Claps and Claptrap: The Analysis of Speaker-Audience Interaction in Political Speeches

Delivery

The delivery of a speech can refer to various forms of body language, such as the use of hand gesture, posture, gaze and facial expression. It can also include the vocal delivery of a speech, for example, tone of voice, loudness, pitch and speech rate. It has long been recognized that delivery plays an important role in oratory. In ancient Rome, the treatises of both Cicero (55BCE/2001) and Quintilian (c.95CE/2015) included a number of observations on the use of gesture. Notably, the term “gestus” for Quintilian referred not only to actions of the hands and arms, but also to movements of the rest of the body.

But can we be more specific about the role of delivery in speech-making? Atkinson (1984, p. 84) argued that audiences are much more likely to applaud if a rhetorical device is accompanied by appropriate delivery. Heritage and Greatbatch (1986) coded a sample of speeches formulated in one of the seven basic rhetorical devices in terms of the degree of “stress”. Stress was evaluated in terms of whether the speaker was gazing at the audience at or near the completion point of the message, whether the message was delivered more loudly than surrounding speech passages, or with greater pitch or stress variation, or with some kind of rhythmic shift or accompanied by the use of gestures. In the absence of any of these features, the message was coded “no stress”. One of these features was treated as sufficient for a coding of “intermediate stress”, while the presence of two or more features was categorized as “full stress”. Over a half of the “fully stressed” messages were applauded, only a quarter of the “intermediate” messages attracted a similar response and this figure fell to less than 5% in the case of the “unstressed” messages. Thus, Heritage and Greatbatch supported Atkinson’s view that delivery increases the chance of a rhetorical device receiving applause (see also Bull, 1986).

An alternative viewpoint has been put forward by Bull and Wells (2002). These authors proposed that delivery tells us whether or not the rhetorical device is an applause invitation. So, for example, a speaker may deliver a three-part list, each item accompanied by a hand gesture, and receive tumultuous applause. But if the speaker continued to gesture after the third item, and/or took a visible and/or audible intake of breath, this would suggest that the list was not intended as an applause invitation. Notably, not every rhetorical device receives applause, something for which Atkinson’s analysis never provided an account. In contrast, Bull and Wells’ (2002) proposal that delivery is integral to applause invitation can be used to account for both why some rhetorical devices receive applause, and not others.

The Content of Speech

Of course, audiences do not simply applaud rhetoric, they also respond to the content of a political speech. Atkinson (1984) never sought to deny this. He conducted an analysis of the kind of content that received applause, and found that predominantly it took the form of what he called ingroup praise (praise of your own party), and outgroup derogation (criticism of others). In a sample of applauded statements using rhetorical devices, 95% made favourable references to specific individuals, favourable references to “us” and unfavourable references to “them” (Atkinson, 1984, p. 44). Atkinson took the view that audiences are much more likely to applaud if applaudable speech content is expressed in an appropriate rhetorical device.

A content analysis was also conducted by Heritage and Greatbatch (1986). 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 of political message made up over 81% of all the applauded messages in their sample (see also Bull & Wells, 2002).

Heritage and Greatbatch (1986) also analysed external attacks in further detail. Whereas 71% of external attacks expressed in one of the seven rhetorical devices were applauded, only 29% of external attacks without rhetorical devices received applause. Thus, while Heritage and Greatbatch acknowledged that applause is clearly related to certain types of speech content, they concurred with Atkinson’s view that such content is much more likely to receive applause if couched in appropriate rhetorical devices. Thereby, speakers may also be seen to facilitate their interaction with the audience, given that there are strong normative expectations that audience members should applaud at party political conferences.

However, what this analysis does not address is the role played by speech content in the absence of applause invitations. In one study, instances were identified from leader speeches at British party political conferences where collective applause occurred in the absence of any of the seven rhetorical devices described above (Bull, 2000). In every case, the applause occurred in response to statements of political policy, that is, what the leader proposed to do if returned to power. Thus, for some messages, speech content may be so significant that it will receive applause in the absence of rhetorical devices.

The following example comes from a speech by Tony Blair (1 October, 1996), his last party conference speech before he became Prime Minister. Applause is indicated by lower case crosses, louder applause by upper case crosses (following Atkinson, e.g., 1984).

