, will be quite different from the practising of sentence-based material. Furthermore, the target items, if they are strategies in the broader sense, that is to say not necessarily lexico-grammatical items, may not be visible or apparent at all to the learner. Finally, the types of production need, as best as possible, to mirror what conversation is really like. All of this does not necessarily imply always presenting entire conversations (it is often difficult in any case to define what a ‘complete’ conversation is) or requiring learners to reproduce them. However, conversational extracts need to be long enough to illustrate the unfolding of the strategy, to enable the process of authentication referred to earlier and to give learners enough context to practise the strategy naturally.

For this reason, it can be suggested that to better practise conversation, we adhere to a different set of methodological concepts, better termed as the three Is: Illustration–Interaction–Induction (Carter and McCarthy Reference Carter and McCarthy1995), as summarised in Table 1.1

The second stage, interaction, is the hub around which the whole process of teaching conversation revolves. It is the crucial practice stage where two aspects of knowledge and skill are honed: noticing and awareness of the target features and knowledge of their nature and use. Language awareness as an integral feature of language learning has been promoted by a number of scholars over the last couple of decades (Van Lier Reference Van Lier1998; Clennell Reference Clennell1999; O’Keeffe and Farr Reference O’Keeffe and Farr2003). In particular, Hughes (Reference Hughes2002: 59–61) argues that a language-awareness-based methodology should be judged not by the quantity of speech produced by learners but by the depth of understanding achieved of why speakers routinely make the choices we observe them making in natural settings. This does not imply that the stages of the method are viewed as purely receptive; the very act of interacting with and practising the target elements is in itself conducive to better inductive skills and increased awareness.

Categorising Items and Skills to Be Taught and Practised

Although we suggest that conversational skill encompasses two aspects, what we termed the broader picture of turn-taking, topic management, openings, closing, etc. alongside the specific linguistic repertoires of the lexicon and grammar of speaking, the two should be seen as complementary and working in harmony to create interaction. That said, many common strategies may be described as techniques for which there is no identifiable set of target language items. Their linguistic realisation depends entirely on the content and topic of the conversation, for example, asking a follow-up question, checking understanding or reacting to what is said by using a word from the previous speaker’s turn (A: She works as a stress counsellor. B: A stress counsellor?).

However, many conversation strategies are often realised through particular grammatical conventions, or involve particular vocabulary items, or both. In some senses this facilitates practice, since what teachers and learners are most familiar with as material for practice are grammatical configurations and lexical combinations. Conventionally, as mentioned earlier, in the conversation class, practice is often carried out as freer speaking on topics generated by the syllabus, by a unit or text in the coursebook, by the socially sensitive teacher or by the students themselves. Freer speaking is difficult if there has not been sufficient preparation or, we argue, insufficient practice taking the students from controlled activities to more open-ended scenarios. But what exactly is it that needs to be practised and how can the material be organised?

McCarten (Reference McCarten2007, Reference McCarthy2010), McCarthy and McCarten (Reference McCarthy, McCarten, Mishan and Chambers2010, Reference McCarthy, McCarten, Hyland, Chau and Handford2012) categorise conversational strategies and the lexical and grammatical items used to realise them for the purposes of building a pedagogic syllabus into four main areas, which are listed with examples below.

1. 1 Managing the conversation as a whole includes starting and ending conversations, changing topics, revisiting earlier topics, etc. Conversation practice is not just practice in talking about one’s job, one’s holidays, one’s family, etc., it is also a question of practising the organisation of such talk in interaction.

2. 2 Constructing your own turn includes features such as taking time to think, hedging, and expressing attitude and stance. It also includes the linear assembling of spoken utterances in real time and strategies such as elaboration and reiteration. Most crucially, it involves linking one’s turn to the previous speaker’s turn to create continuity and confluence and, to that end, paying careful attention to how one opens one’s turn (Tao Reference Tao, Meyer and Leistyna2003; McCarthy Reference McCarthy2010; Evison Reference Evison2013). Practising such linking will be important and should by no means be seen as a skill to be reserved for the advanced levels. Even the most elementary-level conversation can and should be linked cohesively and coherently, one turn to another, to be considered a successful act of communication rather than a succession, however accurate, of lexico-grammatical constructions. For example, the use of Well to answer a yes–no question in a non-yes–no fashion, or simply to take time to think, can easily be illustrated and practised at early elementary level.

