A RT I C L E

603

Arguing against absent arguables: a study of audience participation in political discourse

N I C K L L E W E L LY N

Discourse Studies Copyright © 2006 SAGE Publications. (London, Thousand Oaks, CA and New Delhi) www.sagepublications.com Vol 8(5): 603–625. 10.1177/1461445606064832

U N I V E R S I T Y O F WA RW I C K

A B S T R A C T Based on the analysis of interaction during a public meeting, this article considers how people argue in sequential environments where direct interaction is precluded. The meeting in question was organized so the turns of audience speakers and local authority representatives were produced during different periods; initial actions and their oppositions, counters, etc., could be separated by anything up to 25 minutes. The article describes how speakers adapt their language practices to construct arguing turns and series of actionopposition pairs in social settings thus organized. KEY WORDS:

arguments, conversation analysis, public meetings

Introduction In previous work, conversation analysts have explored how people produce arguments; describing how they start (Maynard, 1985; Hutchby, 1992a), evolve (Coulter, 1990; Hutchby, 1992b), escalate (Goodwin and Goodwin, 1987) and terminate (Dersley and Wootton, 2001). Overwhelmingly, studies have concentrated on interactional environments where the turns of interlocutors are serially and sequentially adjacent; where ‘arguable actions’ (Maynard, 1985: 3) are immediately followed by some counter, which is itself opposed leading to the development of an argument (see Garcia, 1991, as an exception). The present article contributes to the understanding of argument by describing how people dispute, oppose and reject arguable actions – instead – in sequential environments where this condition does not hold, where time and conversational space are inserted between the turns of protagonists. The analysis is of a public meeting held in London (UK) in 2002. This was organized by the London Borough of Haringey to discuss proposals for resolving traffic problems in the suburb of Muswell Hill. A large number of people attended

604 Discourse Studies 8(5)

the meeting, approximately 400, which was convened in a local church. The audience was almost entirely opposed to the council’s plans, which it was argued would lead to more cars driving through the (at that point) quiet backstreets of the suburb. The meeting itself was managed in a novel fashion. In contrast to ordinary conversation, audience contributions and their responses were produced in different periods. A series of audience contributions would be ‘collected’ by a chair; in a separate period councillors would respond in a single extended monologue. Once an ‘answering’ monologue had been completed, the chair would solicit the next series of audience contributions. And so on. The meeting in question consisted of four discernable Q-A cycles. The smallest of which contained five audience turns (done in 5.26s) followed by a 1.43s monologue; the longest consisted of 12 audience turns (done in 19.03s) followed by a 7.49s monologue. How might actors assemble series of sequentially responsive social actions in such settings? Consider what would be involved. Suppose an initial audience turn in questioning period A. This period would have to end before a Local Authority (LA) speaker could respond to the audience contribution in answering period A. Suppose, during the LA speaker’s turn, the initial audience contribution was heard to be addressed what then? Answering period A would have to come to an end before a new audience speaker could ‘pick up’ the line of debate and respond to it in the next questioning period B. For the sequence to develop a stage further, questioning period B would then have to end before a further LA speaker could respond in answering period B. And so on. In the main, arguments are understood to involve ‘adjacent, directly addressed exchanges of oppositional turns’ (Garcia, 1991: 819), an initial opposition is immediately followed by some counter, which is itself rebuked leading to the development of an argument. In these terms, ‘collecting questions’ eliminates argument. But, and importantly, it does not stop people producing the basic building blocks of arguments, series of interlocking ‘action-opposition’ pairs (Hutchby, 1996). The type of sequence in Figure 1 is perfectly possible. In Garcia’s data, for instance, where turns were separated by long periods and addressed to a third party (mediator), interlocutors still produced the basic action-components of argument sequences; accusations were followed by denials, for example, ‘I had nothing to do with it’ (Garcia, 1991: 829). Actors display an ability to produce and recognize sequentially adjacent activities despite their separation into different periods. One task for applied research is to 1 2 3 4 5 FIGURE

LA1: AUD1: LA2: AUD2 LA3

Assertion Counter. Rejection. Counter. Rejection.

1 . The sequence type to be examined

Llewellyn: Organizing audience participation in political discourse 605

describe how such sequences are produced, thereby detailing the consequences of institutionally specific turn-taking conventions for the practical methods through which members of society dispute, contest and oppose what has gone before. One obvious practical problem concerns how turns are ‘tied together’ (Sacks, 1992: LC1, 150) in settings where initial actions do not open-up discrete slots where responses should be produced. In Figure 1, how does AUD2 show they are countering LA2’s turn and not some other previous contribution; how does LA2 show they are rejecting AUD1’s counter of LA1’s original assertion? In an environment where multiple ‘arguables’ are floating about, how do participants establish which turn is being opposed? As an analyst Garcia (1991) demonstrates an ability to do this; she identifies a denial to an accusation produced some 30 minutes previously. But she does not explicate how she did this. What features did that denial have that made it recognisable as a rejection of that accusation? Of course, this is not primarily an analyst’s problem. First and foremost, it is a practical problem for people tasked with hearing and producing responses in such environments. At one point in the present data, for example, an LA speaker produces at least 10 ‘responses’ (to previous audience contributions) in a row; 10 times he faced the practical problem of how to display a connection between his ‘response’ and a previous audience contribution. So one matter is how this was done, given such turn-taking conventions, how do interactants make it apparent that there is something akin to ‘a conversation going on’ (Sacks, 1992: LC1, 150)? A second notable characteristic of ‘collecting questions’ is that the ‘cut and thrust’ that normally characterizes argument (Hutchby, 1992b) is replaced by something quite different. Instead, the action components of arguments are produced within monologues. Speakers are given slots – with defined boundaries – where they can assemble extended complaints, accusations or counters. In the empirical materials, LA speakers never once spoke during a ‘questioning’ period. Audience turns were only ever ‘interrupted’ by outbreaks of audience applause. A further question thus concerns the implications of this for the design and overall shape of oppositions and arguing turns. Speakers were able to build their turns over the course of three to four minutes. In the context of audience participation television chat shows, Hutchby (1999: 253) suggests that contributions took the form of either three part ‘recursive’ (position taking, accounts, recapitulation) or ‘progressive’ (preface, account, position statement) presentations. Such rhetorical forms seem to handle practical problems that arise where audience members are ‘given the floor’. They bring turns to ‘hearable completion’, enabling the host/chair to move the talk on and the overhearing audience, if they wish, to produce a collective response. The present article aims to describe consequences of separating turns into different periods for the practical methods through which members of society dispute, contest and oppose what has gone before. As virtually all studies of argumentation, Garcia (1991) aside, have been based on settings that allow for

606 Discourse Studies 8(5)

direct interaction, how separating the turns of disputing parties shapes spoken discourse remains unclear. As an interactional model for public consultation, does ‘collecting questions’ effectively stop debate and contestation? If so, broader questions will be raised about the assumptions underpinning interactional provisions for public consultation.

