The Mistiming of Applause in Political Speeches

JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 Bull, Noordhuizen / THE MISTIMING OF APPLAUSE THE MISTIMING OF APPLAUSE IN POLITICAL SPEECHES PETER BULL University of York MEREL NOORDHUIZEN University of Amsterdam According to Atkinson, speakers at political meetings invite applause through rhetorical devices, which indicate when and where applause is appropriate. Hence, speech and applause are characterized by a high degree of synchronization. Thus, incidences of unsynchronized applause are of considerable theoretical interest. An analysis of such mismatches is reported based on six speeches delivered by the three leaders of the principal British political parties to their respective party conferences in 1996 and 1997. Only a mean 61% of applause incidences were fully synchronized with speech. Four principal types of mismatch were identified: isolated applause, delayed applause, interruptive applause, and applause interrupted by the speaker. Mismatches occurred when rhetorical devices were ineffective. They also occurred in the absence of rhetorical devices when applause was not invited. It is concluded that Atkinson’s theory of rhetoric needs to be modified to take account of a distinction between invited and uninvited applause. According to a theory proposed by Atkinson (e.g., 1983, 1984a, 1984b), a limited range of rhetorical devices such as contrasts and three-part lists are consistently effective in evoking applause from audiences. Atkinson further observes that applause is typically closely synchronized with speech, which he cites as strong evidence that rhetorical devices function as signals indicating when and where is an appropriate place to applaud (Atkinson, 1984a, pp. 33-4). Thus, in testing the validity of this theory, incidences where applause is not fully synchronized with speech are of considerable theoretical interest. If such incidences occur with considerable frequency, this would suggest at the very least that rhetorical devices do not function as effectively as Atkinson has argued or, more fundamentally, that they are not as important as Atkinson proposes. Such failures in synchronization between speech and audience applause are termed mismatches and form the focus of the study reported in this article. AUTHORS’ NOTE: The authors would like to thank Pam Wells for conducting the second reliability study. They would also like to thank the anonymous reviewers for their instructive comments on an earlier version of this paper. Correspondence should be addressed to Peter Bull at the Department of Psychology, University of York, YO10 5DD, United Kingdom. JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY, Vol. 19 No. 3, September 2000 275-294  2000 Sage Publications, Inc. 275 276 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 To date, the research evidence on rhetorical devices in political speeches has undoubtedly demonstrated what is in effect a strong positive correlation with collective applause. Heritage and Greatbatch (1986) analysed the full televised output from the 1981 British party political conferences (Conservative, Labour, and Liberal), giving a total of 476 speeches. They investigated seven basic rhetorical formats (contrasts, lists, puzzle-solution, headline-punchline, combinations, position taking, and pursuits), which they found to be associated with more than two thirds of all the collective applause that occurred during speeches. Of these seven basic formats, the contrast and the list were by far the most effective: Contrasts were associated with no less than 33.2% of the incidences of collective applause during speeches, lists with 12.6%; hence, almost half the collective applause was associated with Atkinson’s two principal rhetorical devices. Atkinson’s research was based on the detailed qualitative analysis of specific examples of rhetorical devices; consequently, it was always open to the criticism that his choice of examples (and speeches) might have been highly selective. Given the comprehensive sampling employed in the Heritage and Greatbatch study, this possible criticism would appear to be unfounded. To further refine their evaluation of Atkinson’s research, Heritage and Greatbatch tested and confirmed the hypothesis that rhetorical devices were more likely to be associated with collective applause than with isolated clapping (on the part of only one or two audience members). Similarly, Bull (1986) found in a content analysis of a speech by Arthur Scargill, President of the National Union of Mineworkers, that collective applause was associated more frequently with speech using rhetorical devices, whereas isolated applause occurred more frequently in their absence. In seeking to explain the role of such devices, Atkinson argues that audiences are more likely to respond to statements that are constructed in such a way as to both emphasise and highlight the content, and which project a clear completion point for the message in question. Emphasis naturally calls attention to passages to which the speaker attaches particular significance, but Atkinson argues that emphasis alone is rarely sufficient to ensure a response. Projectability is also important because audience members must decide not only if they will applaud but when to applaud; if the speech is constructed in such a way as to indicate appropriate applause points, this assists the audience in coordinating their behaviour. According to Atkinson (1984a, p. 18), the use of rhetorical devices is in the interest of the audience because it helps them applaud together rather than risk exposure to public ridicule and humiliation by applauding in isolation. Atkinson further observes that applause is typically closely synchronized with speech. In the vast majority of cases, it begins to get under way either just before or immediately after a completion point, Bull, Noordhuizen / THE MISTIMING OF APPLAUSE 277 such as the end of a sentence (Atkinson, 1984a, p. 33). Similarly, Heritage and Greatbatch (1986, p. 112) found that most applause was typically initiated within 0.3 seconds of the precipitating item. Applause also reaches its maximum volume