There was nothing in Blair’s use of the phrase …to join a trade union…” to suggest that this was a completion point. Indeed, given that he followed it with “…and if…”, it seems likely that his intention was to continue. Nor did his delivery suggest that he had reached a completion point; he was not gesturing, and he continued to look straight ahead at the audience. It is of course possible that the audience mistakenly anticipated a completion point after “…join a trade union”, but given the strong traditional association of the Labour Party with trade unionism, it is much more likely that audience members were interrupting in order to endorse Blair’s support of the right to join a trade union, hence, thus, their applause seemed a direct response to the content of speech.

Uninvited Applause

In the example above, the applause for “…to join a trade union” seemed not only to be a direct response to the content of speech, it also seemed to be uninvited. That is to say, Blair was not making use of any of the rhetorical devices described above, nor did his delivery suggest an applause invitation.

Both invited and uninvited applause were analysed in 15 speeches from the annual British party conference speeches (Bull & Wells, 2002). To identify uninvited applause was relatively unproblematic; reliability between the two raters was 0.94. Most of the applause was judged as invited (86% of all applause incidents), the remaining 14% was uninvited. However, in contrast to the previous study (Bull, 2000), there were incidents of uninvited applause associated with rhetorical devices, but where the delivery did not suggest the device was intended as an applause invitation (75% of all incidents of uninvited applause).

For example, the following extract comes from a speech by William Hague (7 October, 1999), at that time Leader of the British Conservative Opposition. “What annoys me most about today’s Labour politicians is not their (A) beliefs – they’re entitled to those – (B) but their sheer, unadulterated hypocrisy. They (A) say one thing and (B) they do another”. In this extract, Hague used the rhetorical device of a contrast twice in quick succession (“beliefs” are contrasted with “hypocrisy,” “saying one thing” is contrasted with “doing another”). However, Hague also showed a very clear and visible intake of breath following the phrase “they do another,” which suggested that his intention had been to continue, and that he had not been seeking applause at that point. Hence, the applause which occurred after “…they do another” was judged to have been uninvited and interruptive.

Thus, from this perspective, uninvited applause may occur not only as a direct response to the content of the speech, but also through a misreading of rhetorical devices as applause invitations, when the associated delivery suggested that the politician intended to continue with his speech.

Rhetorical Devices

The seven rhetorical devices originally identified by Atkinson (1984) and Heritage and Greatbatch (1986) (referred to subsequently as the seven traditional devices) were all identified from British political speeches. But these devices may not be characteristic of political oratory worldwide, they may be specific only to British political culture.

In the two studies of Japanese elections (Bull & Feldman, 2011; Feldman & Bull, 2012), the seven traditional devices could all readily be identified, but only accounted for a small proportion of the collective applause (29% in the 2005 election, 26% in the 2009 election). Hence, it was found necessary to devise seven new categories of rhetorical device. These additional categories are listed and defined below, together with illustrative examples (based on Bull and Feldman, 2011).

Audience Responses

All the early interactional research on political speeches was focussed principally on applause (e.g., Atkinson, 1984). But of course audiences do other things beside applaud. They may, for example, cheer or laugh. In the two studies of Japanese politicians, laughter and cheering were analysed, as well as applause. Although applause was the predominant form of audience response in the 2005 election (59% of responses), there was also a substantial proportion of laughter (25%) and cheering (16%) (Bull & Feldman, 2011). In the 2009 election, there was almost as much laughter (39%) as applause (40%), while cheering was just 9% (Feldman & Bull, 2012).

In addition, analyses were conducted of what are termed aizuchi (Bull & Feldman, 2011; Feldman & Bull, 2012). In Japanese, this term refers to what in English has been called listener responses (e.g., Dittmann & Llewellyn, 1967), signals used to indicate continued listener attention and interest. Aizuchi are considered reassuring to the speaker, showing that the listener is active and involved in the discussion. Common aizuchi are “hai”, “ee”, or “un” (yes, with varying degrees of formality), “sō desu ne” (that’s how it is, I think), “sō desu ka” (is that so?), “hontō”, “hontō ni” or “honma” (really). In these Japanese speeches, the use of phrases such as “Don’t you think so?”, “Wouldn’t you agree with me?” is directly comparable to the way in which speakers request aizuchi in ordinary conversation. However, audiences typically responded to such phrases not with aizuchi but with applause. Actual aizuchi responses were relatively infrequent (only 3.3% of all affiliative responses in the study of the 2009 election; Feldman & Bull, 2012), but those that did occur were typically in response to the speaker requesting agreement (75% of all aizuchi responses). Specifically, these took the form of hai, tadashii desu and honto desu (“yes, this is true”), machigai nai and sono ori desu (“you are correct”), atarimae and tozen (“naturally, obviously”), and tashika ni (“certainly”).

