In the theater of political style: Touches of populism,pluralism and elitism in speeches of politicians

Introduction

Currently, in most modern Western democracies, classifications of politicians are problematic. For instance, in the United States, political opponents Donald Trump and Bernie Sanders both criticize the same political elite and agree on subjects such as trade and infrastructure. Similarly in France, extreme right Marine Le Pen and extreme left Jean-Luc Mélenchon both want to change French policy radically, proposing to leave the European Union (EU). Similarities such as these are often explained as ‘populist’ in contrast with ‘normal’ or ‘pragmatic’ politics (Canovan, 1999). These politicians would share a populist style but at the same time differ in ‘full’ ideology, because populism is not a full ideology itself, but a political style or ‘thin-centered ideology’ (Mudde, 2004: 544). ¹

However, this definition of populism poses a contrast between ideology and political style that does not match the observation that ideologies today are not as clear-cut as they have seemed in the past, given the above sketched similarities between left and right wing politicians. Based on theories of ideology and style, I suggest that style should not be seen as something apart from ideology, but as a modern, fragmented shape within which ideological thought is expressed. It is based more on ‘common sense’ than on ideological theory (Van Dijk, 1998, 2015) and can be seen as a ‘tag-cloud’ of issues scattered over the (new) media. This cloud of what I call ‘ideological style’ is held together by a politician’s ‘social style’ and ‘rhetorical style’: the former is used to connect with a group of voters sharing the same style; the latter aims at persuading people who do not share that style (yet). The three styles together form the ‘political style’ of a politician. This redefinition of political style, splitting it up into ideological, social and rhetorical, is grounded on style theories from diverse disciplines (political science, political philosophy, rhetoric, sociology and sociolinguistics) (i.e. Ankersmit, 1996; Coupland, 2007; Giddens, 1991; Hariman, 1995; Moffitt, 2016; Pels, 2003).

This article further examines the nature of ‘ideological style’, or the part of political style that concerns the expression of ideas. Regarding it as a loose collection of ideological meaning, this article adapts critical discourse studies (CDS), since this approach mainly focuses on revealing hidden ideological meaning in discourse. With the help of the discourse-historical approach to CDS (Reisigl and Wodak, 2015; Wodak, 2015), this article offers a set of tools for ideological style analysis of political performances. These tools are based on theories of political performance (Wodak, 2009), on ideological theory (Thompson, 1990; Van Dijk, 1998, 2015) and Goffman’s (1956) theory on the presentation of the self in everyday life.

Five ‘core’ traits of ideological style are identified where politicians in modern Western democracies can be compared. Labels of three main styles are adopted from the historical context (elitism, pluralism and populism) and defined with the help of political research (i.e. Bachrach 1971, 1980; Blokland, 2011; Dahl, 2006; Mudde, 2004).

Using this method for speech analyses of candidates in the 2016 US primaries (Trump, Sanders and Clinton) results in unique combinations of elitism, pluralism and/or populism. These combinations form individual ideological style profiles, contributing to a better understanding of the performances of these three politicians.

Ideology changing form

Due to media developments, the political environment has dramatically changed over recent decades. This has had a great impact on political style. According to Street (2003), today’s politicians can be compared to film stars: ‘Skills of performance are skills in self-presentation, in style. Style matters to politics, just as it does to other cultural forms’ (p. 95). In today’s so-called ‘audience democracy’, a politician is a media expert, a person who excels in communication through the modern media. In this type of democracy, ‘a candidate […] must not only define himself, but also his adversaries. He not only presents himself, he presents a difference’ (Manin, 1997: 222).