very quickly, typically after the 1st second, again suggesting that the audience is able to project the appropriate place to applaud (Atkinson 1984a, p. 24). Atkinson (1984a, pp. 33-34) writes as follows: If displays of approval are seldom delayed for more than a split second after a completion point, and frequently start just before one is reached, it means that speakers must be supplying their audiences with advance notice as to precisely when they should start clapping. Otherwise, it is quite impossible to see how anyone would ever be able to respond so promptly. One major potential criticism of all this research is that it fails to take account of the role of speech content. Atkinson, Heritage, and Greatbatch are well aware of this. Atkinson (1984a) points out that there are a restricted number of messages that can be regarded as applaudable: favourable references to persons, favourable references to us, unfavourable references to them. He found that in one sample, these three categories accounted for 95% of the bursts of applause at British political party conferences (Atkinson, 1984a, p. 44). Heritage and Greatbatch have also performed a number of content analyses. They, too, found that applause was reserved for a relatively narrow range of message types. Specifically, these were external attacks (statements critical of outgroups such as other political parties), general statements of support or approval for the speaker’s own party, internal attacks (criticisms of individuals or factions within the speaker’s own party), advocacy of particular policy positions, commendations of particular individuals or groups, and various combinations of these message types. In total, these categories made up more than 81% of all the applauded messages in the speeches under investigation. With regard to specific message types, they found that audiences were nearly three times as likely to applaud criticisms of policies and their proponents than to applaud constructive policy recommendations. Heritage and Greatbatch further found that the likelihood of these message types receiving applause was greatly increased by rhetorical formatting. This influence was particularly strong in the case of positive assertions, which, as has been noted above, were less likely to be received with applause. In the case of negative attacks, rhetorical devices were generally about twice as effective in evoking applause, whereas in the case of positive assertions, rhetorical devices were between three and five times more effective. Heritage and Greatbatch conclude that although applause is clearly related (not surprisingly) to certain types of speech content, nevertheless the chance of that speech 278 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 being applauded is greatly increased if it is expressed using an appropriate rhetorical device. Although Atkinson, Heritage, and Greatbatch integrate the role of content into their analysis, one issue that they fail to address is how and why collective applause occurs in response to nonrhetorically formatted statements. It is possible that in these cases it is speech content alone that is sufficient to evoke applause. Furthermore, if, as Atkinson argues, rhetorical devices are important in assisting the audience to coordinate their applause, the question arises as to how such coordination occurs when nonrhetorically formatted statements are applauded. Heritage and Greatbatch note that slightly more than two thirds of the collective applause in their sample of speeches was associated with their seven rhetorical devices. They also made extensive comparisons of applauded and nonapplauded statements (based on 41.75 hours of broadcast speeches containing an estimated 20,000 sentences) to provide evidence that rhetorical formatting increases the likelihood of a statement receiving applause. But neither Atkinson nor Heritage and Greatbatch have ever conducted a detailed analysis of how and why collective applause occurs in response to nonrhetorically formatted statements. To address this issue, a further analysis was conducted of political speeches, focussed on statements that received collective applause without rhetorical formatting (Bull, 2000). Fifteen such statements were culled from the speeches delivered by the three leaders of the principal British political parties to their respective party conferences in the autumn of 1996. Applause to these statements was typically not synchronized with speech: In the majority of cases, it started either well before a completion point or was actually interruptive. A content analysis was also conducted that showed in every case the applauded statement constituted a statement of political policy—that is to say, the audience could be seen to be responding not to rhetorical devices but to the political content of the speech. Bull (2000) argued that some aspects of speech content may be so emotive or so politically significant that they will be applauded, irrespective of whether they are rhetorically formatted. How does the audience coordinate its applause in response to such statements? Clayman (1993) has proposed that there are two principal ways in which an audience can coordinate its behaviour. There is what he calls independent decision-making, whereby individual audience members may act independently of one another yet still manage to coordinate their actions—through, for example, applause in response to the rhetorical devices identified by Atkinson, Heritage, and Greatbatch. There is also what Clayman calls mutual monitoring, whereby individual response decisions may be guided, at least in part, by reference to the behaviour of other members. For example, once it becomes evident that Bull, Noordhuizen / THE MISTIMING OF APPLAUSE 279 some members of the audience are starting to applaud, this drastically alters the expected payoff for other audience members: The fear of responding in isolation will be reduced; conversely, not applauding can also increasingly become an isolating experience. Clayman further observes that responses organized primarily by independent decision making should begin with a burst that quickly builds to maximum intensity