The target data The analysis focuses mainly on a sequence sketched below. Characterizing this as an ‘argument sequence’ is not unproblematic, as mentioned above. The topic of the talk was the council’s approach to consultation on a road closure. A skeletal outline is presented below.1A great many details have been omitted, including those pertaining to speech production (intonation, etc.), gross audience reactions and shouted contributions. Only a fraction of each turn is presented. Q&A1

T1/0.00/Q3/ AUD1:

T2/4.00/A4/ LA1:

T3/7.00/

Q&A2

LA2:

T4/11.40/Q3/ AUD2:

T5/16.33/A3/ LA1:

Q&A3

T6/25.15/Q6/ AUD3:

T7/26.50/Q7/ AUD4:

‘where will the council draw the line for the consultation. . .’ ‘my guess is. . .most weight to people who live in the local area’ ‘it’s unrealistic to expect everyone in the ward can have their say’ ‘I do disagree with the speaker that. . .this is a serious issue for the whole area. . .’ ‘I agree the issue of the whole solution not in parts I’m afraid it’s not’ ‘what you want to do rather than what people in the whole of muswell hill want ‘It’s not fair to. . .shift things around and then sit back and sought of say. . .

It is worth briefly fleshing out the data. Initially, AUD1 asks where the line will be drawn on the consultation; making the point that people who live three miles away (her example is people in ‘Enfield’) will not be effected by the proposed changes. In turn two, LA1 closes his turn by arguing these matters are yet to be worked out. Initially, he partially agrees with AUD1’s position (‘you are quite right that’), before adding a clause (‘but others will have a keen interest in what is going on’). His turn is produced despite audience heckles and loud collective disaffiliative noise. Upon the completion of LA1’s turn, which is greeted with roughly 10 seconds of loud disaffiliative noise, LA2 self-selects to make a position statement; that it is

Llewellyn: Organizing audience participation in political discourse 607

‘unrealistic’ to expect that ‘everyone’ should have a say. Instead, there should be ‘neighbourhood-based consultation’. This turn is produced despite a range of similar audience reactions. At a number of points LA2 is simply holding the floor. By warrant of its position in discourse, LA2’s turn can be heard ‘in light of ’ both AUD1’s question and LA1’s initial response. It seems to address AUD1’s question, whilst being more concrete than LA1’s initial ‘guess’. For the sequence to be examined, this is the ‘antecedent event’ (Eisenberg and Garvey, 1981: 51), the turn which is subsequently opposed. LA2’s ‘spontaneous’ contribution is the final act of question and answer period one. It was the only spontaneous LA action of this type. The chair now moves to collect a new series of audience questions. The third turn of this new series is produced by AUD2, who opposes a position attributed to LA2; that the best way to conduct a public consultation is by splitting people into groups based on where they live. AUD2 argues that everyone should be brought together in a single meeting, where differing points of view can be discussed. No more ‘questions’ in this period are produced on this topic. Following two more questions, the chair asks LA1 to produce ‘snappy’ answers. In what is hearable as the third answer in this short monologue (1m 47s), LA1 responds to AUD2. Initially, he partially agrees with AUD2’s point about the ‘whole area’, before then disagreeing with the gist of his argument. He makes the point that whenever you close a road a very large number of people oppose such plans, whilst a very small number support it. Whilst LA1’s turn is indexed to AUD2’s previous contribution, quite where it takes the argument is unclear. How does this basic fact (that whenever you close a road a very large number of people oppose such plans, whilst a very small number support it) undermine AUD2’s argument? Is the argument that consultation itself is a waste of time? Following LA1’s answers, the chair solicits more questions. Nearly 19 minutes after LA2’s ‘arguable utterance’ (Maynard, 1985: 3), AUD4 takes a turn, the seventh question in a series of 12. Initially, AUD4 aligns with the immediately prior audience turn (T6 by AUD3). He then produces a turn which can be heard to oppose LA1’s previous contribution (T5 in the sequence), produced approximately 10 minutes previously. In producing his turn, AUD4 shows an understanding of where LA1’s contribution takes the argument, that is, that he is ‘sitting back and saying well that’s just life’. As with those turns produced by AUD1, 2 and 3, upon its completion – and at various points during its production – AUD4’s turn meets with spontaneous audience applause. This is the end of the sequence, in the next answering period LA3 makes no hearable effort to respond to AUD4, in a monologue lasting nearly eight minutes.

Analysis In the following sections, three major features of the target sequence are considered: 1) how people tied their talk to previous utterances; 2) the matter of

608 Discourse Studies 8(5)

selective response; and 3) how speakers encouraged and accommodated gross audience reactions to their arguing turns. Taken together, these points provide an initial account of how separating the turns of protagonists into periods impacts upon the way in which people oppose, reject and counter what has gone before. INVOKING THE ARGUABLE

In the present setting, speakers who wished to respond to an utterance faced the practical problem of invoking that utterance, so the speaker of that utterance and the overhearing audience could hear what was being done. The audience could not know which turn was about to be addressed. If their social activity was to be recognized speakers had to display which turn they were responding to. This section examines this work. How it was done, where and with what consequences? Initially, it is useful to consider a minimal way of doing this, where the speaker does little to clarify a) which turn is being addressed and b) how it is being addressed. In extract one LA3 is in the midst of an answering monologue. Initially, he marks the closure of one response, before starting upon something new. (1) 1 2 3 4 5 6 7 8 9 10 11

LA3:

[Q&A3:LA3] the local authority will have to °work° wi:th (.4).hh speed enforcement was: huh ra:ised an’that is a p’ticular issue (.) generally ((continues 20.0s)) speed enforcements remains the responsibility of °the° metropolitan police service .h and huh