Considering the strong connection between mass communication and the emergence of ideologies, this new mediated environment changes the very core of ideology. In the late 18th century, mass media were a condition for the rise of traditional ideologies ² (Gouldner, 1976: 91–117; Thompson, 1990: 82–85). As the media were a print form, the newly formed ideologies were bound to written theory; not only was verbal logic required to form coherent arguments, the editors of newspapers and publishing houses belonged to an intellectual class who favored thoroughly deliberated theories. This changed in the second half of the 20th century, when the dominant media shifted from printed to televised and online media. The newer media allow a looser assemblage of ideological content where images are as important as words. This assemblage does not need a profound coherence because its content is spread in fragments over television and the Internet. This ‘ideological style’ is mostly based on what Van Dijk (1998: 103–107) describes as ‘common sense’ – socially shared representations, based on immediate experiences and on simplified scholarly knowledge, as communicated through mass media. As a consequence, people no longer need to be educated to a level of ideological competence in order to judge politicians; they can judge them based on political style and taste (Pels, 2003: 50).

In sum, within the past few decades communicated ideological meaning has changed from verbal long stories or theories into visual/auditory tag-clouds or styles. This does not mean that traditional ideologies are necessarily ‘dead’, but rather that they have a new strong competitor in ideological style. It is still difficult to make sense of these new forms of communicated ideological meanings because people are used to looking for a coherent long story or theory. To see a new sort of coherence, ideological styles must be examined alongside traditional ideologies.

One similarity between traditional ideologies and ideological style is that both can be defined as ‘the interface between social structure and social cognition’ (Van Dijk, 1998: 8). It means that it serves as the basis for shared representations of a group. Furthermore, according to Van Dijk, analysis of ideological meaning has to consider three levels: first, these representations consist of a cognitive frame in which issues are regarded (cognitive level); second, ideology is always shared by groups (group level); third, ideology is always expressed in language or non-verbal communication (discourse level). Applied to ideological style, it results in its definition as ‘the stylized manner in which ideological thought is expressed in discourse and group behavior’. To understand how ideological style is politically used, the concept needs to be regarded in the broader context of political style.

Three layers of political style

Long before the emergence of political ideologies, political styles were acknowledged. In classical Rome, Cicero distinguished two styles: the sober Attic style based on ancient Greek rhetoric, and the bombastic Asiatic style developed in Roman cities in Asia Minor. The Attic style appeared more sincere because of a ‘lack’ of rhetoric, while the abundance of rhetorical elements of the Asiatic orators was thought to be more powerful. Cicero preferred the Asiatic style in front of the people at the Forum. For a more educated audience at the Senate, he typically favored the sober Attic style (Cicero, 2013: [46 BC]; Kennedy, 1999: 48–49).

Traditionally, political style was merely a matter of persuasion: following Aristotle’s Rhetoric, politicians had to adjust their performance according to their audience in order to be effective. For instance, a young audience demanded a different style than an older one, and educated listeners a different style than those less literate (Aristotle, 1984: [360–355 BC], 1390 a 25).

Today, however, sociologists and sociolinguists might frame the two Roman styles as ‘the styling of social identities’ or as ‘self-identities’ adjusted to different ‘lifestyle sectors’ (Coupland, 2007; Giddens, 1991). Within modern sociological theories, identity is not unitary, fixed and stable, but plural and co-constructed in interaction with people, things and spaces. This identity building is a creative process in which people express both their social identity and their search for social roles (Coupland, 2007: 106–108).

Contemporary political scientists, on the other hand, frame political style as ‘aesthetic representation’; within the context of modern representative democracy (in contrast to Cicero’s direct democracy), political styles are regarded as creative fillings of the ‘aesthetic gap’ between the representative and the represented. Because politicians can never be identical to the people they represent, they have to bridge the difference creatively, like a painter paints a portrait (Ankersmit, 1996: 49–56; Pels, 2003: 49). Ankersmit and Pels differentiate between a ‘mimetic’ and an ‘aesthetic’ approach to democracy; a mimetic approach denies the gap and is therefore at odds with the representative nature of democracy. In an aesthetic approach, politicians seek connection with a group of people through a shared style and taste, without suggesting being identical (Ankersmit, 1996: 49–56; Pels, 2003: 49).

Although the notion of the ‘aesthetic gap’ is valuable and is for a large part in accordance with the notion of social style, there is one dissention: in social theory there is no such thing as a not-stylish (mimetic) approach: all social behavior, including politics, is inevitably stylish (Giddens, 1991: 81). In this regard, the mimetic approach can be seen as an anti-style, denying the existence of political styles – just another way to fill the ‘aesthetic gap’.