as many audience members begin to respond together, whereas mutual monitoring in contrast should result in a staggered onset as the initial reactions of a few audience members prompts others to respond. This staggered onset was typical of much of the applause in response to the nonrhetorically formatted statements observed by Bull (2000), especially in examples of applause that did not occur at a completion point. Clayman’s analysis was actually based on booing, and he concluded (1993, p. 124) that “clappers usually act promptly and independently, wheras booers tend to wait until other audience behaviours are underway.” The different characteristics of applause observed by Bull (in press) suggest that both independent decision making and mutual monitoring may be involved in clapping: mutual monitoring in applauding nonrhetorically formatted statements and independent decision making in applauding rhetorically formatted statements. On this basis, it was proposed that two distinct types of applause may be identified in political speeches, referred to respectively as invited and uninvited applause. Applause occurring in response to the rhetorical devices identified by Atkinson, Heritage, and Greatbatch might be referred to as invited applause. It is invited in the sense that the speaker uses the rhetorical structure of the speech to indicate when and where applause is appropriate; according to Atkinson, it is typically closely synchronized with speech. Uninvited applause is typically not associated with rhetorical devices, nor is it necessarily synchronized with speech; it may be initiated by a small section of the audience, and there may be a substantial delay before it is taken up by the rest of the audience (staggered onset). Uninvited applause seems to occur in response to specific aspects of speech content, which certain members of the audience wish to endorse and which may be taken up by a wider section of the audience or by the audience as a whole. Thus, whereas the synchronization of speech and applause plays a central role in Atkinson’s analysis, it is not a problem for the concept of uninvited applause because it can occur through the process of mutual monitoring described by Clayman. In testing the validity of Atkinson’s theory of rhetoric, incidences where applause is not fully synchronized with speech are thus of considerable theoretical interest. Yet, to date, no systematic investigation of such mismatches in political speeches has been carried out. The purpose of this study was to obtain some basic information on mismatches and to relate that information to 280 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 Atkinson’s theory of rhetoric and the proposed distinction between invited and uninvited applause. At first sight, it would appear that there are at least four ways in which lack of synchronization may occur between speech and audience applause. One form of mismatch is isolated applause (claps from one or two people), which Heritage and Greatbatch (1986) distinguish from collective applause. Given that isolated applause does not involve a coordinated response from the audience as a whole or a substantial section of it, isolated applause may be regarded as a mismatch. Another form of mismatch may occur if there is an extended silence between the speaker’s utterance and audience applause. Silence suggests that the speaker was expecting applause, but for some reason the audience failed to respond appropriately. Hence, there is a failure of synchronization between speaker and audience, just as an extended silence in conversation may also be considered awkward. Just as incidences of simultaneous speech in conversation can be regarded as a failure in turn-taking, incidences of applause simultaneous with the politician’s speech may also be regarded as mismatches. Such incidences may be initiated either by the speaker or the audience. The audience may interrupt the speaker by applauding; this can be regarded as a mismatch, whether or not the speaker completes what he was saying. Conversely, the speaker may interrupt the audience’s applause by starting to speak again before the applause has subsided. However, brief overlaps where the audience starts to applaud just before the speaker completes an utterance or the speaker resumes the speech as the audience applause dies down would not be regarded as mismatches because they suggest that either the speaker or the audience are anticipating a completion point—just as a brief overlap in conversation between one speaker and another would not be regarded as interruptive. It should, however, be noted that interruptions of applause by the speaker differ in certain important respects from the other three types of mismatch. In particular, whereas the other three categories are focussed on the audience’s behaviour, this is the only category that deals with the speaker’s behaviour. Furthermore, as Atkinson (1985) pointed out in his analysis of refusing invited applause, a charismatic orator by speaking into the applause may create an impression of overwhelming popularity, struggling to be heard while at the same time inhibiting and frustrating the audience’s wish to applaud. When the speaker does finally allow the audience an opportunity to respond, they may be literally bursting to applaud, thereby the speaker is seen as receiving a rapturous reception. As such, interruptions of applause by the speaker may be strategic, inciting the audience to ever greater enthusiasm. However, there is no reason to believe that every incidence of the speaker interrupting applause is necessarily strategic in Bull, Noordhuizen / THE MISTIMING OF APPLAUSE 281 the way that Atkinson describes. Audience applause can simply go on for so long that the speaker has to interrupt to continue with the speech. Thus “speaker interrupts audience applause” is regarded as a mismatch, but one of a special kind. The above analysis would suggest that there are at least four different ways in which mismatches between speech and applause can occur. In the light of the above discussion, the first three categories (isolated applause, delayed applause, audience interrupts speaker) are referred