Another notable feature of Japanese audience responses was the total absence of what is termed isolated applause. This occurs when only one or two people clap, in contrast to collective applause, either from the audience as a whole, or from a substantial proportion of it. Isolated applause has been noted in several studies of British political speeches (e.g., Heritage & Greatbatch, 1986; Bull, 1986). For example, in one study, 5% of all the applause in six party political conference speeches was judged to be isolated (Bull & Noordhuizen, 2000). In the US, individualized responses are even more pronounced. In the study of 11 speeches from the 2012 American presidential election, individualized audience responses were observed throughout, with a constant flurry of isolated applause and encouraging individual verbal remarks, mostly interruptive (Bull & Miskinis, 2015). Such responses were never observed in the analyses of Japanese political meetings. In Japan, all the audience responses were collective, that is to say, people applaud, laugh, cheer or produce aizuchi together (Bull & Feldman, 2011; Feldman & Bull, 2012).

The collective US audience responses were coded into applause, cheering, laughter, chanting and booing. There were also a noticeable proportion of responses that could not be assigned to any category because of their infrequency and variance, hence there was a sixth category of others. These include unison verbal remarks, for example “Yes!”, “Amen”, and empathetic sighs. Overall, in marked contrast to both Japan and the UK, it was found that cheering was by far and away the most frequent response (63% of total audience responses); applause accounted for only 12% of all responses.

Another distinctive feature of audience responses in the US was the occurrence of booing. In one study of 178 speeches from the US presidential campaign of 1980, the occurrence of booing was noted as well as that of applause (West, 1984). In another study, Clayman (1993) gathered data on booing from a wide variety of public speaking environments, including US Presidential debates, Congressional floor debates, TV talk shows, and British party conference speeches. Clayman observed that booing occurs quite differently from applause. Booing is typically preceded by a substantial delay or by some other audience behaviour (such as heckling, jeering, clapping or shouting), or by both of these in combination. It was further observed that booing can follow affiliative responses (such as applause and appreciative laughter) as often as it follows disaffiliative responses. In this context, booing can be seen as a reaction to such affiliative responses, indicating that support for the speaker is not universal; furthermore, this reaction can be seen as decidedly competitive.

From this analysis, Clayman (1993) proposed that there are two principal ways in which an audience can coordinate its behaviour, referred to as independent decision-making, and mutual monitoring. In independent decision-making, individual audience members may act independently of one another yet still manage to coordinate their actions, for example through applause in response to rhetorical devices. In mutual monitoring, individual response decisions may be guided, at least in part, by reference to the behaviour of other members.

Thus, once it becomes evident that 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, while conversely not applauding can also increasingly become an isolating experience. 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. “Clappers usually act promptly and independently, while booers tend to wait until other audience behaviours are underway” (Clayman, 1993, p. 124).

In contrast, in the study of speeches from the 2012 US presidential election, Bull and Miskinis (2015) identified two distinctive types of booing: disaffiliative (the audience boo the speaker), and affiliative (the audience align with the speaker to boo a political opponent). Overall, in the 11 speeches, there were 48 instances of booing (45 affiliative, 3 disaffiliative). All 48 instances of booing were preceded by rhetorical devices characteristically used to invite applause. Hence, this raises the question as to whether booing (like applause) can be regarded as an invited response.

An example of affiliative booing can be seen in the following statement from a speech by Obama in Ames, Iowa (29 August, 2012): “Last week my opponent’s [i.e., Romney’s] campaign went so far as to write you off as a lost generation. That’s you according to them”. When the audience booed this statement, they could be seen not as attacking Obama, but as aligning themselves with Obama against Romney.