In this interpretation, I define political style as ‘a stylized political performance purposed to fill the “aesthetic referential gap” with a political identity’. This identity is partly constructed through the expression of ideological thought and partly through expressions reflecting the style of certain social groups in order to connect with these groups. Additionally, political style is about persuading people who are different and do not think alike (yet). Therefore, political style consists of three components:

In practice the three styles are blurred. Nevertheless, for analytical purpose a distinction is advantageous; it allows seeing that politicians creatively combine ideological styles with social and rhetorical styles. For instance, many politicians with a populist ideological style choose a matching popular social style, but this is not mandatory. Pim Fortuyn, for instance, a well-known Dutch politician, mixed mainly a ‘populist’ ideological style with a clear ‘elite’ social style: driving a Jaguar, wearing expensive Italian suits and coquetting with his former title of ‘Professor’. This elite social style did not make him ideologically an ‘elitist’; it just gave him a unique stylized political profile that was ideologically, socially and rhetorically attractive. It is a creative way to ‘fill the aesthetic gap’ and a solution to the problem that all modern populist politicians face – having to be ‘ordinary’ and ‘extra-ordinary’ at the same time (cf. Moffitt, 2016: 55).

This example implies that for the analysis of political style, three different style labels are needed: social, ideological and rhetorical labels. Existing political style labels refer either to rhetoric, using classical concepts such as logos, ethos and pathos in combination with modern rhetorical theory (i.e. Charteris-Black, 2009; Schoor, 2015), or to social styles, such as the four styles of Hariman (1995). Although Hariman’s realist, courtly, republican and bureaucratic styles do refer to political surroundings (such as classical Rome, African autocracy and European contemporary bureaucracy), they do not refer to different styles within the political context of the analyzed performances. This makes labels ‘circular’, as De Haan (2000: 238) argued; they allow no further conclusions than that there are personal (read social) differences in style. Such labels will not enhance insight into the political context; they can even obstruct it. Ankersmit proposed to use political style labels borrowed from painting or music styles; however, these labels also lack a link with the political context (Ankersmit, 2000, 2003; De Haan, 2000).

For insight into the political context, ideological style labels are needed. One known style label referring to the current political context is the ‘populist style’, but it is unclear which other styles it relates to. Mudde (2004) compares populism with ‘pluralism’ and ‘elitism’, while Moffitt (2016) opposes it to the ‘technocratic style’, a style combining expert knowledge with good manners and a belief in progress. Canovan (1999), on the other hand, opposes populism to a ‘pragmatic style’ favoring practical solutions over visionary beliefs. To decide which labels apply, it is necessary to develop the political and ideological style concept further.

Political style and the theater concept

Thus far, political style has not been analyzed systematically. The theater concept can be of help here because it is a common metaphor for politics and social behavior and has facilitated the way people think of both (cf. Lakoff and Johnson, 1980). Earlier, De Haan (2000: 246–248) pointed at the usefulness of the theater to comprehend political style as it expresses its aesthetic and interactive character. Recently, Moffitt (2016: 49) used the theater to explain the performative character of populism as political style.

Traditionally, the theater concept has been used by rhetorical scholars for centuries, visible in concepts such as ‘decorum’ with remarks on tone and gesture (Hariman, 1995: 180; Kennedy, 1999: 110). Since the 19th century, political observers have also used the theater metaphor to express the theatrical character of revolutionary political situations (Fitzgerald, 2015). During the second half of the 20th century, political theorists adopted the metaphor to stress the mediated character of politics. Television evoked the image of politics as a broadcast show that voters at home could follow more directly than ever (Edelman, 1988; Green, 2010; Manin, 1997). An objection to the theater concept is that the metaphor would depict politics as an ‘elite play’ in which the audience can only passively watch. However, this underestimates the active role of the audience/spectators; according to Green, a ‘gaze’ includes a judgment. According to Edelman, this judgment is crucial for the acceptance of an issue as a political problem or not (Edelman, 1988: 32; Fitzgerald, 2015; Green, 2010: 128–129).