LA3 invokes an audience turn in two ways. He begins his response by glossing the previous turns topical domain (‘speed enforcement was raised’, lines 3–4). He then makes a point that speed enforcement is a matter for the police, not for the local authority. The relevance of this point, for the previous audience contribution, is not explicitly mentioned or even alluded to. When he has finished this point, he moves on to another response about ‘walk on buses’. In this case, how might it be possible to find a) which audience contribution was being addressed and b) how it was being addressed? As an analyst armed with a recording, it was possible to go over the previous talk in detail. When this is done it becomes apparent that, during the sixth audience contribution of the previous ‘questioning period’, AUD3 makes the argument that the council’s plans are not ‘common-sense’ (see extract 7 below). He reasons they will not be

Llewellyn: Organizing audience participation in political discourse 609

able to ‘enforce’ the speed limits being proposed. AUD3’s was the only turn on this topic produced during this period. Hence, it might be reasoned that LA3 was addressing this turn. When the turns of LA3 and AUD3 are brought together (in a transcript), it is also apparent how LA3 is responding to AUD3. He seems to be attend to – by rejecting – a presupposition contained within AUD3’s turn. AUD3 argued the council was unable to enforce speed limits, LA3 states speed limits are enforced by the police. But would this be apparent as the action happened? These turns were separated by conversational space (six audience contributions and approximately six LA responses) and time (over 20 minutes). At no point does LA3 display the relevance of ‘his point’ (about speed enforcement) for the audience contribution, he does not characterize the audience contribution in any way. He moves from ‘glossing a topic’ to ‘making a point’ without any intermediary stage that might clarify the sequential relevance of what he is saying. In terms of memory alone, there are clear challenges. In the main, and in contrast to LA3, speakers did a great deal of work to display both what they were responding to and how they were responding to it. Consider the following. (2) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

CHR:

AUD2:

AUD:

[Q&A2. Q3:T4] blue arm at the back there ((pointing)) (6.0) ((microphone passed)) hu::h could I just say that I- I do disagree with the speaker thatin fact the best form of consultation is huh (.) splitting (.) people up into gr:oups.h as ((name)) said at the beginning that- (.) this is a serious issue for the (.2) a:rea >and in fac [t there’s no point< [xxxxxxxxxxxXXXXXXX

In this instance, AUD2 labels his social activity as a sequentially responsive activity, an ‘opposition’ (he says ‘I disagree’, line 5) and – at least minimally – he invokes ‘the speaker’ he is arguing with (line 7). He then glosses the position he is attributing to ‘the speaker’ (that ‘the best form of consultation is splitting people into groups’, lines 9–10). It is worth reflecting on this work in a little more detail. Formulating talk as a disagreement with a previous ‘speaker’ – whilst signalling how a turn is responding to a previous utterance – cannot clarify what or who is being opposed. Where

610 Discourse Studies 8(5)

turns are serially adjacent (back to back) such items can do this work, people will presume the immediately previous turn is being opposed. In the present setting such items cannot perform this function, as there were always numerous speakers and numerous potential arguables. As such, efforts to gloss or characterize a position are of vital significance, if people are to recognize a speaker’s social activity. As a method for doing this AUD2’s gloss trades for its sense on the audience’s ability to recall a previous turn, in much the same way that AUD3’s did above (extract 1). Somebody just coming into the meeting would struggle to find what was meant by ‘splitting people into groups’. It only makes sense – it would seem – as a characterization of LA2’s previous point about the benefits of ‘neighbourhood based consultation’. In the previous answering period, LA2 suggested separate meetings should be held. It would seem that AUD2 is invoking this turn. But again, to find such connections audience members would have to remember LA2’s turn and hear ‘splitting people into groups’ that way, that is, as a gloss of LA2’s turn. This method of invoking an arguable (glossing a previous position) is also apparent below. Initially, LA1 is ending an answer. The indexical ‘that’ (line 1) invokes the topic (‘a point about shopkeepers’) of the response that is being brought to a close (with ‘that’s done’, lines 1–2). There is then a brief pause, before the speaker starts upon something new. (3) 1 2 3 4 5 6 7 8

LA1:

[Q&A2. A3:T5] so I- I- I- think that’s °done° (.) I a:gree the issue of the whole solution not in parts (.) ar: I- I’m afraid it’s not as s-s- simple as thth- spe- (.) ah- as you suggest ((continues))

Again, LA1 labels his turn as a socially responsive action (an ‘agreement’, line 2), before glossing a position (‘the issue of the whole solution and not in parts’, lines 3–4) attributed to an audience speaker (‘sp-’ is hearable as the repaired beginning of ‘speaker’, line 7). The overall project of the turn then becomes apparent, LA2 signals his upcoming opposition to AUD2 with the markers ‘I’m afraid that’ and ‘it’s not as simple as’ (lines 5–6). In this case, how might the audience hear that ‘the issue of the whole solution and not in parts’ is a gloss of AUD2 argument position (his disagreement with LA1, extract 2)? One way is through a degree of ‘rhetorical cohesion’ (Gruber, 1998: 475) between AUD2’s turn and LA2’s response. LA2 recycles, albeit imperfectly, a rhetorical device used by AUD2, namely ‘ (.2) a:rea’ (extract 2, line 15). In LA2’s turn this becomes ‘the whole solution’ (lines 3–4). Following this, AUD3 (T6 in the target sequence) then potentially carries

Llewellyn: Organizing audience participation in political discourse 611

this on, with the phrase ‘in the whole of muswell hill want’ (see extract 7 below). This contrast of ‘whole verses parts’, where ‘whole’ – in each case – is given a noticeable emphasis, would appear to be a further method of signalling the relevance of some talk for an ongoing series of turns. A second source of evidence, as with extract one, is that LA1 is attentive to presuppositions contained within AUD2’s turn. He argues that AUD2 is supposing these matters are ‘simple’; one purpose of the present meeting – LA1 argues – is to resolve them. Crudely, those rebuttals, rejections and counters which met with the heartiest affiliative or disaffilative reactions, were those which invoked previous turns in a more complete fashion, compared with the above, typically within ‘you say X, what about Y’ type devices (see Hutchby, 1992a: 675–82). Consider the following example. (4) 1 2 3 4 5 6 7 8 9 10 11 12