3. 3 (Good) listenership, as described by McCarthy (Reference McCarthy, Reppen, Fitzmaurice and Biber2002, Reference McCarthy2003), involves responding appropriately to show both understanding and engagement. Central to conversation is the quick alternation of roles – speakers become listeners and vice versa. How one behaves as a listener, what one does and says, is just as important as what happens when one is the speaker. This will be a significant feature of practice.

4. 4 Taking account of others refers to the language choices speakers make depending on who their interlocutors are: friends and family or strangers, peers and equals or superiors, and this category addresses issues of formality, politeness and the projection of new and shared knowledge. Choices have to be made in real time, requiring considerable skill and, once again, practice.

In terms of linguistic components, the language required to realise many strategies can include single-word discourse markers as well as chunks or fixed expressions of varying length, and it is often possible to identify or create a coherent if not closed set of linguistic items that can realise a particular discourse strategy (a feature reported on in the early seminal work on discourse analysis by Sinclair and Coulthard Reference Sinclair and Coulthard1975). Examples (not an exhaustive list) of strategies within each category above and attested lexico-grammatical realisations, derived from corpus evidence, are given in Figure 1.1

Figure 1.1 Examples of conversation strategy types and possible lexical realisations

Depending on the level of students, any set of items can be reduced or expanded accordingly. For example, ‘taking time to think’ can be taught with a simple use of Well or Um. Responding to news with That’s … can be expanded with the addition of a wide range of positive adjectives (cool, lovely, awesome, etc.), and can be contrasted with reactions to negative news (That’s awful/dreadful/so sad, etc.); single-word exclamatives can also be added (e.g. Wow! Gosh!, etc.)

Other strategies may involve the use of grammatical structures, examples of which are presented in Hughes and McCarthy (Reference Hughes and McCarthy1998), depending on the context of the conversations and relationships of speakers. For example, Taking account of others (see Figure 1.1) may involve the strategic use of a past tense form of a verb when asking a favour politely of a boss or professor – as opposed to a close friend or spouse, e.g. I was wondering if I could ask you a question (see Carter and McCarthy Reference Carter and McCarthy2006: 604 for further examples). Forms like these can be viewed either as aspects of grammatical choice or as lexico-grammatical chunks. Either way, the goal of practice is to move ultimately to a level of automaticity that is the hallmark of fluency.

What we have called good listenership is an area of particular significance in relation to practice. The cognitive demands of speaking in class are often so great that students have little time to think about appropriate responses to their partners’ utterances and focus understandably on their role as possible or certain next speaker. And yet the evidence of spoken corpora as regards conversation is that listeners are anything but passive receptors who simply process incoming messages for comprehension. This is a basic human behaviour common to all languages, though there may be cultural differences in styles of response (e.g. on Spanish response tokens, see Amador Moreno et al. Reference Amador Moreno, McCarthy and O’Keeffe2013). McCarthy (Reference McCarthy, Reppen, Fitzmaurice and Biber2002, Reference McCarthy2003) identified a core list of common non-minimal response tokens in English used by listeners to show simultaneously comprehension and engagement and to contribute to the flow of the conversation. These response tokens represent a mezzanine level between simple back-channel utterances encouraging the current speaker to continue such as mm, uhuh, yeah (see Gardner Reference Gardner1997) and extended, content-rich turns as the listener assumes the role of speaker. Non-minimal tokens include common adjectives and adverbs such as good, fine, right, definitely and absolutely and show engagement without taking over the floor. Practice in the non-minimal response tokens can be given from the earliest stages of learning, for example, basic lexical items such as Good, Great! and simple phrases such as That’s nice/awful. Developing automaticity through such practice lowers the cognitive stress levels of taking the turn. It not only acts as a spur to richer, more natural and fluent interaction but also functions as a monitoring system for the teacher. If the students are proffering inappropriate responses, then the chances are that there is a comprehension problem somewhere that needs addressing. Thus, systematic practice in good listenership can provide useful evidence of ‘listening comprehension’ as well. In the world outside of the classroom, we show we have understood not by answering comprehension questions but by responding appropriately, and good practice in the classroom can reflect those natural conditions more effectively. Finally, listener behaviour illustrates the fact that in typical conversations, all speaker turns apart from the very first opening one are responses to previous utterances (see Tao Reference Tao, Meyer and Leistyna2003 on the functions of turn-openings). One could, hopefully without too great a risk of oversimplification, state that the teaching of conversation is the teaching of how to respond, and therefore that appropriate response should be a recurring, core feature of practice in the conversation class.