Besides for politics, the theater metaphor is used for social life, most notably by Goffman (1956). In his theory on the presentation of the self in everyday life, all people perform a play in their interactions, and most of the plays are standardized so that observers can easily understand a situation (Goffman, 1956: 16). Wodak (2009) elaborated further on Goffman’s theory, focusing on political situations. Like Goffman, Wodak discerns a ‘frontstage’ and a ‘backstage’: at the frontstage, the political performance takes place before an audience. This frontstage requires a ‘setting’, for instance parliament or a television broadcast, and a ‘personal front’. This personal front has two aspects: ‘appearance’ and ‘manner’. Appearance refers to items that reflect social status, and the manner in which actors conduct themselves (Wodak, 2009: 9–10). Although Wodak does not use the words ‘political style’, personal front clearly equals political style as it refers to politicians’ appearances (a dress code, the entourage of a politician, his car, etc.) as well as to their manner (the way they speak, what they say, whom they are friendly with and whom not, etc.). Where Wodak (2009: 23, 36, 56) focuses on the backstage, this article concentrates on the frontstage to define the traits political performances consist of, by comparing frontstage elements such as stage, play and audience with elements of a political performance.

When applied to modern representative democracies, the picture is as follows: in the political theater, actors (politicians) perform a play called ‘democracy’. The set is the location of a speech or debate: in parliament, on television and so on. The stage represents sociohistorical context: the country, available technology and historical events. The roles of the actors differ: being a parliamentarian or part of the government. Because the democratic play does not allow just one director (this would be an autocracy), all politicians are their own director, more or less co-directed by their party. The script is offered by traditional ideologies, but actors can also choose to improvise their performance or use multiple scripts of different ideologies.

Without mandatory scripts or directions, all actors produce their own performance. This is their political style. The stage, being the sociopolitical context, alters in time, and accordingly the audience changes and uses different technology. Because the play resembles a television contest, the audience endorses or declines the actors during the play, calling them to leave or to adjust their appearance. Also, the theater (politics as a whole) and the play (democracy) are under constant critique, or the ‘gaze’ of the audience. Media have the role of ‘super-audience’; because they control the spotlights, they draw attention to one actor or another, and because they have the loudest voice, they also impact public opinion.

Defining core traits of ideological style

The political theater as sketched here offers further insight into the elements constituting political style if these elements are linked to the frontstage characteristics defined by Goffman. Furthermore, four context levels must be reckoned with (Reisigl and Wodak, 2015: 93). Table 1 shows these context levels in relation to Goffman’s frontstage elements, along with those of the political theater.

OPEN IN VIEWER

Table 1.

Analysis of the ‘frontstage’ of politics related to Goffman’s frontstage elements and to Reisigl and Wodak’s four context levels.

	Theater metaphor	Politics	Goffman: frontstage =setting +personal front	Wodak’s context levels a
	Performance	Political style = rhetorical+ social + ideological style	Personal front (= manner +appearance)	1, 2, 3 and 4
1	Lines	The speeches that are uttered	Manner	1
2	The acting itself	Non-verbal communication, movements, gestures, etc.	Manner	1
3	Props and costumes	Dress code (formal or informal), entourage (i.e. security), etc.	Appearance	1, 3
4	Roles	Member of parliament or government, opposition or coalition, candidate in elections or not, etc.	Appearance	3
5	The theater	Politics	Setting	2, 3
6	The play	‘Democracy’ consisting of multiple acts, such as a parliamentary debate, elections, governing	Setting	2, 3
7	The actors	Politicians	Setting	2, 3
8	The audience (interactive)	The people/voters, the media (super-audience)	Setting	2, 3
9	The set	Institutional surroundings/location: parliament, government, party gathering, television broadcast, etc.	Setting	2, 3
10	The stage	The sociohistorical context: the country, historical events, available technology	Setting	2, 4

a

Context levels: 1 = text level; 2 = discourse/intertextual level; 3 = social variables and institutional frames; 4 = social political history.