AUD:

AUD: AUD

[Q&A3] I- I find it very worrying (.) that TFL are talkin’ about (.2) talking t’people from enfield (.) and from haringey (.) whereas one of the major areas (that will be) affected is a very large chunk of ↑barnet (.) xxxxxxxxxxx [xxxxxxxxxxxxx [hu:h (.) s- so I- I think

As in the previous examples, AUD frames his talk as a responsive social activity by stating he finds a previous contribution ‘very worrying’ (line 1). He then reformulates the arguable action – attributed to TFL – in the context of a contrast structure that generates a sense of scepticism (Drew, 1990) about the merits of TFL’s plans. These types of rhetorical devices (‘TFL say X, what about Y’) might be especially well-suited to this type of interactional environment, precisely because they characterize two contrasting arguing positions. The result is something about which people can take a stance, without having to remember a previous turn. Consider a further example, this time involving an LA speaker. Here LA1 begins a new response by glossing a previous turn’s topical domain (‘there was an issue about traffic lights’, lines 2–3). At this point he could – as LA3 did in extract one – have simply made his point about traffic levels. But he does something additional, before making his point he re-presents the contrastive audience argument. The character of his social activity as an opposition is thus ‘very apparent’, evidenced – potentially at least – by the onset of a collective disaffilitative response prior to completion.

612 Discourse Studies 8(5) (5) 1 2 3 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

LA1

AUD: AUD:

[Q&A2: A4] and that’s all I’d say (.) there was an issue about the traffic lights an’th an’th an’the increase in traffic °(flows)° (.) .h I’m entirely sure th’speaker was right that the level of traffic after the *lights were put* in would have ↑increased .hh what I’m saying now that the level of traffic now (.) now that the lights have been in for some while and ( ) and the levels of traffic historically are at broadly similar lev [els to ( ) [N [O! [zzzzzzzzzzzzzz zzzzzzzz

Turns designed in this fashion appeared to be very effective in this setting, perhaps because they resolve problems (of memory and interpretation) peculiar to this setting, by making reference to both sides of an argument. As will be considered latter, and as seen above, such formats were also effective in organizing (affiliative and disaffiliative) audience reactions. People index back to previous turns (or do not) ‘for all practical purposes at hand’. If a speaker wishes to produce an opposition, a previous turn has to be invoked or reformulated in such a way that makes it ‘opposable’; a position has to be alluded to that can be challenged, rebutted or rejected in some way, upon some grounds. The grounds for the rejection have to be apparent in the way the turn is invoked or reformulated. In such a setting, it might have been imagined that speakers would routinely reformulate previous turns in mischievous ways; to make a position sound ridiculous or foolish (Hutchby, 2001). Remarkably enough perhaps, this did not happen. In the following extract, it might be suggested that AUD4 characterizes LA1’s turn (extract 12) in a slightly mischievous fashion, deploying attributed speech and – perhaps – an extreme formulation of LA1’s arguing position. (6) 1 2 3

AUD4:

[Q&A3. Q7:T7] to shift things aro:und (.) and then sit back and sought of say (.) well

Llewellyn: Organizing audience participation in political discourse 613 4 5 6

>y’know< that’s just life well I’m afraid it’s not just life (.) it’s not

In this instance, AUD3 is not claiming that LA had really said ‘that’s just life’. This is clearly a characterization of a position that AUD4 is attributing to LA1. And, it is clearly a characterization that is ripe to be treated sceptically. But even here, it is difficult to suggest that AUD4’s characterization is entirely mischievous. He is responding to LA1’s turn where he argued (against AUD2) that ‘whenever you close a road there are a few people in favour of it and a very large number against it’. In terms of AUD4’s reasoning, it is clear to see where the characterization ‘that’s just life’ comes from. This section has considered practical methods for invoking and arguing with a turn produced some time ago. It has considered work that speakers had to do (displaying what they were responding to and how they were responding) that would ordinarily be apparent given the gross positioning of turns in spoken discourse. In the materials, speakers were observed: 1) explicitly addressing how they were being responsive to a previous utterance (‘I disagree’, extract 2; ‘I agree’, extract 3; ‘I find it worrying that’, extract 4); 2) explicitly marking previous turns (‘was raised’, extract 1; ‘disagree with speaker’, extract 2; ‘agree with the issue’, extract 3; ‘TFL talking about’, extract 4; ‘speaker was right’, extract 5); and 3) invoking or re-presenting the target of their opposition. Practical ways of invoking an arguable have been described (glossing a turn’s topical domain, glossing a position, etc.) and contrasted in terms of the interpretative burdens they place upon audiences. SELECTIVE RESPONSE

In the sequential environment of ‘collecting questions’ arguable utterances and their oppositions are separated in conversational space and time. Imaginably, this might have implications for the way in which speakers are constrained (or not) by previous social actions. The present section considers this matter further, exploring two aspects of ‘selective response’ (Garcia, 1991: 829). It would be wrong to suggest (LA) speakers always or often ducked ‘difficult issues’. Indeed, as seen in extract five above, LA speakers often displayed the ‘arguable’ and controversial character of their talk. But at the same time, and despite there prevalence, LA speakers never once responded to an accusation or a complaint. Empirically, LA speakers took issue with such things as proposals and position takings, that is, that traffic levels had increased after traffic lights had been put in (extract 5 above) or that there should be a single meeting where conflicting views might be brought together (extract 3). This is worth pointing out because in ordinary conversation (Sacks et al., 1974) accusations are viewed as the first pair part of an adjacency pair, with denials typically being the preferred response (Atkinson and Drew, 1979) positioned ‘immediately after the accusation, without delay, accounts, or other mitigating techniques’ (Garcia,

614 Discourse Studies 8(5)

1991: 828). In this comparative sense, ‘collecting questions’ appears to be associated with some weakening of sequential constraints that are ‘ordinarily’ established by arguing actions such as accusations. Consider the following empirical example, which takes us back to the response about ‘speed enforcement’. As the extract begins, AUD3 has been speaking for approximately 25 seconds. (7) 1 2 3 4 5 6 7 8 9 10 11 . 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