to as audience mismatches, the fourth category (speaker interrupts audience) as speaker mismatches. To test the validity of these distinctions, it was decided to construct a typology of mismatches based on the detailed analysis of a number of political speeches. On the basis of this typology, an investigation was then conducted of both the frequency and the manner in which mismatches occur, thereby to consider their significance for Atkinson’s theory of rhetoric. METHOD PARTICIPANTS The study was based on six speeches delivered by the leaders of the three principal British political parties to their annual conferences in 1996 and 1997 (before and after the General Election of 1997). The speeches were as follows: 1996 September 24th October 1st October 11th 1997 September 24th September 30th October 10th Paddy Ashdown (Leader of the Liberal Democrats since 1988), 57 minutes 30 seconds Tony Blair (Leader of the Labour Party since July 1994; Prime Minister since May 1997), 65 minutes 37 seconds John Major (Conservative Prime Minister, 1990-1997), 51 minutes 8 seconds Paddy Ashdown, 56 minutes 10 seconds Tony Blair, 59 minutes 4 seconds William Hague (Leader of the Conservative Party since June 1997), 58 minutes 41 seconds APPARATUS The speeches were recorded off-air and analyzed using a VHS format videocassette recorder with slow-motion replay facilities. Transcripts of the speeches were obtained and converted into a word-processing package. The 1997 speeches were downloaded from each party’s Web site; the 1996 speeches were obtained as hard copies 282 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 from the respective party headquarters and put through an optical scanner. The text of each speech was then checked for accuracy against the videorecording and amended accordingly. PROCEDURE Incidences of applause were marked on each transcript. Following Atkinson’s (1984a) notation, the use of small and large crosses was used to represent the changing intensity of applause (e.g., xxxxXXXX), a dash on either side of a cross represented an isolated clap (-x-), and several in a row represented hesitant or spasmodic clapping (-x-x-x-). Two parts of a contrast were identified with the letters A and B. The three elements of a three-part list were numbered from 1 to 3. Nonverbal and vocal features of delivery were also noted if they were considered to affect the judgment of a mismatch. A typology of mismatches was devised from detailed analysis of the 1997 speeches by Paddy Ashdown and Tony Blair. Four categories of mismatch were distinguished, and all six speeches were coded in terms of this typology. In settling on the final categories of mismatch for all six speeches, any disagreements were resolved by discussion. To test the reliability of this coding system, two studies were conducted. The first was intended to assess the identification of mismatches, the second the coding of the four mismatch categories. In the first study, the two authors independently coded all incidences of applause in the four speeches by Hague (1997), Blair (1996), Major (1996), and Ashdown (1996) as either synchronized or unsynchronized. In the second study, an additional rater independently coded all the mismatches from those same four speeches in terms of the four categories in the mismatch typology. RESULTS TYPOLOGY OF MISMATCHES The results of the two reliability studies were as follows: identification of mismatches .87 (Phi coefficient) and coding of the four categories in the mismatch typology .77 (Cohen’s k; Cohen, 1960). It was found that all instances of mismatches observed in all six speeches could be subsumed within the four mismatch categories. The full mismatch typology is presented below. It is based on a distinction between three different types of audience mismatch (isolated applause, delayed applause, audience applause interrupts speaker) Bull, Noordhuizen / THE MISTIMING OF APPLAUSE 283 and a fourth category of speaker mismatches (speaker interrupts audience applause). It should be noted that these four categories are not mutually exclusive. For example, isolated applause may not only be interruptive; it may also be delayed. AUDIENCE MISMATCHES Isolated applause. Refers to claps by one or two people, as distinct from collective applause by all of the audience or a substantial section of it. Isolated applause is always coded as a mismatch unless it constitutes part of a wider collective audience response (e.g., isolated claps that occur amidst collective laughter would not be regarded as a mismatch). Delayed applause. There is a discernible silence between the end of the speaker’s utterance and the onset of applause (typically of at least 1 second). Audience applause interrupts speaker. Applause is regarded as interruptive if the onset occurs at a point where it is not possible to project the completion point of the speaker’s utterance. If the audience start to applaud at a possible completion point but it is clear that the speaker intended to continue, this would also be regarded as interruptive. If the audience starts to applaud just before the speaker completes an utterance (where the completion point is projectable), this is regarded not as a mismatch but as an overlap. Brief overlaps suggest that the audience is anticipating a completion point, just as a brief overlap in conversation between one speaker and another is not typically regarded as interruptive (Roger, Bull, & Smith, 1988). SPEAKER MISMATCHES Speaker interrupts audience applause. If the speaker resumes the speech before the end of the applause, this is regarded as a mismatch unless the audience applause is already dying down, when it would be regarded as an overlap. In the latter case, it would appear that the speaker is anticipating the end of the applause not interrupting it. It should be noted that deciding whether it is the audience who interrupts the speaker or whether the speaker interrupts the audience depends who has the turn at that point. If the onset of applause occurs before the speaker has finished the utterance, then it is regarded as an interruption by the audience. But if the speaker has already relinquished the turn to audience