In Goffman’s terminology, political style, involving manner and appearance, consists of elements 1–4 as shown in Table 1: speech, non-verbal communication, dress code and other entourage elements, and political role. The former two elements concern the text level, while the latter two mostly belong to the social variables and institutional frames.

This means that political style can be described through observation of elements 1–4. By comparing these descriptions with social styles from the historical context, the social style can be revealed. In order to apprehend the rhetorical style, language use (element 1) can be analyzed with the help of rhetorical theory. Finally, in order to define the ideological style, the four style elements have to be interpreted in relation to ideological differences.

Traditionally, the sociohistorical context (element 10, the stage) was the focus of political researchers to define the differences between politicians on ideological issues. However, politicians can also differ in the style in which they frame the context: they can look at it as being ‘in crisis’ or as ‘under control’. In the same manner, politicians can differentiate themselves ideologically on other elements of the political setting (elements 5–10) because all these components possess a discourse component besides their extra-linguistic/institutional nature. For instance element 5, ‘politics’, obviously has an institutional side (an institutional rule is, for instance, the prohibition of physical fights); but within this institutional frame some politicians see politics as ‘cooperation’, while others see politics as ‘conflict’. This differentiation is possible for all mentioned elements (5–10 in the table). Even element 9, ‘location’, is for some a respected institution and for others a ‘fake parliament’.

So besides the traditional ideological differences on political issues, politicians can deliberately make an ideological difference to other aspects of a political performance. An ideological style analysis should therefore look at elements 5–10 in Table 1, forming the core traits of ideological style: ‘politics’, ‘democracy’, ‘people’, ‘politicians’ and the ‘context’. Because ‘location’ is very context-specific and directly connected with other traits, it is integrated in the other traits.

So far, this article has discussed three levels and five core traits of ideological style but is not useful as an analysis instrument yet. These traits and levels have to be combined with labels derived from the political context. Before picking any label from an abundance of seemingly applicable ‘-isms’, the maximal ideological differences on the five traits should be logically subtracted. This deductive approach guarantees that the right labels are selected to most accurately match ideological style. Logically, the largest possible ideological differences are as follows: ⁴

Earlier in this article, populist, elitist, pluralist, technocratic and pragmatic styles were mentioned as existing political style labels. Relating these labels to the five traits of ideological style reveals whether these styles are to be seen as ideological styles or rather as social or rhetorical styles.

Comparing known descriptions of populism with the five traits indisputably positions populism at one extreme end of all traits: populists see (1) the people as ‘good’, (2) politicians as ‘bad’, (3) democracy as ‘people’s self-government’, (4) politics as conflict and (5) the historical context as ‘in crisis’ (Akkerman et al., 2014; Hawkins et al., 2012; Moffitt, 2016; Rooduijn, 2014). So evidently populism settles as ideological style.

Furthermore, comparing the five style traits with the technocratic style (Moffitt, 2016: 46) suggests that the latter refers to a social style. This style associates itself directly with a social group of ‘the experts’, while its features only partly, and more indirectly, refer to the ideological style traits. As a social style, it is better to oppose the technocratic style to a popular (social) style rather than to a populist (ideological) style.

The same applies to the pragmatic style of Canovan (1999): it associates itself primarily with a social group of hands-on entrepreneurs and solution seekers, only resembling the ideological style traits of ‘politics’ and ‘context’. As a social style, it is better to oppose this style to a visionary (social) style rather than to a populist (ideological) style.

Elitism and pluralism, both described by (Mudde, 2004: 543–544) as oppositions to populism, directly refer to all five traits. Theoretical descriptions of elitism situate it at the other end of the ideological style scale, opposite to populism (Akkerman et al., 2014; Bachrach 1971, 1980; Mackie, 2009; Mudde, 2004; Schumpeter, 2008 [1942]). Descriptions of pluralism show that pluralism positions itself exactly at the middle of the scale on the five traits (Blokland, 2011; Dahl, 2006). This means that populism, pluralism and elitism settle as ideological styles, as demonstrated in Table 2.