Differences between Traditional Practice and Practising Conversation

Traditionally, practice exercise types for grammar and vocabulary mostly rely on the fact that specific target items can be said to be correct or incorrect in terms of formal lexico-grammatical rules. For example, very often only one form of a grammatical paradigm is correct, as in the case of third-person singular simple present verbs, and there are spelling rules and irregular forms to learn for plural nouns. Word order issues when taught and practised are often somewhat rigidly defined, such as interrogative forms with auxiliary do/does/did, etc., or the placement of adverbs in relation to verbs and their objects.

The practice of many of the items which realise conversation strategies cannot rely on such formal constraints. They often take the form of fixed expressions such as at the end of the day (used to conclude or summarise), I guess so (used to converge, accept or agree) or If I remember rightly (used to hedge an assertion) and cannot be said to belong to a paradigm (e.g. as a convergent non-minimal response, the form is always I guess so, not He/She guesses so). Others such as the vague category markers are semi-fixed, such as and all that kind of thing/sort of thing/sort of stuff, and speakers have a freer choice as to which version to use. Equally, in relation to conventional sentence syntax, spoken language often displays freer word order, so an item like you know, I guess, or I suppose can occur almost anywhere in an utterance, and the verbs lose their written-language requirement of an obligatory object. Similarly, common adverbs such as still, even, probably, definitely often occur utterance-final in conversation (e.g. I can’t find my keys still. / I don’t know whether they have a passport even). Final position is not obligatory, but in writing it is extremely rare and is likely to be marked down in formal written examinations. Consequently, typical practice activities that ask students to put an item in the correct place in a sentence are difficult, if not impossible, to create for natural conversation.

Speaking and Writing: Uneasy Companions

Many of the most popular coursebooks over recent years have tended to use predominantly written texts as presentation vehicles and as a starting point for both writing and speaking practice. As a result, the models of grammar taught have tended to be based on written rather than the spoken norms identified by corpus linguists such as Biber et al. (Reference Biber, Johansson, Leech, Conrad and Finegan1999) and Carter and McCarthy (Reference Carter and McCarthy1995, Reference Carter and McCarthy2006). Until recently, scant attention had been paid to the practice of strategic uses of grammatical structures, such as the use of the historic present in storytelling (Schiffrin Reference Schiffrin1981) or past continuous verbs in speech reporting for the purposes of projecting newsworthiness (see McCarthy Reference McCarthy1998: ch. 8; Carter et al. Reference Carter, McCarthy, Mark and O’Keeffe2011: 468).

The practice of some spoken language forms poses a pedagogical dilemma by the very fact that the vehicles for their presentation and practice must often necessarily be written down. They will inevitably contain examples of informal, conversational usage that would be inappropriate in many types of written English, for example the vague expression and stuff. Perfectly acceptable syntactic phenomena in spoken language are frequently regarded as incorrect usage in written language or marked down in written examinations. An everyday example is the use of there is with a plural complement (There’s three restaurants in that street), a form which has almost become an educated standard in informal, unrehearsed speaking. A further example is so-called left and right dislocated structures (a metaphor which is anyway dependent on the Western convention of left-to-right script) as in the examples below (see Carter et al. Reference Carter, McCarthy, Mark and O’Keeffe2011: 236–237), which enable speakers to organise, announce and evaluate topics.