[Q&A3, Q6:T6] AUD3/T6: it doesn’t take a genius to work out that if you introduce traffic calming ( ) or ( ) twenty miles an hour as someone suggested earlier (.) or thirty miles an hour which is th- which I believe is the law (.) then you must en↑force it ((continues 30 seconds)) AUD3/T6: meet the findings that you think that will support what you want to do (.) rather than what people in the whole of muswell hill= (.) AUD: xx [xxxxxxxXXXXXXXXXXX= AUD3: = [want to do and thAUD: XXXXXXXXXXXXXxxxxxx AUD3: ( = AUD: xxxxxxx ((5.0)) AUD3: you’re) shifting one problem to another part of ( ) muswell hill and it’s not democ- democracy its just it its its not democracy at all (.) its not even showing an interest in the ↑people (.) AUD: absolut [ely AUD: [xxxxxXXXXXXXXXXXXX

In his turn, AUD3 makes more than one point, which was a feature of over half of all audience turns in the meeting. He starts with ‘speed enforcement’ before moving to discuss the ‘consultation’. A series of accusations are produced

Llewellyn: Organizing audience participation in political discourse 615

that the council will only consider findings that support what they want to do, is not taking an interest in the people and is acting undemocratically. Some time later, LA3 responds by addressing a point about ‘speed enforcement’. Consider a second example, this time going back to AUD2’s turn. Having stated that he disagrees with LA2’s point (about the benefits of neighbourhood based consultation), AUD2 goes on to make a separate complaint/accusation, that LA speakers are ‘talking at, rather than listening to’ the audience. (8) 1 2 3 4 5 6 7 8 9 10

AUD2:

AUD:

[Q&A2, Q3:T3] (.) but the thing that I’m really disappointed about tonight (.) is that it seems to me that the panel are more intent at actually talking at us (.) than actually ↑listening t [o us [xxxxXXXXXXXxx xxxxxxx (5.5)

When LA2 responds, he responds to a ‘point’ about the ‘whole solution and not in parts’. He responds to a gloss of the more benign of two points raised by a previous speaker. Not all accusations were ‘second points’ (as in extracts 7 and 8). Where audience turns consisted solely of complaints or accusations, LA speakers simply did not respond. A whole range of argument moves were ignored. AUD4, for example, follows on from AUD3. He produces an extended turn that consists of strongly voiced complaints and accusations, that the council’s actions are ‘not fair’, that they are ‘setting one part of the community against another’ and that the consultation is not being done ‘properly’ (see extract 16 below). It received no hearable response in the period that followed. Of course, such a claim trades on the researcher’s ability to identify ‘absent answers’ and it is worth considering how this might be done. In ordinary interaction, not responding (at all) to an utterance that warrants a response such as a question, a counter assertion or a rebuttal is ‘noticeable’ – in no small part – because initial actions open up slots where the sequentially appropriate response should go (Atkinson and Drew, 1979). Of course, this does not apply in the present data. Consequentially, vital resources for identifying absent responses are removed. In addition, during the meeting it was common for numerous audience contributions to address the same topic, for example, increased traffic flows. When LA speakers invoked a previous turn by glossing its topical domain (there was a point about increases in traffic) it becomes extremely difficult to establish exactly which turn was being addressed. In such contexts what might aid people’s ability to identify missing answers. In this regard, a useful place to look is the only cycle in the corpus where an

616 Discourse Studies 8(5)

audience contributor comes back to complain that their contribution did not receive a response. In this instance, the following local contingencies were in play: 1) AUD’s contribution was the first contribution in the smallest ‘five question’ cycle; 2) the ‘answering’ section of the sequence begins with the chair’s request for ‘snappy answers’ and is the shortest in the corpus (1m 43s); and 3) whilst ‘answering’ LA1 clearly does the work – described above – of beginning and ending each ‘answer’ and invoking, in various ways and to different extents, previous contributions. (9) 1 2 3 4

LA1:

1 2 3

LA1:

1 2 3 4

LA1

(10)

(11)

[Q&A2: A1] yeah huh (.) t- the i:ssue of the ho:me zone obviously must be something that [Q&A2:A2] . . .point about shopkeepers so I- I- I- think that’s huh done (.) I agree the. . . [Q&A2: A4] and that’s all I’d say (.) there was an issue about the traffic lights an’th an’th an’the

Given these conditions, the analyst can – with confidence – establish the actions that constitute the full QA2 sequence.2 There are four clear responses to five AUD turns, two of which contained two points. Having done this it is noticeable that neither of the points raised by the first AUD speaker receives a response and neither does AUD2’s second complaint/accusation. More significantly this ‘finding’ was produced, perhaps using different methods perhaps similar ones, by the first AUD speaker herself. She comes back to complain. It is worth considering how this was done. (12) 1 2 3 4 5 6 7 8 9 10 11

LA1: AUD1: CHR: AUD1: CHR:

AUD1: CHR: AUD1:

[End Q&A1] we need to proceed (.) zzzzzzz(2.2) [zzzzzzzzzz= [can I ask] =zzzzzz(5.0) [zzzzzzz [can I ask for more questions [ now (.) we can= [I didn’t ( =we can. ] )]

Llewellyn: Organizing audience participation in political discourse 617 12 13 14 15 16 17 18 19 20 21

AUD1: CHR: AUD1: CHR:

(.4) I didn’t get an answer (.2) your question: was- ab [o:ut [((re states the question)) I think to a certain extent that has already been answered we might as well move on

In this setting, the absence of a response would be only apparent right at the very end of an answering phase, which would be signalled by the speaker (sitting down, glancing to the chair) or by the chair intervening in the speakers discourse. This is a fairly inhospitable slot for uninvited audience actions. It was typically occupied by audience noise or by the chair’s efforts to ‘move things along’. In this instance, LA2 closes by moving from the microphone and sitting down. This is met by mild disaffiliative buzzing. About 2 to 3 seconds after LA2’s close, the chair attempts to move the meeting on. This begins in the midst of the audience buzzing (lines 3 then 5). AUD, if she wishes to complain, can either start in the midst of the audience noise, or in the midst of the chair’s discourse. As it transpires, she interjects at the first transition relevant place in the chair’s discourse (following ‘more questions’, line 7). Both speakers then ‘drop-out’, with AUD coming back to re-do her complaint. Having made the complaint, it is noticeable that the chair ‘hedges’ (line 18–21) in her rejection of the complaint. She might be unable to remember whether it had been answered, or she might be implying that AUD’s concerns had actually been addressed, perhaps by a response to another question. Either way, she is unable to claim that AUD’s question had definitively been addressed. In this cycle both the analyst and AUD could identify a missing answer, although the chair’s turn orients to ambiguities even here. When the local conditions that enable such close monitoring are not in place, this becomes more difficult. Consider the long cycle where answering takes approximately eight minutes. First, 12 rather than five audience contributions were ‘collected’. Second, numerous contributions addressed the same topic (two on consultation, three on public transport, two on traffic flows). Consequentially, when the speaker began some response by glossing a topical domain it was unclear which contribution was being addressed. Third, on four different occasions the speaker started upon a topic without any clear interactional warrant, without identifying a previous turn. Where such conditions are in play, monitoring talk for its responsiveness to specific contributions becomes very difficult, even for an analyst working with a recording.