applause and then tries to resume the 284 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 speech before the applause has died down, it is regarded as an interruption by the speaker. PROPORTION OF MISMATCHES Table 1 shows the frequency with which different types of applause occurred in the six speeches analysed in this study. These results show that mismatches between speech and applause are a common occurrence. Across all six speeches, audience mismatches accounted for a mean 29.2% of applause events, speaker mismatches for 12.9%. Only a mean 61% of applause incidences were fully synchronized with speech. By far, the most frequently occurring type of mismatch is applause where the audience interrupts the speaker (M = 17.8%), followed by incidences where the speaker interrupts the audience applause (M = 12.9%). Isolated applause (M = 4.7%) was the least frequently occurring type of mismatch. It should also be noted that there is considerable variability between the speakers in the extent to which they show different types of mismatch, especially with regard to incidences where the speaker interrupts the applause (SD = 18.8%) and incidences where the audience interrupts the speaker (SD = 10.9%). Furthermore, there is considerable variability between the speakers in the extent to which mismatches occur, from John Major where 80.3% of his applause events were synchronized with speech to Paddy Ashdown in whose speeches only 33.3% and 50.9% of applause events were synchronized with speech (SD for all speeches = 17.3%). In evaluating the significance of mismatches for Atkinson’s theory of rhetoric, it was not sufficient simply to present frequency counts for each type of mismatch. Consideration needs also to be given to the ways in which each of the four principal types of mismatch are related to rhetorical devices, and for this reason, a qualitative analysis was conducted, the results of which are reported below. HOW MISMATCHES OCCUR Each of the four major types of mismatch can occur for a variety of reasons. Audience Mismatches: Isolated Applause Misreading of cues. The speaker employs a rhetorical device, but some members of the audience fail to project the completion point accurately. The following example from William Hague can be seen in Heritage and Greatbatch’s terms as a puzzle-solution. Bull, Noordhuizen / THE MISTIMING OF APPLAUSE 285 Table 1 Frequency of Different Types of Applause Audience Mismatches Isolated Applause 1996 Paddy Ashdown Speaker Mismatches Audience Delayed Interrupts Applause Speaker Total Speaker Interrupts Audience Total Synchronous Incidents Applause of Applause 0 (0) 2 (2.4) 2 (2.6) 6 (16.7) 4 (4.7) 7 (9.2) 8 (22.2) 27 (31.8) 6 (7.9) 13 (36) 32 (37.6) 15 (19.7) 16 (44.4) 4 (4.7) 0 (0) 12 (33.3) 50 (58.8) 61 (80.3) 3 (5.7) Tony Blair 4 (5.1) William Hague 8 (12.1) M (%) 4.7 SD (%) 4.2 3 (5.7) 1 (1.3) 5 (7.6) 7.5 5.2 7 (13.2) 21 (26.9) 3 (4.5) 17.8 10.9 13 (24.5) 26 (33.3) 16 (24.2) 29.2 7.4 15 (28.3) 1 (1.3) 0 (0) 12.9 18.8 27 (50.9) 52 (66.7) 50 (75.8) 61 17.3 Tony Blair John Major 1997 Paddy Ashdown 36 85 76 53 78 66 Note. Percentages are in parentheses. Totals cannot be derived directly from individual scores because one incident of applause may involve more than one category of mismatch. And what about the minimum wage? Wasn’t it John Prescott who said that of course a minimum wage destroys jobs; any silly fool knows that he said [PUZZLE] Well, now Margaret Beckett is planning to introduce one [LAUGHTER] [-x-] [2 seconds] Apparently not every silly fool knows that! [SOLUTION] [xxXXXXXXXXXxx] [LAUGHTER] In this example, the statement of one leading Labour politician (John Prescott) is used to pose a puzzle but also to imply that another leading Labour politician (Margaret Beckett) is a silly fool (solution). The isolated applause after the first part of the solution seems to be a misreading of cues because the statement (“Well, now Margaret Beckett is planning to introduce one”) could well have been taken as the solution 286 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 to the puzzle: There is already the clear implication that according to John Prescott’s statement, Margaret Beckett is a silly fool. Hague’s 2-second pause at this point might also be taken to indicate that this is the solution. However, the pause might also have been for dramatic effect, and it is only when Hague goes on to complete the solution that it wins extensive collective applause. Failure of rhetoric. In some instances, the speaker employs a rhetorical device but fails to win collective applause from the audience. The example below from William Hague might also be seen in Heritage and Greatbatch’s terms as a puzzle-solution device. This government is like a hologram [PUZZLE] It looks so real. But when you reach out to touch it, there’s nothing there [SOLUTION] [1.5 seconds -x-] However, Hague’s delivery of this solution is somewhat confusing. The pause suggests that he might have been expecting applause at this point, but he also keeps his right hand raised as if he wishes to continue speaking. Speaker overshoots completion point. The speaker employs a rhetorical device but fails to leave the audience sufficient opportunity to applaud. In the next example, William Hague is clearly using a contrast, but by continuing to speak (“The prospect . . . ”), the applause is not allowed to develop. A The great danger is not that Britain will be left behind in Europe B but that Europe will be left behind in the world [x-x-x-x] The prospect the prospect of a single currency may present this country with one of the most momentous decisions in its history. Isolated applause in the absence of rhetorical devices. In some instances, isolated applause seems to occur in response to the content of the speech alone. So in the following extract from John Major, the isolated applause seems to be a direct reaction to the reference to “grammar schools in every town” before any rhetorical formulation by Major. And if parents want grammar schools in every town well then so do I and they shall [xx] have them Bull, Noordhuizen / THE MISTIMING OF APPLAUSE 287 Audience Mismatches: Delayed Applause Poorly constructed