An example of disaffiliative booing comes from a speech delivered by Romney to a predominantly hostile audience at a conference of the National Association for the Advancement of Colored People in Houston (NAACP), Texas (11 July, 2012). “If our goal is jobs, we have to stop spending a trillion dollars more than we take in every year. And so, I am gonna eliminate every non-essential programme I can find. [HEADLINE] That includes Obamacare. [PUNCHLINE] And I’m gonna work to reform…” In this example, the booing started when Romney stated his opposition to a focal election campaign topic, that of Obamacare [which required everyone to buy health insurance from January 2014]. The booing can be clearly understood as disaffiliative, because of the accompanying shouts directed at Romney of “Get off the stage”, “no” and “shame” (Bull & Miskinis, 2015).

In this example, the booing is clearly associated with a headline-punchline device, customarily used to invite applause. Given that the booing was clearly disaffiliative (it was directed at Romney), the audience response might be regarded as uninvited (i.e., Romney was inviting applause, but instead got booed). However, this interpretation is seriously open to question. Notably, Romney was subsequently widely criticized by Democrats for staging a political stunt in which he deliberately invited booing to make himself look tough in the eyes of hard-line Republicans – as the man who was standing up to Obamacare, who was prepared to put himself on the line even to a predominantly hostile audience (see: “Romney Says He Knew He’d Be Booed at NAACP”; “Mitt Romney On NAACP Obamacare Booing: If People Want More Free Stuff They Should Vote For Obama”). From this perspective, this example of disaffiliative booing may be regarded as invited response. In a sense, Romney appears rather like a stage villain in a pantomime, inviting the audience’s disapproval. Thereby, he might enhance his standing not with the audience in the conference hall but with an audience elsewhere, namely, that of hard-line Republicans (Bull & Miskinis, 2015).

Overall, in the other 10 speeches by Obama and Romney, 7% of audience responses took the form of affiliative booing, most frequently associated with the rhetorical device of negative naming (55% of devices associated with affiliative booing) (Bull & Miskinis, 2015). Negative naming does also occur in British political speeches, but as a form of applause invitation (Bull & Wells, 2002). However, affiliative booing has never been observed in any of the analyses of British political speeches (e.g., Atkinson, 1984; Bull, 2006; Heritage & Greatbatch, 1986), it seems to be distinctive to US political culture. Notably, neither negative naming nor booing (affliative or disaffiliative) were observed in either of the studies of Japanese speeches (Bull & Feldman, 2011; Feldman & Bull, 2012).

Most of the audience responses discussed above were affiliative, that is to say, the audience are invited to align with the speaker. Even booing can be affiliative, as shown in the analysis of US speeches. Of course, booing can also be disaffiliative, for example, when Romney was booed in his speech in Texas (Bull & Miskinis, 2015). But so too can applause, cheering and laughter. For example, an audience may slow hand clap, or cheer or laugh at a pratfall by the speaker. From this perspective, it is not the responses themselves that are intrinsically affiliative or disaffiliative, but how they are used and in what context.

Individualism and Collectivism

In what way can these cross-cultural differences in speaker-audience interaction between the US, UK and Japan be conceptualized? In his highly influential cultural dimensions theory, the Dutch social psychologist Geert Hofstede (e.g., Hofstede, 2001; Hofstede et al., 2010) famously distinguished between what he called collectivist and individualist cultures. A collectivist society is defined as one in which “people from birth onwards are integrated into strong, cohesive in-groups, which throughout people’s lifetime continue to protect them in exchange for unquestioning loyalty” (Hofstede et al., 2010, p. 92). In a collectivist culture, people tend to view themselves as members of groups (e.g., families, tribes or nations), and usually consider the needs of the group to be more important than the needs of individuals. Most Asian cultures, such as Japan, tend to be collectivist, according to Hofstede (e.g., Hofstede, 2001; Hofstede et al., 2010).

Conversely, an individualist society is defined as a culture in which “the ties between individuals are loose; everyone is expected to look after him or herself and his or her immediate family” (Hofstede et al., 2010, p. 92). Thus, the emphasis is on personal freedom and achievement at the possible expense of group goals, resulting in a strong sense of competition. Social status is awarded to personal accomplishments, and all actions that make an individual stand out.