OPEN IN VIEWER

Table 2.

Populism, pluralism and elitism compared on five core traits.

Ideological style/traits	Populism	Pluralism	Elitism
1. The people	The people are one and good/capable(exclusion of ‘others’)	The people are multiple groups with different interests(inclusion)	The people are one and bad/ignorant(exclusion of the people)
2. Politicians	Politicians are incapable and corrupt(exclusion of politicians)	Politicians are, like all people, likely to do good but also likely to abuse power(inclusion)	Politicians are proficient and good(exclusion of the people)
3. Democracy	Democracy is self-government of the people(power to the people)	Democracy is majority rule, respecting minority rights(power to no one)	Democracy is electing an elite government(power to the elite)
4. Politics	Conflict	Negotiation(conflict as well as cooperation)	Cooperation
5. Political context	Negative: focus on what is lost(crisis)	Factual: focus on facts(alert)	Positive: focus on what has been reached(at ease)

I suggest that with populism, pluralism and elitism, the basic scale of ideological styles being set: elitism and populism holding the end positions and pluralism occupying the middle. However, both populism and elitism are not only exclusively practiced within generally accepted democratic boundaries (Moffitt, 2016: 140) – undemocratic forms of elitism and populism form the real ends of the scale as ‘extreme elitism’ and ‘extreme populism’. Additionally, two mixed forms of populism/pluralism and elitism/pluralism can be inserted to keep the scale balanced and get a better view of the differences between pluralist styles. Questions might arise because the scale levels a mixed elitist/populist score into an average pluralist score. This is accurately so, because a mixed elitist/populist score inevitably means that the politician is a pluralist, albeit performing in a stylized manner.

With a seven-point style scale (varying from extreme populism to extreme elitism), five style traits (the people, politicians, democracy, politics and the context), and three levels (cognitive, group and discourse level), the structure of the analysis tool is ready. It consists of five rubrics on three levels, describing five styles on each trait (the mixed styles are not described separately). ⁵ Based on this structure, a computer-based analysis instrument for the detection of ideological style elements in speeches is developed: the ‘PEP-index’, named after the initials of populism, elitism and pluralism. In the following section, I give examples of an analysis using the PEP-index.

Ideological style of Trump, Sanders and Clinton

The PEP-index is used for speech analysis of US politicians Bernie Sanders, Donald Trump and Hillary Clinton, encompassing two speeches each from the primaries in April 2016. The complete analysis is too long to include; ⁶ however, one trait (the people) is discussed in more detail, while the other traits are only examined in relation to the general results. The analysis was focused on the whole video performance. Partly captioned video recordings of the speeches were retrieved from the Internet ⁷ and transcribed. To define the style, speech fragments were compared with the five rubrics of style traits, of which Rubric 1 (The People) is displayed in Table 3. Meaningful fragments (sentences or short paragraphs) were labeled with codes referring to elements of the rubrics. For instance, the code (1D/PO) refers to a populist score at the discourse level (D) of style trait 1 (The People; see Table 3).

OPEN IN VIEWER

Table 3.

Rubric of ideological style trait 1: ‘The People’.