There is thus a risk that students will transfer spoken usages to their written work unless the presentation and practice are accompanied by warnings about their appropriate use.

Correct Choice versus Appropriate Choice

In traditional grammar (and vocabulary) practice, the target item must be chosen or used – correctly – and the practice vehicle, often a sentence, generally provides all the contextual cues students need (see Cook Reference Cook2001 for an interesting discussion of example sentences). However, so much of conversational practice involves the student in making a choice as to whether to use an item at all, particularly items that express attitude and stance. In natural conversation, only the speaker can decide whether to use the sentence-adverb obviously to signal information he or she takes to be clear/shared, or I guess as a hedge, or a present-tense verb to dramatise the narrative when telling a story. In the case of grammar, Carter and McCarthy (Reference Carter and McCarthy2006: 6–7) have referred to this as the ‘grammar of choice’ as opposed to the grammar of rules. Practice, therefore, has to give the student the flexibility to choose whether to use an item or not and the strategic non-use (or avoidance) of an item can pose an issue for the teacher in deciding whether the item has been truly learned. This also raises a problem which Buttery and Caines (Reference Buttery, Caines, Tono, Kawaguchi and Minegishi2010) have referred to as ‘opportunity of use’: how does one create natural opportunities for students to use the desired target items in the real-time and unplanned world of conversation where interaction is as important as, or even more important than, the transaction of information? Creating opportunities of use may require flexibility of the time devoted to an activity, pre-planning of language to be used in an activity and opportunities to rehearse, reflect and redo activities.

We have already discussed above the importance of good listenership: the language of response to show understanding and engagement in the conversation. This is an area which epitomises the importance of appropriate responses, and not just lexico-grammatically accurate ones. At elementary levels this can involve simple reactions to good or bad news using adjectives, or expressions with That’s + adjective, or a limited range of other response tokens (Really? I bet, etc.). An example of how listenership might be practised at a low level is shown in Figure 1.2, which comes from a lesson that teaches the strategy of responding to narratives, in this case stories about things that went wrong. In Part A, students listen to a narrative and at each pause, which replicates a response opportunity for an interlocutor, they choose one of two predetermined responses, only one of which is appropriate at that point. This more controlled phase of the activity set may serve teachers as a check on students’ comprehension of the story, as we intimated earlier. In Part B, students listen again and write their own responses, exercising a freer choice of items. They can of course also say them out loud. A possible preparation for (or even follow-up to) this activity could be to ask students to suggest in class the range of responses they have learned which are possible in each case to generate and recycle the repertoire of language items to choose from. Finally, in Part C, students get a chance to take turns telling and responding to a narrative as a free practice phase. They are given the choice of retelling the story they just listened to in their own words in case they are unable to summon up a simple or appropriate incident to tell from memory.

Figure 1.2 Listenership activities from McCarthy et al. (Reference McCarthy, McCarten and Sandiford2014: 91)

A more complex response type is where target forms cross turn boundaries as in co-constructed clausal configurations (see Tao and McCarthy Reference Tao and McCarthy2001; Clancy and McCarthy Reference Carter and McCarthy2015), where more than one speaker contributes items to a single proposition, as in the example below.