Individualism and collectivism have been conceptualized in terms of self-construals by Markus and Kitayama (1991). Thus, many Asian cultures have concepts of self that insist on the fundamental relatedness of individuals to each other, on harmonious interdependence. In contrast, Americans neither assume nor value such overt connectedness. Individuals seek to maintain their independence from others, by discovering and expressing their own unique attributes. On the basis of a range of large-scale questionnaire studies, the Hofstede Centre (https://www.geert-hofstede.com) has provided ratings of different cultures on individualism and collectivism. Whereas the US and the UK have one of the highest rated levels of individualism in the world (91 and 89 respectively), Japan scores just 46 (scores range from 0 to 100).

Hofstede’s cultural dimensions theory (e.g., Hofstede, 2001; Hofstede et al., 2010) has been highly influential, but is also highly controversial. A number of criticisms have been summarized by Shaiq, Khalid, Akram, and Ali (2011) as follows. For example, one common objection is that Hofstede overgeneralises to whole communities from individual assessments, in particular, that he based his analyses on the data from only one commercial company (IBM). Another criticism is that whereas his unit of analysis is the nation state, cultures may be fragmented across groups and national boundaries. A further criticism is that the research work is too old, and cannot be effectively implemented in an era of rapidly changing environment, international convergence and globalization. One of Hofstede’s most vocal critics has been McSweeney (2002), who regards Hofstede’s research methodology as fundamentally flawed, dismissing it as “a triumph of faith – a failure of analysis”.

Despite these criticisms, it is notable that cross-cultural differences in speaker-audience interaction can be readily subsumed within Hofstede’s dimension of individualism-collectivism. Thus, in the two studies of general elections in Japan (Bull & Feldman, 2011; Feldman & Bull, 2012), all audience responses were collective, there were no incidents of either isolated or uninvited applause. Nor were there any incidents of either negative naming or booing, which might be regarded as disruptive to group harmony, and hence to the interconnectedness regarded as such a distinctive feature of collectivist societies (e.g., Markus & Kitayama, 1991). Furthermore, rhetorical devices were predominantly explicit, thereby making it clear to the audience when it would be appropriate to respond. Arguably, if people in a collectivist society want to respond together en masse, it is helpful if speakers provide the audience with clear guidance as to what is expected (Bull & Miskinis, 2015). Implicit devices can be confusing, and result in applause which may be uninvited (Bull & Wells, 2002), or asynchronous, delayed or interruptive (Bull & Noordhuizen, 2000).

In the United Kingdom, audience responses are typically collective, but incidents of isolated applause have also been observed (e.g., Heritage & Greatbatch, 1986; Bull & Noordhuizen, 2000). So too have incidents of uninvited applause, which can occur either through a misreading of rhetorical devices (Bull & Wells, 2002), or as a direct response to speech content (Bull, 2000). The occurrence of both isolated and uninvited applause suggests that British audiences are more individualistic than those in Japan. Furthermore, rhetorical devices are predominantly implicit, which arguably shows greater respect for individual autonomy, thereby allowing audience members greater freedom of action as to whether or not to respond (Bull & Miskinis, 2015). Again, this would be more consistent with an individualistic society.

In the US, rhetorical devices are also predominantly implicit; arguably it is thus possible to speak of an Anglo-American rhetorical style. However, audience responses are much more varied than the UK and Japan. Not only do US audiences applaud, cheer and laugh, they also chant and boo. Furthermore, although they respond collectively, they also respond on an individual basis, and there is a constant flurry of encouraging individual verbal remarks audible throughout all the speeches, which are both uninvited and interruptive (Bull & Miskinis, 2015). In that respect, US audiences are highly individualistic, in marked contrast to the pronounced collectivist audience behaviour observed in the two studies of general elections in Japan (Bull & Feldman, 2011; Feldman & Bull, 2012). However, it should perhaps be noted that this apparent individualism might also reflect response norms from another context, namely, that of Evangelical/Pentecostal religious responses to sermons and prayers during services, where people are believed to be “moved by the Spirit” to express utterances about the proceedings. Although they respond individually, they might thereby be seen as following group norms.

Political Speech-Making as Dialogue

The Cross-Cultural Context of Speaker-Audience Interaction