rhetoric. The following extract comes from Tony Blair’s (1997) speech in which he uses a three-part list, but one which is poorly constructed. And parents will have to play their part (1) There will be home school contracts for all pupils I say sign them (2) There will be new measures to tackle truancy and disruptive children new homework requirements Support them (3) And when a school disciplines a child why not back the teacher [1 second] [xxxxxxxxXXXXXXXXXXXXX] In this example, it is clear that Blair is inviting applause not only because he uses the rhetorical device of a three-part list but also because when the applause fails to come, he stands there silently nodding his head (after “why not back the teacher”) until the audience start to applaud. Whereas the first two items of the list use the same rhetorical structure in which Blair describes what is going to happen (“There will be . . . ”) and then makes a recommendation as to the appropriate course of action, the third item departs from that format (“And when a school disciplines a child . . . ”). Because of this lack of symmetry, the completion point may not be fully projectable. Absence of rhetorical devices. In the absence of a rhetorical device, the audience may not know that the speaker was expecting applause; hence, there is a pause while he waits for them to applaud the statement. In this example from William Hague’s speech, the language is highly emotive, but there are no obvious signals in the rhetorical structure that he is inviting the audience to applaud. His delivery at this point is confusing: The long pause at the end of the sentence suggests that he might have been expecting applause, but simultaneous with the onset of the applause, he opens his mouth as if to start speaking again, only to close it as the applause gathers pace. Education needs the Conservative Party now. For if Labour gives way to dogmatic attacks on choice and betrays our nation’s children then they will face the implacable opposition of the Conservative Party and deserve the condemnation of the country. [1.5 seconds] [xxxxxxxXXXXXX] 288 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 It should be noted that although delayed applause is typically treated as an indication of lack of synchrony between speaker and audience, there was one notable exception in this corpus of data. In the following example from Paddy Ashdown’s 1997 speech, delayed applause seems entirely appropriate. The extraordinary response to the death of Diana was in part at least a cry for a (1) more caring (2) more decent (3) and more compassionate society [second][xxxxxxxxXXXXXXX] The speech was delivered on 24th September, less than a month after the death of Princess Diana in a car accident (31st August, 1997), and in not responding immediately to Ashdown’s three-part list, the audience applause seems not to be unsynchronized with speech but to be showing respect for the dead princess. Audience Mismatches: Audience Interrupts Speaker Misreading of cues. The speaker uses a rhetorical device (or devices), but the audience misjudges when the completion point occurs. In the next example, Blair (1997) uses a complex rhetorical device that involves both a headline-punchline and two three part-lists (the second list forming the third item of the first list). But the audience appears to project the completion point at the third item of the first list, starting to applaud after “it can be done”; this is understandable, given that the first two items seem to imply that “it can’t be done.” And you know when people say to me sorry that’s too ambitious sorry it can’t be done I say [HEADLINE]: (1) This is not a sorry country (2) We are not a sorry people (3) It can be done if we have the (1) will and (2) courage and (3)determination to [xxxxxxXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX] do it [XXXXXXXXXXXXxxxxxxxx] Poorly constructed rhetoric. Rhetoric may be poorly constructed, such that the audience does not applaud at the appropriate completion point. The following is an example from Tony Blair’s 1997 speech of what Atkinson (1984a, p. 49) refers to as projecting a name, where the speaker identifies someone in the audience for commendation. Bull, Noordhuizen / THE MISTIMING OF APPLAUSE 289 However, as Atkinson (1984a , pp. 54-57) points out, there are a number of ways in which such namings can go wrong such that the applause does not occur at the right moment. In the example below, the naming occurs too soon. Blair names Neil Kinnock before expressing his tribute to him so that the audience applaud twice—once after hearing Neil Kinnock’s name and a second time after hearing the tribute to him. Although Blair pauses after stating Neil Kinnock’s name, the rising intonation and the tribute which follows strongly suggest that he intended to continue at this point and that the applause was interruptive. And let me pay thanks to those that led our party before me. To Neil Kinnock [xxxXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXxxxx] the mantle of Prime Minister was never his but I know that without him it would never have been mine [xxxxxXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXxxxxxxxxxxx] Failure of rhetoric. Applause does not occur immediately in response to a rhetorical device, so the speaker continues as the audience starts to applaud; hence, the applause becomes interruptive. In the following example, Ashdown (1996) uses what Heritage and Greatbatch term a position taking. Because the audience does not applaud immediately, he continues with his speech just as the applause gets under way. Now my fear my urgent fear is this. That we’ll see we shall see an election, and maybe a change of government— but we shall not see a change of direction. We Britain shall still be starved of clear vision, a commitment to change, the courage to face up to what must be done. It is the very first crucial role of this Party our primary role to ensure that that does not happen [POSITION TAKING] With the [xxxxxxxxxXXXXXXXXXxxxxxxx] With the Liberal Democrats strong in the next Parliament Britain will face the challenges that confront us Speaker overshoots completion point. The speaker employs a rhetorical device but fails to give the audience an opportunity to applaud. In the following example, Blair (1997) uses a three-part list to which the audience responds with collective applause. Blair continues the speech with the phrase “Now people . . . ” almost exactly at the same time as the applause begins but then breaks off; hence, he is interrupted by the applause. 