618 Discourse Studies 8(5) TURN COMPLETION : MANAGING AUDIENCE REACTIONS

The analysis now shifts from beginnings to closings. The matter of how speakers make the upcoming completion of their monologues ‘recognizable’ is interesting enough, but in a normative environment where the majority of audience turns received some applause turn closings have added significance. Those producing accusations, complaints, etc., might wish to ‘get the audience on their side’, to show they are expressing shared grievances. In the analysis of the talk of professional politicians and speech-makers, wellknown studies (Atkinson, 1984; Heritage and Greatbatch, 1986) have described rhetorical devices which, by warrant of the particular clarity with which they project completion, organize the seemingly spontaneous onset of gross audience reactions. In the present materials, we see lay rather than professional speakers utilizing just these devices to bring their arguing turns to a close. In the following data, AUD2 brings his turn to a close with a complaint/accusation (that the panel is talking ‘at’ the audience, but not listening to them) in the form of a ‘contrast structure’. (13) 1 2 3 4 5 6 7 8 9 10

[Q&A2. Q3:T4] AUD2: (.) the thing that I’m really disappointed about tonight (.) is that it seems to me that the panel (.) are more intent at actually talking at us (.) than actually ↑listening t [o us AUD: [xxxxXXXXXXXxxxxxxxx xxxxxxx (5.5)

Such devices work, Heritage and Greatbatch (1986: 122–3) argue, by projecting their completion ‘through a process in which audience members match the unfolding second half of a contrast against its first half ’. Consequentially, most – though not all – contrast structures that mobilize audience support are ‘rhythmically balanced and contain similarities of length, content and grammatical structure’ (Heritage and Greatbatch, 1986: 123). This is apparent in AUD2’s talk, which closely mirrors Heritage and Greatbatch’s examples (their extracts 3 and 4). There is a clear rhythmic quality, arising from the emphasis placed upon ‘at us’ (line 6) in the first part and then on ‘to us’ (line 7) in the second. Grammatically, the term ‘actually’ appears in both parts. At the level of rhetoric numerous items can be heard together ‘talking/listening’ and ‘at us/to us’. The projected completeness of the device is apparent in the audience reaction, as in Heritage and Greatbatch’s examples (1986) applause starts before the speaker is through. Not all closings were that smooth; many were full of glitches and hesitations. Consider the following example, where AUD3 is making a series of complaints

Llewellyn: Organizing audience participation in political discourse 619

and accusations about the consultation; that the council is shifting ‘the problem’ from one area to another, that the consultation is undemocratic and that the council is not ‘showing an interest in the people’. (14) 1 2 3 4 5 6 7 8 9 10 11 12

AUD3:

AUD: AUD:

[Q&A3. Q7:T6] you’re) shifting one problem to another part of ( ) muswell hill and it’s not democ- democracy its just it its its not democracy at all (.) its not even showing an interest in the ↑people (.) absolut [ely [xxxxxXXXXXXXXXXXXX

As the extract begins, AUD3 has been talking for just over 60 seconds. At line three, he attempts to construct a two-part rhetorical device (‘its not democracy it’s just’), but fails. The line about democracy is then recycled, with an added ‘at all’ (line 6), but this time AUD3 does find a second part (‘it’s not even showing an interest in the people’, lines 6–8). Whilst this is not a contrast structure, both parts are negative, they are rhythmically balanced and contain similarities of length, content and grammatical structure. In this case, two aspects of the onset of the audience applause are worth noting. First, it starts up after the completion of the speaker’s turn, not in the midst of that completion, which may reflect the production problems discussed above. Second, following a micro-pause, a lone audience member produces a strongly affiliative reaction, (‘absolutely’, line 10), which potentially helps organize the collective response. This also appears in Heritage and Greatbatch’s (1986) data, where it is common for audience members to shout out ‘hear hear’, acting as a bridge between the turn completion and the gross audience reaction. No turn by an LA speaker generated applause. Whilst collective reactions to LA speakers were entirely disaffiliative (Clayman, 1991), LA speakers still deployed devices and practices that clearly projected the upcoming completion of their turns. In some instances, LA speakers even seemed to be ‘baiting’ the audience, by anticipating (and helping organize) collective disaffiliative responses. Consider the completion of LA1’s response, where he is rejecting, albeit in ways that are at least ambiguous, AUD2’s turn. (15) 1 2 3 4

LA1:

[Q&A2. A4:T6] huh (.) whenever you close a road I believe I said this all to you before and I believe ( ) last time

620 Discourse Studies 8(5) 5 6 7 8 9 10 11 12 13 14 15 16

LA2: AUD:

you all shouted back at me but I’ll try again.hhh whenever you close a road there are a few people in favour of it (.) and a very large number against it (.) tha [t is that is the that [zzzzzzzzzzzzzzzzzZZZZZ is the truth ZZZZZZzzzz ((approx 7.0))