Speaker A adds to his own first turn Some teachers were only interested in Ancient Civilisations with two which comment clauses, at the same time agreeing with and mirroring the which comment clause added by speaker B (which is nice). B’s comment and A’s follow-up, both framed as which-clauses, are examples of listenership and choosing how to construct one’s turn working in harmony. The which-clauses are not obligatory (B could have just said That’s nice and A could have added Yeah, it’s good, and I liked it personally). The exchange is also an example of confluence at work, i.e. the ability of one speaker to link their turn to another’s. In this perspective, fluency is seen as a property of conversational exchanges, rather than the individual speaker’s contribution (McCarthy Reference McCarthy2010). The practice of items like these places greater cognitive demands on students in that they not only have to formulate the content of their own turn but must also attend to what their practice partner says, as occurs naturally in real-life conversations outside of the classroom. This kind of activity requires more collaborative effort between students than traditional single sentence-based activities or activities where students exchange set-piece information or attempt to fill artificially generated information gaps.

Some Principles of Good Conversation Practice

A Further Example of Practice

Figure 1.3 shows a matching activity from a CEFR A2 level course (for details of the CEFR levels, see Council of Europe 2001). Prior to this activity, students have listened to, answered questions about, read and practised a conversation between two students, one of whom is overworking and not getting enough sleep. The target strategy is keeping the conversation going by commenting on what the other speaker says and asking a follow-up question. This is an openly realised strategy where students are learning and practising a technique, not a set of forms to realise the strategy. It has been ‘illustrated’ and has also been identified and explained. Students have then gone on to ‘interact’ with the conversation by finding the various examples of the strategy in it, an example of the practice of language awareness as discussed earlier. The activity focuses on one sub-topic from the theme of the lesson (sleep). This begins the controlled practice phase.

Figure 1.3 Phase 1 of a conversation strategy activity from McCarthy et al. (Reference McCarthy, McCarten and Sandiford2014), Level 2: 26

Students are first directed to match typical everyday comments about sleep with appropriate responses that show different realisations of the strategy. The items are within the students’ language abilities in that they recycle simple present verbs, adverbs of frequency and basic time expressions. The responses also recycle response tokens from previous lessons (Really? Me too, That’s good). Students first match partly on the basis of lexical clues (e.g. item 2, weekends, with (f), on Sundays; item 5, in lectures, with (a), professor). This is a controlled practice phase of the material and can be said to be test-like practice in that only one answer is correct. However, the challenge is within students’ abilities, and in doing the activity they gain further and repeated exposure to different realisations of the strategy. By reading these out with a partner, they gain further practice in phonological articulation of the items. Teachers can also ask students to suggest and then use other possible responses as a stepping-stone activity to a freer, more personalised stage.

The activity is subsequently phased from controlled to freer practice. The main value of the activity as practice comes in the next stage (Figure 1.4), where students use the numbered comments as the basis of statements about their sleep habits. In pairs, they then take turns making their true comments and responding in the manner that is modelled. In doing so, the practice should lead to our third methodological principle, ‘induction’, i.e. the understanding of, the appropriation of, and the personalised use of an item or strategy through repeated exposure to and manipulation of it in different settings.

Figure 1.4 Phase 2 of a conversation strategy activity from McCarthy et al. (Reference McCarthy, McCarten and Sandiford2014), Level 2: 26

An important criterion is that the practice material, i.e. the utterances and conversation extracts that act as the practice items should be representative of conversation in the sense that they should be examples of the kinds of things people say about any given topic or in a given situation. Furthermore, students should be able to recognise these as realistic and thus ‘authenticate’ them. The term authentic has different interpretations in regard to teaching materials. In one sense it describes the presentation texts in materials and is defined by Wallace (Reference Wallace1992: 145) as a real-life text ‘not written for pedagogic purposes’. In the case of conversation materials, we might amend this to include language that has not been spoken for pedagogic purposes. The considerable issues and challenges in using unedited authentic conversations in teaching materials are discussed in McCarthy (Reference McCarthy2010) and will not be re-examined here. However, another view of authenticity from Widdowson (Reference Widdowson1998) refers to the authentication of language, that is, the language is perceived as authentic and adopted as such by the learner. However much we may think a piece of material is authentic, without the learner’s own authentication of it through meaningful and personalised practice, the learning objective may well fail in certain respects, especially those relating to motivation (see also McCarten Reference McCarthy2014). Griffiths and Keohane (Reference Griffiths and Keohane2000: 1–2), based on their classroom experience, report that personalisation is strongly linked to motivation and that personalised activities help in making language learning memorable. Personalisation is also seen as a crucial element in promoting intrinsic motivation by Ushioda and Dörnyei (Reference Ushioda, Dörnyei, Gass and Mackey2012), while Guilloteaux and Dörnyei (Reference Guilloteaux and Dörnyei2008) list personalisation among the teacher practices which support task motivation in their classroom-observational study. A further consideration with regard to judging authenticity is whether the learner’s use of the language is perceived by others as authentic – that is, natural and fluent – as a result of learning features of naturally occurring conversation, as is suggested by Hasselgreen (Reference Hasselgreen2004).