290 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 (1) Teacher training will be reformed (2) Headteachers will have a proper qualification (3) And poor teachers will go Now people . . . [xxxxxxxxxxxxxxXXXXXXXXXXxxxxxx] And I’ll say why. People say my job is pressurised. So is teaching Absence of rhetorical devices. Interruptive applause may occur in the absence of rhetorical devices, in which case the audience seems to be responding directly to the content of the speech. In the following example from Blair (1996), the audience interrupts with applause to endorse Blair’s commitment to the principle of trade union representation. We will be part of the European Social Chapter as every other government Tory or Labour is in the rest of Europe. And there will be a right for any individual to join a trade union and if . . . [xxxxXXX XXXX] SPEAKER MISMATCHES: SPEAKER INTERRUPTS AUDIENCE The speaker may interrupt audience applause either successfully or unsuccessfully. Successful interruption. In the following example from Ashdown (1997), the applause seems to be unenthusiastic, so Ashdown continues with the speech. Now I know I know not in many of your council chambers but in Westminster at least socialism and all it stood for has been consigned to a quiet burial in an unmarked grave [xxx-xxxx-xx-x] But it isn’t just old party boundaries that are shifting. Ideas are finding new homes too. On other occasions, the applause seems to be so enthusiastic that Ashdown has to curtail it to continue with the speech. In response to the following extract from Ashdown’s (1997) speech, the applause continues for a full 24 seconds before Ashdown eventually intervenes to continue. Bull, Noordhuizen / THE MISTIMING OF APPLAUSE 291 So where shall we be, we Liberal Democrats, as this historic game is played out in the months and years ahead? Some say that we should be satisfied with our local strength and concede that Westminster will always be a side-show for us. Some say that we should be content to be a good conventional opposition, and that is enough. Well no doubt our opponents would like to see us satisfied with such limited ambitions. But I am not. And I hope that you will not be, either. I have bigger ambitions for the Liberal Democrats. I accept no glass ceilings for this Party. [xxxxxxXXXXXXXXXXXXXX] Unsuccessful interruption. In the following example from Ashdown (1997), he attempts to interrupt the applause unsuccessfully (“So . . . ”) then waits for the applause to die down before continuing with the speech. A Where we should cooperate we will do so wholeheartedly B Where we must oppose we will do so unflinchingly [xxxxXXXXXXXXXXXXXXXXXXXXXXXXXXXxxxxxxxxxxx] So So here’s my prayer for the Parliament ahead here’s my prayer for the next 4 or 5 years DISCUSSION Mismatches in audience applause are relatively commonplace during political speeches according to the results of this study. Across all six speeches, audience mismatches accounted for a mean 29.2% of applause events. This contrasts with Atkinson’s statement that “displays of approval are seldom delayed for more than a split second after a completion point, and frequently start just before one is reached” (Atkinson, 1984a, p. 33). Speaker mismatches also accounted for a mean 12.9% of applause events. Only a mean 61% of applause events across all six speeches were found to be fully synchronized with speech. Of the four principal types of mismatch identified in this study, only one (isolated applause) is discussed in any detail by Atkinson, Heritage and Greatbatch. Given that isolated applause occurred the least frequently of the four types of mismatch (M = 4.6% of applause incidences), it is perhaps not surprising that Atkinson’s research underestimates the frequency of mismatches. A detailed qualitative analysis showed that each of these four principal types of mismatch may occur either in the presence or absence of rhetorical devices. Mismatches associated with rhetorical devices can occur in a variety of ways. If the speaker does not allow the audience sufficient time to applaud at a possible completion point and continues 292 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 to speak, the audience applause may become interruptive. If the rhetorical device is complex, the audience may start to applaud before the speaker has reached the projected completion point, thereby interrupting the flow of the speech. If the rhetorical device is poorly constructed, applause may be delayed until the audience recognizes that a completion point has been reached, or interruptive, if the audience anticipates the completion point because it has not been projected sufficiently well. Each of the above forms of mismatch may take the form of either collective or isolated applause. Finally, the speaker may interrupt applause that occurs in response to a rhetorical device. All the above forms of mismatch show different ways in which the rhetorical devices identified by Atkinson, Heritage, and Greatbatch may break down. However, it is possible to accommodate these different forms of mismatch within Atkinson’s existing theoretical framework if one accepts that speakers may differ in the skill with which they deploy these devices. If the speaker uses poorly constructed devices, if the rhetorical structure is complex, if the speaker’s timing is poor, this may all result in incidences of mistimed applause. But this does not undermine the concept of rhetorical devices; it simply means that speakers do not always use them to full effect. Of much greater significance for Atkinson’s theory are mismatches which occur in the absence of rhetorical devices. In the case of delayed applause, it is possible that if the speaker pauses for the audience to respond, the pause itself may function as a kind