Whilst LA1’s c-structure (‘a few people in favour of it and a very large number against it’, lines 8–11), is rhythmically and grammatically balanced, it fails to generate an immediate disaffiliative response, which Clayman (1991) suggests is typical. He found disaffiliative collective reactions tend to be dispreferred (delayed onset, comparatively slow volume increases, etc.). It might even be argued that LA1’s re-completion (‘that is the truth’, lines 13–15) is attending to the absence of such a response, given the work LA1 has already done to make one relevant. This may show something of the flexibility of re-completions, which have typically been examined ‘searching for applause’ (Heritage and Greatbatch, 1986: 133). As a final example, and to highlight something of the sophistication of these lay practices, consider the following example of a ‘post-response-initiation recompletion’ (Atkinson, 1984: 397; Hutchby, 1997).What is this? A common way in which people bring a monologue to a ‘hearable completion’, is to recapitulate an earlier point (Hutchby, 1999). In environments where applause might be expected, such recapitulations may appear to be ‘actively pursuing applause’ (Heritage and Greatbatch, 1986: 133). But as Atkinson notes, in specific sequential positions – for example, just after some initial applause – such recapitulations can have a quite different character, they can exaggerate the sense of support for a speaker, ‘insofar as early response onset is hearable as a display of greater than usual enthusiasm for what a speaker is saying’ (Atkinson, 1984: 397). (16) 1 2 3 4 5 6 7 8 9

[Q&A3. Q8:T7] AUD4/T7 ((20 seconds into turn)) . . .and frankly it’s just not (.) fa:ir (.2) ((continues approx 20 seconds)) if you are going to have consultation then do it PROPERLY (.) do not set one part of the community against

Llewellyn: Organizing audience participation in political discourse 621 10 11 12 13 14

AUD: AUD:

anoth [er (.) it’s= [xxxxxxxxxxxxxx= =just NOT FAIR =xxxxxXXXXXXXXXXXXXXXx xxxxxx (10.0)

This would appear to be a particular smooth example. From lines 4–6, a very clear contrastive device is used, which is effective in organizing the immediate onset of applause just prior to the completion of ‘another’ (line 9). Following a micro-pause, the speaker re-completes his earlier phrase exactly, that ‘it is just not fair’ (initially lines 2–3, then lines 10–12), entirely in overlap with the applause. In the present materials there is thus some considerable overlap between practices of argumentation and those of public speaking. Accusations, complaints and so forth are brought to a close with devices that review and restate opposed positions (particularly apparent in extracts 6 and 7), whilst simultaneously co-ordinating collective audience displays.

Discussion This article has made an initial attempt to describe the consequences of ‘collecting questions’ for the way in which members of society dispute, reject and oppose something that has gone before. It contributes to previous studies of ‘real time’ disputation and generates practical insights into work involved in administering new institutions of British local governance. As a way of managing talk, ‘collecting questions’ might appear to be a relatively minor or technocratic device. But it is difficult to imagine a stronger way of minimizing ‘debate’ than precluding direct interaction, inserting conversational space between the turns of participants and ensuring any two people can only ever ‘interact’ once. As might be imagined, separating turns into periods was found to be associated with a weakening of sequential constraints ‘ordinarily’ established by actions such as accusations, disagreements and complaints. At one extreme, as Garcia (1991) found, a whole range of initial actions received no response (not an evasive response, simply no talk at all). Numerous – specifically interactional – reasons for this have been considered: 1) limited to one turn, audience members would make multiple points. LA speakers would then respond ‘selectively’, avoiding the more confrontational aspects; 2) depending on a range of local contingencies (number of questions collected, their topical domain, etc.), it became very difficult to observe ‘no response’, that is, absolutely no talk following a first pair part (Sacks et al., 1974); 3) even where audience members did manage to identify such absences, they faced a range of problems pursuing any complaint. Perhaps more significantly, where LA speakers did address a specific audience turn, they responded not to audience turns per se, but to their own treatments of those turns (to glosses,

622 Discourse Studies 8(5)

characterizations and reformulations), again securing them an unusual amount of latitude in the construction of their talk. Garcia’s (1991) argument, crudely, is that this all aids dispute resolution precisely because parties can explore each others perspective without interactional antagonisms getting in the way; selective response is viewed as facilitating ‘the resolution of conflict’ (Garcia, 1991: 830). In important ways, this would be a very generous reading of the present data. ‘Collecting questions’ certainly enabled LA speakers to avoid a series of difficult issues, but without any comparable mechanism for resolution. On numerous occasions, for example, audience members argued the council’s approach to consultation was divisive. But at no point was this addressed by any LA speaker. Whether this aids conflict resolution or is a source of underlying tension must be an open question. Like mediation, collecting questions reduces possibilities for escalation, but because there are no frameworks for resolution, where conflict arises it may ‘maintain’ rather than eliminate ‘group boundaries’ (Coser, 1964, quoted in Garcia, 1991: 818). Despite various constraints, the article has described how audience members adapted their language practices to accomplish the action-components of arguments and – potentially – series of interlocking action-opposition pairs (Hutchby, 1996). In this regard, audiences were not entirely passive recipients of such interactional provisions. At the beginning of their turns AUD speakers found ways to display the action status of their upcoming talk and ‘recall’ previous utterances. This work was successful to the extent speakers were able to produce adjacency pairs (Sacks et al., 1974) despite the separation of talk into different periods. Assertions were countered (LA2 and AUD2) and counters were rejected (AUD2 and LA1). Whilst ‘argument’ may depend upon the speech exchange system of ordinary conversation (Garcia, 1991), the present data suggest the basic components of arguments can be produced outside arguments’ ‘natural home’. The research has also pointed to an intersection between the practices of argument and those of public speaking more generally. Perhaps most impressive, was the way in which lay speakers ‘finished off ’ their complaints, accusations, etc., by deploying rhetorical formats that reviewed and contrasted ‘both sides’ of an argument whilst simultaneously cueing collective audience displays. In line with the findings of Hutchby (1999), ‘lay’ members of the public were found to be skilled orators, able to actively incorporate applause into the design of their turns, evidenced by such phenomena as ‘post response initiation re-completion’ (Atkinson, 1984). Finally, questions might be raised about whether such interactional arrangements support or undermine the stated aims of these consultative forums; to re-connect community and council (DETR, 1998). Recapping, it has been shown that: 1) ‘collecting questions’ enabled LA speakers to ignore audience turns; 2) that unable to respond directly to LA speakers, audiences would heckle and collectively ‘hiss’ LA speakers; and that 3), as the identities of

Llewellyn: Organizing audience participation in political discourse 623

contributors were never recorded, responses tended to be addressed to collectives or parties such as ‘the panel’ (extract 8) or ‘you’ (extract 15). In a range of ways, interactional provisions seemed to actively contribute to the reproduction of existing antagonisms, identities and group boundaries. To the extent this has been shown, the article is a further demonstration of the relevance of turntaking provisions for the reproduction of social structures (Boden and Zimmerman, 1991). NOTES

1.