As such the practice phases support students in assembling their own comments and exchanges, which can be modelled on or draw heavily from the practice items. The practice items constitute a springboard for personalisation, enabling students to use them as a basis for saying something true about themselves, their experiences, family and friends or views. This means practice items should not include many (if any) references to third-party fictional names and events or too much assumed shared knowledge between speakers or events that cannot easily be related to the here and now or adapted to the students’ own experience or world knowledge. The items should be generic in the sense that students should be able to see how to adapt them easily by making simple changes, e.g. subjects (My brother to My friend Marco), verb forms (affirmative to negative) and adverbs (sometimes to never).

Classroom Considerations

Another principle is to use practice activity types that are familiar to teachers and students and therefore easy and straightforward to set up and manage. Many of the concepts involved in conversation strategies might be new to teachers, as well as students (e.g. adding a which comment clause to a previous speaker’s comment and the so-called dislocated structures mentioned above). Where the concepts are new to teachers, it is important to make their explanation and practice simple, familiar and practicable so that the teaching and practice are achievable and successful. Students should equally be able to see what is required of them. Traditional activity types such as gap-fill, matching and choosing correct options, can all be effective ways to achieve more controlled practice phases in the teaching of non-traditional conversational features such as co-constructed clauses, follow-up questions, the use of particular chunks and so on. Figure 1.5 shows a familiar activity type of a conversation gap-fill, but the target items – which comment clauses – may be less familiar.

Figure 1.5 Conversation strategy practice of which comment clauses from McCarthy et al. (Reference McCarthy, McCarten and Sandiford2012), Level 1: 24

Conclusion

Throughout this chapter, we have argued that conversation is not the same as the broader notion of ‘speaking’ and that its special qualities require special consideration from a number of angles. We may summarise these special considerations here in the form of 10 key criteria for the teaching and practice of conversation:

1. 1 Language description: Conversation skills, strategies and language characteristically used to realise them can be identified thanks largely to corpus research.

2. 2 The strategies and language can be coherently organised for teaching purposes.

3. 3 The methodology for teaching conversation cannot rely on traditional PPP steps, especially those involving decontextualised sentences as vehicles of presentation.

4. 4 Practice in conversation involves more than working towards accuracy and/or complexity. The goals of conversational practice are fluency, confluence and successful social interaction.

5. 5 The methodology proposed and exemplified is best seen as founded on the three Is: Illustration, Interaction and Induction.

6. 6 Illustration involves presenting conversational extracts in the contexts in which they naturally occur, and based on the evidence of natural spoken corpora.

7. 7 Interaction refers not just to the interaction between learners in activities, but also to learners’ interacting with the text/transcript by noticing, finding and coming to conclusions about linguistic features in conversation.

8. 8 Induction refers to learners’ adoption of a linguistic feature through repeated exposure to it in illustrative and practice material they encounter.

9. 9 Fluency practice can and should be principled by observing criteria such as being within learners’ linguistic comfort zone and including representative examples of naturally occurring talk as the basis for learners’ authentication and induction of target items and strategies.

10. 10 Above all, to be successful, practice should afford learners overt opportunities to personalise the language, allowing them to use the target item in saying what is true for them about their world and engaging confidently in confluent conversational activity in the L2 as they might in any other language they speak.