of clumsy invitation to applaud. But when interruptive applause occurs in the absence of rhetorical devices, there is no reason to believe that in any sense the speaker is inviting applause. It can be more easily understood in the context of Bull’s (2000) proposal that a second type of applause can be distinguished in the context of political speeches, referred to as uninvited applause. Bull’s (2000) analysis included all instances of interruptive collective applause to nonrhetorically formatted statements in three of the speeches analysed here (those from 1996). From a content analysis of these statements, he argued that the applause was initiated by the audience specifically in response to statements of policy; thus, the audience was applauding the content of the speech, regardless of the fact that these messages were not rhetorically formatted. Whereas Heritage and Greatbatch (1986, p. 146) state that “audience agreement may be a necessary condition for the generation of applause, but it is not generally a sufficient one,” Bull’s (2000) analysis of uninvited applause in response to specific policy statements would suggest that audience agreement alone can indeed be a sufficient condition for collective applause. The concept of uninvited applause can also be extended to incidences of isolated applause. Bull (2000) has argued that collective uninvited applause need not be synchronized with speech but may be characterized by a staggered onset, through what Clayman (1993) terms mutual Bull, Noordhuizen / THE MISTIMING OF APPLAUSE 293 monitoring. In this instance, the initial reactions of a few audience members may prompt others to join in, resulting in collective applause. From this perspective, isolated applause in the absence of rhetorical devices may be seen as the response of a few audience members to some aspect of the content of the speech, but one which fails to draw in sufficient numbers of the audience to turn it into collective applause. Indeed, the speaker may interrupt such applause to move on with the speech. Thus, the results of this study are consistent with Bull’s (2000) proposal that two types of applause can be distinguished in political speeches, namely, invited and uninvited applause. In effect, audience applause can be seen to occur as a result of an interaction between speaker, message, and audience; however, the interaction can be seen to vary according to whether the applause is invited or uninvited. In the case of invited applause, the speaker encodes applaudable content in rhetorical structures, which if decoded appropriately will result in a high degree of synchrony between speech and applause. Mismatches may occur if the rhetorical structures are poorly encoded or if the audience misreads the signals. In the case of uninvited applause, members of the audience are not responding to invitations to applaud but are initiating applause in response to specific aspects of speech content. Thus, frequent mismatches might be expected to characterize uninvited applause; interaction occurs in terms of the audience’s response to speech content and also between different sections of the audience if the applause is taken up through mutual monitoring. In summary, this analysis of mismatches and the earlier analysis of nonrhetorically formatted statements (Bull, 2000) when taken together can be seen as highlighting a principal weakness in Atkinson’s research, that is, a failure to examine negative examples that might prove inconsistent with his theory. Whereas Atkinson, Heritage, and Greatbatch have produced a great detail of data to show what is in effect a strong positive correlation between rhetorical devices and collective applause, they do not report analyses of nonrhetorically formatted statements that receive collective applause nor of instances where applause is not synchronized with speech. Analyses of such negative instances reported here and elsewhere (Bull, 2000) are consistent with the view that applause is not only invited by speakers at political meetings through rhetorical devices but may also be uninvited in the absence of such devices. Furthermore, it may be unsynchronized with speech, initiated by a few members of the audience in direct response to speech content and sometimes taken up by the audience as a whole. In short, applause in political speeches may not only be invited through rhetorical devices but may also be an uninvited, spontaneous reaction to the content of the speech itself. 294 JOURNAL OF LANGUAGE AND SOCIAL PSYCHOLOGY / September 2000 REFERENCES Atkinson, J. M. (1983). Two devices for generating audience approval: A comparative study of public discourse and text. In K. Ehlich & H. van Riemsdijk (Eds.), Connectedness in sentence, text and discourse (pp. 199-236). Tilburg, Netherlands: Tilburg papers in Linguistics. Atkinson, J. M. (1984a). Our masters’ voices. London: Methuen. Atkinson, J. M. (1984b). Public speaking and audience responses: Some techniques for inviting applause. In J. M. Atkinson & J. C. Heritage (Eds.), Structures of social action: Studies in conversation analysis (pp. 370-409). Cambridge: Cambridge University Press. Atkinson, J. M. (1985). Refusing invited applause: Preliminary observations from a case study of charismatic oratory. In T. van Dijk (Ed.), Handbook of discourse analysis (Vol. 3, pp. 161-181). London: Academic Press. Bull, P. E. (1986). The use of hand gesture in political speeches: Some case studies. Journal of Language and Social Psychology, 5, 103-118. Bull, P. E. (2000). Do audiences applaud only “claptrap” in political speeches? An analysis of invited and uninvited applause. Social Psychological Review, 2, 32-41. Clayman, S. E. (1993). Booing: The anatomy of a disaffiliative response. American Sociological Review, 58, 110-130. Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37-46. Heritage, J., & Greatbatch, D. (1986). Generating applause: A study of rhetoric and response at party political conferences. American Journal of Sociology, 92, 110-157. Roger, D. B., Bull, P. E., & Smith, S. (1988). The development of a comprehensive system for classifying interruptions. Journal of Language and Social Psychology, 7, 27-34.