The People (1)	Cognitive level (1C)	Group level (1G)	Discourse level (1D)
Extreme elitism (XEL)	The elite are superior; the people are inferior.	Discrimination and/or violence toward the people. The elite have special perks.	Verbal threats or incrimination of the people. Justification of criminal behavior of the elite.
Elitism (EL)	The people are a homogeneous group > exclusion of the in-group: the in-group is ignorant and tends to make poor choices. The out-group (the elite) knows what is right.	Elitists have a distance from the people; they are not a part of the people. Identification with an (elite) out-group who knows what is right (exclusive: not everybody belongs to the people).	The people are referred to as ‘them’. Elitists say what they think is good for the people. They speak to or over the people with arrogance; the people do not understand politics.
Pluralism (PL)	The people are a heterogeneous group > inclusion. The people consist of different groups that unite if interests coincide and compete if interests collide.	Pluralists identify with a part of people. They know their own group and communicate with other groups (inclusive: everybody belongs to the people).	The people are referred to as ‘me’, ‘you’ and ‘we’ in the sense of ‘working together’, not ‘being the same’. Pluralists speak as part of one group to other groups. They speak with the people.
Populism (PO)	The people are a homogeneous group > exclusion of the out-group. The in-group are the ‘real people’; they are good and know what is right.	Populists identify with the people; the populist is ‘one of us’ and knows what is right ‘for us’. Distance from the out-group, who are unfair and/or do not know what is right (exclusive: not everybody belongs to the people).	The people are referred to as ‘we’ in the sense of populists being ‘the voice of the people’, saying what the people think, in the language of the people. The out-group is ‘them’.
Extreme populism (XPO)	The people are superior; out-group(s) are inferior people.	Segregation from and discrimination of out-groups. Deportation/exclusion of out-groups from the country.	Justification of discrimination and deportation. Accusing of the out-group of lying.

Source: All five rubrics are available at the Open Science Framework: osf.io/9esxm

Although the PEP-index allows a quantitative approach, the performed analysis was restricted to qualitative inquiry. For each style element, fragments were selected, studied and compared to reveal the overall style of each particular style element. Only speech fragments adding new information were included; for instance, as soon as it was clear that a politician spoke with ‘the voice of the people’ (1D/PO), the focus was no longer on this feature but on fragments that would suggest other styles on element 1D. If two different styles were detected on one element, an estimation of the quantity was taken into consideration. To be confident that this approach was sufficient, a quantitative analysis has been performed of one complete speech. Because this analysis resulted in the same style score as the restricted qualitative search, the rest of the speeches have been analyzed based on the latter approach.

Conclusions of the analysis

The overall results of the analysis are displayed in Figure 1, including the other four ideological style traits, not elaborated on here. The figure shows that Trump’s style can be characterized as ‘populist’, Sanders’ style as mixed populist-pluralist and Clinton’s as mixed elitist-pluralist. Looking at their styles in more detail contributes in many ways to a better understanding of the political situation, of which I mention two.

OPEN IN VIEWER

Figure 1.

Ideological style on five traits and on average of three candidates in the US primaries, April 2016.

First, Trump and Sanders share a populist profile; however, on four of five traits they differ in sub-categories. They only come close in their conception of ‘politics’: both see politics as conflict. Yet there is another similarity: they both blame their political opponent for dividing the people (Trump blames Obama and Sanders blames Trump). This can explain why they both aroused strong feelings among their followers; they both instilled the belief that their followers were the only ones seeing it right. This is more powerful than the pluralist standpoint that different groups always have different interests or the elitist position that the people do not really understand politics anyway.

Second, looking closer at the difference between Clinton and Sanders shows that, despite both being Democrats, their style difference is larger than that between Sanders and Trump. They differ most in their conception of politics: one seeing it as conflict and the other seeing it as working together. Only on one trait (the people) does Clinton score not as an elitist. However, the question is whether the audience believes Clinton’s bond with the people, because although she says that she is ‘one of us’ on a discourse level, her performance on the group level tells that she is far too busy for that. This inconsistency could be an explanation for the media describing her appearance often as unauthentic.

Conclusion and discussion

This article defines political style as ‘a stylized political performance purposed to fill the “aesthetic referential gap” with a political identity’. While social identities are never simple, the constructed political identity is particularly complex because it is assembled through a blend of three styles: besides a social style (the co-construction of identities in interaction with others), it encompasses an ideological style (the fragmented expression of ideological thought) and a rhetorical style (language aiming at persuasion).

This article focuses further on ideological style. Through conceptual analysis with the help of the theater metaphor, five traits of ideological style are defined: the people, politicians, democracy, politics and context. Comparing ideological differences on these traits with existing style labels shows that ‘populism’, ‘pluralism’ and ‘elitism’ position themselves on all five traits to form a balanced scale upon which ideological style can be measured.