2.

Notes on the presentation of the sequence. Reading left to right, Q&A1 denotes question/answer period 1, Q&A2 period 2 and so on. T1, T2, T3, etc., denotes turns 1, 2 and 3 in the sequence. The numbers which follow are approximate relative timings. Q3/AUD1 denotes that the turn was the third question produced in that questioning period. A4/LA1 denotes that the ‘answer’ was hearable as the fourth in the answering monologue of LA1. AUD1 denotes an audience speaker; LA denotes a local authority speaker. The full cycle can be mapped as follows: No 1

Selected Start 0 7

Finish 1.02

Length 55

1.12 1.57

1.21 2.05

1.45 2.40

24 35

4 5

Question Phase (a) Impact of closure. (b) Traffic calming. Shopkeepers. (a) Consultation. (b) About meeting. Survey findings. Home zones.

2.56 4.19

2.57 4.23

4.10 5.30

1.13 1.07

No 1 2 3 4

Answer Phase Home Zone Shopkeepers ‘Whole community’ Survey

Times (approx) 6.24–50 6.50–55 6.55–7.30 7.44–8.07

2 3

From Question Approx. 1m Approx. 5m. Approx. 4m. Approx. 3m.30s.

REFERENCES

Atkinson, J.M. (1984) ‘Public Speaking and Audience Responses: Some Techniques for Inviting Applause’, in J.M. Atkinson and J. Heritage (eds) Structures of Social Action, pp. 370–410. Cambridge: Cambridge University Press. Atkinson, J.M. and Drew, P. (1979) Order in Court: The Organisation of Verbal Interaction in Judicial Settings. Basingstoke: Macmillan. Boden, D. and Zimmerman, D.H. (1991) Talk and Social Structure: Studies in Ethnomethodology and Conversation Analysis. Cambridge: Polity. Clayman, S.E. (1991) ‘Booing: The Anatomy of a Disaffiliative Response’, American Sociological Review 58(February): 110–30. Coulter, J. (1990) ‘Elementary Properties of Argument Sequences’, in G. Psathas (ed.) Interactional Competence, pp. 181–204. Washington, DC: International Institute for Ethnomethodology and Conversation Analysis and University Press of America. Dersley, I. and Wootton, A.J. (2001) ‘In the Heat of the Sequence: Interactional Features Preceding Walkouts from Augmentative Talk’, Language in Society 30: 611–38.

624 Discourse Studies 8(5) DETR (1998) Modern Local Government: In Touch with the People. CM 4014. London: DETR TSO. Drew, P. (1990) ‘Strategies in the Contest between Lawyer and Witness in Cross Examination’, in J. Levi and A.G. Walker (eds) Language in the Judicial Process, pp. 39–64. Plenum: New York. Eisenberg, A. and Garvey, C. (1981) ‘Children’s Use of Verbal Strategies in Resolving Conflicts’, Discourse Processes 4: 149–70. Garcia, A.C. (1991) ‘Dispute Resolution without Disputing: How the Interactional Organization of Mediation Hearings Minimizes Argument’, American Sociological Review 56(December): 818–35. Goodwin, M.H. and Goodwin, C. (1987) ‘Children Arguing’, in S. Philips, S. Steel and C. Tanz (eds) Language, Gender and Sex in Comparative Perspectives, pp. 200–48. Cambridge: Cambridge University Press. Gruber, H. (1998) ‘Disagreeing: Sequential Placement and Internal Structure of Disagreements in Conflict Episodes’, Text 18(4): 467–503. Heritage, J. and Greatbatch, D. (1986) ‘Generating Applause: A Study of Rhetoric and Response at Party Political Conferences’, American Journal of Sociology 92(1): 110–57. Hutchby, I. (1992a) ‘The Pursuit of Controversy: Routine Scepticism in Talk on Talk Radio’, Sociology 26(4): 673–94. Hutchby, I. (1992b) ‘Confrontation talk: Aspects of ‘’Interruption’’ in Argument Sequences on Talk Radio’, Text 12(3): 343–71. Hutchby, I. (1996) Confrontation Talk: Arguments, Asymmetries and Power on Talk Radio. Mahwah, NJ: Lawrence Erlbaum. Hutchby, I. (1997) ‘Building Alignments in Public Debate: A Case Study from British TV’, Text 17(2): 161–79. Hutchby, I. (1999) ‘Rhetorical Strategies in Audience Participation Debates on Radio and TV’, Research on Language and Social Interaction 32(3): 243–67. Hutchby, I. (2001) ‘’’Oh’’, Irony and Sequential Ambiguity in Arguments’, Discourse and Society 12(2): 123–41. Maynard, D. (1985) ‘How Children Start Arguments’, Language in Society 14: 1–30. Sacks, H., Schegloff, E. and Jefferson, G. (1974) ‘A Simplest Systematics for the Organization of Turn-Taking for Conversation’, Language 50(4): 696–735. Sacks, H. (1992) Lectures on Conversation. Oxford: Basil Blackwell. APPENDIX TRANSCRIPTION SYMBOLS

(.7) (.) = [] .hh hh (( )) : ! ()

Length of a pause. Micro-pause. A latching between utterances. Between adjacent lines of concurrent speech indicates overlap. Inbreath. Outbreath. Non-verbal activity. Sharp cut-off. Stretching of a word. Denotes an animated tone. Unclear fragment.

Llewellyn: Organizing audience participation in political discourse 625

: CAPITALS >< <> :: Word zz zzZZ xx xxXX

Quiet utterance. Noticeably louder. The talk in-between is quicker. The talk in-between in slower. Rising or fall intonation. Underline indicates speaker emphasis. Audience buzzing. Buzzing becoming louder. Audience applause. Applause becoming louder.

is a lecturer in the Industrial Relations and Organizational Behaviour group at Warwick Business School, University of Warwick. His research examines talk and interaction in institutional settings. He has publications in journals such as Sociology and Organization Studies. Current projects include how selection panels make recruitment decisions. A D D R E S S : Industrial Relations and Organizational Behaviour Group, Warwick Business School, University of Warwick, Coventry CV4 7AL, UK. [email: [email protected]]

N I C K L L E W E L LY N