Although the performed analyses in this article explain important parts of the performances of Trump, Sanders and Clinton, to get a more complete picture ideological style analysis should be complemented with analysis of rhetorical and social styles. Rhetorical styles can be defined using classical rhetorical theory complemented with modern metaphor theory, as applied by Charteris-Black (2011, 2014) and Schoor (2015). This way the link between rhetorical and ideological styles can be studied further, in the same manner that Hariman (1995) linked rhetorical features to his four political (social) styles.

Furthermore, to define social style, labels can be abstracted from the social context. As starting points, Moffitt’s technocratic style (and its opposite, the popular style), Canovan’s pragmatic style (and its opposite, the visionary style) and the earlier mentioned elite style or notions such as ‘globalist’ versus ‘localist’ can be used to develop a more elaborate range of political social styles. As mentioned earlier, the number of social styles can vary endlessly, but most important is that these styles refer to the social context the analysis focuses on. It would be interesting to see how specific social styles relate to ideological style combinations of politicians.

Although the performed analysis of US presidential candidates only shows two mixed style forms (populism/pluralism and pluralism/elitism), it does not deny the existence of the third combination: elitism/populism. Returning to the earlier mentioned Pim Fortuyn, although he was generally seen as a populist, he sometimes used a mixed populist/elitist ideological style: ‘In my campaign not only politics get a whipping, I also come down hard on the citizen. Worthless politicians but also worthless citizens’ (cited in Pels, 2003: 52). This populist/elitist mix makes sense considering his earlier mentioned ‘elite’ social style.

Fortuyn is not the only one using a mixed populist/elitist style; it resembles a traditional Dutch social style: the ‘Calvinistic reverend’ ⁸ dooming the people as well as politicians for not listening to God. It is, for instance, the style of Reverend Zandt in 1946, a parliamentarian for the Dutch Reformed Party (SGP):

Dit woord ‘vooruitstrevend’ is thans bij de groote massa zeer in trek. Het streelt het oor van menigen kiezer. Het is de stroop, waarmede men de kiezers als vliegen bij menigten vangt. (Currently the word ‘progressive’ is very popular amongst the masses of the people. It is music in the ear of many voters. It is the syrup that is used to catch the voters like masses of flies. (Handelingen, 9 July 1946, p. 68; translated by author)

This quote is taken from a investigation into ideological and rhetorical style changes in Dutch politics, for which the PEP-index was originally developed. Although this research is still in progress, first analyses show that the PEP-index is useful in both the United States and the Netherlands, and therefore probably in other modern representative democracies as well. This suggests that the instrument is suitable for national and international comparison of styles over a longer range of time, as well as for more detailed analysis presented in this article; further research has yet to show whether this is the case.

While this approach assembles theories from different disciplines, the guidelines offered by CDS made it possible to avoid eclectic failures such as mixing up seemingly similar concepts (i.e. Weiss and Wodak, 2003). A considerate interdisciplinary approach not only introduces new and insightful perspectives to the political realm, it also provides an awareness of the multivalence of concepts such as ‘ideology’ and ‘style’.

It might be argued that populism, elitism and pluralism cannot be split up into sections; it is the combination of features that makes them what they are. Nevertheless, as politics is an art, as suggested by political and rhetorical theorists (i.e. Ankersmit, 1996; Hariman, 1995; Pels, 2003), mixing styles is to be seen as a creative act. Political ‘actors’ artistically combine theater costumes of different styles – choosing a populist hat with an elitist gown and pluralist shoes. This sectional view on style provides more meaning than ‘pure’, unalterable labels of populism, elitism and pluralism, mostly because politicians seldom see themselves as elitists or populists, but they might admit that they have populist or elitist traits.

In this article, this approach shows its value through the comprehensive view of the separate style traits and the possibility to compare politicians on such a detailed level. It helps identify precisely which traits a particular style consists of. This is essential to know because when it comes to performing on the political stage, it is often the smaller, cursory stylistic detail that makes the difference.